Split (3)

train · 24.8k rows

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
A fundamental question in human susceptibility to bacterial infections is to what extent variability is a function of differences in the pathogen species or in individual humans . To focus on the pathogen species , we compared in the same individual the human adaptive T and B cell immune response to multiple strains of two major human pathogens , Staphylococcus aureus and Streptococcus pyogenes . We found wide variability in the acute adaptive immune response induced by various strains of a species , with a unique combination of activation within the two arms of the adaptive response . Further , this was also accompanied by a dramatic difference in the intensity of the specific protective T helper ( Th ) response . Importantly , the same immune response differences induced by the individual strains were maintained across multiple healthy human donors . A comparison of isogenic phage KO strains , demonstrated that of the pangenome , prophages were the major contributor to inter-strain immune heterogeneity , as the T cell response to the remaining “core genome” was noticeably blunted . Therefore , these findings extend and modify the notion of an adaptive response to a pathogenic bacterium , by implying that the adaptive immune response signature of a bacterial species should be defined either per strain or alternatively to the species’ ‘core genome’ , common to all of its strains . Further , our results demonstrate that the acquired immune response variation is as wide among different strains within a single pathogenic species as it is among different humans , and therefore may explain in part the clinical heterogeneity observed in patients infected with the same species . Large intra-species variability exists in bacterial genome content . The gene collection found in all members of a species is defined as the essential ‘core’ genome , while genes that are found only in some strains are termed the ‘accessory’ genome . Gene families found within a species as a whole are considered the pangenome [1 , 2] . As a result , various strains of a bacterial species are not equally pathogenic due to dissimilarity in virulence factor expression , among others [3] . Strains within a species are known to express a unique combination of virulence factors and superantigens [4] , many of which are carried by prophage [5 , 6] , constituting a portion of the strain accessory genome . Though some of these virulence factors were shown to induce a robust T cell activation [7] , it is unclear whether and how various strains within a species differentially affect aspects of the adaptive immune response . To address this: ( i ) blood samples from healthy donors were used to rule out primary or secondary immune deficiency as a cause for heterogeneity ( ii ) we evaluated the adaptive immune response of proliferating cells following stimulation with heat killed bacteria [8 , 9] and , ( iii ) the acute adaptive immune response 4 days after initial stimulation was examined , thereby simulating the acute adaptive response to a pathogen early after its encounter . We first assessed the T cell responses to 16 different heat killed Staphylococcus aureus ( S . aureus ) strains , either methicillin sensitive ( MSSA ) , resistant ( MRSA ) or vancomycin resistant ( VISA/VRSA ) . CD4 T cell proliferation and Interferon-gamma ( IFNγ ) expression by proliferating cells , denoting S . aureus-specific cells , demonstrated broad heterogeneity in response to the various staphylococcal strains within the same blood donor ( Fig 1A and 1B ) . Further , results from 10 additional unrelated donors showed that the heterogeneity and relative intensity of T cell responses ( proliferation , IFNγ expression ) by the 16 strains is maintained across different donors ( Fig 1C top for 16 strains , and bottom left for 8 representative strains with high , intermediate and low T cell proliferation , and S1 Table ) . Therefore , strains that were relatively weaker ( e . g . , USA 600 , USA 100 ) or stronger ( e . g . , Newman , NRS111 ) inducers of T cell proliferation or IFNγ expression kept their relative intensity across all donors ( Fig 1B and 1C , and S1 Fig , S2 Table ) . Furthermore , this T cell response heterogeneity was also seen within different strains of Streptococcus pyogenes ( Fig 1G ) . We next evaluated other adaptive immune functions e . g . , B cell proliferation and IgG expression by proliferating cells , using the same donor as in Fig 1A in response to the same 16 S . aureus strains . To decrease bias between experiments , the PBMCs from the culture well that were stained for T cells ( Fig 1A ) were simultaneously co-stained for B cell markers and analyzed by FACS . Again , we observed wide heterogeneity in the intensity of B cell proliferation and IgG expression in response to the various strains ( Fig 1D and 1E ) . Additionally , strains that induced relatively strong T cell responses ( Fig 1A , e . g . , NRS111 , USA300 ) either maintain the same intensity of B cell responses ( Fig 1D , e . g . , USA300 ) or had a low B cell response ( e . g . , NRS111 ) and vise versa ( e . g . , USA600 maintained low , and USA100 increased ) . Moreover , as with the T cell responses , the relative intensity of the B cell response to the 16 strains was maintained across the same 10 donors ( Fig 1F , S2 and S3 Figs and S3 and S4 Tables ) . Comparing T and B cell response intensities to the 16 strains across donors , demonstrated a unique combination of adaptive immunity activation by different strains of the same S . aureus species ( Fig 1H and 1I ) . Indeed , USA600 , USA100 , Mu50 and Newman , induce low/low , low/high , high/low and high/high combinations of T/B cell proliferation , respectively ( Fig 1H and S5 Table ) . The combination of the adaptive effector molecules IFNγ/IgG ( Fig 1I and S6 Table ) has a similar low/high and high/low induction pattern ( for USA600 and Mu50 , respectively ) , however with Newman and USA100 there is some “spread” showing higher variability in IgG expression . However , USA500 induces a high/high IFNγ/IgG expression pattern with low heterogeneity ( S4 Fig and S7 Table ) . Despite the fact that all the strains we used carry a combination of virulence factors and at least one superantigen ( the lab strain RN4220 has no superantigens , [10] ) , their net effect on the adaptive immune response is strikingly heterogeneous . Thus , these findings provide strong evidence that various strains of a species are markedly varied in the acute adaptive immune response they induce . To further understand the potential implication of the adaptive immune response heterogeneity to various strains of a species , we evaluated its possible relevance to immune protection . Th17 cells are crucial in preventing S . aureus infections and a low Th17 cell count is responsible for recurrent infection by this bacterium [9 , 11 , 12] . Further , Th1 and MRSA-specific IFNγ+ CD4 T-cell responses were shown to be essential for the control of initial and recurrent MRSA infections in HIV-infected people [13] . Therefore , we next assessed the S . aureus-specific Th response , focusing on two of the strains in Fig 1 that showed wide differences in their adaptive immune responses ( Newman and USA600 ) . Using cell surface markers ( CxCR3 , CCR6 , CCR4 ) to identify the Th subsets [14] , we found a dramatic difference in S . aureus-specific Th1 , Th17 , and Th1/Th17 subset responses to Newman compared to USA600 ( Fig 2A , 2B and 2C ) . USA600 had at least a 10-fold lower response in both mean values of IL17A ( and IL17F ) as well as IFNγ-expressing cells among the CD3+CD4+ proliferating cells ( Fig 2D and S5 Fig ) . This was accompanied by a similar trend in expression of their master transcriptional regulators in the proliferating cells , i . e . , RORγt and Tbet , respectively ( Fig 2E ) . Further , in contrast to the robust Th17 response described before in response to S . aureus [8] , our findings indicate that the Th response intensity is strain dependent . Apparently , most of this effect is due to prominent differences in CD4 proliferation in response to USA 600 compared to Newman ( Fig 1A , 1B and 1C ) , differences that take place despite the presence of superantigen genes in both strains [4 , 15] . In fact , the observed 10 fold lower response with USA600 occurs despite coding for at least 7 superantigens , while Newman contains only one [4 , 15] . This suggest that its not the presence or number of superantigens or specific virulence factors that contribute to the observed difference but rather the net effect of the combination of these molecules ( unique to each strain ) that dictate the final result on the adaptive immune response . Of note , similar variability in Th1/Th17 was also observed with other S . aureus strains ( e . g . , NRS111 had a 7-fold higher response in values of IL17A as well as IFNγ-expressing cells among the CD3+CD4+ proliferating cells compared to USA700 ) . Since the effect on T cells may be affected by the donor’s previous exposure to S . aureus , we used murine CD4+T cells to identify the difference in the strain effect . We found a statistically significant difference of almost two-fold in T cell proliferation in response to Newman vs USA600 ( S6 Fig ) . We again verified our results with a different bacterial species . The adaptive immune response against S . pyogenes includes a robust Th1 response , and IL-17A was shown to be necessary for S . pyogenes clearance [16 , 17] . Comparing S . pyogenes strains from two M serotypes , i . e . , M1 and M3 , demonstrates a >10-fold difference in mean strain-specific IL-17A or IFNγ CD3+CD4+ expressing cells ( Fig 2F ) , which in large part may be attributed to effects on proliferation ( Fig 1 ) . Serotype M6 strain MGAS10394 demonstrates intermediate counts of IL17A ( S7 Fig right ) or IFNγ ( S7 Fig left ) expressing CD3+CD4+ cells , that did not differ from either serotypes M1 strain SF370 or M3 strain MGAS315 , respectively . However , by performing a joint analysis of the two cytokine-expressing cells ( IFNγ+IL17A- , IL17A+IFNγ- ) we could identify the heterogeneity of M6 compared to the other two strains , showing significant differences in all post-hoc pairwise comparisons ( M1vs M3 , M1 vs M6 , M3 vs M6 , Fig 2G ) . This suggests that combined analysis of several variables may better reveal strain heterogeneity . Thus , our results from two different species strongly suggest that individual strains within a species may differ prominently in the intensity of the necessary protective Th response . To determine the source of the heterogeneity in the adaptive response to the various strains of a species , we focused our attention on those parts of the bacterial genome that may account for inter-strain differences , i . e . , the accessory [1 , 2] genome . To this end , we compared the adaptive immune response to the Newman wild type ( WT ) strain of S . aureus to that of Newman strain TB4 deleted of all four of its prophages ( ϕNM1-4 ) [18] . Regardless of the presence of plasmids , pathogenicity islands or other mobile genetic elements , prophages are a major part of the accessory genome in this species; therefore , their removal more closely reflects the “core” genome of Newman , and other strains within the S . aureus species . Newman strain TB4 induced a blunted T cell response ( Fig 3A and 3B ) , with a striking 10-fold reduction in the means of Newman-specific Th1 and Th17 cell counts from the same donor population ( Fig 3D and 3E ) . This suggests that antigens encoded within the accessory genome induce most of the acute Th1 and Th17 responses , considered to be necessary for protection against S . aureus infection [9 , 11–13] . Using the same approach we compared the adaptive immune response to wild type S . pyogenes M1 ( SF370 ) vs its complete phage KO ( CEM1ΔΦ ) [19] . Similar to S . aureus Newman and its KO , most of the T cell responses to the S . pyogenes strain M1 originate from genes found in the M1 accessory phage genome ( Fig 3F and 3H ) . However , in contrast to Newman , a prominent portion of B cell activation is also attributed to the net effect of the accessory genome ( Fig 3G ) . This suggests that the influence of the accessory genome is not confined to T cell responses alone , but rather , depending on the species , both arms of the adaptive immune response may be regulated by the bacterial accessory genome . Moreover , the heterogeneity of the adaptive immune response to various strains of a species may imply that there is no single immune signature ( or Th ) that represents a bacterial species . However , defining the adaptive immune response to a species by its “core genome” , whether practical or not , may be a more suitable alternative for comparing species ( see Fig 3I ) . Can our observed immune response variability to different strains explain part of the inter-individual differences found in response to infection by a specific species , and if so what might this mean clinically ? Comparing the Coefficient of Variation of T cell proliferation across donors to that across strains ( Fig 4A ) and of three other immune adaptive read outs ( B cell proliferation , IFNγ , and IgG expressing cells , Fig 4B , 4C and 4D ) demonstrates that the contribution of inter-strain variability to the final immune response is at least as large as the contribution of inter-individual variability ( see Components of Variance in S8 Table ) . Moreover , the magnitude of the observed difference in Th17 and Th1 cell induction following stimulation with either Newman or USA600 ( Fig 2 ) is as large as the difference previously described in S . aureus-infected patients with primary [9 , 12] or secondary [13] immune deficiency conditions , respectively , compared to controls . Taken together , the latter two findings further suggest that the clinical heterogeneity observed during infection by a bacterial species may be contributed by the wide variability in the acute adaptive immune response we found in response to individual strains within a species . These results might further fuel the debate between those that emphasize the contribution of human immune variation in determining susceptibility to infections [20] and those who support bacterial genome versatility and diversification as the cause . In the current study , we evaluated the net effect of the human adaptive immune system 4 days after stimulation , in an attempt to simulate , within the limits of the in-vitro system , the way the adaptive immune system would react to the various strains during an acute infection . The observed wide inter-individual differences in the adaptive responses to an acute infection may represent a combination of the variability in the activation of memory cells ( from previous exposure/s to the bacteria ) , and the activation of new effector cells from naïve cells upon the new encounter with the pathogen . In either case we are evaluating the net adaptive effect of these two populations ( naïve and memory ) and the combined differential activation by the various strains . In conclusion , this study demonstrates that the prominent adaptive immune heterogeneity in response to various strains of a bacterial species , as well as the large difference in induction of what is considered to be protective Th1/Th17 immunity [9 , 11–13] , is contributed by the bacterial accessory genome known to contain a unique combination of virulence determinants and superantigens . Thus , the net immune effect of a strain’s accessory genome “added” to the basic immune response to its "core" ( common to all strains of a species ) becomes the total unique response to a strain . However , the exact role that the additional response plays towards patient outcome is at this time unknown . Blood from adult healthy donors was obtained at the Rockefeller University hospital . Peripheral blood mononuclear cells ( PBMCs ) were isolated with Ficoll-Paque Plus ( GE Healthcare ) , after which the cells were stained with Carboxyfluorescein succinimidyl ester ( CFSE ) and stimulated with heat-killed bacteria in the presence of anti human CD28 ( eBioscience , clone 28 . 2 ) . 4 days later the cells were restimulated for 6 hrs before harvesting and evaluating for proliferating cells . The cultured CFSE stained cells were incubated with Brefeldin A ( Biolegend ) concomitant with their restimulation . 6 hrs later the cells were harvested and stained . For T cell and B cell activation cells were harvested and stained as follow: with Aqua Live/Dead ( Life Technologies ) anti-human CD3 ( PerCP5 . 5 ) , CD4 ( Alex700 ) , CD19 ( PE ) , IgG ( BV421 ) . The cell were fixed and permeabilized using BD Cytofix/Cytoperm kit according to the manufacturer’s protocol , after which intra-cellular staining was done with anti-human IFNγ ( APC ) ( all fluorochrome-labeled antibodies from Biolegend ) . FACS analysis was gated on Live CD3+CD4+CFSElow for proliferating T cells , or Live CD3-CD19+ CFSElow for B cell proliferation . Cells were analyzed by FACS with BD LSRII . The following cell surface markers for T helper ( Th ) were used CxCR3 ( BV421 ) , CCR4 ( APC ) , and CCR6 ( PE ) . LIVE CD3+CD4+ cells were gated on CXCR3+CCR4–CCR6– cells ( defined as Th1 ) , CCR6+CCR4+CXCR3– ( defined as Th17 ) , CCR6+CXCR3+CCR4– ( defined as Th1/Th17 ) , and on CFSElow for proliferation . The following fluorochrome-labeled antibodies were used for intra cellular staining for cytokines or transcription factors: IL-17A ( BV421 ) , IFNγ ( APC ) , Tbet ( APC ) , all from Biolegend , IL17F ( PE , from eBioscience ) , RORγt ( BD Pharmingen ) . FACS analysis was done with BD LSRII at our core facility . We used 16 strains of S . aureus . All are clinical isolates ( except the lab strain RN4220 ) that are part of our lab collection obtained from NARSA . The Streptococcus pyogenes strains that were used are part of our laboratory’s Lancefield collection . S . aureus and S . pyogenes strains were grown over night at 37°C in Tryptic soy ( BD ) or Todd-Hewitt broth plus 1% Yeast extract , respectively , after which 1:100 dilution of each strain was grown to OD600 0 . 5 in 20 ml of media . This OD was chosen because the bacteria are at their maximal proliferation stage and before secretion of virulent factors and superantigens [21] . To remove residual virulent factors in the media , the 20 ml of bacterial growth was adjusted to 50 ml with PBS , centrifuged and the bacterial pellet was washed with 50 ml of PBS , then resuspended in PBS to 1010 bacteria per ml ( based on serial dilution and plating ) and were heat killed at 80°C for 1 hr . All strains were maintained in aliquots at -20°C before use . Statistical analysis was performed with the Prism software ( GraphPad ) . Data represent means ± SEM values , and significance was assessed by nonparametric Mann Whitney test or Friedman test with post-hoc Dunn’s multiple comparison tests . MANOVA test for bivariate analysis was performed with R statistical software . Coefficient of Variation and Components of Variance approach were used to estimate the contribution of donors and strains to the total variability .	We address a fundamental question regarding the variability in human susceptibility to bacterial infection . Recent attention has been directed to human immune variation as the major element in infection susceptibility , however , our study focuses on the contribution of the pathogenic bacteria . To separate the contribution of bacterial versus human variability , we compared the acute adaptive immune response ( both T and B cells ) induced by several strains within two human disease-causing bacterial species in the same individual , and verified in 10 other humans . Our results clearly demonstrate that different strains of a species vary widely in the acute adaptive immune response they induce . Most importantly , using mutant strains , we show that the observed variability is a function of the differences found in the bacterial accessory genome , i . e . , its lysogen . Previous publications generally base their results on a single strain as the representative of a species . Our findings extend and modify this notion by implying that the adaptive immune response signature to a pathogenic species should be defined either per strain or alternatively to the species’ ‘core genome , common to all of its strains . The results , while novel and unexpected , may set the stage to ultimately better predict patient disease outcome .	[ "Abstract", "Introduction", "Results/Discussion", "Materials", "and", "methods" ]	[ "blood", "cells", "bacteriology", "medicine", "and", "health", "sciences", "immune", "cells", "pathology", "and", "laboratory", "medicine", "pathogens", "immunology", "microbiology", "staphylococcus", "aureus", "streptococcus", "pyogenes", "bacterial", "genetics", "microb...	2018	Strains of bacterial species induce a greatly varied acute adaptive immune response: The contribution of the accessory genome
Telomerase is a ribonucleoprotein complex involved in the maintenance of telomeres , a protective structure at the distal ends of chromosomes . The enzyme complex contains two main components , telomerase reverse transcriptase ( TERT ) , the catalytic subunit , and telomerase RNA ( TR ) , which serves as a template for the addition of telomeric repeats ( TTAGGG ) n . Marek's disease virus ( MDV ) , an oncogenic herpesvirus inducing fatal lymphoma in chickens , encodes a TR homologue , viral TR ( vTR ) , which significantly contributes to MDV-induced lymphomagenesis . As recent studies have suggested that TRs possess functions independently of telomerase activity , we investigated if the tumor-promoting properties of MDV vTR are dependent on formation of a functional telomerase complex . The P6 . 1 stem-loop of TR is known to mediate TR-TERT complex formation and we show here that interaction of vTR with TERT and , consequently , telomerase activity was efficiently abrogated by the disruption of the vTR P6 . 1 stem-loop ( P6 . 1mut ) . Recombinant MDV carrying the P6 . 1mut stem-loop mutation were generated and tested for their behavior in the natural host in vivo . In contrast to viruses lacking vTR , all animals infected with the P6 . 1mut viruses developed MDV-induced lymphomas , but onset of tumor formation was significantly delayed . P6 . 1mut viruses induced enhanced metastasis , indicating functionality of non-complexed vTR in tumor dissemination . We discovered that RPL22 , a cellular factor involved in T-cell development and virus-induced transformation , directly interacts with wild-type and mutant vTR and is , consequently , relocalized to the nucleoplasm . Our study provides the first evidence that expression of TR , in this case encoded by a herpesvirus , is pro-oncogenic in the absence of telomerase activity . Telomerase is a multi-component ribonucleoprotein complex . One of its main functions is the maintenance of telomeres , a protective structure at the termini of linear chromosomes . The telomerase complex consists of two essential core components , telomerase reverse transcriptase ( TERT ) and telomerase RNA ( TR ) , which serves as a template for the catalytically active subunit in the elongation of telomeric repeats ( TTAGGG ) n at the end of chromosomes [1] . TR contains four structural domains , which are highly conserved regions ( CR ) in all vertebrates: I ) the pseudoknot ( core ) domain , containing the template sequence ( CR1 ) ; II ) the H/ACA box and III ) the conserved region ( CR ) 7 domain , both of which are essential for TR stability and localization; IV ) the CR4-CR5 domain , which is required for efficient TR-TERT complex formation , hence telomerase activity and processivity [2] , [3] . An essential structure within the CR4-CR5 domain is the P6 . 1 stem-loop . Base pairing of the P6 . 1 stem is completely conserved in all vertebrates . Disruption of the base paring of the P6 . 1 stem was shown to interfere with proper TR-TERT interaction and resulted in absence of telomerase activity in vitro and in vivo [3]–[5] . In addition , the P6 . 1 stem-loop was shown to interact with conserved sequences of the template region CR1 , which also plays a critical role in the catalytic activity of the telomerase complex [5] . Telomerase activity is tightly regulated and varies amongst cell types . While it is commonly up-regulated in germ-line , stem and cancer cells , it is absent in most somatic cells [6] . The absence of telomerase activity often leads to progressive telomere shortening , known to initiate cellular senescence and irreversible cell cycle arrest . Several tumor-inducing viruses have evolved strategies to evade and subvert this mechanism of cellular senescence , mainly via the up-regulation of TERT , which was shown to be the limiting factor of telomerase activity in some organisms , such as the human and the chicken [7] , [8] . It has been suggested that up-regulation of TERT expression and provision of more active telomerase increases the proliferative potential of persistently infected cells , which in turn might be beneficial to accumulate genetic alterations and transformation after infection [8] . One of the most remarkable viruses with respect to the efficiency of the induction of fatal tumors is Marek's disease virus ( MDV ) , a lymphotropic alphaherpesvirus , that causes Marek's disease ( MD ) in chickens , characterized by neurological disorders , immune suppression and , primarily , malignant T cell lymphomas [9] . The rapid onset of MD-induced lymphomas , as early as 2 weeks post-infection , and high tumor-induced mortality ( 90–100% in susceptible animals ) , suggests a direct involvement of virus-encoded oncogenes in the process . The major MDV oncogene , meq , encodes a basic leucine zipper ( bZIP ) transcription factor ( TF ) that was shown to interact with Rb , cdk2 and p53 , proteins involved in cell-cycle control , and several cellular TFs including c-Jun , c-Fos and c-Myc , an oncogene known to regulate TERT expression [10] , [11] . In addition , and as a unique feature , the MDV genome harbors two copies of its own TR subunit , termed viral TR ( vTR ) , that shares 88% sequence identity with chicken TR ( chTR ) , contains all four conserved structural TR domains , and was likely acquired from the chicken genome [2] . The 180-kbp linear , double-stranded DNA genome of MDV consists of a long ( L ) and short ( S ) unique region ( UL and US ) flanked by terminal ( TRL or TRS ) and internal ( IRL or IRS ) inverted repeats . Both vTR copies are located in the repeats flanking the UL , TRL and IRL . Besides the presence of vTR in the TRL and IRL , MDV also contains two sets of tandem repeats in very close proximity to the genomic termini that represent perfect telomeres [12] . vTR is expressed during both lytic and latent MDV infection . It is functionally active and was shown to more efficiently induce telomerase activity in vitro when compared to its cellular homologue , chTR [13] , [14] . Although dispensable for lytic replication in vitro and in vivo , vTR is required for efficient MDV-induced tumorigenesis , as MDV mutants lacking both copies of vTR were severely impaired in lymphoma formation and dissemination [14] . Recent reports suggest that both TERT and TR may also have roles in tumorigenesis aside from their role in the maintenance of telomere length in rapidly dividing cells [7] , [15] , [16] . For example , human TR has been shown to restrain activity of ATR , a factor in the DNA damage response pathway , in a telomerase-independent fashion allowing the survival of cells after cellular stress such as UV radiation [17] . Furthermore , knockdown of TR in human cancer cells induced rapid changes in the global gene expression profiles that were independent of telomere maintenance and DNA damage responses . Induced changes in expression levels included genes involved in cell cycle progression ( Cyclin G2 and Cdc27 ) and adhesion ( integrin αV ) , that may have an effect on MDV pathogenesis and tumorigenesis as well [15] . Similarly , expression of vTR in the chicken fibroblast DF-1 cell line that does not exhibit telomerase activity , induced a 2-fold increase of integrin αV expression , suggesting a telomerase-independent function for vTR [14] , [15] , [18] . One potential interaction partner of vTR is ribosomal protein L22 ( RPL22 ) , previously shown to interact with human TR [19] . Besides associating with ribosomes , RPL22 is also involved in the development of T-cells [20] , [21] , the target of MDV transformation . Epstein-Barr virus ( EBV ) , a herpesvirus that shares many pathobiological similarities with MDV , encodes two small RNAs , termed EBER-1 and EBER-2 , that contribute to tumor formation and are highly abundant in latently infected cells [22] . EBER-1 was shown to interact with RPL22 and the interaction resulted in relocalization of RPL22 from the nucleolus to the nucleoplasm . The interaction of EBER-1with RPL22 is associated with enhanced potential for cellular proliferation [22] , [23] . In order to elucidate whether MDV vTR has functions that are independent of telomere maintenance and its presence in the telomerase complex , we mutated the P6 . 1 stem-loop present in CR4-CR5 of MDV-encoded vTR . The mutation was shown to efficiently abrogate vTR-mediated telomerase activity in vitro . In vivo studies analyzing MD incidence , tumor development and dissemination confirmed that vTR serves functions that are both dependent and independent of the formation of an active telomerase complex . In addition , we identified RPL22 as an interaction partner of vTR and show that it is relocalized upon vTR expression . To our knowledge , the data presented here provide the first in vivo evidence that a TR executes functions important for tumor formation that are independent of telomerase activity and likely depend on the alternate usage of RPL22 in the transformation process . To ensure that the disruption of the vTR P6 . 1 stem-loop , as previously shown for cellular TRs , efficiently abrogates vTR-TERT interaction and consequently telomerase activity , we performed gel-based telomere repeat amplification protocol ( TRAP ) assays . Base pairing of the P6 . 1 stem-loop was disrupted by mutating base pairs ( bp ) 295–298 of vTR from 5′-AGAG-3′ to 5′-UCUC-3′ ( Fig . 1A ) . In order to confirm the absence of telomerase activity via TRAP assay , in vitro transcription was used to generate various vTR's ( Fig . 1B ) that were used in the TRAP assays: wild- type ( wt ) vTR , vTR containing the P6 . 1 mutation ( Fig . 1A ) , or , as a negative control , vTR containing a mutation in the template sequence ( AU5 ) resulting in the addition of ( TATATA ) n repeats that are not amplified in the TRAP assay . Functional chTERT protein was obtained by in vitro transcription of a synthetic cDNA followed by translation using a rabbit reticulocyte lysate system ( Fig . 1C ) . In order to reconstitute the telomerase complex , chTERT was incubated with vTR variants or actin control RNA and telomerase activity analyzed by TRAP assays . While TRAP products were readily detected with wt vTR confirming earlier results [13] , telomerase activity was undetectable when vTR with the P6 . 1 mutation was used , as was evident from the absence of TRAP products in reactions containing P6 . 1mut . Similarly , addition of vTR with a template mutation ( AU5 ) or negative control RNA to the TRAP reaction did not result in telomere elongation ( Fig . 1D ) . Although clearly detectable , few TRAP products were obtained with the vTR-TERT combination . The relatively low activity of reconstituted vTR-TERT compared with the positive control TR could be due to the low TERT levels generated by in vitro transcription/translation , the lack of accessory telomerase factors , or a high protein content of the reticulocyte lysates known to reduce TRAP product generation [24] . Our results clearly demonstrated , however , that the introduced mutation within the vTR P6 . 1 stem-loop completely abrogates the formation of an active telomerase complex . To determine whether the established tumor-promoting function of vTR is dependent on the formation of an enzymatically active telomerase complex , we manipulated the P6 . 1 stem-loop in pRB-1B , an infectious BAC clone of the highly oncogenic RB-1B MDV strain ( Fig . 2A ) [25] . Base pairing of the P6 . 1 stem-loop was disrupted by mutating base pairs ( bp ) 295–298 of vTR , as described above , via two-step Red-mediated mutagenesis [26] ( Fig . 1A ) . Two rounds of identical mutagenesis allowed the desired alteration of both copies of the diploid vTR gene within the MDV genome , and the resulting mutant infectious clone was termed pP6 . 1mut . In addition , a revertant BAC clone ( pP6 . 1rev ) was generated in which the original sequence was restored in both alleles . All clones were confirmed by PCR , DNA sequencing and multiple restriction fragment length polymorphism analyses ( RFLP ) to ensure the integrity of the genome ( Fig . 2B ) . In order to confirm that the mutation did not revert during any of the experimental procedures , DNA of stock viruses used for infection of the animals and viral DNA obtained from tumor cells were analyzed by nucleotide sequencing , which demonstrated that the vTR mutants were genetically stable throughout the experiments . In order to investigate the effect of the P6 . 1 stem-loop mutation on virus replication in vitro , wt pRB-1B , pP6 . 1mut and pP6 . 1rev BACs were transfected into chicken embryo cells ( CEC ) resulting in the reconstitution of recombinant viruses termed vRB-1B , vP6 . 1mut and vP6 . 1rev . Multi-step growth kinetics revealed that replication of vP6 . 1mut was unaffected in vitro when compared to that of wt vRB-1B or vP6 . 1rev ( Fig . 3A ) . In addition , mutation of the P6 . 1 stem-loop had no effect on the plaque sizes induced by the vP6 . 1 virus mutant ( Fig . 3B ) . These findings were consistent with previous data on vTR deficient viruses , which had shown that vTR is dispensable for lytic virus growth in vitro [14] . Since efficient lytic replication in vivo is considered a prerequisite for efficient lymphomagenesis , we analyzed the replicative potential of the various mutant viruses in the natural host . We infected 1-day-old chickens and monitored virus levels by qPCR using DNA isolated from whole blood obtained by wing vein puncture until 28 days post infection ( dpi ) . MDV is present in peripheral blood mononuclear cells ( PBMC ) and qPCR analyses showed that vP6 . 1mut replicated in those cells to levels that were comparable to those of wt vRB-1B or vP6 . 1rev ( Fig . 4A ) . The results were again consistent with published data on the lytic replication of vTR deficient viruses , which were shown to be fully capable of robust lytic replication [14] . The observed dispensability for lytic replication of vTR-TERT interaction and vTR-mediated telomerase activity in general can be explained by the fact that the initial virus production in chicken B and T cells does not require long-term survival of the host cell or host cell proliferation . Survival of the latently infected host cell is , however , a prerequisite for , or consequence of transformation and tumor formation in general . From the results of the experiments on lytic replication of the P6 . 1mut viruses we concluded that viruses containing the P6 . 1 stem-loop mutation are capable of efficient replication in cultured cells in vitro as well as in the target cells in vivo . Therefore , vTR-TERT interaction mediated by the P6 . 1 stem-loop and , therefore , telomerase activity is dispensable for MDV replication in vivo . We have previously shown that MD lymphoma formation was significantly reduced in the absence of vTR [14] . To address whether the observed reduction is dependent on the interaction of vTR with TERT , we performed two independent animal experiments in which we monitored the temporal occurrence of virus-induced lymphoma in chickens infected with vRB-1B , vP6 . 1mut or P6 . 1rev . In a first animal experiment , we established that abrogation of vTR-mediated telomerase activity markedly delayed the onset of MD lymphomas ( Fig . 4B ) . We observed that the time until development of tumors and occurrence of MD in 50% of the infected animals ( MD50 ) was increased from 36 dpi in vRB-1B-infected chickens ( n = 5 ) and 32 dpi in vP6 . 1rev-infected animals ( n = 8 ) up to 46 dpi in vP6 . 1mut-infected birds ( n = 10 ) . In a second animal experiment , the clinician examining inoculated chickens was blinded to eliminate subjectivity . In agreement with the results of the first animal experiment , MD50 was significantly delayed in vP6 . 1mut infected chickens ( 49 dpi , n = 22 ) when compared to vP6 . 1rev expressing wt vTR ( 32 dpi , n = 20 ) ( p = 0 . 0012 ) . We hypothesize that the observed delay in the development of lymphomas is caused by curtailing telomerase activity mediated by vTR and , consequently , the absence of enhanced telomere maintenance . Such enhanced telomere maintenance , which was shown in MDV-infected animals [27] and is thought to play an important role for the survival of rapidly dividing MDV-transformed cells early in the transformation process , is probably mediated mainly by an interaction between vTR and cellular chTERT . In the absence of the P6 . 1 stem-loop , the interaction can no longer occur , and , therefore , the pool of transformed cancer stem cells surviving the initial crisis may be reduced . It was notable , however , that , in contrast to viruses lacking vTR [14] , all animals infected with v6 . 1mut succumbed to MD before termination of the experiment , indicating that vTR has functions independent of the formation of an active telomerase complex . From the results of the animal experiments we conclude that the rapid onset of MD observed in chickens infected with wt MDV ( vRB-1B ) or the vP6 . 1rev virus is dependent on telomerase activity that involves vTR-chTERT interaction . Lymphoma formation , however , and fatal disease outcome are efficient even in the absence of enzymatically active telomerase . MDV-induced tumor formation and metastasis were previously shown be significantly reduced in the absence of vTR [14] . In addition , our earlier findings of integrin αV up-regulation mediated by vTR alone suggested that malignant lymphoma dissemination may be a result of the action of vTR that is independent of vTR-TERT interaction [14] . To address whether animals infected with the P6 . 1mut virus , where vTR-TERT complex formation is absent and , hence , more non-complexed vTR is available , would corroborate these earlier findings . We enumerated the gross lesions in infected birds during necropsies on animals that had succumbed to infection . Consistent with our earlier results and the hypothesis that lymphomagenesis and metastasis could be largely determined by vTR action alone , disruption of the P6 . 1 stem-loop led to a significant increase in the number of solid lymphomas in chickens infected with the vP6 . 1mut virus when compared to vP6 . 1rev-infected chickens ( p = 0 . 0016 ) . All vP6 . 1mut-infected animals developed gross tumors in at least three organs ( Fig . 5A ) . Furthermore , the average number of organs with solid lymphomas was mildly albeit significantly increased from 3 . 1 in vP6 . 1rev to 4 . 0 in vP6 . 1mut ( p = 0 . 0381; Fig . 5B ) . We concluded , therefore , that efficient tumor dissemination observed in vP6 . 1mut-infected animals supported the results of a previous study suggesting that vTR is involved in increased lymphoma dissemination and metastasis [14] . In order to address whether telomerase activity was affected in tumor cells derived from vP6 . 1mut-infected animals , we performed quantitative CY5 gel-based TRAP assays as described by Herbert and coworkers [28] . The experiment showed that telomerase activity was not affected in primary tumor cells derived from vP6 . 1mut-infected animals when compared to tumor cells recovered from animals infected with parental vRB-1B virus , suggesting that endogenous TR in transformed T-cells can compensate for the telomerase activity mediated by vTR . In addition , we analyzed established , clonal LCLs derived from animals infected with vRB-1B or virus containing the P6 . 1 stem-loop mutation . vP6 . 1mut-derived cell lines exhibited telomerase activity comparable to those transformed with wild-type vRB-1B . To address whether vTR contributes to telomerase activity during MDV transformation in vitro , we performed TRAP assays using CU91 , a retrovirus-transformed T-cell line obtained from chickens with the same genetic background ( B19B19 ) as the animals from which cell lines after infection with vRB-1B or vP6 . 1mut were derived . Similarly , the CU210 cell line was used , which was generated by superinfection of CU91 with MDV strain RB-1B [33] . Latent MDV infection in CU210 and , hence , vTR expression , did not increase telomerase activity when compared to the parental CU91 cell line , which showed higher telomerase levels when compared to the MDV-derived cell lines . The high telomerase activity of MSB-1 , an MDV-transformed and highly passaged LCL , suggested that such serial passage might select for increased telomerase activity that likely contributes to a profound transformation phenotype that is reflected by very robust proliferation observed for MSB-1 [29] . Taken together , our data suggested that vTR does not contribute to telomerase activity in MDV-transformed tumor cells , which further lends support to the hypothesis of telomerase-independent functions of vTR in the development and dissemination of lymphoma . As previously reported , EBV transformation mediated by EBER-1 is dependent on its interaction with RPL22 [22] . In order to determine if wild-type and/or mutant P6 . 1 vTR interact with RPL22 , we performed biotin-RNA pull-down assays . vTR , vTR P6 . 1 , chTR and EBER-1 were found to precipitate RPL22 , while biotin-labeled β-actin control RNA did not ( Fig . 5C ) . EBER-1 showed the strongest interaction that was 5 . 1-fold stronger than that determined for vTR ( Fig . 5D ) , potentially because it contains three independent RPL22 binding sites [30] . The interaction of chTR with RPL22 was reduced by 2 . 0-fold , indicating that cellular TR does not interact as strongly as that encoded by MDV . vTR P6 . 1 showed a 1 . 9-fold increase in precipitated RPL22 when compared to wild-type vTR . This apparently enhanced interaction could be caused by a conformational change of the P6 stem-loop structure , that exhibits high similarity to the EBER-1 stem-loop 3 that is known to interact with RPL22 [30] . In addition , the abrogation of vTR-TERT interaction in MDV infected cells could increase the amount of free vTR available for RPL22 interaction , which may provide an explanation for the increased number of solid tumors found in vP6 . 1mut infected animals . Finally , to address whether vTR expression has an affect on the localization of RPL22 , we determined RPL22 localization by confocal microscopy in vTR-transfected cells . HeLa cells were co-transfected with an expression plasmid encoding RPL22 that is C-terminally tagged with mRFP ( RPL22-mRFP ) and plasmids expressing EBER-1 , vTR , P6 . 1mut or chTR . In cells transfected with RPL22-mRFP alone or together with empty vector , RPL22-mRFP localized almost exclusively to the nucleolus as described previously [31] . As shown in previous reports , EBER-1 expression induced relocalization of RPL22; however , RPL22 was not completely absent from the nucleoli , which could possibly be attributed to lower amounts of EBER-1 plasmid DNA used here when compared to earlier reports ( Fig . 6C ) . A relocalization of RPL22 quite similar to that following EBER-1 expression was also detected for vTR , P6 . 1mut and chTR , suggesting that over-expression of viral as well as cellular TR affects RPL22 subcellular distribution ( Fig . 6C–E ) . Cells with a nucleolar localization of RPL22 were quantified to confirm that the relocalization is a general and not isolated event . As in cells transfected with EBER-1 , the number of cells with nucleolar RPL22 localization was clearly reduced after co-expression of vTR , P6 . 1mut or chTR . Under the conditions used here , efficiency of relocalization of RPL22 was comparable between EBER-1 and the vTR and chTR constructs ( Fig . 6 ) , which may suggest that EBER-1 and vTR serve similar purposes in the process of transformation of human and chicken lymphocytes . In this report , we demonstrate that the herpesvirus telomerase RNA , vTR , has at least two functions in virus-induced lymphomagenesis . One of its functions is dependent on vTR-TERT interaction , while the other is independent of the formation of an active telomerase complex . The rapid onset of lymphoma formation seems dependent on vTR-mediated telomerase activity because a delay in the development of tumors was observed when vTR-TERT interaction was abrogated . The documented increase in telomerase activity mediated by the presence of vTR in complex with TERT when compared to the presence of cellular TR likely plays an important role in the initial establishment and maintenance of MDV-transformed cells . It may , therefore , facilitate the development of lymphomas by increasing the pool of candidate tumor stem cells ( Fig . 7 ) . Functions of vTR that are independent of telomerase activity , however , are needed later in the process and influence homing of tumor cells to various organs , seeding and metastasis . These processes are likely a consequence of TR-mediated gene regulation [15] and the interaction of vTR with RPL22 suggests an alternative mechanism involved in transformation that may be similar to that demonstrated for EBV EBER-1 ( Fig . 7 ) . In conclusion , our study demonstrates that TR is directly involved in tumor formation in vivo , in a fashion that is independent of its function as an integral component of an active telomerase complex . All animal work was conducted at Cornell University according to national regulations . The animal care facilities and programs of Cornell University meet the requirements of the law ( 89–544 , 91–579 , 94–276 ) and NIH regulations on laboratory animals , and are in compliance with the Animal Welfare Act , PL 279 . The College of Veterinary Medicine at Cornell University is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care . All experimental procedures were in compliance with approval of Cornell University's Institutional Animal Care and Use Committee ( IACUC , internal approval number: 2002-0085 ) . MDV transformed lymphoblastoid T cell lines ( LCL ) were generated as described previously and cultivated in RPMI medium 1640 plus 10% FBS and 8% chicken serum at 41°C in a humidified atmosphere of 5% CO2 [32] , [33] . The MSB-1 cell line was kindly provided by Mark S . Parcells ( University of Delaware , Newark , DE ) whereas the CU91 and CU210 [34] cell lines were kindly provided by Karel A . Schat ( Cornell University , Ithaca , NY ) . CECs were prepared from specific-pathogen-free embryos and maintained as described previously [35] . Recombinant viruses were reconstituted in CECs by CaPO4 transfection of purified BAC DNA as described previously [36] , [37] . The lox-P-flanked mini-F sequences within the infectious clones were removed by cotransfection with a Cre recombinase expression vector ( pCAGGS-NLS/Cre ) [36] . Removal of the mini-F sequences was ensured by analyzing recombinant virus stocks by analytic PCR as described previously [36] . Virus propagation as well as determination of virus growth kinetics and plaque sizes were performed as described previously [38] . pP6 . 1mut and pP6 . 1rev were generated by two-step Red-mediated recombination [26] , [36] . Primers used for the mutagenesis are given in Table 1 . SPF P2a ( MHC: B19B19 ) chickens were inoculated intra-abdominally with 500 to 2 , 000 plaque-forming units at day 1 of age and housed in isolation units . All experimental procedures were conducted in compliance with approved Institutional Animal Care and Use Committee ( IACUC ) protocols ( internal approval number: 2002-0085 ) . Chickens were evaluated for symptoms of MDV-induced disease on a daily basis and examined for gross tumors when clinical symptoms were evident . DNA was extracted from whole blood and MDV genomic copies were determined by qPCR [39] , [40] . Briefly , MDV DNA copy numbers were detected using primers and probe specific for the ICP4 locus and normalization was achieved using chicken inducible nitric oxide synthase ( iNOS ) genome copies . vTR was amplified from pRB-1B and subsequently cloned into the PstI and XbaI sites of the pUC119 plasmid resulting in plasmid pUC119-vTR . The T7 promoter was inserted at the 5′ end of vTR via a 5′ overhang in the vTR-T7-for primer . vTR was also cloned into the EcoRI and BamHI sites of the pCMS-EGFP plasmid resulting in plasmid and pCMS-vTR . Mutation of the template ( AU5 ) and the P6 . 1 stem-loop was done based on pUC119-vTR and pCMS-vTR by Phusion Site-Directed Mutagenesis ( Finnzymes Inc . ) according to the supplier's instructions and resulting in pUC119-vTR-AU5 , pUC119-vTR-P6 . 1 and pCMS-vTR-P6 . 1 respectively . Chicken TERT ( chTERT ) was obtained as a synthetic , codon-optimized sequence from GenScript ( Piscataway , NJ USA ) , PCR amplified including an upstream Kozak sequence and inserted into pcDNA3 . 1/V5-His TOPO ( Invitrogen ) containing a 5′ T7 promoter , resulting in plasmid pcDNA-chTERT-His . Chicken TR ( chTR ) and RPL22 was amplified from chicken DNA and inserted into pcDNA3 . 1/V5-His TOPO , resulting in pcDNA-chTR and pcDNA-RPL22-His . For the expression of fluorescently labeled RPL22 , we amplified mRFP from pmRFP-1 [41] and inserted it into the XhoI site of pcDNA-RPL22-His resulting in pcDNA-RPL22-mRFP . Oligonucleotides used for amplification are given in Table 1 . vTR variants , chTR , EBER-1 , or β-actin were transcribed using the Maxiscript T7 kit ( Ambion ) following the manufacturer's instructions where the linearized plasmids pUC119-vTR , pUC119-vTR-AU5 , pUC119-vTR-P6 . 1 , cDNA-chTR , pSG5-EBER-1 ( a kind gift of Dr . Rona Scott , Louisiana State University Health Science Center , Shreveport , LA ) and pTRI-β-actin ( Ambion ) served as templates . Biotin-labeled RNAs were generated using the biotin RNA labeling mix ( Roche ) . chTERT and RPL22 were transcribed via the mMESSAGE mMACHINE T7 Kit ( Ambion ) according to the supplier's recommendation using linearized pcDNA-chTERT-His or pcDNA-RPL22 as templates . RNAs were purified via the RNeasy Kit ( Qiagen ) , analyzed on a 2% denaturing agarose-formaldehyde gel , and quantified with a NanoDrop 1000 ( Thermo Scientific ) . In vitro transcribed chTERT-His or RPL22-His RNA was used for in vitro translation using the Rabbit Reticulocyte Lysate System ( Promega ) according to the manufacturer's protocol . chTERT-His and RPL22-His expression was analyzed by western blotting , using a mouse anti-5xHis antibody ( Qiagen ) . In vitro-transcribed vTR variants ( 1 µg ) were incubated with 1 µL of in vitro translated chTERT for 1 h at 30°C to reconstitute the telomerase complex . Telomerase activity was subsequently determined using the TRAPeze gel-based telomerase detection kit S7700 ( Chemicon ) following the manufacturer's instructions or the CY5 gel-based TRAP assay as described by Herbert and coworkers [28] . 4 µL in vitro translated RPL22-His was mixed with 3 nmol biotin labeled vTR , vTR P6 . 1 , chTR , EBER-1 or β-actin control RNA and incubated in binding buffer ( 150 mM NaCl , 50 mM Tris pH 7 . 0 , 0 . 1% Tween20 , 1 µg of tRNA , 0 . 5 mM DTT , 0 . 5 mM PMSF ) containing 10 µg tRNA for 1 h at 37°C . 20 uL EZview Strepavidin beads ( Sigma ) were washed with binding buffer and added to the setup . After binding occurred for 1 h at RT , supernatant was collected and beads washed 7 times with binding buffer containing 1 µg tRNA . Precipitated and unbound protein was analyzed by western blotting , using a mouse anti-5xHis antibody ( Qiagen ) . 5×104 HeLa cells seeded on coverslips in 24-wells were transfected using Lipofectamin 2000 ( Invitrogen ) with 200 ng pcDNA-RPL22-mRed and 500 ng of either empty vector ( pCMS-EGFP ) , pCMS-vTR , pCMS-vTR P6 . 1 , pCDNA-chTR or pcDNA-RPL22 . At 24 h after transfection , cells were examined using an SP5 confocal microscope system ( Leica ) . Images were taken and RPL22 localization evaluated in at least 160 individual cells per sample . Significant differences in MD incidence were determined using the Wilcoxon rank-sum test ( Fig . 4C ) . Significant differences in tumor distribution were determined using Chi-Square test ( Fig . 5A ) . Significant differences in mean tumor incidences were determined using Student's t test ( Fig . 5B ) .	The enzyme complex telomerase , with its two main components telomerase reverse transcriptase and telomerase RNA , plays an important role in telomere maintenance . Perturbation of telomere length regulation can ultimately result in cellular senescence ( telomere shortening ) and is also observed in tumor cells ( increased telomere maintenance ) . Recent studies suggest telomerase RNAs can function independently of the telomerase complex and promote tumor development independently of telomere maintenance . Here we demonstrate that vTR , a herpesvirus-encoded telomerase RNA , serves two distinct functions in MDV-induced tumor formation . vTR has its first function early after infection , when it is part of the telomerase complex and contributes to the survival of rapidly dividing transformed cells . The second function of vTR is independent of telomerase action and essential for formation of solid lymphomas and metastasis . This latter function is likely a consequence of vTR-mediated gene regulation that is at least in part controlled by its interaction with and relocalization of RPL22 , a cellular factor involved in T-cell development and virus-induced transformation . Taken together , our study demonstrates that telomerase RNA encoded by a herpesvirus is directly involved in tumor formation in vivo in a fashion that is largely independent of its function within an active telomerase complex .	[ "Abstract", "Introduction", "Results/Discussion", "Materials", "and", "Methods" ]	[ "virology/persistence", "and", "latency", "virology/virulence", "factors", "and", "mechanisms", "virology/animal", "models", "of", "infection", "virology", "virology/viruses", "and", "cancer" ]	2010	Herpesvirus Telomerase RNA(vTR)-Dependent Lymphoma Formation Does Not Require Interaction of vTR with Telomerase Reverse Transcriptase (TERT)
Here we present the development and implementation of a genome-wide reverse genetic screen in the budding yeast , Saccharomyces cerevisiae , that couples high-throughput strain growth , robotic RNA isolation and cDNA synthesis , and quantitative PCR to allow for a robust determination of the level of nearly any cellular RNA in the background of 5 , 500 different mutants . As an initial test of this approach , we sought to identify the full complement of factors that impact pre–mRNA splicing . Increasing lines of evidence suggest a relationship between pre–mRNA splicing and other cellular pathways including chromatin remodeling , transcription , and 3′ end processing , yet in many cases the specific proteins responsible for functionally connecting these pathways remain unclear . Moreover , it is unclear whether all pathways that are coupled to splicing have been identified . As expected , our approach sensitively detects pre–mRNA accumulation in the vast majority of strains containing mutations in known splicing factors . Remarkably , however , several additional candidates were found to cause increases in pre–mRNA levels similar to that seen for canonical splicing mutants , none of which had previously been implicated in the splicing pathway . Instead , several of these factors have been previously implicated to play roles in chromatin remodeling , 3′ end processing , and other novel categories . Further analysis of these factors using splicing-sensitive microarrays confirms that deletion of Bdf1 , a factor that links transcription initiation and chromatin remodeling , leads to a global splicing defect , providing evidence for a novel connection between pre–mRNA splicing and this component of the SWR1 complex . By contrast , mutations in 3′ end processing factors such as Cft2 and Yth1 also result in pre–mRNA splicing defects , although only for a subset of transcripts , suggesting that spliceosome assembly in S . cerevisiae may more closely resemble mammalian models of exon-definition . More broadly , our work demonstrates the capacity of this approach to identify novel regulators of various cellular RNAs . The coding portions of most eukaryotic genes are interrupted by non-coding introns which must be removed prior to the translation of their messenger RNAs ( mRNA ) . Removal of introns from pre–mRNAs is catalyzed by the spliceosome , a large and dynamic ribonucleoprotein complex comprised of five small nuclear RNAs ( snRNAs ) and at least 100 proteins [1] . Much of our knowledge about the components that comprise the spliceosome as well as their mechanisms of action has been derived from experiments using the powerful genetic tools available in the budding yeast , Saccharomyces cerevisiae . Indeed , although the RNA2 – RNA11 genes originally identified in Hartwell's forward genetic screen preceded the discovery of splicing [2] , the mechanistic characterizations of these genes , since renamed PRP2 – PRP11 , underlie current models of the splicing pathway . Importantly , because the core components of the spliceosome are highly conserved between budding yeast and humans , the mechanistic details derived from work in yeast have been instrumental in understanding mechanisms of pre–mRNA splicing in higher eukaryotes . The modern view of pre–mRNA splicing acknowledges the integrated role of the spliceosome with several other aspects of RNA processing . Whereas the historical view of splicing envisioned a cascade of temporal events initiated by transcription , followed by polyadenylation , and finalized with splicing and export of mRNAs from the nucleus , it is now clear that these pathways are not independent from one another but rather are functionally coupled . Strong evidence in both yeast and higher eukaryotes demonstrates that recruitment of the spliceosome to intron-containing transcripts occurs co-transcriptionally [3]–[6] , mediated at least in part by physical associations between the C-terminal domain ( CTD ) of RNA polymerase II and the U1 snRNP [7] . A growing body of evidence also indicates that the landscape of chromatin modifications encountered by transcribing polymerase molecules can dictate the activity of the spliceosome at various splice sites . For example , recent work has identified an enrichment of methylated lysine-36 in the histone H3 protein specifically within exonic sequences , suggesting a possible mechanism for facilitating the identification of intron-exon boundaries [8] , [9] . Similarly , the rate of transcription by RNA polymerase II , which can be impacted by chromatin marks , has also been shown to be critical for dictating alternative splicing decisions [10] . Furthermore , it is also clear that splicing is coupled to downstream steps in RNA processing . For example , the yeast Ysh1 protein [11] , [12] , which is the homolog of CPSF73 , the mammalian endonuclease required for 3′ end processing , was originally identified as Brr5 in a cold-sensitive screen for mutants defective in pre–mRNA splicing [13] . Consistent with this observation , recent evidence suggests that transcriptional pausing near the 3′ end of genes is a critical component of pre–mRNA splicing efficiency [14] . Despite the increasing evidence of the interconnectivity of these pathways , in many cases the mechanistic details which underlie these functional relationships remain unclear . Our understanding of these mechanistic connections would benefit from a more complete understanding of the complement of factors through which splicing is connected to these cellular processes . A variety of recent genome-wide approaches have provided important insights into the connections that exist between the spliceosome and other cellular processes . Two powerful approaches , Synthetic Genetic Array ( SGA ) Analysis [15] and Epistatic MiniArray Profiling ( E-MAP ) [16] , leverage genetic tools available in yeast to systematically generate millions of double-mutant strains and then carefully quantitate their cellular fitness to determine an interaction score for every pair-wise mutation . On the basis of strong positive or negative genetic interaction scores these approaches have been successfully used to infer functional relationships between many cellular pathways , including several with pre–mRNA processing [17] , [18] . Simultaneously , improvements in proteomic methodologies have enabled the direct analysis of protein complexes in organisms as diverse as humans and yeast , allowing for an assessment of all of the stably-bound proteins involved in pre–mRNA splicing in many organisms [19] , [20] . While the combination of these and other approaches has provided a global picture of many of the cellular factors that influence the splicing pathway , either directly or indirectly , an important question remains about the functional significance of these factors in the splicing of specific transcripts . Indeed , it has long been known that certain transcripts require the activity of unique accessory factors to facilitate their splicing [21] . Moreover , recent work supports the idea that different transcripts can have a greater or lesser dependence upon the activity of core spliceosomal components for their efficient splicing [22] , [23] . Here we present the results of a novel approach that complements the genetic and physical approaches of others by allowing for a direct functional assessment of nearly every gene in the S . cerevisiae genome in the pre–mRNA splicing process . For this work , we developed automated methods that enabled the isolation of total cellular RNA from about 5500 unique strains , each of which contained a mutation in a single gene , and all of which were examined during exponential growth in liquid medium . Using a high-throughput quantitative PCR ( QPCR ) assay , the relative cellular level of nearly any RNA can be readily determined in the background of each of these strains . By assessing the levels of several different pre–mRNA species , we were able to identify not only those factors which are necessary for the splicing of many transcripts , but also factors that are specifically required for the splicing of a subset of intron-containing genes . Whereas our study specifically examines the levels of several cellular pre–mRNAs , the approach described herein can be easily adapted to study the level of nearly any RNA molecule of interest under a wide variety of cellular growth conditions . To identify the comprehensive network of cellular factors that lead to a change in splicing efficiency , we developed a high-throughput reverse genetic screen that allowed us to readily assess changes in pre–mRNA levels in the background of ∼5500 Saccharomyces cerevisiae strains , each of which contained a mutation in a single gene . The library of strains contained deletions of non-essential genes [24] as well as conditional mutations in essential genes [25] , accounting for mutational access to over 93% of known yeast genes . Using a liquid-handling robot , protocols were developed ( see Materials and Methods ) that allowed for the simultaneous collection of each of these strains under exponential growth conditions in liquid medium in 384-well plates . Total cellular RNA was isolated robotically from each of these strains using a phenol extraction protocol [23] followed by a glass-fiber purification step [26] . After converting this RNA into cDNA using a random-priming strategy , QPCR was used to directly measure the level of a given RNA species within each strain . Because of the inherent variability between the samples in the cell collection , RNA isolation , and cDNA synthesis steps , the levels of six different RNA species were measured in each of the samples in order to calculate a normalization constant . On the basis of this normalization constant , the relative level of virtually any cellular RNA species can be determined in each of the mutant strains . As an initial test of our approach we sought to identify the full complement of factors involved in pre–mRNA splicing by determining the relative levels of unspliced U3 small nucleolar RNA ( snoRNA ) present in each of the mutant strains . The U3 snoRNA is unique in the S . cerevisiae genome in that it is the only known non-coding RNA that is interrupted by a spliceosomal intron [27] . Nevertheless , the U3 transcript has been widely used historically as a splicing reporter , owing to its relatively high basal expression level and the strong accumulation of U3 precursor levels observed in the background of canonical splicing mutants [13] , [28] , [29] . As shown in Figure 1 , the U3 precursor levels are unaffected in the vast majority of the strains examined , with levels varying by less than 1 . 35-fold from one another for 95% of the strains . Indeed , only ∼200 of the ∼5100 strains that passed our quality filters ( see Materials and Methods ) showed a change in the relative U3 precursor levels of more than ∼30% from the median value ( ∼0 . 35 in log2-transformed space ) , consistent with our expectation that mutations in most genes will have little or no effect on cellular pre–mRNA splicing efficiency . The tight distribution of relative U3 precursor levels seen within this dataset demonstrates the high precision with which these measurements can be made , and suggests a low false discovery rate for our approach . To characterize the data generated by this approach we sought to define the biological significance of those strains that showed increased levels of U3 precursor . As an initial analysis , we examined the U3 precursor levels in those strains containing mutations in known splicing genes . Using the GO PROCESS: RNA Splicing as a guide [30] , a total of 71 strains in our library were classified as containing mutations in canonical splicing factors ( Figure S1 ) , of which 68 passed our quality filters for the U3 precursor dataset ( see Materials and Methods ) . A strong overrepresentation of these splicing factors can be seen within the set of strains showing an enrichment of U3 precursor ( Figure 1A ) . Of the 68 strains containing splicing mutations that passed our quality filters: 53 are found within the top 200 strains ( p = 9 . 28E-64 , Fisher's exact test ) ; 38 are found in the top 50 strains ( p = 1 . 33E-66 ) ; and the top 14 strains all belong to this list ( p = 1 . 27E-27 ) . Taken together , these data argue strongly that the candidates identified by this approach will be characterized by a high true positive discovery rate . By contrast , out of the 68 strains containing mutations in known splicing factors for which we obtained high quality data , 15 failed to show an enrichment of U3 precursor levels in this dataset , suggesting either that mutations in these genes don't cause an increase in U3 precursor levels ( true negative ) , or that our approach incorrectly failed to detect the accumulation of unspliced U3 ( false negative ) . To better resolve these possibilities we chose to more completely examine the global splicing fitness of some of these strains using splicing-sensitive microarrays . For every intron-containing gene in the genome , these custom-designed microarrays contain at least three probes ( Figure 2A ) that allow us to distinguish between spliced and unspliced isoforms [31] . We used these microarrays to assess the global splicing defects of four mutants: two canonical splicing mutants that showed strong U3 precursor accumulation ( snt309Δ and lsm6Δ , Figure 2B ) , and two that showed little or no accumulation ( mud2Δ and cus2Δ , Figure 2C ) . As expected , and consistent with previous work from others [32] , the snt309Δ and lsm6Δ strains demonstrate a broad splicing defect , with most intron-containing genes displaying an increase in precursor levels accompanied by a decrease in the amount of spliced mRNA . By contrast , the global splicing profiles of the mud2Δ and cus2Δ strains are markedly different . In the cus2Δ background , few intron-containing genes display a splicing defect: very little precursor accumulation is observed , and there is little if any detectable loss in mature mRNA . The mud2Δ mutation does cause a splicing defect for some intron-containing genes , whereas little change in splicing efficiency is seen for many others . Notably , as seen in Figure 2D , the microarrays of both the snt309Δ and lsm6Δ strains show a strong accumulation of U3 precursor levels , whereas the mud2Δ and cus2Δ strains show almost no accumulation , consistent with our QPCR screen results . It is worth noting that in our experience the behavior of the U3 transcript differs from the other intron-containing genes in that every splicing mutation we have examined that causes an increase in the U3 precursor levels also results in an increase in the total level of U3; the reason for this apparent discrepancy is currently under investigation . Nevertheless , these microarray data demonstrate that our failure to detect an increase in U3 precursor levels in the mud2Δ and cus2Δ strains does not represent a failure of the approach , but rather that these are true negative results . To better assess the total complement of genes that can impact the splicing of any precursor transcript , we chose to expand our analysis by measuring the precursor levels of several additional intron-containing genes . We chose to examine four ‘canonical’ intron-containing genes ( RPL31B , UBC13 , TUB3 and TEF5 ) that vary in terms of intron size , transcriptional frequency , biological function , and the presence or absence of an intron-encoded snoRNA . In spite of these differences , these transcripts are similar to one another in so much as they each contain splice site and branch point sequences that conform to consensus sequences . In addition to these four genes , we chose to examine two intron-containing genes ( YRA1 and REC107/MER2 ) that are known to be poorly spliced under standard growth conditions [21] , [33] , [34]; as such , we expected the behavior of these two transcripts to be distinct from the efficiently spliced transcripts . For all six of these genes , the precursor levels were measured in all ∼5500 strains . As an initial analysis of this data set , we considered the behavior of the 71 strains containing mutations in spliceosomal components ( Figure 3 ) . As expected , precursor accumulation can be detected for each of the canonical intron-containing transcripts in the background of nearly all of the splicing mutations . While all four canonical precursors accumulate in the mud2Δ background , consistent with our microarray data , no precursor accumulation is detected for any of these transcripts in the cus2Δ strain ( Figure 3B ) . In addition , several of the splicing mutants that failed to cause an increase in the U3 precursor levels do cause a splicing defect for these other transcripts . Importantly , the behavior of the Rec107 and Yra1 pre–mRNAs within this subset of strains differs significantly from that seen for the canonically spliced transcripts . Splicing of the Rec107 pre–mRNA shows a strong accumulation in the upf1Δ and upf2Δ strains ( Figure 3C ) , consistent with its known degradation via the nonsense-mediated decay pathway [35] . Because the Rec107 pre–mRNA does not engage the spliceosome during vegetative growth [21] , no precursor accumulation is expected in strains containing spliceosomal mutations . Likewise , the Yra1 pre–mRNA shows a strong accumulation in the edc3Δ strain [34] , consistent with its previously characterized cytoplasmic degradation pathway . The failure to detect Yra1 pre–mRNA accumulation in strains containing spliceosomal mutations presumably reflects the inherently high levels of unspliced Yra1 transcript present in a wild type cell . Taken together , these data strongly support the capacity of this approach to successfully identify mutations that impact pre–mRNA splicing with low false positive and false negative rates of discovery . To expand our analysis beyond previously characterized splicing factors , we sought to identify novel mutations that caused an increase in precursor levels in most , if not all , of our canonical intron-containing genes . By determining the rank order of precursor accumulation in each strain for each of the five canonical splicing substrates ( U3 , Rpl31b , Tef5 , Tub3 , and Ubc13 precursors ) , a composite rank order of each strain was calculated as the average of these independent measurements ( Figure 4 ) . Remarkably , while the majority of the mutations examined cause little or no change in precursor levels of these four transcripts , the subset of mutations which do cause detectable increases in precursor levels is larger for some of the coding mRNAs than was seen for U3 . Interestingly , although there is variation in the number of strains that cause pre–mRNA accumulation of the different transcripts , with Tub3<Tef5<Ubc13∼Rpl31b , strong overlap can nevertheless be identified across the four transcripts . For example , the majority of the strains that cause an increase in the Tub3 pre–mRNA also display an increase in the pre–mRNA levels of the other three transcripts . By contrast , many strains cause a strong accumulation of the Ubc13 and Rpl31b pre–mRNAs without causing a significant change in the Tub3 or Tef5 pre–mRNA levels . Because the absolute levels of the Rpl31b , Tef5 , Tub3 and Ubc13 pre–mRNAs are significantly lower than the U3 precursor levels in most strain backgrounds ( Table S1 ) , we considered the possibility that these results reflected a technical artifact associated with measuring the cellular levels of low abundance RNA species in certain strain backgrounds . Importantly , however , the relative levels of the Rec107 pre–mRNA , whose normal cellular level is similar to these other pre–mRNAs , is largely unchanged in the vast majority of the strains examined ( Figure 4 ) . Likewise , an analysis of the cellular levels of the Faa1 mRNA , an intronless gene whose transcript abundance is of a similar magnitude as the Rpl31b , Tef5 , Tub3 and Ubc13 pre–mRNAs , also shows a nearly constant level in all of the examined strains , further suggesting that there is no inherent bias in detecting low level transcripts . Finally , the Yra1 pre–mRNA , which is inefficiently spliced and has a higher endogenous level than most pre–mRNAs , also shows very little change in the examined strains . Taken together , these results strongly support the conclusion that the levels of the Rpl31b , Tef5 , Tub3 and Ubc13 pre–mRNAs are increased in these strains . Because our approach , as described so far , directly measures the cellular levels of precursor RNA but does not directly determine the efficiency of splicing per se , those mutations which cause an increase in the precursor levels could be doing so simply by increasing the transcriptional frequency of these genes rather than by directly impacting their splicing . To distinguish this possibility from a true splicing defect , we chose to directly calculate the splicing efficiency of the Tef5 transcript by measuring the total cellular level of Tef5 mRNA by QPCR in each strain and using this value to calculate the ratio of unspliced∶spliced RNA in the cell , a classical measure of splicing efficiency . Consistent with a splicing rather than transcriptional cause for precursor accumulation , the measured levels of total Tef5 transcript showed little variation across nearly the entire set of strains ( Figure S2 ) . Indeed , nearly every strain that showed an increase in Tef5 pre–mRNA levels also showed a decrease in the splicing efficiency of the Tef5 transcript ( Figure 4 ) , suggesting that those mutations affect the splicing of this transcript rather than its transcription . These results strongly suggest that the increased pre–mRNA levels observed in these strains largely reflect changes in pre–mRNA splicing . To assess the functional significance of the strains displaying increased pre–mRNA levels , we sought to rule out the possibility that mutations which cause a change in overall cellular fitness might indirectly lead to a decrease in overall splicing efficiency . To test this , we compared our precursor accumulation levels with recently described strain fitness data calculated for each of the 5000 non-essential genes [36] . This comparison yielded no correlation between precursor accumulation and cellular fitness ( Figure S2 ) , suggesting that cellular growth rate alone is insufficient to explain the observed increase in pre–mRNA levels . While the precursor accumulation seen for each of the canonical transcripts in the known splicing mutants lends strong empirical support for the overall robustness of our approach , additional analysis was needed to assess the statistical significance of the data we generated . Towards this end , we employed a statistical approach originally developed for analysis of microarray data called Significance Analysis of Microarrays [37] , or SAM ( see also Materials and Methods ) . We chose this software because , conceptually , the data generated by our QPCR approach are orthogonal to those from a microarray experiment: whereas a microarray experiment examines the behavior of thousands of mRNAs in a single strain , here we examine the behavior of a single RNA in thousands of different strains . Because similar concerns regarding multiple hypothesis testing apply to both types of data [38] , we used this software as a tool for assessing the quality of our data . The results of our SAM analysis were consistent with the qualitative results seen in Figure 4 , in so much as the number of strains causing a statistically significant increase in the levels of each precursor species varied depending upon the precursor mRNA in question . A total of 224 strains caused a statistically significant increase in the Rpl31b pre–mRNA levels , 209 strains caused a significant increase in Ubc13 pre–mRNA levels , 146 strains caused a significant increase in U3 precursor levels , 83 strains caused a significant increase in Tef5 pre–mRNA levels , and 78 strains caused a significant increase in Tub3 pre–mRNA levels . The complete list of SAM-identified strains for each RNA species is provided in Table S2 . Importantly , many of the SAM-identified strains are found to cause a significant enrichment of the precursor levels of all five of these RNAs , including the majority of strains with mutations in canonical splicing factors . Interestingly , for some of the species examined , a small number of strains were identified which showed decreased levels of precursor RNA . In certain instances these reflected expected outcomes: a large decrease in the Ubc13 precursor was identified in the ubc13Δ strain , for example . However , in other cases these may indicate important biological phenomena . For example , both the xrn1Δ and the tfg2Δ strains cause a significant decrease in the U3 precursor levels . We have previously shown that deletion of the Xrn1 nuclease paradoxically leads to decreases in many precursor RNAs [39] , although the mechanism by which this occurs remains unknown . Likewise , it is unclear whether the decreased precursor resulting from deletion of the TFIIF component Tfg2 reflects an overall decrease in transcription of this gene , or whether this in fact reflects increased splicing efficiency perhaps resulting from a decreased transcription elongation rate [10] . To better characterize the factors that impact pre–mRNA splicing , we examined our lists of SAM-identified candidates for factors that are not canonical components of the spliceosome . As an initial approach , we asked whether any functional categories of proteins were statistically overrepresented within this set of strains . For this analysis , we ordered the strains according to the largest precursor accumulation that they affected for any of the RNA species . We then used the GO::Term Finder program [40] to identify overrepresented classes of genes . As expected , when considering the 50 strains that caused the largest precursor accumulations , a strong enrichment for splicing factors was seen with 30 out of 50 strains containing mutations in genes belonging to the GO PROCESS: RNA Splicing category ( p = 1 . 3E-40 with Bonferroni correction ) . Interestingly , when the top 100 strains are considered , significant enrichment can also be seen for strains with mutations in factors belonging to the GO PROCESS: Chromatin Remodeling category , with eight different mutants causing precursor accumulation ( arp5Δ , arp8Δ , bdf1Δ , npl6Δ , rsc2Δ , rsc9-ts , vps72Δ , and yaf9Δ; p = 1 . 5E-03 ) . Expanding our analysis to the top 200 candidates increases the enrichment of this category to include twelve factors ( adding arp6Δ , swc5Δ , swr1Δ , and taf14Δ; p = 1 . 8E-04 ) . Interestingly , within the top 200 candidates , significant enrichment is also seen for the GO PROCESS: RNA Catabolic Process category , with 13 different factors being present ( ccr4Δ , dis3-ts , dbr1Δ , kem1Δ , lsm2-ts , lsm6Δ , lsm7Δ , prp18Δ , rrp6Δ , rtt101Δ , ski3Δ , ssn2Δ , and upf3Δ; p = 8 . 5E-03 ) . Whereas some of these factors , such as lsm2-ts , lsm6Δ , lsm7Δ , and prp18Δ are known to directly function in pre–mRNA splicing , the identification of many of these factors presumably reflects their defects in degradation pathways for unspliced pre–mRNAs . One of the top factors we identified that bridges chromatin remodeling with transcription initiation is the bromodomain factor Bdf1 . Bdf1 is a member of the SWR1 complex and , along with its homolog Bdf2 , has been shown to interact with the TFIID component of RNA polymerase II [41] . Moreover , BDF1 and BDF2 have been demonstrated to be genetically redundant with one another . Whereas our SAM analysis indicated that the bdf1Δ caused a statistically significant accumulation of most of the canonical precursor species in our experiments , the bdf2Δ strain showed little or no detectable increase in the levels of any of the precursors tested ( Figure 5A ) , and was not considered by SAM analysis to be significant for increases in any of the precursor RNAs . To better characterize the global splicing profile of these two mutants , we again turned to our splicing-sensitive microarrays . Remarkably , a dramatic splicing defect can be seen in the bdf1Δ strain for most intron-containing genes , as evidenced by an increase in the precursor transcript levels with a concomitant decrease in the mature and total transcript levels ( Figure 5A ) . By comparison , the bdf2Δ mutation has almost no effect on cellular splicing , strongly corroborating the specific identification of Bdf1 in our screen . To better assess the mechanism by which Bdf1 impacts pre–mRNA splicing , we monitored U1 snRNP recruitment in the background of wild-type , bdf1Δ , and bdf2Δ strains using chromatin immunoprecipitation coupled to QPCR ( ChIP-QPCR ) . As seen in Figure S3 , these experiments show that the deletion of Bdf1 but not of Bdf2 decreases the occupancy of U1snRNP at several intron-containing genes , suggesting impairment of co-transcriptional spliceosomal recruitment in the bdf1Δ strain . A more comprehensive ChIP-Seq experiment will be required to fully characterize the global landscape of genes impacted by the deletion of Bdf1 and further characterize the roles of Bdf1 and Bdf2 in transcription and splicing . We also chose to further examine several factors that our screen identified that are more classically connected with chromatin remodeling . The lower panels of Figure 5B and 5C show the locations within our U3 precursor dataset of all of the strains containing mutations in components of the SWR1 complex , and the RSC complex , respectively . Notably , mutations in many but not all of the components of these complexes cause a splicing defect of the U3 transcript . Moreover , each of the five precursor species that we examined shows a slightly different susceptibility to the different components of these complexes . We chose to examine the global splicing defects of strains containing mutations in two of these components: Vps72 , a member of the SWR1 complex; and Rsc9 , a member of the RSC complex . Splicing-sensitive microarrays of the vps72Δ and rsc9-ts strains , respectively , reveal a splicing defect in each strain ( Figure 5B and 5C ) . However , unlike the bdf1Δ strain , the vps72Δ and rsc9-ts strains cause a splicing defect in only distinct subsets of intron-containing genes . Interestingly , the affected subsets of transcripts are neither completely overlapping nor completely independent of one another; rather the microarray data are consistent with our QPCR data in suggesting that mutations in specific chromatin-modifying components can result in aberrant splicing of specific pre–mRNA transcripts . While an ontology-based approach can successfully identify entire pathways that display enrichment , we were also interested in considering those factors which showed strong pre–mRNA accumulation but whose functional categories were not statistically over-represented at the top of our dataset . Remarkably , while the GO PROCESS: RNA 3′ end Processing wasn't significantly overrepresented as a category within our dataset ( 9 out of top 200 , p = 0 . 09 ) , several strains with mutations in factors belonging to this category resulted in a strong , statistically-significant accumulation of multiple precursor species . Included among these were: yth1-ts , a zinc-finger containing protein that is the homolog of human CPSF-30; cft2-ts , the homolog of human CPSF-100; and fip1-ts , a component of the polyadenylation factor PF I . To further examine the global splicing defects of each of these mutants , microarrays were performed comparing mutant and wild type behavior after shifting them to both elevated and reduced temperatures . Of the three mutants , the profile seen in the yth1-ts mutation most closely resembles a canonical splicing defect , with more than half of the genes showing an increase of precursor and loss of mature RNA ( Figure 6A ) . Interestingly , the splicing defect is strongest at reduced temperatures even though this strain has only a subtle low-temperature growth defect ( not shown ) . By comparison , neither the cft2-ts nor the fip1-ts strains showed a strong splicing defect at low temperature ( not shown ) , but each mutant was characterized by an unusual phenotype at elevated temperatures . As seen in Figure 6B and 6C , two distinct types of behavior are seen in the cft2-ts and fip1-ts mutants , respectively , that are largely defined by whether or not the affected transcript encodes a ribosomal protein gene ( RPG ) . For a subset of the non-RPG transcripts a canonical splicing defect is apparent , consistent with our QPCR results . Interestingly , the subset of affected non-RPG transcripts is different between the two mutant strains . By comparison , nearly all of the RPG transcripts show a dramatic increase in both the mature and total mRNA levels , with little or no detectable change in precursor levels . The strong increases caused by these mutants suggest that the RPG transcripts may be subject to regulatory control at their 3′ ends . Interestingly , while it has long been known that RPG introns are , in general , longer than non-RPG introns [42] , whereas the second exons of RPGs tend to be shorter than non-RPGs [5] , we nevertheless find no strong correlation between either intron or second exon length and the strength of the splicing defect seen for these 3′ end mutants ( data not shown ) . The mechanisms by which these 3′ end factors impact pre–mRNA splicing are currently under investigation . In considering the mechanisms by which candidate factors may be functioning , we sought to determine whether any of the candidates we examined might be indirectly affecting pre–mRNA splicing by changing the cellular levels of known spliceosomal components . Although our splicing-sensitive microarrays were designed primarily to interrogate the splicing status of the ∼300 intron-containing genes in S . cerevisiae , they also contain probes against all ∼6000 protein-coding genes and ∼200 RNA genes , including the spliceosomal snRNAs . Figure S4 shows the relative RNA levels for each of the canonical spliceosomal components , including the snRNAs , in the background of each of the different strains we examined , as determined from our microarray analyses . While these results positively recapitulate the expected changes ( for example , the decreases in Snt309 and Lsm6 mRNA levels in the snt309Δ and lsm6Δ strains , respectively ) , with only a few exceptions , most spliceosomal components appear unchanged in most of the mutants we examined . Importantly , the transcript encoding the Mud1 protein showed dramatic mis-regulation in both the yth1-ts and cft2-ts strains , increasing by more than 10-fold in each background . To test whether Mud1 overexpression might be causing the splicing defects observed in these strains , a strain was constructed where the wild type Mud1 transcript was encoded on a high-copy plasmid . As seen in Figure S5 , in spite of the over 30-fold increase in Mud1 levels in this strain , there is no detectable change in pre–mRNA splicing . Therefore , while the mis-regulation of Mud1 levels in these 3′ end mutants suggests that , similar to its human homolog , Mud1 levels in yeast may be subject to negative regulation via its 3′ end processing [43] , it nevertheless appears that the splicing defect observed in these strains is not a consequence of Mud1 overexpression . Interestingly , several of the strains , including bdf1Δ , yth1-ts , cft2-ts , and fip1-ts , showed an ∼2-fold increase in the levels of both the U1 and U2 snRNAs . Although spliceosomes function as an equimolar complex of all five snRNAs , the total cellular levels of the snRNAs vary: in yeast , the U2 snRNA is the most abundant [44] , while in mammals the U1 snRNA is most abundant [45] . While recent work demonstrates the cellular defects associated with decreased levels of snRNA [46] , it is less clear whether increases in their levels will impart a defect on global splicing . Nevertheless , because each of these strains shows a similar increase in these snRNA levels but distinct splicing defects , it seems unlikely that the changes in snRNA levels alone can explain the observed splicing phenotypes . However , it is not inconceivable that small changes in levels for one or more of these transcripts could lead to the observed splicing defects . As such , additional work will be necessary to determine the functional consequences of these mutations . Here we present the results of a global survey designed to identify the full subset of cellular factors in the budding yeast , Saccharomyces cerevisiae , that impact the efficiency of pre–mRNA splicing . As a complement to other recently described genetic and physical genome-wide approaches , in this work we have developed an approach that allows for a direct readout of the accumulation of specific RNA species in the background of thousands of different mutant strains . An important strength of a genome-wide screen such as this is its unbiased approach . By directly measuring the splicing efficiency of endogenous transcripts , this method avoids bias generated using reporter constructs . Moreover , the ability to examine numerous different transcripts allowed us to distinguish the natural variation in the spliceosomal factors that are required for efficient splicing of different intron-containing transcripts . Indeed , by systematically examining the precursor levels in the background of each strain , mutations can be identified which result in a change in splicing efficiency regardless of their previously described functions . In the work described here , mutations in scores of genes with no previously known role in splicing were identified , some of which impacted the splicing of all five canonical transcripts examined and some of which impacted only a subset of them . While some of these factors have been further characterized and discussed here , many have not ( see Table S2 ) . To be sure , as is the case with all genetic screens , it is impossible on the basis of these screen data alone to ascribe a direct role for any of these candidate factors in the splicing pathway . Rather , the identification of these different factors can be seen as generating a rich dataset from which hypotheses can be generated and tested for their mechanistic underpinnings . Beyond known splicing factors , the most highly over-represented set of factors identified in this work function in chromatin remodeling . One particularly interesting mutation that was identified was the bdf1Δ mutant . In budding yeast , Bdf1 has been demonstrated to play a role precisely at the interface of transcription initiation and chromatin remodeling . Based in part on its physical interaction with the Taf7 subunit of TFIID , yeast Bdf1 has been proposed to function as the missing C-terminal portion of the higher eukaryotic TAFII250 [41] , the largest subunit of the TFIID complex . More recently , it has become clear that Bdf1 interacts with Swr1 and functions in recruiting the entire SWR1 chromatin remodeling complex to nucleosomes . A recent genome-wide study demonstrates that Bdf1 is enriched on the +1 and +2 nucleosomes of actively transcribed genes [47] , and that it coincides with the localization of Vps72 , another component of the SWR1 complex , and another component which was identified in our screen ( Figure 5 ) . Remarkably , we demonstrate here that the splicing of nearly every intron-containing gene is negatively affected in a bdf1Δ strain , and that the quantitative defect seen in this mutant rivals that seen for canonical splicing mutants . Given its role in global gene expression , one possible explanation for our results in the bdf1Δ strain is that the transcription of some key splicing factor is repressed by this mutation , causing a decrease in splicing efficiency . Indeed , early work on Bdf1 from the Séraphin lab suggested a role in global transcription , including transcription of the spliceosomal snRNA genes [48] . However , our microarray analyses show essentially normal RNA levels of all known splicing components in the bdf1Δ strain ( see Figure S4 ) . Moreover , our microarray data assessing the snRNA levels themselves are entirely consistent with Séraphin's original observations and demonstrate that none of the five wild type snRNAs are decreased in cellular level during growth at 30°C in the bdf1Δ mutant; rather , there are subtle increases in the U1 and U2 snRNA levels . Importantly , our ChIP-QPCR experiments in the bdf1Δ strain demonstrate a decreased occupancy of the U1 snRNP on all four intron-containing genes we tested , suggesting the intriguing possibility that Bdf1 plays a direct role in connecting pre–mRNA splicing with chromatin remodeling and transcription initiation . In considering such a role for Bdf1 , it is important to note that the yeast BDF1 gene has a close sequence homolog in the BDF2 gene . These two genes are genetically redundant , in so much as both single gene deletions are viable but the double mutant bdf1Δ/bdf2Δ is lethal . Moreover , it has been shown that these two genes evolved from a single ancestral gene following a whole-genome duplication event [49] . Yet surprisingly , unlike the bdf1Δ strain , the bdf2Δ strain showed no signs of a splicing defect either in our screen or when examined by splicing sensitive microarrays . Moreover , unlike the bdf1Δ strain , there was no apparent decrease in U1 snRNP ChIP-QPCR signal in the bdf2Δ strain . In considering a mechanism whereby Bdf1 connects transcription initiation and chromatin remodeling with pre–mRNA splicing , it is worth noting that , unlike human genes , the majority of yeast genes do not contain an intron . As such , co-transcriptional recruitment of the spliceosome is unnecessary for most yeast genes . We are intrigued by the possibility that , in the time since the duplication event , Bdf2 has evolved to a point where it retains the capacity to recruit RNA polymerase but has lost the ability to efficiently connect splicing with transcription . Such a scenario would explain the differences observed between the bdf1Δ and bdf2Δ microarrays and U1 snRNP ChIP-QPCR data . It would also likely explain the previously published results that Bdf1 shows higher sequence conservation with the C-terminal domain of human TAFII250 than does Bdf2 [41] . Given such a model for the divergence of Bdf1 and Bdf2 functions , the differences in protein sequence between these two proteins may prove informative for deciphering the mechanism of Bdf1 activity . In addition to the over-representation of factors marking the 5′ end of genes , our screen identified a number of factors involved in the 3′ end processing of mRNAs . Splicing-sensitive microarrays confirm a broad splicing defect in a mutant of Yth1 , the homolog of human CPSF30 , and transcript-specific splicing defects in mutants of Cft2 and Fip 1 , the homolog of human CPSF100 and a component of the polyadenylation factor complex PF I , respectively . In higher eukaryotes , components of the 3′ end processing machinery have been shown to physically associate with components of the U2 snRNP [50] and U2AF65 [51]; moreover , in vitro studies demonstrate a functional link between the pre–mRNA splicing and 3′ end processing pathways [52] . The interactions between these two pathways in mammalian systems have led to the proposal that the 3′ end machinery plays an important role in terminal exon definition . Whereas the exon-definition model for mammalian spliceosome assembly posits that internal exons are defined by interactions between U1 and U2 snRNP components across an exon [53] , definition of terminal exons is achieved by interactions between the 3′ end processing machinery and the U2 snRNP ( Figure 7A ) , imposing a functional connection between the pathways . Yet because of the relatively short length of S . cerevisiae introns , and the limited number of genes that are interrupted by multiple introns , splicing in yeast has long been considered to proceed through a model of intron-definition . Nevertheless , the Keller lab recently demonstrated that some conditional alleles of YSH1/BRR5 lead to a decrease in splicing efficiency [54] . Our demonstration here of pre–mRNA splicing defects in the background of additional mutants in 3′ end processing mutants suggests the intriguing possibility that some of the basic interactions that facilitate exon-definition in higher systems may also be present in budding yeast . Indeed , further characterizing the mechanism by which these 3′ end processing factors are affecting splicing in yeast may provide important insights into the mechanisms by which exon-definition is accomplished in higher eukaryotes . An important strength of an approach such as this is the genome-wide perspective that it provides . Figure 7B shows a model of an idealized transcript along with the functional location of a subset of the factors that have been examined in our screen . It is striking to note that many of the factors identified here function both during transcription initiation ( Bdf1 and others ) and termination ( Yth1 , Cft2 , Fip1 , and others ) , thereby defining the beginning and ends of the first and last exons , respectively . In this work , we have identified not only those factors whose disruption leads to a functional defect in splicing efficiency , but in many cases the specific transcripts whose splicing is affected . More broadly , the work presented here demonstrates the feasibility of quantitating cellular RNA levels in the background of large mutant strain collections . While our current approach examined splicing efficiency in the context of optimized growth conditions , a similar approach could be applied to identify factors necessary for efficient splicing under varying cellular or developmental growth states . Likewise , although our work focused on the levels of several pre–mRNA species , this methodology should be directly applicable to assessing the levels of nearly any cellular RNA of interest . All experiments were performed using haploid strains . To assess the function of non-essential genes , the mat a version of the haploid deletion library from Open Biosystems [55] was used ( referred to herein as non-essential strains ) . Likewise , to assess the function of essential genes , a collection of strains provided by the Hieter lab [25] was used ( referred to herein as essential strains ) . In addition , a collection of strains containing previously characterized mutations in core spliceosomal components was used ( from here on considered a part of the essential strains set ) . A complete list of the strains used in this work is included in Table S3 . Unless otherwise indicated , all strains were grown in rich medium supplemented with 2% glucose ( YPD ) [56] . When appropriate , strains were recovered from frozen glycerol stocks on solid medium supplemented with 200 µg/ml G418 grown at either 30°C ( non-essential strains ) or 25°C ( essential strains ) . A manual pinning tool ( V&P Scientific , cat . #: VP384FP6 ) was used to transfer cells from solid medium into 384-well microtiter plates ( Greiner BioOne , cat . #: 781271 ) for growth in liquid media . Liquid cultures were grown in an Infors HT Multitron plate shaker at 900 rpm with 80% constant humidity . Breathable adhesive tape ( VWR , cat . #: 60941-086 ) was used to seal the plates and reduce evaporation . Because the growth rates of the strains being used vary significantly [36] , an approach was developed to enable the systematic collection of a similar number of rapidly dividing cells for each strain . An initial liquid culture was grown in 384-well plates for two days , allowing nearly all strains to reach saturation . Because all of the strains being used are derived from a common parental strain , the cell density for each of these strains is nearly identical at saturation , allowing us to effectively ‘normalize’ the cell numbers . Using a liquid handling robot ( Biomek NX ) , 2 µl of saturated culture were used to inoculate a fresh 150 µl of YPD . This culture was allowed to grow for four hours , an amount of time which is sufficient to allow all strains to exit lag-phase and begin exponential growth , but not so long as to result in a large variation in cell densities among the strains ( Figure S6 ) . For the non-essential strains , all growth was conducted at 30°C . For the essential strains , the initial growth was done at 25°C ( a permissive temperature for all strains ) , but the saturated cells were back-diluted into plates containing media pre-warmed to 30°C ( a non-permissive temperature for many , but not all , of the strains ) and allowed to continue growing at 30°C for four hours . For both the non-essential and the essential strains , two independent biological replicates were initiated from each saturated plate . After four hours of outgrowth , cells were harvested by centrifugation at 4000×g for five minutes . The cell pellets were flash frozen in liquid N2 and stored at −80°C until further processing . Isolation of total cellular RNA was performed using custom protocols written for a Biomek NX liquid handling system . To each frozen cell pellet collected as described above , 50 µl of Acid Phenol: Chloroform ( 5∶1 , pH<5 . 5 ) and 25 µl of AES buffer ( 50 mM sodium acetate ( pH 5 . 3 ) , 10 mM EDTA , 1%SDS ) were added . The plates were sealed with plastic CapMats ( Greiner BioOne , cat . #: 384070 ) and vortexed for five minutes at top speed on a plate vortex . The plates were incubated for 30 minutes in a water bath at 65°C with intermittent vortexing . After incubation , the plates were spun for one minute at 1000×g . An additional 35 µl of AES buffer was added to each well , and after mixing the organic and aqueous phases were separated by centrifugation for five minutes at 3000×g . Using a slow aspiration speed , 40 µl of the upper phase containing the RNA were robotically transferred to a new 384-well microtiter plate . The transferred aqueous phase was mixed with 3 volumes of RNA Binding Buffer ( 2 M Guanidine-HCl , 75% isopropanol ) and passed through a 384-well glass fiber column ( Whatman , cat . #: 7700-1101 ) by centrifugation for two minutes at 2000×g . The column was washed twice with two volumes of Wash Buffer ( 80% ethanol , 10 mM Tris-HCl ( pH 8 . 0 ) ) , followed by a final dry spin for two minutes at 2000×g . To remove any contaminating genomic DNA , 5 µl of DNase Mix ( 1× DNase Buffer , 0 . 25 units of DNase I ( Promega ) ) was added to each well and incubated at room temperature for 15 minutes . After the incubation , 80 µl of RNA Binding Buffer was added to each well of the 384-well glass fiber plate and spun as before . After washing and drying as above , 15 µl of sterile water was added to each well of the glass-fiber plate to elute the RNA into a clean 384-well microtiter plate ( Greiner BioOne , cat . #: 781280 ) . In general , this procedure yielded about 1 µg of total cellular RNA from each cell pellet . The quality of the RNA produced by this protocol is equal to our conventionally purified samples , and the effectiveness of the DNase treatment is demonstrated in Figure S7 . Total cellular RNA was converted into cDNA in 384-well microtiter plates . Of the 15 µl of RNA purified as described above , 10 µl were used in a cDNA synthesis reaction that had a total volume of 20 µl and which contained 50 mM Tris-HCl ( pH 8 . 3 ) , 75 mM KCl , 3 mM MgCl2 , 10 mM DTT , 0 . 5 mM each dNTP , 5 µg dN9 primer , and 60 ng M-MLV RT . Reactions were incubated overnight at 42°C . The cDNA reactions were diluted 30-fold with water , giving a final concentration of ∼1 ng/µl based on the initial RNA concentration , and used without any further purification as templates in high-throughput quantitative PCR ( QPCR ) reactions . The QPCR reactions were performed in a reaction volume of 10 µl , containing 5 µl of template ( ∼5 ng of template ) , 10 mM Tris-HCl ( pH 8 . 5 ) , 50 mM KCl , 1 . 5 mM MgCl2 , 0 . 2 mM each dNTP , 0 . 25× SYBR Green , 5% DMSO , 0 . 7 ng Taq DNA polymerase , and 250 nM forward and reverse primers . The sequences of the primers used for each targeted RNA are shown in Table S4 . Standard curves were generated consisting of 4-fold serial dilutions of genomic DNA and covering a range of 1 . 6×105 molecules . Each primer pair was well-behaved , showing an amplification efficiency of between 86% and 97% ( Figure S8 ) . Two technical replicates were measured for each biologically independent sample , generating four independent measurements for each of the ∼5500 mutant strains . On the basis of standard curves generated using QPCR , relative nanogram quantities were calculated for every RNA transcript within each of the ∼5500 strains tested . To assess reproducibility , coefficients of variation ( CV ) were determined for each primer pair and each strain . The vast majority of these were highly reproducible , both overall and on a per plate basis . As an initial quality filter , we chose to exclude any samples for which the CV was greater than 0 . 25 . Because no simple mechanism exists to normalize for variability in each of our experimental steps , we instead chose to measure the levels of six different RNAs in each of the samples and use these to determine a composite normalization value to account for the overall yield in our procedure . The six RNAs were: U1 snRNA , Scr1 ( SRP ) RNA , Tef5 mRNA , Tub1 mRNA , Srb2 mRNA and Faa1 mRNA . These RNAs were chosen because their biological functions are diverse and their cellular levels vary over a broad range ( ∼300-fold , Table S1 ) . For both independent biological replicates of every strain , a composite normalization constant , , was calculated according the following formula:For each primer pair , represents the relative nanogram quantity calculated for an individual sample . Similarly , represents the median value determined for a given primer pair on an individual QPCR plate run . Because of the subtle variations that are apparent from one plate run to the next , we found that this per plate normalization using gave us the most robust data . By determining the ratio of for every primer pair , a relative abundance of total RNA can be calculated for every sample . As seen in Figure S9 , a histogram of values follows a normal distribution in log2 space with a variance of 1 . 5 units . A second filter at the level of values was introduced which allowed for the filtering of samples with very low amounts of cDNA . For strains that passed this filter , the normalized levels of a given RNA were determined according to the following formula:The relative amount of RNA in a given strain was then determined according to the following formula:For each primer pair , represents the median value of the normalized RNA levels determined within a given QPCR plate . For each biological replicate of every strain , both the and the values are available through Gene Expression Omnibus ( GEO ) using accession number GSE34330 . To determine the subset of strains that cause a statistically significant increase or decrease in precursor levels , we employed the Significance Analysis of Microarrays , or SAM , program [37] . While this software was originally designed for the analysis of microarray data , a significance analysis of our QPCR data is subject to similar concerns regarding multiple hypothesis testing . For each RNA , SAM analysis was performed on the four values , comprised of both technical and biological replicates that were generated for each of the ∼5500 strains . For each transcript , a one class SAM analysis was performed where the Δ value was adjusted to minimize the false discovery rate ( FDR ) , yielding the following values: for the U3 precursor using Δ = 0 . 983 , FDR = 0 . 045; for the Tub3 pre–mRNA using Δ = 0 . 91 , FDR = 0; for the Rpl31b pre–mRNA using Δ = 1 . 061 , FDR = 0 . 003; for the Ubc13 pre–mRNA using Δ = 0 . 978 , FDR = 0 . 002; and for the Tef5 pre–mRNA using Δ = 0 . 99 , FDR = 0 . The candidate non-essential deletion strains were grown to saturation in YPD at 30°C , then back diluted in 25 ml cultures in flasks at a starting A600∼0 . 2 and allowed to grow at 30°C until they reached an optical density of between A600 = 0 . 5 and A600 = 0 . 7 . The candidate essential strains were initially grown at 25°C in YPD , then shifted to the indicated temperatures for 15 minutes after they reached an optical density of between A600 = 0 . 5 and A600 = 0 . 7 . In parallel with the collection of the mutant strains , wild type isogenic controls were grown and collected under the same conditions as the mutant strains examined . Total cellular RNA samples were isolated , converted into cDNA , and fluorescently labeled as previously described [31] . All microarrays were performed as two-color arrays comparing mutant and wild type strains , each grown under identical conditions . Both raw and processed microarray data are available through GEO using accession number GSE34330 . The U1C-Tap bdf1Δ and U1C-Tap bdf2Δ strains were generated by deleting the appropriate genes in the background of the U1C-Tap strain from Open Biosystems [57] using standard techniques . The strains were grown at 30°C in rich medium supplemented with 2% glucose ( YPD ) until they reached an optical density of A600∼0 . 7 . The chromatin was cross-linked with 1% formaldehyde for 2 minutes at 30°C . Glycine was added at a final concentration of 125 mM and the cultures were left shaking for another 5 minutes . Cell pellets from 50 ml of culture were collected by centrifugation at 1620×g for 3 minutes , then washed with 25 ml ice-cold 1× PBS and the pellet stored at −80°C . The pellets were resuspended in 1 ml Lysis buffer ( 50 mM Hepes pH 7 . 5 , 140 mM NaCl , 1 mM EDTA , 1% TritonX-100 , 0 . 1% Na-deoxycholate supplemented with protease inhibitors ) and lysed in the presence of 500 µl 0 . 5 mm glass beads in a beat beater . The lysate was collected by centrifugation at 1000×g for 1 minute , and then pre-cleared by spinning for 15 minutes at 14000 rpm in a tabletop centrifuge at 4°C . The pellet was re-suspended in another 1 ml of Lysis buffer and the chromatin was sheared to an average size of 300 bp ( range 100–500 bp ) by means of a Bioruptor sonicator . The sample was clarified by 2 cycles of centrifugation at 14000 rpm for 15 minutes in a tabletop centrifuge at 4°C and the resultant chromatin solution frozen and stored at −80 C . From the chromatin samples a 1% Input sample was retained , and then each sample was split equally between a Mock IP and an IP sample . The IP samples were incubated with 5 µl 0 . 5 mg/ml anti-Tap Antibody ( Thermo Scientific , CAB1001 ) . After 2 hours at 4°C on a rotator , 25 µl of protein A/G-agarose resin ( #Sc-2003Santa Cruz ) was added to all samples and they were further incubated for another 2 h at 4°C . The resin was washed twice with 1 ml Lysis buffer , twice with 1 ml Wash buffer ( 10 mM Tris-HCl , 25 mM LiCl , 0 . 5% NP-40 , 0 . 5%Na-deoxycholate , 1 mM EDTA ) supplemented with 360 mM NaCl , twice with 1 ml Wash buffer , and finally twice with 1 ml TE . The first wash was a brief one , followed by a 15 minute incubation of the samples on a rotator at 4°C for the second wash . In between washes , the resin was collected by short spins at 2000 rpm in a tabletop centrifuge . The resin was resuspended in 100 ul Elution buffer ( 50 mM Tris-HCl pH 8 . 0 , 5 mM EDTA , 1% SDS ) and the immunoprecipitated material was eluted from the beads by incubating at 65°C for 30 minutes with occasional tapping . To reverse crosslinks , the IPs and the 1% Input samples were incubated overnight in a 65°C water bath . The next day , the samples were treated with 12 . 5 µl 20 mg/ml Proteinase K solution and incubated at 42° for 2 h . The DNA was then purified by using a Cycle Pure Kit ( Omega Bio-Tek , D6492-01 ) following the manufacturer's instructions and eluted in a final volume of 120 µl . Quantitative real-time PCR was performed on a Roche Light Cycler 480 machine as described above , using the 1% Input sample to generate a standard curve for each of the primer pairs we used . For the primers used in the screen , the sequences are available in Table S4 . The primers for the different regions of actin gene and the PMA1 gene are the same as previously published [5] . For each sample the Mock IP value calculated as percent input was subtracted from the IP value ( in percent input ) . Then , a fold enrichment value was calculated , by dividing these values by the PMA1 value . An overexpressing plasmid containing a full-length copy of the Mud1 gene including ∼500 bp up- and down-stream of the ORF was transformed into BY4741 ( Open Biosystems ) . This strain and a control strain containing the empty vector were grown in 25 ml minimal media until they reached an optical density of A600∼0 . 5–0 . 6 . RNA isolation was performed as previously described [31] , and cDNA synthesis and Q-PCR were performed as described above . The primer sequences are found in Table S4 .	The coding portions of most eukaryotic genes are interrupted by non-coding regions termed introns that must be excised prior to their translation . The excision of introns from precursor messenger RNA ( pre–mRNA ) , is catalyzed by the spliceosome , a large macromolecule composed of both RNA and protein components . Several studies have uncovered connections between pre–mRNA splicing and other RNA processing pathways such as the remodeling of chromatin structure , transcription , and processing events that take place at the 3′ end of the transcript . To date , however , the full complement of factors that function to couple splicing to other processes in the cell remains unknown . Here , we have developed a novel screening methodology in the budding yeast , Saccharomyces cerevisiae , that allowed us to individually examine nearly all of the ∼6 , 000 genes to determine which factors functionally impact splicing . We identified mutations in components that function at either the 5′ or 3′ end of a gene . Most of these components have previously established roles in other aspects of gene expression , including chromatin remodeling and cleavage and polyadenylation processes , and their identification here provides the first evidence for their roles in coupling these pathways .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "genome", "analysis", "tools", "chromosome", "biology", "gene", "expression", "genetics", "biology", "genomics", "genetics", "and", "genomics" ]	2012	A Quantitative, High-Throughput Reverse Genetic Screen Reveals Novel Connections between Pre–mRNA Splicing and 5′ and 3′ End Transcript Determinants
Large spatial and temporal fluctuations in the population density of living organisms have profound consequences for biodiversity conservation , food production , pest control and disease control , especially vector-borne disease control . Chagas disease vector control based on insecticide spraying could benefit from improved concepts and methods to deal with spatial variations in vector population density . We show that Taylor's law ( TL ) of fluctuation scaling describes accurately the mean and variance over space of relative abundance , by habitat , of four insect vectors of Chagas disease ( Triatoma infestans , Triatoma guasayana , Triatoma garciabesi and Triatoma sordida ) in 33 , 908 searches of people's dwellings and associated habitats in 79 field surveys in four districts in the Argentine Chaco region , before and after insecticide spraying . As TL predicts , the logarithm of the sample variance of bug relative abundance closely approximates a linear function of the logarithm of the sample mean of abundance in different habitats . Slopes of TL indicate spatial aggregation or variation in habitat suitability . Predictions of new mathematical models of the effect of vector control measures on TL agree overall with field data before and after community-wide spraying of insecticide . A spatial Taylor's law identifies key habitats with high average infestation and spatially highly variable infestation , providing a new instrument for the control and elimination of the vectors of a major human disease . Vector-borne pathogens contribute to 17% of the global human disease burden [1] . Chagas disease or American trypanosomiasis , one of the World Health Organization's "neglected tropical diseases , " is caused by the protozoan Trypanosoma cruzi . It is transmitted mainly by diverse triatomine bug species associated with selected wild , peridomestic and domestic habitats in the Americas . The major vectors of human Chagas disease thrive in human dwellings and peridomestic structures housing domestic animals . ( Peri ) domestic populations of the major vector Triatoma infestans differ widely depending on the specific local habitat and host species [2] . Here " ( peri ) domestic" refers to structures that are domestic or peridomestic . We show that Taylor's law ( TL ) , which we describe below , describes well the average and variance of habitat-specific relative population sizes of T . infestans and three other vector species of T . cruzi in all ( peri ) domestic habitats . The data result from 33 , 908 habitat searches for triatomine bugs in four areas of Argentina from 1993 to 2010 before and after the large disturbance caused by community-wide insecticide spraying directed to suppress ( peri ) domestic infestations with T . infestans . One area , well described by TL , had moderate insecticide resistance that caused vector control failures . We determine the effect of insecticide spraying or the history of chemical control interventions on the values of the parameters of TL and describe some implications of TL for vector control and surveillance . The present paper may be the first to demonstrate the connection of TL with any aspect of Chagas disease , and in particular with the population densities of the insect vectors of the disease . Large spatial and temporal fluctuations in the population density of living organisms have profound consequences for biodiversity conservation , food production , pest control and disease control , especially vector-borne disease control . In empirical studies of insects and many other species , the sample variance v and the sample mean m of population counts or other measures of population density or abundance approximate well a linear relationship on log-log coordinates , log10 v ≈ a + b * log10 m [3] , which is mathematically equivalent to the power law v ≈ 10amb . From a mathematical point of view , the slope b of TL is the proportional ( or percentage ) rate of increase in the variance for a given infinitesimal proportional ( or percentage ) increase in the mean . The slope b has been interpreted as an index of spatial aggregation because purely random ( Poisson ) distributions of individuals have a variance equal to the mean and therefore would be expected to generate TL with b = 1 , while some distributions in which the variance grows faster than in proportion to the mean would be expected to generate TL with b > 1 . Variations in habitat suitability and other ecological mechanisms could also generate TL with b > 1 . ( See Future research in the Discussion . ) Although Taylor was not the first to publish empirical examples of the above linear relationship , he made it widely known [4 , 5] , and it is usually called Taylor's law ( TL ) among ecologists , or fluctuation scaling or large-number scaling among physicists [6] . More than 1000 papers have been published on TL and its applications to hundreds of species and many fields besides ecology [6] , including weekly cases of measles in 366 communities in England and Wales pre- and post-vaccination [7] , the aggregation of parasite individuals within host individuals ( not including any parasites , vectors , or hosts related to the transmission of Chagas disease ) [8 , 9] , human population densities [10] , crop yields [11] , prime numbers [12] and tornado outbreaks [13] . TL can be generated by many different models ( e . g . , [6 , 7 , 14–16] ) . TL has important applications in the management of agricultural pests and fisheries . When TL is valid , TL can be used to design more efficient sampling schemes to estimate pest density and decide whether to spray pesticides or release natural enemies in a timely fashion [17 , 18] . TL provides a stopping rule for fixed precision sampling of fisheries , permitting reduced sampling effort [19] . The uses of TL to identify unusual variability in crop yields [11] and plan more efficient control measures by recognizing the heteroskedasticity of population densities at different mean densities are potentially valid for controlling the insect vectors of major human infectious diseases , including malaria , dengue , Chagas disease , sleeping sickness and the leishmaniases . However , literature searches in Pubmed and Google Scholar ( October 16 , 2017 ) using "Taylor’s law" ( or "Taylor’s power law" ) combined with "malaria mosquito ( or Anopheles ) " , or "dengue mosquito ( or Aedes aegypti ) " , or "tsetse fly ( or Glossina fly ) " , or "Chagas vector ( or Triatoma ) " , or "Leishmaniasis sandfly ( or Lutzomyia or Phlebotomus ) " , identified no paper on Chagas disease vectors and TL and only a few papers which mainly used TL for sample size determination of malaria and dengue mosquitoes [20–29] . Two widely tested forms of TL are a temporal TL and a spatial TL . In a temporal TL , n populations labeled i = 1 , … , n are followed over time , and the sample mean size ( averaged over time ) mi of population i and the sample variance of population size ( over time ) vi of population i are calculated separately for each population i . Each population is represented by one dot associated with population i on a plot of log10 vi ( vertical axis ) as a function of log10 mi ( horizontal axis ) . If the dots fall approximately along a straight line , the data support a temporal TL . In a spatial TL , which we pursue here , different populations of a species are grouped into different categories . In this article , each category will be a different habitat in which Chagas vectors may be found , such as a chicken coop or a goat corral . Habitats are labeled h = 1 , … , H , where H is the number of different habitats . The mean mh and the variance vh of population sizes over all sites of habitat h ( e . g . , over all chicken coops in a community ) are calculated and log10 vh is plotted as a function of log10 mh , with one data point for each habitat h . If the H dots fall approximately along a straight line , the data support a spatial TL . We adopt some conventions of language . We use "areas" to refer collectively to the four geographical locations , Amamá , Olta , Figueroa , and Pampa del Indio , where studies and control efforts were conducted . We use "habitat" for a category of individual places that were surveyed for bugs . For example , chicken coops are one habitat , goat corrals are another habitat , and cow corrals are a third habitat . We use "site" for a particular exemplar of a habitat , such as a particular chicken coop , or a particular goat corral . A "house compound" consists of a domicile for people and near-by buildings for human use and corrals for animals . Each such domicile and building is one site . As indicated above , " ( peri ) domestic" habitats include all such structures . Here we demonstrate that TL describes the spatial distribution ( in different sites of a habitat ) of four of the vector species of Chagas disease . When the means and variances of the number of each species of vector are computed over sites separately for each habitat in a community of house compounds , TL is confirmed with high accuracy and consistency over time and under diverse control procedures . The slope b of TL does not deviate significantly from the range 1 < b < 2 . We develop simple mathematical models to help interpret and extend this primary empirical finding . We suggest some practical consequences and potential uses of TL in Chagas disease vector control . The full implications of TL for Chagas disease vector control remain to be worked out in future research and practice and are not the primary objective of this paper . Finally , we suggest some future research . The data come from four large research projects in the Argentine Chaco region where Chagas disease was endemic . These projects aimed primarily to control the major vector Triatoma infestans , but also included observations of other local triatomines not considered as the main control targets . The surveys were conducted in well-defined rural areas of Olta ( municipalities of General Belgrano and Chamical , of the province of La Rioja , western Argentina ) , Figueroa and Amamá ( Figueroa and Moreno departments , respectively , of the province of Santiago del Estero , northwestern Argentina ) , and Pampa del Indio ( General San Martín department , of the province of Chaco , northeastern Argentina ) . The studies were organized spatially in a hierarchy with five levels: Argentine Chaco region; four study areas within the region; villages within each area; house compounds ( defined above under "Purposes" ) within each village; and sites within each house compound . The details of each area are described extensively in Detailed Methods in S1 Text . Fig 1 ( A ) maps the areas of these studies and Fig 1B , 1C and 1D illustrates the hierarchy of villages , house compounds , and sites . Table 1 summarizes the quantities of the data collected . Suppose that Taylor's law ( TL ) describes well the relation between the mean and the variance of relative population density of a single vector species in the habitats of a study area before the house compounds ( including all ( peri ) domestic structures ) are sprayed with insecticides to kill the vectors . What would we expect to be the effect of spraying ? Specifically , would we expect TL to hold after spraying ? If so , what if any connection should we expect between the intercept and slope of TL before and after spraying ? Here we propose two simple models to answer these questions . In the Results , we will compare some of the predictions of these models with observations . It is not necessary to follow the mathematical details to understand the models' predictions or the empirical results . S1 Text gives mathematical proofs . Both models use the same general notation . Suppose there are H > 2 habitats , such as chicken coop; open shed; oven; piled materials; cow corral; latrine/bathroom; etc . These habitats are labeled h = 1 , 2 , … , H . Let B ( h ) be a random variable representing the number of vectors of a single species ( not all Triatoma species combined ) in the various sites of habitat h in the study area Before spraying , and let A ( h ) be a random variable representing the number of vectors in the various sites of habitat h in the study area After spraying . Because of the gap in time between the survey before spraying and the survey after spraying , the set of sites of a given habitat before spraying may differ from the set of sites of that habitat after spraying . The population mean ( or expectation ) of B ( h ) will be written E ( B ( h ) ) and the population variance , Var ( B ( h ) ) ; likewise , the population mean E ( A ( h ) ) and population variance Var ( A ( h ) ) of A ( h ) . We assume the population mean and population variance exist and are positive . For the vectors before spraying , the log-log form of TL using the population mean E ( B ( h ) ) and the population variance Var ( B ( h ) ) instead of the corresponding sample mean m and sample variance v is log10 Var ( B ( h ) ) = a + b log10 E ( B ( h ) ) . This linear form is mathematically equivalent to the power-law form of TL , namely , Var ( B ( h ) ) = C[E ( B ( h ) ) ]b , with C = 10a . The value of the slope b in the log-log form of TL is identical to the value of the exponent b in the power-law form , hence we use the same notation b and we refer to b interchangeably as the slope or the exponent . The value of the intercept a in the log-log form is related to the value of the coefficient C in the power-law form by 10a = C or log10 C = a , hence we use different words and symbols ( intercept a versus coefficient C ) . We assume that spraying reduces the relative population density of the vector , and does not increase it . Taylor [3 , 4] estimated that different insect species had different slopes b . He proposed that the slope b was a characteristic unique and specific to each species , meaning that a given species always had the same value of b and no other species had that value of b . Others observed that Taylor had not considered the sampling variability in estimates of b . Two different species with identical population values of b could have different point estimates of b due to sampling fluctuations alone , not necessarily because of any statistically significant difference in b between them . Moreover , two different species with different population values of b could have identical or statistically indistinguishable point estimates of b due to sampling fluctuations alone . Moreover , Yamamura [45] and Cohen et al . [46] showed by examples that the physical scale on which population density was measured in a spatial TL could substantially affect the value of b , so that b was not necessarily an invariant species-specific characteristic . The question of whether the slope b is a species characteristic ( under some range of circumstances ) has practical significance for Chagas disease vector control , because it implies further questions . For a given species , must b be measured anew in every new field situation ? Must b be measured independently for different vector species ? S3 Table gives the Welch test statistic , df , and P for testing the null hypothesis of no significant difference between every two estimates of the slope b in Table 2 ( excluding 4 of the 83 estimates , as we now describe ) . All these studies shared a similar ecology in the Gran Chaco and similar sampling methods . S3 Table and the following analyses omit two regression estimates in Table 2 from Amamá for T . guasayana in the core postintervention surveys of October 1993 and May 1999 , for which df = 0 and 1 , respectively , and the two estimates for Olta with three species combined , pre- and postintervention , leaving 79 = 83 - 4 species-specific estimates of b . There are 3 , 081 = 79×78/2 distinct pairs of estimates of b . Of these 3 , 081 comparisons , 206 , or 6 . 7% , were less than 0 . 01 , nearly seven times more than the approximately 31 = 3 , 081×0 . 01 P values less than 0 . 01 that would be expected on average by chance alone under the null hypothesis that all underlying values of b are identical . Because each P value describes the difference of two slopes b and the same b occurs in multiple comparisons with all other values of b , we refrain from attempting to assign a probability to the chance that 3 , 081 dependent trials with marginal probability of success equal to 0 . 01 would yield 206 cases with P < 0 . 01 in the Welch test that two b values are equal . Instead , we simply infer that at least some of the differences in b were probably not due to sampling variation alone , and we focus attention on where the frequency of significant differences in b occurred most frequently . Table 3 summarizes the number of cases where the Welch test's P < 0 . 01 for each pairing of each of the four species of Triatoma with itself and with each of the other three species . According to the diagonal elements of Table 3C , the percentage of P values that were less than 0 . 01 exceeded 1% by a factor of 2 . 8–4 . 7 when the slope for each of the four species was compared with the slope for the same species in another study . Each diagonal element was less than the percentage directly above it or directly to its right , meaning that there were fewer intraspecific than interspecific significant differences in slope . Each species resembled itself in TL slope b more than it resembled any other species , when we compared pairs of distinct studies . The largest percentages of P values that were less than 0 . 01 occurred in the comparison of b values of T . guasayana with those of the other three species . Overall , T . guasayana had b values that differed the most from those of the three other species , and overall the slope b of TL varied less within any of the four species than interspecifically . The b values from T . sordida and T . garciabesi hardly differed significantly more often interspecifically ( 3 . 2% ) than they varied intraspecifically ( 2 . 8% for both species ) . With remarkable precision , Taylor's law ( TL ) described the relationship , among multiple habitats , of the variance of relative bug population density in sites of a given habitat to the mean of relative bug population density in the same sites of the given habitat . The log variance was well approximated as a linear function of the log mean in four studies that differed in geographic location , methods of bug control , time since spraying , degree of insecticide resistance , and principal bug species . The adjusted R2 had median 0 . 94 and lower and upper quartiles 0 . 91 , 0 . 96 in 79 single-species field surveys with df > 1 in Table 2 . The slope b of the linear relationship of log variance to log mean differed most between Triatoma guasayana and the other three species , T . infestans , T . sordida , and T . garciabesi . Interspecific differences exceeded intraspecific differences in slope b . Slopes of TL generally , but not always , rejected Poisson-distributed vectors with different means in different habitats and were consistent with substantial spatial aggregation or differences in habitat suitability: the median slope b was 1 . 48 and the lower and upper quartiles were 1 . 35 , 1 . 63 in these 79 field surveys ( Table 2 ) . Only four of these field surveys had b > 2 , and none of these slopes was significantly greater than 2 . Only four of these field surveys had b < 1 , and none of these slopes was significantly less than 1 . Thus none of the 79 TL field surveys significantly rejected the general pattern that 1 < b < 2 , as has been widely found in other studies . The first inequality , 1 < b , implies that , when different habitats were compared , the variance in their bug populations increased faster than in proportion to their mean bug populations . The second inequality , b < 2 , implies that , when different habitats were compared , the coefficient of variation ( standard deviation divided by mean ) decreased as the mean number of bugs per habitat increased . Model 1 usefully highlights deviations from the predictions that follow from its very simple assumptions . The increased slope of TL for T . infestans observed after spraying in Olta , Figueroa , and Pampa del Indio indicates that spraying lowered mean vector abundance in some habitats more than in others , and hence was not uniformly effective across habitats . This heterogeneity in the effectiveness of spraying is also evident in the plots of log-mean-abundance-after-spraying as a function of log-mean-abundance-before-spraying ( Fig 5 , Fig B in S1 Text , Fig C in S1 Text ) . This evidence rejects the classic assumption made by vector-control agencies that insecticide spraying is uniformly effective . Heterogeneity in the effectiveness of spraying calls for targeted , improved vector control , possibly including appropriate environmental management measures . This issue is of paramount importance for the goal of large-scale Chagas disease vector elimination under the aegis of regional intergovernmental programs such as the Southern Cone and Central America Initiatives created in the 1990s and ongoing [47 , 48] . Model 1 has relevance beyond attempts to control Chagas disease vectors . As part of a study of the ecological consequences of sudden oak death in northeastern United States temperate forests , oak trees were censused and measured in 12 plots in 2007 and again in 2010 in Black Rock Forest , Cornwall , New York . Cohen et al . [49] fitted TL successfully to the counts of oak trees in 2007 before a major intervention in 2008 ( killing of some oak trees by girdling ) and again in 2010 . The girdling of the oaks was analogous to spraying the bugs . The acceptable fits of TL before and after intervention were consistent with prediction 4 of model 1 , and the absence of statistically significant evidence of a change in slope as a result of the intervention was expected from prediction 5 of model 1 . Cohen et al . [49] offered no explanation of why the intervention did not significantly change the slope or destroy the fit of TL . Model 1 offers an explanation , or at least a phenomenological description . Taylor's law identifies key habitats with high mean and variance of infestation before or after insecticide spraying . TL has not previously been used to discover highly variable infestations by any insect vector , to the best of our knowledge . In general , a habitat with exceptionally high variance , given its mean , would likely have a high likelihood of a vector ( here bug ) outbreak , and might deserve special attention for control by spraying or environmental alteration . A habitat with exceptionally low variance , given its mean , would be a habitat with a relatively stable endemic bug population . For example ( Fig 2A ) , the spatial variance of T . infestans in chicken houses ( ckh ) in the periphery of Amamá is at or very near the lower limit of the 95% CI given the rather large spatial mean abundance of bugs in chicken houses in the periphery . Even with no spraying of insecticides , one might expect chicken houses with relatively stable chicken populations to harbor spatially consistent bug populations . TL is useful for practical control efforts also for habitats that are not outliers from the log-log regression . For example , in the graphs of TL based on T . infestans abundance by habitat in Amamá , the data points at the upper right of the graph were granaries and sheds ( i . e . , open sheds with a thatched roof , frequently used as a storage area , with wide variation in stored contents and usage between households , sometimes with chickens nesting there ) . Overlooking or failing to inspect or treat adequately granaries and sheds could have large effects on bug persistence and subsequent house reinfestation rates . People rarely permit insecticide spraying of granaries with stored corn ( except when the granaries are empty ) . The preintervention mean bug population size of granaries was large in Figueroa and a high proportion of granaries was infested [2] . Granaries were again at the top ( upper right ) of the log-log regression before and 5 months after spraying in Figueroa . Open sheds do not get much attention as potentially important habitats for bug control , perhaps because of the wide variability in bug counts between individual sheds . Granaries and open sheds had greater means and variances of bug abundance than chicken coops , storerooms , and domiciles . The latter three habitats usually take all the attention of bug control personnel . Domiciles , storerooms and kitchens tend to appear on the upper end of the log variance-log mean regressions . Because they concentrate competent hosts , these three habitats concentrate nearly all of the T . cruzi-infected bugs in any study area in the Gran Chaco , especially domiciles where all human-vector contacts occur ( e . g . , [50] ) . The same argument applies to other habitats that receive less attention because they are located outside of domestic compounds in areas less disturbed by human activities . Examples include orchard fences where cavies and other rodents thrive and chickens sometimes nest [51] , and abandoned constructions covered by vegetation and used by free-ranging goats , sheep or other domestic or sylvatic mammals [52] . Such areas are very rarely inspected for bugs or sprayed by bug control staff on the assumption that they are rarely infested ( i . e . , implicitly assuming that such habitats have an exceptionally large variance of bug population density ) and because substantial time , effort and insecticide would be needed to spray them rigorously . Over the years , we collected several examples of such highly infested , rarely detected , habitats that would fall in the upper right corner of the graph of TL and would severely hamper any serious elimination campaign . In the practical control of Chagas disease vectors , researchers and bug control people historically have shown little interest in levels of statistical precision or in using sampling theory . In the case of T . infestans , perhaps one reason was that the main initial program goal was to eliminate the vector completely from most of its range , not to control it or estimate its relative population density , and to treat all infested houses completely and homogeneously . Another possible reason was the crude method of detecting or sampling bugs , which is still used . Bug control programs declare house compounds infested or not . Though they may count the number of bugs per unit of search effort at the level of the house compound ( distinguishing domestic from peridomestic habitats ) , they do not use this information other than for saying "many bugs or few bugs" at the community-wide level . Failure to use sampling theory in past practice does not diminish the present and future need to consider TL and its implications for certifying the interruption of transmission of T . cruzi to humans at a district- or state-wide scale , for example . Figs 2–4 , 6 and 7 make very clear which habitats have the largest means and variance of bug populations and how habitats respond to insecticide applications . Sometimes a habitat stays at the extreme right ( with large mean and large variance of relative population size ) after interventions while other habitats jump to the left extreme or remain there before and after interventions . That a habitat stays at the upper extreme after insecticide applications demonstrates the vulnerability of vector control through insecticides and the need for better chemical control or alternative interventions . The habitat "other" appears several times with high mean abundance and high variance . Because "other" habitats are hard to classify , they may not be identified as important for insecticide applications . This problem is particularly acute for Triatoma dimidiata and other species where selective control is frequent ( i . e . , only domiciles , or only domiciles and chicken coops , are sprayed , for example ) . Neglected "other" habitats may be sources of recurrent infestations . This collection of surveys and trials shows that insecticide-based vector suppression in several study areas across the Argentine Chaco has been far from uniformly successful over the 28 years from 1993 to 2010 . The graphical display of means and variances by habitat may provide useful quantitative guidance for improving insecticide-based vector suppression . The relative population densities of Chagas vectors reported here strongly confirm TL but do not identify the mechanisms that may generate TL , such as insect behavior versus habitat suitability ( measured by bug birth rates and death rates in different habitats ) . How important are selective dispersal and migration ( insect behavior ) versus differences between habitats in their average suitability for bugs and in their variation ( across sites ) of suitability ? Control measures that transitorily lower the suitability of a habitat , such as residual insecticide spraying , could in principle stimulate adaptive behavioral responses by insects ( e . g . , excito-repellency ) , which may reduce the effectiveness of spraying at the level of the house compound or the village . Thus understanding the mechanisms that produce TL is important . In these studies , postintervention surveys occurred from 4–6 months to 12 months after insecticide spraying . Hence the observed relative bug population density postintervention included immigration , local recruitment ( new births of bugs ) , and local survival postintervention . Models 1 and 2 of habitat-specific persistence represent only the last of these processes , and therefore cannot be expected to account precisely for our field observations . Models 1 and 2 assume that insecticide spraying does not reduce to zero the average or the variance of relative population density in sites of a habitat . They are models of successful partial control but not models of successful local elimination of bugs from all sites of any habitat . When the ultimate program goal of insecticide spraying is local elimination ( suppression ) of bugs from all sites of a habitat rather than control , these models may be helpful in identifying less vulnerable habitats using TL . It would be useful to supplement these models with an empirical summary and theoretical model of the fraction of sites of each habitat from which spraying eliminated the vectors . If the number of bugs per site of a habitat were described by the negative binomial distribution , then the mean and variance would imply a fraction of sites with zero bugs . Efforts have been made to link TL with the mean and the variance of the negative binomial distribution [8 , 52–57] , but it has recently been recognized that TL cannot hold with constant parameters at the same time that the negative binomial distribution holds with a constant scale parameter and a changing probability parameter ( [9] , p . E50 ) ( see S1 Text for further details ) . A practically important topic for further research is finding a useful way to estimate the fraction of sites of a habitat with no bugs , when the mean and variance of bug abundance obey TL , as here . The units of analysis in this paper are habitats such as chicken coops , kitchens , or granaries . By contrast , vector control programs use house compounds as units to calculate village-wide infestation rates . Additional empirical and theoretical analyses are needed to link habitats with house compounds and village-level summaries of infestation . Moreover , since the goal of vector control is to reduce or eliminate human T . cruzi infections , it would also be highly desirable to explore the potential of TL to shed light on the distribution of T . cruzi infection in vectors , humans and other animal hosts .	Chagas disease , or American trypanosomiasis , is mainly transmitted to humans by insects that dwell in human habitations and structures closely associated with human habitations , such as kitchen out-buildings , chicken coops , goat corrals , and grain storage bins . Widespread in the Americas , the disease causes chronic illness and often eventual death . No vaccines exist . Available drugs may cause undesirable adverse effects and do not prevent re-infection . Efforts at suppressing the disease have been directed at eliminating the principal insect vector species from human dwelling compounds . Effective insecticide spraying requires finding out where the insects are . Both the average and the variance of the relative number of insect vectors of each species in each habitat are relevant to control efforts . We demonstrate here that the spatial distribution of the insect vectors of Chagas disease obeys a previously unrecognized pattern , known in ecology as Taylor's law ( TL ) : in different habitats , the variance of vector relative numbers is approximately a power function of the mean of vector relative numbers . TL identifies key habitats with high average infestation and highly variable infestation , providing a new instrument for the control and elimination of the vectors of a major human disease .	[ "Abstract", "Introduction", "Methods", "and", "materials", "Results", "Discussion" ]	[ "medicine", "and", "health", "sciences", "ecology", "and", "environmental", "sciences", "tropical", "diseases", "vertebrates", "chickens", "parasitic", "diseases", "animals", "habitats", "neglected", "tropical", "diseases", "population", "biology", "infectious", "disease"...	2017	Chagas disease vector control and Taylor's law
Phenotypic resistance describes a bacterial population that becomes transiently resistant to an antibiotic without requiring a genetic change . We here investigated the role of the small regulatory RNA ( sRNA ) RyhB , a key contributor to iron homeostasis , in the phenotypic resistance of Escherichia coli to various classes of antibiotics . We found that RyhB induces phenotypic resistance to gentamicin , an aminoglycoside that targets the ribosome , when iron is scarce . RyhB induced resistance is due to the inhibition of respiratory complexes Nuo and Sdh activities . These complexes , which contain numerous Fe-S clusters , are crucial for generating a proton motive force ( pmf ) that allows gentamicin uptake . RyhB regulates negatively the expression of nuo and sdh , presumably by binding to their mRNAs and , as a consequence , inhibiting their translation . We further show that Isc Fe-S biogenesis machinery is essential for the maturation of Nuo . As RyhB also limits levels of the Isc machinery , we propose that RyhB may also indirectly impact the maturation of Nuo and Sdh . Notably , our study shows that respiratory complexes activity levels are predictive of the bacterial sensitivity to gentamicin . Altogether , these results unveil a new role for RyhB in the adaptation to antibiotic stress , an unprecedented consequence of its role in iron starvation stress response . The emergence and spread of bacterial multi-resistance to antibiotics has become a major health issue in the last decades , urging for the development of new anti-bacterial molecules and for a better understanding of the molecular mechanisms at work behind bacterial resistance [1 , 2] . While acquired resistance mechanisms ( acquisition of genes or mutations that confer resistance ) have long been the main focus of attention , less is known about “phenotypic” resistance , which is the process in which a bacterial population becomes transiently resistant to an antibiotic without requiring a genetic change [3–5] . For instance , this kind of resistance has been associated with specific processes such as stationary growth phase , persistence and metabolic changes , reinforcing the idea that the environment encountered by the pathogen is a key determinant for antibiotic susceptibility [6] . Change in utilization of iron-sulfur ( Fe-S ) cluster biogenesis machineries in Escherichia coli gives a striking example of phenotypic resistance [7] . Fe-S clusters are ubiquitous and ancient cofactors used in a plethora of biological processes , such as metabolism and respiration [8 , 9] . In E . coli , Fe–S clusters are formed and brought to target proteins thanks to two dedicated biogenesis systems: the so called “housekeeping” Isc machinery , which homologs are found in mitochondria of eukaryotic organisms , and the stress-responsive Suf system , in which homologs are found in chloroplasts of plants [10 , 11] . These systems are responsible for the maturation of more than 150 Fe-S cluster containing proteins in E . coli , notably numerous proteins contained in the main respiratory complexes I ( Nuo ) and II ( Sdh ) [12–14] . Strikingly , it was shown that impairment of the E . coli Isc machinery enhances resistance to aminoglycosides , a well-known class of antibiotics that target the ribosome [7] . This resistance is due to a deficiency in the maturation of the respiratory complexes in isc mutants , which in turn leads to a decrease in the proton motive force ( pmf ) that is essential for aminoglycosides uptake [15] . Consistently , the Suf machinery was shown to maturate inefficiently the Fe-S cluster containing proteins of the respiratory complexes , although the molecular reason for this still remains unclear [7] . Overall this study predicted that an environmental signal that induces the switch from Isc to Suf should induce a transient resistance to aminoglycosides . Iron starvation is one such signal as it decreases the expression of the isc operon and enhances that of the suf operon . The small RNA RyhB was proposed to participate to this transition [16] . RyhB is one of the most studied sRNAs to date in E . coli [17–19] . RyhB is regulated by Fur , the main regulator of Fe-homeostasis in many bacteria and is expressed during iron starvation [20 , 21] . When iron becomes limiting in the medium , RyhB base-pairs and represses the translation of more than 100 mRNA targets that encode non-essential iron-utilizing proteins , thus engaging an “iron sparing” response and redirecting iron consumption in the cell [19] . Notably , RyhB was shown to base-pair to the iscRSUA mRNA [16] . RyhB induces the degradation of the 3’ part of the mRNA that contains iscSUA , in this way limiting Isc levels , while the 5’ part that encodes iscR remains stable . While RyhB repression of isc expression is rather modest , it may play a more important role by indirectly contributing to the Isc to Suf transition . Indeed , IscR is the major regulator of Fe-S clusters homeostasis and is itself a Fe-S cluster protein maturated by Isc [22] . Likewise inhibition of Isc functioning is predicted to yield to accumulation of IscR in its apo-form , which actually acts as an activator of the suf operon [23] . Note that the suf operon is also under Fur repression , which is alleviated under iron limitation . Iron homeostasis has been shown to modify the sensitivity of bacteria to a number of antibiotics , although the molecular basis behind this is not always clear [24] . Here we asked if the sRNA RyhB could participate in phenotypic resistance to various antibiotics during iron starvation . We found that RyhB is necessary to induce gentamicin phenotypic resistance in low iron conditions . By further investigating the mechanism by which RyhB controls this phenotypic resistance , we show that RyhB controls entry of aminoglycosides in the cell by inhibiting the activity of the two pmf-producing respiratory complexes Nuo and Sdh . We first investigated whether RyhB has any role in resistance against different classes of antibiotics during iron starvation . To mimic iron starvation , we treated the medium with 250 μM of dipyridyl ( DIP ) , a strong iron chelator . We chose this concentration of DIP because it is known to induce RyhB synthesis [25 , 26] . Growth of both the WT and ryhB mutant strains were slightly affected by depleting iron from the medium , but importantly , doubling time of the ryhB mutant was identical to that of the WT strain ( S1 Fig ) . This observation precludes any difference in antibiotic sensitivity between strains to be attributed to difference in growth properties . We then performed antibiotic killing assays by growing wild type ( WT ) and ryhB mutant cells in LB medium added or not with DIP . Antibiotics were added when cells reached early exponential phase ( OD600 = 0 . 2 ) and the number of survivors was determined by counting the number of colony forming units ( c . f . u ) after 3 hours of incubation . Four different major classes of antibiotics were tested: aminoglycosides ( gentamicin ) , β-lactams ( ampicillin ) , fluoroquinolones ( norfloxacin ) , and tetracycline . Iron chelation did not protect cells against tetracycline ( Fig 1C ) . In contrast , adding DIP to the medium protected the WT and ryhB mutant strains against toxicity of ampicillin , norfloxacin and gentamicin ( Fig 1A and 1B ) . The protective effect of iron deprivation for these antibiotics has already been observed and its underlying cause has been greatly debated [7 , 24 , 27 , 28] . As cells were protected independently of ryhB , we did not pursue these antibiotics further . In contrast , WT cells were protected against gentamicin when DIP was added to the medium , but this protection effect was lost when cells were mutated for ryhB . ( Fig 1D ) . This result thus suggested that RyhB is involved in the protection of bacterial cells against aminoglycosides during iron starvation . Next , we performed gentamicin kinetic killing assays by counting the number of WT or ryhB survivors at different time intervals after adding gentamicin . In this experiment , both the WT and ΔryhB strains showed the same profile when grown in LB ( Fig 1E ) . In both cases , the majority of the cells were rapidly killed after 1 h 30 min of incubation with gentamicin ( 5 logs of killing ) . Again , addition of DIP to the medium had a ≈ 4 log protective effect against gentamicin on WT cells as early as 1 h 30 min post addition of the antibiotic . Cells then remained mainly resistant to gentamicin during the course of the experiment . In contrast , the ryhB mutant gradually became as sensitive as cells grown in the absence of DIP ( see 4 h 30 min time point ) , although killing kinetics were slightly slower than in presence of iron . Finally , effect of RyhB on gentamicin efficacy during iron starvation was estimated by defining minimum inhibitory concentration ( MIC ) values for both WT and ryhB mutants in presence of increasing concentration of gentamicin , with or without DIP . MIC value for the WT strain almost doubled when the cells were grown in the presence of DIP ( S2 Fig ) . In sharp contrast , MIC values of the ryhB mutant were the same in the presence or absence of DIP . Altogether , these results indicated that RyhB is needed for the phenotypic resistance of E . coli to gentamicin in low iron condition . Entry of aminoglycosides is dependent on the proton motive force ( pmf ) mainly produced directly by respiratory complex I and indirectly by the respiratory complex II , respectively encoded by the nuo and sdh operon [12 , 15 , 29] . Thus , based on our previous study [7] , one hypothesis was that RyhB induced resistance was due to an inhibitory effect on the activity of these two complexes that would block entry of gentamicin in the cell . To test this hypothesis killing assays were run with a strain deleted for both respiratory complexes ( Δnuo Δsdh ) . As expected , this mutant was resistant to gentamicin ( Fig 2 , left panel ) [7] . Adding DIP to the medium somewhat increased by 1 log the survival of the nuo sdh mutant , suggesting that pmf might be even more decreased in these conditions . Nevertheless , deleting ryhB from this strain did not increase its sensitivity to gentamicin during iron starvation ( Fig 2 , right panel ) indicating that the observed DIP enhancing sensitivity of the ryhB mutant was dependent on nuo and sdh . We further assessed the implication of each of the respiratory complexes by testing the sensitivity of the Δnuo and Δsdh single mutants , deleted or not for ryhB ( S3 Fig ) . The nuo simple mutant was almost completely resistant to gentamicin in presence of DIP , whether ryhB was present or not . In contrast , the sdh simple mutant became somewhat more sensitive ( 1 log ) when ryhB was deleted from the chromosome . We conclude from these results that while both complexes are needed for full sensitivity of ryhB mutants to gentamicin , Nuo seems to be slightly more important than Sdh . Results above suggested that RyhB inhibits the activity of both respiratory complexes during iron starvation . To test this , we measured Nuo and Sdh specific enzymatic activities in WT and ryhB mutant strains grown in the presence or absence of the iron chelator DIP . In the WT strain grown in the presence of DIP , Nuo activity was reduced down to 25% as compared with the WT grown in the absence of DIP ( Fig 3A ) . In contrast , Nuo activity was only modestly reduced in the ryhB mutant grown in presence of DIP . The same pattern was also observed for Sdh activity ( Fig 3B ) . Altogether , these results confirm that RyhB represses the activities of both Nuo and Sdh complexes in medium deprived for iron . RyhB inhibition of Sdh and Nuo activities may be due to the repression of the synthesis and / or of the maturation of the complexes . Expression of sdh has already been shown to be repressed by RyhB [20 , 30] . In contrast , although pointed out in global approaches , RyhB regulation of nuo genes expression still awaited investigation [17 , 31–33] . Using the RNA-fold software ( http://unafold . rna . albany . edu ) , we could predict a base-pairing in between RyhB and the 5’ un-translated region of the first gene of operon , nuoA [34] . This base-pairing involves 21 nucleotides ( nt ) of RyhB and includes the ribosome-binding site ( RBS ) and the start codon of nuoA ( Fig 4A ) . Overexpression of ryhB on a plasmid decreased the activity of a PBAD-nuoA-lacZ fusion of about 4-fold , as compared to cells transformed with an empty vector ( Fig 4B ) . In addition , the PBAD-nuoA-lacZ activity was decreased by 2-fold when WT cells were treated with DIP . This was in sharp contrast with the isogenic ryhB mutant strain for which activity remained the same in presence or absence of DIP ( Fig 4C ) . We then tested the biological relevance of the predicted base-pairing by introducing point mutations in the PBAD-nuoA-lacZ chromosomal fusion , giving rise to the nuoAmut-lacZ fusion ( G86C and C87G; Fig 4A ) . In contrast to the WT nuoA-lacZ fusion , RyhB overexpression was no longer able to repress activity of the nuoAmut fusion ( Fig 4D ) . We then introduced compensatory mutations in the pRyhB plasmid that should restore base-pairing to the mutated , but not to the WT , nuo-lacZ fusion , giving rise to pRyhBmut . As seen in Fig 4D , overexpression of RyhBmut failed to fully repress the WT nuo-lacZ fusion but was able to repress nuoAmut-lacZ fusion . Altogether these results strongly suggest that RyhB represses nuo expression by base-pairing on the mRNA upstream nuoA . We then evaluated the effect of this repression on protein levels by performing Western blot analyses against NuoG , a protein of the complex . Strikingly , NuoG protein levels decreased steeply , about 3-fold , when the WT strain was grown in presence of DIP ( Fig 4E ) . This phenotype was suppressed in the ryhB mutant , confirming the in vivo inhibition of Nuo synthesis by RyhB . As a control and to compare sdh regulation to nuo , we performed a series of similar tests on a sdhC-lacZ fusion . We saw that RyhB overexpression repressed the expression of the fusion by more than 10-fold ( S4A Fig ) . In addition , the WT fusion was also strongly inhibited when cells were grown in presence of DIP but not when ryhB was deleted ( S4B Fig ) . Identical conclusions were reached from analyzing SdhB protein levels by performing Western blots ( S4C Fig ) . These experiments thus confirm the regulation of sdh by RyhB . Biogenesis of Fe-S clusters by the Isc machinery has been shown to be key for full Nuo and Sdh activity and their associated pmf production . The iscSUA mRNA is a known RyhB target [16] . Therefore , we asked if RyhB-mediated repression of the iscSUA genes bears any consequence on maturation of Nuo and Sdh . We first checked that RyhB repressed isc expression in our conditions by following levels of the IscS protein after treatment with DIP in a WT and in a ryhB mutant strain ( S5 Fig ) . Interestingly , DIP treatment did not seem to affect levels of IscS in the WT strain . However , in agreement with the study of Desnoyer et al . [16] , levels of IscS rose by a two-fold factor in the ryhB mutant after 90 minutes of DIP treatment . These results thus confirm that RyhB limits the expression of Isc during iron starvation , albeit modestly , counteracting IscR alleviation of repression at the level of the Pisc promoter . We then measured Nuo and Sdh specific activities in strains deleted for suf ( deletion of the whole operon ) or for isc ( iscUA deletion mutant ) , with or without ryhB . We first checked that deleting ryhB from the isc and suf mutants did not perturb growth in presence of DIP ( S6 Fig ) . Growth of the iscUA mutant in LB was slightly slower than the WT strain , as expected from the literature , and was not affected by adding DIP to the medium ( S6A Fig ) [35] . Introducing a secondary ryhB mutation did not change growth of the iscUA mutant . In sharp contrast , growth of the suf mutant was severely affected in presence of DIP ( S6B Fig ) . This was expected since the suf mutant was shown to be essential for growth in defined media containing higher doses of DIP ( ≥300 μM ) [36] . Interestingly , introducing a ryhB mutation in the Δsuf background slightly improved growth , perhaps suggesting that the alleviation of isc repression in the suf mutant may partially restore Fe-S cluster homeostasis . We then measured Nuo activities of the different mutants grown until mid-exponential phase ( OD600 = 0 , 6 ) . In agreement with the literature [7] , Nuo activity was decreased more than 5 fold in an isc mutant wherein the Suf machinery alone is responsible for Fe-S biogenesis ( Fig 5A ) . Nuo activities of the isc ryhB mutant remained low in iron-deprived conditions . Nuo activity of the Δsuf strain was comparable to that of the WT and DIP treatment inflicted the same drop inactivity in both strains . Strikingly however , deleting ryhB in the Δsuf mutant almost completely restored Nuo activity of cells grown in low iron condition . These data strongly suggest that the contribution of Isc to maturation of Nuo complex is of paramount importance , even in growth conditions limited in iron availability . The situation was slightly different for Sdh . Deleting isc severely affected activity of Sdh in presence or absence of iron . Further deleting ryhB from this strain marginally restored Sdh activity , indicating that a significant level of Sdh maturation can be controlled by Suf . In sharp contrast to Nuo however , activity of Sdh was not restored when ryhB was deleted in the suf mutant ( Fig 5B ) . These results thus suggest that Isc cannot ensure maturation of Sdh in low iron conditions . The fact that we could see Nuo activity during iron starvation in the ryhB mutant was surprising as it suggested that there is Isc dependent de novo biogenesis of Fe-S clusters , at least for this complex , in iron limiting conditions . To further test this hypothesis , we measured Nuo activity in cells treated with DIP ( 250 μM ) for one hour . Treatment with DIP induced a slight delay in growth that was identical for both WT and ryhB mutant cells ( Fig 6A ) . In untreated cells , Nuo activity increased with growth as already reported ( Fig 6B ) [37] . In sharp contrast , WT cells treated with DIP for one hour did not show any increase in Nuo activity , while in the ryhB mutant Nuo activity showed a two-fold increase ( 40 to 80 nmol /min /mg of protein ) . Likewise , levels of NuoG protein increased in ryhB mutant cells treated with DIP , but not in the WT strain ( Fig 6C and 6D ) . These experiments thus support the idea that both de novo synthesis and maturation of Nuo take place in ryhB cells treated with DIP . Incidentally , we note that while there is a four-fold increase in the quantity of Nuo proteins in ryhB mutant cells after one-hour treatment , there is only a two-fold increase in Nuo activity . This thus strongly suggests that while Isc is able to ensure maturation of Nuo in iron depleted conditions , it is not as efficient as in iron replete conditions . In order to better appraise the role of Fe-S clusters maturation inhibition by RyhB in the resistance to gentamicin , we performed sensitivity assays in strains containing only one of the two Isc or Suf Fe-S biogenesis machineries . As previously shown , the isc mutant was fully resistant to gentamicin in LB ( Fig 7A ) [7] . This phenotype remained unchanged when DIP was added to the medium , whether RyhB was present or not ( Fig 7A ) , thus showing that the slight Sdh activity observed in these conditions ( Fig 5 ) is not sufficient to render the cells sensitive to gentamicin . In sharp contrast , introducing a ryhB mutation restored sensitivity of a suf mutant strain when grown in presence of DIP ( Fig 7B ) , which is in agreement with the restoration of Nuo activity in this strain under these conditions . As Nuo and Sdh activities are crucial for gentamicin sensitivity , we investigated if we could correlate both the levels of complexes enzymatic activity with that of resistance to gentamicin . Strikingly , there was an almost linear correlation between Nuo or Sdh activities of each strain and its sensitivity to gentamicin ( S7A and S7B Fig ) . For instance , strains displaying the lowest Nuo activities were the most resistant to gentamicin , and vice versa . Phenotypic resistance can take place when environmental conditions change as adaptive molecular responses modify cellular physiology , giving rise to a transient resistance state . Here , we show that the sRNA RyhB is a major contributor of E . coli phenotypic resistance to gentamicin in iron limiting conditions . Aminoglycosides uptake depends upon pmf , which is produced by the activity of respiratory complexes I ( Nuo ) , and , indirectly , by complex II ( Sdh ) . RyhB negatively regulates synthesis of both respiratory complexes . RyhB may also impact activity of Nuo indirectly by limiting the levels of Isc , which we show to be essential for its maturation ( i . e . acquisition of Fe-S clusters ) ( Fig 8 ) . Our model strengthens the role of the pmf-producing respiratory complexes in entry of aminoglycosides . Fe-S biogenesis maturation of the complexes was earlier pointed out as the main factor for resistance [7] . By identifying here that the nuo mRNA is targeted by RyhB in addition to sdh , we show that synthesis of the respiratory complexes is also key in this process . As early as 2005 , the nuo mRNA was suspected to be a target of RyhB as the operon was down-regulated when the sRNA was over-expressed , [17] . The nuo mRNA was also more recently found associated with Hfq and RyhB in a global study of sRNA-mRNA interactions [33] . We here could predict and confirm a direct base-pairing of RyhB to the nuo mRNA at the level of the UTR of nuoA , the first gene of the operon . This base-pairing occurs close to the ribosome binding site of nuoA , which strongly suggests that RyhB represses expression of nuo by occluding binding of the ribosome , and subsequently degradation of the mRNA [38] . The nuo mRNA is very long ( about 15 kb ) and comprises 14 genes , which makes it one of the longest mRNAs regulated by a sRNA to our knowledge . Importantly , in addition to the effects seen on nuoA expression by using beta-galactosidase assays as a read-out ( Fig 4 ) , we could also observe repression at the level of the NuoG protein by using Western blots analysis ( Fig 4E ) . The structural nuoG gene lies more than 5 kb away from the RyhB/Nuo base-pairing site . It will thus be interesting to investigate how far downstream the nuo operon RyhB repression propagates . Respiratory complexes are high iron consumers , with a total of 12 Fe-S clusters for Nuo and Sdh in E . coli . Thus , their repression by RyhB is in line with its role in installing an iron sparing response when iron becomes scarce [17 , 19] . Before our results , one could have imagined that RyhB represses Nuo and Sdh expression in order to limit accumulation of inactive apo-complexes in iron scarce conditions . However , both protein levels and activity of Nuo are restored in a ryhB mutant in iron-deprived medium indicating that maturation of respiratory complex I is possible under these conditions . These results strongly suggest that RyhB inhibits synthesis of Nuo Sdh to preclude respiratory complexes to divert iron from other essential processes . We here show that Isc is essential for Nuo maturation when iron is depleted in the ryhB mutant ( Fig 3A and Fig 5A ) . In contrast , maturation of Sdh was only partially restored in the ryhB mutant in presence of DIP ( Fig 3B ) and , perhaps more surprisingly , this activity did not seem to be dependent on Isc but rather on Suf ( Fig 5B ) . More investigation is needed to understand the molecular basis for the difference in between Isc and Suf substrates preference . In any case , our results also clearly show that Nuo activity is more important than that of Sdh in installing a phenotypic resistance to gentamicin ( S2 Fig ) . This may relate to pmf production by Nuo and Sdh . Indeed , Nuo , but not Sdh , directly translocates 4 protons across the membrane while both indirectly contribute to pmf production by passing electrons to cytochrome oxidase [12 , 29] . Our experiments clearly show that there is de novo synthesis and biogenesis , at least of Nuo complexes , during iron starvation in the ryhB mutant ( Fig 6 ) . We also confirm that the maturation of this respiratory complex depends on the Isc machinery . Taken together , these results indicate that the Isc machinery can be functional during iron starvation and mature at least the Nuo complex . In agreement with previous results from the Massé laboratory [16] , we also show that RyhB limits levels of the Isc machinery during iron starvation . Thus , a tempting hypothesis is that RyhB inhibits indirectly Nuo activity by limiting its maturation by Isc . However , given that RyhB effect on Isc is relatively modest , we cannot exclude that producing Nuo complexes alone , even while keeping Isc synthesis repressed by RyhB , may be sufficient to restore Nuo activity . Furthermore , the situation might be even more complex as iron depletion is likely to modify to different extent both the levels and activity of all of the proteins intervening in this process , namely Isc , Suf , Nuo and Sdh . Thus , fully testing the hypothesis that RyhB-mediated reduction of Isc synthesis will bear an effect on Nuo/Sdh activity will require a thorough assessment of both the concentration and the activity of all components cited above . For the time being , we consider the hypothesis of an indirect control of Nuo maturation as a likely contribution to the RyhB dependent phenotypic resistance we observed . Nevertheless , the fact that Isc is able to maturate Nuo in iron deprived conditions may seem contradictory with previous studies that have shown the Suf system to be essential during iron starvation [36] . In agreement with that idea , we have seen that deleting ryhB partially suppressed the growth defect phenotype of a suf mutant grown in the presence of DIP ( S6 Fig ) . However , growth of this mutant is not restored to wild type-like levels , indicating that while when overexpressed Isc may promote Fe-S cluster biogenesis , during iron starvation , it is not as efficient as the Suf machinery and thus explaining the need for a second iron limitation stress responsive system . Our study puts RyhB on the focus among a growing number of sRNAs that have been directly or indirectly linked to antibiotic resistance [39–41] . However , in most of these cases phenotypes were derived from overexpression of the sRNAs and such situations might not be relevant to physiological conditions . For instance , 17 out of 26 E . coli sRNAs that were assessed in a systematic manner against a variety of antibacterial effectors were shown to affect sensitivity to antibiotics when overexpressed , but few showed any phenotype when mutated [42] . A most spectacular case is represented by the role RyhB could play in the bacterial persistence of uropathogenic E . coli to different classes of antibiotics , among which gentamicin [43] . Persistence is a phenomenon in which a fraction of the bacterial population enters a metabolically inactive state that enables it to survive exposure to bactericidal antibiotics [44] . It was proposed that ryhB mutants would induce less persister cells because they display increased ATP levels and altered NAD+ / NADH ratios . In the light of our results , we believe these effects could also be explained by the fact that ryhB mutants display higher levels of Nuo , Sdh and Isc and therefore are more metabolically active , but also more prone to uptake the antibiotic . It is noteworthy that these experiments were conducted in rich medium not devoid for iron , and after long treatment with antibiotics ( four days ) , which may explain low induction of RyhB in only a small percentage of bacterial cells that would then be able to resist antibiotics treatment in a persister-like manner . RyhB homologs and paralogs are found in multiple other bacterial species , which suggests that many bacteria outside of E . coli may share the resistance mechanism that we describe here [45] . In particular , other pathogenic bacteria such as Yersinia , Shigella or Salmonella possess not only RyhB homologs , but also the Isc and Suf system and rely on Nuo and Sdh for respiration on oxygen [46 , 47] . RyhB has also been implicated in promoting sensitivity to colicin IA , which is not an antibiotic in a narrow sense , but a bacteriocin secreted by other species to outcompete bacteria sharing the same niches [48] . In addition , RyhB has been shown to be involved in the virulence of Shigella dysenteriae by repressing the major virulence regulator virB , and the sRNA may be associated with the virulence of Yersinia pestis , as the expression of its two RyhB homologs ( RyhB1 and RyhB2 ) increases in the lung of infected mice [49 , 50] . Altogether , these data point out for a major role for RyhB in escaping antibacterial action . All strains used in this study are derivatives of E . coli MG1655 and are listed in S1 Table . Strains were grown in LB broth ( Difco ) , containing various concentrations of 2 , 2’-dipyridyl ( DIP ) ( Sigma ) when stated . Transductions with P1 phage were used for moving marked mutation as described previously in [51] . The plac and pRyhB plasmids used in this study are described and have been transformed as previously described in [52] . All oligonucleotides used are listed in S2 Table . Starting from overnight cultures in LB , strains were diluted 1/100 time in fresh medium containing or not DIP and grown aerobically at 37°C with shaking until OD600 ≈ 0 . 2 . At this point , antibiotics were added to the cells ( gentamicin: 5 μg / mL; ampicillin: 5 μg / mL; tetracycline: 5 μg / mL and norfloxacin: 25 ng / mL ) . After 3 h cells were taken , diluted in PBS buffer and spotted on LB agar plates and incubated at 37°C for 16 h . Cell survival was determined by counting the number of colony-forming units per mL ( c . f . u . / mL ) . The absolute c . f . u at time-point 0 was of ≈ 5 x 107 cells / mL in all experiments . The MIC were determined as previously described [53] . Briefly , each antibiotic containing-well ( with 0; 2 , 5; 3 , 75; 5; 6 , 25; 7 , 5; 8 , 75; 10; 11 , 25; 12 , 5; 13 , 75; 15 μg / mL of gentamicin respectively ) of a 96-well micro-titer plate was inoculated with 100 μL of a fresh LB bacterial inoculum of 2 × 105 c . f . u / mL . The plate was incubated at 37°C for 18 h under aerobic conditions . OD600 for each well was then determined by measuring the absorbance on a Tecan infinite 200 . MIC was defined as the lowest drug concentration that exhibited complete inhibition of microbial growth . The PBAD-nuoA-lacZ and PBAD-sdhC-lacZ fusions were constructed and recombined in PM1205 strain , as previously described [25] . Briefly , sequences corresponding to nuo or sdh genes starting from its +1 transcriptional start up to 30 nucleotides downstream of the ATG codon were amplified using oligonucleotides PBAD-nuoA-F or PBAD-sdhC-F , and lacZ-nuoA-R or lacZ-sdhC-R , respectively . PCR amplifications were carried out using the EconoTaq DNA polymerase from Lucigen . The purified PCR products were then electroporated into strain PM1205 for recombination at the lacZ site . Recombinants carrying the desired fusions ( SC005 and SC009 ) were selected on LB plates devoid of NaCl and containing 5% sucrose , 0 , 2% arabinose and 40 μg / mL X-Gal ( 5-bromo-4-chloro-3-indolyl-D-galactopyranoside ) . Blue colonies were chosen , and the resulting fusions were sequenced using oligonucleotides lacI-F and Deep-lac . Overlap PCR was used to introduce point mutation in the fusion . The two PCR products corresponding to the sequence upstream and downstream of the desired mutation were amplified by PCR with oligonucleotides nuoAmut-F and Deep-lac , and LacI-F and nuoAmut-R containing the desired mutation and using genomic DNA from the SC005 strain as a template . The two PCR products were then joined by an overlap PCR using oligonucleotides lacI-F and Deep-lac . The resulting PCR products were purified , electroporated in strain PM1205 and sequenced as described above . For point mutations in the pRyhB plasmid , the pRyhB plasmid was first purified from a WT ( dam+ ) E . coli strain , and then amplified by PCR with oligonucleotides RyhBmut-F and RyhBmut-R , containing the desired mutation . The native plasmid was eliminated from the resulting PCR product by Dpn1 enzyme digestion for 1 h at 37°C . Plasmids containing the desired mutation were then purified and transformed in SC005 and SC0026 strains . Overnight cultures of different strains were diluted 1/100 times in fresh medium in culture flasks containing ampicillin and IPTG ( isopropyl ß-D-1thiogalactopyranoside ) or DIP when indicated . After ≈ 7 hours of growth 100 μL of cultures were dispatched in 96 wells microtiter plates ( triplicates for each conditions ) . Absorbance at 600 nm was measured in a microtiter plate reader ( Tecan infinite 200 ) . Then , 50 μL of permeabilization buffer were added in each well ( 100 mM Tris HCl pH 7 , 8; 32 mM Na2HPO4; 8 mM EDTA; 40 mM Triton ) and the microtiter plate was incubated for 10 minutes at room temperature . O-Nitrophenyl-β-D-galactopyranoside ( ONPG ) was added to the solution and appearance of its degradation product was immediately determined by measuring the absorbance at 420 nm on a Tecan infinite 200 during 30 minutes . The specific activities were calculated by measuring the Vmax of the OD420 appearance divided by the OD600 . Values were then multiplied by 100000 , a coefficient that was chosen empirically to approximate Miller units . The Nuo and Sdh enzymatic activities were determined as previously described [54 , 55] . Briefly , overnight cultures of the strains of interest were diluted 1/100 times in fresh LB medium containing or not 250 μM of DIP and grown at 37°C with shaking until they reached OD600 ≈ 0 . 6 . Cultures were pelleted by centrifugation ( 11 000 G , 10 min at 4°C ) and washed in phosphate buffer ( 50 mM pH 7 , 5 ) . Cells were then lysed at the French press and 100 μL were immediately frozen in liquid nitrogen before determining Nuo activity . Nuo enzymatic activity was determined at 30°C by monitoring the disappearance of the specific Deamino-NADH ( DNADH ) substrate at 340 nm every 5 s during 10 min at 30°C in a spectrophotometer . For Sdh activity determination , lysate samples from French press were pellet by centrifugation ( 11 000 G , 10 min at 4°C ) and the supernatant was used for membrane fraction preparation by ultracentrifugation at 45 000 G at 4°C during two hours . Pellets were then resuspended in phosphate buffer and kept in liquid nitrogen for later Sdh activity measurements . The enzyme was first activated by incubation in 50 mM Tris-HCl ( pH 7 . 5 ) , 4 mM succinate , 1 mM KCN for 30 min at 30°C . The enzymatic activity was measured in the membrane fraction by monitoring Phenazine EthoSulfate ( PES ) -coupled reduction of dichlorophenol indophenol ( DCPIP ) at 600 nm , in a reaction containing 50 mM Tris-HCl ( pH 7 . 5 ) , 4 mM succinate , 1 mM KCN , 400 μM PES and 50 μM DCPIP . The specific activities were calculated by measuring the Vmax divided by the protein concentration in total extracts evaluated by absorbance at 280 nm . Total extracts and membranes preparation prepared for Nuo and Sdh activities were used for quantification of Nuo and Sdh protein levels , respectively . Total protein levels were determined by measuring absorbance at 280 nm on a spectrophotometer . Same amount of total protein level were migrated on poly-acrylamide gels Tris-gly Sodium Dodecyl Sulfate ( Novex 4–20% Tris-Glycine Mini Gels ) then , transferred on nitrocellulose membrane using Pierce G2 Fast Blotter ( 25 V , 1 , 3 mA , 7 min ) . Protein level were detected by incubating the membrane with α-NuoG , α-SdhB , or α-IscS ( 1/1000 ) antibodies from rabbit and then by an α-rabbit antibody ( 1/1000 ) coupled with Hrp peroxidase . Signals were detected by chemiluminescence with Pierce ECL Western blotting system on an ImageQuant LAS 4000 camera . Quantification of protein levels was determined by measuring the specific signal intensity of the bands corresponding to Nuo , Sdh or IscS proteins with the ImageJ software . Intensities were normalized using an unspecific band detected by the same antibody .	Understanding the mechanisms at work behind bacterial antibiotic resistance has become a major health issue in the face of the antibiotics crisis . Here , we show that RyhB , a bacterial small regulatory RNA , decreases the sensitivity of Escherichia coli to the antibiotic gentamicin when iron is scarce , an environmental situation prevalent during host-pathogen interactions . This phenotypic resistance is related to the activity of the respiratory complexes Nuo and Sdh , which are producing the proton motive force allowing antibiotic uptake . Altogether , this study points out to a major role for RyhB in escaping antibacterial action .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "antimicrobials", "cell", "physiology", "medicine", "and", "health", "sciences", "drugs", "microbiology", "operons", "antibiotic", "resistance", "mutation", "antibiotics", "pharmacology", "dna", "molecular", "biology", "techniques", "research", "and", "analysis", "methods...	2019	A small RNA controls bacterial sensitivity to gentamicin during iron starvation
Accurate inference of molecular and functional interactions among genes , especially in multicellular organisms such as Drosophila , often requires statistical analysis of correlations not only between the magnitudes of gene expressions , but also between their temporal-spatial patterns . The ISH ( in-situ-hybridization ) -based gene expression micro-imaging technology offers an effective approach to perform large-scale spatial-temporal profiling of whole-body mRNA abundance . However , analytical tools for discovering gene interactions from such data remain an open challenge due to various reasons , including difficulties in extracting canonical representations of gene activities from images , and in inference of statistically meaningful networks from such representations . In this paper , we present GINI , a machine learning system for inferring gene interaction networks from Drosophila embryonic ISH images . GINI builds on a computer-vision-inspired vector-space representation of the spatial pattern of gene expression in ISH images , enabled by our recently developed system; and a new multi-instance-kernel algorithm that learns a sparse Markov network model , in which , every gene ( i . e . , node ) in the network is represented by a vector-valued spatial pattern rather than a scalar-valued gene intensity as in conventional approaches such as a Gaussian graphical model . By capturing the notion of spatial similarity of gene expression , and at the same time properly taking into account the presence of multiple images per gene via multi-instance kernels , GINI is well-positioned to infer statistically sound , and biologically meaningful gene interaction networks from image data . Using both synthetic data and a small manually curated data set , we demonstrate the effectiveness of our approach in network building . Furthermore , we report results on a large publicly available collection of Drosophila embryonic ISH images from the Berkeley Drosophila Genome Project , where GINI makes novel and interesting predictions of gene interactions . Software for GINI is available at http://sailing . cs . cmu . edu/Drosophila_ISH_images/ GINI first extracts the gene expression pattern from each image using a computer version driven image analysis pipeline [15] . These expression patterns are spatially aligned and normalized to enable spatial comparison of gene expression across multiple images . Next , the expression patterns are represented by suitable standardized features through a process called “triangulation” , followed by feature normalization and selection . Since each gene may have a different number of images in the data , each gene can now be represented by a “bag” or a set of feature vectors - one feature vector per image . Thus , our task is to estimate the gene network , given bags of images per gene ( Figure 1 ) . We cast the problem of estimating a gene interaction network as the task of estimating the graph structure of a Markov random field ( MRF ) over the genes . The underlying graph encodes conditional independence assumptions between the genes , that is , two genes are said to not interact in the network if their gene expressions are conditionally independent of each other , conditioned on the expression of all other genes in the network . We employ multi-instance kernel technique using different order statistics to compute similarity between bags of images . We then estimate a sparse network of gene interactions by modeling the data as a multi-variate multi-attribute Gaussian , and estimating the sparse inverse covariance matrix of the model . A schematic diagram of the system pipeline can be seen in Figure 2 . GINI is a bioimage informatics system based on a computer vision pipeline for ISH micro-image processing and a statistical learning algorithm for network inference . The main contributions of this work are summarized below . First , the image analytic pipeline used by GINI offers a rigorous and universal approach to extract a standardized representation of spatial patterns of gene expressions . Comparing to the popular SIFT features [24] , which is based on detecting interest points with heavy assumptions on object shape , texture , and other physical properties originally meant for natural objects , our approach is more suitable for ISH staining in Drosophila embryos which do not resemble natural objects and require preservation of overall spatial shape and overall intensity information in a canonical way for intra-gene normalization and inter-gene comparison . Second , GINI infers a network that enjoys the Markov network property: it gives globally consistent conditional-independency interpretation for every edge , and therefore is biologically more meaningful . It is known that marginal correlation ( as often used in estimating an ad hoc network ) , which is computed for every gene-pair in isolation ( i . e . , ignoring all other genes in the system ) , confounds direct and indirect dependencies , and could result in a clique-like dense graph or subgraph among genes that are not directly dependent , but have a long-distance interaction . Studying conditional independencies in a network allows us to predict interactions between a pair of genes in the context of other genes , allowing a distinction to be made between direct and indirect relationships between the genes , and reducing false positives . Third , our formulation based on Gaussian Markov random field and multi-instance kernel for the GINI network is convex , hence the globally optimal estimator of the network is computed , no approximations are involved . Furthermore , under suitable conditions , our graphical model learning algorithm is sparsistent , i . e . as the amount of available data increases , the algorithm is statistically guaranteed to predict the correct interactions between the genes . While Bach et . al . [27] have previously proposed learning the structure of graphical models from data using Mercer kernels , their approach is based on a non-convex local greedy search to find edges in the graph . Our approach represents the first work that uses Mercer kernels and Gaussian Graphical Models to predict kernelized graphical models using a convex formulation . Finally , with the GINI system , we were able to systematically perform a genome-scale network learning and analysis of the genes expressed during 2 time points of Drosophila embryogenesis recorded by ISH imaging from BDGP [16] . In both time points , we find that the GINI networks are modular and scale free , which are properties predicted to hold true in gene interaction networks . Further , different regions of the networks are enriched for spatial annotations , thus GINI is able to cluster spatially similar genes . The hubs of the networks , i . e . , the genes with the largest number of predicted interactions are functionally enriched for important cellular functions . We demonstrate that the networks predicted by analyzing microarray data does not have either spatial or functional enrichment , thus these results could not have been obtained by analyzing microarray data . To the best of our knowledge , GINI represents one of the first efforts to reverse engineering gene networks from ISH image data . In both extensive simulation studies and empirical biological analysis , we demonstrate the effectiveness of GINI in predicting networks , and show that the statistical assumptions behind GINI are reasonable , and the biological analysis enabled by GINI merits close examination and further exploration . We first show how GINI estimates a gene network , when each gene has only one image . The next subsection extends the GINI algorithm to deal with multiple images per gene . Let denote the set of genes being studied , so that is the gene , where , and is the number of features extracted per image . Each feature represents the gene expression in a spatial location of the embryo . Note that algorithms that analyze microarray data typically treat samples drawn from different time points as independent samples [28] , even though expressions of the same gene across time is expected to be auto correlated . We similarly assume that the different spatial features are independent of each other . The spatial independence assumption has also been implicitly made by [29] , [30] while modeling transcription networks in Drosophila embryos . In the results section , we use simulated data to demonstrate that this assumption does not affect the accuracy of the algorithm significantly . By modeling the gene interactions as invariant across the spatial locations in the embryo , we can assume that each feature is independently and identically drawn ( i . i . d . ) from the same distribution . Inferring gene interactions is then equivalent to modeling the dependence between the expression values of different genes at the same spatial location . Expression of the genes in each spatial location is assumed to be drawn from some ( multi-variate ) distribution , independent of all other spatial locations . Each spatial feature ( ) may be modeled as a vector of length , with capturing the expression value of the gene in this location . This gives us independent samples with which the parameters of the underlying distribution may be learned . Formally , let each spatial location be drawn independently from a multi-variate Gaussian , where is the mean vector , and is the positive semi-definite covariance matrix between the genes . In a multivariate Gaussian distribution , the entry of the inverse covariance matrix is zero if and only if the corresponding genes are conditionally independent given the rest of the graph . Thus , the non-zero entries of the inverse covariance matrix correspond to edges in the corresponding Gaussian Markov random field , giving rise to the gene interaction network . The Gaussian Markov random field is also known as a Gaussian graphical model ( GGM ) [31] . Since we expect a small number of interactions per gene , the estimated graph must be sparse , i . e . the number of non-zero entries of the inverse covariance matrix must be small . Thus , the gene interaction network may be estimated by learning a Gaussian distribution from the observed images , such that the inverse covariance matrix is sparse . The mean of the Gaussian is estimated by the observed sample mean , ( 1 ) Then , the inverse covariance matrix can be estimated by minimizing the negative log-likelihood of the data , over all possible positive semi-definite matrices . To enforce sparsity , the norm of , which counts the number of non-zero elements , is added to the negative log likelihood . Since optimizing the norm is non-convex and NP hard , the norm is used as a convex relaxation to the norm . The norm of a matrix is the sum of the absolute values of the elements of the matrix , and also enforces sparsity in the solution . Adding the norm regularization also ensures that the minimizer of the objective function exists , and is well defined . Thus , our objective function is ( 2 ) where is the second moment matrix about the mean ( 3 ) is a tuning parameter , by which we determine the strength of the penalty . As we increase the value of , we increase the penalty on the absolute values of , and hence , the graph induced by becomes more sparse . The edges in the graphical model are then estimated as ( 4 ) Multiple images of the same gene at the same time point should have the same gene expression pattern . However , in practice , the expression patterns in different images may differ considerably , for three main reasons . Firstly , there is a wide interval of time considered as a single time point while collecting such data . For instance , the BDGP data divides embryonic development into 6 time stages . The last stage 13–16 corresponds to development of the embryo 9 . 3 to 15 hours after fertilization , which represents more than a third of the time taken for embryonic development . Hence , the true gene expression pattern may be dynamic within the time period of a single development stage , and the gene expressions captured for the same gene at the same time may not look similar to each other . Secondly , we might expect that for any organism for which ISH data is collected , there will necessarily be some ambiguity in how the development stage of the organism is labeled by human annotators . Finally , noise in the expression patterns due to excessive staining , lighting conditions and similar other reasons will also be observed . For all of the above reasons , any network-learning algorithm should leverage the existence of multiple images per gene per time point in improving its estimates of gene similarity . The problem of multiple images per gene is reminiscent of multi-instance learning [35] , [36] . Multi-instance learning is a form of supervised learning , where instead of labeling each instance , a bag of instances is labeled . A popular solution to the multi-instance problem is to define a multi-instance kernel , that can compute the similarity between bags of instances . Let be a collection of order statistics of the set , for example , mean , median , minimum , maximum etc . In dimensions , is computed on each dimension independently , to form a vector of order statistics . If we use order statistics , then the length of will be . The similarity between gene with a set of images and gene with images can then be computed as ( 5 ) where is an appropriate kernel function between vectors and . Such a kernel is called the statistic kernel . The choice of the order statistics used in the kernel depends on the data collection procedure of the ISH . One concern in ISH data is that images may be overstained . In such a scenario , the median may be an appropriate choice of order statistic . If over-staining is not a concern , the maximum statistic may be more appropriate to ensure that information about presence of gene expression is not lost . For the BDGP data , we use the covariance kernel , and the mean statistic . The choice of using a single statistic to represent information from multiple images was due to the presence of noisy images in the data set . Thus , ( 6 ) Thus , our choice of kernel is equivalent to computing the mean similarity of all pairs of images in bags and . This specific kernel is also known as the normalized set kernel , and has been shown to perform very well in multi-instance classification [37] . Any kernel function may be written as the dot product in some higher dimensional feature space , i . e . [38] . Hence , if we assume that the data is drawn from a distribution such that is a zero-mean Gaussian , we can learn the gene interaction network by treating as the sample covariance matrix . Since estimating the inverse covariance matrix by solving equation 2 requires only the sample covariance matrix and not the data itself , we can kernelize it by using the kernel matrix defined in equation 6 as the required sample covariance matrix . Thus , the objective function is ( 7 ) which can be solved as discussed in the previous section . We convert the ISH images into canonical feature vectors suitable for analysis by our algorithm described above in a three-step manner . First , the precise expression pattern found in each image is extracted and aligned spatially to make all images spatially comparable . Next , each image is represented by a feature vector using Delaunay triangulation . Finally , features are normalized and feature selection is performed to extract meaningful features , that can be then used to compute the multi-set kernels to obtain gene similarity and learn the gene network . Putting everything together , we conclude the method section with a summary of the GINI system for network inference from ISH images . Each ISH image is converted into a standardized expression pattern using , and then triangulated to extract a low-dimensional spatial feature vector . Next , feature values are normalized , uninformative features are removed , and genes with insufficient information available are rejected . Finally , the multi-set kernel is used to compute the similarity between the bags of image vectors available for each gene , and the gene network is estimated using Equation 7 . The algorithm is summarized in Algorithm 2 . GINI assumes that the gene expression in each triangle can be assumed to be independently drawn from a multi-variate Gaussian . However , the true gene expression in adjacent spatial locations is correlated and not independent . To verify that this dependence of adjacent samples does not affect the accuracy of the estimated network , we simulate synthetic data where the underlying network is known , but the data points are not independent of each other , and test whether GINI can recover the correct network in such a scenario . The data samples depend on each other via a parameter that captures degree of dependence between data samples . When , all data samples are drawn i . i . d . from the known distribution . As increases , data samples are drawn from the same distribution , but they depend on the adjacent samples . For a high-dimensional distribution , it is not feasible to test if the data is truly Gaussian . However , a consequence of Gaussianity is that for each gene , the gene expression can be expressed as a weighted linear sum of the expression values of a few other genes , which form the edges of the network . To test if this assumption holds true in ISH data , for each gene , we fit a linear regression between the gene and its neighbors found by GINI and look at the absolute value of the error i . e . the mean absolute difference between the predicted and the known gene expression . When the maximum expression value is 1 , for more than 90% of the genes we looked at , the absolute error was less than 0 . 02; 99 . 5% of all genes had absolute error less than 0 . 05 , confirming that the GINI generative model explains the ISH data . We also confirm that the prediction error is not systematic with respect to the spatial location . For each gene , we compute the prediction error ( residue ) when the gene is predicted by regressing it on its neighbors . For each spatial location , we plot the mean residue at that location for all genes . As can be seen in Figure 5 , there is no systematic bias in the spatial positions that are hardest to predict for any gene . Before running our algorithm on a large sized dataset , we construct an artificial small data set to verify the results . We input 12 images , shown in Figure 6 ( a ) from 6 genes to the GINI algorithm ( each gene has 1–3 images in the data set ) . With , 4 edges are predicted in the network , shown in Figure 6 ( b ) . As can be seen , the three genes hunchback ( hb ) , four-jointed ( fj ) , and Blimp-1 , which are expressed in the dorsal , ventral and procephalic ectoderm , are connected in a single cluster . Similarly , the genes organic anion transporting polypeptide 74D ( Oatp74D ) and bicoid ( bcd ) are connected by an edge , since both show expression in the foregut and the anterior endoderm . Finally , the expression of sloppy paired-1 ( slp1 ) was considered to be sufficiently different from the other genes , hence it is not connected to any other gene in the network . Thus , the gene interaction network found by GINI can be verified to be reasonable for the above small data set . We now turn our attention to the ISH images from the Berkeley Drosophila Genome Project data set . We have obtained around 67400 ISH images of 3509 protein-coding genes from the BDGP data released in September 2009 , captured at key development stages of embryonic development . Each image captures embryonic gene expression of a single gene using RNA in-situ hybridization . Each image was labeled manually with the age of the embryo , categorized into six distinct embryonic stages : 1–3 , 4–6 , 7–8 , 9–10 , 11–12 , and 13–16 . Genes are also annotated with ontology terms from a controlled vocabulary of around 295 terms , describing the unique embryonic structures in which gene expression is observed during the various stages of embryonic development . analyzes these image automatically , rejecting unsuitable images , to produce 51593 expression patterns of 3347 genes . As proof of concept , we focus on images viewed from a lateral perspective from two development stage ranges of this data : 9–10 and 13–16 . For the stage 9–10 , we have 2869 expression patterns of 2609 genes , and for stage 13–16 , we have 6350 expression patterns of 3258 genes . We extracted features as described in the methods section . For each development stage , we ran a separate analysis . Using a value of 0 . 775 for stage 9–10 , we ran GINI and obtained a network having 258 genes , and 516 interactions ( edges ) between them . For the development stage 13–16 , we used , and obtained a network with 1202 genes and 3666 interactions between them . The value was selected for each network by running GINI for 21 values between 0 . 5 and 1 , and picking a value such that the mean-degree for the network is reasonable ( approximately 2–3 ) - see Supplementary Figure S1 for a plot that shows how the number of edges in the network decreases as increases . Some of the interactions predicted by GINI have already been reported in the literature . For example , in the network for stage 9–10 , GINI predicts that DCP-1 ( CG5370 ) , an effector caspase which is involved in apoptosis , will interact with the thread gene ( CG12284 ) , a known inhibitor of apoptosis protein [43] . GINI also predicts that Snf5- related 1 ( CG1064 ) interacts with echinoid ( CG12676 ) , both of which are known to be involved in epidermis development , muscle organ development , as well as imaginal disc-derived wing vein morphogenesis . In the 13–16 development network , GINI predicts that the capping protein beta gene ( CG17158 ) interacts with the Glycogen phosphorylase gene ( CG7254 ) , and Tpc1 ( CG6608 ) interacts with CG2812 , which has been previously reported in [44] . The next five subsections do a detailed analysis of the 2 networks . A network is said to be scale free if its degree distribution asymptotically follows a power law . That is , the fraction of genes that have at least interactions with other genes is ( 10 ) where is the scale free parameter , and is the normalization constant . It has been hypothesized that gene regulatory networks are scale free [10] . We looked at the characteristic of our interaction networks by plotting the number of interactions per gene ( Figure 7 ) , and found that the networks found by GINI are scale free . The parameter obtained is 2 . 3 and 2 . 5 for the 9–10 and 13–16 networks respectively , which corresponds well to the values observed for a large variety of power law graphs . The scale free nature of the network was found to be independent of the tuning parameter of the algorithm . Unlike the gene regulatory network obtained for Human-B cells [10] , we found that the scale-free nature of the gene network we obtain has a good fit , without observing a deviation from the expected at low connectivity values . However , this could be a side-effect of the larger number of genes they analyzed . Using spectral clustering , we construct 12 regions or clusters within each network , and visualize the five biggest clusters of each of the networks in Figure 8 . All 12 clusters in both networks are very well separated . The ratio of within-cluster edges to total number of edges is 70% and 87% for the 9–10 and 13–16 development stage networks respectively , indicating that the estimated networks are highly modular . From a biological perspective , different parts of gene networks may be responsible for different pathways or biological functional components of the cell , thus modularity is a good prediction for real interaction networks . Given the scale-free nature of the network , a small number of the genes have a large number of interactions . We analyze the Gene Ontology functions of the genes having the largest number of interactions , i . e . the hubs of the network . The question we wish to address is: if we pick the top 5% of the genes having the maximum connectivity with other genes , what kind of functional enrichment do these genes have ? Our background population is of the 2609 and 3258 genes for which we have at least one ISH image describing its expression for the 9–10 and 13–16 stages respectively . We use the hypergeometric test , with Bonferroni correction used to correct for multiple hypothesis tests [45] . As can be seen in Table 1 , we observe enrichment of a wide variety of functions that are essential to cell growth and functioning , including metabolic processes , cellular respiration , transport of electrons and ions , protein modification , ribosome biogenesis etc . Next , we examine a few high-degree hubs in the two networks in detail , along with their neighborhood genes in the networks . Figure 9 shows the hub neighborhood for two genes in the 9–10 development stage network . CG3969 is a Activated Cdc42 kinase-like gene known to be involved in protein phosphorylation [46] and cell death [47] , and CG9984 ( TH1 ) is known to be involved in regulation of biosynthetic process [48] and nervous system development [49] . Both genes interact with many genes having functions related to the primary metabolic process , and single-organism cellular process . In stage 13–16 , we examine the hub neighborhood of CG5904 and CG6501 . The mitochondrial ribosomal protein CG5904 has been previously predicted to be a structural constituent of ribosome [50] , and we find that it interacts with many genes involved in the ribosome biogenesis . Gene CG6501 ( Ns2 ) has been previously predicted to be involved in phagocytosis , engulfment [51] , and ribosome biogenesis [46]; CG6501's neighborhood has multiple genes that are also involved in ribosome biogenesis and single-organism cellular process . Each gene in the BDGP data has been labeled manually by annotations describing the spatial gene expression , using 295 annotation terms . We expect that since the gene interaction network is constructed via spatial similarity , genes that are connected to each other in the network will have similar spatial annotation terms . To test this , we cluster the gene network using spectral clustering [52] into 12 clusters , and analyze the enrichment of each cluster for annotation terms using the hypergeometric test , with Bonferroni correction used to correct for multiple hypothesis tests . In the gene network for the 9–10 stage , 11 of the 12 clusters are enriched for 63 total annotation terms ( Figure 10 ) . The only cluster not showing any enrichment in the 9–10 stage network is also the smallest cluster , having only 4 genes . For example , in cluster 8 , 92% of the genes have expression in the ventral nerve cord primordium P3 , while only 8% of the genes in the data have expression in this region . Similarly , 73% of the genes in cluster 11 have expression in the trunk mesoderm primordium , while only 16% of the genes in the data have expression in this region . For the 13–16 stage network , all 12 clusters are enriched for a total of 81 enrichments , a part of which is visualized in Figure 10 . Tables S1 and S2 in the supplementary material report the complete enrichment analysis . We learn a network from microarray data collected by the BDGP project over 12 time points in embryonic development [16] , over the same genes that are being studied in the 9–10 and 13–16 networks , using covariance between the microarray expression as the kernel . We find that the overlap in edges between the 2 networks is very small , only 1% of the edges are common to both networks . If we assume that spatial expression annotations are a proxy for functional enrichment , then we can check if the microarray network is enriched for the spatial annotation terms . Figure 14 shows that the percentage of enriched clusters in the microarray network is small , independent of the number of clusters analyzed . We can also test functional GO enrichment of the hubs of the network . Table 2 shows that the hubs of the microarray network for stage 13–16 are enriched for only a single function , where 4 of the 145 hub genes are involved in the “aromatic compound catabolic process” , while the microarray data network for stage 9–10 has no enrichments . Thus , we find that the network learned from ISH images is clearly different from a network learned from microarray data . The ISH image network is enriched for spatial annotation terms , as well as functional enrichment of the hubs of the network , which does not hold true for the microarray network . This suggests that analyzing ISH images could support different scientific conclusions , which should be studied in greater detail . GINI predicts gene interaction networks by analyzing Drosophila embryo ISH images . While the experiments above have been reported on the ISH data from BDGP , the GINI algorithm can be applied to all image data , by suitably modifying only the image processing pipeline . Using synthetic and image data , we establish that GINI fits the ISH data well , with low error residues , and that it can learn the true network correctly even if the data is not completely i . i . d . The analysis of the BDGP data shows that the hubs of the predicted gene interaction network are enriched for essential cellular functions , and that different regions of the interaction network are enriched for different combinations of annotation terms describing the gene expression . Thus , the predicted gene interaction network is capturing essential spatial and functional information about the expression pattern of the genes . We found that the gene interaction network learned from ISH images differs significantly from a network learned from microarray data . The current work focuses on extracting gene networks from spatial data . The next step is combining information from multiple time stages to improve predictions , thus learning spatial-temporal gene networks . The problem of time-varying networks has been studied extensively for microarray data , by using different statistical penalties to estimate the network . For example , Ahmed et . al . [22] construct time varying networks by using a temporally smoothed -regularized logistic regression formulation , while Danaher et . al . [53] propose a fused lasso and group lasso based approach to combine information across time . Extensions of such algorithms for image data require stronger assumptions on data quality , such as having the same number of genes and image quality across time . Further , certain development stages may be less informative than others; for example , very few genes are active at development stage 1–3 , and expression data from this stage is not as informative as expression data from development stage 13–16 , when the embryo is much more mature . Developing algorithms that can account for such variations in data quality , while combining information across time , remains an interesting future direction to explore .	As high-throughput technologies for molecular abundance profiling are becoming more inexpensive and accessible , computational inference of gene interaction networks from such data based on well-founded statistical principles is imperative to advance the understanding of regulatory mechanisms in various biological systems . Reverse engineering of gene networks has traditionally relied on analysis of whole-genome microarray data; here we present a new method , GINI , to infer gene networks from ISH images , thereby enabling exploration of spatial characteristics of gene expressions for network inference . Our method generates a Markov network , which encapsulates globally meaningful statistical-dependencies from vector-valued gene spatial patterns . In other words , we advance the state-of-art in both the usage of richer forms of expression data , and the employment of principled statistical methodology for sound network inference on such new form of data . Our results show that analyzing the spatial distribution of gene expression enables us to capture information not available from microarray data . Such an analysis is especially important in analyzing genes involved in embryonic development of Drosophila to reveal specific spatial patterning that determines the development of the 14 segments of the adult fly .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[]	2013	GINI: From ISH Images to Gene Interaction Networks
Pathogens have developed diverse strategies to infect their hosts and evade the host defense systems . Many pathogens reside within host phagocytic cells , thus evading much of the host immune system . For dimorphic fungal pathogens which grow in a multicellular hyphal form , a central attribute which facilitates growth inside host cells without rapid killing is the capacity to switch from the hyphal growth form to a unicellular yeast form . Blocking this transition abolishes or severely reduces pathogenicity . Host body temperature ( 37°C ) is the most common inducer of the hyphal to yeast transition in vitro for many dimorphic fungi , and it is often assumed that this is the inducer in vivo . This work describes the identification and analysis of a new pathway involved in sensing the environment inside a host cell by a dimorphic fungal pathogen , Penicillium marneffei . The pakB gene , encoding a p21-activated kinase , defines this pathway and operates independently of known effectors in P . marneffei . Expression of pakB is upregulated in P . marneffei yeast cells isolated from macrophages but absent from in vitro cultured yeast cells produced at 37°C . Deletion of pakB leads to a failure to produce yeast cells inside macrophages but no effect in vitro at 37°C . Loss of pakB also leads to the inappropriate production of yeast cells at 25°C in vitro , and the mechanism underlying this requires the activity of the central regulator of asexual development . The data shows that this new pathway is central to eliciting the appropriate morphogenetic response by the pathogen to the host environment independently of the common temperature signal , thus clearly separating the temperature- and intracellular-dependent signaling systems . Host immune systems actively survey and attempt to kill invading pathogens , so for pathogens to successfully infect a host the pathogen must be able to evade or tolerate these systems . A number of pathogens enter phagocytic cells and primarily reside within these cells to avoid the host's immune defense system . To continually reside within phagocytic cells of the immune system without disrupting their integrity , pathogens such as fungi which can grow in a filamentous , multicellular hyphal form , must be able to produce a uninucleate yeast growth form . The ability to switch between the filamentous and yeast forms is a tightly regulated process known as dimorphic switching . Dimorphism has been shown to be a critical pathogenicity determinant . Penicillium marneffei exhibits dimorphic switching and hence can grow in two distinct cellular forms; multicellular hyphae and unicellular yeast . P . marneffei is the only known Penicillium species which is dimorphic and the switch between growth forms is regulated by temperature [1] . At 25°C , in the saprophytic growth phase , P . marneffei grows as multinucleate , septate , branched hyphae . These hyphae produce conidia , the infectious agent , from specialized multicellular structures termed conidiophores . When switched to 37°C , P . marneffei undergoes a developmental process termed arthroconidiation . Cellular and nuclear division become coupled , double septa are laid down and hyphae fragment at these septation sites to liberate uninucleate yeast cells which subsequently divide by fission [1] . The yeast cells are the pathogenic form and multiple yeast cells are seen in the pulmonary alveolar macrophages and peripheral blood mononuclear cells of infected individuals [2] . P . marneffei infection is likely to occur through inhalation of the conidia produced by the filamentous saprophytic form [2] . It has been proposed that the conidia bind to laminin in the bronchoalveolar epithelia via a sialic acid-specific lectin [3] , [4] . The conidia are then ingested by host pulmonary alveolar macrophages where they germinate into unicellular yeast cells which divide by fission . Therefore the ability to produce infectious propagules such as asexual spores ( conidiation ) in the saprophytic growth state and the capacity upon infection to switch between a multicellular hyphal growth form and a unicellular yeast pathogenic form are both crucial for pathogenicity . Polarity establishment is necessary for the differentiation of distinct cell types during development . The Rho GTPases Cdc42 and Rac act as molecular switches to localize or activate proteins associated with polarized growth . The CDC42 homologue in P . marneffei , cflA , is required for germination of conidia at both 25°C and 37°C , polarized growth and division of hyphae at 25°C and for polarized growth of yeast cells at 37°C [5] . The P . marneffei genome also encodes a second Rac-like Rho GTPase , cflB . Similar to cflA , cflB is required for the polarized growth and division of hyphae at 25°C [6] . However , unlike cflA , cflB plays a key role during asexual development ( conidiation ) at 25°C and is not required for the polarized growth of yeast cells at 37°C [6] . In Saccharomyces cerevisiae , the Rho GTPase Cdc42p activates the p21 activated kinases ( PAKs ) Ste20p and Cla4p [7]–[11] . P . marneffei possesses both STE20 and CLA4 homologues; pakA ( STE20 ) and pakB ( CLA4 ) . Characterization of pakA in P . marneffei has shown that this gene is essential for conidial germination at 37°C and polarized growth of yeast cells , acting downstream of both a heterotrimeric G protein and Cdc42 pathway [12] . ΔpakA and pakA strains containing a mutation in the conserved Cdc42/Rac Interactive Binding ( CRIB ) domain ( pakAH108G ) fail to germinate into pathogenic yeast cells in vivo [12] . This study describes the characterization of the second PAK in P . marneffei , PakB . The pakB gene is expressed during hyphal growth and asexual development at 25°C and is essential for the generation of these 25°C-specific cell types . Deletion of pakB results in yeast-like morphology and the inappropriate production of yeast cells at 25°C . Deletion of the primary regulator of asexual development , brlA , in the ΔpakB strain results in suppression of this inappropriate yeast cell production suggesting that these yeast cells are dependent on the conidiation program . PakB is also essential for yeast morphogenesis during infection but not in vitro . Macrophages infected with the ΔpakB strain exhibit highly branched , septate , hyphal cell growth but no yeast cells . These results suggest that the developmental pathways regulating conidiation at 25°C and yeast cell production at 37°C share a number of regulatory components including PakB and that the developmental outcome of each pathway is regulated in part by the mode of cellular division . A previous low stringency hybridisation screen of a P . marneffei genomic library using an Aspergillus nidulans sequence with strong homology to S . cerevisiae Ste20p yielded five positive clones , which fell into two classes based on restriction enzyme digestion patterns [12] . Sequencing of a cloned fragment from one of these classes ( pKB5751 ) revealed strong sequence homology to STE20-like PAKs and the gene within this clone was subsequently named pakA [12] . A fragment from the second class of clones was also subcloned ( pKB4904 ) and sequencing revealed strong sequence homology to CLA4-like PAKs from Candida albicans ( 48% identity , 58% similarity ) , Ashbya gossypii ( 49% identity , 57% similarity ) , Ustilago maydis ( 53% identity , 61% similarity ) and S . cerevisiae ( 48% identity , 57% similarity ) ( Figure S1 ) . The gene within this clone was named pakB . The homology is to a large extent restricted to the CRIB and kinase domains where , for example , PakB shows 74% and 76% identity , respectively , to the same domains in U . maydis Cla4 . The predicted PakB protein exhibits 37% identity and 46% similarity to PakA . The predicted PakB protein is 733 amino acids in length and contains a PH domain at 81–191 , a Cdc42/Rac Interactive Binding ( CRIB ) domain ( also called PBD for p21-Rho-binding domain ) at positions 195-256 and a predicted kinase domain at 443–712 ( http://pfam . sanger . ac . uk/ ) . An 11 amino acid sequence in the non-catalytic C-terminal region of S . cerevisiae Ste20p has been shown to be required for interaction with Ste4p , the beta subunit of a heterotrimeric G protein , during pheromone signaling [13] , [14] . This region in S . cerevisiae Cla4p has also been shown to interact weakly with Ste4p [13] . The consensus sequence SSLφPLI/VXφφβ ( where X is any residue , φ for A , I , L , S or T and β is for basic residues ) is also found in P . marneffei PakA ( 619–629 ) . RNA was isolated from vegetative hyphae grown for 2 days in liquid medium at 25°C , asexual development ( conidiation ) cultures grown for 4 days on solid medium at 25°C and yeast cells grown for 6 days in liquid medium at 37°C . The level of pakB expression varied substantially depending on cell type . The level of pakB transcript was highest during vegetative hyphal growth at 25°C , lower during asexual development at 25°C and barely detectable during vegetative yeast growth at 37°C relative to the benA control ( Figure 1A ) . P . marneffei infection is believed to occur by inhalation of conidia , which bind to the laminin in the bronchoalveolar epithelium [2]–[4] . Conidia are then ingested by pulmonary alveolar macrophages and germinate directly into uninucleate yeast cells which proliferate within the macrophage [1] . To examine if pakB is expressed during infection , RNA was isolated from yeast cells derived either from LPS activated J774 murine macrophages at 37°C infected with wildtype conidia 24 hours post-infection or from yeast cells incubated in macrophage growth media at 37°C for 24 hours ( Materials and Methods ) . Substantial levels of pakB expression were detected in cells isolated from infected macrophages suggesting that pakB expression is induced during infection ( Figure 1B ) . Low levels of pakB expression was detected in yeast cells derived from the macrophage medium control ( Figure 1B ) . To investigate the localization of PakB , a triple HA tag was inserted into a non-conserved region of PakB between the CRIB and kinase domains . The pakB+ HA construct was co-transformed with the barA+ gene into the P . marneffei strain G487 ( niaD pyrG areA− ) . Transformants were selected for glufosinate resistance and confirmed by Southern blot analysis of genomic DNA ( Materials and Methods ) . Anti-HA immunostaining was performed on two of the pakB+ HA strains after 4 days growth at 25°C . PakB was observed concentrated at the hyphal apex ( Figure 2A–B ) and also localized to all of the cell types of the conidiophore ( Figure 2C ) . PakB was particularly concentrated at the phialide to conidium interface and around the periphery of newly formed , but not old , conidia ( Figure 2D ) . In addition , PakB was co-localized at nascent septation sites presenting either as a single band colocalised with calcofluor stained septa ( Figure 2E ) , two bands on either side of the calcofluor stained septa ( Figure 2F ) or two spots on either side of the calcofluor stained septa ( Figure 2G ) . PakB was not observed at older septa ( data not shown ) . To investigate the localization of pakB during infection , LPS activated J774 murine macrophages were infected with the pakB+ HA strains and anti-HA immunostaining and calcofluor staining were performed 24 hours post-infection ( Materials and Methods ) . PakB was localized around the cell periphery ( Figure 2H–I ) . PakB was not localized either at nascent septation sites prior to , or immediately after , cell wall deposition ( indicated by calcofluor staining ) ( Figure 2H ) . Weak localization at septation sites could be observed prior to cell separation ( Figure 2I ) . PakB was localized at , and adjacent to , the division site during cell separation ( Figure 2I ) . A construct in which a region encompassing the pakB coding sequence ( +323 to +2618 ) was replaced with the pyrG selectable marker was used to create a pakB deletion strain . P . marneffei strain G487 ( niaD pyrG areA− ) was transformed with this construct and pyrG+ transformants selected . Genomic DNA from the PyrG+ transformants was screened by Southern blotting to identify strains which possessed a restriction pattern consistent with replacement of pakB by pyrG at the genomic locus ( data not shown ) . To generate a ΔpakB pyrG− strain , a ΔpakB::pyrG+ deletion strain was plated on medium containing 5-fluoroorotic acid ( 5-FOA ) ( Materials and Methods ) . This strain was cotransformed with plasmids containing pakB+ and pyrG+ genes and co-transformants confirmed by Southern blot analysis . The ΔpakB pakB+ transformants contained 2–8 copies of pakB . After 10 days at 25°C , wildtype P . marneffei grows as polarized vegetative hyphae which bear asexual structures ( conidiophores ) . Colonies appear fuzzy and the surface is green due to the presence of pigmented asexual spores ( conidia ) on conidiophores ( Figure 3A–B ) . The ΔpakB strain produced compact , mucoid yeast-like colonies after 10 days growth at 25°C which resemble the yeast colonies produced by the wildtype at 37°C ( Figure 3A versus C ) . Despite this the ΔpakB strain conidiated upon longer incubation at 25°C ( 14 days ) . Conidiophore structures were visible under higher magnification , however , these were unevenly dispersed over the yeast-like colonies and not as profuse as in wildtype ( Figure 3B ) . Transformation of the ΔpakB strain with pakB+ ( ΔpakB pakB+ ) completely restored the wildtype phenotype ( Figure 3 ) . Deletion of pakB resulted in yeast-like growth at 25°C ( Figure 3A ) . To investigate the molecular basis underpinning this phenotype RT PCR analysis was performed to see if it correlates with a decrease in the expression of hyphal specific genes or an increase in the expression of yeast specific genes . RNA was isolated from both the wildtype and the ΔpakB strain grown as vegetative hyphae for 2 days in liquid medium at 25°C and as yeast cells for 6 days in liquid medium at 37°C and used for RT PCR with primers for a number of cell type specific genes ( Canovas and Andrianopoulos , unpublished ) . In wildtype , the 2E11 probe was expressed specifically in hyphae at 25°C and not in yeast cells at 37°C ( Figure 1C ) . In contrast , the 2E11 transcript was not detectable in the ΔpakB strain at 25°C ( Figure 1C ) . In wildtype , the 2E4 transcript was expressed at a low level at 25°C and expression was greatly increased at 37°C ( Figure 1D ) while expression of 5B10 in wildtype was not detectable at 25°C but was high at 37°C ( Figure 1E ) . Both 2E4 and 5B10 transcripts were highly expressed in the ΔpakB strain at 25°C ( Figure 1D and E ) . The amount of 2E4 and 5B10 transcript was also slightly increased in the ΔpakB strain at 37°C ( Figure 1D and E ) . This suggests that in the ΔpakB strain both a decrease in the expression of hyphal specific genes and an increase in the expression of yeast specific genes may contribute to the yeast-like growth phenotype at 25°C . In order to characterize the cellular basis behind the yeast-like colonial morphology of the ΔpakB strain at 25°C , scanning electron microscopy ( SEM ) was performed on both wildtype and ΔpakB strains after 10 days growth at 25°C ( Materials and Methods ) . Wildtype colonies appeared as a mass of entangled , branched hyphae which radiated from the centre of the colony in a polarized fashion ( Figure 4A ) . The edges of the colony were not clearly defined due to countless hyphae which extended great distances from the colony periphery ( Figure 4A ) . In contrast , the colonies of the ΔpakB strain were compact and therefore the colony edges were distinct ( Figure 4A ) . The hyphae of the ΔpakB strain were more branched than wildtype and often invaded the agar surface ( Figure 4B ) . Invasive growth is a common characteristic of wildtype arthroconidial growth at 37°C ( not shown ) . In contrast to wildtype , in which yeast cells are never seen at 25°C , individual yeast cells were also observed around the ΔpakB colony periphery ( Figure 4C ) . The wildtype and ΔpakB strains were also grown on the standard 37°C media , brain heart infusion ( BHI ) , for 10 days at 25°C and for 5 days at 37°C . SEM of the colony surface showed wildtype colonies at 25°C were comprised of large bundles of branched hyphae ( Figure 4D ) . At 37°C , wildtype undergoes arthroconidiation , a developmental process where hyphae fragment to liberate yeast cells which subsequently divide by fission . After 5 days at 37°C , the wildtype colony was comprised of a mass of fragmented hyphae and yeast cells ( Figure 4D ) . At 25°C , the colony surface of the ΔpakB strain resembled that of wildtype at 37°C ( Figure 4E ) . The ΔpakB strain also appeared to produce more yeast cells than wildtype at 37°C ( Figure 4E ) . To investigate further the molecular mechanisms underlying the deregulation of yeast cell morphogenesis in the ΔpakB strain at 25°C , the wildtype , ΔpakB and ΔpakB pakB+ strains were grown for 4 days at 25°C , stained with calcofluor and Hoechst 33258 and observed microscopically ( Figure 5 ) . Wildtype P . marneffei grows as highly polarized , branched , septate hyphae ( Figure 5 ) . Apical cells are multinucleate whereas subapical cells are predominately uninucleate unless dividing ( Figure 5B ) . In all conditions examined , the ΔpakB pakB+ strains were indistinguishable from wildtype ( data not shown ) . The ΔpakB strain exhibited a compact colony morphology where hyphae were tightly packed and highly branched ( Figure 5A ) . Apical cell branching , which was not observed in wildtype , was frequently observed in the ΔpakB strain ( Figure 5A ) . Septa were present in the ΔpakB strain ( Figure 5A ) . In contrast to wildtype , the ΔpakB strain exhibited an increase in the number of nuclei per subapical cell compartment ( Figure 5B ) . To quantify this increase , the number of nuclei per cellular compartment was recorded for 100 cells on three separate occasions . Only 20 . 5±2 . 04% and 20 . 4±2 . 72% of wildtype and ΔpakB pakB+ subapical cells contained more than one nucleus , respectively . In contrast , 64 . 2±1 . 18% of subapical cells of the ΔpakB strain contained more than one nucleus . To examine if deletion of pakB affects actin distribution , immunostaining using mouse anti-actin was performed on the wildtype and ΔpakB strains . In the wildtype at 25°C , actin is localized as cortical actin spots along the hyphae and concentrated at nascent septation sites and the hyphal apex ( Figure 5C and D ) . Actin was normally distributed in the ΔpakB strain ( Figure 5C and E ) . Wildtype P . marneffei begins asexual development after 4 days growth at 25°C , with the production of a specialized aerial stalk from which differentiated cells are produced sequentially in a budding fashion: metulae bud from the stalk , phialides bud from metulae and uninucleate conidia bud from phialides . To investigate if the deletion of pakB results in aberrant asexual development at 25°C , wildtype , ΔpakB , ΔpakB pakB+ and ΔcflB strains were grown for 14 days at 25°C and were examined by SEM ( Figure 6 ) . Deletion of the P . marneffei RAC homologue , cflB , results in conidiation defects at 25°C [6] . Numerous conidiophores were observed on the surface of wildtype colonies ( Figure 6 ) . The ΔpakB pakB+ strain was indistinguishable from the wildtype ( data not shown ) . The ΔpakB strain also produced conidiophores in which all conidiophore cell types were observed , however a large number of conidiophores in the ΔpakB strain had abnormally large conidia ( Figure 6 ) . In addition , more than one conidium per phialide was rarely observed and the site of conidium to phialide attachment was not as constricted as in wildtype ( Figure 6 ) . In contrast to the wildtype , ΔpakB and ΔpakB pakB+ strains , conidiophores in the ΔcflB strain could not be readily distinguished but presumptive conidia of varied size were noted , similar to the ΔpakB strain although not as extreme ( Figure 6 ) . A unique phenotype of the ΔcflB strain is that lysed conidiophore structures in which the metulae and phialides have ruptured or deflated were frequently observed ( Figure 6 ) . To investigate the cellular defects underlying the aberrant conidiophores of the ΔpakB strain , calcofluor staining was performed on wildtype , ΔpakB and ΔpakB pakB+ strains after 14 days growth at 25°C ( Materials and Methods ) . The ΔpakB conidiophores displayed septation defects in conidiophores ( Figure 7A–B ) . In contrast to wildtype conidiophores , in which two separate chitin disks can be observed at the phialide/conidium and conidium/conidium junctions , only one septum , no septa or incomplete septa were observed at the phialide to conidia cell boundaries in ΔpakB conidiophores ( Figure 7B ) . To assess any defects in conidia that may be the result of the aberrant conidiophore morphogenesis in the mutant strains , calcofluor staining was performed on conidial suspensions of the wildtype , ΔpakB , ΔpakB pakB+ and ΔcflB strains ( Materials and Methods ) . The conidia produced by the wildtype and ΔpakB pakB+ strains were homogeneous in size and showed uniform calcofluor staining around the cell periphery ( Figure 7C and data not shown ) . A proportion of conidia produced by the ΔcflB strain displayed a size increase , however , calcofluor staining remained uniform ( Figure 7C ) . In contrast , conidia from the ΔpakB strain differed greatly in size and showed uneven calcofluor staining ( Figure 7C ) . In addition , ΔpakB conidial preparations contained numerous yeast cells ( Figure 7C ) . Nuclear staining revealed that these yeast cells were uninucleate ( data not shown ) . The germination and colony forming ability of conidia from the wildtype , ΔpakB , and ΔcflB strains was determined to investigate the potential consequences of the aberrant morphogenesis . The germination kinetics were measured by counting the number of ungerminated versus germinated conidia ( conidia with a visible germ tube ) in a population of 100 in three independent experiments after 15 hours in liquid medium at both 25°C and 37°C ( Table 1 ) . Despite the large increase in conidial size , conidia of the ΔpakB strain germinated well at both 25°C and 37°C and actually showed a slight increase in germination compared to the wildtype control ( Table 1 ) . In contrast , the conidia of the ΔcflB strain showed a reduction in germination compared to the other strains ( Table 1 ) . It was also evident that ΔpakB conidia prematurely extended secondary germ tubes . To quantify this , the number of conidia with 1 , 2 or 3 or more germ tubes was counted in a population of 100 in three independent experiments ( Table 2 ) . After 15 hrs at either 25°C or 37°C , the ΔpakB conidia showed a significant increase in germ tube emergence ( Table 2 ) . The ability of single conidia to form colonies was also assessed after 5 days at both 25°C and 37°C ( Table 3 ) . The colony forming units were measured by counting the number of colonies arising from 100 plated conidia in three independent experiments ( Table 3 ) . Consistent with the germination data , the ΔpakB strain showed wildtype viability ( Table 3 ) . The ΔcflB strain showed a reduction in the ability to form colonies suggesting that the previously observed ungerminated conidia were inviable rather than delayed ( Table 3 ) . When examining asexual development in the ΔpakB strain , a number of conidiophores were observed in which the individual cell types had become detached ( data not shown ) . As the ΔpakB strain inappropriately produced yeast cells at 25°C ( Figure 4C and 7C ) we hypothesized that these yeast cells may arise from an inappropriate switch from conidiation to arthroconidiation . To investigate this possibility , a ΔpakB ΔbrlA double mutant was generated ( Materials and Methods ) . brlA encodes the primary regulator of asexual development which is necessary and sufficient for asexual development . The ΔbrlA mutant produces aerial stalks but is unable to produce the various budded cell types of the conidiophore ( Borneman and Andrianopoulos , unpublished ) . Wildtype , ΔpakB , ΔbrlA and ΔpakB ΔbrlA strains were grown on ANM + ( NH4 ) 2SO4 for 14 days at 25°C . Cell suspensions were made in 0 . 005% Tween 80 solution and filtered through Miracloth to remove hyphal cells . As expected , conidia but no yeast cells were observed in wildtype suspensions while the ΔpakB strain produced conidia of varying size and numerous yeast cells . No conidia or yeast cells were observed in the ΔbrlA strain . Interestingly , no conidia or yeast cells were observed for the ΔpakB ΔbrlA strains suggesting that the yeast cells observed in the ΔpakB strain arise from the conidiation program ( data not shown ) . RNA was isolated from wildtype and ΔbrlA strains grown for 4 days on solid medium at 25°C to investigate if brlA is required for pakB expression during asexual development . Expression of pakB during asexual development was similar in both wildtype and the ΔbrlA strains showing that pakB expression during conidiation is independent of brlA ( data not shown ) . Given the inappropriate production of yeast cell at 25°C , the effects of pakB deletion on in vitro yeast cell morphogenesis were assessed . Wildtype P . marneffei produced compact , mucoid , yeast colonies after 5 days growth at 37°C in vitro and both the ΔpakB and ΔpakB pakB+ strains were indistinguishable from wildtype ( Figure 3C ) . To examine yeast cell morphogenesis in vitro , the wildtype ( pakB+ ) , ΔpakB , ΔpakB pakB+ and ΔcflB strains were inoculated on agar-coated BHI slides and incubated for 5 days at 37°C . Wildtype conidia germinate at 37°C to produce polarized arthroconidiating hyphae , an intermediary cell type which is specific to in vitro yeast morphogenesis and which is not manifested in macrophages . Nuclear division and septation become coupled in arthroconidiating hyphae , double septa are laid down and fragmentation occurs along this plane to liberate uninucleate yeast cells which consequently divide by fission . Numerous yeast cells were observed for the wildtype strain after 5 days at 37°C ( Figure 8A ) . Similarly , the ΔpakB , ΔpakB pakB+ and ΔcflB strains produced abundant yeast cells comparable to wildtype and these yeast cells were also uninucleate ( Figure 8A ) . Thus PakB appears to play no role in vitro . This is consistent with the expression data under this condition in which the pakB transcript was barely detectable ( Figure 1A ) . The wildtype , ΔpakB and ΔpakB pakB+ strains also produced numerous yeast cells after 6 days growth at 37°C on Sab and ME yeast medium , comparable to those produced on BHI ( data not shown ) . In order to examine the role of conserved domains of PakB in the phenotypes observed , mutant alleles were generated which altered the CRIB ( H204G and Δ195–256 ) and putative GBB domain ( Δ719–729 ) . The equivalent CRIB mutations in S . cerevisiae Ste20p prevents the Cdc42p interaction ( H345G ) or generate a constitutively active protein which bypasses the requirement for Cdc42p ( ΔCRIB ) [8] . Deletion of the GBB domain prevents Ste20p interaction with Ste4p , therefore affecting activation of the MAPK cascade [13] . These alleles , in addition to the wildtype allele , were placed under the control of the inducible xylP promoter , which is only expressed in the presence of xylose . The ΔpakB pyrG− strain was transformed with these constructs and transformants were directly selected for pyrG+ and confirmed by Southern blot analysis of genomic DNA ( Materials and Methods ) . The xylP ( p ) pakB , xylP ( p ) pakBH204G , xylP ( p ) pakBΔCRIB and xylP ( p ) pakBΔGBB strains were grown on media with or without 1% xylose for 10 days ( Figure 9A ) or for 4 days on agar-coated slides for microscopic observation ( Figure 9B–D ) . As expected , all strains exhibited the ΔpakB phenotype at 25°C on non-inducing medium ( Figure 9A–B ) . On inducing medium , expression of the xylP ( p ) pakB construct completely restored the wildtype phenotype ( Figure 9A and C ) . Expression of the xylP ( p ) pakBH204G allele partially restored the wildtype hyphal phenotype such that colonies were more filamentous ( Figure 9A ) and hyphae were less tightly packed with less apical branching ( Figure 9B–C ) . In contrast , strains expressing either the xylP ( p ) pakBΔCRIB or xylP ( p ) pakBΔGBB allele were indistinguishable at the colonial level on non-inducing and inducing medium ( Figure 9A ) . Both the xylP ( p ) pakBΔCRIB and xylP ( p ) pakBΔGBB strains displayed compact , tightly packed hyphae exhibiting apical branching similar to ΔpakB , however , the xylP ( p ) pakBΔCRIB strains were also hyperbranched and an increase in yeast cells around the colony periphery were observed ( Figure 9C–D ) . These yeast cells appeared to be dividing by fission ( Figure 9D ) . To investigate the effect of the mutant alleles on asexual development at 25°C , these strains were grown for 14 days at 25°C and examined by SEM . All strains exhibited the ΔpakB phenotype on non-inducing medium . Unlike the ΔpakB pakB+ strain , whose conidiophores were indistinguishable from wildtype , the xylP ( p ) pakB strain did not show full complementation on inducing medium; conidiophores mainly consisted of a single phialide and conidiophores with more than one conidium per phialide was rarely observed ( data not shown ) . This suggests that the formation of complex multicellular conidiophores is sensitive to pakB expression levels ( Figure 1A ) , either due to overexpression of PakB in conidiophore cell types or poor induction of the xylP promoter in conidiophore cell types . On inducing medium , the xylP ( p ) pakBH204G , xylP ( p ) pakBΔCRIB and xylP ( p ) pakBΔGBB strains were indistinguishable from the ΔpakB strain ( data not shown ) . To assess whether the H204G , ΔCRIB or ΔGBB mutations affect PakB localization , HA-tagged pakBH204G , pakBΔCRIB and pakBΔGBB constructs were generated and co-transformed with the barA+ gene into the P . marneffei strain G487 ( niaD pyrG areA− ) . Transformants were selected for glufosinate resistance and confirmed by Southern blot analysis of genomic DNA ( Materials and Methods ) . Anti-HA immunostaining was performed on two strains of each genotype after 4 days growth at 25°C . Like wildtype , both the PakBH204GHA and PakBΔGBBHA proteins were concentrated at the hyphal apex , however , the PakBΔCRIBHA protein was not ( data not shown ) . The PakBH204GHA , PakBΔGBBHA and PakBΔCRIBHA proteins showed wildtype localization in conidiophores and at nascent septation sites ( data not shown ) . pakB was not expressed in vitro on BHI at 37°C ( Figure 1A ) but expression was induced during macrophage infection ( Figure 1B ) . Yeast cells produced in vivo are shorter and rounder than those produced in vitro and develop directly from conidia rather than via arthroconidiating hyphae . To examine if induced expression of pakB can effect the changes in yeast cell morphology observed in vivo , the wildtype , ΔpakB , xylP ( p ) pakB , xylP ( p ) pakBH204G , xylP ( p ) pakBΔCRIB and xylP ( p ) pakBΔGBB strains were grown on BHI with or without 1% xylose for 5 days ( Figure 10 ) . As expected , all of the strains were indistinguishable from wildtype and the ΔpakB strain on non-inducing medium ( Figure 10A ) . On inducing medium , the xylP ( p ) pakB strains ( copy numbers 2–12 ) were indistinguishable from the wildtype and the ΔpakB strains ( Figure 10B ) . Therefore an increase in pakB expression does not explain the production of rounded yeast cells in vivo . On inducing medium , the xylP ( p ) pakBH204G , xylP ( p ) pakBΔCRIB and xylP ( p ) pakBΔGBB strains produced yeast cells which were rounder and greatly reduced in length compared to both the ΔpakB and xylP ( p ) pakB strains and a number of yeast cells were produced which appeared to be dividing in a budding manner , similar to that observed during conidiation at 25°C ( Figure 10B–C ) . The yeast cells produced by the xylP ( p ) pakBΔCRIB strains were also often curled ( Figure 10B ) . To investigate if pakB is required for yeast growth during infection , LPS activated J774 murine macrophages were infected with wildtype , ΔpakB and ΔpakB pakB+ conidia and observed microscopically 24 hours post-infection ( Materials and Methods ) . Calcofluor staining was performed to allow visualization of fungal cell walls . After 24 hours , numerous yeast cells dividing by fission were observed in macrophages infected with wildtype ( pakB+ ) ( Figure 8B ) or ΔpakB pakB+ conidia ( data not shown ) . In contrast , the majority of macrophages infected with ΔpakB conidia contained highly branched , septate , hyphal cells but no yeast cells ( Figure 8B ) . A small proportion of macrophages infected with ΔpakB conidia contained large , swollen conidia ( Figure 8B ) . To quantify this difference , the number of cells containing at least one septum was recorded for approximately 100 cells on three separate occasions . In contrast to wildtype and ΔpakB pakB+ , in which 17 . 2±3 . 70% and 15 . 3±1 . 74% of cells in infected macrophages contain septa , 71 . 1±2 . 24% of ΔpakB cells contain septa . This indicates that pakB is required for yeast cell division during infection . To determine if the role played by PakB in vivo is during the initiation or maintenance of yeast-like growth , macrophages were also infected with wildtype , ΔpakB and ΔpakB pakB+ yeast cells ( Materials and Methods ) . Similar to infection with conidia , after 24 hours numerous yeast cells were observed in macrophages infected with wildtype or ΔpakB pakB+ yeast cells whereas macrophages infected with ΔpakB yeast cells contained only septate hyphal cells ( Figure S2 ) . To assess if cellular morphology affects the oxidative state of the host , superoxide production was detected by Nitrotetrazolium Blue Chloride ( NBT ) staining of macrophages 24 hours post-infection ( Materials and Methods ) . No difference in superoxide production was observed in macrophages infected with either wildtype or the ΔpakB strain ( data not shown ) . In addition , the presence or absence of superoxide did not affect the growth phenotype of the ΔpakB strain ( data not shown ) . To examine whether the presence of host extracts is sufficient to induce the morphological switch , lysed macrophage extracts were added to wildtype , ΔpakB and ΔpakB pakB+ conidia ( Materials and Methods ) . After 24 hrs at 37°C , all strains were growing as hyphae indicating that the addition of lysed macrophage extracts is insufficient to induce yeast-like growth ( data not shown ) . To observe if being intracellular is sufficient to induce the morphological switch , macrophages infected with wildtype , ΔpakB and ΔpakB pakB+ conidia were incubated at 25°C . Unlike at 37°C , after 24 hrs all strains grew as hyphae indicating that it is a combination of host and temperature signals which induces yeast-like growth ( data not shown ) . To investigate if the mutant alleles can complement the phenotype of the ΔpakB strain in vivo , pakB+ , pakBH204G , pakBΔCRIB and pakBΔGBB constructs were co-transformed with the barA+ gene into the ΔpakB strain and transformants were selected for glufosinate resistance and confirmed by Southern blot analysis of genomic DNA ( Materials and Methods ) . These strains exhibited identical phenotypes to the equivalent xylP overexpression strains grown on inducing medium at 25°C ( data not shown ) . LPS activated J774 murine macrophages were infected with conidia from these strains and examined microscopically after 24 hours ( Materials and Methods ) . Calcofluor staining was performed to allow visualization of fungal cell walls . The number of cells containing a septum was recorded for approximately 100 cells , in four transformants of each genotype , in three independent experiments ( Table S1 ) and significant differences assessed by two-level nested analysis of variance ( ANOVA ) ( Table S2 ) . ANOVA showed there was no significant difference between the wildtype , ΔpakB pakB+ and ΔpakB pakBH204G strains but there was a significant difference between these strains and the ΔpakB , ΔpakB pakBΔCRIB and ΔpakB pakBΔGBB strains . There was also a significant difference between the ΔpakB pakBΔCRIB strains and either the ΔpakB pakBΔGBB or ΔpakB strains . No significant differences were detected between transformants of the same genotype ( Table S2 ) . After 24 hours , numerous yeast cells dividing by fission were observed in macrophages infected with wildtype , ΔpakB pakB+ or ΔpakB pakBH204G conidia ( Figure 8C ) of which 17 . 2±3 . 70% ( wildtype ) , 19 . 5±0 . 86% ( ΔpakB pakB+ ) and 22 . 2±1 . 3% ( ΔpakB pakBH204G ) of cells contained at least one septum . In contrast , macrophages infected with either of the ΔpakB , ΔpakB pakBΔCRIB or ΔpakB pakBΔGBB strains contained both septate yeast and hyphal cells ( Figure 8C ) and the number of septa was substantially higher than the wildtype ( wildtype 17 . 2±3 . 70%; ΔpakB strain 71 . 1±2 . 24%; ΔpakB pakBΔCRIB 42 . 6±1 . 9%; ΔpakB pakBΔGBB 62 . 2±1 . 99% ) . To assess whether the H204G , ΔCRIB or ΔGBB mutations affect PakB localization during macrophage infection , anti-HA immunostaining was performed on the PakBH204GHA , PakBΔCRIBHA and PakBΔGBBHA strains after 24 hours post-infection of LPS activated J774 murine macrophages ( Materials and Methods ) . The PakBH204GHA , PakBΔGBBHA and PakBΔCRIBHA proteins showed wildtype localization during macrophage infection ( data not shown ) . The Rho type GTPases CDC42 and Rac are known regulators of PAKs and interact via the CRIB domain . As the ΔpakB pakBH204G and ΔpakB pakBΔCRIB strains showed either no effect or only partial deregulation of yeast morphogenesis during infection , the in vivo phenotype of cflAG14V , cflA120A and ΔcflB mutants was assessed [5] , [6] ( Materials and Methods ) . After 24 hours , numerous yeast cells dividing by fission were observed in macrophages infected with wildtype ( pakB+ ) ( Figure 8B ) . In contrast , ΔcflB conidia in infected macrophages remained predominately ungerminated ( Figure 8B ) ( 76 . 2±6 . 00% for ΔcflB compared to 13 . 3±4 . 86% for wildtype ) . Unexpectedly , numerous yeast cells with wildtype morphology were observed in macrophages infected with the cflAG14V and cflA120A mutants ( Figure S3 ) . These results suggest that any interaction between CflA and PakB is non-essential for in vivo morphogenesis and that the defects in yeast cell morphology observed in the cflAG14V and cflA120A mutants in vitro [5] are circumvented when growing inside host cells . Many pathogens reside within host phagocytic cells where they evade much of the host immune system . For dimorphic fungal pathogens a central attribute which facilitates this immune system avoidance is the capacity to switch from a multicellular hyphal growth form to a unicellular yeast form and it has been demonstrated that blocking this transition abrogates pathogenicity [15] , [16] . Host body temperature ( 37°C ) is the clearest ex vivo inducer of the hyphal to yeast transition in many dimorphic fungi and it is often assumed that this is the inducer in vivo . Here we show that pakB , which encodes the second p21 activated kinase in the dimorphic pathogen P . marneffei , is strongly upregulated upon phagocytosis by macrophages and is essential for yeast morphogenesis but not growth in macrophages . In contrast PakB plays no role in yeast morphogenesis at 37°C in vitro . This clearly places PakB in a new signalling pathway which responds to host cell inductive signals , not temperature , and is necessary for intracellular yeast morphogenesis and consequently pathogenicity . The mechanism by which PakB controls yeast cell morphogenesis inside host cells is unique . The Rho type GTPases CDC42 and Rac are known regulators of PAKs in many eukaryotes , interacting via the CRIB domain of these kinases . Consistent with this PakB and CflB orthologues from a number of fungal pathogens have been shown to either physically or genetically interact , yet none of these systems represent intracellular pathogens . The P . marneffei ΔcflB ( Rac ) mutant strain fails to germinate in macrophages and mutation of the PakB CRIB domain ( pakBH204G ) does not recapitulate this phenotype nor that of the ΔpakB mutant . In strains where the entire CRIB or predicted Gβ binding domains are deleted , there is a partial deregulation of morphogenesis leading to the production of both yeast and hyphal cells in macrophages . Previous studies that have shown that CflA is required for correct yeast cell morphogenesis in P . marneffei during in vitro growth , potentially implicating CflA as a regulator for PakB . However the morphology of yeast cells in CRIB domain pakB mutants , both in vitro and inside macrophages , is essentially wildtype and cflAG14V and cflA120A mutants produce wildtype yeast cells in vivo showing that any interaction between CflA and PakB is non-essential for in vivo morphogenesis and that the defects in yeast cell morphogenesis observed in the cflAG14V and cflA120A mutants in vitro may be due to a general defects in morphogenesis . In dimorphic fungi which have a predominant yeast phase such as C . albicans and Y . lipolytica , mutation of the pakB orthologue ( CLA4 ) leads to defective hyphal formation and invasive growth , and consequent changes in pathogenicity for C . albicans [17] , [18] . For fungi whose predominant growth phase is hyphal and pathogenic phase is yeast , mutation of pakB has the opposite effect . The simplest explanation for these opposing effects is that pakB orthologues control a conserved fundamental cellular process which is recruited by the organism-specific regulatory systems for either the yeast-hyphal or hyphal-yeast dimorphic switch . Based on the results described here , and discussed below , this process is likely to be the regulation of both morphogenesis and septation during cellular division . Dimorphic fungal pathogens must regulate both the transition and maintenance of two vegetative growth states , multicellular hyphal and unicellular yeast , as well as other morphogenetic programs such as asexual development [19] . This capacity is lost when the pakB gene is deleted from P . marneffei . The pakB gene is specifically expressed during hyphal growth and asexual development at 25°C and is required for the generation of these 25°C specific cell types . Loss of pakB leads to defects in polarised growth of hyphal cells and these defects partially overlap with those for cflA ( CDC42 ) and cflB ( Rac ) mutants . In contrast , the second PAK in P . marneffei plays no role in hyphal cell morphogenesis [12] . Furthermore , loss of pakB leads to inappropriate production of yeast cells at 25°C and it might be argued that PakB is either required to promote hyphal growth or to negatively regulate yeast growth , and that yeast growth is the default state . In support of this hypothesis , a number of hyphal specific genes showed decreased expression while yeast specific genes showed increased expression in the ΔpakB strain . Studies of the P . marneffei transcriptional co-repressor tupA lead to a similar conclusion about the default growth state [20] . Whilst the nature of the default state is likely to be correct , it is clear that pakB plays a more fundamental role beyond hyphal morphogenesis . The inappropriately produced yeast cells in the ΔpakB strain are not derived from vegetative hyphal cells but from differentiating conidiophore cells . Deletion of the primary regulator of asexual development which is not expressed in vegetative hyphae , the C2H2 Zn finger transcription factor gene brlA , abolished production of yeast cells at 25°C in the ΔpakB strain . This suggests overlap in the morphogenetic mechanisms controlling conidiation at 25°C and yeast cell production at 37°C and supports previous studies in P . marneffei and other dimorphic fungi [16] , [20] , [21] , [22] , [23] . More importantly it points to the underlying mechanism being the control of the mode of cellular division . One of the major differences between vegetative hyphal cells and differentiated conidiophore cells is that the former divide by septation ( analogous to fission but without cell separation ) while the latter divide by budding , so a possible explanation is that PakB is required for the correct execution of budding division during conidiation in P . marneffei and loss of PakB defaults to a fission mode of division with cell separation in conidiophore cell types; as occurs in P . marneffei yeast cells at 37°C . In support of this hypothesis , deletion of the CLA4 homologue in U . maydis results in yeast cells which separate by fission , instead of the normal budding mode at the distal tips of yeast cells , with constriction occurring at a centrally located septum [24] . Although this situation differs from P . marneffei , as it is occurring in undifferentiated cell types , it suggests that fission is acting as the default mechanism for division when budding is aberrant . These results suggest that budding is a derived mode of division and this is consistent with our understanding of the molecular mechanisms which underlie budding , as septation ( fission ) follows the isotropic growth phase during the budding program . Asexual development is a developmental program in which the hyphal-based apical growth and septation mode of growth switches to acropetal budding division without cell separation and finally basipetal budding division with cell separation , with the concomitant switch to uninucleate cells . Many dimorphic fungi produce yeast cells that divide by budding while P . marneffei produces yeast cells that divide by fission . In all of these instances , the coupling of septation to cell separation determines whether yeast or hyphal ( or pseudohyphal ) cells will be generated . The pakB deletion strain is able to undergo asexual development , albeit less profusely than wildtype , producing swollen conidia with an enlarged conidium attachment site , more akin to the septa of vegetative hyphal cells . The septa separating the other conidiophore cell types are also aberrant . This suggests that PakB activity regulates the constriction , and possibly formation , of septa during conidiogenesis and is supported by the localization of PakB to conidiophores where it is particularly concentrated at phialide to conidium boundaries . Despite the abnormal morphology of the ΔpakB conidia , they were still able to germinate normally at both 25°C and 37°C suggesting that it is only the final stages of conidial separation which are affected in the ΔpakB mutant . This is similar to the rice pathogen Magnaporthe grisea in which deletion of the CLA4 homologue , CHM1 , results in irregularly shaped conidia with reduced constriction at the conidium attachment site but is unlike the rye pathogen Claviceps purpurea , in which the Δcla4 strain is completely unable to sporulate [25] , [26] . Loss of pakB completely blocks the link between septation and cell separation in P . marneffei during intracellular growth , leading to the formation of hyphal cells . This is consistent with the localization of PakB to septation sites in vivo where it is specifically localized to septa only after cell wall deposition had occurred suggesting that PakB is required for cell separation rather than septation . The ΔpakB phenotypes are consistent with the role of CLA4 homologues in regulating cytokinesis . S . cerevisiae , C . albicans and U . maydis CLA4 mutants are also unable to undergo normal cytokinesis during budding [18] , [24] , [27] , [28] . In S . cerevisiae , CLA4 is known to regulate the activity of Lte1p by localisation to the bud cortex , an important event in the Mitotic Exit Network ( MEN ) , and loss of Lte1p leads to delayed cytokinesis [29] , [30] . P . marneffei and U . maydis do not have clear LTE1 orthologues and loss of pakB resulted in increased numbers of nuclei per cell compartment inconsistent with reduced or delayed mitotic exit . Furthermore , given the role of CLA4 orthologues in cytokinesis , it may have been expected that deletion of P . marneffei pakB would also result in septation defects in vegetative cells at 25°C . However , ΔpakB hyphae displayed normal calcofluor stained septa and actin ring formation . Normal calcofluor-white stained septa were also present in the CLA4 deletion strains of C . purpurea , M . grisea and A . gossypii [25] , [26] , [31] . Interestingly , the ΔAgcla4 mutation in A . gossypii results in the absence of almost all actin rings [31] . The localization of wildtype PakB as a cap at the hyphal apex is very similar to that observed for the A . nidulans polarisome component SpaA [32] . The polarisome is a complex of polarity determining proteins originally identified in S . cerevisiae which include Bni1p ( formin ) , Spa2p ( scaffold ) , Bud6p and Pea2p . In S . cerevisiae Ste20p directly phosphorylates Bni1p [33] . Similar to the ΔpakB mutant , deletion of A . nidulans sepA ( BNI1 ) , spaA ( SPA2 ) or budA ( BUD6 ) results in apical branching [32] , [34] . In contrast , deletion of P . marneffei pakA ( STE20 ) does not result in apical branching and PakA co-localizes with actin to discrete spots concentrated at the hyphal apex which is inconsistent with polarisome morphology [12] . Thus , in contrast to S . cerevisiae , it is the Cla4p homologue PakB and not the Ste20p homologue PakA which activates the polarisome during hyphal growth . The CRIB domain is required for Ste20p localization via interaction with Cdc42p in S . cerevisiae . Both the STE20ΔCRIB and STE20H345G mutations result in reduced localization to sites of polarized growth [8] as does the equivalent mutation ( pakAH108G ) in P . marneffei [12] . In contrast , the PakBH204GHA and PakBΔCRIBHA proteins exhibited wildtype localization , with the exception of reduced localization to the hyphal apex in the latter , suggesting that the CRIB domain is largely non-essential for localization . This may represent a new paradigm for PAK function in non-yeast fungi and the mechanism by which PakB is correctly localised remains to be determined . P . marneffei , like many other dimorphic fungal pathogens , has two p21-activated kinases which are responsible for a variety of signaling and morphogenetic activities . The pakA gene encodes a Ste20p-like PAK which is essential for polarity establishment during conidial germination and polarised growth of the pathogenic yeast cells at 37°C such that conidia from ΔpakA strains fail to germinate upon phagocytosis by macrophages [12] . Based on genetic interaction studies it was shown that PakA lies downstream of CflA . However , pakA is not required for germination , polarised hyphal growth or asexual development at 25°C and it was postulated that this role may be filled by pakB [12] . Based on the data presented here , it is clear that pakB fulfils the predicted role of a polarity determinant during hyphal growth and asexual development but plays no role in conidial germination at 25°C , suggesting that this process , unlike its counterpart at 37°C and in vivo , may be PAK-independent . Formal proof of this hypothesis will require the generation of a ΔpakA ΔpakB double mutant . Unexpectedly , pakB plays a critical role in the formation of yeast cells in host cells , instead producing highly branched , septate , hyphal cells , but not in vitro . Upregulation of pakB expression in P . marneffei isolated from macrophages as opposed to in vitro cultured yeast cells shows that PakB activity is likely to be regulated at both the expression level as well as post-translationally . Identifying these host cell specific signals is the important next step in understanding how pathogens sense and respond to their hosts . P . marneffei genomic DNA was isolated as previously described [21] . Southern and northern blotting was performed with Amersham Hybond N+ membrane according to the manufacturer's instructions . Filters were hybridized using [α-32P]dATP labeled probes by standard methods [35] . RNA was isolated from 2161 ( wildtype ) vegetative hyphal cells grown at 25°C for 2 days in liquid medium , from asexually developing cultures grown on solid medium at 25°C for 4 days and from yeast cells grown at 37°C for 6 days in liquid medium . RNA was isolated from yeast cells derived either from LPS activated J774 murine macrophages at 37°C infected with wildtype conidia 24 hours post-infection or from yeast cells incubated in macrophage growth medium ( complete DMEM ) at 37°C for 24 hours . Macrophages were infected as described below . RNA was also isolated from the ΔpakB strain after 2 days growth at 25°C in liquid medium and after 6 days at 37°C in liquid medium . RNA was extracted using TRIzol Reagent ( Invitrogen ) and a MP FastPrep-24 bead beater according to the manufacturer's instructions . RNA was DNaseI treated ( Promega ) prior to RT PCR analysis and a no cDNA synthesis control was performed to ensure no DNA contamination was present . Expression of pakB , brlA and the 25°C or 37°C specific probes 2E11 , 2B10 and 2E4 was determined by RT PCR ( Invitrogen Superscript III One-Step RT-PCR with Platinum Taq ) using the primers: pakB-Y41 ( 5′- ACGGTGCGGTCGGAAAGA-3′ ) , pakB-Y42 ( 5′- CTCCTTACGCACGGGCTG-3′ ) , brlA-FF6 ( 5′- CATTCCCACAACCGATGACT-3′ ) , brlA-FF7 ( 5′-CATACCTGGCGAGATCCACT-3′ ) , 2E11-DD1 ( 5′-TTATTGTTGGCATTGGCG-3′ ) , 2E11-DD2 ( 5′-TTATTGTTGGCATTGGCG-3′ ) , 2B10-CC53 ( 5′-CGGTGCCGTACACAGGTATT-3′ ) , 2B10-CC54 ( 5′-TTGATTTCAGGGCGGAGTAG-3′ ) , 2E4-DD3 ( 5′-ATCCATCCCCCGTGAAGC-3′ ) , 2E4-DD4 ( 5′-GCCGACACGAAGTGATCC-3′ ) , benA-F58 ( 5′-GCTCCGGTGTCTACAATGGC-3′ ) and benA-F59 ( 5′-AGTTGTTACCAGCACCGGAC-3′ ) . A range of cycle numbers was used to ensure the amplification was in the exponential phase and benA was used as an input RNA control . Previously , the A . nidulans STE20 homologous sequence was used to screen a P . marneffei genomic library ( constructed in λGEM-11 ) at low stringency ( 50% formamide , 2xSSC , 37°C ) [12] . A 4 kb SacII/XhoI hybridizing fragment from a second positively hybridizing clone was subcloned into SacII/XhoI digested pBluescript II SK+ ( pKB4752 ) . This clone did not contain the entire pakB ORF so a 5 . 8 kb PstI hybridizing fragment was also subcloned into PstI digested pBluescript II SK+ ( pKB5794 ) . To generate a clone containing the entire pakB ORF , a 3 . 2 kb SacI fragment from pKB4752 was cloned into SacI digested pKB5794 , this generated a 8 . 6 kb pakB clone ( pKB4904 ) . Double stranded sequencing was performed on 4 . 2 kb of clone pKB4904 and analyzed using SequencherTM 3 . 1 . 1 ( Gene Codes Corporation ) . A pakB deletion construct ( pKB6019 ) was generated by cloning a 1 . 8 kb PstI fragment from a 2 . 9 kb HindIII SK+ subclone ( pKB5746 ) into PstI digested pAB4626 ( pyrG+ ) , followed by cloning a 3 kb BglII/XbaI fragment from pKB4904 into BamHI/XbaI ( generating pKB6019 ) . This resulted in pyrG+ flanked by 3 . 2 kb of 5′ and 1 . 8 kb of 3′ pakB sequence , and deleted from +323 to +2618 . To introduce the H204G mutation into pakB , inverse PCR using the mutagenic primers N64 ( 5′- GTCGGTTTCGATCCCAAGACT-3′ ) and N65 ( 5′- GTGGACACGACCGCTGAAATTGG-3′ ) was performed on a 4 kb BgllII/StuI pakB pLitmus 29 subclone ( pKB5964 ) . To introduce the ΔCRIB mutation , inverse PCR using the mutagenic primers S53 ( 5′- CGGGATCCCATTCCTGGGCACCGTTCGT-3′ ) and S54 ( 5′- CGGGATCCATGCGGGAACAGAACCCTCA-3′ ) was performed on pKB5964 . This deletes from +595 to +783 of pakB ( amino acids 195-256 ) . The ΔGBB mutation was introduced by inverse PCR using the mutagenic primers AA7 ( 5′- AATGGAGGACAGTAAAAAGCC-3′ ) and AA8 ( 5′- TCGACTACAGCCCATTTTCA-3′ ) on pKB5964 . This mutation deletes from +2628 to +2661 of pakB ( amino acids 719-729 ) . The promoter was added to these constructs by cloning a 1 . 8 kb NcoI/BgllII fragment into the NcoI/BgllII sites , generating pKB7306 , pKB7171 and pKB6932 , respectively . The integrity of the constructs was confirmed by sequencing . The xylP ( p ) pakB+ pyrG+ ( pKB7024 ) , xylP ( p ) pakBH204G pyrG+ ( pKB7025 ) , xylP ( p ) pakBΔCRIB pyrG+ ( pKB7026 ) and xylP ( p ) pakBΔGBB pyrG+ ( pKB7027 ) constructs were generated by PCR using the primers S51 ( 5′- CGTCTAGATGAACCCTGGACCAGCCCCG-3′ ) and S52 ( 5′- CCATCGATTACTGTCCTCCATTCTTCTT-3′ ) , on pKB6941 ( xylP ( p ) pakB+ pyrG+ ) , pKB7306 ( xylP ( p ) pakBHG pyrG+ ) , pKB7171 ( xylP ( p ) pakBΔCRIB pyrG+ ) and pKB6932 ( xylP ( p ) pakBΔGBB pyrG+ ) , digesting the PCR product with XbaI/ClaI and cloning into XbaI/ClaI pDAP2 ( xylP ( p ) pyrG+ ) . The HA tagged constructs were generated by replacing the 406 bp SacII/HindII fragment of pakB with the SacII/HindII fragment of pKB4693 ( 3x HA tag ) in plasmids pKB6941 , pKB7306 , pKB7171 and pKB6932 , generating pKB7116 ( pakB+ HA ) , pKB7118 ( pakBH204G HA ) , pKB7172 ( pakBΔCRIB HA ) , and pKB7064 ( pakBΔGBB HA ) . This deletes from amino acid 328–343 . This region is a poorly conserved region between the conserved CRIB and kinase domains and includes the first intron . To test for functionality , the pKB7116 ( pakB+ HA ) plasmid was co-transformed with the barA+ gene into the ΔpakB strain . Transformants were selected for glufosinate resistance and confirmed by Southern blot analysis of genomic DNA . Transformation of the ΔpakB strain with the pakB+ HA plasmid ( pKB7116 ) complemented the ΔpakB phenotype . Strains used in this study are shown in Table 4 . P . marneffei FRR2161 , cflAG14V , cflA120A and ΔcflB have been previously described [5] , [6] . Transformation was performed using the previously described protoplast method [21] . The ΔpakB strain ( ΔpakB::pyrG+ ) was generated by transformation of strain G487 ( niaD− pyrG− areA− ) with linearized pKB6019 and selecting for pyrG+ . The ΔpakB pyrG− strain was isolated by plating the ΔpakB strain ( ΔpakB::pyrG+ ) on medium containing 1 mg mL−1 5-fluoroorotic acid ( 5-FOA ) supplemented with 10 mM γ-amino butyric acid ( GABA ) and 5 mM uracil to select for the loss of the pyrG marker . The strain is unable to grow in the absence of 5 mM uracil . The ΔpakB pakB+ , ΔpakB pakBH204G , ΔpakB pakBΔCRIB and ΔpakB pakBΔGBB strains were generated by cotransformation of the ΔpakB pyrG strain with plasmids containing the appropriate mutant allele and pMT1612 ( barA+ ) . The ΔpakB xylP ( p ) pakB+ , ΔpakB xylP ( p ) pakBH204G , ΔpakB xylP ( p ) pakBΔCRIB and ΔpakB xylP ( p ) pakBΔGBB strains were generated by transformation of ΔpakB pyrG− with the appropriate mutant allele and directly selecting for pyrG+ . HA tagged strains were generated by cotransformation of G487 ( niaD− pyrG− areA− ) with the appropriate allele and pMT1612 ( barA+ ) . Southern blot analysis was used to confirm cotransformation and to determine the plasmid copy number . The ΔbrlA ΔpakB double mutant was generated by transformation of strain G526 ( ΔpkuA niaD− pyrG− areA− ) ( K . Boyce and A . Andrianopoulos , unpublished ) with linearized pAB5229 ( A . Borneman and A . Andrianopoulos , unpublished ) and selecting for pyrG+ . Deletion of brlA was confirmed by Southern blot analysis . A ΔbrlA pyrG− strain was isolated by plating the ΔbrlA strain on medium containing 1 mg mL−1 5-FOA supplemented with 10 mM GABA and 5 mM uracil . The ΔbrlA pyrG− strain was transformed with linearized pKB6019 and ΔbrlA ΔpakB double mutants selected for pyrG+ . Deletion of pakB was confirmed by Southern blot analysis . At 25°C strains were grown on A . nidulans minimal medium ( ANM ) or on synthetic dextrose ( SD ) medium supplemented with 10 mM ammonium sulphate ( ( NH4 ) 2SO4 ) as a sole nitrogen source [36] , [37] . At 37°C strains were grown on Brain Heart Infusion ( BHI ) medium ( 3 . 7% brain heart infusion ) or Sabouraud ( Sab ) medium ( 1% mycological peptone , 2% D-glucose ) , malt extract ( ME ) medium ( 0 . 5% mycological peptone and 3% malt extract ) or SD medium supplemented with 10 mM ( NH4 ) 2SO4 . The xylP ( p ) strains were grown on carbon-free ANM plus 10 mM ( NH4 ) 2SO4 supplemented with either 1% glucose or 1% xylose at 25°C and on BHI ±1% xylose at 37°C . J774 murine macrophages ( 1×105 ) were seeded into each well of a 6 well microtitre tray containing one sterile coverslip and 2 mL of complete Dulbecco's Modified Eagle Medium ( complete DMEM: DMEM , 10% fetal bovine serum , 8 mM L-glutamine and penicillin-streptomycin ) . Macrophages were incubated at 37°C for 24 hours before activation with 0 . 1 µg mL−1 lipopolysaccharide ( LPS ) from E . coli ( Sigma ) . Macrophages were incubated a further 24 hours at 37°C , washed 3x in phosphate buffered saline ( PBS ) and 2 mL of complete DMEM medium containing 1×106 conidia or yeast cells ( grown at 37°C for 8 days in liquid BHI medium ) was added . A control lacking conidia or yeast cells was also performed . Macrophages were incubated for 2 hours at 37°C ( to allow conidia or yeast cells to be engulfed ) , washed once in PBS ( to remove free conidia ) and incubated a further 24 hours at 37°C . Macrophages were fixed in 4% paraformaldehyde and stained with 1 mg mL−1 fluorescent brightener 28 ( calcofluor - CAL ) to observe fungal cell walls . The numbers of germinated conidia was measured microscopically by counting the numbers of germinated conidia ( conidia with a visible germ tube ) or yeast cells in a population of approximately 100 fungal cells in macrophages . The numbers of cells with septa was measured microscopically by counting in a population of approximately 100 cells . Three independent experiments were performed . Mean and standard error of the mean values were calculated using GraphPad Prism3 . To analyze the significance of the septation results , two-level nested analysis of variance ( ANOVA ) was performed on the data to test if the percentage of septate cells was significantly different among genotypes and also between transformants of the same genotype ( http://udel . edu/~mcdonald/statnested . html ) ( Table S2 ) . Superoxide production was detected by addition of a 0 . 05% Nitrotetrazolium Blue Chloride ( NBT ) ( Sigma ) solution ( in 0 . 05 M sodium phosphate buffer ( pH 7 . 5 ) ) to macrophages 24 hours post-infection . Macrophages were incubated for 1 hour and observed by differential interference contrast ( DIC ) microscopy . To examine whether the presence of host extracts is sufficient to induce the morphological switch , LPS activated J774 murine macrophages in complete DMEM were lysed by freezing at −70°C for 20 minutes and slow thawing . The lysed extracts were added to wildtype , ΔpakB and ΔpakB pakB+ conidia and incubated for 24 hrs at 37°C . Minus macrophage controls were also performed for comparison . P . marneffei strains were grown on slides covered with a thin layer of solid medium , with one end resting in liquid medium [21] . Strains were grown on ANM medium supplemented with ( NH4 ) 2SO4 at 25°C for 2 or 4 days . To observe conidia and yeast cells , asexually developing plates were harvested into 0 . 005% Tween 80 solution and filtered through Miracloth to remove hyphae . To observe conidiophores , asexually developing plates were scraped onto a coverslip containing 5 µL 0 . 005% Tween 80 solution . At 37°C strains were grown on BHI medium for 5 days or Sab and ME medium for 6 days . xylP ( p ) strains were grown on ANM plus 10 mM ( NH4 ) 2SO4 ±1% xylose at 25°C and on BHI ±1% xylose at 37°C . For germination experiments , approximately 106 spores were inoculated into 300 µL of SD plus 10 mM ( NH4 ) 2SO4 and incubated for 15 hours at 25°C or 37°C . The rates of germination were measured microscopically by counting the numbers of germinating conidia ( conidia with a visible germ tube ) in a population of 100 cells . The number of conidia with 1 , 2 or 3 or more germ tubes was counted in a population of 100 cells . Three independent experiments were performed . Mean and standard error of the mean values were calculated using GraphPad Prism3 . Immunofluorescence microscopy for examination of the actin cytoskeleton was performed with a mouse C4 monoclonal anti-actin primary ( Chemicon International , Inc . ) and an ALEXA 488 rabbit anti-mouse secondary antibody ( Molecular Probes ) . PakB HA immunofluorescence localization was performed with 3F10 rat monoclonal anti-HA primary ( Boehringer Mannheim ) and an ALEXA 488 goat anti-rat secondary antibody ( Molecular Probes ) using standard protocols [38] . Immunofluorescence microscopy controls using only primary or secondary antibodies were performed to confirm the specificity of the antibodies . Slides were examined using differential interference contrast ( DIC ) and epifluorescence optics for antibody fluorescence , cell wall staining with fluorescent brightener 28 ( calcofluor - CAL ) or nuclear staining with Hoechst 33258 and viewed on a Reichart Jung Polyvar II microscope . Images were captured using a SPOT CCD camera ( Diagnostic Instruments Inc ) and processed in Adobe PhotoshopTM . For scanning electron microscopy ( SEM ) , strains were grown on ANM or BHI for 10 days at 25°C or BHI for 5 days at 37°C . xylP ( p ) strains were grown on carbon-free ANM plus 10 mM ( NH4 ) 2SO4 supplemented with either 1% glucose or 1% xylose at 25°C for 14 days . Agar cubes containing the fungal biomass were fixed with 2 . 5% glutaraldehyde in PBS buffer for 2 hours , washed 3 x in PBS and postfixed with 1% osmium tetroxide for 2 hours . Samples were then washed 3 x in PBS and ethanol dehydrated by washing in increasing concentrations of ethanol . Samples were dried in a Balzers CPD 030 Critical Point Dryer and gold coated in an Edwards S150B Gold Sputter Coater . Samples were examined with a Philips XL30 FEG Field Emission Scanning Electron Microscope .	Dimorphic fungal pathogens pose significant health and agricultural problems worldwide . These fungi have the capacity to switch between a multicellular hyphal growth form and a unicellular yeast growth form . Often one form is pathogenic , found in infected hosts , and the other is not . Many dimorphic fungal pathogens of humans produce the yeast form during infection and this form resides within host phagocytic immune cells , where it can tolerate killing by these cells and is not exposed to the acquired immune system . Inhibiting the pathogen's ability to switch growth forms has been shown to block pathogenicity . This study identifies a pathway used by the fungal pathogen to sense the host and switch to the appropriate growth form . This study provides new insights into the molecular mechanisms which are important for pathogenicity and may identify factors which can be targeted to block the ability of the pathogen to successfully reside within host cells .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "cell", "biology/morphogenesis", "and", "cell", "biology", "infectious", "diseases/fungal", "infections", "cell", "biology/microbial", "growth", "and", "development", "genetics", "and", "genomics/gene", "function", "microbiology/microbial", "growth", "and", "development", "...	2009	In Vivo Yeast Cell Morphogenesis Is Regulated by a p21-Activated Kinase in the Human Pathogen Penicillium marneffei
The detection of epistatic interactive effects of multiple genetic variants on the susceptibility of human complex diseases is a great challenge in genome-wide association studies ( GWAS ) . Although methods have been proposed to identify such interactions , the lack of an explicit definition of epistatic effects , together with computational difficulties , makes the development of new methods indispensable . In this paper , we introduce epistatic modules to describe epistatic interactive effects of multiple loci on diseases . On the basis of this notion , we put forward a Bayesian marker partition model to explain observed case-control data , and we develop a Gibbs sampling strategy to facilitate the detection of epistatic modules . Comparisons of the proposed approach with three existing methods on seven simulated disease models demonstrate the superior performance of our approach . When applied to a genome-wide case-control data set for Age-related Macular Degeneration ( AMD ) , the proposed approach successfully identifies two known susceptible loci and suggests that a combination of two other loci—one in the gene SGCD and the other in SCAPER—is associated with the disease . Further functional analysis supports the speculation that the interaction of these two genetic variants may be responsible for the susceptibility of AMD . When applied to a genome-wide case-control data set for Parkinson's disease , the proposed method identifies seven suspicious loci that may contribute independently to the disease . With the development of modern human and medical genetics , it has been widely accepted that genetic variation plays an important role in the pathogenesis of genetic inherited diseases [1] . The identification of causative genetic variants therefore becomes the primary step towards the understanding of genetic principles underlying these diseases . For Mendelian diseases in which an individual genetic variant in a single gene is both sufficient and necessary to cause a disease , classical statistical approaches such as linkage analysis [2]–[5] and association studies [6] , [7] have shown remarkable successes in the identification of causative genetic variants . Nevertheless most common diseases are complex ones that are supposed to be caused by multiple genetic variants , their interactive effects , and/or their interaction with environment factors [7] , [8] . The detection of such interactive effects therefore plays a key role in the understanding of these diseases . The interactive effects of multiple genetic variants underlying complex diseases are often referred to as epistasis or epistatic interactions . Recent advances in biomedical studies have been confirming the contribution of epistasis to complex diseases . For example , Tiret et al reported synergistic effects of polymorphisms in the angiotensin-converting enzyme and the angiotensin-II type 1 receptor gene on the risk of myocardial infarction [9] . Ritchie et al identified the association of a high-order interaction among four polymorphisms in three estrogen-metabolism genes with breast cancer [10] . Williams et al reported the influence of a two-locus interaction between polymorphisms in the angiotensin converting enzyme and the G protein-coupled receptor kinase on hypertension susceptibility [11] . Tsai et al identified the association of a three-locus interaction among polymorphisms in renin-angiotensin system genes with atrial fibrillation [12] . Cho et al reported the association of a two-locus interaction between polymorphisms in the uncoupling protein 2 gene and the peroxisome proliferator-activated receptor gamma gene with Type 2 diabetes mellitus [13] . Martin et al reported the influence of a two-locus interaction between polymorphisms in KIR3DL1 and HLA-B on both AIDS progression and plasma HIV RNA [14] . With these examples , epistasis between multiple genetic variants is now widely believed to be the causative pattern of human complex diseases . In order to detect epistasis , a number of multi-locus approaches have been developed . For example , Hoh et al proposed a trimming , weighting , and grouping approach that used the summation of statistics on the basis of single-locus marginal effects and the Hardy-Weinberg equilibrium ( HWE ) for hypothesis testing [15] . Nelson et al proposed a combinatorial partitioning method ( CPM ) that exhaustively searched for a combinatory genotype group that had the most significant difference in the mean of the responding continuous phenotype [16] . Culverhouse et al proposed a restricted partitioning method ( RPM ) which modified CPM by ignoring partitions that combined individual genotypes with very different mean trait values [17] . Millstein et al proposed a focused interaction testing framework ( FITF ) in which a prescreening strategy was developed to reduce the number of tests [18] . Chatterjee et al used Turkey's 1-degree-of-freedom model to detect interacting loci from different regions [19] . Ritchie et al proposed a multifactor-dimensionality reduction ( MDR ) method in which exhaustive search was performed to detect combinations of loci with the highest classification capability [10] . Although these methods have shown their successes in association studies for small scale candidate genes [10] , [15]–[19] , their effectiveness for large scale case-control data has not yet been validated . Besides , most of the methods rely strongly on exhaustive search for combinations of multiple loci . This search strategy , though feasible when the number of candidate genetic variants is small , can hardly be computationally practical for large scale or whole-genome association studies in which the number of candidate genetic variants is typically very huge . For example , a study on Age-related Macular Degeneration ( AMD ) has genotyped more than 100 thousand single nucleotide polymorphism ( SNP ) markers for 96 patients and 50 unaffected people [20] , and a recent genome-wide association study on Parkinson's disease has genotyped more than 400 thousand SNP markers for 270 patients and 271 unaffected people [21] , [22] . With such dense SNPs being genotyped , methods based on exhaustive search are computationally impractical due to the vast number of possible combinations of the SNP markers . The main challenge for genome-wide association studies is therefore to design computational approaches that are capable of avoiding the “combinatorial explosion” curse to identify epistatic interactions . A recent breakthrough in genome-wide epistasis mapping is the introduction of the Bayesian epistasis association mapping ( BEAM ) method [23] that integrates a Bayesian model with the Metropolis-Hastings algorithm to infer the probability that each locus is associated with the susceptibility of a specified disease . BEAM classifies SNP markers into three types: SNPs unassociated with the disease , SNPs contributing to the disease susceptibility independently , and SNPs influencing the disease risk jointly with each other . However , the genetic models for complex diseases could be far more complicated than that proposed by BEAM . For example , the disease-associated SNPs that jointly influence the disease risk may be further divided into subgroups , in which a SNP interacts with other SNPs in the same subgroup , but not with those in the other subgroups . This situation could be very common in real data , making BEAM ineffective in the exploration of true interactive effects of multiple loci . To overcome this limitation , in this paper , we give an explicit presentation of “epistasis” and define “epistatic modules” as basic units of disease susceptibility loci . On the basis of this notion , we put forward a Bayesian marker partition model to explain the observed case-control data and further generalize this model to account for the existence of linkage disequilibrium ( LD ) between genetic variants . To facilitate the identification of epistatic modules , we develop a Gibbs sampling strategy with a reversible jump Markov chain Monte Carlo ( RJ-MCMC ) procedure to simulate the posterior distribution that genetic variants belong to the epistatic modules and further resort to hypothesis testing to screen out statistically significant modules . In contrast to most of the existing methods that entirely or partially rely on exhaustive search for combinations of loci , the proposed approach , named epiMODE ( epistatic MOdule DEtection ) , natively identifies interactive loci ( epistatic modules ) without enumerating their combinations , thereby being capable of detecting interactive effects of multiple loci from a vast number of genotyped genetic variants . We systematically compare the proposed approach with three existing methods on seven simulated disease models . The results show the superior performance of our approach over the other methods . We further apply the proposed approach to a genome-wide case-control data set for Age-related Macular Degeneration ( AMD ) that contains more than 100 thousand SNPs genotyped for 96 cases and 50 controls [20] and successfully identify two SNPs that are known to be associated with the disease . Besides , the results also suggest that two other SNPs ( rs1394608 and rs3743175 ) may have interactive effects on the susceptibility of the disease . We also apply the proposed approach to a genome-wide case-control data set for Parkinson's disease ( 400 thousand SNPs genotyped for 270 cases and 271 controls ) [21] , [22] and identify seven SNP markers that may be associated with the disease . The concept of epistasis implies that the phenotypic effect of one locus is dependent on one or more other loci . Nonetheless the definitions of epistasis in biology and statistics are not exactly consistent . Even from the statistical perspective only , researchers have different understandings of epistasis [24] , [25] . Considering these inconsistencies , it is necessary to first give a clear definition of epistasis , for the purpose of developing a computational method for identifying multiple loci that contribute to the disease susceptibility . In this paper , a locus stands for a SNP . A genotype stands for a set of two alleles ( one inherited from father and the other from mother ) at a locus and has three possible values: homozygosity of common alleles , homozygosity of minor alleles , and heterozygosity . A combinatory genotype represents the genotype of a combination of multiple loci . For a combination of t loci , the number of all possible combinatory genotypes is 3t . The penetrance of a combinatory genotype is the probability/risk that an individual with this combinatory genotype is affected , given the combinatory genotype of the multiple loci . We first assume that all loci are in linkage equilibrium , also known as independent , and then we generalize the definitions to the situation with linkage disequilibrium between multiple loci . Let be the set of all L loci under investigation , and be the set of all s disease susceptibility loci that determine the disease risk . For any two subsets , S1 and S2 , of S ( , , and ) , their penetrance given the combinatory genotypes and respectively , can be described aswhere represents the penetrance of a given combinatory genotype , a combinatory genotype of the multiple loci , and the function denoting how combinatory genotypes determine the disease penetrance . For any given combinatory genotypes of S1 and S2 , ifis always true , the relationship between the two subsets of loci S1 and S2 is defined as “independently contributing” to the disease . Otherwise , the relationship between S1 and S2 is defined as “epistasis . ” Particularly , the relationship between a set of loci and a null set is defined as epistasis . A set of loci is an “epistatic module” if and only if the relationship between S1 and its complement , , is “independently contributing , ” that is , for any given genotype and , and the relationship between any subset of S1 , , and its complement is epistasis . Obviously , the set of disease susceptibility loci S consists of one or more epistatic modules . We further verify that there is no overlap between any two epistatic modules , and epistatic modules are independent in both case and control populations ( Text S1 ) . In genome-wide association studies where the SNPs are quite dense , it is common that a SNP may be in LD with other SNPs . To account for this situation , we define a group of SNPs that are in LD with each other as an “LD set” and extend the above definition of epistatic modules by replacing individual loci with LD sets . Note that with this extension , all properties of epistatic modules remain unchanged , as long as we treat an LD set as an individual locus in the previous derivation . The mechanism how a number of susceptibility SNPs contribute to the disease risk through epistatic modules is shown in Figure 1 . The disease risk is determined by a number of epistatic modules , each of which contributes to the disease independent of the others . An epistatic module is composed of one or more susceptibility SNPs , each of which may be in LD with some other SNPs , forming an LD set . A disease susceptibility SNP , together with the SNPs that are in LD with it , relies on other disease susceptibility SNPs or LD sets in the same epistatic module to affect the disease susceptibility . An epistatic module cannot be further divided into smaller epistatic modules; hence epistatic modules are the smallest genetic units that independently influence the disease risk . Suppose that in a population-based case-control study , cases and controls are genotyped at a number of L SNP markers . The genotypes for cases and controls are represented as and , respectively , where and denote the genotypes of the i-th patient and the j-th unaffected individual at the L markers , respectively . With the understanding of epistatic modules , the L markers can be partitioned into modules M0 , M1 , … , MS , with M0 containing markers unlinked to the disease and M1 to MS being epistatic modules . Let ( ) be an indicator of the assignment of the i-th marker into one of the modules , and be a vector representing the assignments for all of the L markers . Obviously , has possible values . Let lm be the number of markers falling into the m-th module ( ) . We have that . Let Dm and Um be the genotypes of the sets of markers that belong to the m-th module in the case and the control populations , respectively . Obviously , we have that when and for the case population , and when and for the control population . With these concepts , the problem of finding markers that have epistatic interactions on the disease risk is equivalent to a problem of assigning the markers to epistatic modules . Particularly , the assignment for a marker can be done by first calculating the probability of the observed data given a certain marker partition pattern and then obtaining the posterior probability that the marker belongs to each module using some sampling strategy . For a clear presentation , we first derive a Bayesian model that assumes independence between SNPs and then generalize the model to account for the existence of LD sets . The module M0 consists of markers that are unlinked to the disease . Therefore , markers in D0 ( the case population ) should follow the same distribution as those in U0 ( the control population ) . Let , , be the probabilities of occurrence of the three possible genotypes for the i-th marker in M0 , and be the vector that is composed of all probabilities of genotypes of the l0 markers belonging to M0 . Let and be the number of individuals that have the k-th genotype at the i-th marker in the case and the control populations , respectively . The joint distribution of the observed genotypes D0 and U0 , given the partition I and the parameters can then be written as ( 1 ) Following the Bayesian approach , we assume that every ( ) follows a Dirichlet distribution with the hyper-parameter , that is , . Integrating out in Equation ( 1 ) , we obtain ( 2 ) where is the Gamma function . For an epistatic module Mm ( ) containing lm SNPs , there are a total of combinatory genotypes . Let and be the probabilities of occurrences of all combinatory genotypes in the case and the control populations , respectively . Let and be the numbers of occurrences of the k-th combinatory genotype in the case and the control populations , respectively . The distributions of Dm and Um , given the parameters and , can be written asrespectively . Assuming that and follow Dirichlet prior distributions with hyper-parameters and , respectively , we integrate out and and obtain ( 3 ) and ( 4 ) As the distributions of Dm and Um are independent , we havePutting the above likelihood functions together , we have the posterior distribution of I , given the observed genotypes , as The prior distribution need to be determined in advance . For simplicity , we assume that the partition of the loci are independent , and for each locus , without prior knowledge , the probability that it belongs to the m-th module is ( and ) . With these two assumptions , we have . Note that when prior knowledge that can be used to infer the relationship between a locus and the disease risk is available , the corresponding could be updated accordingly . We assume that all Dirichlet hyper-parameters are equal to 0 . 5 unless otherwise specified . We use a first-order Markov model to account for the situation in which a set of SNPs are in LD with a disease susceptibility SNP in an epistatic module , say , an LD set . For a clear presentation , we refer to the disease susceptibility SNP as the core SNP and SNPs in LD with it as peripheral SNPs . Given a core SNP , the likelihood of the genotypes of a peripheral SNP in the case population iswhere is the probability that the peripheral SNP has the k-th genotype conditional on that the core SNP has the j-th genotype , and is the number of cases for which the core and peripheral SNPs have the j-th and k-th genotypes , respectively . Assuming Dirichlet priors with hyper-parameters for , we integrate out and obtain the posterior distribution of the genotypes of the peripheral SNP in the case population conditional on the core SNP as Suppose that in a module with SNPs , there are core SNPs and peripheral SNPs ( ) . Let be the set of peripheral SNPs that are in LD with the c-th core SNP ( ) . We have that the intersection of any two of these sets is empty , while the union of all these sets contains all peripheral SNPs . The posterior distribution of the genotypes of the set of peripheral SNPs in the case population conditional on the c-th core SNP is given by ( 5 ) where ( , , ) are Dirichlet hyper-parameters , and is the number of cases for which the c-th core SNP has the j-th genotype , and the i-th peripheral SNP has the k-th genotypes . Putting Equations ( 3 ) and ( 5 ) together , the likelihood of the genotypes in the case population iswhere pcore is given by Equation ( 3 ) as Similarly , by replacing the case population with the control population , the likelihood of the genotypes in the control population can be obtained as where is given by Equation ( 4 ) asand is given by ( 6 ) where ( , , ) are Dirichlet hyper-parameters , and is the number of controls for which the c-th core SNP has the j-th genotype , and the i-th peripheral SNP has the k-th genotypes . Finally , the likelihood of observing both the case and the control populations is given by ( 7 ) We also assume that all Dirichlet hyper-parameters are equal to 0 . 5 unless otherwise specified . The Bayesian marker partition model described above assumes independence between SNPs that are unlinked to the disease . Nevertheless the existence of LD may make distributions of genotypes of these SNPs dependent . In the model discussed above , there is no specific module for these linked disease-unassociated SNPs . As a result , these SNPs could be partitioned into some epistatic modules and negatively affect the correct partition of these modules . We therefore propose the use of LD modules to account for the existence of LD between disease-unassociated SNPs . Although the distributions of genotypes for markers in LD are dependent in both the case and the control populations , as those for markers in epistatic modules , the underlying principle between LD markers and epistatic modules are quite different . For LD markers , the distributions of genotypes are almost the same for the case and the control populations , while for epistatic modules the distributions of genotypes are different between the case and the control populations . In order to incorporate this understanding into the Bayesian partition model , we assume that other than the S epistatic modules , there further exist T LD modules , labeled by {} , in each of which loci are in strong LD with each other . We also use a first-order Markov model to account for LD between the SNPs in an LD module . For an LD module ( ) , we assume that there exists a core SNP c , and the distributions of genotypes of all other ( peripheral ) SNPs in this LD module depend on the genotype of this core SNP . Let be the set of the peripheral SNPs that are in LD with the core SNP . Using similar reasoning as for the epistatic modules , we obtain that is derived with a similar way as Equation ( 2 ) and is given bywhere and are the numbers of individuals that have the k-th genotype at the core SNP in the case and the control populations , respectively . is derived with a similar way as Equation ( 6 ) and is given bywhere are Dirichlet hyper-parameters , and and are the numbers of individuals for which the core SNP has the j-th genotype , and the i-th peripheral SNP has the k-th genotypes in the case and the control populations , respectively . We also assume that all hyper-parameters are equal to 0 . 5 unless otherwise specified . With LD modules being incorporated , the posterior distribution for the generalized indicator vector under the generalized Bayesian model is then The posterior distribution of the partition I given by the above Bayesian partition model suggests the following Gibbs sampler ( 8 ) where and . In order to calculate this sampler in an efficient way , we computefor , and then obtainWith this sampler , a Gibbs sampling algorithm can be performed as follow . In order to calculate the Gibbs sampler , i . e . , Equation ( 8 ) , we need to partition SNPs in epistatic and LD modules into core SNPs and peripheral SNPs , say , to obtain structures of the modules . Besides , the numbers of modules ( S and T ) are also unknown . We will address these two questions in the following two sections . Given a set of SNPs in an epistatic module , we need to partition the SNPs into non-overlap LD sets . For each LD set , we need to assign a core SNP . The partition of LD sets , together with the assignment of a core SNP for each LD set , is referred to as the structure of an epistatic module . A naïve method for obtaining the structure of a module is to exhaustively search for all possible structures of the module and then select the one with the maximum likelihood . Specifically , for an epistatic module Mm ( ) containing SNPs , there are ways for selecting the core SNPs , corresponding to the different ways of selecting non-empty subsets from the SNPs . Furthermore , in the case that the number of core SNPs is , the number of ways for associating the rest peripheral SNPs to the core SNPs is , since each peripheral SNP can be assigned to one of the core SNPs , and the assignments are mutually independent . Obviously , the number of all possible structures of an epistatic module grows rapidly , making the exhaustive search strategy practical only when the module contains a small number of SNPs . We therefore propose the following sampling approach to search for a reasonable module structure when the exhaustive search strategy is hard to apply . For an epistatic module with SNPs , in which are core SNPs , and the rest are peripheral ones , we index the core SNPs by numbers from 1 to , and we index the peripheral SNPs by numbers from to . We further introduce an indicator vector , representing the status of all SNPs in the module . In this vector , ( ) means that the i-th SNP is a core SNP , and means that the i-th SNP is a peripheral SNP of the k-th core SNP . Consider a peripheral SNP indexed by i ( ) . The posterior distribution of the indicator , given the rest of the indicators and the observation and , can be written aswhere the likelihood function can be calculated in a similar way as Equation ( 7 ) . Assuming equal prior probabilities for all possible structures of the module , the above posterior distribution suggests the following Gibbs sampler for the peripheral SNP , ( 9 ) Consider a core SNP indexed by i ( ) . There are two situations: ( 1 ) the core SNP has some peripheral SNPs , and ( 2 ) the core SNP has no peripheral SNPs . In the former case , we need to fix the indicator . In the latter case , a Gibbs sampler can be obtained as ( 10 ) where we exclude the situation in which the core SNP becomes its own peripheral SNP . The above Gibbs samplers suggest the following sampling strategy: To further reduce the computational burden , we propose the following forward and backward strategies that are very economy in terms of computation time . In the forward strategy , we consider three situations of adding a SNP into an existing epistatic module . First , the SNP is itself a core SNP , and there are no other SNPs in LD with it . Second , the SNP is in LD with an existing core SNP , and this core SNP remains unchanged . Third , the SNP is in LD with an existing core SNP , but this core SNP needs to be updated as the added SNP . To deal with the first case , we try to create a new LD set to include the new SNP as the core SNP in constant time complexity . To deal with the second case , we try to add the new SNP as a peripheral SNP to every existing LD set in linear time complexity , proportional to the number of existing LD sets . To deal with the third case , we try to add the new SNP as the core SNP and downgrade the previous core SNP to a peripheral SNP for every existing LD set in linear time complexity , also proportional to the number of existing LD sets . Finally , we compare likelihood values of resulting structures of the above efforts and select the structure with the highest likelihood as the new module structure . In the backward strategy , we also consider three situation of removing a SNP from an existing epistatic module . First , the SNP is in LD with a core SNP . Second , the SNP is itself a core SNP with no other SNPs in LD with it . Third , the SNP is a core SNP with some other SNPs in LD with it . The first and second cases can be dealt with in constant time complexity . The third case can be exhaustively searched for the new core SNP in linear time complexity , proportional to the number of SNPs in LD with the removed SNP . By comparing the likelihood values of these three cases , we can obtain a new structure for the module . The exhaustive search strategy can provide optimal module structures , but its computation time is acceptable only when a module contains a small number of SNPs . The sampling strategy takes uncertainty in the partitioning process into consideration and can alleviate the computational burden when a module contains a large number of SNPs . The forward and the backward strategies can greatly reduce the computational burden and offer sub-optimal module structures . To achieve a reasonable trade-off between the computational burden and the optimality of module structures , we also propose a hybrid strategy in which we mainly perform the forward and the backward strategies and periodically apply the exhaustive search or the sampling methods . According to our experience , the hybrid strategy is much faster than the exhaustive search and the sampling methods and can yield similar results as the other two methods in most cases . Therefore , we suggest the use of the hybrid strategy . Similar to epistatic modules , we need to also assign a core SNP for each LD module . However , the situation is quite simple for obtaining structures for LD modules , because an LD module has only one core SNP , and thus the number of possible structures for an LD module is equal to the number of SNPs in the module . In the exhaustive search strategy , we can search for the core SNP in linear time complexity , proportional to the number of SNPs in the module . In the forward strategy , we consider the situation of adding a SNP into an LD module , and determine the structure by comparing the likelihood values of two cases: ( 1 ) the added SNP is a peripheral SNP , and ( 2 ) the added SNP is the core SNP . This can be done in constant time complexity . In the backward strategy , we consider the situation of removing a SNP from the module . If the removed SNP is not the core SNP , we simply remove it . In the case that the deleted SNP is the core SNP , we select a new core SNP from the previous peripheral SNPs by exhaustive search , which can be done in linear time complexity , proportional to the number of SNPs remaining in the module . Since the exhaustive search strategy is straightforward and already computationally economy ( linear complexity ) , we simply apply the exhaustive search strategy to obtain structures for LD modules . With the module structures being obtained , we are now able to calculate the Gibbs sampler defined by Equation ( 8 ) . We assume that the numbers of epistatic modules ( S ) and LD modules ( T ) are already known in the Gibbs sampling strategy for marker partitioning . Nevertheless the values of S and T are usually unknown in real applications . To address the uncertainty of S and T , we adopt a reversible jump Markov chain Monte Carlo ( RJ-MCMC ) procedure [26] as follows . With a sufficient number of the above RJ-MCMC sampling procedure being repeated , the Markov chains for S and T could achieve their stable distributions . In our studies , we use , , and . The RJ-MCMC procedure samples the posterior distributions of the numbers of epistatic and LD modules , while the Gibbs sampling algorithm gives us the posterior probability that a locus belongs to a module and enables us to sample the indicators with the use of their conditional distributions in a sequential way . Starting from an initial ( random ) assignment of the indicators , the Gibbs sampling procedure simulates a Markov chain whose stationary distribution follows the distribution of the indicator vector . When the Markov chain reaches its stationary distribution after a number of burn-in iterations , we record candidate epistatic modules and their posterior probabilities . The posterior probability of an epistatic module represents the strength that the module is associated with the disease and thus can be directly used to make statistical inference . For example , biologists can select epistatic modules with top posterior probabilities for further functional analysis or biological experiments . Nevertheless , the statistical significance of epistatic modules might be more desired by geneticists . We therefore provide in the following parts a permutation test method and a “selection-testing-correction” approach for assessing the statistical significance of candidate epistatic modules . The URL for the software presented herein is as follows: http://bioinfo . au . tsinghua . edu . cn/epiMODE In order to verify the capability of the proposed approach in the detection of epistatic interactions in real genome-wide association studies , we apply epiMODE to an Age-related Macular Degeneration ( AMD ) data set [20] , which contains 103 , 611 SNPs genotyped with 96 cases and 50 controls . The authors of the original paper reported that two SNPs , rs380390 and rs1329428 , were believed to be significantly associated with AMD . Our method successfully indentifies both of the two SNPs through the identification of an epistatic module that included these two SNPs ( two more SNPs are also indentified in the same epistatic module , and the posterior probability of the module is above 0 . 9 , see Figures 6 and 7 ) . The nominal p-values for rs380390 and rs1329428 are 1 . 75×10−6 and 3 . 64×10−6 , respectively , according to the Chi-squared test with two degrees of freedom . Our method also indentifies two novel SNPs , rs1394608 and rs3743175 , by detecting an epistatic module that includes both loci ( two more SNPs in LD with them are also indentified in the same epistatic module , and the posterior probability of the module is greater than 0 . 9 , see Figures 6 and 7 ) . The nominal p-values for these two SNPs are 8 . 81×10−5 and 1 . 76×10−3 , respectively , according to the Chi-squared test with two degrees of freedom . Note that the p-value for the combination of rs1394608 and rs3743175 is 7 . 39×10−7 , while the p-value for the combination of rs380390 and rs1329428 is only 1 . 84×10−5 , according to the Chi-squared test with eight degrees of freedom . The distributions of the combination of rs1394608 and rs3743175 in cases and controls are shown in Figures 8A and 8B , respectively . According to Chi-squared tests with four degrees of freedom , these two SNPs are independent in controls ( p-value = 7 . 50×10−1 ) and dependent in cases ( p-value = 3 . 27×10−3 ) . We also infer the genotype frequencies of the combination of these two SNPs according to their distributions in controls and the Hardy-Weinberg equilibrium ( HWE ) , and we further infer the penetrance for the combination of these two SNPs according to their distributions in cases and the inferred genotype frequencies , as shown in Figure 8C ( see Text S1 for details of inferring genotype frequencies and the penetrance ) . The penetrance of genotypes of rs1394608 differs stronger from that of rs3743175 , suggesting that rs1394608 may be the dominant locus for disease susceptibility . Specifically , the homozygote TT of rs1394608 is responsible for disease risk significantly higher than the heterozygous and the other homozygous genotype . However , the effect of rs1394608 is strongly regulated by rs3743175 , especially for the homozygous genotype TT of rs1394608 . The penetrance for the combination genotype TT/CC ( for rs1394608 and rs3743175 , respectively ) , is 9 . 64×10−2 , significantly larger than the penetrance for the combination genotype TT/CT and the penetrance for the combination genotype TT/TT . Odds ratio values in table 2 also give similar results . From the above analysis , we infer that the relationship between the combination of SNPs rs1394608 and rs3743175 and the disease risk is a classic epistatic interaction , in which one dominant variant locus ( rs1394608 ) is regulated by the other locus ( rs3743175 ) . In the following part , we perform functional analysis of these two SNPs . AMD is the primary cause of irreversible visual loss in the Western world [29] . The clinical hallmark of AMD is pathological extracellular deposits in retinal called drusen . Previous single-locus studies have identified the complement factor H ( CFH ) and the HtrA serine peptidase 1 ( HTRA1 ) as two major risk genes for AMD [30]–[32] . Despite the complex etiology of AMD , no significant epistasis has been identified by BEAM in the genome-wide case-control data used in this study . The most significant epistatic effect we identified is between SNPs rs1394608 and rs3743175 . Interestingly , there is another SNP rs2828155 with exactly the same genotype distribution as rs3743175 among all case/control samples ( rs2828155 is also detected by epiMODE in the same module with rs1394608 and rs3743175 , see Figures 6 and 7 ) ; therefore the epistasis may also exist between rs1394608 and rs2828155 . rs1394608 resides within the intron of SGCD , a gene located on chromosome 5q33-34 , which has been implicated in AMD [33] , [34] and predispose to drusen formation [35] . SGCD is the delta subunit of the sarcoglycan complex , a component of the dystrophin-glycoprotein complex , linking the cytoskeleton to the extracellular matrix . The sarcoglycan complex involves in plasma membrane deposition , and the co-expression of SGCD and SGCB ( beta subunit ) is responsible for delivery to and retention of sarcoglycan complex at the cell surface [36] . Defects in SGCD are the cause of limb-girdle muscular dystrophy type 2F ( LGMD2F ) and dilated cardiomyopathy 1L ( CMD1L ) . The detected SNP rs1394608 , together with all 16 SNPs of strong LD ( r2>0 . 8 ) within 1 Mb neighboring region , are all significantly associated with the expression of FBLN1 ( p-value<1×10−7 , according to [37] ) , a gene belongs to the fibulin family of extracellular matrix proteins . Other members ( FBLN3 , FBLN5 , FBLN6 ) of the family have been associated with AMD [38]–[41] , and various evidences support that FBLN1 may also play a role in AMD [39] , [42]–[45] . Specifically , FBLN1 can act as a cofactor for the matrix metalloprotease ADAMTS1 and play important roles in the degradation of proteoglycans by ADAMTS1 during pathological conditions induced by inflammatory processes [46] . Therefore variants in SGCD may lead to AMD in a similar way to HTRA1 , which may regulate the degradation of extracellular matrix by facilitating access of other degradative matrix enzymes , such as matrix metalloproteinases to their substrates [47] . rs3743175 resides within the intron of SCAPER/ZNF291 , a gene located on chromosome 15q24 . Iyengar et al [34] have also identified linkage signal for a marker in the nearby locus 15q21 . A weak linkage signal on chromosome 15q has also been observed in another full-genome scan [48] . Further , translocation of 15q24 had been found in a patient with visible disc drusen [49] . Sequence analysis [50] identified in SCAPER an unstable non-coding tandem repeat , an important form of mutation responsible for several neurological , neurodegenerative and neumuscular disorders [51] . The detected SNP rs3743175 has the strongest association with the expression of the gene itself ( p-value = 9 . 5×10−14 ) . The mechanism for the epistasis between rs1394608 in SGCD and rs3743175 in SCAPER is unclear . We speculate that SCAPER may exert influence to SGCD susceptibility through the regulation of aging process , as evidences show both genes are involved in cell cycle regulation and DNA repair . eQTL analysis [37] shows that rs1394608 is significantly associated with the expression of RAD9B ( p-value<1×10−7 ) , a novel component of the 9-1-1 cell-cycle checkpoint response complex [52] , while rs3743175 is significantly associated with the expression of SCAPER , a novel regulator of cell cycle progression [53] . We also find that the DNA repair gene ERCC6 , which plays roles in the aging process and predisposes to AMD [54] , is significantly ( p-value<1×10−4 ) co-expressed with SGCD and SCAPER across more than 40 human tissues [55] . Co-expression analysis also finds that both SGCD and SCAPER are significantly correlated with MASP1 ( p-value<1×10−2 ) and MASP2 ( p-value<1×10−5 ) , activators of the complement pathway . Together with the report of synergic effect between ERCC6 and CFH in predisposing AMD [54] , the above analyses suggest clues for the link between the aging component and the immune component in the etiology of AMD . The third SNP rs2828155 locates in an about 4 Mb intergenic region in chromosome 21q21 . 1 , a region has also been implicated in AMD [35] . ADAMTS1 and ADAMTS5 lie about 4 Mb downstream of the SNP . There is possibility that rs2828155 may regulate the expression of these two enzymes , and then the epistasis between rs2828155 and rs1394608 is more straightforward: rs2828155 regulates the enzyme ADAMTS1 and rs1394608 regulates FBLN1 . As FBLN1 can act as a cofactor of ADAMTS1 and plays an important role in the degradation of proteoglycans by ADAMTS1 during pathological conditions induced by inflammatory processes [46] , it is possible that rs2828155 and rs1394608 have epistatic effect in AMD . Linkage signals for AMD from the two loci have been detected in the same linkage scan [35] . In summary , our association study suggests the existence of epistasis in AMD , while the functional analysis provides new insights for the understanding of the epistasis from the biological point of view . Certainly , further work , especially experimental verification of the above epistasis , is necessary in order to confirm the roles of the identified SNPs and their epistasis in AMD . We further apply our approach to a genome-wide case-control data set of Parkinson's disease [21] , [22] , which contains 408 , 803 SNPs genotyped with 270 cases and 271 controls . With the use of epiMODE , we identify 12 independent contributing markers with posterior probabilities of associations greater than or equal to 0 . 9 . The p-values for these markers , obtained by Chi-squared tests with two degrees of freedom , are shown in Table 3 , which suggest that 7 out of the 12 markers are statistically significantly . The original paper [21] only tests SNPs that give successful genotypes in more than 95% samples . As a result , the significant markers identified by our method are all excluded . In our analysis , we run our method on the original data without discarding any SNP . The fact that no interaction effect is detected may be partly due to the disease model itself , which may have no strong interaction effects . Another reason may be the missing genotype problem that aggravates the insufficiency of the sample size in mapping epistatic effects . In the detection of a k-locus interaction , if the genotype missing rate is ( ) for each locus , the expected percentage of samples that could be used is only , which decreases fast with k , the number of loci in the interaction , and makes the power for detecting high-order interactions even lower . In this paper , we explicitly define epistatic modules as basic genetic units that influence the disease susceptibility and put forward a Bayesian marker partition model to explain the observed case-control data . We develop a Gibbs sampling strategy to simulate the posterior distributions that markers belong to epistatic modules and further resort to hypothesis testing to screen out statistically significant modules . We extensively assess the effectiveness of the proposed epiMODE approach . In simulation studies , epiMODE significantly outperforms all other methods . In the application to the Age-related Macular Degeneration ( AMD ) data , epiMODE successfully identifies two loci that are known to be associated with the disease , and suggests the epistatic interaction of two other loci . In the application to the Parkinson's disease data , epiMODE identifies seven loci that might contribute to the disease susceptibility . The success of the proposed approach can be attributed to a combination of several aspects . First , with the explicit definition of epistatic modules , our approach is able to capture patterns of epistatic interactive effects of multiple loci in a native way . Second , the incorporation of LD information between markers in epistatic modules greatly improves the power of our method in the detection of indirect subtle associations . Third , the introduction of LD modules further minimizes the possibility of assigning disease-unassociated loci into epistatic modules . Fourth , the Gibbs sampling strategy is effective in obtaining the posterior distributions that disease-associated loci belong to epistatic modules . Finally , the native identification of interactive effects of multiple loci ( epistatic modules ) instead of enumerating combinations of SNPs makes our approach capable of and suitable for dealing with large scale case-control data . Our marker partition model is proposed from the Bayesian perspective , and thus a natural advantage of this model is its capability of incorporating prior biological knowledge about individual SNP markers , such as their locations ( e . g . , coding region , promoter region , etc . ) , genotype frequencies , and LD information . Nevertheless , some parts in our approach are not formulated from the pure Bayesian perspective , mainly for the consideration of reducing the computational burden . Structures of epistatic modules are represented by results of searching procedures ( exhaustive/sampling/greedy ) rather than averaged over all possible structures according to their posterior probabilities because this part is heavily used by the up-level Gibbs sampling algorithm . The standard Chi-squared test with Bonferroni correction is also much more computationally economy than the permutation test method . Such efforts for greatly reducing the computational burden are necessary in handling large scale case-control data sets , which may contain more than 500 , 000 SNPs and have been very common in recent genome-wide association studies . Certainly , the proposed approach can further be improved from the following directions . First , the assumption of equal values for all Dirichlet priors in the Bayesian marker partition model is obviously for the purpose of seeking for simplicity . Although different priors do not yield very different results , careful selection of priors is still worth investigating . One possibility is to systematically minimize the impact of priors using techniques such as prior annealing [56] . Another possibility is to select priors that reflect existing biological knowledge , such as the rich genotype frequency and LD information from the International HapMap Project [57] , [58] . Second , although our experience suggests that the current Gibbs sampling strategy works well , some sophisticated sampling strategy such as the “split-merge” algorithm [59] might be incorporated to further improve the efficiency of the sampling strategy . Finally , currently we do not formulate the marker partitioning procedure from the viewpoint of mixture models , such as the Dirichlet process ( DP ) mixture model . Since a DP mixture model assumes an infinite number of mixture components and thus provides more flexibility in controlling the complexity of the model [60] , [61] , it would be interesting to explore the possibility of incorporating the DP mixture model into genome-wide association studies .	Although genome-wide association studies ( GWAS ) have been quite popular due to recent advances in low-cost genotyping techniques , most of the reported studies only analyze single-locus effects because traditional multi-locus methods are not computationally practical in the detection of epistatic interactive effects of multiple loci . Here , on the basis of a rigorous definition of epistatic modules that describe interactive effects of multiple loci , we take advantage of a Bayesian model with a properly designed Gibbs sampling strategy to facilitate the detection of such modules . We confirm via extensive simulation studies that the proposed method , named epiMODE , is not only feasible in detecting multi-locus effects but also more powerful than three representative methods on seven disease models . We apply the proposed method to an Age-related Macular Degeneration ( AMD ) data and discover that a combination of two loci—one in the gene SGCD and the other in SCAPER—might be associated with AMD . Considering its advantages , we suggest that the proposed method be applied to more GWAS data for the detection of multi-locus interactive effects .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "computational", "biology/population", "genetics", "genetics", "and", "genomics/complex", "traits", "genetics", "and", "genomics/disease", "models", "genetics", "and", "genomics/genetics", "of", "disease", "mathematics/statistics", "genetics", "and", "genomics/population", "g...	2009	Epistatic Module Detection for Case-Control Studies: A Bayesian Model with a Gibbs Sampling Strategy
Systemic sclerosis ( SSc ) is a rare systemic autoimmune disease characterized by skin and organ fibrosis . The pathogenesis of SSc and its progression are poorly understood . The SSc intrinsic gene expression subsets ( inflammatory , fibroproliferative , normal-like , and limited ) are observed in multiple clinical cohorts of patients with SSc . Analysis of longitudinal skin biopsies suggests that a patient's subset assignment is stable over 6–12 months . Genetically , SSc is multi-factorial with many genetic risk loci for SSc generally and for specific clinical manifestations . Here we identify the genes consistently associated with the intrinsic subsets across three independent cohorts , show the relationship between these genes using a gene-gene interaction network , and place the genetic risk loci in the context of the intrinsic subsets . To identify gene expression modules common to three independent datasets from three different clinical centers , we developed a consensus clustering procedure based on mutual information of partitions , an information theory concept , and performed a meta-analysis of these genome-wide gene expression datasets . We created a gene-gene interaction network of the conserved molecular features across the intrinsic subsets and analyzed their connections with SSc-associated genetic polymorphisms . The network is composed of distinct , but interconnected , components related to interferon activation , M2 macrophages , adaptive immunity , extracellular matrix remodeling , and cell proliferation . The network shows extensive connections between the inflammatory- and fibroproliferative-specific genes . The network also shows connections between these subset-specific genes and 30 SSc-associated polymorphic genes including STAT4 , BLK , IRF7 , NOTCH4 , PLAUR , CSK , IRAK1 , and several human leukocyte antigen ( HLA ) genes . Our analyses suggest that the gene expression changes underlying the SSc subsets may be long-lived , but mechanistically interconnected and related to a patients underlying genetic risk . Genome-scale gene expression profiling of systemic sclerosis ( SSc ) skin has identified distinct intrinsic molecular subsets ( inflammatory , fibroproliferative , and normal-like ) within the subset of patients diagnosed with diffuse cutaneous SSc ( dSSc ) based upon the extent of skin involvement . These subsets are identified by an intrinsic gene analysis [1] that shifts the focus to differences between patients rather than patient biopsies . The inflammatory subset is characterized by increased expression of genes associated with inflammation and extracellular matrix ( ECM ) deposition , while the fibroproliferative subset is characterized by increased expression of genes associated with cell proliferation [1] , [2] . Biopsies from patients in the normal-like subset show gene expression most similar to healthy control skin biopsies . The presence of three distinct molecular SSc subsets within patients diagnosed with dSSc underscores the molecular heterogeneity of SSc . However , it is unclear whether the subsets represent distinct diseases with different etiologies or whether they represent disease progression . To address this question , we identified the conserved molecular pathways characteristic of each subset that are reproducible between different datasets from multiple patient cohorts , and examined the connectivity of these genes and SSc-associated polymorphic genes in a predicted functional network . SSc is a rare disease without validated disease progression markers and no known cure . SSc affects between 49 , 000–276 , 000 Americans; one in three patients dies within 10 years of diagnosis [3] . The rarity of SSc makes this disease an excellent test case for a genomic meta-analysis to understand disease mechanism . Using this approach , we have begun to understand the molecular and clinical complexity of SSc . Our findings may assist in generating patient-specific therapies [4] and delivering real-time quantitative feedback regarding therapeutic response during clinical trials [4] , [5] . High-throughput gene expression data has demonstrated that genes that function together are almost always co-expressed and thus highly correlated with each other [6] , [7] . Indeed , the expressed genes in a biopsy form a co-expression network , where genes serve as nodes and correlations as links between genes . ( The language of networks is technical and beyond the scope of this paper . The supporting information ( S1 Text ) contains a glossary of keywords that are used in a technical sense in the main body of the paper . ) This observation forms part of the basis for “network medicine” [8] , [9] . The co-expression network contains groups of highly correlated genes that represent the biological processes at work in the tissue [6] , [10] . These groups of genes can be found using a variety of procedures for co-expression clustering ( e . g . [6] , [10] ) . All of these procedures group highly correlated genes together , i . e . they partition the genome into non-overlapping groups of genes with similar expression patterns sometimes called modules . The output of co-expression clustering is a data-driven partition of the expressed genes in the genome . As a result of technical differences in data acquisition protocols as well as true biological variation ( patient heterogeneity , treatment ) , the exact modules identified in one SSc dataset differ somewhat from those identified in another dataset , although the same pathways , biological processes , and many core genes are found in each dataset . We developed a tool to compare gene co-expression modules derived from multiple disparate datasets to identify the modules that are reproducibly expressed in each dataset . We then derived gene sets from the overlaps between conserved modules across datasets . These “consensus clusters” are the gene clusters that are conserved across all datasets . A naive approach to solving this problem is to simply intersect the “intrinsic gene lists” derived for each of the cohorts [1] , [4] , [11] . The methodological issue with this approach is that these lists are derived under a large multiple hypothesis testing burden , and although the same biological processes and some genes are found reproducibly , the gene sets do not exactly recapitulate across data sets [11] . This simple intersection approach would be much too conservative and consequently exclude many biologically important genes . Our alternative approach is to consider “modules first” . In brief , our goal is to identify the modules that are conserved across datasets first and then extract the consensus genes as those that are consistently assigned to those modules . This transfers the multiple testing burden onto the much smaller list of modules and allows genes to be included in the consensus even if they do not achieve extremely high statistical significance in all datasets simultaneously . We developed this idea into a novel data mining procedure called Mutual Information Consensus Clustering ( MICC ) to identify conserved gene expression modules across multiple gene expression datasets . Consensus clustering is a set of techniques from computer science and bioinformatics that refers to strategies for extracting robust clusters from an ensemble of partitions . Typically this is done using a large ensemble of partitions . Early work focused on weak clustering algorithms and consensus clustering was used to “boost” the weak partitions into an aggregate , consensus partition [12] . In bioinformatics , consensus clustering algorithms have been developed to aggregate ensembles of partitions that are derived from data resampling [13] . These techniques have in common that they do not “trust” a particular partition from one of their clustering algorithms . Here , we use a strong clustering algorithm called weighted gene co-expression network analysis ( WGCNA ) . Our ensemble of partitions is the collection that we obtain from having multiple , clustered datasets from independent cohorts . While WGCNA extracts meaningful signals in each data set , the potentially interesting modules in one dataset are not precisely replicated in all others . Mutual information [14] provides a rigorous criterion by which modules from different datasets can be said to have significant overlap ( i . e . are conserved ) and allows one to identify when the available information between two partitions is exhausted . Mutual information is a sum of positive and negative contributions from each pair of modules across datasets , and MICC automatically disregards all overlaps that do not contribute positively to the total mutual information , thus giving an objective measure of conserved gene expression that is both comprehensive and parsimonious . As such , MICC is a metaclustering procedure that “clusters the clusters” [12] , but does not produce a complete partition . Instead , it generates only a partition of the subset of the genome that has strongly conserved gene co-expression . Using previously published gene expression data from skin biopsies from patients with SSc recruited at three independent academic centers [1] , [4] , [11] and new samples analyzed as part of this study ( Table 1 ) , we identified the consensus clusters that were present in all datasets . Due to the unbiased nature of high-throughput screening , these datasets contain information about SSc-specific biology as well as the general biology of skin . We showed that MICC yields consensus clusters that are biologically specific . At the level of the whole transcriptome , we demonstrated that the consensus clusters are enriched for hubs defined by co-expression network analysis . We then filtered the consensus clusters down to those that were intrinsic subset-specific . The existence of the intrinsic subsets is a robust observation in each of these studies and the consensus clusters associated with them provide a rigorous picture of the core gene set underlying the subsets . The comprehensive and concise annotation of the conserved differential gene expression that we developed suggests that the intrinsic subsets represent pathophysiological states of one disease . Our major findings include the following: 1 . We show that the subset-specific consensus clusters are part of a gene-gene network and for the first time to our knowledge , demonstrate putative connections between the intrinsic gene expression subsets of SSc and SSc-associated genetic polymorphisms identified by candidate and genome-wide association studies ( GWAS ) . 2 . We provide additional unbiased data to support the hypothesis that immune system activation is an early event and plays a central role in SSc pathogenesis as SSc risk alleles are linked to the immune system nodes of our network . 3 . The consensus gene-gene network provides insights into genes that may be central to the major disease processes and identifies genes and pathways that may connect these major groups of genes . 4 . We show a link between the inflammatory and fibroproliferative patient groups through a shared TGFβ/ECM subnetwork , suggesting a theoretical path by which these gene expression subsets may be linked . Collectively , these findings demonstrate that MICC is a powerful tool that identifies the reproducible signals in gene expression data across multiple datasets and shows how they may relate to the genetic polymorphisms associated with SSc . Weighted gene coexpression network analysis ( WGCNA ) [10] is a gene co-expression clustering procedure that automatically detects the number of modules in a dataset and removes outlier genes . WGCNA performed on a single SSc skin dataset ( Milano et al . [1] , S1 Data file ) , demonstrates the complexity of comparative studies across multiple datasets ( Fig . 2 ) . The molecular subsets termed ‘SSc intrinsic subsets’ were first identified by Milano et al . [1] . The Milano et al . dataset is the best characterized dataset showing the SSc intrinsic subsets that has been analyzed to date . These data were clustered into modules by WGCNA , and the resulting modules were summarized by their first principal component or module eigengene ( Fig . 2 ) . Module eigengenes are a one-dimensional summary of the gene expression within a module that captures the bulk of the variance within that module . To identify those modules that were intrinsic subset-specific , we performed Kruskal-Wallis tests on the module eigengenes with groups defined according to intrinsic subsets . Of the 54 total modules , 23 had a significant p-value for association with the intrinsic subsets ( all p<0 . 05 after Bonferroni correction; Fig . 2 shows six representative examples ) . These 23 modules comprise approximately 40% of the expressed genes in the genome ( Table 2 ) , demonstrating that the intrinsic subsets found in SSc skin are defined by deregulation of a very large fraction of the expressed genes in any given cell . Gene expression for six significant modules along with their corresponding module eigengenes demonstrates clear association with the intrinsic subsets ( Fig . 2 ) . These modules are enriched for broad functional categories previously associated with SSc , including chemokine signaling , NFκB signaling , RAS-RAC signaling in the inflammatory subset , and cell cycle processes in the fibroproliferative subset [1] , [4] , [11] . To identify the core set of genes reproducibly found in each SSc intrinsic subset , we performed WGCNA on two additional SSc skin gene expression datasets: Pendergrass et al . [11] and an expanded version of Hinchcliff et al . [4] ( Data files S2–S3 ) . The Hinchcliff data are from an ongoing clinical trial of mycophenolate mofetil ( MMF ) . Preliminary data have been published [4] , but we also analyzed data for an additional 82 unpublished samples from that trial ( the full data available from NCBI GEO at GSE59787 ) . A summary of the three cohorts , including the expanded Hinchcliff cohort , is available in Table 1 . Each of the three datasets had approximately 60 modules . The module eigengenes were tested for association with the intrinsic subsets , and it was found that , within a dataset , between 18% and 67% of the modules were subset-specific ( Table 2 ) . This shows that for each dataset a substantial fraction of modules was associated with the intrinsic subsets and , that 4 , 004–10 , 373 genes out of the approximately 19 , 500 in the human genome were in differentially regulated modules associated with the intrinsic subsets ( Tables 2 and 3 ) . The gene co-expression modules represent biological processes that are active in skin and some are reflective of disease pathogenesis . To determine which processes were conserved across all three datasets , we constructed the information graph for the three separate WGCNA partitions of the genome ( Fig . 3A ) . The information graph is a network where a node in the network is a module from one dataset , and a link between modules indicates that the overlap between those modules is significantly larger than would be expected at random . In other words , an edge represents conservation of a significant part of a module across two datasets ( see Materials and Methods for a detailed discussion of module overlap scores ) . Triangles in the information graph correspond to a significant three-way overlap of modules or , equivalently , a module conserved across all three datasets . We enumerated all triangles in the information graph to identify all such conserved modules . There were 157 triangles and approximately 2000 genes in their corresponding triple overlaps . Most ( 129 out of 178 ) of the modules across all SSc datasets are present in at least one triangle , i . e . most co-expression modules had a significant portion co-expressed in the other datasets ( Table 2 , bottom row ) . This indicates that the WGCNA-derived modules are reproducible features of SSc gene expression . Nine of the triangles had all three nodes ( modules ) significantly associated with the subsets ( five inflammatory , four fibroproliferative; see below and Fig . 3 ) . The consensus genes are hubs in the gene-gene co-expression networks . To see this , we noted that module eigengenes represent hubs in the gene-gene correlation network [15] . A module eigengene does not correspond to an actual gene , but rather represents a theoretical gene that is most central in the module . Therefore , genes that are highly correlated to their module eigengene are more central within their module . We calculated the correlation of each gene to its corresponding module eigengene ( S1 Fig . ) . The density of these gene-eigengene correlations is shown for all genes in the genome ( blue curve ) and for only the consensus genes ( red curve ) . The consensus genes are significantly more correlated with their module eigengene than randomly selected genes are with their module eigengene , indicating that the consensus genes are significantly enriched for hub genes in their ( dataset-specific ) co-expression network . This is a useful positive control for the MICC method because it shows that the consensus genes are enriched for “hubness” in the SSc co-expression network and thus MICC finds genes that have salient network features . While most modules are partially conserved between the three datasets and many of them are intrinsic subset-specific , not all intrinsic subset-specific modules are conserved across all datasets ( S4 Data file ) . To find the conserved , intrinsic subset-specific modules , we noted that the information graph has groups of triangles with considerable mutual edge-sharing ( Fig . 3A ) . Many of the triangles in the information graph overlap and form communities of triangles ( Fig . 3A , S2 Fig . ) . This was intriguing because it opened up a broader interpretation of “consensus cluster” . If the information graph had been a disconnected collection of single triangles , this would have implied that there was a one-to-one mapping between the modules from different datasets . Instead , a single module from one dataset gets broken into pieces in the other datasets . The community structure of the information graph indicates what we have known from many prior microarray studies , namely that specific groups of genes are commonly expressed together and that the aggregate set of genes underlying these multiple co-expression clusters constitutes the truly conserved processes in SSc [1] , [6] , [16] . We detected communities in the information graph using a variant of clique percolation [17] , a network community detection procedure that , in this case , explicitly identifies communities of triangles ( S1 Text ) . Clique percolation identified 26 communities , 13 of which were single , isolated triangles , while the rest were groups of more than one triangle ( Fig . 3A , S2 Fig . ) . To derive a gene set associated with a community in the information graph , we took all modules within the community , computed their union within datasets , and computed their intersection across datasets ( S3 Fig . ) . In this way , we captured all genes whose co-expression was conserved across the three datasets . ( A mathematical description of this procedure is presented in the Materials and Methods . ) We termed these community-derived gene sets consensus clusters ( CCs ) . Using g:Profiler [18] , we found that the consensus clusters are enriched for many biological processes ( summary in Table 3; raw data in S5 Data file ) present in both healthy and SSc biopsies . For example , CCs 1 , 4 , 5 , 7 , 8 , and 11 are enriched for basic metabolic and cellular processes , while CC 12 showed enrichment for keratinocyte-specific processes ( Table 3; Fig . 3A , cyan ) . These consensus clusters show that MICC extracts biologically coherent sets of genes that are known to be active in skin as consensus clusters . This provides an additional positive control for the MICC method . More importantly , CC3 and CC9 showed enrichment for processes implicated in SSc ( Table 3; S5 Data file ) . CC3 was enriched for response to interferons , B cell receptor signaling , monocyte chemotaxis , and TGFβ and PDGF signaling , as well as ECM remodeling processes . CC9 showed enrichment for cell cycle and cell proliferation processes , as well as integrin interactions with fibrin . Note that CC3 and CC9 both show enrichment for distinct ECM-related molecular processes . These data are consistent with the analysis of experimentally derived pathway signatures [19] . The consensus clusters CC3 and CC9 map to the major intrinsic subsets previously described [1] . We tested every module for association with the intrinsic subsets ( see Materials and Methods ) and we constructed a “heatmap” of the triangles in the information graph by dataset ( Fig . 3B ) . The rows were ordered by community membership and the columns were ordered by dataset . We concatenated each of these plots so all subsets , datasets , and consensus clusters can be viewed simultaneously . Only consensus clusters 3 and 9 were enriched for SSc intrinsic subset specificity . Consensus cluster 3 contained modules that are almost all significantly expressed at high levels in the inflammatory group of patients ( Fig . 3A , purple nodes; Fig . 3B ) . Consensus cluster 9 contains modules that are almost all significantly expressed at high levels in the fibroproliferative group ( Fig . 3A , red nodes; Fig . 3B ) . We also included tests for all SSc biopsies versus healthy controls to determine if there were any consensus clusters that were generally conserved across all SSc biopsies . There were no consensus clusters that were enriched for all SSc versus healthy controls , which illustrates quantitatively SSc heterogeneity . Furthermore , there were no consensus clusters that were consistently expressed at low levels in any of the subsets . Some consensus clusters are enriched for a subset in some of the datasets , but are not replicated across all datasets ( Fig . 3B ) . For example , CC2 is expressed at high levels in the inflammatory subset and low levels in the proliferative subset in Milano , but neither of the other datasets . Inflammatory-specific CC3 is expressed at low levels in the proliferation subset in Milano and in the normal-like subset in Pendergrass and Hinchcliff , and is expressed at high levels in all SSc versus healthy controls in Pendergrass and Hinchcliff only . Similarly , CC9 , which is proliferative-specific , is expressed at low levels in the inflammatory subset in Milano only . These observations demonstrate that genes with increased expression should be the focus in SSc . The biology of CC3 and CC9 show the processes common to the intrinsic subsets that have been observed across multiple gene expression datasets: inflammation , cell interactions with ECM , and cell proliferation ( Table 3 ) . To determine if there was a more interconnected relationship between these conserved processes ( such as genes related to specific cell types ) than could be gained from an ontological annotation analysis like g:Profiler , we used CC3 and CC9 as a query gene set for the IMP gene-gene interaction Bayesian network ( IMP ) ( Fig . 4 ) [20] . IMP is a gene-gene interaction network developed using a large compendium of high-throughput biological data including all publicly available microarray data that predicts the probability that pairs of genes have a co-expression interaction . A list of genes is imported into IMP , and a list of high-probability interactions between the genes on the imported list and ( up to 50 additional genes in ) the rest of the genome is generated . IMP is completely agnostic to SSc-specific biology and reports predicted interactions that are based on the preponderance of evidence across all publicly available gene expression data . As our query , we pooled the two consensus clusters CC3 and CC9 to discover possible molecular links between the inflammatory and fibroproliferative intrinsic subsets . We added polymorphic genes from genome-wide association studies ( GWAS ) , as well as genes from candidate gene studies that have been replicated in at least one follow up study ( see Materials and Methods; S6 Data file ) . In addition , we added four genes that are putative predictors of Modified Rodnan Skin Score ( MRSS ) , a widely used clinical measure of skin fibrosis [5] . The output network from IMP was dominated by one large interconnected network that had five distinct subnetworks ( Fig . 4; S7–S9 Data files ) . The five molecular subnetworks were each enriched for a distinct biological process: interferon response , M2 macrophage activation , adaptive immunity , ECM deposition and remodeling and TGFβ signaling , and cell proliferation . One subnetwork was dominated by interferons and interferon-inducible genes ( Fig . 4 , top middle; S9 Data file ) . The interferon subnetwork contained genes solely from the inflammatory consensus cluster ( Fig . 4 , purple nodes ) . This subnetwork contained the interferon inducible genes IFI16 and IFI44 , the latter of which is a putative biomarker of fibrosis [5] . This subnetwork also contains the polymorphic interferon regulatory factor genes IRF5 , IRF7 , and IRF8 . A second subnetwork contained genes characteristic of M2 macrophage activation ( Fig . 4 , bottom left; S9 Data file ) . The genes in this network , which include major histocompatibility complex ( MHC ) class II genes with SSc-associated polymorphisms , are derived primarily from the inflammatory consensus cluster , implicating macrophages as mediators of inflammation . Polarized macrophages can broadly be categorized as “classically activated” ( M1 ) or “alternatively activated” ( M2 ) , although it is important to recognize that macrophage polarization encompasses a broad spectrum of activation states . M1 macrophages may be elicited through stimulation with IFN-γ and LPS , are microbicidal , and promote Th1-mediated immune responses . In contrast , M2 cells , which mediate immune suppression , may be activated by various stimuli , including IL-4 and/or IL-13 , which are elevated in SSc sera [21] , [22] . Genes associated with M2 activation , including CX3CR1 [23] , IL10R [24] , and HLA-DMB [25] , were consistently expressed in this subnetwork , in accordance with previous studies that found increased M2-polarized macrophages in SSc skin compared to healthy skin [26] . As M2-polarized cells regulate vascularization and are a potent source of TGFβ , PDGF , and inflammatory cytokines [27]–[29] , activated M2 macrophages may play a role in mediating fibrosis and inflammation in SSc . A third molecular subnetwork contained genes related to adaptive immunity ( Fig . 4 , top left; S9 Data file ) . There are relationships to both B and T cells in the genes in this subnetwork . Two chains of the T cell receptor complex are represented: CD3G ( gamma chain of the T cell receptor ( CD3 ) ) and CD247 ( the zeta chain of the T cell receptor ) , which contains SSc-associated polymorphisms . The IL-12 pathway , which mediates Th1 cell differentiation and activation [30] , [31] , is represented through IL12RB2 . Binding of IL-12 to IL12RB2 on activated T cells initiates a signal transduction cascade that results in activation of STAT transcription factors , including STAT4 [32] ( also represented in this subnetwork ) , which regulate T cell signaling and immune activation [33] . Aberrant expression of IL12RB2 has been reported in autoimmune and infectious diseases [34] , [35] , implicating this gene as an important regulator of inflammation and immune defense . B cell receptor activation and signaling are also represented in this subnetwork . DOCK10 expression is up-regulated in B cells by pro-inflammatory IL-4 [36] , and BANK1 and BLK are B cell proteins that have polymorphisms associated with SSc . Both LYN and CSK appear in this subnetwork and are directly connected to each other . The tyrosine kinase LYN , which plays a critical role in down-regulating B cell activation and mediating self-tolerance [37] , [38] , is phosphorylated by CSK [39] . Polymorphisms in CSK have been linked to both SSc and systemic lupus erythematosus ( SLE ) and are associated with aberrant B cell signaling [40] . CSK also associates with Lyp [41] , which is the product of the tyrosine phosphatase PTPN22 . The PTPN22 gene also contains an SSc-associated polymorphism . Mutations in PTPN22 that interfere with its ability to bind to CSK also interfere with both B and T cell receptor activation [42] , [43] . Moreover , mutations in PTPN22 have been reported in a variety of other autoimmune diseases , including SLE , rheumatoid arthritis , and type 1 diabetes [44] . Negative regulators of B and T cell activation such as SOCS2 and SOCS3 , are included in this network . SOCS3 has been shown to directly inhibit IL-12-induced STAT4 activation [45] . The co-occurrence of pro- and anti-inflammatory signals in this subnetwork is notable and is likely because our data are derived from whole skin biopsies ( see Discussion ) . The fourth molecular subnetwork contained TGFβ pathway genes ( which have long been implicated in the activation of fibrosis in SSc [46] , [47] ) and ECM structural proteins ( Fig . 4 , bottom middle ) . This TGFβ/ECM subnetwork contained genes from both the inflammatory and fibroproliferative consensus clusters ( Fig . 4 , red and purple nodes; S9 Data file ) . We also found expression of genes associated with Notch signaling such as NOTCH4 , which contains SSc-associated polymorphisms , and with the epithelial-mesenchymal transition ( EMT ) such as LATS2 . Alternatively activated macrophages are known to produce large quantities of TGFβ in SSc pulmonary fibrosis [29] , suggesting that the M2 macrophage subnetwork could drive activation of the TGFβ/ECM subnetwork . The final molecular subnetwork contained cell cycle/cell proliferation genes , which were primarily from the fibroproliferative consensus cluster ( Fig . 4 , right ) . The expression of proliferation genes is commonly observed in cancer [2] , [7] , [48] and their presence in the gene expression data of SSc was a surprising and unexpected finding [1] . The large and densely interconnected subnetwork of genes in Fig . 4 ( right , red nodes ) was composed almost exclusively of cell cycle-regulated genes including AURKA/B , CCNA2 , CCNB1 , CHK1 , and DHFR [7] . This subnetwork was conserved and showed increased expression in the fibroproliferative subset of patients across all three cohorts , and constituted the core gene expression signature in that subset of patients ( Fig . 4 , red nodes ) . Therefore , the cell proliferation signature of the fibroproliferative subset of patients first observed in Milano et al . [1] is a conserved feature of SSc across three independent cohorts from three separate clinical centers . This molecular subnetwork has connections to each of the other four subnetworks ( interferon , M2 macrophages , adaptive immunity , and TGFβ/ECM ) suggesting that cell proliferation in SSc skin is modulated by the inflammatory and ECM remodeling processes in skin . IMP predicts that the genes linked to SSc-associated polymorphisms ( 30/41 total ) and the putative MRSS biomarker genes of Lafyatis and co-workers ( 4/4 ) have interactions within this large component of the molecular network ( Fig . 4 ) . Polymorphisms in IRF5 , IRF7 , and IRF8 were linked to the interferon subnetwork . IRF7 is also differentially expressed in the inflammatory subset . The polymorphisms associated with human leukocyte antigen ( HLA ) alleles predominantly have interactions with the M2 macrophage subnetwork of genes . Polymorphisms in and differential expression of NOTCH4 were linked to the TGFβ/ECM subnetwork . The same was true for the MRSS biomarker genes; IFI44 was linked to the interferon subnetwork; SIGLEC1 was linked to the M2 macrophage subnetwork; and both COMP and THBS1 were linked to the TGFβ/ECM subnetwork . These results suggest that prediction of worsening skin disease requires sampling genes from each molecular subnetwork . The molecular network contains genes that are hubs ( i . e . highly connected nodes ) of the subnetworks . Interferon-induced protein 44 ( IFI44 ) is a hub of the interferon subnetwork . It has conserved high expression across all three of our cohorts in the inflammatory subset and is one of the most highly connected genes in the interferon subnetwork ( Fig . 5 , top right ) . IFI44 is predicted to have co-expression interactions with several other interferon-inducible and interferon-regulating genes , including IFI16 , IRF7 , IFITM2 , ISG20 , GBP1 , and TRIM22 . Allograft Inflammatory Factor 1 ( AIF-1 ) is a hub of the M2 macrophage . AIF1 is consistently highly expressed in the inflammatory subset across all three SSc skin cohorts and is one of the most highly connected genes in the M2 macrophage subnetwork ( Fig . 5 , bottom left ) . In the molecular network ( Fig . 5 , bottom left ) , AIF1 has many links including: ITGB2 , a binding partner of the monocyte marker ITGAL , and MHC class II genes HLA-DMB , HLA-DPA1 , and HLA-DQB1 . In addition , AIF1 has connections to chemokine receptors CCR1 and CX3CR1 , which are connected to chemokines CX3CL1 ( fractalkine ) and CCL2 ( MCP-1 ) . The tyrosine kinase gene LYN is a hub of the adaptive immunity subnetwork ( Fig . 5 , top left ) . LYN has predicted edges with four polymorphic genes in this subnetwork: BLK , BANK1 , CSK , and GRB10 . LYN also has connections to the polymorphic , bridge genes PLAUR and LCP2 ( see below ) , and suppressors of cytokine signaling genes SOCS2 and SOCS3 . The conserved finding of high expression of LYN in the inflammatory subset and its centrality within the adaptive immune subnetwork suggests that LYN plays a key role in the adaptive immune component of SSc in skin . Fibrillin-1 ( FBN1 ) is a hub of the TGFβ/ECM subnetwork ( Fig . 5 , bottom right ) . High expression of FBN1 is conserved across the inflammatory subset of all three cohorts of SSc skin , and FBN1 is highly connected within the TGFβ/ECM subnetwork of the molecular network ( Fig . 5 , bottom right ) . The TGFβ/ECM subnetwork includes genes that primarily show high expression in the inflammatory subset but also includes genes that are highly expressed in the fibroproliferative group , thus providing a putative molecular link between the two groups . FBN1 has predicted connections to many genes whose increased expression is conserved , including: pro-fibrotic genes including COL1A2 , COL5A2 , and elastin ( ELN ) , CTGF , SPARC , THBS1 , THBS4 , COMP , TNC and ECM remodeling and wound response genes LOX , NNMT , and FBLN5 . In addition , FBN1 has connections with growth factor genes and receptors such as HTRA1 and NOTCH4; cell adhesion genes CDH11 and LAMA4; as well as the complement system gene C1S . In addition to containing discrete subnetworks , the molecular network also shows genes that bridge the subnetworks ( Fig . 6 ) . These genes are of particular interest because they have predicted connections between multiple , distinct subnetworks . The primary reason for using CC3 and CC9 simultaneously as queries to IMP was to identify possible molecular connections between the core molecular processes of the inflammatory and fibroproliferative intrinsic subsets . The bridge genes live at the interfaces between the subnetworks that constitute these core molecular processes . The genes CXCR4 and LCP2 are the major connections between the adaptive immunity subnetwork and the M2 macrophage subnetwork ( Fig . 6 ) . LCP2 ( SLP-76 ) , which modulates T cell activation [49] , has predicted interactions with AIF1 and IL10RA in the M2 macrophage subnetwork and to SOCS2 , SOCS3 , STAT4 , LYN , and CSK in the adaptive immunity subnetwork ( see edges extending from LCP2 in Fig . 6 ) . The chemokine CXCR4 has predicted interactions with the cytokines/chemokines IL10RA , CX3CR1 , CCR1 , and the polymorphic CCR6 in the M2 macrophage subnetwork ( Fig . 6 ) . CXCR4 has predicted interactions with SOCS3 and JAK3 in the adaptive immunity subnetwork . GRB10 contains an SSc-associated polymorphism and is also expressed at high levels in the inflammatory subset ( see blue GRB10 node , Fig . 6 ) . GRB10 is part of a complex path from the adaptive immune subnetwork to the M2 macrophage subnetwork that includes genes containing pleckstrin homology domains including PLEKHO1 , PLEKHO2 , CYTH4 and ADAP2 . The major connection between the M2 macrophage subnetwork hub AIF1 and the interferon subnetwork hub IFI44 is through RAC2 . RAC2 encodes a member of the Rac family of signaling molecules and has multiple predicted interactions with both the interferon subnetwork and the M2 macrophage subnetwork ( Fig . 6 ) . In the interferon subnetwork ( Fig . 6 , upper middle ) , RAC2 connects to CTSC ( cathepsin C ) , IFITM1 and IFI16 , as well as the Rho GTPase related genes ARHGDIB and RAB31 . In the M2 macrophage subnetwork ( Fig . 6 , lower left ) , RAC2 connects to ITGB2 , the actin cytoskeleton related proteins LCP1 and COTL1 , and GMFG . COTL1 is also related to leukotriene biosynthesis through a known interaction with ALOX5 . These diverse interactions suggest that RAC2 is involved simultaneously in macrophage motility , leukotriene biosynthesis , and interferon signaling . The major bridges between the M2 macrophage subnetwork and the ECM subnetwork are THY1 ( CD90 ) and CD14 ( Fig . 6 , lower left ) . THY1 connects to SIGLEC1 , MXRA5 and COL1A2 . THY1 mediates adhesion of leukocytes and monocytes to endothelial cells and fibroblasts [50] , may also have a role in lung fibrosis ( a major complication of SSc ) ; THY1 knockout mice have increased lung fibrosis [51] , [52] . CD14 is a cell surface protein mainly expressed by macrophages , is inducible by and connected to AIF1 [53] . It also has connections to the polymorphic genes TLR2 and HLA-DRA ( Fig . 6 , lower left ) . PLAUR ( UPAR ) contains a putative SSc-associated polymorphism , is a member of the interferon subnetwork , and has numerous links with the ECM , M2 macrophage , and the adaptive immunity subnetworks ( Fig . 6 ) . PLAUR encodes the plasminogen activator , urokinase receptor protein and is a pleiotropic gene at the interface of ECM remodeling , as a component of the fibrinolysis system , and in both adaptive and innate immune processes , including monocyte migration [54] . PLAUR is inducible by proinflammatory cytokines IL1β and TNFα . PLAUR connects to the tyrosine kinase LYN , the hub gene of the adaptive immunity subnetwork , and to the integrin gene ITGB2 in the M2 macrophage subnetwork . It is also connected to the polymorphic genes TNFSF10B and TNFAIP in the interferon network and to TPM4 , INHBA , THBS1 , and CCL2 in the ECM subnetwork . The centrality of PLAUR within the consensus gene network suggests that PLAUR may be a key mediator of inflammatory and ECM remodeling signals in SSc skin . The proliferation subnetwork has predicted interactions with the inflammatory and ECM subnetworks . The most pronounced connection is between the ECM subnetwork and the cell proliferation subnetwork through TGFβ pathway genes ( Fig . 6 ) . The TGFβ pathway is known to modulate cell proliferation . There are multiple paths from the TGFβ pathway genes TGFB3 and TGFBR2 to the cell proliferation subnetwork through the polymorphic genes IRAK1 and PXK , which have predicted interactions with the serine/threonine kinases LATS2 , WNK4 , and PRKAA1 . Serine/threonine kinases are well known to be important regulators of cell proliferation and they are bridges between the ECM subnetwork and cell proliferation network . Two of the consensus clusters were SSc subset-specific ( Fig . 3B ) . These clusters contain the key gene expression abnormalities in SSc that are conserved across all three cohorts . The consensus clusters are enriched for inflammatory process Gene Ontology terms , as well as TGFβ signaling , PDGF signaling , and cell proliferation ( Table 3 ) . Most ( 30 out of 41 ) of the genes with replicated SSc-associated polymorphisms are predicted to interact with genes in the consensus clusters; 28 out of 30 of these interact in the immune ( interferon , M2 macrophage , and adaptive immunity ) and TGFβ/ECM subnetworks ( Fig . 4 ) . The inflammatory-specific consensus cluster also contains the genes FBN1 and AIF1 . Previous work implicates FBN1 in SSc pathogenesis , as a duplication of FBN1 causes fibrosis in the Tsk1 mouse [58] and a point mutation in FBN1 causes the fibrotic phenotype in the Stiff Skin Syndrome mouse [59] . Fibrillin-1 forms a matrix of elastic microfibrils that provide a scaffold for elastins and collagens , and a means for sequestering matricellular growth factors . Mouse embryonic fibroblasts expressing the Tsk1 mutant FBN1 have altered microfibril morphology that results in increased collagen deposition [58] . While polymorphisms in FBN1 might cause dosage effects that result in fibrosis in some models ( e . g . Tsk1 ) , it is possible that chronic inflammation causes chronic high expression of FBN1 to similar effect in humans . Rare polymorphisms in FBN1 have been associated with SSc in some subpopulations [60]–[62] . Similarly , AIF1 is implicated in SSc disease progression . A SNP in AIF1 has been implicated in anticentromere antibody ( ACA ) positive SSc [63] . Moreover , AIF-1 is interferon-inducible , constitutively expressed in macrophages [64] , and plays a role in vasculogenesis and endothelial cell proliferation and migration [65] . In the Sclerodermatous Graft-Versus-Host Disease ( sclGVHD ) mouse model of SSc , AIF1 was found to be highly expressed in skin [66] and to induce fibroblast and monocyte chemotaxis [53] . AIF1 has many predicted interactions with chemokine receptors CCR1 and CX3CR1 , which are connected to chemokines CX3CL1 ( fractalkine ) and CCL2 ( MCP-1 ) . The genes CX3CL1 and CCL2 are M1 and M2 macrophage-related genes respectively [67] and are chemotactic for monocytes , macrophages , and T cells [68] , suggesting enhanced recruitment of inflammatory cells to this subnetwork . A recent study of a mouse model of SSc demonstrated that both CCR2 and CX3CR1 regulate skin fibrosis , further implicating these mediators in the pathogenesis of SSc [69] . In addition , CCL2 has been shown to induce M2 macrophage polarization [70] , which may result in persistent M2 activation . The repeated and conserved finding of high AIF-1 levels in the inflammatory subset and its tight connection to innate immune mediators of inflammation suggest it may be involved in enhanced macrophage chemotaxis and activation in SSc skin . LYN , a hub of the adaptive immunity subnetwork , modulates B cell activation and plays a role in self-tolerance . B cell signaling has been implicated in SSc development and progression , as B cells have been shown to play a role in both the development of autoantibodies and cutaneous fibrosis in the Tight Skin 1 ( Tsk1 ) mouse model of SSc . Notably , LYN is overactive in response to overexpression of CD19 in this model [71] . Thus , LYN may play a role in the autoimmune component of SSc in human patients . The consensus gene network ( Figs . 4 and 6 ) also implicates genes as bridges between the subnetworks . These notably include the polymorphic genes PLAUR , IRAK1 , PXK , and GRB10 . In addition , we find differentially expressed genes straddling the subnetworks including RAC2 and LCP2 . The interconnections between the subnetworks present possible molecular paths through which these processes interact . The finding that most SSc-associated polymorphisms are associated with immune system mediators suggests that the initial events in SSc are likely to be immune-regulated and to involve interferon activation ( Fig . 7 ) . The immune response in SSc likely differs from a normal response because of predisposing genetic variants in these and associated genes . This may lead to the secondary recruitment of macrophages via RAS-RAC signaling ( Fig . 7 ) . We predict that the interferon network suppresses cell proliferation , given the clear distinction between the inflammatory and fibroproliferative subgroups . This inference is based on known interferon biology and not on the network itself . In contrast , it is possible that the ECM network stimulates cell proliferation through the TGFβ pathway and serine/threonine kinases IRAK1 , LATS2 , WNK4 , and PRKAA1 . In this model , inflammatory gene expression creates a balancing feedback loop that modulates fibroproliferative gene expression ( Fig . 7 ) . A major strength of the IMP network and its data integration capabilities derives from its ability to provide a more detailed picture of SSc development and progression compared with more conventional approaches . For example , while all of the purple nodes in Fig . 4 are highly expressed in the inflammatory group across all data sets , the IMP network provides information regarding gene-gene interactions in addition to expression data . In this example , the IMP network indicates which subnetworks correspond to discrete processes ( interferon , M2 macrophages , ECM , and adaptive immunity ) and which interactions are mediated through the network . Thus , we gain insight by recognizing that the interferon component is distinct from the M2 macrophage component , despite their co-expression and known interdependence . The value of the IMP network is as much in the connections that are not present as those that are . Since the original publication of the intrinsic subsets , two important questions have been central to their interpretation and their clinical relevance: First , can a patient's subset change over the course of their disease ? And second , can the subsets predict therapeutic response ? Pendergrass et al . [11] demonstrated that a patient's subset is stable over time scales of 6 to 12 months . This means either that patients never change subsets and the intrinsic subsets are effectively distinct diseases , or that the subsets are long-lived states of the same disease . Our analysis shows that the inflammatory and fibroproliferative subsets share a molecular network containing TGFβ pathway genes and ECM component genes , suggesting that inflammatory patients may transition to the fibroproliferative subset , perhaps in response to successful immunosuppressive therapy . Indeed , immunosuppressive therapy has not been widely successful for treatment of SSc [72] . On the other hand , fibroproliferative biopsies still have some activation of the TGFβ/ECM network despite the absence of the inflammatory signature ( Fig . 4 ) . The connection of the subsets through the TGFβ/ECM subnetwork indicates that the fibroproliferative subset shares a common pathway with the inflammatory subset and that the fibroproliferative subset is tied to chronic TGFβ activation and ECM deposition . Thus , based on the molecular network , it is possible that immunosuppressive therapy can move patients to the fibroproliferative subset rather than restoring their gene expression to that of healthy skin . Our data from an ongoing MMF clinical trial and analysis of mouse models of SSc suggests that gene expression changes precede clinical changes [4] , [66]; therefore gene expression could act as a readout for the effectiveness of a drug . This idea should be rigorously tested in clinical trials that carefully monitor gene expression in patient skin biopsies . The pathogenesis of SSc has been enigmatic , but a number of genetic risk factors have been identified by genome-wide association studies and candidate gene studies . Three of these polymorphic genes , NOTCH4 , IRF7 , and GRB10 , are in the inflammatory consensus cluster , and hence are consistently differentially expressed in the inflammatory subset ( Fig . 4 ) . This suggests that these may be cis-acting alleles and demonstrates the need for candidate gene studies to determine if differential expression is genetically driven in a subset of patients . The IMP functional network predicts that twenty-five of the remaining forty-one polymorphic genes interact with genes from the inflammatory consensus cluster ( Fig . 4 ) . Rather than being scattered evenly across all of the subsets or unrelated to any of the consensus genes , the risk alleles are overwhelmingly related to the inflammatory subset . The genetic studies , however , did not stratify their patients by intrinsic gene expression subset . The studies were carried out as case versus control or case versus case , when stratified by autoantibody status or other clinical outcomes . Risk alleles associated with a particular gene expression subset have not been reported . We reemphasize the fact that we found no consensus clusters that were differentially regulated in all SSc vs . healthy control biopsies . These data support the hypothesis that the subsets are related to disease progression and that SSc starts with immune activation , perhaps in response to an environmental trigger [73] , [74] . The SNPs associated with SSc would then likely be risk factors for an aberrant immune response to this trigger . Should such a model be correct , we are still left with the question of why we have different subsets that generally show little or no correlation with disease duration . The simplest explanation for this result is that patients progress through the gene expression subsets at dramatically different rates and that our measures of disease duration are currently inadequate . Another possibility is that any given patient transitions between these intrinsic gene expression groups in a dynamic manner that we do not observe using serial skin biopsies across 6–12 month time interval . This would mean that cross-sectional studies of patients would still capture all subsets while maintaining a weak correlation to disease duration . We think this is unlikely because serial biopsies are generally found in the same subset . The final possibility is that the subset a patient stays in , and the duration in which they remain , is dependent on many outside and as yet poorly characterized factors . These could include environmental stimuli that trigger an inflammatory response , or genetic factors that determine the rate at which one progresses through the mechanistic stages of SSc . It is possible that patients in each intrinsic subset have a different set of predisposing genetic polymorphisms or similar environmental triggers . This can only be addressed if we can look for genetic risk factors in a cohort of patients stratified by gene expression subset for genetic risk alleles . There may be genetic risk factors that cause a patient to “stall” at particular point along the progression from inflammatory to proliferative to normal-like . Genetic modifiers of the molecular links in the consensus gene network ( Fig . 4 ) might hold the key to showing why many patients go into spontaneous remission while others experience rapid clinical progression , and indeed , our network analysis suggests candidates for explaining this ( Figs . 6 , 7 ) . For example , IRAK1 and PXK are polymorphic genes that exist on paths in the network between the TGFβ/ECM network and the cell proliferation network . This strongly argues for future studies that test their possible roles in TGFβ-modulated cell proliferation , with particular attention to their roles in influencing other serine-threonine kinases that modulate the cell cycle . The presence of antinuclear autoantibodies in patient serum is a widely used biomarker of SSc . To date the intrinsic subsets have shown no clear association with autoantibody status [1] , [4] , [11] , which is consistent with a model by which the subsets represent disease progression . Several genetic polymorphisms are associated with autoantibody status ( S9 Data file ) , including BLK and BANK1 , which are related to ACA- and ATA-positive SSc respectively . These B cell proteins are already attractive candidates for autoantibody production , as they are directly associated with the cells that produce the antibodies , but our network analysis also shows that they are functionally related to adaptive immune genes that are highly expressed in the inflammatory subset . A primary role of bioinformatics in complex diseases is to pare down the possibilities to a coherent set of candidates for future study . The many risk alleles for SSc each have modest odds ratio and the final picture of SSc will likely lie in the interactions between various risk factors , but the number of possible interactions between these combinatorial factors is prohibitively large . It is here that the network approach may be most useful in delineating candidates for interaction studies . We might speculate , for example , that SSc results from the presence of multiple , functionally distinct alleles , but that it does not matter what gene is mutated as long as the mutation has a particular functional outcome . The predicted interactions in the network suggest which alleles might be functionally related and which might be distinct from each other , as the alleles either cluster within a subnetwork or straddle the subnetworks . This report is limited by our utilization of whole skin biopsies , which are complex mixtures of cells , and in that the studies were observational . The use of whole skin means that we cannot directly ascribe gene expression to specific cell types . For example , we infer that the M2 macrophage subnetwork is related to that cell type based on the coherent expression of monocyte markers and cytokines related to M2 polarization of macrophages . Our study is therefore hypothesis generating . Mechanistic studies will be needed to evaluate the existence of the molecular links suggested by the network analysis . Nevertheless , our analyses place the intrinsic subsets as a possible readout of SSc pathology . The consensus gene expression of the subsets implicates a number of molecular mechanisms that have been associated with SSc and suggests functional roles for a large fraction of the replicated SSc-associated polymorphisms . We demonstrate that the core molecular processes of the inflammatory and fibroproliferative subsets are molecularly connected to each other . This suggests the possibility that SSc subsets may be dynamic and interconnected . The analysis of prospectively collected human samples in this study was approved by the Committee for the Protection of Human Subjects at Dartmouth College ( CPHS#16631 ) and by the IRB review panel at Northwestern Feinberg School of Medicine ( STU00004428 ) . All subjects in the study provided written consent , which was approved by the IRB review panels at Dartmouth College and Northwestern Feinberg School of Medicine . This study used data from three previously published cohorts ( Table 1 ) . Each of the studies is available from NCBI GEO at the following accession numbers: Milano et al . ( GSE9285 ) , Pendergrass et al . ( GSE32413 ) and Hinchcliff et al . ( GSE45485 ) . We used an expanded version of the Hinchcliff dataset that contained an additional 12 SSc patients , 1 healthy control and 1 morphea patient beyond what was included in Hinchcliff et al . [4] ( GSE59785 ) . Each of the three study cohorts contained patients with SSc defined using the 1980 ACR criteria . Specifically , all patients met the American College of Rheumatology classification criteria for SSc [75] and were further characterized as the diffuse ( dSSc ) , or the limited ( lSSc ) subsets . Limited SSc patients had 3 of the 5 features of CREST syndrome , or had Raynaud's phenomenon with abnormal nail fold capillaries and scleroderma-specific autoantibodies . All three studies used Agilent Technologies 44 , 000 element DNA microarrays representing the full human genome . All samples were processed and all microarrays hybridized in the Whitfield lab providing consistency between the datasets . The DNA probes between these datasets are identical and thus were indexed using the same probe identifiers allowing direct mapping from one data set to another without significant loss of data . Microarray data from each cohort were Log2 Lowess-normalized and only spots with mean fluorescent signal at least 1 . 5 greater than median local background in Cy3- or Cy5- channels were included in the analysis . Genes with less than 80% good data were excluded . Since a common reference experimental design was used for all cohorts , each probe was centered on its median value across all arrays . Data were multiplied by -1 to convert them to Log2 ( Cy3/Cy5 ) ratios . The three cohorts were clustered into coexpression modules using the WGCNA procedure . We used the WGCNA R package available on the Comprehensive R Archive Network ( http://cran . r-project . org ) and described in [10] . We used the default parameters for running the software except that we used the “signed” network option and a soft thresholding parameter d = 12 . These parameters are described in depth in [10] , [15] . Genes that were classified as outliers were discarded from further analysis . To each pair of modules from different datasets we associate an overlap score W . Specifically , if Ci is a module in , say , Milano et al . and Cj a module in Pendergrass et al . , then we definewhere N is the total number of genes in the genome . The W-scores can be interpreted as edge weights in a module-module network ( the information graph ) . This network encodes the mutual information between the WGCNA-derived genomic partitions . We computed the W-scores between each pair of modules across all three datasets by the above formula and set the small and negative W-scores below a threshold to zero ( S4 Fig . ) . A mathematical derivation of the relationship between the W-scores and mutual information and a detailed description of the thresholding procedure are available in the supporting information ( S1 Text ) . The resulting 3-partite information graph was mined for consensus clusters . Since triangles in the information graph represent a module conserved across all three datasets , we clustered the information graph using a variant of triangle percolation [17] , which is a community detection procedure designed to find sets of modules that are members of many triangles together . Specifically , from the information graph we constructed an auxiliary graph , called the triangle graph , and detected communities in the triangle graph by greedy modularity maximization [76] . A description of the construction of the triangle graph is available in the supporting information ( S1 Text ) . We define a final consensus cluster as all of the genes that are contained in a module from the community for each of the three data sets community . Note that triangle percolation allows for overlapping communities in the underlying information graph . For example , the inflammatory consensus cluster and the keratinocyte consensus cluster overlap by one module ( Fig . 3A ) . This is one of MICC's strengths because it does not require a whole module from one dataset to be associated with only one consensus cluster . To derive a gene set associated to the consensus cluster , we took all modules within that community , computed their unions within their dataset , and then computed their intersection across datasets . In symbols , let Comm denote a set of modules that form a community in the information graph ( e . g . the dotted circles Fig . 1B and the colored nodes of Fig . 3A ) . Let MComm , PComm , and HComm denote respectively the sets of Milano , Pendergrass and Hinchcliff modules within Comm . Let m , p , and h denote modules in the Milano , Pendergrass , and Hinchcliff data sets respectively; note that these are sets of genes . We associate a gene set CCComm with the community Comm through the following formula:We call CCComm the consensus cluster associated with the community Comm and it consists of all genes that are present in a module from each data set within the community . The elements of CCComm are the consensus genes . It is clear by definition that the consensus clusters are nonoverlapping even though communities can share modules . This is because a gene needs to be present in a module in the community from each of the three data sets . Since modules do not overlap within data sets , consensus clusters cannot either . To determine if a WGCNA-derived module was significantly differentially regulated in a subset , we performed one-tailed Wilcoxon rank sum tests . Specifically , we computed the module eigengene of each module by first normalizing the gene expression so that each gene expression vector had Euclidean length 1 . The module eigengene is the first principal component of the normalized gene expression vectors within the WGCNA module . The module eigengene is a one-dimensional summary score for the module's gene expression across all biopsies . To determine if the module was significantly up- or down-regulated in a particular subset , we determined if the median of the module eigengene for that subset was above or below that of the whole population , and then performed a one-tailed Wilcoxon rank sum test to determine the significance of the median being above or below that of the population as a whole . We used the subset assignments reported in the previous papers describing these datasets [1] , [4] , [11] . We used Bonferroni corrections for multiple comparisons . There were 178 modules in total across the datasets . In Table 2 , we corrected for 178×3 tests for each of the subset-specificity tests . In Fig . 3B , we corrected for 178×4 tests because we included tests for all non-normal-like SSc versus normal-like SSc and healthy controls ( see also S4 Data file ) . The IMP Bayesian network is available through an online interface at ( http://imp . princeton . edu ) . To build our network , we queried IMP with four gene sets: inflammatory and fibroproliferative consensus genes derived from the consensus clusters , SSc-associated polymorphisms ( as described below ) , and the four gene MRSS biomarker reported in [5] . IMP provides export of the subnetwork corresponding to the query genes as a weighted edge list ( a three-column table indicating which genes are connected and with what probability ) . IMP automatically thresholds the probabilities at 0 . 5 and exports the network with up to an additional 50 genes that provide extra context for the query genes . In our case , the 50 genes were predominantly cell cycle genes . This is probably because the cell cycle is heavily studied in the microarray compendium from which IMP was built . In that case , IMP would be highly confident about predicting interactions between the fibroproliferative genes and other cell cycle genes . We developed in-house Matlab and R scripts to transform the edge list data into the Graph Exchange Format ( gexf ) , which allows for manipulation in Gephi , an open source network visualization program [77] . Data files S7-S8 contain post-processed networks and Data files S10-S11 provide R data and code snippets for manipulating the network programmatically . We collected genes with SSc-associated polymorphisms from the literature and curated them according to the following criteria . We included polymorphic genes that were reported in genome-wide association studies of SSc [78]–[82] , from a recent study using the Immunochip platform [83] and from case-control candidate gene studies that were replicated in at least one other study [54] , [84]–[105] . This resulted in a list of 41 polymorphic genes ( S6 Data file ) .	Systemic sclerosis ( SSc ) is a rare autoimmune disease characterized by skin thickening ( fibrosis ) and progressive organ failure . Previous studies of SSc skin biopsies have identified molecular subsets of SSc based upon gene expression termed the inflammatory , fibroproliferative , normal-like , and limited intrinsic subsets . These gene expression signatures are large and although the biological processes are conserved , the exact list of genes can vary across datasets due to random variation , as well as minor differences in the composition of the study cohorts ( e . g . early vs . late disease ) . We developed a computational tool to identify the consensus genes underlying the subsets across heterogeneous data and characterized the biological role of the consensus genes in SSc in order to obtain a systems level perspective of the SSc subsets . Our analysis reveals a complex network of genes connecting two of the major SSc intrinsic subsets , inflammatory and fibroproliferative . Many genetic loci associated with SSc risk show connections with the consensus genes of the intrinsic subsets , indicating that differential expression of genes defining the subsets may be related to genetic risk for SSc , thus for the first time placing the genetic risk factors in the context of , and showing putative relationships with , the intrinsic gene expression subsets .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "systems", "biology", "genetics", "biology", "and", "life", "sciences", "genomics", "computational", "biology" ]	2015	Systems Level Analysis of Systemic Sclerosis Shows a Network of Immune and Profibrotic Pathways Connected with Genetic Polymorphisms
During sprouting angiogenesis in the vertebrate vascular system , and primary branching in the Drosophila tracheal system , specialized tip cells direct branch outgrowth and network formation . When tip cells lumenize , they form subcellular ( seamless ) tubes . How these seamless tubes are made , shaped and maintained remains poorly understood . Here we characterize a Drosophila mutant called ichor ( ich ) , and show that ich is essential for the integrity and shape of seamless tubes in tracheal terminal cells . We find that Ich regulates seamless tubulogenesis via its role in promoting the formation of a mature apical extracellular matrix ( aECM ) lining the lumen of the seamless tubes . We determined that ich encodes a zinc finger protein ( CG11966 ) that acts , as a transcriptional activator required for the expression of multiple aECM factors , including a novel membrane-anchored trypsin protease ( CG8213 ) . Thus , the integrity and shape of seamless tubes are regulated by the aECM that lines their lumens . Biological tubes transport nutrients , respiratory gases , metabolic wastes , and secretions that are essential for viability . To function properly , these tubes must achieve and maintain their proper shapes . Defects in tube shape can result in severe organ malfunction , such as in polycystic kidney and liver diseases , and in hereditary hemorrhagic telangiectasia [1 , 2] . Most biological tubes are comprised of polarized epithelial or endothelial cells that shape their apical domains to build and expand a tube lumen . To carry out their specialized functions , tubes come in different architectures , shapes and sizes [3] . In multicellular tubes , an epithelial sheet surrounds an extracellular lumenal space with intercellular junctions connecting the cells to each other to form a selectively permeable barrier . Tubes can also be unicellular: an individual cell may wrap around a lumenal space and seal into a tube through the formation of self-junctions ( autocellular tube ) , or may generate an internal subcellular lumenal space unbounded by cell junctions ( seamless tube ) . Seamless tubes are found across the animal kingdom , from invertebrates to vertebrates . The most extensively characterized seamless tubes are those of the nematode excretory system , the vertebrate vascular system , and the fly respiratory ( tracheal ) system . Within the duct and canal cells of the C . elegans excretory system , seamless tubes form by two distinct mechanisms–by fusion of membranes bridged by autocellular junctions , or by fusion of intracellular vacuoles into an internal subcellular tube [4–6] . In the vertebrate vasculature , seamless tubes form within endothelial tip cells that lead the migration of new branches and mediate anastomoses [7–10] . In a striking parallel , tracheal tip cells in Drosophila lead outgrowth of primary branches and form seamless tubes that either are blind-ended ( terminal cells ) or mediate anastomoses ( fusion cells ) [11] . During embryonic stages , terminal cells will form a single gas-filled seamless tube; however , during larval life terminal cells will emanate dozens of subcellular branches that ramify on internal tissues and organs , with each branch containing a blind-ended seamless tube that serves as the final interface for gas exchange [12] . Despite their ubiquity in the animal kingdom , the mechanisms regulating seamless tube morphogenesis remain poorly understood . Branching morphogenesis has been examined in a broad range of model systems , with multiple mechanisms implicated in the shaping of tubes [13–34] . Among these are mechanisms in which the apical extracellular matrices ( aECMs ) that line the tube lumen play a key role . The aECM is a heterogeneous three-dimensional extracellular matrix comprised of polysaccharides , proteoglycans , and glycoproteins . Lumenal matrices are common to tubular organs [35–37] and are thought to regulate multiple steps of lumen morphogenesis; however , their mechanisms of action remain , in general , poorly defined . The lumenal matrix protein GP-135/Podocaylxin is an early marker of apical domains in MDCK cysts [38 , 39] and is required for epithelial polarity , suggesting a role for lumenal matrix factors in lumen initiation . After apical membranes have been established , they must separate and expand to create a lumenal space . Matrix components are thought to promote lumen expansion by inhibiting adhesion between apical membranes [40 , 41] and/or creating an extensively glycosylated lumenal surface that promotes the osmotic influx of water to form an expandable gel-like matrix [42–46] . Finally , lumenal matrix components can regulate lumen growth by influencing apical membrane morphogenesis and cell shape [47–53] or by influencing cell rearrangements [52 , 54] . Perhaps the most thoroughly studied aECM is that of the multicellular tube ( the dorsal trunk ) in the embryonic Drosophila tracheal system [55] . The most abundant component of insect aECM is chitin [56] . During embryonic development , a transient chitin filament is first deposited within the lumen of the dorsal trunk [57–59] . This filament regulates multiple aspects of tube shape , including length and diameter . In mutants disrupting chitin synthesis , tracheal lumens expand in an uncoordinated manner across the tracheal epithelium , resulting in a cystic lumen that has alternating areas of tube constriction and dilation [57–59] . Later , embryonic tracheal tubes elongate , to create a continuous network , as neighboring tracheal hemisegments fuse together . Coinciding with expansion of the apical domain to accommodate the increase in dorsal trunk tube length , the physicochemical properties of the chitin filament are altered through the action of the chitin deacetylases Vermiform and Serpentine . This covalent modification of chitin filaments is required to restrict axial tube growth [49 , 50 , 60]–dorsal trunk tubes in mutants lacking both deacytlases become excessively elongated and convoluted . The chitin filament expands in concert with the apical membrane , suggesting the two structures are physically coupled [61] . Through this physical coupling , the chitin filament may create a resistance force counterbalancing axial apical membrane growth [61] . During seamless tube growth , single epithelial cells must dramatically increase their surface area/volume ratio through a dramatic increase in the size of the apical compartment [10 , 62] , yet the continuity and shape of these lumens must be maintained across this relatively large surface area . How seamless tube diameter and shape are initially determined and then maintained is not known; likewise , the role of the aECM in seamless tube morphogenesis is unknown . The aECM is ideally positioned to coordinate apical membrane morphogenesis along the length of seamless tubes , either by providing mechanical support and by actively influencing apical membrane dynamics . Indeed , during seamless tube elongation in the duct cells of C . elegans , a transient lumenal matrix maintains lumen integrity by promoting apical membrane expansion [41] . In Drosophila terminal cells , the majority of seamless tube growth occurs during larval stages , after the chitin filament has been cleared and replaced by a more mature aECM called the cuticle [63] . Initial cuticle deposition is followed by tracheal gas-filling [64 , 65] . The tracheal cuticle is a multi-layered aECM organized into a series of circumferential folds , called taenidia [12 , 66] . Our understanding of how aECMs , such as the Drosophila cuticle , mediate outside-in signaling in tubular epithelia is greatly hindered by a grossly incomplete understanding of their molecular constituents and their functions . As a result , the outside-in signaling pathways linking aECM components to intracellular regulators of seamless tube morphogenesis remain entirely unknown . Roles for aECMs in tube morphogenesis span multiple organs and tissues from invertebrates to vertebrates , indicating an evolutionarily ancient reliance on aECMs to ensure proper lumen shape . An integrated model for lumen morphogenesis , therefore , requires a better understanding of the molecular components of lumenal matrices and their functions during tubulogenesis . We find that the tracheal mutant , ichor ( ich ) , previously identified in a forward genetic screen [67] , compromises seamless tube shape and integrity by disrupting the aECM . We determined that ich mutations are loss of function alleles of the uncharacterized gene CG11966 , which encodes a putative zinc finger transcription factor . Expression of ich coincides spatially and temporally with cuticle production , and ich is essential for aECM assembly , including of the lumenal matrix lining terminal cell lumens . Further , we show that Ich functions as a transcriptional activator in terminal cells and identify aECM components whose expression is Ich-dependent . Taken together , these data suggest a role for Ich in coordinating the assembly or modification of the aECM , and demonstrate that the aECM plays a crucial role in seamless tube shape and integrity . A single mutation in the gene ichorous , here renamed ichor ( ich ) to avoid confusion with the vertebrate zinc finger transcription factor , ikaros , was identified on the basis of a terminal cell-specific gas filling defect ( Fig 1A and 1A’ , S1E–S1G’ Fig ) in a genetic mosaic screen [67] . In contrast to neighboring heterozygous control terminal cells , GFP-labeled terminal cells homozygous for ich ( see Materials and Methods ) lack gas-filled seamless tubes . This phenotype indicates that ich terminal cells are defective in some aspect of tube formation and/or liquid clearance . To determine whether ich terminal cells lacked lumens , or instead contained fluid-filled lumens , a fluorescent reporter , lum-GFP , was utilized and found to be secreted into a liquid-filled lumenal space [67] . To further characterize the seamless tube defects , we expressed a genetically encoded chitin reporter [68] in heterozygous control cells ( ich206/+ ) and GFP-labeled ich206 clones ( Fig 1B and 1C” ) . Third instar larvae were heat-killed and examined in whole-mount preparations . In control terminal branches ( Fig 1B–1B” ) , ChtVis-Tdtomato was observed bounding gas-filled lumens as well as in the fluid-filled tips of terminal branches ( Fig 1B’ and 1B” ) , indicating ChtVisTdTomato labels the cuticle-lined gas-filled lumens as well as fluid-filled lumens . In contrast , ChtVis-Tdtomato in ich terminal cells outlines fragmented and discontinuous lumens ( Fig 1C–1C” ) . Discontinuities were present throughout the terminal cell; for example , in proximal regions connecting the terminal cell to the neighboring stalk cell ( Fig 1C ) , and in distal regions near the growing tips of the cell ( Fig 1C” ) . Staining against a lumenal membrane antigen ( anti-Wkd peptide ) [69] showed that in contrast to control ( 2FRT ) terminal cells , which contained patent tubes with locally uniform morphology ( Fig 1D’ ) , ich terminal cells exhibited discontinuous membrane-bounded lumens ( Fig 1E’ and 1F’ ) with an irregular cystic appearance . Other defects were noted at a lower penetrance ( see S1A , S1A’ and S1D–1D” Fig ) . Although seamless tubes first form about 13 hours after egg lay ( ael ) [11] , late stage ich embryos ( 16 hrs ael and older , S1B and S1C Fig ) did not exhibit defects in initial seamless tube lumen extension , indicating that the tube discontinuities must arise after embryogenesis is complete . Collectively , these data suggest that ich disrupts multiple aspects of tube morphogenesis in terminal cells , including seamless tube growth , shape , and integrity . From the original screen [67] , only a single allele of ich ( ich206 ) was isolated . To facilitate mapping and gene characterization , an EMS mutagenesis and non-complementation screen was performed to recover additional ich alleles ( see Materials and Methods ) . We recovered a single new allele ( ich543 ) that exhibited identical terminal cell defects ( Fig 1F–1F’ , Fig 2F ) . Genetically , ich206 and ich543 behaved as strongly hypomorphic or amorphic alleles ( see Materials and Methods ) ; both were recessive embryonic lethal , and indistinguishable in homozygous and hemizygous conditions ( S1J Fig ) . Using standard procedures ( see Materials and Methods ) we mapped ich to CG11966 ( Fig 2A ) . The CG11966 open reading frame encodes a 592 amino acid protein with two C2H2-type zinc fingers and a predicted nuclear localization signal ( NLS ) ( G473-Q483 ) . The ich206 allele carries a non-sense mutation ( coding for Q107→Stop ) upstream of the two C-terminal zinc fingers , while the ich543 allele carries a missense mutation ( coding for a H582→Y ) resulting in the substitution of a tyrosine residue for a histidine that is required for zinc ion coordination in the second zinc finger . Combined with our genetic analysis , this suggests that at least the second zinc finger domain is essential for Ich function . Further establishing the identity CG11966 as ich , we showed that terminal cells homozygous for a small chromosomal deletion [70] uncovering CG11966 ( as well as oskar , skap , and CG11964 ) ( Fig 2B’B’ ) show the ich phenotype . Likewise , expression of a dsRNA ( hereafter ich RNAi ) targeting the CG11966 transcript induced terminal cell-specific tube discontinuities ( Fig 2C– 2D’ ) . Finally , full-length and flag-tagged ich cDNAs were used to generate transgenic strains ( see Materials and Methods ) , and expression of wild type CG11966 under control of pan-tracheal btl-GAL4 [71] was able to rescue seamless tube continuity in ~67% of ich mutant clones ( Fig 2E and 2F ) . Interestingly , expression level-dependent gain of function defects in morphology , including terminal cell pruning ( Fig 2E and S2C–S2D’ Fig ) , were also observed . To better understand ich function during tracheal development , we sought to characterize its expression pattern during embryogenesis . Though our genetic mosaic analysis indicated ich is required cell-autonomously in the trachea , previously published RNA in situ data [72] do not show ich expression in the trachea , suggesting ich may be weakly expressed . We used a LacZ enhancer trap , P{PZ}l ( 3 ) 05652 , at the ich locus to better characterize ich expression in the trachea and other tissues . From 9–15 h ael , the ich::nLacZ transcriptional reporter was expressed specifically in ectodermally-derived epithelia , including the trachea , that secrete cuticle ( S3 Fig ) . This expression pattern was consistent with a role for Ich in aECM assembly and/or modification , which was further supported by the observation that ich transcripts are expressed during cuticle deposition in the pupal wing epithelium [73] . To investigate further , we asked whether ich embryos , like embryos with known defects in the production of a mature aECM ( cuticle ) , would appear grossly inflated in cuticle preparations–the “blimp” phenotype , which reflects an increased elasticity ( compare Fig 3A and 3A’ ) [74] . Both ich206 and ich543 embryos exhibited mild “blimp” defects in cuticle preps ( Fig 3B’ ) as well as reduced sclerotization of the head skeleton and ventral epidermal denticles ( Fig 3B’ and 3C’ ) , which are also characteristic of aECM defects . These defects were milder than those observed in embryos mutant for chitin synthase , suggesting that ich embryos are not completely chitin deficient . In the tracheal system , a chitin filament is generated prior to the mature cuticle , and plays an essential role in tube expansion [57–59] . We find that the production of the chitin filament is unaffected in ich embryos ( S4 Fig ) , consistent with ich not being required for chitin synthesis . Though Ich is dispensable for tracheal specification , branching , and lumen growth during embryogenesis , pan-tracheal depletion causes a liquid-clearance defect and breaks in the multicellular dorsal trunk tubes in first instar larvae ( S1H and S1I’ Fig ) . Since tracheal cuticle assembly is required for gas-filling [64] and tracheal tube integrity [12 , 75] , we next visualized the ultrastructure of the tracheal cuticle in larval terminal cells using transmission electron microscopy ( TEM ) . Wild-type terminal branches have locally uniform lumens lined by a cuticle organized into taenidia ( Fig 3D ) . By contrast , terminal branches in ich-depleted cells have irregularly-shaped lumens devoid of taenidia and instead occluded with electron-dense material ( Fig 3E ) . The electron-dense material may represent mis-assembled matrix material . These phenotypes indicate that Ich is required for lumenal matrix assembly in seamless tubes and suggest that seamless tube shape requires a properly assembled lumenal matrix . Insofar as Ich is dispensable for chitin filament formation , we suggest that Ich functions in the trachea as a specific regulator of the mature aECM ( cuticle ) . Since the ich locus had not been previously characterized , we next sought to address Ich molecular function . To test the hypothesis that Ich acts as a transcription factor , we first examined the subcellular localization of a functional ( Fig 2E ) Flag-tagged isoform of Ich . UAS-Flag-Ich was expressed in terminal cells using the GAL4/UAS two-component system [76] . To limit over-expression phenotypes , we utilized a tubGAL80ts transgene [77] and shifted temperature to permit Flag-Ich expression only in third instar larvae ( Fig 4A–4A” ) . As predicted for a transcription factor , Ich localizes at steady-state to terminal cell nuclei ( Fig 4A’ and 4A” ) . To test whether Ich functions as a transcriptional activator , we assayed the rescuing activity of an Ich-VP16 chimera . The chimeric protein consisted of the putative DNA binding domain of Ich ( the Ich zinc fingers ) fused to the transcriptional activation domain of the viral protein VP16 [78] ( Flag-VP16-IchDBD , Fig 4B ) . An exogenous NLS was introduced to ensure nuclear localization of the Flag-VP16- IchDBD chimera ( Fig 4C’ ) . Expression of Flag-VP16-IchDBD in ich206 terminal cells restored seamless tube continuity in 84% of cells ( n = 46 cells ) ( Fig 4C” , Fig 2F ) . This result indicates that the essential terminal cell function of Ich is as a transcriptional activator . Consistent with this hypothesis , a fusion of the GAL4 DNA binding domain to full-length Ich ( GAL4DBD-Ich ) induced transcriptional activation of an artificial UAS-Luciferase reporter construct in S2 cells [79] . If Ich functions primarily as a transcriptional activator , then we predicted that ectopic recruitment of repressive machinery to Ich target genes in terminal cells would phenocopy ich loss-of-function . We expressed a chimeric Ich consisting of the Engrailed transcriptional repressor domain [80] fused to the Ich zinc fingers ( EnR- IchDBD-Flag , Fig 4B ) , and found that EnR- IchDBD-Flag localized to the nucleus of terminal cells ( S2B and S2B’ Fig ) and induced ich-like tube discontinuities in 98% of cells ( n = 42 terminal cells ) ( compare Fig 4D and 4D’ with Fig 4E and 4E’; Fig 2F ) . Importantly , the phenotypes caused by overexpression of EnR- IchDBD-Flag are distinct from those caused by overexpression of full-length Ich ( see S2 Fig ) . The best-characterized component of the insect aECM is chitin [66 , 74 , 81–83]; however , its requirement in seamless tubes has not been directly examined . While ich is not required for the chitin filament in embryonic development , we wished to determine if the ich seamless tube defects in terminal cells derive from the loss of chitin in the mature aECM . We turned to mutations in kkv , which encodes the sole tracheal chitin synthase . The short-of-breath ( sob ) alleles of kkv , were identified by their gas-filling defects ( S5A and S5A’ Fig ) [67] , but had not been more thoroughly examined . We now report that the transient chitin filament normally present in the embryonic tracheal lumens cannot be detected in kkvsob404 and kkvsob483 embryos ( S5D–S5G’ Fig ) , which displayed the cystic dorsal trunk defect ( S5H–S5L Fig ) characteristic of kkv1 embryos [57 , 58] . All 6 sob alleles carry point mutations predicted to disrupt the kkv coding sequence ( Fig 5A ) and two of these seemed likely to be null for kkv function . The kkvsob404 allele carries a non-sense mutation predicted to truncate the open reading frame after 147 amino acids ( Fig 5A ) and the kkvsob483 allele disrupts a conserved nucleotide in the splice donor site just upstream of the second intron , predicted to result in a S48→R missense mutation followed immediately by an in-frame stop codon . By genetic assays ( S5H–S5L Fig , see Materials and Methods ) , kkvsob404 behaved as a strong hypomorphic or null allele , whereas kkvsob483 exhibited recessive antimorphic properties ( or else carries a second-site modifier ) . If a loss of chitin synthesis underlies the ich tube defects in terminal cells , then chitin-deficient terminal cells should phenocopy ich mutants . The apical membrane in kkvsob terminal cells was cystic ( cysts marked by arrowheads ) , showing irregular contours ( Fig 5B , 5B” and 5C’ ) , in contrast to the smooth , locally uniform apical membrane of control terminal branches ( Fig 1D and 1D’ ) . In addition , kkv terminal branches contained apical membrane discontinuities , albeit with limited penetrance and expressivity ( Fig 5C’ and 5D’ ) . The penetrance of these discontinuities varied with the kkvsob allele used ( Fig 5E ) : whereas kkvsob404 terminal cell clones exhibit discontinuities in ~12% of terminal cells examined ( n = 38 ) , kkvsob483 clones exhibit discontinuities in ~52% of terminal cells examined ( n = 28 ) . This difference in penetrance is consistent with the analysis of allele strengths described above ( S5H–S5L Fig ) . Additionally , while multiple terminal branches per cell showed discontinuities in ich terminal cells , discontinuities in kkvsob terminal branches were typically limited to a single branch per cell . We conclude from these studies that a chitin-based aECM is required for seamless tube shape , and that it also contributes to tube integrity while not being absolutely essential . Interestingly , this suggests that other Ich-dependent aECM components may play chitin-independent roles in the maintenance of tube integrity . Since the transcriptional targets of Ich regulation are not known , we sought to identify candidate genes . To focus our search , we utilized two previously published RNAseq data sets , one from the modENCODE consortium [84] and one from the Adler lab , which generated a systematic RNAseq data set across the stages of cuticle deposition in the pupal wing [73] . We reasoned that ich target genes would be co-expressed with ich , and might be enriched in the overlap of the ich co-expression clusters from each data set . We manually selected 4 genes , each expressed in the embryonic trachea , to examine for ich-dependent expression: ectodermal ( ect ) , osi18 , osi19 , and CG8213 . Ect is an apically secreted protein expressed exclusively in chitin-secreting epithelia ( S6A–S6D Fig ) [85] . Depletion of ect in the pupal wing epithelium causes defects in cuticle assembly [73] , suggesting Ect is a structural component of chitin-based cuticles . Although ich is dispensable for tracheal expression of ect ( S6B and S6C–S6F Fig ) , ich is required for full ect expression in the foregut primordium and the epidermis . Like all Osiris family members ( 24 in total ) , Osi18 and Osi19 are small proteins with predicted signal peptides and transmembrane domains [86] . Gene expression studies suggest Osiris family members may play a role in aECM assembly [73 , 87] , perhaps by affecting one or more steps of membrane trafficking [88] . Expression of osi18/19 mRNA initiates ~10h ael specifically in the trachea . Consistent with a role for Osi18/19 in tracheal aECM maturation , osi18/19 expression is coincident with aECM secretion and modification ( S7 Fig ) . Wild-type ich function is essential for the expression of both osi18 and osi19 . Embryos mutant for ich ( ich206/Df ( 3R ) osk ) exhibit a delay in the onset of osi18/osi19 expression in the trachea ( compare Fig 6A with 6D , 6B with 6E , Fig 6G with 6J and 6H with 6K ) . By ~15h ael osi18 and osi19 mRNA are readily detected in the dorsal trunks of ich206/Df ( 3R ) osk embryos ( Fig 6F and 6L ) although expression in other tracheal branches remains severely reduced as compared to control siblings ( compare Fig 6C with 6F and 6I with 6L ) . That osi18 and osi19 expression in the dorsal trunks eventually becomes strong in the absence of ich , suggests that other transcriptional regulators expressed in the dorsal trunk can compensate . Consistent with this hypothesis , recruitment of repressive transcriptional machinery to ich-regulated loci through the tracheal expression of EnR-IchDBD-Flag confers a complete loss of dorsal trunk osi19 expression ( compare Fig 6N with 6M ) , suggesting that Ich is part of a network of transcription factors controlling gene expression during maturation of the tracheal aECM . The ich candidate target gene , CG8213 , encodes a transmembrane protease . RNAi studies in the pupal wing epithelium implicate CG8213 in cuticle assembly [73] . The expression pattern of CG8213 during embryogenesis is highly reminiscent of ich ( Fig 7A–7D ) . To determine whether Ich is required for CG8213 expression in embryonic trachea , ich206/ich543 embryos and sibling controls ( ich206 or ich543/+ ) were examined by in situ hybridization . While sibling controls ( Fig 7F and 7F’ ) showed the expected wild type pattern , ich206/ich543 embryos showed a loss of both pan-tracheal ( arrowhead Fig 7G ) and hindgut ( arrowhead Fig 7G’ ) expression of CG8213 . Foregut expression ( Fig 7G’ ) of CG8213 persisted in ich mutant embryos , and thus is independent of Ich function and also serves as a convenient internal control . Tracheal expression of CG8213 in ich206/ich543 embryos was restricted to a small number of cells at the posterior spiracle , indicating that Ich is required ( either directly or indirectly ) for the induction of CG8213 expression in the trachea . As an independent confirmation of this result , we found that pantracheal expression of the dominant negative Ich transgene ( EnR-IchDBD-Flag ) produced a tissue-autonomous reduction in CG8213 expression . Indeed , while control embryos showed prominent tracheal expression of CG8213 by ~11h ael ( arrowhead , Fig 7H ) , embryos with pan-tracheal expression of EnR-IchDBD-Flag showed a loss of CG8213 transcript exclusively in the trachea . As in ich206/ich543 mutants , tracheal expression of CG8213 persisted only in the posterior spiracles ( arrowhead , Fig 7I ) . We next sought to determine whether loss of CG8213 could account for the ich terminal cell defects . Terminal cell-specific knockdown , using either of two independent RNAi lines , induced numerous apical membrane discontinuities along each branch ( Fig 8B–8B” ) , as well as seamless tube cysts ( arrowheads . Fig 8B” ) . In light of these findings , we rename CG8213 lumens interrupted ( lint ) . In addition , ~21% of terminal cells ( n = 33 ) exhibited a dramatic decrease in branch number , with the remaining terminal branches containing only isolated inclusions of apical membrane ( Fig 8C and 8C’ ) , suggesting a defect in the addition of apical membrane in these terminal branches—similar to the most severe lumen defect observed in ich terminal cells ( S1B and S1B’ Fig ) . We confirmed the specificity of these RNAi phenotypes by targeting the lint locus region using the CRISPR/Cas9 system ( Fig 8D ) [89] . Though we did not isolate alleles that deleted the entire coding region by targeting both gRNA sites , we isolated a lint mutant allele ( lintΔ4 . 64 ) that behaves genetically as a loss-of-function ( see Materials and Methods ) . We found that lintΔ4 . 64 terminal cells clones exhibited discontinuous lumens ( Fig 8D , 8E and 8G ) similar to those observed in RNAi experiments , in clones homozygous for a non-complementing , recessive lethal transgene insertion at lint ( Fig 8F and 8G ) , and in ich clones . We propose that Lint acts downstream of Ich to regulate seamless tube growth and/or maintenance . To further test if lint is the essential target of Ich regulation in tracheal morphogenesis , we asked whether expression of a lint cDNA could bypass a tracheal requirement for ich . Overexpressing lint in otherwise wild-type terminal cells caused liquid-clearance defects ( S8A and S8A’ Fig ) , but did not impact terminal branching or lumen morphogenesis ( S8B and S8B’ Fig ) with the exception of a few small tube dilations observed in some , but not all , cells ( arrowheads in S8B’ Fig ) . Expression of lint in ich206 terminal cells restored seamless tube integrity and shape at a low very low frequency ( 2% of clones , n = 103 , P = 0 . 5 , one-sided Fisher’s exact probability test; S8E Fig ) . The limited rescue observed was expression level-dependent , as no rescue was observed when larvae were raised at 25°C ( n = 53 clones; S8E Fig ) , but the elevated transgene expression observed at 29 oC led to 2% rescue ( n = 103 clones , S8D and S8E Fig ) . Based on these results , we conclude that Ich regulates terminal cell aECM assembly , at least in part , by promoting ( directly or indirectly ) the expression Lint , a secreted protease that may act in the processing of lumenal matrix proteins . While extracellular matrices coating the basal aspects of epithelia have long been known to play key roles in organ development and morphogenesis , any potential role of extracellular matrices lining the apical or lumenal aspects of epithelia ( aECM ) have , by comparison , been largely neglected . Critical gaps in knowledge include: what the components of the aECM are , how they interact with each other and with the apical membrane during assembly , how they are modified over time , and how the aECM regulates morphogenesis and is altered under varying physiological conditions or in disease . In this study we identify an essential role for the aECM in maintaining the shape and integrity of seamless tubes ( S9 Fig ) . Tracheal terminal cells lacking the transcriptional activator Ichor , or its downstream target , Lumens Interrupted , exhibited discontinuous seamless tube fragments , in which the segments of tube that remained intact showed pronounced cystic dilations . TEM analysis demonstrates that aECM organization is compromised in ichor deficient larva . Further supporting the hypothesis that it is the aECM factors downstream of Ichor that are relevant to seamless tube morphogenesis , we show that the elimination of chitin , the most abundant aECM component in insects , is sufficient to compromise seamless tube shape and , to a lesser degree , integrity . Our findings further strengthen the conclusion , prompted by previous studies in nematodes and flies , that the aECM is important for maintaining epithelial integrity . In C . elegans embryos , the so-called sheath at the apical surface of the epidermis maintains epidermal integrity during cell shape changes that elongate embryos along their anteroposterior axis [90–93] . In the Drosophila trachea [54] and C . elegans excretory system [91] , lumenal matrices form scaffolds that maintain the integrity of intercellular contacts joining unicellular tubes . Do lumenal matrices regulate seamless tube integrity in vertebrates ? Endothelial cells secrete a lumenal matrix known as the glycocalyx , which has been implicated in lumen expansion in multicellular blood vessels [40] . In endothelial tip cells , blood flow is required to expand seamless tubes [9 , 10] . Interestingly , during this initial lumen growth , seamless tube stability is quite labile , undergoing temporary collapse during changes in blood pressure [9 , 10] . The endothelial glycocalyx , which promotes lumen expansion through electrostatic repulsion of lumenal surfaces [40] , is still maturing at the onset of blood flow [94] . It is possible that an immature glycocalyx contributes to seamless tube collapse during vessel anastomosis . Alternatively , the glycocalyx , a known mediator of shear stress response in endothelial cells [95] , could help transduce hemodynamic mechanical signals needed to promote the inward blebbing of lumenal membrane in endothelial tip cells [10] . It would be interesting to determine whether an intact glycocalyx is a prerequisite for stable lumen expansion in endothelial tip cells . The mechanisms by which the aECM promotes tube integrity in seamless tubes are not known . We favor two non-exclusive possibilities: first , we propose that the aECM may serve as a scaffold capable of dissipating tension along the apical membrane , and second , we propose that the lumenal matrix may play a more direct role in regulating the growth of seamless tubes . In Drosophila terminal cells that ramify over areas of 100s of square microns , seamless tubes are likely under considerable tensile stress . These tensile forces are far from static , changing in both magnitude and direction during tube growth as well as during larval locomotion . Against such dynamic tensile strain , tube integrity must be maintained for proper gas-exchange . In larger tube types ( autocellular and multicellular ) , cell junctions are thought to serve as a critical counter force resisting tensile stress; however , along the length of seamless tubes there are no cell junctions . The aECM , with its physical connections to the apical membrane , is ideally positioned to distribute the forces along the length of the tube and in that manner perhaps avoid local forces strong enough to fragment the tube . We also think it likely that the aECM plays a more informative role in regulating tube growth . In gain-of-function experiments , Ich overexpression arrested the growth of seamless tubes in terminal cells without perturbing lumen continuity . These data suggest that seamless tube growth in terminal cells is sensitive to the levels of Ich effectors , possibly including lumenal matrix factors . A model in which a lumenal matrix merely maintains expanded terminal cell lumens cannot readily account for defects in apical membrane biogenesis , suggesting that components of the aECM may help to organize a membrane platform promoting the growth of the apical domain in terminal cells . Indeed , in loss of function studies , the most severely affected ich or lint terminal cells showed scattered inclusions of apical membrane , consistent with a failure in the addition of new apical membrane to the growing blind end of terminal cell seamless tubes . This would be consistent with previous work demonstrating that the aECM and the cortical actin meshwork underlying the apical membrane are in communication [58 , 96 , 97] . The question of how aECM change and mature over time is also one that applies broadly across tubular organs and phyla . For example , in order for the mammalian lungs or the Drosophila tracheal system to become functional , their lumens must transition from fluid-filled to gas-filled . Conserved mechanisms mediate this transition , such as liquid and salt reabsorption via epithelial Na+ channels [98–102] . The transition from fluid-filled to gas-filled lumens also entails secretion and modification of a lumenal matrix . The ichor mutant has a specific defect in tracheal tube maturation without affecting earlier tracheal morphogenesis . Others [64 , 75 , 103] have reported that mature aECM assembly is required for tracheal liquid clearance and/or gas-filling . In our mosaic analysis , we found ich to be required only in larval terminal cells but to be dispensable for tracheal specification , branching , and lumen morphogenesis during embryogenesis . On the other hand , pan-tracheal knockdown of ich blocked tracheal gas-filling and resulted in one or more breaks in the dorsal trunks of first instar lavae , suggesting a tissue autonomous but non-cell autonomous role of Ich in larger tubes ( presumably reflecting lumenal secretion of Ich target genes ) . Thus ich is broadly required during the maturation of the trachea at the end of embryogenesis or shortly after hatching . This phenocritical phase for ich coincides with tracheal aECM maturation ( cuticle deposition ) and fits with our EM data demonstrating that aECM is disorganized in the trachea of larvae deficient for ich . The mature tracheal aECM is thought promote the liquid-to-gas transition by promoting de novo gas bubble generation via cavitation [103] . By this process , a column of lumenal liquid must be ruptured at some point to generate a gas-bubble , requiring a decrease in the pressure gradient required to rupture the liquid column at the liquid-cell interface ( also called tensile strength ) . Both the hydrophobic envelope layer of the cuticle , as well as its convoluted taenidial folds , at the surface-liquid interface are predicted to decrease the tensile strength of the lumenal fluid [103] . Tracheal gas-filling takes place with temporal and spatial stereotypy [17] , implying the execution of an equally stereotyped developmental program . We propose that Ichor helps ensure this timely transition by initiating the transcriptional activation of a suite of genes promoting mature aECM assembly . Maintaining gas-filled respiratory lumens in insects and mammals is most critical in the finest lumenal spaces , such as lung alveoli and tracheal terminal cells . In the mammalian lungs , the fine lumens of alveoli impose high surface tension at the air-liquid interface , threatening collapse of these air-filled lumens . To prevent the collapse of these fine lumens , specialized pneumocytes secrete a matrix of phospholipids and proteins that form a lumenal surfactant matrix [104] . Failure to clear lumenal lung liquid or secrete the surfactant matrix underlies lethal respiratory distress syndromes in neonates . The fine lumens of terminal cells exhibit variable regions of fluid-filling , especially at the tips of terminal branches where the lumens are finest [67 , 105] . Analogous to the alveolar spaces of mammalian lungs , the fine lumenal spaces of Drosophila terminal branches may impose high surface tension in a surface-associated aqueous phase , causing the collapse of gas-filling in limited portions of terminal branches . An Ich-dependent lumenal matrix may be required to limit the expansion of these fluid-filled regions throughout the terminal cell , by reducing the surface tension in a lumenal-surface-associated aqueous phase . Modifications to a lumenal matrix are a conserved process underlying the functional maturation of tubular organs . In the Drosophila trachea [57–59 , 81] and C . elegans excretory system [41 , 91 , 106 , 107] transient aECMs give way to stable ones at the completion of embryogenesis . However , the developmental signals controlling such matrix reorganizations remain poorly understood . In the trachea , this process presumably entails the upregulation of enzymatic processes , such as proteolysis and chitin hydrolysis , at the end of embryogenesis to clear the chitin filament . Endocytosis is then required to clear material from tracheal lumens [17] . We propose that Ichor and its as-yet-unknown cooperating transcription factors ( see above ) help initiate stable aECM assembly in the trachea by transcriptionally activating a suite of genes , such as Lint , involved in cuticle biogenesis . While zinc finger transcription factors are abundant throughout the animal kingdom , Ich does not have clear orthologs beyond insects; however , the aECM targets whose expression Ich regulates may be more closely conserved . The domain organization of Lint , for example , is similar to that observed in a family of vertebrate type II transmembrane serine proteases ( TTSPs ) . Vertebrate TTSPs are defined structurally by an N-terminal single-pass transmembrane domain separating a short cytosolic domain from a larger extracellular domain containing a trypsin protease domain at its C-terminus [108] . Genome sequencing in mouse and humans has identified ~20 TTSPs in mouse and humans . Lint shows 32–37% identity in its C-terminal trypsin protease domain to multiple mouse and human TTSPs , including TMPRSS11f , ST14 ( Matriptase ) , TMPRSS9 , and mouse TMPRSS11C . Lint differs from canonical TTSPs in that it lacks the variable array of protein-protein interaction domains found in their extracellular domains . But , like Lint , vertebrate TTSPs such as Matriptase are expressed broadly in epithelia , suggesting a conserved role for secreted transmembrane proteases in the development or physiology of epithelial tissues [108] . Indeed , membrane-anchored proteases have a conserved role in the assembly and maintenance of epithelial barriers . Lint is required for cuticle assembly in the pupal wing , where it is proposed to play an in instructive role , either directly or indirectly , in organizing the multi-layered aECM [73] . Similarly , in mice , TTSPs such as Matriptase [109 , 110] and TMPRSS11f [111] are essential for the formation and/or function of an extracellular barrier in the mammalian cornified envelope . The cornified envelope is a complex meshwork of terminally differentiated non-living keratinocytes , called corneocytes , that are chemically cross-linked with a heterogeneous array of structural proteins and lipids . Together , these structures form a water-impermeable barrier preventing unregulated inside-out and outside-in passage . The cornified envelope serves the same physiological function as the extensively chemically cross-linked and hydrophobic insect cuticle . The murine Matriptase homolog is thought to play a role in the secretion of extracellular matrix material needed to maintain epidermal barrier integrity [110] . Our work implicates Lint-mediated lumenal matrix assembly in the integrity of seamless tubes . Interestingly , Matriptase has been implicated in neural tube closure and angiogenesis [112] , suggesting a possible conserved role for membrane-anchored serine proteases in tubulogenesis . To generate GFP-labeled mitotic clones in the tracheal system , the following stocks were used: ywFLP122; btl>GFP; FRT82B tubGAL80 and ywFLP122; btl>GFP , Wkd-mKate2; FRT82B P{tubGAL80} and ywFLP122; btl>moeABD-GFP; FRT82B P{tubGAL80} and ywFLP122; FRTG13 P{tubGAL80}; btl>CD8-GFP to generate MARCM clones ( Lee and Luo , 2001 ) or ywFLP122; btl>GFP , DsRednls;FRT82B UAS-GFPi ( Ghabrial et al , 2011 ) ; btl>GFP; FRT2A FRT82B and btl>GFP; FRT82B kkvsob stocks are described in [67] . An FRT82B ich206 sr e ca/TM3 stock was generated by recombining an FRT82B ich206 chromosome [67] with an isogenic FRT82B cu sr e ca chromosome . FRT82B ich543/TM2 was recovered from a non-complementation screen ( see below ) . UAS-ChtVisT-TdTomato was a generous gift from Paul Adler’s lab [68] . w; Df ( 3R ) osk was a generous gift from Paul MacDonald’s lab ( U . Texas at Austin ) [70] and was recombined onto the FRT2A FRT82B chromosome . The P{PZ}l ( 3 ) 05652 insertion was provided by the Bloomington Stock Center ( Bloomington , IN , USA ) and was recombined onto the FRT2A FRT82B chromosome . For RNA interference experiments , the following strains [113] were obtained through the Bloomington Stock center: y1 sc* v1; P{y[+t7 . 7] v[+t1 . 8] = TRiP . HMS02762}attP2 carrying a dsRNA against ich under UAS control; y1 sc* v1; P{y[+t7 . 7] v[+t1 . 8] = TRiP . HMJ22360}attP40 and y1 sc* v1; P{y[+t7 . 7] v[+t1 . 8] = TRiP . HMC04037}attP40 , expressing independent dsRNAs against lint . ru1 th1 st1 Df ( 3R ) 3-4/TM3 was provided by the Bloomington Stock center . A recessive lethal transgene insertion at lint , MI04680 , was obtained from Bloomington Stock Center and recombined onto a FRT40AFRTG13 chromosome provided by Bloomington Stock Center . The following GAL4 drivers were used: btl-GAL4 [71] , SRF-GAL4 ( a gift from Mark Metzstein , U . Utah , Salt Lake City , UT , USA ) , and drumstick-GAL4 [114] ( Bloomington Stock Center ) . For making germline clones of ich , the w*; FRT82B P{ovoD1-18}/st1 βtub85DD ss1 es/TM3 [115] was obtained from Bloomington Stock Center . An isogenic w1118 wild-type strain was obtained from Bloomington Stock Center . Positively marked clones were generated in the tracheal system as previously described ( Ghabrial et al , 2011 ) . Briefly , embryos were collected for 4h at 25C , then heat-shocked 1h at 38°C to induce FLPase expression . Embryos were aged at 25C until the wandering third instar unless otherwise indicated . Mosaics larvae were identified using a Leica MZ16F fluorescence stereomicroscope . Germline clones were generated using the dominant female-sterile technique [115]: briefly , hsp-FLP122; FRT82B P{ovoD1-18}/FRT82B ich206 sr e ca second and third instar larvae were heat-shocked twice for 2h at 38°C , over the course of two days . Females with mosaic germlines were crossed to FRT82B ich206/TM3 twi>GFP males , embryos were collected and maternal/zygotic null animals identified by lack of GFP expression . To facilitate mapping of the ich locus , ethylmethanesulfonate ( EMS ) -induced mutations were screened for non-complementation of the original ich206 recessive lethal allele [67] . ~50 Males from an isogenic btl>GFP; FRT2A FRT82B strain were fed 25 mM EMS in a sucrose solution overnight . These males were then mated to ~100 btl>GFP; Pr hs-Hid/TM3 P{tubGAL80} females [67] . Eggs were collected over the course of several days , then heat-shocked to kill any hs-Hid animals . Approximately ~1000 mutagenized FRT2A FRT82B chromosomes were then screened for the presence of a recessive lethal mutation ( s ) that failed to complement the tester chromosome carrying the ich206 allele . One additional allele , ich543 , was recovered . This allele was also for complementation with Df ( 3R ) osk , which uncovers ich , and with P{PZ}l ( 3 ) 05652 , a hypomorphic P-element induced allele of ich . The ich543 allele failed to complement all of these independently derived ich-deficient chromosomes . FRT82B kkvsob483/TM3 twi>GFP , FRT82B ich206/TM3 twi>GFP , or FRT82B ich543/TM3 twi>GFP adults were inter-crossed and eggs collected on apple juice agar plates for 6h at 25°C . Eggs were aged further for 16h at 29°C . Hatched first instar heterozygous control larvae were picked and placed in 1:1 methanol: heptane . Mutant embryos were sorted under a fluorescence stereomicroscope by the absence of GFP expression and dechorionated in 50% bleach for 1 . 5 min . Embryos were devitellinated in 1:1 methanol: heptane and washed 3–4 times in 100% methanol . Methanol was replaced with 0 . 1% PBS-Tween-20 ( PBS-Tw ) and embryos/larvae were allowed to settle for 10min at room temperature . Cuticles were expanded in PBS-Tw at 65°C for 20 min . PBS-Tw was replaced by 50 μl Hoyer’s medium . Embryos/larvae were incubated in a 2:1:1 mixture of Hoyer’s mountant: lactic acid: dH2O for 16-24h at room temperature and were then mounted on slides . Preparations were visualized by phase contrast using an Evos FL Auto Imaging microscope . The entire coding sequences of CG11966 ( RE65372 ) and CG8213 ( LD43328 ) were obtained as cDNAs from the Drosophila Genomics Resource Center ( DGRC , Bloomington , IN , USA , NIH grant 2P40OD010949 ) . To generate pUAST-ich , EcoRI and KpnI restriction sites were added during PCR amplification ( see S1 Table ) . The amplicon was directionally subcloned into the pUAST vector ( Brand and Perrimon , 1993 ) . To generate pUAST-CG8213 , the full-length cDNA insert was subcloned into pUAST at the EcoRI and XhoI restriction sites . To generate Flag-tagged Ich , the EcoRI site , start codon , and Flag epitope were added to the forward primer , and a KpnI site was included in the reverse primer ( see S1 Table ) . The Ich-Flag PCR product was TA cloned ( TOPO-TA cloning kit , Invitrogen ) and then subloned into pUAST ( EcoRI , KpnI ) . Sanger sequencing was used to verify the final constructs ( S1 Table ) . Overlap extension PCR [117] was used to generate the Flag-VP16-IchDBD and EnR-IchDBD-Flag fusions ( for primers , see S1 Table ) . The following plasmid templates were used in the PCR strategy: pActPL-VP16AD ( Addgene #15305 ) [78] , en cDNA clone LD16125 ( DGRC ) , and ich cDNA clone RE65372 ( DGRC ) . EcoRI and KpnI restriction sites were added during PCR . PCR amplicons of the expected size were TA cloned ( TOPO TA kit , Invitrogen ) and subsequently subcloned into pUAST ( EcoRI , KpnI ) . A start codon , flag sequence , and SV40 nuclear targeting signal were added to the VP16 transcriptional activation domain during PCR . An SV40 nuclear targeting sequence , followed by a flag sequence and stop codon were added to IchDBD during PCR for the EnR-IchDBD-Flag fusion . Transgenic strains for UAS-ich , UAS-Flag-ich , UAS-Flag-VP16-IchDBD , UAS-EnR-IchDBD-Flag , and UAS-CG8213 were generated by P-element transformation ( Rainbow Transgenic Services , Camarillo , CA , USA ) . Optimal gRNA target sites in the Drosophila melanogaster genome ( release 6 ) were selected using the default settings of the flyCRISPR Optimal Target Finder web program ( http://tools . flycrispr . molbio . wisc . edu/targetFinder/ ) . Genomic DNA from a w1118; FRT40A FRTG13 stock ( Bloomington Stock Center ) was isolated using a Qiagen DNeasy Blood and Tissue kit . Genomic DNA flanking candidate gRNA sites was amplified and sequenced to rule out gRNA sites containing single-nucleotide polymorphisms . A site upstream of the transcription start site ( GCCATGGACACCAACTGATTCGG ) and a site within the 3’ UTR ( GCATTTCAAACGACATTCGCCGG ) common to all spliceoforms were chosen . 5’ phosphorylated oligonucleotides were synthesized by Integrated DNA technologies ( Coralville , IA , USA ) according to [89] . 5’ gRNA plasmids were designed according to [89] . Single colonies of pU6-BbsI-gRNA [89] transformants were selected for plasmid isolation ( GenElute Minipreparations , Sigma Millipore ) and validated by sequencing . A y1 M{nos-Cas9}; FRT40AFRTG13 stock was constructed using the y1 M{nos-Cas9} stock [118] ( Bloomington Stock Center ) expressing the Cas9 nuclease from a germline promoter . y1 M{nos-Cas9}; FRT40AFRTG13 embryos were injected with a mixture of two gRNA plasmids by Best Gene , Inc . ( Chino Hills , CA , USA ) . Single GO’s were outcrossed to al1 dpov1 b1 p1 Bl1 c1 px1 sp1/SM1 ( Bloomington Stock Center ) flies and single G1’s were back-crossed to establish individual lines , which were tested by genetic complementation against multiple deficiencies uncovering CG8213 , including Df ( 2R ) Np5 In ( 2LR ) w45-32n and Df ( 2R ) H3E1 ( Bloomington Stock Center ) , as well as a recessive lethal Minos insertion at CG8213 ( y1w1; CG8213MI04680 ) . PCR ( see S1 Table for primers ) analysis suggested that the 5’ gRNA site was targeted , while the 3’ gRNA site was left intact . However , the exact nature of the lesion remains unclear due to difficulties in PCR amplifying gDNA flanking the 5’ gRNA site from mutants . Mosaic larvae were identified by direct fluorescence using a Leica MZ16F fluorescence stereomicroscope ( Leica Microsystems ) . To visualize terminal cells in wholemount specimens , third instar larvae were placed in a drop of 60% glycerol in 1X PBS , then heat-killed at 70C for ~12s , and flattened under a coverslip . Larvae were then imaged using direct fluorescence and Brightfield optics using a Leica DM5500 B upright or A Leica DM6000 inverted widefield epifluorescence microscope ( Leica Mirosystems ) . Images were acquired using either a Leica DFC360FX camera or a Hammamatsu Orca-R2 Digital CCD camera ( C10600 , Hamamatsu Photonics ) . Z-stacks were captured and processed by deconvolution using Leica Advanced Fluorescence Application Suite ( Leica Microsystems ) . For most images , a single deconvolved z-slice is shown , except where projected z-stacks were used to capture whole-cell detail . To analyze terminal cell lumen ultrastructure by TEM , SRF>eGFP , ich RNAi first instar larvae were subjected to a high pressure freezing/freeze substitution protocol as described in [63] . SRF>eGFP and btl>GFP larvae were processed as a representative wild-type controls . 45–70 nm-thick sections were imaged at 125 keV using a Hitachi 7200 electron microscope . RNA in situ hybridizations were performed on St . 13-16 embryos according to the protocol described by [119] . Briefly , DIG-labeled RNA probes were synthesized from PCR templates amplified from cDNA clones for full-length ect ( RE01075 ) , CG8213 ( LD43328 ) , osi18 ( RE07882 ) , and osi19 ( RE01054 ) obtained from the DGRC . Primers including T3 and T7 promoters were designed according to [120] . See S1 Table for description of PCR primers . PCR products were resolved by agarose gel electrophoresis and purified using glassmilk ( MP Biomedicals ) . in vitro transcription reactions were performed using T3 ( anti-sense ) and T7 ( sense ) DNA-dependent RNA polymerases ( Promega corp . , Madison , WI ) . RNA was labeled with Digoxigenin ( Roche Applied Science , Indianapolis , IN ) . Quality of in vitro transcription products was assessed by agarose electrophoresis and the RNA precipitated using ethanol . Eggs were collected on apple juice plates for 6-7h at 25C , then aged for ~16h at 18°C . Embryos were manually sorted under a fluorescence stereomicroscope to assess genotype ( TM3 twi>GFP balancer ) . Embryos were dechorionated in 50% bleach for 1 . 5 min and fixed for 25 min in a formaldehyde and heptane mixture . Embryos were devitellinated using a 1:1 heptane:methanol mixture . Embryos were then processed for probe hybridization according to [119] . ~50 ng of DIG-labeled RNA probe was diluted in 100 μl hybridization buffer . RNA probe signal was detected by an Alkaline Phophatase ( AP ) reaction using nitroblue tetrazolium ( Roche Applied Science ) and bromochloro indoyl phosphate ( Roche Applied Science ) in 0 . 1M Tris-Cl , 0 . 1 M NaCl , 0 . 05M MgCl2 in 0 . 1% PBS-Tween-20 , pH 9 . 5 . For CG8213 , ect , and osi19 probes , the AP reaction was developed for 4 . 5h at room temperature , while for osi18 probes , the AP reaction was developed for ~16h at room temperature . Embryos were mounted in 60% glycerol in 1X PBS and imaged using Brightfield microscopy . For cDNA rescue experiments , the significance of categorical frequency data was determined using Fisher’s exact probability tests ( http://vassarstats . net ) . one-sided P values were reported under the assumption that rescue conditions will deviate from mutant conditions in one direction . For frequency data with more than 2 possible outcomes , Chi-Square tests were used .	Biological tubes adopt a variety of shapes to carry out their functions . In addition to multicellular tubes , single epithelial or endothelial cells build unicellular lumens lined by an apical membrane devoid of cell junctions , or seams . Such seamless tubes are highly conserved from invertebrates to vertebrate organs , but the factors regulating their formation and maintenance remain poorly understood in any system . Using a forward genetic approach in the Drosophila tracheal ( respiratory ) system , we have characterized a mutant called ichor , which compromises the integrity and shape of seamless tubes in tracheal terminal cells . We demonstrate that Ichor promotes seamless tube integrity and shape by transcriptionally activating genes required to assemble an extracellular matrix ( cuticle ) lining the lumens of terminal cells . The cuticle has long been thought function as a passive exoskeleton , but this work demonstrates that the cuticle contains signals regulating seamless tube growth and/or maintenance . All tubes are lined by apical extracellular matrices , but their composition , assembly , and functions are poorly understood . By characterizing effectors acting downstream of Ichor , we can systematically identify factors controlling all three processes in a model lumenal matrix .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "chitin", "medicine", "and", "health", "sciences", "cloning", "alleles", "developmental", "biology", "respiratory", "system", "materials", "science", "molecular", "biology", "techniques", "embryos", "trachea", "macromolecules", "cellular", "structures", "and", "organelles...	2018	An Ichor-dependent apical extracellular matrix regulates seamless tube shape and integrity
The vegetative insecticidal proteins ( Vip ) , secreted by many Bacillus thuringiensis strains during their vegetative growth stage , are genetically distinct from known insecticidal crystal proteins ( ICPs ) and represent the second-generation insecticidal toxins . Compared with ICPs , the insecticidal mechanisms of Vip toxins are poorly understood . In particular , there has been no report of a definite receptor of Vip toxins to date . In the present study , we identified the scavenger receptor class C like protein ( Sf-SR-C ) from the Spodoptera frugiperda ( Sf9 ) cells membrane proteins that bind to the biotin labeled Vip3Aa , via the affinity magnetic bead method coupled with HPLC-MS/MS . We then certified Vip3Aa protoxin could interact with Sf-SR-C in vitro and ex vivo . In addition , downregulation of SR-C expression in Sf9 cells and Spodoptera exigua larvae midgut reduced the toxicity of Vip3Aa to them . Coincidently , heterologous expression of Sf-SR-C in transgenic Drosophila midgut significantly enhanced the virulence of Vip3Aa to the Drosophila larvae . Moreover , the complement control protein domain and MAM domain of Sf-SR-C are involved in the interaction with Vip3Aa protoxin . Furthermore , endocytosis of Vip3Aa mediated by Sf-SR-C correlates with its insecticidal activity . Our results confirmed for the first time that Sf-SR-C acts as a receptor for Vip3Aa protoxin and provides an insight into the mode of action of Vip3Aa that will significantly facilitate the study of its insecticidal mechanism and application . Microbial insecticides , as substitutes for chemical pesticides , are alternatives for insect control in crops . Bacillus thuringiensis ( Bt ) is the most extensively used biopesticide worldwide due to its ability to produce insecticidal crystal proteins ( Cry and Cyt toxins ) [1–3] . The classical pore-forming model is the widely accepted mode of action of the three-domain crystal protein ( 3d-Cry ) [1] . A signaling pathway model of the Cry toxin’s action has also been reported [4 , 5] . Recently , Fengjuan et al . showed Cry6Aa could trigger the Caenorhabditis elegans death by necrosis signaling pathway [6] . In spite of differences , all three models agree that binding to host specific receptors is a key step in the process involved in cytotoxicity . Several types of receptors for Cry toxins have been reported , such as aminopeptidase N ( APN ) , the cadherin-like proteins , alkaline phosphatases , and ABC transporter [1 , 7 , 8] . Bt has been used successfully to control many crop pests by transgenic plant or traditional spray approaches , however , many pests are not sensitive to Cry toxins and a number of cases of insect resistance to Cry toxins have been reported as a result of laboratory or field selections [1–3] . Vegetative insecticidal proteins ( Vip ) , which are produced by Bt during its vegetative stages , share no sequence or structural homology with known Cry proteins , and have a wide spectrum of specific insecticidal activity , especially against lepidopteran pests [9] . Vip3 toxins have a different insecticidal process compared with Cry proteins , indicating they are likely to complement and extend the use of Bt insecticidal proteins . A synergistic effect of the toxins in Spodoptera frugiperda , Spodoptera albula , and Spodoptera cosmioides larvae was observed when Vip3Aa and Cry1Ia10 were combined [10] . Moreover , reports showed that transgenic cotton and corn co-expressing Vip3A and Cry1Ab , or Vip3A and Cry1Ac , provided excellent control of target insect species [3 , 11–14] and no cross-resistance between Vip3A and Cry proteins was observed [3 , 11 , 12] . However , compared with Cry toxins , studies on the insecticidal mechanisms of Vip3A are scarce . Lee et al . proposed pore forming as the principal Vip3A mode of action [15] . Our previous work demonstrated that Vip3Aa induces apoptosis in cultured S . frugiperda ( Sf9 ) cells [16] . Recently , Hernandez-Martinez et al . found that Vip3Aa could induce apoptosis in Spodoptera exigua midgut epithelial cells [17] . Reports also showed that Vip3A can not bind to the APN and cadherin-like protein [15] . Instead , it binds to proteins of susceptible insect’s midgut , which are distinct from the known Cry receptors [15 , 18] . So far , almost nothing is known on Vip definite receptors except for their molecular weight . Previously , we and Singh et al . have found Vip3A protoxin has cytotoxicity to S . frugiperda cells ( Sf9 cells and Sf21 cells ) [16 , 19] . Therefore , we speculated that there are receptors for Vip3Aa in Sf9 cells membrane . In this study , to identify the receptors of Vip3Aa protoxin , we analyzed the Sf9 cells membrane proteins that bind to the biotin labeled Vip3Aa , via the affinity magnetic bead method coupled with nano-HPLC electrospray ion trap mass spectrometry ( HPLC-MS/MS ) . We paid more attention to the scavenger receptor class C like protein ( Sf-SR-C ) from the 70 identified proteins due to class C scavenger receptors ( SR-C ) are membrane proteins and have only been identified in insects [20] . We investigated whether Sf-SR-C is the receptor of Vip3Aa both in vitro and ex vivo . Furthermore , we detected which domain of Sf-SR-C participates in the interaction with Vip3Aa , and validate whether Sf-SR-C mediates the internalization of Vip3Aa , since we observed the presence of Vip3Aa in the cytoplasm of Sf9 cells . Our data confirmed Sf-SR-C acts as the receptor of Vip3Aa , demonstrated the complement control protein ( CCP ) domain and MAM domain of Sf-SR-C interact with Vip3Aa , and further revealed that endocytosis of Vip3Aa mediated by Sf-SR-C correlates with its insecticidal activity . These results will significantly promote the study and application of Vip3Aa . In addition to the significant virulence effect of Vip3Aa to S . frugiperda Sf9 cells [16 , 19] , we also found that Vip3Aa-RFP ( a fusion protein of Vip3Aa protoxin and red fluorescence protein ) could bind to the Sf9 plasma membrane as shown by colocalization with FITC-phalloidin and internalize in endosomes , while the RFP itself could not ( Fig 1A and S1A Fig ) . Thus , to identify the receptors of Vip3Aa protoxin in Sf9 cells , biotin labeled Vip3Aa ( Bio-Vip3Aa ) ( S1B Fig ) was incubated with the extracts of Sf9 cell membrane proteins , immunoprecipitated with Streptavidin Mag Sepharose , and detected by Coomassie brilliant blue staining ( Fig 1B , a ) . The rest of the bands were analyzed by HPLC-MS/MS after the band corresponding to Vip3Aa was excised ( Fig 1B , b ) . Protein sequence database searching of the MS/MS spectra revealed that the bands represented 70 proteins ( S1 Dataset ) , which included 33 ribosomal proteins ( in which ribosomal protein S2 had been reported as an interacting partner protein of Vip3A by Singh et al . [19] ) and 37 other proteins , including Sf-SR-C ( S1C Fig ) . At present , class C scavenger receptors ( SR-C ) have only been identified in insects [20] and only described in Drosophila melanogaster [21 , 22] . First , the SR-C like gene was cloned from the cDNA of Sf9 cells and named as the Sf-SR-C gene ( GenBank accession no . KX925839 ) . We then purified the extracellular sequence of Sf-SR-C ( aa 20–558 ) ( Sf-SR-C-N ) with a glutathione-S-transferase ( GST ) tag ( GST-SR-C-N ) ( Fig 1C ) . A GST pulldown assay demonstrated that GST-SR-C-N could bind to Vip3Aa-Flag , but could not bind to the control protein Cry1Ac ( Fig 1D ) . To assess the binding affinity between Vip3Aa protoxin and GST-SR-C-N , we used a microscale thermophoresis assay ( MST ) in which biomolecular interactions are quantitated by examining the motion of the molecules along a microscopic temperature gradient induced by an infrared laser [23 , 24] . The estimated dissociation constant ( Kd ) was 190 ± 75 nM ( Fig 1E ) . To further test whether full-length Sf-SR-C can interact with Vip3Aa , Sf-SR-C was then transiently expressed in Sf9 cells with a V5 tag after ligation into plasmid pIZT/V5-His ( pIZT-SR-C ) ( Fig 1F ) . Immunoprecipitation analysis using the anti-V5 antibody showed that Vip3Aa-Flag could be co-immunoprecipitated with Sf-SR-C-V5 ( Fig 1G ) . In the control experiment , we could not detect Cry1Ac after it was incubated with the lysate of Sf9 cells transfected with pIZT-SR-C ( Sf9-pIZT-SR-C cells ) . Ligand blotting was used to detect the specific binding of Sf-SR-C to Vip3Aa . As shown in Fig 1H , Vip3Aa-Flag could bind to Sf-SR-C-V5 and excess Vip3Aa ( 200-fold ) competed for Vip3Aa-Flag binding with Sf-SR-C-V5 , which further indicated that Vip3Aa and Sf-SR-C can bind specifically . In addition , via the affinity magnetic bead method , immunoblotting revealed that the Sf-SR-C-V5 from the lysate of Sf9-pIZT-SR-C cells could interact with biotin-labeled Vip3Aa-Flag ( S1D Fig ) . In contrast , Sf-SR-C-V5 could not interact with control biotin labeled ChiB-flag ( Chitinase B secreted by Bt ) . These results indicated Vip3Aa protoxin can interact with Sf-SR-C in vitro . To verify the role of Sf-SR-C in Vip3Aa protoxin binding to Sf9 cells in more detail , we attempted to generate Sf9 cells in which the expression of endogenous Sf-SR-C gene was reduced . Two plasmids , pIZT-SRi1 and pIZT-SRi2 , which can generate fragments of double-stranded RNA ( dsRNA ) from the Sf-SR-C gene ( S2A and S2B Fig ) [25] , were stably transfected into Sf9 cells , respectively , which resulted in the generation of Sf-SRi1 and Sf-SRi2 cell lines . As the quantitative real-time reverse transcription PCR ( qRT-PCR ) result shown in Fig 2A; the expression level of the Sf-SR-C gene was markedly reduced in the Sf-SRi1 and Sf-SRi2 cells compared with the Sf9 cells and the cells stably transfected with pIZT/V5-His ( Sf-pIZT cells ) . Consistent with this , a CCK-8 cytotoxicity assay results showed that the cytotoxic effects of Vip3Aa on the Sf-SRi1 and Sf-SRi2 cells were also clearly reduced compared with those on Sf9 cells and Sf-pIZT cells ( Fig 2B ) . Next , we carried out co-localization assays to detect the interaction between Vip3Aa and Sf-SR-C . After treating the Sf9 cells with Vip3Aa-RFP for 6 h , we monitored the Vip3Aa and Sf-SR-C distribution using immunofluorescent staining . The anti-Sf-SR-C-N polyclonal antibody and Alexa Fluor 488-conjugated anti-rabbit antibody were used to show the location of Sf-SR-C in Sf9 cells . As shown in Fig 2C , most of the dots of Vip3Aa-RFP were co-located with Sf-SR-C , especially in the dots that were Sf-SR-C-rich . In the control experiment that the anti-GST polyclonal antibody was used , we detected almost no green fluorescence . We also observed that Vip3Aa-RFP has almost no affinity for Drosophila S2 cells ( S2 cells ) ( S3A Fig ) . We therefore cloned the gene of Sf-SR-C into plasmid pAc5 . 1/V5-HisB ( pAc-Sf-SR-C ) and transiently transfected it into S2 cells ( S2-Sf-SR-C cells ) to examine the specific interaction of Vip3Aa and Sf-SR-C in S2 cells . The ribosomal S2 protein of Sf9 cells ( Sf-S2 ) was also heterologously expressed into the S2 cells ( S2-Sf-S2 cells ) as a control ( S3B Fig ) . The Dylight 488 conjugated anti-V5 antibody was used to show the heterologously expressed protein in S2 cells . After treating the S2 cells with Vip3Aa-RFP for 12 h , immunofluorescent staining showed that Vip3Aa-RFP could clearly bind to the S2-Sf-SR-C cells , and the dots of Vip3Aa-RFP were co-located with the dots that were rich in Sf-SR-C ( Fig 2D and S3C Fig ) , similar to the phenomenon that Vip3Aa-RFP binds to Sf-SR-C in Sf9 cells . In contrast , we didn’t detect the interaction between Vip3Aa-RFP and S2-Sf-S2 cells , nor did we find the binding of RFP to S2-Sf-SR-C cells . In addition , the cytotoxicity assay showed Vip3Aa protoxin is more toxic to S2-Sf-SR-C cells than to S2-Sf-S2 and S2 cells ( Fig 2E ) ( The transfection efficiency was about 30% ) . Taken together , these results revealed that Sf-SR-C could also interact with Vip3Aa protoxin ex vivo . Vip3Aa has a high affinity for IOZCAS-Spex-II-A cells ( Spodoptera exigua cells ) ( S3A Fig ) and shows a significant toxic effect to S . exigua [10] . We also cloned two partial sequences with similarity to the Sf-SR-C gene from the total cDNA of S . exigua cells ( GenBank accession no . KY829113 and MF969248 ) . Therefore , we attempted to use ingestion of bacterially expressed dsRNA to reduce the expression of the S . exigua larvae midgut SR-C gene ( Se-SR-C ) to detect whether it affected the toxicity of Vip3Aa to the larvae . The vector pET-Se-SRi , which expresses a partial dsRNA of the Se-SR-C gene ( S2C Fig ) , was transformed into bacterial strain HT115 ( DE3 ) , which lacks RNase III activity to express dsRNA fragment of Se-SR-C ( HT-pET-Se-SRi ) [26] ( S4A Fig ) . The vector pET-Hypi , which expresses a partial dsRNA of a hypothetical protein ( Hyp ) ( GenBank: PCG66164 . 1 ) , and the blank plasmid pET28a were transformed into the HT115 strain as control ( HT-pET-Hypi and HT-pET28a ) . The qRT-PCR results showed that after continuous feeding of the S . exigua larvae with the strains for 7 days ( Fig 3A and S4B Fig ) , the transcription level of the Se-SR-C gene of the larvae fed with the HT-pET-Se-SRi strain was effectively inhibited compared with the control ( Fig 3B ) . The larvae were then exposed to Vip3Aa and the strains for another 5 d to detect the survival rate . The bioassay results shown in Fig 3C indicated that the mortality rate of the larvae in which the Se-SR-C gene was silenced was clearly lower than that of the control , which suggested that reducing the expression of the Se-SR-C gene in S . exigua lavae decrease their sensitivity to Vip3Aa . Vip3A has high insecticidal activity against Lepidopteran rather than Dipteran [9] . To further examine the interaction of Sf-SR-C with Vip3Aa in an insect that is insensitive to Vip3Aa ( S4C Fig ) , we constructed transgenic Drosophila that overexpressed Sf-SR-C using the esg-Gal4 tub-Gal80ts system . In this system , the esg-Gal4 driver is mainly active in the midgut cells of Drosophila and Gal4 is under the control of a temperature sensitive Gal80 that allows the conditional induction of the UAS-linked Sf-SR-C gene [27] . After culturing at 25 °C for 4 d , the fly strains were shifted to 29 °C ( Gal4 ‘‘on” ) or 18 °C ( Gal4 ‘‘off” ) for 3 d ( Fig 3D ) . The about 2-day-old larvae were then treated with Vip3Aa or dialysis buffer for 48 h and the survival rates were detected ( Fig 3D and S4D Fig ) . As shown in Fig 3E , the Drosophila larvae that overexpressed Sf-SR-C in their midgut ( esgts>SR-Cvk33 ) ( Red ) had a significantly higher mortality rate after exposure to Vip3Aa compared with the control group , which was treated with dialysis buffer . In the group of esgts ( Green ) and UAS-SR-Cvk33 ( Purple ) , which could not express Sf-SR-C , Vip3Aa showed no obvious toxicity to the larvae compared with the control . Moreover , shutdown of the expression of Sf-SR-C in the gut epithelia of the larvae ( 18 °C treated ) eliminated the toxicity of Vip3Aa to the larvae ( Blue ) . These results further indicated Sf-SR-C is the receptor for Vip3Aa , which causes the death of sensitive insects . From BLASTP analysis , we found that the protein sequence of Sf-SR-C was not similar to the SR-C from D . melanogaster ( dSR-CI ) ( only about 27% sequence identity ) . However , the extracellular sequence of Sf-SR-C has four structural domains that are similar to dSR-CI , including the CCP , MAM , somatomedin B , and Ser/Thr rich domains ( Fig 4A ) . To further investigate which domain of Sf-SR-C mainly participates in the interaction with Vip3Aa protoxin , the extracellular sequence of Sf-SR-C was divided into three parts ( SR-F-1 , SR-F-2 , and SR-F-3 ) ( Fig 4A ) and expressed as fusion proteins with GST . Dot blotting analysis revealed that GST-SR-F-1 ( aa 20–138 ) , which contains the CCP domain ( aa 26–76 ) , and GST-SR-F-2 ( aa 139–320 ) , which is the MAM domain , could bind to Vip3Aa-Flag , while GST and GST-SR-F-3 ( aa 321–558 ) could not ( Fig 4B ) . Furthermore , excess Vip3Aa ( 500-fold ) competed for Vip3Aa-Flag binding with GST-SR-F-1 and GST-SR-F-2 ( Fig 4B ) . Moreover , pulldown experiments also revealed that GST-SR-F-1 and GST-SR-F-2 could directly interact with Vip3Aa-Flag ( Fig 4C ) . These results indicated that the binding of SR-F-1 and SR-F-2 with Vip3Aa-Flag was specific . GST-SR-F-1 contained regions other than the CCP domain; therefore , we further purified Sf-CCP ( CCP domain of Sf-SR-C ( aa 20–80 ) with a His-tag ) to detect the interaction with Vip3Aa-Flag , and the Dm-CCP ( CCP domain ( aa 20–80 ) of dSR-CI ( GenBank: U17693 . 1 ) ) was used as a control . Both dot blotting analysis ( Fig 4D ) and pulldown assays ( Fig 4E ) verified the physical interaction between Sf-CCP and Vip3Aa-Flag . The results also showed that the Dm-CCP domain could not bind to Vip3Aa-Flag , which further validated the specific binding between Vip3Aa and Sf-SR-C . Furthermore , MST was also applied to assay the binding affinity of Vip3Aa protoxin with Sf-CCP and MAM domains ( Fig 4F and 4G ) . The determined Kd values were 2 . 19 ± 1 . 55 μM and 463 ± 117 nM , respectively . These results certified the CCP and MAM domains of Sf-SR-C could bind to Vip3Aa protoxin . As Figs 1A and 2C showed above , we observed red dots in the cytoplasm of Sf9 cells after exposing them to Vip3Aa-RFP , which suggested the internalization of Vip3Aa . We first used several inhibitors of endocytosis to test whether Vip3Aa-RFP could enter the Sf9 cells via endocytosis [28 , 29] . As shown in Fig 5A and 5B , dynasore , which is an inhibitor of dynamin , could significantly inhibit Vip3Aa-RFP entry into Sf9 cells . The known macropinocytosis inhibitors , amiloride , cytochalasin D , LY294002 , and wortmannin also reduce the number of red dots inside the Sf9 cells . However , among two inhibitors of clathrin-mediated endocytosis ( chlorpromazine and monodansylcadaverine ) and two inhibitors of clathrin-independent endocytosis ( nystatin and cholesterol-oxidase ) , only monodansylcadaverine could reduce the number of red dots in Sf9 cells; the others had no effect on the number of Vip3Aa-RFP dots in the Sf9 cells compared with the control . These results suggested Vip3Aa enter Sf9 cells through dynamin-dependent and macropinocytosis-related endocytosis . One of the main functions of scavenger receptors ( SRs ) is endocytosis [20] . Thus , we hypothesized that Vip3Aa enters the Sf9 cells via endocytosis mediated by Sf-SR-C . To further verify Sf-SR-C mediated the internalization of Vip3Aa , purified anti-Sf-SR-C-N polyclonal antibodies were incubated with the Sf9 cells for 1 h and the cells were then co-incubated with Vip3Aa-RFP for another 6 h . The results showed that the number of red dots in the cytoplasm of Sf9 cells was reduced visibly after treatment with the anti-Sf-SR-C-N polyclonal antibody , while there are many red dots in cells treated with anti-GST polyclonal antibodies ( Fig 6A and 6D ) . We also quantified the number of red dots in the Sf-SRi1 and Sf-SRi2 cell lines . Compared with the Sf-pIZT cells , the internalization of Vip3Aa-RFP also reduced markedly in the Sf-SRi1 and Sf-SRi2 cells ( Fig 6C and 6D ) . Furthermore , because we found that Vip3Aa can bind to GST-SR-F-1 and GST-SR-F-2 , Vip3Aa-RFP combined with an excess of GST-SR-F-1 and GST-SR-F-2 ( 20-fold ) were exposed to Sf9 cells , respectively . The competitive binding assay showed the amount of red dots in the Sf9 cells treated by Vip3Aa-RFP and GST-SR-F-2 was significantly decreased compared with the control cells treated with Vip3Aa-RFP and GST ( Fig 6B and 6D ) . However , in the case of GST-SR-F-1 , such phenomenon did not occur , which suggested the MAM domain might play more critical role in the internalization of Vip3Aa than the CCP domain . These results indicated the Sf-SR-C mediates the internalization of Vip3Aa via endocytosis . The above results showed that silencing of Sf-SR-C gene could clearly reduce the toxicity of Vip3Aa to Sf9 cells ( Fig 2B ) and also reduce the amount of Vip3Aa entering into Sf9 cells ( Fig 6B and 6D ) , which suggested the amount of Vip3Aa entering cells is directly related to its toxicity . Therefore , we carried out further experiments to verify this speculation . Firstly , we have demonstrated that the endocytosis inhibitor dynasore could significantly inhibit the internalization of Vip3Aa , without affecting the binding of Vip3Aa to Sf9 cells ( Fig 5A and 5B and S5 Fig ) . Through cytotoxicity assay ( Fig 7A ) , we further found that dynasore ( 4μM ) markedly decreased the toxicity of Vip3Aa to Sf9 cells while reducing the entry of Vip3Aa into cells . Dynasore alone did not affect the survival of Sf9 cells . In addition , in the Fig 5A and 5B , we found that the DMSO ( 0 . 1% ) had a tendency to increase the number of Vip3Aa into Sf9 cells . So we explored the highest concentration of DMSO that did not cause toxicity to Sf9 cells . As shown in Fig 7B and 7C , we found that when the concentration of DMSO increased to 0 . 6% ( v/v ) , it could clearly increase the number of Vip3Aa entering Sf9 cells . Moreover , the cytotoxicity assay also showed that DMSO increased Vip3Aa's toxicity to Sf9 cells while promote the internalization of Vip3Aa ( Fig 7D ) , and DMSO alone had no obvious toxicity to Sf9 cells . These results further demonstrated that the internalization of Vip3Aa is directly related to its toxicity . Taken together , our results indicated the induced mortality of Vip3Aa in Sf9 cells correlated with its endocytosis mediated by Sf-SR-C . Vip3Aa proteins have been studied for more than 20 years since they were first found by Estruch et al . in 1996 [30] . They are considered as novel insecticidal toxins secreted by Bt because they have different insecticidal properties compared with known Cry toxins and have a broad insecticidal spectrum within Lepidoptera [9] . To date , more than 138 Vip proteins have been found and were divided into four categories according to the classification of Bt Toxin Nomenclature Committee [31] . However , there has been no report of a definite receptor for Vip toxins up to now . In this paper , via HPLC-MS/MS , 70 potential binding proteins of Vip3Aa , including ribosomal protein S2 and actin , were identified ( S1 Dataset ) . Singh et al . identified ribosomal protein S2 as a toxicity-mediating interacting partner protein of Vip3A in Sf21 cells [19] . However , as an intracellular protein , S2 protein is not likely to be a receptor of Vip3A . That maybe why Singh et al named it interacting partner protein , not a receptor . Our results also showed Vip3Aa could not bind to the S2-Sf-S2 cells , which heterologously expressed Sf-S2 into the S2 cells , and had no obvious cytotoxicity to them ( Fig 2D and 2E ) . It suggests that S2 protein is not a receptor for Vip3Aa . Actin was identified as a novel Cry1Ac binding protein in Manduca sexta midgut through proteomic analysis [32] . For the same reason , it is unlikely that this protein is serving as a receptor for Vip3Aa . We speculate that the Vip3Aa may interfere with the function of the ribosome and actin after entering the cells . Estruch et al . mentioned that a 48-kDa protein from Agrotis ipsilon with homology to tenascins may act as the receptor of Vip3A in their patent [33] . However , they did not provide any supporting data and no subsequent reports proved their speculation . Furthermore , consistent with previous reports [9] , we did not find the receptors for Cry toxins such as APN or cadherin-like proteins in the 70 proteins we identified , which suggested Vip3Aa share no binding sites with Cry toxins . Moreover , we also demonstrated that Cry1Ac could not bind to Sf-SR-C ( Fig 1D and 1G ) . These results further strengthen the viewpoint that Vip3 toxins and Cry toxins have different mechanisms of action , which makes Vip3 toxins good candidates for combination with Cry toxins in transgenic plants to prevent or delay insect resistance and to broaden the insecticidal spectrum . In this study , we provide in vitro , ex vivo , and bioassay evidences for the first time confirming that the SR-C-like protein Sf-SR-C from Sf9 cells is the receptor for Vip3Aa . Scavenger receptors are cell surface receptors that typically bind multiple ligands and promote the removal of non-self or altered-self targets . SRs are classified into 10 classes [20] . At present , the vast majority of SRs have been identified and studied in mammals; however , SR-C have only been found in insects and have only been described in Drosophila [21 , 22] . In mammalian cells , SRs play a crucial role in maintenance of host homeostasis , and also participate in host immune responses and metabolism . They can recognize and bind to a broad spectrum of ligands , including modified and unmodified host-derived molecules or microbial components [21 , 34] . However , researchers also found that pathogens have evolved mechanisms to subvert SRs’ function to infect host cells [34] . For example , hepatitis C virus [35] , enterovirus 71 ( EV71 ) [36] , and coxsackievirus ( CVA7 , CVA14 and CVA16 ) [37] utilize class B receptors to infect host cells . Chlamydia trachomatis uses the lipid transfer activity of SR-B1 for survival in host cells [38] . Even the class B scavenger receptor CD36 , which has been implicated in the clearance of several bacterial and protozoan pathogens , has been reported to be co-opted by mycobacteria [39] . In D . melanogaster , SR-CI was identified as the recognition receptor for acetylated low-density lipoprotein [22] and bacteria [40] . In addition , Philips et al . found that Peste in D . melanogaster ( a CD36 homolog ) is involved in the uptake of mycobacteria into host cells [41] . In this study , we provide another example , in which the bacterial toxin Vip3Aa can exploit Sf-SR-C of Sf9 cells to kill host cells . In addition , the Vip1 and Vip2 proteins which were first found in Bacillus cereus are regarded as binary toxins . Vip1 is speculated as the binding component and triggers endocytosis , and Vip2 enters the cell and exerts its toxic effect [9] . Vip3Aa has no sequences similarity to Vip1 or Vip2 , however , our results certified Vip3Aa can entry into the Sf9 cells by itself via the endocytosis mediated by Sf-SR-C . In insects , there has been little research into the mode of action of SR-C . However , in mammals , one of the main functions of SR proteins is endocytosis , which can trigger a series of signaling pathways [21 , 34] . More interestingly , SR function is increasingly linked to apoptosis in a wide variety of cell types . Binding of fucoidan ligand by the macrophage SR-A1 triggers endocytosis by caveolae-dependent pathways , which stimulates apoptosis via a p38 MAPK and JNK dependent intracellular signaling pathway [42] . In vascular cells , thrombospondin-1 activation of SR-B2 triggers downstream signaling through p38 MAPK and caspase dependent pathways with increased apoptosis [43] . In addition , SR-E1 function is linked to apoptosis in the endothelium , vascular smooth muscle cells , macrophages , epithelial cells and neurons . [44 , 45] . As mentioned above , some pathogens can utilize the function of SRs to invade the host cell . Some toxins secreted by bacteria can also entry into host cell via endocytosis to exert their toxic effects . Diphtheria toxin , an exotoxin secreted by Corynebacterium diphtheriae and causes the disease diphtheria in humans , is believed to enter toxin-sensitive mammalian cells by receptor-mediated endocytosis and inhibit protein synthesis of host cells [46 , 47] . In this way , it acts as a RNA translational inhibitor and results into cell apoptosis . Receptor-mediated endocytosis is required for efficient expression of toxicity . Once endocytosis was inhibited , the cytotoxicity of diphtheria toxin was blocked accordingly [46 , 47] . Our results indicated that the toxicity of Vip3Aa to Sf9 cell correlated with its endocytosis mediated by Sf-SR-C ( Figs 5 , 6 and 7 ) . It suggested that internalization is essential for Vip3Aa to exert its toxic effects . Endocytosis is mentioned in Cry5- Caenorhabditis elegans system . In that case , however , endocytosis is a protection strategy utilized by worms to against the toxin Cry5 [48] . As to the “signal transduction” model , endocytosis is not an indispensable step [4] . In newly found “necrosis” model , Cry6A toxin is also internalized into intestinal cells , but cell death induced by Cry6Aa does not depend on the apoptotic mechanism [6] . These further implied the mode of action of Vip3Aa toxins different from that of Cry toxins . However , the more detailed mechanisms of how Vip3Aa kills Sf9 cells after interacting with Sf-SR-C and the follow-up connection with our previous results that Vip3Aa can induce apoptosis in Sf9 cells [16] , are complex and interesting and will require further study . In cytotoxicity assays , we used Vip3Aa protoxin . However , we found that the purified Vip3Aa protoxin was unstable . As shown in S1B Fig , in the biotin labeled Vip3Aa ( lane 3 ) , we can see the emergence of the activated Vip3Aa like protein ( about 66 kDa ) . After incubating Vip3Aa protoxin with Sf9 cells , western blotting revealed that the activated Vip3Aa like protein was also apparent in the medium ( S6 Fig ) . Therefore , it is difficult to exclude the existence of activated Vip3Aa in the process of toxicity testing . Lee et al . have already demonstrated that the activated Vip3Aa has the pore formation activity . In contrast , the full-length Vip3Aa protein was unable to form pores [15] . They proposed formation of ion channels as the principal mode of action of activated Vip3Aa . However , in this work , we have demonstrated that full-length Vip3Aa could bind to the Sf-SR-C receptor and endocytosis of Vip3Aa correlates with its toxicity . In addition , the activated Vip3Aa protein was considered to correspond to the C terminus of the Vip3Aa protoxin ( from amino acid 199 to the end ) [9 , 33] . We also purified the activated Vip3Aa protein ( Vip3Aa-199 ) ( S7A Fig ) . Through cytotoxicity assay , we found that although the activated Vip3Aa is also toxic to Sf9 cells , it is obviously less than that of Vip3Aa protoxin ( S7B Fig ) . So we think that the Vip3Aa protoxin plays a major toxic role in cytotoxicity assays . Furthermore , we found that endocytosis of Vip3A was almost completely inhibited after treated with dynasore ( Fig 5B ) . Meanwhile , the toxicity of Vip3Aa was decreased clearly but not reduced accordingly ( Fig 7A ) . It implied that endocytosis is critical for Vip3Aa to exert its toxic effects , but it may not be responsible for all the toxicity of Vip3Aa . Recently , Tabashnik et al . proposed a new model for Bt mode of action named “dual model” , where both the protoxin and activated Cry toxin forms can kill insects , with each form exerting its toxic effect via a different pathway [49] . This contrasts with what “classical model” in which protoxins are inactive . Whether Vip3Aa protoxin and the activated toxin use the different mechanism of action and whether Vip3A have other mechanisms of insecticide need further study . To date , SR-C has only been described in Drosophila . The present study cloned and identified another SR-C gene in S . frugiperda . Moreover , we also cloned two other fragments from S . exigua cells , which have high sequence and structural homology with Sf-SR-C . This indicated that SR-C also exists in S . exigua and further extends the range of SR-C in insects . Our results showed that SR-C can be detected in Sf9 cells and S . exigua cells , as well as in Drosophila cells . However , only the former two types of cells have high affinity for Vip3Aa ( Fig 1A and S3A Fig ) . We hypothesized that subtle differences in the sequence and the three-dimensional structure of the protein might influence their interaction with Vip3Aa . Consistent with our conjecture , Vip3Aa-Flag can bind to Sf-CCP but not to DM-CCP ( Fig 4D and 4E ) . Furthermore , from the sequence alignments , we found that Sf-SR-C has no sequence and structural homology with known proteins from vertebrates , as well as with known Cry toxins receptors . The results presented here provide a plausible molecular basis for the lack of toxicity of Vip3A toxins toward non-target insects and vertebrates , and strongly support its use as a safe biopesticide . In addition , because SRs play a crucial role in innate immunity and in the pathogenesis of various diseases in mammals [34] , our study might extend our understanding of SR-C proteins and provide other avenues for studying host-pathogen interactions . What’s more , although reducing the expression of the Se-SR-C gene clearly reduced the toxicity of Vip3Aa to the larvae compared with that of the control ( Fig 3C ) , the effect sizes between the larvae of Se-SR-C gene silencing and the control groups were not as obvious as expected . This implied that there may be other receptors for Vip3Aa contributing to the overall toxicity . Just like several receptors for Cry toxin have been discovered [1 , 7 , 8] , some reports also found Vip3Aa could bind to different molecular weight proteins in the brush border membrane vesicles of susceptible insects , such as the 55 , 65 , 80 , 100 and 110 kDa proteins [15 , 50–52] , which further indicated the existence of different kinds of receptors for Vip3Aa . Moreover , we have identified 36 other proteins besides the ribosomal proteins and Sf-SR-C from the extracted Sf9 cell membrane proteins which could interact with the Vip3Aa . Whether or not there are other receptors play roles , sequentially or simultaneously , in killing insect process , further in-depth studies are needed . In conclusion , the present study identified and confirmed Sf-SR-C as the receptor for Vip3Aa , proved the CCP and MAM domains of Sf-SR-C interact with Vip3Aa , analyzed the binding specificity between Vip3Aa and Sf-SR-C , and certified Sf-SR-C mediate the internalization of Vip3Aa via endocytosis . Our results contribute to the understanding of the mode of action Vip3Aa and significantly facilitate the further study of its insecticidal mechanism and application . E . coli DH5α for plasmid constructions and E . coli BL21 ( DE3 ) for protein purification were cultured at 37 °C in lysogeny broth ( LB ) or agar . Bt9816C was previously isolated and maintained in our laboratory for generation of Vip3Aa [53] . The Drosophila S2 cells , S . frugiperda Sf9 cells and S . exigua cells ( IOZCAS-Spex-II-A ) were maintained and propagated in Sf-900 II SFM ( Invitrogen ) or SFX-Insect ( HyClone ) culture medium at 27 °C . Spodoptera exigua and Drosophila strains were used for the bioassays . Drosophila genotypes used were: esgts: esg-Gal4 , tub-Gal80ts , UAS-GFP/cyo; Tm2/Tm6B . UAS-SR-Cvk33: SP/Cyo; 10×UAS-SR-Cvk33/Tm6B . esgts>SR-Cvk33: esg-Gal4 , tub-Gal80ts , UAS-GFP/Cyo; UAS-SR-Cvk33/Tm2 . To ectopically express Sf-SR-C in Drosophila , the primers pJF-Sf-SR-C-F and pJF-Sf-SR-C-R were used to clone the Sf-SR-C gene , which was then recombined with the linearized vector pJFRC2-10XUAS-IVS-mCD8::GFP ( Addgene plasmid #26214 ) using a ClonExpress II One Step Cloning Kit ( Vazyme ) . Transgenic lines were established through microinjection of the transgene DNA into embryos of PhiC31-mediated chromosome-integrated Drosophila strains PBac {y[+]-attP-3B} VK00033 [54] . Primary antibodies: Mouse anti-Flag ( Cell Signaling 8146 ) , rabbit anti-V5 ( Cell Signaling 13202 ) , rabbit anti-His ( Cell Signaling 12698 ) , anti-Sf-SR-C-N polyclonal antibodies were generated by immunizing rabbits with purified GST-SR-C-N . Secondary antibodies: goat anti-mouse IgG-HRP conjugate ( Santa Cruz sc-2005 ) , goat anti-rabbit IgG-HRP conjugate ( Cell Signaling 7074 ) , rabbit anti-GST ( Polyclonal , Bioss bs-2735R ) . The primary antibodies and secondary antibodies were used at 1: 1000 for western blotting . For immunostaining assays , Anti-V5-Dylight 488 conjugate ( Invitrogen MA5-15253-D488 , 1:200 ) and Alexa Fluor 488 goat anti-rabbit IgG ( Cell Signaling 4412 , 1:200 ) were used . Inhibitors: Dynasore ( dynamin inhibitor , TargetMol T1848 , 7 . 5μM ) , chlorpromazine ( clathrin-mediated endocytosis inhibitor , Millipore 215921 , 40μM ) , monodansylcadaverine ( clathrin-mediated endocytosis inhibitor , Sigma D4008 , 150μM ) , nystatin ( sequesters cholesterol , Millipore 475914 , 20μM ) , cholesterol-oxidase ( oxidize cholesterol , Millipore 228230 , 4unit/ml ) , amiloride ( macropinocytosis inhibitor , Millipore 129876 , 150μM ) , Cytochalasin D ( macropinocytosis inhibitor , Millipore 250225 , 500nM ) , LY294002 ( broad PI ( 3 ) K inhibitor , Cell Signaling 9901s , 50μM ) , and wortmannin ( broad PI ( 3 ) K inhibitor , Cell Signaling 9951s , 2μM ) . Cells were treated with the inhibitors for 1 h at 27°C before Vip3Aa-RFP was added . For the expression of Vip3Aa protoxin , the vip3Aa gene was cloned in pET-28a ( + ) vector ( Novagen ) using oligonucleotide primer Vip-F and Vip-R ( plasmid pET-Vip ) , resulting in a His6 fusion . For the expression of Vip3Aa-RFP , the vip3Aa gene and rfp gene were amplified using oligonucleotide primer pairs Vip-RFP-Up-F and Vip-RFP-Up-R , and Vip-RFP-Do-F and Vip-RFP-Do-R , respectively . Then the two gene fragments were ligated into the pET-28a ( + ) vector ( Novagen ) using a pEASY-Uni Seamless Cloning and Assembly Kit ( TransGen ) after digesting the vector with NcoI and XhoI ( plasmid pET-Vip-RFP ) , resulting in a His6 fusion . The plasmids used to express Vip3Aa-Flag ( pET-Vip-flag ) was constructed similarly to plasmid pET-Vip-RFP , using oligonucleotide primer Vip-flag-F and Vip-flag-R . The plasmids used to express RFP ( pET-RFP ) , Sf-CCP ( pET-Sf-CCP ) and Dm-CCP ( pET- Dm-CCP ) were constructed similarly to plasmid pET-Vip , using oligonucleotide primer pairs RFP-F and RFP-R , Sf-CCP-F and Sf-CCP-R , and Dm-CCP-F and Dm-CCP-R , respectively . For the expressing of Sf-SR-C-N fused with glutathione-S-transferase ( GST ) , the Sf-SR-C-N gene was amplified using primer SR-C-N-F and SR-C-N-R . The amplification product was inserted into the pGEX-6P-1 ( GE Healthcare ) vector using a pEASY-Uni Seamless Cloning and Assembly Kit ( TransGen ) after digesting the vector with BamHI and XhoI ( plasmid pGEX-SR-C-N ) . The plasmids used to express SR-F-1 ( pGEX-SR-F-1 ) , SR-F-2 ( pGEX-SR-F-2 ) , and SR-F-3 ( pGEX-SR-F-3 ) were constructed similarly to plasmid pGEX-SR-C-N , using oligonucleotide primer pairs SR-F-1-F and SR-F-1-R , SR-F-2-F and SR-F-2-R , and SR-F-3-F and SR-F-3-R , respectively . Plasmids were transformed into E . coli BL21 ( DE3 ) ( Invitrogen ) for expression and purification [55] . His-tagged proteins was purified by using cOmplete His-Tag Purification Resin ( Roche ) , whereas GST- tagged proteins was purified by using GST-Sepharose affinity column ( GE Healthcare ) . The purified protein was dialyzed in buffer containing 25 mM Tris-Hcl ( pH 8 . 0 ) , 150 mM NaCl and 5% glycerol at 4 °C with three buffer changes . The purified Vip3Aa is used for cytotoxicity assays . All the primers and plasmids used in this study are shown in S1 and S2 Tables . MST was used to determine the binding affinity between Vip3Aa protoxin and Sf-SR-C protein fragments . Briefly , purified proteins were dialyzed into 25 mM Hepes ( pH 7 . 5 ) , 150 mM NaCl , and 0 . 05 ( v/v ) % Tween-20 . The purified Vip3Aa was labeled with the Monolith NT Protein Labeling Kit ( Cat # L008 ) according to the supplied labeling protocol . 10 nM labeled Vip3Aa were incubated with 0 . 3 nM to 10 μM Sf-SR-C protein fragments for 20 min at RT respectively . Samples were then loaded into standard treated capillaries and analyzed with a NanoTemper Monolith NT . 115 Pico ( NanoTemper Technologies GmbH , Munich , Germany ) at 25°C . Furthermore , the laser power was set to 10% and the LED power was set to 60% . Normalization of the fluorescence signal and fitting to the Hill equation were performed using the software MO Affinity Analysis v2 . 2 . 2 ( NanoTemper ) . For each sample , the whole procedure was performed three times to yield independent triplicates . Total RNA was isolated from cultured cells or S . exigua midgut using RNAiso Plus ( Takara ) . cDNA was synthesized using a Transcriptor High Fidelity cDNA Synthesis Kit ( Roche ) . Quantification of the cDNA was carried out using SYBR Premix Ex Taq II ( Takara ) and analyzed by using StepOne software ( Applied Biosystems ) as previously described [55] . The actin gene acted as the endogenous control . The relative amount of cDNA was calculated according to the 2−ΔΔCT method [56] . Data were analyzed from three independent experiments and are shown as means ± SD . Plasmids used for Sf-SR-C gene silencing experiments were constructed as described by Katsuma et al . [25] . Fragments of the Sf-SR-C gene ( nucleotides [nt] 294 to 803 , dsRNA1s ) and 400 bp from the complementary strand of the Sf-SR-C gene ( nt 693 to 294 , dsRNA1as ) were amplified by using the primer sets SRi1-Up-F and SRi1-Up-R ( for dsRNA1s ) or SRi1-Do-F and SRi1-Do-R ( for dsRNA1as ) . dsRNA1s was designed to be 110 bp longer than the dsRNA1as . dsRNA1s and dsRNA1as were inserted in tandem into the pIZT/V5-His vector using a pEASY-Uni Seamless Cloning and Assembly Kit ( TransGen ) after digesting the vector with KpnI and AgeI ( pIZT-SRi1 ) . In the same way , we constructed pIZT-SRi2 using the primer sets SRi2-Up-F and SRi2-Up-R for dsRNA2s ( nt 1081–1590 ) or SRi2-Do-F and SRi2-Do-R for dsRNA2as ( nt 1480–1081 ) . We generated stable Sf-SR-C gene silencing Sf9 cells lines by transfection with pIZT-SRi1 or pIZT-SRi2 using the Cellfectin II reagent ( Invitrogen ) and PLUS Reagent ( Invitrogen ) . At 2 d after transfection , zeocin ( 500 μg/mL ) was added into the medium . Two to three weeks after drug selection , we examined the expression level of the Sf-SR-C gene by qRT-PCR analysis by using the primers SR-RT-F and SR-RT-R . The vectors pIZT-SR-C , pAc-SR-C , and pAc-Sf-S2 , which were used to express the Sf-SR-C or Sf-S2 , were transfected into Sf9 cells or S2 cells to express the targeted proteins using Cellfectin II reagent and PLUS Reagent . The plasmid pET-Se-SRi and pET-Hypi were constructed as the pIZT-SRi1 by using the primer sets pET-SRi-Up-F and pET-SRi-Up-R for dsRNA3s ( nt 1–870 ) , pET-SRi-Do-F and pET-SRi-Do-R for dsRNA3as ( nt 718–1 ) , Hypi-Up-F and Hypi-Up-R for dsRNA4s ( 620bp ) , and Hypi-Do-F and Hypi-Dp-R for dsRNA4as ( 500bp ) . Then the dsRNA3s and dsRNA3as or dsRNA4s and dsRNA4as were inserted in into the pET28a vector . The pET-Se-SRi and pET-Hypi were transformed into the HT115 ( DE3 ) strain , which lacks RNase III activity for dsRNA expression , as described by Tian et al . [26] . The purified Vip3Aa protoxin was labeled with biotin using an EZ-Link Sulfo-NHS-SS-Biotinylation Kit . ( Thermo Scientific ) . The membrane proteins of Sf9 cells were extracted using a ProteoExtract Transmembrane Protein Extraction Kit ( Novagen ) . Streptavidin Mag Sepharose beads ( 50 μL ) ( GE Healthcare ) were washed and incubated with 0 . 2 mg biotin labeled Vip3Aa ( Bio-Vip3Aa ) for 1 h at 4 °C and washed three times with TBS to remove unbound proteins . The Vip3Aa tagged beads were then incubated with 1 mL of extracted Sf9 cell membrane proteins for 3 h at 4 °C and washed five times with washing buffer ( TBS + 2 M urea ) . The precipitants were boiled with SDS loading buffer and subjected to SDS-PAGE . After cutting out the band representing Vip3Aa , the remaining bands were sent for LC-MS/MS ( tandem mass spectroscopy ) analysis . The targeted sample was resolved by SDS-PAGE and transferred onto a Polyvinylidene fluoride ( PVDF ) membrane ( Millipore ) . Primary antibody and HRP-coupled secondary antibody were used to detect the sample . The membrane was visualized using Immobilon Western chemiluminescent HRP Substrate ( Millipore ) . Cells were collected and lysed in 0 . 5 ml radio immunoprecipitation assay buffer ( Cell Signaling 9806S ) for 30 min on a rotor at 4 °C . After centrifugation at 12 000× g for 15 min , the lysate ( 30μL ) was co-incubated with Vip3Aa-Flag ( 10 μg ) for 2 h at 4 °C . The sample was immunoprecipitated with 5 μL anti-V5 antibody overnight at 4 °C , and 40 μL of protein G agarose beads ( Santa Cruz ) were washed and then added for additional 4 h . Thereafter , the precipitants were washed five times with washing buffer ( 3 . 2 mM Na2HPO4 , 0 . 5 mM KH2PO4 , 1 . 3 mM KCl , 135 mM NaCl , pH 7 . 4 ) , and the immune complexes were boiled with loading buffer for 6 min and then analyzed by western blotting . Five microliter of different regions of the Sf-SR-C protein ( 0 . 1 nmol ) were dotted onto a PVDF membrane . After blocking with 5% skimmed milk in phosphate buffer solution with 0 . 05% tween-20 ( PBST ) , the membrane was incubated in Vip3Aa-flag ( 100 nM ) for 1 h at room temperature , and washed at least three times using PBST . Vip3Aa without Flag-tag ( 500-fold excess ) was used in the competition assays . The following steps are consistent with western blotting . Ten microliter of Sf9-pIZT-SR-C cells lysate were subjected to SDS-PAGE and then transferred to PVDF membrane . After blocking with 5% skimmed milk in PBST , the membrane were incubated in Vip3Aa-flag ( 100 nM ) for 2 h at room temperature , and washed at least three times using PBST . Vip3Aa without Flag-tag ( 200-fold excess ) was used in the competition assays . The following steps are consistent with western blotting . Different parts of the Sf-SR-C protein fused with ( GST ) ( 0 . 4 nmol ) were incubated with GST-Sepharose affinity beads at 4 °C for 3 h and then washed three times with PBS to remove unbound proteins . Beads were incubated with Vip3Aa-flag ( 100 nM ) and washed five times with PBS . The precipitated components were boiled with sample buffer for 10 min and analyzed by western blotting . Cell viability assays were performed using the CCK-8 Counting Kit ( Dojindo ) . Briefly , cells with a density of 5 × 104 cells per ml were seeded into 96-well culture plates separately . After overnight incubation , the cells were treated with Vip3Aa protoxin ( 50 μg/mL ) for 48 h . WST-8 reagent was then added to each well . After incubating at 27 °C for 2 h , the absorbance was measured in microplate reader ( PerkinElmer ) at 450 nm . Treatment with sterile dialysis buffer was used as a control . All tests were performed in triplicate and were repeated at least three times . Cell viability ( % ) = average absorbance of treated group / average absorbance of control group × 100% . Cells were grown to 60–80% confluence in Laser confocal culture dishes . After treatment , cells were washed three times with PBS to remove unbound ligands , and fixed with freshly prepared 4% paraformaldehyde at 37 °C for 30 min . For co-localization experiments , cells were then permeabilized ( 0 . 2% Triton X-100 ) and immunostained ( primary and secondary antibodies were diluted in 5% skimmed milk powder ) . Cellular cortical actin and nuclei were labeled for 30 min with fluorescein isothiocyanate ( FITC ) -phalloidin ( Sigma ) and DAPI ( Sigma ) respectively . Cell images were captured using a Zeiss . LSM710 confocal microscope . S . exigua: Drosophila: Experiments were performed at least three times independently . All statistical data were calculated with SPSS software . ( v . 22 . 0 ) . For comparisons of the means of two groups , two-tailed t test was used . For comparisons of multiple groups with a control group , one-way ANOVA method was used . Significance of mean comparison is annotated as follow: ns , not significant; P<0 . 05; P<0 . 01; **P<0 . 001 .	Bacillus thuringiensis Vip3A has potential in control of Lepidopteran pest and has been used in transgenic plants . However , studies of the insecticidal mechanisms of Vip3A are rare , and none of their definite receptors have been reported so far , which seriously restricts the study of its insecticidal mechanism and application . This work identified and confirmed the scavenger receptor class C like protein ( Sf-SR-C ) acts as the receptor of Vip3Aa protoxin , demonstrated that Sf-SR-C mediates the toxicity of Vip3Aa to Sf9 cells in an internalized manner . These results extend our understanding of SR-C proteins in insects and explain the specificity of Vip3Aa insecticidal activity , which strongly support it as a safe biopesticide . More importantly , it suggests the insecticidal mechanism of Vip3Aa different from the well-known “pore formation” model , “signal transduction” model , as well as newly found “necrosis” model of Cry toxins , which will significantly promote the relevant study of Vip3Aa . Last but not least , because scavenger receptors play a crucial role in innate immunity , our results provide relevant insights into host-pathogen interactions .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "toxins", "pathology", "and", "laboratory", "medicine", "cell", "processes", "animals", "toxic", "agents", "toxicology", "plasmid", "construction", "animal", "models", "membrane", "proteins", "developmental", "bi...	2018	Scavenger receptor-C acts as a receptor for Bacillus thuringiensis vegetative insecticidal protein Vip3Aa and mediates the internalization of Vip3Aa via endocytosis
Natural killer ( NK ) cells are circulating lymphocytes that play an important role in the control of viral infections and tumors . Their functions are regulated by several activating and inhibitory receptors . A subset of these receptors in human NK cells are the killer immunoglobulin-like receptors ( KIRs ) , which interact with the highly polymorphic MHC class I molecules . One important function of NK cells is to detect cells that have down-regulated MHC expression ( missing-self ) . Because MHC molecules have non polymorphic regions , their expression could have been monitored with a limited set of monomorphic receptors . Surprisingly , the KIR family has a remarkable genetic diversity , the function of which remains poorly understood . The mouse cytomegalovirus ( MCMV ) is able to evade NK cell responses by coding “decoy” molecules that mimic MHC class I . This interaction was suggested to have driven the evolution of novel NK cell receptors . Inspired by the MCMV system , we develop an agent-based model of a host population infected with viruses that are able to evolve MHC down-regulation and decoy molecules . Our simulations show that specific recognition of MHC class I molecules by inhibitory KIRs provides excellent protection against viruses evolving decoys , and that the diversity of inhibitory KIRs will subsequently evolve as a result of the required discrimination between host MHC molecules and decoy molecules . Natural killer ( NK ) cells constitute 5–25% of the lymphocytes circulating in human peripheral blood [1] . Being part of the innate immune response , they play an important role in the defense against viral infections and in tumor surveillance [2] . In contrast to T and B cells , NK cells do not use somatic gene rearrangements to generate a diverse repertoire of cells expressing unique receptors . Instead , they sample a subset of receptors from a repertoire of activating and inhibitory receptors encoded by the germline . Individual NK cells express several inhibitory and activating receptors that recognize , among others , major histocompatibility complex ( MHC ) class I and MHC class I related molecules as their ligands [3] . The interaction between these receptors and ligands generates signals that either allow the NK cell to attack target cells or prevent it from harming healthy tissue . Several viruses down-regulate the expression of host MHC class I molecules , and since these molecules are often inhibitory ligands of NK cell receptors , loss of their expression on the infected cell induces NK cell activation . This mechanism by which NK cells attack MHC-class I deficient cells was coined by Kärre et . al [4] as “missing-self” detection . In humans there are two main receptor families contributing to missing-self detection . The inhibitory receptor CD94/NKG2A binds to complexes of the human leukocyte antigen ( HLA ) -E , presenting peptides derived from the leader sequences of HLA-A , -B , and -C [5] , [6] . In this inhibitory interaction both receptor and ligand are highly conserved , and the down-stream effects are remarkably similar in different individuals [7] . In contrast , killer immunoglobulin-like receptors ( KIR ) , recognizing the highly polymorphic HLA-A , -B , and -C molecules , can be both inhibiting and activating , are very diverse , and rapidly evolving [8] . Engagement of either inhibitory KIR or NKG2A inhibits the activity of an NK cell , preventing target cell lysis . Phylogenetic studies have shown that the CD94/NKG2 system is relatively old , and that the KIR genes have evolved more recently [9] . Thus , there are two NK cell receptor systems , one conserved and one highly diverse , detecting abnormalities in MHC expression on cell surfaces . KIRs are encoded by a large family of genes exhibiting a remarkable variability in gene content and allelic polymorphism . The KIR complex in humans contains up to 14 KIR genes and pseudogenes [10] that are arranged into two main groups of haplotypes , A and B , differing in size , gene content , function , and disease association [11] . Since the MHC and KIR loci are on different chromosomes ( in humans , on chromosome 6 and 19 , respectively ) , a tremendous number of possible receptor-ligand combinations exists on the population level . Moreover , KIR-HLA interactions are rather specific , with four mutually exclusive epitopes on HLA molecules ( A3/11 , Bw4 , C1 and C2 ) so far identified as inhibitory KIR ligands [12] . KIR interactions with HLA-C are sensitive to polymorphisms at distal positions [13] and to bound peptides [14] , affecting KIR binding , and with that the functionality of NK cells . It is widely accepted that the fine specificity and vast diversity of B and T cell receptors per individual render each host the capacity to recognize many different pathogens , and to distinguish them from healthy tissue . But how does the specificity and much smaller diversity of NK cell receptors per individual contribute to the host's survival ? If missing-self detection were the main function of inhibitory KIRs , and since this can also be achieved by the conserved receptor NKG2A , why have these more recent NK cell receptors evolved to become specific , polymorphic , and diverse ? Specific KIR alleles have been associated with particular infections such as HIV , HCV , cerebral malaria , and with several pregnancy disorders [15]–[21] . Indeed , population genetic studies have suggested that a high degree of KIR diversity is necessary for surviving epidemic infections and population bottlenecks [22] , but no explicit evolutionary mechanism selecting for novel KIR alleles has been proposed so far . Why polymorphic KIRs would be required to just detect MHC down-regulation remains puzzling . Cytomegaloviruses ( CMV ) and other viruses from the herpes family have large genomes that encode for a series of immuno-evasive mechanisms , targetting key molecular steps necessary for a successful immune response [23]–[25] . Particularly important for the evasion of NK cell surveillance are MHC-I like molecules that can engage inhibitory NK cell receptors , like the mouse CMV ( MCMV ) encoded glycoprotein m157 binding to Ly49 receptors [26] , [27] , and the human CMV ( HCMV ) UL18 engaging the inhibitory leukocyte immunoglobulin-like receptor LIR–1 [28] . Not all of these evasion strategies have been elucidated yet , and it remains unclear whether m157 and UL18 are the only decoy molecules evolved by herpes viruses . Recent studies have revealed a strong imprint in the KIR repertoire of CMV seropositive individuals [29] , [30] , suggesting that additional CMV evasion mechanisms interacting directly with KIRs ( e . g . novel decoy molecules yet to be identified ) exert a strong selection pressure . We investigated whether the presence of viral decoys like MCMV m157 or HCMV UL18 can drive the expansion of specific , inhibitory NK cell receptors , such as KIRs . We performed our study with an agent-based computer model of co-evolving hosts and viruses . Our results show that specific MHC recognition by inhibitory KIRs provides excellent protection against viruses evolving decoy molecules , and that diversity in the receptor system can be a consequence of this specific interaction between MHC and KIR molecules . If detection of MHC class I is the main function of inhibitory KIRs , we expect that KIRs do not need to be specific nor diverse because missing self detection can be achieved by a limited set of monomorphic receptors . To address this , we analyzed the effects of KIR specificity on populations infected with a virus that is capable of MHC down-regulation to escape T cell response . The immune escape of this mutant was modeled by decreasing the probability of clearing the infection ( from 85% to 70% ) , resulting in a better spread of the virus and a larger fraction of individuals becoming chronically infected ( Fig . 1 B , and Fig . 2 A red line ) . We screened the average population size as an indicator of the hosts' protection after an infection . By comparing simulations with degenerate KIRs with those with specific KIRs , we observed significant differences in population sizes ( from 4300 in to 4100 in , , Mann-Whitney U test ) . Although the effect of KIR specificity on protection against an MHC down-regulating virus was small , it clearly indicated that hosts with highly specific KIR–MHC interactions were more vulnerable than those having degenerate KIRs ( Fig . 2 A , B , E ) . Why is a high KIR specificity disadvantageous during an infection with an MHC down-regulating virus ? Since degenerate KIRs ( i . e . ) are likely to recognize any MHC in the population , these receptors are perfectly capable of detecting the presence ( and hence absence ) of MHC molecules within one individual . But if the KIR-MHC interaction is specific enough ( i . e . ) , the chance of a KIR to recognize any MHC within the same individual is small , impeding the host to detect MHC down-regulation , i . e . missing-self . Thus , the potential to recognize the absence of MHC molecules , and with it to clear the infection , decreases with a higher specificity of KIR-MHC interactions . Note that the inability of a specific inhibitory KIR to recognize missing-self is independent of the education process we implemented in the model . These results were consistent in all simulations we ran for each specificity setting ( , Fig . 2 E red line ) , confirming our reasoning that for missing-self detection , inhibitory NK cell receptors do not need to be specific . To avoid elimination by the host immune response , viruses like CMV code decoy MHC molecules that can engage inhibitory NK cell receptors [26] , [27] . As KIR specificity did not have a large effect on missing-self detection , we wondered whether high KIR specificity can be an adaptation to a CMV like virus . In our model , a virus down-regulating the MHC expression in one individual , can randomly select one of the MHC molecules of its host , incorporate it in its “genome” , and express it as a decoy protein in the current and subsequent hosts . While in the current host this decoy is always successful , in the subsequent hosts its success will depend on the specificity of the KIRs . Viruses carrying successful MHC decoys can escape the immune response of both T and NK cells . The fitness cost of a host infected with one of these successful viruses was modeled by decreasing the probability of clearing the infection to zero ( Fig . 1 B ) . Thus , each individual with KIRs recognizing a foreign viral decoy like self MHC , became chronically infected in the model . The better adaptation of a decoy virus compared to the MHC down-regulating virus was reflected in a higher fraction of chronically infected individuals and in a lower population size ( Fig . 2 C ) . But , opposite to what we observed with the virus down-regulating MHC , the effect of KIR specificity was drastic . The average population size increased from 2500 individuals in a degenerate system to 4100 in a very specific system ( , Mann-Whitney U test ) . Populations having specific KIR-MHC interactions were thus much better protected than those with degenerate or cross-reactive KIRs ( Fig . 2 C , D , E ) . Why is a highly specific KIR-MHC interaction advantageous in this CMV like infection ? To protect the host , KIRs face the challenge to detect MHC down-regulation but not recognize the viral decoy masking MHC down-regulation . As seen in the previous section , a host with degenerate KIRs always has a large repertoire of licensed KIRs , and therefore always succeeds in detecting missing-self . But because of the same low specificity , the KIRs within that individual are expected to also recognize foreign decoy molecules as self MHC . On the other hand , a specific KIR system results in a smaller repertoire of licensed KIRs per individual , impeding the host's ability to detect missing-self ( see Table 1 and previous section ) . However , because of their high specificity , it is also unlikely for any licensed KIRs to recognize foreign decoys . Therefore , a decoy virus typically fails to escape NK immune responses , allowing the infection to be cleared . Again , these results were consistent in all simulations we ran for each specificity setting ( , Fig . 2 E black line ) , showing that KIR specificity helps protecting individuals against viruses evolving MHC-like molecules . We next studied the effect of specificity on the evolution of KIRs . To estimate the diversity of KIR molecules in the population , we calculated the Simpon's Reciprocal Index ( SRI ) [33] . The SRI is a diversity measure that is equal to the total number of KIR alleles if they are all equally distributed in the population , whereas the SRI is lower than that in a population where some alleles dominate ( described in Material and Methods ) . This measurement of diversity has the advantage that it is not sensitive to fluctuations in the frequencies of rare KIR alleles in the population . KIR polymorphism remained low in populations having degenerate and cross-reactive KIRs ( i . e . ) , whereas it increased significantly in populations with specific KIR-MHC interactions ( Fig . 3 A–B ) . Why is there only selection for diversity in those populations having specific KIRs ? Since every host needs to recognize at least one of its MHC-molecules to have a licensed KIR repertoire , and this is guaranteed with degenerate KIRs , there is hardly any selection pressure in these populations to evolve novel KIR molecules . But with specific KIRs , individuals do not always recognize their own MHC and hence they are more vulnerable during the infection with an MHC down-regulating virus . The chance of recognizing self MHC is higher in individuals carrying two different haplotypes of inhibitory KIRs . Therefore , heterozygous hosts have an advantage over homozygous hosts , an effect that becomes larger with increasing KIR specificity . We conclude that this “heterozygous advantage” is the main selection pressure driving the evolution of novel KIR haplotypes , and that the selection pressure is largest in populations with specific KIR-MHC interactions . We argued that heterozygous advantage drives the selection of novel KIR molecules in populations with specific KIR-MHC interactions ( i . e . ) . Yet , the KIR diversity differed significantly between simulations with MHC down-regulating and decoy viruses for exactly the same specificity ( , and , Mann-Whitney U test , Fig . 3 C ) . This result was surprising , because the heterozygous advantage in the likelihood of recognizing self MHC should be equally strong in both types of infection . To address the possible mechanisms underlying this result , we studied the KIR molecules that were being selected after an infection with the CMV-like virus . First , we analyzed the specificity of the KIR molecules . Although the specificity threshold was fixed , some KIRs happened to recognize more MHC molecules than others . In populations with highly specific KIR-MHC interactions ( e . g . ) , the initial haplotype was composed of 5 KIRs , each of them recognizing a different number of MHC in the population ( Fig . 4 A , B ) . This distribution remained constant until the mutant viruses emerged . When the CMV-like virus was introduced , there was a clear selection for those KIR molecules that were most specific , i . e . KIRs recognizing only one MHC molecule in the population ( Fig . 4 B ) . Similarly , populations infected with an MHC down-regulating virus evolved more cross-reactive KIR molecules ( Fig . 4 A ) . Thus , although the specificity threshold was fixed , the system exploited the stochastic variation in cross-reactivity among KIR molecules , evolving towards the specificity that rendered most protection ( Fig . 4 C , D ) . During an infection with decoy viruses , this selection started already in populations having intermediate specific KIRs ( i . e . ) . Surprisingly , a higher specificity was achieved by haplotypes implementing duplicate KIR genes , which effectively decreased the number of loci ( Fig . S2 ) . The fact that the evolution of an even higher specificity affects the heterozygote advantage ( Fig . S1 ) explains the variation in KIR diversity between the infections with MHC down-regulating and decoy viruses ( Text S-S ) . If evolution selects for the most specific KIRs to protect against viruses evolving decoys , the challenge of recognizing self MHC is even larger . Is there any mechanism that allows for a higher chance of detecting self despite a high KIR specificity ? To address this question , we studied the KIR haplotypes before and after the infection with a decoy virus , and again measured the number of recognized MHC per haplotype . In populations having specific KIRs ( e . g . ) , a randomly generated initial haplotype recognized an average of 8 MHC molecules , reflecting the expected cross-reactivity of its KIRs . Upon infection with decoy viruses , more specific haplotypes evolved due to the evolution of more specific KIR molecules . At the end of the simulation , approximately 80% of the haplotypes recognized only five different MHC molecules in the population ( Fig . S3 ) ; a surprising result because this property is only expected for 45% of the randomly created KIR haplotypes with . Hence , there was a clear selection for haplotypes that overlapped as little as possible in their MHC recognition , while keeping the highest specificity per KIR molecule . By evolving such “orthogonal” haplotypes , the paradox of recognizing as many MHC in the population as possible without detecting foreign decoy molecules was solved . Together , our analysis suggests that , if the specificity of KIR increases , it becomes beneficial to have more loci to be able to detect missing-self , which provides an explanation for the observed polygeny in the KIR complex . The exact evolutionary advantage of the highly diverse KIRs has remained intriguing , especially because MHC class I detection , i . e . “missing-self” detection , can also be achieved by a limited set of monomorphic receptors . Our results show that for simple detection of MHC down-regulation , degenerate KIR molecules are advantageous , while a specific KIR-MHC interaction protects hosts against viruses evolving decoy molecules . In the presence of viruses expressing decoy molecules , the KIR became very specific , while at the same time the number of recognized MHC molecules per haplotype was maximized . The more specific the system becomes , the stronger the selection pressure on hosts to carry two different KIR haplotypes . As a result of this heterozygote advantage , KIR haplotypes evolve a high degree of diversity . The results with viruses evolving decoy molecules depend strongly on the implemented MHC dependent NK education process , as we allow only for the “licensed” KIRs to participate in the immune response . Inhibitory receptors for MHC class I are very important for the education , repertoire development , and response of NK cells , and there is indeed good evidence that the failure to engage inhibitory receptors during development results in peripheral NK cells that are hyporesponsive [32] , [34]–[37] . Yet , recent studies [38]–[40] showed that NK cell populations that cannot ligate their inhibitory receptors–either because they are unlicensed cells , or because they have been transferred into a different MHC class I environment–respond in a normal inflammatory manner after viral infection . This response of unlicensed NK cells appeared to be even more robust and protective than that of licensed NK cells . In all these studies , the NK cells were stimulated via their activating receptors . However , we only considered inhibitory receptors , modeling “functionality” as the capacity of the mature NK cells to respond to cells in which the expression of self MHC class I is decreased . MHC independent mechanisms for NK cell activation , such as activation via cytokines , is implicit in the model , and is taken into account in the overall probability of clearing the infection . Also note that we do not model the KIR molecules on individual NK cells , but define which KIRs are licensed in a host's whole repertoire . The KIR system has evolved unusually rapidly , resulting in different levels of specificities across species . While KIRs in rhesus macaque have a broad specificity , orangutans , chimpanzees , and humans have evolved more specific KIR systems [41] , [42] . KIR recognition in humans is restricted to at least four epitopes ( HLA-A11 , -Bw4 , -C1 , and -C2 ) , where HLA-C1 and -C2 have the highest avidity . The evolution of these particular MHC epitopes have left an imprint on the evolution of the KIR system . This is clearly shown in the differences of the KIR haplotypes starting from old world monkeys to humans [43] . Humans , chimpanzees , bonobos , gorillas , orangutans , and rhesus macaques share four lineages of KIR genes , which expanded approximately 35–40 million years ago [31] . Within these lineages , each species has independently evolved different numbers of KIRs with either an inhibiting or activating function , and their emergence is related to the evolution of their ligands . The expansion of lineage III KIRs in orangutans , chimpanzees , and humans is associated with the emergence of the C1- , and later with the C2 epitopes . In contrast , rhesus macaques have expanded lineage II KIR genes , corresponding to their complex MHC system , which is composed of several subsets of differentially expressed MHC-A , and -B genes in the absence of an MHC-C locus . Here , we do not model the four KIR epitopes present on several MHC alleles , but have randomly made MHC molecules . The composition of KIR haplotypes , as well as the evolution of novel MHC molecules , is not the focus of this manuscript . Our main question is why KIRs have evolved a degree of specificity , and our approach clearly reveals that specificity has a selective advantage because of its protective effect against CMV-like viruses evolving MHC decoys . Some of the high specificity values used in our model might seem contradictory to the small number of MHC epitopes identified as main KIR ligands . Yet , all results presented here are already obtained at specificity values , which corresponds to a recognition of 20% of MHC molecules in the population ( see Table 1 ) , and is in agreement with the four MHC epitopes that have been identified so far in human KIRs . Hosts are exposed to multiple challenges during their life span , and the immune system has evolved to respond to all of them rather than adapt to only one particular virus . For simplicity , we here consider only one type of infection at a time . Nevertheless , CMV seems to have an important role in NK cell mediated immunity . Recent studies revealed that there is a strong imprint in the NK cell repertoire of CMV seropositive individuals because a particular subset of NK cells with “self-specific” inhibitory KIRs is expanded [29] , [30] . Furthermore , it has been shown that CMV plays an important role in viral driven evolution of NK cell receptors in mice [44] , [45] . Mice possess the Ly49 receptor system , which is functionally similar to KIR but evolutionary and structurally different . The Ly49 receptors exhibit also high genetic diversity and also have mouse MHC-class I molecules as ligands . Mouse strains that are resistant to MCMV carry an activating receptor , Ly49H , binding to the “MHC-class I decoy” m157 with high affinity . Mice susceptible to MCMV lack the Ly49H gene but possess the inhibiting receptor Ly49I also binding strongly to the m157 glycoprotein . Because Ly49H evolved from its inhibitory homologue , Ly49I [46] , it seems that the m157 induced immune pressure led to the evolution of a new activating NK cell receptor , conferring resistance to the virus . Our results agree with this data , showing that a CMV encoded MHC-like decoy imposes a selection pressure to drive the evolution of novel NK cell receptors . Fighting pathogens and successful reproduction are two crucial functions for survival . By their contribution to immune defense and reproduction , KIRs reveal various selection pressures imposed on NK cells , emphasizing the importance of diversity for surviving population bottlenecks and infections . For these reasons , it may seem intuitive that receptor diversity is beneficial for viral control . But we have seen that the mere detection of missing self is achieved best with degenerate KIRs . Our agent-based model provides a solid explanation for one selection pressure driving the evolution of specific KIRs , namely viruses expressing MHC decoys . This does not need to be the only explanation , and our findings call for further studies into other possible mechanisms . The evolution of the specificity and number of loci per haplotype , as well as the evolution of activating receptors or other viral strategies , should now be integrated in our model to address additional questions . We developed an agent-based model consisting of two types of actors ( hosts and pathogens ) and three types of events ( birth , death , and infection ) . The basic time step of the model is one week , during which we run through all hosts in a random order and confront them to one of the randomly chosen events . Hosts age over time , and after each time step , their age , infection state , and infection type is updated . The cycle is repeated for many hosts generations to model the long-term evolution . All model parameters are fully described in Table 2 . The following is a detailed description of the actors and the events: The model was initialized with a host population of 4500 hosts , with a random age between 1 and 70 years . Gene pools for MHC and KIR alleles were created at the start of each simulation . The pool of MHC consisted of 14 alleles according to the most frequent HLA-C alleles in the European population ( dbMHC Anthropology [49] ) . For each MHC allele , ten different KIR were randomly generated , which could bind to the MHC with a specificity of at least , resulting in a KIR pool of “functional” 140 alleles . To create the initial genome of each host , MHC and KIR genes were randomly drawn from the pools . The individuals were initialized with the same KIR haplotype , but with different MHC genes . The Simpson's Index is a measurement of diversity that can be interpreted as the probability that two randomly chosen molecules from two random hosts in the population are identical . The lower the Simpson's Index , the higher is the diversity of molecules in the population , and the reciprocal of the Simpson's Index [33] defines a “weigthed” diversity . This diversity measure has the advantage over the total number of unique KIR molecules because it is less sensitive to fluctuations in molecule numbers caused by random neutral drift . For instance , if all molecules are equally frequent in a population , the SRI score is equal to the number of alleles in the population . A population dominated by a single molecule will have an SRI score close to 1 . The SRI was calculated as follows: , where is the fraction of the molecule over all KIR molecules in the population , and is the total number of unique KIR molecules . The probability that a host having a heterozygous diploid genome recognizes its own MHC molecules is defined by , where is the probability that a KIR recognizes a random MHC molecule in the population ( which depends on , see Table 1 ) , and is the number of KIR loci . The expected number of licensed KIR is determined by . In our model each host has a genome consisting of one MHC locus and five KIR loci , i . e . , hence that individual will recognize its own MHC molecules with a chance , and the expected number of licensed KIR for the same individual will be . The expected protection against a decoy virus , i . e . the probability of not recognizing the viral protein as self MHC molecule , depends on the size of the licensed KIR repertoire , , and is described by . Heterozygous advantage is defined as: , where and represent the probability of recognizing self MHC molecules for a heterozygote and a homozygote individual , respectively . We obtained by measuring the fraction of MHC molecules detected by a single KIR haplotype . To obtain we measured the fraction of recognized MHC by all pairwise combinations of KIR-haplotypes . The population has heterozygote advantage if . Values of HA for different , , and are given in Table 1 . The model was implemented in the C++ programming language . The value of was varied in a range from one to ten . For each value , ten simulations were performed for 2000 centuries .	Human natural killer ( NK ) cells patrol peripheral tissue , monitoring changes on the surface of body cells . They express a network of activating and inhibitory receptors called the killer immunoglobulin-like receptors ( KIRs ) . The main ligands of inhibitory KIRs are MHC class I molecules , which present viral peptides to other immune cells . Several herpes viruses interfere with MHC expression , and when a virus down-regulates MHC class I , NK cells loose an inhibitory signal , become activated and kill the infected cell . The KIR family has a large genetic diversity . However , for the recognition of “missing” MHC molecules this diversity seems redundant as one set of receptors should be sufficient . To study why the KIR system has evolved such a high complexity , we developed an in-silico model , simulating the evolution of populations infected with a herpes-like virus . Next to down regulating MHC-I molecules , these viruses are able to escape the NK cell response by expressing MHC-decoys engaging the inhibitory KIRs . We show that specific KIR-MHC interactions protect against viruses expressing decoys . Because of the provided protection , specific inhibitory KIRs have an evolutionary advantage , giving rise to a high level of diversity . We propose that herpes-like viruses evolving decoys affect in the evolution of KIRs .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2013	Virus Encoded MHC-Like Decoys Diversify the Inhibitory KIR Repertoire
Gene duplication provides much of the raw material from which functional diversity evolves . Two evolutionary mechanisms have been proposed that generate functional diversity: neofunctionalization , the de novo acquisition of function by one duplicate , and subfunctionalization , the partitioning of ancestral functions between gene duplicates . With protein interactions as a surrogate for protein functions , evidence of prodigious neofunctionalization and subfunctionalization has been identified in analyses of empirical protein interactions and evolutionary models of protein interactions . However , we have identified three phenomena that have contributed to neofunctionalization being erroneously identified as a significant factor in protein interaction network evolution . First , self-interacting proteins are underreported in interaction data due to biological artifacts and design limitations in the two most common high-throughput protein interaction assays . Second , evolutionary inferences have been drawn from paralog analysis without consideration for concurrent and subsequent duplication events . Third , the theoretical model of prodigious neofunctionalization is unable to reproduce empirical network clustering and relies on untenable parameter requirements . In light of these findings , we believe that protein interaction evolution is more persuasively characterized by subfunctionalization and self-interactions . Gene duplication is readily accepted as a primary mechanism for generating organismal complexity . Phenomena proposed for the fate of gene duplicates include neofunctionalization and subfunctionalization . Neofunctionalization posits that the functional redundancy intrinsic to initially identical gene duplicates releases one duplicate from selective pressure . While under neutral selection one of the duplicates can accumulate random mutations and potentially acquire novel and beneficial functions [1] . Subfunctionalization states that both gene duplicates acquire mutations resulting in each duplicate assuming a complementary subset of the ancestral gene's original functions [2] . Gene duplication and subsequent neofunctionalization and subfunctionalization have straightforward analogs in models of protein interaction network ( PIN ) evolution . With proteins as nodes , edges between proteins represent physical interactions and serve as an indication of protein function . Proteins with identical sets of interacting partners are presumed to have identical functions . Gene duplication is modeled by copying a protein node in the network along with its interactions . Neofunctionalization and subfunctionalization are modeled by the gain and loss of interactions respectively . This straightforward representation has made PINs an attractive target for the study of evolution . Both neofunctionalization and subfunctionalization have been shown to occur in protein interaction analyses of extant species . Since paralogs are by definition related by gene duplication , the similarities and differences between the interactions of paralogous pairs have been used to elucidate the role of neofunctionalization and subfunctionalization in the fate of gene duplicates . Wagner [3] , [4] noted that an interaction between a paralogous pair forms by one of two methods: either the duplication of a self-interacting protein ( Figure 1 ) , or a de novo interaction forming between the pair sometime after duplication . Wagner's analysis of three Saccharomyces cerevisiae interaction datasets revealed that the vast number of interacting duplicate pairs were not themselves self-interacting . Therefore , the absence of homomeric interactions in interacting paralogous pairs suggested that these interactions formed de novo ( i . e . , neofunctionalization ) . Extrapolating the probability of an interacting paralogous pair to the entire network , Wagner estimated that Saccharomyces cerevisiae adds between 108 and 294 . 5 interactions de novo every million years . Wagner also compared the age of paralogs to the number of shared interaction partners . Wagner found that , except for the most-recently duplicated genes , duplicate pairs have lost on average from 85 to more than 90 percent of their shared interactions depending on their age and the dataset examined [4] . The rapid loss of common interacting partners between duplicates strongly suggests that subfunctionalization occurs quickly after duplication . A more recent study using similar methods measured 93% shared interaction loss in yeast [5] . He and Zhang also found evidence of rapid subfunctionalization followed by a prolonged period of neofunctionalization in Saccharomyces cerevisiae protein interactions [6] . They reasoned that the set of nonredundant interacting partners shared between paralogous pairs should remain constant over time if subfunctionalization occurs without neofunctionalization . They ascertained that the set of nonredundant partners increased with the age of the paralogous pair , indicating the presence of neofunctionalization ( Figure 2 ) . Neofunctionalization and subfunctionalization also appear in theoretical models of protein interaction evolution . The first model combining both neofunctionalization and subfunctionalization came from Solé and colleagues in 2002 [7] . Their duplication and diversification model iteratively duplicates a random gene and its interactions , followed by probabilistically deleting copied interactions ( subfunctionalization ) and adding new interactions ( neofunctionalization ) . A number of topological measures were found to be consistent between both the network produced by their model and observed Saccharomyces cerevisiae protein interactions , including connectivity , clustering coefficient , power-law degree exponent , and path length . The two most common high-throughput assays used to determine yeast protein interactions , yeast two-hybrid ( Y2H ) assays and affinity purification with mass spectrometry ( AP-MS ) , have limited ability to discern self-interactions . In Y2H assays , self-interacting baits interact together and self-interacting prey interact together reducing the concentration of bait/prey interactions with respect to their heterointeracting counterparts . Additionally , the GAL4 binding domain binds DNA as a dimer [10] , [11] , allowing homomeric bait pairs to dimerize with each other instead of prey ( Figure 3 ) [12] , [13] . Large-scale TAP-MS studies [14]–[16] report no homomeric interactions due to a lack of endogenous ( untagged ) homomeric mates to discern from the affinity tagged protein [17] . Other large-scale AP-MS studies [18] use small epitope tags . The epitope tagged homomer very nearly overlaps with its endogenous mate in the MS spectra making the flagged homomer difficult to discern from its unflagged mate . For example , only a single homomeric interaction among 3 , 617 reported interactions was identified by Ho and colleagues in 2002 [18] using the FLAG epitope tag . Examination of the physical data supports a higher proportion of homomers than yeast two-hybrid and AP-MS studies indicate . First we compiled a set of non-redundant structures containing Saccharomyces cerevisiae protein complexes from The Protein Data Bank ( PDB ) [19] . We then cross-referenced these structures to the iPfam database of PDB protein interactions [20] ( see Methods ) . A tally of identical proteins self-interacting across different polypeptide chains confirms the ubiquity of self-interacting proteins . There are 207 non-redundant yeast structures containing 210 Saccharomyces cerevisiae proteins , 149 of which ( 71% ) are self-interacting . Similarly , the BRENDA enzyme database [21] contains 102 Saccharomyces cerevisiae enzymes with specific hetero- and homomeric k-mer counts ( monomer , dimer , trimer , etc . ) . Self-interacting enzymes ( k-mers with k≥2 ) accounted for 60% of the Saccharomyces cerevisiae enzymes . At the protein complex level , Pereira-Leal et al . [22] found that 90% of the structures in the Protein Quaternary Structure database [23] include homomeric interactions , and other studies also identify a high proportion of homomeric interactions [8] , [9] . By contrast , in high-throughput yeast two-hybrid studies by Uetz et al . and Ito et al . detected homomeric proteins in only 4 . 6% and 6 . 6% respectively of the proteins included in their core interaction sets . Additional evidence supports widespread duplication of self-interacting proteins . Zhang et al . found that , of nine tested attributes , homology was one of four attributes showing substantial predictive value for predicting co-complexed pairs of proteins [24] . Additionally , interactions within paralogous families are much more likely than within randomly-formed families ( P<10−6 , see Methods ) . The wide disparity between the frequency of paralogous versus random interactions indicate that some process other than the random , de novo addition of interactions which characterize neofunctionalization is at work . Duplication of homomers is a more parsimonious explanation than neofunctionalization for the interaction evolution between paralogous proteins . Underrepresented self-interactions in interaction data have not been previously realized , leading to erroneous assertions . Wagner [4] identified 31 interacting paralogous pairs from Y2H assays ( gathered from Uetz et al . , and Ito et al . [25] , [26] ) , and 13 interacting paralogous pairs from non-Y2H assays ( gathered from MIPS [27] ) . In 34 of these 44 interacting paralogous pairs , neither protein of the pair had a self-interaction . Looking for an evolutionary explanation for the presence of the 34 paralogous interactions , Wagner reasoned that either the 34 paralogous pairs ( i . e . , 68 proteins ) lost their ability to self-interact , or that the 34 interactions appeared de novo sometime after duplication . Wagner concluded that the most parsimonious explanation was 34 interactions gained de novo , rather than 68 lost self-interactions . This reasoning led Wagner to postulate that of the other combinations of self- and paralogous- interacting pairs , de novo interaction gain accounted for all but two pairs in which both protein members self-interacted and interacted with each other ( as in Figure 1B ) . Using the number of putative de novo gains as a metric , Wagner extrapolated to arrive at the ubiquitous 108–294 . 5 de novo interactions gained per million years of evolution . Once assay biases are considered as an alternative to evolutionary loss in explaining the absence of self-interactions among Wagner's paralogous pairs , the opposite conclusion is reached: paralogous interactions are more parsimoniously explained by duplicating homomers , not de novo interaction gain . Complementary degenerative mutations intrinsic to subfunctionalization take the form of complementary interaction loss in its network analog . One interaction from each pair of redundant interactions may be lost , but He and Zhang [6] reasoned that in the absence of neofunctionalization , the union of the duplicates' interacting partner sets will remain unchanged over time . Figure 2A features a portion of the methodology used by He and Zhang to test this . They compared the ages of gene duplicate pairs to the union of their interacting partner sets . Contrary to what they believed subfunctionalization alone would show , they found that the union size increased with the age of the duplicate pair . Neofunctionalization was credited with the increase in the number of interacting partners . This argument fails to recognize that the interacting partners evolve as well . Gene duplication and subfunctionalization occur among all genes concurrently with the paralogous protein pair under study . Figure 4A shows a typical gene duplication scenario followed by neofunctionalization as proposed by He and Zhang . Figure 4B shows that the increase in interaction partners over time attributed to neofunctionalization is readily explained by gene duplication occurring elsewhere in the network . After gene duplication , each additional interacting partner acquired by the duplicate pair over time may simply result from an interacting partner undergoing gene duplication . We validated the role subsequent duplications play in increasing the number of interacting partners by counting interacting partners of gene duplicates both before and after accounting for subsequent duplications . Saccharomyces cerevisiae gene duplicates were binned into four different age groups based on genome-wide gene trees developed from 19 fungal genomes ( drawn from revised data provided on Web site associated with Ref . [28] , see Methods ) . Figure 5 shows the phylogenetic nodes which correspond to the age bins gene duplicates were placed into . Interacting partners of gene duplicates were then tallied and plotted according to their age bin ( Figure 6 ) . Before considering subsequent duplications , the number of interacting partners of gene duplicates increases with the age of the duplicate , consistent with the findings of He and Zhang [6] . Once interactions associated with subsequent gene duplications are removed , interacting partner counts show little change over time ( see Methods ) . Another observation is that under concurrent gene duplication , the interacting partners of a duplicate pair should be enriched in paralogs born of subsequent duplications . This is illustrated in Figure 4B . The four interacting partners in frame B4 are two pairs of paralogs which arose via gene duplications subsequent to the original duplication in frame B1 . We sought this evidence in the interacting partners of each duplicate pair present in the both combined datasets [29] , [30] used by He and Zhang [6] and the physical interactions from BioGrid [31] ( see Methods ) . As we expected , the interacting partners of duplicate pairs are significantly enriched with paralogs born of subsequent duplications . The mean proportion of interacting partners which are paralogous in the He and Zhang dataset is 0 . 029 ( P<10−6 , random expectation 0 . 0014 ) and 0 . 042 ( P<10−6 , random expectation 0 . 0025 ) in the BioGrid data . Theoretical models of PIN evolution reproduce characteristics of observed interaction networks while honoring aspects of biological evolution . In 2002 , Solé et al . introduced a “duplication and diversification” model which established the relevance of gene duplication and interaction gain and loss to PIN evolution [7] . The following year Vázquez and colleagues published an alternative model of PIN evolution which includes interaction loss due to subfunctionalization , but does not include neofunctionalization [32] . The common feature of both models is subfunctionalization . That is , both models include a parameter specifying the probability of losing ( or retaining ) interactions to protein partners shared by both the progenitor and progeny genes . The models differ in the method through which new interactions are formed in the network . A second parameter of the Solé et al . model controls the probability of forming new interactions from the newly duplicated gene to each extant gene in the network . A second parameter of the Vázquez et al . model controls the probability of forming a new interaction from the newly duplicated gene to the progenitor gene . Essentially , the difference between these two models can be characerized as neofunctionalization versus homomeric duplication ( i . e . , duplicating a self-interacting gene ) . This difference reflects the dichotomy we've established and therefore deserve additional attention . We have quantified this dichotomy using the topological measure C , the clustering coefficient [33]:T is the number triangles ( three fully-connected nodes ) , and Γ is the number of connected triples ( a node connected to an unordered pair of other nodes ) . The clustering coefficient is a relavant measure for two reasons . First , gene duplications , subfunctionalization , neofunctionalization , and homomeric duplication each produce a measurable change in the number of triangles and connected triples which comprise the clustering coefficient . Second , protein interaction networks have been found to have high clustering coefficients relative to random networks [3] , [7] , [34]–[36] . Table 1 shows that the clustering coefficients for several Saccharomyces cerevisiae datasets are a factor of 5 , 10 , and more above that of equivalent random networks . We seek to identify those evolutionary events which contribute to a high clustering coefficient . The change in clustering coefficient resulting from simple gene duplication , ΔCsimple ( i . e . , duplicating a node and its interactions without regard to subsequent interaction loss ) , occurs locally . The change can be defined in terms of the progenitor's ( p ) triangles ( tp ) and degree ( kp ) , and the degree of the progenitor's neighbors ( kg , g = 1 . . kp , see Figure 7 ) . Because ΔCsimple is restricted to the neighborhood around the duplication progenitor , the majority of duplication scenarios can be modeled by considering only small subnetworks . We enumerated all connected networks ( i . e , all non-isomorphic networks with a single component ) having three to nine nodes . This produces 273 , 191 networks containing a total of 2 , 445 , 434 nodes . Each node in every network was duplicated and the clustering coefficient before and after was measured . In 1 , 864 , 851 ( over 76% ) of the possible duplications ΔCsimple<0 ( Figure 8A ) . In other words , most simple gene duplications decrease the clustering coefficient . Table 2 shows the change in clustering coefficient for enumerated networks as the number of nodes considered increases . We then incorporated a complementary loss probability into our simple gene duplications in the enumerated networks to quantify the impact subfunctionalization has on the clustering coefficient . Subfunctionalization generates an even greater proportion of duplications reducing the clustering coefficient . Figure 8B and 8C show the effect subfunctionalization has on the clustering coefficient in the enumerated networks . The preponderance of enumerated network duplications which reduce the clustering coefficient suggest that additional evolutionary mechanisms beyond that produced by simple gene duplication and subfunctionalization are required to achieve a high clustering coefficient . Indeed , the black lines in Figure 9 show that networks evolved via simple duplication and different degrees of subfunctionalization produce clustering coefficients lower than their random equivalents . The high clustering coefficients relative to equivalent random networks observed in empirical data are unattainable using a simple duplication and subfunctionalization network model . Solé et al . extend simple duplication and subfunctionalization by adding a probability α of adding a de novo interaction from a gene duplicate to each of the existing genes in the network . This probability is defined as: where N is the number of nodes currently in the network and β is a constant reflecting the expected number of de novo interactions added to each gene duplicate [7] ( see Discussion ) . The value of β ( that is , the frequency of neofunctionalization ) can be selected to achieve any desired clustering coefficient . In the extreme , new interactions could be added exhaustively driving the clustering coefficient towards one , that is , the clustering coefficient of a completely-connected network . However , the neofunctionalization model adds random interactions , which drives the clustering coefficient towards random expectation . We updated our simple duplication and subfunctionalization model to include neofunctionalization as implemented in the Solé et al . model . Figure 9A shows the model for various values of β . Biologically plausible β generate too few new interactions and are unable to appreciably affect the topology of the simple duplication model . The value of β derived in Solé et al . [7] is 0 . 16 . The resulting clustering coefficient ( blue line ) is indistinguishable from the simple duplication model . At β = 1 . 6 ( red line ) , the networks and their random equivalents are nearly the same . Increasing β to 16 and 50 ( brown and green lines respectively ) increases the clustering coefficient but also increases the clustering coefficient of its random equivalent . These extreme values for β highlight the close relationship between the neofunctionalization model and its random equivalent . The random edges inherent to neofunctionalization drive the clustering coefficent toward random expectation . At β = 16 and β = 50 , each gene duplicate adds an average of 16 and 50 additional interactions respectively which is biologically untenable . In order to achieve higher clustering coefficients , additional triangles must be added to the network while minimizing the number of triples added to the network . Gene duplication alone can increase the number of triangles if the duplicate is a self-interacting protein . Figure 8A shows that a self-interacting protein increases the clustering coefficient ( ΔChomomer>0 ) of the enumerated networks in 2 , 246 , 876 ( almost 92% ) of possible duplications . In fact , ΔChomomer is always greater than ΔCsimple for an equivalent duplication ( a proof of this can be found in the Supporting Information , Text S1 ) . To contextualize the ability of homomeric duplication to increase the clustering coefficient , we updated our simple duplication model to include homomeric duplication as defined in Vázquez et al . [32] . Note that Vázquez et al . use the term heteromerization . Figure 9B shows that the model produces clustering coefficients markedly higher than those of their random equivalents . It is notable that the probabilities sampled for Figure 9B produce substantially high clustering coefficients despite being much lower than the proportion of homomeric proteins we reported on earlier in structure and enzyme datasets ( 71% and 60% respectively ) The structure and enzyme probabilities are omitted from Figure 8B simply because the higher clusterings they produce result in uninformative lines which are nearly vertical on the plot . Despite the increase in clustering coefficient due to homomeric duplication , the random equivalent networks remain virtually identical to the simple duplication random equivalent , reflecting the modest effect a single edge added by homomeric duplication has on the number of edges and hence on the expected ( i . e . , random ) number of triangles . Gene duplication , neofunctionalization , subfunctionalization , and homomeric duplication each uniquely affect the clustering coefficient . Only homomeric duplication achieves clustering coefficients appreciably higher than clusterings in equivalent random networks . Biological network research is particularly sensitive to dataset biases [41] . Identified correlations between topology and essentiality have been challenged for relying on small-scale assay data which are more frequently the focus of interesting ( i . e . , essential ) genes [42] , and topological inferences of underlying networks have been questioned due to the incomplete sampling of biological assays provide [36] , [43] . The dearth of homomeric interactions in data produced from Y2H and AP-MS assays is another bias which was not previously recognized and needs to be accounted for . The line of reasoning establishing the ubiquity of neofunctionalization was based on such biased data . Wagner based his conclusions on an assumption that the lack of homomeric interactions was a true characteristic of the data . Failure to account for homomeric biases continues to affect evolutionary inferences . Recently , Presser et al . [44] determined that many more self-interacting proteins existed prior to the whole-genome duplication event ( WGD ) in Saccharomyces cerevisiae evolutionary history than are observed today . This determination was accompanied by a discussion about evolutionary causes underlying the loss of self-interactions from the WGD to today . Once the lack of self-interactions is recognized as a result of assay artifacts and not a true characteristic of the data , a simpler conclusion can be drawn: Saccharomyces cerevisiae had many self-interacting proteins prior to the WGD , and continues to have many self-interacting proteins today . Another line of reasoning establishing the ubiquity of neofunctionalization was based on the neighbor sets of duplicated proteins . When inference relies on the neighbors of protein duplicates , accurate estimates require recognizing that those neighbors are also subject to duplication . This omission resulted in He and Zhang's erroneous conclusions . He and Zhang are not alone in failing to recognize this . Concurrent and subsequent duplication has been universally ignored in estimating the rate of subfunctionalization , that is , the proportion of conserved interactions among gene duplicates [4] , [5] , [45]–[47] ( Figure 10 ) . The probability of interaction conservation is estimated by dividing the number of interacting neighbors of both members of a paralogous pair by the total number of neighbors between the pair . If the duplication event which produced the paralogous pair predates the duplication of any of its interacting neighbors , estimates of conservation of interactions are underestimated . Equivalently , estimates of interaction loss are overestimated ( Figure 10 ) . We found that simple duplication and subfunctionalization are unable to produce clustering coefficients observed in empirical protein interaction networks . Neofunctionalization is also ineffective at increasing the clustering coefficient unless untenably high values of β are used . The clustering coefficients resulting from these high values of β are bound closely to the clustering coefficients of random equivalent networks , contrary to observed networks . We found that producing high clustering coefficients with low clusterings in random equivalents as observed in empirical protein interaction networks requires the duplication of self-interacting proteins . A discrepancy remains between our observations and those of Solé et al . [7] . They reported that at β = 0 . 16 their duplication and diversification model generated clustering coefficients consistent with Saccharomyces cerevisiae protein interaction networks . As illustrated in Figure 9A , we found that the same parameter value produces clustering coefficients much lower than observed protein networks and lower than equivalent random networks . This discrepancy is resolved upon further examination . First , the clustering coefficient Solé et al . report for Saccharomyces cerevisiae is taken from a 2001 study [3] which in turn calculated the value based on high-throughput yeast two-hybrid data generated in 2000 [25] . In the intervening years the available protein interaction data has increased tremendously and has resulted in combined datasets with better coverage of the yeast interactome [30] , [40] . It is known that an incomplete sample of a highly-clustered network produces a clustering coefficient lower than the actual network [48] . Therefore as the coverage of the sample increases , the clustering coefficient of the sample should increase as well , eventually reaching that of the actual network when the sample reaches total coverage . The observed clustering coefficients we report in Table 1 are significantly larger than 2 . 2×10−2 , the observed clustering coefficient cited by Solé et al . So although the Solé et al . model produces clustering coefficients consistent with a 2000 dataset , it is low when compared to the more complete datasets available today . A second discrepancy lies in the choice of random equivalent networks . Solé et al . note that their model produces a clustering coefficient roughly 10 times higher than random networks . The random networks they compare against are Erdős-Rényi random graphs which produce a Poisson degree distribution . This degree distribution is quite different than the power law degree distribution of protein interaction networks [49] . A more appropriate network comparison is against a network having an identical degree distribution , but with the edges randomized [50] , [51] . Once equivalent random networks are employed , the reported 10-fold increase in clustering coefficient over random disappears . In fact , at β = 0 . 16 as published by Solé et al . , the model produces clustering coefficients lower than equivalent random networks . It is also useful to look beyond the topologies produced by theoretical models of homomeric duplication and neofunctionalization to the parameters of the models themselves . The Solé et al . model simulates neofunctionalization by forming de novo interactions between the newly-created duplicate and each of the other proteins in the network with probability α . If α is assigned a constant , gene duplicates will acquire an ever-increasing number of interacting partners as the network grows . For example , for α = 0 . 10 , a duplicated gene in a 10-gene network will acquire one interacting partner on average . By the time the network grows to 100 genes , a gene duplicate will acquire 10 interacting partners on average . In order to maintain an average connectivity consistent with observed biological networks , α is adjusted downward as the network grows . Solé et al . 's duplication and diversification model calculates α as proportional to the inverse of the number of nodes currently in the network [7] . This parameterization is difficult to justify biologically . It requires a locally occurring phenomena to be cognizant of a global property of the system , in this case the total number of proteins . By contrast , homomeric duplication models have no such restriction . The model introduced by Vázquez et al . [32] utilizes a simple constant for the probability of adding an interaction between the progenitor and progeny genes ( i . e . , the probability that a self-interacting protein was duplicated ) . In other words , gene duplicates are oblivious to the global state of the system . Solé et al . 's neofunctionalization model and Vázquez et al . 's homomeric duplication model have also been compared in other venues . A study which used machine learning classification to compare seven network evolution models ( including Vázquez et al . and Solé et al . ) to the Drosophila melanogaster PIN found that the Vázquez et al . model produced networks closest to the Drosophila PIN [52] . Model validation of homomeric duplication was also performed by Ispolatov et al . [53] who found that the Vázquez et al . model generated clique distributions consistent with those observed in the Drosophila PIN . The inability of models featuring neofunctionalization to produce a clustering coefficient greater than that of random equivalents , and the absence of a biologically rational method to produce de novo interactions during the evolution of the network argues against the prevalence of neofunctionalization . However , the neofunctionalization model need not be entirely abandoned . Though the neofunctionalization model has little evolutionary inferential efficacy , networks produced from the model have some topological value . The clustering coefficient is just one of several network measures used regularly in network analysis . Producing networks with characteristics consistent with observed PIN topologies is useful in biological network research , and models of both homomeric duplication and neofunctionalization continue to have utility in this regard [48] , [54] . Although we have argued against the ubiquity of de novo interaction gain in protein interaction networks , this does not correspond to a denial of neofunctionalization . There are alternative evolutionary phenomena which may result in new functions , are relevant to protein interactions , and don't necessitate de novo interaction gain between extant proteins . New gene functions may arise through changes in interaction stochiometry or through the formation of new genes formed by exon shuffling , domain insertion , domain loss , domain shuffling , mobile elements , gene fusion , or gene fission [55] , [56] . Gene duplication is generally accepted as a key component of evolution , and protein interactions provide an attractive construct for studying the role of neofunctionalization , subfunctionalization , and homomeric duplication in evolution . Studies of protein interactions derived from empirical data and theoretical models of PIN evolution have regarded ubiquitous neofunctionalization as a requisite feature of post-duplication evolution . We have demonstrated assay limitations and the failure to recognize concurrent gene duplication and subfunctionalization underlie much of the literature which engender neofunctionalization as a prominent factor in protein interaction evolution . Furthermore , biologically implausible parameter requirements and distinctly non-biological clustering characteristics reduce the support theoretical models provide to a ubiquitous neofunctionalization argument . It would be malapropos for us to assert that protein interaction evolution is absent of neofunctionalization . However , we believe de novo interaction gain is not as prevalent as previously thought . We have identified important factors which should be considered in any vetting of evolutionary interaction phenomenon before invoking neofunctionalization as a dominant mechanism . To get structural interactions , we first generated a non-redundant set of Saccharomyces cerevisiae proteins from the Protein Data Bank ( PDB ) [19] . The non-redundant set of protein complexes was identified in a manner similar to Levy et al . [57] . Specifically , for each structure in the PDB containing a yeast protein amino acid chain , create a simple undirected graph where each amino acid chain is an unlabeled node and interactions between different protein chains are edges . Group structures according to shared ( isomorphic ) graph topology . From these build subgroups according to shared sets of Pfam protein domains found in the complex . Further subdivide into subgroups containing the same set of proteins . One member from each of these subgroups is selected to be a non-redundant structure . The selected member is that with the X-ray crystallography structure having the greatest resolution . We then cross-referenced this non-redundant structure set with interacting residue data gathered from version 21 . 0 of iPfam [20] . A protein was identified as self-interacting if there were two molecules ( amino acid chains ) of the protein within a complex that had interacting residues according to iPfam . Enzyme subunit composition was derived from the December , 2007 update of the BRENDA database [21] . BRENDA enzymes with subunit designations of homodimer , dimer , trimer , tetramer , hexamer , octamer , and nonamer were categorized as self-interacting . Monomers and heterodimers were categorized as non-self-interacting . Gene dating ( i . e . , assigning genes to one of T3 , T2 , T1 , T0 as shown in Figure 5 ) was derived from “orthogroup” gene trees from reference [28] . Gene duplications in the gene trees were associated with the phylogenetic nodes in which they occurred . In Figure 6 , black triangles are protein degrees after adjusting for more recent duplications . A black triangle aligned with T3 is the connectivity of a gene duplicate born in T3 after interactions with duplicates born during T2 , T1 , and T0 are removed . Black triangles in T2 have had interactions with duplicates born in T1 and T0 removed . Similarly , T1 black triangles have had interactions with T0 duplicates removed . Duplications within each time period Ti ( i = 1 , 2 , 3 ) , occurred sequentially over a period of evolutionary time and not concurrently . For a given duplication occuring in Ti , on average one-half of the other duplications within Ti occurred subsequent to the given duplication . Therefore , in addition to removing interactions in subsequent time periods as specified above , duplications occurring in the same time period multiplied by 0 . 5 are also removed . Singleton genes , that is genes not associated with any duplication event , are considered to have birthdays preceding T3 in Figure 5 . Singletons interacting with plotted proteins are included in the degree tally , but are not themselves plotted because , by definition , they did not arise during T3 , T2 , or T1 . Each duplication has a progenitor , the ancestral gene , and a progeny , the gene born of the duplication . An issue to be addressed is which gene is the progenitor and which is the progeny . In some cases this is unambiguous . For example , an orthogroup may have three paralogous members: PA , PB , and PC . A common ancestor would have a single gene: PABC . During evolution a duplication event would produce an extant progeny gene ( PA ) and an ancestral progenitor gene ( PBC ) . However , the vast majority of orthogroups contain only two genes . In these cases the duplication event produces two extant genes , making the assignment of progenitor and progeny ambiguous . To address this ambiguity , extant genes pairs produced from duplication events were randomly assigned “progenitor” and “progeny” labels . This random assignment was repeated 100 times and the protein connectivity of the 100 progeny assignments both before and after accounting for subsequent duplications was averaged and plotted as shown in Figure 6 . Duplicate pairs in which both members had degree zero were omitted from the analysis . All duplication events resulting in two extant genes were paired and dated as described above . For each paralogous protein pair born in T3 , the non-redundant set of their neighbors was identified . Paralogous pairs born in T2 , T1 , and T0 were counted as neighbors of the T3 pair if both paralogs of the younger pair were part of the non-redundant set . Paralogous pairs born in T3 were counted at half for the reasons specified above . The equivalent process was used to identify paralogous neighbors of pairs born in T2 and T1 . The P-value represents the number of times a random network with identical topology is at least as enriched in paralogs . To compute the P-value , the gene lables on the network were randomized 106 and the same computation done . As the P-value indicates , none of the randomized networks were as enriched as the empirical networks . Equivalent random networks were generated in order to derive clustering coefficients . Because self-interactions are not included in calculating the clustering coefficient , they were ignored for purposes of creating the random networks . The equivalent random networks used in Table 1 and Figure 9 were generated by rewiring links while preserving the degree distribution [51] . At each iteration a pair of edges were selected at random and one end from each edge was swapped . If the swap created a duplicate edge or a self-interaction the swap was aborted and the next iteration begun . The number of iterations performed was 100E where E is the number of edges in the network . First note that any connected network with N nodes must have a minimum of N−1 edges ( i . e . , a tree ) . All non-isomorphic connected networks with N nodes were determined in two stages . In stage one , a set of N-node trees was built from N−1-node trees established in the previous iteration by adding a node and testing for isomorphism each network generated by adding an edge between the new node and each existing node . Stage 2 follows similarly by iteratively testing networks for isomorphism by adding a single edge to existing N−node networks until N ( N−1 ) /2 edges is reached ( i . e . , the number of edges in a completely connected N-node network ) . The algorithm begins with the two possible 3-node networks , C3 and P3 . Isomorphism is a computationally expensive process . Therefore , isomorphism comparisons were first pre-screened by only evaluating networks with an identical number of edges , nodes , degree distribution , and 2-hop distribution . The algorithm as described in reference [58] was used to determine network isomorphism . Table 2 shows cumulative ΔC of simple duplication and homomeric duplication of the enumerated networks as the number of nodes increases . For the plots in Figure 9 , each network began with a 100-node Erdős-Renyí seed graph . The seed graph was generated by randomly adding edges between the N ( N−1 ) /2 , N = 100 node pair combinations with a probability p = 0 . 04 . We ensured homogeneity by using the same seed graph for each network . Each simulation included simple duplication and subfunctionalization . Figure 9A added neofunctionalization , while Figure 9B added homomeric duplication to simple duplication and subfunctionalization . Simple duplication is defined as randomly selecting an existing node in the network , identifying the set of neighbors the selected node interacts with , and adding a new node to the network which interacts with an identical set of neighbors . Subfunctionalization is defined as removing each interaction from the newly-added node with a given probability . Neofunctionalization is defined as adding an interaction from the newly-added node to each existing node in the network with a given probability β . Homomeric duplication is defined as adding an interaction between the randomly-selected node ( i . e . , the progenitor ) and the newly-added node ( i . e . , the progeny ) with a given probability . Newly-added nodes having no interacting partners after going through the relevant evolutionary processes were discarded . Simulated networks were evolved until they reached 5794 nodes , the putative number of yeast genes . Each line plotted in the figure was based on the mean clustering coefficient of 100 networks for each of 80 loss probabilities: [0 . 20 , 0 . 21 , … , 0 . 99] . That is , each line is the result of 80×100 = 8000 generated networks . In the neofunctionalization plot probabilities 0 . 20 thru 0 . 39 were not calculated for β = 50 nor were probabilities 0 . 20 thru 0 . 22 for beta = 16 due to prohibitive runtime and/or overflow errors in the 32-bit numbers used to store the number of triangles and triples in the growing networks .	Molecular evolution studies have shown that the redundancy intrinsic to gene duplication may allow one gene duplicate to acquire a new function ( neofunctionalization ) or for both duplicates to each assume a subset of the ancestral gene's functions ( subfunctionalization ) . Studies of networks of interacting proteins and models of evolving protein interaction networks have shown that both subfunctionalization and neofunctionalization are widespread in protein evolution . Here , we present evidence that shows that the methods and models that have established neofunctionalization as a ubiquitous force in protein interaction network evolution are flawed and under reexamination support subfunctionalization , not neofunctionalization . We start by reviewing the methods and models that engender prolific subfunctionalization and neofunctionalization in evolution . We then critically approach neofunctionalization . We show that biases in protein interaction assays , failure to consider concurrent and subsequent gene duplications in evolutionary inferences , and an inability of theoretical models to reproduce empirical clustering have all led to neofunctionalization being erroneously identified as a pervasive force in evolution .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "computational", "biology/evolutionary", "modeling", "computational", "biology/systems", "biology" ]	2009	Questioning the Ubiquity of Neofunctionalization
The pathogen Mycobacterium tuberculosis employs a range of ESX-1 substrates to manipulate the host and build a successful infection . Although the importance of ESX-1 secretion in virulence is well established , the characterization of its individual components and the role of individual substrates is far from complete . Here , we describe the functional characterization of the Mycobacterium marinum accessory ESX-1 proteins EccA1 , EspG1 and EspH , i . e . proteins that are neither substrates nor structural components . Proteomic analysis revealed that EspG1 is crucial for ESX-1 secretion , since all detectable ESX-1 substrates were absent from the cell surface and culture supernatant in an espG1 mutant . Deletion of eccA1 resulted in minor secretion defects , but interestingly , the severity of these secretion defects was dependent on the culture conditions . Finally , espH deletion showed a partial secretion defect; whereas several ESX-1 substrates were secreted in normal amounts , secretion of EsxA and EsxB was diminished and secretion of EspE and EspF was fully blocked . Interaction studies showed that EspH binds EspE and therefore could function as a specific chaperone for this substrate . Despite the observed differences in secretion , hemolytic activity was lost in all M . marinum mutants , implying that hemolytic activity is not strictly correlated with EsxA secretion . Surprisingly , while EspH is essential for successful infection of phagocytic host cells , deletion of espH resulted in a significantly increased virulence phenotype in zebrafish larvae , linked to poor granuloma formation and extracellular outgrowth . Together , these data show that different sets of ESX-1 substrates play different roles at various steps of the infection cycle of M . marinum . Mycobacterium tuberculosis , the etiological agent for the disease tuberculosis ( TB ) , is still one of the most dangerous pathogens for global health [1] . Successful infection requires secretion of multiple virulence factors , facilitated by type VII secretion systems ( T7SS ) . Pathogenic mycobacteria have up to five T7SS , called ESX-1 to ESX-5 [2] , of which at least three are essential for growth and/or virulence [3 , 4] . The ESX-1 locus was the first T7SS to be identified . The loss of ESX-1 function in Mycobacterium bovis BCG is considered a decisive factor of attenuation of this vaccine strain [5] . Mouse infection experiments utilizing M . tuberculosis with a partial deletion in ESX-1 showed reduced granuloma formation , the characteristic pathological hallmark of mycobacterial disease [6 , 7] . Similarly , efficient granuloma formation , dissemination of disease and invasion of endothelial cells in the fish-pathogen Mycobacterium marinum is dependent on a functional ESX-1 secretion system [8–10] . More detailed analysis showed that ESX-1 substrates are required for phagosomal membrane rupture [11 , 12] . Thus far , about a dozen different proteins have been identified to be secreted through ESX-1 , which can be divided in three subgroups , the Esx proteins , the PE/PPE proteins and the Esp proteins . Of these substrates , the Esp proteins are ESX-1 specific [13] . The ESX-1 substrates EsxA ( ESAT-6 ) and EsxB ( CFP-10 ) are secreted as an antiparallel heterodimer [14] . Interestingly , the limited structural data available for PE and PPE proteins also show that these proteins form a heterodimer [15–17] . These heterodimers form a four-helix bundle and contain a YxxxD/E secretion motif directly after the helix-turn-helix on one of the Esx proteins and on the PE protein [15 , 18] . The ESX-1 substrate EspB forms a similar four helix bundle with the conserved secretion motif at the same position in the structure and therefore does not seem to require a partner protein [17 , 19] . EsxA and EsxB are most intensively investigated of the different ESX-1 substrates [11 , 20–22] and EsxA is thought to be responsible for ESX-1 related virulence determinants [11 , 21–24] . EspA and EspB have additionally been implicated to be important for virulence [25 , 26] . However , studying the exact role of each substrate is complicated , as deletion of esxA/esxB abolishes secretion of all different Esp proteins [8 , 27] , while espA and espB deletion mutants are unable to secrete EsxA/EsxB [25 , 27] . The ESX-1 secretion system consists of a membrane complex composed of the ESX conserved components ( Ecc ) EccB1 , EccCab1 , EccD1 and EccE1 [28 , 29] , which is stabilized by the MycP1 protein [29] . The ESX-1 secretion system additionally contains the cytosolic accessory components EspG1 and EccA1 . EspG functions as a specific chaperone of cognate PE/PPE substrates [30 , 31] and deletion of espG1 leads to a block in the secretion of PE35/PPE68_1 in M . marinum [31] . Loss of EspG1 in M . tuberculosis caused severe attenuation , both in cell infection and in mice [32] . EccA1 is a cytosolic AAA+ ATPase ( ATPases Associated with diverse cellular Activities ) , which is essential for the EsxA secretion in both M . tuberculosis and M . marinum [33 , 34] . The M . marinum eccA1-null strain has been shown to be attenuated in zebrafish larvae [34] . However , its exact function is not further characterized . In the M . marinum , the genes espG1 ( MMAR_5441 ) and eccA1 ( MMAR_5443 ) are separated in the esx-1 locus by espH ( MMAR_5442 ) . EspH-like proteins are unique for the ESX-1 system . EspD is a homologue of EspH , sharing 55% sequence identity in M . tuberculosis . EspD is encoded by the espACD locus , located more than 260 kb upstream of the ESX-1 gene cluster . Interestingly , M . tuberculosis EspD has a role in stabilizing the intracellular levels of the secreted substrate dimer EspA/EspC [35] . These observations suggest that EspH might function as a molecular chaperone . Here , we study the role of three accessory proteins EspG1 , EccA1 and EspH in M . marinum and could show that mutants in the corresponding genes displayed distinctive and contrasting virulence phenotypes , demonstrating that ESX-1 substrates play different roles in virulence . We additionally identified several potential new ESX-1 substrates . To study the role of accessory ESX-1 proteins EspG1 , EccA1 , and EspH in secretion , we created targeted knocked-out strains for espH and eccA1 and used the previously described espG1 knockout in M . marinum [31] . Deletion of the individual genes had no effect on bacterial growth in 7H9 medium ( S1A Fig ) . However , colonies of the eccA1 mutant appeared dry with a rough-surface , while no phenotypic change was observed for the ΔespG1 and ΔespH colonies . In addition , qRT-PCR on total RNA extractions showed that the different deletions had no polar effect on the transcription of neighboring genes ( S1B Fig ) . Next , secretion analysis was performed using immunoblotting and a set of antibodies directed against known ESX-1 substrates . GroEL2 was included as a loading and lysis control . As a known ESX-1 negative mutant we included the Mvu strain , which has a frameshift mutation in eccCb1 [4 , 36] ( Fig 1B , lane 6 and lane 7 , respectively ) . Our analysis showed that EsxA was no longer secreted in the ΔespG1 strain ( Fig 1B , lane 9 ) , similarly as observed in a previous study from our group [31] , but in contrast to the results obtained in M . tuberculosis [33] . Interestingly , the deletion of espH also resulted in a dramatic decrease in the secretion of EsxA ( Fig 1B , lane 10 ) . Surprisingly , and in contrast to what has been published previously [8 , 34] , we observed that secretion of EsxA was reduced in the eccA1 mutant , but not completely aborted ( Fig 1B , lane 8 ) . Next , we analyzed another ESX-1 substrate EspE ( MMAR_5439 ) , a highly abundant cell surface protein of M . marinum , which can be extracted from the cell surface using the mild detergent Genapol X-080 [37] . The surface localization of the ESX-5 dependent PE_PGRS proteins was included as controls . In the WT strain , EspE was secreted in two forms: a full-length protein of ~ 40 kDa and a putatively processed form of ~ 25 kDa ( Fig 1C , lane 6 ) . Surface localization of EspE was abolished in all the mutant strains ( Fig 1C , lane 7 to lane 10 ) . Notably , while EspE accumulated in the cell pellet of all non-secreting strains , this protein was not detected in the pellet fraction of the espH mutant ( Fig 1C , lane 5 ) , indicating that secretion of EspE was blocked at a different stage as compared to the other mutants . To confirm that the observed secretion defects were caused by the targeted mutations , complementation plasmids were constructed . Two different complementation plasmids were used: the first one includes the genomic region from espF ( MMAR_5440 ) to eccA1 ( MMAR_5443 ) , whereas in the second plasmid only the espG1-espH-eccA1 locus was present . Complementing the knockout strains with either of these plasmids fully restored the secretion of EsxA and EspE in all of the mutants ( Fig 1D and 1E ) . A major discrepancy with previous publications was our finding that EccA1 has a limited effect on EsxA secretion . Previously , Gao et al . showed , using the same M . marinum background strain , that EccA1 is crucial for ESX-1 secretion [8 , 34] . We realized that there is a difference in the growth conditions between the two studies; we used 7H9 medium whereas Gao et al . used Sauton medium [8 , 34] . To test whether the observed differences could be linked to a difference in growth condition , secretion analysis was performed on cultures grown in Sauton medium . Interestingly , whereas the results for ΔespG1 and ΔespH were identical ( Fig 2 , lane 9 and lane 10 , respectively ) , EsxA was no longer secreted in the eccA1 mutant strain ( Fig 2 , lane 8 ) , which shows that the role of EccA1 in EsxA secretion is dependent on the growth condition . The proteome of a number of ESX-1 targeted knockout strains of M . marinum has been determined previously [38] . However , this study did not include an espH mutant and the cell surface proteome was not analyzed . In order to obtain a comprehensive and detailed view , the complete secretomes of our mutant strains , the corresponding complemented strains and both the WT and ESX-1 secretion mutant eccCb1 were analyzed by mass spectrometry . As some ESX-1 substrates are efficiently secreted into the culture supernatant , while others mainly remain attached to the cell surface [37] , cells were grown with or without Tween 80 to study secreted proteins in the medium or the cell surface proteins , respectively . The cell surface proteins were extracted from the bacterial cells using Genapol X-080 . For the ESX-1 secretion ( eccCb1 ) mutant , a massive reduction in the secretion of all known ESX-1 substrates , i . e . EsxA ( MMAR_5449 ) , EsxB ( MMAR_5450 ) , EspB ( MMAR_5457 ) , EspC ( MMAR_4167 ) , EspE ( MMAR_5439 ) , EspF ( MMAR_5440 ) , EspJ ( MMAR_5453 ) , EspK ( MMAR_5455 ) and PPE68 ( MMAR_5448 ) , was observed , both in the cell surface-enriched fractions ( Fig 3A ) and the culture supernatants ( Fig 4A ) . These results are in line with published data [38] . Also the secretion of several other proteins , including the PE protein MMAR_2894 and PPE protein MMAR_5417 , was blocked , suggesting they are novel ESX-1 substrates . This notion is strengthened by the fact that these two proteins are homologous to the PE and PPE protein encoded by the esx-1 locus . For the other proteins that showed reduced spectral counts in the cell surface fractions it is more difficult to draw any conclusion . First of all , the difference in secretion levels are smaller as compared to the known ESX-1 substrates ( Fig 3 ) , but furthermore they lack known characteristics of T7SS substrates , such as the YxxxD/E secretion motif preceded by a predicted helix-turn-helix structure . The espG1 mutant showed similar secretion profiles as the eccCb1 mutant ( Fig 3B and Fig 4B ) , although the secretion of EspB , EspK and EspE seemed to be slightly less severely affected . This suggests that EspG1 is not only required as a chaperone for its cognate PE/PPE substrates , but plays a more central role in the secretion of all ESX-1 substrates . The secretion of all ESX-1 substrates returned to WT levels in the espG1 mutant carrying the pMV361::espF-eccA1 complementation plasmid ( S2A and S2B Fig ) . The secretome profiles of the eccA1 mutant in 7H9 medium showed only a mild reduction of ESX-1 substrates in both cell surface and supernatant fractions ( Fig 3D and Fig 4D ) . For instance , EsxA and EsxB secretion was five and two-fold decreased , respectively , while in the eccCb1 mutant the reduction of both was 10 fold ( Fig 4D ) . The substrates EspE , EspF , EspJ and EspK are more affected by the eccA1 mutation than the other substrates in both protein fractions . In concordance with the data obtained by immunoblotting , the complementation of the eccA1 mutant with pMV361::espF-eccA1 plasmid restored the secretion of all ESX-1 substrates ( S2A and S2B Fig ) . Deletion of espH resulted in a severe reduction of EspE and EspF ( Fig 3C ) , in line with our immunoblot analysis . This reduction was in fact almost complete , both in the fraction of the surface proteins ( determined LC-MS/MS ) and in the bacterial pellet ( determined by immunoblotting ) , which again suggests instability of intracellular EspE/EspF in the absence of EspH . This effect was restored when the complementation plasmid was introduced ( S2A and S2B Fig ) . Interestingly , the effects of the espH deletion on secretion of EsxA and EsxB was only mild as compared to the eccCb1 mutant , while the effects on other ESX-1 substrates , such as EspB , EspK and EspJ were also only minor ( Fig 4C ) . This indicates that ΔespH has a specific secretion defect for a subset of ESX-1 substrates and there is no substrate dependency between EspE/EspF and other Esp proteins . Surprisingly , we also identified some proteins that were present in significantly increased amounts in the cell surface enriched fractions of various mutants . One of these proteins is SecA2 , a cytosolic component of the Sec transport system and proposed to contribute to the virulence of M . tuberculosis and M . marinum [39 , 40] . SecA2 was present in higher amounts in all mutants except the ΔespH , suggesting a link with intracellular accumulation of EspE/EspF . Another intriguing observation is an increase of Mak in the ΔespG1 , ΔespH and the ΔeccA1 ( Fig 3B , 3C and 3D , respectively ) . Mak is a mycobacterial maltokinase whose function is involved in the glycan synthesis from trehalose [41] and considered to be essential for the growth of M . tuberculosis [42] . This could suggest that there is an indirect effect of ESX-1 secretion on the synthesis of the mycobacterial capsule . The observation that EspH mainly affects the secretion of EspE/EspF and that EspE could not be detected in the espH mutant pellet fraction raised the hypothesis that EspH could either regulate the transcription of espE/espF or stabilize EspE/EspF at the protein level . To get more information on the putative function of EspH we used the protein structure prediction program Phyre2 [43] . This analysis showed that part of EspH ( region between amino acid 65 and 135 ) is predicted to share structural similarity to YbaB proteins of Escherichia coli and Haemophilus influenza . Although the sequence identity with these proteins is low ( 15% ) the confidence of the structural homology is very high ( 97% ) . Because YbaB is reported to be a small DNA-binding protein that plays a regulatory role [44] , an effect on transcription regulation could be possible . Therefore , we measured the effect of espH deletion on espE and espF mRNA levels . Because the EsxA secretion was reduced in the espH mutant , esxA mRNA level was checked as well . Total mRNA was extracted from the WT MUSA , eccCb1 mutant and the ΔespH strain , and qRT-PCR was performed using primer sets for espE , espF and esxA . The results showed that the mRNA levels of all three genes were comparable to those of the eccCb1 mutant strain analyzed ( S3A Fig ) . Thus , we could disprove the possibility that EspH regulates espE at the transcriptional level . Next , we studied a direct interaction of EspH with EspE and/or EspF . Based on the high homology of EspE with EspA and EspF with EspC , we speculated that , similarly to EspC/EspA [45] , EspF might be secreted together with EspE . We therefore constructed a plasmid containing espE/espF in which espE was modified to express a C-terminal Strep tag . We also introduced a His tag at the C terminus of EspH in the espG1/espH/eccA1 complementation plasmid . Introduction of both plasmids in the WT and ΔespH mutant resulted in surface localized EspE , as judged by immunoblot analysis of the cell surface extracted protein preparations ( S3B Fig ) . These results show that the addition of the Strep tag to the C terminus of EspE and the His-tag to EspH did not affect the functionality of these proteins in the secretion process . To study the interaction of EspE and EspH , we overexpressed EspE-Strep/EspF and EspH-His in the eccCb1 mutant strain . The ESX-1 secretion system is defective in this strain and therefore EspE and EspH accumulate in the cytosol , which allows their analysis and co-purification . The subcellular localization of EspE and EspH was examined by a subcellular fractionation procedure , showing that EspE-Strep was partially soluble while EspH-His was exclusively present in the soluble fraction ( S3C Fig ) . Next , we used StrepTactin beads to purify Strep-tagged EspE from these soluble fractions . Immunoblot analysis showed that EspE-Strep was efficiently purified . Importantly , EspH-His , appearing as a ~ 25 kDa band , was only present in the elution fractions when expressed in the presence of EspE-Strep ( Fig 5A ) . In contrast , the ESX-1 substrates PPE68 and EsxA were not co-purified and both remained in the flow-through fraction . To confirm this EspE-EspH interaction , a reciprocal pull-down assay was performed using Ni-NTA beads and lysates of the eccCb1 mutant containing EspE-strep/EspF only or EspE-strep/EspF and EspH-His . Immunoblot analysis confirmed the efficient purification of EspH-His ( Fig 5B ) . Using anti-EspE on these samples showed co-elution of endogenous EspE only in the presence of the His-tagged EspH ( Fig 5B ) . Again , PPE68 and EsxA were only found in the flow-through fraction , indicating that they do not bind EspH . In conclusion , these data confirmed that EspH specifically interacts with EspE in the cytosol of M . marinum and this interaction is probably required for EspE secretion . ESX-1 functioning in M . marinum has been associated with lysis of red blood cells [8] . Because of this , the hemolysis assay has been employed as a model for the ESX-1-dependent lysis of ( phagosomal ) membranes [8] . Prior work suggested that the ESX-1 associated membrane lytic activity was mediated by EsxA through its pore-forming activity [21 , 46] . Because the deletion of espG1 , espH and eccA1 differently affected the secretion of EsxA , we examined to what extend these mutant strains were able to disrupt erythrocytes . While we confirmed that our WT strain showed hemolysis ( Fig 6A ) , both the eccCb1 and ΔespG1 mutant strain lost this ability , in line with the absence of ESX-1 substrates in the culture supernatant ( Fig 6A ) . Interestingly , the ΔespH and ΔeccA1 strains were also non-hemolytic , although these strains were still able to secrete EsxA to significant levels ( Fig 6A ) . The defects in hemolysis by the knockout strains were restored when the complemented plasmids were introduced into these mutant strains ( Fig 6B ) . As in the ΔespH and ΔeccA1 mutants mainly the secretion of different Esp proteins are specifically affected , our findings indicate that not a single ESX-1 substrate , such as EsxA , but a combination of different Esp proteins , are responsible for the hemolytic phenotype . To further characterize the function of the different ESX-1 substrate subsets , we used different phagocytic cells to study the ability of the mutant strains to survive and replicate within these cells . Phagocytic cells from mice ( RAW macrophage cell line ) and the protozoa Acanthamoeba castellanii were infected with green fluorescent protein ( GFP ) -expressing mycobacteria and infection levels were quantified by flow cytometry at different time points . As shown before , the eccCb1 mutant was strongly attenuated in both A . castellanii and RAW cells ( Fig 7; [47] ) , showing a 2-fold reduction in the number of infected cells after 24 h . As expected , based on the proteome profiles , the ΔespG1 strain showed an attenuated phenotype similar to the eccCb1 mutant . For the ΔespH mutant , the proportion of infected A . castellanii cells did not change over time ( Fig 7B ) , while in RAW macrophages a slight reduction of infected cells at 24 hpi could be observed ( Fig 7C , p = ns ) . Infection with the ΔeccA1 mutant resulted in an increase of infected cells over time , for both A . castellanii and RAW cells , and was therefore less attenuated as compared to the other mutants ( Fig 7B and 7D ) . Although this strain was able to infect A . castellanii to the same extend as the WT strain , infection with this mutant was not as successful as WT infection in RAW macrophages ( Fig 7A , ns; Fig 7C , p < 0 . 001 ) . Taken together , our data show the importance of espG1 in achieving successful infection of phagocytic cells , while the loss of eccA1 only marginally affects the ability of M . marinum to survive and replicate in a phagocytic host cell . These findings are in line with the proteomic analysis , i . e . the espG1 mutation has a strong effect on secretion of all ESX-1 substrates , while deleting eccA1 only results in a mild secretion defect . EspH , which seems to mainly influence EspE and EspF secretion , is also important for infecting phagocytes , but to a lesser extent than EspG1 . To study whether the individual ESX-1 proteins play a role during infection in vivo , we used the zebrafish larva-M . marinum infection model . Larvae were systemically infected with the fluorescently labeled mutant , complemented and WT strains and infection was analyzed 4-days post infection ( dpi ) by fluorescence microscopy . In addition , L-plastin staining was performed to visualize phagocytic cells in order to study the formation of early granulomas by confocal microscopy . Infection of zebrafish larvae with the ΔespG1 and ΔeccA1 mutant strains resulted in infection levels as expected from the previous experiments , i . e . the ΔespG1 showed a similar level of attenuation as the eccCb1 mutant , while the ΔeccA1 mutant infections were similar to wildtype infection ( Fig 8A , 8D and 8H for ΔeccA1; Fig 8B , 8F and 8J for ΔespG1 ) . Higher magnification of individual infection loci in ΔeccA1 infected larvae revealed recruitment of phagocytic cells and formation of early granulomas comparable to infection with WT ( Fig 8E for WT , n = 12 larvae; Fig 8I for ΔeccA1 , n = 8 larvae ) . In contrast , confocal imaging of ΔespG1 infected fish showed a predominance of single infected macrophages and formation of very small clusters of these infected macrophages similar to infection with the eccCb1 mutant ( Fig 8G for eccCb1 mutant , n = 10 larvae; Fig 8K for ΔespG1 , n = 7 larvae ) . Together , this shows that espG1 , but not eccA1 , plays a major role in early stages of infection in vivo . Moreover , since these strains show a comparable behavior during in vitro and in vivo infections , this indicates functional similarities for these genes in protozoa , mouse macrophages and zebrafish larvae . In contrast to the ΔespG1 and ΔeccA1 strain , the behavior of ΔespH in zebrafish larvae was completely different from its attenuated phenotype in vitro . Systemic infection of zebrafish larvae resulted in an increased bacterial load as compared to WT infection ( Fig 8C; p < 0 . 05 ) . Large bacterial clusters and a phenotype known as cording were seen in fluorescence images ( Fig 8L , arrow ) , especially at higher magnification of individual clusters ( Fig 8L , closed arrow , n = 15 larvae ) . Cording in zebrafish has been associated with extracellular growth [48] . In addition , very limited numbers of intact phagocytic cells and the presence of fluorescent spots suggestive for phagocytic cell debris were observed ( Fig 8L , open arrow ) . These observations raised the question whether this phenotype is still preceded by granuloma formation or if this mutant strain is preventing early granuloma formation by inducing rapidly host cell death . Therefore , larvae were systemically infected with either ΔespH or WT M . marinum as control and monitored daily for 4 consecutive days ( Fig 9 ) . Mycobacteria were phagocytosed by L-plastin positive phagocytic cells at 1 dpi in both groups ( Fig 9A and 9D ) . Subsequently , phagocytic cells were recruited and early granulomas started to form ( Fig 9B and 9E ) . However , at 4 dpi , in larvae infected with the ΔespH strain a strong decrease in phagocytic cells and increase in bacterial growth was observed ( Fig 9C and 9F ) . In the absence of phagocytic cells bacteria were able to show cording in both blood vessels ( Fig 9F , closed arrow ) and tissue ( Fig 9F , open arrow ) . Taken together , the ΔespH mutant seems to have a host-specific or in vivo-specific effect , illustrated by a hypervirulent phenotype seen in zebrafish larvae , but not in cell infections in vitro . Therefore , our data indicates that EspH is not required for initial phagocytosis , recruitment of cells and primary establishment of early granulomas , but this protein , and therefore a subset of ESX-1 substrates , seems to be essential for the maintenance of a stable granuloma . A number of studies have shown that the mycobacterial ESX-1 system plays a pivotal role in mycobacterial pathogenesis [6 , 21 , 27 , 33] . The system affects virulence through secretion of protein effectors with host-modulatory effects . Here , we show that EccA1 is not strictly required for the secretion of ESX-1 substrates . The finding that EccA1 is important for secretion is in line with previous reports [8 , 34] , but the fact that the role of EccA1 is depending on the growth medium is entirely surprising . This difference could also explain the variable results described for the role of EccA1 in EsxA secretion by M . tuberculosis [49] . Of all ESX-1 substrates , EspE , EspF , EspJ and EspK secretion was mostly affected in our eccA1 mutant strain , while secretion of EspB , EsxA/EsxB and PE/PPE was hardly altered . An interesting observation here is the discrepancy between the active secretion of EsxA in the ΔeccA1 strain and at the same time loss of hemolytic activity . Although this observation has been described before , this was always linked to a reduced secretion of EsxA in these strains [8 , 34] . In a recent study , the importance of EsxA in lysing membranes was questioned [50] . Our results also supports an alternative mechanism: we find a strong correlation between ESX-1 functionality and hemolysis , but this correlation is not seen for EsxA secretion . Our finding is in line with several other recent studies , who showed that both EsxA and the cell-surface lipid PDIM are important for phagosomal rupture and escape by M . tuberculosis [51–53] . We propose that it is not the loss of secreted EsxA , but the loss of ( multiple ) surface-exposed Esp proteins that results in hemolytic deficiency . Even though the ΔeccA1 strain lost its ability to induce membrane lysis , virulence in isolated phagocytes and in zebrafish larvae was only mildly affected in our study . This is in contrast with other studies , who described an attenuated phenotype for similar mutants in M . tuberculosis and M . marinum in murine macrophages and zebrafish [8 , 34] . The latter observations were made after a longer incubation time , which might explain the discrepancy with our study . Distinct phenotypes of the eccA1 mutant in different host cells have also been reported in a genome-wide transposon mutagenesis study [47] . Here , transposon insertions in M . marinum E11 eccA1 led to severe attenuation in mammalian phagocytic cells but these mutants were hypervirulent in protozoan cells [47] . This suggests that M . marinum can employ host-specific virulence mechanisms to adapt to different intracellular environments and that EccA1 might be essential for secretion and virulence under specific circumstances or in a subset of specific hosts . The role of EspG as a specific chaperone for the recognition and secretion of cognate PE/PPE proteins has been well established in M . marinum [30 , 31] . Our extracellular proteomic study not only confirms the loss of PE/PPE substrate secretion in the M . marinum ΔespG1 strain , but also reveals the secretion block of other ESX-1 dependent substrates , including EsxA/EsxB . This effect on EsxA/EsxB secretion however was not observed in an M . tuberculosis espG1 knock-out strain [33] . EspG proteins bind specifically to the extended helices of the PPE protein , which are absent in Esx proteins . Therefore , the strong effect of espG1 deletion on Esx ( and also Esp ) protein secretion in M . marinum is likely indirect due to a mutual dependency in secretion among the ESX-1 substrates [27 , 31 , 35] . This co-dependency of PE/PPE and other ESX-1 substrates for secretion is possibly less strict in M . tuberculosis , explaining that mutating espG1 had no effect on EsxA/EsxB secretion in this species . Because of the severe secretion defect of all detectable ESX-1 substrates , the M . marinum espG1 mutant is non-hemolytic and strongly attenuated in macrophage and amoeba , which is in good agreement with previous reports [8] . Furthermore , the loss of espG1 resulted in a strong attenuation in zebrafish , to the same extend as the eccCb1 mutant . Our most significant and surprising results were obtained for EspH . EspH is specific for the ESX-1 secretion system and is highly conserved among pathogenic mycobacterial species , including M . tuberculosis and M . leprae . The latter species has been streamlined into a minimal genome by a process of extensive genome decay . In our study , deletion of this gene abolishes the expression and secretion of two specific ESX-1 substrates EspE and EspF . Furthermore , we could show that EspH specifically interacts with EspE in the cytosol , indicative of chaperone activity . However , the Phyre2 structural prediction program [43] indicated that EspH is shares similarity to YbaB . The first structural study on YbaB strongly indicated that this protein binds DNA as a dimer [44] . However , recent studies indicated that the function of YbaB might be more diverse . One study showed that YbaB is associated with and a target of ClpYQ proteases in E . coli [54] , while another study indicated that overexpression of YbaB enhanced the production of heterologous membrane proteins [55] . Based on the direct interaction of EspH with EspE and that the EspH-like protein EspD stabilizes intracellular EspA/EspC substrates [35] , we propose that these YbaB-like proteins encoded by the esx-1 cluster of pathogenic mycobacteria function as dedicated chaperones for specific ESX-1 substrates . Recently , a study of M . tuberculosis EspL also predicted a high resemblance to YbaB [56] , making it tempting to speculate that EspL may as well function as a dedicated chaperone , for instance the ESX-1 substrates encoded by the adjacent genes EspK or EspB . It becomes clear that multiple chaperones , such as EspG1 , EspD and EspH , are responsible for stabilizing their cognate substrates PE35/PPE68 , EspC/EspA and EspF/EspE , respectively . Interestingly , secretion of other substrates of the ESX-1 system , such as EspB , EspK and EspJ , did not seem to be affected by disruption of the espH gene . A similar phenotype was observed previously in an espA::tn mutant of M . tuberculosis [26] , where secretion of EsxA/EsxB but not EspB was aborted . These results show that interdependence in ESX-1 secretion is not a general feature . Deletion of espH did result in reduced secretion of EsxA/EsxB , which was not due to differences in mRNA levels . This hints towards a possible regulation mechanism between the secretion of the central components EsxA/EsxB and the individual Esp substrate ( pairs ) but not among the Esp proteins themselves . The espH mutant strain showed a loss of hemolytic activity and a reduction of intracellular growth in phagocytic host cells in our study . Strikingly , zebrafish larvae were heavily infected with this mutant strain and showed even hypervirulence at later time points , even though EsxA/EsxB secretion was reduced in this mutant . More detailed analysis showed that initial phagocytosis and primary establishment of an early granuloma was not affected in this mutant . Eventually , a stable cluster of immune cells could not be maintained in larvae infected with the espH mutant , with subsequent extracellular bacterial outgrowth and apparent phagocyte death . The discrepancy between in vitro and in vivo results indicate an essential role for a , yet unknown , host factor involved . It is tempting to speculate that EspE/EspF , the two proteins that are most severely affected by the espH deletion , interact with this host factor in order to induce the homeostatic balance between host and pathogen in developing granulomas . Furthermore , because EsxA and EsxB secretion was diminished , other ESX-1 substrates in addition to these central substrates might be involved in the infection process . A candidate might be EspB , whose secretion was not affected in espH mutant strain , and was shown to be able to facilitate M . tuberculosis virulence in vitro and in vivo in an EsxA-independent way [26] . In summary , this study highlights the complexity of the ESX-1 secretion machinery . We unravel valuable information about the functions of the individual ESX-1 components EccA1 , EspG1 and EspH , all having their unique role in secretion of the different substrate classes . We can conclude that ESX-1 has several different sets of substrates that are involved in distinctive processes required for virulence . All M . marinum strains used in this study were derived from the wild-type strain MUSA [57] . The eccCb1 ( MVU ) strain was previously identified as an ESX-1 secretion mutant with a spontaneous out of frame mutation in eccCb1 [36] and also the knock-out strain espG1 was described before [31] . The knockout strains of eccA1 and espH were generated using the mycobacteriophage approach ( see below ) . All strains were routinely cultured on Middlebrook 7H10 plates or in Middlebrook 7H9 medium ( Difco ) containing ADC supplement or on Sauton medium [58] supplemented with 2% glycerol and 0 . 015% Tween-80 . When required , 0 . 05% Tween-80 and the appropriate antibiotics were added ( 25 μg/ml kanamycin ( Sigma ) and/ or 50 μg/ml hygromycin ( Roche ) . M . marinum cultures and plates were incubated at 30°C . E . coli TOP10F’ was used for cloning experiments to generate the complemented plasmids and was grown at 37°C on LB plates and in LB medium . Different antibiotics were added to the cultures or plates when necessary at similar concentrations as for M . marinum cultures . All DNA manipulation procedures followed standard molecular biology protocols . Primers were synthesized and purified by Sigma . Phusion polymerase , restriction enzymes and T4 DNA ligase were obtained from New England Biolabs ( NEB ) . Macrogen performed DNA sequencing . Bacterial RNA was extracted from various M . marinum strains as described previously [31] and cDNA was synthesized using SuperScript VILO cDNA Synthesis kit ( Thermoscientific ) according to manufacturer protocol . For the PCR mix the SYBR GreenER qPCR SuperMix ( Thermoscientific ) was used according to manufacturer instructions , including the addition of ROX dye reference . qRT-PCR was performed in Applied Biosystems 7500 Fast system . The primer sequences used for qRT-PCR are listed in S3 Table . Controls without reverse transcriptase were done on each RNA sample to rule out DNA contamination . The sigA gene was used as an internal control . An eccA1 and espH knockout was produced in M . marinum MUSA by allelic exchange using a specialized transducing mycobacteriophage as previously described [59] . High phage titers were prepared following the previously described protocol [31] . Subsequently , the M . marinum wild-type strain was incubated with the corresponding phage to create eccA1 and espH knockouts . Colonies that had deleted the endogenous eccA1 and espH were selected on hygromycin plates and tested for sucrose sensitivity , induced by the presence of the sacB gene . The deletion was confirmed by PCR analysis and sequencing . To remove the resistance and sacB gene , a temperature sensitive phage encoding the γδ-resolvase ( TnpR ) ( a kind gift from Apoorva Bhatt , University of Birmingham , UK ) was used , generating an unmarked deletion mutation . M . marinum cultures were grown in 7H9 medium supplemented with ADC and 0 . 05% Tween 80 to mid-logarithmic phase . Bacteria were washed two times and set to OD600 of 0 . 35 in 7H9 medium containing 0 . 2% glycerol , 0 . 2% dextrose and 0 . 05% Tween 80 for overnight growth . Supernatants were filtered using 0 . 2 μm filter , concentrated by trichloric acid ( TCA ) precipitation , washed with acetone and the supernatant pellets were resuspended in solubilisation/denaturation ( S/D ) buffer ( containing 100mM DTT and 2% SDS ) . Bacteria were washed once with PBS . Aliquots were taken for the whole cell lysate preparations and for Genapol X-080 extraction of cell-surface-attached proteins . Bacteria were incubated with 0 . 5% Genapol X-080 in PBS for 30 minutes with head-over-head rotation at room temperature . Genapol extracted supernatants were denatured in S/D buffer . The bacterial pellet and Genapol extracted cells were lysed by bead beating for 1 minute two times after which S/D buffer was added . All samples were boiled for 10 minutes at 95°C before loading on SDS-PAGE . For His-tag pulldown , mycobacterial cultures grown to an OD600 of 1 . 0 were incubated for 1 h with 100 g/ml ciprofloxacin ( Sigma ) , pelleted , washed twice with PBS , and subsequently resuspended in PBS supplemented with Complete protease inhibitor mixture ( Roche Applied Science ) , 1 mM EDTA , and 10 mM imidazole . Cells were broken by two-times passage through a One-Shot cell disrupter ( Constant Systems ) at 0 . 83 kbar . Unbroken cells were spun down by repeated centrifugation at 3000 g , and subsequently the cell envelope and soluble fractions were separated by ultracentrifugation at 100 , 000 g for 1hr . Membrane-cleared lysates of M . marinum expressing proteins of interest were incubated with Ni-NTA agarose beads ( Qiagen ) for 1 h at room temperature with head-over-head rotation . After washing the beads five times with phosphate buffer containing 50 mM NaH2PO4 and 300 mM NaCl , ( pH 8 . 0 ) , supplemented with 20 mM imidazole , bound proteins were eluted three times by incubation with phosphate buffer containing 400 mM imidazole . Immunoprecipitation of strep-tagged proteins was performed using the Strep-Tactin Sepharose kit ( IBA ) , following the manufacturers protocol . Proteins were separated by SDS-PAGE and stained with Coomassie Brilliant Blue G-250 ( CBB; Bio-Rad ) , or transferred to nitrocellulose membranes by Western blotting . The membranes were then incubated with different antibodies followed by enhanced chemiluminescence . Primary antibodies used in this study include: anti- GroEL2 ( CS44 , Colorado state university ) , anti-PE_PGRS antibody ( 7C4 . 1F7 ) [36] , anti-EsxA ( Hyb76-8 ) [60] , polyclonal anti-EspE and anti-PPE68 [61 , 62] . To investigate the cell-surface attached proteome , samples for LC-MS/MS analysis were prepared using the mild detergent Genapol X-080 as previously described [63] . To prepare the secreted material , the M . marinum MUSA wild-type and the studied ESX-1 mutant and complemented strains were grown to stationary phase in 7H9 medium supplemented with ADC , 0 . 2% glycerol and 0 . 05% Tween 80 . The supernatant fractions containing secreted proteins were collected and spun at 2500 × g for an additional 20 min at 4°C and subsequently filtered through a 0 . 2 μm pore size membrane to remove residual cells and cell debris . The filtered supernatants were 20 times concentrated using Amicon Ultra-15 Centrifugal 3 kDa molecular weight cut off membrane at 4°C . The retained proteins were TCA precipitated , pelleted , washed in acetone , dried and resuspended in S/D sample buffer to the corresponding OD of 200 units/ml . All samples were analyzed with SDS-PAGE and CBB staining . Total protein lanes of cell surface and culture supernatant proteins were excised in 3 or 1 fragment ( s ) per lane , respectively , and analyzed by LC-MS/MS as described before [63] . The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD008905 . M . marinum strains were grown in 7H9 medium supplemented with ADC and 0 . 05% Tween-80 till the mid-logarithmic phase . All strains were washed once with PBS and inoculated in 7H9 medium with or without Tween-80 at 0 . 35 OD600/ml and inoculated for 20 hours . Bacteria were collected by centrifugation , washed once in PBS and diluted in fresh DMEM medium without phenol red ( Gibco , Life technologies ) . Bacteria were quantified by absorbance measurement at OD600 with an estimation of 2 . 5108 bacteria in 1 ml of 1 . 0 OD600 . At the same time , defibrinated sheep erythrocytes ( Oxoid-Thermo Fisher , the Netherlands ) were washed five times and diluted in the same fresh DMEM medium . 4 . 2107 erythrocytes and 1*108 bacteria were added for one reaction of 100 μl in a round-bottom 96 well-plate , gently centrifuged for 5 minutes and incubated at 32°C . After an incubation of 3 hours , cells were resuspended , centrifuged and 80 μl of supernatants were transferred to a flat-bottom 96-wells plate and measured at an absorbance of 405nm to quantify hemoglobin release . The mouse macrophage line RAW264 . 7 ( American Type Culture Collection ) was cultured in RPMI 1640 with Glutamax-1 ( Gibco ) supplemented with 10% fetal bovine serum ( FBS; Gibco ) , 100 U of penicillin/ml , 100 μg of streptomycin/ml at 37°C , 5% CO2 . A total of 3 × 107 cells was seeded in T175 flasks ( Corning ) . Acanthamoeba castellanii was seeded in T175 flasks and grown in PYG medium , which is 0 . 4M MgSO4 . 7H2O , 0 . 05M CaCl2 , 0 , 1 M Sodium citrate . 2H2O , 0 . 05M Fe ( NH4 ) 2 ( SO4 ) 2 . 6H2O , 0 . 25M Na2HPO4 . 7H2O , 0 . 25M KH2PO4 in distilled water with 2% proteose peptone ( W/V , BD 211684 ) and 0 . 01% yeast extract . After pH adjustment to 6 . 5 , 2M glucose was added . All bacterial strains were grown until the exponential growth phase , washed with 0 . 05% Tween 80 , spun down and resuspended in RPMI medium . RAW macrophages were infected with a MOI of 5 for 3 hours and incubated at 30°C , 5% CO2 . Cells were washed in RPMI to remove extracellular mycobacteria and either analyzed immediately or incubated for another 21 hours at 30°C , 5% CO2 . A . castellanii ( ATCC 30234 ) infection was performed with a MOI of 1 , 3 , 9 , 27 , 54 , and 108 to determine optimum MOI , for the remaining experiments MOI 3 was chosen . Incubation for 3 hours or 24 hours was done at 30°C , 5% CO2 . Uptake of strains in host cells was quantified for all cell lines with a BD Accuri C6 flow cytometer ( BD Biosciences ) with a 488-nm laser and 585/40-nm filter to detect mEos3 . 1 . A minimum of 5000-gated events was collected per sample per time point , percentage of living cells , percentage of infected cells and median fluorescent intensity per cell was analyzed using BD CFlow software . Injection stocks were prepared by growing bacteria until the logarithmic phase ( OD600 of 0 . 7–1 ) . Bacteria were spun down at low speed for 1 min to remove the largest clumps , washed with 0 . 3% Tween-80 in phosphate buffered saline ( PBS ) and sonicated shortly for declumping . Bacteria were than resuspended in PBS with 20% glycerol and 2% PVP and stored at −80°C . Before use , bacteria were resuspended in PBS containing 0 . 17% ( V/V ) phenol red ( Sigma ) to aid visualization of the injection process . Transparent casper zebrafish larvae [64] were removed from their chorion with tweezers and infected at 1 day post fertilization ( dpf ) via the caudal vein with bacterial suspension containing 50–200 CFU . Injection was performed as described previously [65] . To determine the exact number of bacteria injected , the injection volume was plated on 7H10 plates containing the proper antibiotic selection . At 4 days post infection ( dpi ) larvae were analyzed with a Leica MZ16FA fluorescence microscope . Bright field and fluorescence images were generated with a Leica DFC420C camera . Infection levels were quantified with a custom-made fluorescent pixel counting software . The software is in house developed and freely available under MIT license . Following analysis , larvae were fixed overnight in 0 . 4% ( V/V ) paraformaldehyde ( EMS , 100122 ) in PBS , washed and stored in PBS for immunohistochemistry . All procedures involving Danio rerio ( zebrafish ) larvae were performed in compliance with local animal welfare laws and maintained according to standard protocols ( zfin . org ) . The breeding of adult fish was approved by the institutional animal welfare committee ( Animal Experimental licensing Committee , DEC ) of the VU University medical center . All protocols adhered to the international guidelines specified by the EU Animal Protection Directive 86/609/EEC , which allows zebrafish larvae to be used up to the moment of free-living ( approximately 5–7 days after fertilization ) . In the current study , zebrafish larvae were used between 1 and 5 days post fertilization . Larvae were rinsed with 1% PBTx , ( 1% Triton X-100 in PBS ) , permeated in 0 . 24% trypsin in PBS and blocked for 3 hours in block buffer ( 10% normal goat serum ( NGS ) in 1% PBTx ) . Samples were incubated with anti-L-plastin [1:500 ( V/V ) dilution] in antibody buffer ( PBTx containing 1% ( V/V ) NGS and 1% ( W/V ) BSA ) overnight at RT . Samples were washed with PBTx , incubated for 1 hour in block buffer and stained with an Alexa-Fluor-647 goat-anti-rabbit antibody ( Invitrogen A21070 , 1:400 ) , overnight at 4°C . Confocal analysis was performed on larvae , embedded in 1% low melting-point agarose ( Boehringer Mannheim , 12841221–01 ) in an 8-well microscopy μ-slide ( ibidi ) , Analysis was performed with a confocal laser scanning microscope ( Leica TCS SP8 X Confocal Microscope ) . Leica Application Suite X software and ImageJ software were used to adjust brightness and contrast and to create overlay images and 3D models . Graphs were made using Graph Pad Prism 6 . 0 . Pixel counts were logarithmic transformed; error bars represent mean and standard error of the mean . A one-way ANOVA was performed followed by a Bonferroni’s multiple comparison test to analyze statistical significance . Graphs with results of RAW264 . 7 and A . castellanii infection experiments show percentage-infected cells of total cells; error bars represent mean and standard error of the mean . Data representing the fold change between 3 and 24 hpi was logarithmic transformed . A two-way ANOVA followed by a Sidak’s multiple comparison test was performed for statistical significance .	M . tuberculosis is a facultative intracellular pathogen that has an intimate relationship with host macrophages . Proteins secreted by the ESX-1 secretion system play an important role in this interaction , for instance by orchestrating the escape from the phagosome into the cytosol of the macrophage . However , the exact role of the ESX-1 substrates is unknown , due to their complicated interdependency for secretion . Here , we study the function of ESX-1 accessory proteins EccA1 , EspG1 and EspH in ESX-1 secretion in Mycobacterium marium , the causative agent of fish tuberculosis . We found that these proteins affect the secretion of different substrate classes , which offers an approach to study the roles of these substrate groups . An espG1 deletion broadly aborts ESX-1 secretion and thus resulted in severe attenuation in a zebrafish model for tuberculosis , whereas EccA1 is only crucial under specific growth conditions . The most surprising results were obtained for EspH . This protein seems to function as a molecular chaperone for EspE and is as such involved in the secretion of a small subset of ESX-1 substrates . Disruption of espH showed a dual character: whereas this gene is essential for the successful infection of macrophages , deletion of espH resulted in significantly increased virulence in zebrafish larvae . These data convincingly show that different subsets of ESX-1 substrates play different roles at various steps in the mycobacterial infection cycle .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "bacteriology", "protein", "transport", "blood", "cells", "medicine", "and", "health", "sciences", "fish", "immune", "cells", "pathology", "and", "laboratory", "medicine", "pathogens", "cell", "processes", "immunology", "microbiology", "vertebrates", "animals", "animal"...	2018	EspH is a hypervirulence factor for Mycobacterium marinum and essential for the secretion of the ESX-1 substrates EspE and EspF
Gut immunity is regulated by intricate and dynamic mechanisms to ensure homeostasis despite a constantly changing microbial environment . Several regulatory factors have been described to participate in feedback responses to prevent aberrant immune activity . Little is , however , known about how transcriptional programs are directly tuned to efficiently adapt host gut tissues to the current microbiome . Here we show that the POU/Oct gene nubbin ( nub ) encodes two transcription factor isoforms , Nub-PB and Nub-PD , which antagonistically regulate immune gene expression in Drosophila . Global transcriptional profiling of adult flies overexpressing Nub-PB in immunocompetent tissues revealed that this form is a strong transcriptional activator of a large set of immune genes . Further genetic analyses showed that Nub-PB is sufficient to drive expression both independently and in conjunction with nuclear factor kappa B ( NF-κB ) , JNK and JAK/STAT pathways . Similar overexpression of Nub-PD did , conversely , repress expression of the same targets . Strikingly , isoform co-overexpression normalized immune gene transcription , suggesting antagonistic activities . RNAi-mediated knockdown of individual nub transcripts in enterocytes confirmed antagonistic regulation by the two isoforms and that both are necessary for normal immune gene transcription in the midgut . Furthermore , enterocyte-specific Nub-PB expression levels had a strong impact on gut bacterial load as well as host lifespan . Overexpression of Nub-PB enhanced bacterial clearance of ingested Erwinia carotovora carotovora 15 . Nevertheless , flies quickly succumbed to the infection , suggesting a deleterious immune response . In line with this , prolonged overexpression promoted a proinflammatory signature in the gut with induction of JNK and JAK/STAT pathways , increased apoptosis and stem cell proliferation . These findings highlight a novel regulatory mechanism of host-microbe interactions mediated by antagonistic transcription factor isoforms . The innate immune system of mammals and insects is regulated by intracellular signaling pathways and transcriptional programs that show remarkable signs of evolutionary conservation . Well-known examples are the Toll/Toll-like receptor ( TLR ) , immune deficiency ( IMD ) /tumor necrosis factor-α ( TNF-α ) , JAK/STAT and JNK signaling pathways and their respective downstream transcriptional activators , nuclear factor kappa B ( NF-κB ) , STAT and AP-1 , which regulate innate immune responses in both Drosophila and mammals [1–4] . Pathway activation triggers a vast set of genes that encode effector molecules such as antimicrobial peptides ( AMPs ) and cytokines , which in mammals also support the induction of adaptive immune responses [5] . The underlying regulation is complex , especially in the intestine and other barrier epithelia that are in constant contact with the commensal microbiota . Improper control of the innate immune system and loss of tissue homeostasis can cause inflammation and other autoimmune diseases , and may lead to system failure and early death [1 , 6] . In Drosophila , the IMD and Toll pathways and downstream NF-κB homologs , Relish ( Rel ) [7] and Dorsal-related immunity factor ( Dif ) [8] , are crucial activators of immune genes in response to infection . About 25% of Drosophila immune-regulated genes ( DIRGs ) are , however , expressed independently of these pathways [9] . Additional transcriptional activators [10–13] as well as repressors [14–17] have been implicated in the immune response during specific conditions . However , little is known concerning how such factors compete for the same targets and interact to balance responses and maintain homeostasis . The POU/Oct transcription factor family is a subclass of the homeodomain proteins present in all metazoans [18] and encompasses crucial regulators of developmental decisions , metabolism , immunity and cancer [19] . Human Oct-1 ( POU2F1 ) was originally discovered as a regulator of adaptive immune responses [20] and has more recently been shown to regulate numerous target genes involved in both innate and adaptive immunity , stress resistance , metabolism and cellular proliferation [21] . In a yeast screen for transcriptional regulators of innate immune response genes , we have previously isolated three candidates from the Drosophila POU/Oct family [22] . The Oct-1 homolog , nub , was subsequently found to encode a negative regulator ( Nub-PD ) of NF-κB/Rel-driven immune gene expression in flies [15] . Nub-PD is necessary to control aberrant gene activation as nub1 mutant flies are short-lived and have a severely distorted gut microbiota [23] . Both Oct-1 and Nub regulate target genes by binding to the canonical octamer DNA sequence motif ( ATGCAAAT ) or variants thereof , via their C-terminal POUS and POUH domains [24–26] . Thus , the DNA-protein interaction surfaces appear conserved , further emphasizing the evolutionary relationship of these ancient transcriptional regulators . It has earlier been reported that alternative transcript forms of nub exist [27] . However , functional studies have , to our knowledge , been focused solely on the Nub-PD isoform . In this study , we demonstrate that nub encodes a novel isoform , Nub-PB , which is a strong activator of immune gene expression . Furthermore , both isoforms are expressed in midgut enterocytes and regulate the same immune target genes antagonistically . We show that such tuning of the transcriptional output of Nub target genes is crucial for host immunity , fitness and survival . Recent public annotations indicate two promoters of the nub gene , which in combination with promoter-specific splicing produce two independent proteins , Nub-PB ( 104 kDa ) and Nub-PD ( 65 kDa; Fig 1A ) . Both proteins contain identical C-terminal domains , in which the DNA-binding POUS and homeodomain ( POUH ) are located . Thus , Nub-PB and Nub-PD are expected to bind the same DNA sequences , while abilities for protein-protein interactions are likely to differ due to the distinct N-termini of the proteins . Initial experiments were aimed at detecting and elucidating any immune-regulatory features of Nub-PB . We applied the Gal4-UAS system and constructed a UAS-nub-RB line to drive its overexpression in immune competent tissues ( midgut and fat body ) using the c564-Gal4 driver . To simultaneously detect the ability of Nub-PB to regulate AMP gene expression , we combined c564>nub-RB with a lacZ reporter construct for Cecropin A1 ( CecA1 ) . This resulted in very prominent β-galactosidase activity in the abdominal fat body ( Fig 1B ) . Nub-PB hence appears to function as a transcriptional activator of immune effector genes , in stark contrast to Nub-PD . We carried on by analyzing the global transcriptional profiles of c564>nub-RB in comparison to driver controls ( c564>+ , from hereon referred to as wild type [wt] ) . To enable expression analysis of fat body and gut independently , mRNA from the fly body without head and gut ( “rest” ) and the digestive system ( “gut” ) were extracted and analyzed separately . The raw data were normalized , preprocessed and filtered to remove genes that were not expressed above the detection level ( S1 Table; Methods ) . Unbiased principal component analysis ( PCA ) , correlation analyses and hierarchical clustering showed that the tissue variable ( “rest” versus “gut” ) accounted for the largest distinction , as expected ( PC1 , explains 47% of the variance; Figs 1C and S1A ) . The PCA furthermore indicated that genotype accounted for the second largest separation in the dataset ( PC2 explains 9 . 5%; most apparent in “gut” ) . After adjusting p-values for a false discovery rate of 1% , 1177 ( rest ) and 545 ( gut ) probes indicated significant differential expression ( S1 Table ) , of which 132 were in common for both tissues ( S1B Fig , S2 Table ) . Gene set enrichment analyses ( GSEA ) according to “Biological process” of upregulated transcripts in respective tissue resulted in a single major GO cluster encompassing various aspects of immunity ( Fig 1D and 1E , S3 Table ) . Of transcripts that were found downregulated , the only significant subnode in “rest” was “proteolysis” ( also significant for upregulated probes in “rest” ) , whereas those from gut samples formed GSEA clusters of several overlapping cellular and developmental processes such as cell fate commitment , organ development , Notch signaling pathway and sensory responses along with GO categories of antibacterial defense and transcriptional regulation , suggesting a multifaceted role of Nub-PB in this organ ( S2 Fig , S3 and S4 Tables ) . To detect additional DIRGs , the dataset was manually curated and run against a collection of putative immune genes ( https://lemaitrelab . epfl . ch/resources , accessed February 2015 ) , which yielded 152 DIRGs in total ( Fig 1F , S5 Table ) . Overall , a striking coherence with well-characterized DIRGs typical for activated Toll , IMD and JAK/STAT pathways was observed . In summary , the global expression analysis demonstrated a broad immune activation following nub-RB overexpression , which further indicates that Nub-PB is a transcriptional activator of immune genes . A comparative analysis of the differentially expressed genes in response to Nub-PB overexpression ( this study ) and nub1 mutant flies ( disrupts Nub-PD , but not Nub-PB expression ) [15 , 27] revealed an extensive overlap as 65 immune-regulated genes were upregulated in both transcriptional profiles ( S6 Table ) . Out of the most highly expressed immune-process genes in “rest” ( FC>10 , 26 genes ) , 25 were upregulated at least 2-fold in nub1 mutants . To compare the capacity of Nub-PB and Nub-PD to regulate AMP genes , c564-driven overexpression of either transcriptional regulator was performed in parallel experiments ( Fig 2A ) . In validation of the transcriptional profiling , overexpression of nub-RB promoted strong upregulation of all eight assayed AMP genes ( AttA , AttB , CecA1 , CecC , DptA , Drs , Drsl2 and Drsl3 ) from whole fly extracts . Conversely , overexpression of nub-RD decreased CecA1 , DptA , Drsl2 and Drsl3 significantly compared to controls . We next evaluated the combined effect of the two isoforms in co-overexpression assays . To circumvent developmental and secondary effects , a temperature sensitive ( ts ) c564-Gal4; Tub-Gal80 driver line was applied . This resulted in expression levels of AttA , CecA1 , CecC and DptA similar to those in the control cohort ( Fig 2B ) , indicating that the two Nub isoforms overall neutralize each other’s activity . Co-expression did not dampen the increased expression of Drsl3 , whereas Drsl2 was partially restored to control levels , suggesting that Nub-PB induces these targets with greater affinity than the corresponding repression mediated by Nub-PD . We conclude that the Nub isoforms are able to regulate the expression of several AMP genes in an antagonistic manner . It has previously been shown that Nub-PD is a repressor of Rel-target genes in the absence of infection [15] . The observation that Nub-PB overexpression activates the same gene set ( S6 Table ) prompted us to explore its putative cooperation with Rel . Co-transfection of mbn-2 cells with nub-RB and Rel induced CecA1-luciferase expression significantly stronger ( 42 . 8-fold ) than single transfections ( nub-RB , 18 . 6-fold; Rel , 2 . 2-fold ) , which also indicated a synergistic effect ( Fig 2C ) . We therefore hypothesized that Nub-PB acts as a co-activator of Rel . To investigate this , AMP expression was assayed in whole flies following overexpression of nub-RB in a Rel mutant ( RelE20 ) or wild type background ( Fig 2D–2F ) . To induce a robust IMD pathway response , separate cohorts were subjected to systemic infection with Enterobacter cloacae β12 . Three AMP genes were assayed based on their known dependency of Rel and proximal Nub/Oct-sites: DptA ( Rel-dependent , one Nub/Oct-site ) , CecA1 ( Rel-dependent , several Nub/Oct-sites ) [15 , 28] and Drsl2 ( Rel-independent , one Nub/Oct-site ) [29] . Of note , binding of Nub to the proximal promoter region of DptA and CecA1 has been demonstrated biochemically [15] . Overexpression of Nub-PB was sufficient to drive expression of CecA1 , but not DptA , in absence of Rel and independent of infection status ( Fig 2D and 2E ) . Drsl2 , specifically expressed in the gut via the JAK/STAT pathway [29] was , as expected , not affected by the RelE20 mutant background ( Fig 2F ) . Taken together , the data suggest that Nub-PB can influence AMP gene transcription both independently of , and together with Rel . The proximal promoter region of CecA1 contains a cluster of six Oct/Oct-like sites required for Nub-PD mediated repression of a CecA1-luciferase reporter construct in vitro [15] . To investigate the requirement of this cluster for the regulatory capacity of Nub-PB and Nub-PD in vivo , the expression of different reporter constructs was analyzed in female ( Fig 3 ) and male flies ( S3 Fig ) . As expected , flies carrying the CecA1-lacZ construct with an Oct cluster deletion ( pA10ΔOct; Fig 3A ) displayed a derepressed and hence stronger reporter gene expression in fat body than flies with the complete upstream region ( Figs 3B , 3C , 3H and S3A and S3B ) . Since deletion of the cluster promoted reporter gene expression per se , the incubation time was decreased to circumvent saturation of the response ( Figs 3D–3G and S3C–S3F ) and combined with c564-driven overexpression of Nub-PB ( Figs 3F , 3G and S3E and S3F ) . The full-length pA10 CecA1-lacZ reporter gene responded strongly to Nub-PB overexpression ( Fig 3F and 3H; also shown in Fig 1B ) . Overexpression combined with the Oct cluster deletion did , however , not promote full transcriptional activation ( Fig 3G and 3H ) . Still , an intermediate level of β-gal reporter gene expression was observed , significantly stronger than in the Oct deletion control ( Fig 3E and 3H ) , suggesting that the CecA1 locus may contain additional Nub target sequences outside the pA10ΔOct region . We conclude that both Nub isoforms require the Oct cluster for accurate regulation of the CecA1 gene ( Fig 3I ) . Transcription factor antagonism requires a co-localized expression . We explored the spatial distribution in vivo of nub-RB and RD transcripts from dissected tissues of adult male flies ( Fig 4A ) . Notably , the RNA levels of nub-RD exceeded that of nub-RB in the gut , whereas expression of the two appeared roughly similar in other tissues . To further investigate the spatial expression , we applied transgenic reporter lines specific for either the nub-RB or nub-RD transcript . A dual isoform marker line ( nub-RBGFP , nub-RDAC-62>mCherry ) was applied to visualize the spatial expression of both isoforms in parallel ( S4 Fig ) . Prominent dual fluorescence was observed in larval wing discs , leg discs and foregut , with varying degree of overlap , whereas only nub-RD>mCherry was expressed in the larval brain ( S4A–S4C Fig ) . For adult expression , tissues from nub-RBGFP and nub-RDAC-62>mCherry flies were studied separately . Fluorescence was observed in wing veins and leg joints with either line , whereas mainly nub-RD appeared expressed in the abdominal fat body or adjacent tissues ( S4D–S4I Fig ) . A strong GFP signal was observed throughout the midgut of nub-RBGFP flies ( Fig 4B ) while nub-RDAC-62-driven mCherry expression was present only in the anterior region and faded within days in newly eclosed adult flies ( S4A’ Fig ) . Therefore , additional nub-RD-Gal4 driver lines were screened and especially nub-RDVT6452 drove prominent mCherry expression in the midgut ( Figs 1A and 4B ) . For the purpose of the present work , we conclude that the isoforms are expressed to varying degrees in immune competent tissues and thereby should be able to act as transcriptional antagonists in a competitive manner . In contrast to the observed isoform transcript levels in Fig 4A , Western blot experiments on midgut and carcass extracts using a non-discriminatory Nub antibody yielded an overall stronger band for Nub-PB ( S5A Fig ) [15] . This suggests that additional regulation occurs post-transcriptionally . Since nub-RD mutant flies exhibit chronic immune activation and microbial dysplasia in the gut [15 , 23] , we hypothesized that this isoform is required to suppress aberrant activity of Nub-PB in these tissues . In fact , Nub immunostaining is commonly used to mark the nucleus of gut enterocytes [30] , which suggests that at least one of the isoforms is strongly expressed in this cell type . The enterocyte-specific driver NP1-Gal4ts was therefore applied to drive RNA interference ( RNAi ) of nub-RB ( nub-RB-IR ) in adult flies . This resulted in a roughly two-fold decrease of the targeted transcript accompanied by a significant reduction of Drsl2 , Drsl3 , DptA , CecA1 and upd3 in midgut tissues ( Fig 4C ) . The latter encodes an infection-inducible cytokine and ligand of the JAK/STAT pathway [31 , 32] and was found upregulated 4 . 8-fold in the gut , albeit below the signal threshold in the global transcriptional profile ( S1 Table ) . The RNAi was further confirmed at the protein level , as Nub-PB decreased 2-fold in extracts from whole flies , strongly indicating that enterocytes constitute a major source of this isoform ( S5B Fig ) . Similar effects were observed with c564-Gal4ts in the midgut , but not in the abdominal carcass ( encompassing the fat body ) , suggesting that the endogenous Nub-PB acts as a positive regulator of AMPs specifically in the midgut ( S5C and S5D Fig ) . Conversely , RNAi-mediated downregulation of nub-RD , using a similar assay as in Fig 4C , resulted in an overall increased expression of the same set of genes ( Fig 4D ) . Comparative overexpression of the two isoforms in gut enterocytes confirmed their opposite regulatory effects on immune genes , previously observed in whole flies ( Fig 2A and 2B ) , as Drsl2 and upd3 were strongly up- and downregulated in midgut extracts following nub-RB and nub-RD overexpression , respectively ( Fig 4E ) . CecA1 and Drsl3 levels were increased by nub-RB but unaltered by nub-RD whereas neither overexpression affected DptA significantly . Interestingly , NP1ts>nub-RD decreased the expression of nub-RB 2 . 4-fold , but not vice versa . To investigate whether Nub-PB expression affected the gut bacterial load , we performed a qPCR analysis of the relative 16S rDNA levels from midgut extracts . NP1ts-driven overexpression of nub-RB resulted in a ~56% reduction of bacteria after 24 h ( Fig 4F ) . Conversely , RNAi using the same driver increased bacterial loads by 67% ( Fig 4G ) . The same set of flies displayed striking lifespan phenotypes as nub-RB overexpression significantly shortened median longevity ( females , 19 . 3; males , 20 . 3 days; average from three individual experiments ) relative to controls ( females , 46 . 2; males 45 . 8 days ) , whereas its downregulation increased longevity ( females , 52 . 3; males , 48 days ) ( Figs 4H , 4I and S6A–S6C ) . Overexpression of nub-RD resulted in longer lifespan ( females , 29; males 33 days ) compared to that of nub-RB , but shorter than controls ( S6B and S6C Fig ) . Surprisingly , antibiotic supplementation of the diet enhanced longevity of nub-RB-overexpressing flies ( females 15 . 8% , males 28 . 9% ) but not those of nub-RB RNAi ( females -3 . 1%; no change in males ) , suggesting that the microbial composition , rather than loads influences host lifespan ( Figs 4F , 4G and S6B and S6D ) . The germ-free conditions enhanced female ( 6 . 8% ) , but not male controls . Conversely , germ free nub-RD overexpressing males displayed enhanced longevity at early time points , but were overall not significantly benefited , whereas females were equally long-lived in comparison to conditionally reared flies . Together , these findings demonstrate that Nub isoforms regulate the expression of midgut AMPs in opposite manners and that Nub-PB expression correlates with gut microbiota and host lifespan . To investigate the role of Nub-PB after bacterial challenge , we performed oral infections using Erwinia carotovora carotovora 15 ( Ecc15 ) , a well-characterized bacterium with generally low oral pathogenicity in adult flies . Strikingly , overexpression of nub-RB caused a hypersensitive phenotype as all males and roughly 70% of the females succumbed within a day post infection ( Fig 5A and 5B ) . Interestingly , the median lifespan of the remaining overexpressing female survivors after bacterial exposure was 10 days , suggesting that death occurs primarily due to acute effects . RNAi directed against nub-RB caused a significant female ( ~25% ) , but not male , mortality during the acute stage of infection compared to infected driver control flies . We next explored the induction of the immune response mediated by the combined overexpression of nub-RB and Ecc15 infection ( Fig 5C–5F ) . At three hours post infection ( hpi ) , Drsl2 expression was upregulated by three orders of magnitude relative to the similarly infected driver control cohort , while at 24 hpi , levels were more comparable to those of the uninfected overexpression cohort ( Fig 5C ) . This indicates a rapid and transient hyperinduction of this gene by the combined effect of overexpression and infection . The induction of upd3 and CecA1 was also strongly enhanced by the combined effect of overexpression and infection , with peak expressions occurring later than for Drsl2 ( Fig 5D and 5E ) . Conversely , and in line with Fig 4E , DptA was not affected by nub-RB overexpression with NP1 ( Fig 5F ) . We hypothesized that such a strong effect on the immune response would likely be detrimental for the host , while at the same time enhance clearance of the infection . In line with this , enterocyte expression levels of nub-RB correlated with the relative clearance of Ecc15 compared to controls ( Fig 5G–5J ) . In comparison , overexpression of nub-RD caused a moderate , non-significant decrease in clearance ( S7A and S7B Fig ) . To exclude a confounding effect of feeding rate , a capillary feeding assay was performed where the different genotypes were found to consume similar amounts to controls ( Figs 5K , 5L and S7C ) . Proinflammatory responses in the gut could potentially cause epithelial damage and ultimately cause gut leakage . To test this , we applied the Smurf assay [54] but did not observe any flies turning blue , neither from genotype , infection , nor the combination of both , suggesting that death occurs due to other causes ( Fig 5M ) . We also observed upregulated levels of both isoforms in the midgut of orally infected flies , albeit stronger for nub-RB than nub-RD ( Fig 5N ) . This suggests that the balance is temporarily skewed towards the activating function of Nub-PB during the acute stage of infection . Following recovery on regular fly food , the expression of both isoforms returned to pre-infection levels around 48 hpi , indicating a pattern typical for transiently induced DIRGs . Taken together , these data indicate that Nub-PB is involved in the midgut immune response to ingested Ecc15 and that the activity of this transcription factor requires tight control to avoid detrimental effects on the host . Among the identified IMD/Toll-independent DIRGs in the transcriptional profiles , several targets of the JAK/STAT pathway were induced , such as the gut-specific and infection-inducible Drsl2 , stress-regulated Turandots ( Tots ) and the immunomodulatory cytokine Diedel ( in the fat body ) , suggesting that Nub-PB either acts above , or at the level of , the JAK/STAT pathway ( Fig 1F , S1 and S5 Tables ) [10 , 29] . In Drosophila gut enterocytes , the JNK pathway has been implicated in the regulation of Upd3 , which in turn acts as a ligand for the JAK/STAT pathway in response to bacterial infection and stress [31] . Pathway activation triggers intestinal stem cell differentiation and proliferation to replenish extruded enterocytes [31] . Our observations that the expression of upd3 and Drsl2/3 is regulated by Nub-PB led us to investigate the role of the above-mentioned pathways in this context ( Fig 6 ) . Prolonged nub-RB overexpression for five days resulted in prominent induction of reporters for JNK ( Fig 6A–6D ) , upd3 ( Fig 6E–6F” ) and JAK/STAT ( Fig 6G and 6H ) . This was accompanied by a general disorganization of enterocytes ( Fig 6E’ and 6F’ ) , increased number of mitotic cells ( Fig 6I , 6J and 6M ) and apoptosis ( Fig 6K and 6L ) . Combined overexpression of nub-RB and targeted RNAi against the JNK-homolog basket ( bsk-IR ) in gut enterocytes attenuated the induction of Drsl2 but not upd3 , suggesting a dependency on JNK pathway activity for the former , but not the latter target ( Fig 6N ) . To investigate the role of JAK/STAT , nub-RB was co-expressed together with a dominant negative form of the receptor Domeless ( DomeDN; Fig 7O–7Q ) . Similarly to the findings in Fig 6N , and independent on infection status , this diminished the induced expression of Drsl2 , but not upd3 . Nub-PB might hence act together with the transcription factors of the JNK and JAK/STAT pathways to induce midgut-specific immune genes not typically regulated by NF-κBs . In support of this , Drsl2 was found to contain putative DNA-binding motifs for Nub [15] , AP-1 and Stat92E [33] in the proximal promoter region ( Fig 2F ) . As expected , the expression of nub-RB was similar between the overexpression genotypes and was , in line with observations in Fig 6H , also induced by Ecc15 infection in the driver control line ( Fig 6N and 6Q ) . Together these results indicate that Nub-PB is sufficient to drive most , if not all , of the documented aspects of intestinal immunity and inflammation . We have shown that the large isoform encoded by nub , Nub-PB , is a novel and exceptionally strong transcriptional activator of immune genes in Drosophila . Compared to the major immune regulatory factors , Rel and Dif , Nub-PB can potentially target an even broader set of DIRGs ( Fig 1F ) . This could be explained by the notoriously promiscuous nature of Oct factors in terms of their conformations , dimer formations , protein-protein interactions and post translational modifications [21] . In humans , the Nub homolog , Oct-1 , has been proposed to act as a switchable stabilizer of either repressed , induced or poised states of genes depending on its protein-protein interactions [34] . Furthermore , interactions between human Oct-1 and NF-κB have been demonstrated biochemically [35] . In agreement , we found that Nub-PB and Rel synergistically induce CecA1 transcription ( Fig 2C and 2E ) . Moreover , the Nub binding sites ( Oct sites ) , located in immediate proximity to the κB-sites in the proximal promoter region of CecA1 , were required to both repress [15] and fully induce expression ( Figs 3 and S3 ) , suggesting that the isoforms bind the same motifs to antagonistically regulate the transcription of Rel-target genes ( Fig 3I ) . Several Oct family members encode alternative isoforms [36] . Knowledge of their unique functions is , however , sparse . Drosophila appears to be no exception as both nub and its paralogs pdm2/miti and pdm3 have similar gene organizations and encode promoter-specific isoforms with unknown functions . The underlying mechanism of the antagonism demonstrated in this study remains to be deciphered and is likely under multilayered regulation through isoform expression , mRNA/protein stability and the potential to form homo- and heterodimers . The very short unique N-terminus of Nub-PD implies that the domains required for transcriptional activation , e . g . via protein-protein interaction , are located in the larger N-terminal part of Nub-PB , although prediction analyses of protein functions did not reveal any distinct domains . Nub-PD might act as a passive repressor in a competitive manner , by binding to target sequences and prevent recruitment of additional factors , or alternatively form inactive heterodimers with Nub-PB . In accordance , isoform dimerization of the human POU protein Brn results in its inactivation [37] . Opposing effects of rat Oct-2 isoforms in vitro have also been reported and were dependent on the sequence and position of the octamer motif [38] . Finally , the human Oct-1 locus encodes at least three N-terminally distinct isoforms , which have been reported to act in partly distinct , albeit not opposite , manners [39] . It is hence plausible that distinctly acting , or even antagonistic isoforms of POU/Oct factors are an evolutionarily conserved phenomenon . Intestinal immune and stress responses need to be tightly controlled to avoid excessive damage to host tissues . The gut microbial level and composition in adult flies correlate with host immunity and lifespan [6 , 40 , 41] . Age-related gut dysfunctions have been linked to dysbiosis , chronic inflammation and ultimately host death [42] . In line with these studies , we have previously observed that flies with the nub1 mutation ( disrupts Nub-PD , but not Nub-PB expression ) display chronic immune activation , microbial dysplasia and shortened lifespan [15 , 23] . Importantly , we found that both isoforms are expressed in midgut enterocytes ( Figs 4A–4D and S5B ) and that overexpression of Nub-PB in these cells resulted in overall similar phenotypes as those previously found in nub1 mutants ( Fig 4E , 4F , 4H and 4I ) , suggesting that the function of Nub-PD in this context is to counteract Nub-PB and repress aberrant responses . In fact , prolonged enterocyte-specific overexpression of Nub-PB was sufficient to drive most aspects of immune and inflammatory responses previously recorded during oral infection [31] including JNK and JAK/STAT pathway activation , Upd3 induction and increased gut mitosis and apoptosis ( Fig 6 ) . It is hence plausible that Nub-PB represents a missing node in the Upd3 mediated signaling from enterocytes to intestinal stem cells resulting in subsequent activation of JAK/STAT-driven stem cell proliferation and epithelial renewal [31] , although a recent study demonstrated that a large number of transcription factors could potentially be involved [43] . The shortened lifespan of these flies correlated with a decreased microbiota ( Fig 4F , 4H and 4I ) and enhanced clearance of orally administered Ecc15 , indicating that death is not a direct consequence of bacterial overgrowth but rather occurs due to a hyperactive immune response . In agreement , genetic manipulations to inhibit feedback regulation of the IMD pathway impair host lifespan and survival to infection [44–45] . Also , similar to nub1 mutants [23] , microbial depletion extended longevity in nub-RB overexpressing flies , which furthermore suggests that bacterial exposure could aggravate the inflammatory response triggered by imbalance between Nub isoforms ( S6B Fig ) . Enterocyte-specific and adult-restricted RNAi of nub-RB yielded overall the opposite phenotypes: reduced DIRG transcription , increased level of gut bacteria and enhanced lifespan , suggesting that Nub-PB is both necessary and sufficient to drive these phenotypes ( Fig 4C and 4G–4I ) . Importantly , genes involved in mounting an immune response at all levels from recognition to effectors and cytokines were activated by Nub-PB ( Fig 1F ) . In contrast to Rel-driven immune responses [2] , very few components providing negative feedback regulation were induced by Nub-PB . We hence speculate that endogenous Nub-specific transcription might be regulated via feedback on the Nub isoforms per se , possibly through protein degradation , a typical feature of many transcription factors . In addition , the combined effect of Nub-PB overexpression and infection resulted in immune gene expression levels up to three orders of magnitude above those of infected control flies , which implies that Nub-PB can enhance transcription during infection . Such vast expression levels are likely to impact host tolerance to microbial exposure [46] through loss of homeostasis , generation of self-inflicted damage , stress or even a metabolic collapse , which may explain the shortened life span and hypersensitivity to infection . Interestingly , Oct-1-deficient mouse fibroblasts are hypersensitive to stress [47] , of which some parallels can be drawn to Nub-PB overexpressing , as well as nub1 flies [15 , 23] . We propose that Nub-PD , in analogy with to Oct-1 , acts as a stress sensor to neutralize the activity of Nub-PB , and that a balanced ratio between the isoforms is required to maintain a healthy gut environment . Our finding that two N-terminal isoforms of the Oct-1/Oct-2 homolog , Nub , play antagonistic roles in immune/stress gene transcription provides genetic evidence of a novel switch-like regulation mediated via the same gene . We further suggest that Nub-PB and Nub-PD together form a molecular rheostat that dynamically tunes the transcriptional output to balance responses and efficiently eradicate pathogens while avoiding excessive activation and autoimmune-like reactions . This raises the possibility that Nub protein isoforms also regulate other physiological and developmental processes in opposite directions . Furthermore , the findings highlight a potential need to scrutinize the present view of POU/homeodomain networks by considering the presence of antagonistically operating isoforms , which may radically alter the transcriptional output . It remains to be explored whether similar modes of molecular rheostasis constitute a general and evolutionarily conserved mechanism to ensure flexible adjustment to environmental and developmental cues . The following transgenic fly lines were used in this study . ( A ) w1118;; ( RRID:BDSC_5905 ) was applied as wild type . ( B ) y1w; nubMI05126 ( RRID:BDSC_37920 ) . In this stock , the MiMIC cassette [48] is inserted into the 5’ UTR of nub-RB and the GFP expression derived from the MiMIC cassette is under control of the endogenous nub-RB promoter . ( C ) w; nubAC-62; UAS-mCherry/TM3 ( RRID:BDSC_38418 ) . The nubAC-62 allele carries a Gal4 enhancer trap inserted into the upstream region of nub-RD promoter , and has been recombined with nuclear UAS-mCherry ( Müller and Affolter , personal communication to Flybase ) . ( D ) w; nubGAL4 . K ( RRID:BDSC_42699 ) . The nubGal4 . K line carries a Gal4 reporter driven by a 5 . 3 kb fragment from the nub-RD promoter and approximately 5 kb of upstream sequence[49] . CecA1-lacZ reporter lines: ( E ) pA10 [50] and ( F ) pA10ΔOct , ( this work ) . Lines for overexpression of transcription factors: ( G ) w;; UAS-nub-RD ( 15 ) , ( H ) w; UAS-nub-RB ( this work ) , ( I ) w;; UAS-DomeDN/TM3 , sb1 ( a gift from Nicolas Buchon ) . Lines for RNAi ( J ) w; UAS-nub-RB-IRKK113120 ( RRID:FlyBase_FBst0476872 , no predicted off-targets ) , ( K ) y1 , v1;; nub-RD-IRP{TRiP . JF02973}attP2/TM3 , sb1 ( RRID:BDSC_28338; moved to a w1118 background prior to experiments ) , ( L ) w;; UAS-bsk-IR . Gal4 driver lines for expression in specific tissues: ( M ) w1118; c564-Gal4 ( RRID:BDSC_6982 ) and ( N ) w1118; c564-Gal4; tub-gal80ts for fat body and midgut , and ( O ) w; NP1-Gal4 ( a gift from Bruno Lemaitre ) combined with Tub-gal80ts or ( P ) w; NP1-Gal4; Tub-gal80ts , UAS-GFP ( a gift from Nicolas Buchon ) for midgut enterocytes and ( Q ) w;; nub-RD-Gal4VT006452 ( RRID:FlyBase_FBst0483181 ) . Mutant fly strains: ( R ) w;; RelE20 , es ( RRID:BDSC_9457 ) . ( S ) w; UAS-nub-RB; UAS-nub-RD was constructed from ( G ) and ( H ) ; ( T ) w; UAS-nub-RB; UAS-DomeDN from ( H ) and ( I ) ; ( U ) w[1118]; UAS-nub-RB; UAS-bsk-IR from ( H ) and ( L ) ; ( V ) w1118; c564-Gal4; RelE20 , es from ( M ) and ( R ) ; [23] w1118; UAS-nub-RB; RelE20 , es from ( H ) and ( R ) using the double balancer line ( W ) w1118; if/CyO; MKRS/TM6B , tb1 . ( X ) w;; TRE-eGFP , ( Y ) w;; puc-lacZ/TM3 ( gifts from Ulrich Theopold /Dirk Bohmann ) , ( Z ) w; upd3-lacZ and ( AA ) w; 10xStat92E-GFP ( gifts from Nicolas Buchon ) were recombined to ( I ) and crossed to ( O ) or ( P ) for immunostainings . Flies were maintained on instant potato medium [23] in mixed female/male populations at 25 °C , 60% RH , with a 12 h light/12 h dark cycle . For experimental crosses , flies were reared at 18 °C until at least two days post eclosure , and then switched to 29 °C . Overexpression experiments were carried out following two days of incubation at 29 °C . For RNAi , flies were maintained at 29 °C for 5–7 days prior to experiments . All experiments were performed using 5–10 day old females with the exception of lifespan ( newly eclosed males and females , which were recorded daily ) and survival assays post infection ( 5–10 day old males and females maintained separately ) . For lifespan analysis in germ free conditions , food was supplemented with an antibiotic cocktail [23] . Drosophila mbn-2 cells ( DGRC , cat . no . 147 ) were cultured in Schneider’s medium ( Gibco ) supplemented with 10% fetal bovine serum ( Gibco ) in 5 ml plates at 25 °C , to a cell density of approximately 6–7 × 106 cells/plate . Total RNA was isolated from OrR males with TRIsure ( Bioline ) and used as template for reverse transcriptase ( RT ) . Due to the gene locus size , two RT-PCR reactions were performed in parallel , using coupled AMV RT and Tfl DNA polymerase ( Access RT-PCR system , Promega Biotech AB ) , to amplify the 5’ and 3’ fragments of the nub-RB cDNA separately . The 5’ forward primer was constructed with a NotI site-containing overhang and an internal reverse primer; the reverse with an internal forward primer and a reverse primer with a BamHI site-containing overhang at the 3’-end of the nub-RB cDNA . The two cDNAs , 1603 bp and 1803 bp respectively , which partly overlap and contain a common EcoRV site in exon 5 , were cloned into the pGEM-T easy vector . After DNA sequence verification , the two nub-RB cDNA halves were excised with NotI/EcoRV and EcoRV/BamHI respectively and then ligated into the pcDNA3 . 1 ( - ) vector to create the complete 2883 bp nub-RB coding sequence with short 5’ and 3’ UTRs . Drosophila expression plasmids were created using Gateway Technology ( Invitrogen , Carlsbad , CA , USA ) . Briefly , nub-RB coding cDNA was amplified from pcDNA3 . 1 ( - ) nub-RB using Pfu DNA polymerase ( Thermo Fisher Scientific , Waltham , MA , USA ) according to standard procedures . The purified PCR product was cloned into the pENTR/D-TOPO vector using pENTR Directional TOPO Cloning ( Invitrogen ) followed by recombination of the nub-RB cDNA into the pTW and pAW destination vectors ( obtained from TD Murphy ) using LR Recombination and the LR Clonase enzyme mix ( Invitrogen ) . P element transformation of w1118 flies with pTW-nub-RB was performed according to standard procedures[51] . Transfection of cells with pAW-nub-RB is described below . Plasmids and transgenic fly lines with CecA1-luciferase and CecA1-lacZ ( pA10 ) reporter constructs with the complete CecA1 upstream region and a short 5’ UTR have been described previously [13 , 47] . To create a CecA1-lacZ reporter with the Oct cluster deleted , pA10 was used as template for inverse PCR amplification with phosphorylated primers and cloned as described previously for CecA1ΔOct-luc [15] . The whole CecA1ΔOct-lacZ fragment was thereafter excised by XbaI-XhoI digestion and ligated into the P element vector pW8 plasmid , opened with the corresponding enzymes . P element-transformation of w1118 was carried out according to standard procedures [51] . Transfections were performed with 1 μg of pA10-luc construct and mixed with 1 μg of pAW-nub-RB , 500 ng of pAct-Rel , and 100 ng of Pol III-Renilla luciferase ( Addgene plasmid 37380 ) as internal reference . Carrier DNA was added to reach 10 μg in each sample , and transfections were performed using a calcium phosphate transfection kit ( Invitrogen ) as described previously [15] . Luciferase values were measured by the Dual-luciferase Reporter Assay System ( Promega ) . To stimulate immune activation , peptidoglycan was added in the form of a crude lipopolysaccharide ( LPS ) preparation ( 25 μl of 2 mg/ml ) , 4 h prior to harvest . For analysis of CecA1-lacZ reporter gene expression in transgenic flies , adults were dissected to remove heads and separate the digestive system from the rest of the fly , then fixed in 1% glutaraldehyde in phosphate-buffered saline and stained for β-gal activity using 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside ( X-gal ) as substrate , as described previously [52] . Incubation with substrate was done for 2 h at 37 °C , and in some cases continued for 16 h at 25 °C . To estimate X-gal intensity , 8-bit images were first processed in Adobe Photoshop ( version 2015 CC ) using the grayscale tool to convert all colors to white except blue ( converted to grayscale ) . Mean gray values per fly were subsequently measured in ImageJ ( in the range 0 ( white , no staining ) to 255 ( black ) ) using the freehand tool ( N = 6 ) . Total RNA extractions were carried out using TRIsure ( Bioline ) from adult females ( three flies for whole fly extracts and six dissected tissues , respectively , per replicate ) , followed by DNAse treatment and cDNA synthesis as previously described with a few modifications[15] . Taqman probes/primers were used to measure gene expression according to the manufacturer´s instructions ( Applied Biosystems ) . Primer/Probes: nub-RD ( CG34395-PD ) : Dm01841366_m1 ( Applied Biosystems ) ; nub-RB ( CG34395-PB ) : Dm01812808_s1 ( Applied Biosystems ) . Primer/probes for AMP gene expression were as previously published [15] . Samples were analyzed in biological triplicates or quadruplicates and relative amounts of each target were quantified relative to a set standard curve pooled from all samples in the analysis and finally normalized relative to those of RpL32 . Overexpression of nub-RB , was carried out in flies of the genotype w; UAS-nub-RB/c564-Gal4 and compared with flies carrying the c564-Gal4 driver but no UAS target gene ( w; c564-Gal4/+ ) . Flies were raised at 18 °C and adults reared in mixed sex populations at 25 °C . Female flies were used at 10 days of age , dissected to separate guts and the rest ( flies without guts and heads ) . Total RNA extractions were carried out using TRIsure ( Bioline ) , followed by DNAse treatment with Turbo-DNase ( Ambion ) and purification using RNeasy ( Qiagen ) . Tissues ( 25 per replicate ) from three or four independent pools of flies were analyzed as biological replicates on Drosophila Agilent microarrays . Total RNA ( 150 ng ) was amplified and labeled using Low Input Quick Amp Labeling Kit according to the manufacturer’s instructions . Cyanine 3-CTP-labelled c-RNA ( 1 . 65 ug ) was used for 17 hour of hybridization at 65 °C to the Drosophila ( V2 ) Gene Expression Microarray , 4x44K . The hybridized arrays were washed and scanned with the Agilent DNA microarray scanner G2505C . The fluorescent intensities of the scanned images were extracted and preprocessed using the Agilent Feature Extraction Software ( version 10 . 7 . 3 . 1 ) . Preprocessing of the raw data was done according to the standard analysis pipeline at the Bioinformatics and Expression Analysis Core Facility at Karolinska Institutet , Huddinge , Sweden . In short , Agilent processed signals ( i . e . feature gProcessedSignal ) were imported to Partek Genomics Suite and subjected to quantile normalization . After preprocessing and normalization , the data was filtered to remove genes that were not expressed at detectable levels ( estimated background signal ) . A factorial map of principal component analysis was executed on the whole expressed data using Bioconductor 3 . 1 and R 3 . 1 . Multiple T-tests of the entire dataset were performed in Graphpad Prism 6 and p-value thresholds adjusted using the FDR approach ( Q set to 1% ) . Filtered data ( above detection limit in at least one of the groups compared and below the adjusted p-value threshold ) were further used throughout the extended explorative downstream analysis with a few exception were a subsequent fold change cut-off were additionally applied ( where denoted ) . The GSEA to reveal enriched GO biological processes was performed using Cytoscape ( version 3 . 6 . 0 ) and the plugin Bingo ( version 3 . 0 . 3 ) . The analysis was executed using the hyper-geometric test with Benjamini-Hochberg FDR correction . Venn diagrams were constructed using the web-based software Venny ( version 2 . 1 ) . Female guts were dissected in PBS , pH 7 . 4 and fixed with 4% paraformaldehyde ( PFA ) at room temperature for 1 h . Following two 5 min washes in PBS-T ( 1x PBS +0 . 1% TritonX-100 ) , tissues were incubated in blocking solution ( PBST+ 0 . 5% normal goat serum ) at room temperature for 30 min and probed with primary antibodies ( mouse α-β-gal [1:20; DSHB , RRID:AB_528101]; rabbit α-PH3 [1:300; Millipore , RRID:AB_310177]; rabbit α-cleaved caspase-3 [1:300; Cell Signaling , RRID:AB_2341188] ) at 4 °C over night . The next day , tissues were washed in PBST 4×15 min , incubated with secondary antibodies ( goat α-mouse Alexa594 [1:1000; Invitrogen , RRID:AB_141372]; goat α-rabbit Alexa594 [1:1000; Invitrogen , RRID:AB_141359] ) at room temperature for 2 hours , washed again in PBST 4×15 min and then stained with DAPI ( Sigma ) at room temperature for 10 min . Stained samples were mounted on a glass slide with DABCO ( Sigma ) and confocal images were acquired using a LSM 780 microscope ( Zeiss ) . Carried out as previously described with a few modifications [15] . Briefly , whole fly extracts were prepared by grinding three flies per replicate in 2x standard Laemmli buffer in a 1 . 5 ml microcentrifuge tube using a plastic pestle followed by 10 min heating at 70 °C and 15 min centrifugation at 4 °C to remove debris . Samples were run on 10% polyacrylamide gels ( 125V , 90 min ) , wet transferred to PVDF membranes ( Millipore; 10V O/N followed by 40V for 1 h ) and blocked with 5% non-fat dry milk in Tris-buffered saline , 0 . 1 Tween-20 . Membranes were incubated with rabbit-α-Nub [15] in TBS-T at 4 °C overnight , washed and incubated for 1 h at RT with HRP-conjugated donkey-α-rabbit ( 1:10 , 000; GE Healthcare , cat . no . LNA934V/AG ) . Bands were obtained using SuperSignal West Pico ( Thermo Fisher Scientific ) . After quick stripping ( 5 min protocol using 0 . 3 M NaOH ) , PVDF membranes were blocked and reprobed with mouse α-β-Actin ( 1:10 , 000; Abcam ) followed by HRP-conjugated sheep-α-mouse ( 1:10 , 000; GE Healthcare , cat . no . LNA931V/AG ) as loading control . Erwinia carotovora carotovora 15 ( Ecc15 ) and Ecc15-GFP , kindly donated by Bruno Lemaitre , were cultured in LB medium at 30 °C with shaking [53] . For oral infections , overnight cultures of bacteria were pelleted and resuspended to OD100 in a 1:1 ratio of bacterial medium and Milli-Q H2O with 5% sucrose and 1% isotonic phosphate-buffered saline ( PBS ) , pH 7 . 4 . Infection vials were prepared by depositing 100 μl of bacteria or control solution on a filter paper placed on top of a 1% PBS , 2% agar gel to maintain humidity . Prior to challenge , flies were starved and desiccated in empty vials for 2 h , then briefly anesthetized to allow transfer to infection vials . Flies were maintained on this diet for 1 h ( bacterial count ) or 24 h ( survival analysis ) and subsequently transferred to fresh food vials . For systemic infections , E . cloacae β12 was grown over night at 37 °C with shaking , pelleted and resuspended in PBS at OD 1 . Flies were infected by abdominal injections with approximately 50 nl of bacterial suspension under brief anesthesia . Following all infections , flies were maintained at 29 °C . Smurf assay [54] was initiated at 16 hpi by transferring flies onto regular fly food supplemented with Coomassie Brilliant Blue FCF ( commercial , food grade ) . Flies were maintained up to two weeks post infection on this diet and observed for smurfs daily . For relative 16S rDNA quantification of gut bacteria , midguts from six flies per replicate were dissected and bacterial DNA extracted using the DNA Blood and Tissue ( Invitrogen ) kit according to instructions , including the Gram+ lysis step . For bacterial counts of Ecc-15 post infection , Ecc15-GFP was applied and cultured in LB with carbenicillin ( 50 μg/ml ) . Oral infections were performed as described above . At indicated time points , individual flies were anaesthetized and ground in 100 μl PBS on ice . Ten-fold serial dilutions were added to LA-carbenicillin plates and subsequently incubated over night at 30 °C . GFP-positive bacterial colonies were quantified from seven individual flies per replicate . The experiment was performed three times for conditional rearing , and once under germ free conditions . Carried out as described by Ja et . al with a few modifications [55] . Two females per replicate were placed in a vertically standing microcentrifuge tube with the bottom part excised and sealed with cotton to allow air exchange . Microcapillaries ( 3 . 2 mm , 5 μl; Drummond ) were filled with a 5% sucrose , 5% yeast extract solution placed through a small hole in the cap of the tube . To prevent evaporation , tubes were maintained in a high humidity climate chamber , with quick replacement of microcapillaries every 24 h . Following 48 h entrainment at 18 °C , tubes were placed at 29 °C to induce Gal4-activity . For overexpression , measurements were performed between day 2–3; for RNAi between day 5–6 after the temperature shift . Analyses of two sample means were performed using a two-tailed Student’s unpaired t-test ( Figs 4C , 4D , 5I–5L , 6M and S5B–S5D , S7B and S7C ) . Equal variances between the groups were ensured using an F-test ( p>0 . 05 ) . Multiple comparisons were carried out using a one-way ( Figs 2A–2C , 3H , 4E–4G , 5N and 6N ) , or two-way ANOVA ( Figs 2D–2F , 5C–5G and 6O–6Q ) combined with Tukey’s post hoc test . In Fig 5C–5F , denoted significant differences between cohorts were derived from the interaction between the two factors: 1 ) time post infection and 2 ) nub-RB overexpression as determined by a two-way ANOVA without further post-hoc analysis . In Fig 5N , Dunnett’s test was applied to analyze significant fold changes post infection relative to at 0 hpi ( set as control ) . Lifespan and survival assays were analyzed using Mantel-Cox log-rank test with Bonferroni-corrected thresholds for significance ( p<0 . 05/number of comparisons ) . Normalized qRT-PCR data were log2-transformed in order to model proportional changes prior to statistical analysis . Statistical analyses and graph constructions were carried out in Graphpad Prism 6 . For experiments involving dissections , sample sizes were set to N = 3–4 to allow collections within the designated time point and to minimize sample degradation due to handling . Parametric tests were chosen based on previous experience of the normality of gene expression of the selected targets . Fold changes in expression were quantified relative to the mean value of the control cohorts .	The numerous human diseases caused by aberrations in intestinal immunity and integrity urge a better understanding of the regulatory interactions that balance the output of host-microbe interactions . In this study , we discovered a novel phenomenon of transcriptional antagonism exerted via two isoforms encoded from the same gene . Balanced expression of the two forms was necessary for ensuring normal immune gene expression and maintaining immunological homeostasis in the Drosophila gut . We performed genetic manipulations to skew the balance of these isoforms . This resulted in a dysregulated immune system , changed levels of gut bacteria and altered host lifespan . Moreover , when we overexpressed the activating form flies quickly succumbed to oral bacterial infection despite an enhanced immune response . We suggest that antagonistically acting transcription factor isoforms may constitute a general mechanism for adjusting gene expression in various biological processes .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "medicine", "and", "health", "sciences", "rna", "interference", "gene", "regulation", "immunology", "dna", "transcription", "sequence", "motif", "analysis", "molecular", "biology", "techniques", "epigenetics", "research", "and", "analysis", "methods", "sequence", "analy...	2018	Nubbin isoform antagonism governs Drosophila intestinal immune homeostasis
Evolutionary innovation in eukaryotes and especially animals is at least partially driven by genome rearrangements and the resulting emergence of proteins with new domain combinations , and thus potentially novel functionality . Given the random nature of such rearrangements , one could expect that proteins with particularly useful multidomain combinations may have been rediscovered multiple times by parallel evolution . However , existing reports suggest a minimal role of this phenomenon in the overall evolution of eukaryotic proteomes . We assembled a collection of 172 complete eukaryotic genomes that is not only the largest , but also the most phylogenetically complete set of genomes analyzed so far . By employing a maximum parsimony approach to compare repertoires of Pfam domains and their combinations , we show that independent evolution of domain combinations is significantly more prevalent than previously thought . Our results indicate that about 25% of all currently observed domain combinations have evolved multiple times . Interestingly , this percentage is even higher for sets of domain combinations in individual species , with , for instance , 70% of the domain combinations found in the human genome having evolved independently at least once in other species . We also show that previous , much lower estimates of this rate are most likely due to the small number and biased phylogenetic distribution of the genomes analyzed . The process of independent emergence of identical domain combination is widespread , not limited to domains with specific functional categories . Besides data from large-scale analyses , we also present individual examples of independent domain combination evolution . The surprisingly large contribution of parallel evolution to the development of the domain combination repertoire in extant genomes has profound consequences for our understanding of the evolution of pathways and cellular processes in eukaryotes and for comparative functional genomics . Most eukaryotic proteins are composed of multiple domains , units with their own evolutionary history and , often , specific and conserved functions . The ordered arrangement of all domains in a given protein constitutes its architecture . Protein architecture can also be described in a simplified way as a list of binary domain combinations . While not completely equivalent , both views provide similar insights , and in this manuscript we will predominantly use the latter . Many domains can combine with different partner domains and , as a result , form a wide variety of domain combinations , often even within the same species [1] . Bringing together multiple domains in one protein creates a distinct entity , combining functions of its constituents . The emergence of proteins with new domain combinations is thought to be a major mechanism of evolution of new functionality in eukaryotic genomes [2] , [3] . It is especially important in the evolution of pathways , where physical proximity of domains in multidomain proteins links different elements of the pathway; thus , emergence of a new domain combination may rearrange pathways or processes in the cell [3] . The modular structure of eukaryotic proteins provides a mechanism that promotes differentiation and variation of protein functions despite the existence of only a limited number of domains . One can argue that the ease with which new domain combinations can be created provides eukaryotic genomes with the flexibility/plasticity that horizontal gene transfer and mobile genomic elements bring to prokaryotic genomes . The domain repertoires of most eukaryotes are remarkably similar , both in size as well as in their content [4] , [5] . In contrast , the number of distinct domain combinations found in multidomain proteins shows more variability between different organisms and correlates strongly with organism complexity and lifestyle [6] , [7] . To some extent , the evolution of eukaryotes and major milestones in their phylogeny , leading to the rise of complex organisms , such as humans , can be linked to the emergence of specific protein architectures rather than to the emergence of new domains . This finding has prompted many studies on the evolution of multidomain proteins , especially those that are involved in regulation of pathways such as apoptosis or innate immunity [5] , [8] . Special attention was , for instance , put on “promiscuous” domains—domains that have been found in combination with a particularly large number of partner domains [9]–[11] . Many such domains , for instance the PDZ or SH3 domains , are protein–protein interaction domains that “recruit” proteins to specific signaling pathways . Several groups attempted to infer domain architectures in ancestral genomes to be able to better study the evolution of new functionalities and processes carried out by such proteins 12–14 . The dynamics of domain rearrangements add an additional dimension to the analysis of the evolution of gene families , which , on top of point mutations , deletions , and insertions , can also include gains and losses of entire domains . Proteins can gain ( or lose ) new domains in genome rearrangements , creating ( or removing ) domain combinations [15] , [16] . In particular , the emergence of animals and , even more so , the emergence of vertebrates , have been associated with the appearance of novel domain combinations [2] , [17] . It has been suggested that new domain combinations allowed for functional diversification and contributed decisively to the evolution of complex multicellular systems in animals [2] . However , one can expect that the process of domain shuffling could lead not only to the emergence of completely new domain combinations , but also to the independent emergence of domain combinations already present in other , even distantly related , organisms . In fact , it has been shown that certain domain combinations observed in proteins involved in innate immunity have evolved independently several times [8] . Such proteins are not descendants of an ancestor protein that already contained the domain combination ( s ) in question , but instead evolved independently by parallel evolution [8] . The issue of independent domain combination evolution is not only of great theoretical interest , but also important in the context of comparative functional genomics , in particular for protein function prediction . Proteins with identical architectures are oftentimes ( but sometimes erroneously ) viewed as orthologs with all the resulting expectations as to the similarity of their functions [18] . Yet , if the architectures are the result of two different evolutionary trajectories , they may seem like orthologs but may well not be . At the same time , the question of if and how parallel evolution of domain combinations relates to functional similarity is an important and as of yet only poorly explored one . However , here we mostly address a simpler question , namely that of how common this phenomenon is . The first study addressing this issue , based on then-available 5 eukaryotic , 11 archaeal , and 46 bacterial genomes , claimed that independent evolution of domain combinations is rare and that only 0 . 4% to 4 . 0% of proteins are the result of such evolution [19] . A more recent study based on a larger set of 28 eukaryotic , 15 archaeal , and 53 bacterial genomes suggests that between 5 . 6% and 12% of domain architectures ( which can contain more than two domains and are thus different from domain combinations ) appeared independently more than once ( both between species and within species ) [20] . Another study , using a large dataset of proteins , but not necessarily from complete genomes , showed a more complex picture , with such events being rare in small gene families , but more frequent in large ones [13] . Here , we return to the same question , taking advantage of a now-available , much larger set of 172 complete genomes sampling most ( five out of six ) eukaryotic supergroups . At the same time , we focus our analysis solely on eukaryotic genomes , mostly because the tree of life for bacteria and archaea is not well defined and because of the role lateral gene transfer is likely to play in these kingdoms . We have collected complete sets of predicted proteins for 172 eukaryotic genomes representing five out of six eukaryotic supergroups [21] , [22]—Opisthokonta , Amoebozoa , Archaeplastida , Chromalveolata , and Excavata , with no representatives from Rhizaria [23] . In particular , we analyzed 112 genomes from Opisthokonta , namely 48 Metazoans ( animals ) , 2 Choanoflagellata ( unicellular and colonial eukaryotes , the closest living relatives of metazoa ) , Capsaspora owczarzaki as the sole representative of Filasterea ( unicellular euakaryotes , forming a sister clade to Metazoa and Choanoflagellata [24] ) , and 61 fungi . In addition , one genome , that of Thecamonas trahens , represents apusozoa ( flagellate protozoa , most of which feed on bacteria [25] ) , which are usually grouped with Opisthokonta . Three genomes in our analyses are from Amoebozoa ( amoeboid protozoa ) , 33 from Archaeplastida ( plants and relatives ) , 18 from Chromalveolata ( a large and very diverse group of unicellular euakaryotes ) , and 5 from Excavata ( unicellular eukaryotes , many of which lack traditional mitochondria ) ( see Table S1 for the full list ) . An overview of the current view of the phylogeny of these groups is shown in Figure 1 ( a detailed phylogeny is shown in Figure S1 ) . All proteins from all genomes were analyzed for the presence of protein domains , as defined by the Pfam database ( version 25 . 0 ) [26] and using programs from the HMMER3 package [27] with the “gathering” cutoff scores suggested by the Pfam database ( see the Materials and Methods section for details ) . On average , 76% of all proteins have at least one domain assigned . However , the distribution of the domain coverage is very broad , with outliers at the 40% and 95% level marks ( Table 1 ) . The outliers on the low coverage level include the single-celled rodent malaria parasite Plasmodium chabaudi ( 40% ) , the single-celled ciliate Tetrahymena thermophila ( 44% ) , and the single-celled parasite Trichomonas vaginalis ( 47% ) . The likely explanation for these low coverages is that these genomes contain a large percentage of domains not found in the traditional model organisms , which are therefore not yet included in the Pfam database . The genome of the pufferfish Takifugu rubripes , on the other hand , has the highest coverage level at 94% . For the human genome , this percentage is 85% . Overall , 34 , 778 distinct domain combinations were found in the 172 genomes analyzed here . A total of 22 , 241 of these appear in just one genome , and only 33 ( listed in Table S2 ) are present in all the eukaryotic genomes analyzed here and are generally involved in fundamental processes , such as transport , DNA repair , transcription , and translation . More detailed analysis is presented in the following paragraphs . In order to analyze the genomic domain combination content , we described each multidomain protein as a set of directed binary domain combinations . For example , a protein composed of domains A , B , and C ( listed in the direction from the N-terminus to the C-terminus ) is described as a set of the three binary combinations A∼B , B∼C , and A∼C . We retained information about the domain order , i . e . , domain combination A∼B is not the same as B∼A . Combinations between the same domains were not included in the analysis ( e . g . , a protein with the architecture A-B-B would be decomposed into only one binary combination , namely A∼B ) . The rationale for this is that combinations between the same domains can be a result of local duplication , ancestral descent , or domain fusion and the only approach to distinguish between these would be by explicit phylogenetic analysis of each domain . First , we simply counted how many distinct domains and domain combinations each of the 172 genomes contains . The average number of domains per protein is 1 . 7 for all the genomes analyzed here; however , for animal genomes this number is higher ( 2 . 0 ) . The distribution of domains in multidomain proteins is not uniform , with 1 , 448 domains appearing exclusively in single-domain proteins and 535 appearing only in multidomain proteins ( see Tables S3 and S4 for lists of these Pfam domains ) . The vast majority ( 6 , 040 ) of all domains appears in both single- and multidomain proteins . The number of distinct domains per genomes shows little variance between different organisms ( with the exception of some parasitic species ) and , when including inferred sets for the ancestral species , generally displays a decreasing trend in going from the last common eukaryotic ancestor to large multicellular organisms [5] . In contrast , the number of distinct domain combinations per genome varies more and shows strong correlation with the morphological complexity of their organisms , ranging from 1 , 178 in the free-living unicellular ciliate Paramecium tetraurelia to 2 , 372 in one of the simplest multicellular animals , Trichoplax adhaerens [28] , to 4 , 821 in humans . Examples of individual domain and domain combination counts are shown in Figure 2 ( data for each genome are shown in Table S5 ) . Especially striking is the large difference in domain combination numbers between deuterostomes ( which include vertebrates ) and protostomes ( ecdysozoa and lophotrochozoa [29] ) , ∼4 , 050 versus ∼2 , 650 . It has been suggested that the number of domain combinations in a genome approximately grows with the number of domains squared [10] . Figure 3 shows the average ratio between the sums of the number of distinct domain combinations and the sum of ( the number of distinct domains ) 2 for select groups of organisms . Even with this correction , fungi , as well as Embryophyta ( land plants ) and Chlorophyta ( green algae ) , have proportionally fewer domain combinations than the other groups . On the other hand , Alveolata ( a large and diverse superphylum of single-celled eukaryotes , represented in this work by genomes from ciliates and from the mostly parasitic Apicomplexa ) appear to partially compensate for their limited number of domains by having a comparatively large number of domain combinations ( even larger than for animals if normalized by the domain number squared ) . As mentioned above , in our analysis of 172 genomes , 22 , 241 ( out of 34 , 778 ) domain combinations appear only once and thus are specific to a single species . These species-specific domain combinations are relatively evenly distributed , with 95 out of the 172 analyzed genomes having between 10 and 100 domain combinations that are specific to the individual species ( 17 species have fewer than 10 specific domain combinations , and 60 have more than 100 ) , with a median value of 57 ( see Table S6 for complete data ) . Interestingly , these 95 genomes also include species for which very close relatives have been sequenced , such as human ( with 41 species-specific domain combinations ) and chimpanzee ( 9 ) , or mouse ( 24 ) and rat ( 50 ) . The chordate Branchiostoma floridae ( amphioxus ) , for which no close relative has been sequenced so far , is one of the exceptions , with about 2 , 140 species-specific domain combinations . The next-most-prolific organisms in terms of the number of unique domain combinations are the hemichordate Saccoglossus kowaleskii ( “acorn worm” ) with about 850 species-specific domain combinations , the brown tide heterokont ( stramenopile ) Aureococcus anophagefferens ( ∼760 ) , the purple sea urchin Strongylocentrotus purpuratus ( ∼720 ) ( which , together with S . kowaleskii , is a member of the superphylum Ambulacraria [30] ) , and the sponge Amphimedon queenslandica ( ∼610 ) . In all likelihood , in most cases the large number of unique domain combinations of these organisms is partially due to the fact that they are the sole sequenced representatives of their respective ( sub- ) phyla , and their domain combination counts have to be compared with those of the entire phyla , such as Arthropoda , which has about 1 , 500 clade-specific domain combinations ( based on 12 genomes ) . On the kingdom level , animals clearly have the most clade-specific domain combinations ( about 12 , 800 ) , whereas Embryophyta ( land plants ) and fungi both have less than half that number . Clade-specific domain and domain combination numbers for select taxonomic groups are shown in Figure 4 . Figure 5A shows the distribution of all the 34 , 772 distinct domain combinations encountered in this work between the five “supergroups” covered here , plus Thecamonas trahens ( which is related to Opisthokonta , see Figure 1 ) . More than half ( 57% ) of all encountered domain combinations are exclusively found in Opisthokonta , out of which 43% are specific to Holozoa ( animals and their closest relatives—Choanoflagellata and Capsaspora owczarzaki ) . In contrast , only 14% are specific to fungi , despite fungi having the largest numbers of genomes analyzed in this work . Thirteen percent of all domain combinations are not supergroup-specific , i . e . , they appear in at least two different supergroups . Only 2% of all domain combinations are found in representatives of all five eukaryotic supergroups analyzed in this work ( listed in Table S7 ) . Figure 5B shows the distribution of the 14 , 704 Holozoa-specific domain combinations over various groups of Holozoa . As expected , Chordata have the largest number of clade-specific domain combinations among Holozoa ( 27% ) . Interestingly , with 7% , the single-celled choanoflagellates are an unexpectedly large source of Holozoa-specific domain combinations , especially compared to Lophotrochozoa and Nematoda with 5% and 7% , respectively . About 20% of all Holozoa-specific domain combinations are unspecific within Holozoa at the taxonomic level used in Figure 5B ( for example , appearing in both nematodes and arthropods ) . We also investigated the numbers and types of domain combinations that are not only exclusive to a given clade , but also appear in each genome belonging to this clade . We termed such domains core domain combinations . Examples for selected genomes are shown in Figure 4 ( see Table S6 for counts and Tables S8 and S9 for detailed lists ) . As expected , these numbers are generally small . For clades represented by at least 10 genomes , only vertebrates ( or more precisely , Euteleostomi—“bony vertebrates” ) and land plants ( Embryophyta ) have more than 40 core domain combinations . The majority of all clades have fewer than two core-specific domain combinations ( with the obvious exceptions of extremely closely related species , such as members of the same genus , like Arabidopsis thaliana and Arabidopsis lyrata ) . Table 2 shows the nine metazoan-specific core domain combinations . Interestingly , these nine combinations are all involved in extracellular matrix/cell–cell adhesion ( signaling ) and in transcription regulation . Next , we investigated the evolutionary history of domain combinations—do they tend to appear once and are then inherited by the descendants ( as , for example , is the case for the nine domain combinations listed in Table 2 ) , or do they reappear independently multiple times in different branches of the tree of life ? Because both fusion and fission of domains are relatively simple processes that happen as a byproduct of genome rearrangements , in sharp contrast to the appearance of protein domains themselves [5] , [31] , we used unweighted parsimony for the reconstruction of ancestral domain combinations [12] , [32] . Unweighted parsimony assumes that domain fusion and fission are , on average , equally likely ( in fact , domain fusion appears to be slightly more common than domain fission [33]–[35] ) . Such a parsimony analysis is based on phylogeny of all living organisms , often called the “tree of life . ” Currently , the topology of “tree of life” is still disputed . Here , we use one that follows the newly emerging paradigm according to which eukaryotes can be classified into two larger clades , unikonta and bikonta [21] , with details taken from literature [21]–[25] , [29] , [36] , [37]; however , as we show later , removing controversial sections of this tree leads to quantitatively similar results . We used the Fitch algorithm in conjunction with this tree , and by minimizing the gain–loss sum [32] , we calculated gains and losses for each directed domain combination . The result of this analysis , a phylogenetic tree overlaid with domain-combination gains and losses , is available as Figure S4 ( Figure S2 contains data for individual domains ) . In order to determine whether a given directed domain combination appeared once or multiple times , we counted on how many tree nodes it ( re- ) appeared . For an example , see Figure 6; here , the domain combination KH ( orange rectangle ) ∼DEAD ( green rectangle ) appears in bilaterian animals ( deuterostomes and protostomes ) , as well as in a group of green algae ( Micromonas ) . All the clades in-between lack this combination , even though they possess each of the individual domains . Therefore , the KH-1∼DEAD was likely to have been “rediscovered” two times ( at a “cost” of two gains , as opposed to one gain and nine losses if KH-1∼DEAD were deemed ancestral ) . The results of analyzing each domain combination in this manner show that that a significant number of domain combinations emerged independently multiple times ( Figure 7 ) . From a total of 34 , 778 distinct domain combinations present in the 172 analyzed eukaryotic genomes , 25 , 433 were formed only once , 3 , 486 appeared independently twice , 1 , 683 three times , and 4 , 176 four or more times ( detailed numbers are presented in Table S10; lists of domain combinations are in Table S11 ) . The most frequently reemerging domain combinations are listed in Table 3 . The total number of domain combinations that appeared independently more than once is 9 , 345 , about 27% of the total . However , out of the total of 34 , 778 distinct domain combinations , 22 , 241 appear in only one genome ( i . e . , they are species-specific ) . Accounting for this , we can say that 75% of all recurring domain combinations have evolved independently at least once . The unexpected result of this is that for the majority of the eukaryotes analyzed here ( 154 out of 172 ) , more than 50% of the domain combinations present in their genome can also be found in at least one other eukaryotic genome , not by evolutionary descent but by independent reemergence . For example , 3 , 431 of the 4 , 821 domain combinations found in the human genome ( 71 . 1% ) independently evolved in at least one other species . On average , this ratio is the highest in vertebrates ( 70 . 7% ) and the lowest in Excavata ( 43 . 6% ) . The polychaete worm Capitella teleta has the highest ratio ( 72 . 4% ) and the obligate intracellular parasite Encephalitozoon cuniculi the lowest ( 24 . 4% ) ( detailed numbers are shown in Table S12 ) . To test whether longer proteins are more likely to contain reoccurring domain combinations than shorter ones , we compared the average lengths of proteins that contain reoccurring domain combinations to those that do not . The result is that the average length of proteins containing reappearing domain combinations is 782 residues ( median: 589 ) and is slightly longer than that of proteins with non-reappearing domain combinations that have an average of 712 residues ( median: 534 ) . On the other hand , the average number of domains in these two groups is almost identical ( ∼3 . 3 ) . We also compared the average lengths of domains themselves in those two groups . The results support the observation that repeated domain combinations tend to appear in longer multidomain proteins , but the preference is not very strong . We also investigated how the specific choices of parameters affect these numbers . In particular , we tested a range of cutoff E-values ( from 1e−3 to 1e−15 ) , as well as domain-specific “trusted” and “noise” cutoff scores from the Pfam database ( instead of gathering cutoff values ) [26] . Under any of these conditions , the percentage of independently evolved domain combinations was at least 26% ( see Table 1 ) . Furthermore , we also assessed the effect of two alternative operational definitions of domain combinations . First , we performed our analysis under a model of undirected domain combinations ( in which a domain combination is considered A∼B equivalent to B∼A ) . Second , we ran our analysis under a domain combination model that required domains to be adjacent to be considered a combination ( e . g . , a protein with A-B-C architecture contains domain combinations A∼B and B∼C , but not A∼C ) . Under both of these conditions , the parallel evolution percentage remained around 27% ( at 27 . 3% and 27 . 9% , respectively ) . Since the deep topology of the eukaryotic tree of life , as well as that of some subtrees ( e . g . , the fungal subtree ) , are still topics of ongoing research ( see Discussion ) , we also calculated the percentage of independent domain combination evolution specifically for the Metazoan and Viridiplantae ( green plants ) subtrees ( for the reason that plant and animal evolution is comparatively well understood ) . If only Metazoan genomes were included in the analysis , the resulting parallel evolution percentage was 29 . 8% , whereas for Viridiplantae it was 19 . 3% ( other major subtrees had percentages between these two values ) . One can argue that an unexpectedly high rate of parallel evolution events is due to potential false negatives caused by highly divergent domains . In this scenario , the ancestral domain combination would be wrongly counted as having been lost and replaced by an apparently independently evolving domain combination , while in fact the “new domain” would be simply a divergent version of the old domain . To test this hypothesis , we performed the analysis analogous to one described before , but on the level of Pfam-clans ( groups of domains that are believed to have originated from a common ancestor , but at much earlier point in evolution [38] ) . To do this , we replaced Pfam domains with their matching clan ( for the roughly 47% of Pfam domains that are not members of a clan , we used the domains themselves ) . The resulting rate of parallel evolution of domain combinations was about 42% for all conditions tested ( see Table 1 ) . Therefore , it is unlikely that our results are simply an artifact due to highly divergent domains . The explanation of this high rate is that a proportionally large number of domain combinations that appeared only once are members of large clans . Next , we investigated whether parallel domain combination evolution is equally prevalent in all subtrees of the eukaryotic tree of life or whether some branches differ in their propensity for parallel domain combination evolution . Related to this issue is the question of how large the evolutionary distances between pairs of independently evolved domain combinations are . For this purpose , we calculated the last common ancestor ( LCA ) for each pair of independently evolved domain combinations and then counted for each internal node of the eukaryotic tree of life for how many pairs of independently evolved domain combinations it represents the LCA . These counts were then normalized by the sum of species emerging from each node . The results for the nodes with the highest rates are summarized in Figure 8 . The result is that the most events go back to the split between unikonts and bikonts ( i . e . , to the last eukaryotic common ancestor . The split between deuterostomes and protostomes has the second-highest relative rate of independent domain combination evolution . In general , we noticed that independent domain combination evolution is more prevalent in Opisthokonta , and especially in Metazoans , than in the rest of eukaryotes , even if normalized for the number of genomes . In the following , we present some examples of parallel evolution of domain combinations between animals and fungi , Amoebozoa , and green plants ( see Figures 8 to 10 ) . An example of independent domain combination evolution between animals ( Neoptera , winged insects , in this particular case ) and Dikarya ( that subkingdom of fungi that includes the two major phyla Ascomycota and Basidiomycota ) is shown in Figure 9 . The combination of a NACHT domain [39] with ( a ) Ankyrin repeat ( s ) appeared independently at least twice , once ( in fact , probably more than once ) in Dikarya and once in Neoptera . The nucleotide binding NACHT ( present in the neuronal apoptosis inhibitory protein ( NAIP ) , CIITA , HET-E , and TP1 ) domain is primarily found in proteins associated with apoptosis and innate immunity [40] . Ankyrins are a family of adaptor proteins involved in metazoan cell adhesion [41] , [42] . Despite being present in numerous fungal genomes , the function and role of NACHT and Ankyrin domain proteins is unknown as of this writing . Individually , both the NACHT domain and the Ankyrin repeat are present in almost all eukaryotes ( i . e . , there is no particular clade in which all members lack either one of these two domains ) . The corresponding is true for the following two examples . Another example of parallel evolution is the Amidohydrolase∼Aspartate/ornithine carbamoyltransferase combination that evolved independently in Metazoa and in Dictyostelium ( Figure 10 ) . This combination is found in the mammalian CAD protein , which is a multifunctional protein that performs multiple enzymatic activities in the de novo pyrimidine synthesis pathway [43] . A more-distant parallel evolution example is the evolution of the K Homology ( KH ) ∼DEAD/DEAH box helicase combination that appeared independently in Bilateria and in Micromonas ( a group of green algae ) ( Figure 6 ) . The K homology ( KH ) domain is an evolutionarily conserved domain and is present in a wide variety of nucleic acid–binding proteins . The KH domain binds RNA and can function in RNA recognition [44] . The KH∼DEAD/DEAH box helicase combination is found in the probable ATP-dependent RNA helicase DDX43 [45] , [46] . The role of the protein with this domain architecture in green algae is unknown . All the values presented here depend critically on the number and phylogenetic distribution of genomes analyzed , the size of the domain database , and the significance thresholds ( and sensitivity ) for domain assignments . We evaluated the effects of these on our results , especially in the light of many earlier papers reporting results of somewhat similar analyses being contradictory to our results . To understand these apparent discrepancies , we have repeated our analyses using a reduced number of genomes , different domain recognition thresholds , and smaller domain databases mimicking the approaches used in earlier works . The results from these analyses confirm that the differences between the earlier and the current results stem mostly from the increase in the number of analyzed genomes and in the size of the domain databases . For instance , using only five genomes resulted in a reemergence percentage similar to the estimates presented in ( [19] and [20] ) . Clearly , increasing the number of the genomes in the analysis results in a large increase of this percentage . We can expect that the value reported here is a lower estimate of the real value , and with a significant growth of the number of completed eukaryotic genomes , this value could grow even further . Similarly , we can show that other differences between our results and that of the previous analyses are mostly due to the changes in the number of genomes and the size of the domain database . For instance , analysis of five eukaryotic genomes and domain definitions from the SCOP 1 . 53 database [47] led to the estimate that 80% of all eukaryotic proteins are multidomain proteins [48] ( similar numbers were reported in Liu ) , while our results suggest that this number is around 32% ( Table 1 ) . Two reasons are likely to contribute to these discrepancies . First , here we used much-more-stringent cutoff values than the unrealistically low E-value of 10−2 used in [48] . But even performing our analysis with an E-value cutoff of 10−2 instead of the domain-specific “gathering” thresholds results in a multidomain protein percentage of 52 ( and a protein match range of 52% to 97% ) , which is still lower than reported in [33] and [49] . This effect is due to the growth of the domain databases over the last 10 years—the SCOP database has more than doubled during that time—and the specific bias in the order in which domains are added to databases such as SCOP or Pfam . For instance , central and highly promiscuous domains [10] , such as kinase , PH ( Pleckstrin homology ) , PDZ , SH3 ( Src Homology 3 ) , and AAA ( ATPases Associated with diverse cellular Activities ) , have been studied and , as a consequence , added to the domain databases earlier than rare and less-central domains . Confirming this trend are two more-recent studies based on seven eukaryotic genomes in which the percentage of eukaryotic multidomain proteins is estimated to be 65% [49] . Unfortunately , we cannot completely exclude the effects of erroneous gene models . To partially address this problem , we performed our analysis under both inclusion as well as exclusion of the one genome with the most unusual domain combinations ( that of the amphioxus Branchiostoma floridae ) . Furthermore , we performed our analysis on various subsets of all available eukaryotic genomes that are believed to be of high ( er ) quality ( results not shown ) . In both cases , the effect on the resulting parallel evolution rate of 27% was negligible . There are several simplifications made in our model that likely lead to underestimating the number of independently emerging domain combinations . First , since our analysis is not based on domain trees ( evolutionary trees built for specific domains ) , our results do not take into account parallel domain combination evolution within a genome , i . e . , between paralogs in large protein families ( this is in contrast to the study performed in [20] ) . Second , it has previously been shown that domain fusion is more likely than domain fission [33]–[35]; thus , emergence of the same domain combinations is more likely than repeated loss of large ancestral sets of domain combinations undergoing primarily domain fission . Third , the results presented here depend to some degree on the number of genomes analyzed . Performing our analysis with a reduced number of genomes ( results not shown ) results in a smaller reemergence percentage ( similar to the lower estimations in [19] and [20] ) ; therefore , it is expected that with even more completed eukaryotic genomes , this percentage will grow further . Finally , we would like to point out while the tree of life shown in Figure 1 ( and in detail in Figure S1 ) is still disputed; this is mainly due to uncertainty regarding the placement of Rhizaria . Since our analysis does not include any genomes from this group , this controversy has no bearing on the results presented here . The second controversy concerns the placement of haptophytes ( a phylum of algae ) , which in the model used here are considered part of Chromalveolata , but which according to recent results might form a clade with Archaeplastida [24] . In our analysis , haptophytes are represented by only one genome , Emiliania huxleyi , the placement of which on the tree of life has no measurable effect on the results presented here ( data not shown ) . Our analysis shows that the number of distinct domain combinations per genome varies greatly between different groups of species and increases systematically with their complexity . This increase matches the intuitive meaning of “complexity” as related to differentiation between cell types in an organism , which typically results from the interactions between multidomain regulatory processes . The main result presented in this paper , namely the fact that at least 25% of all known and 75% of all recurring domain combinations have evolved independently , is less intuitive . On one hand , it is an obvious effect of the plasticity of eukaryotic genomes , with genome rearrangements constantly reshuffling existing domain combinations . On the other hand , it is interesting that this apparently random process leads to repeated reemergence of the same domain arrangements . Given that the genomes analyzed in this work contain a total of 8 , 023 distinct domains , it would allow the formation of about 64×106 distinct directed domain combinations . And yet in the genomes analyzed here , we observed a total of only 34 , 778 domain combinations , which corresponds to only about 0 . 05% of the theoretical maximum . Therefore , we can speculate that the process of domain recombination is not entirely random and that organisms evolved some mechanisms that constrain the process of domain recombination in such a way that the chances of harmful , nonsensical arrangements are decreased . Here , we can only speculate about possible mechanisms to implement such constraints , but , for example , this could be achieved via the specific distribution of transposable elements and/or chromosomal locations of preferred recombination “hot spots . ” The number of times many domain combinations emerged independently is even more significant when viewed from the perspective of individual species . Over 70% of the domain combinations present in the human genome , and about 70% for all vertebrates , have evolved independently in other species at least once . This apparent discrepancy between the global and per-species averages is caused by a large number—over 22 , 000 , unique , species-specific domain combinations , which , while rare ( about 130 on average , with a median of 57 ) in individual species , add up to a large percentage over all species . One can argue that we are seeing two types of domain combinations: “universal , reemerging domain combinations” and “clade–specific , non-reemerging domain combinations . ” One might speculate that domains that tend to appear in independently evolved domain combinations could be functionally different from those that make up combinations that only appeared once . This seems not to be the case , though— preliminary studies using a variety of methods and tools ( such as Gene Ontology term enrichment analysis ) indicate that there is no significant correlation between domain function and the tendency of domains to appear in independently evolved domain combinations . Similarly , strong correlation between domain “promiscuity” [10] and presence in reemerging domain combinations could not be observed . On the other hand , the modeling of structures of several specific cases of independently emerged domain combinations indicates that surface features of individual domains could be dramatically different , suggesting dissimilar functions [8] . This interesting issue definitely requires more in-depth analysis . Observations presented in this paper have important consequences in interpreting similarities and differences between genomes of distantly related organisms . Usually , discovery of a protein with known domain architectures in newly studied species is taken as an argument for evolutionary conservation of function of these proteins . This is of particular importance when attempting to transfer protein function from distantly related model organisms , such as from the ecdysozoans Drosophila melanogaster and C . elegans , to vertebrates , such as humans . The high rate of independent domain combination evolution between protostomes and deuterostomes ( the second-largest rate; see Figure 8 ) is yet another reason for interpreting results from such model organisms with caution [18] . Besides estimating the rate of independent domain evolution , we also assessed the number of clade-specific domains and domain combinations . All branches of life ( at all levels ) have unique domain combinations ( combinations not shared with other branches ) . Due to unequal sampling , it is difficult to compare these numbers . Nevertheless , some issues are worth mentioning . While , as expected , animals have the largest number of unique domain combinations ( ∼12 , 800 , based on 48 genomes , compared to ∼4 , 800 in fungi based on 61 genomes and ∼3 , 700 in green plants based on 33 genomes ) , within animals there appears to be little-to-no correlation between the number of unique domain combinations and morphological complexity . For example , mammals have ∼400 unique domain combinations from 10 genomes , whereas Arthropoda have roughly three times that number ( ∼1 , 500 from 12 genomes ) . Clearly , the number of unique domain combinations does not explain the complexity of mammals . In this context , we introduced the concept of clade core domain combinations , combinations exclusively found in each genome of a given clade . It can be argued that such clade core domain combinations provide fundamental and distinguishing functionality for the organisms of a clade . For example , animal core domain combinations are all involved in extracellular matrix/cell–cell adhesion functions and in transcription regulation functions and are thus strongly correlated with the development of multicellular organisms . In summary , our results stress a recurring theme—namely , that evolution is an exceedingly dynamic , and seemingly random , process . New domain combinations are being created and recreated throughout evolution . Each group of organisms ( and probably even each organism ) has their own solution , based on a partially shared set of building blocks ( domains ) to solve shared biochemical and regulatory needs . As more and more genomes are being sequenced , we expect the percentage of independent domain combination evolution to grow even more . In fact , we expect that , with sufficient data available , the following paradigm of evolution at the domain level will emerge . Major clades ( such as animals ) have a relatively small set of distinguishing core domain combinations that are essential and defining for members of that clade ( such as developmental programs and cell–cell adhesion for animals ) . Outside of these hierarchical sets of core domain combinations ( such as for eukaryotes , animals , and vertebrates ) , all domains are randomly undergoing reshuffling , and the vast majority keep reemerging and disappearing both over species space and over time , with the exception of various small sets of core domain combinations . Protein predictions for 172 completed eukaryotic genomes were downloaded from a variety of sources ( for details , see Table S1 ) and analyzed for domain content against the hidden Markov models ( HMMs ) from the Pfam domain database ( version 25 . 0 ) using hmmpfam from the HMMER software package ( version 3 . 0 ) [26] , [27] . For score thresholds , we primarily used the per-domain “gathering” cutoff bit scores ( “GA2” ) from the Pfam database ( these cutoffs are used by Pfam to determine which sequences get included in Pfam full alignments ) . Domains associated with viruses , transposons , and bacteriophages were ignored . For overlapping domains , only the domain with the lowest E-value was retained . The domains of multidomain proteins were decomposed into all possible pairs of directed binary combinations; combinations between identical domains were ignored . Based on these preprocessing steps , lists of domains and domain combinations were created for each genome analyzed and then mapped onto corresponding external nodes of the eukaryotic evolutionary tree ( see Figure S1 ) . The presence and absence of domains and domain combinations for each internal tree node were inferred under unweighted parsimony using the Fitch algorithm for domain combinations [32] , [50] and using the Dollo parsimony for individual domains [5] , [51] . This allowed us to count how many times each domain combination appeared . In order to test the robustness of the results , we performed the analyses with various parameters . For example , we tested filtering the predicted domains by E-values ranging from 10−3 to 10−15 and/or filtering using the lists of domain-specific score cutoff values used by the Pfam database ( “trusted , ” “gathering , ” and “noise” cutoffs , with trusted cutoffs being the most stringent ) . Furthermore , we tested the effects of ignoring overlapping domains . We were unable to find a combination of these settings that would significantly change the numbers presented here and invalidate our conclusions . The preprocessing steps , the unweighted Fitch parsimony , the Dollo parsimony , and basic ancestral GO term analyses were performed by software of our own design [52] .	Most proteins in eukaryotes are composed of two or more domains , evolutionary independent units with ( often ) their own individual functions . The specific repertoire of multidomain proteins in a given species defines the topology of pathways and networks that carry out its metabolic and regulatory processes . When proteins with new domain combinations emerge by gene fusion and fission , it directly affects topology of cellular networks in this organism . To better understand the evolution of such networks we analyzed a large set of eukaryotic genomes for the evolutionary history of known domain combinations . Our analysis shows that 70% of all domain combinations present in the human genome independently appeared in at least one other eukaryotic genome . Overall , over 25% of all known multidomain architectures emerged independently several times in the history of life . The difference between a global and species specific picture can be explained by the existence of a core set of domain combinations that keeps reemerging in different species , which are accompanied by a smaller number of unique domain combinations that do not appear anywhere else .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "biology", "computational", "biology", "evolutionary", "biology" ]	2012	This Déjà Vu Feeling—Analysis of Multidomain Protein Evolution in Eukaryotic Genomes
Polycomb ( PcG ) regulation has been thought to produce stable long-term gene silencing . Genomic analyses in Drosophila and mammals , however , have shown that it targets many genes , which can switch state during development . Genetic evidence indicates that critical for the active state of PcG target genes are the histone methyltransferases Trithorax ( TRX ) and ASH1 . Here we analyze the repertoire of alternative states in which PcG target genes are found in different Drosophila cell lines and the role of PcG proteins TRX and ASH1 in controlling these states . Using extensive genome-wide chromatin immunoprecipitation analysis , RNAi knockdowns , and quantitative RT–PCR , we show that , in addition to the known repressed state , PcG targets can reside in a transcriptionally active state characterized by formation of an extended domain enriched in ASH1 , the N-terminal , but not C-terminal moiety of TRX and H3K27ac . ASH1/TRX N-ter domains and transcription are not incompatible with repressive marks , sometimes resulting in a “balanced” state modulated by both repressors and activators . Often however , loss of PcG repression results instead in a “void” state , lacking transcription , H3K27ac , or binding of TRX or ASH1 . We conclude that PcG repression is dynamic , not static , and that the propensity of a target gene to switch states depends on relative levels of PcG , TRX , and activators . N-ter TRX plays a remarkable role that antagonizes PcG repression and preempts H3K27 methylation by acetylation . This role is distinct from that usually attributed to TRX/MLL proteins at the promoter . These results have important implications for Polycomb gene regulation , the “bivalent” chromatin state of embryonic stem cells , and gene expression in development . The paradigmatic view of PcG repression is derived from the analysis of its role in the regulation of Drosophila homeotic ( HOX ) genes for which PcG genes were first discovered ( for review see [1] ) . The expression pattern of HOX genes is set in the very early embryo by segmentation gene products , which determine the embryonic domains in which each HOX gene is active or repressed . The PcG proteins present in the early embryo do not prevent this initial activation but , when , shortly thereafter , the segmentation gene products disappear , they maintain the repressed state throughout development . While the analysis of HOX gene regulation gave the impression that PcG repression is all-or-nothing and , once established , is permanently maintained it is clear now that many other genes are also PcG targets in flies as in mammals , that PcG repression can set in at later stages and can be abrogated in the course of differentiation or in specific situations . The genetic evidence , however , shows that both the repressed and the non-repressed state tend to be inherited through successive cell cycles . The functions closely associated with the maintenance of the non-repressed state are those of the trithorax and ash1 genes . These functions are not responsible for transcriptional activation per se , which still requires the appropriate enhancers and their binding factors , rather they are important to antagonize PcG repression and therefore to maintain a non-repressed “open” chromatin state that renders the target gene available for activation [2] , [3] . How do TRX and ASH1 create a state resistant to PcG repression ? Both TRX and ASH1 are SET domain proteins reported to have histone methyltransferase ( HMTase ) activity . Complexes containing TRX have been found to methylate histone H3K4 [4] . In budding yeast , Set1 , a close relative of TRX , is a component of the COMPASS complex , which is recruited to the 5′ region of transcription units and methylates H3K4 to promote transcriptional elongation [5] , [6] . Similar complexes containing mammalian TRX orthologues MLL1 and MLL2 have been biochemically characterized [7] , [8] . Mammalian genomes as well as Drosophila encode several H3K4 methylating enzymes . Complexes containing true mammalian Set1 orthologues Set1A/B are responsible for most of the H3K4 methylation [9] , [10] . It is thus likely that TRX and MLL1/2 complexes are specialized versions recruited to a subset of genes that includes PcG target genes . TRX and its mammalian counterparts are cleaved in two moieties by the threonine aspartase Taspase1 , of which the C-terminal moiety contains the catalytic SET domain [11]–[13] . In the case of MLL1 , the two cleavage products have been found associated together [14] . ASH1 has been variously reported to methylate in vitro H3K4 , H3K9 , H4K20 [15] or H3K36 [16] . The hallmark of a locus repressed by PcG mechanisms is its residence within a broad chromatin domain enriched in trimethylated H3K27 , often including more than one gene [17]–[21] . PcG repression and H3K27 trimethylation are initiated and maintained by one or more Polycomb Response Elements ( PREs ) that appear as narrow regions , within the methylation domain , that simultaneously bind several PcG proteins [17]–[19] . In mammalian embryonic stem cells , many PcG target genes have been reported to bear both repression-associated marks and H3K4me3 [22] . These genes have been said to assume a bivalent state with some transcriptional activity taking place even in the presence of PcG complexes . Upon differentiation , these bivalent states may become resolved into fully active states , with no PcG proteins , or fully repressed states , with PcG binding and extensive H3K27me3 domains . What controls the balance between activity and repression and whether bivalent states occur in Drosophila is not known . Certain Drosophila genes have been found to bind PcG proteins while remaining functional and sustaining transcriptional activity [19] . These findings raise questions about the relationship between PcG complexes and associated chromatin marks and transcription , Trithorax Group ( TrxG ) proteins and their associated chromatin marks . In this work , we have exploited the genomic approach to seek answers to these questions . By comparing PcG/TrxG and transcriptional landscapes in three Drosophila cultured cell lines of different origin we have asked what range of chromatin states can be assumed by PcG target genes , whether the binding of PcG proteins constitutes the default state and how the chromatin landscape and the binding of TRX and ASH1 is involved in changes and in the stable maintenance of alternative states . We then asked whether the differences in PcG binding in the three cell lines were reflected by changes in transcriptional activity of the underlying genes and/or in the binding of TrxG proteins . In the analyses that follow , we used the presence of RNA Pol II at the Transcription Start Site ( TSS ) and H3K4me3 at position +500 bp as criteria to distinguish transcriptionally active from inactive genes ( for details , see Text S1 and Figure S4 ) . We first consider the Bithorax Complex ( BX-C ) , containing the three homeotic genes Ubx , abd-A and Abd-B . In two cell lines , BG3 and D23 , all three genes are PcG-repressed: they bind PcG proteins at the known or presumed PREs and the entire BX-C domain of more than 320 kb is highly enriched for H3K27me3 ( Figure 1B , Figure S5 ) . In Sg4 cells , Ubx and abd-A are repressed but Abd-B is transcriptionally active , lacks H3K27me3 and at least two of its known PREs , Fab-7 and Fab-8 , lack PcG binding ( Figure 1A ) . Abd-B has five different promoters . In Sg4 cells four are active , while the one furthest upstream is inactive , binds PcG proteins and is enriched for H3K27me3 [19] . Consistent with this , peaks of Pol II and of H3K4me3 are found at the active Abd-B promoters , as well as in the Abd-B downstream region , known to produce non-coding RNAs [23] ( Figure 1A ) . These peaks are absent in BG3 and D23 cells where the repressive marks predominate ( Figure 1B , Figure S5 ) . We next determined the relationship of TRX and ASH1 to the state of activity of the BX-C genes using two sets of antibodies for each of the N-ter TRX , the C-ter TRX and the ASH1 polypeptides ( see Text S1 and Table S9 for details ) . When assayed with C-ter antibodies , TRX is found at the known and presumptive PREs of the BX-C region , forming sharp and distinctive peaks irrespective of the transcriptional activity of the target genes , consistent with previous observations [17] , [18] , [24] , [25] . The TRX C-ter antibodies also detected TRX at transcriptionally active promoters in the Abd-B region of Sg4 cells but not in BG3 or D23 cells or at the repressed promoters of Ubx and abd-A ( Figure 1 , Figure S5 ) . The TRX N-ter distribution shows surprising differences . It is found together with TRX C-ter at PREs and at active promoters but it is also associated with the broad region that includes Abd-B and its downstream non-coding transcribed region in Sg4 but not in BG3 cells ( Figure 1 ) . A related function is implied by the distribution of ASH1 . No ASH1 was found in the entire BX-C region of BG3 or D23 cells , where all the genes are PcG-repressed , but in Sg4 cells ASH1 was extensively associated with the entire region occupied by TRX N-ter ( Figure 1 ) . These findings suggest that full derepression results in massive loss of H3K27me3 , dissociation of Polycomb proteins from PREs and formation of an extensive ASH1 domain . They also point to a specific function of N-ter TRX that does not involve HMTase activity but extends with ASH1 over the entire domain that would otherwise be enriched for H3K27me3 . TRX was found at all PREs of the Bithorax-Complex . Of 170 computationally defined PREs in Sg4 cells ( see Text S1 and Table S6 for details ) 94% bind TRX C-ter , indicating that the presence of TRX at PREs is a common feature of PcG-repressed genes . The N-ter TRX antibody is generally weaker in ChIP-chip experiments and often does not reach the two-fold enrichment cut-off . Despite this , 39% of computational PREs also bind TRX N-ter , suggesting that the TRX complex at PREs contains both TRX moieties . Furthermore , looking at the entire catalogue of potential PREs , we found that TRX binds at essentially all known or presumptive PREs , whether or not they were also occupied by PcG proteins . That is , irrespective of the repressed or active state of the associated genes , PREs are also TREs ( Trithorax Response Elements ) . The presence of an ASH1/TRX N-ter domain at the derepressed Abd-B locus is consistent with the antagonistic genetic interactions between PcG and ash1 or trx . Are the ASH1/TRX N-ter domains generally characteristic of PcG target loci in the derepressed state ? A survey of the Sg4 genome revealed 56 ASH1 domains half of which were broad: ranging from 10 . 7 to 77 . 5 kb in length ( Figure 2A ) . 79% of the ASH1 binding regions also bound TRX N-ter . An excellent overall correlation between the binding levels of the two proteins within ASH1/TRX N-ter domains ( Figure 2B ) suggested that their binding is interdependent . To test this directly we knocked down ASH1 or TRX in BG3 cells to ∼20% of wild type by RNAi ( Figure S6B ) and assayed the chromosomal distribution of ASH1 and TRX N-ter by ChIP/chip . The knock-downs of both protein resulted in substantial reduction of their binding to chromosomes ( Figure 2C and 2D ) . The knock-down of TRX also reduced ASH1 binding to an extent comparable with that seen in RNAi against ASH1 itself ( Figure 2E ) . Strikingly , RNAi against ASH1 led to marked loss of TRX N-ter from the broad ASH1/TRX N-ter domains but not from presumptive PREs ( Figure 2F ) . We therefore conclude that binding of ASH1 and TRX N-ter within broad domains is interdependent while association of TRX with PREs is not dependent on ASH1 . We next examined the correlation between transcriptional activity of PcG target genes and binding of ASH1/TRX N-ter . When we looked at genes that are PcG-repressed in one cell line but acquire Pol II at TSS and H3K4me3 at position +500 in another cell line , in 39 of 47 cases ( 83% ) derepression was accompanied by binding of ASH1 , indicating strong correlation . We then asked if regions that bind ASH1/TRX N-ter in Sg4 cells but no longer do so in BG3 cells generally acquire PcG and H3K27me3 and vice versa . From 44 regions that bind ASH1 and TRX N-ter in Sg4 cells , 29 no longer bind these proteins in BG3 cells and seven of these simultaneously acquire PC and H3K27me3 ( Figure 3A ) . Conversely when the state of PcG targets detected in Sg4 cells was examined in BG3 cells , half of the cases that lost PcG binding acquired at the same time an ASH1/TRX N-ter domain ( Figure 3B ) . The reciprocal comparison between BG3 and Sg4 cells gave similar results . Although our current data set does not provide conclusive proof that all genomic ASH1/TRX N-ter domains correspond to PcG targets , we conclude that the chromatin state of the active Abd-B locus is not exceptional and that in multiple instances the loss of PcG and H3K27me3 in one of the cell lines was accompanied by the appearance of an ASH1 domain and derepression , in turn suggesting that broad binding of ASH1/TRX N-ter is likely a general mark of the derepressed state of PcG targets . The comparison of PcG and trxG landscapes in different cell lines showed that in half of the cases the loss of PcG and H3K27me3 was not associated with the gain of an ASH1/TRX N-ter domain and vice versa ( Table S7 ) . Strikingly , in these cases TRX binding to the PRE was also absent . In 31 of 36 cases ( 86% ) , genes that lost PcG and H3K27me3 without acquiring ASH1/TRX N-ter did not display marks of transcriptional activity in the absence of PcG repressive marks . This chromatin state is vividly exemplified by the hedgehog ( hh ) locus in D23 cells ( Figure 3E ) . hh encodes an important signaling protein essential for morphogenesis in flies and mammals , whose control by PcG mechanisms is well established [26] . Consistent with this , hh is a PcG target region in Sg4 ( Figure 3D ) and BG3 ( not shown ) cells . However , in D23 cells this locus was completely devoid of both PcG and TrxG proteins but remained transcriptionally inactive ( Figure 3C ) . Another example of this kind is tiptop ( tio ) , also an essential developmental gene . In this case PcG regulation of tio was evident in the BG3 line ( Figure S7B ) but in Sg4 ( Figure S7A ) and D23 cells ( not shown ) the gene is devoid of both PcG and TrxG proteins . As in the case of hh , the gene remained inactive in the absence of PcG proteins ( Figure 3C ) . We call this chromatin state the “void” state since it is characterized by the absence of PcG/TrxG proteins or chromatin marks . The lack of transcriptional activity of target genes in the “void” state indicates that mere absence of PcG proteins is not sufficient for expression of the target gene . Conversely , lack of transcriptional activity or of active chromatin marks is not sufficient for binding of PcG proteins to a PcG target gene . It is clear that neither binding of PcG proteins nor of TRX constitutes the default chromatin state of PcG target genes . The function of TRX and ASH1 is to antagonize PcG repressive activities when the target gene is in the active mode . Genetic evidence indicates that , in the absence of PcG repression , TRX and ASH1 are dispensable for the expression of PcG target genes , suggesting that their role is different from that of general transcription activators [3] . Consistent with this , our genome-wide mapping indicates that ASH1 and TRX are not general transcription factors . Thus , while Sg4 cells have 5771 active transcription units , they contain only 56 ASH1 domains and 618 TRX binding regions , implying that transcription of the vast majority of genes does not involve ASH1 or TRX . Our data also argue against the idea that TRX is a functional ortholog of the yeast SET1 protein , which is generally recruited to active promoters via its interaction with RNA Pol II and is responsible for trimethylation of H3K4 on promoter-proximal nucleosomes [6] . Most active promoters bearing the H3K4me3 mark have no TRX binding ( Figure 2A ) and we did not detect any decrease in the overall level of H3K4me3 after ten cell generations of RNAi treatment reducing the level of TRX to 20% of wild type ( Figure 4A ) . A similar knockdown of E ( Z ) produced prominent reduction in global levels of H3K27me3 ( Figure 4B ) . H3K4 methylation by TRX does not explain its anti-repressive activity: H3K4 methylation by other methyltransferases occurs at all active loci . Furthermore , TRX N-ter lacks the SET domain and therefore has no methyltransferase activity . ASH1 does have a SET domain but our results show that ASH1-TRX binding regions are not domains of H3K4 methylation . A histone modification that would be clearly antagonistic to PcG mechanisms is H3K27 acetylation ( H3K27ac ) . In fact all ASH1/TRX N-ter domains are co-extensive with domains of H3K27ac ( Figure 4D , Figure 1A ) which reach both upstream and downstream of target genes . This modification is not exclusive for ASH1/TRX N-ter domains and is also found at 66% of transcriptionally active genes that lack ASH1 . In these cases , however , the H3K27ac distribution is narrow ( compare Figure 4E and 4F ) and largely confined to a prominent peak 450 bp downstream of the promoter . The histone acetyltransferase ( HAT ) activity responsible for the broad H3K27ac distribution in ASH1/TRX N-ter domains appears to be specific: H3K9ac in these domains is limited to a peak around 450 bp downstream of the promoter and is no different from that found at other active genes ( Figure S8 ) . Coextensive binding of ASH1/TRX N-ter and H3K27ac suggests that HAT activity responsible for H3K27 acetylation may constitute a part of ASH1 or/and TRX N-ter protein complexes . To explore this possibility we examined the effect of TRX knock-down on the level of H3K27ac within ASH1/TRX N-ter domains . The knock-down of TRX results in simultaneous loss of ASH1 thus the effect of loss of both proteins was assayed in this experiment . Only a modest decrease in H3K27ac was found ( Figure 4C ) , despite the strong reduction in TRX N-ter and ASH1 ( Figure 2E ) . We conclude that the level of H3K27ac within ASH1/TRX N-ter domains is functionally related to the binding of TRX or ASH1 but is not directly linked to their levels . Consistent with the idea that ASH1 is a marker of anti-repressive , TrxG-dominated chromatin states , there is little overlap between regions bound by ASH1 and domains characterized by H3K27me3 and PcG proteins ( ∼6% ) and , unlike TRX , ASH1 is not generally recruited to PREs in the repressive state . However , exceptional sites that bind both ASH1 and H3K27me3 are very instructive ( Table S8 ) . One of these corresponds to the Psc gene , which encodes a core component of the PRC1 complex . In all cell lines examined , this gene has prominent PcG binding and H3K27me3 but remains functionally active and produces PSC protein . A prominent peak of Pol II is found at the TSS of Psc and of the adjacent , closely related and divergently transcribed Su ( z ) 2 gene . The presence of ASH1 and TRX N-ter in a narrow region at the 5′ end of Psc ( Figure 5D ) supports the idea that these proteins mark alleviation of transcriptional repression at PcG target genes . It also suggests that PcG and TrxG proteins at this locus are not mutually exclusive , although it remains possible that Psc expression cycles on and off . Another region of this kind contains engrailed ( en ) and invected ( inv ) , two genes encoding related homeodomain factors with important roles in metazoan development . Their regulation by PcG mechanisms is well known [27] . In Sg4 and D23 cells , both genes are contained within a PcG domain and lack Pol II at their promoters ( Figure 5A , Figure S9 ) . However , in BG3 cells , in addition to PcG and H3K27me3 , both genes show pronounced binding of ASH1 and TRX N-ter around the 5′ ends of their transcription units as well as prominent peaks of Pol II at promoters ( Figure 5B ) . In agreement with this , both genes produce elevated levels of mRNA in BG3 cells relative to Sg4 ( Figure 5C ) . The above examples indicate that , while most PcG/TrxG target loci in cultured cells are dominated by either the PcG or TrxG function , some genes display both simultaneously , or at least they can alternate from one state to the other . We call such a condition a “balanced” state and speculate that in this state PcG and TrxG proteins act in concert with other positive and negative regulators of transcription , keeping one epigenetic activity from “overcoming” the other . To what extent are the alternative chromatin states described above dependent on the relative cellular levels of PcG and TrxG proteins ? Do these levels directly determine their binding equilibrium at target genes ? To address these questions we looked for changes in the genomic distributions of PC , TRX N-ter , ASH1 , H3K27me3 and H3K27ac in BG3 cells subjected to RNAi against TRX or PC . The knock-down of TRX led to a substantial drop in the amount of TRX N-ter and ASH1 bound to transcriptionally active PcG targets ( Figure 2C and 2D ) . This , however , caused no redistribution of either PC or H3K27me3 , leading us to conclude that the overall relationship of PcG and ASH1/TRX N-ter binding to target genes is not governed by a simple competitive equilibrium . Consistent with the lack of changes in the distributions of PcG proteins , the expression of target genes in the “active” chromatin state was not altered ( Figure 6D ) , indicating that once a stable chromatin state is established , at least over 9 generations in this cell culture , the expression of active PcG target genes is not very sensitive to the levels of associated ASH1 and TRX N-ter . While depletion of TRX had modest general effect on the levels of H3K27ac at active PcG targets , four regions showed exceptionally strong loss of K27 acetylation ( Figure 6A ) . Curiously , all four resided in the “balanced” state in untreated BG3 cells but , despite the initial presence of PcG proteins and H3K27me3 , the nearly complete loss of H3K27ac after TRX RNAi did not cause more binding of PcG or H3K27me3 . Expression was modestly reduced at all four regions ( Figure 6C ) but not completely repressed ( Figure 6B ) . We conclude that , while the level of H3K27ac in ASH1/TRX N-ter domains is positively correlated with the extent of transcriptional activity , it is not its primary determinant . These observations also suggest that H3K27ac is not the only anti-PcG mark . Similar to TRX knock-down , the reduction of PC levels to 20% of wild type ( Figure S6 ) did not change the chromatin states of the majority of the target genes . Despite the lack of global effect , we found eleven exceptional PcG-repressed loci which , upon PC knockdown , acquired binding of ASH1 , TRX N-ter , H3K27ac and a corresponding marked reduction of H3K27me3 , i . e . they switched their chromatin state from “fully repressed” to “balanced” ( Figure 7A and 7B ) . When expression of six of these loci was checked by RT-qPCR , switching of the chromatin state was paralleled in all cases by strong increase in transcription ( Figure 7D ) . In marked contrast , the expression of randomly chosen PcG target genes whose repressed chromatin state did not change remained constant ( Figure 7E ) . We interpret these findings to indicate that in BG3 cells the transcriptional activators of the exceptional 11 PcG target genes are available but at levels below the threshold necessary for derepression under normal PC levels . PC knockdown lowers the threshold for the amount of activator required to switch to a transcriptionally active chromatin state . We also conclude that derepression of PcG target genes is tightly linked to the formation of an ASH1/TRX N-ter domain and appearance of the H3K27ac mark . In flies , trx mutations act as suppressors of Pc mutations . We thus wondered whether the derepression of target genes caused by the reduction of PC levels in the cell culture is also sensitive to the level of TRX . Comparison of the effects of single PC knockdown versus simultaneous knockdown of PC and TRX on the expression levels of Abd-B and mirr genes indicates that this is indeed the case . Double knockdown of PC and TRX resulted in two- to four-fold lower transcription than knockdown of PC alone ( Figure 7C , Figure S10 ) , suggesting that , for a given dose of activator , the response of the target gene depends on the relative levels of PC and TRX . Key to our current understanding of PcG mechanisms is the fact that , while PcG proteins are present in most kinds of cells , the decision whether or not to repress a target gene depends crucially on whether that gene had been repressed in the previous cell cycle . This effect is responsible for the epigenetic maintenance of the repressed state and associated chromatin modifications . Similarly , through the action of TRX and ASH1 , a PcG target gene that had not been repressed tends not to become repressed in the subsequent cell cycle and remains susceptible to transcriptional activators . By comparing PcG/TrxG and transcriptional landscapes in three lines of Drosophila cultured cells we found that the full repertoire of chromatin states that PcG target genes can assume is not limited to the repressed state dominated by PcG mechanisms and the transcriptionally active state governed by TrxG proteins but in addition includes transcriptionally active “balanced” states subjected to simultaneous or at least rapidly alternating control by both PcG and TrxG proteins , and a transcriptionally inactive “void” state lacking both PcG and TrxG control ( Figure 8 ) . Thus , although PcG mechanisms first achieved fame for producing stable long-term silenced states in Drosophila homeotic genes , it is clear that , in the general case , PcG states are not necessarily stable nor long-term . Our results establish clearly that in robust PcG target regions ( i . e . Class I PcG target regions ) PcG and TrxG regulation are tightly coupled . Considering the role of MLL1 in the regulation of HOX genes [28] and the similarity between PcG complexes in flies and mammals , we expect that the same holds true for mammalian cells . It is possible that PcG and TRX recruitment to PRE/TREs share some DNA-binding proteins or DNA motifs . It will be important to determine whether PcG and TRX bind simultaneously or alternate over time . The nature of the TRX complex that binds to PRE/TREs remains enigmatic . To date the only TRX complex characterized biochemically is TAC1 , purified from Drosophila embryos [29] . It is said to contain uncleaved full length TRX , anti-phosphatase Sbf1 and histone acetyltransferase dCBP . We can detect no uncleaved TRX in the nuclei of cultured cells ( Y . B . S . , T . G . K . and V . P . , unpublished ) indicating that the TRX bound at PRE/TREs of repressed genes does not represent TAC1 . Proteolytically cleaved human orthologs of TRX , MLL1 and MLL2 have been purified as part of complexes similar in composition to the yeast COMPASS [7]–[8] , [14] . Although the PRE/TRE binds both parts of the cleaved TRX , it lacks some COMPASS components ( Y . B . S . , T . G . K . and V . P . , unpublished ) and lacks H3K4 trimethylation , suggesting that it involves a different complex whose composition is yet to be characterized . Consistent with genetic evidence , the presence of ASH1 and TRX at PcG target regions is linked to their transcriptional activity . However the two proteins show important differences in their behavior: binding of ASH1 is limited to transcriptionally active ( fully derepressed or balanced ) PcG targets and is not detected at completely repressed target loci . TRX is more complex . Both N-ter and C-ter parts of the protein associate with PREs regardless of the transcriptional status of their target genes and bind in the vicinity of TSS specifically when a target gene is transcriptionally active . In addition , the N-ter moiety of TRX together with ASH1 forms broad domains that encompass transcriptionally active PcG target genes . The different behavior of N-ter and C-ter parts of TRX may account for the discrepancy between reports of the co-localization of TRX and PcG proteins at many chromosomal sites [30] or PREs [17] , [18] , [24] , [25] and reports claiming that TRX binds exclusively to transcriptionally active target genes [31]–[33] . The different accounts are due to the use of anti-TRX antibodies specific to different parts of the protein . Whether the C-ter or N-ter specific antibodies were used , the number of TRX bound regions detected in our experiments is small compared to the number of active genes . This argues against a general role for TRX in transcription and is consistent with the limited number of regions detected on polytene chromosomes by various antibodies directed against C-ter or N-ter TRX [30] , [31] . In marked contrast to these observations , Schuettengruber et al . [34] have recently reported exclusive association of N-ter but not C-ter moiety of TRX with TSS of most active transcription units in the chromatin of embryonic cells . The same report also asserted that in embryonic cells TRX C-ter but not TRX N-ter is bound at PREs . Remarkably the TRX N-ter specific antibody used by Schuettengruber et al . [34] is reportedly the same as one of the two used in our experiments [31] . While we cannot exclude the possibility that the behavior of the N-ter moiety of TRX in embryos is totally different from that in cultured or salivary gland cells , we suspect that more likely the preparation of the antibody used by Schuettengruber et al . [34] cross-reacted with some general transcription factor particularly abundant in embryonic cells . This emphasizes the importance , even the necessity , of using two or more independent antibodies to verify genome-wide ChIP results . The RNAi knockdown experiments show that the broad binding of ASH1 and TRX N-ter within transcriptionally active PcG target regions is interdependent . This is consistent with the reported dissociation of ASH1 from polytene chromosomes of the salivary gland cells subjected to TRX RNAi [32] and the severe reduction of TRX N-ter binding to polytene chromosomes of ash1 mutant larvae [31] . Despite interdependence in binding there is no compelling evidence that ASH1 and TRX N-ter are in the same protein complex . Although an interaction between TRX and ASH1 has been reported , it was said to require the intact SET domain of TRX , which is absent from its N-ter moiety [35] . We have not found TRX N-ter and ASH1 to co-precipitate from nuclear extracts ( K . O . and V . P . unpublished ) , strengthening the impression that the two peptides do not interact directly . A histone mark associated with ASH1/TRX N-ter domains is H3K27ac . Acetylation of H3K27 can antagonize PcG activity by competing with the placement of the H3K27me3 mark , which in our model is needed for effective contact of the PRE complex with the promoter , as well as for stable PcG binding . In fact , targeting a histone H3 acetylase to a PRE is sufficient to prevent the epigenetic maintenance of repression [2] . A recent study by Tie et al . [36] indicates that in Drosophila the HAT responsible for bulk acetylation of H3K27 is CREB-binding protein ( dCBP ) , encoded by the nejire gene . Direct association of both TRX and ASH1 with dCBP was previously reported [29] , [37] . Consistent with this , the TRX knock-down experiment , which also impairs ASH1 binding , shows that either or both proteins promote acetylation of K27 in the chromatin of active PcG targets . It also shows , however , that the level of H3K27ac is not directly related to the amount of ASH1 or TRX N-ter bound . It is possible that a small amount of TRX and/or ASH1 remaining on the chromosomes after RNAi depletion is sufficient to target enough HAT activity to maintain nearly normal levels of H3K27ac . Alternatively H3K27 acetylation may be produced by a process that is not mechanistically linked to ASH1 or TRX but is promoted by the two . A global reduction of immunostaining of polytene chromosomes with anti-H3K27ac antibodies and a global elevation of immunostaining with anti-H3K27me3 antibodies in trx mutant larvae was recently reported [36] , [33] . We did not detect any global changes in either H3K27ac or H3K27me3 levels in our experiments . It is possible that the effects are generally weak and could only be detected on polytene chromosomes that consist of thousands of chromatin fibers bundled together . We note , however , that the reported changes of H3K27 acetylation and trimethylation levels involved numerous chromosomal sites , most of which , according to our data , do not stably bind PcG or TrxG , suggesting that in these cases the effect of trx mutation may have been indirect . Our microarray data show that narrower peaks of H3K27ac are also found near numerous active promoters , at genes not known to be PcG targets . We speculate that a role of H3K27ac at these promoters may be to antagonize dimethylation of H3K27 , which is abundantly distributed throughout the genome [38] and may have a general negative effect on transcription . In Drosophila cultured cells most PcG target genes are either completely repressed or fully derepressed or entirely devoid of both PcG and TrxG regulation . However in about 5% of cases , exemplified by the Psc-Su ( z ) 2 or inv-en loci , we see that binding of PcG complexes does not result in complete transcriptional silencing and can coexist with binding of ASH1/TRX N-ter . Whether a PcG target gene is capable of and will assume a “balanced” state may depend on the nature of the PRE , its binding complexes , and the promoter of the target gene . More likely , however , the major determinants are the nature and concentration of activators and repressors that act in concert with PcG/TrxG . Consistent with this idea , in imaginal disc cells , which are controlled by much more complex regulatory networks than cultured cells , simultaneous presence of both PcG and TrxG proteins at the transcriptionally active PcG target genes appears to be more common [18] , [39] . Interestingly a common feature of the “balanced” chromatin state in both cultured and imaginal disc cells is the confinement of ASH1/TRX N-ter binding to the regions immediately around the promoters . We take this as a hint that the formation of a broad ASH1/TRX N-ter domain starts in the vicinity of the TSS . Much has been said about the “bivalent” state , i . e . containing both PcG repressive marks and transcriptional activity marks , as characteristic of genes in mammalian embryonic stem cells [22] . In Drosophila , in cases such as those of the Psc-Su ( z ) 2 or inv-en loci , the balanced action of PcG and TrxG results in chromatin states similar in appearance to the “bivalent” state . We suppose that , like the “balanced” chromatin state of Drosophila PcG targets , the “bivalent” domains of embryonic stem cells would be associated with both PcG proteins and mammalian orthologs of TRX and ASH1 . PcG target genes may also assume a state lacking both PcG and TrxG proteins . The fact that in several instances the same locus resides in this “void” chromatin state in cultured cells of completely different origin argues against it being a product of genomic aberrations . Several experimental observations also indicate that the “void” state is not a peculiarity of cultured cells . Thus , in salivary glands the hh gene lacks PcG binding and H3K27me3 but remains transcriptionally inactive [40] , as in D23 cells . More recently a comparison of embryonic and imaginal disc cells showed that in many cases lack of PC binding was not accompanied by transcriptional activity [41] . The void chromatin state might be simply interpreted as a derepressed region that is transcriptionally inactive because the needed activator is absent . However , this does not explain why in this state TRX is also absent from the PRE , implying that neither PcG nor TRX binding is the default state of the PRE or that some specific condition prevents the recruitment of both . We have not detected other known repressive marks such as H3K9 methylation at these sites ( unpublished observations ) . In line with the finding that the lack of PcG repression in the “void” chromatin state does not automatically lead to the activation of a target gene is the observation that PC knock-down elicits a very specific genomic response . Remarkably Sex combs reduced ( Scr ) , Antennapedia ( Antp ) and Abd-B , the HOX genes whose derepression in heterozygous +/Pc− flies gives the famous Polycomb phenotype , are also among the genes most sensitive to PC knockdown in BG3 cells . The sensitivity of these genes cannot be explained by intrinsically poor recruitment of PcG proteins as Abd-B and Scr are controlled by multiple strong PREs capable of robust repression when placed next to a reporter gene [42]–[43] . We suppose that the reason for the differential sensitivity to PC levels lies in the availability of the corresponding transcriptional activators . We propose that in the BG3 cells the transcriptional activators of sensitive PcG target genes are present but at levels insufficient to override repression under normal conditions . The knockdown of PC lowers the threshold required for derepression . We suggest that a general role of PcG and TrxG mechanisms is to modulate the constraints on the levels of transcriptional activators required to switch the expression of PcG target genes . This concept helps to explain why , despite the implication of the PcG system in the control of all morphogenetic pathways , the reduction of PcG levels during differentiation of mammalian cell lineages [44] or tissue regeneration in flies [45] results in the execution of very specific genomic programs . The description of cell lines and culturing conditions are detailed in Text S1 . The source , amount and specificities of antibodies used for ChIP are indicated in Figure S11 , Figure S12 , and Table S9 . ChIP , hybridization to Drosophila tiling arrays v1 . 0 ( Forward ) ( Cat# 900587; Affymetrix ) and primary data processing were done as described ( [19] , Text S1 ) . To derive the genomic binding profile for a given protein , microarray hybridizations of the DNA from at least two independent ChIP experiments and two matching chromatin inputs were used ( Table S10 ) . The results were visualized with the Integrated Genome Browser ( Affymetrix ) . Details of computational definitions of bound regions and comparisons of data sets are described in Text S1 and Table S12 . For all analyses D . melanogaster Apr . 2004 ( dm2 ) genome assembly and genome annotation version 4 . 3 were used . Total RNA from 5×106 cells was isolated using Trizol ( Invitrogen ) and 3 µg were used for random primed synthesis of cDNA with First Strand Synthesis Kit ( GE Healthcare ) . The control reaction , omitting reverse transcriptase , was always run in parallel . cDNA was purified with QIAquick PCR Purification Kit ( Qiagen ) and eluted with 100 µl of elution buffer . qPCR was done with the Mx3000P instrument ( Stratagene ) in total volume of 10 µl containing 2 . 5 µl of cDNA solution , 1xSYBR Green PCR Master Mix ( ABgene ) , 100 nM of corresponding primers and 100 nM of ROX as a reference dye . The sequences of primers are given in Table S11 . Serial dilutions of genomic DNA were used to make the standard curve . The amount of cDNA for a gene of interest in the given preparation was expressed as a fraction of RpL32 cDNA . The RNAi was performed as described [46] with minor modifications detailed in Text S1 . Microarray data are available from GEO under accession numbers GSE18100 , GSE18176 and GSE18177 .	Polycomb ( PcG ) regulation has been thought to produce stable long-term gene silencing . Genomic analyses in Drosophila and mammals , however , have shown that it targets many genes that can switch state during development . Here we analyze the repertoire of alternative states in which PcG target genes are found in different Drosophila cell lines . In addition to the known repressed state , PcG targets can be found in a transcriptionally active state characterized by formation of an extended chromatin domain enriched in features that antagonize PcG repression . These features include the ASH1 protein , the N-terminal , but not C-terminal moiety of TRX , and histone H3K27 acetylation . ASH1/TRX N-ter domains and transcription are not incompatible with repressive marks , sometimes resulting in a “balanced” state modulated by both repressors and activators . Often , however , loss of PcG repression results instead in a “void” state , lacking transcription , H3K27acetylation , and other marks of active chromatin or binding of TRX or ASH1 . The propensity of a target gene to switch states depends on relative levels of PcG , TRX , and activators .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "molecular", "biology/chromatin", "structure", "molecular", "biology/histone", "modification", "molecular", "biology/transcription", "initiation", "and", "activation", "genetics", "and", "genomics/functional", "genomics", "molecular", "biology/bioinformatics", "genetics", "and", ...	2010	Alternative Epigenetic Chromatin States of Polycomb Target Genes
Previously , we showed Leishmania donovani Ufm1 has a Gly residue conserved at the C-terminal region with a unique 17 amino acid residue extension that must be processed prior to conjugation to target proteins . In this report , we describe for the first time the isolation and characterization of the Leishmania Ufm1-specific protease Ufsp . Biochemical analysis of L . donovani Ufsp showed that this protein possesses the Ufm1 processing activity using sensitive FRET based activity probes . The Ufm1 cleavage activity was absent in a mutant Ufsp in which the active site cysteine is altered to a serine . To examine the effects of abolition of Ufm1 processing activity , we generated a L . donovani null mutant of Ufsp ( LdUfsp−/− ) . Ufm1 processing activity was abolished in LdUfsp−/− mutant , and the processing defect was reversed by re-expression of wild type but not the cys>ser mutant in the LdUfsp−/− parasites . Further LdUfsp−/− mutants showed reduced survival as amastigotes in infected human macrophages but not as promastigotes . This growth defect in the amastigotes was reversed by re-expression of wild type but not the cys>ser mutant in the Ufsp−/− indicating the essential nature of this protease for Leishmania pathogenesis . Further , mouse infection experiments showed deletion of Ufsp results in reduced virulence of the parasites . Additionally , Ufsp activity was inhibited by an anti-leishmanial drug Amphotericin B . These studies provide an opportunity to test LdUfsp−/− parasites as drug and vaccine targets . Leishmaniasis is a spectrum of diseases caused by protozoan parasites belonging to several different Leishmania species . These blood borne pathogens are currently prevalent in 88 countries around the World with an estimated 2 million new cases each year [1] . At present there are no effective vaccines against any of the clinical forms of leishmaniasis . Further drugs against this parasite are becoming limited in their usefulness due to inappropriate use and because of the development of drug resistance against pentavalent antimonials [2] . Recent advances in genome sequencing ushered in post-genomic analysis of Leishmania parasites in terms of parasite biology in the sand fly vector and mammalian host , including host responses [3] . Yet , the parasitic factors involved in pathogenesis associated with any form of leishmaniasis remain to be fully understood , as the parasite virulence is determined by numerous factors . Protein modifications by ubiquitin and ubiquitin-like proteins ( Ubls ) are widely described in eukaryotes [4] . The modification of target proteins by Ubls involves covalent attachment of Ubls to a substrate protein [5] . The best-known consequence of ubl conjugation is the targeting of proteins for degradation by the proteasome [6] . In addition to proteasomal targeting , conjugation by Ubls have been shown to affect a broad range of functions including subcellular localization , endocytosis , membrane trafficking , protein kinase activation , DNA repair , chromatin dynamics and protein-protein interactions [7] . Ubiquitin-fold modifier 1 ( Ufm1 ) that possesses a similar tertiary structure compared to ubiquitin , has recently been identified as a novel protein-conjugation system [8] . Attachment of Ufm1 to its substrate proteins has been shown to follow enzymatic reactions commonly found in many ubl conjugation reactions . Ufm1 is synthesized as a precursor form and processed C terminally by two specific proteases , UfSP1 and UfSP2 in humans [9] . The processed Ufm1 is activated by the E1-like enzyme , Uba5 , and then transferred to an E2 enzyme , Ufc1 . Finally the Ufm1 is covalently conjugated to the substrate proteins via an E3-like enzyme Ufl1 [10] . Studies in mouse revealed that an ER protein named C20orf116 , with unknown function is the substrate protein for mammalian Ufm1 [11] , [12] . Although Ufm1 has been studied in humans , its functions are still not completely understood . Deletion of Uba5 , the Ufm1 activating enzyme resulted in embryonic lethality in mice [12] suggesting that genetic manipulation of some of the Ufm1 associated proteins may not be feasible in mammalian cells . We have recently shown that the human protozoan parasite Leishmania donovani contains the full complement of Ufm1 conjugation reactions that is revealed by the presence of substrate proteins conjugated to Ufm1 [13] . Among the substrate proteins conjugated to Leishmania Ufm1 is a mitochondrial trifunctional protein ( MTP ) involved in the β-oxidation of fatty acids [13] , [14] . Deletion of Ufm1 from L . donovani resulted in loss of β-oxidation of fatty acids due to the absence of conjugation\ of MTP with Ufm1 [14] . These results suggested the importance of Ufm1 mediated conjugation in the physiology of L . donovani . In L . donovani Ufm1 must be processed to remove the 17-aminoacid C-terminal extension to expose the glycine residue important for the conjugation to occur indicating the presence of such processing activity in the parasite [13] . In this report , we have demonstrated the presence of Ufm1 processing activity of LdUfsp in L . donovani and this activity can be inhibited by antileishmanial compound Amphotericin B . Deletion of Ufsp affects this Ufm1 processing activity and results in reduced survival of amastigotes as was previously seen with Ufm1 deletion [14] further suggesting the importance of the role of Ufsp mediated Ufm1 processing in the parasite survival . Reduction in virulence of Leishmania Ufsp null mutant implies that this novel strain can have important applications as a live attenuated vaccine candidate as has been demonstrated by us and others using gene deleted Leishmania parasites and additionally aid identify drug candidates that target Ufsp and mimic the loss of Ufsp function in Leishmania [15] . To our knowledge Leishmania Ufsp1 null mutant is the first of its kind and Ufsp null mutant can be useful not only in the understanding of Leishmania pathogenesis but also in other organisms including humans . Leishmania donovani promastigotes ( strain 1S , WHO designation: MHOM/SD/62/1S ) were grown in M199 medium containing 10% heat inactivated fetal bovine serum . Promastigotes were transfected by electroporation and selected for growth in medium containing Nourseothricin ( 100 µg/ml ) . Recombinant Ufsp−/− parasites reexpressing wild type or mutant Ufsp were similarly selected in medium containing Geneticin ( G418; 100 µg/ml ) . These drug-resistant cells were used in all subsequent experiments . Axenic amastigotes of wild type or the mutant lines were generated following a published protocol [16] , [17] . For the analysis of enzymatic activity encoded by the putative LdUfsp , FRET-based fluorescent probes were prepared following the methodology described by Tatham and Hay [18] with brief modifications . Briefly , coding sequences containing either wild type LdUfm1 or the mutant LdUfm1 in which C'terminal Gly is altered to Ala were amplified with flanking BamHI and HindIII sites . The resulting fragments were ligated into the BamHI-HindIII digested pHis-TEV-30a-YFP-SUMO-1-ECFP plasmid , thus replacing the SUMO-1 . The ligation was confirmed by nucleotide sequencing of the bacterial expression plasmids . The plasmids were used to transform E . coli ( BL21-PLysS , Invitrogen ) . Recombinant fusion proteins containing the wild type or the mutant LdUfm1 were prepared as described [18] . For preparing the recombinant protein encoded by the putative LdUfsp , the open reading frame corresponding to 427–1518 bp was PCR amplified and ligated into pEXP5-CT-Topo plasmid vector . The authenticity of the plasmid was confirmed by nucleotide sequencing . BL21-pLysS bacterial cells were transformed with this expression plasmid and the recombinant Ufsp protein was purified under native conditions . This recombinant protein was used for preparing polyclonal antibodies in rabbit ( Spring Valley Labs , Sykesville , MD ) . For fractionation of Ufm1 processing activity , extracts were prepared from the Leishmania cultures either wild type , Ufsp−/− or Ufsp−/− parasites re-expressing Ufsp ( 2–3×109 cells ) in a buffer containing 25 mM Tris HCl pH 8 . 0 , 150 mM NaCl , 1%NP-40 . The extracts were passed through a strong anion exchange column ( Q-column , Pierce Biotechnology ) and fractions were collected in a buffer containing increasing salt concentration ( 0 . 2–2 M NaCl ) following the protocol suggested by the manufacturer . The fractions were tested for the Ufm1 processing activity in a FRET based cleavage assay and the positive fractions were pooled . These pooled fractions were dialyzed against a buffer containing 25 mM Sodium acetate pH 5 . 5 . These were passed through strong cation exchange column ( S-column , Pierce Biotechnology ) and fractions were collected in a buffer containing increasing salt concentration ( 0 . 2–2 M NaCl ) . The fractions were dialyzed against PBS and used in FRET based activity assays . Cell fractionation for localization studies was done using digitonin reagent as described previously [19] . The activity of the recombinant LdUfsp was measured in an assay buffer containing 50 mM Tris-HCl , pH 7 . 5 , 150 mM NaCl , 5 mM β-mercaptoethanol , 0 . 1 mg/ml bovine serum albumin ( BSA ) . The recombinant enzyme ( 0 . 5 µg ) and the FRET substrate ( 250 nM per well ) were dispensed into the wells of a black 96 well clear bottom plate and followed by single readings at 480 and 530 nm at the start of the reaction and at the end of 60 minute incubation at 37°C in a Spectramax M5 fluorescent plate reader . The reduction in the fluorescence at 530 nm ( Δ530 ) corresponding to YFP was measured after subtracting the reduction in 530 nm due to incubation at 37°C for 60 min in wells containing the probes alone . A reduction in the YFP fluorescence would indicate that the cleavage of Ufm1 has occurred and the emissions from the C′ terminal ECFP can no longer excite YFP fused at the N′ terminal end . For co-immunoprecipitation analysis , 1×108 Leishmania amastigotes were lysed in 1 ml of NET buffer ( 150 mM NaCl , 1 mM EDTA , 10 mM Tris–HCl , pH 7 . 5 , 1% Nonidet P-40 with protease inhibitor cocktail ) and the lysate was centrifuged at 12000 rpm for 20 min at 4°C to collect the supernatant . Five µl of anti-Ufsp antibodies were added to 500 µl to the lysate and the mixture was incubated under constant rotation overnight at 4°C . The complexes were immunoprecipitated with 25 µl of Protein-A Sepharose beads by incubation under constant rotation at 4°C for 1 hr . The precipitated complexes were washed five times with ice-cold NET buffer and eluted by boiling for 5 min in SDS sample buffer in the presence of β-mercaptoethanol . The supernatant was subjected to SDS–PAGE and analyzed by immunoblots with anti-Ufm1 antibodies . Human elutriated monocytes were resuspended at 1 . 8×105 cells/ml in RPMI medium containing 10% FBS and macrophage colony-stimulating factor ( 20 ng/ml , ProSpec , Israel ) , plated at 0 . 5 ml/well on eight-chamber Lab-Tek tissue-culture slides ( Miles Laboratories ) and incubated for 9 days for differentiation into macrophages . The macrophage infection experiments were performed essentially as described earlier [19] . 5- to 6-wk-old female BALB/c mice were infected via tail vein with 3×106 stationary-phase wild type , LdUfsp−/− or LdUfsp−/−+UfspWT parasites . Infected mice were sacrificed after different periods of infection parasite load was measured in spleens and livers from the infected mice by limiting dilution assay . The drug resistance markers nourseothricin was used to obtain LdUfsp−/− . To generate the targeting construct , a 793 bp fragment from the 5′ region and an 735 bp fragment from the 3′ region partly overlapping with the LdUfsp open reading frame were amplified by PCR using L . donovani genomic DNA . The 5′ flanking fragment included part of the N′ terminal open reading frame ( 515 bp from the starting methionine ) encoding LdUfsp in addition to the UTR . This was done due to the limited amount of 5′ flanking sequence ( 452 bp ) that separates LdUfsp from the upstream ORF . Similarly , the 3′ flanking fragment included part of the LdUfsp open reading frame ( 385 bp from 1137–1521 ) encoding LdUfsp due to the limited amount of 3′ flanking sequence ( 591 bp ) that separates LdUfsp from the downstream ORF . The primers used to amplify 5′ flanking fragment included restriction sites HindIII and BamHI . Similarly , the primers added SpeI and XbaI sites to the 3′ flanking fragment . The drug resistance marker nourseothricin was amplified with primers that add BamHI and SpeI to the open reading frame . These DNA fragments were subcloned into the pCR2 . 1-Topo vector and the nucleotide sequence was determined to ensure fidelity . The plasmid containing the 5′ flanking fragment was digested with HindIII/BamHI , gel purified and ligated into a similarly digested plasmid containing nourseothricin . The resultant plasmid , containing both the 5′flanking region and the drug resistance markers was digested with SpeI/XbaI and the 3′flanking fragment isolated by SpeI/XbaI digestion was ligated into these sites . The authenticity of the final plasmid was confirmed by DNA sequencing . For the purpose of transfection , the targeting construct was prepared by digestion with HindIII/XbaI , which cuts out a linear fragment containing the Ufsp 5′ flanking sequence , the nourseothricin gene and the Ufsp 3′ flanking sequence . The fragment was gel purified and used in transfection . Total genomic DNA was isolated from promastigotes with the Wizard genomic DNA purification kit ( Promega Biosciences ) , following the method suggested by the manufacturer . The DNA was digested with restriction endonuclease BglI or XhoI and separated on 1% agarose gels . Southern blot analysis of the resolved DNA was done as described previously using a 32P-labelled 620 bp partial LdUfsp coding sequence corresponding to 516–1136 bp of the Ufsp open reading frame as a probe [14] . To restore Ufsp expression in the LdUfsp−/− parasites , the LdUfsp ORF was first PCR amplified using a LdUfsp containing plasmid as template and the following a forward primer: 5′-ACTAGT ATG GAG GAT GTC GTG ACC GGC GTT GC-3′ , and a reverse primers: 5′- ACTAGT TCA CTT GTC ATC GTC GTC CTT GTA GTC TCG AGG ATC GAA CAG GTC AAC GCG TGG-3′ that amplify a wild type Ufsp coding sequence or a C327S variant of Ufsp in two independent amplification reactions . These oligos introduced a FLAG epitope tag at the C′ terminus of the recombinant protein . The amplified product was subcloned into the pCR2 . 1-TOPO cloning vector . The fidelity of the cloned sequence was verified by nucleotide sequencing . The SpeI insert was ligated into the SpeI site of the pKS-Neo vector [20] and the recombinant plasmids , pKSNeo-LdUfspWT or pKSNeo-LdUfspC>S was transfected into the LdUfsp−/− promastigotes as described previously [19] . Transfected promastigotes were selected with minimal dose of G418 ( 20 µg/ml ) . Leishmania axenic amastigotes were stained with a solution containing acridine orange and propidium iodide and viable cells were counted using Cellometer instrument . Immunofluorescence assay was performed essentially as described in [13] . Leishmania parasites were fixed in 2% p-formaldehyde-PBS for 5 min and allowed to adhere to poly-L-Lysine treated glass slides . After blocking with 5% BSA-PBS , the cells were incubated with either anti-Ufsp antibodies or pre-immune serum . Cellular localization of endogenous Ufsp was detected using anti-rabbit IgG-Alexa488 conjugate antibodies . Nucleic acids were stained with DAPI . Cell fractionation was performed as described previously [19] . Briefly , Leishmania cells were washed three times in 15 ml MES buffer ( 20 mM MOPS , pH 7 . 0 , 250 mM sucrose , 3 mM EDTA ) . The cell pellet was resuspended in 0 . 2 ml MES buffer containing 1 mg/ml digitonin and protease inhibitor cocktail ( Roche Applied Science ) . The suspension was incubated at room temperature for 5 minutes and centrifuged at 10 , 000 g for 5 minutes . The resulting supernatant was collected as a cytosolic fraction , and the heavy membrane pellet enriched for mitochondria was resuspended in phosphate buffer ( 20 mM sodium phosphate , pH 7 . 0 , 3 mM EDTA ) . The polyclonal antibodies against Ufm1 , Ufc1 , Ufsp , TatD proteins were prepared in our laboratory and were described previously [13] , [21] . Student's t-test was used to test the statistical significance of the observed differences . Protein modifications mediated by the Ubls in the parasitic organisms such as Leishmania are of considerable interest . Our previous studies analyzing the Ufm1 mediated protein conjugation studies in Leishmania donovani demonstrated the presence of an Ufm1 conjugation system in this parasite . Further , evidence for an Ufm1 processing reaction to generate a conjugatable Ufm1 by a C- terminal hydrolase activity has been demonstrated in Leishmania by the absence of Ufm1 processing when an Ufm1G>A mutant was expressed in Leishmania [13] . Therefore , we searched the genome databases of Leishmania infantum to find if sequence homologs of human Ufm1 specific proteases such as Ufsp1 ( containing 217 amino acid residues ) and Ufsp2 ( containing 416 amino acid residues ) involved in the processing of human Ufm1 [9] are present in Leishmania . Studies of human Ufsp have also shown the deconjugation activity of Ufsp i . e . , cleaving of Ufm1 from its substrate proteins [9] . Database searches of the L . infantum and other trypanasomatid genomes using human Ufm1 processing peptidases as query sequences revealed the existence of a putative Ufm1 specific protease in L . infantum ( www . tritrypdb . org LinJ . 34 . 3830 ) . Protein sequence alignments using ClustalW ( http://www . ebi . ac . uk/Tools/msa/clustalw2/ ) revealed that the overall amino acid sequence similarity between mouse , human and trypanosomatid Ufsp is relatively low ( ≤25% ) ( Fig . 1 ) . However , similar to human deubiquitinating enzymes and Ubl specific proteases , trypanosomatid Ufsp molecules have highly conserved cysteine and histidine residues that form a “Cys box” and a “His box” ( Fig . 1 boxed with arrows ) . Unlike humans that have Ufsp1 and Ufsp2 , only a single copy of the Ufsp gene is found in Leishmania genome . Structural analysis of the putative Leishmania Ufsp protein revealed that it is similar to Ufsp2 of mouse and humans . Together , the sequence homology , conservation of catalytically important amino acid residues and presence of Ufm1 processing activity in Leishmania as reported in our previous studies characterizing Ufm1 conjugation associated enzymes [13] indicated that Leishmania genome contains a gene encoding Ufsp . Therefore , we explored the biochemical characteristics of Ufm1 processing activity of this putative Ufsp protein encoded in Leishmania genome . To investigate the protease activity that is responsible for processing the C-terminal extension of Ufm1 , we prepared FRET associated activity probes based on the protocol developed by Tatham and Hay [18] . Fluorescence based activity assays offer several advantages over the more commonly used gel based approaches in terms of sensitivity and allow quantitative assessments . Recombinant YFP-Ufm1-ECFP and the mutant version i . e . YFP-Ufm1G>A-ECFP were purified from bacteria and SDS-PAGE analysis showed the fusion proteins as a single band on the gel ( Fig . 2A ) . To test the fluorescence properties of these probes and to verify if the recombinant proteins allow the FRET activity to occur , the probes were analyzed on a fluorescent plate reader . Analysis revealed that the fluorescent probes when excited at 405 nm , result in two emission lines; one at 480 nm corresponding to ECFP and another at 530 nm corresponding to YFP ( Fig . 2B ) . Storage of the fluorescent probes at −80°C did not result in measurable deterioration of FRET activity ( data not shown ) . To test if the recombinant protein encoded by the putative LdUfsp possesses the Ufm1 processing activity , the purified recombinant Ufsp protein was incubated with the fluorescent probes YFP-Ufm1-ECFP and YFP-Ufm1G>A-ECFP . Purification of recombinant protein encoded by the partial open reading frame resulted in approximately 43 and 39 kDa bands on a SDS-gel ( Fig . 3A , lane 2 indicated by arrow heads ) . Recombinant human Ufsp2 has been shown to undergo spontaneous cleavage [22] indicating that similar processing might be occurring in LdUfsp ( short fragments in Fig . 3A ) . Ufm1 substrate cleavage results showed that the putative Ufsp protein indeed possesses Ufm1 processing activity ( Fig . 3B ) . This cleavage was not detected in presence of the mutant probe ( Ufm1G>A ) indicating that the cleavage of Ufm1 by the putative Ufsp is specifically occurring at the glycine residue ( Fig . 3C ) . To further verify the specificity of the Ufm1 cleavage activity by the Ufsp , we performed the activity assays using an unrelated recombinant protein ( rTbEndoG , a nuclease produced in E . Coli [19] ) purified in a parallel experiment as a source of the enzyme . Results showed that this unrelated protein did not cleave the fluorescent Ufm1 substrate indicating that any contaminant protein in the bacterial lysate as an unlikely source for the observed Ufm1 cleavage with Ufsp ( Fig . 3B , C ) . To test whether this in vitro Ufm1 processing activity of the recombinant Ufsp can be inhibited by the anti-parasitic drugs that are commonly in use , and thus to verify if Ufsp activity is amenable to inhibition we tested a limited set of the anti-parasitic drugs in our assay . We tested Nifurtimox , Amphotericin B , Melarprosol , and Geramin following the doses most commonly used in in vitro studies [23]–[26] . Results showed that Nifurtimox , Melarprosol , and Geramin did not inhibit Ufm1 processing activity at the concentrations tested ( low , medium and high , Fig . 3D ) . However , Amphotericin B showed inhibitory effects on the Ufm1 processing activity and this inhibition correlated with the dose used ( Fig . 3D ) . To test if the observed inhibition is also accompanied by binding of Amphotericin B to the recombinant Ufsp , we performed binding activity assays using biotinylated Ufsp ( Fig . 3 F ) on a streptavidin sensor on an Octet instrument . Binding of Amphotericin B to the biotinylated Ufsp produces an optical interference pattern that is recorded by the instrument . Results showed that Amphotericin B can bind to the biotinylated Ufsp protein ( Fig . 3E ) suggesting a pharmacological basis for the observed inhibitory activity . To investigate if the putative Ufsp is indeed the in vivo source for the Ufm1 cleavage activity and to further probe the in vivo functions of the Ufsp in L . donovani , we generated a L . donovani Ufsp null mutant by homologous recombination ( Fig . 4A ) . L . donovani promastigotes were transfected by electroporation and the two alleles of LdUfsp were replaced by recombination with a targeting construct containing nourseothricin marker flanked by DNA fragments corresponding to 5′ and 3′ untranslated regions ( UTR ) of LdUfsp . Southern hybridization of genomic DNA isolated from the LdUfsp null mutants with a 32P labeled probe corresponding to the targeted region of Ufsp revealed that both alleles of LdUfsp were lost in the null mutant ( Fig . 4B LdUfsp−/−lanes probed with LdUfsp ) and were replaced by the nourseothricin marker ( Fig . 4B LdUfsp−/−lanes probed with NAT ) . Loss of LdUfsp expression in the LdUfsp−/− parasite was confirmed by Western blot analysis using anti-Ufsp antibody ( Fig . 4C , LdUfsp−/− lane ) . This antibody recognizes the full length protein of LdUfsp with a calculated mass of 56 kDa with an apparent mobility of ∼70 kDa ( Fig . 4C ) . LdUfsp expression was restored by transfecting the null mutant cells with the pKSNeo vector containing the coding sequence of either wild type Ufsp or a mutant Ufsp where the catalytic cysteine was altered to a serine . ( Fig . 4C , LdUfsp−/−+UfspWT , LdUfsp−/−+UfspC>S ) . Immunoblotting with an anti-tubulin antibody ( anti-chicken tubulin , Sigma ) revealed equal loading of protein ( Fig . 4C , anti-α-tubulin ) . The in vitro biochemical assays using FRET based activity probes indicated that the putative Ufsp has the Ufm1 processing activity . To examine if this Ufsp is indeed the protease responsible for the Ufm1 processing in vivo , we fractionated the cell extracts obtained from the L . donovani wild type parasites . For this purpose we utilized fractionation on strong anion columns followed by testing of Ufm1 processing activity . We reasoned that by such fractionation we should be able to concentrate the Ufm1 processing activity into few select fractions reproducibly . Results showed that by using FRET based activity probes we were able to fractionate Ufm1 processing activity from the parasite lysates ( Fig . 5A ) . We resolved the fractions obtained on SDS-PAGE and immunoblot with an anti-Ufsp antibody showed that the protein is enriched in fractions where we observed the peak Ufm1 processing activity ( Fig . 5A lower panel ) . By following an identical fractionation scheme , a comparison of the fractions collected from L . donovani wild type , LdUfsp−/− and LdUfsp−/− mutants re-expressing Ufsp should allow us to examine whether in vivo Ufm1 processing activity was lost in the LdUfsp−/− mutant . This should also allow us to demonstrate that there are no other cellular sources capable of Ufm1 processing activity . Results showed that fractions from wild type parasites corresponding to 0 . 3–0 . 4 M NaCl possess the peak Ufm1 processing activity indicated by the reduction in the 530 nm emission when these fractions were incubated with the YFP-Ufm1-ECFP substrate ( Fig . 5B ) . In contrast , similar fractions from the LdUfsp−/− mutant did not show any reduction in the 530 nm emission indicating the intactness of the YFP-Ufm1-ECFP substrate , thus absence of Ufm1 processing ( Fig . 5B ) . On the other hand , the Ufm1 processing activity was significantly restored in the fractions from the LdUfsp−/− mutants re-expressing Ufsp indicating that indeed Ufsp is the source of Ufm1 processing activity and also that fractionation of exogenously expressed Ufsp follows that obtained from wild type cells . To further demonstrate that the observed Ufm1 processing activity in the fractions obtained from wild type and Ufsp−/− cells re-expressing Ufsp is specific and dependent on the presence of glycine residue at the C′ terminus of the Ufm1 , we performed the Ufm1 processing activity assays with the fractions ( Fractions 2 , 3 and 4 ) obtained from the wild type , Ufsp−/− and Ufsp−/− cells re-expressing Ufsp in presence of the mutant probe YFP-Ufm1G>A-ECFP ( Fig . 5C ) . Results showed that the cleavage did not occur at a random residue ( as there is no glycine at the C′ terminus ) in the mutant probe incubated with positive fractions obtained in the previous experiment ( Fig . 5B ) . Further , to test whether the endogenous Ufm1 remains unprocessed in the Ufsp−/− mutant cells and therefore results in a small shift in the molecular weight , we performed an immunoblot from the whole cell lysates prepared from the wild type and Ufsp−/− mutants and probed with an anti-Ufm1 antibody . Results suggested that indeed the Ufm1 remains as unprocessed that is detectable as a slightly higher molecular mass in the Ufsp−/− mutants compared to the wild type cells ( Fig . 5D ) . The slightly higher molecular weights upper band that is reactive to the anti-Ufm1 antibodies is evident in wild type and the Ufsp−/− mutant lanes . This band is identical in both lanes with no apparent shift indicating that this is likely a non-specific reactive band . In order to characterize the molecular interaction between Ufm1 and the processing enzyme Ufsp , in L . donovani , a co-immunoprecipitation assay was performed with an anti-Ufsp antibody and the blots were probed with an anti-Ufm1 antibody ( Fig . 6 ) . Results showed that Ufm1 does interact with Ufsp as revealed by the presence of a ∼12 kDa band corresponding to Ufm1 in wild type Leishmania ( Fig . 6; WT lane ) and not in the Ufsp−/− , indicating the specificity of the molecular interaction between Ufm1 and Ufsp proteins ( Fig . 6 ) . Together , these results demonstrated that Ufm1 can interact with Ufsp protein . This result is consistent with our previous observation that Ufm1 is processed by the Ufsp in Leishmania . Previously we have shown that proteins mediating Ufm1 conjugation such as Uba5 , Ufc1 as well as a majority of Ufm1 are localized in the mitochondria [13] . Since the Ufm1 processing is the first step in the Ufm1 conjugation cascade , we wanted to investigate the cellular distribution of endogenous LdUfsp protein in Leishmania . To this end , antibodies against L . donovani Ufsp protein were used in immunofluorescence studies using wild type and Ufsp−/− cells . Immunofluorescence assays revealed that endogenous LdUfsp has a diffuse localization in the cytoplasm of L . donovani . ( Fig . 7 , top panel ) . The Ufsp−/− mutants did not show any reactivity with the anti-Ufsp antibodies ( Fig . 7 , bottom panel ) . No background reactivity was obtained with the pre-immune serum in these experiments ( data not shown ) . To further confirm the localization of the Ufsp in L . donovani , we performed cell fractionation studies followed by immunoblotting . Results showed that Ufsp is predominantly localized in the mitochondrial fraction but also present in cytosolic fraction in considerable amount ( Fig . 7B lanes M and C ) . The authenticity of fractionation was confirmed by using antibodies for mitochondrial ( Ufc1 ) and cytosolic ( Tat-D ) markers . Results showed that Ufc1 and Tat-D markers were only detected in mitochondrial and cytosolic fractions alone respectively ( Fig . 7B lanes M and C ) . These results further suggest that most of the components of Leishmania Ufm1 conjugation pathway including Ufsp are localized in mitochondria . We have previously shown that Ufm1-deficient L . donovani parasites show defects in fatty acid metabolism in the amastigote stage of the parasite and since the fatty acid metabolism is necessary for energy generation , we wanted to analyze what effect loss of Ufm1 processing in the Ufsp−/− mutants have on growth in vitro as well as ex vivo in human macrophages . Survival of the axenic amastigotes was analyzed by counting viable cells over a period of 7 days ( Fig . 8A ) . The data showed that the Ufsp−/− parasites failed to grow as amastigotes after 3 days in culture ( Fig . 8A ) . Re-expression of wild type Ufsp restored the growth ( Fig . 8A ) . The growth was not restored when a mutant Ufsp ( UfspC>S ) was expressed in the Ufsp−/− background ( Fig . 8A ) . The growth of promastigotes was not affected in Ufsp−/− mutants ( data not shown ) . Next , we examined their growth in macrophages ex vivo . To this end , in vitro differentiated human macrophages were infected with stationary phase cultures of wild type , LdUfsp−/− and Ufsp re-expressing either wild type ( UfspWT ) or mutant ( UfspC>S ) promastigotes ( Fig . 8B ) . The results at 6 h post-infection showed that the percentage of macrophages that are infected with the parasites was similar with all four cell types . These macrophage cultures were subsequently examined at 1 , 2 , 3 , 4 , 5 , 6 and 7 days post-infection , and the percentage of infected macrophages was calculated . After 4 days the Ld Ufsp−/− and Ld Ufsp−/−+UfspC>S expressing cells start to show significant decline in growth ( Fig . 8B ) , whereas wild type control cells and add-back cells with wild type LdUfsp continue to grow inside macrophages ( Fig . 8B ) . By day 6 , LdUfsp−/− and Ld Ufsp−/− with UfspC>S expressing cells dropped to <1 parasite/macrophage . To further test the virulence of LdUfsp−/− parasites in vivo , we conducted mouse infection experiments and measured the parasite burdens at defined time intervals . Results showed that deletion of Ufsp leads to reduced parasite growth and by 10 weeks post infection , parasite burdens reached to non-detectable levels in LdUfsp−/− infected liver and spleens ( Fig . 8C and D ) whereas in the mutants parasites in which Ufsp is added back ( Ufsp−/−+UfspWT ) the parasite growth reached comparable levels to that of wild type parasites . These results indicate that lack of Ufm1 processing in LdUfsp−/− could result in growth reduction in amastigotes as was observed by us in the Ufm1−/− parasites [14] . Ubiquitin-like protein modifiers ( Ubls ) although share the β-grasp fold structure among the various members , regulate a strikingly broad set of cellular processes including proteolysis , endocytosis , membrane trafficking , protein kinase activation , DNA repair , autophagy and chromatin dynamics in eukaryotic cells [4] , [27] . The dynamic process of addition and removal of ubl molecule to its substrates is widely conserved throughout in evolution in organisms from yeast to humans and more recently demonstrated even in prokaryotes [28] . Therefore investigation of the various Ubls and their functions in parasitic protozoa assumes importance as identification of ubl conjugation and deconjugation reactions uniquely found in the parasites might provide novel targets for developing inhibitors that block these activities . Inhibition of LdUfsp by Amphotericin B in our studies is of particular interest since this demonstrates the feasibility of developing novel inhibitors against Ufsp . However , there are serious side effects of this drug treatment in patients . Since our results showed the binding of the drug to Ufsp , inhibition of Ufsp activity and essential nature of this activity for Leishmania pathogenesis it will be important to explore whether non-toxic regimens of Amphotericin B such as liposomal form of Amphotericin B [29] can also inhibit this activity and molecules analogous to Amphotericin B could be tested as inhibitors against Ufsp to support their use in treatment . Ufm1 in Leishmania is unusual in its structure compared to the mammalian Ufm1 . Leishmania Ufm1 has a 17 amino acid long C′ terminal extension where as the human homolog has only 2 residues beyond the C′ terminal glycine implies that the active sites of the Ufm1 processing enzyme from mammalian and Leishmania are likely to be different in terms of their three dimensional conformation . Such possibility makes further studies on Leishmania Ufsp structure function highly attractive . Low overall homology between human and Leishmania Ufsp proteins ( ∼25% Fig . 1 ) makes it feasible to identify small molecular compounds that selectively inhibit parasitic Ufsp activity and could help in development of novel drugs . Previous studies in Plasmodium have demonstrated the utility of such approaches in identifying subtle differences in the plasmodium specific ubiquitin and Nedd8 hydrolase UCHL3 enzymatic activities [30]–[32] that may be amenable to inhibition by novel inhibitors . More recently , Plasmodium SUMO-specific protease ( PfSNEP1 ) has been shown to possess a unique cleavage sequence preference compared to human SUMO-specific protease and this difference was exploited in identifying inhibitors specific to Plasmodium [33] . Such studies in Leishmania parasites are warranted since the parasites belonging to Leishmania species cause considerable mortality and resistance against the available drugs renders the therapies ineffective . Previous studies to identify the Ubl processing protease activities relied on chemically modifying the Ubl into an electrophile in an active-site directed reaction and co-precipitating the processing activity using this probe followed by identification by proteomic methods [9] , [34] . In the present study we utilized genetic methods to generate an Ufsp null mutant that not only allowed us to demonstrate the Ufm1 processing activity biochemically but also allowed us to investigate the consequence of this deletion to the parasite survival . Deletion of Ufsp resulted in abolition of Ufm1 processing demonstrating that Ufsp is specific to Ufm1 and other protease activities can not compensate for the loss of Ufsp . This is also demonstrated by the re-expression experiments in which the growth of the Leishmania amastigotes is restored only when a wild type Ufsp is expressed but not in case of the UfspC>S mutant . This is consistent with the absence of Ufm1 processing activity in the UfspC>S mutant suggesting that a functional Ufm1 processing activity is necessary for the survival of the parasite . Ubls such as Ubiquitin , SUMO and NEDD have been identified in medically important parasitic protozoa and are being investigated in these organisms [30] , [31] . Recently , Ubl targets that are unique to parasitic protozoa have been described . For instance , several proteins including metacaspase-3 , thymidine hydroxylase and histone acetyl transferase were found to be potential sumoylation targets in Trypanosoma cruzi [35] , [36] . Arguably , sumoylation of thymidine hydroxylase and histone acetyl transferase , DNA and chromatin modification proteins might represent regulation of transcriptional activity in these parasites . Similarly , SUMOylation of paraflagellar rod protein has been shown to be essential for flagellar homeostasis in T . cruzi [37] . These unique modifications by SUMO might represent novel functions in the protozoan parasites such as T . cruzi . Previously we demonstrated the existence of Ufm1 conjugation pathway in L . donovani [13] . To understand the significance of post-translational modifications in regulating Leishmania differentiation mediated by the Ufm1 conjugation , we generated Ufm1 knockout mutant parasites , LdUfm1−/− and characterized the function of the Ufm1 pathway and its role in Leishmania pathogenesis [14] . Deletion of Ufm1 from L . donovani resulted in loss of β-oxidation of fatty acids due to the absence of conjugation of MTP with Ufm1 [14] . Amastigote stage specific growth defects in the Ufm1−/− studies suggested the importance of conjugation in amastigote stage thus Ufm1 C-terminal processing activity of the enzyme Ufsp might be an important target for inhibition . In this report , we demonstrated the Ufm1 processing activity of LdUfsp both biochemically and by genetic analysis and that LdUfsp is the cellular source of Ufm1 processing activity . LdUfsp−/− parasites grow normally as promastigotes but are growth defective in the amastigote stage , suggesting the importance of LdUfm1 mediated conjugation in the virulent form of the parasite . LdUfsp−/− parasites showed reduced survival in macrophages and also in Balb/C mice that are highly susceptible to Leishmania infection . Metabolic deficiency in the β-oxidation of fatty acids previously observed in LdUfm1−/− parasites coupled with the observation that no compensatory activity for the absence of Ufsp is apparent in these mutants suggests that reduced survival of Ufsp−/− parasites could be primarily due to intrinsic defects in the mutant parasites likely in the cell division by altering the Ufm1 function which has an indirect effect on the function of Ufsp protein [14] . However , enhanced parasite killing due to anti-microbial activities of macrophages such as increased NO production could also be contributing to the elimination of the Ufsp−/− mutants from the infected mice . This is consistent with our earlier observations where we showed that Ufm1 null mutant had amastigote stage specific growth defects [14] . In our previous studies , we have shown that Ufm1 , Uba5 and Ufc1 are localized to the mitochondria [13] . The presence of Ufsp in mitochondria further supports the idea that Ufm1 pathway is of mitochondrial origin in Leishmania . In conclusion , our study demonstrates for the first time role of the Ufm1 processing enzyme Ufsp in Leishmania pathogenesis that has not been described in other organisms . Studies of Ubl mediated modifications in T . cruzi , T . brucei and L . donovani show evidence that ubl mediated modifications can be parasite specific and distinct from those observed in human Ubls . Studies of Ufsp mediated modifications in parasitic protozoa are likely to lead to a deeper understanding of their contribution to parasite pathogenesis . Further , since Ufsp deficiency does not affect promastigote stages , other components of the Ufm1 conjugation system such as Uba5 and Ufc1 may also be targets for developing novel inhibitors provided their unique functions in Ufm1 conjugation are demonstrated conclusively . It is of particular importance that growth of an Ufsp knock out mutant is attenuated in the amastigote stages; as such stage specific attenuation makes it a promising candidate to be tested as a live attenuated vaccine . Moreover , specific inhibition of Ufsp activity by antileishmanial compound raises the possibility that this enzyme can be exploited as a drug target . In addition , the studies described here with Ufsp in Leishmania could help in the understanding of its function in other organisms .	Ubiquitin and ubiquitin like proteins ( Ubls ) and the enzymes that mediate the conjugation/deconjugation reactions have not been well studied in protozoan parasites despite their widely recognized importance in a broad range of cellular functions in eukaryotes . We have previously reported that Ufm1 has distinct protein targets and cellular localization in the human parasite Leishmania donovani and deletion of Ufm1 in L . donovani adversely impacts the pathogenesis suggesting that Ufm1 associated enzymes could be exploited as drug targets . Using sensitive FRET based activity probes we identified the Ufm1 processing peptidase Ufsp in L . donovani . In addition , we show that deletion of Ufsp specifically reduces the survival of amastigotes , the parasite stage that is present in the humans thus altering the pathogenesis . Studies showing inhibition of Ufsp activity by anti-leishmanial drug further suggests that Leishmania Ufsp can serve as a novel target for pharmacological intervention for this parasite that causes deadly disease .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "parasitology", "biology", "microbiology" ]	2014	Deletion of Ubiquitin Fold Modifier Protein Ufm1 Processing Peptidase Ufsp in L. donovani Abolishes Ufm1 Processing and Alters Pathogenesis
Huntington's disease ( HD ) is one of several neurodegenerative disorders caused by expansion of CAG repeats in a coding gene . Somatic CAG expansion rates in HD vary between organs , and the greatest instability is observed in the brain , correlating with neuropathology . The fundamental mechanisms of somatic CAG repeat instability are poorly understood , but locally formed secondary DNA structures generated during replication and/or repair are believed to underlie triplet repeat expansion . Recent studies in HD mice have demonstrated that mismatch repair ( MMR ) and base excision repair ( BER ) proteins are expansion inducing components in brain tissues . This study was designed to simultaneously investigate the rates and modes of expansion in different tissues of HD R6/1 mice in order to further understand the expansion mechanisms in vivo . We demonstrate continuous small expansions in most somatic tissues ( exemplified by tail ) , which bear the signature of many short , probably single-repeat expansions and contractions occurring over time . In contrast , striatum and cortex display a dramatic—and apparently irreversible—periodic expansion . Expansion profiles displaying this kind of periodicity in the expansion process have not previously been reported . These in vivo findings imply that mechanistically distinct expansion processes occur in different tissues . Huntington's disease ( HD ) is a genetically determined neurodegenerative disorder , the onset of which is known to depend upon the length of glutamine-encoding CAG-repeat sequences lying within the Huntingtin ( HTT ) gene [1] . Humans may develop the disease if they have more than 36 repeats and disease onset usually starts during mid-life . An inverse relationship has been shown between CAG repeat length and age of onset in HD [2]–[5] . Additionally , somatic instability in human cortex has recently been shown to be a good predictor of disease onset [6] . Children with 108–256 CAG repeats are reported to show disease onset from one and a half years to six years of age [7] . Trinucleotide repeat ( TNR ) instability varies between organs in a variety of neurodegenerative disorders which are caused by expansion of CAG repeats in a coding gene , with the greatest instability observed in the brain [8]–[11] . In HD , striatum tissue shows the most severe neuropathology , followed by cortex . CAG length expansion is correlated with neuropathology and probably precedes the onset of symptoms [12] . The CAG repeat length is unstable in most cell types of the brain , but neurons tend to show the greatest mutation lengths in both humans and mice [13]–[15] . Meanwhile , minimal expansion is considered to occur in many other somatic tissues . TNR sequences may form slipped strands during replication or repair , creating loops or hairpins , which protrude from the DNA duplex [16] . In the earliest model for repeat expansion the DNA polymerase forms slip-outs on the nascent strand leading to small-scale repeat expansion in repetitive sequences [17] . Loops of repeat-containing DNA are believed to cause either expansions or contractions during replication , when the slip-out occurs in the nascent or template strand , respectively [18] , [19] . Several models have been suggested to explain TNR expansion during replication , such as folding of the lagging strand template into a hairpin , stalled replication forks and the orientation of the TNR in the genome , as well as the location of the origin of replication , as shown in several experiments in bacteria , yeast and human cells ( Reviewed in [20] ) . More recently , a pertinent role of DNA repair proteins in CAG repeat expansion has been demonstrated in vivo . In particular , deletion of the mismatch repair ( MMR ) proteins , Msh2 and Msh3 [21]–[24] has been shown to abolish age-dependent somatic CAG repeat expansion in mouse models for HD . MMR has also been shown to be involved in TNR expansion in mouse models of myotonic dystrophy ( DM1 ) [25]–[27] . Furthermore , the age-dependent expansion of TNR sequences in somatic cells was shown to be modified by the base excision repair ( BER ) 8-oxoguanine DNA glycosylase ( Ogg1 ) in the R6/1 mouse model , demonstrating that there may be a link between oxidative DNA damage and TNR instability [13] . The flap endonuclease 1 ( FEN1 ) , which removes 5′-flaps during replication [28] and is involved in long-patch BER [29] is also implicated in expansion . Secondary TNR structures have been shown to inhibit FEN1 activity [30] . In addition to flap endonuclease activity , the EXO [31] and GAP activities of FEN1 have been shown to contribute to the resolution of TNR secondary structures in vitro [32] . Recently , it was shown that the stoichiometry of BER proteins , such as Ogg1 , polymerase β and FEN1 , may contribute to the tissue-selectivity of somatic HD CAG repeat expansion [33] . Nevertheless , the processes causing this expansion remain poorly understood , particularly in mammalian systems , although the formation of secondary DNA structures within the repeat sequence is thought to underlie the process [34] . Here we present evidence for two distinct modes of somatic expansion identified by the analysis of CAG repeat fragments from 103 HD R6/1 mice; a continuous slight expansion in tail , lung , heart and spleen , and a dramatic periodic expansion in striatum , and cortex , which we also compare to the expansions observed in liver . The continuous expansion process is shown here to conform to a bi-directional , forward-biased model that represents the occurrence of multiple short – tending towards unitary – CAG repeat insertions and deletions , at random moments as the mouse ages . In contrast , the dramatic expansion seen in brain tissues demonstrates a periodicity centred around seven repeats , which correlates with the stochastic insertion of stable TNR segments of consistent length . Meanwhile liver tissue shows a comparable average increase in CAG repeat length to striatum , but with a much weaker inclination to exhibit periodicity , tending more towards a continuum . This suggests either a much less controlled insertion length when compared to expansions in striatum , or that liver tissue undergoes both types of expansion simultaneously . We also present discursive models for these two expansion mechanisms . Identification of these two independent modes of expansion , in particular the tight mechanistic control implicit in the expansions within neuropathologically relevant tissues , increases our understanding of the tissue-dependent progress of HD . This brings us a step closer to inferring the in vivo mechanisms of the molecular components involved , by showing that only a limited selection of the existing models for expansion are able to explain the age-dependent CAG repeat expansions we observe . In order to understand the mechanisms underlying somatic CAG instability , 42 R6/1 HD exon 1 transgenic mice were sacrificed at either 10 or 21 weeks of age , whereupon tail , heart , lung , spleen , liver , cortex and striatum samples were taken for analysis of HD CAG repeat length . A tail biopsy at 3 weeks of age represents the reference level of CAG repeats present near birth in all tissues for each mouse [35] ( Figure S3 ) . Thus changes in the CAG composition of tissues in an individual mouse could be compared over a 7- or 18-week period . A slight expansion was observed in tail ( Figure 1A ) , whereas cortex and striatum demonstrated a dramatic and periodic expansion process , with no significant difference between genders ( Figure 1B and 1C ) . Liver demonstrated an equally rapid , but apparently more continuous expansion . Heart , lung and spleen displayed a slight expansion that was identical to tail ( Figure S11 ) . A parallel dataset from 61 hHD+/−Fen1E160D/E160D mutant mice , in which flap endonuclease activity of FEN1 is reduced to ∼20% [36] was included in the study . Reduced FEN1 endonuclease activity did not affect the rate of CAG repeat expansions measured in any tissues , implying that this mutation did not affect any role FEN1 plays in expansion . The two datasets were qualitatively and quantitatively identical with regard to the following analysis in all organs tested and were therefore combined , such that our analysis covers observations across two HD genotypes , reinforcing the ubiquity of the results . Individual fragments from tail fit well to a normal distribution and are thus described by the mean ( μ ) and standard deviation ( σ ) of the curves fitted to raw data ( Figure 2A , 2B; see Methods ) . Expansion within the population of 59 mice ( 10 week old mice excluded ) is clearly shown ( Figure 2C ) by the relative difference in μ of the 21-week and 3-week groups . The median expansion found in 59 tails of 21-week mice was 1 . 97 CAG triplets ( Figure 2E ) . In addition , σ of individual tail data sets is shown to increase from 1 . 98 triplets at 3-weeks to 2 . 87 triplets at 21-weeks ( Figure 2D ) . The increasing σ is not an artefact of PCR errors , as is demonstrated in Figure 3A . Having made these observations , it is necessary to consider them in the context of potential models for expansion , in order to fully investigate their implications and attempt to parameterize the processes involved . A continuous increase in both mean and standard deviation can be generally accounted for by multiple stochastic unitary ( single CAG-repeats ) extension and contraction events on the CAG tracts within the sample ( Figure 3B ) . A full discussion of the potentially applicable models and considerations is presented as supporting information ( Figure S8 , Text S1 , and Videos S1 , S2 , S3 , S4 , S5 , S6 , S7 , S8 ) and we confine ourselves here to a simple application of the most probable hypothesis , yielding upper estimates for expansion and contraction rates . Assigning probabilities to non-simultaneous unitary expansions and contractions , and respectively , allows the measured temporal change in the mean ( ) and variance ( ) of tail samples to be defined by ( 1 ) and ( 2 ) . ( 1 ) ( 2 ) Using a time interval ( ) of one day , the measured expansion of 1 . 97 repeats and the concomitant increase in average standard deviation from 1 . 98 to 2 . 87 , the values of and are calculated to be 0 . 026 and 0 . 010 respectively . This gives a maximum expectation of ∼0 . 036 ( pe+pc ) events per repeat tract per day . In contrast to fragment data from tail , heart , lung and spleen , raw data from 10- and 21-week cortex and striatum tissue show a peak retained at the 3-week repeat level alongside an age-dependent number of periodically spaced subsequent peaks ( Figure 4A , Figures S1 , S2 , and S10 ) , to which a series of normal distributions were fitted ( see Methods ) . Knowing that the relative areas of overlapping distributions define the proportion of each mean CAG repeat length present ( demonstrated with mixed samples and serial dilutions of a 21-week striatum sample in Figures S4 and S5 ) , we infer – on account of the regularity of the intervals between neighbouring peaks at both 10- and 21-weeks of age – that expansion involves a proportion of the brain tissue undergoing insertions of consistent-length CAG repeat fragments over time . If expansion events inserted CAG fragments of uncontrolled length , the clarity of subsequent peaks would be lost . Likewise if different cell-types within one sample expanded at different rates , one would expect a continuum of peak separations in the collected accumulated data from many mice , and would have little reason to expect a consistent periodicity between peaks at 10-weeks and 21-weeks . This argues for the stochastic step-wise insertion of CAG fragments with an average length μb−μa ( Figure 4A ( 10wk and 21wk ) and 4B ) by a mechanism that may recur within the same cell . To measure the periodicity seen in brain tissues , we compiled histograms of all intervals between identifiable peaks , binned by size , from the individual cortex and striatum samples of mice aged 10- and 21-weeks ( Figure 4C ) . The measured intervals are clearly shown to be distributed around a peak at 7 CAG repeats , with a mean length of 7 . 14 ( σ = 1 . 78 with a cut-off for doublet measurements set at intervals ≥12 ) . The median interval of 7 also confirms that this expansion process is centred around the insertion of 7-repeat fragments into the CAG tract , although the width of the starting distribution indicates insertion of 5 to 9 repeat-fragments ( see Figure S9 and Videos S1 , S2 , S3 , S4 , S5 , S6 , S7 , S8 for further discussion ) . The relative sizes of peaks in 21-week striatum ( see Methods ) imply that on average , a total of ∼10 , 000 7-repeat insertions occurred in each dramatically expanding striatal sample ( on our timescale probably mainly within neurons; however , glial cells also undergo expansion and not all neuronal cells are guaranteed to show expansion [14] , [15] , giving a periodic expansion probability estimate of 0 . 018 events per repeat tract per day . The fact that efficiency of PCR amplification of longer CAG tracts is reduced , may result in some measure of underestimation with this value . This is ∼70% of the estimated probability for unitary expansion events in tail , however the 7-repeat average insertion size renders the resulting expansion more dramatic in striatum . In contrast , liver data while showing comparable levels of average expansion to brain tissues , lacks a clear signature of periodicity ( see Figure 1D ) tending towards bimodality , with a more continuously located second peak . This would imply a much less controlled insertion length during the expansion process , or possibly a combination of expansion mechanisms . Having shown data and analysis to define these two distinct modes of expansion , we place our findings into the context of existing literature , in order to develop reasonable hypothetical models for these two types of expansion . The continuous expansion we observe shows a progressive increase in CAG tract length , which is comparable with the expansions observed in fibroblasts derived from an adult HD mouse [37] . In the debate regarding the relative roles of replication and repair in TNR instability , these results present an interesting question , since the continuous expansion process occurs in organs containing dividing cells . A process occurring with the previously calculated expectation rate of ∼0 . 036 events per repeat tract per day on the ∼360 nucleotide CAG segment of the 2 . 5 gigabase mouse genome , would correlate to ∼250 , 000 genome-wide events per cell per day . This is well above the upper estimate for daily DNA damage events , making it unlikely that these are the sole initiator of expansions in tail tissue . Several replication-based models for TNR expansion in dividing cells [18] , [19] have been proposed . The potential for replication to be entirely responsible for this expansion is considered in detail elsewhere ( Figures S7 , S8 , and Text S1 ) and is considered to be unlikely , particularly in light of the fact that lymphoma tissues ( with necessarily higher replication rates ) isolated from several mice showed no increase in TNR instability ( Figure S7 ) . Therefore we infer that other mechanisms must also prevail , and propose that slipped strand structures [16] , [38] generated by out of register rehybridization of CAG repeats [27] during transcription , or genome maintenance , may spontaneously form unstable loops or cruciforms which may subsequently stabilize by migrating apart . Similar small loops can also be formed by polymerase slippage during replication [17] . We propose that two separate pathways may repair these loops , leading to single repeat expansions or contractions ( Figure S8 ) . Further research is needed to resolve the specific origin of this mode of expansion . A hypothesis for periodic expansion is also presented ( Figure S9 ) . The previously calculated periodic expansion probability of 0 . 018 events per repeat tract per day would correlate to ∼100 , 000 genome-wide events per cell per day . This lies in the vicinity of upper estimates for accumulated oxidative DNA damage [39] . It is therefore possible that DNA damage contributes as catalyst for periodic expansions in brain tissues . However , oxidative damage is not sufficient to trigger somatic instability [33] . Of particular interest here , are the potential molecular components that could repeatedly generate a regularly sized repeat insertion that is dominantly seven repeats in length . We have therefore chosen to briefly review the relevant literature in search of further insight . Theory suggests that CAG flaps ranging from 4 to 16 triplet repeats in length can form thermodynamically unstable hairpin structures with an even number of repeats [40] . However , under physiological conditions , 6 triplet repeats have been shown to form hairpins irreversibly [41] , [42] . This implies that a progressively generated triplet repeat flap can stabilise into a 6-repeat hairpin at the free end that could be cleaved by FEN1 causing no net expansion [43] ( Figure S9 ) . However , the presence of metastable intermediates during flap generation may allow the flap length to increase beyond 6 repeats before a stable structure is formed , thereby producing a hairpin with an overhanging CAG at the 5′-end . This free CAG repeat can hybridise back to the DNA duplex , and it has been shown by Liu et al . [31] that such hybridisation facilitates bubble formation followed by ligation and expansion . In a few instances , two or even three free CAG repeats in the 5′-end of the hairpin would produce a periodicity of eight or nine CAG repeat steps . In addition , the size of the hairpin could vary with a few repeats . However , this occurs more rarely , as observed in Figure 4C . After gap filling and ligation , a loop of excess CAG repeats in one strand would be produced ( Figure S9 ) that can be recognized by the MMR complex Msh2-Msh3 with high affinity [44] . This binding might further stabilize the CAG loop and additionally explain why Msh2-Msh3 is necessary for large expansions to occur in striatum [21]–[24] , although the role of MMR in causing TNR expansion is not understood . Moreover , Msh2-Msh3 function ceases due to impaired ATPase activity on loops exceeding 16 nucleotides in lengths [45] - probably due to the presence of A•A mispaired bases in the loop [22] - which may explain why the MMR system fails to repair longer extraneous CAG loops . Subsequently , a nick generated on the strand opposite to the loop could result in faulty repair of the CTG strand along the CAG slip-out , causing expansion by a repair process independent of MMR [46] ( Figure S9 ) . Thus , oxidative damage on the CTG strand could result in base excision repair ( BER ) and Ogg1 moderated expansion , which is also in concordance with the proposed ‘toxic oxidation cycle’ by Kovtun et al . , 2007 [13] . We therefore propose that a coincidental cooperation can occur between MMR and BER in cases where a long CAG flap is able to stabilize itself , which can form the basis of the consistent periodic insertion we observe . It should be pointed out that it is possible that other stabilized structures , such as loop-outs or a stabilizing interaction with one of the many proteins and complexes that are in contact with the DNA may also serve to cause the observed periodicity . The majority of the potentially applicable models are covered in the literature [13] , [24] , [27] , [34] . Further work is necessary to resolve this completely . Cell proliferation in neurons and glial cells has been observed in the subependymal layer adjacent to the caudate nucleus in human HD brains [47] . However , polymerase slippage usually forms small expansions [17] , and the repetitive uniformity of the periodic expansion makes it unlikely that polymerase slippage is responsible for the dramatic expansions seen in cortex and striatum . Furthermore , the lack of periodically spaced peaks containing fewer repeats than the mice were born with – as would be expected from a bi-directional process - means that 7-repeat hairpin-based contraction events occur either at a negligible rate in comparison to expansions , or do not occur at all . During the 18 week period , tail-type expansion events are not evident in striatum since they would obfuscate latter periodic peaks ( Figure S6 ) . Therefore , the two expansion mechanisms seem to be either entirely independent – not occurring simultaneously in the same cell – or that if they do share common elements , they progress along mutually exclusive pathways . However , it is important to notice that both expansion mechanisms must be dependent on proteins from the MMR complex , since expansions are eliminated in all tissues in either Msh2 or Msh3 nulls in HD [21]–[24] , and also in DM1 [25] , [26] . To date , we have not managed to define the individual cell types that are specific to these expansion mechanisms . It will be of great interest to compare the expansions observed in animal models to those in cultured HD mouse fibroblasts , as a way of identifying the cell specificity of these modes of expansion . The liver may be particularly interesting in this regard , since this tissue potentially exhibits a mix of both types of expansion mechanisms . This could be attributed to different modes of expansion occurring in different cell types . Indeed , instability of the DM1 CTG•CAG repeat is known to occur in liver hepatocytes with high DNA ploidy [48] . The question also arises as to why FEN1 did not influence CAG repeat expansion in the organs tested . One might expect a difference , since a recent in vitro study has shown that FEN1 , together with long-patch BER of long repeat sequences by polymerase β , promotes expansion by facilitating the ligation of hairpins formed by strand slippage [49] . However , FEN1 flap activity has shown to be circumstantial , with much lower activity in the striatum than in the cerebellum of R6/1 mice [33] . In yeast the capture of flap structures by FEN1 , rather than the endonuclease activity , is the most important function of FEN1 in preventing TNR expansion [50] . EXO- and GAP activity of FEN1 are also involved in in vitro triplet repeat expansion in yeast [32] , and these activities are probably not influenced by the Fen1E160D/E160D mutation . Therefore , it seems that the 20% endonuclease activity [36] of the Fen1E160D/E160D mutation does not affect the rate of CAG repeat expansions . In concordance with our finding Fen1 did not control instability of ( CTG ) n* ( CAG ) n repeats in a knock-in mouse model for DM1 [51] . It is worth considering briefly why this periodicity has not been described before , since there are numerous potential reasons . One possibility could be that the mice used in the present study exhibit more instability due to environmental factors [52] or genetic background [53] . Perhaps the periodic signal becomes more disperse in older mice used elsewhere; degrading the quality of the data , and that the age-range , as well as the relatively long starting CAG lengths , of our samples is optimal for observing this periodicity . Another possibility is simply that later versions of the GeneMapper system are more sensitive , in comparison to the GeneScan method applied in some older studies , allowing us to see more detail . While periodicity has not featured in other studies of similar tissues and disease models , it is difficult to state unequivocally that it was unobservable in their data . The small volume of data presented in articles , uncertainty over the precise PCR conditions used and the simple fact that periodicity was not the focus of these investigations can be sufficient cause for this phenomenon to have been previously overlooked . There is some variability among replicate striatum samples as shown in Figure S10 . This could be explained by sampling error or polymerase slippage in early PCR cycles . Sampling error is however unlikely to be the reason behind the periodicity as explained in Figure S12 and Text S2 . So far , we have only studied periodicity in the R6/1 mouse model and without specific studies of other HD CAG mice models the generality of our data is unknown . Yet , the R6/1 transgenic mouse is a widely accepted and commonly used model for human HD that exhibits a progressive neurological phenotype that exhibits many of the features of HD [54] . HD CAG repeat instability has shown to be similar in humans and mice , with the longest expansion lengths occurring in striatum , followed by cortex , and little expansion in cerebellum and most other tissues [8] , [14] , [35] . In addition , the HD CAG repeat length appears to be expand most in neuronal cells rather than glial cells in both species [14] , [15] . Due to the long starting CAG repeat length in the transgenic mouse , the model may be most relevant as a model for juvenile HD . Given the stated similarities , there are grounds to suspect that the mechanisms of expansion are identical in mice and humans , only occurring at an accelerated pace in the mouse on account of the long repeat tract . In this case , the mouse model would function as a good model for human cases with mid-life age of onset as well . However , periodicity has not previously been reported in HD patients . A possible explanation could be that the repeat length in the R6/1 mouse is longer than the repeat length that has been analyzed in any HD human tissue with regard to somatic instability . The genomic localization of the randomly integrated HD gene fragment in R6/1 mice might modulate the CAG stability . Furthermore , CAG instability in HD patient brain cannot be analyzed at an early time-point that would allow for a direct comparison to the R6/1 data . Importantly , CAG repeat expansion in human cortex is associated with an earlier age of disease onset , in addition to the role of the constitutional CAG repeat length [6] . This implies that there are disease modifiers that influence somatic instability , and conversely , factors that determine somatic instability which may modify disease pathogenesis . It is therefore critical to understand the mechanism of the different expansion processes and the factors involved . To summarize , we present two different mechanisms of somatic CAG repeat expansion; a continuous bi-directional expansion in tail , lung , heart and spleen tissues , and a dramatic periodic expansion centred around 7-repeat insertions in striatum and cortex . Further experiments are needed to determine whether the 7-repeat step-size is independent of CAG tract length and species and it remains to be shown whether these two models can explain the expansions observed in other brain tissues and organs , as well as in humans . Nevertheless , these results provide significant new insights into in vivo expansion mechanisms , which may also be relevant to other triplet repeat disorders . All animal experiments have been approved by the local and national animal - and are carried out according to the regulation by FELASA ( Federation of European Laboratory Animal Science Associations ) . B6CBA-Tg ( HDexon1 ) 61pb/J mice of the R6/1 line [54] with ∼119 CAG repeats in exon 1 of the HTT gene , were purchased from The Jackson Laboratories and either interbred or crossed with C57BL/6J Fen1E160D/E160D mice [36] . The mice were fed Rat and Mouse No . 3 breeding diet ( Special Diet Services ) and tap water ad libitum . At 3 weeks of age the mice were anesthetized by i . p . injection of a combination of Midazolam ( Dormicum “Roche” ) and Fentanyl/Fluanisone ( Hypnorm ) solutions , and tail biopsies were taken . DNA from tail biopsies of the first two generations were lysed as described [36] and DNA isolated using standard NaCl precipitation or phenol/chloroform procedure . At 10 and 21 weeks of age the mice were sacrificed by cervical dislocation . The organs were harvested , frozen on dry ice and stored at −70°C . During dissection of striatum we lost 9 samples . DNA from all tissues and tails from the F2 and F3 generation was isolated according to the DNeasy Blood & Tissue kit ( Qiagen GmbH , Germany ) . DNA from tail biopsies of 3-week old mice were used for genotyping . HD mice were genotyped with forward 5′-cggctgaggcagcagcggctgt-3′ and reverse 5′-gcagcagcagcagcaacagccgccaccgcc-3′ PCR primers [54] according to the Advantage GC 2 PCR Kit & Polymerase Mix ( Clontech , CA ) . The Fen1+/+ and/or Fen1E160D/E160D knock-in allele was genotyped as described [36] . CAG repeats were sized by PCR with primers 5′-FAM-atgaaggccttcgagtccctcaagtccttc-3′ and 5′-ggcggctgaggaagctgagga-3′ according to [54] with slight modifications . Approximately 75ng of genomic DNA ( this approximates to DNA extracts from ∼10 , 000 cells ) was amplified with AmpliTaq Gold DNA polymerase with PCR Buffer II , 1 . 25 mM MgCl2 ( Applied Biosystems , CA ) , and 2 . 5 mM dNTPs ( GE Healthcare ) . The cycling conditions were 94°C for 10 min , 35 cycles of 94°C for 30 sec , 64°C for 30 sec , 72°C for 2 min , and a final extension at 72°C for 10 min . The FAM-labeled PCR products were mixed with GeneScan - 600 LIZ Size Standard and HiDi Formamide ( Applied Biosystems ) and run on an ABI 3730 Genetic Analyzer ( Applied Biosystems ) . Sizing of the PCR fragments was performed by using the GeneMapper Software Version 3 . 7 ( Applied Biosystems ) . All raw data was processed through a masked Nelder-Mead simplex fitting method , optimising free parameters of standard deviation , mean and amplitude to fit consecutive normal distributions sufficient to account for ≥98% of the total area of the raw data set . In the case of tail data , only a single normal distribution was required . These optimised parameters were returned as the means ( μ ) , and standard deviations ( σ ) , which were used to define the TNR lengths present in each data set . CAG repeat tracts were flanked by sequences 86 bp in length as verified by sequencing . Thus , the mean number of CAG triplets ( μt ) present in a fragment analysis sample with a measured mean ( μm ) is defined by μt = ( μm−86 ) /3 . When analysing the periodicity present in striatum and cortex data , standard frequency analysis methods are not suitable , therefore our peak fitting method was used to fit consecutive normal distributions to the raw data ( Figure 4A ) . We were unable to perform fitting analysis on 25 striatum samples due to the quality of the PCR product . The means ( μa , μb etc ) and relative areas ( Aa , Ab etc ) of each peak were calculated ( Figure 4B ) . The intervals between neighbouring means ( S1 , S2 etc ) were also recorded at both 10 and 21 weeks . The area ( A ) of each peak was used to estimate the number of cells containing triplet repeats that had expanded with a step size defined by the separation interval ( S ) . The average area of the first peak in all 21-week data ( Aa from Figure 4B ) was used to estimate the proportion of non-expanding cells in the striatum as 54% ( σ = 15 . 9 ) , implying that approximately 45% of each striatum sample underwent periodic expansions within the first 21 weeks . By comparison , the proportion of dramatic expanding tissue observed in cortex samples was ∼20% . Previous work has shown that the dramatic expansion observed in the striatum of adult mouse brain tissue largely occur in the neuronal cells [15] , [14] , although slower expansion can be observed in glial cells and we used this to approximate the level of expansion in neuronal cells in combination with an estimate of the average number of expansion events that were measured in 21-week mice .	Huntington's disease ( HD ) is a genetically determined neurodegenerative disorder identified by the presence of a mutation for a long series of CAG repeats ( >36 repeats ) in the Huntingtin ( HTT ) gene . Longer repeat sequences cause disease onset at a younger age . The mutation encodes an expanded glutamine tract within the huntingtin protein . This enlarged polyglutamine fragment in the protein leads to the formation of the huntingtin aggregates that are observed in HD brains . The stretch of CAG repeats expands with age in affected brain areas , increasing the length of the polyglutamine tract , and is believed to amplify the effect of the disease . Several HD mouse models display phenotypes relevant to the human disease . We have investigated the rate and modes of expansion in striatum , cortex , and tail in transgenic R6/1 mice . Tail was included as a stable tissue , however we observed a small continuous expansion of CAG repeats in tail tissues . In brain tissues , we identified a periodic expansion process consisting of predominantly seven repeat steps . Our findings point towards a very controlled molecular mechanism as the cause of expansion in the most severely affected tissues , which may provide useful targets that can be used to inhibit disease development .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "genetics", "and", "genomics/disease", "models", "genetics", "and", "genomics/genetics", "of", "disease", "biochemistry/replication", "and", "repair", "mathematics/statistics", "genetics", "and", "genomics", "molecular", "biology/dna", "repair" ]	2010	Continuous and Periodic Expansion of CAG Repeats in Huntington's Disease R6/1 Mice
Autozygosity occurs when two chromosomal segments that are identical from a common ancestor are inherited from each parent . This occurs at high rates in the offspring of mates who are closely related ( inbreeding ) , but also occurs at lower levels among the offspring of distantly related mates . Here , we use runs of homozygosity in genome-wide SNP data to estimate the proportion of the autosome that exists in autozygous tracts in 9 , 388 cases with schizophrenia and 12 , 456 controls . We estimate that the odds of schizophrenia increase by ∼17% for every 1% increase in genome-wide autozygosity . This association is not due to one or a few regions , but results from many autozygous segments spread throughout the genome , and is consistent with a role for multiple recessive or partially recessive alleles in the etiology of schizophrenia . Such a bias towards recessivity suggests that alleles that increase the risk of schizophrenia have been selected against over evolutionary time . Schizophrenia is a highly ( . 70– . 80 ) heritable [1] neurodevelopmental disorder that has a lifetime prevalence of ∼0 . 4% [2] . As with most complex disorders , the specific genetic variants that account for a majority of the heritability of schizophrenia remain to be discovered . Two primary factors may explain the difficulty in identifying risk variants . First , the results of genome-wide association studies ( GWAS ) make it clear that a very large number of genes contribute to schizophrenia risk , and the overall population risk attributable to any one risk variant must be small [3] . Second , although common causal variants almost certainly play an important role in the genetic etiology of schizophrenia [4] , [5] , it is likely that the frequency distribution of schizophrenia risk alleles is biased towards the rare end of the spectrum [5] . Both of these factors are consistent with selection keeping schizophrenia risk alleles with the largest effects rare , such that no single allele can contribute much to population risk . If schizophrenia risk alleles have been selected against across evolutionary time ( have been under “purifying” selection ) , another prediction is that schizophrenia risk alleles will be biased towards being recessive . This bias , called directional dominance , occurs in traits subject to purifying selection because selection more efficiently purges the additive and dominant alleles with the strongest effects , leaving the remaining pool of segregating alleles more recessive than otherwise expected [6] . Directional dominance has traditionally been inferred from observations of inbreeding depression , the tendency for offspring of close genetic relatives to have higher rates of congenital disorders and lower fitness [7] . Fitness traits such as survival , reproduction , resistance to disease , and predator avoidance tend to show more inbreeding depression than traits under less intense selection [8] . Interestingly , there are numerous reports of inbreeding effects on human complex traits such as heart disease [9] , hypertension [10] , osteoporosis [11] , cancer [12] , and IQ [13] , [14] . Studies that have investigated inbreeding effects on schizophrenia using pedigree data suggest that close ( e . g . , cousin-cousin ) inbreeding is a risk factor [15] , [16] , [17] , [18] , [19] , [20] , although three studies have failed to find the predicted effect [21] , [22] , [23] . However , close inbreeding cannot be a major contributor to schizophrenia risk in industrialized countries given its rarity ( <1% of marriages ) [24] and the modest increase in the odds of schizophrenia among highly inbred offspring ( ∼2- to 5-fold ) [15] , [16] , [17] , [18] , [19] . Nevertheless , inbreeding is a matter of degree; when distant relatives are considered , everyone is inbred to some degree . It is likely that the parents of the vast majority of people alive today share a common ancestor within ∼15 generations [25] . Although such “distant” inbreeding would be prohibitively difficult to detect from pedigrees , it can leave signals in the genome that are detectable using genome-wide single nucleotide polymorphism ( SNP ) data . The inbreeding coefficient of an individual , F , is defined as the probability that two randomly chosen alleles at a homologous locus within an individual are identical by descent ( IBD , identical because they are inherited from a common ancestor ) [26] . Homozygosity arising from the inheritance of two IBD genomic segments is termed autozygosity . Most estimates of F assume that marker data are independent , and provide an aggregate measure of homozygosity at measured variants across the genome [27] . Recently , however , several investigators have used runs of homozygosity ( ROHs; long stretches of homozygous SNPs ) to infer autozygososity , and have investigated whether the proportion of the genome that exists in such ROHs , Froh , predicts complex traits [28] , [29] , [30] , [31] , [32] , [33] , [34] , [35] . Of several alternative estimates of F , including F estimated by treating markers independently and F estimated from pedigree information , Keller , Visscher , and Goddard [25] recently concluded that Froh is optimal for inferring the degree of genome-wide autozygosity and for detecting inbreeding effects . However , given the small variation in genome-wide Froh in unselected samples ( e . g . , SD ∼ . 005 ) , large sample sizes ( e . g . , >12 , 000 ) are necessary to detect inbreeding depression for likely effect sizes in samples not selected for recent inbreeding [25] . Studies investigating the effects of Froh on human complex traits with samples sizes <3 , 000 and that failed to find significant inbreeding effects [28] , [33] , [34] , [35] , [36] are likely to have been underpowered . That said , the only study of Froh in schizophrenia [29] found a very large inbreeding effect , but the effect was observed in a small sample ( n = 322 ) and was significant only for ROHs caused by common haplotypes . The present study uses imputed SNP data from 17 schizophrenia case-control datasets ( total N = 21 , 844 ) that are part of the Psychiatric GWAS Consortium ( PGC ) [3] , [37] to investigate whether Froh is associated with higher risk of schizophrenia . We also use an ROH mapping approach to investigate whether specific areas of the genome are predictive of case-control status when autozygous . This study represents the largest investigation to date on the potential consequences of autozygosity as estimated using Froh , and may help elucidate the genetic architecture and natural history of schizophrenia . We regressed case-control status on Froh separately in each of the 17 datasets using logistic regression , controlling for potential confounding factors such as population stratification and SNP quality metrics ( see Methods ) . Figure 2 shows the estimated change in odds of schizophrenia for every 1% increase in Froh and the 95% confidence intervals from these 17 logistic regression equations , and Figure S2 shows the same results from an analysis conducted on the raw ( non-imputed ) SNP data . It should be noted that confidence intervals are symmetric on the log odds scale but asymmetric on the odds ratio scale shown in Figure 2 and Figure S2 . As indicated by the confidence intervals , there was a great deal of variability in the estimates of the Froh-schizophrenia association , and none of these 17 odds ratios significantly differed from one . Nevertheless , 13 of the odds ratios were greater than one ( i . e . , consistent with autozygosity being a schizophrenia risk factor ) while 4 were less than one , a result inconsistent with chance ( exact binomial test , p = 0 . 025 ) . More formally , using a mixed linear effects logistic regression model that treated dataset as a random factor ( which also controlled for SNP platform because dataset was nested within each platform ) , the overall association between schizophrenia and Froh in the combined sample was highly significant ( β = 16 . 1 , z = 3 . 44 , p = 0 . 0006 in the imputed data , and β = 17 . 98 , z = 3 . 89 , p = 0 . 0001 in the raw data ) . A slope of Froh on schizophrenia of 16 . 1 is interpreted as saying that for every 0 . 01 increase in Froh , the odds of schizophrenia are multiplied by , or increased by 17% . Several secondary analyses were undertaken to explore the robustness and generality of the Froh-schizophrenia association . There was no evidence that the Froh-schizophrenia association differed significantly between datasets ( = 0 . 253 , p = 0 . 88 ) , and the association remained highly significant in 17 models that removed one dataset at a time . To understand if this association was sensitive to the covariates included in the model , we ran additional models that controlled for no covariates , various combinations of covariates , and dataset-by-covariate interactions . In all of these models , the association between Froh and schizophrenia remained significant . We also found that our conclusions were insensitive to the SNP threshold used to define ROHs; the association between Froh and schizophrenia remained relatively unchanged and significant for all SNP thresholds of ≥40 consecutive SNPs in both the imputed ( Figure 3 ) and raw ( Figure S3 ) data . Finally , both common ROHs ( β = 28 . 5 , z = 2 . 51 , p = . 012 ) , which arose from haplotypes that were observed often in the data , and uncommon ROHs ( β = 20 . 4 , z = 3 . 29 , p = . 001 ) were predictive of case-control status ( see Methods ) . Copy number variant deletions can create apparent ROHs in SNP data . We could not systematically catalog the overlap between deletions and ROHs in the full dataset because deletion information is not available on the entire sample . However , Levinson and colleagues [41] identified 501 , 890 deletions ( using their “broad” criteria ) in the MGS2 dataset ( n = 5 , 163 ) , comprising about one-fourth of the total sample used here . The median length of a deletion in the MGS2 dataset was ∼10 kb , whereas the median length of a ROH was ∼2 , 000 kb , suggesting that very few deletions would be long enough to qualify as ROHs . Consistent with this expectation , we found that only 10 of 6 , 480 ROHs in the MGS2 dataset were possible deletions using the algorithm described by McQuillan et al . [31] , which called a ROH a “possible deletion” if its total length was <500 kb after removing deletion regions from ROHs . The percentage of ROHs thus classified ( 0 . 15% ) was similar to the percentage ( 0 . 30% ) reported by McQuillan et al . [31] . This percentage is too small to have a meaningful impact on our results , because when we removed a larger percentage of ROHs that were identified as being the largest schizophrenia risk factors ( see below ) , the Froh-schizophrenia association remained highly significant . We conclude that ROH results reported above are due to autozygosity rather than hemizygosity . A reverse-causation explanation of the Froh-schizophrenia association is possible: people who have a higher “load” of schizophrenia risk alleles ( and who transmit this risk to offspring ) may be more likely to mate with a relative . This counter-explanation to the causal interpretation of the Froh-schizophrenia relationship is less likely if the relationship holds not only for close inbreeding , but also for autozygosity caused by distant and almost certainly unintended inbreeding ( arising from common ancestors who lived many generations ago ) . One way to investigate this issue is to remove positive outliers on Froh and reassess the Froh-schizophrenia relationship . We reran models after dropping a ) two individuals with Froh>0 . 125 , the approximate equivalent of half-sibling inbreeding ( β = 15 . 57 , 95% CI ( β ) = [25 . 0 , 6 . 14] , z = 3 . 24 , p = 0 . 001 ) ; b ) 15 individuals with Froh>0 . 0625 , the approximate equivalent of cousin-cousin inbreeding ( β = 15 . 13 , 95% CI ( β ) = [26 . 1 , 4 . 25] , z = 2 . 73 , p = 0 . 006 ) ; c ) 56 individuals with Froh>0 . 03125 , the approximate equivalent of half-cousin inbreeding ( β = 8 . 43 , 95% CI ( β ) = [21 . 43 , −4 . 55] , z = 1 . 27 , p = 0 . 20 ) ; d ) 942 individuals with Froh> . 005 , consistent with elevated levels of distant inbreeding ( β = 5 . 17 , 95% CI ( β ) = [34 . 84 , −24 . 50] , z = 0 . 34 , p = . 73 ) ; and e ) 6 , 101 individuals with Froh scores above the mean level of Froh ( β = 66 . 91 , 95% CI ( β ) = [139 . 2 , −5 . 4] , z = 1 . 81 , p = . 07 ) . To test whether the change in significance after dropping outliers was due to the Froh-schizophrenia association being stronger for individuals with high levels of autozygosity , we included a quadratic term ( Froh2 ) in the regression model . In contrast to the highly significant linear term of Froh , the quadratic term of Froh was non-significant ( p = . 09 ) , suggesting that the effect of autozygosity is linear across the range of Froh observed here . The simple approach—dropping outliers—to distinguishing the effects of distant versus close inbreeding is problematic for two reasons . First , Froh is naturally extremely right-skewed ( Figure 1 and Figure S1 ) , even in large , simulated populations where close inbreeding is disallowed [25] , and so dropping even a small number of outliers greatly reduces the variation in Froh , decreases the statistical power to detect an association , and degrades the precision of point estimates . Indeed , there is no evidence that the schizophrenia-Froh association changes as outliers are removed , because the original point estimate ( β = 16 . 1 ) is contained within every confidence interval above . Thus , the results from dropping outliers demonstrate that the Froh-schizophrenia relationship is not driven by a few highly inbred individuals , but do not allow us to distinguish the effects of distant vs . close inbreeding . Second , individuals with high Froh can arise by chance from the accumulation of many paths of distant inbreeding [25] , and are not necessarily the products of close inbreeding . For example , the distribution of lengths of observed ROHs among individuals with Froh>0 . 0625 is more consistent with inbreeding from common ancestors living ∼6 generations ago than with first cousin inbreeding ( Figure 4 ) . An alternative and more robust approach for assessing the relative importance of distant versus close inbreeding is to compare the effects of short versus long ROHs . We defined Froh<5 Mb as the proportion of the autosome in ROHs of length <5 Mb and Froh>5 Mb as the converse , with 5 Mb chosen as the threshold because the variances of Froh<5 Mb and Froh>5 Mb were equal . An autozygous segment spanning <5 Mb should originate from a common ancestor ≥10 generations ago on average [41] . The effect of Froh<5 Mb ( β = 27 . 6 , z = 2 . 23 , p = 0 . 026 ) was similar to the effect of Froh>5 Mb ( β = 24 . 3 , z = 2 . 01 , p = 0 . 044 ) , consistent with the hypothesis that autozygosity arising from distant inbreeding is about as much of a schizophrenia risk factor as autozygosity arising from more recent common ancestors . The top of Figure 5 shows the −log10 p-values for the 5 , 742 logistic regressions predicting case-control status from ROHs at each 500 kb bin along the autosome . No regions reached genome-wide significance although two ( 1p13 . 2 and 3p24 . 1 ) exceeded the “suggestive significance” threshold . Table 2 shows the twelve genes located in these two regions along with their potential functional significances . Neither region has been previously implicated in linkage analyses [42] , copy number variant analyses [43] , or GWAS meta-analyses [3] of schizophrenia . After recalculating Froh with the two suggestively significant regions removed , results of the burden analysis remained essentially unchanged , showing that these regions have only a minor influence on the overall Froh-schizophrenia association and suggesting that the effect of autozygosity is diffused across the genome . The bottom of Figure 5 shows the frequencies of ROHs occurring at each 500 kb bin across the autosome . With one exception , less than 1 . 5% of the sample had an ROH at each region . The exception occurs in the Major Histocompatibility Complex region in 6p21 . 3 , where 15 . 5% of the sample had an ROH . This high number of ROHs is explained by the low recombination and long , common , geographically-specific haplotypes that occur here [44] , [45] . Inbreeding has had a central place in population genetics since its inception , but until recently , the effects of inbreeding could only be investigated from careful analysis of pedigrees and only for close inbreeding . SNP data allows investigation into the effects of potentially very distant inbreeding in non-selected samples , and allows insight into where the signal comes from in the genome . However , unless samples are specifically selected based on inbreeding , very large samples are required to reliably detect effects of autozygosity due to the low variation between individuals in their levels of autozygosity . The present investigation used SNP data from a large sample to conclude that autozygosity is a risk factor for schizophrenia . If the relationship between Froh and schizophrenia is due to directional dominance , such that schizophrenia risk alleles are more recessive than otherwise expected , this suggests that alleles that increase the risk of schizophrenia have been under negative selection ancestrally . Full methods are given elsewhere [3] . Briefly , 9 , 388 schizophrenia cases and 12 , 456 controls were collected with institutional review board approval from 17 sites in 11 countries ( Australia , Bulgaria , Denmark , Germany , Ireland , Netherlands , Norway , Portugal , Sweden , United Kingdom , and Unites States of America; see Table 1 ) . As is typical in the field , individuals with schizophrenia or schizoaffective disorder were included as cases [49] [50] . The quality of phenotypic data was verified by a systematic review of data collection methods to ensure consistency between sites . The initial set of samples and SNPs passed common GWAS QC procedures [3] . In particular , we removed a ) one individual from any pair of individuals who were related with pi-hat >0 . 2 , b ) individuals with non-European ancestry as determined by principal components analysis; c ) samples with SNP missingness >0 . 02; or d ) samples with genome-wide heterozygosities >6 standard deviations above the mean . SNPs were excluded if they a ) deviated from Hardy-Weinberg equilibrium at p<1×10−6; b ) had missingness >0 . 02; c ) showed a minor allele frequency difference to HapMap CEU>0 . 15; or d ) had a missingness difference between cases and controls >0 . 02 . On average the QC processes excluded 15 individuals ( 0–100 ) and 38K SNPs ( 5K–160K ) per dataset . The number of SNPs per dataset after QC varied between 250K and 680K ( Table 1 ) . Six different SNP platforms ( Affymetrix 500K , 5 . 0 , and 6 . 0 chips along with the Illumina 317K , 550K , and 650K chips; Table 1 ) were used across the 17 datasets . Differences across platforms in SNP densities , frequency distributions , LD patterns , and missingness led to variation in ROH statistics across datasets . For example , the DK dataset contains 280K SNPs after LD pruning ( 1 SNP per 11 kb ) whereas the UCL datset contains 156K SNPs after LD pruning ( 1 SNP per 21 kb ) . ROHs therefore would have to be about twice as long in the UCL dataset to qualify , which induces artifactual noise in ROH statistics due to platform effects . This issue is not circumvented by using an ROH threshold based on length rather than number of SNPs; in this case , half as many homozygous SNPs in a row would be required to call an ROH in the less dense dataset . In both cases , the type-I and type-II error rates of autozygosity detection differ systematically between datasets . To overcome these issues , we imputed dosages for 1 , 252 , 901 autosomal SNPs in each dataset using BEAGLE [38] and HapMap3 as the reference panel [3] . We converted imputation dosages to best-guess ( highest posterior probability ) SNP calls because ROH detection algorithms require discrete SNP calls . Because typical imputation QC thresholds can lead to a high number of missed ROHs , we used extremely stringent imputation QC thresholds that have been shown to achieve accuracy rates similar to those in genotyped SNPs [39] . In particular , we removed 854 , 566 imputed SNPs with dosage r2<0 . 90 in any dataset ( the dosage r2 is equivalent to MACH's r2 measure described in [51] ) , that had a dosage r2<0 . 98 or >1 . 02 in the overall sample , or that had MAF<0 . 05 , leaving 398 , 325 high-quality imputed SNPs . Because only ∼100K SNPs are use to make ROH calls ( see below ) , we could afford to lose a large number of imputed SNPs from QC procedures . ROHs called from imputed data were less variable across platform and across datasets in terms of basic descriptive statistics , in the effects of potential artifacts ( e . g . , SNP missingness rates and excess heterozygosity on Froh ) , and in their associations with schizophrenia . We therefore report results on ROHs called from imputed data . However , results for the ROHs called from raw data were similar , and are shown in Figures S2 and S3 . Of three programs investigated ( PLINK , GERMLINE , and BEAGLE ) , a recent investigation by three of the authors of the current report [40] concluded that PLINK ( using the –homozyg commands ) optimally detected autozygous stretches and maximized power to detect an effect of autozygosity on a phenotype . In particular , the authors recommended: a ) pruning for strong LD ( removing any SNPs having a multiple R2>0 . 90 with all other SNPs in a 50 SNP window ) , which reduced false autozygosity calls by removing redundant markers in SNP-dense regions and by making SNP coverage more uniform; and b ) defining ROHs as being ≥65 consecutive homozygous SNPs with no heterozygote calls allowed [40] . We used these recommendations to detect ROHs in all analyses , although to ensure that we did not miss potential effects of autozygosity , we report on results from the specific ROH threshold ( number of homozygous SNPs in a row ) that minimized the p-value of the Froh-schizophrenia association ( see Figure 3 and Figure S3 ) . This threshold was 65 SNPs-in-a-row ( spanning ∼2 . 3 Mb ) in the imputed SNP data and 110 SNPs-in-a-row ( spanning ∼1 . 7 Mb to ∼3 . 2 Mb depending on the dataset ) in the raw data . It should be noted that results were relatively insensitive to the specific threshold chosen ( Figure 3 and Figure S3 ) . Finally , to ensure that no ROH crossed a region of low SNP density ( e . g . , a centromere ) , we also required that ROHs have a density greater than 1 SNP per 200 kb , and we broke an ROH in two if a gap >500 kb existed between adjacent homozygous SNPs . ROHs can also be categorized by their frequency ( how often a particular haplotype creates ROHs at a given location ) . We used PLINK's –homozyg-group and –homozyg-match arguments to understand whether uncommon ROHs or common ROHs were particularly predictive of case-control status , defining ROHs in a given region as “uncommon” when they allelically matched with 16 ( the median ) or fewer other ROHs in the combined data; all other ROHs were defined as “common . ” For each individual , we summed the total length of all their ROHs in the autosome and divided by the total SNP-mappable autosomal distance ( 2 . 77×109 bases ) to derive Froh , the proportion ( 0 to 1 ) of the autosome in ROHs . Froh was used as the predictor of case-control status in ROH burden analyses . Froh can be influenced by confounding factors like population stratification ( e . g . , if background levels of heterozygosity or autozygosity differed by ancestry ) , low quality DNA leading to incorrect SNP calls , and heterozygosity levels that vary across plates , DNA sources , etc . To control for the effects of stratification , we included the first 20 principal components based on ∼30K SNPs genotyped in all datasets . We also controlled for the percentage of missing calls in the raw SNP data and excess heterozygosity as these track the quality of SNP calls [52] . Using simulations , Keller et al . [25] showed that the ability of Froh to accurately estimate autozygosity is negligibly affected by statistically controlling for excess heterozygosity , and therefore doing so should have minimal effect on results when genotyping error rates are low , but may help elucidate effects of ROHs when such errors are present . We regressed case-control status on Froh separately in each of the 17 datasets using logistic regression , controlling for the potential confounders discussed above . We then employed a mixed linear effects logistic regression model ( using the lme4 package in R version 2 . 11 ) to estimate the overall effect of Froh across datasets , treating dataset as a random factor . This also controlled for SNP platform because dataset was nested within each platform ( controlling for platform was statistically redundant in a model also controlling for dataset ) . To understand whether any genomic area was predictive of case-control status , we divided the autosome into 5 , 742 segments of length 500 kb each . At each segment , an individual was scored as either having a ROH that partially or completely overlapped the segment or not . We performed 5 , 742 logistic regressions , regressing case-control status on whether or not individuals had an ROH in each segment , controlling for covariates described above . To derive a genome-wide significance threshold corrected for multiple testing , we permuted case-control status within the 17 datasets and reran the 5 , 742 logistic regressions , preserving the most significant result of each permutation . We repeated this permutation 1 , 000 times . The 50th most significant p-value was the genome-wide significance threshold and the 100th most significant p-value was the “suggestive” genome-wide significance threshold .	Inbreeding occurs when genetic relatives have offspring . Because all humans are related to one another , even if very distantly , all people are inbred to various degrees . From a genetic standpoint , it is well known that inbreeding increases the risk that a child will have a rare recessive genetic disease , but there is also increasing interest in understanding whether inbreeding is a risk factor for more common , complex disorders such as schizophrenia . In this investigation , we used single-nucleotide polymorphism data to quantify the degree to which 9 , 388 schizophrenia cases and 12 , 456 controls were inbred , and we tested the hypothesis that people whose genome shows higher evidence of being inbred are at higher risk of having schizophrenia . We estimate that the odds of schizophrenia increase by ∼17% for every 1% increase in inbreeding . This finding is consistent with a role for multiple recessive or partially recessive alleles in the etiology of schizophrenia , and it suggests that genetic variants that increase the risk of schizophrenia have been selected against over evolutionary time .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "genome-wide", "association", "studies", "medicine", "psychoses", "mental", "health", "population", "genetics", "schizophrenia", "biology", "evolutionary", "genetics", "psychiatry", "haplotypes", "genetics", "human", "genetics", "evolutionary", "biology", "genetics", "and",...	2012	Runs of Homozygosity Implicate Autozygosity as a Schizophrenia Risk Factor
We describe a consanguineous Iraqi family in which affected siblings had mild mental retardation and congenital ataxia characterized by quadrupedal gait . Genome-wide linkage analysis identified a 5 . 8 Mb interval on chromosome 8q with shared homozygosity among the affected persons . Sequencing of genes contained in the interval revealed a homozygous mutation , S100P , in carbonic anhydrase related protein 8 ( CA8 ) , which is highly expressed in cerebellar Purkinje cells and influences inositol triphosphate ( ITP ) binding to its receptor ITPR1 on the endoplasmatic reticulum and thereby modulates calcium signaling . We demonstrate that the mutation S100P is associated with proteasome-mediated degradation , and thus presumably represents a null mutation comparable to the Ca8 mutation underlying the previously described waddles mouse , which exhibits ataxia and appendicular dystonia . CA8 thus represents the third locus that has been associated with quadrupedal gait in humans , in addition to the VLDLR locus and a locus at chromosome 17p . Our findings underline the importance of ITP-mediated signaling in cerebellar function and provide suggestive evidence that congenital ataxia paired with cerebral dysfunction may , together with unknown contextual factors during development , predispose to quadrupedal gait in humans . The hereditary ataxias comprise a diverse groups of disorders characterized by loss of balance and coordination . They are classified as autosomal recessive , autosomal dominant , X-linked , and mitochondrial , and are clinically and pathogenetically diverse [1] , [2] . The identification of genes associated with individual forms of ataxia has gone a long way towards the identification of factors responsible for development , homeostasis , and function of the cerebellum , which is the organ primarily affected in most forms of ataxia . Recently , a form of ataxia characterized by quadrupedal gait has been described in several families . A syndrome of nonprogressive cerebellar ataxia and mental retardation associated with inferior cerebellar hypoplasia and mild cerebral gyral simplification was initially identified in patients with a disorder termed “dysequilibrium syndrome” ( MIM 224050 ) . This disorder was found to be due to a 199 kb deletion on chromosome 9p24 encompassing all of the very low-density lipoprotein receptor gene ( VLDLR ) and part of the poorly characterized gene LOC401491 [3] . Homozygous premature-truncation-codon ( PTC ) mutations in VLDLR were subsequently found to be associated with severe ataxia , cerebellar hypoplasia , dysarthria , and severe mental retardation . Affected persons walked on all four extremities . Although the affected persons could stand upright and even walk bipedally , they preferred quadrupedal walking [4] , [5] . An unrelated Iranian family with a PTC mutation in VLDLR was subsequently reported in which affected persons had mental retardation , strabismus , short stature , disturbed equilibrium , and walking disability , but no tendency towards quadrupedal gait [6] . A further unrelated family with cerebellar hypoplasia , mental retardation , and quadrupedal gait demonstrated linkage to a locus on chromosome 17p [7] . The observation that some , but not all , VLDLR mutation carriers walk on all four extremities raised the question of whether quadrupedal gait is a functional adaptation that can be seen in congenital ataxia syndromes depending on unknown internal or external influences . Although many mammals and other animals can stand or walk bipedally for shorter or longer periods of time , the manner in which humans do so is unique to our species . Our upright posture is the result of constant skeletomuscular adjustments of posture . In contrast , other primates make frequent use of “holds” during locomotion . In effect , we are always a second or two away from falling down . Human bipedalism and higher cognition are closely integrated . Especially running places high requirements on cognition , such as the need to incorporate information about uneven parts of the ground ahead . Nonhuman primates such as gorillas can only run for short distances , and human children do not develop adult competence in walking and running until about the age of seven years [8] . In the current work , we report on mutations in the CA8 gene in a consanguineous family from Iraq . Affected family members displayed a syndrome of ataxia and mild mental retardation ( AMMR ) and ambulate on all four extremities ( quadrupedal gait ) . In contrast to the previously reported cases associated with VLDLR mutations and with a locus on chromosome 17p , affected persons have only mild mental retardation . Including CA8 as described in the present report , there are now three gene loci , mutations in which are associated with ataxia and cerebral defects , whereby some , but not all affected persons display quadrupedal gait . We therefore suggest that congenital ataxia together with cerebral deficits may , together with other , currently unknown external or internal factors , predispose to quadrupedal gait in humans . We identified an Iraqi family with a syndrome characterized by ataxia and mild mental retardation ( AMMR ) . The healthy parents are first cousins , and four of eight sibs are affected . The parents claimed that the affected persons never learned to crawl on their knees as most infants do , but ambulated from infancy on with their legs held straight with a “bear-like” gait . They also claimed that attempts to teach the children to walk on two legs with crutches or other supports failed . Our own observations confirm that four of the affected children as adults are predominantly quadrupedal ( see Video S1 ) . They walk with straight legs , placing weight on the palms of their hands . Although the affecteds are able to walk on two legs for several steps , they tend to tumble into a quadrupedal position quickly , complain of lack of balance and occasionally fall from a sitting position . Affected persons were noted to have mild mental retardation and dysarthric , slurred speech , but there were no other symptoms such as retinopathy or pyramidal signs . Table 1 provides an overview of the clinical features observed in the four affected siblings encoded using terms of the Human Phenotype Ontology [9] . We used a genome-wide linkage approach to identify the genetic basis of AMMR in this family . We established significant genetic linkage to a 6-centiMorgan interval on chromosome 8q12 with a lod score of 3 . 01 ( Figure 1 ) This interval spans nucleotides 58 , 881 , 724–64 , 607 , 419 on chromosome 8 ( March 2006 , UCSC hg18 assembly ) and contains 17 protein-coding genes , including an obvious candidate , carbonic anhydrase-related protein VIII ( CA8 ) , which is abundantly expressed in cerebellar Purkinje cells [10] . Carbonic anhydrases are a family of monomeric zinc metalloenzymes that catalyze the reversible hydration of CO2 . However , CA8 lacks one of the three histidine residues required for binding to the zinc ion and thus has no catalytic carbonic anhydrase activity [11] . The waddles ( wdl ) mouse is a spontaneous animal model with ataxia and appendicular dystonia without morphological abnormalities in the central or peripheral nervous system . The wdl mouse was shown to harbor a 19-bp deletion in Ca8 that leads to rapid degradation of mutant Ca8 mRNA and to the almost complete lack of detectable Ca8 protein [12] . Although the morphology of the cerebellum appeared normal by confocal microscopy [12] , subsequent investigations revealed abnormalities of parallel fiber-Purkinje cell synapses in the cerebellum together with defects in excitatory transmission [13] . We therefore sequenced all 8 exons of CA8 and identified the homozygous transition c . 298T>C , predicted to lead to the substitution of a serine by a proline residue ( S100P ) , in all affected individuals ( Figure 2 ) . c . 298T>C segregated correctly with the disease in the family and was not found in 200 population-matched controls , thus making a previously undescribed polymorphism unlikely . In addition , we found no other nonsynonymous sequence changes in any of the other 16 genes in the mapping interval , including the TTPA gene , mutations in which are found in ataxia with vitamin E deficiency [14] ( AVED ) . The sole known biochemical function of CA8 is to inhibit inositol 1 , 4 , 5-triphosphate ( IP3 ) binding to IP3 receptor 1 ( ITPR1 ) [15] . ITPR1 plays a critical role in the modulation of intracellular calcium ( Ca2+ ) signaling [16] . Like CA8 , ITPR1 is also highly expressed in Purkinje cells . Mutation of Ip3r1 underlies ataxia in mice and mutations in ITPR1 have been identified in spinocerebellar ataxia 15 in humans [17] . Interestingly , mutant ataxin-3 , which is the protein product of the gene mutated in spinocerebellar ataxia type 3 ( SCA3 ) , was shown to bind to ITPR1 and thereby cause a destabilization of neuronal calcium signaling [18] . These observations suggest that disturbances of ITPR1-mediated calcium signaling may be an important and common phenomenon in hereditary ataxias . We therefore investigated whether S100P affects CA8-ITPR1 binding using recombinant full-length CA8 and a region of ITPR1 containing the known CA8-binding domain [15] . No detectable difference was observed between wildtype and mutant CA8 constructs using blot-overlay assays ( Figure 3 ) . Therefore , since the wdl mouse displays a hypomorphic Ca8 mutation [12] , we hypothesized that S100P might affect protein stability . HEK cells were stably transfected with Flag-tagged wildtype and S100P mutant CA8 constructs using a system that allows single-copy targeted integration of a tetracyclin-inducible construct . This allowed us to compare the stability of mutant and wildtype CA8 mRNA and protein at several different levels of expression . Following selection of stable clones , the expression of wildtype and mutant CA8 was examined using quantitative realtime PCR , and the quantity of wildtype and mutant protein was examined by Western blots . Whereas mRNA levels for the mutant CA8 were not different from those of the wildtype construct ( Figure 4 ) , mutant CA8 protein was strongly reduced , being only barely detectable by Western blotting ( Figure 5A ) . The level of mutant CA8 protein could be partially rescued using MG132 , a proteasome inhibitor ( Figure 5B ) . Taken together , these results indicate that S100P causes a reduction of protein stability owing to accelerated proteasomal degradation , which could be related to protein misfolding or other factors . The cerebellum is a complex neurological structure , containing more than half of the brain's total number of neurons . Cerebellar networks show long-term synaptic plasticity , which indicates that experience-dependent adaptive and learning processes are a salient feature of cerebellar function . Most afferent information enters the cerebellum via climbing fibers ( CF ) and mossy fibers , which excite the Purkinje cells indirectly through the parallel fiber ( PF ) pathway . Binding of CA8 to ITPR1 inhibits IP3 binding to ITPR1 by reducing the affinity of the receptor for IP3 [15] . This is one of several factors that modulate the ability of ITPR1 to rapidly release calcium stores from the endoplasmatic reticulum [19] . Mice with mutations in either Ip3r1 or Ca8 do not display cerebellar atrophy , but rather both show neurophysiological defects [20] , [21] . Modulation of intracellular calcium is important for a number of cerebellar functions such as long-term depression [22] . Therefore , we speculate that the consequences of a CA8 mutation may involve improper modulation of the ITPR1 with resultant functional and/or developmental defects in the cerebellum . S100P leads to a proteasome-mediated reduction in protein stability in our in vitro assay . Loss of protein stability is a common mechanism of missense mutations associated with human disease [23] . We suggest that it is plausible that the mutation leads to a reduction of the amount of CA8 in the cerebellum of the affected individuals , which might lead to a similar defect as that observed in the waddles mouse , in which Ca8 is nearly undetectable in cerebellum owing to a 19 bp deletion in the gene [12] . Mutations at three loci ( CA8 , VLDLR and the yet-to-be discovered gene at 17p ) have now been found to be associated with quadrupedal locomotion in humans , although not in all affected individuals . The clinical picture of the disorders associated with mutations at the three loci is similar but shows important differences . The affected persons of the family described in this work showed a relatively mild degree of mental retardation . Seizures have not been observed in the family described in this study , but did represent a characteristic clinical feature of the family in whom linkage to a locus on chromosome 17p was reported [7] . Table 1 gives an overview of the clinical features encountered in the three forms of ataxia which have been observed with quadrupedal gait . Given the variable incidence of quadrupedalism in individuals with mutations in the same gene , we think it probable that contextual factors during development – either internal or external – contribute to this particular phenotypic outcome [24] . As one possibility , we note that ataxia associated with mutations at all three loci is congenital and also associated with mental retardation , which is not generally a feature of other hereditary ataxias , such as Joubert syndrome [25] or AVED [14] . Thus , perhaps it is only when congenital ataxia is coupled to a certain kind of malfunction of the cerebral cortex that individuals are likely to remain walking on all fours . The family was genotyped using the Affymetrix GeneChip® Human Mapping 10 K Array . The maximum expected lod score for a first cousin marriage with 4 affected children is approximately 3 . 01 . The two-point and multipoint linkage analyses ( Genehunter [26] , Allegro [27] and Merlin [28] ) were performed assuming a fully penetrant autosomal recessive trait with a disease frequency of 0 . 001 and no phenocopies . Analyses were performed using the easyLINKAGE and easyLINKAGE-Plus tools [29] , [30] . All PCRs were performed using High-Fidelity Taq Polymerase ( Invitrogen ) . Two large amplicons comprising the full-length human CA8 transcript and a region of ITPR1 containing the known CA8-binding domain [15] ( amino acids 1387–1647 ) of ITPR1 were amplified by PCR from human fibroblast cDNA . A CA8 amplicon corresponding to positions +91 to +1190 ( numbering based on GenBank entry AY075022 ) was amplified using the primers 5′- tgcactcacactgcggttca-3′ ( f ) and 5′- aagggcattataggaccact-3′ ( r ) . An ITPR1 amplicon corresponding to positions +4361 to +5270 ( numbering based on GenBank entry L38019 ) was amplified using the primers 5′-ctcatgtaccacatccactt-3 ( f ) and 5′-ccttaatgcagagcttctct-3′ ( r ) . We first produced CA8 and ITPR1 constructs for investigation of protein-protein binding . A recombinant CA8 construct corresponding to the complete coding region of human CA8 was then amplified by nested PCR with the primers 5′-gtatGGATCCgatggcggacctgagcttca-3′ ( f ) and 5′- cttgCTCGAGctactgaaatgcagctctaa-3′ ( r ) . Similarly , a recombinant ITPR1 was amplified with the primers 5′- gtatGGATCCgaagaatgtctacacagaga-3′ ( f ) and 5′-cttgCTCGAGttatttcacatttccttctggcgt-3′ ( r ) . The resulting fragments were subcloned into the plasmid pFastBac HT A ( Invitrogen ) using the BamH1 and Xho1 restriction sites ( capitalized in the primer sequences ) . pFastBac HT A adds a hexa-histidine tag and an rTEV protease cleavage site to the N-terminus of the expressed protein . The c . 298T>C ( S100P ) mutant CA8 construct was prepared by using the GeneTailor Site-Directed Mutagenesis kit ( Invitrogen ) using the primers 5′- tcaaaatcagttcttCcgggaggaccattgc-3′ ( f ) and 5′-aagaactgattttgacttcaggataacctga-3′ ( r ) according to the manufacturer's protocol . Additional Flag-tagged CA8 constructs were produced for investigation of the influence of S100P on protein stability . PCR amplification of the full-length wild-type or S100P mutant CA8-pFastBac HT A constructs was used to subcloned to the BamH1 and Xho1 sites of pcDNA3 ( Invitrogen ) using the primers 5′-gtatGGATCCatggcggacctgagcttca-3′ ( f ) and 5′-tgCTCGAGctacttgtcatcgtcgtccttgtagtcctgaaatgcagctct-3′ ( r ) . The reverse primer introduces the Flag tag sequence DYKDDDDK to the C-terminus of the CA8 protein . pcDNA5/FRT/TO Flp-In constructs were then obtained by subcloning Flag-CA8 ( WT ) and Flag-CA8 ( S100P ) from pcDNA3 . 1 into the pcDNA5/FRT/TO vector using the BamH1 and Xho1 restriction sites . All resulting recombinant vectors were verified by sequencing . Spodoptera frugiperda ( Sf9 ) cells were maintained in serum-free Insect-Xpress medium ( Biowhittaker ) supplemented with 5% fetal bovine serum , 10 U/ml penicillin , and 10 µg/ml streptomycin , and cultured at 27°C . The production of the protein was performed according to the protocol of the Bac-to-Bac Baculovirus expression system ( Invitrogen ) . Briefly , the recombinant pFastBac HT A vectors were transformed into E . coli DH10Bac cells ( Invitrogen ) . The isolated positive recombinant bacmids were verified by PCR and used to transfect Sf9 insect cells for viral particle formation . Following three rounds of amplification of baculovirus , the virus titer was determined by plaque assay . Small scale time-course expression experiments were conducted in 6 well plates to optimize the multiplicity of infection ( MOI ) and expression time . Bands corresponding to the size of recombinant protein were visualized on Western blots by detection with monoclonal anti-His6 antibody ( Novagen ) . For large-scale production of protein , 5×108 Sf9 cells were infected at an MOI of 1 plaque-forming unit ( pfu ) /cell with recombinant baculovirus . After about 48 hours cells were harvested and resuspended in 10 ml lysis buffer ( 10 mM Tris-HCl , 10 mM Imidazol , 25× EDTA free protease inhibitor cocktail ( Sigma-Aldrich ) , 0 . 1 mM PMSF ) , sonication was performed on ice for 2×15 s at 50% maximum energy output and was cleared by centrifugation at 12 , 000 rpm for 30 min . Purification of the recombinant protein was performed as described before [31] . Concentration of the recombinant polypeptide was determined with the BCA protein assay ( Pierce ) according to the manufacturer's instruction . For the blot overlay assay , recombinant wildtype and S100P mutant CA8 were separated by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis ( SDS-PAGE ) and then transferred onto a polyvinylidene difluoride ( PVDF ) membrane ( Immobilon-P transfer membrane , Millipore ) . A recombinant fibrillin-1 fragment [32] was used as a control . Non-specific binding sites were blocked with dilution buffer ( ProFound Far-Western Biotinylated Protein∶Protein Interaction Kit , Pierce ) , according to the manufacturer's instructions , and incubation was performed for 2 h at room temperature with 2 µg/ml of the biotinylated ITPR1 fragment in dilution buffer and was followed by washing the PVDF membrane three times with PBS including 0 . 025% Tween 20 . For detection , incubation was performed with streptavidin-horseradish peroxidase conjugate for 3 h at room temperature . Membranes were washed six times in 1× PBS including 0 . 025% Tween 20 . Membranes were then incubated in the UnBlot substrate working solution ( ProFound Far-Western Biotinylated Protein∶Protein Interaction Kit , Pierce ) . Hyperfilm ECL chemiluminescence films ( Amersham Pharmacia ) were exposed according to the manufacturer's instructions . To generate stable tetracycline-inducible Flp-In T-REx cells , 0 . 5 µg pcDNA/FRT/TO encoding Flag-CA8 ( WT ) or Flag-CA8 ( S100P ) ] or empty pcDNA/FRT/TO and 1 µg pOG44 were cotransfected into Flp-In T-REx 293 cells ( 1×106 ) using Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's directions . 48 hours after transfection cells were reseeded at less than 25% confluence , after 4 hours , the medium was changed to medium supplemented with 100 µg/ml hygromycin B ( Invitrogen ) and 15 µg/ml blasticidin . After two weeks , hygromycin-resistant colonies were picked and sub-cultured . Selection of positive colonies was performed by immunoblotting as described below . Tetracycline inducible Flip-In T-REx 293 cells were grown in the absence or presence of 1 µg/ml tetracycline for 24 hours , total RNA of those cells were extracted with NucleoSpin RNA II ( Macherey-Nagel , Duren , Germany ) according to the manufacturer's protocol . 1 µg RNA from cells was reverse-transcribed using the Superscript First Strand synthesis kit ( Invitrogen ) . Quantitative RT-PCR was performed on an ABI Prism 7500 Sequence Detection System using SYBR-Green PCR Master Mix at 50°C for 2 min , 95°C for 10 min , then 40 cycles at 95°C for 15 s , and at 60°C for 1 min . The forward primer for CA8 is complementary to a sequence at the C-terminal region of CA8 , 5′-cttgtggaaggctgtgatgg-3′ , and the reverse primer is complementary to a sequence of the Flag tag , 5′-ttgtcatcgtcgtccttgtag-3′ . The primer pair for amplification of GAPDH was: 5′-ctcaacgaccactttgtcaagctca-3′ , 5′-ggtcttactccttggaggccatgtg-3′ . To detect the expression of CA8-Flag WT and S100P protein , tetracycline-inducible Flip-in T-REx cells seeded in 24 well plate ( 2×105 ) were induced by 0 . 1 µg/ml or 1 µg/ml tetracycline ( Sigma-Aldrich ) . To inhibit the proteasome , cells were incubated in proteasome inhibitor MG132 ( Sigma-Aldrich ) at different concentrations ( 0 . 4 µM , 2 . 0 µM ) in the presence of 1 µg/ml tetracycline . After 24 hours cells were washed twice with PBS and lysed with 100 µl of 1× SDS-sample buffer ( 50 mM Tris-HCL pH 6 . 8 , 10% glycerol , 2% SDS , 1% 2-mercaptoethanol , 0 . 02% Bromphenol blue ) . After brief sonication on ice , 10 µl aliquots were separated on 12% SDS-gel and electrophoretically transferred to a PVDF membrane . Immunodetection was performed by using mouse monoclonal anti-Flag ( 1∶2000 , Sigma-Aldrich ) as primary antibody , followed by subsequent incubation with HRP ( horseradish peroxidase ) -conjugated rabbit anti-mouse secondary antibody ( 1∶2000 , Calbiochem ) . Parallel Western blots were probed with a mouse anti-GAPDH antibody ( 1∶5000 , Abcam ) as a loading control . Protein expression was quantified using the ImageJ program [33] . A CA8 protein multiple alignment from Human ( Homo sapiens; NP_004047 ) , dog ( Canis familiaris; XP_544094 ) , cow ( Bos taurus; NP_001077159 ) , horse ( Equus caballus; XP_001496523 ) , mouse ( Mus musculus; NP_031618 ) , rat ( Rattus norvegicus; NP_001009662 ) , opossum ( Monodelphis domestica; XP_001368351 ) , chicken ( Gallus gallus; XP_419221 ) , frog ( Xenopus ( Silurana ) tropicalis; NP_001011213 ) , trout ( Oncorhynchus mykiss; NP_001118116 ) , zebrafish ( Danio rerio; NP_001017571 ) , and sea urchin ( Strongylocentrotus purpuratus; XP_795365 ) was prepared with ClustalX [34] and sequence logos were prepared using texshade [35] .	We identified a homozygous missense mutation ( S100P ) in the gene encoding carbonic anhydrase VIII in a consanguineous Iraqi family in which affected siblings had mild mental retardation and congenital ataxia characterized by quadrupedal gait . The affected persons walk on their hands and feet with their legs held straight with a “bear-like” gait . Our results show that the mutation S100P induces proteasome-mediated degradation with a severe reduction of the level of CA8 protein . The waddles ( wdl ) mouse , a spontaneous animal model with ataxia , was previously shown to harbor a 19-bp deletion in Ca8 that leads to an almost complete lack of detectable Ca8 protein , resulting in abnormalities in cerebellar synaptic transmission . Therefore , we speculate that the reduction in CA8 protein concentration associated with the S100P mutation could result in similar pathophysiological effects . With the current report , alterations at three gene loci ( CA8 , VLDLR , and a yet-to-be discovered gene on chromosome 17p ) have been reported to be associated with quadrupedal gait . It is unknown whether quadrupedal gait is related to specific molecular abnormalities or is an adaptive response to ataxia in some circumstances . However , we note that ataxia associated with mutations at all three loci is congenital and also associated with mental retardation , which is not generally a feature of other hereditary ataxias .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "genetics", "and", "genomics/gene", "discovery", "genetics", "and", "genomics/medical", "genetics" ]	2009	CA8 Mutations Cause a Novel Syndrome Characterized by Ataxia and Mild Mental Retardation with Predisposition to Quadrupedal Gait
Ubiquitination , a post-translational modification , mediates diverse cellular functions including endocytic transport of molecules . Kaposi's sarcoma-associated herpesvirus ( KSHV ) , an enveloped herpesvirus , enters endothelial cells primarily through clathrin-mediated endocytosis . Whether ubiquitination and proteasome activity regulates KSHV entry and endocytosis remains unknown . We showed that inhibition of proteasome activity reduced KSHV entry into endothelial cells and intracellular trafficking to nuclei , thus preventing KSHV infection of the cells . Three-dimensional ( 3-D ) analyses revealed accumulation of KSHV particles in a cytoplasmic compartment identified as EEA1+ endosomal vesicles upon proteasome inhibition . KSHV particles are colocalized with ubiquitin-binding proteins epsin and eps15 . Furthermore , ubiquitination mediates internalization of both KSHV and one of its receptors integrin β1 . KSHV particles are colocalized with activated forms of the E3 ligase c-Cbl . Knock-down of c-Cbl or inhibition of its phosphorylation reduced viral entry and intracellular trafficking , resulting in decreased KSHV infectivity . These results demonstrate that ubiquitination mediates internalization of both KSHV and one of its cognate receptors integrin β1 , and identify c-Cbl as a potential E3 ligase that facilitates this process . Ubiquitination has been linked to diverse cellular functions including directing protein recycling [1] , [2] . Addition of four or more ubiquitin moieties ( polyubiquitination ) provides the necessary signal for targeting protein for proteasomal degradation . Ubiquitination of target cell surface membrane proteins triggers its internalization and endocytic sorting in addition to proteasomal degradation . For examples , monoubiquitination , the addition of single ubiquitin moieties , of epidermal growth factor receptor ( EGFR ) cytoplasmic tail induces its internalization and transport to the lysosome for degradation [3]–[5] while ubiquitination of membrane receptors β2 adrenergic receptor and interleukin 2 receptor β chain is required for their correct internalization and processing [6]–[8] . The two arms of the ubiquitin/proteasome system are separate , yet closely linked [9] . Ubiquitination is initiated by the conjugation of two ubiquitin molecules to the E1 activating enzyme . The ubiquitin molecules are then transferred to the E2 conjugating enzyme in an ATP-dependent reaction . E2 enzyme interacts with a specific E3 partner and transfers the ubiquitin molecules to the target protein [9] . The ubiquitinated proteins are subsequently transported to the proteasome where the ubiquitin chains are cleaved off by deubiquitinating enzymes ( DUBs ) releasing the ubiquitin molecules to reenter the cycle [2] . Mammalian cells express two E1 activating enzymes , up to 40 E2 conjugating enzymes , and hundreds of E3 ligases [10] , [11] . Kaposi's sarcoma-associated herpesvirus ( KSHV ) primarily utilizes the clathrin-mediated endocytosis pathway to enter host cells [12] , [13] . KSHV is a large enveloped DNA virus that uses cell surface molecules as its receptors , which includes heparan sulfate [14] , integrin α3β1 [15] , integrin α×β3 [16] , xCT [17] and DC-SIGN [18] , to initiate entry into target cells . The possible role of the ubiquitin/proteasome system during the internalization of KSHV and its cognate receptor ( s ) has yet to be examined . In this study , we investigated the role of the ubiquitin/proteasome system during the entry and intracellular trafficking of KSHV and one of its cognate receptors , integrin β1 . We detailed the requirements for proteasome function and de novo ubiquitination for efficient viral entry into endothelial cells and intracellular trafficking to the perinuclear regions . Inhibition of proteasome function by several chemicals reduced viral entry and intracellular trafficking , and caused an accumulation of viral particles in an early endosomal compartment . As a result , KSHV infectivity was reduced . Inhibition of proteasome function resulted in an enrichment of ubiquitinated proteins and a depletion of the cellular pool of free ubiquitin . Inhibition of E1 activating enzyme reduced internalization of KSHV as well as its receptor integrin β1 . The E3 ligase c-Cbl was activated during KSHV entry . Knock-down of c-Cbl or inhibition of its activation blocked KSHV entry and intracellular trafficking , resulting in reduced viral infectivity , suggesting that this E3 ligase may mediate the internalization , and endosomal transport and sorting of KSHV-containing endosomes . We also report the involvement of the E3 ligase Rabex5 , and the endocytic adaptor proteins epsin and eps15 , during KSHV entry into endothelial cells . These results suggest that ubiquitination is required for KSHV entry into endothelial cells and during its endosomal sorting and intracellular trafficking . To investigate the role of proteasome activity during KSHV entry and intracellular trafficking , primary human umbilical vein endothelial cells ( HUVEC ) were pretreated with chemical inhibitors of the 26S proteasome , MG132 or epoxomicin ( EPOX ) , for 1 hr prior to KSHV infection . MG132 is a peptide aldehyde that potently inhibits the transition state of the proteasome [19] . EPOX selectively and irreversibly binds to the catalytic β subunits of the proteasome and effectively inhibits multiple proteolytic activities [20] . At 4 hr post-infection ( hpi ) , the total numbers of viral particles docked at each nucleus or in a whole cell revealed by staining for small capsid protein Orf65 were quantified . Inhibition of proteasome activity with either MG132 or EPOX had no effect on the total numbers of cell-associated viral particles including those that were associated with plasma membrane ( Figure S1 ) , indicating that the proteasome inhibitors do not affect the attachment of viral particles to the cells . However , inhibition of proteasome activity with either MG132 ( Figure 1A–B ) or EPOX ( Figure 1C–D ) resulted in significantly fewer Orf65+ viral particles docked at the nuclear membranes at a dose-dependent fashion compared to cells treated with DMSO alone . Under these experimental conditions , we did not observe any noticeable cytotoxicity of the inhibitors to the cells based on propidium iodide ( PI ) staining ( data not shown ) . To determine whether proteasome inhibitors actually blocked the infectious pathway of KSHV entry and intracellular trafficking , we stained the cells for the expression of LANA protein ( Orf73 ) for evidence of successful infection . Inhibition of proteasome function significantly reduced the numbers of LANA-positive cells at 48 hpi ( Figure 2A and B ) . At the highest used doses of MG132 ( 96 µg/ml ) and EPOX ( 500 nM ) , LANA-positive cells were reduced from 96 . 86% of the untreated cells to 0% and 3 . 49% , respectively ( Figure 2B ) . These results indicate that proteasome activity is necessary for successful KSHV intracellular trafficking to the nuclei . We next sought to identify where viral particles were trapped in cells treated with proteasome inhibitors . HUVEC were pretreated with the indicated inhibitors for 1 hr and then inoculated with KSHV in the presence of the inhibitors . At 4 hpi , cells were stained for viral capsids ( Orf65 ) in red , the cell membrane ( AlexaFluor 647 WGA ) in white , and nuclei ( DAPI ) in blue . Confocal laser-scanning microscopy was used to acquire Z-stack images . Deconvolved images were analyzed in three-dimensional ( 3-D ) to determine the numbers of viral particles in the extracellular , membrane-bound , cytoplasmic , or nuclear spaces . The spot detection function was used to identify and quantify the number of Orf65+ viral particles in a cell . Surface contours were generated to enable the visualization of cell membranes and nuclei ( Figures 3A–C , and Videos S1–S4 ) . There is a modest but statistically significant increase in the numbers of viral particles retained at the plasma membranes following proteasome inhibition ( Figure 3D ) . In addition , a greater proportion of viral particles were retained in the cytoplasm in the proteasome-inhibited cells compared to untreated cells ( Figure 3E ) . As expected , fewer viral particles reached the nuclear membrane ( Figure 3F ) . The increased number of viral particles in the cytoplasm suggests that inhibition of proteasome activity might lead to an arrest in the maturation of viral particle-containing endosomes . To identify the specific cellular compartment where the viral particles were arrested , HUVEC were inoculated with KSHV and fixed at 4 hpi . In addition to Orf65+ viral particles , cells were stained for either EEA1 ( Figure 4 , and Videos S5–S8 ) or LAMP1 ( Figure 5 , and Videos S9–S12 ) to identify the early endosomal compartment or the late endosome/lysosomal compartment , respectively . Confocal z-stack images were acquired and subjected to 3-D colocalization analysis . Proteasome inhibition resulted in more viral particles colocalized with the early endosome marker , EEA1 , in the entire cells ( Figure 4B ) , or viral particles associated with plasma membrane ( Figure 4C ) and cytoplasm ( Figure 4D ) , compared to untreated controls . In contrast , fewer viral particles were colocalized with LAMP1 , the late endosome/lysosome marker , in the entire cells ( Figure 5B ) , or viral particles associated with plasma membrane ( Figure 5C ) and cytoplasm ( Figure 5D ) , following proteasome inhibition . These results suggest that proteasome function is required for the maturation of virus-containing endosomes from the early to late endosome , and that endosomal maturation is necessary for successful viral trafficking to nuclei . Proteasome function per se may not directly regulate internalization of KSHV particles . Inhibition of proteasome function results in the accumulation of ubiquitinated proteins , which may reduce the pool of free ubiquitin available for de novo ubiquitination events , implying that ubiquitination of KSHV and/or its receptors could be the mediator of KSHV internalization . To further understand how proteasome inhibition prevents viral entry and intracellular trafficking , we investigated the role of two adaptor molecules that link clathrin-mediated endocytosis and ubiquitination . Epsin and Eps15 are endocytic adaptor proteins that recognize ubiquitinated cargo through their ubiquitin-interacting motifs ( UIMs ) [21] , [22] . Binding of Epsin to ubiquitinated EGFR is required for its translocation into clathrin-coated pits [22] . Binding of Eps15 to ubiquitinated Cx43 mediates its internalization through endocytosis [21] . Since KSHV enters HUVEC primarily through clathrin-mediated endocytosis , we hypothesized that endocytic adaptor proteins such as Epsin or Eps15 may play an important role during KSHV entry . As shown in Figure 6A , most KSHV particles were colocalized with both Epsin and Eps15 during infection of HUVEC . However , we also observed small numbers of KSHV particles that were not colocalized with the adaptor proteins , possibly reflecting the transient nature of the interaction . The detection of these non-colocalized particles indicates that the observed colocalization is specific but not due to random distribution of the viral particles . These results suggest that KSHV utilizes clathrin-mediated endocytosis to enter cells , and either KSHV or its receptor ( s ) is subjected to ubiquitination during the viral internalization process . Although the proteasome primarily functions as a protein degradation system , inhibition of its functions has deleterious effects on many other cellular activities , including endocytosis [23] , [24] . We hypothesized that inhibition of proteasome function results in an accumulation of ubiquitinated proteins and thus a reduction in the cellular pool of ubiquitin monomers available for de novo ubiquitination reactions . To test this , we treated endothelial cells with either MG132 or EPOX for 1 hr , before the inoculation with KSHV . Cells were analyzed for ubiquitin by immunoblots at different times post-treatment . Both polyubiquitinated proteins and free ubiquitin were detected in the DMSO-treated samples regardless of KSHV infection ( Figure 6B–C ) . The typical smear of higher molecular weight ubiquitinated proteins was visible in all lysates tested , though there was an obvious increase in the amount of ubiquitination in the samples treated with proteasome inhibitors . However , low molecular weight free ubiquitin was not detected in samples treated with proteasome inhibitors , indicating that inhibition of proteasome function reduces the availability of ubiquitin monomers . Integrin α3β1 is one of the cellular receptors for KSHV entry into cells [15] . As expected , 3-D colocalization analysis revealed significant levels of colocalization between KSHV particles with integrin β1 ( Figure 7A ) . To determine the specificity of the observed colocalization , we examined the colocalization of integrin β1 with transferrin , cholera toxin B and rhesus rhadinovirus , all of which enter cells by endocytosis without using integrin β1 as a receptor [25] , [26] . In contrast to the perfect colocalization of KSHV , less than 10% of RRV , transferrin or cholera toxin B was colocalized with integrin β1 ( Figure S2 ) . Integrins are transmembrane glycoproteins composed of α and β chains . Integrins are constitutively endocytosed and recycled [27] , [28] . While integrins are ubiquitinated [29] , the role of ubiquitination in integrin endocytosis is not well understood . As a known receptor for KSHV entry into cells , we hypothesized that ubiqutination regulates the internalization of integrin β1 and KSHV . UBEI-41 is an inhibitor of the E1 activating enzyme that potently and specifically blocks the first step of the ubiquitination reaction [30] . Uninfected cells had low level of ubiquitinated integrin β1 , which was significantly increased following KSHV infection ( Figure 7B ) . Treatment with UBEI-41 almost completely abolished the level of ubiquitinated integrin β1 in both uninfected and KSHV-infected cells . To further understand the role of ubiquitination during the internalization of integrin β1 , we labeled live cells with a monoclonal antibody against integrin β1 to track its internalization from the plasma membrane . As shown in Figure 7C , integrin β1 was internalized and localized to the perinuclear region in untreated cells . Following treatment with UBEI-41 , it was retained at the plasma membrane ( Figure 7D ) , suggesting integrin β1 internalization requires ubiquitination . The results from Figure 6B suggest that depletion of free ubiquitin upon inhibition of proteasome function may be the mechanism that inhibits the entry and intracellular trafficking of KSHV particles . Since inhibition of E1 activating enzyme activity prevents integrin β1 internalization , we expect KSHV entry and intracellular trafficking to be also affected . Indeed , inhibition of E1 activating enzyme and ubiquitination reduced the number of viral particles that successfully entered the cells and trafficked to the nuclei ( Figure 7E–F ) . However , UBEI-41 had no detectable effect on the total numbers of viral particles per cell including those that were associated with plasma membrane ( Figure S1 ) . Thus , the inhibitor did not affect the attachment of viral particles to the cells . PI staining did not detect any cytotoxicity to the cells at all the concentrations used ( data not shown ) . Together , these results indicate that ubiquitination is required for efficient viral entry and intracellular trafficking . E3 ubiquitin ligases are the final effectors in the ubiquitination cascade . E3 ligases contain the substrate recognition motifs that provide the target specificity for ubiquitination . Approximately 1 , 000 E3 ligases have been identified in humans . Among them , Rabex5 is an E3 ligase involved in endocytosis and endosomal maturation by specifically targeting Rab5-positive early endosomes [31] , [32] while the E3 ligase c-Cbl ubiquitinates integrin complexes [33] . c-Cbl is activated through phosphorylation by the Src family kinases principally at tyrosine residues 700 , 731 , and 774 [34]–[36] . Thus , Rabex-5 and c-Cbl are candidate E3 ligases that may potentially mediate KSHV internalization . We therefore sought to investigate the roles Rabex-5 and c-Cbl in KSHV entry into endothelial cells . As shown in Figure 8A–B and Figure S3 , only a minimal amount of KSHV particles ( 6% ) were colocalized with Rabex5 during KSHV infection . In contrast , a significant fraction of KSHV particles ( 70% ) were colocalized with c-Cbl . Furthermore , KSHV particles were also colocalized with two activated phosphorylated forms of c-Cbl , at 40% with phospho-tyrosine 700 , and at 25% with phospho-tyrosine 774 . To directly examine the role of c-Cbl in KSHV entry and intracellular trafficking , we performed loss-of-function analyses . Knock-down of c-Cbl significantly reduced the number of KSHV particles that successfully entered the cells and trafficked to the nuclei ( Figure 9A–C ) . In contract to c-Cbl , knock-down of Rabex-5 had no effect on KSHV entry and intracellular trafficking ( Figure 9A–C ) . Consistent with these results , knock-down of c-Cbl but not Rabex-5 almost completely abolished KSHV infectivity as shown by the numbers of LANA-positive cells at 48 hpi ( Figure 9D and Figure S4 ) . c-Cbl is activated through phosphorylation by the Src family kinases , and Src activity can be blocked by chemical analogs of protein phosphatase 1 ( PP1 ) [37] , which inhibits the E3 ligase activity of c-Cbl [38] . As expected , treatment of HUVEC with increasing doses of PP1 analog inhibited viral entry and intracellular trafficking , resulting in significantly fewer KSHV particles reaching the nuclei ( Figure 10A–B ) . Under these experimental conditions , we did not observe any cytotoxicity to the cells based on PI staining ( data not shown ) . To verify that PP1 treatment prevents c-Cbl phosphorylation , immunoblot analysis on HUVEC lysates collected following treatment with PP1 was performed ( Figure 10C ) . Exposure of cells to either 100 ng/ml EGF or KSHV increased c-Cbl phosphorylation at tyrosines 700 and 774; however these phosphorylation modifications were abrogated in the presence of PP1 analog . Our results clearly demonstrate the essential roles for both ubiquitination and proteasome functions during KSHV entry and intracellular trafficking in endothelial cells . Although it is the first time shown for KSHV infection , the ubiquitin/proteasome system is known to regulate either entry or gene expression of several other viruses , including equine infectious anemia virus [39] , influenza virus [40] , [41] herpes simplex 1 [42] , murine coronavirus [43] , and vaccinia virus [44] . We have found that the total numbers of KSHV particles reaching nuclei were significantly reduced upon suppression of proteasome function , suggesting that virus entry and intracellular trafficking was arrested at a step prior to docking at nuclei . Consistent with these results , we found that proteasome inhibitors reduced KSHV infectivity indicating that they indeed affected the KSHV infectious pathway . Three-dimensional analysis of infected cells revealed that inhibition of proteasome activity resulted in a modest increase in viral particles retained at the plasma membrane , and a significant accumulation of viral particles in the cytoplasm . Furthermore , cells treated with proteasome inhibitors had increased numbers of viral particles retained in the EEA1+ early endosomal compartments . Cells treated with proteasome inhibitors had higher numbers of EEA1+ particles at the membrane , and higher numbers of EEA1+ particles in the cytoplasm as compared to cells treated with DMSO . In contrast , proteasome inhibition reduced the total number of KSHV particles in the LAMP1+ late endosome/lysosome compartments . Lower numbers of LAMP1+ particles were observed both in the cytoplasm and perinuclear regions . These results are in agreement with the study conducted with influenza virus , which showed that proteasome activity is required for progress from the early to late endosomes during its entry of cells [40] . While the proteasome system is required for the entry and intracellular trafficking of several viruses [39]–[44] , the mechanism that mediates this process remains unclear . One of the primary functions of proteasome is degradation of proteins , an intracellular event . For viruses that enter cells via endocytosis , the entry process is initiated by the attachment of infectious virions to the cellular receptors at the surface of the cell followed by internalization of viral particles and receptors . How proteasome function affects the internalization and intracellular trafficking of virions from outside the cell to the nuclei is not immediately obvious . However , it is known that internalization of EGFR and other membrane receptors requires the ubiquitination of the cytoplasmic tail . Following ubiquitination , endocytic adaptor proteins Epsin and Eps15 will recognize the ubiquitinated region of EGFR and facilitate its movement into the clathrin-coated pits . Since KSHV enters cells primarily through clathrin-mediated endocytosis [12] , [13] , we hypothesized that Epsin and or Eps15 are involved in the entry process . Indeed , 3-D colocalization analysis revealed that both Epsin and Eps15 are associated with KSHV particles , suggesting that either the viral particles or their receptors are ubiquitinated . In addition , when proteasome function is inhibited , the level of free ubiquitin is greatly diminished , which may explain how virus internalization and intracellular trafficking are suppressed in this context . Viral particles require cellular structures and processes to enter cells , and specific membrane-bound receptors to facilitate their attachment , internalization , and trafficking through the intracellular space to the final destination . KSHV has evolved to use members of the integrin family , mainly integrin α3β1 and integrin αvβ3 as its receptors [15] , [16] . Integrins are heterodimers composed of α and β subunits . We observed a high percentage of colocalization of KSHV particles with integrin β1 . Integrin β1 is targeted for ubiquitination following inoculation with KSHV , which is prevented by treatment with an inhibitor of the E1 activating enzyme . In addition , live-labeling of cells with antibody against integrin β1 revealed that when ubiquitination is prevented , integrin β1 is retained at the plasma membrane . Finally , entry and intracellular trafficking of KSHV was greatly reduced in endothelial cells treated with the inhibitor of E1 activating enzyme , UBEI-41 , in a dose-dependent manner , proving that ubiquitination itself directly mediates viral entry and trafficking . Together these results indicate that inhibition of proteasome function causes an arrest at an early stage of KSHV entry into cells , which is supported by the observations of increased numbers of viral particles at the membrane and cytoplasm , and increased colocalization of viral particles with EEA1 . These results suggest that ubiquitination may be required for the initial steps of internalization from the plasma membrane and maturation of the early endosome . In fact , ubiquitination may be required at multiple steps of the entry process , from the formation of the clathrin-coated pits and internalization , to endosomal maturation , and finally docking at the nuclear membrane . Further studies should delineate the possible roles for ubiquitination during post-early endosome stages . We have examined E3 ligases that may mediate KSHV internalization and maturation of KSHV-containing endosomes . EEA1 is a docking/tethering protein that binds to membranes containing Rab5 GTPase , thus effecting endosomal fusion and maturation [45]–[47] . Rab5 activation effected by GDP to GTP conversion is regulated by a guanine nucleotide exchange factor ( GEF ) , Rabex-5 . Rabex-5 is autoubiquitinated [48] , and is also an E3 ligase with a N-terminal ubiquitin-binding domain that mediates cargo targeting to the early endosomal membrane [31] , [49] . The fact that we have observed an accumulation of viral particles in an EEA1+ early endosomal compartment in conditions when ubiquitin is unavailable , as well as colocalization of viral particles with Rabex-5 is intriguing . Although the level of colocalization was minimal ( 6% of KSHV particles ) , the infection was unsynchronized , and the assay was performed at a single time point , at 4 hpi . However , we did observe substantial colocalization with the E3 ligase c-Cbl , and more specifically , we observed colocalization with the activated phosphorylated forms of c-Cbl , phospho-Y700 and phospho-Y774 . Consistent with these results , knock-down of c-Cbl significantly inhibited viral entry and intracellular trafficking , and prevented KSHV infection of the endothelial cells . Furthermore , when c-Cbl activation was prevented by inhibition of Src , KSHV entry and intracellular trafficking was dramatically reduced . Immunoblot analysis confirmed that Src inhibition did in fact prevent phosphorylation of c-Cbl at tyrosine residues 700 and 774 during KSHV infection . These results are in agreement with a recent study that demonstrated the important role of c-Cbl tyrosine phosphorylation for KSHV entry into human dermal microvascular endothelial cells via the macropinocytic pathway [50] . The same group also investigated the ubiquitination of myosin IIA during KSHV infection , and found that c-Cbl may be mediating the formation of membrane blebs through its E3 ligase activity . However , our results have demonstrated the role of the ubiquitin proteasome system during clathrin-mediated KSHV endocytic entry into endothelial cells . We have independently identified c-Cbl as a mediator of KSHV entry into endothelial cells , confirming its essential roles as both an adaptor molecule and as an E3 ligase in clathrin-mediated endocytosis . Nevertheless , it remains possible that ubiquitination and other E3 ligases are involved at later stages of endosomal maturation . Mammalian cells express hundreds of E3 ligases . The functions of most of these other ligases have not been well-studied . The identification of an E3 ligase that specifically mediates ubiquitination leading to viral entry could certainly provide a potential therapeutic target . Early passage of HUVEC were obtained from Clonetics , Lonza and maintained in complete EBM-2 culture media ( Allendale , NJ ) . KSHV-infected BCP-1 cells established from the peripheral blood mononuclear cells of a PEL patient [51] were maintained in culture in RPMI1640 containing 10% fetal bovine serum ( FBS ) . Doxycycline-inducible iSLK cells , provided by Dr . Jae Jung at the University of Southern California , were maintained in DMEM supplemented with 10% FBS , 250 µg/ml of G418 , 1 , 200 µg/ml of hygromycin , 1 µg/ml of puromycin , and 1% penicillin-streptomycin solution . To induce virus production from BCP-1 cells , cells grown to log-phase were serum-starved overnight in RPMI1640 . FBS was added back to the culture media to a final concentration of 10% , together with 12-O-tetradecanoyl-phorbol-13-acetate ( TPA ) at 30 ng/ml and sodium butyrate at 200 µM . At 2 days post-induction , the cells were washed and the media was replaced with fresh RPMI1640 containing 10% FBS without TPA or sodium butyrate . The culture medium was collected at 6 days post-induction . To obtain virus from iSLK cells , cells were induced with 1 µg/ml doxycycline and 1 mM sodium butyrate in DMEM supplemented with 10% FBS and the culture medium was collected 4 days later . To concentrate virus , the culture medium was centrifuged first at 5 , 000× g for 30 min to eliminate cell debris and then passed through a 0 . 45-µm filter , followed by centrifugation at 100 , 000× g for 3 h . The final pellet was dissolved in the culture medium overnight and was adjusted to a desired volume . Undissolved debris was eliminated by centrifugation at 5 , 000× g for 10 min . All the procedures for virus concentration were carried out at 4°C . The concentrated virus preparations were aliquoted and stored at −80°C for later experiments . For cell infection , HUVEC were seeded onto glass cover slips overnight to achieve 70–80% confluency . For assays using chemical inhibitors , cells were pretreated with the inhibitors in EBM-2 media for 1 hr prior to infection . The cells were then inoculated with the virus preparation and incubated for the indicated times in the presence of the inhibitors , fixed in 2% paraformaldehyde and processed for immunostaining of KSHV particles . A monoclonal antibody isotype IgG2a ( clone 6A ) to KSHV small capsid protein ( Orf65 ) was used to stain KSHV particles [52] . Rabbit antibodies to EEA1 , integrin β1 , epsin , and eps15 were purchased from Abcam ( Cambridge , MA ) . A rabbit polyclonal antibody against LAMP1 was purchased from Sigma Life Science ( St . Louis , MO ) . Rabbit monoclonal antibodies against c-Cbl , and pY700 c-Cbl were from Epitomics ( Burlingame , CA ) . A rabbit antibody against pY774 c-Cbl was from Cell Signaling ( Danvers , MA ) . Mouse monoclonal antibodies to Cbl and Rabex-5 from Santa Cruz Biotechnology ( Santa Cruz , CA ) were also used for Western-blot detection . A monoclonal antibody against ubiquitin ( P4D1 ) was from Santa Cruz . A rat anti-LANA monoclonal antibody was purchased from Abcam ( Cambridge , MA ) . Secondary antibodies AlexaFluor 568 goat-anti-mouse IgG1 and IgG2a , AlexaFluor 647 goat anti-mouse IgG1 , AlexaFluor 488 goat anti-rabbit , and AlexaFluor 568 goat-anti-rat were from Molecular Probes , Invitrogen ( Carlsbad , CA ) . DAPI was from BioChemika Ultra , Sigma . Chemical inhibitors of proteasome function MG132 and EPOX were purchased from Sigma . The inhibitor of E1 activating enzyme UBEI-41 was obtained from Biogenova ( Rockville , MD ) . Chemical inhibitor of Src kinases pp1 was from Sigma . All chemical inhibitor stock solutions were prepared according to the manufacturer's directions . c-Cbl , Rabex-5 and control shRNA lentivirus particles were purchased from Santa Cruz . HUVEC grown to 40–50% confluency in 6-well plates were pretreated with 5 µg/ml polybrene for 30 min and infected with the specified lentivirus particles . The infected cells were centrifuged at 2 , 000 rpm for 1 hr to facilitate virus entry , and then incubated overnight . The virus were removed on the second day , replaced with growth medium and cultured for another 3 days before inoculation with KSHV or lysed for Western-blotting detection of protein expression . For immunofluorescence analysis , cells were incubated with each primary antibody for 1 hr and then the appropriate secondary antibody conjugated to Alexa Fluor 488 , 568 , or 647 ( Invitrogen ) , all at 1∶100 dilution . After washing with PBS , cells were stained with DAPI and mounted onto glass slides with FluorSave ( Invitrogen ) . For quantification of entry and intracellular trafficking of viral particles to the nuclei , images were acquired using a Zeiss Axiovert 200M epifluorescence microscope equipped with a 63× oil immersion objective ( Carl Zeiss Microimaging Inc . , Thornwood , NY ) . Images were acquired for at least 5 fields of view per coverslip to allow counting of Orf65+ viral particles docked at nuclei . For colocalization experiments , images were acquired with an Olympus FV1000 scanning confocal microscope equipped with a 60× NA 1 . 42 oil immersion objective ( Olympus Life Science , Center Valley , PA ) . Z-stacks were acquired at 0 . 25 µm per slice by sequentially scanning and in 8 bit color depth . Theoretical×and y axis resolution was 0 . 2 µm and 0 . 2 µm , respectively . Olympus FV1000 software was used to generate cross-sectional images and 3D-projection images ( Olympus Life Science ) . Fluorescence images were acquired sequentially , using a 405-nm laser line with emission at 461 nm for DAPI; a 488-nm laser line with emission at 520 nm for Alexa Fluor 488; and a 543-nm laser line with emission at 603 for Alexa Fluor 568 , and a 633-nm laser line with emission at 668 for AlexaFluor 647 . Voltage , gain , and offset were adjusted to prevent bleed-through . Images were assembled using Adobe Photoshop CS3 version 10 ( Adobe Systems Incorporated , San Jose , CA ) . Z-stack images were first deconvolved with AutoQuant deconvolution software using the adaptive point spread function ( Media Cybernetics , Inc . , Bethesda , MD ) . Deconvolved images were then analyzed with Imaris 3-D image analysis software ( Bitplane , Zurich , Switzerland ) . The threshold for each channel was automatically calculated by the program using the Pearson's coefficient approach following orthogonal regression analysis on the image's scatterplot . To measure colocalization of endosomal markers with KSHV particles through the x , y , and z planes , the red channel ( Orf65 ) was masked to create a new channel that encapsulates the 500 nm region of the image around the center of each viral particle . The masked channel was then used to determine viral particle colocalization with each of the endosomal markers ( EEA1 or LAMP1 ) as well as for epsin , eps15 , Rabex-5 , c-Cbl , Y700-c-Cbl , and Y774-c-Cbl . The same procedure was also used to determine colocalization of viral particles with integrin β1 and the plasma membrane labeled with AlexaFluor647 wheat germ agglutinin ( WGA ) in the far-red channel as well as viral particles that were colocalized with nuclei stained with DAPI in the blue channel . The total number of colocalized pixels ( voxels in 3D ) was counted for cells that had a minimum of five viral particles per cell . To avoid an overrepresentation of colocalization , only one colocalization event was counted for each viral particle . To detect ubiquitin , HUVEC were pretreated for 1 hr with the indicated chemical inhibitors of proteasome function then inoculated with KSHV . Cell lysates were collected at the indicated time post-infection and subjected to SDS-PAGE and immunoblotting to detect ubiquitin . For analysis of c-Cbl phosphorylation , HUVEC were serum-starved for 4 hr to reduce basal levels of c-Cbl phosphorylation , and pretreated with increasing concentrations of PP1 analog for 1 hr prior to inoculation with KSHV . Cells treated with 100 ng/ml EGF were included as a positive control for c-Cbl phosphorylation . Lysates were collected and subjected to SDS-PAGE and followed by immunoblot detection for total c-Cbl and the indicated phosphorylated forms of c-Cbl . HUVEC cells were pretreated with DMSO or 5 µm UBEL-41 for 1 hr prior to inoculation with KSHV for 1 hr . Cells were washed 3 times with cold PBS , harvested by centrifugation and suspended in lysis buffer containing 20 mM Tris pH 7 . 4 , 100 mM NaCl , 1% NP-40 , 1 mM EDTA , 1 mM EGTA and protease inhibitor cocktail ( sigma ) . Lysates were sonicated at 25% efficiency for 3 times , each for 5 sec , and centrifuged for 10 min at 10 , 000 g at 4°C . The lysates were pre-cleared with protein G beads for 1 hr and incubated with mouse anti-integrin β1 for 2 hr at 4°C . The protein G beads were used to precipitate the immune complexes overnight 4°C with rotation , and the immune complexes were analyzed by SDS-PAGE and Western-blotting to detect ubiquitin . Results of analysis of viral particles trafficking to nuclei in the presence of chemical inhibitors were expressed as the mean ±s . d . Data were analyzed using t-test , analysis of variance , and Mann-Whitney rank-sum tests where appropriate , with p<0 . 05 considered as significant using SigmaPlot 11 . 0 ( Systat Software , Inc . , San Jose , CA ) . Cells bearing a minimum of five viral particles per cell were included in the analysis . The distributions of viral particle localization are summarized in box and whisker plots to represent the median values ( middle lines ) , the 75th and 25th percentiles ( opened boxes ) , and the 90th and 10th percentiles ( short lines ) . Outliers outside the 90th and 10th percentiles are represented as black dots .	Ubiquitination , a post-translational modification , mediates important cellular functions including endocytic transport of molecules . Kaposi's sarcoma-associated herpesvirus ( KSHV ) is a gammaherpesvirus linked to the development of Kaposi's sarcoma , an endothelial malignancy commonly found in AIDS patients , and several other malignancies . KSHV enters endothelial cells primarily through clathrin-mediated endocytosis . In this study , we show that the proteasome activity is required for KSHV entry into endothelial cells and intracellular trafficking to nuclei . Inhibition of proteasome activity reduced KSHV infectivity and led to the accumulation of KSHV particles in EEA1+ early endosomal vesicles . Furthermore , we show that ubiquitination mediates the internalization of both KSHV and one of its receptors integrin β1 . KSHV particles are colocalized with ubiquitin-binding proteins epsin and eps15 , as well as activated forms of the E3 ligase c-Cbl . Knock-down of c-Cbl or inhibition of its phosphorylation blocked KSHV entry and trafficking , thus preventing KSHV infection of endothelial cells . Together , these results illustrate the essential role of ubiquitination during the internalization of KSHV and its cognate receptor integrin β1 . The identification of an E3 ligase that mediates the ubiquitination of KSHV and its cognate receptor integrin β1 leading to viral entry provide a potential therapeutic target for this oncogenic virus .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "medicine", "biology" ]	2012	The Ubiquitin/Proteasome System Mediates Entry and Endosomal Trafficking of Kaposi's Sarcoma-Associated Herpesvirus in Endothelial Cells
Trypanosomatid parasites are notorious for the human diseases they cause throughout Africa and South America . However , non-pathogenic trypanosomatids are also found worldwide , infecting a wide range of hosts . One example is Trypanosoma ( Megatrypanum ) theileri , a ubiquitous protozoan commensal of bovids , which is distributed globally . Exploiting knowledge of pathogenic trypanosomatids , we have developed Trypanosoma theileri as a novel vehicle to deliver vaccine antigens and other proteins to cattle . Conditions for the growth and transfection of T . theileri have been optimised and expressed heterologous proteins targeted for secretion or specific localisation at the cell interior or surface using trafficking signals from Trypanosoma brucei . In cattle , the engineered vehicle could establish in the context of a pre-existing natural T . theileri population , was maintained long-term and generated specific immune responses to an expressed Babesia antigen at protective levels . Building on several decades of basic research into trypanosomatid pathogens , Trypanosoma theileri offers significant potential to target multiple infections , including major cattle-borne zoonoses such as Escherichia coli , Salmonella spp . , Brucella abortus and Mycobacterium spp . It also has the potential to deliver therapeutics to cattle , including the lytic factor that protects humans from cattle trypanosomiasis . This could alleviate poverty by protecting indigenous African cattle from African trypanosomiasis . Human health is intimately linked to animal health through the impact of infectious agents on livestock productivity and their potential for zoonosis [1] . Indeed , animal borne disease represents the major source of both emergent and resurgent pathogens in humans , this affecting communities in both the developed and developing world . One major source of such zoonotic infections is cattle , which threaten human health in the developed world through their capacity to transmit bacterial infections including Escherichia coli , Salmonella spp . , Campylobacter spp . , Brucella spp . and mycobacteria . In the developing world , livestock are also a reservoir for Human African Trypanosomiasis ( HAT ) caused by Trypanosoma brucei rhodesiense . This parasite , and Trypanosoma brucei gambiense , is closely related to Trypanosoma brucei brucei , which cannot infect humans . The basis of this host restriction is that human serum contains a trypanolytic component of high-density lipoprotein , ApoLI , that kills T . brucei brucei [2] , but to which the human infective parasites have evolved resistance [3] , [4] . Although trypanosomes are important causes of human and animal disease , many species are non-pathogenic . One of these is Trypanosoma ( Megatrypanum ) theileri , a cosmopolitan parasite of bovids that infects most cattle worldwide [5]–[11] . Although one report describes reduced milk yield in three infected cows [12] , the ubiquity of infection with this organism in cattle herds suggests that it has no significant impact on health or productivity in healthy animals , while cases of disease in immuno-compromised animals are sufficiently rare as to merit individual case reports [13]–[15] . The parasite is transmitted in the contaminated faeces of tabanid flies and gains entry into the host through breaks in the skin or via contamination of the oral mucosa [16] . Thereafter , it lives extracellularly , being sustained at a low level ( ∼100 organisms/ml ) for the life of the host . Being non-pathogenic , systemic , related to the genetically well-characterised T . brucei , and sustained long-term , we conceived that T . theileri would make a suitable protein delivery system in cattle , able to generate immunity to expressed antigens . As a naturally non-pathogenic kinetoplastid , T . theileri offers considerable advantages over alternative vaccine delivery systems that comprise engineered , attenuated pathogens such as Salmonella [17] , [18] , Mycobacterium [19] , [20] , E . coli [21] , Vibreo cholera , Listeria or Shigella [22] . Furthermore , being maintained over long periods at low level , T . theileri offers the potential to generate sustained immune responses of greater efficacy than conventional vaccination approaches and to deliver therapeutic proteins of benefit to bovine health or to limit the zoonotic potential of cattle borne diseases . To develop T . theileri as a protein delivery system , the conditions for its axenic in vitro growth and genetic manipulation were established . To achieve this , a T . theileri isolate , originally identified as a contaminant of a primary bovine reticulocyte culture , was cultured under various conditions , optimal and sustained growth being achieved using a semi-defined medium containing 50% conditioned media from a bovine cell culture . In this medium , cell densities of 1×105 – 2×106 cells/ml were achieved during routine passage ( Figure S1 in Text S1 ) . Under these conditions , the position of the kinetoplast ( a specialised mitochondrial genome in trypanosomatid parasites; [23] ) varied in relation to the cell nucleus but this was not clearly dependent upon the cell culture density ( Figure 1A , Figure S1 in Text S1 ) . To generate stable transfectants , bi-cistronic and tri-cistronic expression constructs were developed that comprised a drug selectable marker gene and a reporter gene , this being integrated into the small subunit 18S ribosomal RNA gene locus ( Figure 1B ) . Here , RNA polymerase I-mediated read-through transcription can drive efficient gene expression , transcripts being processed and capped via trans splicing of a T . theileri-specific spliced leader ( SL ) RNA sequence , matching the situation in other kinetoplastids [24] , [25] . This SL sequence was identified by 5′RACE of T . theileri transcripts and matched the determined SL sequence of one previously isolated T . theileri sample , T . theileri D30 ( Figure S2 in Text S1; [26] ) , this matching the trypanosomatid consensus . In order to drive effective gene expression , RNA processing signals derived from T . theileri were used . Since almost no molecular information for these organisms was available , we isolated T . theileri intergenic sequences using degenerate primers able to amplify between the coding regions of well conserved genes ( i . e . paraflagellar rod , tubulin , actin genes ) predicted from the analysis of other kinetoplastid genomes to be tandemly arranged ( Figure S3 in Text S1; [27] ) . The resulting expression constructs were transfected into T . theileri via nucleofector technology and selected using blasticidin or phleomycin drug selection . Unlike other kinetoplastid organisms , G418 and hygromycin B were not effective for selection in T . theileri , wild type cells exhibiting high levels of resistance to these drugs . This may have developed in natural T . theileri populations through long-term use of aminoglycoside antibiotics for the treatment of infections such as those responsible for bovine mastitis . In order to develop the system for the optimal delivery of proteins to the bovine host , the potential to target expressed heterologous proteins to distinct cellular locations in T . theileri was investigated . For secretion , fusion proteins were created comprising the N-terminus of the T . brucei BiP/GRP78 protein ( PMID:8227199 ) ( N-BiP ) , which in T . brucei targets proteins for release into the extracellular milieu [28] . For cell surface expression , signals providing a C-terminal GPI-addition sequence were also added ( N-BiP-GPI ) , whereas for internal expression , native proteins were expressed . Quantitative expression analysis of a chloramphenicol acetyl transferase ( CAT ) reporter protein confirmed protein expression and accurate targeting , with the percentage of protein released to the cell medium being 70% for the N-BiP fusion , 7% for the N-BiP-GPI fusion and 15% for the native protein , the latter probably including protein released from dead cells in the culture ( Figure 1C ) . Although the overall expression efficiency was reduced considerably ( 10 to 50-fold ) by the inclusion of heterologous targeting signals ( N-BiP , N-BiP-GPI ) , this could be compensated somewhat by the expression of two copies of the reporter gene in tandem , generating a tri-cistronic construct ( Figure 1C ) . To evaluate T . theileri as a vaccine delivery system , cells were engineered to express the Bd37 antigen from the cattle pathogen , Babesia divergens [29] . In general , protection against Babesia is thought to require components of the innate and adaptive immune systems ( Reviewed in [30] ) . Effector arms of the adaptive immune system include antigen-specific antibodies , which have been hypothesized to target both infected erythrocytes and extracellular merozoites . There is also evidence that antibody-dependent cell-mediated cytotoxicity ( ADCC ) plays a role in parasite control during acute infection [31] , [32] . A recombinant version of the Bd37 protein has been shown to stimulate antibody-directed protective immunity to B . divergens when administered in the context of an adjuvant [33] , [34] . In our study , Bd37 protein was targeted for internal expression , surface localization or extracellular release by the expression of unmodified Bd37 , N-BiP-Bd37-GPI or N-BiP-Bd37 fusion proteins . Since anti-Bd37 antibody was not effective in immunfluorescence assays , localisation was assessed for surface and secreted expression by incubating non-permeabilised or 0 . 1% TRITON X100 permeabilised cells with an antibody against T . brucei BiP , this recognising the N-terminal BiP component of the expressed fusion protein . In wild type cells , a signal was detected in permeabilised cells ( Figure 2 D–F ) , reflecting the recognition of T . theileri BiP protein by antibody raised against T . brucei BiP . However , as expected , non-permeabilised cells showed little reactivity ( Figure 2 A–C ) . In contrast , in transgenic T . theileri expressing Bd37 , a strong punctate signal was detected in non-permeabilised cells for the N-BiP-Bd37-GPI fusion ( Figure 2 M–O ) , whereas N-BiP-Bd37 cells exhibited a somewhat weaker signal concentrated close to the kinetoplast ( Figure 2 G–I ) . This supported a surface-associated expression for the N-BiP-Bd37-GPI protein and flagellar pocket associated signal ( the route for protein secretion in trypanosomatids ) for N-BiP-Bd37 , although staining of flagellar pocket-proximal vesicles cannot be excluded . Transcripts generated from the expressed reporter constructs were of the expected size , with the polyadenylation sites used for mRNAs derived from each of the constructs being approximately coincident with each other and with the site of beta tubulin polyadenylation ( Figure 3 and Figure S4 in Text S1 ) . Having established protein localisation at different cellular locations , each T . theileri cell line was inoculated intravenously into groups of cattle ( internal expression , n = 6 animals; secreted expression , n = 5 animals; surface expression , n = 6 animals ) , these being maintained in a fly-free , high containment facility . Of these , 5 of the 17 animals had been found to contain a pre-existing natural T . theileri population by screening blood using a specific , nested PCR assay for the T . theileri tubulin gene array ( Figure 4A ) . Regardless of this , the transgenic T . theileri established in all animals within one week , these being detectable by nested PCR specific for the Bd37 gene present within the integrated expression construct ( Figure 4B ) . In each of the experimental groups , the transgenic T . theileri was sustained in all animals throughout the 12 weeks after inoculation . This was without any adverse effects on animal health being detected . Moreover , the transgenic T . theileri could be recovered after several weeks of growth in cattle and was found to maintain Bd37 gene expression ( Figure 5 ) . To assess whether the inoculated animals were able to generate a specific immune response to the delivered antigen , serum samples were tested weekly for Bd37-specific IgG responses by ELISA . Under all treatment conditions , sero-conversion was observed , i . e . at least a 3-fold increase in ELISA OD compared to pre-immunisation control sera from the same animal ( Figure 6A–C ) . Interestingly , the route of antigen expression affected both the sero-conversion frequency and the resulting antibody titre , with sero-conversion occurring in 4/6 animals receiving the cytosolically expressed antigen , 3/6 animals receiving the surface expressed antigen , and 5/5 animals receiving the secreted antigen . Moreover , the secreted antigen produced a significantly higher end-point serum titre than either the surface expressed antigen ( Kruskal-Wallis one way ANOVA , p<0 . 05 ) or the cytosolically expressed antigen ( p<0 . 001 ) ( Figure 6D ) , with antibody levels continuing to increase for 60 days post inoculation before plateauing and remaining high for at least a further 24 days ( Figure 6E ) . Inoculation of a further 1×106 cells expressing the secreted antigen on week 8 had no detectable stimulatory effect on the levels of antibody generated ( vertical arrow , Figure 6C , E ) . To compare the antibody responses generated by transgenic T . theileri with conventional immunisation approaches , the antibody titres were assessed with respect to those established in a prime-boost vaccination study using the same antigen . This demonstrated that the levels of anti-Bd37 antibody stimulated by the transgenic T . theileri delivery system were equivalent to those generated by the same antigen delivered in conventional adjuvant , producing titres that are protective against Babesia in challenge trials ( dashed line , Figure 6D ) . Furthermore , the immune response generated to the T . theileri-secreted Bd37 could effectively compete with a monoclonal antibody able to confer passive immunity to Babesia infection in a gerbil model ( Figure 6F ) , demonstrating the recognition of common , potentially protective epitopes . Combined , these assays confirm that T . theileri represents an effective antigen delivery system able to generate sustained immune responses equivalent to , or exceeding , those generated by standard vaccination and at levels known to be protective against a target pathogen . Our work describes a naturally non-pathogenic eukaryotic organism , T . theileri , engineered to deliver antigens and therapeutic proteins to cattle . Importantly , T . theileri is a flexible and adaptable system for the targeted expression of individual , or cocktails of multiple , heterologous proteins simultaneously . This allows for the targeting of different pathogens , or multiple defined antigens of a specific pathogen , providing an effective approach to limiting the transmission of disease from livestock to man , as well as maintaining bovine health and maximising bovine productivity . This system has many advantages over traditional vaccine delivery technologies , including the possibility of oral delivery , a natural route of T . theileri infection , and the potential for sustained immune stimulation through prolonged infection without adverse consequence on productivity or health . It also provides an explicit example of an unanticipated applied benefit derived from the extensive research investment in the basic biology of kinetoplastid pathogens . A number of kinetoplastid parasites have been successfully transfected to drive ectopic expression of endogenous genes or heterologous expression of reporter genes[35] . In general , these have been used for the analysis of gene function in the transfected organisms or for protein expression in vitro . In particular , the kinetoplastid protozoa T . brucei , Crithidia fasciculata , Leishmania amazonensis and Leishmania tarentolae have each been engineered as eukaryotic vehicles for the expression of proteins which are appropriately post-translationally modified , that retain biological activity , or which are suitable for meaningful inhibition studies for drugs targeted against kinetoplastids [36]–[40] . In each case , the basic organisation of the expression vectors employed was similar . However , there is no predictable cross functionality of RNA processing signals . For example , in one study , transfection of a range of kinetoplastid protozoa with a common expression construct developed in Leishmania major resulted in reporter gene expression in other Leishmania spp . , but not in T . brucei or T . cruzi [41] . The specificity of RNA processing signals among kinetoplastids necessitated that we isolated intergenic sequences from T . theileri . To our knowledge , no protein coding gene sequence have been characterised in these organisms to date . Despite this , we were able to make use of the tandem arrangement of several conserved housekeeping genes in trypanosomatids to isolate intergenic sequences between alpha and beta tubulin genes , beta and alpha tubulin genes , actin genes and the PFR genes . Conservation in these genes allowed successful amplification between genes using degenerate oligonucleotides . In each case , the predicted protein sequences were highly similar to those in other kinetoplastids over the sequenced region , this being reinforced by the reactivity of monoclonal antibodies specific for alpha tubulin , PFRA and BiP in T . brucei with T . theileri ( Figure 1A , Figure 2 and data not shown ) . Extensive protein coding similarity is also evident from a whole genome analysis of T . theileri that we have recently completed ( our unpublished observations ) . Despite this similarity in coding regions , there was no recognisable similarity in the intergenic regions of each gene between species . This emphasised the requirement to use endogenous intergenic sequences in the developed expression constructs . Indeed , transient transfection of T . theileri with a reporter construct generated for use in T . brucei did not result in detectable heterologous gene expression ( unpublished observations ) . Cloning and sequencing of 5′RACE products derived from the actin gene transcript identified the SL sequence common to all trypanosomatid mRNAs . Interestingly , the sequence of the leader RNA did not match the SL sequence of a previous T . theileri isolate ( K127 ) , which was reported to exhibit a 1 nucleotide distinction from the sequence identified on the majority of trypanosomatid parasites . Instead , the leader sequence identified in the isolate of T . theileri used in our study matched exactly the trypanosomatid consensus and shared with T . theileri D30 . This suggests variation in this sequence amongst different T . theileri isolates . Previous approaches have used kinetoplastid parasites as vehicles for protein expression in vitro , or for potential medicinal use via expression of biomolecules from attenuated pathogens in vivo . In these , and other cases of pathogen-based vaccine vectors ( for example , using attenuated Salmonella , Mycobacterium , E . coli , V . cholera , Listeria or Shigella ) , the vehicles have a potential to revert to pathogenicity raising issues of long-term efficacy and safety . Moreover , recombination with non-attenuated pathogens can generate vaccine escape mutants , as observed after pneumococcal vaccination in the USA [42] . By exploiting an already non-pathogenic organism ( T . theileri ) for protein expression in its natural host , the possibility of reversion to , or unanticipated , pathogenicity is greatly reduced . Supporting this , the T . theileri isolate used in our studies has been maintained in tissue culture for over 6 years and yet , when inoculated into cattle , generated no ill effects and was sustained at low abundance throughout a 12-week trial . The transgenic parasites could also establish and be sustained in the context of a pre-existing natural infection . This is an important consideration since T . theileri is almost ubiquitous in cattle herds worldwide with , for example , 44%–80% incidence levels reported in New York State and up to 93% in Louisiana [43] . A survey of local adult cattle in Edinburgh revealed that all animals tested ( 16/16 ) were positive for T . theileri infection by the nested PCR assay for tubulin intergenic sequence ( data not shown ) . T . theileri has clear potential as a flexible vaccine vehicle , able to target a wide range of pathogens , including viruses ( e . g . Foot and Mouth Disease Virus , Bovine Viral Diarrhea Virus ) , bacteria ( e . g . E . coli , Salmonella spp . , Campylobacter spp . , Brucella spp . and mycobateria ) , ectoparasites ( e . g . ticks , mites , lice ) as well as pathogenic eukaryotic parasites ( Babesia spp . , Neospora caninum , Coccidia spp . , Trichomonas spp . and helminths ) . However , T . theileri also has potential to deliver therapeutic proteins such as antimicrobials systemically in cattle , able to limit the impact of bacterial infections , for example . Furthermore , we propose that it may be possible to avoid the need to engineer transgenic cattle to express ApoLI or mutant variants [44] , [45] , allowing indigenous and disease-resistant African livestock to combat pathogenic trypanosome infection via T . theileri directed ApoL1 expression . Such an approach would require that the ApoL1 is expressed in a form that is non-immunogenic and active in cattle . Nonetheless , this could provide a simple and cost effective route to limiting the effects of trypanosome infection on the productivity or zoonotic potential of indigenous African cattle . This has the potential to increase their productivity and eliminate an important reservoir for human disease , alleviating poverty and disease in afflicted regions . Animal trials in this manuscript were reviewed and approved through the ethical review committee at Moredun Research Institute ( Experiment approval no . E28/10 ) . Experiments were carried out under a UK Home Office Licence ( PPL 60/4044 ) in accordance with the UK Animals ( Scientific Procedures ) Act , 1986 . Bovine conditioned media was produced by growing Madin-Darby Bovine Kidney ( MDBK ) cells in Eagle's Minimum Essential Medium with Earle's Balanced Salt Solution and sodium bicarbonate ( Sigma , M2279 ) supplemented with 1% MEM non-essential amino acids ( Invitrogen , 11140 ) , 1% L-Glutamine solution ( from 200 mM solution , Sigma , G7513 ) , and 10% FCS until confluence ( 2–3 days ) . The MDBK-conditioned media was then harvested and filtered prior to use . T . theileri were cultured in 50% HMI-9 medium [46] supplemented with 20% FCS and 10% Serum+ and 50% MDBK-conditioned media as described above . The plasmid backbone used for construction of all expression vectors was derived from the pGemT Easy plasmid ( Promega ) . All elements of the vector were prepared through PCR of the noted sequence , followed by cloning into the vector backbone , and construction was accomplished through the insertion of appropriate restriction sites at the 5′ and 3′ ends of each segment . The segments used for the construction of the expression cassettes and the corresponding primers used in this study include: the 5′ fragment of the T . theileri SSU rRNA gene ( SSU5-ApaI-For and SSU5-AvrII-Rev ) , the 3′ fragment of the T . theileri SSU rRNA gene ( SSU3-PacI-For and 3SSU-Rev-XmaI ) , the trans-splicing addition site from the actin IR sequence ( splice-AvrII-For and splice-FseI-Rev ) , the Bd37 coding sequence ( Bd37-Core-F-FseI or Bd37-Core-F-AvrII and Bd37-Core-F-AvrII and Bd37-Core-R-HindIII ) , the CAT coding sequence ( CAT For FseI or CAT For XhoI and CAT Rev XbaI or CAT Rev AscI ) , the blasticidin resistance cassette ( BSDKpn-F and BSDBgl-R ) , the beta-alpha tubulin IR sequence ( ba-tub-AscI-For or ba-tub-BglII and ba-tub-KpnI-Rev or ba-tub-PacI ) , the T . brucei BiP N-terminal fragment sequence ( BiP-For-FseI or BiP-For-FseI-SpeI and BiP-Rev-XhoI ) , and the GPI addition signal ( GPI-For-HindIII and GPI-Rev-AscI ) . In fusion protein cassettes , care was taken to preserve the reading frame and appropriate start and stop codons were provided . The plasmids containing the expression cassettes were then digested with appropriate restriction enzymes to liberate the plasmid backbone from the linear expression cassette , which was isolated via gel electrophoresis and gel purification . The linear cassette was then purified by ethanol precipitation and resuspended in 5 ml of TE buffer ( 1 mM Tris-HCl ( pH 8 ) and 0 . 1 mM EDTA ) . From a culture of logarithmically growing T . theileri parasites , 10 ml of culture at ∼5×105 cells/ml were used for each transfection . Cells were centrifuged at 1000 × g , for 10 minutes at room temperature and resuspended in 1 ml of sterile PBS to wash the cells . Cells were then re-centrifuged and resuspended in 100 µl Ingenio transfection buffer ( Mirus Bio ) . The cells were added to the prepared linear DNA and transferred to a cuvette for the Nucleofector II electroporation device ( Lonza ) . Transfection was done with Nucleofector program X-001 ( recommended for mouse CD8+ T cells ) and cells transferred to a culture flask containing 10 ml of pre-warmed media and incubated for 24 hrs at 37°C , in 5 % CO2 . Cells were then transferred to media containing the selective drug concentration ( 10 µg/ml of blasticidin ) and plated using a variety of dilutions into a 24-well cell culture plate . After selection clones were recovered under normal cell culture conditions . The localisation of the BiP-Bd37 fusion protein was determined using 2% paraformaldehyde fixed cells permeabilised , or not , with TBS: 0 . 1% Triton X-100 for 2 minutes . Cells were quenched for 30 minutes with TBS:0 . 1% glycine and then blocked for 1 hr with TBS: 1%BSA . Cells were then reacted with anti-BiP antibody ( a gift of Jay Bangs , University of Wisconsin ) diluted 1∶200 in TBS:1%BSA , washed three times with TBS and then incubated with anti-rabbit FITC conjugated antibody ( Sigma ) diluted 1∶100 inTBS:1%BSA . Prior to mounting in MOWIOL , cells were stained with 4′ , 6-diamidino-2-phenylindole for 5 minutes to visualize cellular DNA . Calves ( 6-weeks old ) were injected intravenously with 1×105 T . theileri parasites . Blood samples were taken weekly into EDTA-containing tubes ( Becton-Dickinson ) and DNA extracted . DNA extraction from whole blood samples was performed as follows: 1 ml of blood was mixed thoroughly with 0 . 5 ml of RBC lysis buffer ( 0 . 32 M sucrose , 10 mM Tris-HCl pH 7 . 5 , 5 mM MgCl2 , 0 . 75% Triton X-100 ) in a microfuge tube . The samples were then centrifuged at 14 , 000 rpm for 1 minute to pellet all cells and the supernatant removed . The pellets were repeatedly resuspended and recovered from 0 . 5 ml aliquots of RBC lysis buffer until no red blood cells were present . The resulting pellets were resuspended in 100 µl of lysis buffer ( 50 mM KCl , 10 mM Tris-HCl pH 8 . 3 , 2 . 5 mM MgCl2 , 0 . 1 mg/ml gelatin , 0 . 45% NP40 , 0 . 45% Tween-20 , 60 µg/ml proteinase K ) and kept at 55°C for 60 minutes . The samples were then incubated at 95°C for 10 minutes prior to storage at −20°C until use . Nested PCR reactions were designed to amplify either the T . theileri β-α tubulin intergenic sequence ( to identify any T . theileri population ) or the Bd37 coding sequence ( specific for the vaccine vehicle ) . The primers used were as follows: Tub Diagnostic F1: 5′-AGTAGCAACGACAGCAGCAGT-3′ Tub Diagnostic R1: 5′-GTAAAGTGTTTGAAGAAGAGCTCG-3′ Tub Diagnostic F2: 5′-CGATTCTCTTCGCCTGTTTGT-3′ Tub Diagnostic R2: 5′-ACTAACCGCGACCAAAGAAGT-3′ Bd37 Diagnostic F1: 5′-GCTCACAGGAGCCAGCAGCGG-3′ Bd37 Diagnostic R1: 5′-CCAGAGCTTTGAGATTAGCTGGTA-3′ Bd37 Diagnostic F2: 5′-ACGCAGCAAGGTGGTGCGAA-3′ Bd37 Diagnostic R2: 5′-AGCAAGGCCTCACCGCCCTTGGC-3′ Each 25 µl reaction contained the following components: 5 µl of template , 1X PCR buffer , 0 . 2 mM of each dNTP , 1 . 25 mM MgCl2 , 0 . 4 µM of each primer and 0 . 25 U GoTaq Flexi DNA Polymerase ( Promega ) . The first stage PCR reactions were heated to 95°C for 5 minutes , followed by 35 cycles of denaturation at 95°C for 30 seconds , annealing at 60°C for 45 seconds and elongation at 72°C for 45 seconds . Following the final cycle , the reactions were elongated for a further 4 minutes . The second stage nested PCR reaction was conducted using the same conditions with 5 µl of the first reaction as template . The resulting products were electrophoresed on a 1% Tris-acetic acid-EDTA agarose gel , stained with ethidium bromide and visualized . Serum samples were prepared from 10 ml of fresh blood and stored at −20°C until use . Recombinant ( E . coli ) expressed His-tagged Bd37 antigen was diluted to 5 µg per ml in coating buffer ( 0 . 01 M sodium carbonate pH 9 . 6 ) , and 100 µl were added to each well of microtiter plates and incubated overnight at 37°C . The coating buffer was removed and 200 µl blocking buffer ( 10% horse serum in 10 mM PBS ) were added , and incubated at 37°C for 60 minutes . The plates were washed 3 times with 200 µl washing buffer ( 10 mM PBS , pH 9 . 6 with 1% Tween-20 ) . Bovine serum samples were diluted in blocking buffer as appropriate , and 100 µl were incubated in the coated and incubated at 37°C for 60 minutes . For standard ELISA plates were washed and incubated with HRP-conjugated anti-bovine IgG antibody diluted 1∶1000 in blocking buffer and incubated at 37°C for 60 minutes . For competitive ELISA assays , 100 µl of mouse monoclonal anti-Bd37 antibody ( 1 mg/ml diluted 1∶1000 in blocking buffer ) were added to each well and incubated at 37°C for 60 minutes . Plates were washed and incubated with HRP-conjugated anti-mouse IgG antibody diluted 1∶1000 in blocking buffer and incubated at 37°C for 60 minutes . Plates were washed , treated with TMB supersensitive substrate ( Sigma ) , stopped with 2M sulphuric acid and the OD was measured at 450 nm in an ELISA plate reader . To analyse serum antibody titres , a Kruskal-Wallis one way ANOVA was used .	Single-celled parasites of the order Kinetoplastida are responsible for devastating diseases of humans and animals , including African trypanosomiasis , Chagas' disease and leishmaniasis . However , there are also many species of trypanosomatids that do not cause disease and are distributed globally . One example is Trypanosoma ( Megatrypanum ) theileri , which is restricted to bovids and ubiquitous in cattle herds worldwide . This organism is maintained extracellularly in the blood and tissues long-term without any observed ill effects on host health or productivity . Using knowledge of gene expression and protein trafficking in pathogenic trypanosomatids , we have successfully developed , from first principles , Trypanosoma theileri as a delivery system for vaccine antigens and therapeutics . Procedures for the growth , transfection and heterologous gene expression of T . theileri have been developed , and the delivery of a vaccine antigen derived from Babesia divergens evaluated in vivo . Our results demonstrate the ability of T . theileri to be used as a flexible and easily manipulated protein delivery system suitable for the control of cattle pathogens and cattle-borne zoonoses . In one notable application , we propose that the system could allow the expression of serum trypanolytic factors in cattle , with the potential to alleviate poverty in Africa through the killing of pathogenic trypanosomatids in livestock .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "medicine", "infectious", "diseases", "veterinary", "diseases", "molecular", "cell", "biology", "veterinary", "parasitology", "immunity", "immunology", "biology", "microbiology", "veterinary", "science", "parasitology" ]	2011	Targeting Cattle-Borne Zoonoses and Cattle Pathogens Using a Novel Trypanosomatid-Based Delivery System
Successful infection of the host requires secretion of effector proteins to evade or suppress plant immunity . Secretion of effectors in root-infecting fungal pathogens , however , remains unexplored . We previously reported that Verticillium dahliae , a root-infecting phytopathogenic fungus , develops a penetration peg from a hyphopodium to infect cotton roots . In this study , we report that a septin ring , requiring VdSep5 , partitions the hyphopodium and the invasive hypha and form the specialized fungus-host interface . The mutant strain , VdΔnoxb , in which NADPH oxidase B ( VdNoxB ) is deleted , impaired formation of the septin ring at the hyphal neck , indicating that NADPH oxidases regulate septin ring organization . Using GFP tagging and live cell imaging , we observed that several signal peptide containing secreted proteins showed ring signal accumulation/secretion at the penetration interface surrounding the hyphal neck . Targeted mutation for VdSep5 reduced the delivery rate of secretory proteins to the penetration interface . Blocking the secretory pathway by disrupting the vesicular trafficking factors , VdSec22 and VdSyn8 , or the exocyst subunit , VdExo70 , also arrested delivery of the secreted proteins inside the hyphopodium . Reduced virulence was observed when cotton roots were infected with VdΔsep5 , VdΔsec22 , VdΔsyn8 and VdΔexo70 mutants compared to infection with the isogenic wild-type V592 . Taken together , our data demonstrate that the hyphal neck is an important site for protein secretion during plant root infection , and that the multiple secretory routes are involved in the secretion . Pathogens secrete effector proteins as molecular weapons to evade or suppress plant immunity . Most effectors are small secreted proteins [1 , 2 , 3] , and in many cases , the expression of these genes is induced by infection , helping the microbe to successfully colonize on the surface or inside of the host [1] . Studies of the secretion system have revealed diverse manners for pathogen effector translocation into their host . Fungi secrete different effectors at different infection stages from stage-specific compartments at the host-pathogen interface [4] . Before penetrating host cells , some effector proteins are focally secreted from appressorial penetration pores and may function to suppress early plant defense responses , as in case of several Colletotrichum species [4 , 5] . After invasive hypha developed , effectors in Colletotrichum orbiculare accumulate at the pathogen-plant biotrophic interface , a ring-like region around the neck of the biotrophic primary hypha [5] . Effectors of some plant fungal pathogens are putatively translocated into the host cell , where they interact with cytoplasmic or nuclear R proteins [6] . For example , Magnaporthe oryzae has a highly localized structure to accumulate cytoplasmic effectors secreted by invasive hyphae , known as the biotrophic interfacial complex ( BIC ) , which forms at the tip of the initially filamentous hypha in the host cell [7 , 8] . Secretion of effectors to the host is also important for soil-borne fungal and oomycete pathogens , such as Verticillium dahliae and Phytophthora sojae , for successful infection [9 , 10] . Transit of many oomycete or fungal effectors to host cell depends only on the RXLR motif or other host-entry motifs of the effectors and host molecules , but not pathogen-encoded machinery [11 , 12] . However , however , the mechanism ( s ) by which root-infecting fungal pathogens secrete secretory protein remains unknown . V . dahliae causes vascular wilt disease [13 , 14] and infects more than 200 host species worldwide , including many economically important crops , such as lettuce , cotton and tomato [15] . V . dahliae contains more than 100 small cysteine-rich potentially secreted proteins [3] . So far , only two effectors , Ave1 and Vdlsc1 , have been functionally studied . Ave1 contributes to fungal virulence in the absence of its corresponding R protein ( Ve1 ) [9]; it is a small secreted protein containing 134 amino acids ( aa ) with four cysteines [9] , however , its secretion process has not been studied . Vdlsc1 suppresses salicylate-mediated innate immunity in planta [10] . Vdlsc1 is an unconventionally secreted protein as it lacks an N-terminal signal peptide that can direct the protein to the conventional secretory pathway [10] . Therefore , whether the soil-borne V . dahliae , a root-infecting phytopathogenic fungus , have a highly localized structure to secret secretory effector proteins remains unexplored . We recently identified and provided the molecular features of the infectious structure , the hyphopodium , in V . dahliae [16] . We demonstrate that V . dahliae NADPH oxidase B ( VdNoxB ) is required for local reactive oxygen species ( ROS ) production during infection , and ROS-Ca2+ signaling in the hyphopodium plays key roles in regulating polarized penetration peg formation and pathogenicity in V . dahliae [16] . In M . oryzae , the Nox2 ( NoxB ) -NoxR complex spatially organizes a heteroligomeric septin ring at the appressorium pore [17 , 18] . Septins , small morphogenetic guanosine triphosphatases ( GTPases ) , are thought to reorient and reorganize the cytoskeleton to determine cell shape [19] . Septin ring scaffolds a toroidal F-actin ring and recruits and organizes the exocyst to the appressorium pore where the penetration peg emerges [17 , 18 , 20] . In this study , to explore whether the secretion of effectors in V . dahliae could be associated with the penetration peg , we first verified that VdNoxB was required for the cytoskeletal organization of a septin ring at the penetration peg and its derived hyphal neck in V . dahliae . We observed that the septin-ring-organized hyphal neck acts as a functional fungus-host penetration interface for the delivery and secretion of signal peptide-containing secretory proteins . Using gene functional analyses , we further showed that VdSep5 , the conventional fungal ER-to-Golgi secretion pathway , the endosome-mediated transport , and the exocyst complex are involved in the delivery of secretory proteins to the penetration interface . Verticillium dahliae infection requires the development of an infectious structure , hyphopodium , in which the NADPH oxidase catalytic subunit VdNoxB is specifically expressed to regulate formation of the penetration peg to pierce the cell wall [16] . To further understand the infection process , we used FITC-WGA ( FITC-conjugated wheat germ agglutinin ) to label the fungal hyphae of wild-type V . dahliae strain V592 to assess cellophane penetration and root infection . We observed repeated development of hyphopodium for penetration inside the cellophane and roots ( Fig 1A and 1B , S1 and S2 Movies ) . Transmission electron microscopy ( TEM ) images showed that after penetration , the invasive hypha grew , and the penetration peg acted as a hyphal neck that partitioned the hyphopodium and invasive hypha and came into close contact with the cellophane ( Fig 1C ) or the host ( Fig 1D ) , forming the fungus-host penetration interface . WGA staining showed various uneven hyphal neck lengths after piercing the cellophane or root cell wall ( Fig 1A and 1B ) . The length of the hyphal neck reflects the length of the penetration peg and might be dependent on the piercing sites of different cells , such as the root epidermis or cortical cells , and the piercing angle . Repeated development of hyphopodium inside the cellophane was observed using V592 expressing GFP-tagged VdNoxB under the native promoter ( S2A Fig ) [16] . Inside the cellophane , the GFP signal was observed in flattened irregular hyphopodia and at the tips of the penetration pegs ( S2B Fig ) . The VdΔnoxb mutant , in which VdNoxB was deleted , was incapable of producing a penetration peg ( S2C Fig ) . Together , our results demonstrate that V . dahliae infection requires VdNoxB-dependent , repeated development of the hyphopodium and penetration peg for each cell wall penetration , and the penetration peg-derived hyphal neck connects the hyphopodium to the invasive hypha and marks a site of close fungus-host penetration interface contact . To explore the specific features of the penetration peg and its derived hyphal neck , we first examined whether the cytoskeleton protein septin plays a role in determining penetration peg morphogenesis in V . dahliae . The V . dahliae homolog of Septin5 was identified ( S1B Fig ) and named VdSep5 . VdSep5-GFP was expressed in the V592 and VdΔnoxb mutant . In the wild-type V592 hyphopodium during penetration peg induction , confocal laser scanning microscopy ( CLSM ) observation revealed a continuous funnel-shaped VdSep5-GFP fluorescent signal from the base of the hyphopodium , outlining the curved contact area between the hyphopodium and cellophane membrane ( Fig 2A , plane 1 . 5 μm ) , to the central protruded zone ( refer to the hyphopodium pore ) where the penetration peg was initially developed and spread throughout to its tip ( Fig 2A , from plane 2 . 7 to 4 . 5 μm ) . In contrast , in the VdΔnoxb hyphopodium , the VdSep5-GFP signal outlined the curved contact area between the hyphopodium and the cellophane membrane but without the central septin ring at the base of the hyphopodium ( Fig 2B ) . In wild-type V592 , compared with the widest part of the VdSep5-GFP signal area ( Fig 2A , plane 1 . 5 μm ) , the diameter of the VdSep5-GFP ring at the hyphopodium pore ( Fig 2A , plane 2 . 7 μm ) was reduced approximately 58% , while there was no clear reduction of the diameter of the VdSep5-GFP signal area in the VdΔnoxb mutant , which is deficient in penetration peg formation ( Fig 2C ) . After cellophane piercing and invasive hyphal growth , we observed the compact septin ring signal retained in the hyphal neck ( Fig 2D ) . These results demonstrate that VdNoxB is required for VdSep5 organization of the cytoskeleton to determine the morphogenesis of the penetration peg and its derived hyphal neck . After infecting Arabidopsis root with wild-type V . dahliae V592 , a VdSep5-GFP ring was also observed at the hyphal neck partitioning the hyphopodium and invasive hyphae ( Fig 2E ) . Two VdSep5-GFP rings were observed in the first and second hyphal necks in two CLSM planes within the same scanning view ( Fig 2F , 3 . 6 μm for the first penetration and 13 . 2 μm for the second penetration ) , verifying the requirement of multiple penetrations for each new cell wall to reach the vascular bundle . Together , our results clearly demonstrate that septin-ring organization accompanies V . dahliae penetration of either cellophane or plant roots , and VdNoxB plays a role in the organization of the septin ring at the penetration peg and its derived hyphal neck . Remarkably , once established , the VdSep5-GFP ring was retained at and framed the hyphal neck , forming the fungus-host penetration interface . Consistent with previous findings that septin scaffolds a toroidal F-actin ring at the appressorium pore in M . oryzae [17 , 18] , we also observed that F-actin was organized as a ring structure at the hyphal neck in either cellophane or root by live-cell imaging of V592 expressing LifeAct-GFP ( S3 Fig ) . Next , we investigated whether the septin-ring-organized hyphal neck , in addition to its piercing role , could act as a functional fungus-host penetration interface for the delivery of secretory proteins . The arsenal of potentially secreted proteins in plant pathogens includes key pathogenicity molecules that are generally referred to as effectors ( small cysteine-rich proteins , <400 amino acids ( aa ) and ≥4 cysteine residues ) [3] . Because the well-known Ave1 secreted effector has not been identified in the cotton isolate V592 , we selected three small cysteine-rich proteins ( SCP ) , VDAG_08085 ( 194 aa , 6 Cys , named VdSCP8 ) , VDAG_00902 ( 375 aa , 16 Cys , VdSCP9 ) and VDAG_05717 ( 205aa , 4 Cys , VdSCP10 ) , for analysis in this study . Each of these SCPs has an N-terminal signal peptide predicted by the SignalP 4 . 1 server ( S4A Fig ) [21] . Among them , VdSCP9 is a LysM domain-containing protein . The LysM effector family contains relatively conserved secretory proteins that are known to play significant roles in the pathogen-host interaction [3 , 22 , 23 , 24] . VdSCP8 was identified by liquid chromatography-mass spectrometry ( LC-MS ) of the V592 culture filtrate , and VdSCP10 was one of the potential pathogenicity genes in our previous screening of the T-DNA insertional mutant library [25] and confirmed by the targeted gene replacement mutant of VdSCP10 ( S5 Fig ) . Transcript levels of these SCP genes were first examined using quantitative RT-PCR ( qRT-PCR ) . The expression levels of VdSCP9 and VdSCP10 , but not VdSCP8 , were significantly up-regulated at 4 days post-inoculation ( dpi ) of V592 on cellophane and at 2 dpi on cotton roots ( S4B Fig ) . To observe the localization of these SCPs , VdSCP8-GFP , VdSCP9-GFP and VdSCP10-GFP were expressed under the native promoter . Only VdSCP8-GFP fluorescence was detectable as a ring signal at the penetration zone on cellophane ( S4C Fig ) . Neither VdSCP9-GFP nor VdSCP10-GFP fluorescence was observed on cellophane . Therefore , these GFP fusion proteins were constructed under the oliC promoter . The V . dahliae small effector VdIsc1 ( 190 aa , 1 Cys ) , which lacks a signal peptide and exhibits characteristics that lead to unconventional secretion [10] , was also fused to GFP as a control . After V . dahliae invasion into cellophane , VdSCP8-GFP , VdSCP9-GFP and VdSCP10-GFP , but not the control VdIsc1-GFP , showed ring signals surrounding the penetration zones ( Fig 3A ) . From a picture of the penetration at an incline , the VdSCP10-GFP ring signal was clearly observed surrounding the hyphal neck linking the hyphopodium to the invasive hypha ( Fig 3B ) . We also fused the signal peptides of SCPs to GFP and found that SPVdSCP8-GFP , SPVdSCP9-GFP and SPVdSCP10-GFP also showed ring signals outside the plasma membrane of the penetration zones ( S4D Fig ) . To detect whether the SCP-GFP signal rings overlapping with the septin ring , VdSep5-RFP was co-expressed with VdSCP8-GFP in V592 , and the results showed that the VdSep5-RFP ring was inside the VdSCP8-GFP ring ( Fig 3C ) . Similar results were obtained for VdSCP9-GFP and VdSCP10-GFP , demonstrating that signal peptide-containing SCPs accumulate and/or are delivered to the hyphal neck for secretion . Together , these data suggest that the hyphal neck made up a fungus-host penetration interface for the delivery and/or exportation of secretory proteins . Next , we inoculated strains of V592 expressing the GFP-tagged SCPs on Arabidopsis roots . VdSCP8-GFP accumulation was first observed at approximately 2 dpi . A strong VdSCP8-GFP signal ring was observed at the hyphal neck , which partitioned the hyphopodium and the invasive hypha ( Fig 4A and 4B ) . Some weak signals were also observed inside the invasive hyphae ( Fig 4A and 4B ) . VdSCP9-GFP accumulation was first observed at 1 dpi . One weak and one strong VdSCP9-GFP signal ringwas observed , respectively , at the first and second hyphal neck in two CLSM planes in the same scanning view ( Fig 4C at planes of 1 . 8 μm and 4 . 8 μm ) . This observation is consistent with the requirement of the repeated development of the hyphopodium for each cell wall penetration during the colonization of V592 from the root surface to the vascular bundle ( Fig 1B ) . The VdSCP10-GFP ring signal was observed at approximately 6 dpi . In the upper plane of the hyphopodium ( 0 μm ) , a weak VdSCP10-GFP speckle signal was observed at the periphery of the hyphal cell and on two sides of a septa ( Fig 4D ) . A stronger signal was observed at the apex of the hyphopodium ( Fig 4D ) . The clear and strongest VdSCP10-GFP ring signal was observed at and throughout the hyphal neck using a series of continuous scanning planes ( Fig 4D ) . Taken together , our data demonstrate that successful invasion of plant roots and cellophane by V . dahliae has the common phenomenon of signal peptide-containing secretory protein recruitment at the hyphal neck for effective secretion through this fungus-host penetration interface . The detectable ring signals for each secretory protein at different time points suggest that their synthesis and/or rate of delivery were different , revealing a complex process for successful infection in plant roots by V . dahliae . We then examined whether the directional ring-shaped accumulation of the small secretory proteins was derived from dynamic secretion towards the penetration interface . Fluorescence recovery after photobleaching ( FRAP ) was performed with the V592 strain expressing VdSCP10-GFP on cellophane at 6 dpi . We photobleached VdSCP10-GFP fluorescence at the penetration site and then monitored the fluorescence recovery over time . After near complete elimination , fluorescence recovered within 27 min ( Fig 3D and 3E ) . The fluorescence recovery time for VdSCP8-GFP and VdSCP9-GFP was 22 min and 18 min , respectively ( Fig 3F ) . These data suggest that secretory proteins were continuously synthesized and/or delivered to the penetration interface . To detect whether septin also plays a role in the delivery of secretory proteins to penetration interfaces , the targeted gene knockout mutants VdΔsep5 and VdΔsep3 were generated in wild-type V592 ( S6A and S6B Fig ) . The VdSep3 homologous sequence was searched from V592 based on a BLASTP search using MoSep3 and the VdLs . 17 database ( S1A Fig ) . The VdSep3 knockout mutant strain exhibited a reduced hyphal growth rate on PDA medium compared with V592 ( S6D Fig ) , and developed an abnormal hyphopodium on cellophane without smooth swelling at the end of branching hypha ( S6F Fig ) that was incapable of forming a penetration peg to pierce the cellophane ( S6E Fig ) . This result demonstrates that VdSep3 plays roles in hyphal growth and hyphopodium development . In contrast , the VdSep5 knockout mutant strain exhibited a normal growth rate on PDA medium ( S6D Fig ) but developed fewer hyphopodia on cellophane ( S6F and S6G Fig ) and displayed greatly delayed penetration of the cellophane compared with V592 ( S6E Fig ) , demonstrating that VdSep5 plays a role in hyphopodium development . Consistently , both the VdΔsep3 and the VdΔsep5 mutant showed reduced virulence on cotton plants ( S6H and S6I Fig ) . The reintroduction of Psep3:VdSep3:Ttrpc and Ptef:VdSep5-GFP:Ttrpc restored the hyphal morphologies and cellophane penetration abilities , as well as the pathogenicity ( S6D , S6E , S6H and S6I Fig ) , confirming the targeted gene deletion . Our results suggest that VdSep5 plays an important role in the initiation of hyphopodium formation , whereas , VdSep3 is more important for development of the hyphopodium . To observe the localization of secretory protein in the VdΔsep5 mutant , the targeted gene knockout mutant was generated in VdSCP10-GFP-expressing V592 strain . VdSCP10-GFP secretion in the VdSep5 deletion mutant was assessed . In contrast to the remarkable VdSCP10-GFP signal ring surrounding the hyphal neck in the wild-type V592 ( Fig 5A ) , the VdSCP10-GFP signal was observed in both the hyphopodium and hyphal neck in the VdΔsep5 mutant , in which either the hyphopodium or the hyphal neck was stained with FM4-64 ( Fig 5A and 5B ) . The VdSCP10-GFP signal at the hyphal neck in the VdΔsep5 mutant was clearly reduced compared with that in the wild-type V592 background ( Fig 5B ) . The average signal intensity of VdSCP10-GFP in the hyphal neck of VdΔsep5 was approximately 78% of that in V592 ( Fig 5C ) . The FRAP assay showed 86% recovery of VdSCP10-GFP fluorescence at the penetration interface within 97 min ( Fig 5D and 5E ) , which was significantly longer than the recovery time of 27 min for the wild-type V592 . Three FRAP tests on cellophane showed a similar delayed in secretion . These results demonstrate that VdSep5 plays a role in mediating the delivery of secretory proteins to the penetration interface , in addition to its functions in hyphopodium development and cortical structure organization of the penetration peg and hyphal neck . We next investigated the role of vesicular traffic in the delivery of secretory proteins to penetration interfaces . SNAREs function as key elements in membrane fusion [26 , 27 , 28] . The R-SNARE Sec22 is important for modulating transport between the ER and the Golgi apparatus [29] . The Qc-SNARE Syn8 in S . cerevisiae and M . oryzae ( MoSyn8 ) localizes at endosomes and/or late endosome/prevacuolar compartments ( PVCs ) [26 , 28] . To identify functional proteins in the secretion of V . dahliae , homologous sequences were searched in V592 based on a BLASTP search using MoSec22 and MoSyn8 and the database for VdLs . 17 , designated VdSec22 and VdSyn8 , respectively ( S1C and S1D Fig ) . Targeted gene knockout mutants VdΔsec22 and VdΔsyn8 were generated ( S7A and S7B Fig ) . Both mutants exhibited growth defects with a reduced vegetative hyphal growth rate; VdΔsyn8 also showed reduced melanin production ( S7D Fig ) . The reintroduction of VdSec22 and VdSyn8 under the control of each native promoter recovered the growth ability and hyphal morphologies ( S7C and S7D Fig ) , confirming the targeted gene deletion of VdSec22 and VdSyn8 . VdSCP10-GFP secretion in the VdΔsec22 and VdΔsyn8 mutants was assessed . VdSCP10-GFP expressed under the oliC promoter was transformed into VdΔsec22 and VdΔsyn8 mutants . The single copy insertion strains determined by Southern blot were used for further analysis ( S8 Fig ) . In contrast to the remarkable VdSCP10-GFP signal ring in wild-type V592 ( Fig 6A ) , the VdSCP10-GFP signal was observed in both the hyphopodia and hyphal necks in both deletion mutant strains ( Fig 6B and 6C ) . The VdSCP10-GFP signal in the hyphal neck in both mutants was also clearly reduced , and most of the VdSCP10-GFP signal rings were overlapping with or inside the plasma membrane compared with that in the wild-type V592 background ( Fig 6B and 6C ) . The average signal intensity of VdSCP10-GFP in the hyphal neck of VdΔsec22 and VdΔSyn8 was approximately 54% and 70% , respectively , of that in V592 ( Fig 6E ) , suggesting that ER-Golgi transport is a predominant route of transport of SCPs . The FRAP assay was also performed with VdSCP10-GFP-expressing VdΔsec22 and VdΔsyn8 mutants . Fluorescence was recovered after approximately 52 min and 45 min in VdΔsec22 and VdΔsyn8 mutants , respectively ( Fig 6F ) , which was much longer than the recovery time of 27 min observed for wild-type V592 . Our data demonstrate that VdSec22 and VdSyn8 play roles in mediating the delivery of secretory proteins to the penetration interface . To further determine whether the VdSCP10-GFP signal in the hyphopodium and hyphal neck was due to decreased transport from the ER to the Golgi apparatus in the mutant strains , a VdSCP10-GFP-expressing VdΔsec22 mutant on cellophane was stained with ER-Tracker Blue-White DPX . The VdSCP10-GFP signal was observed to overlap with the ER in the hyphopodium ( S9 Fig ) , suggesting that the deletion of VdSec22 resulted in retention of VdSCP10-GFP in the ER . Taken together , our data demonstrate that the transport route between the hyphal ER and Golgi apparatus and endosome-mediated transport are involved in protein secretion toward penetration interfaces . The exocyst was discovered as a tethering complex that mediates the initial encounter of arriving exocytic vesicles with the plasma membrane [30] . The exocyst complex is an evolutionarily conserved doctameric protein complex comprising Sec3 , Sec5 , Sec6 , Sec8 , Sec10 , Sec15 , Exo70 , and Exo84 [31 , 32] . To test the role of the exocyst in the accumulation of small secretory proteins at the penetration interface in V . dahliae , two predicted exocyst components , VdSec8 and VdExo70 , were identified ( S1E and S1F Fig ) . VdSec8-GFP and VdExo70-GFP were expressed under either their native promoter or the oliC promoter and introduced into V592 . Similar localization profiles were observed for both GFP-tagged proteins under either the native or the oliC promoter; however , the GFP signal derived from the native promoter was weak , and thus the fluorescence signals derived from the oliC promoter were photographed . Both VdSec8-GFP and VdExo70-GFP were observed as a crescent structure at the growing tip of vegetative hyphae ( S10A Fig ) . VdSec8-GFP and VdExo70-GFP were organized as a ring at the base of the hyphopodium that was observed before penetration peg formation on either cellophane or Arabidopsis root ( S10B and S10C Fig ) . After the development of invasive hyphae , VdSec8-GFP was organized at the hyphal neck on either cellophane or root ( S10D Fig ) . Together , our data demonstrate that the exocyst is active at the base of the hyphopodium and the hyphal neck . To characterize the localization relationship between the exocyst complex and VdSep5 , VdSep5-RFP was transformed into the V592-expressing Polic:VdSec8-GFP:Ttrpc strain . Red septin rings were observed in all 20 observed hyphal necks , and VdSec8-GFP signal rings were observed in 14 of the detected septin rings . The corresponding linescan confirmed the co-localization of VdSec8-GFP and VdSep5-RFP ( S11 Fig ) . To further study the role of exocyst subunits on secretory protein accumulation at the penetration interface , we tried to knockout VdExo70 and VdSec8 in V592 . VdΔexo70 mutants carrying the VdExo70 deletion were obtained ( S7B Fig ) , but the deletion of VdSec8 was not successful , in agreement with a previous study in which M . oryzae exocyst-encoding gene knockouts generated only Δsec5 and Δexo70 mutants [20 , 33] . Thus , the failure to delete VdSec8 was possibly due to the lethality of the absence of Sec8 in filamentous fungi including V . dahliae and M . oryzae . The VdΔexo70 mutant exhibited growth defects with a low growth rate on PDA medium ( S7D Fig ) . The reintroduction of Polic:VdExo70-GFP:Ttrpc into the VdΔexo70 mutant recovered the growth ability and hyphal morphologies ( S7C and S7D Fig ) , confirming the targeted disruption of VdExo70 . VdSCP10-GFP was then transformed into the VdΔexo70 mutant and incubated on cellophane for hyphopodium induction . The single copy insertion strains determined by Southern blot were used for further analysis ( S8 Fig ) . The VdSCP10-GFP signal was observed inside of the hyphopodium in the VdΔexo70 mutant ( Fig 6D ) . Weak signals were observed in the hyphal neck , but most of them overlapped with the FM4-64-stained plasma membrane ring ( Fig 6D ) . The average intensity of the green fluorescence ring of VdSCP10-GFP in the VdΔexo70 mutant was approximately 65% of that in V592 ( Fig 6E ) . These results demonstrate that VdExo70 plays a role in secreting VdSCP10-GFP out of the hyphal neck . The FRAP assay on cellophane was also performed with VdSCP10-GFP in the VdΔexo70 mutant . The fluorescence recovered after 63 min ( Fig 6F ) , which was significantly longer than the recovery time of 27 min determined for wild-type V592 . Taken together , our results demonstrate that exocyst components also organize at the hyphal neck and take part in the delivery of secretory proteins to penetration interfaces . To explore the roles of proteins involved in the secretion pathway in the pathogenicity of V . dahliae , we inoculated VdΔexo70 , VdΔsec22 and VdΔsyn8 mutants on cotton plants and found a significant ( P< 0 . 05 ) reduction in the disease index for the three mutants ( Fig 7A and 7B ) . The loss of virulence was restored when the VdΔexo70 , VdΔsec22 and VdΔsyn8 mutants were complemented with Polic:VdExo70-GFP:Ttrpc , Psec22:VdSec22:Ttrpc or Psyn8:VdSyn8:Ttrpc , respectively ( Fig 7A and 7B ) . The significant loss of pathogenicity for the VdΔexo70 , VdΔsec22 and VdΔsyn8 mutants was presumably consistent with their inefficient secretion of effector-related secretory proteins , which are required for successful fungal pathogen infection by evading or suppressing host plant immunity . Therefore , attributed to critical roles in the efficient secretion of secretory proteins at fungus-host penetration interfaces , VdExo70 , VdSec22 and VdSyn8 play important roles in the pathogenicity of V . dahliae . Penetration of the intact cuticles of the host is a very important step for successful infection by phytopathogens , either for leaf- or root-infecting fungal pathogens , such as M . oryzae and the anthracnose disease-causing Colletotrichum species or V . dahliae [4 , 5 , 16 , 17 , 34] . The foliar fungal pathogen M . oryzae forms conspicuous melanized appressoria with an average diameter of 8 . 0 μm when it inflates to full turgor and develops penetration pegs with an average diameter of 780-nm to breach the hydrophobic , waxy leaf cuticle [18 , 35] . In contrast , we found herein that the root-infecting fungus V . dahliae developed hyphopodia with an average diameter of 3 . 4 μm and formed a penetration peg with an average diameter of 1 . 3-μm , suggesting that less pressure is needed for V . dahliae to breach the root cuticle ( Fig 1C and 1D ) . We observed that the nature of the interface between V . dahliae hyphae and the host is the penetration peg-derived hyphal neck , in which a septin ring was organized . In M . oryzae , septins are found to provide the cortical rigidity and membrane curvature necessary for protrusion of the rigid penetration peg to breach the leaf surface [18] . Similarly , we found that the septin ring framed a recognizable cytoskeletal region of the hyphal neck in which F-actin was also organized as a ring structure , partitioning the hyphopodium and invasive hypha on both cellophane and roots . On cellophane , we also observed a funnel-shaped septin structure prior to invasive hyphal growth , suggesting that V . dahliae septins also function in the membrane curvature necessary for protrusion of the penetration peg at the base of the hyphopodium . In the VdNoxB knock out mutant , the VdSep5-GFP signal at the base of the hyphopodium suggests that septins provide membrane curvature , but the mutant strain failed to show protrusion of the penetration peg in the absence of VdNoxB . Previous studies in yeast and in fungal pathogen Aspergillus fumigatus suggest the importance of septin phosphorylation/dephosphorylation in controlling septin assembly [36 , 37] . In yeast , Rts1 , a protein phosphatase 2A ( PP2A ) subunit , regulates septin dephosphorylation during telophase , and this dephosphorylation contributes to cytokinesis [36] . Dephosphorylation of the core septin , AspB , in a PP2A-dependent manner also impacts hyphal septation in A . fumigatus [37] . In animals , PP2A is a well-known tumor suppressor . ROS accumulation in cancer cells causes nitration and inactivation of PP2A , which interferes with the interaction of Bcl-2 with the PP2A catalytic core , leading to increased phosphorylation and antiapoptotic activity of Bcl-2 [38] . We recently reported that V . dahliae VdNoxB is required for local ROS production during infection and plays key roles in regulating polarized penetration peg formation [16] . Together with the regulated synthesis of ROS by M . oryzae Nox complexes directly control septin and F-actin dynamics [17] , and the septin ring assembles in a kinase Chm1-dependent manner [18] , we speculate that fungal Nox-dependent ROS might also play a role in inactivation of PP2A-like phosphatase , leading to increased Chm1-dependent septin phosphorylation , which is key for controlling septin assembly . We speculate that septins also provide membrane curvature for polarity determination during penetration peg development on roots , although funnel-shaped septin signal was barely observed in the hyphopodium-penetration peg on the infected root , presumably due to a fast piercing process on the roots . The targeted gene deletions of VdSep5 or VdSep3 exhibited defects in hyphopodium and/or hyphal development , suggesting that core V . dahliae septins also act cooperatively to form heteroligomers during hyphal growth and infection . This result is consistent with previous observation in M . oryzae that septins formed rings at the neck of nascent appressoria and a wider range of structures in hyphae and during invasive growth , including bars , gauzes , collars and rings [18] , in addition to an appressorium pore-located large septin ring [18] . Nevertheless , our data demonstrate the requirement for VdNoxB-dependent ROS in the regulation of cytoskeleton septin ring remodeling at the base of the hyphopodium , leading to rapid polarized growth of the penetration peg in V . dahliae . Each occurrence of penetration requires septin ring organization at the penetration peg and hyphal neck , supporting that successful colonization of extracellular hyphae to the vascular bundle requires repeated development of the hyphopodium and penetration peg , which repeatedly form penetration interfaces between V . dahliae hyphae and the host . Plant infection by pathogens involves the deployment of effector proteins that suppress plant immune responses and facilitate proliferation of the pathogen within plant tissues [30 , 35] . The delivery of effectors has been shown by extra-invasive hyphal membrane ( EIHM ) and BIC in the first-differentiated bulbous invasive hyphae in M . grisea [7 , 33] . In C . higginsianum , sequential delivery of host-induced effectors by the appressorium pore and intracellular hyphae has been observed [4] . In C . orbiculare , the accumulation of effectors occurred in the ring-like region around the neck linking the penetration peg to the biotrophic primary hyphae [5] . In this study , we found that during the penetration of cellophane or plant roots , the tested SCP-GFP and SP-GFP accumulated on the penetration interfaces , indicating a general role of the penetration interface as an active secretory protein delivery zone in V . dahliae . The FRAP assay revealed the dynamic accumulation of SCPs at the penetration interfaces . Secretion of the three SCPs into the hyphal neck is likely not dependent on the biological host . However , the cellophane membrane was used to mimic the hydrophobic niche for induction of appressoria in M . grisea [39] and hyphopodia in V . dahliae [16] . Together with the identification of VdSCP8 by LC-MS of the V592 culture filtrate without any treatment , and transcripts of VdSCP9 and VdSCP10 , but not VdSCP8 , were induced upon incubation of V . dahliae on both cellophane and roots , we speculate that both VdSCP9 and VdSCP10 are probably in planta-expressed secretory proteins in V . dahliae . Moreover , the targeted gene deletion of VdSCP10 caused a significant decrease in virulence toward cotton plants , suggesting that VdSCP10 may function as an effector to suppress plant immune responses . Although the LysM effector family contains relatively conserved secretory proteins that are known to play significant roles in the pathogen-host interaction [3 , 22 , 23] , it has been recently reported that deletion of the VdSCP9 homologous core LysM protein , Vd4LysM , in V . dahliae strain JR2 , did not compromise virulence during infection in Arabidopsis , tomato or Nicotiana benthamiana [24] . Whether VdSCP9 and VdSCP8 function as effectors to suppress plant immune responses or facilitate proliferation of V . dahliae within plant tissues requires further investigation . Remarkably , the ring signals of the tested SCPs were outside and around the hyphal neck and septin ring ( Fig 3 ) , and they were reduced in the hyphal neck in the VdΔsep5 mutant . These observations indicate that septins are not only required to organize the hyphal neck to form a fungus-host interface but also participate in the delivery and exportation of secretory proteins . Phytopathogenic fungi express numerous small proteins that possess classical N-terminal signal peptides that direct them to the endoplasmic reticulum ( ER ) [3 , 4 , 40] The three signal peptide-containing SCPs , but not the unconventional secretion protein VdIscI , accumulated around the hyphal neck , suggesting that secretion into the penetration interface depends on ER processing . The retention of VdSCP10-GFP in the ER of the VdΔsec22 mutant demonstrates the importance of transport between the ER and the Golgi apparatus in secretory protein delivery to penetration interfaces in V . dahliae . In M . oryzae during the invasion of rice cells , ER-to-Golgi trafficking is involved in the secretion of apoplastic effectors by EIHM [33] . The Δsec22 mutants of C . orbiculare also show a decreased accumulation of effectors at biotrophic interfaces [5] . Similar to M . oryzae and C . orbiculare [5 , 41] , the absence of Sec22 weakens the virulence of V . dahliae ( Fig 7 ) , suggesting that conventional ER-Golgi transport has a conserved function in the transport of some pathogen secretory proteins to interact with host molecules . Endosomes participate in endocytosis and secretion during fungal infection in the host [28 , 42] . The long-distance retrograde motility of early endosomes is necessary to perceive plant cues and trigger the transcription of effector-coding genes during plant infection by the pathogenic fungus Ustilago maydis , which regulates effector production and secretion during host cell invasion [43] . Syn8 in M . oryzae is involved in the secretion of BIC-localized AVR proteins but not the apoplastic effector ( Bas4 ) in planta [28] . In VdΔsyn8 mutants , the retention of VdSCP10-GFP in the hyphopodium and inside the hyphal neck , suggests that the induction of SCP delivery in V . dahliae requires cues from the fungus-contacting surface and that VdΔsyn8 mutants prevent the perception of information from the contact surface , and resulting in VdSCP10-GFP retention . Thus , effective delivery of secretory proteins during fungal infection in the host requires Syn8-mediated transport/cue-sensing via endosomes . The final steps of the secretory pathway , which occur in the vicinity of the plasma membrane , are regulated by an array of small GTPases , the exocyst tethering complex , and SNARE proteins [20 , 30 , 44] . Co-localization of VdSec8-GFP and VdSep5-RFP and the absence of VdSep5 or VdExo70 to impair the delivery of secretory proteins to the penetration interface also support septin-dependent assembly of the exocyst in V . dahliae . Together with previous reports demonstrating that cytoplasmic effector accumulation in BICs of M . oryzae also required the exocyst components Exo70 and Sec5 [33] , we assume that the effective delivery of secretory proteins during infection of plant hosts requires the exocyst coupled with SNARE proteins , such as Sec22 and Syn8 , to tether vesicles loaded with secretory proteins to the plasma membrane . In summary , we provide evidence that hyphopodium-specific VdNoxB -regulated penetration peg formation accompanied by cytoskeletal organization of the septin-ring , form a fungus-host interface that functions as a site for the dynamic delivery of secretory proteins . The exocyst , VdSec22-mediated transport between the ER and Golgi apparatus and VdSyn8-mediated transport/cue-sensing via endosomes are involved in the secretion of secretory proteins , possibly including effectors , towards the interfaces ( Fig 8 ) . We assume that the fungal infectious structures function as key signaling hubs during plant infection and are the apparatus that not only breaches host cells but also generates unique interfaces for the secretion of fungal secretory proteins and associated regulatory components . The virulent defoliating V . dahliae isolate V592 from cotton that originated in Xinjiang , China , was used in this study . This isolate and its transformants were stored at –80°C and cultures were reactivated on potato dextrose agar ( PDA ) medium at 25°C in the dark . The conidia for the infection assays were cultured in liquid Czapek-Dox medium . Hyphae for microscopic observation were incubated on M0 medium with urea modified as NaNO3 [45] . For plant infection , cotton plants ( ‘Xinluzao No . 16’ ) were used in infection assays to evaluate the effect of V . dahliae isolate V592 and transformants on virulence using our laboratory’s unimpaired root-dip inoculation method , as described in our previous research [25] . Disease progression was recorded after 3 weeks of incubation . The infection assay for transformants was repeated three times . The symptoms were evaluated , and the disease grade was classified as follows: 0 ( no symptoms ) , 1 ( 0–25% wilted leaves ) , 2 ( 25–50% ) , 3 ( 50–75% ) and 4 ( 75–100% ) [10] . The data were analyzed using the Student’s t-test . Nucleic acid extraction and fungal transformation have all been previously described [25] . Single copy insertion was confirmed in transformants which were used to analyze the fluorescence intensity . To generate the knockout plasmids pKOVdSCP10 , pKOVdSep3 , pKOVdSep5 , pKOVdSec22 , pKOVdSyn8 and pKOVdExo70 , upstream and downstream genomic sequences were amplified with the primers shown in S1 Table . The upstream and downstream genomic sequence pairs were inserted into a position flanking the hygromycin resistance cassette of vector pGKO-HPT with the USER enzyme to generate knock-out plasmids , and transformation was performed as previously described [46] . All of the GFP fusion constructs or RFP fusion constructs ( next part ) were generated by the infusion cloning method based on homologous recombination using the ClonExpress II or ClonExpress MultiS kit ( Vazyme , China ) . The primers are shown in S1 Table . In each case , the primers contain a 15-20-bp overlap with adjoining fragments to allow the assembly of fragments by homologous recombination . To select transformants with G418 , the pSUL-NEO binary vector was created by insertion of the G418 resistance cassette amplified with Neo-F/R primers from pKOV21 into XbaI-digested pSULPH-GFP [25] . For convenient expression of the GFP fusion protein under the constitutive Tef promotor and TrpC terminator , we generated a binary vector pSUL-NEO-Tef-EKGFP-TrpC using the following steps: ( 1 ) a 0 . 7-kb 3GA-EGFP fragment , amplified with primers 3GAGFP-F ( bearing three repeats of nucleotides encoding‘GA’ ) and GFP-Nt-R from pNPP9 [47] , was cloned into the EcoRI/NotI sites of pNPP94 [48] , resulting in pNPP94-3GAGFP; ( 2 ) we amplified the Ptef:3GAGFP:Ttrpc fusion from pNPP94-3GAGFP using primers psul-ppn-HindIII-F and psul-ppn-EcoRI-R and recombined with the product HindIII/EcoRI-linearized pSUL-NEO to generate plasmid pSUL-NEO-Tef-3GAGFP-TrpC; ( 3 ) to separate GFP and the protein of interest , we introduced a linker with five repeats of nucleotides encoding the ‘EAAK’ motif [49] into EcoRI-digested pSUL-NEO-Tef-3GAGFP-TrpC , generating plasmid pSUL-NEO-Tef-EKGFP-TrpC . To generate the Ptef:VdSep5-GFP:Ttrpc construct , VdSep5 was amplified from the cDNA of V592 and fused into BamHI/EcoRI-linearized pSUL-NEO-Tef-EKGFP-TrpC . To generate a binary vector that included the Tef promotor and TrpC terminator but lacked GFP , we amplified the Tef promoter and TrpC terminator from pNPP94 using primer psul-ppn-HindIII-F and psul-ppn-EcoRI-R and recombined the product with HindIII/EcoRI-linearized pSUL-NEOto generate pSUL-NEO-Tef-TrpC . The Pnoxb:GFP-VdNoxB:Ttrpc construct was generated by cloning the GFP fragment between 2 kb upstream of the start codon and the genomic sequence of VdNoxb . The three sequences were fused into HindIII/EcoRI-linearized pSUL-NEO-Tef-TrpC . To generate the of Ptef1:LifeAct-GFP: Ttrpc fusion construct , we amplified Ptef1: LifeAct-GFP from pAB261 [50] and integrated into HindIII/ EcoRI-linearized pSUL-NEO-Tef-TrpC . To generate the C-terminal GFP fusion construct under the oliC promoter , the primer pair olic-HindIII-F and olic-BamHI-R were used to amplify the template pNAH-Grx1-roGFP2 plasmid [51] , and the resulting PCR products were fused into HindIII/BamHI-digested pSUL-NEO-Tef-EKGFP-TrpC to generate pSUL-NEO-oliC-EKGFP-TrpC . To generate VdSCP8-GFP , VDSCP9-GFP , VDSCP10-GFP , VdSec8-GFP and VdExo70-GFP constructs with the oliC promoter , genomic sequences were amplified and fused into BamHI/EcoRI-linearized pSUL-NEO-oliC-EKGFP-TrpC . To generate VdSCP8-GFP , VdSCP9-GFP , VdSCP10-GFP , VdSec8-GFP and VdExo70-GFP constructs with the native promoter , genomic sequences spanning 1 . 5–2 kb upstream of the start codon were amplified and fused into HindIII/EcoRI-linearized pSUL-NEO-Tef-EKGFP-TrpC using homologous recombination . To co-express the RFP fusion construct with the GFP fusion construct , a nourseothricin resistance cassette was amplified using the primer pair nat-F/R with the pAL6-LifeAct plasmid as template [52] and fused into XbaI/XhoI-digested pSULPH-GFP to generate pNat-GFP . We amplified the Tef promoter and TrpC terminator from pNPP94 using the primers psul-ppn-HindIII-F and trpC-xbaI-R , and we recombined them with HindIII/EcoRI-linearized pNat-GFP to generate pNat-Tef-TrpC . To study the localization relationship between VdSep5 and VdSCP10-GFP , we constructed Ptef:VdSep5-RFP:Ttrpc which was selected by nourseothricin . First , we cloned the RFP sequence from plasmid pAL6-LifeAct [52] by EK-RFP-F and RFP-R into EcoRI-digested pNat-Tef-TrpC , resulting in pNat-Tef-EKRFP-TrpC . A VdSep5 cDNA fragment was fused into BamHI/EcoRI-digested pNat-Tef-EKRFP-TrpC , resulting in Ptef:VdSep5-RFP:Ttrpc fusion . Transformants were selected in the presence of nourseothricin ( 50 μg/mL ) . For complementary VdΔsep3 , VdΔsec22 and VdΔsyn8 mutants , the corresponding genomic sequences , including 1 . 5–2 kb upstream of the start codon , were amplified with the primers listed in S1 Table and fused into HindIII/EcoRI-linearized pNat-Tef-TrpC by homologous recombination . Transformants were selected in the presence of nourseothricin ( 50 μg/mL ) . The Ptef:VdSep5-GFP:Ttrpc construct and Polic:VdExo70-GFP:Ttrpc were used to complementVdΔsep5 and VdΔexo70 mutants , respectively . Total RNA was isolated from frozen mycelium collected from M0 medium cultured for 3d . V . dahliae cDNA was reverse-transcribed using SuperScript® III ( Invitrogen ) . Before reverse transcription , residual DNA was removed from the total RNA using gDNA wiper ( Vazyme ) . cDNA was reverse transcribed using HiScript II Q RT Supermix ( Vazyme ) , and qRT-PCR was performed using ChamQ SYBR qPCR MasterMix ( Vazyme ) with the Bio-Rad CFX96 Real-Time system . The transcription levels of the target genes were quantified relative to the constitutively expressed elongation factor 1-α of Verticillium dahliae ( VdElf ) . The gene-specific primers are listed in S1 Table . Biological replicates were performed three times . To observe the infection of V . dahliae , A . thaliana roots were immersed in a conidial suspension ( ~105 conidia/mL in water solution ) for 10 min and then transferred onto a 0 . 75% agar plate at 25°C in the dark . To observe the protein localization of V . dahliae on cellophane , conidia were placed on cellophane and incubated at 25°C . The mycelium grown on cellophane for 3–9 days was used for protein localization assays . To compare the secretory difference between V592 and VdΔsec22/VdΔsyn8 mutants , the fungi were collected from the outer zone of the colony at the earliest time point for most V592 hyphal necks with ring signals . The fluorescence intensity data were analyzed using the Student’s t-test . Small pieces ( ~0 . 5 cm2 ) of cellophane with mycelium at the margin of the fungal colonies were cut with a scalpel and mounted in water . Images were obtained under a confocal laser microscope ( Leica TCS SP8; Leica Microsystems ) with a 100×oil immersion objective lens . The excitation wavelengths and emission filters were as follows: 488 nm/band-pass 500 to 550 nm for GFP , 561 nm/ band-pass 570 to 670 nm for RFP and FM4-64 , and 405 nm/band-pass 400 to 600 nm for ER-Tracker . Confocal images were captured with a Leica hybrid detector and analyzed with Leica LAS AF software . For each microscopy-based experiment , at least 20 images with three biological independent samples were observed for each micrograph to make conclusions . Each experiment was repeated at least twice . For TEM observation , V . dahliae-infected cotton root and V . dahliae on cellophane were fixed immediately in 2 . 5% glutaraldehyde , buffered with PBS ( pH 7 . 4 ) at 4°C overnight , washed with the same buffer four times and post-fixed with 1% osmium tetroxide for 1 h . Dehydration was then performed in an acetone series ( 50% , 75% , 85% , 95% , 100% ) , and the slices were embedded in Spurr’s resin mixture . Ultrathin serial sections ( 70 nm thickness ) were cut from resin blocks , followed by uranyl acetate staining , and observed with a JEM-1400 electron microscope . For plasma membrane staining , FM4-64 ( ThermoFisher ) was used according to the manufacturer's protocol . For ER-Tracker staining , cultures were incubated at 30°C for 30 min with PBS containing 1 μM ER-Tracker™ Blue–White DPX ( Molecular Probes ) that had been pre-warmed at 30°C for 30 min , washed once with fresh PBS without the dye , and subjected to microscopic observation [53] . Next , 100 μg/mL FITC-conjugated wheat germ agglutinin ( FITC-WGA , Sigma ) was used to stain the fungal cell wall . FRAP analyses was carried out with fungi on cellophane under a spinning disk confocal microscope ( UltraVIEW VoX , Perkin Elmer , Beaconsfield , Buckinghamshire , UK ) equipped with a Yokogawa Nipkow CSU-X1 spinning disk scanner , Hamamatsu EMCCD 9100–13 , and Nikon TiE inverted microscope with the Perfect Focus System . We used the UltraVIEW PK Device to photobleach GFP . For the FRAP analyses , the specific region of interest ( ROI ) covering the entire fluorescence in the ring was selected for bleaching . Twenty bleaching iterations were performed using a 488 laser power of 60% . Image scans were obtained with 15% 488 laser power before and after bleaching . For quantitative analyses , the GFP fluorescence recovery curves were measured as the mean intensity of the ROI pixels , normalized using the using Volocity software ( Perkin Elmer ) , and graphed using Microsoft Excel .	Pathogens secrete effector proteins as molecular weapons to evade or suppress plant immunity . However , the mechanism ( s ) by which root-infecting fungal pathogens secrete secretory effector proteins remains unexplored . We previously reported that Verticillium dahliae , a root-infecting phytopathogenic fungus , forms a specialized infection structure known as a hyphopodium that develops a penetration peg to pierce plant roots . In this study , we observed that after penetration , the penetration peg-developed hyphal neck , partitioning the hyphopodium and invasive hypha , came into close contact with the host , forming the fungus-host penetration interface . NADPH oxidase B ( VdNoxB ) regulated the cytoskeletal organization of the septin ring at the hyphal neck . Importantly , the penetration interface was a preferential site for secretion of signal peptide-containing proteins . Septin plays an important role in the efficient delivery of secretory proteins to the penetration interface . Moreover , the conventional fungal ER-to-Golgi secretion pathway , endosome-mediated transport and the exocyst complex are involved in the delivery of secretory proteins to the penetration interface . Together , our data demonstrate that the V . dahliae infection structure functions as a key signaling hub during plant infection and is the apparatus that not only breaches host cells but also provides a unique interface for the secretion of fungal effectors .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "protein", "transport", "cell", "processes", "light", "microscopy", "fungal", "structure", "fungi", "plant", "science", "microscopy", "septins", "plant", "pathology", "research", "and", "analysis", "methods", "mycology", "proteins", "fluorescence", "recovery", "after", ...	2017	Secretory proteins are delivered to the septin-organized penetration interface during root infection by Verticillium dahliae
Dengue virus ( DENV ) is a mosquito-borne flavivirus , and symptoms of infection range from asymptomatic to the severe dengue hemorrhagic fever/dengue shock syndrome ( DHF/DSS ) . High viral loads correlate with disease severity , and both type I & II interferons ( IFNs ) are crucial for controlling viral replication . We have previously reported that signal transducer and activator of transcription ( STAT ) 1-deficient mice are resistant to DENV-induced disease , but little is known about this STAT1-independent mechanism of protection . To determine the molecular basis of the STAT1-independent pathway , mice lacking STAT1 , STAT2 , or both STAT1 and STAT2 were infected with a virulent mouse-adapted strain of DENV2 . In the first 72 hours of infection , the single-deficient mice lacking STAT1 or STAT2 possessed 50–100 fold higher levels of viral RNA than wild type mice in the serum , spleen , and other visceral tissues , but remained resistant to DENV-induced death . In contrast , the double-deficient mice exhibited the early death phenotype previously observed in type I and II IFN receptor knockout mice ( AG129 ) , indicating that STAT2 is the mediator of the STAT1-independent host defense mechanism . Further studies demonstrated that this STAT2-dependent STAT1-independent mechanism requires the type I IFN receptor , and contributes to the autocrine amplification of type I IFN expression . Examination of gene expression in the spleen and bone marrow-derived macrophages following DENV infection revealed STAT2-dependent pathways can induce the transcription of a subset of interferon stimulated genes even in the absence of STAT1 . Collectively , these results help elucidate the nature of the poorly understood STAT1-independent host defense mechanism against viruses by identifying a functional type I IFN/STAT2 signaling pathway following DENV infection in vivo . Interferons ( IFNs ) play a key role in the defense against viruses and intracellular bacteria [1]–[3] . The type I ( α/β ) IFN receptor is a heterodimer consisting of the IFN-α/β R1 and R2 chains , binding all of the closely related IFN-α subtypes , as well as IFN-β . The type II ( γ ) IFN receptor is a distinct heterodimer , consisting of the IFN-γ R1 and R2 subunits , and it binds IFN-γ . Both IFN receptors signal through the Janus kinase-signal transducer and activator of transcription ( JAK-STAT ) pathway [4] , [5] in which JAKs phosphorylate STATs , which then translocate to the nucleus to induce the expression of IFN-stimulated genes ( ISGs ) [6] . Type I IFN signaling activates STAT1 and STAT2 to heterodimerize and associate with interferon regulatory factor-9 ( IRF9 ) to form IFN-stimulated gene factor-3 ( ISGF3 ) , which specifically binds IFN-stimulated response elements ( ISREs ) within antiviral gene promoters . Type II IFN signaling activates STAT1 , which homodimerizes and binds to DNA at γ-activated sequence ( GAS ) elements . Studies using mice with targeted disruption of individual IFN receptor and STAT genes have provided valuable insights into how type I and II IFNs function during infection by various pathogens . Mice lacking type I IFN receptor are highly susceptible to infection by viruses such as vesicular stomatitis virus ( VSV ) , lymphocytic choriomeningitis virus ( LCMV ) , vaccinia virus ( VV ) , Semliki Forest virus ( SFV ) , and Theiler's murine encephalomyelitis virus ( TMEV ) [3] . In contrast , mice deficient in type II IFN receptor are unable to control infection by intracellular bacteria such as Listeria monocytogenes and Mycobacterium tuberculosis , but are able to mount an effective response against VSV and SFV [3] . Similarly , STAT2-deficient mice exhibit reduced responsiveness to type I IFN and are more susceptible to viral infection [7] , whereas STAT1-deficient mice are defective in their response to both type I and II IFNs and are highly sensitive to both viral and intracellular bacterial infections [8] , [9] . These studies highlight the critical roles STAT proteins play in protection against viral infection , and why STAT proteins are targeted by viruses to evade antiviral responses . The flaviviridae family includes dengue ( DENV ) , yellow fever ( YF ) , West Nile ( WNV ) , and Japanese encephalitis ( JEV ) viruses , and represent a group of pathogens that cause significant morbidity and mortality in humans . Several studies have demonstrated that flaviviruses interfere with antiviral responses by targeting STAT1- and STAT2-mediated signaling . Infection with either WNV or DENV inhibits IFN-mediated STAT1 activation in vitro , including primary human DC cultures [10]–[14] . JEV and Kunjin virus ( a subtype of WNV ) infection blocks IFN-α-induced phosphorylation of both STAT1 and STAT2 in multiple cell lines [10] , [11] , [15] . The antagonists of the antiviral response are primarily the nonstructural proteins of the flaviviruses , which are expressed during infection and required for replication . The expression of NS4B from WNV , YF , or DENV is sufficient for inhibition of IFN-β-induced STAT1 activation in Vero cells , and DENV NS2A and NS4A can contribute to this function [13] , [14] . Similarly , the individual expression of multiple Kunjin virus nonstructural proteins can inhibit IFN-α-mediated STAT2 activation in Vero cells [11] . In addition , DENV NS5 binds and targets STAT2 for proteasome-mediated degradation [16]–[18] , leading to the loss of STAT2 expression observed during DENV infection in vitro [19] . Although STAT1 is a critical component of both type I and II IFN-mediated antiviral responses , STAT1-deficient mice are more resistant to infection with Sindbis virus or murine cytomegalovirus ( MCMV ) than type I and II IFN receptor-deficient ( AG129 ) mice [20] , suggesting that IFN receptors also signal through STAT1-independent mechanisms [21]–[23] . Previously , we have demonstrated that STAT1-deficient 129/Sv mice are resistant to DENV–induced disease , whereas AG129 mice succumb to DENV infection , indicating both STAT1-dependent and STAT1-independent pathways protect against primary DENV infection [24] . However , little is known about the mechanism ( s ) responsible for STAT1-independent protection against viral infections . In this study we used mice lacking STAT1 , STAT2 , both STAT1 and STAT2 , or both STAT1 and the type I IFN receptor to identify STAT2 as a required mediator of the STAT1-independent antiviral response to DENV infection via the type I IFN pathway . In STAT1-deficient mice , STAT2-dependent mechanisms are responsible for the induction of type I IFN and the expression of a subset of ISGs . These data demonstrate the importance of IFN-dependent , STAT1-independent protection against viral infections in vivo , and define key components of protective immunity against DENV . Mice lacking type I and type II IFN receptors ( AG129 ) , but not their wild type or single-deficient counterparts , are highly susceptible to DENV infection and display an early death phenotype when infected with the DENV serotype 2 ( DENV2 ) strain S221 [25]–[27] . In agreement with published studies showing that type I and type II IFNs restrict DENV infection in mice [24] , [26] , 12 of 13 AG129 mice died by day 8 following infection with 1010 genomic equivalents ( GE ) ( ≈2×105 PFU ) of the DENV2 strain S221 [28] , [29] ( Figure 1A ) . Although the STAT1 protein is a critical component of both type I and II IFN receptor pathways , mice lacking STAT1 are able to clear DENV infection [24] . Because STAT2-dependent activity has been previously demonstrated downstream of both type I and type II IFN receptors in the absence of STAT1 [30] , [31] , we hypothesized STAT2 was involved in an IFN-dependent STAT1-independent mechanism of protection against DENV . To assess the contribution of STAT2 to the STAT1-independent control of DENV infection in vivo , STAT1−/−/2−/− double knockout mice were generated by intercrossing single-deficient animals , and also infected with 1010 GE S221 . At this dose of virus , 90–100% of wild type and congenic single-deficient mice lacking either STAT1 or STAT2 survived DENV infection . However , all mice with a combined deficiency of STAT1 and STAT2 displayed the early death phenotype observed in AG129 mice and succumbed to DENV infection within 4–6 days , demonstrating the importance of STAT2 to the STAT1-independent mechanism of protection against DENV infection in vivo ( Figure 1A ) . The median survival of STAT1−/−/2−/− mice was significantly lower than AG129 at this virus dose ( p = 0 . 0013; ** ) , indicating the absence of both STAT1 and STAT2 renders these mice more susceptible to DENV-mediated disease than the combined loss of type I and II IFN receptors . To test the limits of this STAT1-independent protection against DENV , the single-deficient mice were infected with 10-fold ( 1011 GE ) and 100-fold higher ( 1012 GE ) doses of S221 and monitored for survival . Survival proportions of STAT1−/− mice challenged with 1011 GE DENV remained the same as those infected with 1010 GE DENV , whereas 4 of 5 STAT1−/− animals receiving the highest dose of virus ( 1012 GE ) were moribund by day 10 ( Figure 1B ) . In contrast to STAT1−/− mice , all STAT2−/− animals infected with the equivalent doses of DENV survived beyond 30 days , indicating that although the STAT1-independent antiviral response is effective against DENV in vivo , it is less robust than the antiviral response in the absence of STAT2 . STAT2 function has been primarily linked to the type I IFN receptor , where it is well-documented to form the ISGF3 complex with IRF9 and STAT1 . However , recent data have indicated that STAT2 functions not only downstream of type I IFN , but also type II IFN [30] , [32] , and type III ( λ ) IFN signaling [33] , implicating type I , type II , and type III IFN receptors as potential candidates for mediating STAT1-independent , STAT2-dependent signaling during DENV infection . Because we observe the early death phenotype in AG129 mice , which retain the type III IFN receptor , we reasoned STAT2 activation was primarily occurring downstream of either the type I or type II IFN receptor . Therefore , to determine which IFN receptors are involved in the STAT1-independent pathway , double-deficient mice lacking STAT1 and either the type I IFN receptor ( IFNAR ) or type II IFN receptor ( IFNGR ) were generated by intercrossing the single-deficient mouse strains . Following intravenous infection with 1010 GE of S221 , 100% of STAT1−/−/AR−/− mice developed early lethal disease , whereas none of the STAT1−/−/GR−/− mice succumbed to infection ( Figure 1C ) , demonstrating that the survival phenotype of STAT1−/−/AR−/− mice recapitulates that of STAT1−/−/2−/− mice . This result indicates that the STAT1-independent , STAT2-dependent mechanism of protection against DENV infection in mice is mediated by the type I IFN receptor . To evaluate how STAT1 and STAT2 deficiency impacts the control of DENV infection , wild type , STAT1−/− , STAT2−/− , STAT1−/−/2−/− , and STAT1−/−/AR−/− mice were infected intravenously with 1010 GE of S221 , and viral RNA levels in various tissues were measured via qRT-PCR at 6 , 12 , 18 , 24 , and 72 hours after infection . Although minimal viral RNA was detected in each strain at 6 hours post-infection , increasing viremia was observed in all strains except wild type by 12 hours post-infection . At 12 , 18 , and 24 hours post-infection , the high viremia observed in the single- and double-deficient strains ( all p<0 . 0005 ) demonstrates that the combined function of both STAT1 and STAT2 is required for effective control of viral replication ( Figure 2A ) . Previous studies have shown that the spleen is an initial site of DENV replication in mice [28] , [34] , and at 18 and 24 hours post-infection , levels of DENV RNA in the spleen of each knockout mouse strain were significantly higher than wild type ( 14- to 27-fold increase; all p<0 . 0005 ) ( Figure 2B ) . Although each single-deficient mouse strain possessed similar levels of viral RNA as STAT1−/−/2−/− mice at 12 and 18 hours post-infection , the levels of virus in the serum and spleen were 6–7 fold higher in STAT1−/− mice than STAT2−/− mice by 24 hours . This suggests that early control of DENV replication requires the combined function of STAT1 and STAT2 , but STAT2-independent mechanism ( s ) begin to restrict replication by 24 hours post-infection . At 72 hours post-infection , viremia in both STAT1−/− and STAT2−/− mice was reduced relative to the 24-hour time point ( STAT1−/− 30-fold , p<0 . 0001; STAT2−/− 33-fold , p = 0 . 0019 ) , and the levels of viremia were similar between these two strains . In contrast to the single-deficient mice , viremia in STAT1−/−/2−/− mice increased 25-fold between 24 and 72 hours after infection . By 72 hours , viral RNA levels in the spleen had decreased significantly in all mouse strains , but remained at least 40-fold higher in STAT1−/−/2−/− mice as compared to the single-deficient strains . In the liver , kidney , and small intestine , the later sites of viral replication in DENV-infected mice [28] , [34] , significantly higher ( 100 to 10 , 000-fold ) levels of viral RNA were observed in the double-deficient mice at 72 hours post-infection as compared with the single-deficient animals ( Figure 2C ) . Viral burden in STAT1−/−/AR−/− mice was equivalent to STAT1−/−/2−/− mice in all tissues at 72 hours post-infection ( data not shown ) . In wild type mice , minimal or no DENV RNA was detected in these organs . In the single-deficient mice , viral RNA levels in STAT1−/− and STAT2−/− animals were similar in the liver , kidney , and small intestine , although viral load in the kidney was slightly higher in STAT2−/− mice ( 4-fold; p = 0 . 0105 ) . Collectively , these data demonstrate that efficient control of early viral replication requires the combined action of STAT1 and STAT2 . However , each can function independently of the other to limit viral replication later during infection . Only the combined absence of both STAT1 and STAT2 ablates the antiviral response , and results in unrestricted viral replication . Our previous studies have shown that viral load is directly linked to high levels of circulating tumor necrosis factor ( TNF ) , which contributes to the early death phenotype observed in S221-infected AG129 mice [27] , [35] . To compare STAT1−/−/2−/− mice to single-deficient and wild type mice , serum TNF levels were measured by ELISA . Both STAT1−/− and STAT2−/− mice possessed similar levels of TNF in the serum at 72 hours post-infection , which were 4–5 fold higher than that detected in wild type mice ( Figure 2D ) . However , levels of serum TNF in the single-deficient strains remained significantly lower ( >20-fold ) than in STAT1−/−/2−/− mice , indicating that control of viral replication by STAT1 or STAT2 alone results in reduced TNF expression during DENV infection , and is consistent with the survival studies where only double-deficient animals were susceptible to DENV-induced death . Based on the survival data indicating that STAT1-independent protection is mediated via type I IFN signaling , we next examined type I IFN levels in infected mice , as expression of type I IFN is partially dependent upon a positive feedback mechanism [2] . Specifically , type I IFN levels in wild type and single-deficient mice were compared to STAT1−/−/2−/− mice , which are not expected to possess the positive feedback loop . Serum levels of IFN-α and IFN-β in wild type , STAT1−/− , STAT2−/− , and STAT1−/−/2−/− mice were measured by ELISA between 0 and 24 hours post-infection . High levels of serum IFN-α were present in wild type mice by 12 hours post-infection , and these levels diminished between 18 and 24 hours after infection ( Figure 3A ) , consistent with our previous observations in wild type C57BL/6 mice following DENV infection [28] . However , robust IFN-α production in both STAT1−/− and STAT2−/− mice was delayed until 18 hours following infection , and STAT1−/− mice exhibited decreased expression of IFN-α compared with STAT2−/− mice . Serum IFN-α levels in STAT1−/−/2−/− mice followed the same kinetics as STAT1−/− mice , including increased expression at 18 hours post-infection . However , despite similar levels of viremia detected in both strains at 24 hours post-infection ( Figure 2A ) , IFN-α in STAT1−/− mice remained elevated but in STAT1−/−/2−/− mice it decreased to the level at or below the limit of detection ( p<0 . 0001 ) . Similar trends were observed for IFN-β production following DENV infection ( Figure 3B ) , including delayed induction in the knockout strains relative to wild type mice , and elevated IFN-β levels in STAT1−/− but not STAT1−/−/2−/− mice at 24 hours . Together , these results demonstrate that the combined function of STAT1 and STAT2 is required for maximum early induction of type I IFN in response to DENV infection . The presence of either STAT1 or STAT2 alone is also sufficient to drive type I IFN expression , but with delayed kinetics when compared to an intact IFN signaling pathway . Additionally , the difference in IFN levels between STAT1−/− and the double knockout mice at 24 hours shows that a STAT2-dependent mechanism contributes to type I IFN expression in the absence of STAT1 . To further define the mechanisms by which type I IFN receptor-STAT2 signaling protects against DENV infection in vivo , we next examined gene expression in the spleens of S221-infected mice using the Type I IFN-related RT2 Profiler PCR Array ( SABiosciences/QIAGEN ) . RNA from whole spleens was analyzed and fold-induction calculated by comparing infected mice to naïve mice of the corresponding strain . Of the 84 genes in the array , 31 ( 36 . 9% ) , 16 ( 19 . 0% ) , 22 ( 26 . 2% ) , and 7 ( 8 . 3% ) genes , respectively , were induced 3-fold over naïve in wild type , STAT1−/− , STAT2−/− , and STAT1−/−/2−/− mice at 12 hours post-infection ( Figure 4A ) , and the number of genes upregulated was similar at 24 hours post-infection for each mouse strain . Genes that were induced in STAT1−/−/2−/− mice identified ISGs that were upregulated independently of either STAT1 or STAT2 . These ISGs included Ifna2 , Ifna4 , and Ifnb1 , which were induced to high levels in all four strains at 12 hours ( >1000-fold ) and 24 hours ( >100-fold ) post-infection ( Table S1 ) , and likely represent early IRF3-mediated gene expression [36] . A number of ISGs were identified that were induced >3-fold over uninfected controls in STAT1−/− mice , but were not induced in STAT1−/−/2−/− mice ( Figure 4A ) , revealing that a STAT2-dependent mechanism can drive the expression of multiple ISGs in the absence of STAT1 . In addition , several genes that were weakly upregulated in STAT1−/−/2−/− mice ( Cxcl10 , Mx2 , Isg15 ) were induced 6- to 28-fold higher in STAT1−/− mice than in the double knockout , suggesting that the STAT1-independent pathway heavily contributes to the regulation of these genes as well . Genes upregulated in STAT1−/− mice were then compared to wild type and STAT2−/− mice at 12 hours ( 12 genes ) ( Figure 4C ) and 24 hours ( 12 genes ) ( Figure 4D ) post-infection . At the 12 hour time point , more than half of the genes in this group displayed at least 3-fold higher induction in wild type mice than in STAT1−/− mice . However , by 24 hours post-infection , the gene induction levels observed in wild type and STAT1−/− mice were similar ( <2-fold ) except for the highly elevated Oas1a gene in STAT1−/− mice . This increased upregulation in STAT1−/− mice at 24 hours is consistent with the elevated levels of type I IFN observed at 24 hours post-infection ( Figure 3 ) . Two genes upregulated in STAT1−/− mice , Mx1 and Eif2ak2 ( PKR ) , were not induced in STAT2−/− mice , suggesting that STAT2 is required for their expression in response to DENV infection . Conversely , Gbp1 and Gbp2 were induced more efficiently in STAT2−/− animals than in STAT1−/− mice , and several additional genes in this array were induced in STAT2−/− and wild type but not STAT1−/− mice at both timepoints . ( Table S1 ) . Collectively , these results demonstrate that a STAT1-independent pathway regulates the expression of ISGs in a STAT2-dependent manner during DENV infection in vivo . In order to perform more mechanistic studies , bone marrow-derived macrophages ( BMMs ) were isolated from wild type and knockout mice , and evaluated in vitro . BMMs were chosen because cells of the monocyte/macrophage lineage are presumed to be one of the major cell types that support DENV replication in humans and mice [37]–[39] . BMMs were infected with S221 ( MOI = 5 ) , and gene induction at 12 and 24 hours post-infection was evaluated using the RT2 Profiler PCR Array ( Table S2 ) . A majority of genes induced in the spleen were also upregulated in BMMs from each strain ( 29/31 genes in wild type; 13/16 in STAT1−/−; 5/7 in STAT1−/−/2−/− ) , indicating that gene regulation of ISGs is similar between the spleen and BMMs following DENV infection ( Figure 4B ) , and validating the use of BMMs as a suitable in vitro model for these studies . To confirm activation of STAT2 in the absence of STAT1 , phosphorylation and nuclear localization of STAT2 was examined in BMMs . Type I IFN signaling activates STAT2 via phosphorylation of a tyrosine residue ( Y689 ) , which is required for its association with STAT1 and incorporation into the transcriptionally active complex ISGF3 [40] . BMMs isolated from wild type , STAT1−/− , STAT1−/−/2−/− , and STAT1−/−/AR−/− mice were stimulated with recombinant IFN-β for 15 minutes , and the phosphorylation status of STAT2 was examined via Western blot ( Figure 5A ) . As expected , both STAT1 and STAT2 were phosphorylated in wild type cells . Phosphorylation of STAT2 was observed in STAT1−/− but not STAT1−/−/AR−/− cells , although the basal level of STAT2 was lower in both STAT1−/− and STAT1−/−/AR−/− cells than in wild type cells . These results show that STAT2 activation can occur in the absence of STAT1 following type I IFN stimulation . In addition to type I IFN , cytokines produced during DENV infection are able to activate JAK-STAT signaling , including IL-2 , IL-4 , IL-6 , IL-10 , and IFN-γ [35] , [41] , [42] . To examine STAT2 activation in the context of DENV infection , BMMs were incubated with serum taken from wild type mice infected with S221 , and analyzed by Western blot ( Figure 5B ) . Similar to recombinant IFN-β , the infected mouse serum induced both STAT1 and STAT2 phosphorylation in wild type cells . Phosphorylation of STAT2 was detected in STAT1−/− but not STAT1−/−/AR−/− BMMs , indicating that activation of STAT2 in these cells is primarily dependent upon type I IFN signaling . To verify that phosphorylation of STAT2 occurs in the absence of STAT1 during DENV infection , BMMs were infected with S221 and STAT phosphorylation was examined at 12 , 18 , and 24 hours post-infection ( Figure 5C ) . Consistent levels of STAT1 phosphorylation were detected at all time points in wild type cells , whereas levels of phosphorylated STAT2 were highest at 12 hours after infection and decreased over time . Phosphorylated STAT2 was also detected at the 12 hour and 18 hour time points in STAT1−/− cells , but not in STAT1−/−/AR−/− cells . To further confirm STAT1-independent activation of STAT2 , wild type and STAT1−/− BMMs were incubated with serum from either naïve or S221-infected wild type mice for 1 hour , and nuclear localization of STAT2 was examined using immunofluorescence ( Figure 5D ) . Although the staining pattern within the nucleus appeared different between the two strains , STAT2 was observed in the nuclei of both wild type and STAT1−/− cells . Taken together , these results indicate that STAT2 is activated via type I IFN receptor signaling during DENV infection , even in the absence of STAT1 . In response to type I IFN signaling , the STAT1:STAT2:IRF9 ( ISGF3 ) complex translocates to the nucleus to bind ISREs found in the promoters of many ISGs . To determine whether STAT2 is directly involved in the transcriptional regulation of ISGs in the absence of STAT1 , chromatin immunoprecipitation ( ChIP ) experiments were performed using a STAT2-specific antibody . Gene targets Oas1a , Oas1b , and IRF7 were chosen from ISGs that were expressed in wild type and STAT1−/− BMMs , but not in STAT1−/−/2−/− double deficient cells as determined by the PCR array analysis ( Table S2 ) . DNA from S221-infected wild type , STAT1−/− and STAT1−/−/2−/− BMMs was harvested at 12 and 24 hours post-infection , and quantitative PCR was performed on anti-STAT2 precipitated DNA using primer pairs specific to ISREs located in the promoters of these genes . Following DENV infection , increased binding of STAT2 to all three gene promoters was observed in wild type and STAT1−/− BMMs when compared to naïve cells ( Figure 6A–C ) . STAT2 binding to the Oas1a promoter in STAT1−/− cells was only significantly enriched at 24 hours post-infection , while Oas1b and IRF7 had significant enrichment at both time points . In particular , the binding activity of STAT2 to the Oas1b promoter in STAT1−/− cells almost matched that of wild type cells ( Figure 6B ) . Nearly undetectable DNA enrichment was observed for all three genes in STAT1−/−/2−/− cells , providing evidence of the antibody's specificity for STAT2 . Taken together , these results demonstrate that STAT2 binds to the promoters of ISGs in the absence of STAT1 , providing further evidence of a direct role for STAT2 in the regulation of antiviral responses . STAT1 was originally described as a transcription factor essential for type I and II IFN signaling , and initial studies described STAT1−/− cells and mice as unresponsive to both type I and II IFN [8] , [9] . However , later studies in STAT1−/− mice infected with Sindbis virus or MCMV revealed the existence of a STAT1-independent antiviral mechanism [20] . Similarly , we have demonstrated that a STAT1-independent IFN response confers protection from DENV infection in mice [24] . Given that both type I and type II IFN-mediated responses restrict DENV infection in vitro and in vivo [25] , [26] , [43] , [44] , we were interested in defining the mechanism of this STAT1-independent protection . In this study , we demonstrated that the combined loss of both STAT1 and STAT2 in mice resulted in early death following DENV infection . These results recapitulated the early death phenotype of AG129 mice and implied that the increased DENV susceptibility in STAT1−/−/2−/− mice was due to the loss of type I and II IFN signaling . Conversely , both STAT1 and STAT2 single-deficient mice survived infection with DENV at this same dose of virus , suggesting there are sufficient compensatory mechanisms that protect against DENV infection in the absence of either one of these STAT proteins . While the susceptibility of the STAT1−/−/IFNAR−/− double knockout mice to DENV infection indicated that STAT2 was functioning downstream of type I IFN signaling , the survival of STAT1−/−/IFNGR−/− mice confirmed the potency of STAT2-mediated protection . IFNAR−/−/IFNGR−/− ( AG129 ) mice do not survive infection , yet STAT1−/−/IFNGR−/− mice remain healthy , demonstrating that STAT2 alone provides sufficient signaling via the type I IFN pathway for a functional antiviral response , despite the absence of both STAT1 and type II IFN signaling . However , STAT1−/− mice challenged with a 100-fold higher dose of virus succumbed to DENV infection , indicating that STAT2-mediated protection can be overcome . In contrast , STAT2−/− mice remained healthy at the highest dose tested , most likely due to an intact type II IFN signaling pathway , and indicating that functional STAT1 plays a larger role in anti-DENV responses than STAT2 . This claim is supported by the viral load data , where tissue viral burden in STAT2−/− mice remained slightly lower than their STAT1−/− counterparts at multiple timepoints following infection . Although little difference was observed in tissue viral load between the single deficient STAT mice and the double knockouts during the first 24 hours following infection , both STAT1 and STAT2 were able to function independently of each other to provide significant protection in all tissues by 72 hours post infection . Consistent with the kinetics of type I IFN expression , the viral load data suggest that antiviral responses in the single deficient mice are delayed compared to wild type mice . This delayed response is likely due to the inability to form ISGF3 ( STAT1:STAT2:IRF9 ) , the primary transcription complex associated with type I IFN signaling , resulting in suboptimal activation of the type I IFN signaling pathway . Despite the inability to form ISGF3 , STAT1−/− mice still induced a subset of ISGs in response to DENV infection that were not upregulated in STAT1−/−/2−/− mice . The PCR array data , although far less comprehensive than genome-wide analysis , provided evidence that the STAT1-independent pathway mediates protection through the expression of antiviral genes . One concern with the PCR array approach in whole spleens is that total RNA was isolated from a heterogeneous cell population , and differences in cell type distribution between strains or naïve versus infected mice could complicate analysis . To address this , PCR arrays were performed using BMMs from the different knockout strains . Although the kinetics and efficiency of infection are likely to be different in vivo versus in vitro , the overall pattern of ISG induction in BMMs was similar to the spleen , confirming that our gene expression data were not biased by using whole tissue . Furthermore , flow cytometry analysis of both naïve and infected mouse spleens revealed that little difference exists in cellular composition between STAT1−/− and STAT1−/−/2−/− mice ( Figure S1 ) , whose relative differences in gene induction were compared to identify STAT1-independent ISGs . Prior to this work , several key studies demonstrated that STAT2 can function in the absence of STAT1 . Hahm and colleagues showed that dendritic cell maturation is impaired in a IFN-β-dependent , STAT2-dependent manner following infection of pluripotent bone marrow cells by measles virus ( MV ) and LCMV [31] . Although this was the first description of a STAT2-mediated STAT1-independent phenotype , the focus of this work was on an immunosuppressive rather than an antiviral phenotype . Sarkis and colleagues first demonstrated that STAT2 was required for IFN-α-induced expression of several antiviral genes in cultured liver cells , including Mx1 , PKR , and Isg15 [45] , independently of STAT1 expression . However , the authors inferred this STAT2-dependent pathway was liver-specific since they could not carry these observations into cells other than human hepatoma cells . Similarly , Lou and colleagues also described STAT1-independent gene induction of RIG-G by STAT2 in response to type I IFN , in an in vitro system that required the over-expression of IRF9 [46] . The present study extends these bodies of work by demonstrating that STAT2 can mediate antiviral responses independently from STAT1 through direct association with ISG promoters , including antiviral genes that are otherwise regulated by ISGF3 , not just ISGs that have been demonstrated to be STAT2-dependent . To our knowledge , this is the first report of STAT2 functioning downstream of type I IFN signaling in the absence of STAT1 , which results in a functional antiviral response in vivo . ChIP analysis confirmed the association of STAT2 protein with antiviral gene promoters , implying that STAT2 has a direct role in transcriptional regulation , but the exact mechanism of how STAT2 mediates IFN signaling remains to be determined . Although it possesses a strong transactivation domain , STAT2 binds DNA poorly , suggesting that STAT2 does not function alone [47] . IRF9 ( p48 ) provides both DNA binding and sequence specificity functions to the ISGF3 complex [47] , and its involvement has been implicated in STAT1-independent mechanisms [45] , [46] . Therefore , the most likely candidate for an alternative STAT2 complex includes IRF9 . In addition to ISGF3 , several complexes containing STAT2 have been described , including STAT2:3 and STAT2:6 heterodimers , and STAT2:2 homodimers [47]–[49] . However , further biochemical studies will be required to identify the components of this alternative transcriptional complex that functions in the absence of STAT1 . The promoter sequence element that binds this alternative complex also remains unknown , and likely depends upon other binding partners . The ISGs identified by PCR array in this study represented a subset of ISGs induced in wild type mice , suggesting a novel binding element is unlikely . A more comprehensive study of gene modulation other than the PCR array , such as full microarray or ChIP-seq , should shed light on how ISRE , GAS , and other promoter elements are regulated in the absence of STAT1 . It is important to consider the difference between mouse and human STAT2 when interpreting the results of our studies . While the other STAT proteins have a high degree of sequence homology between the two species , the C-terminal transactivation domains between human and mouse STAT2 are completely divergent , so it is unknown whether mouse and human STAT2 function similarly when taken out of the context of ISGF3 . However , in vitro studies have suggested mouse STAT2 is functionally analogous to human STAT2 in ISGF3 formation , ISRE activation , and biological response to type I IFN [50] . The relevance of species-specific STAT2 has recently been demonstrated in terms of DENV infection: the NS5 protein of DENV binds and targets human STAT2 for proteasomal degradation to inhibit type I IFN responses [16] , [18] . Ashour and colleagues have recently demonstrated that this targeting is species specific , and murine STAT2 is resistant to NS5-mediated degradation [17] . In the studies presented here , basal expression of STAT2 was unaffected in BMMs infected with DENV , but the level of infection of our BMMs may be as low as 1% ( unpublished observation ) , suggesting NS5 has little direct effect upon STAT2 in our culture system . In this study , we have identified STAT2 as a key component of the STAT1-independent mechanism of protection against DENV infection in mice , and demonstrated that both STAT1 and STAT2 possess the ability to independently limit the severity of DENV pathogenesis . For many viruses , inhibition of STAT-mediated signaling is a major mechanism to evade antiviral responses . Our data suggests that DENV-mediated inactivation of STAT1 function alone is not sufficient to neutralize antiviral responses , emphasizing the importance of DENV mechanisms to specifically target host STAT2 function . Increasing evidence suggests that the relative ability of flaviviruses to subvert STAT signaling , including DENV , WNV , JEV , and Kunjin viruses , may be a contributing factor to their virulence [51]–[54] . In conjunction , epidemiological studies have revealed that selection for virulent DENV strains occurs in both humans and mosquitoes [55]–[57] . A more detailed understanding of how DENV causes specific pathologies , even in an immunodeficient setting ( as in STAT1−/−/2−/− mice ) , may offer additional insight regarding prevention of human DENV-induced disease from both current and newly emerging strains . 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 US Public Health Service Policy on Humane Care and Use of Laboratory Animals , and the Association for Assessment and Accreditation of Laboratory Animal Care International ( AAALAC ) . All experimental procedures were pre-approved and performed according to the guidelines set by the La Jolla Institute for Allergy and Immunology Animal Care and Use Committee . Generation and preparation of DENV2 strain S221 has been described previously [28] , [35] . Briefly , S221 is a triple-plaque-purified clone of D2S10 , which was generated by alternative passage between AG129 mice and C6/36 mosquito cells . Viral stocks used for these studies were grown in C6/36 mosquito cells and concentrated by ultracentrifugation [34] . Genome equivalents ( GE ) were quantified by real-time reverse transcription-PCR ( RT-PCR ) and infectious titer determined by plaque assay on BHK-21 cells , as previously described [34] . Wild type ( WT ) 129/Sv/Ev were purchased from Taconic Farms . STAT1−/− 129/Sv/Ev mice bearing a deletion in the DNA binding domain of the STAT1 gene were obtained from Dr . Joan Durbin ( Ohio State University , Columbus , OH ) . STAT2−/− 129/Sv/Ev mice were obtained from Dr . Christian Schindler ( Columbia University , New York , NY ) . All mice were bred and housed under specific pathogen-free conditions at the La Jolla Institute for Allergy and Immunology ( LIAI ) . For all experiments , sex-matched mice at 5 to 6 weeks of age were infected intravenously with 1010 GE ( ≈2×105 PFU ) of S221 diluted into 200µL PBS with 5% FCS . For survival studies , mice were sacrificed when moribund or at the first signs of paralysis . Mice were euthanized via isoflurane inhalation and blood was collected by cardiac puncture . Mice were perfused with 60mL phosphate-buffered saline ( PBS ) and tissues collected into RNALater ( QIAGEN ) . Tissues were homogenized for 3 minutes at 4°C in 1mL tissue lysis buffer ( QIAGEN RLT Buffer ) using a Mini-beadbeater-8 ( BioSpec Products ) . Liver and small intestine homogenates were diluted 5-fold and homogenized for 3 additional minutes . RNA was isolated from tissue homogenates using the RNeasy minikit ( QIAGEN ) and from 30µL serum using the QIAmp viral RNA minikit ( QIAGEN ) according to manufacturer's instructions , eluted and stored at −80°C until analysis . Quantitative RT-PCR to detect DENV and 18S was performed as previously described [34] . Viral load is expressed as GE per mL in serum , or GE normalized to copies of 18S in tissues . Serum from infected or naïve animals was analyzed using either a mouse TNF-α ELISA Ready-Set-Go kit ( eBioscience ) or mouse IFN-α ELISA and IFN-β ELISA ( PBL Technologies ) , all according to the manufacturers' instructions . The IFN-α ELISA kit detects all subtypes of IFN-α according to the manufacturer . To generate bone marrow-derived macrophages , bone marrow from femur and tibia of 5- to 7-week old mice were isolated and cultured for 6–7 days in RPMI 1640 ( 10% fetal calf serum , penicillin , streptomycin , HEPES , sodium pyruvate , 55µM β-mercaptoethanol ) in the presence of 100ng/mL murine macrophage stimulating factor ( M-CSF ) ( PeproTech ) . On day 6–7 of culture , cells were treated with either 1000U recombinant murine IFN-β ( PBL Biomed ) for 15 minutes , 100µL serum obtained from DENV2-infected wild type mice for 60 minutes , or infected with S221 ( MOI = 5 ) , followed by harvesting of the cells at 0 , 12 , 18 , or 24 hours post-infection . Immediately following treatment , bone marrow-derived macrophages ( 106 cells ) were lysed in Western lysis buffer ( 25mM Tris-HCl , 150mM NaCl , 1% NP40 ) in the presence of protease inhibitor cocktail ( Sigma ) , Phosphatase Inhibitor Cocktail Set II ( Calbiochem ) , and 1mM okadaic acid ( Sigma ) . Samples were resolved on 8% SDS-polyacrylamide gels , transferred overnight to PVDF ( Millipore ) at 4°C , and probed overnight with the following antibodies: anti-STAT1 and anti-P701-STAT1 ( Cell Signaling ) , anti-STAT2 ( Millipore ) , anti-P689-STAT2 ( Santa Cruz Biotech ) , and anti-beta-actin ( Sigma ) . Blots were incubated with HRP-conjugated secondary antibody ( Pierce ) and visualized with Western Lightning Plus-ECL ( Perkin Elmer ) . Splenic RNA was harvested from spleens as described above . Bone marrow-derived macrophage RNA was harvested using Trizol reagent ( Invitrogen ) and further purified with the RNeasy minikit ( QIAGEN ) . To remove genomic DNA contaminants , 1µg total RNA was treated with 0 . 5 units DNAse ( Invitrogen ) for 15 minutes at 20°C and then 8 minutes at 70°C to inactivate DNAse enzyme . cDNA was then generated using the RT2 First Strand kit ( SABioscience ) according to the manufacturer's protocol . Newly synthesized cDNA was loaded onto RT2 Profiler ( SABioscience ) 384-well PCR array plates ( PAMM-016 ) and amplified on the LightCycler 480 PCR system ( Roche ) for 40 cycles . The resulting threshold cycle values ( Ct ) were uploaded onto the SABioscience website ( http://www . sabiosciences . com/pcr/arrayanalysis . php ) and relative levels of gene expression ( fold differences ) were calculated using the provided software . Ct values ≥35 cycles were considered undetectable . Bone marrow-derived macrophages were seeded on glass coverslips and treated with infected mouse serum for 1 hour . Following treatment , cells were fixed in 3% paraformaldehyde for 30 minutes , permeabilized in 0 . 1% Triton X-100 for 10 minutes , and blocked in 20% purified goat serum ( Pierce ) for 1 hour . Cells were washed , incubated with anti-STAT2 antibody ( obtained from Dr . Christian Schindler ) followed by DyLight 649-labeled goat anti-rabbit IgG ( Jackson Immunoresearch ) . Nuclei were counterstained with DAPI , and viewed using a Marianas deconvolution fluorescence microscope ( 3i ) . ChIP assays were performed essentially as described [58] . Bone marrow-derived macrophages ( 3×106 cells ) were fixed in 1% formaldehyde and lysed in 50mM Tris-HCl , pH 8 . 0 , 10mM EDTA , 1% SDS on ice for 5 minutes . Lysates were sonicated and diluted with 9 parts 50mM Tris-HCl , pH 8 . 0 , 167mM NaCl , 1 . 1% Triton X-100 , and 0 . 11% sodium deoxycholate . 10µg of DNA was used for each immunoprecipitation . Rabbit polyclonal antibody specific for murine STAT2 ( Schindler ) was used to precipitate STAT2-DNA complexes . Immunocomplexes were bound to protein G-agarose beads ( Pierce Thermo Scientific ) , washed 5 times , and eluted by incubating beads at 65°C overnight in 10mM Tris-HCl , 5mM EDTA , 300mM NaCl , 0 . 5% SDS . Eluted DNA was treated with RNAse A and proteinase K , and purified using the QIAquick PCR Purification kit ( QIAGEN ) . Quantitative SYBR Green PCR was performed on the LightCycler 480 PCR System ( Roche ) for 45 cycles . Primer sequences used: Oas1a ( + ) AAACCCCAAGAAAGCCAGAT , Oas1a ( − ) CTCCCAGCCTAGCTGAAATG; Oas1b ( + ) CTGTTCAGAAGCCCTAACGC; Oas1b ( − ) AGGTCAGCACAGAAGCTGGT; IRF7 ( + ) TGGGATCTGAGTAAGGGTCG; IRF7 ( − ) GCCAAGGTGGCTGTAGATGT . Mice were infected i . v . with 1010 GE of S221 . Single-cell suspensions were prepared from naïve and infected spleens harvested at 12 and 24 hours post-infection with Spleen Dissociation Medium ( Stem Cell Technologies ) ( n = 3 per group ) . The following antibodies were used: Alexa Fluor 647-conjugated anti-B220 ( HI30; Biolegend ) and anti-F4/80 ( Cl:A3-1; Biolegend ) ; eFluor 450-conjugated anti-CD4 ( RM4-5; eBioscience ) and anti-Gr-1 ( RB6-8C5; eBioscience ) ; fluorescein isothiocyanate-conjugated anti-CD3ε ( 145-2C11; eBioscience ) and anti-CD11b ( M1/70; BD Pharmingen ) ; phycoerythrin-conjugated anti-CD11c ( N418; eBioscience ) ; phycoerythrin-indotricarbocyanine-conjugated anti-CD8α ( 53-6 . 7; BD Pharmingen ) ; and peridinin chlorophyll protein-cyanine 5 . 5-conjugated anti-B220 ( RA3-6B2; BD Pharmingen ) and anti-CD49b ( DX5; Biolegend ) . Samples were collected on FACS LSR II ( BD Biosciences ) and were analyzed with FlowJo software ( TreeStar ) . Live cells were counted with a Vi-CELL XR ( Beckman Coulter ) . For all studies , data were analyzed by Prism 5 software ( GraphPad Software ) . For survival studies , Kaplan-Meier survival curves were analyzed by log rank test . Viral load , cytokine ELISA , and ChIP data were analyzed using the unpaired t test .	Dengue virus ( DENV ) is a mosquito-borne pathogen present in the tropical and sub-tropical regions of the world , and an estimated 2 . 5 billion people are at risk of infection . Interferon ( IFN ) mediated innate responses in the host are critical for limiting viral spread following DENV infection . We have previously demonstrated that mice lacking STAT1 , a key mediator of both type I and II IFN responses , are not susceptible to DENV-mediated disease . In this study , we sought to determine the mechanism responsible for protection against DENV disease in the absence of STAT1 . Using knockout mice , we identify STAT2 as the protein that mediates type I IFN signaling during DENV infection in the absence of STAT1 . The resulting antiviral response includes amplification of type I IFN and the expression of interferon stimulated genes . These data suggest DENV infection is especially sensitive to STAT2-mediated antiviral responses in vivo , and provide novel insights towards how IFNs protect against viral infections .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "virology/virulence", "factors", "and", "mechanisms", "virology/viral", "replication", "and", "gene", "regulation", "infectious", "diseases/neglected", "tropical", "diseases", "virology/animal", "models", "of", "infection", "virology", "immunology/innate", "immunity", "infect...	2011	STAT2 Mediates Innate Immunity to Dengue Virus in the Absence of STAT1 via the Type I Interferon Receptor
Yellow fever ( Yf ) outbreak was recently reported in South Omo of Southern Ethiopia . This area was also highly affected by Yf outbreak in the 1960s . However , there is no reliable information on the level of community knowledge attitudes and practices about the disease in the area . The objective of the current study was to assess level of community knowledge , attitudes and practices about Yf . Between March and May 2017 , a community-based cross-sectional survey was conducted in two districts of the South Omo area . During the survey , 612 randomly selected adults were interviewed about Yf using structured questionnaire . Out of the 612 study participants , 508 ( 83 . 0% ) reported that they heard about Yf which is locally known as “a disease that causes vomiting blood” . Most ( 90 . 4% ) of the study participants also said that Yf is different from malaria . Two hundred thirteen ( 41 . 9% ) participants said that Yf can be transmitted from a patient to another person , while only 80 ( 37 . 6% ) mentioned that the disease is transmitted through mosquitoes bite . Out of 333 ( 65 . 7% ) study participants who believed that Yf is a preventable disease , 280 ( 84 . 1% ) mentioned vaccine as a preventive method . The majority believed that the disease is a killer ( 97 . 2% ) and a newly emerging ( 69 . 4% ) . Among the total of 612 study participants , 221 ( 36 . 1% ) were considered as having a high level of overall knowledge of Yf . Having educational level above 7th grade ( AOR = 3 . 25 , 95% CI: 1 . 39 , 7 . 57 , p = 0 . 006 ) and being resident of Bena-Tsemay district ( AOR = 1 . 77 , 95% CI: 1 . 12 , 2 . 78 , P = 0 . 014 ) were significantly associated with having a high level of overall knowledge of Yf . Agro-pastoralism as an occupation compared to farming was associated with having a low level of overall knowledge of Yf ( AOR = 0 . 51 , 95% CI , 0 . 33 , 0 . 79 , P = 0 . 003 ) . The findings indicate that most of the study community members had a low level of overall knowledge of Yf , especially about its cause , mode of transmission and preventive methods . Thus , there is a need to increase people’s knowledge and practices regarding the cause , mode of transmission and preventive methods like avoiding mosquitoe breeding sites beside vaccination through various strategies like disseminating information through community health extension workers and community leaders in the study area . Yellow fever ( Yf ) remains a major public health problem in Africa since the 1930s , especially in the endemic areas of equatorial rain forest , the moist savanna and the dry savanna areas [1] , despite the availability of effective vaccine for this disease . Previously , it was estimated that Yf causes 200 , 000 cases and 30 , 000 deaths annually , of which over 90% occurred in Africa [1] . According to a recent estimate , there were 130 , 000 Yf cases and 78 , 000 deaths in Africa for the year 2013 [2] . The disease is considered as one of the common re-emerging diseases in South America and many African countries like Democratic Republic of Congo , Sudan , Cameroon , Chad , Senegal , Côte d’Ivoire , Uganda , Sierra Leone , Ethiopia and Angola , where there is a low vaccine coverage or vaccine had waned [2–5] . In Ethiopia , large and small outbreaks of Yf occurred repeatedly since the 1960s and more recently , between November 2012 and October 2013 Yf outbreak occurred and resulted in many cases and deaths in South Omo area , southern Ethiopia , the same area which was highly affected by the outbreak in the 1960s [6–8] . The virus is transmitted to humans through the bite of widely distributed different species of the Aedes and Haemagogus mosquitoes . Thus , studies suggest that all areas in Africa where environmental conditions are suitable for mosquitoes breeding can be considered as areas at high risk of transmission; and the resurgence of Yf will continue unless vaccination is supported by effective mosquitoes control [9 , 10] . Moreover , it is suggested that control of mosquito-borne diseases like Yf requires effective participation of the local community [11] . Hence , assessing information on what a community knows about Yf would contribute to the efforts to design appropriate control strategies in addition to increasing access for vaccine . In this study , knowledge , attitudes and practices of local community about Yf was assessed in South Omo area , Southern Ethiopia , where Yf outbreak was recently reported . The study protocol was approved by the Institutional Review Board ( IRB ) of the Aklilu Lemma Institute of Pathobiology ( ALIPB ) , Addis Ababa University . The aim of the study was explained to each of the participant and verbal consent was obtained because most of the study participants were illiterate . Each participant was interviewed independently and the collected information was kept confidential . The study was conducted in South Omo Zone , one of the 13 zones in Southern Nations , Nationalities and Peoples' Region ( SNNPR ) . The Zone is located at about 750 km to the south of Addis Ababa and borders with Kenya on the south , a country which reported repeated outbreaks of arboviruses . The zone has eight districts with a total population of 573 , 435 , and the majority of the inhabitants are practicing agro-pastoralism . Detailed information on the study Zone and the study population has been described elsewhere [12] . Among the six districts where Yf outbreak occurred between 2012 and 2013 , two adjacent districts ( Bena Tsemay and Debub Ari ) were purposely selected for the present study based on the reported number of cases and deaths from the two areas [8] . Debub Ari district is located around Jinka town , approximately at 15 km to the North of Jinka . It has 50 small administrative units ( kebeles ) with a total population of 237 , 988 . Among the 50 kebeles , four kebeles namely , Arkisha , Aykamer , Geza and Shepi were purposely selected for the present study based on the recent occurrence of Yf outbreak in the area . Bena-Tsemay district is found at 42 km to southeast of Jinka . The district has 32 kebeles with a total population of 74 , 853 . Among the 32 kebeles , three kebeles namely , Luka , Goldia and Shaba-Argemenda were purposely selected for the present study based on the report of Yf outbreak occurrence . More detailed information on the study districts including a map showing the affected areas by the Yf outbreak has been described elsewhere [8] . Between March and May 2017 , a community-based cross-sectional survey was conducted in the selected kebeles of the two districts . To our knowledge , there was no previous information on the level of community knowledge about Yf in the study area . Thus , assuming that 50% of adults will have high level of knowledge about Yf with 95% confidence in the estimate , 5% degree of accuracy , design effect of 1 . 5 and 90% response rate , a minimum sample size of 634 study participants would be required for the study . Prior to data collection , a list of all the households in the selected kebeles was obtained from each of the respective kebele leaders . Based on the number of households in each of the selected kebeles , the pre-estimated sample size ( 634 ) was proportionally distributed . The required number of participants from each kebele was selected using systematic random sampling . The participants were eligible if they were residents of that kebele , age over 18 years , a husband/wife ( or the responsible person ) in the selected households , apparently healthy and volunteer to be interviewed . Structured questionnaire was developed in English , based on information from available literatures regarding community knowledge , attitudes and practices about arboviruses from previous studies in other countries [13 , 14] . The questionnaire was translated into Amharic and pre-tested for clarity and acceptability in the study districts . During pre-testing , additional information was gathered and some of the questions were modified . The participants were interviewed about the public health importance , cause , mode of transmission , clinical symptoms , treatment , and preventive methods of Yf in their own local language ( Ari , Bena and Tsemay ) by trained health extension workers who were selected from each of the study kebeles . Each interview was made by a house–to-house visit . Information on the socio-demographic characteristics of the participants was also included in the questionnaire . The collected data were double-entered into a data entry file using EpiData software , V . 3 . 1 . The data were exported to Stata version 11 for statistical analysis . Pearson chi-square test was used to evaluate the statistically significant difference in the level of knowledge of signs and symptoms as well as sources of information about Yf , mode of transmission , practices of people and attitudes towards Yf between male and females . Bivariate and multivariable logistic regression analyses were performed to explore associations of socio-demographic characteristics of the study participants with increased odds of having higher levels of overall knowledge of Yf and to quantify the degrees of association using odds ratio . P-values below 5% were considered as indicators of statistical significance . The overall knowledge of the study participants about Yf was assessed using the following eight main questions: ( 1 ) able to mention jaundice and/or vomiting blood as the severe sign/symptoms of Yf , ( 2 ) able to identify that Yf is different from malaria , ( 3 ) able to know that the treatment for Yf is different from malaria treatment , ( 4 ) able to know that Yf is transmitted from a patient to another person through mosquitoes bite , ( 5 ) able to mention that a mosquito which transmits malaria is different from a mosquitoes that transmit Yf , ( 6 ) able to mention that Yf can be transmitted from monkeys to human through mosquitoes bite , ( 7 ) able to mention that Yf is preventable by vaccine , and ( 8 ) able to know that Yf has a vaccine . Response to these questions were added together to generate overall knowledge score ranging from 0 to 8 . A score of one was given to correct response , zero being used for incorrect/do not know response . Then , the response was categorized into a high , {those who scored 5 ( 60% cut-off point ) and above} and a low , {score 4 ( 50% ) and below } overall knowledge of Yf as previously described [15] . Community’s practices regarding preventive methods of Yf were assessed by asking questions such as Yf is a preventable disease , preventive methods for Yf and are you/your family vaccinated for Yf ( Table 4 ) . Gender difference in the proportion of answering correct response to each practice of question was evaluated using Pearson chi-square test . Similarly , study participants were asked questions such as “Yf is a public health problem” , “Yf is a newly occurred disease” , “Yf affects all age groups and Yf is a killer disease” ( Table 5 ) to assess their attitudes . The association of gender and attitudes towards the disease was evaluated using Pearson chi-square test . Table 1 shows the socio-demographic characteristics of the study participants . A total of 612 participants ( 55 . 9% males , age range from 18 to 87 years , mean age 33 . 36 years ) participated in the study from the two areas , with a response rate of 96 . 5% . Among the study participants , 388 ( 63 . 4% ) were recruited from the Debub Ari and the majority ( 65 . 2% ) of the participants were illiterate . Table 2 shows communty’s knowledge of signs/symptoms of Yf and their sources of information . Out of the 612 study participants , 508 ( 83 . 0% ) reported that they heard about Yf which is locally known as “a disease that causes vomiting blood” . The study participants reported that they heard about the disease from individuals who were sick from Yf ( 43 . 3% ) , followed from health workers ( 34 . 1% ) , friends ( 29 . 9% ) and they have heard/seen a person who died of Yf ( 23 . 8% ) . A larger proportion of male participants had information about Yf compared to female participants ( 86 . 5% vs 78 . 5% , X2 = 6 . 89 , p = 0 . 01 ) . The most commonly mentioned signs and symptoms of the disease were vomiting blood ( 84 . 0% ) , fever ( 45% ) and headache ( 43 . 8% ) . Regarding the cause of the disease , 112 ( 22 . 3% ) responded that they knew its cause . However , only 8 ( 7 . 8% ) mentioned a virus as the cause of Yf , while others mentioned bacteria/germ ( 42 . 2% ) or other factors ( 50 . 0% ) . Significantly higher proportion of participants from Debub Ari mentioned vomiting blood as the main symptoms of Yf compared to participants from Bena Tsemay ( 92 . 1% vs 69 . 0% , X2 = 44 . 39 , P<0 . 001 ) . Similarly , a significantly higher proportion of participants from Debub Ari reported jaundice as the symptom of Yf compared to those from Bena Tsemay ( 43 . 1% vs 24 . 6% , X2 = 16 . 44 , P< 0 . 001 ) . On the other hand , a high proportion of participants from Bena Tsemay area mentioned bloody diarrhea as the main symptom of Yf compared to those from Debub Ari ( 56 . 7% vs 18 . 2% , X2 = 76 . 08 , p< 0 . 001 ) . Majority ( 90 . 4% ) of the study participants said that Yf is different from malaria and its treatment also different from malaria treatment ( 72 . 5% ) . Relatively , a higher proportion of males ( 76 . 0% ) than females ( 67 . 3% ) reported that the treatment for Yf is different from the treatment for malaria ( X2 = 5 . 25 , P = 0 . 07 ) . Community’s knowledge regarding mode of transmission of Yf is summarized in Table 3 . Less than half ( 41 . 9% ) of the study participants said that Yf can be transmitted from a patient to another person . Among the 213 individuals who said that the disease can be transmitted , only 80 ( 37 . 6% ) mentioned mosquitoes bite as a mode of transmission from person to person . More than half ( 55 . 9% ) of the study participants thought that the disease can be transmitted from a patient to another person through breathing . A higher proportion of participants from Bena Tsemay mentioned mosquitoes bite as mode of transmission of the disease compared to those from Debub Ari ( 50 . 6% vs 28 . 2% , X2 = 11 . 02 , p = 0 . 001 ) . Few study participants also reported that they ever heard that this disease can be transmitted from monkeys to a person through mosquitoes bite or through drinking water contaminated with monkeys feces . Table 4 shows community’s practices regarding prevention of Yf as reported by the study participants . More than half ( 65 . 7% ) of the participants thought that Yf is a preventable disease . Among those who believed that Yf is a preventable disease , majority ( 84 . 1% ) mentioned vaccine as a preventive method . Most of the participants from Bena Tsemay believed that Yf is preventable through vaccine compared to those from Debub Ari ( 90 . 7% vs 79 . 9% , X2 = 6 . 88 , P = 0 . 01 ) . The participants mentioned stagnant water as the main breeding site for mosquitoes , and suggested avoiding stagnant water and use insecticide sprays as preventive methods of mosquitoes breeding . However , during the survey the research team did not ask the study participants /has not observed whether they practice avoiding stagnant water and use insecticide sprays as preventive methods of mosquitoes breeding . Table 5 shows attitudes of the study participants about the public health importance of Yf . Majority ( 86 . 2% ) reported that Yf is a public health problem in their area , and 69 . 4% thought that Yf is a newly occurred disease in the area . A higher proportion ( 78 . 8% ) of participants from Debub Ari considered Yf as a newly emerged disease compared to those from Bena Tsemay ( 52 . 8% ) ( X2 = 37 . 24 , P< 0 . 001 ) . Some of the participants suggested high rain fall ( 13 . 1% ) and drought ( 4 . 5% ) as the factors for the occurrence of the disease , while most ( 65 . 3% ) of them had no idea about the factors contributing to its occurrence . The majority also believed that the disease affects all age groups ( 93 . 3% ) , it is a killer ( 97 . 2% ) and not easily treatable ( 62 . 3% ) . A larger proportion ( 99 . 1% ) of participants from Debub Ari said that Yf is a killer disease compared to those from Bena Tsemay ( 93 . 9% ) ( X2 = 13 . 04 , P = 0 . 001 ) . Almost all ( 99 . 4% ) of the participants mentioned that if they suspect themselves or their families for Yf , they will visit health facility very soon . Table 6 shows the overall knowledge of the study participants about Yf . Among the 612 study participants , 221 ( 36 . 1% ) were considered as having a high level of overall knowledge about Yf . Having Educational level of at least 7th grade was significantly associated with a having high level of overall knowledge of the disease ( COR = 2 . 6 , 95%CI , 1 . 27 , 5 . 34 , P = 0 . 009 and AOR = 3 . 25 , 95%CI , 1 . 39 , 7 . 57 , p = 0 . 006 ) . Similarly , being resident of Bena-Tsemay district was associated with having a high level of overall knowledge of Yf compared to residents of Debub Ari ( AOR = 1 . 77 , 95% CI , 1 . 12 , 2 . 78 , P = 0 . 014 ) . Agro-pastoralism as an occupation compared to farming was associated with having a low level of overall knowledge of Yf ( COR = 0 . 65 , 95% CI , 0 . 45 , 0 . 94 , p = 0 . 022 , and AOR = 0 . 51 , 95% CI , 0 . 33 , 0 . 79 , P = 0 . 003 ) . Although Yf is becoming one of the most re-emerging mosquito-borne viral diseases in many African countries including the present study area , the findings of the present study showed that people living in endemic areas do not have adequate knowledge about its cause and mode of transmission though they consider it as one of the killer diseases . Although most of the study participants acknowledged vaccination as the main preventive method of Yf , some of the study participants complained that they did not get adequate information about the specific vaccine/for what disease they are vaccinated which affects the positive attitudes of individuals toward Yf vaccine . Thus , there is a need to increase people’s knowledge and practices regarding the cause , mode of transmission and preventive methods like avoiding mosquitoe breeding sites beside vaccination through various strategies like disseminating information through community health extension workers and community leaders .	Yellow fever is becoming one of the most important re-emerging mosquito-borne viral diseases in many African countries despite the availability of an effective vaccine . Hence , assessing information on what a community knows about Yellow fever would contribute to the design of appropriate control strategies in addition to increasing access for vaccine . In this study , we assessed knowledge , attitudes and practices of local community about Yellow fever in South Omo area , southern Ethiopia , where outbreaks have occurred repeatedly since the 1960s . We found that the study community members had low knowledge about the cause and mode of transmission of the disease though they knew that it is a killer and affects all age groups . More than half of the study participants believed that the disease can be transmitted from a patient to another person through breathing . In the present study area , providing information to community members through community health extension workers regarding the role of mosquitoes in the transmission of this disease , and teaching what to do to minimize mosquitoes bite in understandable way would be helpful to increase their awareness about Yellow fever .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "body", "fluids", "pathology", "and", "laboratory", "medicine", "vomiting", "immunology", "tropical", "diseases", "parasitic", "diseases", "animals", "vaccines", "preventive", "medicine", "physiological", "processe...	2018	Community knowledge, attitudes and practices on Yellow fever in South Omo area, Southern Ethiopia
Prion strains are characterized by differences in the outcome of disease , most notably incubation period and neuropathological features . While it is established that the disease specific isoform of the prion protein , PrPSc , is an essential component of the infectious agent , the strain-specific relationship between PrPSc properties and the biological features of the resulting disease is not clear . To investigate this relationship , we examined the amplification efficiency and conformational stability of PrPSc from eight hamster-adapted prion strains and compared it to the resulting incubation period of disease and processing of PrPSc in neurons and glia . We found that short incubation period strains were characterized by more efficient PrPSc amplification and higher PrPSc conformational stabilities compared to long incubation period strains . In the CNS , the short incubation period strains were characterized by the accumulation of N-terminally truncated PrPSc in the soma of neurons , astrocytes and microglia in contrast to long incubation period strains where PrPSc did not accumulate to detectable levels in the soma of neurons but was detected in glia similar to short incubation period strains . These results are inconsistent with the hypothesis that a decrease in conformational stability results in a corresponding increase in replication efficiency and suggest that glia mediated neurodegeneration results in longer survival times compared to direct replication of PrPSc in neurons . Prion diseases are a group of transmissible , fatal neurodegenerative diseases , which include Creutzfeldt-Jakob disease in humans , bovine spongiform encephalopathy in cattle , and scrapie in sheep . The prion agent is comprised mainly , if not entirely , of PrPSc which is an abnormal isoform of the host encoded prion protein , PrPC [1] , [2] , [3] , [4] , [5] , [6] . Prion propagation is thought to occur in a two-step process where PrPSc first binds to PrPC followed by a conformational conversion of PrPC to PrPSc [7] , [8] , [9] . This conversion results in a change in physical properties of PrPC that include an increase in β-pleated sheet content , decreased solubility in non-denaturing detergents and increased resistance to proteolytic degradation [3] , [10] , [11] . Prion strains are operationally defined by characteristic incubation periods and neuropathological features that are maintained upon experimental passage [12] , [13] . The distribution of PrPSc in organs and neuronal populations can differ between strains , suggesting that PrPSc has a distinct strain-specific cellular tropism [14] , [15] , [16] , [17] . The initial uptake of PrPSc by different cell-lines appears to be independent of the particular strain [18] , [19] and suggests that cellular factors are responsible for prion strain tropism [17] , [20] , however , this has not been confirmed in vivo [21] . Prion strain diversity may be encoded by unique strain-specific conformations of PrPSc [15] , [22] , [23] , [24] , [25] , [26] . Consistent with this , strain specific differences in the molecular weight of PrPSc following limited PK digestion , the relative resistance of PrPSc to degradation by PK , the relative alpha helical and beta sheet content of PrPSc , the resistance of PrPSc to PK digestion in increasing concentrations of a protein denaturant ( i . e . conformational stability ) , and the aggregation state of PrPSc have been observed [15] , [22] , [27] , [28] . The mechanisms underlying how strain-specific conformations of PrPSc result in the distinct biological properties of disease are poorly understood . The published reports on the relationship between the conformational stability of PrPSc and the length of the incubation period of disease between prion strains are contradictory . In murine prion strains and during adaptation of synthetic prions , a decrease in the conformational stability of PrPSc correlates with a corresponding decrease in the incubation period [5] , [24] , [29] , [30] . One explanation for this observation is that a decrease of PrPSc stability increases PrPSc fragmentation resulting in an increase in agent replication that produces a correspondingly shorter incubation period [31] , [32] , [33] . Consistent with this , a decrease in Sup35 fiber stability corresponds to an increased rate of fibril fragmentation in yeast prions [33] , [34] . These data contrast with what has been observed in hamster-adapted prion strains . Short incubation period prion strains have PrPSc that is conformationally more stable compared to PrPSc from strains with a relatively longer incubation periods in hamsters [27] . However , a direct comparison between PrPSc replication rate and conformational stability has not been investigated . Both the prion strain and the cell type infected can influence the processing of PrPSc . Studies of sheep infected with different prion strains , either naturally or experimentally , have identified strain-specific patterns of PrPSc truncation in both neurons and glia [35] , [36] . Within a given strain the PrPSc truncation pattern can differ between glia and neurons suggesting that factors in addition to the conformation of PrPSc contribute to PrPSc truncation . While it is thought that replication in neurons is more important to disease development compared to glia , the effect of strain-specific processing of PrPSc in these cell types is less clear [37] , [38] , [39] . To better understand the strain specific relationship between the agent and the host , we evaluated PrPSc amplification efficiency , conformational stability of PrPSc , and susceptibility of PrPSc to endogenous proteolytic processing in vivo in several cell types , of eight hamster-adapted prion strains . Our data indicate that short incubation period strains have correspondingly more efficient replication , a higher conformational stability , and intrasomal accumulation of PrPSc in neurons compared to long incubation period strains . These data suggest that the relationship between agent replication and clearance influence the progression of disease . Brain tissue from hamsters at terminal disease infected with either the HY TME , 263K , HaCWD , 22AH , 22CH , 139H , DY TME or ME7H agents was digested with proteinase K and 250 µg equivalents were analyzed by Western blot ( Figure 1 ) . Western blot analysis of PrPSc indicated that the unglycosylated PrPSc glycoform of each strain migrated at 21 kDa with the exception of DY PrPSc , which migrated at 19 kDa ( Figure 1A , Table 1 ) . The abundance of PrPSc was determined for each strain ( n = 4 ) and there was less than a 25% difference in the abundance of PrPSc per µg brain equivalent for each prion strain analyzed ( Figure 1B ) . To determine if differences exist in the rate of PrPSc replication between strains , protein misfolding cyclic amplification ( PMCA ) was performed on eight hamster-adapted prion strains . Brain homogenates were prepared from animals at the clinical stage of disease or from an uninfected ( mock ) negative control and serial 10-fold serial dilutions of these homogenates were analyzed by Western blot prior to ( Figure 2A and 2C ) or after one round of PMCA ( Figure 2B and 2D ) . PMCA reactions that were initially seeded with 500 to 5×10−2 µg eq of HY TME infected brain homogenate resulted in detectable amplification of PrPSc , but amplification was not detected in PMCA reactions seeded with lower concentrations of HY brain homogenate ( Figure 2B ) . One round of PMCA using brain homogenate from DY TME infected animals amplified PrPSc to detectable levels in reactions that were initially seeded with 500 or 50 µg eq , but was not detected in PMCA reactions seeded with lower concentrations ( Figure 2D ) . This was the general trend , as the short incubation period strains HY TME , 263K , and HaCWD resulted in detection of amplified PrPSc in reactions seeded with lower ug eq of brain homogenate compared to the longer incubation period strains 22AH , 22CH , 139H , DY TME , and ME7H ( Figure S1 , Table 1 ) . These data demonstrate that the efficiency of PrPSc amplification corresponds with the incubation period for the prion strains that were analyzed . The conformational stability of PrPSc for each of the eight hamster-adapted prion strains was determined using either SDS or Gdn-HCl to denature PrPSc . The [SDS]1/2 half values ( % w/v ) segregated into two groups corresponding to the incubation period of the strain . The short incubation period strains HY TME , 263K , and HaCWD have a [SDS]1/2 values of 1 . 14±0 . 03 , 1 . 04±0 . 06 and 0 . 78±0 . 02 respectively compared to the long incubation period strains 22AH , 22CH , 139H , DY TME , and ME7H , which have [SDS]1/2 values of 0 . 53±0 . 04 , 0 . 46±0 . 02 , 0 . 50±0 . 01 , 0 . 53±0 . 05 and 0 . 44±0 . 02 respectively ( Table 1 , Figure 3 , Figure S2 ) . Similarly , there was a corresponding decrease in the [Gdn-HCl]1/2 values with an increase in the incubation period . To demonstrate that the reduction in PrPSc was not due to an inhibition of PrPSc binding to the PVDF membrane due to the presence of SDS or Gdn-HCl , the PK digestion step of the conformational stability assay was omitted which resulted in the detection of PrPSc ( data not shown ) . Overall , this data demonstrates that PrPSc from the short incubation period strains is more stable than PrPSc from the long incubation period strains ( Table 1 , Figure 3 , Figure S2 ) . Immunohistochemistry was performed on CNS tissue of hamsters using a panel of six monoclonal anti-PrP antibodies whose epitopes span the length of the hamster PrP protein ( Table 2 ) . Immunohistochemistry using this panel of six anti-PrP antibodies on mock-infected tissue sections containing red nucleus neurons failed to detect PrPSc , indicating the specificity of the antibodies for PrPSc ( Figure S3 ) . PrPSc deposits were detected perineuronally and within the neuropil of the red nucleus with every anti-PrP antibody tested in animals infected in the sciatic nerve with the DY TME agent at clinical disease suggesting the presence of full length PrPSc ( Figure 4A to 4F ) . Within the soma of neurons , the three antibodies whose epitopes are at the N terminus of PrP ( 8B4 , BE12 , and POM3 ) failed to detect PrPSc deposition ( Figure 4A to 4C ) . The three antibodies whose epitopes are located toward the C-terminal region of PrP ( 3F4 , 6H4 , and POM19 ) occasionally identified PrPSc deposition within the soma of these neurons , however , these DY PrPSc deposits appeared diffuse and faint compared to the PrPSc immunoreactivity in the neuropil ( Figure 4D to 4F , Table S1 ) . This same pattern of PrPSc distribution was also observed in VMNs of the lumbar spinal cord , and the neurons in the interposed nucleus , red nucleus and hind limb motor cortex throughout the course of disease and in animals inoculated by the i . c . route at clinical disease ( data not shown ) . The distribution of PrPSc in the red nucleus of hamsters i . c . inoculated with the long incubation period strains 22AH , 22CH , 139H or ME7H was indistinguishable from DY TME agent infected animals ( Figure S4 , Table S1 ) . To extend these studies on a short incubation period strains , we performed PrPSc immunohistochemistry on the red nucleus from animals infected in the sciatic nerve with the HY TME agent at clinical disease to ensure a direct comparison could be made with DY TME agent infected animals . The deposition of PrPSc in the neuropil and soma of neurons was similar to what was observed following infection with the DY TME agent and the other the long incubation period strains using the anti-PrP antibodies whose epitopes are located N-terminal to the HY PrPSc PK cleavage site ( Figure 4 , Panels G–H ) . However , when using antibodies located C-terminal to the HY PrPSc PK cleavage site , intrasomal and perinuclear PrPSc deposition was detected that was similar in intensity to PrPSc deposition in the neuropil ( Figure 4I to 4L , Table S1 ) . This HY TME specific pattern of PrPSc deposition was observed in animals inoculated by either the sciatic nerve or i . c . routes of inoculation at early and late time points post-infection and in the same brain regions that were examined in the DY TME infected animals . This pattern of HY PrPSc truncation was also observed in animals inoculated with the short incubation period strains 263K and HaCWD by the i . c . route at clinical disease ( Figure S4 , Table S1 ) . These data reveal similarities in the PrPSc deposition patterns in the neuropil and somata of neurons of animals infected with either long or short incubation period strains . The absence of PrPSc immunoreactivity using antibodies directed against the N-terminal regions of PrPSc suggests that truncated PrPSc is present in these cells . To investigate this possibility , serial sections of red nucleus from clinically-ill HY TME infected hamsters were processed using either the BE12 or POM3 antibodies whose epitopes are N-terminal and C-terminal to the HY PrPSc PK cleavage site respectively . Fiduciary marks , such as blood vessels and white matter tracts , were used to increase the likelihood that the same neurons were analyzed in both sections . The BE12 antibody detected punctate HY PrPSc deposits in the neuropil and perineuronally in the red nucleus but failed to detect intrasomal PrPSc ( Figure 5A ) . However , the POM3 antibody detected coarse , intrasomal PrPSc deposits in these same three neurons ( Figure 5B , arrowheads ) . Additionally , the intrasomal PrPSc formed large perinuclear aggregates ( Figure 5B ) . This demonstrates that the loss of N-terminal epitopes of PrPSc and the aggregation of the C-terminal PrPSc fragments occurs within the same neuron . The deposition of PrPSc in astrocytes and microglia was investigated using the same panel of anti-PrP antibodies in combination with anti-GFAP and anti-Iba-1 , which label astrocytes and microglia , respectively . As negative controls , reactive astrogliosis , microgliosis or PrPSc immunoreactivity was not detected in mock-infected animals ( Figure S5A to S5F ) . Additionally , non-specific binding of the monoclonal antibodies or the fluorescently conjugated secondary antibodies was not detected ( Figure S5G and S5H ) . The anti-PrP antibodies 8B4 , BE12 , and POM3 failed to detect PrPSc within astrocytes ( Figure 6A to 6C ) , while the antibodies 3F4 , 6H4 , and POM19 detected coarse punctate PrPSc deposits in astrocytes of hamsters infected with the DY TME agent at clinical disease ( Figure 6D to 6F , Table S1 ) . The anti-PrP antibodies POM3 , 3F4 , 6H4 , and POM19 detected PrPSc within astrocytes , while the antibodies 8B4 and BE12 failed to detect PrPSc within astrocytes of hamsters infected with the HY TME agent at clinical disease ( Figure 6G to 6L , Table S1 ) . The same PrPSc truncation pattern detected in astrocytes of HY TME infected animals was also observed in animals infected with the 263K , HaCWD , 22AH , 22CH , 139H or ME7 agents ( Figure S6 , Table S1 ) . The anti-PrP antibodies 8B4 and BE12 failed to detect PrPSc in microglia , while the anti-PrP antibodies POM3 , 3F4 , 6H4 , and POM19 detected coarse punctate PrPSc deposits within these cells ( Figure 7 , Figure S7 , Table S1 ) . Here we show that short incubation period strains have a more stable PrPSc conformation when compared to long incubation period strains . PrPSc conformational stability assays using either Gdn-HCl or SDS as the denaturant found the same relationship between the conformational stability of PrPSc and incubation period of disease indicating that this relationship is independent of the denaturant used ( Table 1 ) . This relationship between PrPSc conformational stability and incubation period is consistent with previous work examining the conformational stability of purified PrPSc from hamster-adapted prion strains [27] . In contrast to what is observed in hamsters , a decrease in the PrPSc conformational stability correlates with a reduction in the incubation period in mice [24] , [29] , [40] . The results in murine systems suggest that decreasing PrPSc stability increases the fragmentation of PrPSc therefore allowing in the generation of more PrPSc surfaces for PrPC to bind resulting in an increased rate of PrPSc formation and subsequently shortening of the incubation period . Consistent with this hypothesis , studies examining Sup35 , PrP , Tau , α-synuclein , and ß-amyloid demonstrate that less stable fibrils have a higher propensity to undergo breakage , thereby creating new seeds for conversion [33] , [34] , [41] , [42] , [43] , [44] , [45] . The PrPSc conformational stability data presented here suggest that conformationally stable PrPSc may also be more susceptible to fragmentation . SDS , like Gdn-HCl , can increase the susceptibility of PrPSc to PK digestion and inactivate the agent [27] , [46] , [47] . Since treatment of PrPSc that is enriched using detergent extraction and ultracentrifugation with SDS results in the disaggregation of PrPSc and the production of smaller PrPSc particles , SDS can affect the aggregation state of PrPSc [32] , [48] . Therefore , the higher concentration of SDS required to increase the susceptibility of PrPSc to PK digestion of short incubation period strains may be due to increased PrPSc particle size compared to long incubation period strains . Short incubation period strains have more efficient PrPSc amplification compared to long incubation period strains . We used PMCA to determine the relative efficiency of PrPSc conversion between hamster strains . We have previously shown that PMCA of HY and DY TME recapitulates the strain-specific properties of PrPSc and faithfully replicates the HY and DY TME agents [49] . In examining the eight hamster strains we found that the efficiency of PrPSc amplification correlated with the strains respective incubation periods , as the strains with more efficiently replicating PrPSc had a shorter incubation period compared to long incubation period strains ( Table 1 ) . This is consistent with cell-free conversion experiments that demonstrated a faster rate of HY PrPSc synthesis compared to the rate of DY PrPSc synthesis [50] . The data presented here also indicate that conformationally more stable PrPSc amplifies more efficiently compared to less stable PrPSc . Interestingly , the short incubation period strain HaCWD has conformationally less stable PrPSc in SDS compared to 263K and HY PrPSc which corresponded with a lower amplification efficiency compared to the two other short incubation period strains . A possible explanation for the increased amplification efficiency of PrPSc from the short incubation period strains is that this PrPSc is more likely to fragment due to its large PrPSc particle size compared to the longer incubation period strains used in this study . Alternatively , a minor subpopulation of PrPSc that is conformationally less stable may be responsible for the highly efficient PrPSc replication that was observed . This conformationally less stable subpopulation may be masked by an excess of conformationally more stable PrPSc that replicates with lower efficiency [32] , [51] , [52] . Strain and cell-specific variations in the proteolytic processing of PrPSc have been observed in both brain tissue and cultured cells [36] , [53] , [54] , [55] , [56] . The results presented here are consistent with these findings and additionally suggest a relationship between the extent of truncation of PrPSc within the soma of neurons and the strains respective incubation periods . The short incubation period strains , HY TME , 263K , and HaCWD , contained a longer portion of C-terminal protein intact and a large punctate deposition of PrPSc within the soma of neurons , compared to the longer incubation period strains suggesting a strain-specific clearance of PrPSc ( Figures 8 , Figure 9 ) . Furthermore , the low immunoreactivity of PrPSc in ME7H infected animals observed with all six anti-PrP antibodies and within all three cell types examined ( Figure 8 , Table S1 ) may represent the more efficient clearance of PrPSc in both neurons and glia for this particular strain and account for its significantly longer incubation period . However , we cannot exclude the possibility that the inability to detect intense PrPSc immunoreactivity in the soma of neurons from animals inoculated with the long incubation period strains is due to a failure of PrPSc transport to the soma . This strain-specific truncation pattern was only observed in neurons , as the same N-terminally truncated PrPSc species was detected in astrocytes and microglia for all strains examined , with the lone exception of the loss of the POM3 epitope from DY PrPSc within astrocytes ( Figure 8 , Figure 9 ) . These data support the hypothesis that direct infection of neurons leads to more rapid death of neurons resulting in shorter incubation periods , compared to indirect neuronal death via infection of astrocytes and microglia [37] , [38] , [57] . The results presented here suggest the strain-encoded relationship between PrPSc replication , stability and processing in neurons is predictive of the incubation period of disease ( Figure 9 ) . Here we show that strains with a short incubation period have conformationally stable PrPSc that replicates efficiently . The fast replication and stable PrPSc may be responsible for the accumulation of PrPSc in the soma of neurons resulting in a shorter incubation period . The long incubation period strains displayed relatively less efficient PrPSc replication and less stable PrPSc . In these strains , the combination of a slower replicating agent and PrPSc that is less stable may result in neurons to be able to more effectively cleared of PrPSc resulting in longer incubation periods . All procedures involving animals were approved by the Creighton University Institutional Animal Care and Use Committee and were in compliance with the Guide for the Care and Use of Laboratory Animals . Sciatic nerve and intracerebral inoculations of the HY or DY TME agents were performed on male Syrian golden hamsters ( Harlan-Sprague-Dawley , Indianapolis , IN ) as previously described [21] . Groups of five hamsters were inoculated in the sciatic nerve or intracerebrally with 1 or 25 µl , respectively , of a 1% ( wt/vol ) brain homogenate from animals at the terminal stage of disease infected with either the HY TME , DY TME , 263K , HaCWD , 22AH , 22CH , 139H , or ME7H agents . Hamsters were observed three times per week for the onset of clinical signs as described previously [58] . Incubation period was calculated as the number of days between inoculation and onset of clinical signs . Tissue from infected and mock-infected hamsters was collected for either immunohistochemistry ( IHC ) or Western blot analysis . For IHC analysis animals were anesthetized with isoflurane and perfused transcardially with 50 ml of 0 . 01 M Dulbecco's phosphate-buffered saline followed by 75 ml of McLean's paraformaldehyde-lysine-periodate ( PLP ) fixative as previously described [21] , [59] . Brain was immediately removed and placed in PLP for 5 to 7 h at room temperature prior to paraffin processing . For Western blot analysis , animals were sacrificed by CO2 asphyxiation , and the brain was rapidly removed and flash frozen and stored at −80°C . Brain tissue and spinal cord tissue were homogenized to 10% w/v in Dulbecco's Phosphate Buffered Saline ( DPBS ) without Ca++ or Mg++ ( Mediatech , Herndon , VA ) containing protease inhibitors ( Roche Diagnostics Corporation , Indianapolis , IN ) by passing the tissue through a 26 g needle , followed by a 30 second incubation in a cup horn sonicator ( Fisher Scientific , Atlanta , GA ) . The tissue was diluted to 5% w/v in DPBS containing proteinase K ( PK ) at a final concentration of 1 U/ml ( Roche Diagnostics Corporation , Indianapolis , IN ) and incubated at 37°C for 1 hour with constant agitation . The PK digestion was terminated by incubating the samples at 100°C for 10 minutes . SDS-PAGE and Western blot analysis were performed as described previously [49] using the anti-PrP antibody 3F4 ( 1∶600; Chemicon; Billerica , MA ) . The blot was developed with Pierce Supersignal West Femto Maximum Sensitivity Substrate according to manufactures instructions ( Pierce , Rockford , IL ) and imaged in the linear range of detection on a Kodak 4000R Imaging Station ( Kodak , Rochester , NY ) and analysis was performed using Kodak Molecular Imaging Software v . 5 . 0 . 1 . 27 ( New Haven , CT ) as described previously [49] . Protein misfolding cyclic amplification ( PMCA ) was performed as previously described [1] , [49] . Briefly , uninfected brain was homogenized to 10% ( w/v ) in ice-cold conversion buffer [phosphate buffer saline ( pH 7 . 4 ) containing 5 mM EDTA , 1% v/v Triton X-100 , and complete protease inhibitor tablet ( Roche Diagnostics , Mannheim , Germany ) ] using a Tenbroeck tissue grinder ( Vineland , NJ ) . The brain homogenate was centrifuged at 500× g for 30 seconds and the supernatant was stored at −80°C . PMCA was performed with a Misonix 3000 sonicator ( Farmingdale , NY ) with the sonicator output set to level 6 with an average output of 156 watts for each sonication cycle . All PMCA reactions were replicated in triplicate . One round of PMCA consisted of 144 cycles of a five-second sonication followed by a ten-minute incubation at 37°C . Before each PMCA round , an aliquot was placed at −80°C as an unsonicated control . Samples seeded with prion-infected brain homogenate were replicated using a minimum of three individual hamster brains to control for variation between animals . Samples containing uninfected brain homogenate in conversion buffer alone were included in every round of PMCA as a negative control . The amplification efficiency was calculated as the reciprocal of the µg equivalent of last dilution of prion-infected brain homogenate that resulted in detectable amplified PrPSc following one round of PMCA . Brain homogenates [7 . 5% ( w/v ) ] were diluted in either SDS ( Fischer Scientific , Atlanta GA ) to a final concentration of 0 , 0 . 25 , 0 . 5 , 0 . 75 , 1 , 1 . 25 , 1 . 5 , 1 . 75 , or 2% ( w/v ) or in Gdn-HCl ( Sigma-Aldrich , St . Louis , MO ) to a final concentration of 0 , 0 . 25 , 0 . 5 , 0 . 75 , 1 , 1 . 25 , 1 . 5 , 1 . 75 , or 2 molar and were immediately heated at 70°C for 10 min . Proteinase K was added to 0 . 0625 U/ml ( Roche Diagnostics , Indianapolis , IN ) and the samples were incubated at 37°C for 15 min while shaking . All samples were brought to 200 µl in DPBS and the concentration of PrPSc was determined using a 96-well immunoassay as described previously [60] . Gdn-HCL and SDS treated samples were performed in quintuplicate . Serial two-fold dilutions of brain homogenate each strain were performed in triplicate to ensure that the PrPSc levels of the SDS or Gdn-HCl treated samples were in the linear range of PrPSc detection . Denaturation curves were generated by dividing the intensity of all samples by the average intensity of the 0% SDS or Gdn-HCl samples . A sigmoidal dose response curve with variable slope was fitted to the standardized values ( Prism statistical software , GraphPad , La Jolla , CA ) . The [SDS]1/2 and [Gdn-HCl]1/2 values is the percentage of SDS or molarity of Gdn-HCl required for a 50% reduction in the PrPSc signal intensity . PrPSc IHC was performed as previously described [21] , [59] . Briefly , 7 µm tissue sections were deparaffinized and incubated in 95% formic acid ( Sigma-Aldrich , St . Louis , MO ) followed by blocking of endogenous peroxidases by immersion in 0 . 3% H2O2 in methanol . Following blocking of non-specific staining with 10% horse serum , sections were incubated overnight with an anti-PrP antibody ( Table 2 ) at 4°C . The sections were then incubated with a biotinylated horse anti-mouse immunoglobulin G conjugate and subsequent incubation with the ABC-horseradish peroxidase elite ( Vector Laboratories , Burlingame , CA ) staining kit . Sections were developed using 0 . 05% w/v 3 , 3′-diaminobenzidine ( Sigma-Aldrich , St . Louis , MO ) in tris-buffered saline containing 0 . 0015% H2O2 and counterstained with hematoxylin ( Richard Allen Scientific , Kalamazoo , MI ) . Microscopy was performed using a Nikon i80 microscope ( Nikon , Melville , NY ) and images were captured using DigiFire camera and ImageSys digital imaging software ( Soft Imaging Systems , GmbH ) and processed using Adobe Photoshop CS2 v9 . 0 . 1 ( Adobe Systems Inc . , San Jose , CA ) . For double immunofluorescence , tissue sections were deparaffinized and treated with 95% formic acid as described above . The tissue sections were blocked with 10% goat serum in tris-buffered saline for 30 minutes at room temperature , followed by overnight incubation at 4°C with the same panel of anti-PrP monoclonal antibodies ( Table 2 ) and anti-glial fibrillary acidic protein ( GFAP; 1∶16 , 000; Dako; Carpinteria , CA ) or anti-ionized calcium binding adaptor molecule 1 ( Iba-1; 1∶500; Abcam; Cambridge , MA ) . Sections were then incubated with both Alexa Fluor goat anti-mouse 546 and Alexa Fluor goat anti-rabbit 488 ( 1∶500; Invitrogen; Carlsbad , CA ) secondary antibodies for one hour at room temperature . Slides were cover slipped using ProLong Gold antifade reagent with DAPI ( Invitrogen; Carlsbad , CA ) . Fluorescent images were captured on a Zeiss LSM 510 META NLO confocal scanning system ( Carl Zeiss Jena; Jena , Germany ) using a Plan Neo 40× 1 . 3-NA DIC oil objective . Excitation of the Alexa Fluor antibodies and DAPI was achieved using an Argon laser at 488 nm , a Helium Neon laser at 543 nm , and a Coherent Chameleon near infrared tunable Ti:Sapphire laser . To increase the signal to noise ratio , each line was scanned 4 times and averaged . The pinhole aperture for each channel was adjusted so that an optical slice of 1 . 0 µm was imaged . In the profile view for each image , the line tool was used to draw an arbitrary line , and the relative fluorescent intensities along that line were compared to determine intracellular staining . Semi-quantitative measurements of PrPSc immunoreactivity was performed as previously described [61] . Briefly , captured images were randomized and the relative magnitude of neuropil , intraneuronal , intra-astrocytic , and intra-migroglial PrPSc immunoreactivity was classified as absent ( 0 ) , slight ( 1 ) , moderate ( 2 ) , or striking ( 3 ) from a minimum of 6 observations by three independent observers . The PrPSc immunoreactivity scores were compared between strains and cell types and analyzed by two-way analysis of variance and Bonferroni post-tests for statistical significance ( p<0 . 05 ) . These tests were performed using the Prism 4 . 0 ( for Macintosh ) software program ( GraphPad Software , Inc . , San Diego , CA ) .	Prion diseases are a group of infectious fatal neurodegenerative diseases that affect animals including humans . This unique infectious agent is the result of a post-translational conformational change of the normal form of the prion protein , PrPC , to an infectious form of the prion protein , PrPSc . Different strains of the infectious agent result in characteristic incubation periods and neuropathological features within a single host species . These strain-specific differences in disease outcome are likely due to strain-specific conformations of PrPSc , though the mechanisms by which different conformation can affect prion strain properties are not understood . The aim of this study was to investigate the relationship between the biochemical properties of PrPSc to the corresponding neuropathological characteristics of eight hamster-adapted prion strains . Our findings indicate that PrPSc from short incubation period strains were more efficiently replicated , had a more stable conformation , and were observed to be more resistant to clearance from the soma of neurons compared to prion strains with a relatively long incubation period . These results suggest the progression of prion disease is influenced by the balance between replication and clearance of PrPSc in neurons .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "infectious", "diseases/prion", "diseases" ]	2011	The Strain-Encoded Relationship between PrPSc Replication, Stability and Processing in Neurons is Predictive of the Incubation Period of Disease
Brazilian spotted fever ( BSF ) , caused by the bacterium Rickettsia rickettsii , has been associated with the transmission by the tick Amblyomma sculptum , and one of its main hosts , the capybara ( Hydrochoerus hydrochaeris ) . During 2015–2019 , we captured capybaras and ticks in seven highly anthropic areas of São Paulo state ( three endemic and four nonendemic for BSF ) and in two natural areas of the Pantanal biome , all with established populations of capybaras . The BSF-endemic areas were characterized by much higher tick burdens on both capybaras and in the environment , when compared to the BSF-nonendemic areas . Only two tick species ( A . sculptum and Amblyomma dubitatum ) were found in the anthropic areas; however , with a great predominance of A . sculptum ( ≈90% of all ticks ) in the endemic areas , in contrast to a slight predominance of A . dubitatum ( ≈60% ) in the nonendemic areas . Tick species richness was higher in the natural areas , where six species were found , albeit with a predominance of A . sculptum ( ≈95% of all ticks ) and environmental tick burdens much lower than in the anthropic areas . The BSF-endemic areas were characterized by overgrowth populations of A . sculptum that were sustained chiefly by capybaras , and decreased populations of A . dubitatum . In contrast , the BSF-nonendemic areas with landscape similar to the endemic areas differed by having lower tick burdens and a slight predominance of A . dubitatum over A . sculptum , both sustained chiefly by capybaras . While multiple medium- to large-sized mammals have been incriminated as important hosts for A . sculptum in the natural areas , the capybara was the only important host for this tick in the anthropic areas . The uneven distribution of R . rickettsii infection among A . sculptum populations in highly anthropic areas of São Paulo state could be related to the tick population size and its proportion to sympatric A . dubitatum populations . Brazilian spotted fever ( BSF ) , caused by the bacterium Rickettsia rickettsii , is the deadliest tick-borne disease of the New World . The disease is endemic in many parts of southeastern Brazil , especially in the state of São Paulo , where 978 laboratory-confirmed cases were recorded from 2001 to 2018 , of which 489 ( 50% ) had a fatal outcome ( official data from the São Paulo State Health Secretary ) . In North America , where the R . rickettsii-caused disease is known as Rocky Mountain spotted fever ( RMSF ) , multiple strains of R . rickettsii ( including less virulent ones ) are known to occur . In contrast , a highly virulent strain prevails in Central and South America , which has been linked to the higher fatality rates of BSF , when compared to RMSF [1] . In addition , the greatest fatality of BSF is also evidenced by its neglected status in Brazil , such as the unavailability in the country of parenteral doxycycline , considered the first-choice medication to treat severe BSF or RMSF presenting vomiting or altered mental status [2 , 3] . During this century , several studies have elucidated key factors in the epidemiology of BSF in southeastern Brazil , where R . rickettsii is transmitted to humans mainly by the tick Amblyomma sculptum . Besides being a competent vector , A . sculptum larvae , nymphs and adults are partially refractory to R . rickettsii infection , and less than half of the infected females transmit R . rickettsii to their offspring ( transovarial transmission ) [4–6] . This fact , associated to the higher mortality and lower reproductive performance of infected ticks , when compared to uninfected mates ( 5 , 6 ) , causes infection of A . sculptum by R . rickettsii in BSF-endemic areas to be very low , usually <1% [7–10] . Within this scenario , mathematical models have indicated that an A . sculptum population is not able to sustain a R . rickettsii infection for successive tick generations without the creation of new cohorts of infected ticks via horizontal transmission on vertebrate rickettsemic hosts ( amplifying hosts ) [11 , 12] . In this case , the capybara ( Hydrochoerus hydrochaeris ) , the largest living rodent in the world , has been pointed out as the major amplifying host of R . rickettsii for A . sculptum in most of the BSF-endemic areas of southeastern Brazil [11 , 13 , 14] . However , it is important to note that the tick Amblyomma dubitatum has also been frequently found infesting capybaras in southeastern Brazil , albeit with no direct role on BSF-epidemiology [7–10 , 14 , 15] . During the last four decades , the state of São Paulo has undergone extensive anthropogenic modifications in its landscape due to a rapid expansion of agricultural crops ( especially sugar cane ) , deforestation , and creation of artificial water bodies [16 , 17] . Such modifications have favored capybara reproduction primarily by higher food availability by agriculture ( e . g . , sugar cane , corn fields ) and because of the local extinction of natural predators ( e . g . , the jaguar Panthera onca ) , in human-modified landscapes , leading to an increment on the extension and density of its populations [17–19] . Because capybara is considered to be the main host for A . sculptum in such landscapes [10 , 15 , 20] , and at the same time an efficient R . rickettsii amplifying host [13] , the increase of BSF incidence in the state of São Paulo during the last three decades has been associated to the afore mentioned anthropogenic modifications [11 , 14 , 21] . While the expansion of capybaras and their ticks have been well recognized in the state of São Paulo during the last decades , many of these human-modified landscapes have remained free of R . rickettsii circulation , despite of the established presence of capybaras and A . sculptum [22–24] . Since the reasons determining the establishment of R . rickettsii in a capybara-sustained A . sculptum population are not well understood , the present study aimed to characterize and to quantify in time and space the tick fauna in capybaras and in the habitats where these rodents occur among different human-modified landscapes in the state of São Paulo , either endemic or nonendemic for BSF . Differences in the tick fauna composition could be one of the possible multiple reasons driving the uneven distribution of R . rickettsii among different A . sculptum populations . In order to confirm the endemic or nonendemic status of each area , we determined the serological profile of the capybaras against a battery of rickettsial antigens , including R . rickettsii . For comparison purposes , we performed the same capybara and tick evaluations in pristine areas of the Pantanal biome of Brazil , where capybaras live in natural habitats in which landscape has suffered only minimal anthropogenic alterations and from where BSF has never been reported . Our results might provide some clues for a better understanding on the main epidemiological characteristics of the BSF-endemic areas associated to capybaras . This study has been approved by the Institutional Animal Care and Use Committee ( IACUC ) of the Faculty of Veterinary Medicine of the University of São Paulo ( approval number 5948070314 ) , in accordance with the regulations/guidelines of the Brazilian National Council of Animal Experimentation ( CONCEA ) . Field capture of capybaras and collections of ticks were authorized by the Brazilian Ministry of the Environment ( permit SISBIO Nos . 43259–6 ) and by the São Paulo Forestry Institute ( Cotec permit 260108–000 . 409/2015 ) . All study areas were inhabited by capybaras , and were classified into the following three epidemiological categories: ( i ) BSF-endemic areas–highly anthropic areas ( human-modified landscape ) in the state of São Paulo , where human cases of BSF have been recently confirmed and the transmissions have been epidemiologically associated with A . sculptum . Three BSF-endemic areas were sampled: Area 1 in the municipality of Piracicaba , Area 2 in the municipality of Americana , Area 3 in the municipality of Araras , all located in transition areas of the biomes Savannah and Atlantic Forest; ( ii ) BSF-nonendemic areas–highly anthropic areas ( human-modified landscape ) in the state of São Paulo , however , with no history of BSF . Four BSF-nonendemic areas were sampled: Areas 4 and 5 in the municipality of Pirassununga , located in a transition area of the biomes Savannah and Atlantic Forest , Area 6 in the municipality of Ribeirão Preto , located in the Savannah biome , and Area 7 in São Paulo municipality , in the Atlantic Forest biome; and ( iii ) natural areas–low anthropic areas ( natural landscape ) in the Pantanal biome , with no history of BSF . Two natural areas were sampled: Area 8 in Poconé municipality , state of Mato Grosso , and Area 9 in Corumbá municipality , state of Mato Grosso do Sul . We sampled capybaras and ticks in 3 BSF-endemic areas , 4 BSF-nonendemic areas , and 2 natural areas ( S1 Table , Fig 1 and S1 Text ) . During 2015–2018 , we performed capture of capybaras in all study areas by using 16 to 20 m2–corrals baited with sugar cane and green corn . Once closed in the corral , every animal was physically restrained by a net catcher and anaesthetized with an intramuscular injection of a combination of ketamine ( 10 mg/kg ) and xylazine ( 0 . 2 mg/kg ) . Under anesthesia , animals were weighed in an electronic balance ( Pesola model PCS0300 , Hatton Rock , UK ) and identified with a subcutaneous microchip ( Alflex model P/N 860005–001 , Capalaba , Australia ) . In Mato Grosso do Sul , corrals were not effective , thus capybaras were captured by anesthetic darting via a CO2-injection rifle ( Dan-Inject model JM Standard , Denmark ) by injecting the same chemicals above . Capybaras were sexed and aged as follows: young ( <10 Kg ) , juvenile ( 10–35 Kg ) , and adult ( >35 Kg ) , following Vargas et al . [25] . From each capybara , we collected blood samples through the femoral artery or cranial vena cava , and the serum was separated by centrifugation and stored frozen at -20°C until serological analysis ( described below ) . Because many capybaras were heavily infested by ticks , we standardized a 3-min random collection of ticks from the entire body of every capybara . During the 3-min period , any tick on sight was collected , regardless of the size or part of capybara body . These ticks were brought to the laboratory , where they were identified to species following current literature [20 , 26 , 27] . After recovering from anesthesia , capybaras were released at the same capture site . Capybara sera were tested by immunofluorescence assay ( IFA ) as described elsewhere [22] using rickettsial crude antigens derived from Vero cells ( provided by the Instituto Adolfo Lutz , São Paulo , Brazil ) infected with each of the following five Rickettsia species known to infect ticks in Brazil: R . rickettsii strain Taiaçu [28] , Rickettsia parkeri strain At24 [29] , Rickettsia amblyommatis strain Ac37 [30] , Rickettsia rhipicephali strain HJ5 [31] , and Rickettsia bellii strain Mogi [28] . In addition , a sixth rickettsial antigen consisted of C6/36 cells ( provided by the Instituto Adolfo Lutz , São Paulo , Brazil ) infected with Rickettsia felis strain Pedreira , was also implemented [32] . In each slide , a serum previously shown to be non-reactive ( negative control ) and a known reactive serum ( positive control ) from a previous study [13] were included . Slides were incubated with fluorescein isothiocyanate-labeled sheep anti-capybara IgG ( produced by the Centro de Controle de Zoonoses , São Paulo City ) . For each sample , the endpoint titer reacting with each of the six Rickettsia antigens was determined . Sera showing an antibody titer to a Rickettsia species at least fourfold higher than the titers observed for the other Rickettsia species were supposed to be homologous to the first Rickettsia species or to a very closely related genotype , as previously determined for several animal species [33–35] , including capybaras [22] . Host questing ticks were collected in each of the nine study areas ( Fig 1 , S1 Table ) during four consecutive years . Our schedule for collection of free-living ticks was based on the seasonal dynamics of A . sculptum , which is known to complete one generation per year , with larvae peaking during autumn , nymphs during spring , and adults during summer [36–38] . Therefore , between May 2015 and January 2019 , ticks were collected in each area during the larval peak ( May-June ) , nymphal peak ( August-September ) and adult peak ( January-February ) of every year . In each area at each time point , a 1 m2 white flannel was dragged over 800 m of animal trails . With this procedure , every dragging event on a given area represented the number of ticks for an 800 m2-sampled area . Collected nymphs and adults were immediately put in plastic vials containing 70% ethanol , except for a few adult ticks that were placed in dry plastic vials and taken alive to the laboratory , where they were kept frozen at -80°C until molecular analysis for Rickettsia ( described below ) . Every time a larval cluster was captured by dragging , the cluster was immediately picked up with a 5 cm-large transparent plastic adhesive tape , which was then stuck on a white paper that was put within a sealed plastic bag and taken to the laboratory . Adult and nymphal ticks were counted individually and identified to species according to [20 , 26 , 27] . Larvae were counted as number of clusters , since it was assumed that each larval cluster represented the offspring of one engorged female [38 , 39] . Larval taxonomic identification consisted of comparing side-by-side individuals of a larval cluster with laboratory-reared larvae of A . sculptum and A . dubitatum , following established criteria [40 , 41] . Host-questing ticks were also collected by dry ice traps following Szabó et al . [39]; however , this method was used only at one time point ( August 2015 ) in each area , and had to be discontinued due to logistic difficulties . In each area , 20 to 40 dry ice traps were set at 10 m intervals along the same trails that were sampled by dragging . Collected ticks were immediately placed in 70% ethanol , and taken to the laboratory for taxonomic identification as described above . Frozen unfed adult ticks , previously collected by dragging in each area , were thawed and individually submitted to DNA extraction by the guanidine isothiocyanate-phenol technique [42] . Extracted DNA samples were firstly tested by a conventional PCR protocol targeting the tick mitochondrial 16S rRNA gene , as previously described [43] , in order to certify successful DNA extraction . Then , viable DNA samples ( those positive by the tick 16S rRNA PCR assay ) were tested by a Taqman real-time PCR assay targeting the rickettsial gltA gene , as described [5] . The sensitivity of this PCR assay was determined to be 1 DNA copy of R . rickettsii [44] . Positive samples by this Taqman real-time PCR were tested by two protocols of conventional PCR , one targeting a 401-bp fragment of the rickettsial gltA gene [44] , and one heminested PCR assay targeting the ompA gene; this latter protocol consisted of a first reaction targeting a 631-bp fragment , and a second targeting a 532-bp fragment , as described [45] . PCR products were DNA-sequenced and the resultant sequences were submitted to BLASTn analyses ( www . ncbi . nlm . nih . gov/blast ) in order to confirm the identity of the Rickettsia species . The proportions of seroreactive capybaras for R . rickettsii were compared between the nine sampled areas by the Chi-square test . Endpoint titers for the six Rickettsia species were compared between BSF-endemic and BSF-nonendemic areas by the Mann-Whitney test . For the ticks collected on capybaras , we calculated the prevalence ( No . infested hosts / No . examined capybaras x 100 ) , and the mean abundance of tick infestation ( total number of collected ticks / number of examined capybaras ) according to [46] in each of the 9 study areas . Density of host-questing ticks was calculated for the total dragged area ( TDA ) . For this purpose , TDA = number of dragging events performed in one area during the four years of study x 800m2 ( considering that each dragging event encompassed an 800m2 area in each of the study areas ) . Then , the tick density ( TD ) , represented by number of host-questing ticks per 1 , 000 m2 , was calculated by: TD = total number of collected ticks / TDA x 1 , 000 m2 . These calculations were applied to the two most abundant tick species , A . sculptum and A . dubitatum . TD was calculated separately for larvae , nymphs , and adult ticks for the whole study period , as well as for the period of larval peak ( all collections during autumn ) , nymphal peak ( all collections during winter ) , and adult peak ( all collections during summer ) . For statistical analyses , we pooled the tick data for each of the three epidemiological categories: ( i ) BSF-endemic areas , ( ii ) BSF-nonendemic areas , and ( iii ) natural areas . The Chi-square test was used to compare the differences between the proportions of ticks on capybaras or host-questing ticks between the three epidemiological categories ( BSF-endemic , BSF-nonendemic , natural areas ) . The Lilliefors test implemented in the PAST 3 . 19 program was used to analyze the normality of the data in order to choose the appropriate statistical test for each situation . Values of mean abundance of tick infestation were analyzed using the Kruskal-Wallis non-parametric test . For all tests , the level of significance was 5% . Analyses were performed by using PAST Version 3 . 19 and BioEstat 5 . 0 . A total of 347 capybaras were captured during the 2015–2018 period . The number of captured capybaras per each of the nine areas varied from 14 to 73 ( mean: 38 . 6 ) . Since we sampled during four consecutive years , some individuals were captured twice among different years; recaptured capybaras represented 0 to 50% of the total number of captures in each area ( Table 1 ) . Removing recaptured animals , the total number of different individuals sampled in this study was 287; however , we considered all recaptures as different units for our analyses of serology and tick infestations ( described below ) , since recaptures occurred in years different from the first capture . The 347 captured capybaras were represented by 94 ( 27% ) males and 253 ( 73% ) females . They were aged as 27 ( 7 . 8% ) young , 70 ( 20 . 2% ) juvenile , and 250 ( 72% ) adults ( Table 1 ) . Among the 347 captured capybaras , sera were collected from 337 , which were tested by IFA against six Rickettsia species ( Table 2 ) . Considering the three epidemiological categories , the proportions of seropositive capybaras for R . rickettsii in the 3 BSF-endemic areas ( 88 to 98% ) were significantly higher ( P<0 . 05 ) than the proportions in the 4 BSF-nonendemic areas ( 14 to 38% ) and in the natural area of Corumbá ( 47% ) ; the proportions for R . rickettsii in the later 5 areas were statistically similar ( P>0 . 05 ) . While the proportion of seropositive capybaras in the natural area of Poconé ( 100% ) was similar ( P>0 . 05 ) to the BSF-endemic areas , the endpoint titers to R . rickettsii were quite different , with much higher values for the BSF-endemic areas ( S2 Table ) . In fact , 56 capybaras of the 3 BSF-endemic areas had endpoint titers for R . rickettsii at least fourfold higher than the titers for the remaining five Rickettsia species , indicating that these capybaras were likely infected by R . rickettsii ( Table 2 ) . Using these same criteria , no capybara from either BSF-nonendemic or natural areas were considered to have been infected by R . rickettsii , whereas 12 , 36 , and 11 capybaras from BSF-endemic , BSF-nonendemic , and natural areas , respectively , were probably infected by R . bellii; and two , one and nine capybaras from BSF-endemic , BSF-nonendemic , and natural areas , respectively , were probably infected by R . amblyommatis . In addition , five capybaras from the natural area of Poconé were likely infected by R . parkeri ( Table 2 ) . While the endpoint titers of the capybaras from the BSF-endemic areas were significantly higher for R . rickettsii than for the remaining five Rickettsia species , in the BSF-nonendemic areas the endpoint titers were significantly higher for R . bellii ( Fig 2 ) . Capybaras in the seven anthropic areas of the state of São Paulo ( BSF-endemic and BSF-nonendemic areas ) were infested by two tick species , A . sculptum and A . dubitatum . Among the two natural areas in the Pantanal biome , A . sculptum was the only species infesting capybaras in the Corumbá area , while A . sculptum , A . dubitatum and Amblyomma triste were found on capybaras in the Poconé area . Tick prevalence on capybaras was 100% in all seven anthropic areas , and 95% in natural areas , where only three capybaras did not have any tick . For data comparison , we excluded the 27 young capybaras ( Table 1 ) because they were usually infested by low number of ticks ( mean abundance: 15 . 3 ticks/capybara ) , when compared to the overall mean abundance of 31 . 8 ticks/capybara among adults and juveniles . The overall mean abundance of ticks was significantly higher ( P<0 . 05 ) in the BSF-endemic areas ( 40 . 9 ticks/capybara ) than in the BSF-nonendemic areas ( 33 . 7 ticks/capybara ) , which was also significantly higher ( P<0 . 05 ) than the mean abundance in the natural areas ( 7 . 7 ticks/capybara ) ( Table 3 and Fig 3 ) . Amblyomma sculptum was the dominant tick species in the BSF-endemic areas , where they represented 85% ( 4 , 091/4 , 821 ) of all ticks collected from capybaras ( Table 3 ) and had the highest mean abundance values ( Fig 4 ) . In contrast , A . dubitatum was the dominant species in the BSF-nonendemic areas , where they encompassed 69% ( 3390/4914 ) of all ticks collected on capybaras ( Table 3 ) and had the highest mean abundance values ( Fig 4 ) . Mean abundance values of either A . sculptum or A . dubitatum were significantly different ( P<0 . 05 ) between BSF-endemic and BSF-nonendemic areas ( Fig 5 ) . In the natural areas , A . sculptum was the dominant tick species ( 88%; 380/429 ) ; however , with a mean abundance of only 6 . 8 A . sculptum ticks/capybara , contrasting to the mean abundance of 34 . 7 and 10 . 2 A . sculptum ticks/capybara in the BSF-endemic and BSF-nonendemic areas , respectively . Host questing ticks were collected along four consecutive years , during the activity peaks of larvae ( autumn ) , nymphs ( winter ) and adults ( summer ) of A . sculptum in each year . In the anthropic area of São Paulo ( area no . 7 ) , dragging was performed only at two instances during the first year ( 2015 ) ; thereafter , this area had to be discontinued from the study due to a highly fatal outbreak of fascioliasis that decimated the capybara population of the area [47] , what certainly impacted the environmental tick burdens in the subsequent years . In the remaining six anthropic areas of the state of São Paulo ( areas no 1 to 6 ) , dragging was not possible only during the adult tick season of the 2016 summer , due to personal problems beyond our control . In the two natural areas ( Poconé and Corumbá ) , dragging was not possible at two occasions in each area due to logistic problems related to road conditions and access to both areas . Overall , dragging was performed at 11 occasions in each of the three BSF-endemic areas , at 11 occasions in two BSF-nonendemic areas ( Pirassununga A and Pirassununga B ) , at 20 occasions in the BSF-nonendemic area of Ribeirão Preto , at 19 occasions in the natural area of Poconé , and at 16 occasions in the natural area of Corumbá . A total of 21 , 670 ticks were collected by dragging in all areas during the study . In the anthropic areas , only two tick species were identified , A . sculptum and A . dubitatum . In the natural areas , the following six tick species were collected: A . sculptum , A . dubitatum , Amblyomma parvum , A . triste , Ambyomma ovale , and Ornithodoros rostratus ( Table 4 ) . In the BSF-endemic areas , the proportions of A . sculptum and A . dubitatum were 92% ( 10 , 425/11 , 305 ) and 8% ( 880/11 , 305 ) , respectively . In contrast , the proportions of A . sculptum and A . dubitatum in the BSF-nonendemic areas were 43% ( 3 , 688/8 , 633 ) and 57% ( 4 , 945/8 , 633 ) , respectively . These proportions were significantly different ( P<0 . 05 ) between the two epidemiological categories . In the natural areas , A . sculptum represented 98% ( 1 , 694/1 , 732 ) of all collected ticks , a proportion significantly different ( P<0 . 05 ) from the two categories of anthropic areas . Tick density ( TD ) values , represented by the number of host-questing ticks per 1 , 000 m2 , were calculated for the two most abundant tick species , A . sculptum and A . dubitatum , in all areas . Grouping all dragging occasions during the four years , TD of A . sculptum larvae , nymphs and adults were higher in BSF-endemic areas than in BSF-nonendemic and natural areas , with some significant ( P<0 . 05 ) differences ( Table 5 and Fig 6 ) . Comparisons of TD values of A . sculptum with those of A . dubitatum revealed significantly higher values ( P<0 . 05 ) for larvae , nymphs and adults of the former species in the BSF-endemic areas . On the other hand , A . sculptum and A . dubitatum had similar ( P>0 . 05 ) larval , nymphal and adult TD among the BSF-nonendemic areas ( S2 Table ) . Because only two A . dubitatum nymphs were collected by dragging in the natural areas ( Table 4 ) , TD values were not statistically compared with A . sculptum in these areas . Grouping the dragging occasions that were performed during only the autumn season ( larval peak ) of the four years , larval TD of A . sculptum in the BSF-endemic areas ( 31 . 8 larval clusters/1 , 000 m2 ) was ≈2 times higher ( P<0 . 05 ) than in the BSF-nonendemic areas ( 15 . 1 larval clusters/1 , 000 m2 ) , and ≈3 times higher ( P<0 . 05 ) than in the natural areas ( 9 . 1 larval clusters/1 , 000 m2 ) . On the other hand , larval TD of A . dubitatum in the BSF-endemic areas ( 2 . 7 larval clusters/1 , 000 m2 ) was about ≈4 times lower ( P<0 . 05 ) than in the BSF-nonendemic areas ( 13 . 8 larval clusters/1 , 000 m2 ) ( Table 6 and Fig 6 ) . During the winter season ( nymphal peak ) of the four years , nymphal TD of A . sculptum in the BSF-endemic areas ( 703 . 9 nymphs/1 , 000 m2 ) was ≈3 . 5 times higher ( P<0 . 05 ) than in the BSF-nonendemic areas ( 199 . 6 nymphs/1 , 000 m2 ) , and ≈9 times higher ( P<0 . 05 ) than in the natural areas ( 86 . 7 nymphs/1 , 000 m2 ) . On the other hand , nymphal TD of A . dubitatum in the BSF-endemic areas ( 72 . 1 nymphs/1 , 000 m2 ) was about ≈3 . 5 times lower ( P<0 . 05 ) than in the BSF-nonendemic areas ( 262 . 3 nymphs/1 , 000 m2 ) ( Table 7 and Fig 6 ) . During the summer season ( adult peak ) of the four years , adult TD of A . sculptum in the BSF-endemic areas ( 311 . 8 adults/1 , 000 m2 ) was ≈6 times higher ( P<0 . 05 ) than in the BSF-nonendemic areas ( 52 . 3 adults/1 , 000 m2 ) , and ≈14 times higher ( P<0 . 05 ) than in the natural areas ( 22 adults/1 , 000 m2 ) . On the other hand , adult TD of A . dubitatum in the BSF-endemic areas ( 12 . 1 adults/1 , 000 m2 ) was about ≈4 times lower ( P<0 . 05 ) than in the BSF-nonendemic areas ( 50 . 9 adults/1 , 000 m2 ) ( Table 8 and Fig 6 ) . During autumn , winter and summer , TD values of larvae , nymphs and adults , respectively , were always higher ( P<0 . 05 ) for A sculptum than for A . dubitatum in the BSF-endemic areas , but at the same time similar ( P>0 . 05 ) between the two tick species in the BSF-nonendemic areas ( Tables 6–8 ) . A total of 8 , 790 ticks were collected by 220 dry ice traps during August 2015 in all areas of the study ( S3 Table ) . In the BSF-endemic areas , the mean number of A . sculptum ticks per trap was 35 . 4 , ≈3 . 5 times higher than the mean number of A . dubitatum ticks per trap ( 9 . 4 ) . In the BSF-nonendemic areas , the mean numbers of A . sculptum and A . dubitatum per trap were similar , 25 . 2 and 23 . 8 , respectively . In the natural areas , we collected on average 6 . 3 A . sculptum/trap and 0 . 3 A . dubitatum/trap , in addition to two other species , A . parvum ( 0 . 1 ticks/trap ) and O . rostratus ( 0 . 2 ticks/trap ) . A total of 216 host-questing adults of A . sculptum [24 from each of 8 sampled areas ( ticks from the BSF-nonendemic area of São Paulo were not included ) ] were tested individually for the presence of rickettsial DNA , but none of them contained rickettsia . On the other hand , rickettsial DNA was successfully amplified in 4 ( 29% ) out of 14 A . parvum ticks from Corumbá , and in 2 ( 17% ) out of 12 A . triste from Poconé . The rickettsial DNA amplified from all four A . parvum ticks was identified as ‘Candidatus Rickettsia andeanae’; i . e . , their gltA and ompA partial sequences were 100% identical to the corresponding sequences of this agent in GenBank ( KF030931 and KF030932 , respectively ) . The gltA and ompA partial sequences generated from the two A . triste ticks were 100% identical to the corresponding sequences of R . parkeri strain Portsmouth ( CP003341 ) . Tick mitochondrial 16S rRNA gene-DNA was successfully amplified from all Rickettsia-negative samples , validating our PCR-negative results . A four-year field evaluation demonstrated marked differences of capybara and environmental tick burdens between the three epidemiological classifications of the sampled areas , namely BSF-endemic , BSF-nonendemic , and natural areas . Among the nine sampled areas , only three were classified as BSF-endemic , based primarily on recent records of human cases of the disease ( S1 Table ) . In order to certify on the presence/absence of R . rickettsii circulation between capybaras and ticks in all nine study areas , we performed serological analyses of capybaras against antigens of the most frequent Rickettsia species that have been reported in Brazil . While cross-reactive antibodies between Rickettsia species are often observed , testing a vertebrate serum against the possible Rickettsia species known to occur in a given area is ideal because often homologous antibody titers are higher than heterologous antibody titers . In some cases , the differences in titers may be great enough ( ≥ fourfold higher ) to differentiate among the rickettsial species potentially stimulating the immune response [33 , 48] . Based on these criteria , the BSF endemic status of areas no . 1 to 3 ( Piracicaba , Americana , and Araras ) was corroborated by endpoint titers at least fourfold higher for R . rickettsii than for other Rickettsia species in many of the tested capybaras . In fact , we have just reported a successful isolation of R . rickettsii from A . sculptum ticks that were parasitizing one of the capybaras that were captured in the BSF-endemic area of Piracicaba , corroborating local circulation of R . rickettsii between ticks and capybaras [10] . As expected , we had no serological evidence of R . rickettsii infection in the four BSF-nonendemic areas of São Paulo state . Actually , what we observed in these areas was serological evidence of other Rickettsia species , especially R . bellii . This result should be related to the predominance of A . dubitatum ticks in these areas , since it has been reported that most of the A . dubitatum populations are infected by R . bellii ( usually at high infection rates ) in multiple areas in the state of São Paulo , including some of the present study [40 , 44 , 49 , 50] . Similarly to the BSF-nonendemic areas , we did not find serological evidence of R . rickettsii circulation in the two natural areas; however , it was interesting to note that 100% of the capybaras from Poconé were seroreactive to both R . rickettsii and R . parkeri , with endpoint titers generally higher for the later . Our findings of R . parkeri-infected A . triste ticks in Poconé supports the serological evidence that some of the capybaras from this area have been infected by R . parkeri , since A . triste ticks were found infesting capybaras in that area . Finally , the few serological evidence of capybara exposure to R . amblyommatis could be related to the recent reports of R . amblyommatis infecting A . sculptum ticks [38 , 51] , including the Poconé area [52] , where we found seven capybaras with endpoint titers fourfold higher for R . amblyommatis . Our tick surveys clearly demonstrated that the BSF-endemic areas were characterized by tick burdens much higher than in the BSF-nonendemic areas , with A . sculptum encompassing the vast majority of the ticks on either capybaras or in the environment . In contrast , there was a predominance of A . dubitatum over A . sculptum in the BSF-nonendemic areas . Considering that both BSF-endemic and BSF-nonendemic areas had similar landscapes , one of the reasons driving the two distinct tick scenarios could be the size of the capybara population of each area . This hypothesis relies on a recent study performed within another highly anthropic area of the state of São Paulo , which in 2006 was not endemic for BSF , had 78 capybaras , and dry ice traps captured a mean of 0 . 7 A . sculptum/trap and 3 . 3 A . dubitatum/trap; in 2012 , the same area had become endemic for BSF , had 230 capybaras ( ≈3 times higher than in 2006 ) , and dry ice traps captured a mean of 33 A . sculptum/trap ( ≈47 times higher than in 2006 ) and 2 . 1 A . dubitatum/trap ( ≈0 . 3 times lower than in 2006 ) [53] . The authors concluded that the emergence of BSF in the area in 2012 was a consequence of the increase of the local capybara population , which in turn , provided the increment of the A . sculptum population . Unfortunately , the numbers of capybaras among the BSF-endemic and BSF-nonendemic areas were not available for comparisons during the present study . Indeed , further studies should be done in order to verify capybara demographic differences among the areas here prospected . Moreover , these studies should also focus on the reproduction rates of capybara groups , since recent mathematical models have proposed that the establishment of R . rickettsii in a capybara-sustained A . sculptum population is dependent on a high reproduction rate of this host species [11 , 54] . The predominance of A . dubitatum over A . sculptum could also have direct implications on the absence of R . rickettsii in BSF-non endemic areas , especially because populations of A . dubitatum have been found naturally infected by R . bellii throughout the state of São Paulo , usually at high infection rates [40 , 44 , 49 , 50] . One study showed that R . bellii-infected A . dubitatum ticks were partially refractory to R . rickettsii , and were not competent to pass R . rickettsii transovarially [55] . Thus , as long as A . dubitatum prevails in one area , R . rickettsii might not be able to establish an infection in either A . dubitatum or A . sculptum . In the case of A . sculptum , our results and the study of [53] showed that R . rickettsii was established only when there was an overgrowth population of A . sculptum , possibly because the proportion of R . rickettsii-infected A . sculptum ticks under natural conditions is always very low ( <1% ) [7–10] . Such assumption might allow us to speculate that any intervention resulting in a drastic reduction of the A . sculptum population would eliminate the R . rickettsii infection from the tick population . Different from the highly anthropic areas of the state of São Paulo , as much as six tick species were collected in the natural areas of the Pantanal biome . Such species richness was somewhat expected in pristine areas of this biome , where several species of medium- to large-sized mammals act as major hosts for ticks , including A . sculptum [20 , 56–60] . While A . sculptum was the dominant tick species in the natural areas , tick burdens were much lower than in the anthropic areas . Such findings highlight the ecological disequilibrium of the anthropic areas , where much higher tick burdens were associated to a single major host species , the capybara . Other factors that could be contributing for the BSF endemic or nonendemic status in A . sculptum-capybara associated areas in southeastern Brazil are inherent to the A . sculptum populations , namely their susceptibilities to R . rickettsii infection . This hypothesis relies on a recent study that compared the susceptibility of R . rickettsii infection among six populations of A . sculptum [61] . The authors showed that there were significant differences among the susceptibilities of the six tick populations , and suggested that it could be another factor driving the uneven distribution of R . rickettsii among the wide distribution of A . sculptum in southeastern Brazil . However , the mechanisms driving these different susceptibilities are yet to be determined . The relatively low number of sampled areas ( nine ) distributed among three contiguous biomes could be considered a main drawback of this study . Indeed , higher number of areas per epidemiological category would imply greater robustness to our results . However , the high similarity of our observations by epidemiological category supports our results . Actually , sampling of areas within three different biomes was chosen for their strong relation with spotted fever core features; capybaras and A . sculptum ticks . In fact , the endemic and non-endemic areas in São Paulo State were intensely anthropized green areas . Although they might have originally been rainforests or savannahs , their natural phytophysiognomies vanished and now share common environmental features characterized by a water source and low grassy areas with relatively few trees , all adequate for capybaras . Similar natural areas maintaining capybara populations had to be found for the control groups as well . Such pristine areas are non-existent in São Paulo State , and the Pantanal biome was the one that provided the most similar features , for example abundant water source and widespread capybara populations . More importantly , all the nine sampled areas are around the center of the wide range of A . sculptum [20] and away from the geographic boundaries of this tick species , precluding negative effects of extreme weather on our results for ticks . The BSF-endemic areas of the state of São Paulo were characterized by overgrowth populations of A . sculptum that were sustained chiefly by capybaras , and decreased populations of A . dubitatum . In contrast , the BSF-nonendemic areas with landscape similar to the endemic areas differed by having lower tick burdens and a slight predominance of A . dubitatum over A . sculptum , both sustained chiefly by capybaras . Higher species richness of ticks ( six species ) was found in the natural areas of Pantanal , although environmental tick burdens were lower than in the anthropic areas of São Paulo . While multiple medium- to large-sized mammals have been pointed out as important hosts for A . sculptum in the Pantanal , the capybara was the only important host for this tick species in the anthropic areas of the present study . The uneven distribution of the presence of R . rickettsii infection among A . sculptum populations in highly anthropic areas of the state of São Paulo could be related to the tick population size and its proportion in relation to sympatric A . dubitatum populations .	Brazilian spotted fever ( BSF ) , caused by the bacterium Rickettsia rickettsii , is the deadliest tick-borne disease of the New World . In southeastern Brazil , where 489 patients succumbed to the disease from 2001 to 2018 , R . rickettsii is transmitted to humans mainly by the tick Amblyomma sculptum , which uses the capybara ( Hydrochoerus hydrochaeris ) as its main host . During 2015–2019 , we captured capybaras and ticks in seven highly anthropic areas of São Paulo state ( three endemic and four nonendemic for BSF ) and in two natural areas of the Pantanal biome . The BSF-endemic areas were characterized by much higher tick burdens on both capybaras and in the environment , with a predominance of Amblyomma sculptum . In the BSF-nonendemic areas , another tick species , Amblyomma dubitatum , outnumbered A . sculptum . In the natural areas , six tick species were found; however , with much lower numbers than in the anthropic areas . The BSF-endemic areas were characterized by overgrowth populations of A . sculptum that were sustained chiefly by capybaras , and decreased populations of A . dubitatum . Results of this study support the idea that any intervention resulting in a drastic reduction of the A . sculptum population shall eliminate the R . rickettsii infection from the tick population , and consequently , prevent new BSF cases .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion", "Conclusions" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "ecology", "and", "environmental", "sciences", "ixodes", "pathology", "and", "laboratory", "medicine", "pathogens", "geographical", "locations", "microbiology", "rickettsia", "animals", "developmental", "biology", ...	2019	Epidemiology of capybara-associated Brazilian spotted fever
Apoptosis is regulated by several signaling pathways which are extensively linked by crosstalks . Boolean or logical modeling has become a promising approach to capture the qualitative behavior of such complex networks . Here we built a large-scale literature-based Boolean model of the central intrinsic and extrinsic apoptosis pathways as well as pathways connected with them . The model responds to several external stimuli such as Fas ligand , TNF-α , UV-B irradiation , interleukin-1β and insulin . Timescales and multi-value node logic were used and turned out to be indispensable to reproduce the behavior of the apoptotic network . The coherence of the model was experimentally validated . Thereby an UV-B dose-effect is shown for the first time in mouse hepatocytes . Analysis of the model revealed a tight regulation emerging from high connectivity and spanning crosstalks and a particular importance of feedback loops . An unexpected feedback from Smac release to RIP could further increase complex II formation . The introduced Boolean model provides a comprehensive and coherent description of the apoptosis network behavior . It gives new insights into the complex interplay of pro- and antiapoptotic factors and can be easily expanded to other signaling pathways . Apoptosis is the prototype of programmed cell death and an essential process in multicellular organisms . It is necessary during embryogenesis , tissue growth , differentiation and homeostasis as a protective mechanism to remove superfluous or malfunctioning cells from the organism [1]–[5] . Errors in cell death regulation can result in diseases like Alzheimer and Parkinson when uncontrolled apoptosis occurs or cancer if apoptosis is repressed [6] , [7] . Apoptosis can be induced by several signal transduction pathways that are tightly regulated and linked to other cellular events such as inflammatory responses and proliferation . The understanding of these signaling pathways is thought to provide novel solutions for the treatment of many diseases . However , a large number of participating components , their complex dependencies and multiple biological stimuli make the analysis of small network parts difficult and often less expressive . Therefore some mathematical models have already been presented covering broader structures . For example Huber et al . presented the web service APOPTO-CELL based on 52 ordinary differential equations [ODEs] to calculate the susceptibility of cells to undergo apoptosis in response to an activation of the mitochondrial apoptotic pathway [8] . The power of ODE based modeling concerning dynamic simulation and system analysis is without controversy . However , the use of ODE models for larger networks is limited due to limited biological data . The parameter identification for ODE models is in the very most cases dependent on quantitative measurements which still are a systems biology bottle neck . Another approach is the use of Petri nets [9] , [10] , however , the required input for parameterization is still relatively high due to the need of defining transition rules . In this study , we present a Boolean network of apoptosis . Boolean or logical networks are well suited to reproduce the qualitative behavior of extensive networks even with a limited amount of experimental data . Boolean logic is the algebra of two values , e . g . “1 and 0” or “true and false” or “on and off” [11] and was first shown to be applicable to electrical relay circuits [12] . Furthermore , it can also be applied to biological systems , and signal flow networks can be described reasonable by a logical approach [13] . The Boolean formalism is especially useful for qualitative representation of signaling and regulatory networks where activation and inhibition are the essential processes [14] . In a Boolean representation , the biological active state of a species can be translated into the “on” state whereas the inactive state is represented by the “off” state . Enzymes play the role of switching other enzymes and genes “on” and “off” . Applying Boolean algebra to a signaling network results in an interaction network , analogous to electrical circuits , which can be conveniently represented by logical interaction graphs . Boolean operations and graphs are described in detail in the Materials and methods section . There are different interesting approaches concerning the specific calculation and simulation of Boolean networks . Chaves et al . presented a hybrid model of the NF-κB module combining Boolean and ODE based modeling [15] . Calzolari et al . analyzed an apoptosis gene network with identical topology but different link strengths chosen by random distribution [16] . For the specific cell type of cytotoxic T lymphocytes Zhang et al . built a Boolean model relating the input antigen stimulation with the output apoptosis [17] . They use an asynchronous updating strategy and show multiple simulations with different updating orders . Recently , Mai et al . presented a Boolean apoptosis model including 40 nodes and connecting two inputs , namely TNF and growth factor , to the output DNA damage [18] . They calculated their network with the impressive number of 10 . 000 random initial states to simulate towards apoptosis or stable surviving . We chose a different approach to avoid some known problems concerning logical models . In this study , the logical steady state [LSS] of variables with a unique LSS for a given input setting is determined . For the computation of LSSs the software tool CellNetAnalyzer [CNA] is used . The propagation of signals through the network is thereby calculated by iterative derivation of partial LSSs for smaller subnets based on already identified partial LSS until no further ones can be found [19] . There is no need to simulate the network many times or to perform statistical analyses . CNA has previously been used to describe and analyze large-scale Boolean models of biological networks . This tool is also useful to predict and verify experimental data , examine the structure and the hierarchy of the system as well as the relevance of its components [19]–[21] . Not least , manual analysis and the identification of network wide dependencies become error-prone for large logical networks . Therefore , construction and analysis of the logical interaction hypergraph model is achieved more reliable in this study using CNA . Special features of the CNA are described in the Materials and methods section since they are used to reveal essential properties of the network structure and thereby deduce biological conclusions on the complex signaling network of apoptosis . The large-scale Boolean network constructed in this study is based on extensive literature research . It simulates apoptotic signal transduction pathways in response to various input stimuli and allows a comprehensive evaluation and analysis of the different pathways ( Figure 1 ) . We considered the intrinsic and extrinsic apoptotic pathways and their crosstalks as well as the survival and metabolic insulin pathways . We show that the extension and refinement of the logical formalism with multi-value logic and so called timescale constants allows the capturing of dynamical features such as threshold behavior , feedback loops and reaction delays and thereby a correct description of the global signaling behavior . The states of several network nodes are experimentally validated for different inputs in order to prove the coherence of the model . In this context a dose dependent effect of UV irradiation concerning apoptosis induction is demonstrated on mouse hepatocytes . Finally , the model is analyzed with regard to its internal connectivity and crosstalks with a special attention on significant feedback loops and gene regulatory effects . The model is a logical interaction hypergraph , which is a connection of logic gates , and comprises 86 nodes and 125 interactions ( Figure 1 ) . Abbreviations and descriptions of the network nodes are given in Text S1 . Text S1 also lists all equations of the model including the respective timescale constants , literature references and organisms from which the information was derived . Due to the number of included interactions in the model we refer to the given literature references for detailed information about the biological processes . There are eight input nodes , namely glucagon , insulin , TNF-α [TNF] , Fas ligand [FasL] , interleukin-1β [IL-1] , UV-B irradiation [UV] and two special nodes for applying Smac mimetics and for simulating type II apoptotic signaling . Smac mimetics are promising reagents that sensitize cells for apoptosis via the neutralization of inhibitor of apoptosis proteins ( IAPs such as XIAP , cIAP1 , cIAP2 , etc . ) [22] , [23] . They are considered as a separate node . The input node ‘Type 2 receptor ligand’ [T2RL] allows simulating apoptosis via the mitochondrial type II pathway in contrast to the type I pathway which proceeds via a direct activation of the caspase cascade [24] . The T2RL node is experimentally represented in this study by human Jurkat T cells treated with Fas ligand . Recently , the type I and type II pathways were shown to operate in the same cell type but under different culturing conditions suggesting that cells are able to switch between both ways depending on external stimuli [25] . However , the molecular mechanism of the switch itself has not yet been uncovered . Therefore , an additional node P representing some unknown protein or mechanism is introduced here to model the switch behavior . Another specialty is the ‘housekeeping’ node , which shall reproduce constitutively expressed genes ( Figure 1 , in green ) . The output node of the model is apoptosis . It was shown that dynamic processes can also be captured in logical networks by introducing time delays to the logical functions [13] . An equivalent function is provided in CNA where processes can be assigned to different timescales . These timescales are constants that specify in which state a certain node can become active . Simulating a network at timescale τ = x means that all interactions with a timescale constant τ≤x are considered , but interactions with timescale constant τ>x are omitted . The apoptosis model contains six timescales {τ} = {0 , 2 , 3 , 4 , 5 , 10} which are not numbered consecutively such that additional timescales can be easily inserted . Rapid , easily reversible signaling effects like phosphorylation that are based on fast protein interactions can thus be separated from long-term effects like gene expression and protein synthesis . However , we use the so called timescale function not only for an approximate discretization of signaling events to time segments but also to separate functional groups of interactions such as feedback loops . As we calculate the logical steady state , no transition rules for any updating strategy have to be assumed which would be afflicted with high uncertainty . There are no disadvantages connected with extensive defining of timescales concerning the simulation of the network . However , every timescale can be used to generate a snapshot of the network and accomplish its separate analysis . So for example , the topology of the network including only early signaling events or the specific influence of feedback loops can be analyzed by assigning separate timescales to them . Overall the introduction of timescales to the logical formalism allows to describe different signaling effects and gene regulatory mechanisms in one unifying model but to analyze them separately . All interactions of the apoptosis model with their respective timescales are listed in Text S1 . The first timescale τ = 0 is reserved for the housekeeping interactions that activate nodes which are constantly active and represent constitutively expressed genes . Timescale τ = 0 contains 7 interactions and symbolizes the state of the cell before stimulation . However , note that interactions of the housekeeping node with other nodes activated later are set to the later timescale . Also the input and output arcs are assigned to τ = 0 ( 11 interactions including multilevel inputs ) . On the second timescale τ = 2 only early TNF signaling events take place which include TNF signal transduction towards the formation of complex I ( 5 interactions ) . The internalization of complex I was described to be slow in comparison to other signaling processes . An additional timescale τ = 3 is assigned to further interactions of the TNF pathway that are required for complex II formation ( 5 interactions ) . 73 interactions referring to signaling transduction events except the early events of the TNF pathway take place at τ = 4 . An additional timescale τ = 5 is introduced to model feedback loops ( 9 interactions ) . Assigning a separate timescale to feedback loops allows their separate analysis which is very reasonable considering their impact on the system . The final timescale τ = 10 is reserved for modeling gene expression in response to signaling events and includes 15 interactions . As an example , some node values for different timescale scenarios after combined stimulation of the apoptosis model with TNF and smac-mimetics are shown in Table 1 . All references underlying the according interactions can be found in Text S1 . Note that the node complex2 is activated by the interaction RIP-deubi+FADD+comp1 = comp2 . The node FADD is set to level 1 by the housekeeping node on timescale τ = 0 . At timescale τ = 2 TNF receptor 1 is activated by the input TNF . The input smac-mimetics activates smac and thereby RIP-deubi at timescale τ = 3 . At timescale τ = 4 for example the complex smac-XIAP is build . In this setting there is no influence by feedback loops on timescale τ = 5 . At timescale τ = 10 upregulation of TRAF2 via NF-κB leads to complex1 formation and thereby to complex2 formation and finally apoptosis . A feedback loop is a circular path in a signed directed graph . In general , feedback loops provide a challenge for Boolean networks as they can lead to oscillations or multistability . In this case no definite logical steady state can be assigned for the affected nodes and they remain unevaluated in CNA . We excluded 13 interactions because of this implication from the logical steady state analysis which are listed and numbered in Text S1 . However , these reactions are included in all other computational analysis and can be included in logical steady state computation again very easily by removing the according check mark in the CNA menu Network Composer . The omission of these 13 interactions has the benefit of getting a definite simulation result for every possible input setting . Of course the network is changed thereby but this is not disadvantageous as the impact of these particular interactions on the logical steady state analysis is biologically not significant as discussed in the following paragraph . Four of these 13 interactions represent positive feedback loops which would even enhance an existing activation status in a dynamical model ( no . 7–9 , 13 ) ; however , the respective node is already in the “on” state in the logical model when the feedback would become active . As the Boolean model is not quantitative the feedback loop would not have an impact on the biological result anyway . Three negative feedbacks excluded from logical steady state computation involve the fine-tuning of C3p17 and Raf activity in a dynamical model , but they do not affect the activation level in the logical model for the same reason ( no . 2 , 5 , 6 ) . Five negative feedback loops govern NF-κB signaling back to its initial configuration and thereby inactivate NF-κB ( no . 1 , 3 , 10–12 ) and the negative feedback loop towards IRS-P inhibits the signal as well ( no . 4 ) . However , the switching off of the network is generally excluded in this model because we restrict ourselves to the critical events of apoptosis . Consequently the respective validation experiments described below are performed in the corresponding time period of the first response of every node . A promising feature of CNA is the possibility to use multi-value logic , which is equivalent to the discretization of the “on” state and was shown to be applicable to logical models of biological systems [13] . Biochemical decisions are often made in increments caused by thresholds that are essential for setting boundaries between different states in living cells . This is especially true for apoptotic processes [26]–[28] . We show here for a comprehensive network that the use of multi-value logic in the description of biological systems allows us to model several distinct active states . Multi-value nodes thereby don't substitute quantitative modeling , but the different node value levels are defined by qualitative properties . This is a general idea of our modeling approach and we name it the functional definition of node values . Assigning different effects to different active states is equivalent to biological threshold behavior . CNA therefore allows the specification of so called non-monotone arcs . In non-monotone interactions multi-value coefficients are assigned to the participating species . Non-monotone interactions can only be active if the specified species coefficients are matched exactly by the species state . For example , consider the two non-monotone interactions 1 A = 1 B and 2 A = 1 C . In this case 1 A will not activate 1 C und also 2 A will not activate 1 B , so the two distinct levels of A can be employed in different further interactions representing different biological effects . By default all nodes have been considered as single-value nodes which only occur with the values 0 or 1 . Notice that the use of multi-value nodes increases the complexity of the interrelations in the network considerably . However , several biological settings could not be realized with single-value nodes and on that condition the domain of some nodes has been expanded . There are 14 non-monotone interactions in the apoptosis network as listed in Text S1 . Non-monotone interactions are involved in the modeling of the FasL pathway , which was reported to show threshold behavior [29] and the modeling of NF-κB mediated upregulation of anti-apoptotic proteins FLIP , XIAP and c-IAPs [30] , [31] . The respective multi-value nodes are FasL , Fas , DISC , FLIP , C8* , C8-DISC , C3p20 , C3p17 , XIAP and c-IAP that occur with the coefficients {0 , 1 , 2} . Additionally , a multi-value node for UV irradiation was added based on own experimental results ( see Figure 2 ) . Overall the steady states of the model reflect the following behaviors , which would not be possible without using multi-value nodes: ( i ) Apoptosis is not reached in the model by FasL in activity state 1 [FasL ( 1 ) ] but by FasL ( 2 ) reproducing the threshold behavior of Fas signaling [26] . However , FasL ( 1 ) activates several nodes in the network , and their influence and crosstalk with other signaling pathways can be analyzed . ( ii ) The nodes of anti-apoptotic proteins FLIP , XIAP and c-IAPs can be set to zero representing a knockout scenario but they also have graded effects in their “on” state . For example , caspase-3 p20 ( 2 ) can be further processed to the highly active caspase-3 p17 form which ensues in apoptosis if XIAP is low abundant as it is represented by XIAP ( 1 ) . However , if XIAP is upregulated to value “2” it prevents processing and activation of caspase-3 p17 . ( iii ) UV ( 1 ) leads to apoptosis whereas UV ( 2 ) does not lead to apoptosis ( see Figure 2 ) . After we built the mathematical model we performed extensive experimental validation . The logical apoptosis model is based on a vast number of different studies , which were performed in different organisms and were in part highly focusing on important details . Here , we show that the behavior that emerges from these particular interactions in the model is coherent with experimental data on the behavior of the whole network . Table 2 shows the model prediction for different proteins and stimuli which are critical for apoptosis represented by the resulting logical steady state values of the model for the final timescale τ = 10 . The model values of the input nodes are given in parentheses in Table 2 and mock is represented by the logical steady state of the model without activation of any input node . In the experiments , two different cell types were used to account for the distinct signaling mechanisms in cells using the type I ( mouse hepatocytes treated with FasL ) and the type II ( human Jurkat T cells treated with FasL representing the T2RL node ) apoptotic pathways . The measured parameters/nodes of the model are: NF-κB-DNA binding and IκB-α degradation for NF-κB-signaling , activated caspase-3 [C3p17] and the mRNA levels of inhibitory proteins c-IAP , XIAP and FLIP for the caspase cascade , Bid as member of the Bcl-2 family , the activation state of c-Jun N-terminal kinase [JNK] and finally apoptosis as an output signal . Note that the different forms of c-IAPs and FLIP are merged to one node in the model , and measured mRNA levels are c-IAP2 and all 3 isoforms of FLIP . The respective stimuli and nodes are also indicated in Figure 1 . Details on the experimental procedures can be found in the section Materials and methods . All model predictions listed in Table 2 were successfully approved on the first try without changing the model , apart from the effect of UV irradiation on the network . We found an unexpected UV dose effect in primary mouse hepatocytes which was included in the model and will be discussed in the next section . First , the system response to FasL , TNF-α and IL-1β will be presented . All measured entities could be experimentally shown to be active or existing , respectively inactive or non-existing as predicted by the model in response to these stimuli . Selected results of FasL , TNF-α and IL-1β stimulation in mouse hepatocytes are shown in Figure 3 . Stimulation with FasL leads to only weak NF-κB activation and hence no significant c-IAP2 and FLIP upregulation . As there is no signaling effect on the subsequent nodes the model shows NF-κB ( 0 ) in this setting according to the functional definition of its node value which is depending on the node's effect on the network . Caspase-3 p17 is highly active . In contrast , NF-κB is clearly activated after stimulation with TNF-α or IL-1β . Accordingly , c-IAP2 and FLIP are upregulated and , as predicted , caspase-3 p17 is not activated . All validation experiments for Table 2 which are not shown in Figure 3 can be found in Protocol S1 . In addition we tested apoptosis induction in Jurkat T cells after stimulation with TNF-α and IL-1β . As expected and predicted by the model these stimuli do not have cytotoxic effects on the cells and the according experiments are documented in Protocol S1 . It is impossible to test every signaling scenario of the presented apoptosis model due to technical limitations and the mere number of nodes . However , the accuracy of the performed validation experiments indicates fundamental correctness and significance of the model . During experimental validation of the model , we found dose dependent NF-κB activation and apoptosis after UV irradiation in primary mouse hepatocytes . Based on the results shown in Figure 2 , two distinct levels for the UV input node were implemented . The updated model version properly reflects the network behavior in response to UV irradiation and is presented here . UV ( 1 ) represents the stimulation of mouse hepatocytes with 300 J/m2 UV irradiation and UV ( 2 ) with 600 J/m2 . Weak UV irradiation leads to weak NF-κB activation and no c-IAP2 and FLIP mRNA upregulation . As there is no signaling effect on the subsequent nodes the model shows NF-κB ( 0 ) in this setting . As a consequence , mouse hepatocytes show significantly increased caspase-3 p17 activity and consequently cytotoxicity due to apoptosis can be observed as expected after UV irradiation . In contrast , the higher dose of UV irradiation leads to strong NF-κB activation and subsequently c-IAP2 and FLIP mRNA is upregulated . This correlates with previous findings showing a marked NF-κB induction after strong translational inhibition and relative resistance to lower doses [32] . The proteins c-IAP2 and FLIP function as anti-apoptotic inhibitors and prevent caspase-3 p17 activity in this setting . Accordingly , cells show less cytotoxicity after strong UV irradiation and the amount of cell death observed in the MTT viability assay is probably caused to a high extent by necrosis when comparing with caspase-3 p17 activity . In addition , we also treated Jurkat T cells with UV irradiation . We did observe apoptosis neither after 300 J/m2 nor after 600 J/m2 and expect the critical apoptotic UV irradiation dose for Jurkats at higher levels . All validation experiments concerning UV irradiation which are not shown in Figure 2 can be found in Protocol S1 . A feedback loop is a circular path in a signed directed graph , and the overall sign of F is determined by the parity of the number of inhibiting and activating arcs [33] . The sign of a feedback loop has great impact on the dynamics of a system . On the one hand , positive feedback loops allow for multistationarity which is required for epigenetic differentiation in biological systems [34]–[36] . On the other hand , negative feedback loops generate periodicity and are essential for maintaining homeostasis [35] , [36] . The total number of positive and negative feedback loops for each timescale is shown in Figure 4A . As CNA searches for feedback loops of arbitrary length in the network the algorithm finds in fact more feedback loops as expected from a superficial look on the network map . Considering the interactions for τ = 5 there are already 26 positive and 9 negative feedback loops . For τ = 10 these numbers increase up to 82 positive and 13 negative feedback loops . This proportion reflects the typical features of apoptosis networks where positive signal amplification and multistationarity are characteristic . In contrast , antiapoptotic mechanisms are rather realized by inhibitory proteins such as XIAP than by negative feedback loops . Interestingly , as shown in Figure 4B , there is an unexpected feedback already for τ = 4 in the network which was not modeled explicitly . The formation of complex II induces activation of caspase-8 which leads to the release of Smac in response to Bid cleavage finally resulting in mitochondrial pathway activation in type II cells . According to our model , Smac could further increase complex II formation by increasing the amount of available RIP-deubi . The biological relevance of this feedback is speculative . However , the topological possibility of a feedback loop in apoptosis signaling upstream of the caspase cascade is fascinating and potentially important . The relevance of feedback loops [37]–[39] and associated affects such as bistability [27] , [40] and oscillations [41] , [42] are a largely discussed topic . The so far analyzed and well known feedback loops are usually consisting of very few molecules [43] , [44] . The analysis of the apoptosis model shows a high number of feedback mechanisms consisting of many interactions building long loops . As the Boolean model is not dynamic it cannot tell whether these structures are biological relevant or take place on an insignificant timescale . However , their further analysis might be promising . In the following section , we discuss the influence of feedback loops and gene regulatory effects on the signaling behavior of the model for τ = 5 and τ = 10 . The relative participation of network components in all feedback loops on the respective timescale is shown in Text S1 . The general tendency of signaling is still maintained for τ = 5 as the apoptosis supporting input nodes mainly participate in positive signaling pathways and vice versa . However , the combination of negative and positive pathways allows for a more differentiated response to input signals . The components of the caspase module are involved in most of the feedback loops for τ = 5 , and their relative participation reaches up to 89% for C3p17 ( Text S1 ) . This high involvement originates from the high connectivity of these nodes with other pathways and is indicative of the important role of caspases , especially caspase-3 , in apoptosis regulation . For τ = 10 we noticed an increased involvement of NF-κB and components of the mitochondrial module in feedback regulation . In particular , Bax participates at 76% ( Text S1 ) . In summary , only a small group of species is involved in most of the feedback loops , but as such this group plays a prominent role in the regulation and determination of the network response to input signals . This small group consists mainly of caspases , mitochondrial proteins and NF-κB signaling components which are important for the robustness of the entire system and indicate their importance in apoptosis execution and control . The regulatory importance of feedback loops is also reflected by the species dependencies for different timescales . The respective dependency matrices are due to their size shown in Figures S1 , S2 , S3 . Until τ = 4 almost only total activation and inhibition processes occur in the network which represents the linear and parallel behavior of the signaling processes ( Figure S1 ) . A comparison with the species dependencies for τ = 5 shows a substantially changed network topology and reveals all species that are influenced by negative feedback loops in their respective pathways but also pathways to which they are connected ( Figure S2 ) . The dependency matrix for τ = 10 finally completes the overall picture of complex and ambiguous relationships in the network showing almost no total activation and inhibition processes anymore but an increased number of ambivalent effects ( Figure S3 ) . The total number of calculated signaling pathways from each start node to the apoptosis node is shown in Table 3 for each timescale . No continuous signaling pathways to the apoptosis node exist for τ≤3 because the caspase activation module is only active for τ≥4 as described before . For τ = 4 all input nodes with apoptosis supporting effects exclusively participate in positive signaling pathways to the apoptosis output node . In accordance , all input nodes with apoptosis inhibiting effects do not show any or only negative pathways . This topology describes a non-regulated cell which would show a linear signaling behavior without the ability to integrate received information and adapt to situations . Additionally , the constraint signaling behavior would render the cell error-prone for the failure of individual molecular species . For τ = 5 feedback loops extend the network topology . Although only nine interactions are added at this timescale their impact is significant and most input nodes already have ambivalent potential to influence the apoptosis node depending on further circumstances . The number of signaling paths from the input nodes to apoptosis finally dramatically increases for τ = 10 by adding gene regulatory effects by the NF-κB node . Concerning the final decision between cell survival and apoptosis the overall network presents itself as highly crosslinked and regulated in a complex manner . High connectivity increases the number of possible pathways between two nodes and the reliability and flexibility of the network to respond to its environment . CNA considers strongly connected components as maximal subgraphs of the interaction graph in which paths between all pairs of nodes exist . The apoptosis model contains two groups of strongly connected components . One comprises the nodes PKC , PKB , PDK1 , PIP3 , PI3K and IRS-P . These nodes are part of the insulin signaling pathway and connected to a feedback loop by PKB . The second group contains 30 nodes , which belong to complex formation in the upper apoptosis signaling ( complex1 , complex2 , TRAF2 , RIP-deubi , comp1-IKK , NIK , C8-comp2 , FLIP ) , caspase cascade ( C6 , C3p20 , C3p17 , C3-XIAP , XIAP , c-IAP , C8* , C9* , BIR1-2 ) , mitochondrial release ( tBid , Bax , Bcl-xl , apopto , Apaf-1 , smac-XIAP , smac , cyt-c ) and NF-κB signaling ( NF-κB , IκB-α , IκB-ε , A20 , IKK* ) . The high connectivity between these nodes is only partially due to the cascading topology of enzyme activation . Furthermore , the involved proteins such as the inhibitor XIAP , several feedback loops and especially the inclusion of NF-κB signaling in this strongly connected subgraph reflect the highly controlled and robust structure of death signaling . As a transcription factor , NF-κB has central role for the network . The anti-apoptotic impact of NF-κB is ensured via the upregulation of survival factors . However , analysis with CNA reveals an even broader influence of the NF-κB node resulting from its high connectivity . There are 34 inhibitors , 27 activators and 8 ambivalent factors affecting NF-κB . In turn , NF-κB is an ambivalent factor for 30 species , an activator for 8 and an inhibitor for 1 . In addition to these highly connected subgraphs crosstalks between individual signaling modules determine the behavior of the network . Amongst others , the model includes the following crosstalks with insulin signaling ( documented with the according interactions in Text S1 ) : ( i ) TNF-α stimulates IRS phosphorylation and thereby inhibits insulin signaling . ( ii ) In response to insulin PKB is activated and phosphorylates Bad . Phosphorylated Bad is sequestered by 14-3-3 proteins and therefore cannot activate pro-apoptotic Bax . ( iii ) PI3K is involved in insulin signaling and also contributes to NF-κB activation via IKK . ( iv ) Raf can be activated via insulin signaling and inhibited by glucagon signaling and active Raf also triggers IKK-dependent NF-κB activation . Also there were two crosstalks explicitly presumed in the modeling process . Smac mimetics were shown to have an apoptosis promoting effect after stimulation with TNF-α [23] and also lead to autocrine TNF-α secretion [45] , [46] . The network reflects this crosstalk as Smac mimetics don't induce apoptosis but promote complex II building via RIP and lower the threshold for C3*p17 activation via sequestering XIAP . Accordingly , while TNF stimulation of the model does not lead to apoptosis as observed in hepatocytes and Jurkat T cells , the combination of TNF and Smac mimetics does . Another crosstalk is based on the antiapoptotic influence of IL-1β via NF-κB [47] . Although FasL ( 2 ) alone leads to apoptosis it does not in combination with IL-1β ( 1 ) in the model . The explicitly and implicitly modeled crosstalk connections in the network also lead to further effects in the model . The resulting value for the apoptosis node is systematically simulated for all double stimulation scenarios and listed in Table 4 . The diagonal shows the resulting apoptosis value for the according single stimulations . One would assume the outcome for two combined stimuli to follow the rules 0+0 = 0 , 1+1 = 1 and 0+1 = 1 . However , there are some aberrations which are highlighted bold in the Table and discussed in the following text . Smac-mimetics lead to apoptosis in combination with FasL ( 1 ) by the same mechanism as discussed above . There are also two other combinations aside from IL-1β which prevent apoptosis after FasL ( 2 ) stimulation in the model . Namely Insulin and TNF have an antiapoptotic effect based on NF-κB activation via Raf and complex-1 respectively . There are also some interesting crosstalks concerning UV stimulation . The antiapoptotic effects of insulin and IL-1β also prevent apoptosis in combination with UV ( 1 ) . However , in combination with TNF apoptosis is still enforced by UV ( 1 ) as smac is released by UV irradiation and counteracts XIAP upregulation . The input combinations of UV ( 2 ) with TNF and FasL ( 1 ) also lead to apoptosis as the latter activate caspase-8 ( 1 ) . In contrast , the combination of FasL ( 2 ) and UV ( 2 ) does not cause apoptosis in the model as the NF-κB activation by UV ( 2 ) is dominant in this setting . In the future we will especially focus on the investigation and expansion of the model regarding further crosstalk effects between distinct pathways as well as on their experimental validation . Unfortunately , this is not trivial as the Boolean model does not give advice how to combine stimuli experimentally concerning timing and dosage . However , the connectivity of subnetworks and single components via crosstalks is helpful information to include all essential interactions when focusing on a smaller subsystem or specific question . We propose to check the Boolean model for important interaction players when modeling a particular signaling pathway or designing biological experiments to elucidate functional relationships . The Boolean approach we use here for modeling apoptosis obviously has a systematic drawback resulting from the reduction on qualitative network behavior . The reaction rates of biological processes and the quantitative amount of molecules cannot be assigned straight forward to model values . Instead careful conversion has to be done for particular cases and biological knowledge of the modeler is of special importance . In return the presented logical model is easy to use and very flexible . Protocol S2 comprises detailed instructions how to start up the apoptosis model ( Protocol S3 ) without any previous knowledge . One can use the apoptosis model for comparison with own results as well as for further analyses . It can be modified and expanded to other cell types , additional pathways or crosstalks . In particular , any kind of knock-out or knock-in scenario can be simulated with the model by setting certain nodes or interactions to the desired value . Subsequently , resulting variations in signaling behavior and the changed network topology can be analyzed . On the other hand CNA can search for minimal intervention sets . Thereby the algorithm computes all possibilities to reach a user-defined network state under user-defined constraints as fixed states or maximum number of interventions . Finally , uncovering sensitive points in the network and failure modes of the system concerning specific questions will provide suggestions for biological experimental design as well as predictions how the system reacts in response to selected challenges . Taken together , the logical model presented here can easily be applied to a broad spectrum of scientific questions concerning apoptosis signaling pathways and their complex crosstalk to other pathways and serve as a helpful and valuable tool in a variety of research aims . Each species of a network is considered to be a node and two nodes are connected by an edge , also called arc , indicating a direct dependency between them . Nodes and edges form a graph . Directed graphs are a subclass of graphs in which the orientation of the edge determines the direction of the signal flow [48] . At the boundaries of an interaction graph sources and sinks can be found . Sources represent inputs and are not influenced by other nodes . Sinks represent outputs and do not influence further nodes . Adding a sign to the edge specifies whether the influence of a node is activating ( positive ) or inhibiting ( negative ) . In signed directed graphs linear connections between two nodes that are not directly connected to each other describe paths which have an overall sign . The sign of the sequence of arcs is negative if the number of arcs with negative sign is odd and positive if the number of arcs with negative signs is even or zero . Feedback loops in the biological sense are regulatory functions that integrate the state of a downstream system variable with a state prior in the path and return an answer which then leads to further enhancement or abortion of the signal . In a graph theoretical sense a feedback loop would involve only one node influencing itself . In this work the term feedback loop is used in the biological sense involving one or more nodes . A feedback loop ends at the same node where it started and no other node is visited twice . The overall sign of a feedback loop is determined by the parity of the number of inhibiting and activating arcs [33] . The sign of a feedback loop has great impact on the dynamics of a system [34]–[36] . Logical counterparts to numerical operations are conjunction , disjunction and complement . This will be explained for the example of two variables A and B . A numerical multiplication is analogous to a logical conjunction expressed by ( A Λ B ) or ( A AND B ) . A conjunction of two statements is true when both statements are true . A numerical addition of A and B is expressed in Boolean algebra as a disjunction ( A V B ) or ( A OR B ) . A disjunction of two statements is true when one of the statements is true ( inclusive disjunction ) . In the presented logical apoptosis model disjunctions are not notated explicitly but are represented by several interactions which can lead to the same result . A numerical negation ( –A ) is expressed by ( A ) , ( NOT A ) or ( ! A ) in Boolean algebra . The complement of a statement is true when the statement is false . Another important property of biological regulatory networks is the participation of two or more species in one interaction whereas in an interaction graph one node influences one other node . A representation of more than one species influencing another can be facilitated by logical AND connections . A graph containing AND connected species is a hypergraph [49] . A hyperarc connects two subsets of nodes . The resulting graph is termed a logical hypergraph [19] . The MATLAB based tool CellNetAnalyzer [CNA] [19] allows construction and analysis of metabolic ( stoichiometric ) as well as signaling and regulatory networks via a graphical user interface . In this study CNA Version 9 . 2 has been used . The network map can be created with external programs , and we used Microsoft Power Point . A Boolean network is represented in CNA as a logical interaction hypergraph that can also be transformed into an interaction graph . Thereby hyperarcs are splitted and parallel arcs can arise which may lead to undesired effects . For example , ( A+B = X , A+C = X ) is converted to ( A = X , B = X , A = X , C = X ) . After transformation of the presented logical apoptosis model four duplicated arcs and ten parallel edges are removed . The interaction graph representation is required for computation of signaling pathways , feedback loops and species dependencies because it unambiguously indicates which nodes are involved in interactions . For logical steady state analysis and minimal intervention sets the logical interaction hypergraph representation is required to capture all constraints and influences included in the model . The CNA algorithm for computation of feedback loops identifies paths with the same start and end node . Additionally , the direction of the edges is considered so that signal flow occurs only in the specified direction . Another necessary property of the calculated circuits is their non-decomposability into smaller circuits in order to fulfill the notion of elementary modes [50]–[52] . Signaling pathways in CNA are calculated in analogy to feedback loops . To identify the influence a species A exhibits on another species B all signaling paths leading from A to B can be computed . A is not influencing B if such a path does not exist . Otherwise the influence of A on B is characterized as follows: A is a total activator/inhibitor of B if only activating/inhibiting paths are found . A is a non-total activator/inhibitor of B if only activating/inhibiting paths are found but a path contains an intermediate node that is involved in a negative feedback loop . A is an ambivalent factor if activating and inhibiting paths are found . For every Boolean network all possible logical steady states [LSSs] can be calculated [53] . In CNA a LSS is computed based on specified initial values and the signal propagation through the network is calculated . There are no interactions with so called incomplete truth tables in the network so that all nodes can be evaluated for every input setting . LSSs can be used to simulate changes in the network structure and analyze the consequences on the signal propagation . The knock-out of a certain gene is represented by deactivation or removal of a species achieved by setting the value of this species to zero . Constitutive expression of a gene can be represented by setting the value of this species to greater zero ( on-state ) . Primary hepatocytes were isolated from 8–12 week old B6 ( C57Bl/6NNrl ) mice as previously described [54] . The use of mice for hepatocyte isolation has been approved by the animal experimental committees and animals were handled and housed according to specific pathogen free ( SPF ) conditions . Cells were plated on collagen-coated tissue culture dishes in William's medium E ( WME , from Biochrom ) supplemented with 10% FCS , 100 nM dexamethasone , 2 mM L-glutamine and 1%-penicillin/streptomycin solution ( all reagents from Gibco ) . Cultivation was carried out as described [54] , following a three step starvation procedure . To allow hepatocytes to attach , cells were kept in a humidified atmosphere at 37°C and 5% CO2 for 4 h . Subsequently , FCS cell culture medium was removed and replaced by serum-free culture medium ( WME supplemented with 100 nM dexamethasone , 2 mM L-glutamine and 1%-penicillin/streptomycin solution ) . Following 4 h incubation in serum-free culture medium hepatocytes were washed three times with starvation medium ( WME supplemented with 2 mM L-glutamine and 1%-penicillin/streptomycin solution ) and further kept for 16–24 h in the same medium . Jurkat T cells ( suspension ) were maintained in RPMI 1640 medium supplemented with 10% FCS and 1%-penicillin/streptomycin . For preparation of total extracts 2×106 cells were centrifugated ( 2150 g , 4°C , 3 min ) , washed with PBS , centrifugated again and 140 µl of lysis buffer ( 136 mM NaCl , 2 mM EDTA , 20 mM Tris/HCl pH 7 . 4 , 10% glycerol , 4 mM benzamidine , 50 mM β-glycerophosphate , 20 mM Na-diphosphate , 10 mM NaF , 1 mM Na3VO4 ) supplemented with protease inhibitors ( 5 µg/ml aprotinin , 5 µg/ml leupeptin , 0 . 2 mM pefablock ) was added . Cell lysis was performed by shaking for 20 min at 4°C and final centrifugation at 20800 g , 4°C for 10 min . For preparation of nuclear extracts 1×106 cells were washed with PBS and collected in Eppendorf tubes . After centrifugation ( 2150 g , 4°C , 3 min ) , the pellet was resuspended using 400 µl buffer A ( 10 mM Hepes/KOH pH 7 . 6 , 15 mM KCl , 2 mM MgCl2 , 0 . 1 mM EDTA pH 8 . 0 ) and incubated on ice for 10 min . Then , the cell suspension was centrifuged ( 2150 g , 4°C , 3 min ) and buffer A was replaced by 200 µl buffer A containing 0 . 2% NP-40 supplemented with Complete protease inhibitors ( Roche Applied Science ) and incubated for exactly 5 min on ice to lyse the cytoplasma membrane . After centrifugation ( 8062 g , 4°C , 2 min ) , supernatants were stored as cytoplasmic extracts and pellets were resuspended in 50 µl buffer C ( 25 mM Hepes/KOH pH 7 . 6 , 50 mM KCl , 0 . 1 mM EDTA pH 8 . 0 , 10% glycerol , Complete protease inhibitors ) and kept on ice . After 5 min , 4 . 5 µl of a 5 M NaCl solution was added and incubated for 30 min with gentle shaking at 4°C . After centrifugation ( 20800 g , 4°C , 10 min ) the supernatant was isolated as nuclear extract . For measuring the activity of the executioner caspases 3/7 DEVDase assay was performed . Primary mouse hepatocytes and Jurkat T cells ( 1×106 cells respectively ) were incubated with TNF-α ( R&D Systems ) 25 ng/ml , IL-1β ( Jena Bioscience ) 50 ng/ml or FasL ( N2A FasL as described in [25] ) 50 ng/ml for 6 h or exposed to 300 J/m2 or 600 J/m2 UV irradiation ( Stratalinker UV crosslinker from Stratagene ) . Then the cell suspension was centrifugated , washed with PBS and homogenized in 50 µl of homogenization buffer . Caspase-3 activity assay was performed exactly as described in [55] using the caspase-3 substrate DEVD-AMC ( Alexis ) at a concentration of 200 nM . Relative fluorescence units ( RFU ) values were calculated via the ratio of average rate of the fluorescence increase and protein concentration determined by Bradford assay ( Biorad ) . To compare different experiments , RFU sample values were referred to negative control ( untreated cells ) . At least three independent experiments were carried out and means of these experiments including the SD are shown . After exposition to the different stimuli for 6 h or to UV irradiation of the aforementioned doses , primary hepatocytes and Jurkat T cells were treated with 1 ml of 0 . 5 mg/ml MTT ( Sigma ) solution in PBS , and incubated at 37°C for 2 h . After observing a color change to purple the supernatant was removed and the crystals dissolved in DMSO . The samples were transferred into a fresh 96-well plate , and the color reaction measured with an ELISA reader at 595 nm . The sample values were referred to untreated control . Again , means of three independent experiments with SD are shown . Please note that the MTT assay only measures viability and does not differentiate between apoptosis and other forms of cell death . Nuclear protein extracts were prepared as described above . Equal amounts of nuclear proteins ( 4 µg ) were added to a reaction mixture containing 20 µg bovine serum albumin , 2 µg poly ( dI-dC ) ( Roche Molecular Biochemicals ) , 2 µl buffer D+ ( 20 mM HEPES , pH 7 . 9 , 20% glycerol , 100 mM KCl , 0 . 5 mM EDTA , 0 . 25% NP-40 , 2 mM DTT , 0 . 1% PMSF ) , 4 µl buffer F ( 20% Ficoll 400 , 100 mM HEPES , 300 mM KCl , 10 mM DTT , 0 . 1% PMSF ) and 100 , 000 cpm ( Cerenkov ) of a P33-labeled oligonucleotide for NF-κB made up to a final volume of 20 µl with distilled water . For competition experiments ( not shown ) the reaction mixture contained a 100-fold excess of the respective non-radioactive labeled oligonucleotide . NF-κB oligonucleotide ( 5′-AGT TGA GGG GAC TTT CCC AGG C-3′ , Promega ) was labeled using [γ33P]ATP ( 3000 Ci/mmol , Amersham Biosciences ) and a T4 polynucleotide kinase ( New England Biolabs ) . After 25 min of incubation at room temperature the samples were resolved through non-denaturing 6% polyacrylamide gel electrophoresis and then the dried gel was exposed to an Imaging Plate ( BAS-MS 2340 , Fujifilm ) overnight which was finally analyzed using a FLA-3000 ( Fujifilm ) . In the figures the resulting images are shown together with the quantified 33P-stimulated luminescence ( PSL ) units of each specific shift . Dimer composition was determined by supershift analysis ( not shown ) using specific antibodies for p65 and p50 NF-κB subunits ( from Santa Cruz Biotechnologies ) . To analyze protein levels in total cell lysates , samples containing 50–70 µg protein were separated by SDS-PAGE ( 12% or 15% gels ) and transferred to a 0 . 45 µm or 0 . 2 µm pore size PVDF ( Roche Applied Science and BioRad , respectively ) membrane . Antigen detection was done using antibodies against P-JNK at 1∶1000 ( Cell Signaling ) , IκB-α at 1∶1000 ( Cell Signaling ) , β-actin at 1∶10000 ( MP Biomedicals ) , Bid at 1∶700 ( gift from David Huang , WEHI ) , XIAP at 1∶2000 ( StressGen ) , appropriate horseradish peroxidase-labeled secondary antibodies ( Jackson ImmunoResearch Laboratories or Cell Signaling ) , and the ECL plus chemiluminescence detection reagent ( Amersham Biosciences ) . Chemiluminescent images were quantified using the LumiImager and the LumiAnalyst Software ( Roche Applied Science ) . Total RNA was isolated using RNeasy Plus Kit ( Qiagen ) and extraction was performed according to the manufacturer's directions . The quantity and purity of RNA was determined by measuring the optical density at 260 and 280 nm . Subsequently , 1 µg of total RNA was converted to single strand cDNA using Quantiscript Reverse Transcriptase ( Qiagen ) resulting in 100 µl diluted cDNA . The analysis of mRNA expression profiles was performed with multiplex quantitative real time PCR . In a 25 µl PCR reaction , 2 µl of cDNA ( corresponding to 20 ng of total RNA input ) was amplified in an Light Cycler 480 ( Roche ) , using 2-fold QuantiTect Multiplex PCR Master Mix ( Qiagen ) , 50 nM primers and 100 nM probe for the 18S rRNA reference gene ( fwd: 5′-CGGCTACCACATCCAAGG-3′ , rev: 5′-CGGGTCGGGAGTGGGT , probe: 5′-TTGCGCGCCTGCTGCCT ) , and 300 nM primers and 100 nM probe for the gene of interest . The following target gene primers and probes were used ( all from Sigma ) : mouse cIAP2 ( fwd: 5′-ACATTTTCCCCACTGTCCATTT-3′ , rev: 5′-CTATCCAGGGGTCATCTCCA-3′ , probe: 5′-ATGCAGACACACTCTGCTCG-3′ ) , human cIAP2 ( fwd: 5′-CTGGAAACAAAGCATTGAAGTCTG-3′ , rev: 5′-GCCATTAGTAAAGAGGTTCTGAGTC-3′ , probe: 5′-CGTCTGTGAGATCCAGGAAACCATGCTTGC-3′ ) , mouse cFLIP ( fwd: 5′-TGCCAGAGTGTGGAGAACAG-3′; rev: 5′-TTACCCAGTCGCATGACAAA-3′; probe: 5′-GGGGGAGGTTATCTACCAAGT-3′ ) and human cFLIP ( fwd: 5′-AGACCCTTGTGAGCTTCCCTAG-3′ , rev: 5′- GCAGCATCTCCTTCTCATCTGTATC-3′ , probe: 5′-AGTGCTTCTTCAACCTGATGGATGACTTCA-3′ ) . The mRNA level for the gene of interest was determined as 2-ΔΔCT and therefore reflects changes relative to unstimulated cells . Cells were treated with TNF-α 25 ng/ml , IL-1β 50 ng/ml or FasL 50 ng/ml for 8 , 3 or 6 h respectively . All experiments were performed at least three times and means of three independent experiments with SD are shown .	Apoptosis is one of the most investigated topics in the life sciences , especially as this kind of programmed cell death has been linked to several diseases . The strong desire to understand the function and regulation of apoptosis is unfortunately confronted with its complexity and its high degree of cross linking within the cell . Therefore we apply the so-called logical or Boolean mathematical modeling approach to comprehensively describe the numerous interactions in the apoptotic network . Classical Boolean modeling assumes that a certain cellular signal is either present ( on ) or absent ( off ) . We use extensions of classical Boolean models , namely timescale constants and multi-value nodes , which allow the model to emulate typical apoptotic features . The mathematical model describes for the first time the numerous relevant interactions and signals that control apoptosis in a single and coherent framework . The logical model of apoptosis provides valuable information about the topology of the network including feedback loops and crosstalk effects . Proper investigation of the mutual interactions between species points towards hubs in the network with outstanding relevance . These species are of special interest concerning experimental intervention as well as drug target search . The model we present here is easy to use and freely available .	[ "Abstract", "Introduction", "Results/Discussion", "Materials", "and", "Methods" ]	[ "cell", "biology/cellular", "death", "and", "stress", "responses", "computational", "biology/systems", "biology", "cell", "biology/cell", "signaling" ]	2009	ON/OFF and Beyond - A Boolean Model of Apoptosis
Microbes are predominantly found in surface-attached and spatially structured polymicrobial communities . Within these communities , microbial cells excrete a wide range of metabolites , setting the stage for interspecific metabolic interactions . The links , however , between metabolic and ecological interactions ( functional relationships ) , and species spatial organization ( structural relationships ) are still poorly understood . Here , we use an individual-based modelling framework to simulate the growth of a two-species surface-attached community where food ( resource ) is traded for detoxification ( service ) and investigate how metabolic constraints of individual species shape the emergent structural and functional relationships of the community . We show that strong metabolic interdependence drives the emergence of mutualism , robust interspecific mixing , and increased community productivity . Specifically , we observed a striking and highly stable emergent lineage branching pattern , generating a persistent lineage mixing that was absent when the metabolic exchange was removed . These emergent community properties are driven by demographic feedbacks , such that aid from neighbouring cells directly enhances focal cell growth , which in turn feeds back to neighbour fecundity . In contrast , weak metabolic interdependence drives conflict ( exploitation or competition ) , and in turn greater interspecific segregation . Together , these results support the idea that species structural and functional relationships represent the net balance of metabolic interdependencies . It is now widely accepted that most polymicrobial communities living in natural environments form spatially structured and surface-attached consortia ( biofilms ) [1] . There has recently been a great interest in investigating how spatial structure may forge and stabilize the complex web of interactions occurring within these multispecies communities , including mutualistic [2] and competitive [3] relationships . Empirical work in multispecies biofilms has acknowledged that species composition affects community structure and species distribution within the biofilm [4] as a result , for example , of mixing species that have distinct monoculture structures [5] , or via metabolic interactions , such as cross-feeding [6] , [7] , [8] , [9] or detoxification of exogenous waste products [10] . The type of carbon source also plays a major role in generating the diversity of spatial arrangements observed in polymicrobial communities , as varying the source of carbon likely alters the metabolic interactions between members of the community . For example , in a two-species biofilm consisting of Burkholderia and Pseudomonas , Nielsen et al . [8] observed that when the two species were competing for a common resource ( non-cross-feeding medium ) , the biofilm consisted of separate microcolonies ( high species segregation ) . In contrast , when the two species were involved in a one-way obligate cross-feeding interaction ( cross-feeding medium ) , the microcolonies consisted of both species ( greater mixing ) . Evolutionary theory has suggested that spatial mixing favours the evolution of mutualism because it keeps mutualistic partners in close proximity , thereby allowing for stronger reciprocity [11] , [12] , [13] , which may in turn facilitate the exchange of metabolites between partners . However , it has also been proposed that , under some conditions , spatial mixing may impair mutualism because of spatial limits on exchange [14] , or because it hinders cooperators' clustering in within-species cooperation [15] . Empirical work on the evolution of microbial cross-feeding mutualisms has also found opposite responses to environment structure . For example , Harcombe [16] provided empirical support for the benefits of spatial structure in the evolution of mutualistic cross-feeding between Salmonella and auxotrophic Escherichia coli . However , another study on the nascent cross-feeding mutualism between Desulfovibrio vulgaris and Methanococcus maripaludis showed that mutualism was initially favoured in a well-mixed rather than static environment [17] . Although the authors suggested that this different response to environmental structure is due to the lack of a direct fitness cost of cooperation in the latter cross-feeding model system [18] , other mechanisms may be at play as well . The spatial separation ( distance ) between species has also been identified as a key factor for the stable coexistence of a synthetic mutualistic bacterial community [19] . While evolutionary ecology has traditionally assumed that structure is a fixed environmental property ( i . e . either structured or well-mixed ) , there has been a recent interest in regarding structure as an emergent property of the aggregate behaviour of individuals [20] ) . Individual-based simulations of microbial growth have started to shed some light on this topic . For example , Nadell et al . [21] explored how physical and biological parameters of bacterial growth in biofilm affect lineage segregation , which in turn determines the fate of within-species cooperation . Using the same framework , it has also been proposed that within-species cooperation can be favoured due to social insulation of cooperators from non-cooperators by a second species [22] . Recently , using a mix of experiments and simulations , Momeni et al . [23] showed that strong inter-population cooperation led to inter-population mixing in microbial communities , and specifically in a pattern of successive layering . Despite this , however , far too little attention has been given to how specific metabolic interactions generate the emergent spatial and functional properties of microbial communities . Our goal here is to address this question by investigating how metabolic constraints of individual species shape the emergent functional relationships and spatial structuring of a two-species community . For this , we focus on a specific type of interspecific metabolic interaction - trading food for detoxification ( for empirical examples see [24] , [17]; for a theoretical approach see [25] ) , and we explore how a partner's need for help ( either detoxification to the producer or food to the cross-feeder ) affects the ecology , spatial structure , and productivity of the two-species community . Using an individual-based modeling ( IBM ) framework that models microbial population growth on a solid surface [26] , our results show that stronger metabolic interdependence generates more mutualism , more interspecific mixing , less sensitivity to initial conditions and enhanced community productivity . The emergence of this metabolism-dependent community structure and functioning is driven by demographic feedbacks , such that providing aid to a mutualistic partner generates a positive feedback on the individual's growth whereas providing aid to a competitor or exploiter generates a negative feedback on the individual's growth . In consequence , demographic feedbacks strengthen mutualistic relationships via increased lineage mixing , and weaken competitive relationships via increased segregation . The ecological outcome of the food for detoxification interaction depends on the balance between costs and benefits of interspecific association . The potential costs are interspecific competition for common nutrients and space , while the potential benefits are food for the cross-feeder and detoxification for the producer . To determine the type of ecological interaction forged between producer and cross-feeder , we measured the net costs and benefits from association [32] , [33] by comparing species growth rates when grown alone with their growth rates when grown in coculture ( see Methods ) . If both species have an increase in growth rate relative to their growth rate in monoculture , the association is mutualistic . If both species have a decrease in growth rate when grown in coculture relative to their growth rate in monoculture , the association is competitive . If one species benefits at the expense of the other , then there is exploitation . Analytical work under the limiting assumption of a well-mixed ( planktonic ) community found that diverse ecological relationships can emerge from a one-way cross-feeding interaction where nutrients are traded for detoxification [25] . Does the same result hold when the environment is spatially structured ? To address this question , we first investigated how the degree of metabolic interdependency ( varying along two species axes ) shapes the ecological relationships between two species . Specifically , we vary by-product toxicity from non-toxic to highly toxic ( variations in metabolite toxicity can occur , for instance , via changes in pH or the type of metabolite produced [34] ) and the degree of cross-feeder obligacy from non-cross-feeder to obligate cross-feeder ( see fig . S1A for a schematic representation of species interactions , and Methods ) . Metabolic interdependency implies that a species' chemical environment is improved in at least one specific dimension by the presence of another species ( for instance , detoxification or provision of a growth substrate ) . However this specific chemical aid does not imply that the recipient gains a net growth advantage from association , as the two species may also compete for other limiting resources ( space and/or nutrients ) . We found that metabolic interdependency gives rise to diverse ecological interactions , ranging from mutualism to competition ( figs . 1A , S2 ) , thus corroborating our previous finding for well-mixed populations [25] . Specifically , mutualism only emerges when the specific help received outweighs the competitive costs endured for both partners . In fig . 1A we have assumed that the two species compete for space and limited nutrients ( unless cross-feeding is entirely obligate ) . We next ask what is the relative contribution of competition for space and nutrients to our results ? Importantly , these two limiting resources are linked as winning the competition for space means getting access to nutrients , and similarly , winning the competition for nutrients means getting access to space . To disentangle their effects we relax nutrient competition ( see schematic fig . S1B ) . As expected , we found that removing competition for nutrients leads to less negative associations , as seen by a shift from competition to exploitation , or from exploitation to mutualism ( fig . 1A , B ) . As toxicity increases , the ability of the producer to compete for a shared nutrient resource is diminished . Therefore , removal of nutrient competition has a disproportionately positive effect on mutualism as toxicity increases . Our definition of mutualism ( [25] , [32] ) implies that the total productivity of the two species community will be greater than the summed productivities of the two species apart . However , our results also show that enhanced community productivity does not itself imply mutualism , as exploitative relationships can also lead to a community gain ( figs . S3 , S4 ) . This is consistent with empirical studies that have documented that resource ( niche ) partitioning via cross-feeding interactions enhances community productivity [35] , [36] , [37] , [38] , with the caveat that enhanced community productivity does not alone dictate a mutualistic relationship . Theoretical modelling has suggested that population segregation ( high relatedness ) can favour within-species cooperation because segregation keeps the benefits of cooperation close to cooperators [21] , [22] , although these benefits are potentially mitigated by enhanced competition among kin [39] , [40] , [41] . Furthermore , it has been suggested that population mixing favours between-species cooperation because it facilitates the exchange of the benefits of cooperation , therefore creating a tension between within-species cooperation and between-species cooperation [22] . In our food for detoxification interaction , the effect of within-species cooperation on population segregation is relaxed , therefore allowing for between-species mutualism to occur under a broader range of conditions . In a recent simulation and experimental study , it has been shown that strong inter-population cooperation leads to inter-population mixing in microbial communities , and specifically in a pattern of successive layering [23] . Based on these observations , we next hypothesized that varying metabolic interdependence would dictate the degree of species intermixing within the two-species community , and in a way that reflects the net costs and benefits of interspecific association . In particular , we would expect that increasing metabolic interdependence would result in higher species intermixing within the biofilm to facilitate trade . We generally found that , as by-product toxicity increases , intermixing increases ( figs . 2 , S5 ) . Similarly , increasing cross-feeder obligacy leads to higher intermixing ( figs . 2 , S5 ) , except in the non-cross feeding medium ( and intermediate to high by-product toxicity ) . The latter scenario likely occurs because the fast growing cross-feeder cells insulate the poorly growing producer cells in separate enclaves , thus leading to greater mixing . The segregation index ( Methods ) provides a global statistic of population structure , but does not reveal the developmental patterning of the two intermixing species or their resulting shared architecture . Videos S1 and S2 illustrate the resulting development and architecture of the two-species community , and highlight that strong mixing is achieved via a striking and emergent branching pattern producing increased inter-digitation and contact surface between interdependent cell lineages . Branching-like patterns within single clonal lineages have been observed previously under conditions of low nutrient availability , due to stochastic variations in a thin active growth layer [21] . The resulting separated ‘towers’ ( observable in fig . 2D ) are mutually repulsive , as growth towards conspecifics increases competition for limiting resources . In contrast , as mutual interdependence increases , demographic movement towards heterospecifics becomes increasingly rewarding , resulting in branching of lineages towards heterospecifics and away from conspecifics , generating a robust and stabilising mixing pattern . It has been recently documented that population intermixing of a yeast obligate cooperative community is robust to a broad range of initial conditions , including initial ratio and densities [23] . In this study , however , the authors assumed that cells were randomly seeded . Given this , we hypothesized that the degree of intermixing at inoculation may influence the ecological and structural relationship of the two species trading food for detoxification , by modulating the establishment of key metabolic and demographic feedbacks . Indeed , increasing segregation at inoculation might have two opposite effects: on the one hand , we would expect the costs of interspecific competition to be delayed , but on the other hand , the benefits of trade would be reduced . To examine this , we repeated the simulations of monoculture , facultative cross-feeding coculture , and obligate cross-feeding coculture , but now the cells were inoculated in two microcolonies of size 30 µm and separated by a distance of 70 µm from each other ( coculture ) or in a single microcolony of size 30 µm ( monoculture ) . The degree of initial intermixing was changed by varying the proportions of producer and cross-feeder in each microcolony but keeping the total number of inoculated cells constant and 1∶1 . This means that , for example , when both microcolonies were inoculated with equal number of cells of producer and cross-feeder type , then they were completely intermixed ( i . e . segregation index , s , equal to 0 , see Methods ) . When one microcolony was inoculated with cells of producer type only and the other microcolony with cells of cross-feeder type only , then they were fully segregated ( i . e . segregation index , s , equal to 1 ) . Note that monoculture simulations were repeated using the same seeding rule to prevent any bias from inoculation crowding effects when we are comparing monoculture and coculture growth . In the absence of metabolic interaction , the two species ( here , differing only in colour ) tend to segregate , independently of initial intermixing ( fig . 3B ) . This agrees with modelling [21] and empirical work on the social amoeba Dictyostelium discoideum [42] showing that spatially structured growth is a passive mechanism that increases relatedness ( or lineage segregation ) . But what happens when the lineages experience metabolic interactions ? Our results suggest that the emergent patterns of lineage mixing ( fig . 2A ) are highly robust against variation in initial inoculum mixing , except when the two species are completely segregated in two separate microcolonies at inoculation ( fig . 3A , fig . S6A–C ) . Indeed , if the two species are strongly interdependent , they are conditioned to mix to grow . Thus , when initially segregated , such strong initial segregation may delay ( fig . S7A , B ) or even prevent ( e . g . when interdependency is too high; fig . S7C ) the structural relationship to be forged . This result also supports the idea that spatial distance between species plays a critical role for the stable coexistence of obligate mutualistic bacterial communities [19] . We also found that the strongly interdependent community shows a strong signature of negative frequency dependent selection ( the rare lineage is favoured ) , ensuring a stable coexistence frequency of around 34% producers , regardless of their initial frequency ( fig . 3C ) . In contrast , the control community is sensitive to the proportion of producers at inoculation ( fig . 3D ) , due to the absence of stabilising mechanisms of interaction . To further understand the demographic drivers of intermixing , we break the demographic feedbacks by modifying both initial segregation conditions and the mass-transfer regime ( by-product diffusion ) . First , we simulated the growth of an initially segregated two-species community and separately tracked the growth rates of cells situated nearer towards or further apart from the heterospecific lineage . We found that when metabolic interdependence is high , the cells that are closer to interspecific cells grow better than the cells that are further away from interspecific cells ( fig . S8A ) . As shown in Video S3 , obligate cross-feeder cells closer to producer cells grow towards the producer cells , i . e . towards the by-product . In turn , this reduces the concentration of toxic by-product in the microenvironment of producer cells that are closer to the obligate cross-feeder , thus favouring the growth of those neighbouring producer cells . This result highlights the importance of demographic feedbacks that follow from growth benefits of trading resources for detoxification in shaping community function and spatial structure . At a more macroscale , the results of demographic feedbacks among mutualists are clear in fig . 3C , where we see the signature of negative frequency-dependence driving the two partners to a stable coexistence point irrespective of initial frequency , and in fig . 4AB where we see an accelerating growth of mutualists with increasing heterospecific proximity and mixing . Furthermore , as observed earlier for intermixed inocula ( fig . 2B and Video S1 ) , the community branching structure emerges as the community grows , but the branching pattern is now- with separated inocula- more pronounced , probably because of reduced space constraints ( fig . 4A , Video S3 ) . The emergence of similar architectures and intermixing statistics between Videos S3 and S1 ( i . e . separate and intermixed inocula , respectively ) highlights the robust community developmental programme that results from strong metabolic interdependencies , which in turn deliver a high-functioning community . Given facultative cross-feeding , the cross-feeder can grow using the shared limiting nutrient ( e . g . glucose ) as well as the producer by-product . When the by-product is weakly toxic both producer and cross-feeder cells that grow closer to interspecific cells grow better than the cells that are further away ( fig . S8B , Video S4 ) , but the disadvantage of cells growing further away is now smaller and mixing is reduced . As by-product toxicity increases , producer cells growing closer to the cross-feeder can even grow more slowly than those further away , despite receiving greater detoxification benefits . The producer cells adjacent to cross-feeding cells suffer due to the increased competition for the shared limiting nutrient ( fig . S8B ) . At a more macroscale , the results of demographic feedbacks among weakly interdependent partners ( fig . 4CD ) can be seen by a negative correlation between the densities of producer and cross-feeder across replicates following lineage contact ( fig . S9B ) , as a stochastic advantage to one lineage spells a cost to the competitor lineage ( generating the increased variance around the mean in fig . 4D ) . In contrast , strong interdependence generates a positive correlation between producer and cross-feeder across replicates following contact ( fig . S9A ) , as a stochastic advantage to one lineage drives further advantages to its partner lineage . The ability to effectively carry out a food-for-detoxification exchange depends ultimately on an effective process of molecular transport from producer to consumer cell . In our final manipulation , we vary the rate of diffusion to explore the importance of mass-transfer processes on the establishment and maintenance of metabolic and demographic feedbacks . We found that when the two species are initially spatially segregated , increasing diffusion improves the performance of both species , due to an enhanced metabolic flux kick-starting the exchange ( figs . 5A , S10A ) . In contrast , when the two species are initially mixed , performances ( lineage growth rates ) are scarcely touched by changes in diffusion over two orders of magnitude , as the initial proximity of the partner lineages assures effective inter-cellular transport even at very low rates of diffusion ( figs . 5B , S10B ) . The effect of diffusion is however very pronounced on the resulting strength of mutualism . When diffusion is very low , mutualism is far stronger simply because the producers are in much more trouble when alone ( fig . 5B ) . In contrast , as diffusion increases , solitary producer colonies suffer less from their byproduct toxicity due to a rapid abiotic removal process , making the net benefit of partnership much weaker ( fig . 5B ) . Together , these results illustrate the important and interacting roles played by initial segregation and diffusion in establishing an effective metabolic exchange , and consequently the emergent function and spatial structure of communities . While it is well acknowledged that spatial structure plays a critical role in shaping the ecological outcome of species interactions , our understanding of how community structure and function emerge from the mechanistic bases of species interactions is still poorly understood . Here , we addressed this question by investigating how metabolic constraints of individual species shape the emergent functional and structural relationships of a two-species microbial community that trades food for detoxification . Specifically , our main findings reveal that mutual interdependence generates a robust and highly stabilising mixing pattern . This happens because demographic movement towards heterospecifics becomes increasingly rewarding , resulting in branching of lineages towards heterospecifics and away from conspecifics . These demographic feedbacks strengthen mutualistic relationships via increased lineage mixing , and weaken competitive relationships via increased segregation . Furthermore , we show that initial mixing and diffusion play a critical role in establishing effective metabolic exchange , and therefore in defining the emergent functional and structural relationships among species . Strong metabolic interdependence is commonly found in syntrophic ( cross-feeding ) relationships [43] , and empirical evidence for the importance of spatial distribution in the functioning of metabolically interdependent syntrophic consortia is growing in the literature [44] , [45] . But , what if mutualism is based on bidirectional cross-feeding rather than a food for detoxification mutualism ? Recent work has suggested that strong inter-population cooperation , in which each strain depends on the provision of an essential metabolite by the other strain , leads to population mixing in a pattern of successive layering ( for yeast see [23] , for E . coli see [46] ) . One potential explanation for this discrepancy in spatial pattern between their findings and ours is the specific nature of the mechanistic interaction ( e . g . bidirectional cross-feeding vs food for detoxification cross-feeding ) . To assess this possibility we ran additional simulations assuming bidirectional cross-feeding instead of food for detoxification cross-feeding and we observed a hybrid result . We still observe a characteristic emergent branching pattern , although now the producer forms a mantle layer at the top of the biofilm ( fig . S11 ) . Understanding the drivers of these distinct spatial patterns is an interesting area of research to be pursued . A striking result in our simulations is the emergent two-species branching structure of communities that exhibit strong interdependence ( Videos S1 , S3 ) . Branching patterns are commonly found in nature ( e . g . neurons , blood vessels , trees ) . In bacteria , branching has been observed in swarming colonies , including Bacillus subtilis [47] and Pseudomonas aeruginosa [48] , [49] , [50] , but what may explain such community architecture here ? Branching seems to emerge because of lineage growth with demographic movement away from conspecifics and towards interspecifics ( helpers ) , thereby maximizing the surface contact area with interspecifics . Video S3 suggests that the first mover is the obligate cross-feeder which branches into regions of high by-product concentrations ( high toxicity for producer ) . This relieves inhibition on the producer , which can now grow until toxicity returns . Here , we have assumed that the facultative cross-feeder is able to use both the common resource and the by-product independently of their concentrations in the environment . This means that the trade-off between the cross-feeder's ability to use both nutrients is fixed , and not under regulatory control . Regulatory control , however , plays a critical role in bacterial metabolism and social dynamics [48] , [51] . One could relax this assumption and allow for regulatory control in our cross-feeding model . For example , common resource vs by-product consumption could be a plastic trait that depends on the local concentration of the by-product . Specifically , one could assume a scenario where the metabolism of the by-product inhibits the uptake of the common resource [52] . While outside the scope of this study , we believe that investigating how metabolic plasticity in resource use affects the structure-function dynamics of interspecific interactions would add to our understanding of mapping metabolism to ecology and structure in polymicrobial communities . Our work looks at interspecific mutualisms that arise due to by-product mutualisms , as the benefit provided to the other species occurs as a result of a trait carrying no immediate , direct cost to the actor [33] , [53] . Additionally , our model assumes that cell movement is purely due to demographic processes of cell growth . This means that there is no behavioural mechanism that preferentially directs help towards a mutualistic partner ( such as in partner choice , [13] , [53] ) or makes an individual preferentially move towards a mutualistic partner . While it is unclear whether partner choice exists in bacteria , motility [54] and chemotaxis are behavioural mechanisms that allow bacterial cells to move towards favourable environments ( e . g . food gradient ) and therefore influence species functional and structural relationships . It would be interesting to see how these mechanisms would affect the functioning and structuring of our food for detoxification association . One would nevertheless expect a similar general structural pattern even when behavioural processes are at play , i . e . mix when the benefits of association outweigh the costs , but segregate when the costs of association outweigh the benefits . Another explicit assumption of our model is that cells are growing on an inert surface and that the nutrient diffuses from the bulk ( above ) into the biofilm . This implies that only the cells that are at the surface of the biofilm are able to access the nutrient and grow . This is a common assumption when using this individual-based framework , but this may not always be the case as in , for example , the gut environment ( see [27] for an individual-based model of host-microbiota interactions where the authors assume bidirectional nutrient gradient ) . Under these conditions , and assuming sloughing of microbial cells , different emergent structures and branching patterns may arise . Our study illustrates how community structural and functional relationships emerge from metabolic signatures of interspecific interaction . Although we focused on a specific mechanism of trade - food for detoxification of a metabolic by-product - we believe that our approach of mapping metabolism into function and spatial organization can be extended to other types of microbial associations . It would be interesting , for instance , to investigate what are the emergent functional and structural relationships of a two-species community trading food for detoxification of an exogenous toxic metabolite ( e . g . antibiotic ) . Also , trading food for detoxification implies that when mutualism emerges , it is intrinsically resistant to interspecific cheating strategies . This conclusion lends greater relevance to our ecological results , however it still leaves open a number of questions on the potential for coevolutionary dynamics within this mutualistic space , for instance towards greater rates of waste production [25] . Finally , we suggest that further research into the interplay between the molecular mechanisms of species interactions and the ensuing population and community dynamics is needed to foster our understanding of how natural microbial communities emerge and are maintained in the first place , as well as predict how they may be affected by environmental perturbations on both ecological and evolutionary timescales [55] . Our model assumes two species , a producer ( A ) of a metabolic by-product ( E ) , and a cross-feeder ( B ) ( see fig . S1 for a schematic representation ) growing on an inert surface . The producer and cross-feeder are ecological competitors for a common limiting nutrient ( R , e . g . glucose ) that diffuses from the bulk ( above ) into the biofilm . The bulk consists of a liquid and well-mixed compartment where the concentration of nutrient ( Rbulk ) is held constant . Thus , the growths of species A , and species B , are a function of the rates of uptake of R in the local microenvironment of A and B , respectively . In addition , the cross-feeder's growth is enhanced by its ability to use the producer waste product E , while the producer's growth is decreased by E ( i . e . toxic waste product ) . Thus the concentrations of R and E vary in space and time due to production/consumption reactions and diffusion . The metabolic reactions and stoichiometric matrix used in the model are described in detail in Table S1 . Briefly , the reaction of transformation of R into E and biomass A ( XA , cell growth of A ) follows a Monod-form kinetic , and E inhibits this reaction via simple inhibition . The reaction of transformation of R and E into biomass B ( XB , cell growth of B ) follows a Monod-form kinetic on R and E , respectively . Also , we assume that the producer has more affinity and is more efficient at using the main nutrient ( R ) than the cross-feeder , such that KR , A<KR , B and YR , A>YR , B , respectively . This may represent , for example , a cost of resource generalism to the cross-feeder [56] . We assume that the obligate cross-feeder ( Bobl ) is specialist on the producer's waste product and incapable of using the limiting nutrient . This means that the two species do not use overlapping nutrients and that the cross-feeder depends on its partner's waste product for growth . Specialization on a partner's waste product of metabolism can occur via mutations [57] or due to an exclusion mechanism in which the metabolism of the waste product inhibits the uptake of the limiting nutrient [52] . In addition , we assumed three facultative cross-feeders . Consistent with previous empirical work we assume that the facultative cross-feeders are able to use both the common limiting nutrient and the metabolic by-product ( see e . g . [36] , [37] , [57] , [58] ) , but differ in their degree of obligacy , varying from strongly dependent ( BfacS ) to intermediately dependent ( BfacI ) to weakly dependent ( BfacW ) on the producer's waste product for growth ( see Table S2 ) . Finally , in the producer- non-cross-feeder ( Bncf ) association there is complete overlap of resource use . Specific parameter values used for the simulations are described in Table S2 , and other simulation parameter values used for the simulations are described in previous work [22] . Unless otherwise stated , we assume cyclic boundary conditions . Inoculation densities are 60 cells in monoculture , and 60 cells of each species ( 1∶1 ) in coculture . This means that the initial density of each individual species is held constant across culture type ( i . e . monoculture and coculture ) , and thereby the total inoculation density of monoculture is half the total inoculation density of coculture ( additive experimental design ) . This approach gives us a measure of how an individual species is affected by diversity only , and not by initial individual species densities . Growth rate is measured as ( Nf−Ni ) / ( tf−ti ) where Ni represents the number of cells inoculated at time 0 ( ti ) , and Nf represents the number of cells at the end of the simulation ( tf ) . Unless otherwise stated , data represent growth after 96 hours , and are the mean of 3 replicates . The segregation index ( s ) is an indicator of species segregation ( or mixing ) within their local neighbourood measured relative to global species frequencies , and is measured as:andwhere segA ( segB ) represents the proportion of species A ( species B ) in the local microenvironment ( i . e . neighbourhood ) , and ( ) is the proportion of species A ( species B ) in the whole population . Note that this way of measuring species segregation in an interspecific population is similar to the relatedness coefficient used in social evolution to measure relatedness within-species [42] , [59] . This intermixing index can also be seen as an indicator of species co-assortment , e . g . , whether species A is more assorted ( or segregated ) with species B than if the two species were distributed randomly ( i . e . when s = 0 ) . The calculation of the proportions of species A and species B in the local environment is adapted from the methodology used in Mitri et al . [22] to measure population segregation in biofilm . In brief , for each individual cell ( ci ) of a given species - i . e . either species A or species B- in a population of N = NA+NB cells we identify all the neighbour cells ( cj ) falling within a neighbourhood distance of a radius of 10 µm . The segregation of each individual cell ci is defined as:where g ( ci , cj ) = 0 if ci and cj belong to different species , or , g ( ci , cj ) = 1 if ci and cj belong to the same species , and Nd is the number of cells falling within the distance of 10 µm . The segregation index segA ( segB ) of species A ( species B ) is then defined as:	Understanding the structure and functioning of polymicrobial communities is a major challenge in biology , as witnessed by the dramatic yet mysterious roles played by the human microbiome in human health . Microbial multispecies communities often show complex spatial structures and patterns of metabolic exchange , yet our understanding of how species spatial and ecological relationships emerge from the metabolic rules of species interactions is still limited . What mechanisms underlie multispecies community self-organization ? In this study , we simulate the growth of a minimal—two species— community and show how the emergent properties of community spatial structure and function depend on the nature of metabolic interactions between the two species . We found that strong mutual need for help ( strong metabolic interdependence ) favours the emergence of mutualism , increased productivity and lineage mixing via striking and highly stable branching patterns . In contrast , when the mutual need for help is low , conflict dominates and the two species tend to segregate . Finally , we show how the emergent species mixing follows from a positive feedback of providing aid to neighbouring helpers .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[]	2013	Metabolic and Demographic Feedbacks Shape the Emergent Spatial Structure and Function of Microbial Communities
A rare subset of HIV-infected individuals , designated viremic non-progressors ( VNP ) , remain asymptomatic and maintain normal levels of CD4+ T-cells despite persistently high viremia . To identify mechanisms potentially responsible for the VNP phenotype , we compared VNPs ( average >9 years of HIV infection ) to HIV-infected individuals who have similar CD4+ T-cell counts and viral load , but who are likely to progress if left untreated ( “putative progressors” , PP ) , thus avoiding the confounding effect of differences related to substantial CD4+ T cell depletion . We found that VNPs , compared to PPs , had preserved levels of CD4+ stem cell memory cells ( TSCM ( p<0 . 0001 ) , which was associated with decreased HIV infection of these cells in VNPs ( r = −0 . 649 , p = 0 . 019 ) . In addition , VNPs had decreased HIV infection in CD4+ central memory ( TCM ) cells ( p = 0 . 035 ) , and the total number of TCM cells was associated with increased proliferation of memory CD4+ T cells ( r = 0 . 733 , p = 0 . 01 ) . Our results suggest that , in HIV-infected VNPs , decreased infection of CD4+ TCM and TSCM , cells are involved in preservation of CD4+ T cell homeostasis and lack of disease progression despite high viremia . In the majority of cases , HIV infection is characterized by high levels of viral replication , progressive loss of CD4+ T cells , and if left untreated , eventual progression to AIDS [1] . Predominant correlates of disease progression include HIV replication , loss of CD4+ T cells , high levels of pro-inflammatory cytokines , cell cycle dysregulation , signs of lymphocytic exhaustion , and up-regulation of type I interferon-stimulated genes ( ISGs ) [2] , [3] . However , there is marked variability in the rate of HIV disease progression , largely due to host immune and genetic factors , as well as virus replication [4] , [5] . Even an apparent relatively stable asymptomatic course of infection is observed in a small subset of patients , termed long-term non-progressors ( LTNP ) [1] , [6]–[9] . The vast majority of LTNP maintain relatively high CD4+ T cell counts and low levels of immune activation by controlling viral replication to very low levels , often through vigorous CD8+ T cell mediated immune responses that occur in individuals with certain “protective” HLA class I alleles including HLA-B57 and HLA-B27 [7] , [8] , [10] . In particular , a striking feature in LTNPs which is associated with protection from disease progression is maintenance of long-lived CD4+ central memory T cells ( TCM ) which are essential for long-term immunological memory [11]–[13] . In addition , Descours et al . recently described that decreased HIV infection of TCM cells in HLA-B57/B27 positive LTNPs was associated with HIV-specific CD8+ T cell responses , and is a mechanism of TCM preservation and lack of disease progression [9] . More recently , a rare population of HIV infected individuals has been described who do not progress to AIDS and maintain high CD4+ T cell counts despite high levels of virus replication for many years ( Viremic Non-Progressors , VNPs ) [14]–[17] . Choudhary et al . first reported this rare phenotype in three HIV-infected individuals who maintained high levels of HIV replication ( 104–105 copies HIV RNA/mL plasma ) , but had low levels of immune activation and preservation of CD4+ T cells [15] . Rotger et al . subsequently studied a larger cohort of six HIV-infected individuals with persistently high-level viremia who did not show any signs of progressive HIV disease [14] . The VNPs in this study were compared to HIV-infected rapid progressors , and several features of the profile of gene expression ( as measured by microarray ) were associated with the VNP phenotype , including lower expression of ISGs [14] . In addition , recent studies have demonstrated that VNPs have high levels of mucosal immune activation , corresponding to low levels of mucosal T regulatory cells [18] . Furthermore , VNPs have CD8+ T cell responses similar to those of HIV-infected individuals with chronic disease progression [19] . Interestingly , VNPs share a common gene regulation profile with SIV-infected sooty mangabeys ( SMs ) , an African natural SIV host species that typically experience a non-pathogenic infection despite high virus replication [20] and are thus reminiscent of the rare VNP phenotype . However , it was unclear from these studies whether the differences between VNPs and rapid progressors were simply a consequence of more advanced disease progression in the rapid progressors . For example , HIV-associated immunodeficiency may allow for more microbial translocation and asymptomatic co-infections ( i . e . cytomegalovirus , Epstein-Barr virus , etc . ) , therefore causing greater innate immune activation rather than being a consequence of it [21] . To address these issues , and to identify novel mechanisms that might explain delayed or even lack of disease progression in VNPs , we conducted a detailed immunologic and virologic characterization of a cohort of HIV-infected individuals that were stringently defined as VNPs as previously defined by Rotger et . al [14] . Of note , both the Choudhary and the Rotger studies [14] , [15] used as controls HIV-infected individuals with canonical or rapid disease progression . In contrast , in the current study we have used as controls a group of HIV-infected subjects with relatively preserved CD4+ T cell counts . Since VNPs and other LTNPs ( including the so-called Elite Controllers , EC ) only represent , in aggregate , a very small subset ( <1% ) of HIV-infected subjects [1] , controls with early HIV infection are statistically predicted to progress to AIDS with time if left untreated ( hence we termed them “putative progressors”; PP ) . The rationale for using this type of control group was to avoid the potential confounding effect of immunological impairments that occur in HIV-infected individuals with signs of overt disease progression and complete CD4+ T cell depletion . Here we found that when compared to PP HIV-infected individuals , VNPs showed decreased infection of central memory CD4+ T cells ( TCM ) and stem cell memory T cells ( TSCM ) by HIV , as well as increased proliferation of memory CD4+ T cells , which was associated with increased CD4+ TCM counts . Importantly , TSCM cells have been proposed to represent a preferred viral niche as they characteristically maintain low sensitivity to HIV cytopathic effects and show enhanced long-term survival and proliferation potential as compared to other memory T cell subsets [22] , [23] , and here we found that the frequency of HIV-infected TSCM cells negatively correlated with the number of TSCM cells . Furthermore , in SIV-infected sooty mangabeys , TSCM are less infected in vivo as compared to pathogenic SIV infection in rhesus macaques [24] . In addition , preservation and decreased infection of TCM in sooty mangabeys has been implicated as a mechanism for protection from disease progression in this model [25] . Thus , the attenuated virus infection in TCM and TSCM cells from VNP individuals may be indicative of at least one mechanism of protection . Taken together , we propose that these novel immunological features of VNPs may underlie maintenance of CD4+ T cells and lack of disease progression in these individuals despite many years of viremic HIV infection . We identified treatment-naïve HIV-infected VNPs for this study using strict criteria: maintenance of CD4+ T cell counts >500/mm3 blood ( average 735 . 0±135 . 7 ) , sustained plasma HIV RNA levels >104 ( average plasma HIV RNA levels 4 . 68±0 . 4 log10 copies/mL plasma ) , and >9 years ( average 19 . 8±4 . 8 ) since initial HIV diagnosis ( Table 1 ) . For comparison , controls were treatment-naïve individuals with confirmed HIV infection ( 2 months to 7 years since the estimated date of infection [26] ) , and with comparable levels of viremia ( average plasma HIV RNA levels 5 . 07±0 . 49 log copies/mL plasma ) and relatively comparable CD4+ T cell counts ( average 602 . 9±173 . 4 CD4+ T cells/mm3 blood ) ( Table 1 ) . These HIV-infected individuals are assumed to progress to AIDS if left untreated at a pace representative of the general HIV-infected population , and were therefore termed “putative progressors” ( PP ) . We chose these controls to minimize confounding by the extent of CD4+ T cell depletion , which could have been a cause rather than a consequence of the immunologic differences observed in prior studies of VNPs . VNPs in our study had been infected for a significantly longer period with a duration of diagnosis of 19 . 8 years ( ±4 . 8 ) compared to 1 . 56 years ( ±02 . 08 ) for PPs ( p<0 . 0001 , Figure 1A ) . Plasma HIV RNA levels tended to be higher in the PPs , despite the shorter length of infection ( p = 0 . 0532 , Figure 1B ) . The percentages of CD4+ T cells among CD3+ T cells were similar between the groups , with a mean of 24 . 7% CD4+ T cells in VNPs and 30 . 5% CD4+ T cells in PPs ( p = 0 . 0503 , Figure 1C ) , however VNPs maintained a higher absolute CD4+ T cell count than PPs ( p = 0 . 0370 , Figure 1D ) . Thus , even though we were successful in sampling HIV-infected controls who had not yet experienced significant peripheral blood CD4+ T cell depletion , the PP controls still had lower CD4+ T cell counts than VNPs . We next sought to determine whether CD4+ memory T cell populations were preserved in VNPs . We measured central memory CD4+ T cells ( TCM; live , singlet , CD3+CD4+CD27+CD45RO+CCR7+ cells ) , stem cell memory CD4+ T cells ( TSCM; live , singlet , CD3+CD4+CD27+CD45O-CCR7+CD95+ cells ) , naïve CD4+ T cells ( TN; live , singlet , CD3+CD4+CD27+CD45O-CCR7+CD95− cells ) and effector memory CD4+ T cells ( TEM; live , singlet , live , singlet , CD3+CD4+CD27−CD45RO+/dimCCR7− cells ) based on previous studies [23] ( representative flow cytometry staining Figure 2D ) . We found that there was no significant difference in the number of CD4+ TEM cells ( Figure 2A ) or CD4+ TCM ( Figure 2B ) in VNPs compared to PPs . However , the number of CD4+ TSCM cells were significantly depleted in PPs compared to VNPs ( p<0 . 0001 , Figure 2C ) , demonstrating that these important regenerative memory cells are preserved in VNPs . To assess whether lack of disease progression in VNPs compared to PPs was associated with decreased immune activation , we measured the frequencies of CD38+HLA-DR+ T cells in peripheral blood . Indeed , expression of the markers CD38 and HLA-DR is increased during chronic HIV infection and correlates with disease progression [27] . Unexpectedly , we found similar levels of T cell activation in VNPs compared to our PP cohort in both CD4+ and CD8+ T cells ( Figure 3A–B ) . To investigate further how VNP maintain normal CD4+ T cell counts despite active virus replication , we next evaluated Ki67 expression , an index of cellular cycling/proliferation . We found that the frequency of total CD4+ T cells that expressed Ki67 was significantly higher in VNPs than PPs in bulk CD4+ T cells ( p = 0 . 0101 , Figure 3C ) , but there was no difference in bulk CD8+ T cell proliferation ( Figure 3D ) . We further assessed proliferation in CD4+ T cell memory subsets by measuring expression of Ki-67 in memory CD4+ T cell subsets ( TEM , TCM , and TSCM ) . We observed a non-significant trend ( p = 0 . 0786 , Figure 3E ) towards higher frequencies of Ki-67+ CD4+ TEM from VNPs compared to levels in PPs . However , we found significantly higher frequencies of Ki-67+ CD4+ TCM and TSCM cells in VNPs compared to PPs ( p = 0 . 0101 , Figure 3F and p = 0 . 0002 Figure 3G respectively ) . Thus , overall , CD4+ T cells in VNPs had increased proliferation compared to PPs . The observation of increased frequencies of Ki-67+ CD4+ TCM and TSCM in VNP raises the possibility that these cells may be potential targets for virus infection as a result of their proliferation state . To directly test this possibility , we next determined the number of HIV DNA copies in flow cytometrically sorted CD4+ T cell subsets , including CD4+ TSCM , TCM , and TEM . We assessed HIV DNA copies by quantitative real time PCR as described in [28] . The levels of viral DNA in CD4+ TEM were similar in both groups ( p = 0 . 4458 , Figure 4 left ) . However , despite over a decade longer duration of untreated HIV infection and higher frequencies of cycling CD4+ T cells , VNPs had significantly lower levels of cell associated HIV DNA in their CD4+ TCM than in CD4+ TCM from PPs ( p = 0 . 0349 Figure 4 middle ) , as well as decreased levels of HIV DNA in TSCM in VNPs compared to PPs ( p = 0 . 0186 , Figure 4 right ) . In addition , while PP have significantly increased infection of TCM compared to infection in TEM ( p = 0 . 0411 , Figure 4 ) , there was not a significant difference between the HIV DNA in TCM and TEM of VNPs . These data are strikingly consistent with the observed low levels of SIVsmm DNA in CD4+ TCM during non-progressive infection of sooty mangabeys and in long-term non progressors [9] , [25] , and suggest that a possible common mechanism underlying the lack of disease progression in HIV-infected VNP , LTNPs , and SIVsmm-infected SMs is the relative resistance of CD4+ TCM from direct virus infection . In addition , VNPs also had decreased infection of TSCM cells , which have recently been described to be essential for long-lived memory reservoirs , and are preferentially infected in progressive HIV infection [22] , [23] . To our knowledge , TSCM infection has not been assessed in other non-progressive cohorts , however , preservation of long-lived memory CD4+ T cells is likely essential for lack of disease progression in VNPs . We next sought to determine what mechanisms may underlie protection of CD4+ memory T cell populations . However , differences were apparent when comparing the number of TCM to the frequency of proliferating memory CD4+ T cells . Indeed , in VNPs , we found a significant positive correlation between the frequencies of CD4+Ki67+ memory cells compared to the number of TCM cells ( r = 0 . 7333 , p = 0 . 0311 , Figure 5A ) . However , in PPs , this correlation did not exist ( r = 0 . 0091 , p = 0 . 9895 , Figure 5B ) , suggesting that proliferation of TCM cells may underlie disease protection after several years of HIV infection in VNPs , but not in PPs , who will ultimately have depletion of these cells if left untreated . This suggests that an increased and/or more efficient proliferation of CD4+ TCM may be a mechanism underlying preserved CD4+ T cells after several years of infection , despite active virus replication . While a similar trend as TCM existed between TSCM cells and frequency of proliferating CD4+ memory cells in VNPs , it did not reach significance ( r = 0 . 5000 , p = 0 . 1777 , Figure 5C ) , while the lack of association was consistent in PPs ( p>0 . 9999 , Figure 5D ) . Given the lack of a correlation between proliferation and the number of TSCM cells , despite the preserved nature of these cells in VNPs , we investigated whether the decreased HIV infection in TSCM cells was associated with the preservation of these cells . Indeed , we found a significant , negative correlation between the frequency of HIV-infected TSCM cells and the overall number of TSCM cells ( r = −0 . 6484 , p = 0 . 0194 , Figure 5E ) . However , of note , this relationship between HIV infection and cell count did not exist in the TCM cell subset ( r = −0 . 2246 , data not shown ) . Thus , these data indicate that a potential mechanism underlying preservation of TSCM cells in VNPs compared to PPs is decreased HIV infection of these essential long-lived cells . Finally , we assessed the role of CD4+ effector memory T cells ( TEM ) , and found that TEM cells had the opposite relationship with proliferation , with a trend towards a negative correlation between TEM cells and proliferating CD4+ memory cells in VNPs ( r = −0 . 4000 , p = 0 . 2912 Figure 5F ) , and a significant negative correlation between TEM cells and proliferating CD4+ memory T cells in PPs ( r = −0 . 6848 , p = 0 . 0347 , Figure 5G ) . Furthermore , we observed a significant correlation between the frequency of activated ( CD38+HLA-DR+ ) CD4+ T memory cells and TEM cells in both cohorts ( r = 0 . 6305 , p = 0 . 0050 , Figure 5H ) . However , no relationship was observed between TEM cells and HIV infection of TEM cells ( r = −0 . 0506 , data not shown ) , nor did we observe any relationship between CD4+ T cell memory activation and the frequency of TCM or TSCM cells ( p = 0 . 4365 and p = 0 . 2611 , respectively , data not shown ) . Taken together , these data indicate that different mechanisms may underlie preservation of CD4+ T cells in VNPs despite several years of infection and high virus replication , with proliferation mainly associated with TCM cells , lack of HIV infection associated with TSCM cells , and activation driving TEM cells . To further identify potential mechanisms underlying the VNP phenotype , we performed microarray analysis on RNA derived from whole blood from five VNP and seven PP HIV-infected individuals . Similarity of the transcriptomic profiles of individual patients was visualized using Principal Components Analysis ( PCA ) . PCA indicated that the total transcriptomic profiles clustered according to their status as either a VNP or PP ( Figure 6A ) indicating that there was a conserved transcriptomic signature specific to each patient cohort . Genes differentially regulated within each patient class were defined as those displaying a fold-change of greater than +/−1 . 5 relative to the reciprocal class , and with an unadjusted p-value of p = 0 . 05 defined by two-sample T-test . Using these criteria , we defined 769 probesets ( 476 higher in VNPs compared to PPs , 293 higher in PPs reciprocally ) representing 499 annotated transcripts that were differentially expressed between subsets ( listed in Tables S1 and S2 , respectively ) . The distribution of genes upregulated in each class using these criteria are depicted in Figure 6B . Hierarchical clustering demonstrated that the genes were consistently expressed across individuals within each patient class ( Figure 6D ) . To summarize biological functions represented within the differentially expressed genes , we tested for enrichment of pathways annotated within the Gene Ontology ( GO ) database ( Figure 6C ) . The pathways demonstrating the highest level of enrichment were those regulating immune processes and cellular motility pathways ( biological adhesion , locomotion , establishment of localization ) . 52 genes within these pathways were annotated as having immune function , and we further sub-classified these genes according to their expression levels in PPs and VNP ( represented in the forest plot in Figure 6E ) . The largest number of genes were in the category ‘immune response” , “leukocyte migration” and “leukocyte activation” –Figure 6G demonstrates individual activation-related genes higher in either phenotype: PPs had elevated expression of IFNAR1 , which encodes for one chain of the Type I IFN receptor , a probeset specific for KIR2-transcripts . We also noted elevated expression of STAT5B in PPs , which signals downstream of IL7 , a cytokine important for maintaining T cell homeostasis ( STAT5B was also the lone gene comprising the “leukocyte homeostasis” category ) ; however we were not able to detect significant enrichment genes downstream of IL7 . In addition , cytokine and chemokine analysis in plasma by luminex demonstrated that neither IL-7 , nor any other cytokines and chemokines measured , were significantly different in VNPs compared to PPs ( data not shown ) . In the “leukocyte migration” category , we noted expression of two chemokines: two independent probesets for CCL23 were upregulated in PPs ( data from one probeset shown in Figure 6G ) . CCL23 is highly homologous to CCL3/MIP1A , however acts as an agonist for CCR1 and has been shown to induce signaling in monocytes and resting T cells , but has not been studied in the context of HIV infection [29] . CXCL3 , which has been demonstrated to inhibit proliferation and induce apoptosis in T cells , was also upregulated in PPs [30] . There were a number of genes that were not incorporated into the GO annotation that have putative immune function ( depicted in “miscellaneous” in Figure 6G ) . Upregulated in PPs was TNFSF15/TL1A , a broadly activating co-stimulatory molecule from the TNF-superfamily associated with several inflammatory disorders ( Figure 6G ) . TNFSF15/TL1A ligation of its cognate receptor , DR3 , has been demonstrated to induced apoptosis in T cell lines in vitro and primary endothelial cells [31] , and recently was implicated in inducing inflammatory expression in CD4+CD161+ T cells in gut inflammation [32] . Also interesting were several leucine-rich repeats which had higher expression in VNPs , and which were recently demonstrated to inhibit activation of the NLRP3 inflammasome [32] . In previous work , we performed a microarray analysis of CD4+ and CD8+ T cells isolated from VNP and found that interferon stimulated genes ( ISGs ) were expressed at higher levels in chronically infected rapid progressor patients compared to VNPs , despite higher viral loads in the latter patient [14] . While a handful of ISGs had increased expression in VNP or PP ( Figure 6F ) , the majority of probes representing ISGs did not display any significant variation , nor did they exhibit enriched expression in either phenotype ( Figure 6H ) , despite the elevated expression of IFNAR1 in the PPs ( Figure 6G ) . To increase the sensitivity of detecting molecular pathways differentially regulated between patient groups , we used gene set enrichment analysis ( GSEA ) [33] , and confirmed that there was not a consistent upregulation of Type I interferon between PPs and VNPs ( Figure 6H ) . We next used GSEA to test for the enrichment of transcripts involved in other immune activation pathways . To identify other potential pathways of immune activation , we selected six genesets from MSIGDB that are related to immune activation and inflammation . Of note , we noted that the pathway representing canonical IL6 signaling ( BIOCARTA ID:M5489 ) was significantly enriched in the PP phenotype ( Figure 6I ) . Further analysis demonstrated that multiple genesets representing overlapping but distinct genes implicated in IL6 signaling were enriched in the PP phenotype relative to VNPs . Within the microarrary data , IL6 had very low but consistent upregulation in PPs ( fold-change = 1 . 18 , p = 0 . 006 , data not shown ) . The observation of multiple genesets demonstrating consistent enrichment in IL6 signaling indicates that PP patients have overall elevated expression of genes in this pathway . However , no differences in IL-6 , or any other cytokines or chemokines measured , were observed in the plasma between PP and VNP patients ( data not shown ) . Thus , the differences in cytokine signaling and/or leukocyte activation observed by microarray may reflect overt immunological dysfunction in the PPs compared to the VNPs , consistent with progressive HIV phenotype [34] , [35] . Viremic Non-Progressors ( VNPs ) are infrequent among HIV-infected individuals and remain clinically asymptomatic and maintain high CD4+ T cell counts despite many years of infection with robust virus replication . Interestingly , these VNPs show striking similarities with SIV-infected sooty mangabeys ( SMs ) , an African “natural” host non-human primate species whose infection is typically non-pathogenic and characterized by healthy CD4+ T cell counts and low immune activation despite high levels of virus replication ( Reviewed in [36] ) . Earlier reports [14] , [15] , [18] , [19] compared VNPs to HIV-infected individuals who had been infected longer , with much greater degrees of CD4+ T cell depletion . To address this issue , we studied a group of stringently defined VNPs and , in contrast to previous studies , we compared them to HIV-infected individuals with similar CD4+ T-cell counts and viral load , mainly in early infection ( i . e . “putative progressors”; PP ) . Of note , comparing subjects mainly in early infection has the potential caveat that the results are influenced by length of infection . However , there is no correlation between years infected and infection frequency in any of the CD4+ T cells subsets measured ( data not shown ) . In addition , previous studies have demonstrated that chronically infected individuals have high levels of HIV DNA and/or reservoir in TCM and TSCM cells , and increased HIV DNA and RNA compared to recently infected individuals , supporting the concept that this comparative cohort of PPs is appropriate [22] , [37]–[39] . Here we demonstrated that VNPs have decreased infection of both stem cell memory and central memory CD4+ T cells by HIV , and also had increased frequencies of Ki-67+ CD4+ T cells . Because the expression of Ki-67 was higher in VNPs than in PPs , while immune activation markers such as HLA-DR and CD38 were expressed at similar levels , we propose that , in VNPs , the combination of increased CD4+ T memory cell proliferation and low levels of direct virus infection of CD4+ TSCM and TCM cells may be reflective of more efficient homeostatic proliferation , a physiologic process induced by loss of lymphocytes , rather than overt activation . Indeed , in non-pathogenic infection of SMs , there is a similar association between TCM count and proliferation [40] . However , whether these factors represent a cause or consequence of lack of progression in VNPs and altered TCM and TSCM infection is unclear . Indeed , in VNPs , proliferation of CD4+ TCM was associated with increased numbers of CD4+ TCM , while the opposite was true in PPs . And while a similar trend existed in TSCM cells , the strongest association with preservation of TSCM cells in VNPs is their lower frequency of HIV infection . Of note , the absolute number of CD4+ T cells that are infected TCM in VNPs compared to PPs loses significance , but maintains a trends towards decreased infection ( p = 0 . 0999 , Figure S2 ) . However , a significant difference between VNPs and PPs is maintained when the absolute number of TSCM cells that are infected is calculated ( p = 0 . 0295 , Figure S2 ) . In this study , we found that VNP HIV-infected individuals were remarkably similar to PPs in terms of markers of systemic immune activation as well as overall profile of gene expression . This finding seemingly is in contrast to previous studies suggesting that VNPs had low immune activation and a profile of gene expression characterized by low ISG expression [14] , [15] . However , we believe that the choice of PP HIV-infected individuals as controls allowed us to identify the immunological features of VNPs that distinguish this rare patient population , avoiding confounding by progressive infection . Collectively , the transcriptomic data have identified that VNPs have elevated expression of several genes associated with immune homeostasis , and conversely , lower expression of inflammatory genes compared to non-VNPs . However , given the disparate associations we observed between CD4+ memory cell subsets and activation ( TEM cells ) , proliferation ( TCM cells ) and HIV infection ( TCM cells ) , RNA studies in whole blood as was performed here may not be as informative , and future studies that sort memory subsets would be of great interest . While these data provide only an associative link between the identified candidates and the VNP phenotype , in combination with the immunophenotyping data , the gene expression adds further support to a model in which HIV infection in VNPs is associated with lower inflammation . In this regard , the current study suggests that the low immune activation and absence of significant ISG up-regulation in VNPs could potentially be a consequence rather than a cause of their preserved immune status . One potential explanation of the VNP phenotype is that these individuals are infected with HIV-1 strains that show intrinsically lower infectivity and/or cytopathicity compared to strains isolated from normal progressors . However , the study by Choudhary et al . indicated that , in organ culture , HIV isolates derived from VNP were as cytopathic as viruses isolated from normal HIV progressors [15] . However , we performed a supplemental study in an additional group of HIV-infected individuals , including several individuals with a phenotype similar to viremic non-progressors ( Table S3 ) to investigate the accessory Nef protein . It has been previously demonstrated that the efficient suppression of T cell activation and apoptosis by Nef-mediated down-modulation of TCR-CD3 may help the infected host to prevent chronic immune activation and CD4+ T cell depletion [41] . However , just like nef alleles from HIV-1-infected individuals with progressive infection , those derived from VNPs were generally unable to remove CD3 from the cell surface ( Figure S1 ) . Overall , the differences in Nef function between VNP and chronic progressor HIV-infected individuals were much more subtle than those established for HIV-1 and SIVsmm Nefs [41] , and it is unclear whether differences in Nef function are a cause or consequence of differences in disease progression . An additional potential mechanism of protection from disease progression in VNPs is CD8+ T cell mediated immunity . In our analysis , while we observed an increase in CD8+ T cell count in VNPs , we did not find an increase in proliferation or associations between CD8+ T cell subsets and proliferation , or HIV levels in CD4+ T cells as we observed for CD4+ T cells ( data not shown ) . In addition , given that virus load is not controlled in plasma , overall CD8+ T cell control is unlikely , and previous studies of viremic controllers demonstrated that CD8+ T cell immunity was not increased [19] . However , in long-term non-progressors with low viremia , HIV-specific CD8+ T cell responses are associated with limited TCM infection , particularly in HLA-B27 and HLA-B57 patients [9] . Indeed , a potential mechanism may exist whereby CD8+ T cells can mount preferential protection against TCM and TSCM infection , and this possibility should be investigated in future work . In addition , while we saw no significant difference in the expression of CCR5 on CD4+ T cells subsets between VNPs and PPs in this study , the role of HIV co-receptors in protection from infection in VNPs should be further investigated . Lastly , another possible mechanism for protection is differential expression of restriction factors in CD4+ T cells subsets of VNPs . Indeed , understanding the mechanisms by which these cells are protected will be crucial in understanding the lack of progression and potential intervention strategies . The observation that VNPs have significantly lower infection of both CD4+ TCM and TSCM than do the same subsets in PPs identifies a novel , potentially crucial mechanism of protection of CD4+ T cell homeostasis in this rare subset of HIV-infected individuals . In addition , it identifies another striking similarity between VNPs and naturally SIVsmm-infected SMs , who also experience a non-pathogenic , immunologically benign infection despite chronic virus replication [20] . Our observation that TCM and TSCM in VNPs harbor less HIV DNA as opposed to PPs is also consistent with another recent report suggesting that VNPs tend to have lower T cell activation than progressors in peripheral blood , yet higher T cell activation in the rectal mucosa , where a much higher proportion of CD4+ T cell have an effector phenotype [18] . Preservation of CD4+ TCM and TSCM from direct virus infection may be of particular importance during HIV and SIV infections , as these cells are longer lived than CD4+ TEM , and proliferation of TSCM feeds the CD4+ TCM cell pool , which in turn is essential to maintain a sufficient number of CD4+ TEM in mucosal tissues [42] . Indeed , previous studies by Okoye et al . have elegantly shown that while CD4+ TEM depletion is the proximate mechanism of immunodeficiency , the tempo of SIV disease progression is largely determined by destruction , failing production , and gradual decline of CD4+ TCM cells [42] . Thus , a shared mechanism based primarily on preserving CD4+ TSCM and TCM cells from virus infection may underlie the lack of disease progression in both VNPs and SIVsmm-infected SMs . Finally , emerging data suggest TSCM cells represent an important niche for replication-competent viral reservoir , especially given their ability to harbor immense amounts of virus when measured on a per cell basis [22] . TSCM cells stably persist in secondary lymphoid organs and provide multipotent and self-renewing potential which allows for the incorporation of abundant virus into other T cell memory phenotypes downstream of proliferating TSCM cells [22] , [23] . Thus , future studies to determine possible mechanisms underlying TCM and TSCM cell resistance to direct virus infection , such as genetic factors , co-receptor regulation , restriction factor expression and viral determinants may provide critical information to better understand how VNPs avoid CD4+ T cell loss and maintain attenuated disease progression . HIV-infected viremic non-progressor ( VNP ) and putative progressor ( PP ) samples were sampled from the UCSF SCOPE and OPTIONS cohorts , respectively . VNPs were defined as having confirmed HIV-1 infection for more than 9 years with sustained plasma HIV RNA levels >10 , 000 copies/ml and maintenance of peripheral blood CD4+ T cell counts >500 cells/mm3 and a CD4% ( of all lymphocytes ) >15% ( Table 1 ) . Recently HIV-infected PPs were defined as having plasma HIV RNA levels >10 , 000 copies/mL , CD4+ T cell counts >400 cells/mm3 and having been initially infected with HIV 2 months to 7 years prior to the index visit ( Table 1 ) . The estimated date of initial HIV infection was calculated according to published algorithms that incorporate “de-tuned” anti-HIV-1 antibody ELISA results [43] , [44] or by a documented sero-conversion window of <6 months . All participants were required to be antiretroviral therapy ( ART ) -naïve . Cryopreserved PBMCs were isolated from whole blood , and stored at the UCSF AIDS Specimen Bank . T cell activation was measured by the UCSF Core Immunology Laboratory , as previously described and optimized [45] . Cryopreserved PBMCs were thawed and stained with the following markers: Aqua Amine Reactive Dye ( Invitrogen , Carlsbad , CA ) , CD3 Pacific Blue , CCR5 PE-CY5 ( BD Pharmingen , San Jose , CA ) , CD38 PE , HLA-DR FITC , ( BD Biosciences ) , CD4 PE Texas Red , and CD8 QDot 605 ( Invitrogen ) . CD4+ T cells were sorted into TCM and TEM subsets using a BD FACS Aria ( BD Biosciences , San Jose , CA ) run by BD FACS DIVA software . TCM were defined as live , singlet , CD3+CD4+CD27+CD45RO+CCR7+ cells . TEM were defined as live , singlet , CD3+CD4+CD27−CD45RO+/dimCCR7− cells , TSCM were defined as live , singlet , CD3+CD4+CD27+CD45O-CCR7+CD95+ cells as previously described and as demonstrated in Figure 2D [23] . Cryopreserved PBMCs were thawed and stained with predetermined optimal concentrations of the following markers: Aqua Amine Reactive Dye ( Invitrogen ) , CD3 AL700 , CD8 PERCPCY5 . 5 , CCR5 PE , CCR7 PE-CY7 , CD95 PE-CY5 and Ki67 FITC ( BD Pharmingen ) , CD4 eFluor450 , CD27 APC-eFluor780 ( eBiosciences , San Diego , CA ) , CD45RO ECD ( Beckman Coulter , Chaska , MN ) . HIV DNA was quantified in CD4+ TCM and TEM subsets using an ABI StepOnePlus real time PCR system ( Life Technologies , Grand Island , NY ) as previously described [28] , [46] . Albumin was used to determine cell number in each reaction , and gag DNA was used simultaneously quantified to determine HIV levels with HIV gag forward primer: GGTGCGAGAGCGTCAGTATTAAG; HIV gag reverse primer: AGCTCCCTGCTTGCCCATA; and HIV gag probe: AAAATTCGGTTAAGGCCAGGGGGAAAGAA . Duplicate reactions were run and template copies were calculated with ABI software . If no viral DNA was amplified from a given cell population , we report half the lower limit of detection , based on twice the number of cells put into each PCR as previously described [28] . RNA extraction and microarray analysis were conducted at the Yerkes NHP Genomics Core Laboratory ( http://www . yerkes . emory . edu/nhp_genomics_core/ ) . Whole blood was from HIV-infected donors was collected into RNA PAXgene tubes ( QIAGEN , Valencia , CA ) and purified as previously described [47] . Purified RNA was assessed by Nanodrop and Agilent Bioanalyzer analysis; all samples had RIN scores>8 . 0 . 100 ng of total RNA was amplified , labeled and hybridized to Affymetrix Human U133 Plus 2 . 0 arrays ( Affymetrix , Santa Clara , CA ) using the NuGEN Ovation RNA Amplification System V2 , Ovation WB Reagent and Encore Biotin Module according to manufacturer's specifications ( NuGEN Inc , San Carlos , CA ) . After hybridization , arrays were washed on Affymetrix FS450 fluidics stations using the NIRAV-WASH protocol and scanned on an Affymetrix 3000 7G GeneChip Scanner . CEL files from individual arrays were preprocessed and normalized by RMA within PARTEK Genomics Suite Software . NUSE and RLE plots were inspected to ensure there were no outlier arrays at the hybridization level . PCA was on the RMA normalized data and determined one VNP patient to be an outlier from both VNPs and PPs , and was removed from downstream statistical analysis . To determine differentially expressed genes , a two-tailed T-test was run using Partek Genomics Suite software ( v . 6 . 13 , Partek Inc , St . Lousi MO ) . Using Benjamini-Hochberg correction for multiple hypothesis testing yielded only two unannotated transcripts detected as differentially expressed between patients . Several genes within the dataset were upregulated or downregulated several-fold , with consistent expression values across multiple redundant probesets , suggesting the BH corrected p-value ( 1 . 8e-06 ) was overly stringent . To prioritize genes with differential expression , we relaxed our gene-filtering criteria to an unadjusted P -value of P = 0 . 05 and a fold-change of +/−1 . 5 between classes . Lists of genes with differential expression between classes using this criteria are contained in Table S1 ( elevated in VNPs ) and Table S2 ( elevated in PPs ) . Gene Set Enrichment Analysis ( GSEA ) was used as a more sensitive method to detect significantly enriched pathways in either the VNP or PP transcriptomes . GSEA software was downloaded ( http://www . broadinstitute . org/gsea/index . jsp ) and run locally using the following parameters: Signal2Noise metric for ranking genes; the dataset and genesets were converted into Gene Symbols; 1000 geneset permutations; redundant probesets were collapsed using the max probeset and the ‘weighted’ enrichment statistic was employed . Microarray data was submitted to the GEO database according to MIAME standards . Statistical calculations were performed as follows: ( i . ) Mann-Whitney non-parametric t test for comparison of VNPs to PPs; ( ii . ) Paired t test for comparison of VNPs to VNPs and PPs to PPs; ( iii . ) Spearman correlation with linear regression was used for all correlative analysis . All statistical analysis was performed using Graph Pad Prism , version 5 . 0 . P values of <0 . 05 were considered significant for nef analysis . The raw data for the microarray analysis has been deposited at the NCBI GEO database ( http://www . ncbi . nlm . nih . gov/geo/ ) under accession #GSE57730 . All SCOPE and OPTIONS cohort samples ( Table 1 ) were obtained after written informed consent and approval by the University of California San Francisco Institutional Review Board . All BRESCIA cohort patient samples ( Supporting data ) were obtained after written informed consent and approval by the University of Brescia Institutional Review Board .	Here we assessed correlates of protection from disease progression in a rare subset of HIV-infected individuals , viremic non-progressors ( VNP ) . These individuals have high viral load for several years . In contrast to the majority of infected individuals , however , these individuals do not progress to AIDS . Here we found this lack of progression was associated with selective preservation of two critical subsets of memory CD4+ T cells , central memory ( TCM ) and stem-cell memory ( TSCM ) cells . Compared to HIV-infected putative progressors , VNPs had higher proliferation of these indispensable subsets of memory cells . In addition , the long-lived CD4+ TCM and TSCM cells in VNPs had decreased HIV infection compared to the less critical effector memory CD4+ T cells , which indicates a possible mechanism by which VNPs maintain their CD4+ T cell pool after several years of infection , and remain free from AIDS progression .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "immunodeficiency", "viruses", "infectious", "diseases", "medicine", "and", "health", "sciences", "immune", "cells", "diagnostic", "medicine", "cell", "biology", "animal", "cells", "clinical", "immunology", "medical", "microbiology", "hiv", "viral", "pathogens", "hiv", ...	2014	Limited HIV Infection of Central Memory and Stem Cell Memory CD4+ T Cells Is Associated with Lack of Progression in Viremic Individuals
Atypical porcine reproductive and respiratory syndrome ( PRRS ) , which is caused by the Chinese highly pathogenic PRRS virus ( HP-PRRSV ) , has resulted in large economic loss to the swine industry since its outbreak in 2006 . However , to date , the region ( s ) within the viral genome that are related to the fatal virulence of HP-PRRSV remain unknown . In the present study , we generated a series of full-length infectious cDNA clones with swapped coding regions between the highly pathogenic RvJXwn and low pathogenic RvHB-1/3 . 9 . Next , the in vitro and in vivo replication and pathogenicity for piglets of the rescued chimeric viruses were systematically analyzed and compared with their backbone viruses . First , we swapped the regions including the 5′UTR+ORF1a , ORF1b , and structural proteins ( SPs ) -coding region between the two viruses and demonstrated that the nonstructural protein-coding region , ORF1b , is directly related to the fatal virulence and increased replication efficiency of HP-PRRSV both in vitro and in vivo . Furthermore , we substituted the nonstructural protein ( Nsp ) 9- , Nsp10- , Nsp11- and Nsp12-coding regions separately; or Nsp9- and Nsp10-coding regions together; or Nsp9- , Nsp10- and Nsp11-coding regions simultaneously between the two viruses . Our results indicated that the HP-PRRSV Nsp9- and Nsp10-coding regions together are closely related to the replication efficiency in vitro and in vivo and are related to the increased pathogenicity and fatal virulence for piglets . Our findings suggest that Nsp9 and Nsp10 together contribute to the fatal virulence of HP-PRRSV emerging in China , helping to elucidate the pathogenesis of this virus . Porcine reproductive and respiratory syndrome ( PRRS ) is characterized by reproductive failure in sows and respiratory diseases in all ages of pigs [1] , [2] . This disease was first reported in the United States in the late 1980s [3] , and in Germany in 1990 , and then this disease became widespread throughout the world [4] , [5] . The causative agent of this disease , PRRSV , was first identified in Europe in 1991 and independently in the USA in 1992 [6] , [7] . PRRSV subsequently spread to the pig-producing countries and areas worldwide [1] , [8]–[10] , resulting in considerable economic loss to the pig industry worldwide [2] , [11] , [12] . PRRSV is an enveloped , positive-strand RNA virus that belongs to the family Arteriviridae , together with equine arteritis virus ( EAV ) , mouse lactate dehydrogenase-elevating virus ( LDV ) and simian hemorrhagic fever virus ( SHFV ) [13] . Based on genetic and antigenic characteristics , two major genotypes of PRRSV , type 1 ( European ) and type 2 ( North American ) , have been identified and share approximately 55–70% nucleotide identity [14] . The phylogenic analyses of the ORF7 sequence show that the European PRRSV can further be divided into three subtypes , pan European subtype 1 and East European subtypes 2 and 3 [15] , and the North American PRRSV can be classified into at least 9 distinct genetic lineages [16] . The PRRSV genome is approximately 15 kb and consists of at least ten overlapping open reading frames ( ORFs ) [13] , [17]–[20] . ORF1a and ORF1b , which occupy approximately three-fourths of the viral genome , encode the replicase polyproteins pp1a and pp1ab , which are autoproteolytically cleaved into at least 15 functional nonstructural proteins ( Nsps ) involved in virus replication and in transcription [21]–[26] . ORF1a encodes Nsp1α/β to Nsp8 , including three important virus proteases , Nsp1 ( papain-like cysteine protease ) , Nsp2 ( chymotrypsin-like cysteine protease ) and Nsp4 ( 3C-like serine protease ) . ORF1b is composed of Nsp9 ( viral RNA-dependent RNA polymerase , RdRp ) , Nsp10 ( RNA helicase ) , Nsp11 ( endoribonuclease ) and Nsp12 . In particular , Nsp9 and Nsp10 are regarded as crucial enzymes for viral RNA sythesis [27]–[29] . Recently , Nsp11 , with a NendoU domain , has been identified as an important participant in viral genome synthesis and in interferon ( IFN ) inhibition [30]–[33] . The ORFs 2 to 7 encode the structural proteins ( SPs ) of the virion , which are involved in virus infectivity , neutralizing antibody elicitation etc . [34]–[37] . The PRRSV strains circulating in the field are biologically , antigenically , and genetically heterogeneous , leading to extreme diversity in the clinical phenotypes and severities induced by the PRRSV infection [38]–[41] . Previous studies , based on genomic sequence comparisons between the parental virulent and attenuated strains , suggested that important determinants associated with virulence or attenuation might be scattered throughout the viral genome [42]–[47] . However , these speculations could not be easily proven , and some genomic mutations may result from the adaptation of the virus strain to propagate in the host cells but may not necessarily relate to in vivo virulence . A study swapped the 5′UTR+ORF1ab or SP+3′UTR region between a highly pathogenic strain MN184 and PRRS MLV and the pathogenicity analyses showed that both the regions from MLV could attenuate MN184 , suggesting that the replicase gene is an important player in viral virulence and attenuation [48] . Another report indicated that PRRSV virulence was multigenic and that the Nsp3–8 regions were the major virulence determinants based on substituting a series of small genomic regions of a highly virulent strain FL-12 with their counterparts from an attenuated vaccine strain PrimePac [49] . An unparalleled , large-scale , atypical PRRS outbreak caused by the highly pathogenic PRRSV ( HP-PRRSV ) was documented in 2006 in China [40] , [50] , [51] and subsequently , in neighboring Asian countries . In contrast with previous Chinese PRRSV strains , the highly virulent PRRSV isolates with fatal pathogenicity for pigs can induce elevated body temperatures ( over 41°C ) and severe clinical presentations , including depression , anorexia , lethargy , rubefaction on the skin , respiratory distress , shivering and diarrhea [40] , [50] . The highest mortality of nursery pigs in the infected farms could reach 100% . In recent years , HP-PRRSV has been recognized as the dominating virus in China [52] . Chinese HP-PRRSV strains share a unique genomic characteristic , namely a discontinued 30-amino-acid ( 30-aa ) deletion within their Nsp2-coding regions [52]–[54] . Our previous studies have confirmed that this deletion is not related to the fatal virulence of HP-PRRSV [40]; however , whether the genome of the virus has the fatal virulence-determining or related region ( s ) remains unknown . A strain belonging to the 3 subtypes of PRRSV genotype 1 , Lena , with increased virulence , occurred in Eastern Europe , and a recent study demonstrated that Lena , with high replication efficiency , caused high fevers , anorexia , depression and severe respiratory problems , similar to Chinese HP-PRRSV [55] , [56] . Thus , analyzing the molecular mechanism associated with the enhanced virulence of PRRSV is helpful for controlling this disease . The objective of this study was to explore the genomic region ( s ) possibly related to the virulence of Chinese HP-PRRSV . We used the infectious clones ( pWSK-JXwn and pWSK-HB-1/3 . 9 ) of a HP-PRRSV strain , JXwn06 , and of a low pathogenic PRRSV ( LP-PRRSV ) but genetically close strain , HB-1/3 . 9 , as backbones and constructed a series of chimeric clones by individually exchanging the corresponding coding region within the genome between the two parental clones . Next , we rescued the viruses , analyzed their growth efficiency in vitro and in vivo , as well as their pathogenicities for piglets , and , finally determined the genomic regions that contribute to the fatal virulence of Chinese HP-PRRSV . The animal trials in this study were performed according to the Chinese Regulations of Laboratory Animals—The Guidelines for the Care of Laboratory Animals ( Ministry of Science and Technology of People's Republic of China ) and Laboratory Animal-Requirements of Environment and Housing Facilities ( GB 14925-2010 , National Laboratory Animal Standardization Technical Committee ) . The license number associated with their research protocol was 20120611-01 , which was approved by The Laboratory Animal Ethical Committee of China Agricultural University . MARC-145 cells and BHK-21 cells were cultured in Gibco Dulbecco's modified Eagle medium ( DMEM ) ( Invitrogen ) , which was supplemented with 10% fetal bovine serum ( FBS ) ( HyClone ) at 37°C under a humid 5% CO2 atmosphere . BHK-21 cells were used for full-length infectious cDNA clone transfection . Pulmonary alveolar macrophages ( PAMs ) , which are the primarily target cells for PRRSV , were prepared from the lung lavage fluid of 5- to 6-week-old healthy piglets free of PRRSV as previously described [6] , [57] . Primary PAMs were maintained in GIBCOTM RPMI 1640 medium ( Invitrogen ) , which was supplemented with 10% FBS , 100 mg/ml kanamycin , 50 U/ml penicillin , 50 mg/ml streptomycin , 25 mg/ml polymixin B and 1 mg/ml fungizone at 37°C , 5% CO2 or were cryopreserved in liquid nitrogen for later use . The full-length infectious cDNA clone plasmids ( pWSK-JXwn and pWSK-HB-1/3 . 9 ) of HP-PRRSV JXwn06 and of low pathogenic ( LP ) PRRSV HB-1/3 . 9 and the rescued viruses ( RvJXwn and RvHB-1/3 . 9 ) were used in this study [40] , [58] . The 5′UTR+ORF1a or ORF1b between pWSK-JXwn and pWSK-HB-1/3 . 9 was swapped using the unique restriction enzymes ( New England Biolabs ) NdeI and NheI or NheI and AscI , respectively . The structural protein ( SP ) -coding regions from one full-length plasmid and flanking segments of corresponding sites from the other full-length plasmid were individually amplified by PCR using the primers listed in Table S1 . Then , a new fragment D of pWSK-JXwn , which contained the SP-coding region of pWSK-HB-1/3 . 9 , and a new fragment E of pWSK-HB-1/3 . 9 , which contained the SP-coding region of pWSK-JXwn were generated using fusion PCR ( Fig . 1A ) . Then , the PCR products were purified , digested by the restriction enzymes ( AscI and PacI for fragment D , AscI and RsrII for fragment E ) , and finally inserted back into the backbones , which were digested by the same restriction enzymes . The chimeric full-length cDNA clones with the pWSK-JXwn backbone were named pWSK-JH1a , pWSK-JH1b and pWSK-JHSP; correspondingly , the chimeric clones with the pWSK-HB-1/3 . 9 backbone were designated pWSK-HJ1a , pWSK-HJ1b and pWSK-HJSP . The strategies of exchanging the nonstructural protein ( Nsp ) -coding regions were similar to the method described above . The target region was amplified from one full-length plasmid and linked with the flanking fragments amplified from the other full-length plasmid by fusion PCR using the primers in Table S1 , namely , a series of new fragments , C and D , of pWSK-JXwn containing the Nsp9- , Nsp10- , Nsp11- , Nsp12- , Nsp9- and Nsp10- or Nsp9- , Nsp10- and Nsp11-coding regions of pWSK-HB-1/3 . 9 . Additionally , a series of new fragments , D and E , of pWSK-HB-1/3 . 9 containing the corresponding regions of pWSK-JXwn were constructed . Then , the fusion PCR products were inserted into the full-length backbone plasmid by using the closest unique enzyme sites ( Fig . 1B ) . The chimeric full-length cDNA clones with the pWSK-JXwn backbone , containing the Nsp-coding region from pWSK-HB-1/3 . 9 , were individually designated pWSK-JHn9 , pWSK-JHn10 , pWSK-JHn11 , pWSK-JHn12 , pWSK-JHn9n10 and pWSK-JHn9n10n11 , and correspondingly , the chimeric clones with the RvHB-1/3 . 9 backbone were designated pWSK-HJn9 , pWSK-HJn10 , pWSK-HJn11 , pWSK-HJn12 , pWSK-HJn9n10 and pWSK-HJn9n10n11 . To rescue the chimeric viruses , each full-length cDNA clone plasmid was separately linearized using the restriction enzyme PacI ( for cDNA clones with RvJXwn as the backbone ) or RsrII ( for cDNA clones with RvHB-1/3 . 9 as the backbone ) . Then , the purified full-length cDNA clone was transcribed and capped using a mMessage high-yield capped RNA transcription kit ( Ambion ) according to the manufacturer's protocol . The purified RNAs were transfected into BHK-21 cells as previously described [40] . At 24 h post-transfection , the supernatants were harvested and serially passaged for three times in MARC-145 cells . The MARC-145 cells infected with the rescued viruses were examined by indirect immunofluorescence assay ( IFA ) using the monoclonal antibody ( McAb ) SDOW17 ( Rural Technologies ) , which is specific for the PRRSV N protein [59] , [60] . To further verify the correctness of the exchanged regions , the RNAs of third-passage chimeric viruses were extracted and subjected to RT-PCR , followed by sequencing . To analyze the in vitro growth characteristics , the MARC-145 cell monolayers or primary PAMs in T-25 flasks were individually infected with each chimeric virus and with their parental viruses at the same multiplicity of infection ( MOI ) of 0 . 01 . The titers of virus in the supernatants at different time points were determined using a microtitration infectivity assay and expressed as 50% tissue culture infective dose per ml ( TCID50/ml ) . Briefly , the confluent monolayers of MARC-145 cells cultured in 96-well plates were incubated with 10-fold serially diluted virus suspensions . After absorption for 1 h at 37°C , the supernatants were removed , and 5% DMEM was added . The plates were incubated for an additional 48 to 60 h , and then the virus titers were determined . Each time point was independently repeated three times . Healthy , 6-week-old , landrace piglets that were free of PRRSV , classic swine fever virus ( CSFV ) , pseudorabies virus ( PRV ) , porcine circovirus type 2 ( PCV2 ) and M . hyopneumoniae infections were obtained from Beijing Center for SPF Swine Breeding & Management . All piglets were confirmed to be negative for PRRSV and PCV2 antibodies , as determined using commercial ELISA kits and using RT-PCR or PCR for viral nucleic acid detection in sera . The animals were raised in the animal facilities at China Agricultural University ( CAU ) . For the first batch of animal trials , forty-five piglets were randomly divided into nine groups . The animals in each group ( n = 5 ) were separately raised in different isolation rooms . Each piglet in each infection group was intranasally administered with 2 ml of each virus containing 2×105 TCID50 ( RvJXwn , RvJH1a , RvJH1b , RvJHSP , RvHB-1/3 . 9 , RvHJ1a , RvHJ1b or RvHJSP ) . Each piglet in the control group was mock-inoculated with the same dose of MARC-145 cell culture supernatant . For the second batch of animal trials , fifty-five piglets were randomly allotted to eleven groups ( n = 5 ) . Each piglet in each infection group was intranasally inoculated with 2 ml of each virus containing 2×105 TCID50 ( RvJHn9 , RvJHn10 , RvJHn9n10 , RvJHn9n10n11 , RvHJn9 , RvHJn10 , RvHJn9n10 or RvHJn9n10n11 ) . The piglets in the control group were mock inoculated with 2 ml of MARC-145 cell culture supernatant . All the survived piglets were euthanized and necropsied on day 21 post-inoculation ( pi ) . The daily observation of clinical signs and measurement of rectal temperatures were performed throughout the experiments . For the second batch of animal experiments , the clinical symptoms of each piglet were monitored and scored . A detailed scoring system is summarized in Table S2 . Meanwhile the piglets were weighed on days 0 , 7 , 14 and 21 pi to calculate the average daily gain ( ADG ) . Necropsy and gross pathological examinations of lungs were immediately performed once the piglets died during the experiment , or the survived piglets were euthanized at the termination of the experiment . The total amount of lungs affected by grossly visible pneumonia were evaluated as the percentage of lesions noted per lobe , and an overall level of gross pathology was determined according to the previously described standard scoring system [61] . Lung tissues were collected , fixed with 4% paraformaldehyde solution at room temperature for 48 h and then processed by routine histopathological procedures . Each sample was examined on two sections . One section was stained with hematoxylin and eosin ( H&E ) for observing pathological changes , and the other section was stained with McAb SDOW17 at 1∶2000 dilutions for detecting PRRSV N antigen positive cells . Lung sections stained with H&E were blindly evaluated by a veterinary pathologist and the scores from 0 to 4 , which accounted for the distribution and severity of interstitial pneumonia were recorded as previously described [61] . The indications for the scores were as follows: 0 = no microscopic lesions; 1 = mild , focal to multifocal interstitial pneumonia; 2 = moderate , multifocal to coalescing interstitial pneumonia; 3 = severe , patchy to coalescing and extensive interstitial pneumonia; and 4 = severe and diffuse interstitial pneumonia . The detection of the PRRSV antigen was executed through a ranked score of 0–4 , which was used to evaluate of the number of positive cells per section taken from each block , as previously described [62] . The indications for the scores were as follows: 0 = no positive cells , 1 = 1–10 positive cells , 2 = 11–30 positive cells , 3 = 31–100 positive cells , and 4 = or >100 positive cells . Serum samples were collected on days 0 , 3 , 5 , 7 , 10 , 14 and 21 pi to examine the viral loads of the infected animals using a microtitration infectivity assay , as described above , and to detect PRRSV-specific antibodies using a commercial IDEXX Herdchek PRRS 2XR ELISA kit . The significant differences of the in vitro experiment and animal trials were analyzed using a one-way or two-way RM ANOVA in the GraphPad Prism ( version 5 . 0 ) software . A t-test was used to estimate the variability among the gross lesion , histopathological and immunohistochemistry scores of lungs . Differences were considered statistically significant at a P value of <0 . 05 and extremely significant at a value of P<0 . 01 or P<0 . 001 . GenBank accession numbers ( http://www . ncbi . nlm . nih . gov/Genbank ) : PRRSV JXwn06 , complete genome: EF641008 PRRSV HB-1/3 . 9 , complete genome: EU360130 Six chimeric viruses were successfully rescued from the chimeric infectious clones constructed by swapping 5′UTR+ORF1a , ORF1b or structural proteins ( SP ) -coding regions between pWSK-JXwn and pWSK-HB-1/3 . 9 plasmids . The MARC-145 cells infected with each chimeric virus were shown to be positive for PRRSV N protein using IFA ( Supplemental Fig . S1 ) . Further sequencing of the replaced regions and their flanking areas of the third-passage viruses confirmed that the substitutions were consistent with the original design and that no additional mutations were introduced during the construction ( data not shown ) . The six rescued viruses were individually designated RvJH1a , RvJH1b , RvJHSP , RvHJ1a , RvHJ1b and RvHJSP ( Fig . 1A ) . The growth kinetics of chimeras in parallel with their parental viruses , RvJXwn and RvHB-1/3 . 9 , were evaluated by infecting MARC-145 cells or primary PAMs . The results showed that RvJH1a and RvHJ1a had lower growth kinetics , whereas RvJHSP and RvHJSP had similar growth kinetics to their parental backbone viruses in both MARC-145 cells and primary PAMs ( Fig . 2 ) , indicating that exchanging the 5′UTR+ORF1a region can impair the replication efficiency of the viruses , whereas substituting the SP-coding region does not affect this efficiency . Importantly , compared with the backbone viruses , RvJH1b showed a reduced replication efficiency in both MARC-145 cells and primary PAMs ( Fig . 2A and B ) , particularly in MARC-145 cells; in contrast , RvHJ1b exhibited an enhanced replication efficiency with obviously higher titers at all the time points , particularly in primary PAMs ( Fig . 2C and D ) , suggesting that the ORF1b of HP-PRRSV is related to its growth efficiency in vitro . To further determine the effect of the exchanged genomic regions on the pathogenicity of PRRSV , animal inoculation experiments were performed with 6-week-old piglets and the pathogenicities of the chimeric viruses were compared with their parental backbone viruses , RvJXwn and RvHB-1/3 . 9 . The observation of clinical symptoms showed that only one piglet in RvJH1a-inoculated group had dyspnoea , and the piglets in the RvHJ1a-inoculated group had no obvious clinical symptoms , similar to the RvHB-1/3 . 9-infected group . In contrast , the piglets infected with RvJHSP developed marked clinical symptoms , similar to those in the RvJXwn-inoculated group , including depression , anorexia , lethargy , rubefaction on the skin and in the ears , respiratory distress and shivering; whereas the animals in the RvHJSP-inoculated group presented some moderate signs until the end of experiment . The piglets in the RvJH1b-inoculated group only showed mild coughing and sneezing , and conversely , the piglets inoculated with RvHJ1b exhibited severe respiratory conditions , similar to the RvJXwn-inoculated group , indicating that the substitution of RvJXwn ORF1b by that of RvHB-1/3 . 9 could remarkably reduce the viral pathogenicity and that the replacement of RvHB-1/3 . 9 ORF1b by that of RvJXwn could enhance the viral pathogenicity . No clinical signs were observed in the control group during the experiment period . Rectal temperature measurements indicated that both the RvJH1a- or RvHJ1a-inoculated piglets had lower temperatures than the parental backbone virus groups; in contrast , the piglets inoculated with RvJHSP had a peak temperature ( nearly 41°C ) on day 10 pi and then hovered approximately 40 . 5°C ( Fig . 3A and B ) . The RvHJSP-inoculated group had similar temperatures to the RvHB-1/3 . 9-inoculated group . Nevertheless , the rectal temperatures of the RvJH1b-inoculated piglets were obviously lower than those temperatures of the RvJXwn-inoculated group with a high fever during the experiment , except for a temperature of 41°C on day 16 pi; in contrast , the rectal temperatures of piglets inoculated with RvHJ1b elevated at 24 h pi with a temperature of over 41°C for couple days compared with the RvHB-1/3 . 9 group ( Fig . 3A and B ) . The mortalities of the inoculated animals were recorded . As shown in Fig . 3C and D , only one piglet died in the RvJH1a-inoculated group , and all animals in RvHJ1a- and RvHB-1/3 . 9-inoculated groups survived . However , all piglets in the RvJHSP- and RvJXwn-inoculated groups died , whereas all piglets survived in the RvHJSP-inoculated group during the experiment period . Most importantly , all piglets survived in the RvJH1b-inoculated group , whereas three piglets died in the RvHJ1b-inoculated group within day 17 to 21 pi , showing that RvJH1b has a remarkably reduced virulence , while RvHJ1b exhibits obviously an increased mortality for piglets . The virus loads in the sera of infected piglets were assayed using a microtitration infectivity assay ( Fig . 4 ) . The data showed that both RvJH1a- and RvHJ1a-inoculated piglets had lower virus loads than their parental virus groups , whereas no obvious difference of virus titers in sera was found between the RvJHSP- or RvHJSP-infected group and its parental virus group . Interestingly , the virus titers in the sera of RvJH1b-inoculated group were significantly lower than those titers of RvJXwn-inoculated group from day 3 to 10 pi ( P<0 . 001 ) ; in contrast , the virus titers in the sera of RvHJ1b-inoculated group were statistically higher than those titers of RvHB-1/3 . 9-inoculated group on days 3 , 5 , and 7 pi ( P<0 . 05 , P<0 . 001 ) . Collectively , our above results clearly demonstrate that ORF1b affects the fatal virulence of HP-PRRSV . To further determine which coding region ( s ) within the ORF1b play crucial roles in the increased virulence of HP-PRRSV , we swapped Nsp9 , Nsp10 , Nsp11 and Nsp12; or Nsp9 and Nsp10 simultaneously; or Nsp9 , Nsp10 and Nsp11 simultaneously between the infectious cDNA clones of RvHB-1/3 . 9 and RvJXwn and successfully rescued twelve chimeric viruses . The MARC-145 cells infected with each chimeric virus were shown to be positive for PRRSV N protein using IFA ( Supplemental Fig . S2 ) . Sequencing analyses were also performed to ensure the correctness of the substitutions ( data not shown ) . These rescued viruses were individually designated RvJHn9 , RvJHn10 , RvJHn11 , RvJHn12 , RvJHn9n10 , RvJHn9n10n11 , RvHJn9 , RvHJn10 , RvHJn11 , RvHJn12 , RvHJn9n10 and RvHJn9n10n11 . To characterize and to compare the growth properties between these chimeras and their parental viruses , their growth kinetics were analyzed by infecting MARC-145 cells or primary PAMs with each virus . In MARC-145 cells , as shown in Fig . 5A , RvJHn9 showed obviously lower virus titers than RvJXwn , with significant differences from 60 h to 96 h pi ( P<0 . 05 , P<0 . 01 ) . Likewise , RvJHn10 had lower virus titers than RvJXwn; however , RvJHn11 and RvJHn12 showed similar growth patterns to their parental virus RvJXwn . In contrast , with an individual Nsp-coding region substitution on the RvJXwn backbone , RvJHn9 had a slower growth rate than the other three chimeric viruses . Moreover , both RvJHn9n10 and RvJHn9n10n11 displayed significantly lower virus titers than RvJXwn at some time points ( P<0 . 05 , P<0 . 01 ) ( Fig . 5B ) . As shown in Fig . 5C , compared with the parental virus RvHB-1/3 . 9 , all four rescued viruses with single region substitution showed significantly lower virus titers at many time points ( P<0 . 05 , P<0 . 01 , P<0 . 001 ) , whereas RvHJn9n10 had slightly lower virus titers than RvHB-1/3 . 9 . Additionally , RvHJn9n10n11 had slightly higher virus yields than RvHB-1/3 . 9 at some time points . Moreover , the peak titers of RvHJn9n10 and RvHJn9n10n11 were both higher than that of RvHB-1/3 . 9 , without a significant difference ( Fig . 5D ) . In primary PAMs , the growth kinetics revealed that RvJHn9 and RvJHn10 had lower virus titers than RvJXwn at the majority of time points without significant differences , whereas both RvJHn11 and RvJHn12 presented a similar growth curve to that of RvJXwn . The data also showed that RvJHn9 exhibited a lower virus yield than RvJHn11 and RvJHn12 for all time points and than RvJHn10 within 12 h to 60 h pi ( Fig . 5E ) . Furthermore , compared with RvJXwn , RvJHn9n10 had a lower replication efficiency with a peak titer delayed by 24-h . Similarly , RvJHn9n10n11 showed obviously lower virus titers than RvJXwn at some time points ( Fig . 5F ) . In contrast , RvHJn9 and RvHJn10 showed higher virus titers than RvHB-1/3 . 9 , without significant differences , whereas both RvHJn11 and RvHJn12 had a similar growth curve to that of RvHB-1/3 . 9 ( Fig . 5G ) . Among the four chimeric viruses with the backbone of RvHB-1/3 . 9 , RvHJn9 displayed a higher virus yield . As shown in Fig . 5H , both RvHJn9n10 and RvHJn9n10n11 had obviously higher virus titers than RvHB-1/3 . 9 , with significant differences from 12 h to 60 h pi ( P<0 . 01 , P<0 . 001 ) ; and moreover their peak titers were 10 times higher than that of RvHB-1/3 . 9 . Our above results demonstrated that the substitution of the Nsp9- and/or Nsp10-coding region ( s ) could obviously affect the virus replication efficiency in vitro , particularly when exchanging both of these regions together , whereas the replacement of the Nsp11- or Nsp12-coding region did not have this effect , indicating that Nsp9 and Nsp10 together are closely related to the replication efficiency of HP-PRRSV in vitro . To further analyze whether the Nsp9- and Nsp10-coding region alone or together is related to the increased pathogenicity of HP-PRRSV in vivo , the pathogenicities of the chimeric viruses RvJHn9 , RvJHn10 , RvJHn9n10 , RvJHn9n10n11 , RvHJn9 , RvHJn10 , RvHJn9n10 and RvHJn9n10n11 in piglets were investigated and compared with their parental viruses . The animals were inoculated with each chimeric virus . Rectal temperatures of the inoculated piglets were daily measured . As shown in Fig . 6A , the RvJHn9-infected piglets had a rising body temperature , with an average of 41°C on day 6 pi , whereas the body temperatures of the RvJHn10-infected group began to rise on day 1 pi , and then reached a peak of 41 . 4°C , similar to those temperatures of RvJXwn-infected group with an average of 41°C on day 2 pi . However , compared with the RvJXwn-infected group , the temperatures of the RvJHn9n10-infected group slowly elevated to a peak of 41 . 1°C on day 14 pi , with a extreme delay and then gradually declined . Similarly , the RvJHn9n10n11-infected group slowly reached its peak on day 10 pi and then rapidly decreased . In contrast , the temperatures of the RvHJn9-infected group started to rise earlier after day 2 pi , and reached over 40°C from day 4 to 12 pi , whereas the RvHJn10-infected group had a slightly lower temperature , compared with the RvHB-1/3 . 9-infected group . Importantly , the temperatures of RvHJn9n10-infected group reached over 40°C on day 2 pi , then elevated quickly to a peak of approximately 41°C on day 6 pi , and maintained beyond 40°C , with a significantly greater rising tendency , compared with the RvHB-1/3 . 9- , RvHJn9- or RvHJn10-infected groups during the first half of the experiment . The RvHJn9n10n11-infected group also exhibited higher temperatures with over 41°C from day 8 to 10 pi ( Fig . 6B ) . No body temperature reaction was observed in the control group during the experiment period . The clinical signs of the infected piglets were observed and scored . Compared with the RvJXwn-infected group , the RvJHn9-infected piglets showed moderate and slower progress of disease . All piglets in the RvJHn10-infected group exhibited similar severe signs to the RvJXwn-infected group , whereas the piglets in both RvJHn9n10- and RvJHn9n10n11-infected groups only displayed transient fever and mild respiratory symptoms with significantly lower clinical symptom scores than the RvJXwn- ( P<0 . 001 ) , RvJHn9- or RvJHn10-infected group ( Fig . 6C ) . In contrast , in the RvHJn9-infected group , one piglet died with severe symptom from day 15 pi onward , and others showed moderate anorexia , sneezing and coughing . No obvious clinical signs were observed in the RvHJn10-infected group , similar to the RvHB-1/3 . 9-infected group with occasional sneezing and coughing . In contrast , the piglets in RvHJn9n10- and RvHJn9n10n11-infected groups showed typical symptoms caused by RvJXwn from day 6 pi onwards , with higher clinical symptom scores than the RvHB-1/3 . 9- , the RvHJn9- or RvHJn10-infected group ( Fig . 6D ) . No abnormal clinical symptoms were observed in the control group during the entire experiment . The body weights of the infected and control piglets were recorded weekly to calculate the average daily gain ( ADG ) . As shown in Fig . 6E , the RvJXwn- , RvJHn9- and RvJHn10-infected groups had significantly lower ADG . In contrast , the RvJHn9n10-infected group had an ADG of approximately 0 . 17 kg , which was significantly higher than that of RvJXwn- , RvJHn9- or RvJHn10-infected group ( P<0 . 05 , P<0 . 01 , P<0 . 001 ) . The RvJHn9n10n11-infected group showed a higher ADG of nearly 0 . 26 kg , with significant differences in comparison with other three groups ( P<0 . 05 , P<0 . 01 , P<0 . 001 ) . As shown in Fig . 6F , the RvHJn9-infected group had a lower ADG than RvHB-1/3 . 9-infected group , whereas the RvHJn10-infected group had an ADG of approximately 0 . 30 kg , similar to the RvHB-1/3 . 9-infected group . Moreover , the RvHJn9n10- and RvHJn9n10n11-infected groups showed a remarkably lower ADG compared with RvHB-1/3 . 9- or RvJHn10-infected group . The ADG of the control group was statistically higher than those ADG values of all infected groups ( P<0 . 001 ) , except for RvHB-1/3 . 9 and RvHJn10 groups . The deaths of the animals in each group were recorded . As shown in Fig . 7 , all piglets in the RvJXwn- or RvJHn10-infected group died , although the survival time of the RvJHn10 group was obviously prolonged . Three piglets in the RvJHn9-infected group died , whereas all piglets inoculated with RvJHn9n10 or with RvJHn9n10n11 survived during the experiment . Additionally , no piglets died in the RvHB-1/3 . 9- and RvHJn10-infected groups , whereas one piglet in the RvHJn9-infected group died on day 20 pi . Importantly , three piglets in the RvHJn9n10-infected group died during day 10 to 18 pi , and two in the RvHJn9n10n11-infected group died on days 10 and 17 pi , respectively . Taken together , these results clearly showed that simultaneously replacing the Nsp9- and Nsp10-coding regions of RvJXwn by RvHB-1/3 . 9 could remarkably reduce its fatal virulence for piglets , and in contrast , replacing of both the Nsp9- and Nsp10-coding regions of RvHB-1/3 . 9 by RvJXwn could significantly enhance its virulence for piglets . Necropsies and gross lung lesion examinations were immediately performed once the inoculated piglets died during the experiment . Similar to RvJXwn-infected piglets , all dead piglets in the chimeric virus-infected groups presented severe interstitial pneumonia with extensive and marked pulmonary edema , hemorrhage and consolidation ( Supplemental Fig . S3A ) . Their mean scores of gross lung lesions showed no obvious differences ( Fig . 8A ) . By the end of the experiment , the survived piglets were euthanized and necropsied . Moderate diffuse lung lesions were observed in RvJHn9-infected piglets , and scattered lung lesions were shown in the RvJHn9n10- and RvJHn9n10n11-infected piglets; in contrast , RvHJn9- , RvHJn9n10- or RvHJn9n10n11-infected piglets showed more severe interstitial pneumonia primarily at the cranial , middle lobes than RvHJn10- or RvHB-1/3 . 9-infected piglets ( Supplemental Fig . S3B ) . Meanwhile , RvHJn9n10n11-infected piglets had a statistically higher mean scores of gross lung lesions than RvHB-1/3 . 9-infected piglets ( P<0 . 05 ) ( Fig . 8B ) . The microscopic lesions of lungs were observed , and PRRSV antigens in lungs were examined by immunohistochemistry . Similar to RvJXwn , the lungs of the dead piglets in the chimeric virus-infected groups during the experimental period exhibited severe histopathological changes characterized by the complete disappearance of lung structure , thickening of the interlobular septal , a number of inflammatory cells and necrotic debris infiltration within both the alveolar spaces and bronchioleshared ( Fig . 8C ) . Moreover , immunohistochemical staining showed that their lungs were full of PRRSV-positive signals , which were generally located in the alveolar and septa macrophages around bronchia , bronchiole , and alveolar septa ( Fig . 9A ) . The average histopathological and immunohistochemical scores of those groups showed no obvious differences ( Fig . 8C and 9A ) . In addition , the lungs of survived piglets in the RvJHn9- , RvJHn9n10- or RvJHn9n10n11-infected group displayed a gradually alleviated histopathological change; some slight microscopic lesions of lungs could be observed in RvHJn10- and RvHB-1/3 . 9-infected piglets . In contrast , the piglets exposed to RvHJn9 , RvHJn9n10 or RvHJn9n10n11 showed increasingly severe histopathological changes ( Fig . 8D ) and PRRSV-positive signals closely correlated with the degree of histopathological lesions in lungs ( Fig . 9B ) . The histopathological and immunohistochemical scores of the RvHJn9n10n11-infected group were both significantly higher than those of the RvHB-1/3 . 9-infected group ( P<0 . 05 ) ( Fig . 8D and 9B ) . No gross or microscopic lesions and PRRSV-positive signals were observed in the lungs of control piglets . These results indicated that RvJHn9n10 and RvJHn9n10n11 caused less severe histopathological lesions of lungs than RvJXwn , whereas RvHJn9n10 and RvHJn9n10n11 could remarkably lead to more severe histopathological lesions of lungs than RvHB-1/3 . 9 . Taken together , the above results indicated that the Nsp9- and Nsp10-coding regions of HP-PRRSV emerging in China contribute to its increased pathogenicity and fatal virulence for piglets . The virus loads in the sera of inoculated animals were examined using a microtitration assay . As shown in Fig . 10A , the virus titers in the sera of the RvJHn9-infected group were lower than those titers of the RvJXwn-infected group after day 3 pi , with significant differences on day 10 pi ( P<0 . 05 ) , whereas the viremia of RvJHn10-infected group was similar to those titers of the RvJXwn-infected group . Moreover the virus loads of the RvJHn9n10-infected piglets were significantly lower than those titers of the RvJXwn-infected group ( P<0 . 001 ) and of the RvJHn9- and RvJHn10-infected groups at the majority of time points ( P<0 . 05 , P<0 . 01 , P<0 . 001 ) . Similarly , a remarkable reduction of virus loads in the RvJHn9n10n11-infected group was observed compared with the RvJXwn-infected group ( P<0 . 001 , P<0 . 05 ) . As shown in Fig . 10B , RvHJn9 had a higher level of virus loads than RvHB-1/3 . 9 in vivo from day 3 to 7 pi , with significant differences on day 3 pi ( P<0 . 001 ) . Additionally , the virus load in the sera of RvHJn10-infected piglets was slightly higher than that of RvHB-1/3 . 9 on day 3 pi ( P<0 . 01 ) , but subsequently declined with a lower level on day 14 pi ( P<0 . 001 ) . In contrast , compared with RvHB-1/3 . 9 , the virus loads of RvHJn9n10-infected piglets were obviously higher , with significant differences on days 3 and 5 pi ( P<0 . 01 ) ; likewise , the RvHJn9n10n11-infected group showed higher virus titers with significant differences on days 3 , 5 , and 7 pi ( P<0 . 05 , P<0 . 001 ) . The above data suggested that the exchanging of both Nsp9- and Nsp10-coding regions of RvJXwn by RvHB-1/3 . 9 could remarkably reduce the virus replication efficiency in vivo , and in contrast , the replacing of the corresponding regions of RvHB-1/3 . 9 by RvJXwn could increase the virus replication efficiency in vivo at a higher level . The specific antibodies against the PRRSV N protein in the sera of infected pigs were measured using a commercial IDEXX ELISA kit . As shown in Fig . 10C , only one piglet in the RvJHn9- or RvJHn10-infected group became seroconverted , whereas the RvJXwn-infected group had four piglets seroconverted on day 7 pi . However , the piglets in the RvJHn9n10- or RvJHn9n10n11-infected groups were seropositive until day 10 pi , and meanwhile , their antibody levels were remarkably lower than those levels of the RvJXwn-infected group . In contrast , the RvHJn10- and RvHB-1/3 . 9-infected piglets were all seropositive on day 10 pi , and one piglet in the RvHJn9- , RvHJn9n10- or RvHJn9n10n11-infected groups was seroconverted as early as day 7 pi . Meanwhile , the antibody level of the RvHJn9n10-infected group was significantly higher than that of the RvHB-1/3 . 9-infected group on days 14 and 21 pi ( P<0 . 001 ) ( Fig . 10D ) . The control group remained seronegative during the entire experiment . These data further distinctly demonstrated that the Nsp9- and Nsp10-coding regions of HP-PRRSV together were closely related to the virus replication efficiency in vivo . Highly pathogenic PRRSV has extensively prevailed in China and has resulted in great economic loss to the pig industry since 2006 [50]–[52] . HP-PRRSV still circulates in the field as the dominant virus , which may increase the diversity of PRRSV due to the possibility of viral recombination [63] , [64] . Our previous studies have confirmed that the discontinued 30-amino-acid deletion in the Nsp2-coding region , which is the molecular marker of HP-PRRSV , is not directly related to its virulence [40] . It is necessary to explore which region ( s ) of the virus genome contribute to the pathogenicity and virulence of HP-PRRSV . In the present study , large fragments of the genome were initially swapped between HP-PRRSV RvJXwn and LP-PRRSV RvHB-1/3 . 9 by reverse genetic operation to analyze the possible contributor of pathogenicity . Our data indicated that swapping the 5′UTR+ORF1a between RvJXwn and RvHB-1/3 . 9 could simultaneously impair the replication efficiency and pathogenicity of the chimeric viruses with either backbone . Meanwhile , the ORF1b showed obvious effects on the viral replication efficiency in vitro and in vivo and on the virulence for piglets , whereas the SP-coding region of HP-PRRSV did not contribute to its fatal virulence . To further explore the essential virulence determinants within the ORF1b region , we individually or simultaneously constructed and rescued twelve chimeric viruses with exchanged Nsp-coding regions . The pathogenicity for piglets and the in vitro and in vivo replication characteristics of the chimeric viruses were systematically investigated . The results showed that the replacement of Nsp9- and Nsp10-coding regions of RvJXwn together by RvHB-1/3 . 9 could greatly attenuate the virulence of RvJXwn . In contrast , the substitution of the corresponding regions of RvHB-1/3 . 9 by RvJXwn could remarkably increase the virulence of the chimeric virus . Therefore , our sufficient findings suggest that Nsp9- and Nsp10-coding region together contribute to the increased pathogenicity and fatal virulence of Chinese HP-PRRSV . Two previous studies focused on exploring the virulence-associated genes or regions of PRRSV by using reverse genetic techniques [48] , [49] . Kwon and his colleagues systematically substituted a series of small regions of the highly virulent PRRSV strain FL-12 by corresponding parts from the attenuated vaccine strain PrimePac . Using the reproductive failure model in sows , their findings indicated that the nonstructural proteins , particularly Nsp3–8 , might play more important roles than structural proteins in the vaccine attenuation [49] . Our present findings showed that exchanging the 5′UTR+ORF1a ( containing Nsp1 to Nsp8 ) could remarkably impair the fatal virulence of RvJXwn for piglets , but did not affect the pathogenicity of LP-PRRSV RvHB-1/3 . 9; however , exchanging the ORF1b region could not only enhance the virulence of LP-PRRSV but also reduce the virulence of HP-PRRSV . Although these two regions were both related to the reduction of virulence , we were more concerned regarding the region that contributed to the increased virulence . In our study , giving increased virulence is more important than decreasing virulence . Kwon and his colleagues also showed that the substitution of Nsp9 alone or Nsp10–12 together could impair the in vitro growth of the chimeric viruses , indicating that these nonstructural proteins may relate to the viral replication rate [49] . Our findings also showed that the replacement of Nsp9 of RvJXwn alone by RvHB-1/3 . 9 could obviously reduce the fatal virulence for piglets , although this replacement was less effective compared with the replacement of Nsp9 and Nsp10 together . In another similar study , the authors interchanged 5′UTR+ORF1ab or SP+3′UTR between highly pathogenic MN184 and vaccine strain PRRS MLV . The results of animal experiment demonstrated that both the region individually contributed to the attenuation of MN184 and that the 5-terminal fragment had greater effect on viral replication and lung lesion scores for pigs than the 3-terminal fragment [48] . Likewise , in our study , we found that ORF1a and ORF1b both had significant effects on the virulence attenuation of RvJXwn; however , the effect of the SP-coding region was not obvious , although the survival time of RvJHSP-infected piglets was prolonged , with 100% mortality of the infected piglets . When comparing the sequence of structural proteins among these viruses , HB-1/3 . 9 shares 98 . 13% nucleotide identity with JXwn06 , which is significantly higher than the homology ( 90 . 0% ) between MN184 and MLV . Thus , the structural protein might relate to the attenuation of MLV , but did not contribute to the increased fatal virulence of HP-PRRSV . This finding should be the major concern for the two different outcomes . In addition to analyzing the virulence-associated determining regions of PRRSV by reverse genetic operations , many studies were also conducted by comparing the genomic sequences between attenuated vaccine strains and their parental viruses [42] , [46] , [65] , [66] . However , their results were different or even conflicted with each other due to strain-specific variation . Most of the above-mentioned previous studies analyzed the virulence of PRRSV by considering how these viruses were attenuated via serial passages in the permissive cells; thus , the possibility that the attenuation was caused by the incompatibility or imperfect matching of viral genes or UTR derived from the two different heterologous parental strains rather than caused by the true attenuation in a virulence determinant should be ruled out . This possibility can also be used to explain why both RvJH1a and RvHJ1a in our study were attenuated compared with their parental backbone viruses . In our study , we chose a different model to determine which region ( s ) within the genome contributes to the fatal virulence of HP-PRRSV by using two natural strains with distinct virulence . RvHB-1/3 . 9 as a LP-PRRSV shares 97 . 4% identity of the entire genome with HP-PRRSV RvJXwn; however , their virulence for piglets is completely different . Our previous studies have shown that RvJXwn is fatal for piglets , whereas RvHB-1/3 . 9 is not fatal [40] . Thus , using this model , we can distinguish the nonspecific effects due to foreign sequence insertion or to nucleotide mutation from the actual functional differences between virus strains; therefore , the virulence contribution can be solidly confirmed once any genes or regions from RvJXwn are able to increase the mortality of the infected piglets by the chimeric viruses with the RvHB-1/3 . 9 backbone . As is well known , the pathogenicity of viruses is complicated and is dependent on multiple factors , such as viral propagation ability in vivo , tissue and/or cell tropism , immune escape and immune modification , as well as secondary bacterial infections . Many studies have indicated that the infection of highly virulent PRRSVs , such as Chinese HP-PRRSV and Lena , subtypes 3 strain of genotype 1 , is usually accompanied by longer periods of viremia , increased severity of clinical signs , increased mortality , and significantly higher viral loads in blood and in tissues [40] , [55] , [67]–[69] . Additionally , different virulent PRRSVs may have different tissue and cell tropism . Our recent research found that HP-PRRSV JXwn06 infection exhibited more extensive tissue tropism for pigs than HB-1/3 . 9 [70] . A recent study showed that the highly pathogenic PRRSV Lena could infect more subtypes of macrophages than the early European strain LV did , which may result in more severe pathological changes [56] . Meanwhile , different virulent strains can induce different level of cytokines . Compared with the low virulence strain CH1a , Chinese HP-PRRSV HuN4 could induce a higher level of inflammatory cytokines , which might be closely associated with its marked damage on tissues and on organs [71] . Additionally , some studies showed that different genotypes ( 1 and 2 ) of PRRSV were able to induce different patterns of IL-10 and TNF-a , which , therefore , caused different outcomes of the infection [72]–[74] . Moreover , HP-PRRSV infection is usually accompanied by severe secondary bacterial infections , primarily including Haemophilus parasuis and Streptococcus suis infections [67] . We also observed that some dead piglets presented visible gross lesions , such as pericarditis and fibrinous pneumonia , which were caused by secondary bacterial infections during the experiment; this finding could be an important factor that exacerbates the illness and accelerates the death of the inoculated piglets . In this study , when Nsp9- and Nsp10-coding regions were exchanged together , RvJHn9n10 had significantly lower replication efficiency in vitro and in vivo and lower virulence for piglets than its parental virus , RvJXwn; In contrast , RvHJn9n10 displayed remarkably higher virus titers in vitro and in vivo and higher pathogenicity for piglets than its parental virus , RvHB-1/3 . 9 . Nsp9 , which contains RdRp , is regarded as a crucial motor for viral RNA replication [27] , and Nsp10 , which encodes RNA helicase , is another key enzyme that directly participates in RNA synthesis [28] , [29] . Nsp9 and Nsp10 of EAV are assembled into a membrane-associated viral replication and transcription complex ( RTC ) [75] , [76] , which mediates both the genome replication and synthesis of a nested set of subgenomic ( sg ) mRNAs . Considering that both of these regions encode the essential key enzymes directly participating in genomic replication and transcription , not surprisingly , our results showing that Nsp9 and Nsp10 of Chinese HP-PRRSV were closely related to the increased replication efficiency are in agreement with their functions . Similar findings have been documented in some important human viruses . In influenza virus , the viral RNA polymerase complex has been proven to be related to the high pathogenicity and in vivo replication efficiency , which promotes optimal growth in the lower respiratory tract and respiratory droplet transmission in ferrets [77] , [78] . Some researchers have found that the replication-associated proteins—NS3 ( RNA helicase ) , NS4B or NS5 ( RdRp ) of dengue virus correlated with its in vivo virulence through affecting viral RNA synthesis , and consequently the degree of central nervous system damage [79]–[81] . Moreover , one single mutation in the NS3 of West Nile virus could remarkably increase its virulence , which is likely due to increased viral replication efficiency [82] . Meanwhile , the interaction between viral RdRp and viral RNA helicase is a universal phenomenon in many viruses , which is necessary for stimulating viral replication efficiency [83]–[86] . However , there is little knowledge regarding the crystal structure and function of PRRSV Nsp9 and Nsp10 . Whether these proteins interact directly or indirectly through the help of other viral proteins or host cellular proteins should be further explored . By aligning the amino acid sequences of the Nsp9 and Nsp10 regions between JXwn06 and HB-1/3 . 9 or among other Chinese HP-PRRSV strains , Vietnam HP-PRRSV , Laos HP-PRRSV , and Chinese LP-PRRSV strains , we found that there were 4 amino acids that were different in Nsp9 and 5 amino acids that were different in Nsp10 between JXwn06 and HB-1/3 . 9 . Interestingly , one characteristic , a conserved amino acid difference in Nsp9 or Nsp10 between all HP-PRRSV and the Chinese LP-PRRSV strains was found , namely the amino acid at the position 544 in Nsp9 is alanine ( A ) in LP-PRRSV , whereas this amino acid is threonine ( T ) in HP-PRRSV , and the amino acid at the position 408 in Nsp10 is lysine ( K ) in LP-PRRSV , whereas this amino acid is arginine ( R ) in HP-PRRSV ( Supplemental Fig . S4 ) . Analyzing the possible roles of these two amino acids in the enhanced replication efficiency and virulence mediated by Nsp9 and Nsp10 will be of significance for HP-PRRSV pathogenesis . Based on the high identity and common conserved amino acid difference of Nsp9 and Nsp10 between those HP-PRRSV strains , Nsp9 and Nsp10 may likely be the fatal virulence determinants for other Asia HP-PRRSV , not only for Chinese ones . The European highly pathogenic PRRSV strain , Lena , has been recognized to exhibit enhanced replication efficiency and cause similar clinical conditions to Chinese HP-PRRSV [55] . Thus , investigating the role of Nsp9 and Nsp10 in the increased pathogenicity of Lena is necessary for elucidating the mechanisms associated with the virulence of Lena , although Lena and Chinese HP-PRRSV belong to distinct genotypes of PRRSV , with lower amino acid identity in their Nsp9 and Nsp10 . Our present study provides a way to analyze the molecular mechanism of the increased virulence of Lena . Our present study demonstrated that Nsp9 and Nsp10 together contribute to the fatal virulence of HP-PRRSV emerging in China . Additionally , our findings also provide direct evidence that the high replication efficiency is the critical factor for HP-PRRSV virulence . Moreover , by giving the high replication capacity to LP-PRRSV , its pathogenicity for piglets could be remarkably increased . However , more detailed studies concerning the molecular mechanism affecting PRRSV replication by Nsp9 and Nsp10 are required to better understand the pathogenesis of PRRSV . To the best of our knowledge , our findings are not only the first unambiguous illumination concerning the key virulence determinant of Chinese HP-PRRSV but also help to elucidate the pathogenesis of this virus and to develop new drugs and vaccines against PRRSV infection in the future . In summary , our present study indicates that i ) ORF1b affects the fatal virulence of HP-PRRSV for piglet; ii ) Nsp9 and Nsp10 together are closely related to the replication efficiency of HP-PRRSV in vitro; iii ) Nsp9 and Nsp10 together contribute to the fatal virulence of HP-PRRSV for piglets; iv ) Nsp9 and Nsp10 together affect the HP-PRRSV growth in vivo .	PRRS is a considerable threat to the pig industry worldwide . A large-scale atypical PRRS caused by highly pathogenic PRRSV ( HP-PRRSV ) that emerged in 2006 has resulted in considerable economic loss to Chinese pig production . The disease is characterized by a high body temperature ( 41°C–42°C ) , morbidity and by mortality of the affected pigs . Although the genomic marker , the 30-amino-acid deletion in its Nsp2-coding region has been previously verified to have no relation to its increased pathogenicity , the genomic region ( s ) associated with the fatal virulence of HP-PRRSV remain unclear . A series of chimeric viruses with swapped coding regions between HP- and LP-PRRSV were constructed , and their growth abilities and pathogenicities in piglets were analyzed . Our results demonstrated that Nsp9 and Nsp10 together contribute to the replication efficiency and the fatal virulence of HP-PRRSV for piglets . Our finding is not only the first unambiguous illumination concerning the key virulence determinant of Chinese HP-PRRSV but it also provides a novel insight for understanding the molecular pathogenesis of this virus and for designing new drugs and vaccines against PRRSV infection in the future .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "veterinary", "microbiology", "biology", "and", "life", "sciences", "veterinary", "science" ]	2014	Nsp9 and Nsp10 Contribute to the Fatal Virulence of Highly Pathogenic Porcine Reproductive and Respiratory Syndrome Virus Emerging in China
A national mapping of Schistosoma haematobium was conducted in Sierra Leone before the mass drug administration ( MDA ) with praziquantel . Together with the separate mapping of S . mansoni and soil-transmitted helminths , the national control programme was able to plan the MDA strategies according to the World Health Organization guidelines for preventive chemotherapy for these diseases . A total of 52 sites/schools were selected according to prior knowledge of S . haematobium endemicity taking into account a good spatial coverage within each district , and a total of 2293 children aged 9–14 years were examined . Spatial analysis showed that S . haematobium is heterogeneously distributed in the country with significant spatial clustering in the central and eastern regions of the country , most prevalent in Bo ( 24 . 6% and 8 . 79 eggs/10 ml ) , Koinadugu ( 20 . 4% and 3 . 53 eggs/10 ml ) and Kono ( 25 . 3% and 7 . 91 eggs/10 ml ) districts . By combining this map with the previously reported maps on intestinal schistosomiasis using a simple probabilistic model , the combined schistosomiasis prevalence map highlights the presence of high-risk communities in an extensive area in the northeastern half of the country . By further combining the hookworm prevalence map , the at-risk population of school-age children requiring integrated schistosomiasis/soil-transmitted helminth treatment regimens according to the coendemicity was estimated . The first comprehensive national mapping of urogenital schistosomiasis in Sierra Leone was conducted . Using a new method for calculating the combined prevalence of schistosomiasis using estimates from two separate surveys , we provided a robust coendemicity mapping for overall urogenital and intestinal schistosomiasis . We also produced a coendemicity map of schistosomiasis and hookworm . These coendemicity maps can be used to guide the decision making for MDA strategies in combination with the local knowledge and programme needs . Schistosomiasis or bilharzia is prevalent in 76 countries and territories in tropical and subtropical regions and is estimated to infect over 200 million people worldwide , causing significant morbidity [1] , [2] . The disease is caused by infection with trematodes of the Schistosoma genus . There are three major species which infect humans: Schistosoma haematobium causing urogenital ( formerly known as urinary ) schistosomiasis , and Schistosoma mansoni and Schistosoma japonicum ( the latter in Asia ) causing intestinal schistosomiasis . The geographical distribution of schistosomiasis is dependent on the presence of suitable intermediate host snails in the aquatic environment in the tropics and subtropics , amongst other factors . Hookworm is one of the major soil-transmitted helminthes ( STH ) , infecting 576 million people worldwide and causing anemia and undernutrition particularly in poor rural settings [3] . Both intestinal and urogenital forms of schistosomiasis and hookworm are known to be endemic in Sierra Leone [4] , [5] , [6] , [7] , [8] , [9] , [10] . In 2008 , a national integrated control program against neglected tropical diseases ( NTDs ) including lymphatic filariasis , onchocerciasis , schistosomiasis and STH was initiated with financial support from the United States Agency for International Development ( USAID ) NTD Control Program managed by RTI International and technical support from Helen Keller International . The programme uses the integrated mass drug administration ( MDA ) strategy according to the preventive chemotherapy ( PCT ) guidelines recommended by the World Health Organization ( WHO ) [11] . To facilitate the planning and implementation of MDA activities , a national mapping survey on prevalence and distribution of schistosomiasis and STH was conducted in 2008 and 2009 [9] , [10] . The results showed that S . mansoni and hookworm were widespread in Sierra Leone , with high prevalence of S . mansoni in Kono , Koinadugu , Kailahun , Kenema and Tonkolili districts and with high prevalence of hookworm across the country . Spatial analysis predicted that there was a large cluster of high risk of S . mansoni infection ( prevalence >70% ) in the north and most of the eastern areas of the country and a large cluster of high risk of hookworm infection ( prevalence >70% ) in the north-eastern part of the country [9] . However , in that survey , urogenital schistosomiasis was not properly diagnosed due to the limited human resources . Urogenital schistosomiasis was first reported from Sierra Leone in 1909 [8] , and since then , numerous foci of S . haematobium have been reported , with varying levels of prevalence [4] , [5] , [8] , [12] , [13] , [14] , [15] . In general , prevalence in Eastern province was relatively high and in Northern and Southern provinces was relatively low . To enable the national integrated NTD control programme to fine tune the praziquantel distribution strategy in each district , a further national survey of urogenital schistosomiasis was conducted before praziquantel distribution in 2009–2010 . A number of methods are available for mapping the codistribution of helminth infections [16] . One of these methods is the production of coendemicity maps [17] . In this paper , we aimed to analyze the newly collected urogenital schistosomiasis dataset to provide the first urogenital schistosomiasis distribution map for the country . Additionally , by combining this map with the previously reported maps on intestinal schistosomiasis and STH , we aimed to estimate the at-risk population of school-age children requiring integrated schistosomiasis/STH treatment regimens according to the coendemicity of these diseases , based on the WHO guidelines . The national NTD control programme is managed and implemented by the Ministry of Health and Sanitation , Sierra Leone . The programme undertook a national survey on prevalence of each NTD in order to plan the implementation strategy . Ethical approval for data collection in school children was obtained from the Ethics Committee of the Ministry of Health and Sanitation of Sierra Leone . Upon arrival at the selected schools , the investigating team met with the community teachers association in each school , and explained the nature of the survey . Informed consent was verbally given by guardians/parents and recorded by the team leader . The verbal consent was approved by the Ethics Committee as literacy rates are low in Sierra Leone . Once data were collected , the results were entered into a database and analyzed anonymously . No personal identity can be revealed upon publication . All participants subsequently received treatment from the national programme . Sierra Leone is divided into 12 rural health districts ( each with 7–16 chiefdoms ) plus rural Western Area ( WA ) and urban WA . The survey was carried out in 2009 in six rural districts ( Bo , Bombali , Kenema , Koinadugu , Kono and Tonkolili ) which qualified for mass praziquantel distribution according to S . mansoni mapping [9] , [10] , and in 2010 in seven coastal districts including rural and urban WAs which did not qualify for mass praziquantel distribution according to S . mansoni mapping . Due to the focal nature of S . haematobium according to the historical data and the programme planning need in Sierra Leone , the survey sites were not selected randomly , but according to the prior local knowledge and local ecological environment in each district , where the villages were thought most likely to have schistosomiasis . To ensure a good spatial coverage of the survey sites within a district , one site was selected from each chiefdom where S . haematobium was suspected to be likely endemic . The number of sites surveyed in each district is shown in Table 1 . The survey was conducted in primary schools . Within each school , 30–50 children aged 9–14 years were enrolled , balancing for sex . The sample size for each survey site was chosen according to the recommendations in the best practice paper [18] , and the WHO guidelines [19] . A total of 52 sites/schools were surveyed and 2293 children ( 1234 boys and 1059 girls ) were examined . The sampling method for S . mansoni and STH mapping sites in 2008 has been described previously [9] . Briefly , the survey sites ( schools ) were selected according to administrative districts ( four schools per district ) using a two-staged random sampling method to avoid two schools being selected from the same chiefdom to ensure a relatively even geographical coverage throughout the country . In each district four chiefdoms were first randomly selected . Within each selected chiefdom , one primary school was randomly selected . In total , 53 schools were selected for survey throughout the country . Approximately 100 children aged 5 to 16 years per school ( range: 36–134 ) were examined . One urine sample was collected from each of 2293 children around midday . Each sample container was labeled with an identification number . For examination , volume of urine samples was measured and urine containers were centrifuged for five minutes [20] . The sediment was transferred onto a glass slide and covered with a cover slip . These were examined under a light microscope , and the number of S . haematobium eggs was recorded and intensity of infection expressed as number of eggs per 10 ml of urine ( eggs/10 ml ) . The data collection for S . mansoni and STHs using the standard Kato-Katz method has already been described in the previous publication [9] . Survey results were entered into Microsoft Excel . Prevalence of infection and corresponding differences between ages and sex were estimated taking into account the clustered nature of the sampling , using the village/school as a primary sampling unit and including adjustments for the probability of sampling and finite population corrections for sampling without replacement in the Stata/SE 10 . 0 statistical package ( StataCorp , College Station , Texas , USA ) . The coordinates of each sample site were recorded using hand-held global positioning system ( GPS ) devices ( available upon request ) . Prevalence at each location was plotted in a geographical information system ( GIS ) ( ArcGIS version 10 . 0 , ESRI , Redlands , CA ) . Electronic data for land surface temperature ( LST ) and normalised difference vegetation index ( NDVI ) for a 5 km×5 km grid cell resolution were obtained from the National Oceanographic and Atmospheric Administration's ( NOAA ) Advanced Very High Radiometer ( AVHRR; see Hay et al . [21] for details on these datasets ) and the location of large perennial inland water bodies ( PIWB ) was obtained from the Food and Agriculture Organization of the United Nations ( http://www . fao . org/geonetwork/srv/en/main . home ) and the distance to PIWB was extracted for each survey location in the GIS . A 5 km resolution population surface derived from the Global Rural-Urban Mapping Project ( GRUMP ) beta product was obtained from the Center for International Earth Science Information Network ( CIESIN ) of the Earth Institute at Columbia University ( http://sedac . ciesin . columbia . edu/gpw/global . jsp ) . Elevation data with a 5 km×5 km grid resolution , generated by a digital elevation model ( DEM ) from the Shuttle Radar Topography Mission ( SRTM ) , were obtained from the Global Land Cover Facility ( http://glcf . umiacs . umd . edu/index . shtml ) . All environmental datasets were linked to survey locations and values at each survey location were extracted in the GIS . The analysis was carried out in two phases ( Figure 1 ) : in the first phase we aimed to quantify the combined schistosomiasis prevalence . We developed a predictive map of S . haematobium prevalence for Sierra Leone using model-based geostatistics . The resulting S . haematobium predictive map was combined with an existing predictive map of S . mansoni prevalence using a probabilistic approach ( see below ) to obtain a combined urogenital and intestinal schistosomiasis map . This map was then categorized based on the prevalence thresholds in the WHO guidelines for praziquantel distribution: low-risk communities for schistosomiasis were defined as those in areas that had combined prevalence of both infections <10% , moderate-risk communities in areas having combined prevalence of both infections 10–50% and high-risk communities in areas having combined prevalence of both infections >50% . This allowed the identification of areas in Sierra Leone by schistosomiasis risk level . In the second phase of the analysis we combined risk maps of schistosomiasis and major STH ( hookworm ) to quantify the population requiring the different WHO recommended treatment regimens for each parasite . We overlaid the combined schistosomiasis map generated in Phase 1 with an existing map of hookworm prevalence to obtain a schistosomiasis/hookworm co-endemicity map . Hookworm was chosen for this analysis because in a recent nationwide parasitological survey , hookworm was the STH with the highest prevalence ( 38 . 5% ) in the country [9] , [10] , and the prevalence of the other major STH ( Ascaris lumbricoides , 6 . 6% and Trichuris trichiura , 1 . 8% ) was too low to warrant risk mapping . The hookworm prevalence map was categorized into the WHO prevalence thresholds that define risk levels for STH infection and appropriate albendazole treatment regimens ( 20–50% and >50% ) . The resulting coendemicity map was then overlaid with the GRUMP population map in the GIS and the population size in areas belonging to a given coendemicity class was then calculated in order to obtain the numbers of individuals at each risk level of both infections . The initial candidate set of predictor variables included population density , LST , NDVI , PIWB and elevation . Fixed-effects binomial logistic regression models of prevalence of S . haematobium infection were developed in a frequentist statistical software package ( Stata version 10 . 1 , Stata Corporation , College Station , TX ) . In the preliminary , non-spatial multivariable model , elevation was not found to be significantly associated with S . haematobium infection risk ( Wald's P>0 . 2 ) and this variable was excluded from further analysis . A quadratic association between LST and prevalence of infection was assessed and was not found to improve model fit using the Akaike's Information Criterion . Residual spatial dependence was investigated using semivariograms using the package geoR of the statistical software R . We developed the model-based geostatistical spatial prediction model [22] for S . haematobium using the Bayesian statistical software , WinBUGS version 1 . 4 ( Medical Research Council Biostatistics Unit , Cambridge , United Kingdom and Imperial College London , London , United Kingdom ) . Several models were tested and all had the covariates plus a geostatistical random effect , in which spatial autocorrelation between locations was modeled using an exponentially decaying autocorrelation function . Statistical notation of Bayesian geostatistical models , spatial interpolation and model validation procedures are presented in an additional file ( Text S1 ) . We used predicted prevalence estimates of S . haematobium and the predicted S . mansoni prevalence estimates from Koroma et al [9] to derive a combined urogenital and intestinal schistosomiasis prevalence estimate . The combined prevalence was calculated using a simple probabilistic model , incorporating a small correcting factor to allow for non-independence of schistosome species following the approach of de Silva and Hall [23] . In brief , when assuming that the probability of infection with one schistosome species is independent of another , the predicted combined probability of having at least one schistosome infection is the simple probability law for the union between two probabilities: ( 1 ) where is the combined urinary and intestinal schistosomiasis prevalence , h is the urinary schistosomiasis prevalence and m is the intestinal schistosomiasis prevalence . This equation was implemented in the GIS and multiplied by a correction factor due to non-independence between both schistosome surveys . Without this correction factor , the predicted combined prevalence of schistosomiasis would be an overestimate . The correction factor was estimated using data from 67 schools with urinary schistosomiasis and intestinal schistosomiasis coinfections in Burkina Faso , Ghana , Mali and Niger , collected between 2007–2008 with support from the Schistosomiasis Control Initiative ( SCI ) [16] , [24] , [25] . Using these data we plotted the difference between predicted and observed combined prevalence against the observed combined prevalence in each school . We found the association to be highly non-linear , negating the use of a simple linear equation to describe the correction factor . We then fitted non-linear parametric functions using a function finder interface freely available on the internet ( www . zunzun . com ) . This online resource allows curve fit to non-linear observational and experimental data by comparing and estimating fit statistics to a library of over 500 non-linear functions . The Python code for curve fitting is available on the Google code repository http://code . google . com/p/pythonequations/ . We used the predicted combined prevalence of schistosomiasis map and the predicted hookworm prevalence estimates from Koroma et al [9] to derive a schistosomiasis/hookworm coendemicity map . This map was overlaid in the GIS by a 2011 GRUMP population map for children aged between 5–15 years old in Sierra Leone , projected from a 2009 GRUMP map to obtain the number of school-age children with schistosomiasis/hookworm coinfections . This projection assumed a population growth rate 2005–2011 of 2 . 60% and proportion of 5–15 years old of 26 . 5% for 2011 ( http://esa . un . org/unpd/wpp/index . htm ) , that was constant across the country . Point prevalence of S . haematobium infection from each survey site is shown in Figure 2 . Across 52 sites surveyed , S . haematobium infection was found in 30 sites , mainly in Bo and Kono districts . The median prevalence was 2% ( inter-quartile range: 0–18 . 6% and minimum-maximum range: 0–56 . 3% ) in all sites and 17 . 7% ( inter-quartile range: 6 . 2–30 . 5% and minimum-maximum range: 2–56 . 3% ) in S . haematobium-positive only sites . Arithmetic mean intensity of infection including all children examined was 3 . 98 eggs/10 ml urine ( 95% CI: 2 . 73–5 . 22 eggs/10 ml ) . There was no significant difference in either prevalence or intensity of infection between boys and girls or between ages ( p>0 . 05 , details not shown ) . Table 1 summarizes the prevalence and intensity of infection in each district surveyed . S . haematobium was found mainly in the northeast half of the country , with a relatively higher level of endemicity in Bo ( 24 . 6% and 8 . 79 eggs/10 ml ) , Koinadugu ( 20 . 4% and 3 . 53 eggs/10 ml ) and Kono ( 25 . 3% and 7 . 91 eggs/10 ml ) districts . Little S . haematobium infection was found in the southern and western coastal districts . In the non-spatial model of S . haematobium infection , the semivariogram of model residuals exhibited strong spatial variation unaccounted for by the variables included in the model ( Figure 3 ) , justifying a model-based geostatistical approach . Spatial model results ( Table 2 ) indicated that there was no clear association between prevalence of S . haematobium and sex , age , LST , NDVI , PIWB or population density . The rate of decay of spatial autocorrelation [Phi ( φ ) ] was 4 . 20 . This indicates that , after accounting for the effect of covariates , the radii of the clusters were approximately 79 km ( note , φ is measured in decimal degrees and 3/φ determines the cluster size; one decimal degree is approximately 111 km at the equator ) . Variance of the spatial random effect ( σ2 ) was 6 . 41 , indicating a strong tendency for spatial clustering . The spatial prediction map showed a large area of moderate to high risk of S . haematobium infection ( prevalence >10% ) in the northeast two-thirds of the country , with clusters of significant risk of infection ( prevalence >30% ) in a region covering the northeast corner and an area spanning across Bo , Kenema and Kono districts of the country ( Figure 4 ) . The predicted prevalence was highest ( prevalence >50% ) in an area shared by the districts of Bo and Kenema and in a small cluster in the north of Koinadugu district . The model showed an acceptable predictive ability with an AUC of 0 . 78 ( 95%CI: 0 . 72–0 . 83 ) . The relationship between the difference between predicted and observed combined prevalence ( dhm ) against the observed combined prevalence ( ohm ) in each school was highly nonlinear ( Figure 5 ) . The best fitting function to the distribution was of the form:which indicates that the overestimation increased by 0 . 09 for every 10% increase in observed combined prevalence on the natural logarithmic scale . The combined schistosomiasis map for Sierra Leone highlights the presence of high risk communities in an extensive area in the northeastern half of the country ( Figure 6 ) . The transition between schistosomiasis risk areas is made in a northeast to southwest direction in the central districts of Bombali , Tonkolili , Bo and Kenema . Low risk areas ( <10% ) occupy most of the coastal and central districts of Sierra Leone . We predicted an extensive geographical overlap between the risk of schistosomiasis and hookworm . Based on the developed schistosomiasis/hookworm coendemicity map we constructed an integrated treatment map ( Figure 7 ) which shows that most communities in the district of Koinadugu and smaller areas in the districts of Kailahun and Kenema will require once annually treatment of praziquantel and twice annually treatment of albendazole . In addition , school-age children in the districts of Moyamba , Bonthe , Pujehun , and few communities of southern Kenema will require twice a year treatment with albendazole . Most of the coastal and central areas of Sierra Leone will require albendazole once a year . We estimated a total of 1 , 845 , 437 school-age children in Sierra Leone are in need of anthelminthic treatment of which 825 , 871 are in need of annual treatment of praziquantel; 56 . 5% ( 466 , 575 ) of those will also need albendazole once a year ( Table S1 ) . We also estimated that 302 , 814 school-age children will require praziquantel once every two years , and 61 . 3% ( 185 , 713 ) of those will also require a dose of albendazole once each year . Finally , we estimate that 716 , 752 school-age children in Sierra Leone will need praziquantel twice during their primary schooling age , and the majority of these ( 68 . 3% ) will also need albendazole once each year . This is the first comprehensive national mapping of urogenital schistosomiasis in Sierra Leone . The results were broadly in line with the previous data which showed that S . haematobium is heterogeneously distributed in the country with significant spatial clustering in the central and eastern regions of the country [8] . The current results confirmed that the population , particularly children , in these parts of the country is not only at risk of S . mansoni infection [9] , [10] but also S . haematobium infection , further justifying the MDA for schistosomiasis in these seven endemic districts . A collection of historical data showed that both urogenital and intestinal schistosomiasis were endemic in overlapping regions in Sierra Leone , but the former was more widely distributed with higher prevalence than the latter [7] , [8] . There was an indication in the 1980s that S . mansoni was spreading in the country because of cross-border population movement and creation of snail habitats due to alluvial mining activities [7] , [12] . The present results showed that , although both species are still endemic in the overlapping areas , S . haematobium has become a much less dominant species than S . mansoni [9] , [10] , and in fact it seems that S . haematobium may have been in decline . It is not clear why this shift in dominance has occurred in the last decades . Studies have shown that male S . haematobium worms are more dominant when competing with male S . mansoni worms when pairing with females in mixed infections [26] , [27] , but this contradicts the current findings . The switch may have been due to ecological reasons ( e . g . snail habitats ) rather than biological interactions between two species . Swamp rice farming was a major factor in the dramatic increase of schistosomiasis in the neighboring country Liberia [28] , and such farming was encouraged in Sierra Leone but was found not to be spreading S . mansoni infection [7] . Mass human population movement in the rural districts may have led to the change in transmission dynamics and pattern causing the switch in species dominance . Mapping of a disease distribution is a key step in planning an integrated national NTD control program . Schistosomiasis is a focal disease , with risk being closely related to the distance to the water sources where the intermediate host snails thrive [29] . Given the nature of the distribution , mapping of schistosomiasis has always attracted discussion in the current integrated control programs . Different strategies have been used in different countries: large scale surveys through as many schools as possible using Lot Quality Assurance Sampling method [30] , [31] , or stratified sampling surveys in selected schools using a geostatistical design and spatial interpolation [32] , [33] have been proposed . Recent comparisons have showed that Lot Quality Assurance Sampling performs better than geostatistical sampling in correctly classifying schools , but at a higher cost per high prevalence school correctly classified [34] . It is always a balancing act between the program needs and the financial resources available when deciding the strategy for schistosomiasis mapping and how many sites to be surveyed . In Sierra Leone , original mapping surveys were designed based on the previous WHO recommendations [9] , [19] . This proved to be insufficient for decision making at sub-district level , therefore , further surveys were conducted as described previously [10] , and in this paper . By combining all the data obtained throughout the country and overlaying the maps from different species , we were able to provide the most comprehensive understanding of distribution of schistosomiasis and hookworm in Sierra Leone and therefore optimal strategies for targeting of MDA . It is noted that the current coendemicity map for schistosomiasis was constructed using data from separate surveys for urogenital and intestinal schistosomiasis . Due to practical reasons , the surveys for the two species were conducted separately . To avoid overestimation or underestimation of the combined schistosomiasis prevalence by simple overlaying of different endemicity maps , we calculated the combined prevalence using a simple probabilistic model [23] . This model assumes independence between infections and this study indicates that in the case of schistosomiasis , this assumption would grossly overestimate the treatment needs by almost a million school-age children . To address this , the report has presented a new method for calculating the combined prevalence of schistosomiasis using estimates from two separate surveys which accounts for the highly non-linear relationship between observed and predicted combined prevalence of schistosomiasis and is therefore a more robust extension of the coendemicity mapping approach presented in an earlier study . Given the situation where the overall prevalence of schistosomiasis cannot be obtained for each community ( which is typical for NTD control programmes in Sub-Saharan Africa ) , such coendemicity maps would provide a very useful tool to inform decisions for planning national MDA . There are certain limitations in this study and the predicted coendemicity map . For the estimated number of school-age children , the population map used was based on the projected population . There may be a significant underestimate of the current population in Sierra Leone as the country underwent a significant population growth after the civil war and this was evident during MDA in the national NTD programme compared with the 2004 national census [35] , [36] . Secondly , large migrations of internally displaced persons as a result of the civil war during 1991–2002 occurred initially from the east and then the north moving further towards the south and west . Many of these internally displaced persons have remained in the WA and other coastal districts post-war . The schistosomiasis cases identified in these surveys , particularly in the WA , may have been imported from more highly schistosomiasis-endemic districts . All children with S . haematobium infection in coastal districts in this survey were confirmed to be from internally displaced families ( Hodges , personal observation ) . Indeed many of these internally displaced children are known to return to more highly prevalent districts during vacations to stay with their extended family there and then return to schooling in the low-prevalent coastal districts . Therefore , schistosomiasis endemicity in the WA and the coastal districts may have been overestimated , as there was no schistosomiasis or evidence of snails in these districts according to the historical data . Thirdly , the non-random selection of sites for S . haematobium surveys , which was based instead on historical data and local knowledge , may have led to overestimation of overall level of S . haematobium endemicity in the country . However , building on the previous S . mansoni and STH mapping [9] , [10] , and due to the specific nature of focal distribution of the disease , such purposeful and non-random sampling provided the national program with practical tools for MDA planning . From the programmatic point of view , the current co-endemicity map should be used in conjunction with not only the local knowledge as described above , but also the overall programme needs when planning MDA in these districts . Implementation of MDA for schistosomaisis and STH in Sierra Leone is performed by different government Ministries ( health and education ) , through different platforms ( community-based and school-based ) , with different donors and different budget time-lines , functioning within different implementation units ( chiefdoms versus districts ) , and overlapping with other NTD programs such as MDA with ivermectin and albendazole for lymphatic filariasis and/or onchocerciasis . In the context of integrated NTD control , planning of MDA for schistosomiasis and STH as indicated in the co-endemicity map needs to be coordinated to avoid repetition and to increase cost-efficiency . In conclusion , the first comprehensive national mapping of urogenital schistosomiasis in Sierra Leone was conducted which showed that S . haematobium is heterogeneously distributed in the country with significant spatial clustering in the central and eastern regions of the country . Using a new method for calculating the combined prevalence of schistosomiasis using estimates from two separate surveys , we provided a robust coendemicity mapping for overall urogenital and intestinal schistosomiasis . We also produced a coendemicity map of schistosomiasis and hookworm . These coendemicity maps can be used to guide the decision making for MDA strategies in combination with the local knowledge and programme needs .	Two forms of schistosomiasis or bilharzia ( intestinal and urogenital ) exist in Sierra Leone . The main control strategy for this disease currently is through mass drug administration ( MDA ) according to the World Health Organization recommended anthelminthic chemotherapy guidelines , and others include snail control , behavior change , and safe water , sanitation and hygiene . Survey on distribution and prevalence of the disease is vital to the planning of MDA in each district . The distribution of intestinal schistosomiasis in the country has been reported previously . The current national survey showed that urogenital schistosomiasis has a specific focal distribution particularly in the central and eastern regions of the country , most prevalent in Bo ( 24 . 6% ) , Koinadugu ( 20 . 4% ) and Kono ( 25 . 3% ) districts . Using a simple probabilistic model , this map was combined with the previously reported maps on intestinal schistosomiasis and the combined schistosomiasis prevalence was estimated . The combined schistosomiasis map highlights the presence of high-risk communities in an extensive area in the northeastern half of the country , which provides a tool for planning the national MDA activities .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "medicine", "disease", "mapping", "infectious", "diseases", "public", "health", "and", "epidemiology", "epidemiology", "infectious", "disease", "epidemiology", "neglected", "tropical", "diseases", "infectious", "disease", "control", "parasitic", "diseases" ]	2012	Combined Spatial Prediction of Schistosomiasis and Soil-Transmitted Helminthiasis in Sierra Leone: A Tool for Integrated Disease Control
Comparative genomics revealed in the last decade a scenario of rampant horizontal gene transfer ( HGT ) among prokaryotes , but for fungi a clearly dominant pattern of vertical inheritance still stands , punctuated however by an increasing number of exceptions . In the present work , we studied the phylogenetic distribution and pattern of inheritance of a fungal gene encoding a fructose transporter ( FSY1 ) with unique substrate selectivity . 109 FSY1 homologues were identified in two sub-phyla of the Ascomycota , in a survey that included 241 available fungal genomes . At least 10 independent inter-species instances of horizontal gene transfer ( HGT ) involving FSY1 were identified , supported by strong phylogenetic evidence and synteny analyses . The acquisition of FSY1 through HGT was sometimes suggestive of xenolog gene displacement , but several cases of pseudoparalogy were also uncovered . Moreover , evidence was found for successive HGT events , possibly including those responsible for transmission of the gene among yeast lineages . These occurrences do not seem to be driven by functional diversification of the Fsy1 proteins because Fsy1 homologues from widely distant lineages , including at least one acquired by HGT , appear to have similar biochemical properties . In summary , retracing the evolutionary path of the FSY1 gene brought to light an unparalleled number of independent HGT events involving a single fungal gene . We propose that the turbulent evolutionary history of the gene may be linked to the unique biochemical properties of the encoded transporter , whose predictable effect on fitness may be highly variable . In general , our results support the most recent views suggesting that inter-species HGT may have contributed much more substantially to shape fungal genomes than heretofore assumed . Gene gain and loss are deemed to be important mechanisms underlying adaptation to different lifestyles across all domains of life . For many prokaryotes this reflects in the relative sizes of the “core genome” , shared by all individuals of a species , and the so called accessory genome that equips the cells for survival in specific environments and can represent as much as 60% of the total genome [1] . This plasticity is generally thought to be related to the relative ease with which prokaryotes are able to discard genes that are not required , as well as to the diversity and effectiveness of mechanisms mediating gene acquisition [2] . Whereas it is commonly accepted that vertical descent with modification as well as gene duplication followed by divergence of the resulting paralogous genes are paramount for the expansion of gene diversity in prokaryotes , horizontal gene transfer ( HGT ) between species follows closely in importance , since it seems to have been a very frequent source of genetic novelty throughout the evolution of both Archaea and Bacteria [2] , [3] . Many HGT events became evident from the comparison of large numbers of genomes because this allowed the clarification of the phylogenetic relationship between genes that appeared to be paralogous upon examination of a single genome . Indeed , some paralog pairs turned out on more detailed phylogenetic examination to be composed of a “resident” gene and a homolog acquired from a different species , constituting instances of so called “pseudoparalogy” [2] . In other cases , the “resident” gene seemed to have been replaced by a homolog acquired from a different species , an event dubbed “xenolog gene displacement” [2] . Compared with bacteria , nuclear gene content in eukaryotes is generally considered less variable . As possible reasons for this , it has been suggested that meiosis may constitute a limitation to innovation involving large portions of the genome , because it imposes the requirement for pairing between homologous chromosomes [1] . Other reason could be the narrower scope of mechanisms available to eukaryotic cells for the incorporation of DNA from the outside and the necessity that genetic alterations affect the germ line . On the other hand , typical eukaryotic mechanisms like phagocytosis and endosymbiosis seem to have facilitated the acquisition of nuclear genes by protist lineages [4] , [5] . As eukaryotic organisms , fungi share some of these limitations but have also distinctive characteristics . Their ability to propagate asexually and the absence of distinction between soma and germ line would be expected to be more permissive with respect to genome changes . In fact , and contrary to earlier assumptions , the large amount of genomic information available for fungi has brought to light important variability within and between closely related species [6] , which is often associated to particular chromosomal locations , like subtelomeric regions [7] . Even so , gene repertoires seem to be much more similar among closely related fungal species than observed for prokaryotes [8] , [9] . In line with this , HGT is still considered to be generally infrequent , although several well-supported events of acquisition of genes from bacteria [10] , [11] and transfers among fungal species have been reported and are contributing to change this view [12]–[15] . While currently available information still falls short of supporting a major role for HGT in the dynamics of fungal genomes , some functional categories of genes stand out as being more prone to be horizontally transferred , as seems to be the case for gene clusters involved in nutrient assimilation [16] , the production of secondary metabolites [17] , [18] and also for membrane transporters [19] , [20] . In fact , evolution of the latter functional class of genes is uncommonly dynamic , as the incidence of duplications and other copy number changing events seems to be particularly high . [21]–[23] . The evolutionary plasticity of transporter genes is possibly linked to the fact that a single gene may have a dramatic ( positive ) effect on organismal fitness , for example by playing an important role in detoxification processes [24] or by providing a decisive competitive edge in the struggle for nutrients [20] , [25] , [26] . The latter is particularly crucial for fungi because , unlike many other eukaryotic microbes , they lost the capacity for phagotrophy and thus rely completely on osmotrophic feeding for growth [20] . In the model yeast Saccharomyces cerevisiae , evolution of the hexose transporter HXT gene family , which belongs to the Major Facilitator Superfamily [27] was studied in detail and a link between expansion of the family and the extant ability of the species to rapidly ferment glucose and fructose was proposed [22] . In this species , hexose transport relies entirely on the HXT transporter gene family that also includes a galactose transporter and two glucose sensors [22] . However , two closely related species , S . uvarum and S . eubayanus ( as well as the derived hybrid and domesticated species S . bayanus and S . pastorianus ) possess an additional gene , FSY1 , encoding a protein specialized in high affinity , specific fructose uptake [28] , [29] . Unlike the Hxt proteins , Fsy1 is a proton symporter that accepts fructose and sorbose as substrates but is unable to transport glucose , a highly unusual feature among hexose transporters [30] . The affinity of Fsy1 for fructose is at least one order of magnitude higher than that of the Hxt transporters and may therefore confer substantial advantage in environments with low fructose concentrations [28] , [31] . However , Fsy1 was also recently shown to operate under certain circumstances in a mode that co-transports more than one proton with one fructose molecule , which is energetically very costly and unprecedented in fungal hexose transporters [31] . We recently proposed that this departure from the normal 1∶1 stoichiometry might represent a “defective” mode of operation of the Fsy1 transporter , observed only when it functions at high glycolytic fluxes [31] . In two of the species where the FSY1 gene occurs naturally ( S . bayanus and S . pastorianus ) , it was shown to be stringently repressed by high fructose concentrations , which normally concur with a high glycolytic flux [32] . Thus , in the presence of abundant fructose , strong repression of the gene prevents Fsy1 from functioning as a fructose driven ATP consuming device with little or no advantage for the cell because fructose transport via the Hxt proteins is effective under these conditions . This constitutes therefore an instance of a single gene that seems to be able to confer on its own appreciable fitness advantages ( fructose scavenging capacity ) as well as disadvantages ( serious deleterious impact on the cell energy metabolism ) , depending on the environmental conditions and genetic background . In addition to the gene first isolated from S . pastorianus [28] , two Fsy1 homologues were functionally characterized in the yeast Kluyveromyces lactis [33] and the filamentous fungus Botryotinia fuckeliana ( Botrytis cinerea ) [34] . Surprisingly however , the Fsy1 homolog most recently characterized originates from S . cerevisiae wine strain EC 1118 and is located within a 17 kb fragment that was horizontally transferred from an unidentified yeast lineage [14] , [35] . Efficient fructose metabolism is important in the wine industry because lack thereof is thought to result in stuck or sluggish fermentations with important economic repercussions [36] . However , a clear link between advantageous properties of wine strains and the presence of the FSY1 gene remains to be established . These findings concerning an HGT event involving Fsy1 , together with the discovery of higher proton∶fructose stoichiometries during Fsy1 operation at high glycolytic fluxes , spurred us to investigate in detail the evolutionary history of the FSY1 gene , in order to uncover its origin and distribution in fungi and to identify lineage specific losses and putative additional events of HGT . Our hypothesis was that the potentially ambiguous effect of Fsy1 on organismal fitness might affect the pattern of inheritance of the gene . For example , it is conceivable that the gene might be rapidly lost if the organism thrives in environments with high fructose concentrations that can be easily taken up by Hxt-like facilitators without energy expenditure . The fitness disadvantage associated with the presence of the FSY1 gene under these circumstances would be even more pronounced if stringent transcriptional repression is concomitantly relieved , allowing expression of the gene at high fructose concentrations , when operation of Fsy1 dissipates at least double the amount of ATP [31] . On the other hand , in environmental conditions with low fructose concentrations and scarcity of other carbon sources , the fructose scavenging capabilities of Fsy1 are expected to have a clear positive effect on fitness . In line with this hypothesis , we found a highly dynamic pattern of FSY1 gene losses and independent acquisitions by HGT , whose frequency is , to the best of our knowledge , unparalleled by any other single gene in fungi . A total of 241 available fungal genomes were surveyed for the presence of Fsy1 homologues using BLAST searches ( Table S1 ) . Although it harbors the canonical sugar transporter signature sequences [27] , the Fsy1 protein can be clearly distinguished from other fungal sugar transporters , which facilitated the identification of 109 Fsy1 homologs used to construct the phylogenetic tree shown in Figure 1 . An E-value of 1e-80 was found to be appropriate to distinguish Fsy1 homologs from other transporters , as shown by the comprehensive phylogenetic tree in Figure S1 . In most cases , microsynteny conservation in the chromosomal regions surrounding the FSY1 gene consubstantiated the orthologous relationship between genes of closely related species identified through BLAST searches ( Figures 2 and S2 ) . Interestingly , in a phylogenetic tree including all the Fsy1 homologues identified ( Figure 1 ) , the proteins encoded by species belonging to the Pezizomycotina ( filamentous fungi ) and Saccharomycotina ( yeasts ) do not segregate strictly according to the sub-phyla to which they belong . This prompted us to compare in more detail the topology of the Fsy1 tree with that of the species phylogenetic tree and phylogenetic network depicted in Figure 3 and S3 , respectively . The tree and network are based on the concatenated amino acid sequences of six RNA Polymerase subunits previously used in a study with a similar phylogenetic scope [37] and are globally in good agreement with the accepted topology for the different groups within the Ascomycota [38]–[41] . Comparison of the topologies of the trees in Figures 1 and 3 brings to light several clear inconsistencies between the two . These inconsistencies are found among the Saccharomycotina and Pezizomycotina Fsy1 homologs , but most notably among Aspergillus species ( Figures 1 and 3 ) . Hence , a global analysis of the phylogenies suggests that several Fsy1 homologs have had disparate evolutionary ancestries , which cannot be reconciled with a pattern of transmission of the FSY1 gene that involves solely vertical inheritance . FSY1 seems to be absent in basal fungal lineages ( namely Microsporidia , Mucoromycotina , Blastocladiomycota , Chytridiomycota ) and in Basidiomycota ( see Table S1 for a complete list of the species surveyed ) . On the contrary , the gene is very common in the Ascomycota with the exception of the earliest derived sub-phylum Taphrinomycotina , where none of the four genomes examined was found to encode a Fsy1 homologue ( Figure 3 ) . In the remaining sub-phyla , Pezizomycotina and Saccharomycotina , FSY1 distribution is patchy , punctuated by losses common to entire lineages ( e . g . the entire order Onygenales ) or limited to a few species within an otherwise Fsy1 harboring clade ( e . g . the order Capnodiales ) . Using the species tree depicted in Figure 3 , we employed an ancestral state reconstruction method to estimate the likelihood of FSY1 being present in ancestors of extant lineages represented by the various internal nodes in the tree . This analysis suggests that the gene may have originated in the Pezizomycotina , being later horizontally transferred to the Saccharomycotina after the divergence of several basal yeast lineages ( Figure 3 ) . Nearly half of the Pezizomycotina species lacking a FSY1 gene belong to the order Onygenales that includes dimorphic human pathogenic fungi such as Paracoccidioides brasiliensis [42] . In fact , none of the genomes from species in the Onygenales was found to encode Fsy1 homologues , while in the sister clade ( Eurotiales ) the reverse situation is observed and the most recent common ancestor ( MRCA ) of both clades is predicted to have possessed the gene ( Figure 3 ) . The MRCAs of the orders Hysteriales and Glomerellales also seem to have lost the gene , although in these cases only a few species were examined in each order . All the remaining orders exhibit at least one FSY1 gene loss with the exception of the Hypocreales and the Pleosporales in which all the genomes examined encode a Fsy1 homologue ( Figure 3 ) . The phylogeny of Pezizomycotina Fsy1 homologues shows in addition that some species are devoid of a cognate FSY1 gene but acquired a FSY1 gene from another lineage , seemingly by HGT . In addition to several cases in the Eurotiales described in detail in the next section , this was observed for two other species . As shown by the species and Fsy1 phylogenies in Figure 1 and Figure 3 , the FSY1 gene from Glomerella graminicola clusters within the Magnaporthales ( event 1 ) , while Geomyces destructans ( Leotiomycetes ) obtained its copy from a lineage related to the Botryosphaeriales ( Dothideomycetes ) ( event 2 , Figures 1 and 3 ) . Both events are supported by topology comparison , using the Shimodaira-Hasegawa ( SH ) test , in which these FSY1 genes are constrained to occupy their expected place in the phylogeny ( P<0 . 01; Figure S4 and Table S2 ) . In addition , loss of synteny is observed in the regions surrounding the FSY1 gene in G . destructans ( Figure S2 ) , which is in line with its acquisition via HGT . The mycorrhizal fungus Oidiodendron maius is the only species examined that exhibits three FSY1 genes ( Figures 1 and 3 , events 3a and 3b ) . This species not only retained its cognate FSY1 copy but also acquired two additional genes by HGT in events that are supported both by topology testes ( P<0 . 01; Figure S4 and Table S2 ) and by synteny analysis ( Figure S2 ) . One xenolog is phylogenetically nested in the Pezizomycotina but it is not possible to identify unequivocally a donor lineage ( Figure 1 and Figure 3 , event 3a ) . The second clusters with high support with the FSY1 gene from G . destructans also acquired by HGT ( see above and Figure 1 , event 3b ) . Both species belong to the Leotiomycetes , but the remaining species examined in this clade ( Amorphotheca resinae , Botryotina fuckeliana , Sclerotinia sclerotiorum and Glarea lozoyensis ) , which are more closely related to O . maius than G . destructans , lack a similar gene acquired by HGT . This could mean that the MRCA of these species was the recipient of the transferred gene that was subsequently lost after the divergence of G . destructans from the other lineages . The Fsy1 homologues found in Cladonia grayi and Exophiala dermatitidis cluster together with the Leotiomycetes with high bootstrap values ( Figures 1 and S5 ) , but the two species belong to the Eurotiomycetes and Lecanoromycetes , respectively ( Figures 3 and S3 ) . These seem to constitute yet two independent HGT events ( numbered 4 and 5 , Figures 1 and 3 ) , which are supported by topology tests ( P<0 . 01; Figure S4 and Table S2 ) , but insufficient sampling prevents identification of a donor lineage . Finally , we identified a very recent inactivation of the FSY1 gene in Cercospora zeae-maydis ( Capnodiales ) caused by a transposon insertion within the coding region . This event seems to be very recent because the gene displays no sign of degradation and the phylogenetic position of the in silico translated protein is in line with the species phylogeny ( Figures 1 and 3 ) . In the order Eurotiales , four species do not encode a cognate Fsy1 ( Aspergillus nidulans , Aspergillus sydowii , Aspergillus versicolor and Eurotium herbariorum ) . Two independent FSY1 gene loss events seem to account for this ( Figure 4A ) . Interestingly , all these species were capable of capturing FSY1 genes new to the lineage ( events 6 and 7 in Figures 1 and 2; Figure 4 ) . To ascertain the nature of this phenomenon , we first asked whether these species might have “accidentally” lost their cognate FSY1 genes as part of gross deletions or chromosomal rearrangements . To assess this , we examined in detail the chromosomal region where , according to the very well conserved microsynteny among species in the Eurotiales , the cognate FSY1 gene should have been located in the three Aspergilli that lack a cognate FSY1 ( Figures 4B and S6 ) . Surprisingly , in A . nidulans , A . versicolor and A . sydowii the chromosomal region in question retains almost perfect synteny with the closest Aspergillus species in the immediate vicinity of the FSY1 gene with the exception of the absence of the latter and the inversion of a flanking gene ( NEO1 , Figures 4B and S6 ) . In E . herbariorum a similar situation is found , but in this case , three genes novel to the region are located at the position occupied by FSY1 in the species that possess a cognate gene ( result not shown ) . Interestingly , and in addition to these events where apparently only FSY1 was lost from this region , a number of rearrangements in Aspergillus carbonarius , Aspergillus aculeatus and Aspergillus fumigatus cause an abrupt loss of synteny in the regions immediately downstream of the FSY1 gene ( Figures 4 and S6 ) . The two types of xenologs found in these four species ( A . nidulans , A . sydowii , A . versicolor and E . herbariorum ) belong to two unrelated phylogenetic groups . The first group encompasses genes that cluster within the Sordariomycetes ( order Hypocreales; Figure 1 ) and were probably acquired from this lineage by the MRCA of A . nidulans , A . versicolor and A . sydowii ( event 6 , Figures 1 and 3 ) , in a single event that is supported by a topology test that rejects the monophyly of Aspergilli FSY1 genes ( P<0 . 01; Figure S4 and Table S2 ) . In accordance to this hypothesis , the genes exhibit a number of introns ( 3 or 4 introns; Figure 1 ) which is more in line with the number of introns found in FSY1 genes from the presumed donor lineage than with the number of introns found in the cognate Aspergillus FSY1 genes ( 1 intron; Figure 1 ) . Inspection of the genomic region in the vicinity of these xenologs in the three species shows an extent of synteny conservation consentaneous with the phylogenetic relation between the species , supporting that the three genes originate from a single HGT event ( Figure 4C ) . The second group of xenologs are found in six species ( Aspergillus kawachii ) and are also closely related to each other ( Figure 1 , event 7 and Figure 4A ) . Notably , these genes are phylogenetically closer to the Saccharomycotina Fsy1 homologs , suggesting that they were acquired from an undetermined lineage in this sub-phylum . This putative HGT event ( event 7 , Figure 1 ) is also supported both by the absence of introns in their FSY1 genes and by the SH test that rejected the alternative hypothesis of monophyly of Aspergilli FSY1 genes ( P<0 . 01; Figure S4 and Table S2 ) . Intriguingly , A . versicolor possesses two xenologs , one of each of the two groups ( Figure 4 ) although it lacks a cognate FSY1 gene . Two lines of evidence suggest that the acquisition of the “yeast-like” xenolog by the six species required multiple lateral transfer events . Firstly , given the phylogenetic position of the species that possess this type of xenolog , a single acquisition would have to have occurred in the MRCA of all Aspergillus species studied and consequently a large number of independent losses would have to be postulated to explain the present distribution . Secondly , no synteny is found in the vicinity of the “yeast-like” FSY1 xenolog , even between sibling species like A . niger , A . acidus and A . brasiliensis ( Figure 4D ) . This could mean that the gene underwent multiple ( sequential ) events of lateral transfer between these species , after it was originally acquired by one of them . However , it should be noted that the “yeast-like” Fsy1 phylogeny is compliant with species phylogeny , raising the possibility that the lack of synteny could be due to the gene being located in a region where synteny is usually absent , like subtelomeric regions . This could hold at least for A . niger , where FSY1 is located at the subtelomere , but not for its close relative A . acidus since the FSY1 gene is located far from the end of the chromosome in this species . In A . versicolor , the “yeast-like” FSY1 is also found in a non-telomeric region that shares synteny with its close relative A . sydowii , except for the presence of the FSY1 gene ( Figure S7 ) . This provides additional evidence that the acquisition of the “yeast-like” FSY1 gene occurred only after speciation . The alternative hypothesis would require two independent events of loss of FSY1 ( in A . nidulans and A . sydowii ) to explain extant FSY1 distribution . The first notable feature concerning Saccharomycotina Fsy1 homologs , is the fact that the few examples found in early diverging yeast lineages cluster much more closely than expected to the Pezizomycotina Fsy1 clade , or are even included in that clade . The latter is the case for the Fsy1 homolog from Lipomyces starkeyi , the most basal yeast lineage included in this study ( event 8 , Figures 1 and 3 ) . The homologs found in Arxula adeninivora ( event 9 ) and the cluster formed by the Fsy1 proteins found in Candida arabinofermentans and in two Ogatea species ( event 10 ) , seem also to be notably distinct from the remaining Saccharomycotina Fsy1 homologs , since their association with one of the groups defined by either of the two sub-phyla is not well-established ( Figure 1 ) . In addition , no synteny is observed in the vicinity of FSY1 in closely related species of this group ( viz . O . angusta and C . arabinofermentans ) , contrary to what is seen for the large majority of closely related species carrying a FSY1 gene ( Figure 2 ) . Together , these observations could suggest that the above-mentioned species acquired a FSY1 gene from the Pezizomycotina by at least three independent HGT events . However , the result of a topology test that compares the topology in Figure 1 with a topology in which the three lineages ( L . starkeyi , A . adeninivora and C . arabinofermentans/O . angusta and O . parapolymorpha ) were constrained to occupy their expected places in the Saccharomycotina phylogeny does not support this hypothesis . Likelihood scores associated with these two alternative topologies were not found to be significantly different using the SH test ( Figure S4 and Table S2 ) . Hence , using currently available data , we cannot exclude the possibility that FSY1 was present in the MRCA of the Pezizomycotina and Saccharomycotina . However , it is worth noting that this possibility contradicts the ancestral state reconstruction results and would require the assumption of at least five independent FSY1 gene losses in basal Saccharomycotina lineages in order to explain extant distribution of the gene . Should on the contrary events 1 , 2 and 3 be the result of HGT , than it remains difficult to state where in the evolutionary history of the Saccharomycotina the FSY1 gene would have been acquired . This is mainly because there is some uncertainty concerning the phylogenetic position of the C . arabinofermentans/Ogatea lineage ( Figure 3 and [43] ) , which is important to answer this question . In summary , we believe it is premature to conclude firmly about the origin of Sacharomycotina FSY1 gene and on whether or not the FSY1 genes in basal lineages have had their origin in HGT events . Further genome sampling in the more basal part of the Saccharomycotina lineage will be required to amend these results . The remaining Saccharomycotina Fsy1 homologs cluster together ( Figures 1 and S5 ) , indicating that they share a common ancestor , the majority of the genes being found in the so-called CUG clade . The CUG clade is a monophyletic clade within the Saccharomycotina in which all the species translate the CUG codon as serine rather than leucine [44]–[46] ( Debaryomycetaceae/Metschnikowiaceae; Figures 1 and 3 ) . In this lineage , the genomes of most species encode a Fsy1 homolog and the Fsy1 phylogeny within the clade is generally in accordance with the species phylogeny . In line with vertical transmission of the FSY1 gene during evolution of the species in the CUG clade is also the observation of conserved synteny in the region surrounding the FSY1 gene in all species , with the exceptions of Scheffersomyces stipitis and of the earliest derived species Millerozyma farinosa ( Figure 2 ) . However , in the Fsy1 phylogenetic tree ( Figure 1 ) , the CUG clade is not monophyletic . This entails that the lineage harbouring Candida albicans seems to be more closely related to the clade formed by the Saccharomycetaceae Fsy1 homologs than warranted by the phylogenetic relationship between these taxa . This could also suggest that the evolutionary trajectory of the FSY1 gene in the Saccharomycetaceae was initiated by an event of HGT from the C . albicans lineage to the MRCA of the Kluyveromyces/Lachancea lineage . However , this putative occurrence ( event 11 , Figures 1 and 3 ) is not supported by topology tests where the CUG clade is forced to be monophyletic ( Figure S4 and Table S2 ) . Finally , according to the species phylogeny , the Fsy1 homologs from Kluyveromyces and Lachancea should be more closely related to each other than to the Fsy1 homologues from Zygosaccharomyces and Torulaspora , but the phylogeny supports the opposite ( Figures 1 and 3 ) . This could imply that the gene was horizontally acquired by the MRCA of Zygosaccharomyces and Torulaspora from the Kluyveromyces lineage ( event 12 , Figures 1 and 3 ) . In fact , while the region where the FSY1 gene is normally located in the Kluyveromyces/Lachancea lineage exhibits very strong synteny between the latter lineage and Torulaspora delbrueckii , FSY1 is absent from that region in T . delbrueckii and appears rather at a subtelomeric region exhibiting synteny between T . delbrueckii and Z . rouxii ( Figure 2 ) . These observations are in line with the cognate FSY1 gene being lost in the MRCA of the Torulaspora/Zygosaccharomyces lineage and with the acquisition by this lineage of a novel gene from the Kluyveromyces branch . The result of a topology test constraining the Kluyveromyces and Lachancea Fsy1 homologs to conform to the species phylogeny contradicts the null hypothesis of vertical inheritance ( P<0 . 05; Figure S4 and Table S2 ) and thus supports the occurrence of this HGT event . Another clear conflict between Fsy1 phylogeny and the species phylogeny concerns the Fsy1 homologs from Cyberlindnera jadinii and Wickerhamomyces anomalus ( event 13 , Figures 1 and 3 ) . Of these two species , only the latter exhibits a putatively functional FSY1 gene because in the former a heavily degenerated pseudogene is found . The W . anomalus Fsy1 homologue is phylogenetically nested within the Debaryomyces branch of the CUG clade ( Figure 1 ) while the established species phylogeny places this species in a basal position relatively to the Saccharomycetaceae clade ( Figures 3 and S3 ) . Therefore , this species , or possibly the MRCA of W . anomalus and C . jadinii presumably also acquired FSY1 by HGT from a donor in the Debaryomyces lineage ( event 13 , Figures 1 and 3 ) . This putative HGT event is also supported by topology tests ( P<0 . 05; Figure S4 and Table S2 ) . The most important instance of FSY1 gene loss involving a large number of species , is observed in the clade that comprises all the yeasts descending from a common ancestor that underwent a Whole Genome Duplication ( WGD ) approximately 100 million years ago [47] . The pattern of distribution of Fsy1 is in line with the generally accepted view that the Zygosaccharomyces lineage is the closest to the WGD ancestor , and with a loss of the FSY1 gene in the WGD ancestor , since apparently none of the post-WGD genomes encodes cognate Fsy1 homologues ( Figures 1 , 2 and 3 ) . The two genes in S . uvarum and S . eubayanus are , together with the gene found in S . cerevisiae strain EC1118 , the sole instances of genomes encoding Fsy1 in post-WGD yeasts ( Figures 1 and 3 ) . The gene in the S . cerevisiae strain was formerly shown to have been acquired by HGT [14] , [35] . Since the WGD ancestor is predicted with high likelihood to have lost its FSY1 gene , it seems probable that the common ancestor of the very closely related S . uvarum and S . eubayanus species also acquired a FSY1 gene by HGT ( event 14 , Figures 1 and 3 ) . The phylogenetic position of the S . uvarum and S . eubayanus genes does not contradict the species phylogeny in this case , which makes it impossible to find support for this event using topology tests . If a HGT event took place , the donor lineage must have been very closely related to the last pre-WGD species , Z . rouxii and T . delbrueckii . The alternative explanation would be that the FSY1 gene was present in the WGD ancestor and was lost independently in all the post-WGD lineages except the Saccharomyces lineage , which does not seem to be parsimonious and is in contradiction with the predictions of the ancestral state reconstruction analysis ( Figure 3 ) . In addition to multiple events of HGT and the loss of FSY1 in the WGD ancestor , several independent FSY1 gene losses have to be postulated since approximately half of the Saccharomycotina species examined do not carry the gene . This is the case for the genus Eremothecium ( Figure 3 ) , in addition to two independent losses that , in the present limited sample , involve only a single species each ( in Candida tenuis and Candida tanzawaensis , Figure 3 ) . Taking both the phylogeny and the synteny analysis into account ( Figure 2 ) it seems likely that loss of the FSY1 gene in C . tanzawaensis was preceded by a translocation of the gene to a different chromosomal location in the MRCA of Scheffersomyces stipitis and C . tanzawaensis . This is suggested by the loss of synteny in the region surrounding the FSY1 gene in S . stipitis , a situation unique among the CUG clade species . Finally , if basal Saccharomycotina lineages acquired the FSY1 gene by vertical inheritance , which , as mentioned above , cannot be excluded in face of available data , at least five additional independent losses have to be postulated in the this region of the Saccharomycotina tree . We noted that five Saccharomycotina species harbour two FSY1 genes . Three of these species ( C . albicans , C . dubliniensis and C . tropicalis ) represent apparently three independent segmental duplication events that took place after speciation , since the paralog pairs encode ( nearly ) identical Fsy1 proteins . A fourth CUG clade species , Millerozyma farinosa , testifies to a different situation since the sequenced strain ( CBS 7064 ) is in fact an interspecies hybrid in the process of resolution [48] . In this case , the two distinct FSY1 genes are located in two heterozygous chromosomes , one of which was acquired from a different , but closely related species [48] . An older duplication seems to have given rise to the two FSY1 genes presently found in A . adeninivorans . The particularly dynamic evolutionary history of FSY1 raises the question of whether the Fsy1 homologue family may contain proteins whose functional properties underwent substantial changes in one or more lineages in the course of evolution . The best-characterized Fsy1 homologue originally cloned from Saccharomyces pastorianus was found to mediate solely the uptake of fructose and sorbose [28] , [49] . Fsy1-mediated glucose transport was undetectable in in vitro assays and was insufficient to support significant growth on glucose of a S . cerevisiae strain devoid of its cognate Hxt transporters [32] , [50] . The other Fsy1 homologues characterized so far exhibit similar properties [33] , [34] . To evaluate the extent of conservation of Fsy1 function in other phylogenetic lineages , we similarly expressed some of the newly identified genes as sole hexose transporter in S . cerevisiae . The Fsy1 homologues tested , indicated in Figure 1 , were all capable of complementing growth of the hxt-null strain on fructose ( Figure 5 ) . In addition , they all appear to operate as H+ symporters , since substrate addition to aqueous cell suspensions of strains expressing the various Fsy1 proteins resulted in all cases in transient alkalinisation of the extracellular medium . Moreover , the results shown in Figure 5 suggest that all of the Fsy1 homologues tested accept sorbose as a substrate in addition to fructose , while none is capable of transporting significant amounts of glucose . Notably , Fsy1- mediated sorbose uptake seems to be particularly vigorous in L . starkeyi , which contrary to C . arabinofermentans , K . lactis and S . uvarum , is capable of growing on sorbose as sole carbon and energy source . Discrepancies in the strength of H+ symport signals observed between the strains expressing the different transporters may simply result from distinct efficiencies in heterologous expression/membrane localization . Therefore , these discrepancies are not informative in what concerns possible functional differences between the transporters when operating in their natural context . In summary , we conclude that Fsy1 function seems to have remained remarkably constant in the course of evolution , possibly with some species-specific adjustments that can be traced back to physiological characteristics of the species to which a particular protein belongs , but apparently without major shifts in substrate preference or mode of operation . This is particularly important to note for the “yeast-like” A . niger Fsy1 and the homologs from L . starkeyi and C . arabinofermentans , all of which are located on long branches of the Fsy1 tree ( Figure 1 ) . Our survey among available fungal genomes brought to light a patchy distribution of FSY1 , a gene encoding a specific high affinity fructose/H+ symporter , in two sub-phyla of the Ascomycota , Pezizomycotina and Saccharomycotina . Two instances of Fsy1 loss in deeper nodes of the phylogeny are very possibly related to important turning-points in lifestyles and genomic make-up of the lineages involved . The first , concerning the Onygenales , is most probably a consequence of a shift from a nutritional association with plants to animals [42] , [51] and the reduction of gene families associated with the metabolism of plant material [51] . It seems plausible that FSY1 is also among the genes that became dispensable in this context . The second instance concerns the ancestor of the Saccharomyces lineage that underwent whole genome duplication ( WGD ) . In this case , a switch to a fermentative lifestyle occurred and the number of hexose transporters increased substantially [22] . Our phylogenies suggest that Fsy1 became dispensable and was lost very soon after the WGD , before the first extant post-WGD lineages diverged , possibly as a result of the WGD itself . In addition , several independent and more recent losses in various lineages were noted , but as far as can be judged from available data ( for S . uvarum , S . eubayanus and A . niger ) , there is no intra-species variation in the presence of the FSY1 gene . Notably , in these three species , as well as in Z . rouxi and T . delbrueckii , the gene is found in subtelomeric regions ( Figures 3 and 4 ) , a genomic location characterized by frequent chromosomal rearrangements and enrichment in transposable elements [52]–[54] . Interestingly , these regions are also known for facilitating the accumulation of genes that are associated with niche-specific adaptations . For example , in S . cerevisiae several genes involved in sugar utilization , as well as the FLO gene family involved in the flocculation process important in brewing , are located in subtelomeric regions [55] , [56] . The identification of instances of horizontal gene transfer is often controversial , mainly because assumptions on their occurrence have been made based on insufficient evidence in the past and the transfer mechanisms are poorly understood [57] . However , sufficient consensus exists that substantial and well-supported incongruences between a gene tree and the accepted species phylogenetic tree is a strong indication of HGT [20] , [58] . Likewise , other types of “character-state discordance” , such as patchy phylogenetic distribution of the genetic element along various lineages and inconsistency in sequence patterns between the gene and the resident genome ( e . g . number of introns ) are also good evidence to support HGT [59] . In the present study , we detected by phylogenetic analyses at least 10 novel and independent events of HGT involving the same gene , FSY1 . Most of the well-supported HGT involving fungal donors and recipients reported so far , concern entire metabolic pathways related with the production of toxins or the assimilation of nutrients , and the acquisition of detoxification or new metabolic capacities [13] , [16]–[18] , [60] , leading in some cases to a rapid emergence of new pathogenic lineages and to successful host specialization [61]–[63] . The present case fits the previously described scenarios , although it involves a single gene that nevertheless is sufficient to confer fructose-scavenging capabilities to the recipient organism on its own . Fungi are very often associated with plants , and these are , in turn , the main natural source of fructose , often as a constituent of poly- or oligosaccharides . For example , the metabolism of sucrose ( a disaccharide composed of glucose and fructose ) plays a very important role in the sugar-partitioning in plant-fungal interactions that occur both in mutualism and pathogenesis [64] . Complex polysaccharides are also a potential source of fructose , and Fsy1 may play a relevant role in the utilization of fructose resulting from degradation of this type of molecule , which will typically yield low concentrations of fructose that can be better assimilated via an uphill transport system . However , both filamentous fungi and yeasts are known to possess other hexose/H+ symporters that accept both glucose and fructose ( and often other monosacharides ) as substrates [64]–[67] . The advantage of having a specific fructose transporter like Fsy1 may lie in the fact that most monosaccharide transporters , including those active at plant/fungal interfaces in mutualistic interactions ( mycorrhiza ) and pathogenesis , preferentially take up glucose [26] , [64] . Consequently , the availability of a specific fructose carrier that is not inhibited by glucose might bring about a considerable increase in efficiency in fructose utilization in environments where both sugars are present . In this respect , it is striking that O . maius , the only species among those included in this study that possesses three FSY1 homologues , is also the only fungus included in the analysis with a primary mycorrhizal lifestyle , suggesting that Fsy1 may be particularly useful in this setting [64] . In addition , given the fact that most Fsy1 homologues described are found in phytopathogenic fungi , it seems plausible that the gene appeared as a specific fructose transporter operating at the plant/fungal interface either in a pathogenic or symbiotic context . This idea should however be revisited once a larger number of genomes in the Ascomycota are available so that the diverse lifestyles are well represented , since there is a bias towards ( phyto ) pathogenicity in the species sequenced so far . The highly unusual number of independent HGT involving the FSY1 gene , the lack of evidence for flanking genes having been co-transferred and the extremely precise “deletions” of the gene in the A . nidulans clade and in E . herbariorum seem to configure a situation of gains and losses targeting this gene specifically . Another intriguing observation that may have some relation with the above mentioned gene losses in the Eurotiales , is the fact that four species ( A . carbonarius , A . aculeatus , N . fischeri , A . fumigatus ) that retain inter-species synteny upstream of the FSY1 gene , completely loose synteny immediately downstream of the FSY1 gene . This suggests that the region in the vicinity of the cognate FSY1 locus was frequently reused as recombination site and may be particularly prone to genomic instability , at least in this lineage . The reason for this is not obvious , but could lie in the presence of an inconspicuous functional element like an origin of replication , which was previously suggested to favour chromosome fragility [6] . Taken together , these results configure an extraordinarily dynamic evolutionary history centred in a single gene , which is to our knowledge unparalleled in fungi . Contrarily to the turbulent pattern of gains and losses , there is so far no evidence for marked functional divergence between Fsy1 homologues heterologously expressed in S . cerevisiae and spanning large evolutionary distances , both in the Saccharomycotina and the Pezizomycotina [33] , [34] . Our results indicate that this also holds for both A . niger Fsy1 homologues , so that acquisition of a second gene by HGT does not seem to be related with a function distinct from fructose transport , at least in this case . The Botryotinia fuckeliana Fsy1 homologue was shown to mediate fructose uptake in a manner that was not inhibited by glucose , although no kinetic parameters for initial uptake rates are available . On the other hand , expression of the Fsy1 homologue from Fusarium verticillioides in S . cerevisiae failed to restore fructose transport in a strain devoid of hexose transporters but it is presently unclear whether this could be due to incorrect subcellular localization of the heterologous protein [68] . Finally , it should be noted that all Fsy1 proteins tested so far are also capable of transporting sorbose , albeit to various extents . While this capacity does not seem to be relevant in the context of several Fsy1-harbouring yeast species that do not grow on sorbose , it was probably another important factor influencing Fsy1 evolution in sorbose utilizing fungi , like the yeast L . starkeyi . Taking into account the biochemical properties determined for Saccharomyces pastorianus Fsy1 , we hypothesize that the signal of the fitness effect imparted by the presence of the gene oscillated in some lineages in the course of evolution , so that FSY1 gene losses have been fixed at one point in time and later “reversed” by the acquisition of a FSY1 homologue by HGT . Although this seems the most likely course of events , we cannot exclude that the acquisition of a second phylogenetic distant ortholog by HGT preceded the loss the cognate copy [69] . Intriguingly , in the Pezizomycotina , HGT seems to have been the preferred mechanism to increase the number of FSY1 genes within a species ( like in A . niger , A . kawachii , A . brasiliensis , A . acidus , A . versicolor , and O . maius ) . On the contrary , in the Saccharomycotina , gene duplications seem to account for all the instances where two genes are found in one species ( three Candida species and probably in A . adeninivorans ) . While in prokaryotes events of HGT are very common and the underlying mechanisms are generally well understood , fewer examples of HGT have been reported in fungi and the possible mechanisms involved are largely unknown [20] , [58] . In other eukaryotic microbes , HGT has been often associated with endosymbiosis events and with phagocytosis , which are not relevant to explain HGT fungi [5] , [20] , [58] , [70] , [71] . Nevertheless , it should be noted that fungi may readily undergo hyphal anastomoses , and that heterokaryon incompatibility usually does not completely prevent cytoplasmic or nuclear exchange [72] , [73] . While numerous studies of HGT in fungi were so far based on surveys that detected mainly genes of bacterial origin [10] , [74]–[77] , this study is centred in a gene encoding a protein with a well-known and unusual phenotype . Most importantly , the FSY1 gene , being a reasonably recent “invention” in the ascomycetes turned out to be a particularly suitable system to study the dynamics of gene gain and loss while disentangling complicated phenomena like xenolog gene displacement and pseudoparalogy . Orthologous relationships were easy to establish in the Fsy1 cohort and most of the suspected HGT events were confirmed by the expected changes in the chromosomal setting of the gene , in all cases that could be investigated , and were reinforced by topology comparisons . This study suggests the possibility that the methods used so far to survey large amounts of fungal genomic data to detect HGT may be missing a significant number of intra-kingdom events . Identification and detailed study of such events seems worth pursuing , since they are very likely to provide invaluable new insights in the evolution of eukaryotic genomes . To distinguish Fsy1 homologues from other fungal sugar transporters , an initial BLASTP [78] search was performed using the Saccharomyces pastorianus Fsy1 protein sequence [GenBank:CAC08232] as query to retrieve putative homologues from GenBank . Sequences with E-values lower than 1e-41 were aligned using a fast iterative method in MAFFT v . 6 . 956 and poorly aligned regions were removed using Gblocks v . 0 . 91b [79] . The final alignment consisting of 520 sequences ( see Table S3 ) was used to construct a maximum likelihood ( ML ) phylogeny in RAxML v7 . 2 . 8 [80] , using the PROTGAMMAWAG model of amino acid substitution and 100 rapid bootstrap replicates . Sequences with E-values lower than 1e-80 formed a well-defined clade representing the entire FSY1 gene family . Additional BLASTP and TBLASTN [81] searches were performed to retrieve Fsy1 homologues from the nr database in GenBank and from fungal genome databases available as of August 2012 . Sequences were retained if their E-values were lower than 1e-80 and aligned over the majority of their extension . For Candida zemplinina PYCC 3044 , about 83 million Illumina paired-end reads were generated with HiSeq 2000 technology and assembled de novo using the GS De Novo Assembler v . 2 . 6 . The final draft assembly consists of 162 scaffolds . We set up a local BLAST database for this genome to search for FSY1 homologues using the abovementioned criteria . This genome has not been released yet but gene sequences used in this study were deposited in GenBank ( Table S1 ) . To construct the species phylogeny ( Figure 3 ) , for which whole genome sequence data is available , we used a previously described approach [37] . Briefly , the amino acid sequences of six RNA polymerase subunits ( Rpa1 , Rpa2 , Rpb1 , Rpb2 , Rpc1 , Rpc2 ) were retrieved from each genome database ( including our genome database of C . zemplinina ) by BLASTP and TBLASTN using Saccharomyces cerevisiae RNA polymerase amino acid sequences as query ( GenBank:P10964 . 2 , GenBank:P22138 . 1 , GenBank:P04050 . 2 , GenBank:P08518 . 2 , GenBank:P04051 . 1 and GenBank:P22276 . 2 , respectively ) . When predicted gene models were unavailable or in cases where proteins were most likely incorrectly predicted , they were re-annotated using AUGUSTUS [82] , which relies on a set of training annotation files from several fungal species to offer more precise gene predictions . Fsy1 and RNA polymerase sequences of Arxula adeninivorans and Yarrowia hispaniensis ( Candida hispaniensis ) were kindly provided by Cécile Neuvéglise ( INRA , France ) . A complete list of fungal taxa , abbreviated species names , genome databases queried in this study and accession numbers of Fsy1 and RNA polymerase proteins is given in Table S1 . Fsy1 and the individual amino acid sequences of six RNA polymerase subunits were aligned using MUSCLE [83] and poorly aligned regions were removed using Gblocks v . 0 . 91b with the following settings: maximum number of contiguous nonconserved positions allowed = 4; minimum length of a block allowed = 10 . To construct the species phylogeny ( Figure 2 ) , RNA polymerase amino acid sequences were concatenated using Concatenator v . 1 . 1 . 0 [84] to produce a final dataset of 160 sequences containing 6398 positions . This sequence alignment was subsequently used to construct a maximum likelihood ( ML ) phylogeny in RAxML v7 . 2 . 8 [80] , using the PROTGAMMAWAG model . Rhodotorula graminis , Cryptococcus neoformans and Ustilago maydis sequences were used as outgroup . The Fsy1 phylogeny ( Figure 1 ) was also inferred by ML in RAxML using the PROTGAMMAIWAGF model and an alignment of 107 amino acid sequences containing 407 positions . In the absence of a good outgroup the tree was rooted at the midpoint ( corresponding to the longest pathway between two operational taxonomical units ) . Branch supports for both phylogenetic trees were determined using 100 rapid bootstrap replicates . In order to examine the degree of phylogenetic conflict within the RNA polymerase concatenated alignment and in an attempt to corroborate the FSY1 phylogenetic tree , phylogenetic networks were generated with the same datasets . The alternative splits were found using the NeighborNet method [85] , and represented as a phylogenetic network using the Splitstree software v4 . 12 . 6 [86] , [87] . Comparative topology analyses were performed for each of the putative HGT events that resulted in a phylogeny that conflicts with the species phylogeny . In each case , Fsy1 sequences thought to have been horizontally transferred were constrained to be placed as in the species tree ( Figure S4 ) . The Shimodaira-Hasegawa ( SH ) test [88] , as implemented in RAxML , was used to determine whether the ML estimate of each of the constrained topologies differed significantly from the ML estimate of the unconstrained Fsy1 topology ( Table S2 ) . The ancestral state reconstruction of Fsy1 was performed in Mesquite v . 2 . 75 [89] . Discrete characters ( presence/absence of FSY1 gene ) were reconstructed using maximum-likelihood [90] with Markov k-state 1 parameter model” ( Mk1 est . ) [91] , which assumes equal probability for changes between states . The species tree inferred from the ML phylogenetic analysis and a character matrix of “absence” or “presence” of Fsy1 in the extant species ( scored as binary characters “0” or “1” , respectively ) , were used as input . For each clade , conservation of synteny blocks encompassing FSY1 was accessed to corroborate orthology and to investigate possible genomic rearrangements . For Saccharomycetes species included in Figure 2 , synteny in the vicinity of the FSY1 gene was inspected using both the Yeast Genome Order Browser ( YGOB , http://wolfe . gen . tcd . ie/ygob/ ) [92] and the Candida Genome Order Browser ( CGOB , http://cgob . ucd . ie/ ) [93] , which are online tools for visualizing gene order and context in several yeast genomes available . FSY1 homologues from Kluyveromyces lactis ( KLLA0E09021g ) and Candida albicans ( CAWG_01680 ) were used as query in YGOB and CGOB , respectively . For Aspergillus species , ortholog clusters were obtained from the Aspergillus Genome Database ( AspGD , http://www . aspergillusgenome . org/ ) [94] using the two A . niger FSY1 homologues ( An15g01500 and An06g02270 ) as query . The visualization tool Sybil was used to navigate the ortholog clusters in their genomic context . For Eurotium herbariorum , which is not included in the present version of the AspGD , a ∼100 kb annotated region encompassing the FSY1 homologue ( Scaffold 7 , 473286–574932 ) was retrieved from the E . herbariorium genome database ( http://genome . jgi . doe . gov/Eurhe1/Eurhe1 . home . html ) and compared using BLAST analyses . For the remaining fungal species , which are not included in these databases , synteny conservation was assessed based on their current annotations as retrieved from their respective genome databases ( see Table S1 ) and confirmed by high scoring BLASTP hits in GenBank . S . cerevisiae hxt-null EBY . VW4000 [50] was used as host for heterologous expression of putative FSY1 homologues . Candida albicans PYCC 3436T , Candida arabinofermentans PYCC 5603T and Lipomyces starkeyi PYCC 4045T were obtained from the Portuguese Yeast Culture Collection ( PYCC , Caparica , Portugal ) and Aspergillus niger ATCC 16404 was obtained from CiiEM , Instituto Superior de Ciências da Saúde Egas , Portugal . Plasmid p415 TEF was obtained from the American Type Culture Collection ( ATCC; Manassas , VA , USA ) . Strains were grown in YPD medium ( 1% w/v yeast extract , 2% w/v bacto-peptone and 2% w/v glucose ) with the exception of S . cerevisiae EBY . VW4000 that was grown in YPM medium ( 1% w/v yeast extract , 2% w/v bacto-peptone and 2% w/v maltose ) . Plasmids containing FSY1 homologues from S . uvarum CBS 7001 , C . arabinofermentans PYCC 5603T , L . starkeyi PYCC 4045T , C . albicans PYCC 3436T and both the “yeast-like” and the cognate FSY1 copies from A . niger ATCC 16404 were constructed by homologous recombination in S . cerevisiae EBY . VW4000 . Intronless FSY1 homologues were amplified by PCR using genomic DNA isolated as previously described [95] . Two sets of primers and two successive PCR amplifications were performed . In the first PCR , “short” primers matching the 5′ and 3′ ends of FSY1 coding sequences were used to increase amplicon specificity and yield . The resulting PCR products served as template in a second reaction , which used “long” primers comprising overhangs ( 38–46 bp ) identical to the 3′ end of TEF promoter and the 5′ end of the CYC1 terminator , to allow for homologous recombination into the p415 TEF plasmid . Preparative PCRs were performed in a final volume of 50 µl with the following components: 1X Long PCR buffer with 15 mM MgCl2 ( Fermentas ) , 0 . 20 mM of each of the four deoxynucleoside triphosphates ( GE Healthcare ) , 0 . 2 µM of each primer , 100–200 ng of genomic DNA , and 2 U Long PCR Enzyme Mix ( Fermentas ) . Thermal cycling consisted of a 3-minute denaturation step at 94°C , followed by an initial round of 10 cycles of denaturation at 94°C for 20 s , annealing for 30 s ( variable temperature ) , extension at 68°C for 2 min , and a second round of 25 cycles increasing the extension time by 1 s in each cycle . A final extension of 10 minutes at 68°C was performed . FSY1 homologues containing introns ( from C . albicans and the cognate copy of A . niger ) were amplified from cDNA . Total RNA was isolated using Trizol ( Invitrogen ) as previously described [95] from cells grown in Yeast Nitrogen Base ( YNB ) with 0 . 5% w/v fructose . First-strand cDNA was synthesized with Super Script III Reverse Transcriptase ( Invitrogen ) according to the manufacturer's instructions and using the “short” gene-specific reverse primers ( see Table S4 ) . The resulting product served as template in a subsequent PCR using the same settings . S . cerevisiae hxt-null EBY . VW4000 was transformed [96] simultaneously with p415 TEF linearized with BamHI and HindIII ( Roche ) and the various gene fragments with the short flanking regions for homologous recombination . Final plasmid constructs , primer sequences and specific annealing temperature are described in Table S4 . Recombinant yeasts harboring FSY1 homologues were grown in liquid YNB ( without aminoacids ) medium with 1% w/v fructose containing uracil , tryptophan and histidine . Cells were grown to mid exponential phase ( OD640 between 0 . 8 and 1 . 2 ) , harvested by centrifugation , washed twice with sterile cold water and resuspended to a final concentration of 20 to 22 mg dry weight/ml . The presence of symport activity was assessed by computer recording the alkalization of an aqueous yeast cell suspension elicited by the addition of fructose or sorbose , using a standard pH meter [65] and a home designed software . The absence of glucose transport was confirmed by the absence of acidification after sugar addition to the cell suspension . Transport of D-[U-14C]fructose was measured according to the procedures described by Spencer-Martins and Van Uden [97] . Kinetic parameters were determined by non-linear regression ( Michaelis-Menten Equation ) using GraphPad Prism ( v5 . 00 for Windows , GraphPad Software , San Diego California USA ) .	Genes are commonly vertically inherited , meaning that they share the evolutionary history of the organisms in which they are found . However , they can also be transmitted between species with overlapping niches , a phenomenon known as horizontal gene transfer ( HGT ) that can occur between closely related species but also between organisms belonging to different domains of life . While HGT is very common in prokaryotes , it has been less frequently reported in eukaryotes , including eukaryotic microbes . In fungi , several instances of genes acquired by HGT from bacteria have been reported , but gene exchange between fungal species is thought to be rare . Here , we describe our findings concerning a single fungal gene that seems to have been transferred between fungi very often . We believe this may be related to the fact that the gene can be both very useful and detrimental for the host , depending on genetic background and environment . Our results suggest that exchange of genes between fungi may happen much more frequently than assumed so far .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "organismal", "evolution", "genomics", "adaptation", "microbial", "evolution", "comparative", "genomics", "hybridization", "biology", "evolutionary", "biology", "genomic", "evolution", "evolutionary", "processes", "evolutionary", "genetics" ]	2013	Extensive Intra-Kingdom Horizontal Gene Transfer Converging on a Fungal Fructose Transporter Gene
Epilepsy is one of the most common signs of Neurocysticercosis ( NCC ) . In this study , spatial and temporal variations in the incidence of hospitalized cases ( IHC ) of epilepsy and NCC in Ecuadorian municipalities were analyzed . Additionally , potential socio-economic and landscape indicators were evaluated in order to understand in part the macro-epidemiology of the Taenia solium taeniasis/cysticercosis complex . Data on the number of hospitalized epilepsy and NCC cases by municipality of residence were obtained from morbidity-hospital systems in Ecuador . SatScan software was used to determine whether variations in the IHC of epilepsy and NCC in space and time . In addition , several socio-economic and landscape variables at municipality level were used to study factors intervening in the macro-epidemiology of these diseases . Negative Binomial regression models through stepwise selection and Bayesian Model Averaging ( BMA ) were used to explain the variations in the IHC of epilepsy and NCC . Different clusters were identified through space and time . Traditional endemic zones for NCC and epilepsy , recognized in other studies were confirmed in our study . However , for both disorders more recent clusters were identified . Among municipalities , an increasing tendency for IHC of epilepsy , and a decreasing tendency for the IHC of NCC were observed over time . In contrast , within municipalities a positive linear relationship between both disorders was found . An increase in the implementation of systems for eliminating excrements would help to reduce the IHC of epilepsy by 1 . 00% ( IC95%; 0 . 2%–1 . 8% ) and by 5 . 12% ( IC95%; 3 . 63%-6 . 59% ) for the IHC of NCC . The presence of pig production was related to IHC of NCC . Both disorders were related to the lack of an efficient system for eliminating excrements . Given the appearance of recent epilepsy clusters , these locations should be studied in depth to discriminate epilepsies due to NCC from epilepsies due to other causes . Field studies are needed to evaluate the true prevalence of cysticercosis in humans and pigs in different zones of the country in order to better implement and manage prevention and/or control campaigns . Humans are the definitive hosts of Taenia solium harboring the intestinal adult tapeworm , which causes taeniasis [1;2] . Humans acquire the tapeworm through consumption of improperly cooked infected pork . The intermediate pig host gets infected by ingestion of parasite eggs , passed in the stool of a tapeworm carrier . The metacestode larval stage establishes in the pig’s muscles , brain and other tissues ( cysticercosis ) [2] . Unfortunately , humans can also serve as dead-end intermediate hosts by accidentally ingesting parasite eggs and developing the metacestode larval stage [2] . In humans , the parasite tends to locate in the central nervous system ( neurocysticercosis ( NCC ) ) causing a variety of neurological symptoms , such as seizures , headache and in many cases epilepsy [3–5] . In developing countries , NCC is often an underrecognized and neglected disease [6;7] . The large variety of clinical signs and symptoms , and the inaccessibility of highly sensitive tests , like Computerized Tomography ( CT ) or Magnetic Resonance Imaging ( MRI ) ( due to high costs involved and the unavailability of neuroimaging facilities ) , have contributed to the underreporting of NCC . Different measures such as education , improvement of household sanitation , changes in meat inspection practices , identification and treatment of tapeworm carriers , mass drug administration and modifications in pig-rearing methods have proven , at least at short term , to be effective in lowering levels of transmission of NCC [8–10] . However , practices like free roaming pig management , clandestine marketing of living pigs and pork , open defecation , and the use of residual waters in irrigation [10;11] , make the disease still prevalent in many regions . In developing countries , T . solium NCC has been found to be the leading cause of acquired epilepsy . In fact , NCC was found to be responsible for at least half or a third of acquired epilepsy cases [12;13] . In Ecuador , an average of 480 and 1670 patients are hospitalized each year because of NCC and epilepsy , respectively [14] . Notification of hospitalized cases for both disorders is mandatory in Ecuador . Tools for diagnosing causes of acquired epilepsy , including parasitic and infectious diseases are available in the country; however , these diagnostic tools are not regularly used because of their high costs , thus , as in other NCC endemic countries , in Ecuador more than 82% of epilepsies do not have definitive diagnosis and the cause remains unknown [14–17] . Symptomatic and asymptomatic cases of NCC have been studied in urban [18–20] and rural [12;21–25] areas of Ecuador in different epidemiological studies . Endemic areas were found mostly in the highlands , where the population has a high exposure to the parasite as measured by specific antibody detection . Up to one third of the population is seropositive in some of these areas [22] . In some endemic communities in these highland areas , human cysticercosis active infections , as measured by antigen detection are present in up to five percent of the population [23] . However , in the past three decades , a decreasing trend in the incidence of hospitalized cases ( IHC ) of NCC has been observed in the country , apparently due to an improvement in sanitary conditions and the use of better diagnostic tools [20] . Epidemiological surveillance data in the Ecuadorian health status reports only include ambulatory cases from the public health care system . The public registers from the Ministry of Public Health are the only official health reports available . The consultations at any level attended in the public health system have been estimated to be around 30% of all the medical consultations in the country [26] . However , for the ambulatory cases the reporting is not always done properly , mainly due to the lack of synchronization among the different types of health care services in Ecuador [27] . Fig 1 explains the reporting system for NCC and epilepsy cases in Ecuador . Additionally , and in a parallel way , the National office of Statistics ( INEC ) collects all information on morbidity registered in hospitals and clinical centres ( from public , private or social security systems ) where patients are attended at the secondary health care level and also in case of emergencies irrespective of their underlying sources . In these registers , both NCC and epilepsy are frequent causes of hospitalization . Thus , epidemiological data generated by the Ecuadorian office of statistics offers the possibility of studying important variables related to the macro-epidemiology of NCC and epilepsy in Ecuador . Here , we use macro-epidemiology in terms of the determinants of disease , including economic , social , and climatological factors into national patterns in risk assessment [28] . In this study we aimed at identifying areas ( municipalities ) with a high IHC of NCC and epilepsy between 1996 and 2008 . In addition , given the fact that the IHC of NCC and epilepsy have been related to several socio-economic , and landscape variables , we evaluated the macro-epidemiology of epilepsy and NCC in Ecuador at the municipality level . Finally , spatio-temporal analysis was implemented in order to investigate the distribution of the IHC of epilepsy and NCC in Ecuador . Ethical approval was not required for this study . All information used in this study came from public sources freely available on the referenced websites . Reports of human cases belonged to the public health surveillance system , the anonymity of clinical histories is guaranteed by legal mandate . The study unit was the municipality . The number of patients hospitalized in different institutions with diagnosis of epilepsy and NCC , from 1996 to 2008 was obtained from hospital morbidity and mortality databases managed by the National Office of Statistics ( INEC ) [14] . This time period was chosen because of the availability of digitized morbidity data ( www . inec . gob . ec ) , and because data on agricultural and life conditions of the population were available for that time period [14;29] . Additionally , we did not consider the data of the period after 2008 because biases in the number of cases were expected given the fact that public health systems increased their coverage and became more accessible and free of charge , including the distribution of parasitic drugs [30] . The cases were identified through the ICD-10 codes for disease classification [31] . In Ecuador , the protocol to declare a patient with NCC is defined by the neurology department of hospitals . Briefly , the diagnosis of NCC follows the directions of the Del Brutto ( 2012 ) criteria . The diagnosis is based on patient’ clinical symptoms and signs ( seizures , headache , dementia , hydrocephalous , among other neurological disorders ) , serology ( detection of antibodies directed to T . solium metacestodes and/or circulating antigens of T . solium metacestodes in serum or cerebrospinal fluid ) and imaging ( CT , MRI ) [32] . In the public sector there exist 39 neuroimaging facilities ( CT-Scans ) the majority of them are in the provincial capital cities ( 24 provinces ) . The private sector also has neuroimaging capacity but the number of scanners is unknown . The database has information about each hospitalized patient with cause-specific morbidity , hospital of attendance , and the place ( parish or municipality ) where the patient lives ( access to official databases are included in S2 ) . All municipalities in the continental part of the country were included in this study ( 217 municipalities ) . The registers of the Galapagos Islands were excluded because they might distort the spatial analysis , although both disorders have also been reported there . Each record was designed to contain the total number of NCC and epilepsy cases , the total population , the year and the geographical coordinates of the centroid of each municipality . Time trends of hospitalized reported cases for both disorders were tested using Negative Binomial regression models . Furthermore , the relationship between the number of hospitalized cases of epilepsy and NCC was evaluated using correlation analysis on Log transformed data . Additionally , data on several explanatory variables were gathered using governmental databases about several socio-economic indices , information from the agricultural census conducted in 2000 , and climatological information from the National Hydrometeorological Service [33] . The information on explanatory variables possibly associated with NCC and epilepsy at municipality level were grouped into different classes . Climatic variables: ( tropical or highlands ) ( ZONE ) , rainfall ( RAIN ) and number of days without precipitations ( DRYD ) . For each municipality , values of RAIN and DRYD were inferred on the basis of 205 weather stations , and ordinary Kriging was used to interpolate values . The similarity with rainfall maps published by the National Hydrometeorological Service ( INAMHI ) let us choose the proper model [33] . Population variables: population number ( POPULATION ) , percentage of indigenous population ( %INDG ) , and percentage of rural population ( %RURAL ) ; Educational level: mean number of years of schooling ( SCHOOL ) , mean number of years of schooling for farmers in the municipality ( FSCHOOL ) , and percentage of farms receiving technical assistance ( TECHASSIS ) ; Sanitary conditions: % of families with piped water ( TUBWAT ) , % of dwellings with systems for eliminating excrements ( EXCR ) , and physicians per 10 , 000 inhabitants ( PHYS ) ; Poverty indices such as: percentage of families with unsatisfied basic necessities ( UBN ) ( Number of people or families that live under poverty conditions with respect to the population in a specific year , referring to the lack of dwelling , health , education and employment [29] ) , and percentage of people under extreme poverty ( EXTPOOV ) ; Livestock: percentage of agriculture land dedicated to pastures ( %GLS ) and pig population ( PIG ) [29] . In Ecuador , the majority ( 58 . 8% ) of the pig population is raised under the traditional husbandry system if we consider smallholder producers in Ecuador as those farmers owning ≤10 pigs [34] . The agricultural office of geographical information systems provided the map showing the political division of the country at municipality level . A description of all the variables considered for this study is presented in Table 1 ( access to official databases are included in S2 ) . Space-time analysis was used to determine whether municipalities with high incidence of hospitalized cases ( IHC ) of epilepsy and NCC are clustered in space and in time [35;36] . The space-time scan software [35] was used to search , and test for significance and identify approximate locations of areas with an increased risk for the occurrence of NCC and epilepsy cases . The analysis was run in SatScan Software v9 . 4 , with case file as the number of hospitalized reported cases , population file as the estimated total number of individuals in each municipality per year and as the coordinate file , the latitudes and longitudes of the centroids of each municipality . The spatial dimension varied from 0 up to 25% of the total number of centroids in the study region . The temporal dimension was established with a maximum of up to 50% of the study period with a time precision of 1 year . Poisson-distribution was used to contrast the number of cases in the scanning of areas . Space-time clustering was assessed by comparing the iRR ( incidence rate ratio ) of epilepsy and NCC IHC within a specific area and time in contrast to an expected iRR of hospitalized NCC and epilepsy cases if their incidences were randomly distributed . The cylinder with the maximum likelihood ratio was selected as the most likely cluster ( Primary cluster ) , and others no overlapping significant clusters were also selected . The significance of identified space-time clusters was tested using the likelihood ratio test statistic and p-values of the test were obtained through Monte Carlo simulations ( 999 ) . The significance was arbitrated at the 5% level . All significant clusters were visualized using QGIS software ( version 2 . 8 ) . Multivariable-count regression models were used to assess the relative contribution of different socio-economic and demographic variables to the IHC of epilepsy and NCC from 1996 to 2008 across the country . For the case of hospitalized NCC cases , an excess of zero cases were observed . For this reason , Zero inflated negative binomial models ( ZINB ) were used to explain the over-abundance of zero cases [37] . A manual forward stepwise procedure was implemented to select the set of independent variables that describe the number of NCC hospitalized cases or the probability of not observing any hospitalized case of NCC using the zinb command in STATA , version 12 software ( StataCorp LP , College Station , Texas ) . The procedure started with the null model ( with no covariates ) , and subsequently , covariates were added and evaluated for their importance . The Akaike Information Criteria ( AIC ) was used as the calibrating parameter and models with lower AIC values and few parameters were preferred . Two models were considered to be significantly different whenever the difference in AIC was greater than 3 [38] . In the forward selection procedure , each covariate was added in either the linear predictor for the count part or in the logit function for the absence of the disease , and the covariate with the best explanation preserved for the second round . If the addition of a covariate improved the model explanation through the reduction in AIC , the variable was captured and the process was repeated until the AIC value could not be reduced further [39] . The significance level for the covariates in the model was set at 0 . 05 . For the case of NCC Vuong’s test was applied in order to evaluate if zero-inflation was more appropriate as compared to the standard negative binomial model . To assess the influence of the selected indicators on the IHC of epilepsy , a Poisson regression model was used , and due to the presence of over-dispersion , Negative Binomial models were also evaluated through a likelihood ratio test for over-dispersion using the function odTest ( in pscl package , in the R software ) to test the null hypothesis that the restriction implicit in the Poisson model is true [40] . The approach of bidirectional elimination of variables was applied for variable selection , which also used the AIC as calibrating parameter . The functions glm . nb and stepAIC , in the MASS package under the R environment were used [41] . In addition , for the study of epilepsy , Bayesian model averaging ( BMA ) was used in order to deal with the uncertainty about the “correct” model [42] . BMA chooses the better model according to the best posterior probabilities among the models using the Occam’s window principle [38] . The inference about the explanatory variables in the best model is expressed as posterior effect probabilities , which indicate evidence of the importance of the effects of each variable in the model . The function bic . glm ( ) in the BMA package of the R software was used [43] with the specification that counts follow a quasi-Poisson distribution , and that variance increases with the square of the mean: an equivalent version of NB regression [44] . No explicit prior distributions for models and model parameters were assumed implying that all models were equally likely . Figs 2 and 3 display the histogram of the cumulative incidence of hospitalized epilepsy ( Fig 2 ) and NCC ( Fig 3 ) cases , during the study period ( between 1996 and 2008 ) over all the municipalities . For epilepsy , the distribution is more uniform compared to that of NCC; however , for NCC there are high proportions of zero cases throughout the municipalities . Fig 4 presents the Incidence of hospitalized cases ( IHC ) with both health problems over time in Ecuador . In the case of NCC , from 1996 to 2008 , there was an overall decreasing trend with around 5 cases per 100 , 000 inhabitants in 1996 to around 3 cases per 100 , 000 inhabitants in 2008 . The decreasing trend was statistically significant ( p<0 . 001 ) ; thus , annually a reduction of 5 . 68% ( IC95%: 4 . 5%-6 . 4% ) in hospital cases is expected . In contrast , for the case of epilepsy , the incidence appeared to be slowly and steadily increasing from 1996 to 2008 . The increasing trend was statistically significant ( p<0 . 001 ) . Annually an increase of 4 . 7% ( % ( IC95%: 3 . 7%-5 . 7% ) in reported cases is expected . It was also observed that as long as the incidence of epilepsy increased through time , the annual IHC of NCC appeared to decrease slowly; and the IHC for epilepsy was always higher than that of NCC . Table 2 presents a list of the top 15 municipalities with the highest IHC of epilepsy and NCC . It can be observed that municipalities with a high IHC of epilepsy did not necessarily have a high IHC of NCC . However , an overall ( all years combined for each municipality ) , highly significant positive linear correlation ( r = 0 . 78 , IC95% ( 0 . 72–1 . 00 ) , p<0 . 0001 ) was found on log ( x+1 ) transformed number of hospitalized cases of epilepsy and NCC reported in each municipality . Similar significant positive linear trends within municipalities were observed for each year evaluated separately . The IHC of epilepsy varied from 0 to 505 cases per 100 , 000 with an average number of 127 . 1 cases . On the other hand , for NCC , the IHC varied from 0 to 282 . 54 , and the average number of cases was 35 . 1 . Out of the top 15 municipalities with high IHC for epilepsy , six also featured amongst the top 15 for municipalities with high IHC for NCC . It should be highlighted that Quito ( capital city ) presented the highest number of patients ( 1829 and 3123 ) in hospitals for NCC and epilepsy during the study period , but the rates were diluted because of the city’s large population size . Four clusters were detected for the iRR of epilepsy when up to a maximum of 25% of centroids were included in the scanning window ( Fig 5 ) . The most likely cluster extended from the center to the northern part of the country , involving municipalities from Tungurahua , Cotopaxi , Napo and Pichincha provinces . This cluster lasted between 2003 and 2008 with an iRR of 1 . 57 in contrast to centroids outside the cluster ( p = 0 . 001 ) . A secondary cluster was located in the Guayaquil municipality in Guayas province in the Southern coast of the country with an iRR of 1 . 69 ( p = 0 . 001 ) . This cluster existed between 2005 and 2008 with 1745 hospitalized cases when only 1070 cases were expected . A third secondary cluster was located in the southern part of the country in municipalities in the provinces of El Oro , Loja Zamora , Cañar , Azuay , Morona Santiago , and Chimborazo . This cluster was the biggest cluster found , and the iRR was 1 . 60 ( p = 0 . 001 ) compared to the municipalities outside the cluster . This cluster existed from 2002 to 2007 . Finally , in the western coast of the country , a secondary cluster covering 11 municipalities of Manabí province was detected in 2008 , which had an iRR of 1 . 62 ( p = 0 . 001 ) . Like for epilepsy , four significant clusters were identified for NCC . A main cluster was found in the southern part of the country , which existed from 1996 to 2001 with an iRR of 3 . 78 ( p = 0 . 001 ) ( Fig 6 ) . Municipalities of El Oro , Loja , Zamora , Azuay and Morona Santiago provinces were part of this cluster within the given time frame . Likewise , in the central-northern part of the country , some municipalities from Imbabura , and Pichincha provinces were part of an important secondary cluster that lasted from 1996 to 2001 . The iRR for the municipalities within this cluster was 2 . 75 ( p = 0 . 001 ) . Additionally , there was a secondary cluster in the municipality of Riobamba in Chimborazo Province in the middle of the country that lasted between 2000 and 2005 . The iRR for this cluster was 3 . 15 ( p = 0 . 001 ) . Finally , besides the last zone , some municipalities from Cotopaxi and Tungurahua provinces were part of an additional cluster that had an iRR of 2 . 09 ( p = 0 . 001 ) and had a relatively high incidence starting in 1996 until 1997 . The analysis of several socio-economic and demographic variables affecting the IHC of epilepsy between 1996 and 2008 is presented in Tables 3 and 4 . Table 3 presents the set of covariates selected by stepwise procedure and Table 4 presents the total set of variables included in this set of potential indicators and their partial contribution according to the BMA methodology . Posterior effect probabilities ( P ( β≠ 0\|D ) ) were included to show the evidence of an effect for each covariate when model uncertainty was incorporated . According to the results , the variables that significantly increased the IHC of epilepsy in municipalities were: the number of physicians per 10 , 000 inhabitants ( PHYS ) iRR = 1 . 045 ( IC95%: 1 . 031–1 . 059 ) , and the percentage of families having piped water ( TUBWAT , not necessarily drinking water ) iRR = 1 . 022 ( IC95%: 1 . 014–1 . 029 ) . On the other hand , the only variable that apparently led to a significant reduction in the IHC of epilepsy in municipalities was the percentage of houses having any kind of system to eliminate excrements ( EXCR ) iRR = 0 . 986 ( IC95%: 0 . 979–0 . 993 ) . The climatic variable ( Zone ) showed that municipalities located in the highlands presented an increased risk of epilepsy in contrast to municipalities located in tropical zones ( iRR = 1 . 02 ( IC95%: 0 . 855–1 . 207 ) ) even though it was not statistically significant at a 5% level . For the BMA variable selection , out of 15 variables evaluated , 3 had substantial evidence ( P ( β≠ 0\|D ) >90% ) of being different from zero ( Table 5 ) ; PHYS , TUBWAT , and EXCR . Overall , the BMA and stepwise selection methods yielded similar results . Table 5 presents the results of the potential indicator factors associated with NCC in hospitals . In the binary part , three covariates were chosen ( EXCR , PIG , %RURAL ) . According to this selection procedure , the covariates that positively influenced the odds of hospitalized NCC cases in the communities were the implementation of systems for eliminating excrements ( EXCR ) ( OR = 0 . 94; IC95%: 0 . 89–1 . 0 ) , and pig population ( PIG ) ( OR = 0 . 999; IC95%: 0 . 999–1 . 0 ) . In contrast , the higher the proportion of rural population in a community the lower the odds ratio of reporting NCC hospitalized cases ( OR = 1 . 073 ( IC95%: 1 . 01–1 . 14 ) ) . On the other hand , for the count model , the variables that positively influenced the number of hospitalized cases were SCHOOL , TUBWAT , TECHASSIS , and %GLS , and the variables that were associated with a decrease in the number of hospitalized cases were Zone ( Tropical ) , EXCR , EXTPOOV , and %RURAL . The temperate zone ( highlands ) had by far higher NCC cases compared to the tropical zones , so that the possibility of having a NCC diagnosis in Andean zone hospitals is higher ( iRR = 5 . 6 times ( IC95%: 3 . 95–7 . 92 ) ) . Variables such as , the schooling in municipalities ( SCHOOL ) ( iRR = 1 . 212; IC95%: 1 . 012–1 . 451 ) , the percentage of dwellings having piped water ( TUBWAT ) , ( iRR = 1 . 020;IC95%: 1 . 001–1 . 031 ) , the proportion of farms with technical assistance ( TECHASSIST ) ( iRR = 1 . 051; IC95%: 1 . 022–1 . 084 ) , and the percentage of land dedicated to pastures ( %GLS ) ( iRR = 1 . 010; IC95%: 1 . 002–1 . 019 ) were positively associated with an increase in the IHC of NCC . In contrast , covariates that negatively affected the number of NCC hospitalized cases in hospitals were: the percentage of dwellings having any system for eliminating excrements ( EXCR ) ( iRR = 0 . 951; 0 . 934–0 . 964 ) , extreme poverty ( EXTPOOV ) ( iRR = 0 . 980; IC95%: 0 . 965–0 . 997 ) ; and %RURAL population ( iRR = 0 . 990; IC95%: 0 . 9812–0 . 999 ) . The implementation of systems for eliminating excrements ( EXCR ) reduced the IHC of NCC in average by 4 . 87% ( IC95%: 3 . 6%–6 . 6% ) . The model chosen for the iRR of epilepsy fitted better than the model without over-dispersion . The likelihood-ratio test for the over-dispersion parameter was significantly different from zero ( p<0 . 001 ) , meaning that Negative binomial regression was preferred over Poisson regression . For the case of NCC , when the zero-inflated negative binomial model was contrasted against the zero-inflated Poisson model , the likelihood-ratio test was significant in favour of taking into account the overdispersion ( different from zero ) ( p<0 . 0001 ) . In the same way , Vuong’s test confirmed that the assumption of zero inflation in the model for NCC was preferred over a model without this assumption ( p = 0 . 02 ) . In the case of epilepsy , according to the spatio-temporal analysis , all significant clusters of municipalities with high incidence of epilepsy existed between 2002 and 2008 . The incidence risk ratio ( iRR ) of the epilepsy clusters were not so high ( they ranged from 1 . 57 to 1 . 69 ) , meaning that epilepsy can be almost classified as an endemic disorder in Ecuador . A possible explanation for the epilepsy clusters since 2002 is that during these years , health facilities could have improved in municipalities leading to a better coverage of hospitalization . Clustered epilepsy municipalities were located in the Sierra region , mainly in the southern and some in the central-northern part of the country . Some of those municipalities were known as endemic for epilepsy and as epilepsy-NCC related zones , with some new clusters in the Sierra region that have not been pointed out before as endemic areas in previous studies [21;23;24;45] . The appearance of this new epilepsy cluster could also be related to the presence of other infectious and non-infectious diseases present in coastal tropical zones [46] , but the evaluation of the newer clusters become urgent as this could be a strong predictor for NCC [4] . Based on two selection methods , there was a positive association between number of physicians and number of hospitalized cases of epilepsy . This effect occurs mainly in provincial capital cities where more health services exist and consequently specialized physicians are available to people . The implementation of systems for eliminating excrements , such as latrines , septic tanks , or sewage systems , seemed to have an impact on the occurrence of epilepsy , which suggests that a part of hospitalized epilepsy cases might be due to NCC or other fecal-related causes of epilepsy [47] . In this study period , the majority of the epilepsy cases were classified as non-specific cases ( ICD codes: G408 and G409 ) , and they were between 81% and 92% every year . As in other places , the majority of epilepsy cases apparently do not have a recognized etiological origin well identified [16;17] . However , not all types of acquired epilepsy can be attributed to NCC [4;16;47] . Other causes related to poverty , such as poor nutrition , neurological sequels due to infectious agents , and head traumas related to occupational accidents and violence can also contribute to the number of hospitalized epileptic cases [47–50] . The other variable was the percentage of dwellings with piped water . Having piped water not necessarily refers to water that is drinkable . According to estimations , in Ecuador , 30% of the piped water in urban zones is not potable water . [51] . In Ecuador , the protocol to declare a patient with NCC depends on the diagnostic capacity of the neurology department , which in many cases , only exists in the main cities [52] . Based on national hospital data , 35 . 5% of the NCC cases were reported by the public sector , 39 . 4% were reported by the private sector and 23% by the systems of social security insurance . Furthermore , the appearance of clinical signs of NCC in patients might occur several months or even years after infection , so that some etiological factors could not have been measured correctly at the beginning of this study . Likewise , it is worth to mention that only a part of the cases of human cysticercosis are symptomatic , and therefore the statistical relationships found in this study are valid only for the hospitalized cases reported . A wider spectrum might be found with the addition of asymptomatic cases and people suffering from chronic headaches who do not often consult physicians , which only can be found in field studies[4;53–56] . NCC clusters in Ecuador apparently appeared earlier compared to epilepsy clusters and the majority of them existed between 1996 and 2001 . Only one of them was identified from 2000 to 2005 in Chimborazo province and in some surrounding municipalities in Bolívar Province . This area presents ideal conditions for the T . solium life cycle , as it is located in the highlands with a high percentage of rural population and lack of basic services . In this cluster , more than 90% of the pig producers are smallholders [14] . According to the last Agricultural National Census carried out in 2000 , 80 . 5% of the smallholders ( ≤10 pigs ) are located in the Sierra region and 18 . 6% in the Coastal region . Porcine Cysticercosis has not been reported in slaughterhouses by the National Veterinary Services ( Agrocalidad ) since 2001 [57] . However , only 30% of slaughtered pigs in these facilities is provided by smallholder . [58] . Based on a zero-inflated negative binomial model , the percentage of rural population in municipalities was associated with a reduction in the IHC of NCC; so that urban zones increased their incidence in contrast with previous studies published [59] . Nowadays , the rural population in Ecuador accounts for less than 38% of the total population , which is a significant reduction compared to previous decades when the rural population was over 50% . Translocated rural communities tend to settle in slum zones of big cities [60] . However , rural and sometimes poor communities might have not been well represented in the data , as they may refrain from hospitalization due to the high costs of diagnosis and treatments [61] . This concern is a limitation of hospital-based registers [50] . In our study only 7% of patients appeared to belong to rural communities , although , if we consider the patients not living in the provincial capital cities this percentage increases to 26 . 7% . The structure of the data makes it difficult to differentiate people from peri-urban zones or slums , or semi-rural towns and communities from urbanized areas . On the other hand , housemaids and food vendors coming from endemic rural zones have a higher chance to be tapeworm carriers and can be at the origin of spreading T . solium infection among the urban population [18;62;63] . Another possible explanation is that traditional livestock systems are still preserved on a small scale in urban slums , although this presumption has not been quantified . The presence of pigs was the most important positively associated with the appearance of hospitalized-symptomatic NCC cases . Industrialization of pig production , in many cases is not responsible for the increase in NCC cases . On the other hand , the presence of free roaming pigs has been associated with an increased risk for the occurrence of cysticercosis [24;64] . In Ecuador , despite 58 . 8% of pigs are raised in traditional production systems it has been estimated , that it only represents nearly 30% ( 50% in year 2000 ) of pork available in markets . [58] . In the negative binomial count model , eight variables were associated with the IHC of NCC . As in the case of epilepsy , the implementation of systems for eliminating excrements was involved in a reduction in the IHC of NCC . More in depth studies are needed to evaluate the real scale of those variables in the macro epidemiology of NCC , although some of them express the lack of quality in offering services . It has been mentioned that the difficulties of identifying the etiology of epilepsy could play an important role in the sub-notification of NCC [21;65;66] . The condition most commonly associated with NCC is epilepsy , but many cases of NCC are asymptomatic or manifest chronic headaches or other neurological disorders [4;54–56] . In T . solium endemic communities in Ecuador an important proportion of acquired epilepsy cases were due to NCC [21;45] . Although extrapolating this quantity to the current reality in the country may be biased; zones with an apparent increase of epilepsy cases may elucidate the origin of new suspected NCC cases [2;4] . Additionally , epilepsy and NCC in developing countries have been reported to be clustered [1;50;67] , but the presence of imported cases has also been mentioned as an important factor in urban zones . NCC underreporting might be due to a misdiagnosis in the epilepsy etiology . However , in our case both disorders were linearly related . This positive relationship is an indicator of an apparent constant relationship between epilepsy and NCC . This relationship has to be further studied and the meaning of this pattern has to be elucidated . Additionally , given that the lack of sewage systems was demonstrated to be associated with an increase in the incidence of both conditions . Increasing the sewage systems could be used as an important control tool to reduce the incidence of the hospitalized cases . The installation of these systems , at municipal level varied from 20 . 6% to 96 . 3% of coverage with a median value of 68 . 5% , so there are still many municipalities that lack basic services . The zone where the municipality was located was one of the principal indicators affecting the IHC of NCC and epilepsy . Municipalities located in temperate zones ( highlands ) had a significantly higher number of hospitalized NCC cases . From the BMA procedure , in the case of epilepsy , the posterior effect had a small probability ( 11 . 8% ) , so we argue that in tropical zones lack of appropriate diagnostic tools and specialized knowledge of health staff might make it difficult to properly identify NCC cases [48] . Likewise , the levels of coverage of basic services is lower in tropical zones of Ecuador [14] . Thus , the presence of the life cycle of T . solium in these zones traditionally considered to be NCC free cannot be ruled out . A limitation of this study was that peri-urban zones where poverty belts of cities are frequently located could not be analyzed separately given that records do not use this residence category for patients . The conditions in peripheral zones can differ from city to city , thus the assumption that the origin of the hospitalized cases in big cities come from peripheral zones should not be extrapolated in all cases . As in other studies based on hospital data , rural communities might not be appropriately represented in the sample . This could be a major limitation of our study , and also because of the asymptomatic human cysticercosis cases , migraine-type and chronic headaches [4;54;56] . However , due to the fact that ambulatory cases do not offer reliable data , hospital data is a better attempt to represent the situation of the disease in a municipality . Another limitation ins this study might be the presence of duplicated cases in the data base . These cases might be due to the fact that a patient was hospitalized more than once , or because some NCC cases were diagnosed as epilepsy before . But given the mentioned limitations , the present results are still reliable due to their apparent representation of the municipalities in Ecuador , and because they are based on the appropriate statistical tools . So given the data constraints , the methods used to identify risk indicators and/or areas based on available data presents valuable results for veterinary and public health sectors at no cost . There is a need to re-evaluate the current situation for both disorders throughout the country as life conditions have been changing over time [26;30;68] . Given the recent changes in the organization of the public health sector , new trends need enough data collection-time to be evaluated again . In conclusion , NCC might still have a relevant presence in Ecuador and might play an important role as a cause of acquired epilepsy in Ecuador [45] . Although the real burden of NCC is still unknown , we found that the hospitalization rate of patients with epilepsy has been increasing in recent years ( Fig 4 ) . Traditional NCC and epileptic endemic zones were recognized as high risk zones even though more recent clusters of both diseases seem to have appeared . Although the lack quality of basic services was related to the IHC in both disorders , one important finding of this study was that the implementation of systems for eliminating excrements helped to reduce the incidence of hospitalized cases of both epilepsy and NCC , which could be used as an indicator strategy for planning control programs . More specific studies linking human NCC with epilepsy and their respective factors in field conditions are needed to evaluate the prevalence of the disease in humans throughout the country and generate data that could be used for estimation of the burden disease .	T . solium neurocysticercosis is considered the most important parasitic disease of the central nervous system in humans; it is estimated to be responsible for at least one third of acquired epilepsies in developing countries . In Ecuador , the relationship between acquired epilepsy and neurocysticercosis remains unclear due to different factors such as , the lack of specialized health care personnel , appropriate diagnostic techniques and the fact that acquired epilepsy is characteristic of many other infectious and non-infectious diseases in the endemic zones of the country . In this study , spatio-temporal information and potential socio-economic indicators were studied for the number of hospitalized neurocysticercosis and epileptic cases in the country in order to locate and characterize important clusters in space and time . This study identified traditional endemic clusters in the highlands for both conditions as well as new clusters appearing in recent years in other zones not considered endemic . Also the incidence of hospitalized cases of epilepsy and neurocysticercosis were significantly higher in urban zones , probably due to a better access to health facilities . The presence of systems for excrement disposal was significantly associated with a reduction in the incident cases for both epilepsy and neurocysticercosis . More studies are needed to evaluate the true prevalence of neurocysticercosis associated epilepsy in humans and cysticercosis in pigs around the country in order to better implement and manage control campaigns .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[]	2015	Distribution and Potential Indicators of Hospitalized Cases of Neurocysticercosis and Epilepsy in Ecuador from 1996 to 2008
Infantile beriberi ( thiamine deficiency ) occurs mainly in infants breastfed by mothers with inadequate intake of thiamine , typically among vulnerable populations . We describe possible and probable cases of infantile thiamine deficiency in northern Laos . Three surveys were conducted in Luang Namtha Province . First , we performed a retrospective survey of all infants with a diagnosis of thiamine deficiency admitted to the 5 hospitals in the province ( 2007–2009 ) . Second , we prospectively recorded all infants with cardiac failure at Luang Namtha Hospital . Third , we further investigated all mothers with infants ( 1–6 months ) living in 22 villages of the thiamine deficiency patients’ origin . We performed a cross-sectional survey of all mothers and infants using a pre-tested questionnaire , physical examination and squat test . Infant mortality was estimated by verbal autopsy . From March to June 2010 , four suspected infants with thiamine deficiency were admitted to Luang Namtha Provincial hospital . All recovered after parenteral thiamine injection . Between 2007 and 2009 , 54 infants with possible/probable thiamine deficiency were diagnosed with acute severe cardiac failure , 49 ( 90 . 2% ) were cured after parenteral thiamine; three died ( 5 . 6% ) . In the 22 villages , of 468 live born infants , 50 ( 10 . 6% , 95% CI: 8 . 0–13 . 8 ) died during the first year . A peak of mortality ( 36 deaths ) was reported between 1 and 3 months . Verbal autopsy suggested that 17 deaths ( 3 . 6% ) were due to suspected infantile thiamine deficiency . Of 127 mothers , 60 ( 47 . 2% ) reported edema and paresthesia as well as a positive squat test during pregnancy; 125 ( 98 . 4% ) respected post-partum food avoidance and all ate polished rice . Of 127 infants , 2 ( 1 . 6% ) had probable thiamine deficiency , and 8 ( 6 . 8% ) possible thiamine deficiency . Thiamine deficiency may be a major cause of infant mortality among ethnic groups in northern Laos . Mothers’ and children’s symptoms are compatible with thiamine deficiency . The severity of this nutritional situation requires urgent attention in Laos . Thiamine ( vitamin B1 ) acts as an important cofactor in metabolism and energy production . It is required for the biosynthesis of neurotransmitters and the production of substances used in defence against oxidant stress [1] . Thiamine deficiency occurs predominantly in populations , in which the diet consists mainly of very poor sources of thiamine such as milled white cereals , including polished rice ( the rich thiamin envelop is removed by polishing ) and wheat flour , and where other key sources of thiamine ( meat , fish , and vegetables ) are infrequently consumed [2] . It is also related to diets that are rich in thiaminase , the natural thiamine-degrading enzyme , which is abundantly present in raw and fermented fish sauce ( a common Asian delicacy ) certain vegetables and roasted insects consumed primarily in Africa and Asia [3] . Thiamine deficiency can develop within 2–3 months from a deficient intake and can cause illness and death [4] . Clinically apparent thiamine deficiency , also known as Beriberi , has historically been described in vulnerable populations such as refugees , prisoners , during times of war [2 , 5] and in developed countries with alcohol abuse or parenteral nutrition with insufficient thiamine intake in adults [6 , 7] . More recently , thiamine deficiency outbreaks were described among young healthy Thai construction workers in Singapore in the 1980s [8 , 9] and among commercial fishermen in Thailand in 2005 [10 , 11] . Worldwide , outbreaks of thiamine deficiency have recently been reported in Ivory Cost jails [10 , 12] , in the Gambia [13 , 14] , among African Union troops in Mogadishu , Somalia [15] , in Brazil [15–18] and the French island of Mayotte [19] . Adult thiamine deficiency was also recently described in Cambodia , China , India , Thailand , Laos and various countries with unbalanced thiamine/thiaminase diets [8 , 19–24] . Cardiac failure associated with thiamine deficiency has also been described in Japanese teenagers consuming excessive sweet carbonated soft drinks , instant noodles and polished rice [25] . Thiamine deficiency , is rarely seen today in infants after decades of strong public health attention [26] . It is an acute disease mainly affecting infants that are breastfed by women with deficient thiamine levels [7] . The onset of symptoms is often very rapid and the fatality rate is very high with death often occurring within a few days from the onset of symptoms . Historically , the clinical features have been categorized into three main types; the pure cardiac form or wet thiamine deficiency , the aphonic form , and the neurologic or dry form [27] . The more severe form is called Shoshin beriberi and presents as cardiac failure and lactic acidosis [28] . Beriberi poses difficult diagnostic issues and can be a missed diagnosis , as the dry or wet forms can mimic critical illness or polyneuropathies . In addition , clinical manifestations such as tachypnea , chest indrawing , tachycardia and cardiomegaly can suggest other diagnoses [29–31] [32] . The content of thiamine in breastmilk is related to the mother’s thiamine status [7] . Post partum thiamine deficiency in refugee mothers was associated with high infant mortality in Karen refugees [2 , 5] . The overall infant mortality rate declined from 183 before the recognition of thiamine deficiency to 78 per 1000 live births afterwards . Thiamine deficiency has also been described in breastfeeding Cambodian and Lao mothers [20 , 33] , and pregnant mothers in China [24] . Recent surveys using whole blood thiamine diphosphate ( TDP ) revealed that thiamine deficiency was associated with cardiac dysfunction and tachypnea in Cambodian infants [32 , 34] . In developed countries infantile thiamine deficiency outbreaks have recently been described . For example in Israel an outbreak was due to thiamine deficient soya formula , with a high fatality rate and long term sequelae [35] . In the French island of Mayotte , deficiency was related to inadequate nutrition [28 , 36] . Infantile thiamine deficiency is periodically reported in intensive care units in babies receiving parenteral nutrition without thiamine or babies with malabsorption receiving prolonged but inadequate vitamin supplements [6 , 28 , 37 , 38] . Infantile thiamine deficiency was described in Laos in the sixties [39] . More recently , cases of young infants with cardiac failure in Mahosot Hospital , Vientiane , suggested the persistence of thiamine deficiency as a cause of infantile mortality [40] . Traditional food avoidance during the post-partum period , nutritional habits , and the high rate of childhood stunting ( 40% ) may all be related to thiamine deficiency [33] . A recent publication revealed that clinically unapparent thiamine deficiency was common among sick infants without overt clinical thiamine deficiency admitted in 2003–2004 [41] . Alarming reports have been received from physicians about the possibility of thiamine deficiency in infants with cardiac failure in northern Laos in recent years [23 , 41] . However , there is insufficient data from outside Vientiane to provide evidence for discussions about thiamine supplementation in the Lao national nutrition strategy . To help fill this gap , we describe possible and probable cases of infant and maternal thiamine deficiency in Luang Namtha province . Luang Namtha province is located in the northwest of the country , bordering Myanmar and China . It is one of the country’s poorest areas with a population of 148 , 797 people , many of whom live in remote mountain villages . Approximately 23 . 4% of women had at least one ante natal care visit during their most recent pregnancy in the province [32 , 42] . The infant mortality in the province ( 112 per 1000 live births ) is one of the highest in the country ( national rate 70 per 1000 live births ) according to the 2005 national census [43] . At the time of this study ( 2008 ) the national infant mortality rate was estimated at 48 per 1000 births [44] . The provincial hospital in Namtha district is the referral centre for all 5 districts and the military hospital . It is a 50-bed hospital with 98 medical and non-medical staff in 2007 . Approximately 1 , 400 outpatients are seen each month and 330 are admitted . It was the only health facility in the province with X-ray and surgery . The 4 other district hospitals are Long , Sing , ViengPhoukha and Nalea . We conducted various surveys in Luang Namtha province as shown in the flow chart ( Fig . 1 ) . First , in a retrospective survey we recorded all infant inpatients with recorded thiamine deficiency in the 5 hospitals in the province between 2007 and 2009 . Second , in a prospective survey we recorded all infants admitted with cardiac failure at the emergency ward of Luang Namtha hospital from March to June 2010 . Third , from these two surveys we identified 22 villages from where the patients originated and then investigated all mothers with infants ( 1–6 months ) living in these villages . We conducted i ) a verbal autopsy of all infants’ deaths and estimated infantile mortality; ii ) a cross sectional survey of all mothers and infants ( 1–6 months ) using a pre-tested questionnaire , physical examination and squat test ( defined by “the inability of the individual to rise from a squatting position , due to weakness then flaccid paralysis of the lower limbs , without assistance” ) [7 , 12] . Characteristics of infants with a discharge diagnosis of thiamine deficiency admitted at the hospitals were retrieved from hospital records . A standardized form was used that included age of infant , main symptoms , treatment received , and response to treatment . An 80 item questionnaire was used in the villages . It included general information on the population ( 8 questions ) , presence of a rice-mill , type of rice consumed , socio-economic characteristics ( 32 questions ) , maternal food avoidance behaviour , food given to the child in the previous 15 days and information on the children , age by day , sex , mode of birth and detailed information regarding the causes of infants’ death ( 40 questions ) . We used definitions of possible or probable thiamine deficiency for mothers and for infants with sudden cardiac failure or death , based on symptoms and response to thiamine treatment [36] . Possible adult thiamine deficiency was defined in a pregnant women or a mother with a child less than 6 months if she presented or had presented during her pregnancy with at least two of the following signs: motor deficits , paresthesia of the limbs ( peripheral numbness , tingling or plantar pain ) , loss of reflexes , signs of heart failure ( jugular swing , cardiac gallop rhythm on auscultation , hepatomegaly ) , associated with a positive squat test ( unable to rise after squatting ) [2] . Possible thiamine deficiency in infants was defined as acute symptoms in previously healthy breastfeeding infants associated with cardiac failure ( tachypnea> 50/min , tachycardia> 170/min , gallop , hepatomegaly> 3 finger's breadth ) or loss of voice . Probable thiamine deficiency was defined if symptoms recovered after thiamine treatment . Death was diagnosed as due to thiamine deficiency for previously healthy breastfed infants with less than two days of illness fulfilling the possible or probable thiamine deficiency definitions above , and as probable if associated with mother’s symptoms of thiamine deficiency . Due to possible misdiagnosis with acute pneumonia ( though infection can precipitate thiamine deficiency [40] ) children with signs of pneumonia ( cough , fever , +- dyspnea ) were excluded if no signs of thiamine deficiency were present in the mother . The final verbal autopsy diagnosis was proposed during a review meeting of all cases by a committee including one pediatrician , one public health advisor , and two physicians . Only consensual diagnoses were retained . Mothers suspected of thiamine deficiency were treated with vitamin B1 tablets 100mg , twice daily for 20 days and infants with suspected thiamine deficiency were treated with vitamin B1 tablets , 30 mg per day for 20 days . Patients with acute symptomatic thiamine deficiency received an intramuscular or slow intravenous injection of thiamine ( 100mg IM for mothers and 50mg for infants ) . Treatment for associated infection , if any suspected , was provided appropriately . Treatment was provided free of charge . Moreover , all families and village populations received information regarding thiamine deficiency prevention . Data was entered into EpiData freeware . All records were crosschecked with the original data sheets . Analysis was carried out with STATA , Version 8 ( Stata Corporation , College Station , TX , USA ) . Chi-squared , Fisher’s exact tests and Student’s t-test were used to compare categorical variables and continuous data , respectively . 95% confidence intervals were calculated for continuous and categorical data . We considered p < 0 . 05 as statistically significant . Infant mortality rates were calculated and compared to the national rate in Laos at the time of survey [44] . The study was authorized by the Lao health authorities . Information about the study was provided in Lao language and translated into the local ethnic language by one volunteer from each village . All participants gave informed oral consent in the presence of one village witness as the majority could not read . The procedure of the study was granted ethical approval by the Lao Medical Ethics Committee . Between 2007 and 2009 , 54 infants with sudden onset of cardiac failure were admitted to the 5 hospitals of Luang Namtha province . This number increased from 9 to 24 per year . The infant clinical characteristics and treatment evolution are presented in Table 1 . Among them , 20 ( 37% ) probably had an associated infection which may have triggered cardiac failure . Of the 54 infants with cardiac failure , 49 ( 90 . 7% ) were cured after thiamine administration , three died ( 5 . 6% ) and two had an unknown status ( 3 . 7% ) . Time of cure was not recorded in patients’ files . Four infants with clinical thiamine deficiency were observed during the prospective study at the provincial hospital from March to June 2010 . All recovered after thiamine administration and were visited in their own village later . We present a typical description of one patient ( Box 1 ) . Of 22 villages visited , 18 ( 81 . 8% ) had an electric rice mill . Of a total of 167 mothers with an infant less than 6 months , 127 mothers and their infants were present and gave consent to be interviewed and undergo a physical examination ( Table 2 ) . All mothers consumed polished white rice , 36 ( 28 . 3% ) had at least one antenatal visit and 28 ( 22% ) reported they received some information on nutrition from health staff during antenatal care . Less than half of the children had received some immunizations ( 60 , 47 . 2% ) ( Table 3 ) . Nearly all mothers ( 125 , 98 . 4% ) respected food avoidance after delivery with a median of 30 days . A third of the mothers ( 45 , 35 . 4% ) reported to have had at least one of their children die . Of 468 live born infants , 50 ( 10 . 6% , 95%CI: 8 . 0–13 . 8 ) infants died during the first year . Based on this survey , the infant mortality rate was 106 per 1000 live births ( 95% CI: 86–128 ) . Thirty-six infants ( 7 . 6% ) died below the age of 6 months . According to mothers , 22 . 8% of infant deaths occurred during the neonatal period while 29 . 5% and 23 . 8% of deaths occurred during the second and third month respectively , and dropped to 11 . 8% during the fourth and fifth months which suggested a plateau of infantile mortality during the first 3 months of life ( Fig . 2 ) . Twenty ( 10 . 6% ) children presented with sudden death compatible with thiamine deficiency . The verbal autopsy suggested that 17 ( 3 . 6% ) infants died of thiamine deficiency , 13 ( 2 . 7% ) as probable and 4 ( 0 . 8% ) as possible thiamine deficiency . Loss of voice was reported in 10/17 ( 58 . 8% ) . A typical patient is presented in Box 2 . For the remaining deaths the verbal autopsy suggested other pathologies ( meningitis , laryngitis , convulsions , neonatal infection ) as probable causes . Women in Laos should be educated about the importance of a diverse diet before and after delivery and how to maintain a sufficient thiamine intake . Pregnant and lactating mothers must be encouraged to eat unpolished rice , prepare their rice avoiding loss of micronutrients by avoiding unnecessarily long soaking , avoid fermented fish paste , and betel nut chewing during pregnancy and breastfeeding periods . Culturally acceptable ways need to be identified to limit postpartum food avoidance . These measures might be challenging in these populations; hence daily thiamin supplements which are affordable could be considered . Health professionals should provide nutritional advices during the precious time of the antenatal visits and should be trained to recognize , prevent and treat early symptoms of thiamine deficiency and to offer thiamine supplements to mothers who are not able to systematically comply with dietary advice . For many years , Xayaboury province ( north-western Laos ) has included thiamine supplementation in the prenatal care program [50] . The midwife in charge of the program reported that thiamine deficiency cases are very uncommon in the province ( Leila Srour , personal communication ) . Educational campaigns which are now focusing on implementing exclusive breastfeeding in Laos must include thiamine deficiency prevention and detection , as an important component of these educational campaigns . The story of a mother in Savannakhet province helps to understand the context . Attending the district hospital in 2004 with a severely malnourished infant ( 6 kg at 1 year ) she was asked why she fed the infant with coffee creamer . She reported that her first four infants had died before 6 months of age . Suspecting her breastmilk being the problem she decided to feed her fifth child with coffee creamer , as she could not afford expensive infant formula ( Leila Srour , personal communication ) . Further research is needed to evaluate which preventive strategies are the most effective to reach mothers living in remote villages . Determination of thiamine concentration in breastmilk and the infants’ thiamine status are still needed [40] . Further research is also needed to assess and prevent the hidden consequences of infant thiamine deficiency , especially neurological development and epilepsy [51 , 52] . A concomitant concern is the need for more accessible and inexpensive tests to evaluate thiamine deficiency , as the current basal erythrocyte transketolase activity ( ETK ) assays remains unavailable in settings where they are most needed [39] . Another concern is to clarify which tests are most useful . Recently the less conventional whole blood thiamine diphosphate ( TDP ) concentrations have been used in the field to assess thiamine deficiency [20 , 34] . This survey has several limitations , including the use of retrospective hospital data and the confounding factor of associated antibiotics together with thiamine administration , the possible recall bias while interviewing mothers regarding the subjective nature of signs such as paresthesia , the lack of laboratory testing for thiamine deficiency , and the use of clinical criteria only to assess thiamine deficiency [2] . Due to time and budget constraints the team could not follow treatment efficacy in the villages and could not validate cases . We adapted our definitions of thiamine deficiency cases from 2 surveys: one hospital-based survey and one epidemiological outbreak survey [36 , 40] . Our hospital case definition did not exclude the presence of fever or suspected sepsis since there is evidence that these conditions contribute to precipitating thiamine deficiency [7 , 41 , 47] . This may have overestimated the thiamine deficiency frequency but the response to thiamine treatment , an important criterion for diagnosis , was positive in all but three . Conversely , retrospective case review did not include the presence of fever or signs of sepsis and we may have underestimated the number of true thiamine deficiency cases in these high risk populations . Finally , we screened 22 villages with suspected thiamine deficiency cases but this strategy cannot provide a representative overview of the situation of thiamine deficiency in the region . This survey suggests that thiamine deficiency is a major cause of high infant mortality among ethnic groups in northern Laos . Prevention of thiamine deficiency and nutritional education should be addressed on a larger population scale , particularly for pregnant and breastfeeding women , their offspring and their families . It should also focus on at risk Asian populations reporting similar low diversity diets , low thiamine intake , thiaminase rich diets and food avoidance during and after pregnancy .	Infantile thiamine deficiency ( beriberi ) , is rarely seen today after decades of strong public health attention . Infantile beriberi occurs mainly in infants breastfed by mothers with inadequate intake of thiamine . There is evidence of the persistence of infantile thiamine deficiency in Vientiane , the capital of Laos , but insufficient data from outside Vientiane to justify a policy recommendation . We describe possible and probable cases of infantile thiamine deficiency in northern Laos using retrospective and prospective hospital data . In addition we conducted a cross sectional survey in 22 villages where the infants originated . Infantile thiamine deficiency was quite common in retrospective and prospective ( hospitals: 54 , villages: 17 ) and cross-sectional surveys ( hospital: 4 , villages: 10 ) . A second peak of infantile mortality was observed before 6 months and was associated with a high infant mortality in the villages , 106 per 1000 live births ( 95%CI: 86–128 ) . A total of 60 pregnant mothers and 70 lactating women showed signs of thiamine deficiency . This situation requires urgent attention in Laos .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[]	2015	Beriberi (Thiamine Deficiency) and High Infant Mortality in Northern Laos
Paracoccidioides lutzii is a new agent of paracoccidioidomycosis ( PCM ) and has its epicenter localized to the Central-West region of Brazil . Serological diagnosis of PCM caused by P . lutzii has not been established . This study aimed to develop new antigenic preparations from P . lutzii and to apply them in serological techniques to improve the diagnosis of PCM due to P . lutzii . Paracoccidioides lutzii exoantigens , cell free antigen ( CFA ) , and a TCA-precipitated antigen were evaluated in immunodiffusion ( ID ) tests using a total of 89 patient sera from the Central-West region of Brazil . Seventy-two sera were defined as reactive for P . brasiliensis using traditional antigens ( AgPbB339 and gp43 ) . Non-reactive sera for traditional antigens ( n = 17 ) were tested with different P . lutzii preparations and P . lutzii CFA showed 100% reactivity . ELISA was found to be a very useful test to titer anti-P . lutzii antibodies using P . lutzii-CFA preparations . Sera from patients with PCM due to P . lutzii presented with higher antibody titers than PCM due to P . brasiliensis and heterologous sera . In western blot , sera from patients with PCM due to P . lutzii were able to recognize antigenic molecules from the P . lutzii-CFA antigen , but sera from patients with PCM due to P . brasiliensis could not recognize any P . lutzii molecules . Due to the facility of preparing P . lutzii CFA antigens we recommend its use in immunodiffusion tests for the diagnosis of PCM due to P . lutzii . ELISA and western blot can be used as complementary tests . Paracoccidioidomycosis ( PCM ) is a mycotic disease caused by species of the genus Paracoccidioides , a group of thermally dimorphic fungi that grow in mycelial form at room temperature and as budding yeasts when cultured at 37°C or in parasitism in host tissues . PCM is limited to Latin American countries , and the most important regions of endemicity are found in Brazil , Colombia , and Venezuela [1] . PCM presents as two major clinical forms: the acute or sub-acute form and the chronic form . In Brazil , PCM is considered the eighth most common cause of death among infectious and parasitic chronic diseases , with a mortality rate of 1 . 45 per million population [2] . However , PCM may be considered a neglected disease because very few regions of this country have official prevention programs . The criteria for a definitive diagnosis of PCM are based on demonstrating the presence of the fungus as multiple budding cells in clinical materials . A culture is relatively difficult to obtain , and it is a slow procedure . As an adjunct to clinical and histological findings , serological tests can help establish a diagnosis . The detection of antibodies in serum has been one of the main tools for the diagnosis of PCM and may be useful for monitoring its evolution and response to treatment . Among the different serological techniques , the double immunodiffusion ( ID ) test is the most commonly used and has a sensitivity of 80 to 95% [3] . Enzyme-linked immunosorbent assay ( ELISA ) has been employed in the serology of various mycotic diseases [4]–[8] . However , ELISA has important limitations due to cross reactivity . In an attempt to improve the specificity of ELISA for the diagnosis of PCM , Albuquerque and Camargo [9] demonstrated that different procedures are inefficient for eliminating all cross-reacting antibodies and obtaining a specific diagnosis . Exoantigens from cultures of P . brasiliensis were studied by immunoblotting , and components were detected using serum from patients with PCM . Anti-P . brasiliensis IgG reacted with four major components of 70 , 52 , 43 , and 20–21 kDa . The 43 kDa glycoprotein ( gp43 ) was the predominant IgG reactive antigen and recognized by 100% of the patient sera , and the 70 kDa glycoprotein was recognized by 96% of the tested sera . Both gp43 and gp70 can be considered to be markers for human PCM [10] . Until recently , PCM was assumed to be caused solely by P . brasiliensis and transmitted to humans by inhalation of fungal propagules from the mycelia phase occurring in nature . Recent publications [11]–[12] support the idea of several cryptic Paracoccidioides species being phylogenetically related . The real incidence of each phylogenetic species and their implication on ecology and clinical practice is difficult to establish because there is a lack of information in the literature concerning the distribution of these entities . The Phylogenetic Species Recognition ( PSR ) method based on genealogical concordance ( GCPSR ) [13]–[14] has been used to detect limits in many pathogenic fungal species [15]–[17] . Using GCPSR methodology and several isolates of P . brasiliensis from different regions of Latin America , Matute et al . [18] concluded that P . brasiliensis is not a monotypic taxon , but a complex of species: phylogenetic species 1 ( S1 ) , broadly distributed through the Latin America [18]–[20] , phylogenetic species 3 ( PS3 ) is restricted to clinical cases in Colombia [18]–[20] , and a few isolates in the clade phylogenetic species 2 ( PS2 ) have been reported [18]–[20] in Brazil and Venezuela . Phylogenetic analysis of 14 genes in 21 isolates revealed that isolate Pb01 cannot be grouped into any of these species and constitutes a new clade [21] . Seventeen new isolates belonging to this fourth cryptic species ( Pb01-like ) have been identified , 16 of which originate in the Central-West region of Brazil , and one from Ecuador [12] . Thus , Teixeira et al . [22] proposed a new species , Paracoccidioides lutzii , which was formerly known as ‘Pb01-like’ strains based on phylogenetic and comparative genomics data , recombination analysis , and morphological characteristics ( Fig . 1 ) . P . lutzii occurs mainly in the Central-West region of Brazil , but it was also recently described in the northern regions [23] . The Central-West region of Brazil is composed of the states of Mato Grosso , Mato Grosso do Sul , and Goiás , which can be regarded as the areas with highest incidence of PCM caused by P . lutzii . In the South and Southeast regions , P . brasiliensis is the predominant species ( Fig . 2 ) . With the introduction of dissimilar species , PCM serology has been challenging for diagnosis . Several efforts highlight the antigenic variability of this complex , which may have led to a substantial number of false negative results . To date , no studies describing the standardization of P . lutzii antigens for PCM serology are available . In order to improve serological parameters for the diagnosis of PCM caused by P . lutzii , the present study presents a strategy for the immunodiagnosis of PCM caused by P . lutzii by testing three types of antigenic preparations using ID , ELISA , and Western blot . This study was approved by the Research Ethics Committee with protocol number CAAE: 17177613 . 6 . 0000 . 5541 by Federal University of Mato Grosso ( UFMT ) and protocol number 1796-10 by Universidade Federal de São Paulo ( UNIFESP ) . Protocol number of Universidade Federal do Mato Grosso do Sul ( UFMS ) : 354 . 989 . All adult subjects provided informed written consent and the study was approved by ethical committee under number 288 . 250/CEP/HUJM/UFMT . Fourteen isolates of P . lutzii were obtained from different regions of Brazil ( Table 1 ) . Five isolates were obtained from the oral or cutaneous lesions of patients from Júlio Muller University Hospital ( HUJM/UFMT ) in Cuiabá , MT , in the Central-West region of Brazil . The clinical materials were collected by Dr . R . Hahn . Two isolates were obtained in the North region ( Pará state ) by Dr . S . H . Marques-da-Silva [23] . Three isolates were isolated in Goiás state ( Central-West region ) and donated by Dr . C . M . A . Soares , and four isolates were donated by Dr . E . Bagagli . P . lutzii was confirmed by genotyping [12] . Fungal DNA was extracted from suspected colonies and subjected to HSP70 gene amplification using the primers HSPMMT1 ( 5′-AAC CAA CCC CCT CTG TCT TG-3′ ) and PLMMT1 ( 5′-GAA ATG GGT GGC AGT ATG GG-3′ ) as described by Teixeira et al . [12] in order to identify an exclusive indel region of P . lutzii . P . lutzii Pb01 was used as a positive control and P . brasiliensis B339 as a negative control . For the ID study , 89 serum samples from PCM patients ( 75 males and 14 females , age range 23–78 years ) were evaluated: 55 from Mato Grosso state ( Cuiabá and region , Central-West region ) and 34 from Mato Grosso do Sul state ( Campo Grande and region , Central-West region ) . All patients presented with the chronic form of the disease and exhibited clinical and laboratory signs of the disease with pulmonary system involvement and mucosal or mucocutaneous lesions . The diagnosis was endorsed by the clinical experience of the physician responsible for the patient . PCM was confirmed in most patients via direct examination of secretions , such as sputum , oral mucosa lesion samples , and biopsies . However , serological ID tests were not positive for all patients when using the traditional exoantigen from P . brasiliensis B339 ( AgPbB339 standard antigen ) . As a gold standard , we used three sera from PCM patients with positive cultures for P . lutzii ( positive control ) determined by genotyping; sera and isolates were obtained by R . Hahn ) . Heterologous sera from patients with histoplasmosis ( n = 15 ) , sporotrichosis ( n = 15 ) , and aspergillosis ( n = 15 ) were also tested . Finally , 15 serum samples from healthy individuals were also studied ( negative control ) . For ELISA , a batch of sera from Central-Western Brazil was selected . Twenty-eight serum samples from patients with PCM due to P . lutzii ( previously confirmed by ID using P . lutzii antigen ) and 28 serum samples from patients with PCM due to P . brasiliensis from São Paulo , southeast region ( previously confirmed by ID using AgPbB339 ) were used in this assay . Among the sera from patients with PCM caused by P . lutzii , three lacked treatment , 16 had 6 to 12 months of therapy , and 9 had 13 to 18 months of treatment . As a gold standard or positive control , three sera from PCM patients with positive cultures for P . lutzii were used . Heterologous sera and negative controls were also studied . For the Western blot assays , 12 serum samples from patients with PCM due to P . brasiliensis and 12 serum samples from patients with PCM due to P . lutzii ( previously positive by ID and ELISA ) were used . Heterologous sera and negative control sera were also studied . Three millimeters of melted 1% agarose ( Sigma A-6877 ) in PBS was poured onto a glass slide ( 75×25 mm ) . The pattern for this micro-ID test consisted of a central well surrounded by six wells , each 3 mm in diameter . The central well located 6 mm ( edge-to-edge ) from the other wells was filled with the antigen solution . Each slide contained two sets of wells . On each slide the two central wells were filled with 10 µl of antigen . Slides were incubated overnight in a moist chamber at room temperature ( 20–25°C ) , and then washed for 1 h in 5% sodium citrate and for 24–48 h in saline . The slides were dried , stained for 5 min with 0 . 15% Coomassie Brilliant blue ( Sigma ) in ethanol∶ acetic acid∶ water ( 4∶2∶4; v∶v ) , and destained in the solvent mixture alone , when necessary . Precipitation bands were recorded by visual observation [30] . Suspected PCM sera were tested first by ID using the traditional P . brasiliensis exoantigen ( AgPbB339 ) and the purified gp43 antigen from P . brasiliensis B339 in order to discriminate sera from patients with PCM caused by P . brasiliensis . The non-reactive PCM sera , which became the focus of this study , were tested with the three different antigenic preparations . Scheme S1 outlines the final strategy to distinguish sera from patients with PCM caused by P . lutzii and P . brasiliensis . During pilot studies , CFA preparations from different P . lutzii strains were tested and examined by checkerboard titration for antibody detection and P . lutzii strain EPM 208 was chosen for this study . The CFA was used in ELISA at 12 . 5 µg/ml to detect circulating anti-P . lutzii antibodies . We selected 28 serum samples from patients with PCM due to P . lutzii previously confirmed by ID using CFA from P . lutzii EPM 208 , 28 serum samples from patients with PCM due to P . brasiliensis previously confirmed by ID using AgPbB339 and gp43 as antigens , heterologous and negative control sera . Serum samples from patients with PCM due to P . lutzii were from Cuiaba-MT and Campo Grande-MS in the Central-West region of Brazil . As a gold standard , we used three sera from PCM patients with positive cultures for P . lutzii . All sera from patients with PCM due to P . brasiliensis and heterologous sera came from Hospital São Paulo , Escola Paulista de Medicina ( São Paulo , SP , Southeast region ) . All sera were divided into aliquots and stored at −20°C . Polystyrene flat-bottomed plates ( Costar; 96-wells ) were coated with CFA ( P . lutzii EPM 208 ) at 12 . 5 µg/ml diluted in 0 . 1 M carbonate buffer ( pH 9 . 6; 100 µl/well ) and incubated for 2 h at 37°C and overnight at 4°C . The remaining binding sites were blocked with PBS containing 0 . 1% Tween 20 ( PBS-T ) and 5% non-fat dry milk ( PBS-T-M ) ( 200 µl/well ) for 4 h at 37°C . After washing three times with PBS-T , diluted serum ( 100 µl/well ) ( 1∶50 to 1∶204 , 800 in PBS-T ) was added for 1 h at 37°C . After washing three times with PBS-T , 100 µl of goat anti-human IgG-peroxidase ( Sigma ) ( 1∶1000 in PBS-T ) was added in each well and incubated for 1 h at 37°C . After three washes with PBS-T , 100 µl of substrate solution ( 5 mg of o-phenylenediamine in 25 ml of 0 . 1 M citrate-phosphate buffer pH 5 . 0 plus 10 µl of 30% H202 ) was added to each well , and the reaction was interrupted after 8 min in the dark by the addition of 50 µl of 4 N H2SO4 . The optical density was read at 492 nm with a Tecan Sunrise 96 well Microplate Reader ( Tecan , Grödlg , Austria ) . The same sera were tested by ELISA using the classical exoantigen from P . brasiliensis B339 ( AgPbB339 ) at 10 µg/ml as determined by checkerboard titration . For the immunoblot study we used CFA from P . lutzii strain EPM 208 ( 10 µg/lane ) and CFA from P . brasiliensis B339 ( 10 µg/lane ) . Sera that reacted with these two different antigens were compared . The gels and reagents for SDS-PAGE were prepared as described previously [31]; a 10% resolving gel and 3% acrylamide stacking gel were used . After electrophoresis , the gels were stained with Coomassie Blue solution ( 0 . 1% Coomassie Blue , 45% methanol , and 10% glacial acetic acid ) , and excess stain was removed with destaining solution ( 10% glacial acetic acid and 10% methanol ) . We selected 12 sera from patients with PCM due to P . lutzii previously confirmed by ID using CFA from P . lutzii strain EPM 208 and 12 sera from patients with PCM due to P . brasiliensis previously confirmed by ID using AgPbB339 and gp43 as antigens . For Western blot , the samples were transferred to nitrocellulose membranes ( Milipore ) according to Towbin et al . [26] . Gel contents were electrotransferred to nitrocellulose membranes at 400 mA for 1 h in a Trans-Blot cell ( Bio-Rad ) containing transfer buffer ( 25 mM Tris-HCl , 192 mM glycine , and 20% methanol [vol/vol]; pH 8 . 3 ) . Free binding sites in the membranes were blocked by incubation for 2 h in 5% ( wt/vol ) non-fat dry milk in PBS-T ( pH 7 . 5 ) . Membranes were sliced vertically and the strips were incubated for 1 h at room temperature with diluted serum ( 1∶500 in PBS-T containing 5% non-fat milk; PBS-T-M ) . The strips were washed in PBS-T-M four times for 10 min each . The membranes were incubated with peroxidase-conjugated goat anti-human IgG ( Sigma ) at 1∶1 , 000 dilution for 1 h and then washed as above . The reactive antigenic molecules were developed by chemiluminescence mix reagents ( Millipore , WBKLS0500 ) onto the membrane which is reactive to the conjugated secondary antibody peroxidase and luminol . To image the blot , we placed the transferred membrane in the transluminator UVITEC imager ( Uvitec Cambridge , United Kingdom ) . The software settings in the Allience 4 . 7 software were set up to take several images from different time exposures , starting at 2 seconds with a total of ten images between 2 seconds . Then we selected the best time exposure from those 10 pictures taken . Performance measured analyzed for each test were: sensitivity , specificity , positive predictive value ( PPV ) and negative predictive value ( PNV ) . The receiver operating characteristics ( ROC ) curve was drawn to determine the sensitivity and specificity for each antigen preparation ( CFA , EXO and TCA ) in ID test . The areas under ROC curves ( AUC ) were calculated to evaluate the diagnostic values of each antigen preparation . We assumed a test without diagnostic power when the ROC curve was linear with AUC of 0 . 5 ( the ROC curve will coincide with the diagonal ) . A powerful test would give an AUC around 1 . 0 , demonstrating the absence of both false positives and false negatives ( the ROC curve will reach the upper left corner of the plot ) . To measure the degree of concordance of the results of the different assays , the kappa statistic and its 95% confidence interval ( 95% CI ) were calculated . Kappa values were interpreted as follows: 0 . 00–0 . 20 , poor agreement; 0 . 21–0 . 40 , fair agreement; 0 . 41–0 . 60 , moderate agreement; 0 . 61–0 . 80 , good agreement; 0 . 81–1 . 00 , very good agreement [32] . A p value of <0 . 05 was considered to indicate statistical significance . All statistical calculations were performed with the MedCalc Statistical Software version 13 . 2 . 0 ( MedCalc Software bvba , Ostend , Belgium; http://www . medcalc . org; 2014 ) . Suspected Paracoccidioides spp . isolates were subjected to HSP70 gene amplification . All clinical isolates had positive amplification and were identified as P . lutzii ( Figure S1 ) . During pilot studies , various P . lutzii CFA preparations were tested by ID; three were not reactive ( Fig . 3A ) . P . lutzii strain EPM 208 was chosen for the CFA preparation based on the good and rapid growth of this strain , as well as the good bands observed on the slide . Fig . 3B shows that antigens ( CFA ) from P . brasiliensis Pb18 and B339 do not react with P . lutzii serum . In ID tests , 59 ( 66 . 2% ) of 89 patient sera were reactive to the classical antigen AgPbB339 , indicating that these sera were from patients with PCM due to P . brasiliensis . The 30 sera that were not reactive were re-tested with P . brasiliensis-gp43 purified antigen , and 13 ( 43 . 3% ) were reactive , indicating that they were also from patients with PCM due to P . brasiliensis . Some of these serum samples were positive only for gp43 when tested with different protein concentrations of gp43 ( Figure S2 ) . The 17 non-reactive sera were re-tested with three antigenic preparations from P . lutzii strain EPM 208 . Ten ( 58 . 8% ) of the sera were reactive to the exoantigen preparation , but a weak precipitated band was observed ( Fig . 3C ) . Only 3 ( 17 . 6% ) of the sera were reactive to TCA-precipitated antigen , resulting in bands of low intensity ( Fig . 3C ) . Figure 3D shows that antigens from P . brasiliensis B339 ( Ag Exo , TCA or CFA ) do not reacted with serum from PCM patient due to P . lutzii . In contrast , all 17 ( 100% ) sera were reactive to the CFA , resulting in sharp bands ( 1 , 2 or 3 bands ) ( Fig . 3E ) . Figure 3F shows that antigen from Exo B339 or CFA B339 do not reacted with serum from PCM patient due to P . lutzii and that antigen from Exo-PI or CFA-PI do not reacted with serum from PCM patient due to P . brasiliensis . Among those 17 reactive sera with CFA P . lutzii , 2 ( 11 . 7% ) of them also reacted weakly with P . brasiliensis CFA antigen , indicating cross-reactivity between these two species . The most sensitive antigenic preparation was CFA-Pl , with a sensitivity 100% ( IC 80 . 5–100 ) and specificity 100% ( IC 94–100 ) , followed by Exo-Pl with a sensitivity 58 . 82% ( 32 . 9–81 . 6 ) and specificity 100% ( IC 94–100 ) and TCA-Pl with a sensitivity 17 . 65% ( 3 . 8–43 . 4 ) and specificity 100% ( IC 94–100 ) . Positive and negative predictive values ( PPV , NPV ) in ID assay were higher for CFA-Pl ( PPV = 100 , IC 79 . 4–100; NPV = 100 , IC 94–100 ) , followed by Exo-Pl ( PPV = 100 , IC 66 . 3–100 , NPV = 89 . 55 , IC 79 . 6–95 . 6 ) and TCA-Pl ( PPV = 100 , IC 15 . 8–100; NPV = 22 . 07 , IC 70 . 3–89 . 2 ) . Heterologous sera and serum samples from healthy individuals ( negative control ) did not react with the tested antigens . We used the area under the ROC curve ( AUC ) to evaluate the discriminatory values of the antigens ( comparing subjects with PCM due to P . lutzii and those without the disease ) . Our results showed that the CFA-Pl antigen afforded better AUC values ( AUC = 1 . 0 , IC 0 . 95–1 . 0 , p<0 . 0001 ) than those for the Exo-Pl ( AUC = 0 . 74±0 . 06 , IC 0 . 68–0 . 87 , p<0 . 0001 ) and TCA-Pl ( AUC = 0 . 58±0 . 04 , IC 0 . 47–0 . 69 , p = 0 . 0641 ) ( Fig . 4 ) . Judging from the deviating performance of different antigen preparations in ID assays we discarded the Exo-Pl and TCA-Pl antigens in subsequent experiments . For ELISA , the CFA from P . lutzii strain EPM 208 showed excellent reactivity for antibody detection using PCM sera ( P . lutzii and P . brasiliensis ) and heterologous sera ( histoplasmosis , aspergillosis and sporotrichosis ) and normal human sera as negative control . Figure 5A shows the individual serum titers of patients with PCM due to P . lutzii ( sera # 1 to 28 ) and P . brasiliensis ( sera # 29 to 58 ) . Among P . lutzii sera , # 1 , 2 , and 6 represent the gold standard sera and sera from patients lacking treatment; 3 , 4 , 5 , and 7 to 19 represent sera from patients who received 6 to 12 months of therapy; and 20 to 28 represent sera from patients with more than 12 months of therapy . Figure 5B shows these same groups of sera reacting with exoantigen from P . brasiliensis B339 ( traditional antigen: AgPbB339 ) . Figure 5C shows the median curve of 28 PCM sera ( P . lutzii ) compared to the median curve of 28 PCM sera ( P . brasiliensis ) and heterologous sera ( histoplasmosis , aspergillosis and sporotrichosis ) , besides the NHS used as negative control . Figure 5D shows the results of the same set of sera but expressed by the titers of each serum . The heterologous sera and NHS always showed much lower titers in relation to the PCM sera . On an overall , we observed that the cross-reactivity with heterologous sera and NHS does not interfere with the interpretation of results with homologous sera and P . lutzii CFA antigen . Most of the sera from patients with PCM due to P . lutzii presented with titers from 1∶12 , 800 to 1∶204 , 800 ( median = 51 , 200 ) , whereas sera from patients with PCM due to P . brasiliensis presented with maximum titers of 1∶6 . 400 ( median = 1∶ 1 , 600 ) , indicating a great difference between the sera . Among the heterologous sera the maximum of cross-reactivity was 1∶1 , 600 for aspergillosis ( median = 1∶200 ) , 1∶800 for histoplasmosis ( median = 1∶200 ) , 1∶400 for sporotrichosis ( median = 1∶200 ) . NHS reacted until 1∶400 ( median = 1∶200 ) . Western blot allowed us to distinguish PCM due to P . lutzii and due to P . brasiliensis , as only sera from patients with PCM due to P . lutzii are able to recognize antigenic molecules from P . lutzii-CFA ( Fig . 6A ) . Sera from patients with PCM due to P . brasiliensis did not recognize any P . lutzii CFA antigens ( Fig . 6B ) . In addition , we used Kappa test to assess the agreement between different serological assays ( ID , ELISA and Western blot ) based on CFA-Pl antigen . Kappa test ( k = 1 ) revealed a perfect agreement between all pairwise comparisons for different assays . In order to verify cross-reactions between both Paracoccidioides species , sera from patients with PCM due to P . brasiliensis and due to P . lutzii were tested against CFA from B339 ( Fig . 6C and 6D ) . The results showed that sera from PCM due to P . brasiliensis recognized only gp43 molecule; sera from PCM patients due to P . lutzii recognized various molecules . Considering the main bands evidenced , the reactivity were almost similar to that obtained with P . lutzii sera and CFA-Pl . The World Health Organization ( WHO ) estimates that more than 1 billion people , one-sixth of the world's population , suffer from one or more neglected diseases . The diseases are most heavily concentrated in low-income nations in Africa and Latin America , and the most common types of neglected diseases are tropical diseases . Many neglected tropical diseases are caused by parasites , which are spread by insects or contact with contaminated water or soil ( http://www . who . int/neglected_diseases/diseases/en/ ) . In this scenario , PCM can be considered a neglected disease . Coutinho et al . [2] analyzed 3 , 181 deaths from PCM in Brazil based on 16 years of sequential data ( from 1980 to 1995 ) . During this period , PCM showed considerable magnitude and low visibility , representing the eighth most common cause of death from predominantly chronic or recurrent types of infectious and parasitic diseases . The study showed that the mortality rate justifies classifying this disease as an important health problem in Brazil . However , no government programs exist for this mycosis , with rare punctual exceptions . For a century PCM was thought to be caused solely by P . brasiliensis , but recent studies [11] , [12] , [18]–[21] , [33] have shown that other species of Paracoccidioides can also cause PCM . These findings reveal that problems in the serological diagnosis of PCM were due to an incorrect use of antigens from the genus , as only antigens of P . brasiliensis ( PS3 ) have been used for this purpose . With the discovery of P . lutzii , it became clear that new antigens from these new species would be needed to obtain more accurate diagnoses . Recently , a fatal case of PCM due to P . lutzii fungemia was reported [34] and other two cases were reported in the North region of Brazil [23] . Until recently , the serological diagnosis of PCM by ID could be accomplished using only the standard exoantigen from B339 , which has a high concentration of gp43 antigen , the immunodominant and specific molecule in ID tests . Early suspicions that this serology was not entirely certain were raised by our group [35] when we observed that an exoantigen produced from an isolate ( Pb550B ) from the Central-West region of Brazil was capable of reacting with a larger number of PCM sera from that region than the standard antigen from an isolate from São Paulo in the Southeast region ( AgPbB339 ) . Nevertheless , when tested with PCM sera from São Paulo , the antigen from the Central-West region of Brazil did not produce satisfactory results . Also , when we used the standard antigen AgPbB339 , it reacted very well with sera from São Paulo , but very poorly with sera from the Central-West region . Analyses of antigens obtained from the B339 and 550B isolates showed that the former produced high levels of gp43 , which was undetected in the 550B filtrate . This variation in gp43 expression likely influences the low reactivity observed in ID tests using sera from patients from the Central-West region of Brazil . At that time , nothing was known about P . lutzii; therefore , we suggested using regional strains to aid in the diagnosis of PCM for the Central-West region . Importantly , B339 belongs to the S3 species , whereas isolate 550B is from the Central-West region of Brazil , where P . lutzii is more frequently found . The main fact that spurred this study was the observation that sera from patients with PCM ( previously proven by direct examination of secretions or histopathological analysis ) were negative by conventional serology using the standard AgPbB339 antigen in ID tests . In order to elucidate the problem with serological diagnosis , we developed a strategy . First , all sera from patients suspected of PCM should be tested against the traditional exoantigen of P . brasiliensis ( AgPbB339 ) and purified gp43 molecule derived from the same standard strain . The reactive sera were then considered to be PCM due to P . brasiliensis . The non-reactive sera were then tested against CFA from P . lutzii EPM 208 . Applying this strategy , we were able to diagnosis 72 patients with PCM due to P . brasiliensis . The remaining non-reactive patients ( n = 17 ) were diagnosed mostly by ID using CFA from P . lutzii and also by ELISA and Western blot . This type of antigen ( CFA ) is very easy to prepare and is very useful in the diagnosis of PCM due to P . lutzii , exhibiting clearly visible precipitation bands . Among these 17 sera , 2 of them also reacted with the traditional exoantigen B339 , but with a weak reactivity , indicating cross reaction between these two species of Paracoccidioides . However , the exoantigen preparation ( from P . lutzii ) also managed to obtain positive reactions , but the precipitation line was very weak , making it difficult to read the slides . In addition , the antigenic preparations obtained by TCA precipitation were useful , but the bands were quite weak . In both later antigens , the sensitivity of the reactions was lower than that obtained with the CFA preparation . Thus , the CFA preparation obtained from P . lutzii EPM 208 was the better antigen for use in ID tests to diagnosis PCM due to P . lutzii from the Central-West region of Brazil . Other P . lutzii strains were tested with similar results; however , among 14 CFA preparations , three were unable to precipitate antibodies in sera from P . lutzii-PCM patients , indicating that antigenic differences exist among these strains . Therefore , more studies are necessary to elucidate the antigenic variability among P . lutzii strains . Another curiosity we observed when we tested CFA preparations from P . lutzii in the ID tests was that better results were obtained when the protein concentrations of the antigen were greater than 1200 µg/ml . Protein concentrations less than 1200 µg/ml resulted in very weak bands , hindering the visualization of precipitation reactions . Machado et al . [36] described CFA as being inefficient for the diagnosis of PCM due to P . lutzii . In ELISA , sera from patients with PCM due to P . lutzii presented with higher antibody titers than those obtained for the sera from patients with PCM due to P . brasiliensis which can perfectly differentiate between patient sera from P . brasiliensis and P . lutzii . Heterologous sera , such as histoplasmosis , aspergillosis and sporotrichosis and normal human sera exhibited very low cross-reactivity and did not interfere with interpretations of the main system . It was expected that the sera from PCM patients due to P . brasiliensis reacted more intensively with P . lutzii antigens because these two species belong to the same genus . Certainly , the antigenic components present in P . lutzii are constituted by specific antigens of this species plus common antigens shared with P . brasiliensis . Probably , antibodies elicited during infection by P . brasiliensis are only generated against antigens from this species whereas antibodies elicited during infection by P . lutzii are generated by specific antibodies against antigens in this species and also against antigens shared with P . brasiliensis . It was also expected that sera from patients with PCM due to P . brasiliensis reacted more intensively than heterologous sera ( histoplasmosis , aspergillosis and sporotrichosis ) , and not almost in the same level . Perhaps , these results may be explained by the presence of common antigens among P . brasiliensis and heterologous fungi , which are reacting in this ELISA system . In relation to the strategy used for immunoblotting in the present study , we found that sera from patients with PCM due to P . brasiliensis do not recognize any antigen from P . lutzii CFA . In contrast , sera from patients with PCM due to P . lutzii were able to recognize antigens from both antigenic preparations ( CFA from P . lutzii or P . brasiliensis ) . Therefore , P . lutzii is antigenically more complex . These findings suggest that P . lutzii has its own species-specific antigens and antigens common with P . brasiliensis . Thus , we can distinguish between sera from patients with PCM due to P . brasiliensis and those with PCM due to P . lutzii . However , the specific P . lutzii antigen still needs to be identified . The reactivity of PCM sera due to P . brasiliensis predominantly to gp43 of its homologous antigen may be explained by the fact of the great immunogenicity of this molecule , so that , the immune response to other minor molecules is not evidenced . Based on our results , we estimate that the incidence of PCM due to P . lutzii is much greater than we imagine , even though the Central-West region of Brazil where the disease is prevalent has not invested in acquiring new tools for a specific diagnosis of this entity . In general , laboratories for the diagnosis of infectious diseases are small , have few resources , and perform a limited number of tests per day . Our strategy meets all of the requirements for use in laboratories for the diagnosis of PCM due to P . lutzii , as well as seroepidemiological studies . In addition , the test does not require special skills and can be used for a small number of samples . Although the cost of ELISA and Western blot is higher than that of ID , they are economical if the costs associated with laboratory personnel , quality control , and reagent storage are taken into account . Surely P . lutzii is not confined to the Central-West region of the country; we have described two cases in the northern region ( Pará State ) [23] , and we have strains of P . lutzii isolated in the Rondônia State ( North region ) and Paraná ( South region ) in our fungal collection . With our proposed strategy for the diagnosis of PCM caused by P . lutzii , many new cases may arise throughout Brazil and other South American countries . Studies are underway in our laboratories to identify the species-specific antigen for P . lutzii in order to use it for simple and specific diagnosis of this mycosis .	Tropical diseases , such as paracoccidioidomycosis , are the most common type of neglected diseases . From 1980 to 1995 , 3 , 181 deaths from paracoccidioidomycosis occurred in Brazil , representing the eighth most common cause of death from predominantly chronic or recurrent types of infectious and parasitic diseases , showing considerable magnitude and low visibility . Paracoccidioidomycosis is traditionally assumed to be caused solely by Paracoccidioides brasiliensis , but a new species , Paracoccidioides lutzii , was discovered in the Central-Western region of Brazil . Thus , new antigenic preparations and tests for an accurate differential diagnosis between these two species appear to be needed . This study aimed to develop new antigenic preparations from P . lutzii isolates to improve the diagnosis of paracoccidioidomycosis . We used patient serum samples predominantly from the Central-Western region of Brazil . Various antigenic preparations were tested , and a cell free antigen derived from P . lutzii was an excellent antigen for serological diagnosis and able to diagnose 100% of sera from patients with PCM due to P . lutzii .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "mycology", "immunoassays", "biology", "and", "life", "sciences", "immunology", "microbiology", "immunologic", "techniques", "research", "and", "analysis", "methods" ]	2014	Serology of Paracoccidioidomycosis Due to Paracoccidioides lutzii
Ebola and other filoviruses pose significant public health and conservation threats by causing high mortality in primates , including humans . Preventing future outbreaks of ebolavirus depends on identifying wildlife reservoirs , but extraordinarily high biodiversity of potential hosts in temporally dynamic environments of equatorial Africa contributes to sporadic , unpredictable outbreaks that have hampered efforts to identify wild reservoirs for nearly 40 years . Using a machine learning algorithm , generalized boosted regression , we characterize potential filovirus-positive bat species with estimated 87% accuracy . Our model produces two specific outputs with immediate utility for guiding filovirus surveillance in the wild . First , we report a profile of intrinsic traits that discriminates hosts from non-hosts , providing a biological caricature of a filovirus-positive bat species . This profile emphasizes traits describing adult and neonate body sizes and rates of reproductive fitness , as well as species’ geographic range overlap with regions of high mammalian diversity . Second , we identify several bat species ranked most likely to be filovirus-positive on the basis of intrinsic trait similarity with known filovirus-positive bats . New bat species predicted to be positive for filoviruses are widely distributed outside of equatorial Africa , with a majority of species overlapping in Southeast Asia . Taken together , these results spotlight several potential host species and geographical regions as high-probability targets for future filovirus surveillance . After more than 40 years , the natural reservoirs of viruses in the genus Ebolavirus remain elusive . Accumulating indirect evidence during this time points to bats as primary suspects because several species have been found positive for filovirus antibodies ( S1 Table ) . Some of these species have also been confirmed as natural reservoirs for another filovirus , Marburg virus [1 , 2] . Three bat species demonstrate the ability to replicate ebolavirus following experimental inoculation [3] , and ebolavirus RNA has been discovered in three , naturally infected species [4] . In contrast to other surveyed mammal species ( great apes , duiker ) , there is little evidence of filovirus-induced morbidity in bats [1] . Such asymptomatic infections make bats more likely to be natural reservoir candidates for ebolaviruses than , for example , great apes ( gorilla and chimpanzee ) , which suffer mortality rates exceeding those observed in human populations [5] . Effective surveillance in the countries most frequently affected by ebolaviruses ( e . g . , Uganda , Democratic Republic of Congo [6] , and the countries affected by the recent outbreak in West Africa [7] ) is hampered by the incredible diversity of species over such a large geographical area . For example , West Africa is recognized as one of the most species-rich regions on Earth , with a large number of endemic species typically present in low densities [8 , 9] . Moreover , there is pronounced seasonality in regions affected by ebolaviruses with wet and dry seasons contributing to fruiting phenology and water availability that combine to influence the movement ecology , breeding , and birth pulses in a number of species , including bats [10–12] . Though wildlife surveillance to date surpasses 30 , 000 individuals collected from hundreds of species , we have yet to isolate live ebolavirus from any African wild species . What other species might be natural hosts of filoviruses in the wild ? To answer this question , we applied a machine learning approach to mine patterns in data on the world’s bat species . Here , we report an intrinsic trait profile that distinguishes seropositive bat species from all others with an estimated 87% accuracy . We identify a rank order of particular bat species whose trait profiles suggest a high probability that they could also be permissive to filovirus infection , and geographic regions where numerous of these potentially novel filovirus hosts co-occur to highlight surveillance targets of candidate reservoir species . For all 1116 bat species , we collected life history , physiological and ecological traits from PanTHERIA [13] , a species-level database of the world’s mammals ( S2 Table ) . We calculated 3 additional , derivative traits from basic morphological and demographic variables: post-natal growth rate ( weaning body mass/neonatal body mass ) ; relative age to sexual maturity ( sexual maturity age/maximum longevity ) ; relative age at first birth ( age at first birth/maximum longevity ) . We added bat family as a series of 18 binary variables to explore the likelihood of taxonomic clustering among carriers . We calculated species density , defined as the richness of mammal species found within a species’ geographic range ( as reported in IUCN [14] ) divided by the total geographic range area for each bat species ( n/km2 ) . We compiled published data on diet and activity patterns [15]; torpor and migratory behavior [16]; and mass-corrected production ( the mean mass of offspring produced per year , normalized by adult body size [15] . Bat species names were standardized using Wilson and Reeder 2005 [17] . Each bat species was assigned a binary code according to its current status ( 0 –not currently known to carry a filovirus; 1 –published evidence; S1 Table ) . For this binary response variable , we applied generalized boosted regression [18–20] , a type of machine learning that seeks to maximize classification accuracy ( in this case , discriminating reservoir status among 1116 bat species ) by learning the patterns of features that distinguish between bats that have tested positive for filoviruses from all other species . Machine learning is particularly well-suited to comparative studies because it does not assume an underlying data distribution [21] , and explanatory power is unaffected by collinearity , hidden interactions , and non-random patterns of missing data common in ecological data sets ( e . g . , those that arise through sampling bias , or when species share similar trait values as a result of phylogenetic relatedness ) [22 , 23] . The model-free approach of machine learning algorithms like generalized boosted regression trees enables superior predictive accuracy based on patterns inherent in data themselves rather than based on a priori assumptions about underlying ecological processes or simple parametric relationships , in proportion to the quantity of information contained in the data . Boosted regression trees generate a series of recursive binary splits for randomly sampled predictor variables . Each successive tree is built using the residuals of the previous best-performing tree as the new response variable . Thus , an ensemble of linked trees is generated where each tree achieves increasingly more accurate classification based on randomly selected variables . In our analyses , we repeated the tree building process several thousand times to create an ensemble classification model of up to 5000 trees . Datasets were partitioned into training ( 80% of all 1116 species ) and test sets ( the remaining 20% ) prior to analysis . We applied 10-fold cross-validation during model building to prevent over-fitting , and permutation procedures to generate relative importance scores for each predictor variable ( S3 Table , which also summarizes tuning parameters , performance metrics ( AUC ) , and complete trait profiles ) . To calibrate performance , we conducted randomized bootstrapped permutation analysis of the species labels ( 500 permutations ) , a procedure referred to as target shuffling in business analytics . We calculated a baseline mean AUC for these permutations ( 0 . 6 ) and corrected our test AUC ( originally AUC = 0 . 97 ) by this baseline to arrive at our corrected test AUC of 0 . 87 = 0 . 97- ( 0 . 60–0 . 5 ) . To investigate the sensitivity of our results to errors and permutations in the covariates , we randomly removed 1% , 5% , 10% , 15% , and 20% of trait values , refit the model , and calculated the Spearman rank-order correlation between scores obtained using the corrupted data and those of our original analysis . This exercise showed the algorithm to be extremely adept at identifying the relative risk among bat species ( rank order ) with up to 5% of data removed ( ρ = 0 . 99 ) and very good with up to 20% of data removed ( ρ = 0 . 90 ) ( S4 Table ) . In our analysis , “unknown” carriers ( 1095 species ) were designated “non-carriers” , labeled as 0 . In the absence of repeated experimental inoculations , a large number of individuals of each species must be sampled before consensus can be reached that a given species is unable to harbor infection . Thus , we adopted this more conservative designation–essentially presence vs . background–to align with our aim of developing models whose baseline classification performance will continue to improve with future discoveries of new filovirus-positive species . Intrinsic features that reflect life history and biology are less susceptible to sampling biases than epidemiological data–for example , public heath and research expenditures are unlikely to influence a species’ age to sexual maturity , or other similar life history features . However , to control for any potential effect of sampling bias on our results , we tallied the number of primary literature citations in the Web of Science ( WOS ) for each bat species in our dataset as a proxy for study effort . Citation count was within the top dozen variables important for predicting filovirus-positive status , but it had low relative importance for prediction accuracy ( S2 Table ) . Removing WOS hits from the analysis did not alter the rank order of variables most important for predicting filovirus-positive bats , confirming that while some filovirus-positive bat species may be better studied than others , studied-ness did not bias the trait profiles generated by our modeling approach . Analyses were performed using the gbm package [19] in R [24] . To identify hot spots of filovirus carriers , we mapped the geographic ranges of all known filovirus-positive bat species ( S1 Table ) , as well as new filovirus carriers in the 90th percentile of model predictions ( S3 Table ) . We also provide maps for species comprising the 95th and 99th percentiles ( S1 Fig ) . All geographic ranges were obtained from the IUCN database of terrestrial mammals [14] and compiled in ArcGIS [25] . From peer-reviewed primary literature , we identified 21 out of 1116 ( ~1 . 9% ) total extant bat species to have tested positive for any filovirus by means of any diagnostic ( i . e . , either serological or molecular assays ) . Approximately half of these species ( n = 11 ) are fruit bats belonging to Family Pteropodidae ( the Old World fruit bats ) , and the other half are primarily insectivorous bats from 4 families ( S1 Table ) . Although fruit bats comprise only about 16% of global bat biodiversity ( 186/1116 extant species ) , we estimate 5 times as many fruit bat individuals have been sampled for filoviruses compared to insectivorous bats ( S2 Table ) , which corroborates on a global scale the surveillance bias recently reported for ebolaviruses in bats of Africa [26] . Using 57 variables describing the biology , life history , ecology , taxonomy , and biogeography of all bat species ( S2 Table ) , our model predicted filovirus-positivity with 87% accuracy , and revealed a trait profile that distinguishes filovirus-positive species from other bats ( Fig 1 , S5 Table ) . In general , filovirus-positive bat species tend to have neonates that are larger at birth and wean at a larger size compared to other bats . This tendency to produce larger offspring was not an artifact of large adult body size . Rather , filovirus-positive bats produce greater biomass for their body size compared to other bat species ( the production variable [27] , Fig 1 ) . The majority of bats have 1 litter per year with a single pup in each litter , but some populations support a second litter in some years ( notably among the Vespertilionidae , the most speciose Family of insectivorous bats , and the Pteropodidae , the Old World fruit bats ) . We found that filovirus-positive species disproportionately display this tendency to have more than a single litter ( pup ) per year [28] . We also observed a bimodal pattern in sexual maturity age for filovirus-positive species , a pattern we conjecture may arise from small species ( insectivorous bats ) displaying earlier ages of sexual maturity compared to the large species ( fruit bats ) in the tropics where reproductive rates of non-hibernating bats decrease with body size [29] . Filovirus-positive species also display a tendency to live in larger population groups ( roosts ) compared to other bats . While group-living affords many benefits , costs of group living include increased pathogen transmission [30] and conspicuousness to predators , including human hunters [31] . Thus , it is possible that species living in large , conspicuous roosts are displaying compensatory effects of faster reproductive rates ( earlier age to sexual maturity [32] or more offspring per year [29] ) in response to increased extrinsic mortality risks conferred by hunting pressure . Overall , our results suggest that even though bats are constrained to a relatively slow life history strategy ( i . e . , long-lived with few offspring per year ) compared to similarly sized mammals , filovirus-positive bat species are those whose life history pace is at the leading edge of these constraints . In addition to traits that may enable bats to be more permissive to filovirus infection at the cellular level [33 , 34] , a life history profile reflecting faster reproductive rates may increase the likelihood of infection persistence through the more rapid replenishment of susceptible young [1 , 28] . Beyond intrinsic fitness components , our analyses revealed that filovirus-positive species exhibit larger geographic ranges containing higher mammal species richness per square kilometer than other bats ( species density , Fig 1 ) . Even after correcting for geographic range size , filovirus-positive bat species overlap with a greater diversity of mammal species per square kilometer , a finding that recapitulates a scientific consensus that there are likely to be multiple natural reservoirs supporting filoviruses such as Zaire ebolavirus [35] . This result corroborates independent studies within Africa predicting the environmental niche of ebolaviruses to span primary tropical rainforest ( continuous tropical rainforests as well as gallery rainforests , which occur along riparian and transitional zones ) [36–38] . But , in a departure from previous studies , our analysis identified several hotspots outside Africa where up to 25 predicted filovirus host species overlap in geographic range ( Figs 2 and 3; S1 Fig ) . Geographic ranges of filovirus-positive bat species are concentrated in sub-Saharan Africa and Southeast Asia , spanning a total of 133 countries ( Fig 2a ) . There is a conspicuous lack of surveillance in the western hemisphere , and to our knowledge there are no published studies reporting the results ( positive or negative ) of filovirus surveillance efforts in North , Central , or South America [26] . Novel bat carriers predicted by our model ( i . e . , those in the top 10% ) are much more widely distributed than expected , with predicted species occurring across Southeast Asia , and Central and South America ( Fig 2b; S3 Table ) . The predictions in the Americas are intriguing because , while New World bats may exhibit the appropriate traits , biogeographical processes may prevent filoviruses from existing in these regions . Indeed , homologous copies of VP35-like and NP-like gene integrations were found in both Old World and New World species of Myotis bats [39] . If filoviruses are discovered in bat species in the Americas , this would call into question the age of the Filoviridae , which , through whole genome analyses , have been estimated to share common ancestry 10 , 000 years ago [40] . Analyses of integrated elements in mammalian genomes , however , suggest filoviruses may be much older [41] . Among the 112 species comprising the 90th percentile probability there are 9 Myotis species ( S3 Table ) . Among these , Myotis ricketti tested seropositive and Myotis fimbriatus tested seronegative for Reston ebolavirus in China [42] . Diagnostic tests of the remaining 7 species have , to our knowledge , never been reported at the species level . A majority of the newly predicted filovirus carriers overlap in Southeast Asia ( Fig 3 ) , with notable hotspots occurring in regions of Thailand , Burma , Malaysia , Vietnam , and northeast India . A recent study reports the negative results of a large survey testing 500 individuals of Pteropus lylei for ebolavirus across 10 roosting sites in Thailand . This study was designed with enough statistical power to detect ebolavirus prevalence as low as 6% [43] . Our model ranked this particular fruit bat species behind 195 other bat species in its probability of filovirus-seropositivity . In particular , it is preceded by three other species commonly found in Thailand–Pipistrellus tenuis , Eonycteris spelaea , and Megaderma lyra , which rank 3rd , 5th , and 7th in a global list of unsurveyed bats predicted to be seropositive ( S3 Table ) . Future surveillance efforts may be streamlined by prioritizing filovirus-testing by species displaying the strongest trait similarities with known filovirus-positive species . Despite numerous suitable bat hosts and ongoing discoveries of novel filoviruses in this region ( e . g . , [44] ) , there are comparatively few reports of disease outbreaks in Asia . Though pigs were identified as possible reservoirs of Reston ebolavirus through routine investigation of syndromic disease , there have been no reports of human disease in this region . One outstanding question for future work is to investigate why there are so few spillover events reported for human and wildlife populations in Southeast Asia compared to equatorial Africa . Whether outbreaks are indeed occurring but on a smaller or less easily detectable scale ( e . g . , as in Ethiopia [45] ) , or whether filovirus strains in this region are fundamentally less virulent to their host species , sorting the competing hypotheses about why filovirus infection dynamics in Africa differ from those in Asia will begin with more targeted surveillance of candidate reservoir species .	Preventing future outbreaks of ebolaviruses in humans and other vulnerable animal populations will require identifying the natural reservoirs of filoviruses . Accumulating indirect evidence points to certain bat species as prime suspects . To guide the search for natural filovirus reservoirs , we mined intrinsic biological data on the world’s bat species to determine what features best predict filovirus hosts compared to bats at large . We report a suite of traits that distinguishes seropositive bat species from all others with an estimated 87% accuracy . We also identify several bat species not currently known to be filovirus hosts whose trait profiles indicate should be surveillance targets . Geographic regions where numerous potential filovirus hosts co-occur ( potential filovirus hotspots ) suggest that filovirus distribution and diversity may be greater than previously thought .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "and", "Discussion" ]	[ "geographical", "locations", "vertebrates", "animals", "mammals", "viruses", "filoviruses", "mathematics", "rna", "viruses", "phylogenetic", "analysis", "molecular", "biology", "techniques", "africa", "fruit", "bats", "discrete", "mathematics", "research", "and", "analysi...	2016	Undiscovered Bat Hosts of Filoviruses
Differential transcription in Ascaris suum was investigated using a genomic-bioinformatic approach . A cDNA archive enriched for molecules in the infective third-stage larva ( L3 ) of A . suum was constructed by suppressive-subtractive hybridization ( SSH ) , and a subset of cDNAs from 3075 clones subjected to microarray analysis using cDNA probes derived from RNA from different developmental stages of A . suum . The cDNAs ( n = 498 ) shown by microarray analysis to be enriched in the L3 were sequenced and subjected to bioinformatic analyses using a semi-automated pipeline ( ESTExplorer ) . Using gene ontology ( GO ) , 235 of these molecules were assigned to ‘biological process’ ( n = 68 ) , ‘cellular component’ ( n = 50 ) , or ‘molecular function’ ( n = 117 ) . Of the 91 clusters assembled , 56 molecules ( 61 . 5% ) had homologues/orthologues in the free-living nematodes Caenorhabditis elegans and C . briggsae and/or other organisms , whereas 35 ( 38 . 5% ) had no significant similarity to any sequences available in current gene databases . Transcripts encoding protein kinases , protein phosphatases ( and their precursors ) , and enolases were abundantly represented in the L3 of A . suum , as were molecules involved in cellular processes , such as ubiquitination and proteasome function , gene transcription , protein–protein interactions , and function . In silico analyses inferred the C . elegans orthologues/homologues ( n = 50 ) to be involved in apoptosis and insulin signaling ( 2% ) , ATP synthesis ( 2% ) , carbon metabolism ( 6% ) , fatty acid biosynthesis ( 2% ) , gap junction ( 2% ) , glucose metabolism ( 6% ) , or porphyrin metabolism ( 2% ) , although 34 ( 68% ) of them could not be mapped to a specific metabolic pathway . Small numbers of these 50 molecules were predicted to be secreted ( 10% ) , anchored ( 2% ) , and/or transmembrane ( 12% ) proteins . Functionally , 17 ( 34% ) of them were predicted to be associated with ( non-wild-type ) RNAi phenotypes in C . elegans , the majority being embryonic lethality ( Emb ) ( 13 types; 58 . 8% ) , larval arrest ( Lva ) ( 23 . 5% ) and larval lethality ( Lvl ) ( 47% ) . A genetic interaction network was predicted for these 17 C . elegans orthologues , revealing highly significant interactions for nine molecules associated with embryonic and larval development ( 66 . 9% ) , information storage and processing ( 5 . 1% ) , cellular processing and signaling ( 15 . 2% ) , metabolism ( 6 . 1% ) , and unknown function ( 6 . 7% ) . The potential roles of these molecules in development are discussed in relation to the known roles of their homologues/orthologues in C . elegans and some other nematodes . The results of the present study provide a basis for future functional genomic studies to elucidate molecular aspects governing larval developmental processes in A . suum and/or the transition to parasitism . Parasitic nematodes are of major socio-economic importance in animals . For example , hundreds of millions of people are infected with geohelminths ( soil-transmitted worms ) , such as blood-feeding hookworms Ancylostoma duodenale and/or Necator americanus , Trichuris trichiura and Ascaris spp . [1] , causing serious adverse effects on human health , particularly in children . Similarly , parasitic nematodes of livestock , such as pigs , also cause substantial economic losses due to subclinical and clinical diseases , with billions of dollars spent annually on the treatment and control of gastro-intestinal nematodes . In addition to the socioeconomic impact that these parasites have , there is potential for the emergence of resistance in them against all of the main classes of ( nematocidal ) compounds used to treat the diseases they cause [2]–[5] . Therefore , there is a significant need to work toward discovering new compounds to control these parasites . Gaining an improved understanding of the molecular basis of parasite development provides such an avenue . Compared with the free-living nematode Caenorhabditis elegans , there is very little information on fundamental molecular aspects of development in parasitic nematodes [6]–[8] . Since the genome sequence of C . elegans was published in 1998 [9] , many aspects of the molecular biology of this nematode have been elucidated . For instance , microarray analyses have been used to examine developmental and gender-enriched gene expression [10] , [11] , and the functions of more than 96% of the C . elegans genes have been assessed by double-stranded RNA interference ( RNAi , or gene silencing; [12] ) [13]–[18] . Comparative analyses of genetic data sets have shown that parasitic nematodes usually share ∼50–70% of genes with C . elegans ( e . g . , [19] , [20] ) . There is similarity in other features ( such as basic body plan and moulting ) between C . elegans and parasitic nematodes , suggesting that some molecular pathways are relatively conserved [8] , [21] . Understanding the pathways linked to basic nematode biology and development could have important implications for finding new ways of disrupting these pathways and thus facilitate the identification of new drug targets . Despite the advances in genomic technologies [7] , [22]–[29] and the study of C . elegans , there is a paucity of information on the genomics of parasitic nematodes of animals , particularly in relation to development . Also considering the major socioeconomic impact of Ascaris and ascariasis in humans and pigs [30]–[32] , several characteristics , including the large size of the adult worm ( providing the opportunity of investigating individual organ systems and tissues ) , the ability to maintain Ascaris in the pig , store eggs and culture larvae in vitro for relatively long periods of time ( months to years ) [32] as well as the discovery that RNAi achieves “cross-species” gene silencing for a selected number of genes [33] , [34] and the imminent genome sequence ( http://www . sanger . ac . uk/Projects/Helminths/ ) all indicate that Ascaris could serve as a powerful model system for investigating reproductive and developmental processes in nematodes . In the present study , Ascaris from pigs was used to study molecules abundantly transcribed in the infective third-stage larva ( L3 ) . Following the oral ingestion of Ascaris eggs by the host ( human or pig ) , L3s are released and then invade/penetrate predominantly the caecal wall [35] to then undergo hepato-pulmonary migration , after which ultimately the adult females and males establish and develop in the small intestine [36] , [37] . The molecular mechanisms linked to host invasion and parasite development are largely unknown . Here , we constructed an L3-enriched cDNA library using the method of suppressive-subtractive hybridization ( SSH ) , explored transcription of a representative subset of molecules by microarray analysis and conducted bioinformatic analyses to characterize these molecules , map them to biochemical pathways and predict genetic interactions based on comparisons with C . elegans and/or other organisms . Experimental pigs ( 8–12 weeks of age ) were purchased from and maintained in the Experimental Animal Center of South China Agricultural University . These pigs were treated humanely , according to the Animal Ethics procedures and guidelines of the People's Republic of China . Adult worms ( males and females ) of A . suum were collected from the small intestines of pigs from an abattoir in Shenzhen , China . Infective eggs and infective L3s of A . suum were produced according to the methods described previously [38] . In brief , eggs from the uteri of adult females of A . suum were collected and incubated at 28°C for 28 days to allow them to develop to infective eggs ( containing infective L3s ) . To obtain pure infective L3s , 7 . 5% v/v sodium hypochlorite was used to treat the larvated eggs at 37°C overnight and then the eggs were shaken with glass-beads; then , the exsheathed L3s and shells were separated by density gradient centrifugation using lymphocyte separating medium ( LSM ) [38] . Following the experimental infection of helminth-free pigs with infective Ascaris eggs as described previously [39] , the L3s from livers and in lungs as well as L4s in intestines were isolated according to an established method [40] . All parasite materials were snap-frozen in liquid nitrogen prior to storage at −70°C . Total RNA was isolated from adult females and males , different larval stages or eggs of A . suum using TriPure reagent ( Roche ) as recommended by the manufacturer . Equal amounts of total RNA from each stage or sex were pooled . The mRNA was isolated using the Oligotex mRNA Kit ( Qiagen ) , following the manufacturer's protocol . SSH was carried out using the PCR-Select cDNA Subtraction kit ( Clontech ) , according to the manufacturer's protocol . In brief , cDNA synthesized from mRNAs from infective L3s was subtracted against cDNA synthesized from the pooled mRNA from all other stages included herein . The SSH library was constructed using infective L3s as the tester and pooled cDNAs from all other stages as the driver . The effectiveness of this subtraction process has already been demonstrated in previous studies [41] , [42] . The cDNA obtained following SSH was cloned into the pGEM-T Easy plasmid vector ( Promega ) and competent Escherichia coli ( JM109 ) transformed . Positive clones , picked randomly ( based on blue/white selection ) , were grown overnight in Luria Bertani ( LB ) medium ( shaking , 37°C ) . Individual inserts were PCR-amplified using “nested primers” 1 and 2R from the Subtraction kit ( Clontech ) and examined by agarose electrophoresis . Clones ( n = 3075 ) from the subtracted library were picked and cultured overnight in LB containing ampicillin ( 1000 IU/ml ) in sealed 96-well blocks . Five µl of culture suspension from each well were transferred into individual wells thermocycling ( 96-well ) plates and the inserts PCR-amplified using primers 1 and 2R . Following a 10 min denaturation step at 94°C , the amplification proceeded for 25 cycles of 10 s at 94°C , 30 s at 68°C and 1 . 5 min at 72°C , with a final extension for 5 min at 72°C . Products were resolved in agarose gels , ethanol precipitated , re-suspended in 16 µl of “spotting solution” ( Shanghai BioStar Genechip , Inc ) to a final concentration of ∼500 ng per µl , before being printed on to glass slides ( in duplicate ) using a robotic arrayer . Sixteen blanks ( using spotting solution only ) and the same number of negative ( irrelevant cDNAs with no relationship to Ascaris ) were also printed on to slides and served as negative controls; β-actin of A . suum served as a positive control to assess the efficiency of labeling and hybridization . The slides were air-dried for 2 h , and cDNA in the spots were cross-linked at 254 mJ . The printed slides were stored at 4°C . The cDNAs produced from total RNA from A . suum eggs , infective L3s , L3s isolated from pig liver or lung , fourth-stage larvae ( L4s ) , adult males or females [as described in the section ‘Construction of the cDNA Library by Subtractive-Suppressive Hybridization ( SSH ) ’] were labeled with cyanine dyes . Cy3 or Cy5-dCTP was incorporated into cDNA produced from 30 µg of total RNA by direct labeling in a reverse transcription reaction using an oligo ( dT ) primer . Labeled cDNA was purified using DyeEx columns ( Qiagen ) . Microarray slides were incubated with a pre-hybridization solution [5×SSC , 1% bovine serum albumin ( BSA ) , 0 . 1% sodium dodecyl-sulphate ( SDS ) ] for 6 h at 42°C . After pre-hybridization , the microarray slides were incubated with ‘pooled’ Cy3 and Cy5-labeled probes in hybridization solution ( 5×SSC , 1% BSA , 0 . 1% SDS ) , in the dark at 42°C for 18 h , and then washed in solution I ( 1×SSC , 0 . 2% SDS ) for 10 min , followed by solution II ( 0 . 1×SSC , 0 . 2% SDS ) for 10 min at 60°C , according to the protocols provided by Shanghai BioStar Genechip , Inc . A “dye flip” was carried out to control for any bias in hybridization signal between the Cy-labeled cDNA probes ( produced for two distinct mRNA populations ) . The slides were dried and scanned ( ScanArray 4000 scanner ) using image acquisition software ( Shanghai BioStar Genechip Inc . ) and a range of laser power and photo-multiplier tube intensities . The mean hybridization signal ( derived from four replicates of the same array ) were corrected for background , normalized [43] , log2-transformed and then subjected to statistical analysis employing the students t-test in a spreadsheet ( Excel , Microsoft , USA ) . The microarray data were analysed for differential cDNA hybridization ( >2 . 0-fold to 114 . 3-fold ) between L3 and each of the other stages ( eggs , lung and liver L3s , L4 , adult female and adult male ) . For a subset ( n = 17 ) of representative ESTs ( rESTs ) , RT-PCR was used to verify the differential transcription recorded by microarray analysis . Double-stranded cDNA was synthesized from total RNA ( separately ) from each stage or sex of A . suum using reverse transcriptase ( Superscript III , Invitrogen ) . Briefly , 5 µg of total RNA were added to 14 µl of H2O and 1 µl of oligo d ( T ) n = 12–18 primer ( 0 . 5 µg/µl ) , heated to 70 °C for 10 min and chilled on ice . First- and second-strand cDNAs were synthesized via the addition of 4 µl of first-strand cDNA buffer ( 250 mM Tris-HCl , pH 8 . 3 , 375 mM KCl and 15 mM MgCl2 ) , 2 µl of 0 . 1 M dithiothreitol , and 1 µl of 10 mM of each dNTP , followed by an incubation at 25 °C ( 10 min ) , 42 °C ( 50 min ) and 70 °C ( 15 min ) . One-tenth of each double-stranded cDNA produced was then used as a template in the PCR . The transcripts were amplified from individual cDNAs by PCR using oligonucleotide primers ( sequences available upon request ) designed to each EST . The PCR amplification of a portion ( 209 bp ) of the β-actin gene ( accession no . BI594141 ) using forward primer ( 5′-CTCGAAACAAGAATACGATG-3′ ) and reverse primer ( 5′- ACATGTGCCGTTGTATGATG-3′ ) , previously determined to be present in all developmental stages and both sexes of A . suum [44] , served as a positive control . Samples without template ( no-DNA controls ) were included in each PCR run . The following cycling conditions were employed: one cycle at 94 °C ( 5 min ) , 94 °C ( 30 s ) , 60 °C ( 30 s ) and 72 °C ( 30 s ) for 30 cycles , followed by a final extension of 70 °C ( 7 min ) . Following the PCR , 5 µl of individual amplicons were resolved in ethidium bromide-stained agarose gels ( 2% ) and then photographed upon transillumination . The relative band intensities were analyzed using UVIsoft Image Acquisition and Analysis software ( UVITEC ) . The specificity and identity of individual amplicons were confirmed by direct sequencing using the same primers ( separately ) as employed for their amplification . Clones from the SSH cDNA library with increased hybridization in microarray analysis to the infective L3 compared with other stages were sequenced using standard technology [45] . The nucleotide sequences have been deposited in the GenBank database under accession numbers ES290984-ES291074 . Following the processing of the sequences ( i . e . , removal of vector sequences , quality assurance and clustering ) , contigs or singletons from individual clusters were subjected to BLASTx ( NCBI: www . ncbi . nlm . nih . gov ) and BLASTn ( EMBL-EBI Parasite Genome Blast Server: www . ebi . ac . uk ) analysis to identify putative homologues in C . elegans , other nematodes and other organisms ( e-value of ≤1e-05 ) . Peptides inferred from ESTs were classified functionally using Interproscan ( available at http://www . ebi . ac . uk/InterProScan/ ) employing the default search parameters . WormBase ( www . wormbase . org ) was interrogated extensively for relevant information on C . elegans homologues/orthologues , including RNAi phenotypic , transcriptomic , proteomic and interactomic data . ESTs with homologues/orthologues in C . elegans and other nematodes were also subjected to analysis employing the KEGG Orthology-Based Annotation System ( KOBAS ) ( www . kobas . cbi . pku . edu . cn ) , which predicts the biochemical pathways in which molecules are involved . The open reading frames ( ORFs ) inferred from selected ESTs with orthologues in C . elegans were also subjected to “secretome analysis” using the program SignalP v . 2 . 0 www . cbs . dtu . dk/services/SignalP/ ) , employing both the neural network and hidden Markov models to predict signal peptides and/or anchors [46]–[48] . Also , transmembrane domains were predicted using the program TMHMM ( www . cbs . dtu . dk/services/TMHMM/; [49]–[51] ) , and subcellular localization inferred employing the program WoLF PSORT ( http://wolfpsort . org/; [52] ) . The method established by Zhong and Sternberg [53] was used to predict the interactions for C . elegans orthologues of the L3-enriched molecules from Ascaris . In brief , interaction , phenotypic , expression and gene ontology data from fruitfly , yeast , mouse and human were integrated using a naïve Bayesian model to predict genetic interactions among C . elegans genes ( [45] , [53]; Zhong and Sternberg , unpublished ) . The predicted networks resulting from the analyses were saved in a graphic display file ( gdf ) format and examined using the graph exploration system available at http://graphexploration . cond . org/ . Images were labeled and saved in the joint photographic experts group ( jpeg ) format . To identify molecules transcribed abundantly in the L3 of A . suum , an enriched cDNA library was constructed by SSH . From a total of 3075 clones from this library , 2921 ( 95% ) were shown to contain an insert ( which could be amplified by PCR ) . From 2671 ( 92% ) of these clones , amplicons representing single bands of ∼400 to 600 bp in size were produced . These latter amplicons were arrayed ( in duplicate ) on to slides and then hybridized with Cy3-labeled L3-cDNA or with Cy5-labeled cDNA from eggs , liver/lung L3s , L4s , adult female or adult male of Ascaris . Dye flip was conducted to verify the hybridization data . Of the 2671 ( duplicate ) spots , 1526 had a significant difference in hybridization between infective L3 cDNA and cDNAs from all other stages or sexes of A . suum , of which 515 had a >2 . 0-fold increased hybridization for the L3 . In order to independently verify the hybridization results in the microarray , a PCR-based analysis of a selected subset ( n = 17 ) clones was conducted using specific primer pairs . Having verified the specificity and identity of individual amplicons by sequencing , PCR results were reproducible ( based on multiple runs on different days ) and ∼94% ( 16 of 17 ) concordant with those of the microarray analysis ( not shown ) . There was complete concordance for representative clones associated with a differential signal of ≥3 . 0-fold in the microarray . The clones linked to the 515 spots representing increased transcription ( >2 . 0-fold ) in infective L3 compared with the other developmental stages or sexes included were subjected to sequencing . The 498 sequences ( length: 550±115 bp ) determined were then subjected to detailed bioinformatic analyses . There were 91 unique clusters ( accession numbers ES290984-ES291074 ) , of which 55 were singletons ( sequences determined once ) . Of 56 molecules ( 61 . 5% ) with significant similarity to sequences other than A . suum in the databases interrogated , 50 ( 54 . 9% ) had C . elegans or C . briggsae homologues , and six had similarity to ESTs already sequenced from ascaridoid and/or other parasitic nematodes , and/or other organisms ( Table 1 ) . A significant proportion ( 38 . 4% ) did not have any similarity in sequence to any organisms for which data are presently available . Comparative analysis specifically against A . suum EST data sets ( n∼42 , 000 ) available in public databases confirmed independently that the majority of molecules ( >60% ) were present exclusively in the infective L3 stage or were orphans . As gene ontology ( GO ) provides a hierarchy that unifies the descriptions of biological , cellular and molecular functions [54] , this approach was employed to predict the classification and gene function of molecules enriched in infective L3 of A . suum . A summary of the GO categories of these molecules is displayed in Fig . 1 . Of the 91 contigs , 32 ( 35% ) could be functionally assigned to ‘biological process’ ( n = 38 ) , ‘cellular component’ ( n = 17 ) and ‘molecular function’ ( n = 64 ) . The most common subcategories were gluconeogenesis ( 13% ) and metabolic process ( 13% ) within ‘biological process’ , extracellular region ( 24% ) within ‘cellular component’ , and catalytic activity ( 11% ) and phosphoenolpyruvate carboxykinase activity ( 8% ) within ‘molecular function’ ( Table S1 ) . A focused KOBAS analysis inferred the 50 C . elegans orthologues/homologues to be involved in apoptosis and insulin signaling ( 2% ) , ATP synthesis ( 2% ) , carbon metabolism ( 6% ) , fatty acid biosynthesis ( 2% ) , gap junction ( 2% ) , glucose metabolism ( 6% ) or porphyrin metabolism ( 2% ) , although 34 ( 68% ) of them could not be mapped to a specific metabolic pathway ( Table 2 ) . Of these 50 molecules , small numbers were predicted to be secreted ( 10% ) , anchored ( 2% ) and/or transmembrane ( 12% ) proteins ( Table 2 ) . Functionally , 17 ( 34% ) of the 50 molecules were associated with ( non-wild-type ) RNAi phenotypes in C . elegans , the majority displaying embryonic lethality ( Emb ) ( 13 types; 58 . 8% ) , larval arrest ( Lva ) ( 23 . 5% ) and larval lethality ( Lvl ) ( 47% ) ( Table 2 ) . Extending this analysis , a relatively complex genetic interaction network was predicted for the 17 C . elegans orthologues ( i . e . , with non-wild-type RNAi phenotypes ) ( see Table S2 ) . Statistically highly significant interactions were predicted for nine of the C . elegans genes; the top five interactors are displayed in Fig . 2 . The gene ontology categories for eight selected C . elegans genes ( F33D11 . 10 , F55A12 . 8 , kin-2 , mec-12 , mup-2 , pab-1 , rpl-22 and T21B10 . 2 ) included: embryonic development , egg hatching , larval development and/or growth . The other categories included: positive regulation of growth rate ( F55A12 . 8 , kin-2 , mup-2 , pab-1 , rpl-22 and T21B10 . 2 ) and gamete generation and locomotory behaviour ( kin-2 , mup-2 , pab-1 and F55A12 . 8 , kin-2 , mup-2 , respectively ) . The C . elegans homologue egl-3 was predicted to be involved in proteolysis ( see www . wormbase . org ) . All nine C . elegans orthologues were predicted to interact directly with a total of 296 ( range: 5–75 ) other genes and , in particular , a direct genetic interaction was predicted between pab-1 and T21B10 . 2 ( Fig . 2 ) . The 296 interactors were associated with embryonic and larval development ( n = 198; 66 . 9% ) , information storage and processing ( n = 15; 5 . 1% ) , cellular processes and signalling ( n = 45; 15 . 2% ) and metabolism ( n = 18; 6 . 1% ) ; the precise function of some of the interactors ( n = 20; 6 . 7% ) is presently unknown ( Table S2 ) . The present study investigated transcripts in infective L3s of A . suum using a genomic-bioinformatic platform . The focus was on comparisons with C . elegans homologues/orthologues , because the entire genome sequence of this nematode is known [9] and because there is a wealth of information on the localization and functionality of its molecules ( www . wormbase . org; http://elegans . bcgsc . bc . ca/knockout . shtml ) . The functions of most genes in C . elegans have been assessed using RNAi ( e . g . , [14] , [15] , [17] , [55] , [56] ) in the hermaphroditic stage , whereas there is a paucity of functional information available for Ascaris and other parasitic nematodes of animals [57] , [58] . Following the microarray analysis of >2500 ESTs from the SSH library , 498 cDNAs inferred to be enriched in the L3 , based on hybridization signal , were sequenced and subjected to comprehensive in silico analyses . Of the 91 clusters of molecules categorized , 50 ( 54 . 9% ) had C . elegans homologues/orthologues with loss-of-function phenotypes could be mapped to key pathways . The statistically significant genetic interactions predicted for 9 of the 50 C . elegans orthologues [namely egl-3 , F33D11 . 10 , F55A12 . 8 , kin-2 , mec-12 , mup-2 , pab-1 , rpl-22 and T21B10 . 2 ( = enol-1 ) ] and the interaction network included genes encoding kinases , alpha-tubulins , enolases , troponin and other named and unnamed proteins . Eight of these molecules ( enol-1 , pab-1 , F33D11 . 10 , rpl-22 , F55A12 . 8 mec-12 , mup-2 and kin-2 ) have known or predicted roles in embryonic and larval growth and development , gamete generation , locomotory behaviour or other biological processes in C . elegans ( see www . wormbase . org ) . The enolase encoded by enol-1 is predicted to play a role in glycolysis , gluconeogenesis , phenylalanine , tyrosine and tryptophan biosynthesis ( cf . [59] ) . Since glucose is the main source for ATP production , the alteration in these key glycolytic enzymes may lead to cellular dysfunction , such as impaired ion-motive ATPase required to maintain potential gradients , operate pumps and maintain membrane lipid asymmetry [60] . Bioinformatic analysis for transmembrane helices ( TMHMM ) and peptide signal sequences ( SignalP ) predicted ENOL-1 to be a non-secreted protein localized to the cytoplasm ( cf . Table 2 ) . Nonetheless , enolases are often detected in the excretory/secretory ( ES ) products of parasitic helminths , including adult A . suum [61] , and appear to play a role in the triggering of nitric oxide production by host cells . The enol-1 orthologue of C . elegans has been predicted to interact specifically with the polyadenylate binding protein gene , pab-1 , inferred to be involved in coordinated gene transcription and expression during normal larval development [16] . Poly ( A ) -binding proteins ( PABPs ) are recognized to be central to the regulation of mRNA translation and stability [62] . Present evidence suggests that the expression of PAB-1 is regulated by an oligo-pyrimidine tract in response to cell growth and relates to coordinated growth regulation in C . elegans [62] . Furthermore , gene silencing of pab-1 and its selected interactors ( see Fig . 2 ) leads to embryonic lethal ( Emb ) , slow growth ( Slo ) and sterile progeny ( Stp ) phenotypes ( see www . wormbase . org ) . Another gene ( F33D11 . 10; EST code 4F10; see Table 2 ) which encodes an ATP-dependent RNA helicase and is associated with embryonic lethal ( Emb ) and larval lethal ( Lvl ) RNAi phenotypes , was shown to be highly transcribed in infective L3s of A . suum . Helicases are involved in a variety of RNA metabolic processes , including translation initiation , pre-mRNA splicing , pre-rRNA processing , rRNA maturation and RNA degradation [63] , and are crucial for life cycle progression , sex determination and early embryogenesis in C . elegans [60] . The high transcription levels of a homologue/orthologue in the L3 of A . suum might suggest a similar role in this ascaridoid . Similarly , the coordination of the expression of a large number of genes is required for normal growth and cell proliferation during larval development . The high transcription level for the ribosomal protein gene homologue rpl-22 ( large subunit family member; EST code 26G12 , see Table 2 ) in the infective L3 of A . suum compared with other developmental stages is likely to reflect the substantial rate of cell growth in this stage [64] . The gene ( F55A12 . 8 , EST code 4G11; see Table 2 ) encoding an acetyl-transferase with a putative ATPase domain , shown to be enriched in the L3 of A . suum , was predicted to interact with 75 other genes all involved in energy production and/or RNA processing ( see Table S2 ) . Several molecules involved in ATP synthesis and mitochondrial pathways ( e . g . , cytochrome oxidase c subunits 1 , 2 and 3 , ADP/ATP translocases , NADH dehydrogenases , ATPases and ATP synthetases ) have been reported previously to be highly represented in the L3 stage of Anisakis simplex [65] , thus supporting the proposal that substantial energy is required for larval development as well as the transition from the free-living to the parasitic stage and the invasion of the host . There is also likely to be a substantial energy requirement for muscle contraction linked to larval motility in A . suum , as the L3s penetrate the caecal wall in the porcine host , before undergoing hepato-pulmonary migration [35] . Accordingly , genes encoding a specialized tubulin expressed in mechanoreceptors ( mec-12 , EST code 13E09 ) and a troponin ( mup-2 , EST code 01G03; see Table 2 ) , both predicted to interact with a total number of 32 tubulin- and myosin-encoding genes , also supported a link to extensive muscle contraction and motility in A . suum L3s . Also , neuroactive peptides are required to regulate the responsiveness of nematode larvae to mechanical stimuli [66] . A homologue encoded by egl-3 was shown to be highly transcribed in the L3 of A . suum; EGL-3 is predicted to be a pro-hormone convertase involved in the maturation of neuropeptides and could be associated with mechano-sensory responses and touch sensitivity linked to the host invasion . A regulatory subunit of a cAMP-dependent protein kinase ( kin-2 , EST code 22H01; see Table 2 ) was predicted to interact with 72 other genes all involved in diverse cellular processes , such as nuclear trafficking , and DNA replication and repair ( see Table S2 ) . Based on gene ontology terms , kin-2 is implicated in gamete generation , growth , larval development , post-embryonic body morphogenesis , signal transduction and/or protein amino acid phosphorylation ( see Table S2 ) . Gene silencing of kin-2 in C . elegans leads to phenotypes , such as larval lethal ( Lvl ) , larval arrest ( Lva ) , body morphology defect ( Bmd ) , dumpy ( Dpy ) , uncoordinated ( Unc ) and sterile progeny ( Stp ) ( www . wormbase . org ) , suggesting that its homologue in A . suum is central to larval maturation . The KOBAS analysis predicted the protein KIN-2 to be involved in the insulin-signaling pathway , previously implicated in controlling the exit from dauer in C . elegans and the activation of L3s of the canine hookworm , Ancylostoma caninum , following exsheathment [67] . In a recent study , Brand and Hawdon [68] were able to inhibit ( with a phosphoinositide-3-OH-kinase inhibitor ) the activation of infective L3s of both of the hookworms Ancylostoma caninum and Ancylostoma ceylanicum via the insulin signaling pathway , thus lending some credence to the hypothesis that this pathway plays an critical role in regulating the transition from the free-living to the parasitic stage [68] . Recently , it has been proposed that transcriptional and feeding responses to serum-stimulation in Ancylostoma caninum are regulated by parallel systems , with the insulin signaling pathway playing a significant role in the ‘resumption of feeding’ in activated larvae [69] . Protein kinases are also likely to be involved in pathways linked to sexual maturation in developing larvae . As already proposed for adult stages of H . contortus [45] , the protein kinase gene cdk-1 is predicted to play a pivotal role in the germline , oogenesis and spermiogenesis pathways of this parasitic nematode . Other protein kinases , such as PEPCK , and phosphatases , were shown herein to be transcribed at high levels in the L3 stage compared with other developmental stages of A . suum ( see Table 2 ) , which is in accordance to findings reported recently for Anisakis simplex [65] . Due to their major regulatory effects in eukaryotic signaling events and regulatory and sensory functions , protein kinases have been considered interesting targets for anti-parasitic drugs [70] . In conclusion , this study has given some interesting insights into early molecular processes in the L3 of A . suum . Approximately 60% of the transcripts enriched in the L3 stage of A . suum have homologues/orthologues in C . elegans . The bioinformatic analyses of selected molecules suggest that a complex genetic network regulates or controls larval growth and development in A . suum L3s , and some of these might be involved in or regulate the switch from the free-living to the parasitic stage . Some caution is warranted in drawing conclusions regarding molecular mechanisms regulating the transition to parasitism in parasitic nematodes from information on C . elegans , as latter is a free-living nematode . Also , while the method of data integration is essential for the reliable prediction of genetic interactions , it might limit the capacity of the approach somewhat to infer nematode-specific interactions . As additional datasets of genes and gene functions become available for various parasitic nematodes , more informed inferences can be made regarding the functions of nematode-specific genes , particularly those involved in the transition to parasitism . The imminent genome sequence of A . suum ( http://www . sanger . ac . uk/Projects/Helminths/ ) should all assist in this endeavour . Also , functional analysis of selected molecules representing selected ESTs identified herein , utilizing gene silencing approaches established recently [33] , [34] , could provide some insights into developmental processes in Ascaris and related ascaridoid nematodes and provide avenues for the development of novel approaches for their control .	In the present study , we constructed a cDNA library enriched for molecules of the infective third-stage larva ( L3 ) of Ascaris suum , the common roundworm of pigs . Using the method of suppressive-subtractive hybridization ( SSH ) , we explored transcription of a subset of molecules by microarray analysis and conducted bioinformatic analyses to characterize these molecules , map them to biochemical pathways , and predict genetic interactions based on comparisons with Caenorhabditis elegans and/or other organisms . The results provide interesting insights into early molecular processes in A . suum . Approximately 60% of the L3-enriched molecules discovered had homologues in C . elegans . Probabilistic analyses suggested that a complex genetic network regulates or controls larval growth and development in A . suum L3s , some of which might be involved in or regulate the switch from the free-living to the parasitic stage . Functional studies of these molecules to elucidate developmental processes in Ascaris could assist in identifying new targets for intervention .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "biotechnology", "genetics", "and", "genomics/genomics", "computational", "biology/transcriptional", "regulation", "computational", "biology/molecular", "genetics", "genetics", "and", "genomics/bioinformatics" ]	2008	Genomic-Bioinformatic Analysis of Transcripts Enriched in the Third-Stage Larva of the Parasitic Nematode Ascaris suum
Global genome nucleotide excision repair removes DNA damage from transcriptionally silent regions of the genome . Relatively little is known about the molecular events that initiate and regulate this process in the context of chromatin . We've shown that , in response to UV radiation–induced DNA damage , increased histone H3 acetylation at lysine 9 and 14 correlates with changes in chromatin structure , and these alterations are associated with efficient global genome nucleotide excision repair in yeast . These changes depend on the presence of the Rad16 protein . Remarkably , constitutive hyperacetylation of histone H3 can suppress the requirement for Rad7 and Rad16 , two components of a global genome repair complex , during repair . This reveals the connection between histone H3 acetylation and DNA repair . Here , we investigate how chromatin structure is modified following UV irradiation to facilitate DNA repair in yeast . Using a combination of chromatin immunoprecipitation to measure histone acetylation levels , histone acetylase occupancy in chromatin , MNase digestion , or restriction enzyme endonuclease accessibility assays to analyse chromatin structure , and finally nucleotide excision repair assays to examine DNA repair , we demonstrate that global genome nucleotide excision repair drives UV-induced chromatin remodelling by controlling histone H3 acetylation levels in chromatin . The concerted action of the ATPase and C3HC4 RING domains of Rad16 combine to regulate the occupancy of the histone acetyl transferase Gcn5 on chromatin in response to UV damage . We conclude that the global genome repair complex in yeast regulates UV-induced histone H3 acetylation by controlling the accessibility of the histone acetyl transferase Gcn5 in chromatin . The resultant changes in histone H3 acetylation promote chromatin remodelling necessary for efficient repair of DNA damage . Recent evidence suggests that GCN5 plays a role in NER in human cells . Our work provides important insight into how GG-NER operates in chromatin . DNA repair is a central facet of DNA metabolism , and nucleotide excision repair ( NER ) is an important component of a complex cellular response that prevents the loss of genetic information caused by DNA damage . Its importance for the repair of ultraviolet ( UV ) light induced DNA lesions is dramatically illustrated in humans who suffer from the autosomal recessive disease xeroderma pigmentosum ( XP ) . Defective NER in these individuals severely predisposes them to sunlight-induced skin cancers [1] . The excision of lesions from non-transcribed regions of the human genome involves the global genome nucleotide excision repair ( GG-NER ) pathway , which in yeast requires the Rad7 and Rad16 GG-NER proteins [1]–[3] . Many of the core enzymatic activities associated with NER have been determined in some detail , but an understanding of how the process functions in relation to chromatin structure is still in its infancy . DNA in eukaryotic cells is packaged into nucleosomes that form as a result of the wrapping of DNA around histone octamers . Higher-order chromatin structures are formed when nucleosomal arrays are further compacted . Chromatin has a major impact on DNA metabolic processes by controlling the functional interaction of proteins with regulatory and other elements in the DNA [4] , [5] . Chromatin remodelling and histone modification are two major mechanisms that contribute to this regulation . Both processes have roles in controlling gene transcription [6] , [7] and in NER [8]–[10] . GG-NER in S . cerevisiae requires both the Rad7 and Rad16 proteins [11]–[13] . Rad16 is a member of the SWI/SNF super-family of chromatin remodelling factors [14] . This superfamily of proteins exhibits ATPase activity that is stimulated by DNA or chromatin [15] , [16] , and all SWI/SNF-like proteins generate superhelical tension in linear DNA fragments via a DNA translocase activity associated with their ATPase function [17] , [18] . The generation of superhelicity in DNA is a common mechanism of SWI/SNF-like chromatin remodelling complexes for altering chromatin structure [17] . We recently reported that a Rad7 and Rad16 containing protein complex also has DNA translocase activity . However , it is unable to slide nucleosomes unlike some SWI/SNF superfamily complexes [19] . Although Rad16 is a member of the SWI/SNF super-family , direct evidence of a role in chromatin remodelling is lacking . In this study we have addressed how GG-NER functions during DNA repair in chromatin in yeast cells . UV irradiation stimulates histone H3 acetylation at lysine 9 and 14 ( K9 , K14 ) and chromatin remodelling , both globally and in the MFA2 gene [8] , [20] . However , these studies were not able to establish the precise relationship between these two events with respect to their effect on NER , nor did they inform on how these UV induced changes were regulated . Recently we showed that UV induced histone H3 acetylation depends on the Rad16 GG-NER protein . Furthermore , constitutively elevating histone H3 acetylation levels in the MFA2 gene suppresses the requirement for Rad7 and Rad16 during GG-NER [10] . Gene regulation of MFA2 involves the yeast general repressor complex Ssn6-Tup1 [21] . Deletion of TUP1 results in constitutively elevated histone H3 acetylation and modified chromatin structure at the promoter of the MFA2 gene [22]–[24] . Remarkably , Rad7 and Rad16 independent GG-NER occurs in the promoter region of MFA2 in TUP1 deleted cells . This suggested that Rad7 and Rad16 might regulate chromatin structure in response to UV damage during GG-NER via the regulation of histone H3 acetylation levels in chromatin . In this report we demonstrate that the GG-NER proteins in yeast promote chromatin remodelling necessary for efficient DNA repair , revealing how this processes is regulated in response to DNA damage . We define a series of UV induced , Rad7 and Rad16 dependent events that control histone H3 acetylation which in turn drives chromatin remodelling necessary for efficient GG-NER in yeast . Histone H3 acetylation status at MFA2 is determined by Rad7 and Rad16 controlling the occupancy of the Gcn5 histone acetyl transferase on chromatin in response to UV irradiation . These UV induced histone H3 modifications are required for chromatin remodelling necessary for efficient GG-NER in the region . Acetylation of histone H3 after UV irradiation depends on the presence of Rad16 and this process is necessary for efficient GG-NER [10] . Figure 1A shows that UV induced histone H3 acetylation ( K9 , K14 ) at the regulatory region of the MFA2 gene also requires the GG-NER factor Rad7 . Therefore Rad7 and Rad16 function in combination to increase histone H3 acetylation levels at MFA2 in response to UV . Since UV induced histone H3 acetylation correlates with efficient GG-NER and elevated levels of histone H3 acetylation at MFA2 suppress the requirement for Rad7 and Rad16 during GG-NER [10] , this poses the question as to how Rad7 and Rad16 control histone H3 acetylation . Rad7 and Rad16 control UV induced histone H3 acetylation at MFA2 and these proteins are not required for GG-NER when histone H3 acetylation is constitutively elevated in the region . We speculated that during GG-NER Rad7 and Rad16 mediate changes in histone H3 acetylation after UV by controlling the accessibility of the histone acetyl transferase Gcn5 , which regulates histone H3 acetylation at MFA2 . To test this we performed Gcn5 chromatin immunoprecipitation ( ChIP ) experiments in the promoter of the MFA2 gene . Figure 1B shows the relative levels of Gcn5 binding at the repressed MFA2 promoter in the absence of UV ( U ) or following UV irradiation at the times indicated ( 0 , 15 and 60 minutes ) in wild type , rad7Δ and rad16Δ strains . In the absence of UV irradiation , background levels of Gcn5 occupancy are detected in all three strains . However , after UV , a rapid increase in Gcn5 occupancy is observed in the wild type , but not in the rad7Δ or rad16Δ strains . In wild type cells , decreasing levels of Gcn5 occupancy at MFA2 were observed with increasing time after UV irradiation and as repair occurred . Therefore in wild type cells Gcn5 occupies the promoter of the MFA2 gene at low levels , resulting in background levels of histone H3 acetylation at MFA2 in the absence of UV . Following UV , a Rad7 and Rad16 dependent increase in Gcn5 occupancy ( Figure 1B ) and histone H3 acetylation ( Figure 1A ) is observed at MFA2 . We measured chromatin changes at MFA2 in TUP1 deleted α-cells where histone H3 is hyperacetylated and where the requirement for Rad7 and Rad16 during GG-NER is abrogated . Tup1 is a component of a repressor complex that regulates gene expression at MFA2 . In α mating type cells where the chromatin is repressed , the deletion of TUP1 correlates with altered chromatin structure in MFA2 and other TUP1 regulated genes [10] , [23] , [25] . To confirm this we compared the MNase sensitive sites in naked DNA and chromatin from wild type and tup1Δ α-cells on both DNA strands of the MFA2 promoter region ( Figure 2A and 2B , Figure S1 , and Text S1 ) . Figure 2A and 2B reveal that MNase digestion is almost identical between tup1Δ α-cell chromatin and naked DNA , whereas chromatin from wild type α-cells exhibits significantly reduced MNase digestion due to protection by the positioned nucleosomes designated N-1 and N-2 . Autoradiograms are shown in Figure S1 . Therefore chromatin structure is altered in TUP1 deleted α-cells . To further explore the effect of histone acetylation on chromatin structure we examined the accessibility of the restriction enzyme RsaI to nucleosomal core DNA . Chromatin was treated with RsaI restriction enzyme and purified DNA was digested using HaeIII . Restriction with HaeIII generated a 599 bp DNA fragment ( Figure 2C ) . A double restriction digest with RsaI and HaeIII of naked DNA generated a smaller fragment of 419 bp ( Figure 2C ) . In wild type α-cells MFA2 is repressed by positioned nucleosomes and RsaI has only limited access to the DNA at its restriction site located within nucleosome N-2 . RsaI digests only 8 . 7±1 . 9% of the total MFA2 fragments ( Figure 2D , Lane 2 ) . However , in wild type a-cells and tup1Δ α-cells ( Figure 2D , Lanes 1 and 5 ) where MFA2 is derepressed , RsaI cuts in both strains to the extent of 60 . 3±1 . 0% and 74 . 5±2 . 2% of the total HaeIII fragments , respectively . Therefore , restriction enzyme sites are masked in chromatin from wild type α-cells , but are accessible in chromatin from wild type a-cells and tup1Δ α-cells . The relationship between chromatin accessibility and histone H3 acetylation status was examined by measuring the histone H3 acetylation levels in the MFA2 promoter in the absence of and following UV irradiation . In Figure 1A , and in Figure 3 , a three fold increased UV induced histone H3 acetylation is observed in wild type α-cells . In the tup1Δα strain an eight-fold elevation in constitutive histone H3 acetylation is observed and no further increase in H3 acetylation is seen following UV irradiation . A similar result was noted in tup1Δrad16Δ α-cells . Intriguingly , in tup1Δgcn5Δ α-cells histone H3 acetylation remains constitutively high , despite the loss of the Gcn5 histone acetyl transferase in this strain . Figure 2D lanes 3 and 4 demonstrate that in RAD16 or GCN5 deleted α-cells chromatin structure remains closed as evidenced by low-level RsaI cutting observed ( 8 . 2%±2 . 3% and 9 . 0%±2 . 6% respectively ) , similar to levels seen in wild type α cells ( Figure 2D , Lane 2 ) . In tup1Δrad16Δ double mutant α-cells , open chromatin structure is retained as high levels of restriction enzyme cutting are observed ( 73 . 1%±3 . 4% ) ( Figure 2D , Lane 7 ) , similar to levels seen in tup1Δ α-cells ( Figure 2D , Lane 5 ) . An open chromatin structure was also seen in tup1Δgcn5Δ α-cells shown in lane 8 ( 75 . 1%±1 . 0% RsaI enzyme cutting ) . This was unexpected , since Gcn5 is deleted in this strain . But the result is consistent with the constitutively high histone H3 acetylation level detected ( Figure 3 ) , explaining the increased chromatin accessibility observed in this strain ( Figure 2D , lane 8 ) . Note that deleting RAD16 in tup1Δgcn5Δ α-cells to create a tup1Δrad16Δgcn5Δα triple mutant strain results in significantly reduced restriction enzyme cutting indicating the presence of a more repressive chromatin structure at the site ( 45 . 2%±3 . 4% RsaI enzyme cutting ) ( Figure 2D , lane 6 ) . Rad7 and Rad16 independent GG-NER occurs in genomic regions where constitutively elevated levels of histone H3 acetylation are observed , such as the promoter of MFA2 in tup1Δ α-cells ( Figure 4 , Figure S2 , and Text S1 ) [10] . The absence of CPD repair at MFA2 in the tup1Δ , rad14Δ mutant proves that repair in the tup1Δ , rad16Δ α-cells occurs unequivocally via Rad7 and Rad16 independent GG-NER [10] . This suggested that Rad7 and Rad16 mediated UV induced histone H3 acetylation is necessary for efficient GG-NER . We examined this by measuring repair of CPDs in the promoter of MFA2 in tup1Δrad16Δ α-cells and in tup1Δrad16Δgcn5Δ α-cells , where the histone acetyl transferase gene GCN5 is deleted ( Figure S2 ) . Figure 4 shows the time taken to remove 50% of the CPDs ( T50% ) from the nontranscribed strand at the positions indicated . As seen previously , GG-NER in tup1Δrad16Δ α-cells , or tup1Δrad7Δα is restored to near wild type levels compared to the lack of repair seen in the rad16Δ α single mutant cells ( Figure 4 ) . Therefore Rad7 and Rad16 are not required for GG-NER at MFA2 when histone H3 acetylation levels are elevated creating an open chromatin structure ( Figure 3 and Figure 2D , lane 5 ) . To determine the significance of UV induced histone H3 acetylation levels and chromatin structure on efficient GG-NER we examined repair in tup1Δrad16Δgcn5Δ α-cells . Figure 4 reveals that loss of hisotne H3 acetylation which causes reduced chromatin accessibility [See Figure 3 and Figure 2D , lane 6] in this triple mutant strain , results in significantly reduced GG-NER in the region of nucleosomes N-1 and N-2 ( see Figure 2A and 2B ) upstream of the transcriptional start site ( Figure 4: open diamonds ) . Repair in a small region in the vicinity of the transcriptional start site is unaffected . Therefore , the Rad7 and Rad16 independent GG-NER observed at MFA2 in TUP1 deleted cells is primarily due to the constitutively elevated levels of histone H3 acetylation and open chromatin structure in the region . We observed only background levels of histone H3 acetylation in the triple mutated strain and this results in a less accessible chromatin structure and reduced NER activity . This might imply that histone acetylation is not solely responsible for chromatin remodeling necessary for NER , because in the absence of detectable histone H3 acetylation , chromatin remains partially ‘open’ . However our observations reveal that histone H3 acetylation does play a significant role in chromatin remodeling necessary for efficient NER . We also noted that in the absence of Gcn5 , histone H3 acetylation at K9 and K14 can still be detected and this acetylation is dependent on Rad16 , since acetylation is lost in the triple mutated strain ( Figure 3 ) . This underscores the significance of Rad16 in controlling histone acetylation status in the region , and demonstrates that redundancy exists with respect to the histone acetyl transferase that can be recruited to the chromatin . These observations are considered in more detail in the Discussion section . Rad16 has two known catalytic functions: a DNA translocase activity associated with the ATPase domain [19] , and an E3 ubiquitin ligase activity associated with the C3HC4 RING domain embedded within the ATPase domain [see Figure 5A] [26] , [27] . We introduced point mutations into each of the catalytic domains of Rad16 to examine their effect on GG-NER . The ATPase activity was tested by mutating the conserved Walker A box catalytic residue lysine 216 to alanine ( K216A ) . This mutation creates an ATPase null mutant [27] . We call this the RAD16 ATPase mutant . We also mutated the RING domain of Rad16 to test the role of the E3 ligase activity . RING domains have conserved cysteine and histidine residues that coordinate two zinc atoms . A conserved hydrophobic residue is also essential for the interaction between the RING domain and specific E2 ubiquitin conjugating enzymes . We made two point mutations in conserved cysteine and histidine residues; cysteine 552 to alanine and histidine 554 to alanine ( C552A , H554A ) . We call this the RAD16 RING mutant . Finally , we tested the effect of mutating both the ATPase and RING domains of Rad16 by introducing these mutations ( K216A , C552A , H554A ) into a single strain . Figure 5B compares the UV sensitivity in each of these strains compared to the parental wild type , and Rad16 deleted strains . The individual RAD16 ATPase and RING mutant strains show intermediate UV sensitivity . Whereas the double mutant strain is as sensitive as the Rad16 deleted strain . These observations confirm previous findings that both the ATPase and RING E3 ligase catalytic activities contribute independently to efficient GG-NER and UV survival [27] . We examined the effect of the RAD16 point mutations on the level of histone H3 acetylation and Gcn5 occupancy at MFA2 . We performed histone H3 acetylation ( K9 , K14 ) ChIP experiments in the promoter of MFA2 . Figure 5C shows the relative levels of acetylated histone H3 at the repressed MFA2 in the absence of UV ( U ) or after UV irradiation at the times indicated ( 0 , 15 and 60 minutes ) . In the absence of UV irradiation , background levels of histone H3 acetylation are detected in all four strains . However , following UV , a rapid increase in histone H3 acetylation is observed in the wild type strain and in the single RAD16 ATPase and RING mutated strains , but not in the RAD16 ATPase , RING double mutant strain , where UV induced histone H3 acetylation is abolished . Similar results were obtained when Gcn5 occupancy was examined in these strains , Figure 5D . Finally , we examined the repair of CPDs at MFA2 in wild type and each of the point mutated strains described above ( Figure 6A , Figure S4 , and Text S1 ) . A typical autoradiogram is shown in Figure S3 . In Figure 6A repair was expressed as the time taken to remove 50% of the CPDs ( T50% ) from the nontranscribed strand at the nucleotide positions indicated . As seen previously , GG-NER in the nontranscribed strand of MFA2 proceeds efficiently in wild type cells ( Figure 6A ) . Mutating either the ATPase domain or the RING domain of RAD16 individually impairs UV lesion removal , but GG-NER continues less efficiently ( Figure S4 and Text S1 ) . This correlates with the near wild type levels of histone H3 acetylation , Gcn5 occupancy ( Figure 5C and 5D ) , and intermediate UV sensitivity ( Figure 5B ) of these strains . GG-NER in the ATPase , RING domain double mutated strain is abolished over almost the whole of the MFA2 promoter region and occurs at a level seen in the RAD16 deleted strain ( Figure 6A and Figure 4 ) . This correlates with the lack of UV induced histone H3 acetylation and Gcn5 occupancy in the region ( Figure 5C and 5D ) , and the high level of UV sensitivity ( Figure 5B ) observed in this strain . These observations demonstrate that the ATPase and RING domains of Rad16 function in combination to regulate UV induced Gcn5 occupancy and histone H3 acetylation status , which ultimately controls chromatin structure at MFA2 in response to DNA damage . In Figure 6B we demonstrate the lack of UV induced chromatin remodelling observed in the ATPase , RING double mutated strain compared to the remodelling observed in the wild type strain using the restriction enzyme accessibility assay described earlier in Figure 2C and 2D . This confirms the importance of chromatin remodelling to the GG-NER process . We've shown that Rad7 and Rad16 proteins are required for UV induced histone H3 acetylation at MFA2 . These GG-NER factors regulate the acetylation status by controlling the occupancy of the histone acetyl transferase Gcn5 at this locus . In unirradiated wild type cells only background levels of Gcn5 are detected at MFA2 , whereas increased Gcn5 occupancy is seen following UV irradiation . This correlates with increased acetylation of histone H3 observed in wild type cells in response to UV . In Rad7 and Rad16 deleted cells no increased Gcn5 occupancy or increased histone H3 acetylation is observed at MFA2 in response to UV . This indicates that both events are Rad7 and Rad16 dependent in wild type cells . Increased histone acetylation levels have long been associated with changes in chromatin structure , particularly with respect to generating an open chromatin structure needed for gene transcription [8] . To address the impact of histone H3 acetylation on chromatin structure at MFA2 in response to UV , we employed two methods: a nucleosome mapping assay , and a restriction enzyme accessibility assay . We examined these events in TUP1 deleted cells since Tup1 is a component of the Ssn6-Tup1 general repressor complex , which regulates gene expression in a range of genes including MFA2 . In α mating type yeast cells MFA2 is repressed , but in TUP1 deleted α-cells histone H3 levels at MFA2 are constitutively elevated which results in an open chromatin structure at MFA2 . We found that cells with elevated levels of histone H3 acetylation as is the case when TUP1 alone , TUP1 , RAD16 and TUP1 , GCN5 are deleted in α-cells also have an open chromatin structure as demonstrated in the restriction enzyme accessibility assay in Figure 2D . We were surprised to detect elevated levels of histone H3 acetylation , and open chromatin structure in TUP1 , GCN5 deleted cells since the histone acetyl transferase Gcn5 known to function at MFA2 in wild-type cells is absent in this strain [8] . We speculate that in GCN5 deleted cells , an alternative histone acetyl transferase can substitute for GCN5 . Significantly , this redundancy is dependent on Rad16 , since in tup1Δrad16Δgcn5Δ triple mutant cells , histone H3 acetylation is reduced to background levels and the open chromatin structure is altered to a more repressed state . We suggest that these observations underscore the significance of Rad16 in regulating the histone acetylation status of chromatin in the region , and indicate that Rad16 determines histone acetyl transferase recruitment to the chromatin . We examined the significance of histone H3 acetylation at MFA2 on lesion removal during GG-NER by measuring repair in TUP1 deleted cells . To determine whether the elevated levels of histone H3 acetylation and open chromatin structure observed in TUP1 , RAD16 deleted cells promotes the repair observed in these cells , we examined repair in the tup1Δrad16Δgcn5Δ triple mutant strain where histone H3 levels are diminished to background levels , and chromatin accessibility is significantly reduced . We found that the near wild type level of repair observed in the TUP1 , RAD16 deleted cells was significantly reduced in the tup1Δrad16Δgcn5Δ mutant cells indicating the importance of histone H3 acetylation and chromatin structure to the repair observed in the region . Despite detecting only background levels of histone H3 acetylation in the triple mutant strain , we still detect a partially open chromatin structure , which results in a reduced but not totally defective NER efficiency . Our findings demonstrate that UV induced histone H3 acetylation is playing an important role in chromatin remodelling during NER , but recognise that other factors might also be influencing the process . Finally , we investigated whether either of the known activities associated with Rad16 was responsible for controlling this series of events . Strains carrying point mutations in the ATPase or the C3HC4 RING domain of Rad16 , or a double mutant carrying both these mutations were examined . Previous studies showed that the Rad16 ATPase mutant has no detectable ATPase function [27] , and the Rad16 RING mutant has no E3 ligase activity . UV survival experiments showed an intermediate UV sensitivity for the Rad16 ATPase and RING domain single mutants , while the double domain mutant showed higher UV sensitivity , similar to that observed in the Rad16 deleted strain ( Figure 5 ) . Therefore both the ATPase and E3 ligase functions of Rad16 are required for efficient GG-NER , in agreement with previous studies [26] , [27] . This observation suggests that a UV induced ubiquitination event , possibly involving a histone or alternatively another NER factor , is likely important in initiating the chromatin remodelling process . It is established that UV induced histone ubiquitination is observed in human cells and is necessary for efficient NER [28] . We showed that both Rad16 domains contribute to efficient GG-NER . Figure S4 shows reduced levels of CPD removal from the nontranscribed strand of the MFA2 promoter in each of the single domain mutant strains , and defective lesion removal only in the double domain mutant strain ( Figure 6A ) . This observation correlates with the level of Gcn5 occupancy and histone H3 acetylation levels observed in these strains ( Figure 5C and 5D ) . Loss of UV induced Gcn5 occupancy and histone H3 acetylation is only observed in the double mutant strain suggesting that the ATPase and RING domains of Rad16 are both required for efficient chromatin remodelling during GG-NER . Figure 6B confirms that efficient GG-NER observed in the wild type strain is dependent on UV induced chromatin remodelling since failure to remodel chromatin in the ATPase , RING double mutant strain results in defective repair . Collectively our results demonstrate that during GG-NER the Rad7 and Rad16 proteins promote efficient repair by regulating histone acetyl transferase occupancy on chromatin in response to UV . This explains how histone H3 acetylation status and chromatin structure is controlled in response to DNA damage , and that this process is necessary for efficient GG-NER . Our results are consistent with a model for UV induced chromatin remodelling in yeast cells described in Figure 7 ( See Text S1 for further discussion ) . It was recently reported that Gcn5 is recruited to sites of UV induced DNA damage in human cells [29] . However , its role in chromatin remodelling was not determined . Our studies provide important insight into how chromatin is remodelled to facilitate efficient DNA repair following UV induced DNA damage in human cells . The details of plasmids and yeast strains used in this study can be found in Text S1 and in Table S1 . Cells were grown in synthetic complete medium with leucine dropout ( SC-leu− ) to mid-log phase ( around 2×107 cells/ml ) . Following mild sonication , cells were plated on SC-leu− agar plates , then irradiated with the germicidal UV lamp at the indicated UV doses . Following irradiation , plates were immediately wrapped in foil and incubated for 3 days at 30°C . Survival was derived from the number of colonies relative to that in the unirradiated control . Experiments were performed in triplicate . This was performed as in Yu et al , [8] with modifications . In brief , proteins were cross-linked to DNA by addition of formaldehyde to 100 ml yeast cells ( about 2×109 cells ) to a final concentration of 1% for 20 min at room temperature . 5 . 5 ml of Glycine ( 2 . 5 M ) was added to stop cross-linking . Cells were lysed by the addition of 0 . 5 ml of glass beads ( Sigma ) , and vortexed for 30 min on a Disruptor Genie at 4°C . The cell lysate was sonicated to generate DNA fragments ranging from 200–500 bps in length . Sonication was carried out using the Bioruptor ( Diagenode ) following the manufacturer's instruction at 4°C , power position “H” , 20 seconds on and 40 seconds off for 6 cycles . 50 µl of pre-washed pan mouse or anti-rabbit IgG Dynabeads was incubated with 2 . 5 µg of mouse anti-Myc ( 9E11 , Abcam ) antibody , or 2 . 5 µl of rabbit anti-acetyl histone H3 ( at K9 and K14 , Upstate Biotechnology ) at 30°C for 30 min , then the antibody bound Dynabeads were subsequently incubated with 100 µl sheared chromatin solution equivalent to 108 cells in a total volume of 0 . 5 ml for 3 hours at 21°C . After elution with pronase buffer ( 125 mM Tris pH 7 . 5 , 25 mM EDTA , 2 . 5% SDS ) from Dynabeads beads , formaldehyde cross-linking was reversed by incubating the eluate at 65°C overnight in the presence of 125 µg of pronase . Finally , DNA was purified with PCR purification kit ( QIAGEN ) . 50 µl of chromatin solution was taken as input control for each sample . Quantitative PCR was performed in real time using iQ SYBR Green Supermix ( Bio-Rad ) and diluted DNA in the Bio-Rad MyiQ . PCR was performed in triplicate for each sample , and melting curves were executed to ensure single PCR products . Primers for amplifying nucleosome N-2 in the promoter region of MFA2 are: These were carried out as described previously [30] Chromatin was prepared as described in Teng et al , [30] with modifications . In brief , cells from 200 ml YPD ( 2–4×109 cells ) were pelleted , washed in cold PBS and 1 M Sorbitol , and spheroplasted in 1 m lysis solution ( 1 M Sorbitol , 5 mM 2-mercatoethanol ) containing 20 mg of Zymolyase-20T per 1 g of cells for 20 min at 30°C . Spheroplasts were washed with cold 1 M Sorbitol , and lysed in 7 ml Ficoll solution ( 18% Ficoll , 20 mM KH2PO4 , pH 6 . 8 , 1 mM MgCl2 , 0 . 25 mM EGTA , 0 . 25 mM EDTA ) per 1 g cell . Collecting nuclei by centrifugation , and washing the pellet with RsaI restriction enzyme reaction buffer , chromatin from 4×108 cells was incubated with 300 units of RsaI for 3 hours at 37°C . Purified DNA from the digest was subjected to a secondary digestion by HaeIII and then resolved on 1 . 5% agarose gel in 1×TAE buffer . Southern transfer of DNA to GeneScreen Plus Hybridization Transfer Membrane ( Perkin Elmer ) preparation was described previously [20] . These were undertaken as described in Teng et al , [20] . Details are available in Text S1 . These were undertaken as described by Reed et al , [31] and Teng et al , [32] . Details are available in Text S1 .	Protection against genotoxic insult requires a network of DNA damage responses , including DNA repair . Inherited DNA repair defects cause severe clinical consequences including extreme cancer susceptibility . Despite detailed mechanistic understanding of the core reactions , little is known about the molecular events that initiate and regulate these processes as they occur in chromatin . We study the conserved nucleotide excision repair pathway in Saccharomyces cerevisiae . This pathway removes a broad spectrum of DNA damages including UV radiation–induced damage . Patients with mutations in genes involved in this process suffer dramatically elevated levels of skin and other cancers . Here we demonstrate how a group of genes involved in repair of transcriptionally silent regions of the genome , a process called global genome repair , modifies chromatin structure following UV irradiation to promote efficient removal of DNA damage from the genome . We show that the concerted action of global genome repair genes combine to regulate histone acetyl transferase accessibility to the chromatin in response to UV damage . In this way , global genome repair regulates histone H3 acetylation status , which ultimately regulates chromatin structure promoting efficient DNA repair in the genome . Our results contribute a significant advance in our understanding of how chromatin is processed during DNA repair .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "nucleic", "acids", "dna", "biology", "molecular", "cell", "biology", "dna", "repair", "molecular", "biology" ]	2011	How Chromatin Is Remodelled during DNA Repair of UV-Induced DNA Damage in Saccharomyces cerevisiae
Vector-borne diseases account for more than 17% of all infectious diseases , causing more than one million deaths annually . Malaria remains one of the most important public health problems worldwide . These vectors are bloodsucking insects , which can transmit disease-producing microorganisms during a blood meal . The contact of culicids with human populations living in malaria-endemic areas suggests that the identification of Plasmodium genetic material in the blood present in the gut of these mosquitoes may be possible . The process of assessing the blood meal for the presence of pathogens is termed ‘xenosurveillance’ . In view of this , the present work investigated the relationship between the frequency with which Plasmodium DNA is found in culicids and the frequency with which individuals are found to be carrying malaria parasites . A cross-sectional study was performed in a peri-urban area of Manaus , in the Western Brazilian Amazon , by simultaneously collecting human blood samples and trapping culicids from households . A total of 875 individuals were included in the study and a total of 13 , 374mosquito specimens were captured . Malaria prevalence in the study area was 7 . 7% . The frequency of households with at least one culicid specimen carrying Plasmodium DNA was 6 . 4% . Plasmodium infection incidence was significantly related to whether any Plasmodium positive blood-fed culicid was found in the same household [IRR 3 . 49 ( CI95% 1 . 38–8 . 84 ) ; p = 0 . 008] and for indoor-collected culicids [IRR 4 . 07 ( CI95%1 . 25–13 . 24 ) ; p = 0 . 020] . Furthermore , the number of infected people in the house at the time of mosquito collection was related to whether there were any positive blood-fed culicid mosquitoes in that household for collection methods combined [IRR 4 . 48 ( CI95%2 . 22–9 . 05 ) ; p<0 . 001] or only for indoor-collected culicids [IRR 4 . 88 ( CI95%2 . 01–11 . 82 ) ; p<0 . 001] . Our results suggest that xenosurveillance can be used in endemic tropical regions in order to estimate the malaria burden and identify transmission foci in areas where Plasmodium vivax is predominant . Vector-borne diseases account for more than 17% of all infectious diseases , and cause more than one million deaths annually . Vectors are living organisms that can transmit infectious diseases between humans or from animals to humans . These vectors are bloodsucking insects , which ingest disease-producing microorganisms during a blood meal from an infected host ( human or animal ) and can subsequently inject them into a new host during a new blood meal . Mosquitoes are among the best-known disease vectors . Three genera are responsible for transmitting a series of life-threatening diseases worldwide , however mostly in tropical areas: 1 ) Aedes: vector of Chikungunya , dengue , Zika , Rift Valley and yellow fever; 2 ) Anopheles: vector of malaria , Saint Louis encephalitis , West Nile fever and several Anopheles A and B orthobunyaviruses and 3 ) Culex: Japanese encephalitis , lymphatic filariasis , Saint Louis encephalitis , Oropouche and West Nile fever [1 , 2] . Many of these diseases are preventable through informed , protective measures . Surveillance is critical for the prediction of future disease outbreaks and epidemics at an early stage , as well as for identifying transmission hotspots . Among vector-borne diseases , malaria remains one of the most important public health problems worldwide . It is estimated that malaria transmission still occurs in 91 countries and territories of the world , and causes an estimated 216 million clinical episodes and around 445 , 000 deaths globally every year [3] . An unknown , and much higher , number of individuals in malaria-endemic areas silently carry malaria parasites , which provides a reservoir for malaria transmission . The challenge of identifying them has been recognized as a major difficulty in regard to the adequate control and eventual elimination of malaria [4] . The majority of mosquito species is hematophagous , and relies on blood from vertebrates for nourishment and reproduction . When engorged , mobility is limited and they can be easily collected via aspiration . Blood meal analysis is then carried out in order to detect the presence of pathogens , a process termed ‘xenosurveillance’ [5 , 6] . Experimentally , mosquitoes can be used as biological syringes’ so we can accurately quantify the presence of viruses or other microorganisms in their midgut and also evaluate the potential role of small vertebrates in the transmission cycle of arboviruses [7] . Such methods have shown , for instance , transmission of human pathogens , such as hepatitis B virus ( HBV ) [8 , 9] and dengue virus ( DENV ) by Culex quinquefasciatus [10] and hepatitis C virus ( HCV ) by Culex pipiens complex [11 , 12] . Field studies have confirmed that vertebrate viral pathogens that are not vector-borne could also be detected in the blood meals of mosquitoes belonging to a variety of taxa [13 , 14] . H5N1 virus sequences were found in blood-engorged mosquitoes collected near a poultry farm during an outbreak of avian influenza in Thailand [15] . To survey DENV and Japanese encephalitis circulation in this same country , an enzyme-linked immunosorbent assay detected virus-reactive antibodies in blood meals collected from culicids irrespective of the mosquito species’ ability as a vector [16] . Nucleic acid sequences from human papillomavirus 23 ( HPV23 ) , human herpesvirus 1 and human papillomavirus type 112 ( HPV112 ) were identified in mosquito samples from San Diego , California [17] . In Central Brazil , DENV-4 was detected in several culicids , especially Cx . quinquefasciatus [18] . In Liberia , An . gambiae bloodfed mosquitoes resting indoors were found positive for human skin-associated microbes ( e . g . Staphylococcus epidermidis and Propionibacterium acnes ) , Epstein-Barr virus ( EBV ) and canine distemper virus ( CDV ) signatures [6] . Human bacterial pathogens such as Rickettsia spp . [19 , 20] , Francisella tularensis [21] and Borrelia [22 , 23] were also found in field-collected culicids . In the case of malaria , clinical , parasitological and serological markers of transmission , as detected in humans , have been routinely and historically used to detect transmission pockets and guide hotspot-targeted interventions . However , in more recent times , interest has also shifted to evaluating transmission through the exploration of the burden of low-density ( formerly described as “asymptomatic” albeit now recognized to be potential causes of more silent , but still important , clinical consequences ) infections in low transmission areas [24–27] . Nevertheless , current malaria surveillance systems can be expensive , laborious and useless for sampling numbers of exposed individuals over space and time . In Liberia , a pathogen surveillance strategy using xenosurveillance has found the presence of P . falciparum in An . gambiae; however , this result was expected since malaria is holoendemic in this region and therefore the mosquito or/and the human upon which it fed may be infected [6] . Fauver et al . [9] have demonstrated the viability of xenosurveillance as a tool for sampling human blood in order to detect circulating pathogens . However , to the best of our knowledge , xenosurveillance of malaria using highly anthropophilic culicids , frequently more abundant than anophelines , has been never validated in regards to its fundamental utility , feasibility in field timeframes and technically relevant concentrations . In this work , we hypothesized that , in this surveillance approach , molecular detection of Plasmodium in blood fed mosquitoes reflect the malaria burden and make malaria surveillance more feasible in tropical localities where Culex is abundant . This study was conducted in the Brasileirinho , Puraquequara and Ipiranga communities , in a peri-urban area of Manaus , Western Brazilian Amazon . According to a census carried out by the field team before the beginning of the study , approximately 2 , 400 inhabitants were living in the study area , of around 72 km2 . The general characteristics of these communities were previously described [28 , 29] . A cross-sectional sampling was performed from January to March , 2014by simultaneously collecting human blood samples and trapping culicids from 233 households located in this area . Sampling included all inhabitants living in these houses that were willing to participate in the study . For each study participant , a questionnaire was completed , containing personal malaria preventive measure information and the participants’ history of malaria episodes in the preceding 30 days ( Table 1 ) . Upon enrolment , a 300 μL blood sample was collected from the participant via finger puncture using Microtainer tubes containing EDTA and sodium fluoride ( Becton Dickinson , NJ , USA ) . In infants , blood was obtained by puncture of the heel or toe . All samples were frozen at -8°C until further processing . In the case of symptoms related to malaria , a thick blood smear was prepared following the World Health Organization guidelines [30] . All participants that tested positive for P . falciparum and/or P . vivax by thick blood smear and/or qPCR were considered as infected by malaria parasites , and received treatment according to the guidelines of the Brazilian Ministry of Health [31] . Pelleted RBCs obtained from participants’ blood were resuspended in PBS and genomic DNA was extracted using a FavorPrep 96-well Genomic DNA Kit ( Favorgen , Ping-Tung , Taiwan ) according to the manufacturer’s instructions . DNA was eluted in elution buffer and stored at -20°C or vacuum concentrated ( Concentrator 5301 , Eppendorf , Hamburg , Germany ) [32] . All DNA samples were subjected to QMAL Taqman qPCR in order to detect any Plasmodium spp . by targeting a conserved region of the 18S rRNA gene [30] . QMAL-positive samples were further analysed by Taqman qPCR assays to detect species-specific sequences of 18S rRNA gene of P . falciparum and P . vivax , as previously described [32 , 33] . PCR assays were carried out on the 7500 Fast Real-Time PCR System ( Applied Biosystems , Foster City , USA ) . In each household , mosquitos were collected using BG-Sentinel traps with BG-Lure attractant and electric entomological aspirators ( Horst Armadilhas Ltda . , Brazil ) . One BG-Sentinel per house was left generally in a bedroom for around 24 hours . Aspiration inside the dwellings lasted approximately 5 to 15 minutes , depending on the number and size of the rooms . In the collection period , average temperature was 27 . 4°C ( 23 . 4°C- 30 . 8°C ) . The captured culicids were transferred by suitable plastic cages to the FMT-HVD Entomology Laboratory on the same day . The mosquitoes were placed on petri dishes arranged on ice , immediately identified in a dormant state and separated by sex and blood-feeding status ( engorged or not ) . Taxonomical identification was made using the keys of Consoli and Lourenço-de-Oliveira [34] , and Faran and Linthicum [35] . After the identification , engorged female mosquitos were stored individually in ethanol 80% v/v at -20°C until DNA extraction . The number of mosquitoes per pool and the maximum post-feeding time in which Plasmodium DNA detection is still possible were determined by using experimentally fed culicids . A colony of Cx . quinquefasciatus was established from immature forms ( larvae and pupae ) of the mosquitoes collected in natural breeding environments on the outskirts of the city . Cx . quinquefasciatus , the most abundant mosquito species found on the outskirts of Manaus [36] , was chosen for this purpose . These were maintained under standard insectary conditions of 27°C , 80% relative humidity , and a 12:12 light/dark cycle until we obtained adult F1 generation , according to Gerberg et al . methodology [37] . 100 to 120 5-day-old female mosquitoes were given a blood meal with P . vivax obtained from a volunteer’s blood , using an artificial membrane feeding system [38] . After feeding , 50 fully engorged females were immediately sacrificed by freezing at -20°C . The remaining fully engorged females were returned to insectary conditions , and subsequent samples of 10 mosquitoes were sacrificed daily from day 1 ( D1 ) to 10 ( D10 ) after a P . vivax blood meal . In order to estimate the maximum number of mosquitos per pool for Plasmodium sp . detection , different proportions of infected blood fed per non-infected blood fed mosquitos were assessed: 1:1; 1:3; 1:5; 1:7; and 1:9 . For this assay , only abdomens were used for DNA extraction and QMAL q-PCR . DNA extraction was performed according to Musapa et al . [39] by using a 5% w/v Chelex 100 Resin ( styrene divinylbenzene copolymer containing paired iminodiacetate ions; Bio-Rad , Hercules , CA , USA ) . All DNA samples were subjected to QMAL Taqman qPCR to detect Plasmodium spp . by targeting a conserved region of the 18S rRNA gene [32 , 33] . The design of the standardized forms , and their scanning , processing and exporting to Excel sheets was performed using Cardiff Teleform v . 10 . 4 . 1 ( Cardiff Software ) . Statistical analysis was performed using Stata 13 . 1 and QGis v . 2 . 18 . 7 . Since houses were used as unit of analysis , information on building conditions , spatial coordinates , demography and infection status in both human and mosquitos were clustered by household . Relative frequency of Plasmodium DNA presence in culicids was determined by the number of positive pools with QMAL q-PCR genus-specific amplification per total number of pools from the same household . Similarly , the human infection rate was calculated as the proportion of QMAL q-PCR genus-specific positive per total of samples assessed for each household . Although the number of copies could be obtained from both tests , the data was dichotomized as positive or negative for analysis purposes . Negative binomial regression was used to assess association between presence of any positive blood-fed Culex in a household and two distinct variables: ( i ) the total number of new infections per time at risk ( incidence ) ; and ( ii ) number of people infected at the time of mosquito collection in the household ( i . e . , household prevalence ) . Additionally , the same models were carried out using the total number of blood-fed Culex mosquitoes as independent variable . Differences were considered statistically significant for p<0 . 05 . Colonised Cx . quinquefasciatus blood feeding on Mus musculus was approved by the FMT-HVD Ethics Committee in Animal Research ( 349 . 211/2013 ) . The human survey was approved by the Brazilian National Committee of Ethics ( 349 . 211/2013 ) . An informed consent form was signed by all study participants or by a parent or legal guardian . Children between 12 and 17 years signed an additional assent form . Malaria patients were treated according the Brazilian Ministry of Health guidelines [31] . A total of 875 individuals were included in the study , and nearly half of them ( 45 . 6% ) were 0 to 20 years old and 50 . 7% were women . A total of 381 ( 43 . 5% ) individuals reported to use long-lasting insecticidal nets in the preceding night and 556 ( 63 . 5% ) reported that their houses had been sprayed with permethrin in the preceding six months . A total of 244 ( 27 . 9% ) individuals reported no previous malaria episode , 252 ( 28 . 8% ) reported 1 to 3 previous episodes and 379 ( 43 . 3% ) reported having experienced more than 3 episodes in their lifetime . Malaria prevalence in the study area was 7 . 7% , ranging from 3 . 2% in the Brasileirinho community to 11 . 2% in the Puraquequara community . In the Ipiranga community , prevalence was 8 . 1% . Only 3 ( 4 . 1% ) of the total 61 detected infections were symptomatic . From the total of households included , 44 . 2% were located in the Ipiranga community , 34 . 76% in the Brasileirinho community and 21 . 03% in the Puraquequara community . The mean number of residents per household was 3 , but ranged anywhere from 1 to 15 . Regarding the household structure , bricks were mostly used for building ( 133; 57 . 1% ) , houses with complete walls ( 221; 94 . 9% ) . 58 ( 24 . 9% ) of the houses presented cracks in the walls . 52 ( 22 . 3% ) houses possessed long-lasting insecticidal nets , 107 ( 45 . 9% ) had been sprayed in the last 6 months and 138 ( 59 . 2% ) had at least one resident who had used long-lasting insecticide-treated nets during the previous night . Fever was documented in at least one resident in 14 ( 6% ) of the households . Thirty ( 12 . 9% ) households had at least one resident who had been using antimalarials in the last 30 days ( Table 1 ) . In the search for the definition of which field collection mosquitoes should be used for the genus-specific qPCR assays , experimental infections were performed using Cx . quinquefasciatus from a pre-established colony . Plasmodium DNA was detectable at D2 post-feeding , when the Plasmodium DNA carriage was reduced to 12 . 5% . For this , we chose to test only visibly blood-fed mosquitoes . Detection using qPCR was possible up to a pool size of 5 mosquitos ( Fig 1 ) . A total of 13 , 374 specimens were captured: 12 , 132 ( 90 . 7% ) by using BG-Sentinel traps and 1 , 242 ( 9 . 3% ) b electric aspirators . Most of the mosquitoes ( n = 12 , 923; 96 . 6% ) were Culex spp . with Cx . quinquefasciatus being the most common species ( n = 12 , 599; 94 . 2% ) . 293 of the culicids ( 2 . 2% ) were not identified . The mean number of culicids per household was 50 , and ranged from 0 to 887 specimens; in 76 . 6% of the households 0–50 culicids were collected , and 6 . 1% presented more than 150 culicids . BG-Sentinel traps ( 12 , 132; 90 . 7% ) were more effective in capturing mosquitoes and culicid diversity was greater ( at least 10 taxa ) when compared to electric aspiration ( at least 6 taxa ) . Cx . quinquefasciatus predominated in the BG-Sentinel traps ( 95 . 3% ) and the electric aspiration ( 83 . 6% ) collections . A proportion of 17 . 7% of the culicids were engorged , ranged from 14 . 5% in BG-Sentinel trapped specimens to 48 . 1% in those collected by electric aspirations ( Chi-square 875 . 09 , p<0 . 0001 ) . For Cx . quinquefasciatus , 16 . 2% of the culicids were engorged , with 13 . 8% of BG-Sentinel trapped specimens and 42 . 0% of culicids from aspirations ( Chi-square 557 . 08 , p<0 . 001 ) ( Table 2 ) . The mean number of engorged culicids per household was 7 . 7 , ranging from 0 to 176 specimen . A total of 198 households ( 75 . 9% ) possessed at least one engorged mosquito in the collected specimens . The majority of the households possessed 1–5 engorged mosquitoes ( 121; 46 . 4% ) . Two household parameters affected the total number of mosquitoes captured indoors ( Table 3 ) : wooden houses ( IRR 1 . 44 , CI95%1 . 04–2 . 01; p = 0 . 029 ) and houses with cracks in the walls ( IRR 1 . 38 , CI95% 0 . 98–1 . 95; p = 0 . 067 ) . Houses located in the Ipiranga community generally yielded less mosquitoes in comparison to houses located in the Puraquequara community ( IRR 0 . 64 , CI95% 0 . 44–0 . 93; p = 0 . 020 ) . The differences were insignificant when only engorged culicids were analysed . However , the number of engorged mosquitoes was moderately associated with the number of total mosquitoes per house ( p<0 . 0001 , R2 = 0 . 37 ) . Overall Plasmodium DNA positivity in engorged culicids was 2 . 9% , specifically , 3 . 2% in mosquitoes collected by electric aspiration and 2 . 7% collected by BG-Sentinel traps ( Chi-square 0 . 1703 , p = 0 . 679 ) . Cx . quinquefasciatus showed a positivity rate of 3 . 4% , with no difference between collection methods ( 3 . 6% and 3 . 3% , respectively; Chi-square 0 . 0292 , p = 0 . 864 ) . Other Culex genus representatives ( Culex sp . ) showed a positivity rate of 3 . 7% , with no differences between collection methods ( 5 . 9% and 2 . 1% , respectively; Chi-square 0 . 8149 , p = 0 . 367 ) . Anopheles darlingi was found with Plasmodium DNA only in BG-Sentinel traps ( 1 . 9% ) . Other culicids did not present Plasmodium DNA positive samples ( Table 4 ) . The frequency of households with at least one culicid specimen carrying Plasmodium DNA was 6 . 4% . Fig 2 indicates that culicids xenosurveillance was efficient in detecting malaria clusters at the study site . Out of the 14 households with at least one positive culicid , 8 ( 57% ) showed an incidence>0 in humans and 31% had a Plasmodium infection incidence >0 . Plasmodium infection incidence ( as expressed in total number of new infections per household/time at risk ) was significantly related to whether any Plasmodium positive blood-fed culicids were found in the same house when both collection methods were combined [IRR 3 . 49 ( CI95% 1 . 38–8 . 84 ) ; p = 0 . 008] and for indoor-collected culicids [IRR 4 . 07 ( CI95%1 . 25–13 . 24 ) ; p = 0 . 020] ( Table 5 ) . Furthermore , the number of people infected in the house at the time of mosquito collection was related to whether there were any positive blood-fed culicid mosquitoes in that house when both collection methods were combined [IRR 4 . 48 ( CI95%2 . 22–9 . 05 ) ; p<0 . 001] or only for indoor-collected culicids [IRR 4 . 88 ( CI95%2 . 01–11 . 82 ) ; p<0 . 001] . The total number of blood-fed culicids was also significantly associated with Plasmodium incidence ( as expressed in total number of new infections per household/time at risk [1 . 03 ( CI95% 1 . 01–1 . 05 ) ; p = 0 . 006] . A large majority of mosquito species are hematophagous and rely on vertebrate blood for nourishment and reproduction . The contact of several anthropophilic and endophagic culicids with human populations living in malaria endemic areas suggests that the identification of Plasmodium genetic material in the blood present in the gut of mosquitoes may be possible and correlated to infection prevalence in the human population . In view of this , the present study investigated the relationship between Plasmodium DNA presence in engorged culicids and the frequency of individuals carrying malaria parasites in the same environment and the suitability of this novel xenosurveillancetool to support malaria transmission surveillance . In this study , Cx . quinquefasciatus and other less frequent Culex species predominated in the mosquito collections . Some mosquito species , such as Cx . quinquefasciatus , have a spatial distribution and abundance , which is strongly dependent on human presence [36 , 40–44] . These mosquitoes are highly anthropophilic , feed frequently , and prefer to blood feed at night and often stay inside human dwellings [45–47] . Cx . quinquefasciatus , for example , are inclined to blood feed in the evening and evening twilight and they use human dwellings as shelter during the day and at night [40–42 , 48] . The feeding preference is for human blood and the behavior of Cx . quinquefasciatus , make this species very susceptible to pathogens present in the human host blood [6] . After the blood meal , the female mosquitoes have limited mobilityand rest on interior walls of houses for several hours , where they can be easily collected via aspiration . After capture , the blood contents present in the mosquito’s gut can be analyzed for the presence of different pathogens . Mosquitoes behave like “flying biological syringes” [36 , 48–52] . As these biological syringes are not competent vectors , replication is not expected [6 , 8 , 11–12 , 53 , 54] and the pathogen needs to be detected before complete digestion in the mosquito gut . During the study , we identified and separated mosquitoes by taxa to demonstrate the different possibilities for finding Plasmodium DNA . However , the identification is labour-intensive , though this may not be required when the proposed xenosurveillance method is used in operational settings . A variable that needs to be taken into account in this xenosurveillance method is the sensitivity of Plasmodium detection in field-collected culicids . Through laser confocal microscopy , the development of P . falciparum in Cx . quinquefasciatus up to 30 hours after experimental infection was observed [55] . However , the authors point out that very few of the parasites in Cx . quinquefasciatus were alive during this period of 30 hours . The results obtained through the follow-up of P . vivax experimental feeding in Cx . quinquefasciatus demonstrated the decay of the DNA detection rates . The detection was only possible up to 48 hours after feeding . Since the mechanism of parasite killing must be sufficiently powerful that Plasmodium is not able to overcome it , working only with visibly engorged mosquitoes is a feasible , less costly and less laborious option , and the limit of five mosquito abdomens per pool demonstrates the operational viability of the technique . The simple cut of the abdomen of the mosquito does not require more time-consuming procedures and provides resource saving of DNA extraction and qPCR consumables , fundamental characteristics for the feasibility of any surveillance tool . The presence of positive blood-fed culicids was significantly correlated to the force of infection and Plasmodium infection prevalence in humans . This correlation is important for combined collection methods and for indoor-collected culicids . In terms of logistics , using only indoor aspiration collection methods appeared to be more reliable . Although BG Sentinel traps are more productive , culicids collected by this method are mostly not engorged . BG Sentinel traps are large and difficult to carry to the field . The transportation requires vehicles with higher freight capacity and a second home visit for trap removal . Differently , the electric aspirators collect a greater proportion of engorged mosquitoes . This technique permits speed and practicality , allowing only one technician to aspire ~15 houses in one morning , in a single visit for culicid capture . Furthermore , the portability of the equipment with low requirement of logistics , allows the accomplishment of the visits even by motorcycle , which is deemed favourable in operational settings . However , in unfurnished and single-room homes , capture was usually poorly productive or even negative , which is probably due to the absence of mosquito resting sites . Considering that indoor aspiration may be considered by some as invasive , there was an expectation of rejection , which was not confirmed . Participants exhibited a good receptivity , even at times when some members of a family were still asleep . In the force of infection analysis , the total number of blood-fed culicids was significantly associated with malaria infections . One explanation might be that some houses in poorer conditions are more permissive to Culex mosquitoes in parallel with higher odds to inhabitants who also have malaria . Unlike other studies , the samples of the present study were obtained in a seasonal period in which there was a lower prevalence of malaria , and in a single round of collections . Thus , the information generated may be important for estimating the circulation of the pathogen in the area , even in periods of low transmission , corroborating the possibility of using this tool as a new malaria indicator . A good spatial correlation between Plasmodium DNA positive mosquitoes and infected humans was observed in the study area , where asymptomatic infection predominates . The existence of asymptomatic individuals who are carriers of Plasmodium sp . in significant proportions , even in a hypoendemic area , shows that the prevalence of infection , when based on the thick smear drops analyzed by microscopy , is underestimated . This indicates the importance of these reservoirs in the dynamics of malaria transmission , and suggests that the submicroscopic parasite may be important for the transmission between the high and low transmission seasonal periods [56 , 57] . Thus , alternatives to conventional surveillance and control measures should be implemented . In this context , the xenosurveillance method is a valuable complement to the official surveillance system . New malaria infection burden surveillance strategies should be simple to implement , technologically uncomplicated , cost-effective and applicable to areas where malaria cases are known to occur . In this study , we found a significant association between blood-fed and Plasmodium DNA positive culicids and malaria prevalence . Our results suggest that xenosurveillance can be used in endemic tropical regions in order to estimate malaria burden and identify transmission foci in areas of a P . vivax asymptomatic carriers . As a perspective , this tool can be applied for the simultaneous surveillance of human pathogens , such as vector-borne , transmission areas overlapping with malaria , while taking advantage of the same collection efforts .	In the era of malaria elimination , novel surveillance strategies are needed in order to achieve and sustain malaria free status in the region . New malaria infection burden surveillance strategies should be simple to implement , technologically uncomplicated , cost-effective and applicable to areas where malaria cases are known to occur . In this context , the xenosurveillance is an important tool , especially when using widely distributed and anthropophilic mosquitoes . Culex mosquitoes do not transmit malaria , but due to their wide dispersion and anthropophilic behavior , may be used for the monitoring of certain vector-borne diseases , such as malaria , whose agent can be easily identified through molecular techniques in the ingested blood . Here , we evaluated a Plasmodium xenosurveillance approach to determine human malaria prevalence in three distinct peri-urban areas of Manaus . For this purpose , we evaluated the presence of Plasmodium DNA in engorged culicids sampled indoors regardless of their role in malaria transmission and compared them with the human infection status in the same households . The spatial correlation between the presence of Plasmodium DNA in mosquitoes was similar to the infection rate in humans which clearly indicates that xenosurveillance can be a complimentary strategy in endemic tropical regions in order to estimate the malaria burden .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "parasite", "groups", "body", "fluids", "plasmodium", "tropical", "diseases", "vector-borne", "diseases", "parasitic", "diseases", "animals", "parasitic", "protozoans", "parasitology", "apicomplexa", "protozoans", ...	2018	Use of anthropophilic culicid-based xenosurveillance as a proxy for Plasmodium vivax malaria burden and transmission hotspots identification
Nipah virus is a zoonotic pathogen that causes severe disease in humans . The mechanisms of pathogenesis are not well described . The first Nipah virus outbreak occurred in Malaysia , where human disease had a strong neurological component . Subsequent outbreaks have occurred in Bangladesh and India and transmission and disease processes in these outbreaks appear to be different from those of the Malaysian outbreak . Until this point , virtually all Nipah virus studies in vitro and in vivo , including vaccine and pathogenesis studies , have utilized a virus isolate from the original Malaysian outbreak ( NiV-M ) . To investigate potential differences between NiV-M and a Nipah virus isolate from Bangladesh ( NiV-B ) , we compared NiV-M and NiV-B infection in vitro and in vivo . In hamster kidney cells , NiV-M-infection resulted in extensive syncytia formation and cytopathic effects , whereas NiV-B-infection resulted in little to no morphological changes . In vivo , NiV-M-infected Syrian hamsters had accelerated virus replication , pathology and death when compared to NiV-B-infected animals . NiV-M infection also resulted in the activation of host immune response genes at an earlier time point . Pathogenicity was not only a result of direct effects of virus replication , but likely also had an immunopathogenic component . The differences observed between NiV-M and NiV-B pathogeneis in hamsters may relate to differences observed in human cases . Characterization of the hamster model for NiV-B infection allows for further research of the strain of Nipah virus responsible for the more recent outbreaks in humans . This model can be used to study NiV-B pathogenesis , transmission , and countermeasures that could be used to control outbreaks . Nipah virus is a member of the family Paramyxoviridae , genus Henipavirus , and was discovered in 1998–99 to be the etiological agent responsible for an outbreak of severe respiratory disease in pigs [1] and encephalitis in humans in Malaysia [2] . All subsequent outbreaks of Nipah virus have occurred in Bangladesh or India , beginning in 2001 , and have occurred on an almost annual basis [3] . Genetic data demonstrate that the isolates from Malaysia ( NiV-M ) and Bangladesh ( NiV-B ) represent two distinct Nipah virus strains [3] , [4] . Nipah virus outbreaks have case fatality rates of up to 100% and there are no approved vaccines or treatments and these viruses have been categorized as a biosafety level 4 ( BSL4 ) agents . Nipah virus differs from other paramyxoviruses in its ability to infect a wide range of mammals including bats [5] , dogs [1] , [6] , horses [7] , pigs [1] , and cats [1] , [8] . Wildlife surveillance at the time of the first outbreaks , along with several subsequent studies , has identified fruit bats of the family Pteropodidae as the natural reservoir of Nipah virus [5] , [9]–[11] . During the first Nipah virus outbreak in Malaysia , NiV-M caused over 265 cases of encephalitis with 105 human deaths , resulting in a case fatality rate of 40% [1] . Common clinical manifestations of Nipah virus infection included fever , headache , respiratory disease , encephalitis and loss of consciousness [12] , [13] . Fatal human cases of NiV-M infection were characterized by pathology involving the respiratory tract , central nervous system ( CNS ) , heart , kidney and spleen [13] . NiV-M infection causes vasculitis characterized by destruction of the endothelium , syncytia formation , thrombosis and necrosis , with infiltration of inflammatory cells throughout affected organs . In the lungs of infected humans , pulmonary edema , alveolar hemorrhage and pneumonia were documented as well as occasional multinucleated giant cells found in alveolar space [13] . During this outbreak , the disease predominantly affected the nervous system with prominent signs of brain stem dysfunction . Magnetic resonance imaging of the brains of infected individuals showed focal lesions throughout the white matter [13] . In a study examining 94 Nipah virus-infected patients in Malaysia , only 6% showed abnormal chest radiographs , and of these , only one presented with a cough [2] . Also , cases of late onset or relapsing encephalitis were documented during the Malaysia outbreak [2] . During the Malaysian outbreak , pigs predominantly showed signs of respiratory disease and were determined to be an intermediate host [14] , [15] . Epidemiologically , reports of infection with NiV-B differ from that of NiV-M infection in several aspects . Clinically , NiV-B infection resulted in a higher percentage of respiratory disease and a higher case fatality rate , reaching up to 100% , compared to NiV-M infection [12] . This disparity could reflect the differences in availability of health care and in reporting [16] . Disparities , however , could also be caused by intrinsic differences in the pathogenicity of NiV-M and NiV-B . NiV-B is transmitted from bats to humans by multiple routes including the ingestion of contaminated date palm sap [17] , and can subsequently be transmitted nosocomially [18] , or by human-to-human transmission [19]–[22] . Common clinical signs and symptoms of NiV-B infection included fever , altered mental status , headaches , cough , and difficulty breathing [16] , [23] . During the Bangladeshi outbreaks , acute respiratory distress was noted in many patients [16] , [24] . Febrile neurologic illnesses were also documented in some outbreaks of NiV-B , with lesions found in the gray and white matter of the brain [23] , [25] , [26] . In one study looking at 92 patients , 69% had difficulty breathing and 62% had a cough [16] . Limited studies have been conducted to describe the pathology in NiV-B infected patients . In contrast to most other paramyxoviruses , Nipah virus has a broad species tropism and there are few suitable animal models that recapitulate human disease . Experimentally cats , guinea pigs , ferrets , pigs , non-human primates , and Syrian hamsters have been shown to support NiV-M viral replication resulting in clinical signs of infection [27]–[29] . The Syrian hamster is the only small rodent model that closely mimics multiple aspects of human disease [8] , [27] , [30] , [31] . When infected intraperitoneally ( i . p . ) or intranasally ( i . n . ) with NiV-M , hamsters develop respiratory disease and/or encephalitis . The pathological changes that occur in the hamster are similar to those described in humans . The presence of vasculitis , necrosis , and inflammation is seen in both the human and hamsters . Viral antigen and disease pathology is observed in lung , kidney and heart tissue [31] , [32] . Similar to human infections that lead to encephalitis , hamsters show antigen positive neurons , necrosis , and vasculitis in the CNS [31] . These similarities in infection between humans and hamsters make the Syrian hamster a suitable model for the study of Nipah virus pathogenesis . This study was designed to compare NiV-B and NiV-M infections in a hamster-derived cell line , followed by a comparison of the pathogenesis and immune responses to infection by both virus strains in the Syrian hamster . Our results demonstrated that hamster cells are permissive for infection by both virus strains , with NiV-M causing increased syncytia formation and cytopathic effect ( CPE ) compared to NiV-B . In vivo , NiV-B infection resulted in a delayed disease progression compared to NiV-M infection . Overall NiV-M is more cytopathic in vitro and causes an accelerated disease in vivo , compared to NiV-B . All work with Nipah virus , potentially infectious materials , and infected hamsters was completed in the BSL4 facility at the Rocky Mountain Laboratories , Division of Intramural Research , National Institute of Allergy and Infectious Diseases , National Institutes of Health . All standard operating procedures applied were approved by the Institutional Biosafety Committee ( IBC ) . All animal experiments were approved by the Institutional Animal Care and Use Committee of the Rocky Mountain Laboratories and performed following the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care , International ( AAALAC ) by certified staff in an AAALAC-approved facility . NiV-B and NiV-M were provided by the Special Pathogens Branch of the Center for Disease Control and Prevention , Atlanta , GA , USA . NiV-M was isolated from a human case ( cerebrum ) in 1999 and passaged on Vero E6 cells a total of four times before used in experiments [33] . NiV-B was isolated from a throat swab of a lethal human infection from Bangladesh in 2004 and passaged in Vero E6 cells a total of three times [4] . Viruses were propagated on Vero E6 cells in Dulbecco's minimal essential medium ( DMEM ) ( Sigma ) supplemented with 10% fetal calf serum , 2 mM l-glutamine , 50 IU/mL penicillin and 50 µg/mL streptomycin ( Life Technologies ) . Supernatants were collected and clarified by low-speed centrifugation and stored in liquid nitrogen . For plaque assays , Vero C1008 ( European Collection of Cell Cultures ) were grown to confluency in 6-well plates . Media was replaced with 250 µL of serial 10-fold dilutions of virus in DMEM and incubated for 1 hr at 37°C , rocking every 15 min . The virus inoculum was replaced with 2 mL of a 1∶1 mixture of 2× minimal essential medium ( MEM ) and 1 . 6% low-melt agarose ( Life Technologies ) . The cells were then incubated for 3 d at 37°C , 5% CO2 before staining with 2 mL of a 0 . 25% crystal violet solution in 10% formalin for 3 hr at room temperature . The stain and overlay were then removed from the wells and the plaques were enumerated . To determine the 50% tissue culture infectious dose ( TCID50 ) , monolayers of Vero C1008 cells were grown in 96-well plates and 100 µL of serial 10-fold diluted samples in MEM containing 2% FBS , were added to the wells . Cells were then incubated for 5 d at 37°C , 5% CO2 and then scored for CPE . Baby hamster kidney cells ( BHK-21 ) from the American Type Culture Collection were propagated in MEM ( Sigma ) supplemented with 10% fetal calf serum , 2 mM l-glutamine , 50 IU/mL penicillin and 50 µg/mL streptomycin ( Life Technologies ) . Nipah virus infections were performed in 48-well plates when cells reached 95–100% confluency . Infections were performed by replacing medium with 250 µL of diluted virus ( multiplicity of infection ( MOI ) of 0 . 1 and 0 . 01 ) in MEM , 2% FBS . After 1 hr , the inoculum was replaced with MEM supplemented with 2% FBS . Supernatants were collected at 1 hr , and 1 , 2 , and 3 days post infection ( dpi ) for virus titration . Cells were stained using the Kwik Diff Kit ( Thermo scientific ) to visualize syncytia according to the instructions of the manufacturer . Cells were monitored for CPE with a light microscope and images were captured using a Nikon DS-Fi1 camera . Groups of 5 to 6 week old female Syrian hamsters ( Harlan ) were inoculated with the indicated doses of NiV-M or NiV-B diluted in sterile DMEM and administered via the i . p . route in a total volume of 500 µL . Control animals received the equivalent volume of sterile DMEM by the same route . Two groups were inoculated i . n . with 105 TCID50 per hamster of either NiV-M or NiV-B diluted in sterile DMEM . Fifty microliters of virus preparation was delivered to each nare using a pipette . Hamsters were weighed and scored daily for clinical signs for two weeks . When signs of disease no longer existed , animals were monitored but no longer weighed . The health of animals was assessed and scored according to the following criteria: 0 = no signs of disease; 1 = ruffled fur; 2 = ruffled fur & weight loss <5%; 3 = ruffled fur , hunched posture & weight loss >5%; 4 = ruffled fur , hunched posture & weight loss >10%; 5 = ruffled fur , hunched posture , weight loss >15% , or encephalitic signs , or hemorrhagic signs , or paralytic signs or dyspnea; 6 = ruffled fur , hunched posture , weight loss >20% , or encephalitic signs , or hemorrhagic signs , or paralytic signs or dyspnea; 7 = death . Euthanasia occurred at a score of 5 and above . At the time of euthanasia , animals were bled ( EDTA- and heparin-treated vacutainer tubes ) via cardiac puncture . Necropsies were performed to collect lung , spleen , heart and brain tissue . Tissues were placed in lysis buffer RLT ( Qiagen ) for RNA extraction , or 10% formalin for histopathology and immunohistochemistry ( IHC ) analysis . Tissues ( 30 mg pieces ) were homogenized in RLT buffer and removed from the BSL4 using approved standard operating procedures . Total RNA was extracted using RNeasy kit ( Qiagen ) , according to the manufacturers' instructions . Whole blood was collected and inactivated in AVL buffer and removed from the BSL4 using approved standard operating procedures . Total RNA was extracted using QIAamp viral RNA kit ( Qiagen ) , according to the manufacturers' instructions . The RNA was quantified on a nanodrop 8000 spectrophotometer ( Thermo Scientific ) . Real-time quantitative RT-PCR ( qRT-PCR ) was performed on a rotor-gene 6000 instrument ( Corbett Life Science ) using QuantiFast probe reagents ( Qiagen ) targeting the NiV-M or NiV-B nucleocapsid protein gene . Primers and probes used were: NiV-B sense ( 5′-GTTCAGGCCAGAGAAGCTAAATTT-3′ ) , NiV-B antisense ( 5′-CCTCTTCGTCGACATCTTGATCA-3′ ) , NiV-M sense ( 5′- GTTCAGGCTAGAGAGGCAAAATTT-3′ ) , NiV-M antisense ( 5′- CCCCTTCATCGATATCTTGATCA-3″ ) , NiV-B probe ( 5′-6FAM-CTGCAGGAGGTGTGCTCATCGGAGG-TAMRA-3′ ) and NiV-M probe ( 5′-6FAM-CTGCAGGAGGTGTGCTCATTGGAGG-TAMRA-3″ ) . qRT-PCR components were used at the concentrations recommended by the manufacturer and 5 µL of RNA was added to each reaction and the following thermocycling parameters were used: 50°C for 10 min , 95°C for 5 min , and 40 cycles of 95°C for 5 s , 60°C for 10 s . Dilutions of RNA extracted from a known titer of each Nipah virus were run in triplicate to generate a standard curve from which sample TCID50 equivalents were extrapolated . Hamster immune gene expression was determined as previously described [34] . Briefly , RNA was extracted from tissues and qRT-PCR was performed as described above using gene-specific primers and probes under multiplex conditions . The fold-change in each gene was calculated by normalizing the change in CT ( ΔCT ) to the CT values for RPL18 ( as an internal reference gene ) for each sample and comparing this to the CT values of uninfected hamsters ( 2−ΔΔCT ) . Tissues were fixed in 10% neutral buffered formalin for 7 d with one volume change , then transferred out of the BSL4 using approved standard operating procedures . Tissues were then placed in cassettes and processed with a Sakura VIP-5 Tissue Tek , on a 12 hr automated schedule , using a graded series of ethanol , xylene , and ParaPlast Extra . Embedded tissues were sectioned at 5 µm and dried overnight at 42°C prior to staining with hematoxylin and eosin ( H&E ) . Specific Nipah virus IHC was performed using an anti-Nipah virus N protein rabbit primary antibody at a 1∶5000 dilution ( kindly provided by L . Wang , CSIRO Livestock Industries , Australian Animal Health Laboratory , Australia ) [35] . The tissues were then processed using the Discovery XT automated processor ( Ventana Medical Systems ) with a DAPMap ( Ventana Medical Systems ) kit . Statistical analyses were performed on the data form the TCID50 and qRT-PCR experiments using a 2-way ANOVA with a Bonferroni's post-test . To determine whether there were significant differences in the time to death between the viruses , we performed a log-rank test . The mean and SEM is represented and significance ( * = p<0 . 05 , = p<0 . 01 and * = p<0 . 001 ) is reported where appropriate . To determine the cellular responses and replication kinetics of the two Nipah virus strains in a hamster cell line , we infected BHK-21 cells with either NiV-M or NiV-B at MOIs of either 0 . 01 or 0 . 1 . As early as 1 dpi , syncytia formation was apparent in all NiV-M-infected cultures . By 3 dpi , and at both MOIs , NiV-M-infected cells showed extensive CPE and nearly complete destruction of the cell monolayer ( Figure 1A ) . NiV-B-infected cells showed little CPE at any of the time points sampled , regardless of the inoculation dose . At 3 dpi , NiV-B-infected cells began to form small syncytia . At both MOIs , NiV-M replicated sooner and reached higher virus titers in the supernatant at earlier time points compared to NiV-B ( Figure 1B and C ) . At the lower MOI , NiV-M reached a titer that was 4 logs higher at 3 dpi than NiV-B ( Figure 1B ) , whereas end titers were similar for both Nipah virus strains at the higher MOI , with a faster progression for NiV-M ( Figure 1C ) . To date , NiV-B infection has not been examined in an animal model . To assess the suitability of the hamster as a model for NiV-B infection , as well as to compare the two strains , hamsters were inoculated i . p . with 10-fold serial dilutions of Nipah virus from 105 to 1 TCID50 ( Figure 2 ) . Animals were evaluated for clinical signs of disease on a daily bases according to a scoring system outlined in the Materials and Methods section . Only one hamster showed abnormal clinical signs on the day prior to euthanasia , which consisted of ruffled fur . All other hamsters did not display abnormal clinical signs until the day euthanasia was necessary . Hamsters challenged with either virus strain showed clinical signs of respiratory distress and/or neurologic dysfunction leading to a score that required euthanasia . Signs of respiratory disease included labored abdominal breathing and hunched posture; neurological dysfunction included imbalance , partial paralysis and seizures . Similar to previous studies with NiV-M , respiratory distress was observed only in animals infected at the highest doses ( 104 and 105 TCID50/animal ) [32] . The majority of animals inoculated with the lower doses of Nipah virus ( 100 through 103 TCID50/animal ) displayed neurologic dysfunction prior to euthanasia . One animal infected with NiV-M at the highest dose ( 105 TCID50 ) and two animals infected with NiV-B ( one inoculated with 104 and one with 105 ) presented with both respiratory and neurologic dysfunction , while the rest of animals had either respiratory or neurological signs of distress that required euthanasia . NiV-M-infected animals showing severe respiratory signs of disease were euthanized between 5–7 dpi , whereas animals displaying neurological disorders were euthanized between days 5–11 . Disease progression in NiV-B-infected animals was generally slower , and animals displaying severe respiratory distress or neurological dysfunction were euthanized on 8–9 dpi or 8–14 dpi for NiV-M and NiV-B infection , respectively ( Table 1 ) . The slower disease progression in NiV-B-infected animals was reflected in the overall survival curves with 80% lethal disease outcome even at the highest dose of infection ( Figure 2 and Table 1 ) . The LD50 for NiV-M and NiV-B was 68 and 528 TCID50 , respectively . At both 103 and 105 TCID50 , there was a statistically significant difference in the time to death between the two virus strains , with death occurring approximately two days later for NiV-B infected animals at each dose ( Table 1 ) . To determine if the delay in survival is associated with the route of infection , we inoculated hamsters i . n . with 105 TCID50 of either NiV-M or NiV-B . The mean time to death was delayed by two days in hamsters inoculated with NiV-B compared to NiV-M ( Figure S1 ) . Both routes of inoculation showed a two-day delay in NiV-B compared to NiV-M , although the mean time to death was later in both virus groups with the i . n . compared to i . p . route . To compare the pathogenesis of NiV-M to NiV-B , groups of hamsters were inoculated with 105 TCID50 of either Nipah virus and tissues were collected on 1 , 3 and 5 dpi for both virus groups , and 7 dpi for NiV-B . Based on the time to death at this dose from our survival experiment , 7 dpi tissues were not collected for NiV-M-inoculated animals for this pathology experiment . Viral RNA was detected using Nipah virus N-specific qRT-PCR ( Figure 3 ) . In NiV-M-inoculated animals , replication was detected at an earlier time point than NiV-B replication . As early as 1 dpi , viral RNA was detected in lungs , brain and spleen tissue of some NiV-M-infected animals . NiV-B-infected animals had detectable levels of viral RNA at 1 dpi in lung tissue of some hamsters , and in the spleen by 3 dpi . Both strains showed an increase in viral RNA over time in the lungs , brain and spleen , with the highest overall titers in the lungs at the last time point sampled . We assessed viremia in hamsters inoculated with either virus by qRT-PCR . Levels of viral RNA were barely detectable and viral RNA was undetectable in some animals at each time point ( data not shown ) . To examine the kinetics of the host immune response to Nipah virus infection , and compare responses between NiV-M and NiV-B infections , the expression level of a subset of cytokine and chemokine mRNAs were examined by qRT-PCR in the lungs , brain and spleen ( Figure 4 ) . Throughout the infection , the largest overall response was seen in the lungs . At 1 dpi , a statistical difference in the upregulation of interleukin-4 ( IL-4 ) , interleukin-6 ( IL-6 ) , tumor necrosis factor ( TNF ) , and interferon-γ ( IFNγ ) was observed between NiV-M and NiV-B infections , with higher expression of these genes in response to NiV-M . A similar result was measured at 3 dpi for IFNγ in the lungs ( Figure 4 ) . At 3 dpi , IL-4 , IL-6 and TNF were upregulated similarly in response to both virus strains and remained upregulated throughout the course of infection . Upregulation of the gene for myxovirus resistance protein-2 ( Mx2 ) in the lungs was detected only at the last time point for both virus strains . IFNγ-induced protein 10 ( IP-10 ) mRNA increased starting at 1 dpi and remained upregulated in the lungs throughout the course of infection , peaking at 3 dpi in NiV-M-infected hamsters and 5 dpi in NiV-B-infected hamsters . IL-4 , IL-6 and TNF were also upregulated in brain ( Figure 4 ) . There was a significant increase in Mx2 transcription in the spleens of NiV-M-infected hamsters at 3 dpi compared to NiV-B . In the brain , IL-4 , IL-6 and TNF were slightly upregulated over control animals . IL-4 , IL-6 , TNF , and IFNγ mRNAs were downregulated in the spleen . To compare the pathology between the two strains , hamsters were inoculated with 105 TCID50 of NiV-B or NiV-M and tissues were examined histologically . Pathology was observed for both infections and was composed of a mild to moderate multifocal , subacute , bronchointerstitial pneumonia with vasculitis on 5 dpi for NiV-M and NiV-B infection ( Figure 5 ) . By 7 dpi in NiV-B-infected animals , the pneumonia progressed to marked , multifocal to coalescing , subacute bronchointerstitial pneumonia with vasculitis , necrosis , edema , and fibrin deposits . The pneumonia in both groups , on day 5 dpi for NiV-M infection and 7 dpi for NiV-B infection , was characterized by effacement of terminal bronchioles and adjacent alveoli by small to moderate numbers of macrophages , neutrophils , lymphocytes and plasma cells . Multifocal vasculitis was observed with disruption of the arterial tunica media by small numbers of neutrophils and lymphocytes . Syncytial endothelial cells were found in affected small to medium caliber vessels . Hamsters from the final time points had moderate to marked lesions in the lungs and demonstrated a loss of pulmonary architecture with replacement by cellular and karyorrhectic debris with small to moderate amounts of hemorrhage , fibrin deposits and edema . IHC revealed viral antigen in alveolar capillary endothelium , small and medium caliber arteriolar endothelium , and in mononuclear inflammatory cells starting at 3 dpi for NiV-M infection and 5 dpi for NiV-B infection ( Figure 5B ) . The presence of viral antigen was strongly associated with areas of inflammation . No pathological changes were observed in the CNS of hamsters infected at the high dose used in the pathology study . Nipah virus is a zoonotic pathogen that causes encephalitis and pulmonary disease with a high case fatality rate and is classified as a category C pathogen by the NIAID's pathogen priority list [12] . Two strains of Nipah virus , NiV-M and NiV-B , have been isolated from geographically and temporally separated outbreaks [4] . Human outbreaks caused by these strains differ in disease progression and epidemiologically [23] . The Syrian hamster has been established as a disease model for NiV-M infection [31] , [32] , but NiV-B infection studies have not been reported for any animal model . The goal of this study was to compare the replication , pathogenesis , and immune response to infection with NiV-M and NiV-B using in vitro and in vivo methods . BHK-21 cells infected with NiV-M showed more severe damage and supported higher virus replication compared to NiV-B-infected cells . Hamsters infected with NiV-B had a delay in disease progression and increased survival rates compared to NiV-M infected animals . In vitro , BHK-21 cells were permissive for infection by both NiV-M and NiV-B . NiV-M replicated to higher titers in the supernatant at an earlier time point , and infection resulted in widespread syncytia formation causing extensive CPE . Widespread CPE was not observed in NiV-B-infected BHK-21 cells , although a few syncytia were present at later time points . The differences observed in virus replication and syncytia formation could be attributed to either higher viral replication causing more syncytia , or more syncytia formation resulting to higher virus load . Differences in replication efficiency including protein production , viral assembly and budding could explain the higher virus production and large number of syncytia observed in NiV-M infected cultures . Conversely , differences in fusion kinetics could account for disparate amounts of syncytia that then lead to variation in virus replication . The high amount of fusion and syncytia formation in cells could result in the infection of cells that were not initially infected at the low MOIs used in our experiments . This could lead to higher overall levels of virus production . Previous studies using paramyxoviruses have demonstrated that viral spread can occur by cell-cell fusion; the surface proteins of Nipah virus are present at the cell junctions and have been shown to initiate fusion and spread of virus [36]–[38] . Slower fusion kinetics could lead to less and slower formation of syncytia observed in NiV-B infected cells . The affinity of Nipah virus glycoprotein to its receptors , ephrin B2 and B3 , as well as the ability of the glycoprotein to trigger the fusion protein could also affect fusion rates . Further experiments need to be completed to examine the relationship between replication and syncytia formation . In our experiments , we chose i . p . as the route of inoculation due to the more uniform disease progression and outcomes described in previous Nipah virus studies [31] . It is likely that i . p . inoculation would more readily allow for the detection of subtle differences between strains that may not be detectable in a less uniform infection route , such as i . n . When infected with NiV-M or NiV-B , hamsters developed clinical signs of disease similar to human infection [2] , [13] , [31] , [39] . The onset of disease and death in hamsters was rapid and occurred between 5–14 dpi , which corresponds to human cases , where symptoms start to develop between 7–10 dpi [15] , [40] , [41] . We observed earlier replication of NiV-M than NiV-B in all organ types sampled , although , once NiV-B RNA was detected , it reached similar values within two days . Earlier replication of NiV-M in tissues corresponded with earlier pathologic changes and accelerated disease and death compared to NiV-B infection . In humans , CNS pathology is documented , but in our comparative pathology experiment , we did not observe pathology in the CNS . This is likely attributed to the high dose of inoculum for the pathology experiment ( 105 TCID50 ) and route of inoculation ( i . p . ) . However , we did observe pathology in the lungs consisting of multifocal subacute bronchointerstitial pneumonia with vasculitis . The pneumonia was characterized by inflammation in the terminal bronchioles and alveoli spaces , necrosis , hemorrhage , fibrin deposits , edema and syncytia in endothelial cells . In human cases , fibrinoid necrosis , vasculitis , pulmonary edema , alveoli hemorrhaging , and syncytia were documented [13] , [15] , [40] . It is probable that hamsters inoculated with this high dose ( 105 TCID50 ) succumbed to infection due to inflammation , edema , and widespread vasculitis in the lungs that caused interstitial pneumonia . Even with low levels of viral antigen , pathology was severe enough to cause a fatal outcome . The typical dose that humans are infected with , as well as the route of infection is not known . In hamsters , both virus strains caused respiratory distress and/or neurological dysfunction in a dose-dependent manner . Based on previous data in hamsters , it is likely that dose and route of infection might play a role in Nipah virus outcome in humans [32] . Disease progression could be altered by the transmission route , which could include fomite [18] , [42] , oral ingestion [17] , [43] , and respiratory droplets [42] , [44] , [45] . In this study , inoculation of hamsters with NiV-B resulted in a delay in disease progression and the LD50 was approximately a log higher compared to NiV-M . However these data are contrary to what has been reported in humans , where NiV-B results in higher case fatality rates compared to NiV-M . Since we did not observe a difference in disease that would explain differences in the epidemiological data for the two Nipah virus strains , factors other than the intrinsic pathogenicity likely contribute to the disparities in the documented epidemiological data . The suboptimal health care , lack of supportive care and inconsistencies in reporting could account for higher documented case fatality rates and differences in disease manifestations during NiV-B outbreaks [16] . Cytokine and chemokine mRNAs were quantitated in the hamsters over the course of infection and several immune genes were upregulated in the lung , brain , and spleen , although there was a slight downregulation of some genes in the spleen . NiV-M induced an earlier and more robust immune response compared to NiV-B , which eventually reached similar levels to hamsters infected with NiV-M . Early TNF activation during NiV-M infection may contribute to recruitment of inflammatory cells , as observed in the lungs of infected hamsters by histopathology . The upregulation of IP-10 in the lungs coincided with lymphocyte recruitment , appearance of vascular damage , and necrosis in the lungs . IP-10 upregulation has been documented in other Nipah virus studies , specifically focusing on endothelial cells [46] , [47] . Teruya-Feldstein et al reported that high levels of IP-10 are found in necrotic tissue and in areas of vascular damage associated with Epstein-Barr virus-positive lymphoproliferative processes in mice [48] . They demonstrated a correlation between IP-10 regulation , tissue necrosis , and vascular damage during viral infection . Similarly , IP-10 is upregulated in the airways of patients with pulmonary diseases such as tuberculosis and plays a role in recruitment of activated T cells [49] . IL-6 gene expression was increased earlier in the lung in NiV-M compared to NiV-B infected hamsters . IL-6 activates T cells [50] and the recruitment of T cells likely contributed to the widespread vasculitis associated with Nipah virus infection and disease . Recruitment of lymphocytes could also be a way for Nipah virus to disseminate throughout the host , as it has recently been published that lymphocytes and monocytes can carry virus without becoming infected and release virus at distant sites from the original infection [51] , [52] . In the lungs , IL-4 was also upregulated , following similar kinetics than IL-6 . IL-4 promotes differentiation of B cells , and is upregulated is indicative of the activation of a Th2 response [53] . However , during disease , specific antibody production would not occur fast enough , since animals succumb to infection before significant antibody production can likely occur . Due to the use of the hamster as a model , we are limited in the amount of reagents available for a detailed examination of the immune response and future work is needed to get a more complete picture of the immune response to Nipah virus infection . In general NiV-M infection caused earlier induction of immune genes which probably corresponds to the earlier pathology observed . It is possible that the strong early immune response in Nipah virus-infected animals might contribute to disease via an immunopathogenic mechanism . In conclusion , there is a delay in NiV-B-induced disease progression compared to NiV-M , specifically in time to death , virus replication , pathology and immune responses . NiV-M is more cytopathic in vitro and more pathogenic in vivo . Viral antigen staining was low in tissues , although the pathologic changes were extensive and the inflammatory response was robust , suggesting disease progression may not only be a result of direct effects of the virus , but likely has an immunopathogenic component . The experimental data presented herein characterizes the hamster as a suitable small animal model for NiV-B infection , showing clinical signs , viral tropism , and pathologic changes similar to those observed in humans . These data are important to further the understanding of Nipah virus infection and pathogenesis . By applying the hamster model for NiV-B this allows for future studies in transmission , pathology and therapeutics , specifically focusing on the Nipah virus strain responsible for recent outbreaks .	Nipah virus causes severe disease in humans and outbreaks have occurred in two geographic regions , Malaysia and Bangladesh , and viruses have been isolated during outbreaks from both of these regions ( NiV-M and NiV-B , respectively ) . The original outbreak of Nipah virus occurred in Malaysia and caused severe encephalitis in humans . All subsequent outbreaks of Nipah virus have occurred in Bangladesh or India and disease has been characterized as having a strong respiratory component . Nipah virus is a public health concern that can cause up to 100% lethality in humans and there is no approved treatment or vaccine . Current research should focus on understanding disease progression and pathogenicity . We compared NiV-M and NiV-B infection and disease progression using the Syrian hamster model . We found that NiV-M is more destructive in cultured hamster cells and has faster onset of cytopathogenicity compared to NiV-B . This is also true in hamsters , where although both viruses are pathogenic and cause a similar disease , pathology caused by NiV-M infection is accelerated . These data show that there is a difference in disease progression between the two strains of Nipah virus and will allow for a more detailed understanding of the events leading to disease caused by these viruses .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "animal", "models", "of", "infection", "viral", "transmission", "and", "infection", "virology", "emerging", "viral", "diseases", "immunology", "biology", "microbiology", "viral", "replication" ]	2013	Comparison of the Pathogenicity of Nipah Virus Isolates from Bangladesh and Malaysia in the Syrian Hamster
The Global Program to Eliminate Lymphatic Filariasis aims to interrupt transmission of lymphatic filariasis and manage morbidity in people currently living with the disease . A component of morbidity management is improving health-related quality of life ( HRQoL ) in patients . Measurement of HRQoL in current management programs is varied because of the lack of a standard HRQoL tool for use in the lymphatic filariasis population . In this study , the psychometric properties of three health status measures were compared when used in a group of lymphatic filariasis patients and healthy controls . The World Health Organization Disability Assessment Schedule 2 . 0 ( WHODAS 2 . 0 ) , the Dermatology Life Quality Index ( DLQI ) , and the Lymphatic Filariasis Quality of Life Questionnaire ( LFSQQ ) were administered to 36 stage II and stage III lymphatic filariasis subjects and 36 age and sex matched controls in Kerala , India . All three tools yielded missing value rates lower than 10% , suggesting high feasibility . Highest internal consistency was seen in the LFSQQ ( α = 0 . 97 ) . Discriminant validity analysis demonstrated that HRQoL was significantly lower in the LF group than in controls for the WHODAS 2 . 0 , DLQI , and LFSQQ , but total HRQoL scores did not differ between stage II and stage III lymphedema subjects . The LFSQQ total score correlated most strongly with the WHODAS 2 . 0 ( r = 0 . 91 , p<0 . 001 ) and DLQI ( r = 0 . 81 , p<0 . 001 ) . The WHODAS 2 . 0 , DLQI , and LFSQQ demonstrate acceptable feasibility , internal consistency , discriminate validity , and construct validity . Based on our psychometric analyses , the LFSQQ performs the best and is recommended for use in the lymphatic filariasis population . Lymphatic filariasis ( LF ) is the second leading cause of disability worldwide , affecting almost 120 million people across 81 countries [1] . Although LF is not curable , the Global Program to Eliminate Lymphatic Filariasis has reduced disease transmission and is working to expand morbidity management programs globally [2] , [3] . Morbidity management is expected to greatly improve health-related quality of life ( HRQoL ) in LF patients because LF is characterized by disfiguring lymphedema and debilitating inflammatory episodes that cause significant disability and social isolation [4]–[6] . In fact , HRQoL and health status assessments have become primary measures of intervention efficacy in LF studies [7]–[11] . Despite the reliance on HRQoL in LF management , no consensus has been made on the most suitable tool for use in the LF population . Generic instruments , including a seven domains and five levels ( 7D5L ) tool , the International Classification of Functioning , Disability and Health checklist , and the World Health Organization Disability Assessment Schedule 2 . 0 ( WHODAS 2 . 0 ) , have been used in the LF population because of their applicability to a number of diseases and the ease of HRQoL comparison among diseases [8] , [12]–[14] . However , most generic instruments are only able to measure 50% of LF-associated problems , often overlooking feelings of fear and embarrassment [15] . Emphasis has now shifted to disease-specific tools that are better able to measure disease-specific factors as well as disease progression [15] , [16] . Despite the call for development of LF-specific tools , only two such tools exist , the LF QoL Tool ( LF-QoL ) and the LF-specific QoL Questionnaire ( LFSQQ ) [7] , [10] , [17] , [18] . Of the two tools , the LFSQQ has been used to measure intervention efficacy and has been tested in a larger patient population . Because generic measures miss relevant HRQoL content and disease-specific tools have only recently been developed , LF HRQoL has been most often measured by the Dermatology Life Quality Index , a skin-specific HRQoL tool [11] , [19]–[21] . However , the DLQI has been found to assess only 24% of disability caused by LF [15] . Given the lack of a standard HRQoL tool for LF and minimal research comparing the various instruments , the aim of this study was to compare the relative performances of three health status measures in LF subjects and a matched control group . The tools chosen included a generic tool , the WHODAS 2 . 0 , a skin-specific tool , the DLQI , and a disease-specific tool , the LFSQQ . Specifically , we examined four instrument psychometric properties: feasibility , reliability , discriminative validity , and construct validity . Ethical clearance was obtained from the institutional review boards at Northwestern University in Evanston , Illinois USA , the Institute of Applied Dermatology in Kasaragod , Kerala India , and the Speciality Hospital in Kannur , Kerala , India . Within the LF population in Kerala , written consent was unable to be obtained because subjects did not want their name associated with the diagnosis of LF due to the disease's cultural stigma . Because of this limitation , verbal consent was obtained from all participants . As detailed in the IRB-approved protocol , the consent document was read aloud to participants and signed by the reader upon participant's approval to signify the participant's consent . In this cross-sectional study , a total of 36 LF subjects were consecutively enrolled from the outpatient units of the Institute of Applied Dermatology in Kasaragod , Kerala , India and the Speciality Hospital in Kannur , Kerala , India . Subjects were enrolled if they had a clinical diagnosis of LF , had never been treated for LF , and were at least 18 years of age . LF staging was performed by physicians based on the International Society of Lymphology's consensus staging system ( Table 1 ) [22] . Thirty-six control subjects were age and sex matched to LF subjects and were recruited from Kasaragod , Kerala . Control inclusion criteria were: no history of LF diagnosis , no blood relation to cases included in the study , and at least 18 years of age . All subjects completed the WHODAS 2 . 0 , DLQI , and LFSQQ in the local language of Malayalam . Because the WHODAS 2 . 0 was not available in Malayalam , the original English version was translated according to standard protocols [23] . Two native Malayalam speakers independently forward-translated the WHODAS 2 . 0 from English to Malayalam . Both translators and the local team assessed the tool's clarity , cultural relevance , and language , and any differing opinions were discussed . Changes to the instrument were made as needed . This version was then back-translated to English and confirmed to be equivalent with the original English version to ensure the validity of the Malayalam version . The DLQI and LFSQQ were available in Malayalam and have been validated for use [7] , [24] . The sequence of instrument administration was randomized to avoid an ordering bias . Demographic information was obtained via a demographic questionnaire . Table 2 lists the domains of each tool and outlines tool domains that cover similar content . The WHODAS 2 . 0 is a generic health and disability assessment tool that describes effects of disease on six domains: cognition , mobility , self-care , getting along , life activities , and participation in society [25] . Disability perception is measured by responses on a 5-point scale from 1 ( no difficulty ) to 5 ( extreme difficulty or cannot do ) . Final scores were calculated using a WHO SPSS 36 version syntax for employed subjects and a WHO SPSS 32 version syntax for unemployed subjects . The WHO SPSS 32 version syntax is identical to the WHO SPSS 36 version syntax with the omission of four questions relating to work ability . Domain and total scores range from 0 to 100 with a higher score indicating greater impairment of health status . The DLQI is a 10-item questionnaire that assesses skin-specific HRQoL through six domains: individual's symptoms and feelings , daily activities , leisure , work and school , personal relationships , and treatment [26] . Questions are scored on a 4-point scale , resulting in a maximum score of 30 , or large negative effect on HRQoL , and a minimum score of 0 , or no effect on HRQoL . The LFSQQ was developed to assess quality of life in LF subjects through seven domains: mobility , self-care , usual activities , disease burden , pain/discomfort , psychological health , and social participation . Each item is scored on a 5-point scale ( no problem , mild , moderate , severe , most severe ) . Total scores are calculated based on the number of questions answered and the raw scores [7] . Scores range from 0 to 100 with a higher score indicating a higher HRQoL . Median scores and interquartile ranges for the WHODAS 2 . 0 , DLQI , and LFSQQ were calculated for subject and control data . Discriminant ability , or the ability of a questionnaire to discriminate between respondent subgroups , was gauged by comparing HRQoL scores between subjects and controls using Wilcoxon Rank Sum tests . Further comparisons were made adjusting for demographic and disease-specific variables using linear regression models . A number of factors account for the overall feasibility of any tool , including cost , completion time , and ease of administration . In this study , feasibility was evaluated by examining the number of missing item responses . Internal consistency is a measure of the consistency of results yielded by items of a single construct . Cronbach's α-coefficient was calculated to determine the internal consistency of the WHODAS 2 . 0 and the LFSQQ . Coefficient values ≥0 . 70 were deemed reliable [27] . Internal consistency values were not calculated for the DLQI domains because domains consisted of only one or two items . A test demonstrates construct validity if it accurately measures the construct it intends to measure . Construct validity was assessed by comparing correlations between related constructs of the three tools using Spearman's rank correlation coefficients . Correlations were regarded as weak if the Spearman's coefficient was less than 0 . 50 , moderate if the coefficient was between 0 . 50 and 0 . 69 , and strong if the coefficient was greater than 0 . 69 . Missing values of tools are shown in Table 5 . Individual domains of the WHODAS 2 . 0 and DLQI showed low missing value rates ( WHODAS 2 . 0: ≤0 . 1% , DLQI: 0 . 0% ) . Missing values for the LFSQQ domains were <7% , except for the usual activities domain ( 32 . 9% ) . Less than 10% of data for total scores of all instruments were missing ( WHODAS 2 . 0 = 0 . 04% , DLQI = 0 . 0% , LFSQQ = 7 . 07% ) . Cronbach's alpha coefficients are presented for the three tools in Table 5 . Whole instrument reliability was highest for the LFSQQ ( α = 0 . 97 ) as compared to the WHODAS 2 . 0 ( α = 0 . 93 ) and DLQI ( α = 0 . 73 ) . Domains of the WHODAS 2 . 0 demonstrated higher internal consistency ( mean α = 0 . 85; range = 0 . 76–0 . 91 ) than the domains of the LFSQQ ( mean α = 0 . 80; range = 0 . 69–0 . 90 ) . All domains of the WHODAS 2 . 0 and all but one domain of the LFSQQ ( pain/discomfort ) showed internal consistency above 0 . 70 , the minimum value of acceptable internal consistency . Descriptive statistics for each tool in LF subjects and the control group are summarized in Table 6 . All three tools demonstrated lower HRQoL in LF subjects as compared to the control group . The LFSQQ domains discriminated best between the two groups as all but one of the domains ( self-care ) yielded p values <0 . 001 . For all three tools , domains relating to mobility , symptoms , and participation in society yielded notable differences between subject and control scores . Although the WHODAS 2 . 0 , DLQI , and LFSQQ discriminated well between LF subjects and the control group , no global tool score was able to discriminate between stage II and stage III lymphedema subjects . LF stage discrimination was only noted with the DLQI symptoms subscale ( p = 0 . 045 ) and the LFSQQ disease burden subscale ( p = 0 . 040 ) . Results of the correlation analysis between the WHODAS 2 . 0 and the DLQI are presented in Table 7 . Correlations between the WHODAS 2 . 0 and DLQI were generally low or moderate , including correlations between corresponding domains such as the WHODAS 2 . 0 life activities domain and the DLQI daily activities domain . The strongest correlation between the WHODAS 2 . 0 and DLQI was noted between the two total scores ( r = 0 . 748 , p<0 . 001 ) . The majority of correlations between WHODAS 2 . 0 and LFSQQ were moderate to strong ( Table 8 ) . Strong correlations were observed between the WHODAS 2 . 0 mobility and participation domains and LFSQQ domains . The total scores of both tools exhibited an especially high correlation ( r = −0 . 912 , p<0 . 001 ) . Mobility and participation subscales of both domains displayed strong correlation with their corresponding counterparts ( r = −0 . 917 , p<0 . 001 and r = −0 . 829 , p<0 . 001 , respectively ) . In addition to the social participation domain of the LFSQQ , the WHODAS 2 . 0 participation domain also strongly correlated with the psychological health dimension of the LFSQQ ( r = −0 . 855 , p<0 . 001 ) . Correlations between the LFSQQ and the DLQI are shown in Table 9 . Total LFSQQ score highly correlated with DLQI total score ( r = −0 . 808 , p<0 . 001 ) . Additionally , LFSQQ subscales related to disease burden , psychological health , and social participation strongly correlated with DLQI total score and two DLQI domains ( symptoms & feelings and leisure ) . Comparisons of the psychometric properties of health status measures are useful in determining the appropriateness of a certain tool in a specific population . In this study , we examined the performances of three HRQoL tools ( WHODAS 2 . 0 , DLQI , LFSQQ ) in the LF population and an age and gender matched control group . Although such comparisons have been performed in other chronic conditions [28]–[31] , to our knowledge , our study is the first of its kind in the LF community . As such , it should be regarded as a preliminary step in bettering our understanding of the specific benefits and disadvantages of these three tools in LF research . For all tools , feasibility was defined by the percentage of missing values per item . As the order of questionnaire administration was randomized , an ordering effect does not explain the LFSQQ's relatively low observed feasibility as compared to the other tools . Instead , the relatively high number of missing values may reflect the irrelevance of certain items to subjects . For example , the usual activities domain of the LFSQQ yielded the most missing values . This domain contained items assessing the effects of disease on activities such as agrarian work , gardening , and cleaning the floors . Because the tool does not allow for subjects to choose a “not applicable” option , subjects may have skipped items unrelated to their lifestyle rather than choose from the Likert scale ranging from “no problem” to “severe” . Questionnaire length does not seem to have an appreciable effect on the number of missing values as the 10-item DLQI performed similarly to the 36–item WHODAS 2 . 0 . Internal consistency was acceptable ( α>0 . 70 ) for all domains examined , except the LFSQQ pain/discomfort subscale ( α = 0 . 69 ) . Internal consistency values were highest among WHODAS 2 . 0 domains , and values were similar to those obtained by the WHO in a global population ( α range: 0 . 84–0 . 98 ) [25] . Discriminant ability analysis demonstrated that the LFSQQ scales distinguished best between LF subjects and the control group . The DLQI and WHODAS 2 . 0 also performed well , and among all tools , the largest differences in mean scores were noted in domains relating to mobility , symptoms , and participation . Although mobility and symptoms may be directly influenced by the physical manifestations of the disease , participation in society may be an indirect effect of the disease . This finding is in agreement with other studies that demonstrate that LF affects quality of life in a more nuanced manner than solely through the disease's physical signs . Kumari et al . found that in addition to the physical burdens of LF , subjects also coped with shame and depression [4] . The psychological effects of social stigma result in delayed treatment as patients are embarrassed to reveal their condition in society [32] . These components of health status are critical to include in HRQoL measures of intervention efficacy because although a patient's physical symptoms may regress with treatment , the psychological stress and disease stigma may persist and can influence HRQoL . Current disability and HRQoL tools do not encompass all psychosocial disabilities that affect the LF patient , and including domains outside the realm of social participation in HRQoL instruments may further highlight the difficulties faced by patients in this component of HRQoL [21] . Despite the ability of the DLQI , WHODAS 2 . 0 , and LFSQQ to discriminate between LF subjects and the control group , no global tool score was able to differentiate between stage II and stage III lymphedema . Of the domains , only the LFSQQ disease burden domain and the DLQI symptoms domain showed significant , albeit weak , association with LF stage . This weak association is most likely because the International Society of Lymphology's consensus staging system is based on the clinical assessment of disease signs rather than disability . Despite the importance of stage discrimination in the use of HRQoL as an outcome measure of intervention efficacy , the abilities of the DLQI , WHODAS 2 . 0 , and LFSQQ to discriminate between LF stages is unknown . Conflicting results exist concerning the ability of the DLQI to distinguish between LF stages . Yahathugoda et al . noted a correlation between DLQI score and lymphedema stage , and a modified version of the DLQI was also found to correlate with LF stage [20] , [21] . In addition , McPherson et al . reported decreased DLQI score following a clinical intervention , but baseline DLQI was only weakly associated with lymphedema grade [11] . Similar to our results , a study conducted in eastern India did not find a correlation between LF grade and DLQI total score [19] . The LFSQQ and WHODAS 2 . 0 have not been studied extensively in regard to their use in discriminating LF stage . Given our findings , the questionnaires in their current form are not sensitive to LF stage and modifications to each of the tools to better correlate with disease stage would improve each instrument's ability to accurately assess health status in LF subjects . Correlations between global scores of all instruments were high , suggesting that the tools cover similar measures of health status . Domains of the LFSQQ and WHODAS 2 . 0 were highly correlated . The mobility and participation domains of the WHODAS 2 . 0 were differentiated by the LFSQQ . As expected , domains of different instruments covering similar content often significantly correlated . However , frequently domains assessing different facets of HRQoL correlated strongly and at times , were more highly correlated than domains of similar names . For example , the life activities domain of the WHODAS 2 . 0 correlated more strongly to the symptoms & feelings domain of the DLQI than the DLQI's daily activities domain . This finding may be related to the depth and breadth of topics covered by each domain . The life activities domain of the WHODAS 2 . 0 encompasses the effect of disease on housework and work/school in depth . The DLQI covers a broader range of topics , including the effects of disease on shopping , looking after the home , gardening , and clothing choices . This difference in extent of domain coverage may have implications on the comprehensiveness of any tool's measure of HRQoL . As a cross-sectional study , there was no assessment of test-retest reliability of the instruments , a study limitation given the growing interest in LF treatment modalities and the increasing reliance on HRQoL tools that accurately measure change over time . In evaluating HRQoL , variations in tool applicability are expected based on differences in culture and lifestyle of targeted populations . In fact , Zeldenryk et al . noted 43 new HRQoL-related constructs in Bangladeshi focus groups that had not yet been described in the literature [33] . This study was conducted in a rural region of southern India , and thus , our results may not be generalizable to the global LF population . Finally , the study LF population was composed of only stage II and stage III lymphedema subjects . In the study region , stage I lymphedema subjects do not often present to clinic because of their minimal health changes and associated disabilities . Including stage I subjects in future studies of HRQoL tool comparison may shed light on correlations between LF stage and HRQoL . Although many tools have been used to assess HRQoL in the LF population , no consensus has been made on the gold-standard tool . Our aim in this study was to delineate the strengths and weaknesses of the WHODAS 2 . 0 , the DLQI , and the LFSQQ . All three tools showed moderately good performance as measured by feasibility , internal consistency , construct validity and discriminant validity . Specifically , the LFSQQ demonstrated the highest overall internal consistency , construct validity , and discriminant validity . Based on our results , the LFSQQ may be the best tool of the three to accurately assess HRQoL in the LF population . However , use of the instrument in its current state is limited by its high missing value rate . We recommend the addition of a “not applicable” option on the LFSQQ to increase tool feasibility . HRQoL largely depends on the cultural behaviors of the population studied . Before the LFSQQ can be used globally , the tool may require further modifications to take into account culturally-relevant lifestyle activities of LF-endemic areas . We recommend validation of the tool in LF populations outside of India to ensure proper application of the instrument . Although our results suggest that the LFSQQ should be used , further research examining additional psychometric properties , respondent burden , ease of comprehension , and differences in clinically-relevant and research-relevant tools is needed to better select a HRQoL for clinical or research use .	Lymphatic filariasis affects approximately 120 million people and is the second leading cause of life-long disability worldwide . Because lymphatic filariasis is one of the World Health Organization's six eradicable diseases , much effort has been placed into reducing transmission of the disease and managing morbidity . Novel interventions frequently use health-related quality of life as an outcome measure to monitor efficacy of the intervention . In an effort to delineate the strengths and weaknesses of health status measures and recommend use of a single tool in the lymphatic filariasis population , we compared the use of three health status tools ( The World Health Organization Disability Assessment Schedule 2 . 0 , the Dermatology Life Quality Index , and the Lymphatic Filariasis Quality of Life Questionnaire ) in lymphatic filariasis subjects and healthy controls in Kerala , India . The Lymphatic Filariasis Quality of Life Questionnaire performed the best by discriminating well between subjects and controls , possessing significant content overlap with the other two tools , yielding a low missing value rate , and being internally consistent . This is the first study to compare health status measures in lymphatic filariasis subjects and provides insight into the use of the tools in quality of life analysis .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "infectious", "diseases", "filariasis", "tropical", "diseases", "(non-neglected)", "neglected", "tropical", "diseases", "parasitic", "diseases", "lymphatic", "filariasis" ]	2014	Comparison of Three Quality of Life Instruments in Lymphatic Filariasis: DLQI, WHODAS 2.0, and LFSQQ
Francisella tularensis causes the disease tularemia . Human pulmonary exposure to the most virulent form , F . tularensis subsp . tularensis ( Ftt ) , leads to high morbidity and mortality , resulting in this bacterium being classified as a potential biothreat agent . However , a closely-related species , F . novicida , is avirulent in healthy humans . No tularemia vaccine is currently approved for human use . We demonstrate that a single dose vaccine of a live attenuated F . novicida strain ( Fn iglD ) protects against subsequent pulmonary challenge with Ftt using two different animal models , Fischer 344 rats and cynomolgus macaques ( NHP ) . The Fn iglD vaccine showed protective efficacy in rats , as did a Ftt iglD vaccine , suggesting no disadvantage to utilizing the low human virulent Francisella species to induce protective immunity . Comparison of specific antibody profiles in vaccinated rat and NHP sera by proteome array identified a core set of immunodominant antigens in vaccinated animals . This is the first report of a defined live attenuated vaccine that demonstrates efficacy against pulmonary tularemia in a NHP , and indicates that the low human virulence F . novicida functions as an effective tularemia vaccine platform . F . tularensis is a highly infectious bacterium that causes tularemia in humans , a disease that has a high mortality rate when acquired through the pulmonary route . F . tularensis is able to survive and replicate within host macrophages , and this ability is essential for its virulence . Within macrophages , F . tularensis escapes from the phagosomal compartment and replicates within the cytosol [1] . Phagosomal escape is mediated by a cluster of virulence genes in the Francisella Pathogenicity Island ( FPI ) that encode a Type VI-like secretion system [2] . F . tularensis acquired through the pulmonary route disseminate to tissues outside the lung , where they replicate to high levels within internal organs such as the liver . Early in infection , F . tularensis appears to induce broad immunosuppression within the host [3] , as proinflammatory cytokine expression is notably repressed [4] and infected cells are unable to respond to TLR-dependent secondary stimuli [5] . F . tularensis subsp . tularensis ( Ftt ) exhibits the highest level of virulence in all mammalian hosts , including humans , and because of the morbidity and mortality associated with disease as well as the potential for aerosol dissemination , it has been designated a category A biothreat agent . A closely-related species , F . novicida ( Fn ) , is considered essentially avirulent for healthy humans and for this reason is exempt from select agent status . There is currently no tularemia vaccine approved for human use . A live attenuated vaccine strain ( LVS ) was derived in Russia by repeated passage of F . tularensis subsp . holarctica ( Fth ) . LVS vaccination can protect against pulmonary challenge with Ftt in rats [6] , rhesus macaques and humans [7] . The LVS genome contains a large number of mutations that distinguish it from other Fth strains , but the primary attenuating mutations appear to be the deletion of a lipoprotein ( FTT0918 ) and a pilus subunit ( pilA ) [8] . Questions of stability , reversion frequency , and levels of protection may prevent LVS from becoming licensed for human use . However extensive studies with LVS have illuminated attributes of protective immunity against tularemia in mice . T-cell mediated immunity has been shown to be critical , but antibodies also appear to play a role; despite this , no specific correlate of protection has been established ( for review of this extensive field , please see [9] ) . The efficacy of LVS suggests that a safe rationally designed live attenuated vaccine would be effective against pulmonary tularemia . Mice have traditionally been the preferred model for tularemia vaccine development , due to ease of use , availability of reagents , and sensitivity to F . tularensis infections . However , tularemia vaccine development has been hampered by the extreme sensitivity of mice to Francisella subspecies , such that it has proven difficult to induce even partial protection against pulmonary Ftt exposure [10] . In fact , both LVS and Fn , which are known to be essentially avirulent in healthy humans , are virulent in mice . Recently , the Fischer 344 rat has been promoted as a better tularemia vaccine model , in that the rat shows similar sensitivities to the various F . tularensis strains as humans [6] , [11] . Non-human primate models of tularemia have also been reported , including rhesus macaques , marmosets , and African green monkeys [12]–[14] . Cynomolgus macaques are also sensitive to pulmonary infection by Ftt , which causes a fatal systemic disease similar to that seen in humans [15] . In the current study , we demonstrate the protective efficacy of a live highly attenuated Fn strain ( Fn iglD ) against pulmonary infection with Ftt , in both rats and cynomolgus macaques . Our results suggest that the live attenuated Fn iglD strain is a vaccine platform that is inherently safer , yet still effective for protecting humans against pulmonary tularemia . The FPI is required for intramacrophage growth and virulence of Francisella spp . [16] , and inactivation of most FPI genes renders Francisella spp . highly attenuated for intracellular growth and for virulence in animals . The FPI protein IglD is required for Francisella phagosomal escape and intramacrophage replication within macrophages [17]–[19] and Francisella iglD mutant strains are highly attenuated for virulence in mice [20] , [21] . We constructed Fn and Ftt iglD mutant strains and confirmed that both were defective for intramacrophage replication and virulence in mice , in contrast to the respective wildtype strains ( Fig . S1; Ftt has 2 identical copies of iglD , and is thus a iglD1 iglD2 mutant ) . In preliminary studies , the Fn iglD strain inoculated intranasally into mice was able to fully protect against subsequent pulmonary challenge with a relatively high dose of the wildtype F . novicida strain ( 103 CFU ) , but was unable to provide any protection against pulmonary challenge with a similar dose of Ftt ( 103 CFU; Table 1 ) . However , the Ftt iglD strain inoculated intranasally into mice was also unable to provide protection against pulmonary challenge with a similar dose of Ftt ( 103 CFU ) . This suggests that the failure of Fn iglD vaccination to protect against Ftt pulmonary challenge in mice is not due to some inherent deficiency in Fn , but rather may be due to the mouse being an inappropriate animal model for tularemia vaccine studies because of the extreme sensitivity of mice to all Francisella subspecies . The Fischer 344 rat has been promoted as a relevant animal model for tularemia vaccine studies , due to the relative sensitivities of the rat to the various Francisella subspecies , which mirror human sensitivities [6] , [11] . Moreover , LVS vaccination of Fischer 344 rats protects against pulmonary exposure to Ftt [6] , [22] . We have previously shown that oral vaccination of Fischer 344 rats with attenuated Fn strains induces comparable levels of protection against pulmonary Ftt challenge as intratracheal vaccination [23] , so the oral route was used in the following experiments . To determine the relative efficacies of Fn and Ftt live vaccine platforms in the rat , we vaccinated Fischer 344 rats ( n = 6 ) orally with either Fn iglD or Ftt iglD ( both at 107 CFU ) and then challenged the vaccinated rats 30 days later with Ftt ( 104 CFU ) delivered intratracheally ( Fig . 1A ) . 5 of 6 Fn iglD vaccinated rats ( 83% ) survived pulmonary Ftt challenge . 3 of 6 Ftt iglD-vaccinated rats ( 50% ) survived pulmonary Ftt challenge . These results demonstrate that there is no disadvantage to utilizing Fn instead of Ftt as the platform for live attenuated vaccines against pulmonary Ftt . Only one mock-vaccinated rat ( n = 6 ) survived pulmonary challenge with Ftt . Measurement of serum antibody levels in vaccinated rats revealed similar levels of Fn- or Ftt-specific antibodies in both groups , constituted by high levels of IgG2a and low levels of IgG1 ( Fig . 1B ) ; this polarized Th1-type response has been reported previously in immunized rats [11] . Since oral vaccination of rats with the Fn iglD live vaccine strain was shown to induce protective immunity against pulmonary Ftt challenge , we determined whether pulmonary vaccination of rats with Fn iglD also induced protective immunity against pulmonary Ftt challenge . Fischer 344 rats were vaccinated intratracheally with Fn iglD at 105 ( n = 4 ) or 107 ( n = 6 ) CFU , and challenged 30 days later with Ftt ( 104 CFU ) delivered intratracheally ( Fig . 2A ) . All rats vaccinated with Fn iglD at 105 CFU ( 100% protection ) and 5/6 ( 83% ) of rats vaccinated with Fn iglD at 107 CFU survived pulmonary challenge with Ftt , whereas only one of four mock-vaccinated rats survived this challenge , demonstrating the efficacy of pulmonary vaccination with Fn iglD to protect against pulmonary challenge with Ftt . Measurement of serum antibody levels in Fn iglD-vaccinated rats ( via pulmonary route ) again revealed a polarized response similar to oral vaccination , with high levels of Fn-specific IgG2a and low levels of IgG1 ( Fig . 2B ) . [11] It is known that a major target of the humoral response to Francisella infection is the O antigen ( OAg ) of the LPS , and that Fn and Ftt express distinct OAgs [24] . LVS expresses an OAg that is indistinguishable from the OAg of Ftt [25] . To determine if a humoral response to OAg was induced in vaccinated rats , we performed Western immunoblot analyses of serum from one of the rats immunized intratracheally with Fn iglD ( 107 CFU ) and compared that to the reactivity of serum from a rat immunized in a previous study by the same route with the same inoculum of LVS [11] , ( Fig . S2 ) . Serum from the rat vaccinated with Fn iglD reacted strongly with purified LPS from Fn , but also with LPS from LVS . In contrast , the serum from an LVS-vaccinated rat reacted strongly with LVS LPS and did not react at all with Fn LPS . Interestingly , the reactivity of both rat sera was predominantly to high molecular weight material , likely the OAg capsule [26] . These results confirm that the OAg of Fn and LVS/Ftt are distinct and demonstrate that in rats , just as in mice [24] , humoral responses directed towards LVS/Ftt OAg do not crossreact with Fn OAg . However there was some recognition of the LVS/Ftt OAg in Fn iglD-vaccinated rats . In order to measure the cellular responses of rats vaccinated with Fn iglD , we vaccinated 3 rats by the intratracheal route at 107 CFU , collected spleens at 28 days post vaccination , and measured IFNγ production upon stimulation with increasing dose of UV-inactivated Fn iglD , ( 105 and 106 CFU ) or the ( irrelevant ) antigen HEL ( Fig . S3 ) . The Fn iglD-vaccinated rats exhibited a dose-dependent increase ( p<0 . 05 ) in cellular responses to Fn iglD , whereas mock vaccinated rats showed no cellular responses to Fn iglD . The cynomolgus macaque is sensitive to pulmonary infection with Ftt [15] , with an LD50 of approximately 1 CFU via the aerosol route . Clinical symptoms of infection in this model include high respiration rates and serum C-reactive protein ( CRP ) levels , with corresponding high bacterial burdens in the lungs and tracheobronchial lymph nodes ( see below ) . The cynomolgus macaque has been proposed as a relevant non-human primate ( NHP ) model for tularemia vaccine development . We determined whether pulmonary vaccination of cynomolgus macaques with Fn iglD induced protective immunity against pulmonary Ftt challenge . 6 cynomolgus macaques were vaccinated via bronchoscopy with Fn iglD at 108 CFU , and an additional 4 control animals were mock vaccinated with PBS . 4 additional NHPs received LVS vaccination . Because LVS vaccination is known to induce protective immunity against Ftt in humans when administered through the skin , these animals were vaccinated by the subcutaneous route to serve as a vaccination standard against which the Fn iglD vaccine could be compared . The Fn iglD strain was well-tolerated in vaccinated NHPs , similar to the LVS vaccine , based on the lack of increase in respiration rate , and low serum CRP levels . Vaccinated and control NHP were challenged 30 days later with Ftt delivered in a head-only aerosol chamber with presented doses of 2500–5000 CFU ( Fig . 3A; presented doses for each NHP given in Table S1 ) . Challenged NHP were monitored for a number of different parameters , including respiration rate , serum CRP levels , and disease symptoms . Mock vaccinated animals eventually exhibited severe disease symptoms that necessitated euthanasia of all 4 animals when moribund , at days 7 ( 2 X ) , 8 , and 13 post Ftt challenge . In contrast , only one Fn iglD vaccinated animal required euthanasia when it became moribund at day 9 post challenge , and all other Fn iglD vaccinated NHPs survived to the end of the study at 30 days post challenge ( 83% protection ) . This demonstrates the efficacy of pulmonary vaccination with Fn iglD to protect against pulmonary challenge with Ftt in a NHP model of tularemia . All 4 LVS vaccinated NHPs also survived to the end of the study at 30 days post challenge ( 100% protection ) . Mock vaccinated NHPs exhibited significantly increased respiration rates and serum CRP levels compared to the Fn iglD- and LVS-vaccinated NHPs beginning 3 days post-challenge with Ftt ( Fig . 3B and 3C ) . Mock vaccinated NHPs also exhibited a trend in increased serum alanine transaminase ( ALT ) , blood urea nitrogen ( BUN ) , lactate dehydrogenase ( LDH ) , and aspartate aminotransferase ( AST ) levels , indicating liver , kidney , and other tissue damage , although these levels did not reach statistical significance over those of the vaccinated NHPs , perhaps due to small sample size ( Fig . S4 ) . 2 of 4 mock-vaccinated animals had detectable bacteremia at days 5 and 6 post challenge , whereas none of the vaccinated animals had detectable bacteremia at any time post Ftt challenge . Bacterial organ burdens determined at autopsy revealed higher bacterial burdens in the spleen , lung , mesenteric lymph nodes , liver , and tracheo-bronchial lymph nodes of mock-vaccinated NHPs than in those of Fn iglD- and LVS-vaccinated NHP ( Fig . 3D ) . In fact , 2/4 LVS-vaccinated and 3/6 Fn iglD vaccinated NHPs had no detectable bacterial burdens in any of the tissues sampled ( limit of detection ∼70 CFU/g; Fig . S5 ) . When organ burdens of the individual Fn iglD-vaccinated NHPs were compared at autopsy , the single animal ( AO8070 ) that succumbed to Ftt challenge ( day 9 ) exhibited a high bacterial burden in the lung compared to the 5 NHPs that survived Ftt challenge ( day 30 ) ( Fig . S5 ) . At the termination of the study , 2 LVS-vaccinated NHPs exhibited elevated bacterial burdens in the lung , elevated serum CRP levels , and elevated respiration rates ( day 30 ) ( Fig . S5 ) , suggesting they may have progressed to terminal disease in an extended study . T cell responses from vaccinated NHPs were evaluated by measuring IFNγ responses of peripheral blood mononuclear cells ( PBMC ) upon stimulation with either Fn iglD or LVS via ELISPOT ( Fig . 4A and 4B ) . PBMCs were collected from all Fn iglD- and LVS-vaccinated NHPs prior to vaccination ( naïve ) and at day 14 post vaccination . Group responses are shown in Fig . 4A & 4B; responses of individual NHPs are shown in Fig . S6 . Neither the Fn iglD- nor the LVS-vaccinated NHP groups had significant increases ( p>0 . 05 t test ) in cellular responses to either LVS or Fn iglD at day 14 post vaccination . PBMCs were also collected from Fn iglD- and LVS-vaccinated NHPs that survived pulmonary challenge with Ftt , 30 days post-challenge . PBMCs were collected from all LVS-vaccinated NHPs , but only successfully collected from three of the five surviving Fn iglD-vaccinated NHPs ( A08036 , A08245 , and A09393 ) . Group responses are shown in Fig . 4A; individual responses are shown in Fig . S6 . The Fn iglD-vaccinated NHP group that survived Ftt challenge showed a significant increase in cellular responses to both LVS and Fn iglD stimulation . In contrast , the LVS-vaccinated NHP group that survived Ftt challenge showed a similar cellular response upon LVS stimulation to that seen prior to Ftt challenge , and no response to Fn iglD stimulation . Individually , all three Fn iglD vaccinated NHPs tested ( AO8036 , AO8245 , AO9393 ) exhibited enhanced cellular responses post-Ftt challenge to both Fn iglD and LVS ( Fig . S6 ) , whereas only one of four LVS-vaccinated NHPs ( AO7746 ) mounted an enhanced cellular response post-Ftt challenge to LVS and Fn iglD . These results suggest that Fn iglD vaccination of NHPs primes T cells that provide a robust response upon challenge with Ftt . The humoral response in vaccinated NHPs was evaluated by ELISA against whole killed bacteria . Total IgG responses to whole cell Fn iglD , LVS , and Ftt were determined for both Fn iglD- and LVS-vaccinated NHPs ( Fig . 4B ) . While the strongest initial response ( Day 14 ) was toward Fn in Fn iglD-vaccinated animals and toward LVS in LVS-vaccinated animals , cross-reactive antibodies to LVS or Fn and Ftt were induced in vaccinated NHPs at day 30 . Increases in serum antibody titers were seen against all three subspecies in Fn iglD-vaccinated NHP after challenge with Ftt . Interestingly , a comparison of the individual serum antibody titers in the Fn iglD vaccinated NHPs ( Fig . S7 ) revealed lower levels of Fn-specific serum IgG 30 days post vaccination in the animal that succumbed to disease ( A08070 ) than in the 5 other animals that survived challenge . This suggests that anti-Fn antibodies may represent a correlate of protection in this model with this vaccine , although further studies with increased sample size are needed to determine this . To determine if a humoral response to LPS OAg was induced in vaccinated NHPs , we performed Western immunoblot analyses of sera from Fn iglD- and LVS-vaccinated NHPs against purified LPS from Fn and LVS ( Fig . 4E ) . The serum from a Fn iglD-vaccinated NHP ( AO8245; Fig . 4E ) reacted strongly with Fn LPS and not at all with LVS LPS , whereas a LVS-vaccinated NHP ( AO8090 ) reacted with LVS LPS and not at all with Fn LPS . These results again confirm that the OAg of Fn and LVS/Ftt are distinct and demonstrate that in NHP , humoral responses directed towards Fn OAg do not crossreact with Ftt ( LVS ) OAg , and vice versa . Moreover , the NHP humoral response in both vaccinated groups appears primarily directed to OAg associated with LPS , and not the OAg capsule , as was seen in vaccinated rats ( Fig . S2 ) . To identify immunodominant humoral protein antigens associated with Fn iglD vaccination in rats and NHP , sera from vaccinated animals was subjected to a Francisella proteome microarray [27] . The 10 most reactive antigens with NHP sera at day 28 post pulmonary vaccination with Fn iglD ( compared to naïve NHP sera ) are listed in Table 2; antigens are listed by the corresponding homologous ORF in Ftt . Comparisons are made to the immunodominant antigens with NHP sera at day 28 post subcutaneous vaccination with LVS , as well as to the immunodominant antigens with rat sera at day 28 post pulmonary vaccination with Fn iglD ( compared to naïve rat sera ) ( from Fig . 2 ) . The immunodominant antigens identified with Fn iglD-vaccinated NHP sera that are also one of the 20 most reactive antigens with the other sera are noted with a “+” ( Table 2 ) . A comparison is also made to immunodominant antigens identified with sera from mice vaccinated with killed LVS delivered with adjuvant intramuscularly from a previous study [27]; this vaccination regimen partially protected mice ( 40% protection ) against challenge with 6 CFU Ftt delivered subcutaneously . It is notable that the top five immunodominant humoral protein antigens recognized by NHP vaccinated with Fn iglD via the pulmonary route are also immunodominant antigens following vaccination with Fn iglD in rats via the pulmonary route , or following vaccination with LVS ( live ) in NHP via the intradermal route or LVS ( killed ) in mice via the intramuscular route . In fact , nine of the top ten immunodominant protein antigens are shared between NHP vaccinated with either Fn iglD or LVS , and mice vaccinated with LVS ( Table 2 ) . Four of these antigens were identified as being secreted or shed during Ftt infection of mice [28] . The immunodominant antigens reactive with either NHPs or rats vaccinated via the pulmonary route with Fn iglD shared the top five antigens in common . All of the vaccinated NHPs and rats in these groups survived pulmonary challenge with >1000 CFU Ftt , with the exception of one NHP ( A08070 ) . While anti-whole cell Fn humoral reactivity was lower in A08070 ( Fig . S7 ) , humoral reactivity to these ten specific immunodominant antigens was not obviously deficient . Further studies with larger sample sizes are needed to determine if specific humoral response ( s ) represent correlate ( s ) of protection against pulmonary Ftt challenge . These results demonstrate that a core set of immunodominant antigens stimulate the humoral response during vaccination , regardless of route , animal host , or Francisella subspecies . Ftt acquired through the pulmonary route leads to serious disease with a high mortality rate in humans . Although pneumonic tularemia caused by natural Ftt infection is relatively rare , this bacterium was investigated as a bioweapon by several government programs , and the potential exists for its illicit use against human populations . Because of this , Ftt has been classified as a select biothreat agent , and efforts are underway to develop an effective vaccine against pulmonary exposure to Ftt . Francisella subspecies are facultative intracellular bacteria that primarily reside within cells in infected animals , and thus T cell-mediated immunity is an important component of protection against tularemia . However , humoral immunity has also been shown to contribute to protection against Francisella infection [29] , [30] . In limited studies , LVS vaccination via scarification of humans provided protection against pulmonary Ftt challenge , but the vaccine strain needed to be live rather than killed [7] . Due to questions regarding phase variation , genetic cause of attenuation , and levels of protection afforded , it is questionable whether LVS will be approved for human usage . LVS still serves as a useful model for the stimulation of protective immunity in various animal models of tularemia , and we have shown here that it stimulates protective immunity against pulmonary exposure to Ftt in cynomolgus macaques . The ability of a live attenuated Francisella strain such as LVS to protect against pulmonary Ftt exposure indicates that a genetically defined live attenuated Francisella strain may constitute the optimal tularemia vaccine , especially since no protective subunit antigens have yet been identified . F . novicida ( Fn ) is closely related to F . tularensis [31]; although it is officially classified as a separate species , it is frequently referred to as a subspecies of F . tularensis because of this close genetic relationship . Fn has generally been discounted as a potential vaccine against Ftt because although vaccination of mice with live attenuated Fn strains can induce good protection against homologous pulmonary challenge with wildtype Fn , it provides no protection against pulmonary challenge with Ftt . However , vaccination of mice with live attenuated Ftt strains also provides little protection against pulmonary challenge with Ftt , as we have shown here , suggesting that the mouse model may not be appropriate for the assessment of vaccine potential due to its extreme sensitivity to Francisella infections . Indeed , mice are highly susceptible to both Fn and LVS infections , despite the low virulence of these strains in humans . We would argue that tularemia vaccine development for humans requires animal models that reflect human sensitivities to the various Francisella species/subspecies . The Fischer 344 rat reflects the relative sensitivities of humans to Francisella infections , in that it is sensitive to Ftt pulmonary infections , but resistant to pulmonary Fn infections ( approximately 104-fold difference in LD50; [11] ) . Importantly , rats that survive Fn infection are protected against subsequent pulmonary challenge with Ftt , demonstrating the efficacy of Fn as a tularemia vaccine platform in this model [11] . In the current study , utilizing the same attenuating mutation ( iglD ) in either the Fn or Ftt background , we showed that vaccination of rats with either Ftt iglD or Fn iglD strain provided protection against Ftt pulmonary challenge . This demonstrates that , at least in this model , there is no disadvantage to utilizing Fn instead of Ftt as the vaccine platform . With a single oral vaccination of Fn iglD high levels of protection ( 83% ) were achieved against pulmonary Ftt challenge . Even higher levels of protection ( 100% ) were achieved against Ftt pulmonary challenge by a single pulmonary ( intratracheal ) vaccination with Fn iglD . We have previously shown that a Fn strain containing a different attenuating mutation ( iglB ) that prevents intramacrophage replication can also protect Fischer 344 rats against pulmonary Ftt challenge when administered by either pulmonary or oral vaccination [23] . The Fn iglD strain used in the current study appears to induce higher levels of protection in rats , although further direct comparative studies would be needed to establish the relative protective capacities of these two potential vaccine candidates . Regardless , the successes of attenuated Fn strains to protect rats against Ftt pulmonary challenge indicate the promise of this platform in tularemia vaccine development . Given their close genetic relatedness with humans , non-human primates are considered to be valuable models of disease , especially for vaccine development . The cynomolgus macaque is susceptible to pulmonary Ftt infection , which results in a fatal systemic disease similar to that seen in humans ( [15]; manuscript in preparation ) . Additionally , we show here that LVS vaccination via the subcutaneous route protects these NHPs against pulmonary Ftt challenge , similar to humans . Importantly , pulmonary vaccination of cynomolgus macaques with a single dose of Fn iglD also provided high levels of protection ( 83% ) against aerosol challenge with Ftt ( >1000 CFU ) . This is the first demonstration of efficacy of a defined live attenuated vaccine strain against aerosol Ftt exposure in a NHP . In this model , indicators of disease progression include increased respiration rate , elevated serum CRP levels , and high bacterial organ burdens . Vaccination of the NHPs with Fn iglD resulted in reduction in all these indicators following pulmonary Ftt challenge , similar to vaccination with LVS . Analyses of the sera from vaccinated animals indicated that the immunodominant protein antigens recognized by NHPs vaccinated with Fn iglD were largely the same ( 9 of 10 ) as those in NHPs vaccinated with LVS , suggesting that humoral immunodominant protein antigens are conserved between Fn and Fth/Ftt . Four of these antigens ( FTT0472 , FTT0975 , FTT1484 , FTT1696 ) were also identified as within the top 25 immunoreactive antigens using the same proteome microarray with convalescent sera from human patients with Ft infections [32] . Four additional of these antigens ( FTT0901 , FTT1103 , FTT1539 , FTT0863 ) were identified by 2-D immunoblotting as immunoreactive with convalescent sera from human patients with Fth infections [33] These immunodominant antigens may provide a guide to tularemia subunit vaccine development in the future . Notably , Fn iglD vaccination of NHPs induced strong reactivity to Fn LPS but no cross-reactivity to LVS/Ftt LPS , suggesting that protection against Ftt infection by this vaccine does not include antibodies against the LPS OAg . For the near term , the Fn iglD strain has several characteristics that make it an attractive tularemia vaccine candidate . First , Fn exhibits low virulence in healthy humans , making it an inherently safer vaccine platform than the high virulence Ftt and Fth strains . Second , because of the inherent low virulence , Fn is exempt from select agent status , unlike Ftt and Fth , which allows for ease of use , transport , genetic manipulation , etc , without need for high level biocontainment facilities . Third , the defined iglD mutation prevents intracellular replication in permissive host cells and virulence in permissive animal models , resulting in a highly attenuated and inherently safer strain . Finally , Fn is more amenable to genetic manipulations than Ftt or Fth [34] , which facilitates the further development of this vaccine platform to enhance efficacy and provide protection against heterologous antigens . This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . Animal protocols involving rodents were approved by the University of Texas at San Antonio Institutional Animal Care and Use Committee ( IACUC ) under protocol MU009 ( RA ) . The animal protocol for NHPs was approved by the Lovelace Respiratory Research Institute IACUC under protocol FY09-126 . LRRI has attending veterinarians and animal care staff that are available 24 hrs a day , 7 days a week to assist in any animal care issues . All study animals were housed individually in primate cages , and food and water were supplied ad libitum except when animals were removed from their cages for study procedures . Harlan Teklad Certified 20% Monkey Diet ( W ) 2050C was fed to the animals daily , and for daily food enrichment , each animal received ¼ cup of fruit or vegetable prepared by enrichment technicians . The Study Director and the Attending Veterinarian discussed the study protocol and agreed upon scientifically appropriate analgesic , anesthetics and tranquilizing drugs prior to submission of the protocol to the LRRI IACUC . All necessary efforts were made to minimize discomfort , distress , pain , or injury to study animals . To ameliorate suffering , NHPs were conditioned to a restraint collar and restraint chair for sampling , and an implanted transponder allowed for non-invasive measurement of body temperature and respiration rate by a hand-held device . NHPs were anesthetized under the guidance of a veterinarian or a registered veterinary technician ( RVT ) to perform the following procedures: general physical examination , collar placement , aerosol challenge and euthanasia . NHPs were kept warm while under anesthesia with delta phase heating pads . All anesthetic doses were determined from the most recent weight , and NHPs were constantly monitored for respiration and recovery by a veterinarian or RVT . Aerosol challenge doses of F . tularensis were delivered to anesthetized NHPs in a head-only exposure chamber , during which they were breathing freely . The method of euthanasia selected for these studies was administration of barbiturate overdose via intravenous or intramuscular injection following notification of and authorization from the Study Director or the Attending Veterinarian . Euthanasia is always administered to individual animals in a separate area out of the sight of other surviving study animals . The Fn iglD strain KKF37 [19] is isogenic with wildtype Fn strain U112 and the Ftt iglD strain KKT8 is isogenic with wildtype Ftt strain Schu S4 . The Ftt iglD strain had both copies of iglD ( iglD1 iglD2 ) inactivated by a Group II intron targeted to iglD , as described in [35] . Francisella strains were grown in tryptic soy broth ( TSB ) ( BD Biosciences ) supplemented with 0 . 1% ( w/v ) L-cysteine ( Fisher Scientific ) and sodium metabisulfite ( Sigma ) , iron sulfate ( Mallinckrodt ) , and sodium pyruvate , all at 250 µg/ml final concentration , or Chamberlain's defined medium [36] . BALB/c mice were inoculated intranasally as described previously [37] . Female Fischer 344 rats were inoculated intratracheally or orally as described previously [23] . After sacrifice , spleens were collected for T cell recall assays , which were performed as previously described [23] . All animal protocols have been approved by the University of Texas at San Antonio Institutional Animal Care and Use Committee and Institutional Biosafety Committee . Male and female cynomolgus macaques ( Macaca fascicularis , Vietnamese origin , approximately 2 yrs old ) were received from Covance ( Alice , TX ) . LVS vaccinees were immunized with ∼1 . 8×108 LVS organisms by the subcutaneous route . Fn-iglD vaccinees were immunized via bronchoscope with a dose of 1×108 CFU . Control animals were untreated . This animal protocol was approved by the Lovelace Respiratory Research Institute IACUC . For challenges , Ftt Schu S4 was nebulized using a Collison MRE-3 nebulizer and delivered to the anesthetized NHPs in a head-only exposure chamber . The aerosol was sampled directly and viable Ftt CFU were confirmed by quantitative bacterial culture . Blood was collected via venipuncture and analyzed by culture on chocolate agar , and serum was analyzed for CRP , LDH , BUN , ALT and AST . At necropsy , tissues were taken and analyzed for the presence of Ftt by quantitative culture . Peripheral Blood Mononuclear Cells ( PBMC ) were prepared from NHP blood by gradient separation ( Lymphoprep , Accurate Chemical and Scientific Corp . ) . For measurement of cellular responses , PBMCs ( 200 , 000/well ) were added to ELISPOT plate wells ( Mabtech #3420M-2HW-Plus ) that had been coated with anti-monkey IFNγ ( 15 µg/mL , clone GZ-4 , 100 µL/well ) . UV-inactivated Fn-iglD ( 2×106 CFU/ml ) or formalin-fixed LVS ( 1×105 CFU/ml ) were then added as stimuli . Plates were incubated for 20 hours ( 37°C; 5% CO2 ) , washed with PBS , followed by the addition of biotinylated mouse IgG1 anti-monkey IFNγ antibody ( clone 7-B6-1; 0 . 1 µg/well ) and incubation at RT for 2 h . Plates were washed with PBS , 100 µl/well streptavidin-horseradish peroxidase ( HRP ) was added ( 1∶1000 dilution in PBS 0 . 5% FCS ) , and plates were then incubated at RT for 1 hour . Plates were washed and HRP substrate ( 100 µl/well ) was added . After 20–40 min of incubation at RT , plates were rinsed with water , dried , and read on a CTL Immunospot reader . Serum ELISAs were run to obtain 50% binding titers to UV-inactivated Fn iglD , LVS , or Ftt ( 106 cells/well ) . Secondary antibody specific for either rat total antibody , IgG1 , or IgG2a ( Southern Biotech , Birmingham , AL ) , or monkey IgG ( KPL , Gaithersburg MD ) was added for 1 hr . After wash , TMB substrate ( BD Biosciences ) was added . Sera were also evaluated by Francisella proteome microarray ( Antigen Discovery Inc . , [27] ) . Sera were analyzed for reactivity against purified Fn or LVS LPS ( kind gift of J . Gunn ) by Western immunoblot , utilizing either anti-rat ( GE Healthcare ) or anti-monkey ( KPL ) HRP conjugate . Each well contained either 50 µg ( NHP ) or 75 µg ( rat ) purified LPS .	Francisella tularensis is a bacterium that causes the infectious disease tularemia . F . tularensis has been developed as a biothreat agent , because it causes high morbidity and mortality when spread by aerosol . There is currently no approved vaccine for human use , making mankind vulnerable to the illicit use of this organism . F . tularensis contains a cluster of genes in the Francisella Pathogenicity Island ( FPI ) that are required for replication inside host macrophages and virulence . In the current study we created a live vaccine strain by inactivating an FPI gene , iglD , in a closely-related species that does not cause disease in humans , F . novicida ( Fn iglD ) . We demonstrate that vaccination with Fn iglD protects against exposure to airborne F . tularensis . Fn iglD vaccination induces antibody and cellular immune responses and protects two different animals , rats and non-human primates , against lethal pulmonary tularemia challenges . These two animal models reflect human sensitivity to F . tularensis . Our results suggest that a vaccine made from the low virulence F . novicida will protect humans against aerosol exposure to this dangerous pathogen .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "infectious", "diseases", "medicine", "and", "health", "sciences", "biology", "and", "life", "sciences", "immunology", "microbiology" ]	2014	Live Attenuated Francisella novicida Vaccine Protects against Francisella tularensis Pulmonary Challenge in Rats and Non-human Primates
Patients with localized tuberculoid and generalized lepromatous leprosy show respectively Th1 and Th2 cytokine profile . Additionally , other patients in both types of leprosy also show a non discriminating Th0 cytokine profile with both interferon-γ and IL-4 . The present study investigated the role of Th17 cells which appear to be a distinct subtype of Th subtypes in 19 tuberculoid and 18 lepromatous leprosy patients . Five healthy subjects with long term exposure to infection and 4 skin biopsies from healthy subjects undergoing cosmetic surgery were used as controls . An array of Th17 related primers for cytokines , chemokines and transcription factors was used in real time reverse transcribed PCR to evaluate gene expression , ELISA for cytokine secretion in the supernatants of antigen stimulated PBMC cultures and flow cytometry for establishing the phenotype of the IL-17 , IL-21 producing cells . IL-17 isoforms showed significantly higher expression and release in supernatants of antigen stimulated PBMC cultures and dermal lesions of healthy contacts and tuberculoid leprosy as compared to lepromatous leprosy ( p<0 . 003 ) . This was further confirmed by Th17 associated transcription factor RORC , cytokines IL-21 , IL-22 , and IL-23 , chemokines MMP13 , CCL20 , CCL22 . Of interest was the association of IL-23R and not IL-6R with IL-17+ cells . The Th17 cells were CD4+ CCR6+ confirming their effector cell lineage . Polarized Th1 cytokines were seen in 3/7 tuberculoid and Th2 cytokines in 5/10 lepromatous leprosy patients . Of importance was the higher association of Th17 pathway factors with the non-polarized Th0 types as compared to the polarized Th1 and Th2 ( p<0 . 01 ) . Our study draws attention to a third type of effector Th cell that may play a role in leprosy . Leprosy caused by Mycobacterium leprae continues to be a public health challenge in developing countries where total number of new cases reported in 2011 was 210 , 075 in spite of multi drug regimen and good public health practices [1] . Of this , South East Asia alone reported the highest number of new cases of 160 , 132 followed next by the Americas with 36 , 832 . The causative agent is non-cultivable by conventional methods and infects mainly man and the nine banded armadillo . Clinical leprosy presents uniquely as a five point clinic-pathological spectrum with polar ( TT ) and borderline tuberculoid ( BT ) leprosy at one end with patients showing paucibacillary , hypoanesthetic , hypopigmented skin patches . At the other end of the spectrum lie polar ( LL ) and borderline lepromatous leprosy ( BL ) with generalized multibacillary skin lesions . In between lies borderline borderline ( BB ) leprosy which is unstable with patients moving towards tuberculoid or lepromatous presentations [2] . M . leprae is the only bacillus that infects peripheral nerves and shows earlier involvement in tuberculoid and relatively later pathology in lepromatous leprosy [3] . It is now well established that subjects with the localized disease have good T cell immunity to the pathogen and poor antibody responses whereas the converse is observed in lepromatous leprosy . The dichotomy between T cell and antibody responses appeared to be in consonance with the concept of Th1 and Th2 subsets of helper T cells with mutually exclusive cytokine signatures [4] , [5] . Thus lepromatous leprosy was associated with IL-4 , IL-5 and absence of IFN-γ in antigen stimulated PBMC cultures as well as skin lesions indicating Th2 polarization and tuberculoid leprosy in contrast was reported to show Th1 responses where IFN-γ , IL-2 were the predominant cytokines [6] , [7] , [8] . However , many patients in both types of leprosy also show concomitant presence of IL-4 and IFN-γ in antigen stimulated PBMC cultures indicating the presence of non polarized Th0 responses to the pathogen [6] , [7] , [9] and greater heterogeneity in cytokine producing cells at clonal level [10] . At present there is no clear consensus on the central mechanisms that underlie the leprosy spectrum or the antigen specific T cell anergy that is associated with lepromatous leprosy . The lack of a suitable animal model that mimics human leprosy limits our understanding of the events prior to the establishment of the clinical spectrum . In addition to Th1 and Th2 , a third subset of effector cells that produce IL-17 has been identified in the past decade and designated as Th17 [11] , [12] . The primary role of this cell appears to be clearance of pathogen which has not been controlled by Th1 and Th2 . Attention was first drawn to the pivotal role of this subset in experimental auto immune encephalitis and subsequently implicated in other infectious diseases [13] such as tuberculosis [14] , [15] , leishmaniasis [16] , and fungi [17] as well as disorders such as cancers [18] , [19] , mucosal bowel disease [20] psoriasis [21] and vitiligo [22] . Some species related differences are evident in Th17 cells of mice and man [23] . The unique cytokine of Th17 is IL-17 which has several isoforms of which IL-17A and IL-17F appear to be important in regulating immune responses in autoimmunity [24] . Moreover other cytokines such as IL-21 [25] , IL-22 [26] are also secreted by Th17 cells along with chemokine CCL20 [27] . In man , the transcription factor , retinoic acid related orphan nuclear hormone receptor C ( RORC ) is associated with Th17 cells and is a homologue of the murine RORγt [28] . Moreover , receptors IL-23R [29] , IL-1R1 [30] , chemokines CCR6 [31] , [32] and CCR4 [32] are also expressed on Th17 cells . Some reports have demonstrated that human Th17 cells are able to produce both IL-17A and IFN-γ . IFN-γ production required IL-12 [33] . Th17 clones were seen to express IL12rβ2 in addition to IL-23R and Th1 related T-bet as well as RORC [34] . With a view to understanding the basis of the leprosy spectrum particularly in patients who showed non Th1 and non Th2 polarization we investigated the role of Th17 cells in tuberculoid and lepromatous leprosy patients at the two ends of the spectrum in both PBMC and chronic inflammatory/granulomatous skin lesions . We also compared the patients with the healthy clinically normal subjects who had been in constant contact with leprosy patients for 1–2 years ( HC ) , using quantitative real time polymerase chain reaction ( qPCR ) for expression of genes in the Th17 pathway , ELISA for quantitation of cytokines and flowcytometry for identification of cell types in PBMC cultures stimulated with sonicated heat killed armadillo derived M . leprae antigen ( ML ) . Of interest was the higher Th17 cell activity observed in HC and borderline tuberculoid subjects ( BT ) as compared to those with polar lepromatous leprosy ( LL ) . Importantly , the association of Th17 cell signatures in leprosy patients was observed in patients who did not show the conventional Th1 and Th2 phenotypes but had a non polarized subset of CD4+T cells ( Th0 type ) which expressed and secreted IFN-γ and IL-4/IL-5 in antigen stimulated PBMC cultures . In this study we considered the two clinical types of paucibacillary tuberculoid ( BT ) and multibacillary lepromatous leprosy ( BL , LL ) as reflecting natural relative resistance and susceptibility respectively to M . leprae infection . In addition , healthy subjects who had been in contact for 1–2 years with leprosy patients and had not developed the disease and considered be resistant to the disease were included as controls . In view of recent reports on Th17 in relation to resistance to intracellular pathogens [13] , [17] and the lack of consensus on the mechanisms underlying anergy in leprosy , we investigated Th17 cell pathways using PBMC and skin lesions from patients and healthy skin from subjects undergoing cosmetic surgery . Expression of the genes in the pathway of Th17 as well as the identification of cell types expressing IL-17 and IL-21 was investigated using quantitative RT-PCR and flowcytometry respectively . ELISA was used to evaluate the secretion of relevant cytokines into the supernatants of ex vivo cultures of antigen stimulated PBMC . The study protocol , informed consent forms in local language and all procedures were approved by the Institutional Ethical Committee of Safdarjung Hospital ( No . 26-11-EC ( 25/31 ) . Written informed consent was obtained from the patients after counseling and prior to taking blood samples and skin biopsies . 37 newly diagnosed leprosy patients without history of anti-leprosy treatment ( 26 males , 11 females aged between 19–60 years ) attending the Leprosy Clinics of the Department of Dermatology , Safdarjung Hospital , New Delhi were included in the study ( Table 1 ) . Leprosy type was determined by clinical and histological criteria on the basis of Ridley-Jopling classification [2] . Study group included 19 borderline tuberculoid ( BT ) and 18 polar lepromatous ( LL ) leprosy patients . Five healthy subjects who were house hold members in contact with leprosy patients for 1–2 years were included as controls ( HC ) since they showed no evidence of disease after continuing exposure to the same environment and in close contact with infected patients . In addition , normal skin samples ( N ) from 4 patients undergoing cosmetic surgery for burns were also included as controls . 10 ml of venous blood was collected in heparinized sterile tubes . PBMC were separated by density gradient centrifugation on Ficoll-Hypaque ( Histopaque , Sigma Aldrich , USA ) after diluting with 1∶1 in RPMI 1640 ( Sigma Aldrich ) as described earlier [35] . Mononuclear cells were isolated by centrifugation at 800×g for 20 minutes; cells were washed three times in sterile 1× HBSS ( GIBCO NY , USA ) and re-suspended in RPMI 1640 . Cell viability as estimated by 0 . 2% trypan blue ( Sigma Aldrich ) ranged from 95–98% . PBMC cultures were undertaken as described previously [35] . In brief , 1 . 5×106 cells/ml suspended in RPMI 1640 ( GIBCO NY , USA ) with 10% pooled human AB serum , 2 mM L-glutamine , 100 units of penicillin ( Alembic Chemicals , India ) and 100 ug streptomycin ( Sarabhai Chemicals , India ) were cultured in sterile flat bottom 24- well plates ( Falcon , USA ) as follows : Cells i ) alone ii ) with 5 ug/ml phytohemagglutinin ( Sigma Aldrich ) iii ) with 10 ug/ml of M leprae sonicated antigen ( ML ) kindly provided by P J Brennan of Colorado State University . Initial studies on 3 each of BT and LL subjects were undertaken where cultures were incubated for 24 , 48 , and 72 h at 37°C in humidified 5% CO2+air . Subsequently all studies were undertaken at 48 h culture period at which time optimum results were obtained ( Figure 1A ) . After harvest , cells were washed as above and stored in RNA later ( Sigma Aldrich ) for gene expression studies or processed for flow cytometry analysis as given below . The paired supernatants were collected and stored at −80°C for ELISA . Sterile 4 mm punch ( Cardiograph Co , Satara , Maharashtra , India ) biopsies from skin lesions were obtained after application of 1% lignocaine ( Kremoint Pharma , Maharashtra , India ) as local anaesthesia . Part of the biopsy was processed in buffered formalin for routine histopathology . The reminder was placed in 1 ml of RNA later and stored at −80°C . Immediately prior to RNA isolation the frozen stored skin biopsies were thawed and the tissue crushed with liquid nitrogen in pestle and mortar . RNA was isolated from both crushed skin tissue and PBMC cultures using RNeasy Mini Kit ( Qiagen , Maryland , USA ) according to the manufacturer's instructions . The isolated RNA was quantified using Nanodrop spectrophotometer ( Nanodrop Technologies , Wilmington , USA ) and purity at 260/280 from 1 . 8 to 2 . 0 was considered to be optimum . The quality of RNA was also checked for 28 s and 18 s RNA by electropherogram using Bio analyzer ( Agilent Technologies , Inc , Singapore ) . Only samples with optimum RNA Integration Number value of ≥7 were used . For cDNA synthesis 1 µg of total RNA was transcribed with RT First strand kit ( SA Biosciences , MD , USA ) . Reactions were performed according to the manufacturer's instructions and the cDNA stored at −20°C till further use . Gene expression was measured in quantitative real-time polymerase chain reaction ( qPCR ) using customized Th17 PCR array ( SA Biosciences , Quiagen Co . CA , USA ) as per the manufacturer's instructions . Duplicate samples of cDNA from each subject was amplified in 96 well plates containing primers for the genes of interest , cytokines and cytokine receptors: IFN-γ , IL-4 , IL-5 , IL-6 , IL-6R , IL-2 , IL-27 , IL-17A , IL-17C , IL-17D , IL-17F , IL-1β , IL-21 , IL-22 , IL-23A , IL-23R; chemokines: MMP13 , MMP3 , CCL 20 , CCL22; signaling molecules and transcription factors: RORC , SOCS1 , STAT3; 5 housekeeping genes β2M , HPRT1 , RPL13A , GAPDH , ACTB , using qPCR . 1 µg of cDNA was used per reaction in each well containing the ready to use PCR master mix and appropriate primers . These were then subjected to qPCR ( ABI 7000 , Applied Biosystems Singapore ) for 2 h . Threshold cycles values were normalized and expressed as ΔCt: mean Ct of gene of interest-mean Ct of 5 housekeeping genes . All reagents were obtained from BD Biosciences , San Diego , CA . and used as per manufacturer's instructions . For intracellular staining , ex vivo cultured cells were incubated with monensin ( BD GolgiStop ) for 8 h prior to harvest to block secretion of cytokine . For surface staining , 0 . 5×106cells/50 ul in staining buffer were incubated with cocktail containing anti human CD3 ( Per cpcy-5 . 5 ) , CD4 ( APC-H7 ) and CD8 ( PE-Cy7 ) along with isotype controls of PE ( mouse IgG1 ) , Alexa Fluor 488 ( mouse IgG1 ) , Alexa Flour 647 ( mouseIgG1 ) for 45 min at 4°C after which cells were washed two times and permeabilized with permeabilizing/fixation solution ( containing saponin/paraformaldehyde ) for 30 m at 4°C . The cells were washed two times and resuspended in Perm/Wash buffer and incubated with anti human IL-17A ( Alexa Fluor-647 ) , IL-17F ( Alexa Fluor-488 ) , and IL-21 ( PE ) , ) at 4°C for 30 min in the dark followed by two washes as before , resuspended in 500 ul . For evaluating phosphorylation of STAT3 , cultured cells were first fixed for 10 min at room temperature , permeabilized as before with appropriate buffer and stained with a cocktail of PE labeled anti mouse STAT3 , anti human IL-17A , CD3 , CD4 and CD8 antibodies . Stained cells were acquired using BD FACS aria flow cytometry and analyzed with BD FACS Diva software . Cytokines were estimated by ELISA ( Ready Set Go , e-Bioscience , San Diego , CA , USA ) in duplicate culture supernatants as per manufacturer's instructions . In brief , 96-well plates ( Nunc , Rochester , NY , USA ) were coated overnight at 4°C with capture biotin conjugated anti human antibodies for each of the cytokines , IL-17A/F , IL-21 , IL-22 , IL-23A , IL-6 , IL-1β , IFN-γ and IL-5 . Plates were washed 5 times , blotted and blocked with assay diluents for 1 h at room temperature . 100 µl/well of culture supernatant was added and plates incubated overnight at 4°C . After washing each well with buffer , appropriate avidin-horseradish peroxidase-conjugated anti-mouse antibody was added and the plates incubated at room temperature for 30 min . After washing as before , color development was undertaken using peroxidase color substrate TMB ( Tetramethylbanzedine ) and the reaction stopped by the addition of 1N H2SO4 . The optical density ( OD ) of each well was read at 450 nm . Nonparametric statistics was performed using Graph Pad Prism version 5 ( GraphPad Software , Inc . , San Diego , CA , USA ) . Data were analyzed using two tailed Mann-Whitney for significance and Spearman tests for correlation coefficient . p≤0 . 05 was considered as statistically significant . The role of transcription factors RORC and STAT3 associated with Th17 cells was explored . Figure 6A shows not only association of Th17 cells with RORC but also its increased expression in both HC and BT as compared to LL patients in the antigen stimulated PBMC cultures ( p<0 . 0007 and p<0 . 0001 respectively ) and the dermal lesions ( p<0 . 02 , p<0 . 01 respectively ) . It also showed strong correlation with IL-17A and IFN-γ in PBMC cultures ( r2 = 0 . 94 and 0 . 84 respectively , p<0 . 0001 ) of both types of leprosy . STAT3 on the other hand showed high expression but did not discriminate between the clinical groups ( Figure 6A ) nor showed correlation with IL-17 expression . With a view to further dissect the role of STAT3 we next studied its phosphorylation status in 3 subjects each of BT , LL and HC . Figure 6B presents representative data on CD3+ gated cells , wherein CD4+ cells of antigen stimulated PBMC showed phosphorylated STAT3 . The highest percentage of cells was in BT ( 85 . 1 to 90 . 58% ) followed in decreasing order by HC ( 45 . 8 to 46 . 9% ) and LL ( 19 . 2 to 21 . 0% ) subjects . More importantly , in all clinical groups STAT3 phosphorylation was associated with >50% of CD4+IL-17+ cells ( Figure 6B ) with BT showing the highest percentage ( 69 . 2 to77 . 8% ) . One of the known promoters of phosphorylation and activation of STAT3 signaling pathway is SOCS1 [37] , [38] which also showed higher expression in BT ( p<0 . 01 ) as compared to LL patients in PBMC cultures . In skin lesions it was higher in BT when compared to normal skin ( p<0 . 004 ) ( Figure 6A ) . In summary , it is evident that RORC was associated with IL-17+ cells and phosphorylated STAT3 rather than total STAT3 expression revealed functional differences in leprosy and was in agreement with the activation by SOCS1 . Cytokines IL-1β , IL-23 and IL-6 reported to influence and sustain Th17 lineage [39] , [40] were investigated . In general , the expression of all cytokines was higher in ex vivo antigen stimulated PBMC as compared to skin lesions reflecting differences due to recall responses in antigen stimulated PBMC as compared to the status of an ongoing in vivo response . IL-1β , showed high expression in all subjects without discriminating the leprosy types in PBMC cultures . Mean ΔCt± SD ranged from 2 . 6±1 . 4 , 3 . 1±1 . 3 , 2 . 3±0 . 37 respectively in BT , LL and HC . It is of interest that BT skin lesions showed statistically significant increase in expression of IL-1β ( p<0 . 008 ) as compared to normal skin . In contrast , IL-23A considered to be important for maintenance of Th17 cells [41] showed significant increase in HC and BT ( p<0 . 01 ) groups as compared to LL ( p<0 . 007 , Figure 7 , Table 2 ) . Moreover , IL-23A was significantly higher in BT ( p<0 . 0003 , Mean ΔCT ±SD ΔCt: 4 . 7±0 . 7 ) as compared to LL ( Mean ΔCt ±SD: 7 . 26±1 . 09 ) indicating differential expression within the two leprosy types . ELISA confirmed , IL-23 protein to be significantly higher in the antigen stimulated PBMC supernatants of BT as compared to LL patients ( p<0 . 001 , Table 2 ) . Surprisingly , the findings in the skin for IL-23A expression did not show discrimination between the clinical leprosy types . In agreement with the above , IL-23R ( Figure 7 ) which is considered to be important for stabilization of Th17 cells also showed increase in both antigen stimulated PBMC and skin of BT group as compared to LL ( p<0 . 001 and p<0 . 009 respectively ) . Moreover , IL-23R showed significant correlation with IL-17A ( r2 = 0 . 58 p<0 . 008 ) , IL-17F ( r2 = 0 . 52 p<0 . 02 ) , IL-21 ( r2 = 0 . 64 p<0 . 003 ) , IL-22 ( r2 = 0 . 58 p<0 . 008 ) and IFN-γ ( r2 = 0 . 59 p<0 . 007 ) . IL-6 expression also showed a dichotomy between dermal lesions and PBMC cultures of BT patients . Whereas it showed decrease in patients as compared to HC ( p<0 . 007 ) , dermal lesions showed increase as compared to the healthy skin ( p<0 . 008 ) . No distinction was seen between the leprosy types . However , IL-6 ELISA showed significantly lower levels in LL as compared to BT ( p<0 . 0001 ) in culture supernatants . Taken together our data supports the greater role of IL-23 and its ligand IL-23R in leprosy as compared to IL-6/IL-6R for IL-17 production in both circulating cells and at the site of the dermal lesions . IL-27 and IL-2 have been reported to negatively regulate Th17 cells [42] , [43] . Paradoxically , the expression of these cytokines was also higher in healthy contacts as compared to leprosy types ( p<0 . 001 ) but did not show differences within the leprosy groups even though expression was lower in LL ( Table 3 ) . These cytokines did not show significant correlation with IL-17 isoforms ( Spearman tests ) indicating that other regulatory factors may be involved in leprosy . As shown in Figure 8 , expression of MMP3 , MMP13 , CCL22 were seen to be highest in PBMC cultures of healthy contacts as compared to leprosy types . The former two chemokines showed statistically significant decrease in BT ( p<0 . 004 , p<0 . 01 respectively ) . CCL22 showed decrease in LL as compared to healthy subjects ( p<0 . 01 ) but did not discriminate the leprosy types . In contrast , skin lesions showed distinct differences in chemokine expression . MMP3 , CCL20 and CCL22 showed increase in BT as compared to LL lesions ( p<0 . 01 , p<0 . 003 , p<0 . 01 respectively ) . Of interest , was the higher expression of MMP3 and CCL20 in BT as compared to normal skin suggestive of a possible role in the trafficking of relevant cells to the sites of tuberculoid leprosy granulomas . We next investigated the relationship of Th17 cells to the Th polarization status in leprosy . The status of Th subsets in individual subjects is provided by Figure 9 showing the fluorescent signal as an indicator of the magnitude of gene expression . Tables 4 and 5 give ΔCt values of IFN-γ and IL-4 to assign Th status . Expression of Th17 related genes in the same subjects was graded arbitrarily from nil ( − ) to ( +++ ) to illustrate more easily their relationship with Th status . It may be observed that IL-17A , IL-17F , IL-21 and IL-22 were expressed more in the non-polarized Th0 group of patients in both types of leprosy ( Table 4 ) than in subjects with Th1 ( p<0 . 01 ) and Th2 ( p<0 . 01 ) subsets ( Table 5 ) . Seven of the 10 BT patients showed Th0 as indicated by the expression of both IFN-γ and IL-4 . The same patients also showed +++ expression of IL-17A , IL-17F , IL-21 , IL-22 and IL-23R . Transcription factor RORC showed moderate ++ expression in Th0 . In contrast , Th1 polarized patients showed low/nil expression ( Figure 9 and Table 5 ) . In lepromatous leprosy 5 each of 10 patients were of Th2 and Th0 . As may be noted the latter group again showed significantly higher expression levels of Th17 related genes as compared to the polarized Th2 ( p<0 . 01 ) . However , it may be pointed out that expression of the Th17 related genes in LL was lower ( ++ ) as compared to BT ( +++ ) in the Th0 groups ( Table 4 ) . RORC again showed lower level of expression as compared to the Th17 cytokines but was similar expression ( + ) in Th0 and Th2 subsets in LL patients . It is of interest that IL-17 was also associated with IFN-γ in tuberculoid leprosy patients who had Th0 cytokine profile suggesting that both cytokines are involved in the T cell immune responses . Consistent with the above findings in BT , 5 HC who had Th0 profile also showed enhanced expression of Th17 related genes . In conclusion , it would appear that Th17 cells form a third Th subset of CD4+ effector cells in leprosy in addition to the Th1 and Th2 types . Th17 cells have emerged as a third subset of Th cells that play an important role not only in autoimmune diseases where they were first described but also in many infectious diseases in man and in experimental models [14] , [15] , [16] , [17] . The present investigation was undertaken on leprosy patients with a view to understanding the immune mechanisms underlying the unique clinical types in the leprosy spectrum as well as in investigating the differences between circulating cells and local immune responses in the lesional skin . In general , expression of IL-17 isoforms in skin lesions was lower than in antigen stimulated PBMC which may be related to the differences in numbers of cells in lesions or the stronger recall responses in circulating cells . Of interest was the finding that healthy household contacts with long term exposure to the patients showed the highest IL-17 expression as compared to the diseased subjects suggesting that IL-17 was an important early adaptive immune response to M . leprae . Within the leprosy spectrum the more resistant paucibacillary form of tuberculoid leprosy had higher IL-17 associated cytokines IL-21 , IL-22 and the transcription factor RORC . The signatures of Th17 pathway genes were further endorsed by cytokine secretion into the supernatants of PBMC cultures ( Table 2 ) . Consistent with this was also the increased expression of IL-23 and IL-23R known to be involved in differentiation of Th17 cells . It appears that IL-23R was more relevant in leprosy than IL-6R reported to function in a similar manner [44] , [45] indicating that the pathway of Th17 may show subtle differences not only in man as compared to the mouse but also in different diseases . STAT3 expression per se was uniformly strong and did not show differences between the 3 clinical groups in the PBMC or skin lesions . Importantly , phosphorylated STAT3 was much higher in tuberculoid as compared to the other two groups revealing functional and signaling differences in them . SOCS1 known to be a regulator of Th17 differentiation also showed significantly increased expression ( p<0 . 01 ) in the resistant form of leprosy and supported the increased activation of STAT3 in tuberculoid leprosy . Chemokines associated with Th17 cells were found to be discriminatory more at the local level in dermal lesions of the two clinical types of leprosy than in PBMC cultures . MMP3 , CCL20 and CCL22 showed significance increase in tuberculoid as compared to the lepromatous lesions adding support to the increased lymphocytes seen in BT granulomas . Due to lack of evidence by histochemistry we were unable to formally establish Th17 lineage of the lesional lymphocytes . Using flowcytometry , we determined that a large proportion of IL-17+ cells in antigen stimulated PBMC cultures were CCR6+ suggestive of CD4+ effector T cell population [36] . Our findings in leprosy are similar to reports in cutaneous leishmaniasis [16] , tuberculosis [15] and inflammatory bowel disease [20] where Th17 cells were shown to regulate the immune responses in man . The emergence of Th17 cells during leprosy reactions of type 2 or erythema nodosum leprosum has been reported to play an important part in the inflammatory component seen in these states known to be associated with lepromatous leprosy [46] . A recent study on Brazilian populations indicated reduced expression of IL-17A in leprosy skin as compared to healthy subjects and attributed it to genetic differences [47] . The differences between our and their findings may be related to differences between the two ethnically diverse populations . That Candida albicans and Staphylococcus aureus induce Th17 cells that produce different cytokines such as IFN-γ or IL-10 in addition to IL-17 would suggest that regulation of these cells is not only pathogen dependant but also due to other factors such as cytokines and transcription factors [48] . In human tuberculosis , predominance of IL-22 over IL-17 has been seen at the site of disease [49] , a feature that was not observed in the skin lesions of leprosy . In our study it is evident that IFN-γ was produced in the same PBMC cultures concomitantly with IL-17 in healthy contacts and tuberculoid leprosy . IL-23A played a more important role than IL-6 for Th17 differentiation in leprosy which is reminiscent of the report on Mycobacterium bovis infection in a murine model where IL-17 was observed from day 1 of infection , was dependant on IL-23 and associated with granuloma formation [14] . We also explored the relationship of Th17 cells to the conventional Th1 and Th2 subtypes of T cells . Earlier reports [8] , [9] , [50] including ours had shown polarized Th1 and Th2 phenotypes in PBMC cultures of tuberculoid and lepromatous leprosy respectively . This had become a paradigm in explaining the converse pattern of CMI to antibody responses observed in leprosy spectrum . However , the finding of non polarized Th0 phenotypes in many patients of both types of leprosy required further explanation [9] . Therefore , we analyzed the status of Th17 pathway against the Th1 and Th2 background . It is evident from Figure 9 and Tables 4 and 5 that Th17 cells were strongly associated with the non polarized Th0 phenotype in both leprosy types and healthy subjects who showed concomitant IFN-γ and IL-4/IL-5 both by qPCR and ELISA of PBMC culture supernatants . It would appear that Th17 cells form a third subset of Th cells and may play an important role in those patients where Th1 and Th2 polarization is not observed . Taken together the data indicates that in addition to IFN-γ producing Th1 cells , CD4+IL-17+ cells have a role in the adaptive immune response in the more resistant form of tuberculoid leprosy and discriminate between the clinical types of leprosy . However , the Th17 distinction between the paucibacillary BT and the multibacillary LL were less obvious in polarized Th1 and Th2 states . It is possible that Th17 lineage may be an alternate pathway for bacillary clearance when the patient is unable to mount Th1 response or when Th polarization has not set in . Supportive of this were the results obtained on subjects exposed to long term infection who continued to be healthy and not develop the disease . It is not possible from this one time point investigation to conclude whether Th17 indicates T cell plasticity or constitutes a stable lineage . It is of interest that , vaccination against Mycobacterium tuberculosis lung infection produced protection which was associated with IL-23A and IL-17 responses where IL-17 preceded IFN-γ responses [51] . Whether Th17 is a rescue pathway or is a stable alternate defense immune mechanism needs further investigation . In the absence of suitable animal models which mimic the leprosy spectrum , investigations during disease states can only provide limited information on the dynamics of an immune response . Nevertheless as far as we know this is the first report in leprosy that provides evidence for the presence of Th17 cells as a third Th subset of the adaptive immune mechanism in subjects who have not developed the conventional Th1 and 2 phenotypes .	Leprosy caused by Mycobacterium leprosy continues to be a public health challenge in developing countries . It manifests as a leprosy spectrum with varied clinical forms of localized ( tuberculoid ) and generalized ( lepromatous ) disease . Consensus on the immunological basis for leprosy spectrum is lacking . T helper subsets of Th1 and Th2 based on mutually exclusive cytokines IFN-γ and IL-4 respectively were considered as a basis for the inverse relationship between T cell functions and antibody responses seen in this disease . However , there are some patients who show Th0 subset which is non discriminatory and produce both cytokines . In the present study , we provide evidence for the presence of the recently described third subset of Th17 in both antigen stimulated PBMC and skin lesions of tuberculoid leprosy . They appear to be associated with Th0 profile and may represent a third type of effector cell in leprosy .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "immunopathology", "immune", "cells", "t", "cells", "immunology", "biology", "immune", "response" ]	2013	CD4+ Th17 Cells Discriminate Clinical Types and Constitute a Third Subset of Non Th1, Non Th2 T Cells in Human Leprosy
Polyomavirus BKV is highly prevalent among humans . The virus establishes an asymptomatic persistent infection in the urinary system in healthy people , but uncontrolled productive infection of the virus in immunocompromised patients can lead to serious diseases . In spite of its high prevalence , our knowledge regarding key aspects of BKV polyomavirus infection remains incomplete . To determine tissue and cell type tropism of the virus , primary human epithelial cells , endothelial cells and fibroblasts isolated from the respiratory and urinary systems were tested . Results from this study demonstrated that all 9 different types of human cells were infectable by BKV polyomavirus but showed differential cellular responses . In microvascular endothelial cells from the lung and the bladder , BKV persistent infection led to prolonged viral protein expression , low yield of infectious progeny and delayed cell death , in contrast with infection in renal proximal tubular epithelial cells , a widely used cell culture model for studying productive infection of this virus . Transcriptomic profiling revealed the activation of interferon signaling and induction of multiple interferon stimulated genes in infected microvascular endothelial cells . Further investigation demonstrated production of IFNβ and secretion of chemokine CXCL10 by infected endothelial cells . Activation of IRF3 and STAT1 in infected endothelial cells was also confirmed . In contrast , renal proximal tubular epithelial cells failed to mount an interferon response and underwent progressive cell death . These results demonstrated that microvascular endothelial cells are able to activate interferon signaling in response to polyomavirus BKV infection . This raises the possibility that endothelial cells might provide initial immune defense against BKV infection . Our results shed light on the persistence of and immunity against infection by BKV polyomavirus . Infection of BK polyomavirus ( BKV ) in humans is widespread with seroprevalence ranging from 60 to over 90% in populations world-wide [1–3] . Seroconversion of BKV occurs in early childhood and the lifelong infection persists asymptomatically in most individuals [2 , 3] . The site and entry route of initial BKV infection and dissemination route of the virus , as well as the mode of transmission remain to be determined . Diseases associated with BKV only affect immunocompromised populations , especially transplant recipients and AIDS patients , thus host immunity against BKV is critical in limiting pathological consequences due to productive infection by the virus . Immune defense against BKV infection involves both humoral and cellular immunity [4–7] , however , our knowledge regarding BKV specific immunity at the molecular level is limited . Two observations suggest that persistence of BKV is maintained in the urinary system . First , BKV is occasionally detected in urine from healthy individuals; second , productive BKV infection is associated with nephropathy and hemorrhagic cystitis in transplant patients . However , several studies indicate a wide spectrum of BKV tissue and cell type tropism . For instance , most of the 60 NCI-60 tumor cell lines are transducible with BKV pseudovirions encapsidating either a GFP or luciferase reporter [8] . In addition , BKV has been associated with HIV-associated salivary gland disease [9–11] , and replication of BKV has been demonstrated in human salivary gland cells [10] . BKV has also been shown to infect human peripheral blood leukocytes and pancreatic cells [12 , 13] , and animal experiments have demonstrated the uptake of blood born BKV virus-like particles by endothelial cells [14] . Finally , HUVEC ( human umbilical cord vein endothelial cells ) support the growth of archetype BKV [15] and BKV infection was detected in vascular endothelial cells of a renal transplant recipient patient exhibiting vasculopathy [16] . Thus , while it is clear BKV sometimes replicates in the urinary system , especially under pathologic conditions , the site of persistence is unknown . Primary human RPTE ( renal proximal tubule epithelial cells ) provide a valuable and widely used in vitro model for understanding BKV infection [17 , 18] . RPTE support productive infection of the virus and many aspects of BKV infection have been studied using this system [17 , 19–23] . Under culture conditions , BKV undergoes a complete productive infection of RPTE , leading to the release of many progeny virions and loss of cell monolayer viability . This is not surprising given that RPTE do not elicit an efficient innate immune response against BKV infection [20 , 23 , 24] . Two separate transcriptomic profiling studies , using microarray and RNA-seq respectively , did not detect changes in gene expression indicative of activated IFN signaling [20 , 23] and cytokine profiling further confirmed a lack of pro-inflammatory responses following BKV infection [24] . In contrast , infection of RPTE with influenza A virus , herpes simplex virus 1 , cytomegalovirus and the closely related human polyomavirus JCV do produce robust antiviral responses indicating that these cells are fully capable of detecting and responding to viral pathogens [23 , 24] . However , BKV is able to trigger an innate immune response , as indicated by the elevated levels of CXCL10 and ZBP1 , released by leukocytes in response to BKV infection [24] . Taken together , these results suggest that RPTE do not mount an effective antiviral response to counter BKV infection , thus renal proximal tubular cells may not be responsible for triggering the acquired immunity that limits BKV productive infection in vivo . Furthermore , the robust productive infection exhibited by RPTE make it unlikely that persistence is maintained in these cells . In this study , we have found that BKV infection of microvascular endothelial cells ( VEC ) induces a significant innate immune response and that BKV infection persists in these cells for several weeks . These results expand our knowledge of the differential susceptibility and response of specific cells to BKV infection , and provide important implications regarding BKV-induced specific immunity and persistence . To survey BKV infectivity in human tissues , we obtained 9 types of primary human cells isolated from the lung and urinary systems through commercial sources . These included epithelial cells from lung , proximal tubule and medulla of kidney , bladder and urethra , fibroblasts from lung and bladder , and microvascular endothelial cells from lung and bladder . Cells were inoculated with BKV stock at a MOI ( multiplicity of infection ) of 1 FFU ( fluorescence forming units ) /cell and expression of BKV T antigens ( TAg ) and VP1 were examined by IF ( immunofluorescence staining ) using specific antibodies . All cell types showed expression of both TAg and VP1 at 2 or 3dpi ( days post inoculation ) ( Fig 1 ) . Co-staining showed overlapping TAg and VP1 signals in infected fibroblasts from both lung and bladder , indicating that most cells expressing TAg also produced VP1 , thus suggesting productive infection . We also observed the cytopathic effect ( CPE ) induced by BKV in each cell type . Six cell types , including epithelial cells from kidney , bladder and urethra , and the fibroblasts from the lung and bladder , underwent near complete cell death by about 14dpi ( Fig 1A and 1C , live cell images ) . Epithelial cells from the lung developed the most rapid CPE , where most cells died and detached before 5dpi ( Fig 1A ) . In contrast , CPE in the inoculated vascular endothelial cells was less obvious and many cells still survived at 14dpi ( Fig 1B ) . These results were confirmed using RPTE , lung epithelial cells and lung endothelial cells from additional donors . Specific donor information and number of donors tested are listed in S1 Table . We conclude that all 9 cell-types tested can be infected by BKV , but the effects of infection on cell viability are less pronounced in endothelial cells . To further understand the limited CPE observed in microvascular endothelial cells , referred to as VEC in the rest of the manuscript , we directly compared the responses to BKV in VEC to those in RPTE , a well-established cell culture model for studying BKV productive infection . RPTE , LVEC ( lung VEC ) and BVEC ( bladder VEC ) were inoculated with BKV and cell number , viral transcription and protein expression , as well as release of progeny virions were monitored at various times post-inoculation . Two RPTE , three LVEC and one BVEC , each from an independent donor , were used in these experiments . In RPTEs from both donors , cell number differences between the mock and BKV inoculated conditions grew larger as the infection progressed from 2 to 14 dpi ( Fig 2A ) . By 14dpi , the cell numbers of mock-inoculated RPTEs increased a modest amount , while cell number of inoculated cells dropped dramatically . In contrast , the reduction in cell numbers following BKV infection was moderate in both LVEC and BVEC ( Fig 2B and 2C ) . We next investigated viral gene expression at the RNA and protein levels . Transcripts for large T antigen ( LT ) or small T antigen ( sT ) , were monitored by RT-qPCR ( Fig 3 ) . LVEC expressed both LT and sT RNAs at levels comparable to RPTE ( Fig 3C and 3D ) . The expression of viral T antigens in LVEC1 was followed for 2 weeks , during which LT and sT expression peaked around 3dpi and then decreased gradually over the next 10 days . At 14dpi , LT and sT expressions decreased about 13 or 8-fold from their respective peak levels . Thus , the timing and the levels of viral early gene expression in both RPTE and LVEC were similar . Examination of LT by immunofluorescence showed that at 2dpi the percentage of LT positive cells was similar in RPTE1 , RPTE2 , LVEC1 , and BVEC but lower in LVEC2 ( Fig 4A ) . However , at 5dpi there were fewer LT-positive cells in all three VECs compared to both RPTEs . Quantification of positive cells was not attempted at 14dpi due to high background in the staining resulting from cell debris . The expression of LT and VP1 in mock and BKV inoculated RPTE2 and LVEC2 was also verified by Western blot , using specific anti TAg and anti-VP1 antibodies ( Fig 4B and 4C ) . Both proteins were readily detectable in inoculated cells at all tested time points . To evaluate viral DNA replication , DNA was isolated and characterized as described in materials and methods . BKV genomic DNA was first detected at 2dpi in inoculated RPTE , and the amount of DNA increased by 5dpi and 14dpi ( Fig 4D , upper panel ) . In contrast , BKV DNA was not detectable until 5dpi in LVEC2 ( Fig 4D , lower panel ) . Furthermore , we observed that the amount of BKV DNA in LVEC2 was lower than that in RPTE1 . Additional experiments showed similar differences between LVEC1 and RPTE2 ( S1 Fig ) . Next , we measured the production of infectious progeny virus for all RPTEs and VECs by titration of the culture supernatants collected at 2 , 5 and 14dpi . The titers of VEC supernatants were consistently lower than those of RPTE supernatants at all time points ( Fig 5 ) . At 14dpi , when the RPTE-inoculated monolayer was destroyed , viral titers from infected VECs were roughly 10 to 270-fold less than those from RPTEs . These results indicate that BKV infection is limited in VEC relative to RPTE . The limited infection and lack of cell death in infected VEC prompted us to monitor infected VEC over a longer period . Inoculated LVEC2 were passaged 7 times over a 3-month period . The control mock cells in this experiment were passaged 8 times in 2 months ( Fig 6A ) . While most mock cells appeared to be enlarged and flattened , a morphology indicative of senescence , the BKV inoculated cells showed less morphological changes . Expression of LT and VP1 was observed in infected LVEC2 at 66dpi ( Fig 6B ) . The presence of infectious particles was confirmed by inoculating RPTE1 with the culture supernatants collected from mock and BKV inoculated LVEC2 at 62dpi ( Fig 6C ) . Independently , BVEC were inoculated and maintained for 8 weeks without passaging , at which point the cells continued to show expression of LT and VP1 ( S2 Fig ) . The culture supernatant of BKV infected BVEC at 8-week post infection ( wpi ) also contained infectious BKV particles as verified by inoculation of RPTE1 . These results demonstrate that BKV infection persists in VEC for at least two months . To examine the influence of infection on global patterns of gene expression , we performed a series of RNA-seq experiments on BKV-inoculated RPTE and VEC . We first tried several methods of RNA isolation using RPTE , in each case comparing mock to infected cells . These included examining gene expression patterns using total RNA , cytoplasmic or nuclear RNA , as well as polysome-associated RNA . Within each experimental condition , we did not observe significant differences in RNA expression using these different methods ( S3 Fig ) . Therefore , we compared gene expression patterns of mock and infected RPTE and VEC using total RNA . We detected very few genes upregulated or downregulated at early time points ( 11 hpi for RPTE1 and 1dpi for LVEC2 ) before significant accumulation of viral transcripts occurs ( S4 Fig ) . At later time points , we identified 670 upregulated genes in RPTE1 while 840 were upregulated in LVEC2 . The complete lists of upregulated genes with corresponding log ratios can be found in S2 and S3 Tables . The lists of upregulated genes were analyzed using DAVID ( https://david . ncifcrf . gov/ ) for functional classification . This analysis uncovered several functional clusters highly enriched for genes involved in cell proliferation , including cell cycle regulation , cell division , DNA replication and DNA repair . For simplicity we refer to them as proliferation genes . The heat map shown in Fig 7 summarizes the changes in the levels of cell proliferation genes for RPTE1 and LVEC2 . Upregulation of most genes appeared to be very similar in both cell types at the two later time points ( Fig 7A ) . Indeed , out of the approximately 260 upregulated proliferation genes in RPTE1 and LVEC2 , 230 were common between the two cell types ( Fig 7B ) . In addition , the level of upregulation of these genes showed a strong positive correlation ( Fig 7C ) . These results demonstrate a robust increase of cell proliferation genes by BKV infection in both RPTE and VEC . We also identified 53 and 97 downregulated genes in RPTE1 and LVEC2 , respectively , at the later time points . Analysis using DAVID showed no significant enrichment for functional clusters . In addition to the cell proliferation genes commonly upregulated in both cell types , functional classification analysis revealed a group of genes involved in type I IFN signaling induced specifically in LVEC2 ( Fig 8A ) . Genes in this cluster include many well characterized ISGs ( interferon stimulated genes ) , such as antiviral effectors [MX1 , RSAD2 ( Viperin ) , BST2 ( tetherin ) , ISG15 , HERC5 , EIF2AK2 ( PKR ) , OAS1 , 2 , 3 and OASL] and pathogen sensors [DDX60 , DDX58 and MB21D1 ( cGAS ) ] , as well as positive regulators of IFN signaling [STAT1 and 2 , and IRF 7 and 9] . Additional genes involved in antigen presentation , including HLA-F , MICA and MICB , PSMB9 ( LMP2 ) , TAP1 and 2 , were also upregulated in BKV inoculated LVEC2 . Elevated levels of MICB and cGAS were also seen in BKV inoculated RPTE1 but the majority of ISGs were not induced in this cell type ( Fig 8A ) . Using RT-qPCR , we monitored the induction of two ISGs in BVEC , IFI44 and OASL , during a 2-week time course . Levels of both transcripts increased dramatically in BKV infected BVEC between 2 and 3dpi ( Fig 8B , middle panel ) . Expression levels of both ISGs appeared to peak at 3dpi , remained elevated until 6dpi and then gradually decreased . Similar trends were observed for ISG56 and OAS1 although the level of induction was less robust . The level of IFI44 was maintained at higher than basal level until about 8dpi . The kinetics of IFI44 and OASL upregulation observed in this experiment are similar to the pattern of LT expression ( Fig 8B , top panel ) , which rose dramatically between 1 and 6dpi and then decreased . A slight upregulation of IFI44 was observed in mock at 5dpi ( Fig 8B , bottom panel ) , but the increase was considerably lower than the levels seen in BKV infected samples . Our results establish that BKV infection in VEC , but not in RPTE , leads to a robust and lasting activation of ISGs . The observed induction of multiple ISGs led us to hypothesize that BKV infection of VEC results in production of IFN that triggers activation of IRF3 and assembly/nuclear translocation of ISGF3 , which subsequently leads to induction of ISGs . We thus tested the levels of various cytokines in culture supernatants from mock and BKV infected RPTE and LVEC at various timepoints . Low levels of IFNβ were found in supernatants from BKV infected LVEC at 3 and 5 dpi ( Fig 9A ) . IFNβ was not detectable at early time points in BKV infected LVEC nor in BKV infected RPTE up to 7dpi . In addition , we identified dramatic increases of the chemokine CXCL10 in BKV infected LVEC and BVEC but not in RPTE using a Luminex assay ( Fig 9B ) . We did not detect production of IFNα or γ . We then examined two key events in the IFN signaling pathway , nuclear translocation of IRF3 and phosphorylation/nuclear translocation of STAT1 using IF . Co-staining with VP1 and IRF3 antibodies showed the scattered presence of nuclear IRF3 in BKV inoculated but not mock LVEC3 ( Fig 10 ) . However , not all VP1 positive cells showed nuclear IRF3 staining . Quantification indicated 54% of nuclear IRF3 positive cells co-stained with VP1 and 27% of VP1 positive cells co-stained with nuclear IRF3 . While low background staining of IRF3 was observed in mock , the signal was excluded from the nucleus ( Fig 10A , top panel ) . Similarly , no nuclear staining of IRF3 was observed in BKV inoculated RPTE1 ( Fig 10B ) . We then examined the presence of activated STAT1 with STAT1-Y701 ( STAT1-phospho-tyrosine 701 ) antibody . While RPTE cells were able to respond to IFNβ treatment by translocating STAT1 to the nucleus ( S6 Fig ) , BKV inoculated RPTE1 showed no signs of STAT1-Y701 ( Fig 11A ) . In contrast , we observed nuclear translocation of phosphorylated STAT1 in all BKV inoculated LVEC3 , independently of VP1 expression ( Fig 11B ) , suggesting that paracrine signaling may be responsible for the uniform activation of STAT1 . These results are consistent with the presence of secreted IFNb in the supernatant of BKV-inoculated LVEC ( Fig 9A ) and further support activation of an antiviral response in BKV infected VEC . Most BKV productive infections are observed in the kidneys of immunosuppressed patients . The fact that healthy individuals occasionally secrete BKV particles in urine also suggests that the virus infects cells of the urinary tract . However , the route of initial BKV infection is unknown and little is known about how cells from other organs respond to BKV exposure . To assess the response of different human cells to this virus , we examined the responses of nine different primary human cell types to BKV inoculation . These included epithelial cells , endothelial cells and fibroblasts from the lung , kidney , bladder and ureter . We found that BKV is capable of infecting all of these cell types as assessed by TAg and VP1 expression . Furthermore , we confirmed that progeny virions are produced in RPTE and VEC cultures . Thus , BKV has the ability to infect a wide range of cell types in both the urinary and respiratory systems . This broad tissue/cell type tropism may facilitate dissemination of the virus within the host body and allow BKV to reach specific sites , such as the kidney , to establish a lifelong persistent infection . All epithelial and fibroblasts cells we tested underwent cell death within two weeks post-infection ( Fig 1 ) . In contrast , VECs from the lung and bladder survived for at least eight weeks after the initial BKV inoculation ( Fig 6 and S2 Fig ) . Since uninfected RPTEs can be maintained for up to 2 months in culture without apparent morphological changes , persistence of BKV infection in VECs is not likely due to intrinsic higher stability and/or greater longevity of primary endothelial cells in culture . In addition , infection of VECs was not severely delayed , since we could detect viral DNA replication and progeny virus production as early as 3 days post-infection . Furthermore , we found that the expression of viral proteins and a low-level production of infectious progeny virions were maintained through several passages of primary VEC cultures . These results are consistent with BKV establishing a persistent infection in endothelial cells . To better understand the underlying causes of the distinct responses by VEC and RPTE to BKV infection , we compared changes in cellular transcriptomics induced by BKV infection in the two cell types by RNA-seq . We observed common upregulation of cell proliferation genes in infected VEC and RPTE . These results concur with the well-established mechanism of LTs of polyomaviruses , which block the cellular retinoblastoma proteins to drive resting cells into the cell cycle and to support viral DNA replication [25–28] . Our results , both indicating upregulation of cell proliferation genes in BKV infected cells and showing little evidence for an active interferon response in BKV-infected RPTE , are in agreement with two previous gene expression profile studies using BKV infected RPTE [20 , 23] . In contrast to RPTE , VEC mounted a robust antiviral response to BKV infection resulting in the upregulation of many ISGs and other immune related genes . Interestingly , our results differ from a previous microarray analysis using total RNA collected from BKV inoculated HUVEC at 40 hpi , in which the authors observed upregulation of just two ISGs [29] . The early timing for RNA collection is likely responsible for the lack of ISG induction in this previous study . Indeed , the robust ISG induction shown by our results was observed at 3dpi ( Fig 8A ) but not 2dpi . Results of RT-qPCR analysis on IFI44 and OASL induction clearly showed that the RNA levels of both ISGs increased dramatically between 2 and 3dpi ( Fig 8B ) . Moreover , IFNβ was not detected until 3dpi in BKV infected LVEC ( Fig 9 ) . Nevertheless , heterogeneity between HUVEC and microvascular endothelial cells has been recognized [30–32] and could contribute to the different outcomes of viral infection . A side by side comparison between the two cell types will accurately determine whether they respond to BKV infection differently . The production of ISGs in response to viral infection is initiated by the production of type I interferon . We detected IFNβ in the media of BKV-infected VECs by ELISA , while uninfected VECs or BKV-infected RPTE produced little or no interferon ( Fig 9 ) . Consistent with these results , we detected phosphorylated STAT1 and nuclear IRF3 in BKV-infected VECs while , in infected RPTE , STAT1 remained unphosphorylated and IRF3 remained cytoplasmic ( Figs 10 and 11 ) . These results indicate that interferon pathway activation by BKV is cell-type specific , as endothelial cells from two different organs and from multiple donors exhibited an interferon response , while renal proximal tubule epithelial cells failed to do so . Several studies have shown that endothelial cells are important for innate immune signaling . For instance , endothelial cells were identified as the master regulators of cytokine storm during infection with influenza virus [33]; and human cytomegalovirus infection led to a robust type I IFN response dependent on the cGAS-STING signaling in primary human endothelial cells [34] . We also detected secretion of CXCL10 by BKV inoculated VECs from both the lung and the bladder . CXCL10 is a pluripotent chemokine derived from endothelial cells . Through binding to its receptor CXCR3 , CXCL10 mediates the recruitment and activation of monocytes , T cells and natural killer cells , and therefore plays critical roles in defense against infectious diseases [35] . Multiple reports have associated CXCL10 to nephropathy and suggested it as a marker for diagnosis and prognosis of the disease [36–40] . Our findings that elevated CXCL10 levels are present in BKV inoculated VEC implicates the potential of these cells as important sources of CXCL10 during pathogenesis of BKV nephropathy . It remains unclear at this point why VECs mount an antiviral response to BKV infection while RPTEs do not . Several possibilities exist regarding the mechanisms involved in their different responses . In order for cells to mount an innate immune response against a viral infection , the entry and/or replication of the virus has to be sensed . Therefore , it is possible that RPTEs lack the machinery for detection of BKV infection and thus no response is triggered . Alternatively , viruses have evolved to counter cellular defense through various ways either by blocking the activities of antiviral effectors or by targeting them for degradation . The different responses of RPTEs and VECs demonstrated in this work may suggest that BKV has the ability to circumvent any potential innate immune responses mounted by RPTEs but not that by VECs . Further investigation testing the above hypotheses will facilitate our understanding of the interactions between BKV infection and host cells at the mechanistic level . Based on the organization of their NCCR sequences , where the replication origin and the early and late promoters reside , polyomavirus BKV can be classified into two main groups , the archetype and the rearranged . The archetype BKV is found commonly among healthy individuals but very difficult to propagate in vitro , whereas the rearranged BKV is predominantly associated with diseases and can be grown efficiently using cell culture systems [1 , 41–43] . The mechanisms underlying NCCR rearrangement and how this rearrangement contributes to BKV pathogenesis are not well understood . To date , most BKV studies have focused on the rearranged type . The Dunlop strain used in this study also belongs to the rearranged BKV group . Recent progress in methods for increasing yield of viral stock have made investigation of the archetype virus more feasible [21 , 44] , and thus provide the opportunity for future experiments aiming at addressing the consequences of archetype BKV infection in endothelial cells , particularly regarding whether that infection also leads to activation of IFN response . In conclusion , we have demonstrated that BKV infection elicits a robust IFN response in human microvascular endothelial cells , suggesting that these cells may play essential roles in triggering and mediating antiviral defense against BKV . Furthermore , we found that BKV infection is limited , but not abolished , in endothelial cells , raising the possibility that viral persistence is maintained in this cell-type , while productive infections linked to pathology occurs mainly in epithelial cells . Further studies will be required to establish the molecular basis for this cell type specific response to BKV infection and the role it plays in disease . Human primary cells were purchased from commercial sources and cultured following the manufacturers’ instructions . Details about cell type , supplier , donor information and number of donors tested , as well as specific media used are listed in S1 Table . Although the cells came from apparently healthy human donors , they were not tested thoroughly for presence of viruses , except for the few highly pathological agents monitored by the supplier . We thus took the effort to identify the presence of potential adventitious agents in the RPTE1 and LVEC2 used for our RNAseq analysis , and found no evidence indicating the existence of such agents . The initial viral stock of BKV was provided by Dr . Michael Imperiale . Inoculation with BKV for infection was carried out following protocols previously described [19] . Briefly , BKV viral stock was diluted in growth media and added to cells precooled to 4°C . Cells were incubated for 1 hour at 4°C with gentle rocking every 15 to 20 minutes . The media containing viral stock were removed after the incubation . Cells were washed once with 1X PBS , supplemented with fresh growth media and then returned to the CO2 incubator . To propagate the virus , RPTE were inoculated at a MOI of 0 . 1 . Most cells died by 12 to 14dpi . The supernatant was collected and frozen at -20°C until processed . For preparation of the viral stock , the supernatant was first cleared by centrifugation at 8000g for 30 minutes . The cleared supernatant was layered on top of a 20% sucrose solution and centrifuged at 25 , 000rpm for 3 hours at 4°C . The virus pellet was resuspended in Buffer A ( 10mM HEPES , pH 8 , 1mM CaCl2 , 1mM MgCl2 and 5mM KCl ) , followed by centrifugation at 13 , 000rpm for 5 minutes at 4°C to remove any particulate debris . This viral stock was then aliquoted and stored at -20°C . We initially prepared viral stock by collecting both the supernatant and cell debris , which was then treated by neuraminidase and detergent following established protocols [19] . These additional steps for extracting viruses from cell debris did not result in any dramatic increase of the virus yield in RPTE cells . We therefore switched to preparing viral stock using only the infection supernatant . Titration of virus was performed by inoculating RPTE1 cells with serial dilutions of either viral stocks or infection supernatants as described above . Cells were fixed at 2dpi and stained with an antibody against LT . Quantification of % cells infected was performed via ImageJ by determination of % LT positive cell within the entire population of cells visualized by DAPI staining . A titer was calculated using the equation below: Titer ( FFU/ml ) =Numberofcells%CellsinfectedDilutionfactorVolume To monitor cell death in RPTE and LVEC , the inoculation was set up in 12-well plates . For each data point , the average cell number and standard deviation were calculated using cell counts from 3 separate wells . Recombinant human IFNβ ( R&D Systems ) was used to treat RPTE at the concentration of 100IU/ml for 3 hours before fixation . The following primary antibodies were used in this study , Pab416 ( LT antibody , mouse monoclonal Ab , 1:100 ) [45] , GaT ( Goat anti-full-length LT protein polyclonal Ab that reacts with both LT and sT , 1:400 ) , anti-VP1 ( Abnova , mouse monoclonal Ab , 1:400 ) , rabbit anti STAT Y701 ( Cell Signaling , 1:400 ) and rabbit anti IRF3 ( Cell Signaling , 1:400 ) . We used the following secondary antibodies , goat anti mouse FITC conjugated ( Sigma , 1:200 ) , goat anti mouse Alexa Fluor 488 or 568 ( Thermo Fisher Scientific/Invitrogen , 1:400 ) , or goat anti rabbit Alexa Fluor 488 ( Thermo Fisher Scientific/Invitrogen , 1:400 ) , chicken anti goat FITC ( Thermo Fisher Scientific ) . Incubation of primary antibodies was performed at 4°C for overnight . Incubation of cells fixed in coverslips with all secondary antibodies was performed for 1 hour at room temperature . All epithelial and endothelial cells were examined by staining with Pab416 and anti-VP1 separately , except for fibroblasts from lung and bladder that were co-stained with GaT and anti-VP1 . Except for the incubation with primary antibodies , all other steps in immunofluorescence were carried out at room temperature . Cells on coverslips were fixed in 4% formaldehyde ( Sigma ) for 20 minutes and washed three times with 1x PBS . For single staining with Pab416 or anti-VP1 , cells were permeabilized with 0 . 1% TritonX-100 for 5 minutes and washed three times with 1x PBS . Coverslips were then blocked with 5% goat serum in 1XPBS for 1 hour . Secondary antibody for these coverslips were goat anti mouse FITC . For co-staining of GaT and anti-VP1 , coverslips were also permeabilized in 0 . 1% TritonX-100 , then blocked with 5% normal chicken serum and 5% new born calf serum . The secondary antibodies for GaT and anti-VP1 co-staining were chicken anti goat FITC and chicken anti mouse TRITC . Co-staining using IRF3 or STAT1-Y701 were performed following instructions from Cell Signaling Technology . The coverslips were washed 3 to 5 times using 1X PBS after primary and secondary antibody incubation . After the final wash , the coverslips were dried and mounted with ProLong Gold Antifade reagent with DAPI ( Thermo Fisher Scientific/Invitrogen ) and imaged using either a Zeiss AX10 microscope with LED camera and a set of fluorescence filters , or a Leica TCS SP5 confocal/multi-photon imaging system . To quantify % LT positive RPTE and VEC at 2 and 5dpi , the inoculation was set up using cells on coverslips in 12-well plates . For each data point , the average cell number was calculated using cell counts from 3 separate coverslips . The QIAprep Spin Miniprep kit ( QIAGEN ) was used to isolate BKV genomic DNA . The mock and BKV inoculated cells ( from 1 well of a 6-well plate/condition ) were harvested in 250μl of buffer P1 and then processed following manufacturer’s instructions . The circular BKV genomic DNA harbors a single BamHI recognition site and was linearized by BamHI restriction digestion . BamHI digested pBKV , a plasmid containing the full-length BKV genomic DNA ( Genbank accession number , KP412983 ) , was used as positive control . The sequence of pBKV was verified by Sanger sequencing and MiSeq . The digested DNA was resolved on 0 . 8% agarose gel . Total RNA from whole cells growing in monolayers was first extracted with TRIzol ( Ambion ) , and then purified with Direct-Zol columns ( Zymo Research ) according to the manufacturer’s instructions . Cytoplasmic and nuclear RNA was isolated using the PARIS kit ( Thermo Fisher Scientific ) following manufacturer’s instructions . Isolation of polysome associated RNA was based on previously published methods [46] and suggestions from J . Woolford’s laboratory ( Carnegie Mellon University , Pittsburgh PA ) . Control or BKV inoculated cells ( MOI = 4 ) were allowed to grow in 10 cm diameter dishes to 2dpi . Ten plates of each condition were used for each extraction . After removing the culture media , ribosomes were frozen on mRNA by adding to each plate 1ml of ice-cold lysis buffer ( Tris HCl 100mM pH 7 . 5; NaCl 100mM; Mg2Cl 10mM; DTT 5mM; NP40 0 . 5% ) containing 100 μg/ml of cycloheximide , and incubating the cells on ice for 5 min . Cells were collected with a scraper and the extracts were kept on ice for the entire duration of the procedure . After adding RNase-free DNase ( 2 μg/ml ) and incubating the mixture at 4°C for 20 minutes , the extracts were centrifugated at 14K , 4° , 5 min to remove debris . Total RNA from a portion of each supernatant ( 0 . 5 ml ) was obtained with TRIzol and Direct-Zol columns ( Zymo Research ) , following the manufacturer’s instructions . The rest of each supernatant was poured over 12 . 5 ml of a sucrose cushion ( 60% sucrose in lysis buffer ) in Ti70 ultracentrifuge tubes , the tubes were filled with lysis buffer , and samples were centrifuged for 25 hours at 27 , 000 rpm , 4°C in a Ti70 rotor ( Beckman Coulter ) . The pellets from centrifugation , containing both ribosomes and polysomes , were resuspended in ( ~0 . 2ml ) of lysis buffer and RNA was immediately extracted with TRIzol and Direct-Zol columns as mentioned above . RT-qPCR analysis was performed at the Genomics Research Core of the University of Pittsburgh with the Power SYBR Green RNA-to-CT 1-Step Kit ( Thermo Fisher ) . All samples were analyzed in triplicate and the values were normalized against those of the corresponding endogenous control ( GAPDH ) . Relative levels indicated in the graphs refer to each specific value relative to the signal observed in mock at 6 hours . The following sets of primers were used to detect the presence of: BKV LT ( 5’- GAGTAGCTCAGAGGTGCCAACC and 5’-CATCACTGGCAAACATATCTTCATGGC ) ; BKV sT ( 5’- GATCTAAAGCTTTAAGGTGCCAACCTATGG and 5’-CATCACTGGCAAACATATCTTCATGGC ) ; GAPDH ( 5’-AAGGTGAAGGTCGGAGTCAA and 5’-AATGAAGGGGTCATTGATGG ) ; IFI44 ( 5’-TGCAGAGAGGATGAGAATATC and 5’- ACTAAAGTGGATGATTGCAG ) ; ISG56 ( 5’-CAACCAAGCAAATGTGAGGA and 5’- AGGGGAAGCAAAGAAAATGG ) ; OAS1 ( 5’-ATAAAAGCAAACAGGTCTGG and 5’-TCTGGCAAGAGATAGTCTTC ) and OASL ( 5’-AGGGTACAGATGGGACATCG and 5’—AAGGGTTCACGATGAGGTTG ) . Paired-end ( 2x 75bp ) RNA-seq strand specific libraries were constructed using enriched polyadenylated RNA by the Genomics Research Core facility at the University of Pittsburgh . The raw reads were analyzed by FastQC ( Andrews S . ( 2010 ) . FastQC: a quality control tool for high throughput sequence data . Available online at: http://www . bioinformatics . babraham . ac . uk/projects/fastqc ) . No quality filtering was necessary based on the FastQC analysis . Gene expression values in TPM were determined with CLC Genomics Workbench 11 using the Ensembl GRCh38 human genome [47] . To determine cellular gene upregulation by BKV infection in the RNAseq data we used the following definition: the expression level of a gene had to be greater than or equal to 10 TPM in a BKV infected sample , and the fold change of BKV infected over mock had to be greater than or equal to 2 ( Log2 BKV ( TPM ) / Mock ( TPM ) ≥ 1 ) . For downregulation , the expression level of a gene had to be greater than or equal to 10 TPM in a mock infected sample , and the fold change of BKV infected over mock had to be less than or equal to 0 . 5 ( Log2 BK ( TPM ) / Mock ( TPM ) ≤ -1 ) . Correlation analysis of gene expression and upregulation was done using the CORREL function in Microsoft Excel . Functional classification was performed using DAVID Bioinformatics Resources 6 . 8 [48 , 49] ( https://david . ncifcrf . gov/home . jsp ) . Heatmaps were generated using the Heatmapper server [50] ( http://www . heatmapper . ca/ ) . To test IFN and additional cytokine production , culture supernatants from mock and BKV infected RPTE and VEC at various time points were collected and stored at -80°C until processing . IFNβ levels were determined with the Human IFN-beta DuoSet ELISA kit ( R&D Systems ) following manufacturer’s instructions . A 38-plex human cytokine Luminex panel ( HCYTMAG-60K-PX38 , Millipore Sigma ) was used to assay secretion of 38 additional cytokines . The collected supernatants were submitted to the Hillman Cancer Center ( UHCC ) Luminex Core Laboratory affiliated with the University of Pittsburgh Medical Center for the procedures .	Infection by polyomavirus BKV is common and mostly harmless in healthy populations but can cause severe damages to kidney and bladder in transplant recipients . The infection by BKV usually occurs in early childhood and persists chronically in the urinary system throughout life . Our data show that this virus has the ability to infect multiple types of human cells along the respiratory and urinary tracts . Furthermore , the infection elicits an immune response in endothelial cells , the type of cells that line the inner surface of the blood vessels . These results provide insights into the distinct cellular responses displayed by different cell types that BKV encounters during infection and spread of the virus within the body , and on innate immune responses against the infection .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "medicine", "and", "health", "sciences", "respiratory", "infections", "endothelial", "cells", "immunology", "bladder", "cell", "processes", "pulmonology", "epithelial", "cells", "cell", "proliferation", "animal", "cells", "proteins", "gene", "expression", "biological", ...	2019	Human polyomavirus BKV infection of endothelial cells results in interferon pathway induction and persistence
The viral family Arenaviridae includes a number of viruses that can cause hemorrhagic fever in humans . Arenavirus infection often involves multiple organs and can lead to capillary instability , impaired hemostasis , and death . Preclinical testing for development of antiviral or therapeutics is in part hampered due to a lack of an immunologically well-defined rodent model that exhibits similar acute hemorrhagic illness or sequelae compared to the human disease . We have identified the FVB mouse strain , which succumbs to a hemorrhagic fever-like illness when infected with lymphocytic choriomeningitis virus ( LCMV ) . FVB mice infected with LCMV demonstrate high mortality associated with thrombocytopenia , hepatocellular and splenic necrosis , and cutaneous hemorrhage . Investigation of inflammatory mediators revealed increased IFN-γ , IL-6 and IL-17 , along with increased chemokine production , at early times after LCMV infection , which suggests that a viral-induced host immune response is the cause of the pathology . Depletion of T cells at time of infection prevented mortality in all treated animals . Antisense-targeted reduction of IL-17 cytokine responsiveness provided significant protection from hemorrhagic pathology . F1 mice derived from FVB×C57BL/6 mating exhibit disease signs and mortality concomitant with the FVB challenged mice , extending this model to more widely available immunological tools . This report offers a novel animal model for arenavirus research and pre-clinical therapeutic testing . Viral hemorrhagic fevers ( VHFs ) are induced by viruses that belong to one of four families , Arenaviridae , Bunyaviridae , Filoviridae , Flaviviridae . The clinical symptoms of hemorrhagic fever vary depending on the severity and etiological agent but generally fever and bleeding are prominent manifestations of the disease . Hemorrhagic fever viruses , including arenaviruses , pose a significant public health threat both as emerging infectious diseases and as potential bioterrorism agents [1] . The majority of viruses in the Arenaviridae family require maximum biosafety containment ( BSL-4 ) for handling which limits access for most researchers . In addition , the available animal models that induce hemorrhagic fever like symptoms require marmosets , hamsters , guinea pigs , primates , or immunocompromised mice [2] , [3] . The lack of a non-immunocompromised mouse model for viral hemorrhagic fever makes it difficult to conduct pre-clinical drug screening . Mice are ideal for use in pre-clinical drug development because of their low cost and the extensive knowledge and reagents available for the species . There is a dire need for VHF therapeutic as there is no FDA approved drug available for hemorrhagic fever disease . Ideally , the clinical course and signs produced in the animal model will parallel those observed in the human disease . The key characteristics of human viral hemorrhagic fever of arenaviral origin are multiorgan infections with hepatocellular necrosis and thrombocytopenia [2] , [4] , [5] . In this article , we report that the FVB strain of mice exhibits extreme susceptibility to hemorrhagic fever-like signs after LCMV-Clone 13 ( LCMV-13 ) infection . FVB mice demonstrate thrombocytopenia , hepatocellular necrosis , petechiae , and death . This is in contrast to the C57BL/6 mouse strain's response to LCMV-13 , which progresses to a chronic wasting disease [6] . FVB mice showed greatly increased ( IL-6 , IL-17 and IFN-γ ) cytokine and ( CXCL1 and MCP-1 and 3 ) chemokine production profiles early after infection compared to C57BL/6 mice and systemic TNF-α during the hemorrhagic phase of the disease . To investigate the underlying mechanism of the FVB pathology , separate groups of mice were either depleted of CD4+ or CD8+cells at time of infection . We found that mice deficient in either CD4+ or CD8+ cells maintained normal liver function and survived LCMV-13 infection . Furthermore , drawing from data produced in previous mouse HFV challenge experiments ( e . g . Ebola or Marburg ) we chose to examine the role of IL-17 responses in this arenavirus model . FVB mice treated to block IL-17 responses at the time of infection exhibited increased survival from LCMV-13 challenge compared to untreated mice . These data , to the best of our knowledge , describe the first mouse model of arenavirus induced hemorrhagic fever and support the possibility that T cell-mediated immunopathology plays a role in the underlying cause of HFV disease . We initially sought to examine the anti-arenavirus activity of a modified PMO chemistry ( PMOplus ) that is complimentary to a sequence conserved within numerous arenavirus isolates , termed AVI-7012 . The PMOplus chemistry has been shown to be efficacious in NHP filovirus lethal challenge models [7] . Conservation of the targeted sequence aligned to the L and S genomic segment and anti-genomic RNA is shown in Table 1 . A similar targeting strategy demonstrated inhibition of arenavirus replication in vitro and in vivo utilizing a PMO conjugated to an arginine-rich peptide to enhance cellular uptake [8] . AVI-7012 ( 4 mg/kg ) administered to C57BL/6 mice either prior to or following infection with LCMV-13 exhibited significant antiviral activity compared to PBS or scramble control treated mice when viral RNA was measured in kidney , spleen , brain and liver tissue ( Figure 1 ) . In pursuit of examining the uptake of PMOplus chemistry in vivo by cells that are known to support LCMV replication we employed a transgenic mouse model , which expresses enhanced green fluorescent protein ( EGFP ) as a positive readout of antisense activity via corrective splicing of the EGFP open reading frame [9] . As is common for many transgenic models this was produced on the FVB background strain of mouse . Four days post-infection , initial observations showed the infected EGFP mice react in an atypical manner compared to C57BL/6 mice at this point in a LCMV-13 infection . EGFP mice exhibited ruffled fur and showed signs of lethargy . Moreover , overt signs of severe disease were apparent 2–4 days later with some but not all of the infected EGFP animals presenting the following signs: mucosal , cutaneous and organ hemorrhaging ( Figure 2 ) and decreased blood pressure , discordination , unresponsiveness , hypothermia , and seizures . The anomalous results observed following this LCMV-13 infection prompted us to carry out a second study without antisense treatment and with standard strains of C57BL/6 and FVB mice using the same virus stock in order to determine if the disease manifestation was due to the EGFP transgene . FVB mice displayed disease signs similar to those of the EGFP mice and on day 6 post-infection , death ensued and by day 8 post-infection only 1 out of 8 FVB mice infected with LCMV-13 had survived ( Figure 3a ) . The FVB survivor did not clear virus , but instead harbored a long-term chronic infection with high viral load detected in multiple organs as late as day 36 post-infection . In agreement with previous reports , C57BL/6 mice showed 100% survival after LCMV-13 infection ( Figure 3a ) . However , infection with high dose of the less pathogenic Armstrong strain of LCMV did not lead to any signs of hemorrhagic disease ( Figure 3b ) . Weight loss in LCMV-13 infected FVB mice commenced after day 3 post-infection and FVB mice plateaued in weight loss at day 6 post-infection at 10 . 2% net weight loss ( Figure 3c ) . The body weight pattern of C57BL/6 mice was strikingly different with a sharp drop in weight of 9 . 8% by day 2 post-infection , then a rebound in weight followed by a dramatic weight loss of 19 . 8% by day 9 post-infection ( Figure 3c ) . FVB mice displayed a pantropic infection with virus being detected in multiple organs . However , even though FVB mice developed a moribund state while C57BL/6 mice did not , viral load was comparable or less in spleens , lungs , kidneys , and livers of FVB compared to C57BL/6 mice ( Figure 3d–e ) . Yet C57BL/6 mice show no clinical symptoms of hemorrhagic-like disease , which indicates the pathology associated with LCMV infection in FVB mice is not solely virus mediated . We further examined the relationship of disease to viral load by targeting virus replication with antisense to Arenavirus 5′ termini , which has been shown to inhibit LCMV replication both in vivo and in vitro ( Figure 1; [8] ) . Figure 1 shows that AVI-7012 can inhibit viral load in C57BL/6 mice , however , when FVB infected mice were treated to reduce viral load they did not exhibit a concomitant increase in survival ( Figure 3f ) Considering the clear signs of hemorrhage in the FVB mice following infection we next sought to assess the hematologic parameters in FVB and C57BL/6 mice infected with LCMV-13 . Similarly to the clinical manifestation to most arenaviral hemorrhagic diseases , platelet count differences were the most striking . FVB mice showed reduced platelet counts with a range of 146–324 K/ml compared to platelet counts of 1318 K/ml and 1216 K/ml for LCMV-13 infected C57BL/6 and naïve FVB , respectively ( Figure 4a ) . Lymphocyte and granulocyte counts were dramatically increased in blood of LCMV-13 infected mice compared to naïve FVB and infected C57BL/6 mice ( Figure 4a ) . However , spleen size was significantly smaller in LCMV-13 infected FVB mice , independent of gender , compared to infected C57BL/6 mice as well as total T cell counts in the spleen ( Figure 4b and c ) . No mortality was observed in either gender at the challenge inoculum of 104 p . f . u . . In accordance with the smaller spleens was the histopathological finding of severe splenic necrosis in LCMV-13 infected FVB mice while infected non-diseased FVB mice displayed no detectable necrosis ( Figure 5a upper panels ) . Liver pathology in infected diseased mice revealed many single cells undergoing necrosis/apoptosis and randomly scattered zones of parenchymal necrosis with associated degenerate neutrophils ( Figure 5a middle panels ) . Histological signs of modest alveolar edema and/or atelectasis were observed in the lungs of the infected FVB diseased mice but were absent in the infected non-diseased mice . Tissues from infected C57BL/6 mice showed no such disease indication except the livers of some C57BL/6 mice displayed rare tiny foci of hepatocellular degeneration and necrosis with scattered neutrophils ( data not shown ) . The following clinical biochemistry parameters were analyzed: alkaline phosphatase ( ALK ) , alanine aminotransferase ( ALT ) , aspartate aminotransferase ( AST ) , calcium , cholesterol , triglycerides , albumin , creatinine , glucose , phosphorous , total bilirubin ( TBIL ) , blood urea nitrogen ( BUN ) , and total protein . LCMV-13 infected FVB mice had increased levels of ALT , AST , TBIL , and BUN compared to control mice indicating severe kidney dysfunction and hepatocyte destruction ( Figure 5b ) . Calcium , cholesterol , triglycerides , albumin , creatinine , glucose , total protein , alkaline phosphatase , and phosphorous were normal in LCMV-13 infected FVB mice compared to controls ( data not shown ) . These data combined demonstrate a pantropic infection leading to thrombocytopenia , cutaneous hemorrhaging , hepatic dysfunction , and ultimate death . LCMV induces a well-characterized immunoregulatory state in most immunocompetent inbred mouse strains , including C57BL/6 mice [10] with subclinical disease signs . That FVB mice progress to hemorrhagic state and succumb implies a role of the immune response in the manifestation of the FVB hemorrhagic disease . To gain some insight into the inflammatory response prior to onset of hemorrhagic fever symptoms , we assayed for systemic cytokines in LCMV-13 infected FVB mice bled at day 3 post-infection and found increased levels of multiple pro-inflammatory cytokines and chemokines ( Figure 6a ) . LCMV-13 infected FVB mice showed increased levels of IL-6 compared to C57BL/6 mice ( 2152 pg/ml versus 187 pg/ml ) , IFN-γ ( 2184 pg/ml versus 6399 pg/ml ) , CXCL1 ( 2530 pg/ml versus 154 pg/ml ) , MCP-1 ( 12342 pg/ml versus 6413 pg/ml ) , and MCP-3 ( 6284 pg/ml versus 1589 pg/ml ) . Strikingly , on day 1 post-infection FVB mice exhibited systemic IL-17A ( 78 pg/ml ) while levels remained undetectable in C57BL/6 mice . At times during severe disease in FVB mice ( day 6–8 post-infection ) , increased levels of systemic TNF-α were found compared to C57BL/6 mice ( Figure 6b ) . TNF-α levels in LCMV-13 infected FVB mice were between 61–92 pg/ml whereas C57BL/6 mice had undetectable levels of TNF-α . In order to investigate the immune component of the FVB-related hemorrhagic disease further , we depleted either CD4+ or CD8+ cells in FVB mice with anti-CD8 or anti–CD4 . Mice treated with anti-CD4 or anti-CD8 antibody at time of infection and 1 day after demonstrated 100% survival up to day 16 post-infection compared to 0% survival for PBS treated FVB mice ( Figure 7a ) . While peak weight loss for anti-CD8 treated mice was similar to PBS treated mice ( 17 . 5+/−0 . 9% for anti-CD8 treated versus 14 . 6+/−0 . 2% for PBS treated ) , anti-CD8 treated mice regained weight after day 10 post-infection and plateaued at ∼10% lost body weight ( Figure 7b ) . While peak weight loss was not as severe in anti-CD4 treated mice ( 13 . 5+/−5 . 8% ) , a similar trend was seen between anti-CD4 and anti-CD8 treated mice in that they began to regain weight around the same time PBS treated mice succumbed to disease . Anti-CD8 treated mice had lower AST ( 2245 u/ml vs 390 u/ml for PBS and anti-CD8 treated mice , respectively ) and ALT ( 1584 u/ml vs 364 u/ml for PBS and anti-CD8 treated mice , respectively ) readings than PBS treated mice indicating increased liver function ( Figure 7c ) . Viral load in liver was similar between anti-CD8 treated and untreated mice ( Figure 7d ) , which again suggests viral replication alone is not causing hemorrhagic disease . Taken together , hemorrhagic disease in this model appears to be caused by a skewed immune response . IL-17 has been shown to play a role in tissue destruction and we identified increased levels of systemic IL-17 early after infection ( Figure 6 ) . Previous data from our lab has shown antisense ablation of IL-17 receptor C ( IL17RC ) can prevent mortality in a mouse model of Ebola hemorrhagic disease . We , therefore , probed the role of IL-17 in arenavirus hemorrhagic disease . A delivery peptide conjugated antisense phosphorodiamidate morpholino oligomer ( PPMO ) was designed to target the splice-donor site of exon 12 of IL17RC ( IL17RC SD12 ) , thereby disrupting the translational reading frame of IL17RC , leading to reduced levels of surface IL17RC . Mice were treated with 7 . 5 mg/kg of IL17RC SD12 at day 0 , day 1 , and day 2 post-infection and monitored for disease symptoms . While only 12 . 5% of animals survived when treated with PBS , 66 . 7% of animals treated with IL17RC SD12 were still living at day 8 post-infection ( Figure 8a ) . Viral load was significantly reduced in liver , lung , kidney , and brain with IL17RC SD12 treatment compared to untreated FVB ( Figure 8b ) . Furthermore , IL17RC treatment protected liver and kidney function as seen by the reduced levels of ALT , AST , total bilirubin , and alkaline phosphatase found in serum ( Figure 8c ) . Combined , these data suggest a role for IL-17 in mediating arenaviral hemorrhagic induced disease in FVB mice . The FVB strain is useful for production of transgenic mice on inbred genetic backgrounds due to robust fecundity , and fertilized eggs contain large and prominant pronuclei facilitating microinjection of [11] . However , few immunological tools are available to probe the factors influencing hemorrhagic disease on the FVB H-2q background . We , therefore , bred C57BL/6 ( H-2b ) and FVB mice to create F1 hybrids . F1 hybrids were infected with a dose range of LCMV-13 and monitored for hemorrhagic disease symptoms . Mice infected with 2×106 p . f . u . showed 100% mortality by day 9 post-infection , 66% mortality with 6×105 p . f . u . , and no mortality at 2×105 p . f . u . ( Figure 9a ) . We further confirmed hemorrhagic disease in F1 hybrids by infecting mice with either LCMV-13 or LCMV-Armstrong . Day 7 post-infection , C57BL/6 mice infected with LCMV-13 and F1 hybrids infected with LCMV-Armstrong demonstrated no clinical symptoms except for weight loss while F1 hybrids infected with LCMV-13 demonstrated a significant decrease in body temperature . F1 hybrids infected with LCMV-13 had a body temperature of 29 . 5+/−0 . 6 degrees Celsius while F1 hybrids infected with LCMV-Armstrong maintained normal body temperature of 36 . 8+/−0 . 4 degrees Celsius ( Figure 9b ) . We then assessed LCMV-specific T cell responses in F1 hybrids . As can be seen in Figure 9d , F1 hybrids infected with LCMV-13 showed a significant reduction in CD44hi CD8 T cell numbers ( 2 . 4+/−1 . 2×106 cells versus 5 . 8+/−2 . 5×106 cells for LCMV-13 versus Armstrong infected F1 hybrids , respectively ) . LCMV-13 infected F1 hybrids also demonstrated reduced LCMV-specific T cells as assessed by both MHC pentamer staining IFN-γ production ( Figure 9e ) . F1 hybrids infected with LCMV-13 had significantly reduced numbers of Db/NP396–404 specific and Db/GP33–41 specific CD8 T cells as assessed by MHC pentamer staining and IFN-γ production , respectively ( Figures 9e ) . These results confirm the findings from Figure 4 showing that mice undergoing hemorrhagic disease have lower T cell numbers late in infection . One of the requirements for FDA approval of antiviral therapeutics is drug testing in accepted animal models that reproduce human disease as closely as possible . It has been thought that one of the fundamental features of LCMV biology is its ability to establish chronic infections in mice . This is in contrast to the disease course of LCMV in rhesus macaques , which succumb to hemorrhagic fever [12] . It has been reported that under certain circumstances LCMV can lead to mortality in mice . The New Zealand Black strain succumbs to a pulmonary disease much like hantavirus pulmonary syndrome [13] . Likewise , C57BL/6 mice infected with a medium dose of LCMV-13 displayed a similar lung pulmonary edema and interstitial mononuclear infiltration as NZB mice and 23% of those mice died [14] . Two other LCMV infection models have been reported to produce mortality with similar disease signs to the FVB model and a possible link to IL-17 production . The earliest was reported by Sarawar et . al , 1994 . Here a subclinical infection of LCMV followed by low dose i . p . exposure to Staphylococcus aureus Enterotoxin B ( SEB ) resulted in a disease characteristic of hemorrhagic toxic shock leading to significant mortality in Vβ8 . 1 transgenic mice [15] . Although , Sarawar et . al would have not been able to measure IL-17 at the time , it has been shown in later studies that SEB will potently induce IL-17 expression in mice [16] . Additionally , large amounts of IFN-γ and IL-6 , along with a transient increase in TNF-α were detected . Recently it has been shown the inflammatory effect of IL-17 on endothelial activation and neutrophil recruitment acts synergistically with TNF-α [17] . Both of these cytokines were produced in significant amounts in the FVB-LCMV hemorrhagic model and could account for the exaggerated inflammatory response in both models . Moreover , prior depletion of T cells gave similar results whereby the lethal effects of the LCMV infection with SEB were also greatly diminished . Although the source and precise role of IL-17 in the FVB-LCMV hemorrhagic model remains to be determined , anomalous production of IL-17 has been reported for mice deficient in T-bet and eomesodermin when infected with LCMV . These mice fail to differentiate LCMV-specific CD8+ killers T cells , required for defense against the virus , but instead produce a CD8+ IL-17-secreting lineage [18] . Upon viral infection , these mice develop a CD8+ T cell-dependent , progressive inflammatory and wasting syndrome characterized by multi-organ infiltration of neutrophils . There is currently no mouse model that demonstrates multiple symptoms of the human disease of arenaviral hemorrhagic fever [2] . The disease of FVB mice infected with LCMV-13 described in this report mimics LCMV disease in macaques and many of the clinical signs of Argentine hemorrhagic fever ( Table 2 ) . However , the disease progression in FVB departs from the sequelae observed for Lassa Fever ( Table 2 ) . Viral load was not a good predictor of disease in our model while disease outcome is often predicted by viremia level in Lassa fever patients [19] , [20] , [21] . The smallest animal model currently used for arenaviral hemorrhagic fever is the guinea pig [2] . While this is a useful model in some respects , the major limitations of using guinea pigs for therapeutic testing and experimentation is the lack of information and reagents available for guinea pig analysis . Using the FVB inbred mouse strain opens up access to the plethora of tools available for mouse research . In addition , the cost and lower biosafety level ( BSL-2+ ) using mice and LCMV-13 , respectively , allows for high throughput screening . That FVB mice display an acute lethal disease after LCMV-13 infection while C57BL/6 mice develop a chronic infection reinforces the role of the host genetic factors in skewing the arenaviral disease course . Early in infection , we found increased levels of proinflammatory cytokines and chemokines in LCMV-13 infected FVB mice compared to C57BL/6 mice , which points to an immune component in the onset of hemorrhagic fever disease . It has been shown that LCMV-13 infected mice that were depleted of platelets develop lethal hemorrhagic anemia that is dependent on virus induced type I interferons [22] indicating a role for proinflammatory response in hemorrhagic disease . In addition , high levels of IL-6 and IFN-γ were found in rhesus macaques that succumb to lethal disease after LCMV infection [12] . One report has suggested that suppression of pro-inflammatory responses is partly responsible for the terminal shock associated with arenavirus infection in guinea pigs [23] . However , results from this study showed increased IFN-γ and MCP-1 in high pathogenic pichinde virus infected guinea pigs day 2 post-infection , indicating an early , robust proinflammatory response . Similarly , many of the proinflammatory cytokines and chemokines that were upregulated in LCMV-13 infected FVB mice early in infection were below C57BL/6 levels late in infection ( data not shown ) . That either CD4 or CD8 antibody treatment prevents death in all LCMV-13 infected FVB mice also supports an immune mediated component in the development of hemorrhagic disease . We believe that the disease is T cell mediated as waiting 3 days after infection before anti-CD8 treatment continued to protect mice from lethality ( data not shown ) . Presumably LCMV has gone systemic by 3 days post-infection and the protective capability of anti-CD8 is T cell depletion rather than a CD8+ LCMV reservoir . The finding that CD4+ and CD8+ T cells are involved in the pathogenisis of LCMV-13 in FVB mice is in striking contrast to the C57BL/6 model . LCMV-13 induces an exhausted T cell phenotype where numerous inhibitory receptors are upregulated on CD8+ T cells , which lead to a chronic wasting disease [24] . However , LCMV-Armstrong infection in FVB mice mimics the disease progression of C57BL/6 with peak weight loss 8–9 days post-infection and then clearance of virus ( Figures 3 and 9 ) . This suggests that high viremia can skew the immune response . South American arenaviruses and Lassa induce splenic and lymphoid necrosis with varying degrees of lymphoid depletion [25] . While our T cell depletion studies point to a T cell-mediated pathogenesis in the FVB model , many signs in the FVB mice are consistent with human arenavirus immunosuppression such as splenic necrosis ( Figure 5a ) , splenic involution ( Figure 4b ) , and reduced T cell numbers ( Figures 4c and 9 ) . Our model suggests that skewing of the T cell response , possibly Th17 , promotes unchecked inflammation , which then leads to splenic necrosis , lymphoid apoptosis , and lymphopenia . Our results are similar to Flatz et . al where mice that have humanized MHC class I develop severe Lassa fever whereas T cell depletion prevents disease [24] . We propose a three-tiered model similar to Flatz et . al where three outcomes are possible 1 ) potent T cell response controls virus ( LCMV-Armstrong ) , 2 ) intermediate T cell response fails to control virus and triggers severe disease ( FVB mice ) , 3 ) depletion of T cells allows persistence of Lassa virus with mild disease ( T cell-depleted FVB mice ) . The FVB mouse , named for its susceptibility to the B strain of Friend leukemia virus , exhibits a predisposition to several viral induced pathologies compared to other mouse laboratory strains [26] . Some examples are the neurological and immunological sequela observed subsequent to infection with either MOMuLV , a retrovirus , Theiler's a picornovirus and Minute virus , a parvovirus [27] , [28] , [29] . In contrast , the C57BL/6 ( H-2Db ) mouse strain is resistant to disease or persistence following infection with these viruses . In the case of Theiler's virus disease , resistance has been linked to the MHC loci [30] . Specifically it has been shown that the FVB×C57BL/6 F1 and FVB H2-Db transgenic mice are resistance to persistent Theiler virus infection and development of inflammatory lesions . This indicates that the H-2Db allele confers dominance over H-2q . Our observation that the FVB×C57BL/6 F1 does not recapitulate resistance to LCMV infection or onset of hemorrhagic disease suggests that H-2Db is not dominant in this condition of viral induced immunopathology . Although it is yet to be determined what immune related gene ( s ) influence the FVB and F1 mice susceptibility to hemorrhagic disease , a probative advantage prevails with the F1 hybrid . While maintaining the FVB-like hemorrhagic disease the F1 possesses an H2-Db immune system , which will allow further dissection of the mechanism due to the availability of immunological reagents . In summary , we have discovered a unique model for arenaviral hemorrhagic fever that could have broad applicability for arenaviral therapeutic development and arenavirus research . While our model does not mimic all hemorrhagic fever symptoms from viruses in the family Arenaviridae , the development of a hemorrhagic fever mouse model with an intact immune system represents a major advancement for arenavirus research and preclinical testing . In addition , our data suggests an immune mediated component to the onset of arenavirus hemorrhagic fever . If there is an immune component to the susceptibility to hemorrhagic fever , as our data suggests , the FVB and F1 hybrid LCMV-13 infection models will provide the tools necessary to decipher what the key factors are in initiating arenaviral hemorrhagic fever . Animal experiments were conducted to comply with the Public Health Service ( PHS ) Policy on the Humane Care and Use of Laboratory Animals , the US Department of Agriculture's ( USDA ) Animal Welfare Act & Regulations ( 9CFR Chapter 1 , 2 . 31 ) , the Animal Care and Use Review Office ( ACURO ) , a component of the US Army Medical Department and Medical Research and Material Command USAMRMC and the United States Government Principles for the Utilization and Care of Vertebrate Animals Used in Research , Teaching and Testing with prior approvals for established protocols from the Oregon State University Institutional Animal Care and Use Committee and ACURO when appropriate . C57BL/6 and FVB mice were purchased from Jackson Laboratories . C57BL/6 and FVB mice were bred to create F1 hybrids at Oregon State University Department of Animal Resources . All mice were housed at Oregon State University Department of Laboratory Animal Resources facility and experiments were conducted according to approved Institutional Animal Care and Use Committee protocols . Mice were used at 5–8 wks of age . Mice were infected with 1–2×106 p . f . u . LCMV-clone 13 or LCMV-Armstrong . Standardized recording of death and disease symptoms was performed on a daily basis . Symptoms of severe disease were hunched posture for more than 24 hours without movement , discordination , and shaking upon movement . For CD8 depletion , mice were treated with 0 . 5 mg clone 53–6 . 7 anti-CD8 antibody at time of infection and 24 hours post-infection . For anti-CD4 depletion , mice were treated with 0 . 3 mg CD4 antibody clone GK1 . 5 . For IL17RC SD12 antisense treatments , mice were dosed with 7 . 5 mg/kg intraperitoneal route on days 0 , 1 , and 2 . IL17RC SD12 sequence is CTG GAC ACA GAG GTT GG . The PMOplus and PPMO compounds used were manufactured as previously described , respectively [7] , [8] . Tissues were weighed and 300 ml of DMEM was added per tube . A stainless steal bead was added and tissues were homogenized in a Tissue Lyser ( Qiagen ) for 3 min at 20 Hz . Kidneys were digested in DMEM+1 mg/ml collagenase for 30 min prior to homogenization . After homogenization , tissues were spun down at 14 k for 5 min and 50 µl sup was taken for RNA isolation with Magmax Blood RNA Isolation Kit ( Ambion ) according to manufacturer's instructions . Viral load of the spleen , kidney , liver , and lung was determined by qRT-PCR amplification of viral GPC [forward primer ( 5′-GCAAAGACCGGCGAAACTAG-3′ ) , reverse primer ( 5′-CGGCTTCCTGTTCGATTTGGT-3′ ) and a taqman probe ( 5′-CCCAAGTGCTGGCTTGTCACCAAT-3′ ) ] . To translate the qRT-PCR results from a cycle threshold ( CT ) value into copy number , a standard curve was generated using PCR product of the GP amplicon . The GP amplicon PCR product was run on a gel , excised and purified , then spectrophotometry was used to quantify a copy number for that CT value . Plaque assays were conducted as previously described [31] . Blood was collected through retroorbital puncture using capillary tubes on anesthetized mice directly before sacrifice . Serum was collected after spinning at 8 k for 5 min . Blood and serum were sent to Charles River Laboratories for CBC with differential and Clinical Chemistry profiles . Serum was assayed for cytokine and chemokines using Flow Cytomix bead system ( Bender Medsystem ) according to manufacturer's protocol . Db/GP33 , Db/GP276 , and Db/NP396 pentamers were purchased from Proimmune ( Sarasota , FL ) . Single cell suspensions were stained with pentamer according to manufacturers protocol and then subsequently stained for CD8 , CD44 , and CD19 antibodies ( BD Biosciences , San Jose , CA ) . Pentamer positive cells were detected on FC500 ( Beckman Coulter , Indianapolis , ID ) and analyzed on FlowJo software ( Ashland , OR ) . For ICS analysis , splenocytes were stimulated overnight with 1 µg/ml of the indicated peptides . BFA ( Ebioscience ) was added 4 hours prior to harvest and cells surface stained for CD44 and CD8 , fixed in cytofix/perm and washed in perm/wash buffer ( BD Biosciences ) . Cells were then incubated with IFNγ antibody ( BD Biosciences ) .	Arenaviruses are carried by rodents , and in South America and West Africa can cause a fatal hemorrhagic fever syndrome in humans . Food , water or household items contaminated with rodent urine can be a source for transmission . General supportive care , anti-fever medication and the antiviral drug Ribaviran are used , however no treatment has proven effective . Due to the lack of small animal models capable of reproducing the human disease , development of an effective therapeutic has been slow . Here we report that a common laboratory arenavirus isolate , lymphocytic choriomeningitis virus , known to cause only a mild infection in humans and a chronic , wasting disease in most laboratory strains of mice produces a hemorrhagic-like disease in the FVB mouse strain . These mice exhibit signs of bleeding , multi–organ involvement , changes in blood diagnostics and mortality indicative of hemorrhagic fever syndrome following infection . We also show that a drug approach to reduce inflammation as a result of immune responses to the virus reduced disease signs and improved survival . Our study provides a small animal model for testing new treatment approaches and points to drug targets that lessen disease severity and improve survival from arenavirus induced hemorrhagic fever .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods", "and", "Methods" ]	[ "immune", "cells", "immunity", "to", "infections", "immunology", "microbiology", "host-pathogen", "interaction", "animal", "models", "model", "organisms", "inflammation", "t", "cells", "biology", "mouse", "immune", "response", "immunopathology", "immunity", "virology" ]	2012	Lymphocytic Choriomeningitis Virus Infection in FVB Mouse Produces Hemorrhagic Disease
Ornithodoros turicata is a veterinary and medically important argasid tick that is recognized as a vector of the relapsing fever spirochete Borrelia turicatae and African swine fever virus . Historic collections of O . turicata have been recorded from Latin America to the southern United States . However , the geographic distribution of this vector is poorly understood in relation to environmental variables , their hosts , and consequently the pathogens they transmit . Localities of O . turicata were generated by performing literature searches , evaluating records from the United States National Tick Collection and the Symbiota Collections of Arthropods Network , and by conducting field studies . Maximum entropy species distribution modeling ( Maxent ) was used to predict the current distribution of O . turicata . Vertebrate host diversity and GIS analyses of their distributions were used to ascertain the area of shared occupancy of both the hosts and vector . Our results predicted previously unrecognized regions of the United States with habitat that may maintain O . turicata and could guide future surveillance efforts for a tick capable of transmitting high–consequence pathogens to human and animal populations . Argasid ticks in the genus Ornithodoros are globally distributed with five known species present in the United States . In the western and midwestern regions of the United States , Ornithodoros coriaceus , Ornithodoros hermsi and Ornithodoros parkeri are distributed in various ecological settings , while Ornithodoros turicata and Ornithodoros talaje are found in arid regions of the southern United States and into Latin America [1 , 2] . Ornithodoros ticks are known vectors of veterinary and medically significant pathogens [3] . The ticks transmit spirochetes that cause relapsing fever borreliosis [4 , 5] , while O . coriaceus , O . parkeri , and O . turicata have also been experimentally infected with African swine fever virus ( ASFV ) [6] . With outbreaks occurring in Central and South America , and the Caribbean , ASFV is considered a global threat to the livestock industry [7] . The distribution of Ornithodoros species , particularly O . turicata in the southern United States , indicates that a considerable opportunity exists for the maintenance of ASFV were the pathogen introduced [8] . Ornithodoros turicata was first described from specimens collected in Guanajuato , Mexico in 1876 [4 , 9] , while collections in the United States include Arizona , California , Colorado , Florida , Kansas , New Mexico , Oklahoma , Texas , and Utah [4 , 8 , 10] . The absence of O . turicata collections between Texas and Florida indicates the Florida population may be geographically isolated [4 , 8 , 10–14] . Moreover , the distribution of the O . turicata throughout regions of North America suggests the potential for considerable adaptability and genetic plasticity . The feeding behavior of O . turicata further indicates the adaptability of the ticks . Ornithodoros turicata are nidicolous nocturnal feeders [4 , 14] that engorge within 60 minutes , and are rarely found attached on their vertebrate hosts [15 , 16] . They colonize peridomestic settings and inhabit burrows , nests , caves , and cavities under outcroppings [4 , 15] . Ornithodoros turicata are also promiscuous feeders and recognized hosts include prairie dogs ( Cynomys spp . ) , ground squirrels ( Spermophilus spp . ) , snakes , cattle , pigs , and the gopher tortoise ( Gopherus polyphemus ) [4 , 15] . Moreover , successful laboratory colonies have been produced by feeding the ticks on mice [17] and chickens ( authors JEL and PDT ) , while reports from Texas associate the arthropods with feeding on canines and humans [18 , 19] . The life cycle of O . turicata contributes to the capability of the tick to serve as a reservoir for pathogens . The arthropods can have up to six nymphal instars and survive extreme periods between bloodmeals [17] . In laboratory studies , the life span of adult ticks was over 10 years when fed regularly , and they endured at least five years between feedings [4 , 20 , 21] . Prolonged periods of starvation in nature were also reported in mark-and-recapture studies [12] . Adults also feed and reproduce multiple times throughout their life time . The longevity and hardiness of O . turicata indicates the importance of understanding ecological factors that may impact the distribution of the ticks . Linkages of biotic and abiotic factors to tick distribution , and risk of exposure to tick-borne pathogens have been most investigated and modeled for surface dwelling ixodid ticks [22 , 23] . Nidicolous argasid ticks present an additional challenge for modeling , as their life history plays out in niches between subterranean microclimates of burrows or landscape cavities ( e . g . , caves ) and the external environment [24] . Tick development , survival , and dispersal are interdependent upon the population dynamics , diversity , resting cycle , and surface activity of their vertebrate hosts [25] . This current study aimed to provide a composite prediction of the distribution of O . turicata in the United States and into north Mexico . We generated locality points for O . turicata at the county-level by evaluating historic databases and performing tick collection efforts . The data were used in conjunction with a maximum entropy species distribution model ( Maxent ) , which offers good performance , robustness , and statistical validation [26–28] . A shapefile depicting the predicted distribution of O . turicata was generated based on the most informative Maxent model . This shapefile was overlaid with the range of the community of hosts in order to obtain the area shared between individual vertebrate species and O . turicata . Our report provides the framework that will guide future tick surveillance efforts and research to further understand host communities and suitable habitat that supports the maintenance of O . turicata . Tick baiting collections from G . polyphemus dens in Florida were performed under permit number 2014–0310 from the Florida Department of Agriculture and Consumer Services . When O . turicata specimens were collected from Wildlife Management Areas in Texas ( WMA ) , approval was obtained from Texas Parks and Wildlife and WMA personnel . The currently known distribution of O . turicata was assessed by performing literature searches using O . turicata and B . turicatae as key words ( n = 47 ) [13 , 19 , 20 , 29–35] . When authors reported specific regions of field collected ticks or locations of human exposure to B . turicatae geographic data points were noted at the county level . Records of O . turicata collections were also obtained from the United States National Tick Collection ( USNTC ) [36] and the Symbiota Collections of Arthropods Network ( SCAN ) ( n = 104 ) [37] , and field studies in Texas and Florida ( n = 7 ) . The catalog of O . turicata collections provided from the USNTC consisted of O . turicata numbers , developmental stages , specimen identification , and locality information . When a municipality , county , or estimated location of collection was provided , manually georeferenced data points were generated using Google Earth [38] . When ticks were collected from a vertebrate host , the level of host taxa was noted . When Ornithodoros ticks were collected at field sites in Texas and Florida , the latitude and longitude were recorded . In the evening , dry ice was placed at the openings of gopher tortoise burrows , coyote ( Canis latrans ) dens , and within 1–2 m of rodent ( Neotoma , Peromyscus , and Sigmodon spp . ) and burrowing owl ( Athene cunicularia ) nests . As ticks emerged from a given location , they were collected and housed in 15 or 50 ml conical tubes with perforated caps , and kept separate according to the coordinates of the den and nest . Upon returning to the laboratory , the arthropods were housed at 27°C and 85% and 95% relative humidity for ticks collected in Texas and Florida , respectively [39 , 40] . They were also inspected by microscopy to determine species collected using taxonomic descriptions [4] . A total of 158 O . turicata localities with no duplicates from the datasets described above were used to build the SDMs . Locality records ( presence only data ) that lacked geographical coordinates ( longitude and latitude ) were georeferenced using GEOLocate v . 3 . 21 [41] . For specimens that only had county level data , GEOLocate was used to obtain geographic coordinates from the centroid of each corresponding county . Twenty data layers containing environmental and altitude variables at 30 arc-seconds ( ≈1 × 1 km2 grid cells ) spatial resolution ( S1 Table ) were acquired from the WorldClim dataset ( http://worldclim . org ) [42] . The WorldClim dataset was generated through interpolation of average monthly data from numerous global weather stations representing a time period from 1950–2000 . To determine whether there were correlations among the environmental variables , a correlation matrix was built using the “Explore Climate: Correlations and Summary Stats tool” within an extension toolkit SDMtoolbox ( http://sdmtoolbox . org ) [43] . Based on the Pearson’s correlation value we eliminated one variable in each pair that was either ≥ 0 . 80 or ≤ -0 . 80 ( S2 Table ) . The variable accepted out of the pair was based on biological importance in relevance to O . turicata . Out of the 20 variables we kept 12 ( Table 1 ) . Localities , environmental , and altitude variables were used with a common machine learning technique , maximum entropy modeling ( Maxent v . 3 . 3 . 3k ) [26] to construct four O . turicata SDMs . Maxent was chosen because of its performance , robustness , and statistical validation [27 , 28] . Moreover , given the limited understanding of O . turicata geographic range and Maxent’s use of presence-only data , this modeling approach was an appropriate first step to increase our understanding of the distribution of the tick . Maxent was set up with the following default settings: algorithm parameters set as auto features , a convergence threshold of 0 . 00001 , a maximum of 10 , 000 background points , a regularization multiplier of 1 , and a logistic output grid format; all other remaining parameters were left in the default settings . In addition to these default settings the number of iterations was modified to 5 , 000 . Response curves were also included for each Maxent model run , which allows one to evaluate how the prediction depends on the variables . Maxent creates two types of response curves , the first shows how the prediction changes with each variable while keeping all other variables in the model ( S1 Fig ) , and the second only utilizes the corresponding variable ( S2 Fig ) . The advantage of the second response curves is that it allows for easier interpretation of environmental variables that may be correlated or slightly correlated . The following four models with 10 subsampling runs using a random test percentage of 20% ( n = 31 ) were constructed with predefined environmental and altitude parameter data layers ( Table 1 ) : 1 ) an average precipitation model , 2 ) an average temperature model , 3 ) an average full model which included all environmental variables and altitude , and 4 ) an average model based on the top-five environmental variables from the full model ( top-five model ) . Maxent calculated the area under the curve ( AUC ) , a value that ranges from 0 . 5 to 1 , providing information on the species’ restricted predicted distribution in relation to the range of predictor variables in the model , but this value does not necessarily measure the fit of the model [44] . An informative AUC score is equal to 1 , while an AUC score of 0 . 5 equates a model performance no better than random . To see which variables contributed the most to the model , Maxent calculated the percent contribution ( PC ) and permutation importance ( PI ) of each variable for all models . The PC value is dependent on the algorithm path that Maxent used to obtain the model while PI depends only on the final Maxent model . Additionally , variable jackknifing for the training gain , test gain , and test AUC was also conducted where each environmental variable was excluded . A model was subsequently created , and the process was repeated until all the variables had been excluded . The second part of the jackknife test created a model using each environmental variable in isolation of one another . Specifically for the top-five model , three additional models were run using: 1 ) the top-five variables from the PC , 2 ) the top-five variables from the PI , and 3 ) the top-five variables from the jackknife results . The highest AUC score among these three models was selected to be our top-five environmental model . All SDMs were processed and visualized in QGIS v . 2 . 4 Chugiak [45] with the geographic area restricted to the United States and Mexico based on collection records available of O . turicata . Five classes of probabilities each with about a 20% interval were given a specific name and color for visual representation of model results: very high probability ( red ) , high probability ( orange ) , moderate probability ( yellow ) , low probability ( green ) , and very low probability ( white ) . A shapefile representing the most informative SDM was drawn in QGIS by including the percentage of low ( < 20% ) to very high probability ( ∼86% ) . To create a shapefile we excluded isolated areas that were not in close proximity from areas > ~20% ( such as the states of Washington , Idaho , Montana , Wyoming , and South Dakota ) . Alabama , Louisiana , and Mississippi were also excluded because previous survey data indicated the absence of O . turicata [46] . This shapefile was then imported into ArcMap 10 . 2 . 2 [47] where the geometry of the shapefile was checked using the “Check Geometry Tool” within the “Data Management Toolbox . ” This tool reported if any errors occurred , such as overlaps and/or no closed connection for polygons . If problems were found with the shapefile it was fixed using the “Repair Geometry Tool” also within the “Data Management Toolbox . ” The shapefile was projected to the map projection of North American Albers Equal Area before the area ( km2 ) of the shapefile was calculated . Host diversity was obtained from the USNTC , where host records were classified to either genus or species . Therefore in cases with taxa at the genus level we included all species known to occur in the most informative Maxent model distribution . In addition , we included 48 suspected hosts because these species are part of the burrow community utilizing the burrows of original excavators . For example A . cunicularia inhabits the burrows of Cynomys ludovicianus ( S3 Table ) [48] . A total of 58 host species were included in this study ( S3 Table ) . Taxonomic names were based on the International Union for Conservation of Nature ( IUCN ) Red List of Threatened Species ( http://iucnredlist . org ) [49] and Wilson & Reeder’s Mammal Species of the World , 3rd edition database [50] . We obtained most of the host species distribution shapefiles from the IUCN Red List of Threatened Species [49] . For Sus scrofa , its distribution was obtained from the National Feral Swine Mapping System , Southeastern Cooperative Wildlife Disease Study . No spatial data were provided in the IUCN Red List of Threatened Species for G . agassizii and G . polyphemus thus they were drawn in QGIS following known species distributions [51 , 52] . In ArcMap all shapefiles were checked and fixed for errors in geometry and projected to North American Albers Equal Area Conic . Some of the hosts’ distribution shapefiles obtained were clipped to the extent of the United States and Mexico . The area of extent ( km2 ) for each host species’ distribution was calculated in ArcMap . Each of the host species shapefiles were intersected with the most informative SDM shapefile in order to obtain new shapefiles showing shared occupancy . For A . cunicularia , because of its nesting phenology , analyses were conducted with its full range and three range subdivisions: year-round , breeding , and winter . In ArcMap we calculated the area ( km2 ) of these shared occupancies for each individual species . Next we calculated the percentage of the area shared for each species as %ofareashared= ( Areaofindividualspp . shareddistributionTotalareaofO . turicatadistribution×100 ) . A total of 158 sample localities were obtained by performing literature searches in the United States National Library of Medicine ( n = 47 ) ( S4 Table ) [18–20 , 29 , 30 , 32 , 33 , 53] , evaluating records provided by the USNTC and SCAN ( n = 104 ) , and field collection studies ( n = 7 ) ( Fig 1 ) . Since tick-borne relapsing fever spirochetes are transmitted by a specific species of Ornithodoros [54] , literature searches identified presumed case reports of infection caused by B . turicatae as determined by visualizing spirochetes in human blood smears , and county localities were included if O . turicata was collected [18 , 19 , 26] . Also , studies reporting molecular evidence of host infection with B . turicatae were used to define the distribution of the tick vector . For example , DNA sequence analysis demonstrated B . turicatae as the etiological agent in domestic dogs , and serological evidence indicated exposure of a human patient in central Texas [32 , 55] . These reports provided additional localities for O . turicata . Moreover , recent field studies from 2012–2014 in Texas and Florida resulted in the collection of O . turicata ( Table 2 ) , and indicated recent evidence of geographic area for this tick species . The four average species distribution models produced by Maxent ( precipitation , temperature , full , and top-five ) show variation in the probability of occurrences for O . turicata within the United States and Mexico ( Fig 2A–2D ) . The average AUC scores for the full , temperature , and precipitation models were 0 . 949 ± 0 . 011 SD , 0 . 925 ± 0 . 013 SD , and 0 . 910 ± 0 . 027 SD , respectively . The average AUC scores for the three top-five models are as follows: top-five environmental variables based on the percent contribution ( PC ) = 0 . 942 ± 0 . 013 SD , the permutation importance ( PI ) = 0 . 939 ± 0 . 010 SD , and the jackknifes 0 . 933 ± 0 . 01 3 SD . We chose the top-five model using the PC as is it had the highest AUC score out of the three . Thus the results to follow of our top-five model will focus on the top-five variables based on the PC . The regularized training gain is similar to a goodness of fit test , and at the start of a run for a given model this value begins at 0 and increases to an asymptote starting with a uniform distribution then gradually increasing the fit . When the model has reached its end the last gain value represents how the model fits around the input presence data . The highest average training gain was 2 . 065 ± 0 . 041 SD for the full model and with the average likelihood of the presence data 7 . 885 times higher than that of a random background pixel ( e2 . 065 = 7 . 885 ) . This average training gain value for the full model was followed by the value for the top-five ( gain = 1 . 927 ± 0 . 050; times higher than random = 6 . 869 ) , temperature ( 1 . 873 ± 0 . 032; 6 . 508 ) , and precipitation ( 1 . 266 ± 0 . 061; 3 . 547 ) . The average test gains followed the same trend with the full model having the highest test gain ( 2 . 021 ± 0 . 012 SD ) followed by top-five ( 1 . 918 ± 0 . 252 SD ) , temperature ( 1 . 666 ± 0 . 187 SD ) , and precipitation ( 1 . 448 ± 0 . 266 SD ) . In the average precipitation model , precipitation of warmest quarter ( BIO18 ) was a major determining factor for percent contribution ( PC ) followed by precipitation seasonality ( BIO15 ) , annual precipitation ( BIO12 ) , precipitation of driest quarter ( BIO17 ) , and precipitation of coldest quarter ( BIO19 ) ( Table 3 ) . As for permutation importance ( PI ) , precipitation seasonality ( BIO15 ) , precipitation of the warmest quarter ( BIO18 ) , and precipitation of the driest quarter ( BIO17 ) , were most important for the full model ( Table 3 ) . The jackknife test of variable importance for the training gain , test gain , and test AUC indicated that the environmental variable with the highest gain when used alone was precipitation of the warmest quarter ( BIO18 ) ( S3A Fig ) . However , the environmental variable precipitation seasonality ( BIO15 ) indicated the greatest decrease in gain when it was omitted , thus appearing to have the most information that is not present in other environmental variables . The variable in the precipitation model that produced the lowest training gain when used alone was precipitation of the coldest quarter ( BIO19 ) . The raster map of the precipitation model covered most of the Great Plains predicting a larger portion of probable regions in the central northern states ( Montana , Nebraska , North Dakota , and South Dakota ) ( Fig 2A ) , an area where no confirmed records exist . In addition , the precipitation model predicted that many of our confirmed western ( i . e . , California , Nevada ) and Florida localities were of low probability ( ~20% ) for O . turicata . Thus , precipitation variables alone do not accurately reflect the distribution of field collected O . turicata , but suggests a geographical gap exists between Florida and Texas populations , and predicts a large area for which currently no records of O . turicata exist . In the average temperature model , PC was highest for maximum temperature of warmest month ( BIO5 ) , followed by mean temperature of driest quarter ( BIO9 ) , annual mean temperature ( BIO1 ) , mean temperature of the wettest quarter ( BIO8 ) , temperature seasonality ( BIO4 ) , annual mean temperature ( BIO1 ) , and mean diurnal range ( BIO2 ) ( Table 3 ) . Based upon PI the most important variables were annual mean temperature ( BIO1 ) and mean temperature of the direst quarter ( BIO9 ) ( Table 3 ) . The jackknife tests of environmental variable importance for the average training gain , test gain , and test AUC showed that max temperature of warmest month ( BIO5 ) represented the most useful information by itself , whereas omitting temperature seasonality ( BIO4 ) had the most information not present in any of the other variables ( S3B Fig ) . Mean diurnal range ( BIO2 ) , when used alone , produced the lowest training gain for this model . The raster map of the average temperature model showed mainly moderate ( ~40% ) to very high probability ( 85% ) in Colorado , Kansas , New Mexico , Oklahoma , and Texas though areas of moderate probability occurred in other states ( Fig 2B ) . The model also predicted an area of low probability ( ≤ 21% ) throughout the majority of Alabama , Florida , Louisiana , and Mississippi . The temperature model also indicates an area of low probability in the Central Valley of California . In the average full model , the highest contributing variables were maximum temperature of the warmest month ( BIO5 ) followed by mean temperature of the wettest quarter ( BIO8 ) , mean temperature of the driest quarter ( BIO9 ) , precipitation seasonality ( BIO15 ) , and precipitation of the warmest quarter ( BIO18 ) ( Table 3 ) . For permutation importance annual mean temperature ( BIO1 ) and annual precipitation ( BIO12 ) were the most important variables . Jackknifing of the average training gain , test gain , and test AUC results indicated that maximum temperature of the warmest month ( BIO5 ) provided the most important training gain , mean diurnal range ( BIO2 ) produced the lowest , and temperature seasonality ( BIO4 ) had the most information that was not present in other variables ( S4A Fig ) . The full model predicted regions of occurrences where the majority of currently known O . turicata localities occur ( Fig 2C ) , suggesting that it may be the most accurate . A gap of probable occurrences remained between Alabama , Louisiana , and Mississippi , which were projected to be areas of very low ( < 21% ) to low ( ~21% ) probability mainly along the coast . New regions of moderate probability ( ~43% ) for the tick included arid areas of northern Mexico ( Coahuila and Tamaulipas ) , South Carolina , and Georgia ( Fig 2C ) . Our average top-five model based on the percent contribution was also generated by incorporating the top-five environmental variables: maximum temperature of the warmest month ( BIO5 ) , mean temperature of the wettest quarter ( BIO8 ) , mean temperature of the driest quarter ( BIO9 ) , precipitation seasonality ( BIO15 ) , and precipitation of the warmest quarter ( BIO18 ) . The average top-five model indicated a geographical gap through Alabama , Louisiana , and Mississippi with very low ( < 21% ) to low ( ~21% ) probability for O . turicata along the coast ( Fig 2D ) . Compared to the average full model , the regions of high- to very-high probability expanded into Arizona , eastern Kansas , Oklahoma , the Texas Panhandle , and the northern Mexican states of Coahuila , Nuevo Leon , and Tamaulipas , while low probability was predicted in Georgia and South Carolina . In this model , both percent contribution and permutation importance indicated that maximum temperature of the warmest month ( BIO5 ) had the most influence ( Table 3 ) . Jackknife test of the average training gain , test gain , and test AUC results showed that maximum temperature of the warmest month ( BIO5 ) was the environmental variable with the highest training gain , which contained more information that was not explained by the other variables ( S4B Fig ) . The lowest gain in this model for an individual environmental variable was precipitation seasonality ( BIO15 ) . A nidicolous-ectoparasite’s distribution is not shaped exclusively by environmental variables , but it also depends heavily on host distribution , diversity , and interactions among species that are excavators , modifiers , and occupants . In historical O . turicata collections , vertebrate hosts were documented when ticks were obtained off a given animal , and the broader range of potential hosts was not considered . Therefore , we evaluated the amount of distributional overlap between likely vertebrate hosts and O . turicata . The average full model was used for the creation of the shapefile because the AUC score was the most informative ( S5 Fig ) . Overlaying this shapefile identified the shared occupancy area between host species and O . turicata . For host species we included known and suspected vertebrate host associations derived from the USNTC database . The total area of the range of O . turicata based on this shapefile was 1 , 752 , 272 km2 ( southeastern range = 167 , 318 , km2; western range = 1 , 584 , 954 km2 ) . Fifty-eight known or suspected host species were found to inhabit the area within the range of O . turicata , as determined by Maxent ( S3 Table ) . Mammalian hosts were the most numerous ( n = 42; 72 . 4% ) , followed by reptilian ( n = 15; 25 . 9% ) and avian hosts were represented only by A . cunicularia ( 1 . 7%; Strigiformes , Strigidae ) ( Fig 3 , S3 Table ) . The group of hosts was quite diverse as they represented a total of 8 orders , 15 families and 17 genera . Mammals comprised 5 orders , 10 families , and 11 genera , with the genus Dipodomys and Neotoma being the most diverse each with 10 species . Reptiles were represented by 2 orders , 4 families , and 5 genera with the genus Crotalus having the most diversity ( 8 species ) . We assessed the overall area occupied between O . turicata and all 58 host species ( S3 Table ) . Most ( 7 of 10 ) of the species that had half or more of their distribution overlapping with the predicted distribution of O . turicata were mammals , with the exception of two reptiles ( Terrapene ornata and Crotalus atrox ) and the single avian host , after considering its year-round range and total range ( Fig 3 ) . Coyotes ( C . latrans ) and badgers ( Taxidea taxus ) , two species known as hosts of O . turicata , had a range that overlapped most of the predicted distribution of O . turicata . Ten of 12 of the host species that have been recorded to be associated with O . turicata had a range overlap of 40% or more with the predicted model , with the remaining 2 known hosts having low values ( < 12% , G . agassizii , G . polyphemus , ) . Kangaroo rats ( Dipodomys ) represent one of the most diverse genera of hosts for O . turicata but most of them had limited distributions and thus low overlaps ( < 10% ) , except for two species , D . ordii and D . merriami , which had moderate to large overlaps , 49 . 1 and 38 . 7% , respectively . Rattlesnakes ( Crotalus ) were the most diverse genus of reptilian hosts showing low ( 8–10% ) to large ( 61 . 5% ) range overlap with the O . turicata distribution model . This study utilized database searches , field collections , and a maximum entropy modeling approach to generate a present-day geographic distribution prediction of localities for O . turicata in the United States and north Mexico . A stringent definition was used to generate georeferenced data points , potentially excluding regions of Latin America where O . turicata may exist . For example , recent case reports have described human infection to relapsing fever spirochetes in Sonora , Mexico and along the Guatemala-Belize border [56 , 57] . However , the reports did not present molecular evidence of B . turicatae infection , nor were ticks collected at the presumed exposure sites , and thus these localities were excluded from our analyses . Additionally , Dr . Oscar Felsenfeld described O . turicata into South America [2] , yet to our knowledge ticks were not collected or morphological features noted to speciate the vectors and the region was omitted . These reports highlight the need to expand research efforts to understand the distribution of O . turicata in the Neotropics , a region where the vectors and pathogens they transmit have been overlooked . The nidiculous life cycle and rapid feeding behavior of O . turicata has resulted in the collection of few specimens , and our understanding of the tick’s distribution is limited . A total of 158 O . turicata localities were obtained for this study , and over 90 collections occurred prior to 1950 . Moreover , habitat information for ticks was sparse and solely at the county level and estimated georeferenced localities . Consequently , additional fine scale studies were not possible . Regardless , the Maxent analyses in this report have provided the framework to initiate field studies to further define the habitat of O . turicata . We envision microecological analyses of localities where ticks have been collected to better define burrow communities . From the four SDMs generated , we hypothesize that the models including all and the top-five environmental variables most accurately predict the distribution of O . turicata . The model that included only the top-five variables predicted the Florida Panhandle and central California as a region of low to very low probability for O . turicata . While we are unaware of current day field collections in central California , we recently obtained O . turicata from gopher tortoise dens as far west in Florida as Apalachicola National Forest , which indicates that the Florida Panhandle is an ecologically suitable region for this tick species . The models that exclusively used temperature or precipitation variables are likely the least accurate . The precipitation model predicted ranges of very low to low probability for O . turicata in regions where ticks have recently been collected , such as the Florida Panhandle and Joshua Tree National Park , California [58] . Additionally , the temperature model may inaccurately predict regions of probable occurrences throughout Alabama , Louisiana , and Mississippi , where a present day gap exists . While Maxent requires presence only data , an extensive evaluation of gopher tortoise dens in Mississippi failed to identify soft ticks [46] , and supports prediction models that suggest most of this region may not sustain O . turicata . The geographical gap between Texas and Florida may occur due to unique current climate conditions associated with this geographic area . Evaluating the original environmental variable inputs used in the Maxent models , this gap area may be a result of low temperatures during the wettest quarter ( BIO8 ) ( Fig 4A ) and high temperatures during the driest quarter ( BIO9 ) ( Fig 4B ) . The response curves for BIO8 showed that low probability is stable until temperatures reach 20°C where probability started to increase then declined ( Fig 5A ) . While the response curve for BIO9 indicated that the probability increased with low temperatures , it then declined at higher temperatures ( Fig 5B ) . In addition , this gap area also receives greater than 200 mm of precipitation during the driest quarter ( Fig 4C ) . Evaluating the response curve with BIO17 as a variable ( Fig 5C ) the probability drops off at 200 mm within this gap region . The precipitation and temperature variables may explain the gap occurring between Texas and Florida , but whether this vicariance event was due to recent changes in the environment is unknown . Another possibility for the disjunction between Texas and Florida is an extinction event in the intervening gap area , or the introduction of O . turicata by long-distance dispersal of a migratory animal . Currently , burrowing owl populations in Florida are predominately year-round residents [59] and are distinct from migrating western burrowing owls [60] . Interestingly , our work in Texas identified nymphal Carios kelleyi , a genus of soft ticks that also rapidly feed as nymphs and adults , on bats as they emerged from the roost . Given similar collections of O . turicata from caves and the indiscriminant feeding behavior of these ticks , bats may have an important role in soft tick dispersal . Clearly , much is unknown regarding the mechanistic explanation of this distribution gap , and further phylogeographic and modeling studies will provide insight toward gene flow between the populations of O . turicata . To comprehend the dispersal of O . turicata and define suitable habitat for the ticks , field studies and additional tick collections are needed throughout the United States . Our study identified new regions to focus upon , including portions of Georgia , North and South Carolina , Kansas , Oklahoma , and Nevada . Moreover , given the geographical overlap of O . turicata with O . parkeri and O . talaje , do the two species of tick share similar habitat and if so does the host community support the fitness of one tick species over another ? The peculiar life-cycle of soft ticks presents unique challenges for understanding their biology , distributions , and the ecology of the pathogens they transmit . In general the ticks are rapid nocturnal feeders and most of their life-cycle occurs in a nest , den , or cave , which serve as a micro-refuge from extreme temperatures , humidity , fire , and predation [61] . Also , they are rarely encountered on the vertebrate host . Therefore , the identification of endemic regions can be difficult and requires the development of non-traditional tick surveillance methods , such as serological assays to detect antibodies against genus or species-specific tick salivary antigens [62–64] . In this report , we included a total of 58 vertebrate host species , but because burrows are frequently in a state of flux with primary excavators , modifiers , and occupants , there are likely more potential hosts . Bloodmeal sources not included in the study encompass species of rodents such as voles ( Microtus ) , mice ( Mus , Onychomys , and Peromyscus ) , cotton rats ( Sigmodon ) , and gophers ( Geomys ) . Other potential vertebrate hosts include members of Mustelidae , small mammals ( badgers , ferrets , and weasels ) that dwell in habitat utilized by burrowing owls [65 , 66] . Reptilian burrow occupants such as racers ( Coluber ) and the gopher snake ( Pituophis catenifer ) should also be evaluated as potential hosts for O . turicata in addition to lizards that reside in burrows and outcroppings . The complex host community composition ( species makeup ) and structure ( relative abundance ) at a given site will likely influence the ecology of soft tick populations and the pathogens they transmit . Within the list of known hosts , several species have a high spatial overlap , as predicted by our model , and are prevalent components of local communities . The indiscriminant feeding behavior of Ornithodoros species and diversity of vertebrate hosts increases the likelihood of supporting the life-cycle of soft ticks . However , less understood is the host’s role in maintaining pathogens transmitted by soft ticks . Ornithodoros species are important vectors of human and veterinary pathogens [19 , 30 , 33 , 67] . Studies suggest that in addition to vertebrates , the ticks could be a reservoir for microorganisms given their long life-span and ability to endure prolonged periods of starvation [20] . Borrellia turicatae , a human and canine pathogen , is maintained throughout the life-cycle of O . turicata [1 , 30 , 33] , and in rodents a full bloodmeal by the ticks is not required for establishing infection [68] . An important step toward understanding the ecology of this pathogen would be to ascertain which additional vertebrate hosts help maintain B . turicatae in nature . Studies in Borrelia burgdorferi indicate that within an ecological setting , community composition , host diversity and competency for the pathogens has profound implications for the risk of Lyme disease in human populations [69] . Presumably , there are patterns that govern the transmission and maintenance of B . turicatae , but the ecology of this pathogen remains vague . Ornithodoros turicata is also one of several North American species found capable of transmitting ASFV [6 , 8 , 70] . ASFV is an ever-present biosecurity concern to domestic swine industries in the Western Hemisphere , and has been spreading globally over the last decade [7] . The Global Alliance for ASF has emerged to address surveillance , prevention , control , and research needs [71] as the virus inflicts severe economic damage and causes nearly 100% mortality in susceptible domestic pig populations [7 , 72] . The expansion and ecological overlap of feral swine and O . turicata in the southern United States , and the longevity and feeding behavior of this soft tick has increased the risk for the emergence of ASFV in North America [73] . Our predictive SDMs will guide future studies to understand the ecology and increase awareness of pathogens transmitted by O . turicata .	Argasid ticks are understudied vectors of significant human and veterinary pathogens . The life-cycle and feeding behavior of the tick poses challenges when attempting to understand the vector’s distribution . These ticks reside in dens , nests , and cave cavities , and are indiscriminant nocturnal feeders . They also engorge within minutes of attachment , and identifying the ticks on a vertebrate host is infrequent . To guide future surveillance studies , we predicted regions of probable occurrences for Ornithodoros turicata , a species capable of transmitting relapsing fever spirochetes and African swine fever virus . Historical databases and published literature were evaluated , and we collected ticks from regions of the United States . Environmental factors linked with known localities of O . turicata were used in a mathematical modeling program , which predicted regions in the United States and north Mexico likely to sustain the ticks . Additionally , vertebrate host ranges were associated with the predictive models , which may indicate how the tick vectors are dispersed . Collectively , these studies identified previously unrecognized regions that could sustain the ticks , and we envision that our work will help guide surveillance and additional research efforts to understand the ecology of pathogens transmitted by argasid ticks .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "united", "states", "invertebrates", "livestock", "medicine", "and", "health", "sciences", "ixodes", "pathology", "and", "laboratory", "medicine", "atmospheric", "science", "pathogens", "geographical", "locations", "vertebrates", "animals", "mammals", "north", "america", ...	2016	Assessment of the Geographic Distribution of Ornithodoros turicata (Argasidae): Climate Variation and Host Diversity
Alveolar echinococcosis ( AE ) is a parasitic zoonosis resembling malignancy due to its clinically silent infiltrative growth , predominately in the liver . The comorbid psychological burden and fear of disease progression in AE patients have hardly been examined to date . The aim of this study was to evaluate depression , anxiety , quality of life , and fear of disease progression in AE patients . In a cross-sectional study , n = 57 AE patients were invited to report on depression ( PHQ-9 ) , anxiety ( GAD-7 ) , somatic symptom load ( SSS 8 ) , trauma symptoms ( PTSS-10 ) , quality of life ( SF-12 ) and on fear of disease progression ( FoP-Q-SF ) using validated psychometric instruments . Furthermore , attachment style was assessed ( RQ-2 ) . N = 47 patients completed the questionnaires ( response rate 82 . 5% ) . Depression , anxiety , and somatic symptom load were above norm sample means , while physical quality of life was below norm sample means . Existing traumatic symptoms were comparable to those in cancer patients , while fear of disease progression even exceeded cancer patient scores . Patients with a secure attachment style showed less pronounced psychological burden than patients with other attachment styles . Adequate , guideline-based depression and anxiety treatment was very rarely installed . The present study revealed remarkable levels of psychological burden in AE patients . In our study sample , we discovered high depression and anxiety levels , a significant reduction of physical quality of life , and fear of disease progression . These results show how important it is for AE patients to be thoroughly assessed with regard to psychological symptoms and mental disorders so that those in need can receive sufficient psychosocial support and treatment according to official guidelines . Alveolar echinococcosis ( AE ) is a parasitic zoonosis . The life cycle of Echinococcus multilocularis predominately runs between foxes and rodents [1] . Humans are generally infected accidentally . The parasite’s target organ is the liver with an infiltrative clinically silent growth pattern , and—in advanced disease—local and distant metastases . For malignancy , AE , thus , is a main differential diagnosis [2] . In early stages of the disease , the lesion ( s ) is ( are ) resected; in advanced stages , further growth is suppressed with benzimidazoles and patients are closely monitored for side effects and complications . Currently , benzimidazoles are the only available drugs to treat AE . It is crucial to note that benzimidazoles merely have a parasitostatic effect and cannot eradicate the parasite in most cases . Due to its nature , AE may be cause of psychological impairment . So far , only one study [3] has focused on the psychological burden of AE patients and discovered that AE patients have a reduced mental quality of life . As far as we are aware , there has not been any other research on comorbid psychological burden and fear of disease progression ( FoP ) in AE patients . Compared to persons who either have a physical or mental health problem , patients who present both a somatic illness and a psychological comorbidity show dramatically reduced outcomes in all important dimensions of health: they have higher mortality rates [4–7] as well as higher levels of symptom burden and impairment in any medical disorder index [8] . In addition , health care costs are significantly higher in patients with both chronic conditions and comorbid mental disorders [9] , although the increased costs are mainly due to medical , and not psychosocial care [10] . Depression and anxiety disorders are the most frequent comorbidities of somatic diseases . Depressive disorders are predominantly characterised by persistent depressive mood , loss of interest or pleasure , and loss of energy [11] . With a life-time prevalence of up to 24% for major depressive episodes [12 , 13] , which puts them among the most common psychiatric disorders . More common in females , depressive disorders are associated with a significant impairment of quality of life [14] and with high psychological strain , including suicidal behavior [15 , 16] . Anxiety disorders , comprising general anxiety , panic or phobic disorders , are also highly prevalent in the general population [17] . They lead to avoidance behavior , social withdrawal , and are highly resistant to spontaneous recovery without specialized treatment [18] . Although not defined as an anxiety disorder , FoP ( or fear of recurrence ) is a well-known phenomenon which has mainly been researched in cancer patients . FoP is defined as an appropriate , rational response to a real physical threat and somatic treatment , which can become dysfunctional , affecting well-being , quality of life , and social functioning [19] . So far , resilience factors protecting affected persons from developing comorbid psychological disorders have not been investigated in AE patients . However , attachment style has been widely shown to be an important resilience factor [20] . Attachment style is defined as a psychological motivational system which stores perceptions , interpretations , and expectations of interpersonal interactions . This system is shaped by repeated biographical interaction experiences and results in an enduring “internal working model” which is activated in situations of need , illness , or other experiences of excessive demand [21–23] . We can distinguish four different attachment styles , namely the secure , the anxious-ambivalent , the anxious-avoidant , and the disorganized attachment style . All four types are composed of the model of the self ( positive vs . negative ) and the model of others ( negative vs . positive ) . The secure attachment style is categorized as the most mature way in which individuals respond to their own interactional needs and their partners’ requirements [24] . The aims of the current cross-sectional study were: ( 1 ) to assess depression , anxiety , somatic symptom burden , symptoms of posttraumatic stress , quality of life , and FoP in AE patients; ( 2 ) to examine differences in psychological burden based on gender , previous surgical interventions , medical leave and the duration of the illness; ( 3 ) to explore the relationship between FoP and attachment style . We hypothesised that ( i ) depression , anxiety , somatic burden , and trauma symptoms would be more pronounced compared with the general population . We thought that the quality of life would be lower in AE patients than in the general population , while FoP would be greater compared to FoP in specific groups of cancer patients ( patients with malignant melanoma , prostate cancer , or breast cancer ) . Further , we estimated that ( ii ) patients of female gender , patients without curative surgery , and patients with a longer duration of illness would present more symptoms of psychological burden . We hypothesized that medical leave due to AE would be negatively correlated with psychological burden . Lastly , we thought that ( iii ) psychological burden ( depression , anxiety , somatic burden , trauma symptoms , and FoP ) would be elevated in patients with insecure attachment styles . The study is cross-sectional . A cohort of AE patients that were diagnosed following the diagnostic criteria [25] of the expert consensus and consecutively recruited during regular patient contacts at the clinic for echinococcosis at the Section of Clinical Tropical Medicine , University Hospital Heidelberg . Since 1999 , 158 patients with AE have been diagnosed , treated , and followed up at the Section of Clinical Tropical Medicine . The clinic for echinococcosis is run in cooperation with the Departments of Diagnostic and Interventional Radiology , Gastroenterology , Parasitology , and Surgery and is a national clinical reference center for echinococcosis . In total , 57 AE patients who were attending the clinic for echinococcosis between June 2016 and May 2017 were informed and invited by the physician in charge to participate in the study . There were no exclusion criteria . All participants provided an informed consent and could withdraw their participation without any disadvantage . All the analysed data were anonymized . However , if requested before enrolment in the study , participants received a personalized summary and detailed explanation of their test results . Furthermore , all participants were invited to receive psychological consultation at the Department of General Internal Medicine and Psychosomatics . The study was approved by the ethics committee of the University Hospital Heidelberg ( ethics application no . S-232/2016 ) . The study was conducted in accordance with the Declaration of Helsinki ( most recent version: Fortaleza , Brazil , 2013 ) . All data were coded and analysed using SPSS ( version 24 ) . Missing values were not included in the calculation . The raw data are displayed by showing mean and standard deviations . Due to a sample size above 30 , we assumed a normal distribution in our data [41] . We chose a significance level of 5% ( two-sided ) . We compared the group of study participants with representative norm samples and other patient groups by using the Welch test ( tested two-sided ) due to the lack of homoscedasticity . For within-group comparisons , we used the Welch test ( tested two-sided ) in order to describe differences in psychological burden ( depression and anxiety scores , somatic symptoms , posttraumatic stress , quality of life , and FoP ) in regard to an individual’s gender , previous surgical intervention , and medical leave due to echinococcosis . We used Pearson’s correlation to assess differences in psychological burden in connection with the duration of the disease , and the number of years since the initial diagnosis . In order to investigate the correlation between the ‘model of self’ and the ‘model of others’ and levels of psychological burden , we also applied Pearson’s correlation coefficients . 47 out of 57 AE patients completed the questionnaires , equalling an overall response rate of 82 . 5% . Table 1 depicts detailed sample characteristics . Table 2 displays the psychological comorbidity of the assessed AE patients . The study participants showed a significantly higher level of depressive symptoms ( p = . 015; 31 ) and general anxiety symptoms than participants of a representative general-population survey ( p = . 015; [29] ) . Table 3 presents the distribution of patients classified by rating their depression and anxiety symptoms as minimal / mild / moderate / severe . In the PTSS-10 questionnaire , 46 AE patients ( 97 . 9% ) scored higher than 12 points , implicating the need for further diagnostic clarification regarding posttraumatic stress disorder . The mean score of the PTSS-10 in AE patients shows that their posttraumatic stress is higher than posttraumatic stress in patients currently undergoing aftercare of malignant melanoma ( p = . 045; [42] ) . Additionally , somatic ailment was significantly more pronounced in AE patients than in the general population ( p < . 001; [33] ) . Participants reported a significantly lower physical quality of life compared to a representative general population survey ( p < . 001 ) , but showed an average mental quality of life ( p = . 110 , [43] ) . Furthermore , AE patients ( female and male ) showed a significantly higher FoP compared to patients with prostate cancer ( p < . 001; [44] ) . A sum score on the FoP-Q-SF above 36 indicates a high level of fear [38] . In the current study , 18 AE patients ( 38 . 3% ) reported levels of 36 or above this cut-off score . We revealed a difference connected to patients’ gender concerning the following three aspects of psychological comorbidity: Male participants ( M = 2 . 80 , SD = 2 . 34 ) showed lower scores in the PHQ-9 ( depression ) than female participants ( M = 6 . 91 , SD = 6 . 00; Welch test: t ( 39 . 525 ) = 10 . 901 , p = . 002 ) ; male participants ( M = 52 . 54 , SD = 9 . 58 ) presented a higher psychological quality of life compared to female participants ( M = 44 . 95 , SD = 10 . 50; Welch test: t ( 36 . 562 ) = 5 . 974 , p = . 019 ) ; regarding FoP , male participants had lower scores ( M = 30 . 89 , SD = 7 . 89 ) than female participants ( M = 36 . 97 , SD = 12 . 14; Welch test: t ( 44 . 862 ) = 4 . 323 , p = . 043 ) . For further analysis , AE patients were divided into two groups , namely cured patients ( definition see above; n = 10 ) and non-cured patients ( critical and uncritical patients; definition see above; n = 37 ) . Cured patients who had had surgical intervention showed no difference in psychological comorbidity ( depression , anxiety , quality of life , somatic symptom burden , posttraumatic stress symptoms , and FoP; all p > . 05 ) compared to non-cured patients . Study participants ( n = 16 ) who reported to have had at least one period of medical leave from work due to AE showed significantly higher FoP ( M = 38 . 53 , SD = 11 . 84 ) than participants who did not have any medical leave ( M = 31 . 14 , SD = 9 . 49; Welch test: t ( 26 . 084 ) = 4 . 593 , p = . 042 ) . No significant correlation was found between the number of years since the initial diagnosis had been made and the psychological burden ( depression , anxiety , somatic symptom burden , posttraumatic stress symptoms , FoP and quality of life; r = - . 17 - . 08; all p > . 05 ) . Table 4 presents attachment styles of AE patients and their correlation with psychological burden . Attachment of the self correlates significantly with anxiety ( r = - . 319 , p < . 05 ) and trauma symptoms ( r = - . 331 , p < . 05 ) and FoP ( r = - . 334 , p < . 05 ) . Attachment to others correlates significantly with depression ( r = - . 305 , p < . 05 ) , anxiety ( r = - . 372 , p < . 05 ) and trauma symptoms ( r = - . 300 , p < . 05 ) . This suggests that AE patients who are more securely attached may be less likely to develop symptoms of depression , anxiety disorders , FoP , or trauma . Even though our research project has the great advantage of being the first study to investigate the psychological burden and quality of life in AE patients , our findings are limited by the small number of participants . Therefore , our study sample differs considerably to norm samples . Nevertheless , this investigation constitutes the largest research on the psychological burden , quality of life , and FoP in AE patients . Our results are further limited by the aspect that all dependent variables were assessed by self-report measures . Furthermore , it is important to note that our outcome may not be generalizable due to the fact that we were only able to discover a connection between the exhibited psychological symptoms and the personality factor attachment style in a cross-sectional design . The present study revealed pronounced psychological burden in AE patients in terms of high levels of depression and anxiety symptoms , a significantly reduced physical quality of life , and high FoP . Our results show how important it is that AE patients are thoroughly assessed with regard to psychological symptoms and mental disorders screening so that sufficient psychosocial support and treatment can be offered to patients in need . Further research should address the long-term development of psychological burden in AE patients , i . e . using follow-up studies to investigate the overall influence of psychological burden on the course of the disease and the impact of sufficient therapeutic support on the disease outcome .	Alveolar echinococcosis ( AE ) is a parasitic zoonosis resembling malignancy due to its clinically silent infiltrative growth , predominately in the liver . For some somatic diseases , comorbid psychological burden predicts the course of disease . However , as far as we are aware , little is known about comorbid psychological burden and fear of disease progression in AE patients . In a cross-sectional study , AE patients were invited to report on depression , anxiety , somatic symptom load , symptoms of posttraumatic stress , quality of life , and their fear of disease progression . Furthermore , we assessed the patients’ attachment style . The present study revealed that depression , anxiety , and somatic symptom load were above norm sample means . According to our findings , symptoms of posttraumatic stress were comparable to those in cancer patients and fear of disease progression even exceeded scores found in cancer patients . Overall , psychological burden was less pronounced in securely attached patients . AE patients rarely received treatment for their depression or anxiety disorders according to German guidelines . Thus , our results show how important it is that AE patients are thoroughly assessed with regard to psychological symptoms and mental disorders , in order to give psychosocial support and supply treatment to those in need .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "neuropsychiatric", "disorders", "medicine", "and", "health", "sciences", "psychotherapy", "anxiety", "disorders", "post-traumatic", "stress", "disorder", "tropical", "diseases", "social", "sciences", "parasitic", "diseases", "health", "care", "mental", "health", "therapi...	2019	Psychological burden and resilience factors in patients with Alveolar Echinococcosis – A cross-sectional study
Spider silk fibers are produced from soluble proteins ( spidroins ) under ambient conditions in a complex but poorly understood process . Spidroins are highly repetitive in sequence but capped by nonrepetitive N- and C-terminal domains ( NT and CT ) that are suggested to regulate fiber conversion in similar manners . By using ion selective microelectrodes we found that the pH gradient in the silk gland is much broader than previously known . Surprisingly , the terminal domains respond in opposite ways when pH is decreased from 7 to 5: Urea denaturation and temperature stability assays show that NT dimers get significantly stabilized and then lock the spidroins into multimers , whereas CT on the other hand is destabilized and unfolds into ThT-positive β-sheet amyloid fibrils , which can trigger fiber formation . There is a high carbon dioxide pressure ( pCO2 ) in distal parts of the gland , and a CO2 analogue interacts with buried regions in CT as determined by nuclear magnetic resonance ( NMR ) spectroscopy . Activity staining of histological sections and inhibition experiments reveal that the pH gradient is created by carbonic anhydrase . Carbonic anhydrase activity emerges in the same region of the gland as the opposite effects on NT and CT stability occur . These synchronous events suggest a novel CO2 and proton-dependent lock and trigger mechanism of spider silk formation . Spider silk fibers contain regions of crystalline and noncrystalline β-sheets , which mediate mechanical stability [1] . In contrast , the soluble spidroins ( dope ) stored in the tail and sac of major and minor ampullate silk glands [2] exhibit unordered and helical conformations [3] . How spiders rapidly convert the dope into a solid fiber at a defined site of the S-shaped duct has been extensively studied [4]–[8] , but major questions are unresolved: First , how is the pH gradient in the gland generated and maintained ? Second , what is the pH at the phase transition in the duct ? The pH in the major ampullate gland has been shown to decrease from 7 . 2 in the proximal parts of the sac to 6 . 3 in the beginning of the duct [7] , but it has also been proposed that the gradient goes from 6 . 9 in the sac to 6 . 3 in the third limb of the duct [6] . Third , how are the terminal domains affected by the conditions in the duct at a molecular level , and in particular , do they , as proposed [4] , [5] , act in similar manners ? Documented pH-dependent effects at a molecular level include that the N-terminal domain ( NT ) dimerizes at pH 6 [9]–[11] , but pH-induced structural changes of the C-terminal domain ( CT ) have only been observed at pH 2 [4] . Here we address these questions and unravel novel physiological mechanisms for regulated spider silk formation . By use of concentric ion selective microelectrodes ( ISMs ) [12] we determined the pH in the major ampullate gland of Nephila clavipes , from the proximal part of the tail to the middle part of the second limb of the duct . Concentrations of CO32− were also determined at locations where pH was high enough to allow reliable measurements , and used to calculate HCO3− concentrations . We found that the pH decreases from 7 . 6±0 . 1 ( n = 11 ) in the proximal tail to 5 . 7±0 . 0 ( n = 6 ) in the second limb of the duct and that HCO3− concentration increases from 5 mM in the proximal tail to 21 mM in the distal part of the sac ( Figure 1 and Table 1 ) . With these values in the Henderson–Hasselbalch equation , the carbon dioxide pressure ( pCO2 ) could be calculated and was found to increase along the gland ( Figure 1 ) . We observed that the intraluminal pH at different locations did not change despite superfusion of the gland with an elevated pCO2 . This indicates that the epithelium of the major ampullate gland does not allow permeation of CO2 , a phenomenon previously described for parietal and chief cells in gastric glands [13] . The concentrations of K+ , Na+ , and Cl− in the sac were determined to be 6 , 192 , and 164 mM , respectively , using concentric ISMs ( Table 1 ) . The observation of simultaneously decreasing pH and increasing HCO3− and CO2 concentrations from the proximal to the distal parts of the gland ( Figure 1 ) suggested that carbonic anhydrase ( CA ) could be involved through catalysing the conversion of H2O + CO2 ↔H++ HCO3− . By use of a histochemical method [14] we could indeed identify abundant CA activity in intracellular vesicles and at the apical cell membrane of the epithelium in the distal part of the major and minor ampullate sacs and ducts , as well as in aggregate gland ducts and tubuliform glands ( Figure 2A–E ) . The site in the major ampullate epithelium where CA was found to emerge ( Figure 2A ) exactly coincides with the location where the glandular epithelium ceases to produce spidroins [15] . To investigate whether CA is responsible for generating and maintaining the pH gradient , we immersed freshly dissected N . clavipes major ampullate glands in buffers containing methazolamide , a membrane-permeable CA inhibitor [16] . Exposure to methazolamide collapsed the pH gradient , and pH levelled out to approximately 7 in the tail and sac . The gradient could subsequently be restored by removing the methazolamide ( Table 2 ) . Thus , the pH gradient in the major ampullate gland is dependent on active CA . Because CA activity was found in the epithelium of the distal major ampullate duct ( Figure 2E ) , where also proton pumps are present [17] , the pH may well continue to drop along the entire duct—that is , below pH 5 . 7 now measured half-way through the duct . This needs to be experimentally verified , as the extremely small inner diameter in the second half of the duct ( <20 µm ) did not allow measurements with the currently used ISMs . To address the third unresolved question—that is , how the terminal domains are affected by the conditions in the duct at a molecular level—we first compared the in vitro structural stability of NT and CT in the broad pH gradient now observed . We studied isolated domains , and it may be that these domains behave differently in their natural context of full-length spidroins . However , we have observed that NT followed by five repeats behaves as the isolated domain in terms of pH-dependent dimerization [11] . Urea and temperature denaturation studies at different pH values were performed for recombinant NT and CT ( Figures 3 and 4 ) . The stability of NT towards urea remained largely unchanged between pH 7 . 5 and 6 . 5 , but was significantly increased between 6 . 0 and 5 . 0 ( Figure 3 ) . We here analyzed a minor ampullate spidroin ( MiSp ) NT , which has not been studied before , but a similar pH effect was recently shown for a major ampullate spidroin ( MaSp ) NT [11] . This indicates that the structural effects now observed are applicable to spidroins from major and minor ampullate glands , in concordance with the observation of CA in major and minor ampullate , aggregate , and tubuliform glands ( Figure 2 ) . A similar effect as seen for stability towards urea was seen for NT thermal stability; that is , it was increased at lower pH ( Figure 4 ) . Dimerization of NT is completed at pH 6 [11] , and the subsequent stabilization of NT dimers between pH 6 and 5 ( Figure 3 ) may result in the firm locking of spidroins into multimers in the distal part of the duct ( cf . , Figure 1 ) . CT , in sharp contrast to NT , was gradually destabilized towards urea ( Figure 3 ) and temperature ( Figure 4 ) when pH was lowered from 7 . 5 to 5 . 0 . Heteronuclear single quantum coherence ( HSQC ) nuclear magnetic resonance ( NMR ) spectra of CT showed a folded structure at pH 6 . 8 , whereas a gradual conversion to an unfolded state was observed at a pH below 5 . 5 , and at pH 5 . 0 , it is completely unfolded ( Figure 5 ) . Moreover , we observed that CT irreversibly converted from α-helical to β-sheet structure upon thermal denaturation at pH 5 . 5 , but not at pH 6 . 5 or 7 . 5 ( Figure 6 and Table S1 ) . The fact that NMR spectroscopy of CT shows an unfolded state at pH 5 . 0 ( Figure 5 ) whereas circular dichroism ( CD ) spectroscopy and urea denaturation shows residual structure at pH 5 . 0 ( Figure 3 ) may be explained by the different CT concentrations ( 0 . 3 mM versus 5 µM ) and recording times ( hours versus minutes ) used . It should also be pointed out that unfolded species should have increased NMR intensities ( and may thus be overestimated relative to folded species ) due to favorable relaxation and dynamic properties and that helical structure ( which is observed by CD ) may be present in the species that are observed as random coil/unfolded by NMR . Denaturation of NT , in contrast , resulted in mainly unordered structure and was reversible at all three pH values ( Figure 6 and Table S1 ) . It may be worth noting that the structural conversion now observed for CT , but not for NT , resembles that seen for the spidroin dope [18] , which may be relevant for the trigger mechanism as discussed below . Next , we used hydrogen-deuterium exchange mass spectrometry ( HDX-MS ) to study the backbone conformational dynamics of CT at pH 7 . 5 to 5 . 5 . No major differences in HDX were seen between pH 7 . 5 and 6 . 5 , but helices 2 , 3 , and 5 showed increased HDX at pH 5 . 5 compared to at pH 6 . 5 ( Figure 7 ) , indicating increased structural flexibility at lower pH . Previous studies of CT [4] , [19] have identified a strictly conserved salt bridge between an Arg residue in helix 2 and a Glu residue in helix 4 . The NMR structure of Araneus ventricosus MiSp CT now studied ( Figure 8 and Table S2 ) is very similar to those of MiSp CT from Nephila antipodiana [19] and MaSp CT from A . diadematus [4] with backbone root-mean-square deviations ( RMSDs ) of 2 . 4 Å and 3 . 4 Å , respectively ( over 202 residues from both chains; see Figure 8 ) . Largest differences are observed for the N-terminal helix , which is shorter , and the C-terminal helix , which is kinked near the C-terminus in the A . ventricosus MiSp CT structure . A salt bridge between Arg38 in H2 and Glu82 in H4 is indeed found in A . ventricosus MiSp CT ( Figure 8 ) . Computational pKa predictions [20] of the available CT structures uniformly suggested that the Glu residue in H2 ( that participates in the saltbridge ) has a pKa ≥6 , making it possible to protonate in the pH interval now observed in the gland , and mutations interfering with this salt bridge greatly destabilize CT [4] , [19] . Our results suggest that protonation of the conserved Glu in H2 is involved in pH-dependent unfolding of CT in spider silk glands , and further experimental studies are warranted to determine exactly what residues are protonated in CT at low pH . Although the NMR structures of several CTs from different spidroins have been solved and their biochemical properties have been studied , the now observed pH responsive behavior of this domain has not been investigated in detail before [4] , [5] , [19] , [21]–[23] . The shared overall fold suggests a conserved function of CT , but the possibility that CT has diverse functions in different silks cannot be excluded and is an important topic for further studies . The conditions now determined for the distal parts of the gland—that is , low pH combined with increasing HCO3− concentration and low CO2 permeability of the gland—imply that pCO2 is elevated along the sac and duct . For MaSp CT , it has been shown that shear forces induce conformational changes that result in increased exposure of nonpolar surfaces [4] , and CO2 interacts mainly with nonpolar regions in proteins [24] , [25] . Therefore , we used the CO2 analogue CS2 [24] to identify potential interaction sites in the NMR structure of A . ventricosus MiSp CT . CS2 interacts specifically with a few , mainly hydrophobic , CT residues distributed in helices 2–4 , of which many are partly buried ( Figure 9A–D ) . NT on the other hand shows weak interactions with CS2 and only at conditions that favor the monomeric form , at pH 7 . 2 and 200 mM salt ( Figure 10 ) , which is characteristic to parts of the gland where pCO2 is low ( Figure 1 ) . In contrast to CT , no specific interactions between NT and CS2 were found at pH 5 . 5 ( Figure 10 ) , suggesting that NT stabilization at low pH ( Figure 3 and Figure 4 ) protects its hydrophobic , buried residues from interacting with CO2 . Amyloid fibrils are β-sheet polymers formed from ( partly ) unfolded proteins in a nucleation-dependent reaction and are found in tissue deposits associated with disease but also in some functional protein aggregates [26] . Amyloid fibrils share similarities with the β-sheets of spider silk and have been observed in the distal third of the spinning duct by electron microscopy ( EM ) , and it was proposed that the spidroin repetitive parts are responsible for the amyloidogenic behavior [27] . The poly-Ala segments of spidroins need to rapidly form β-sheet structure in silk formation , although Ala is highly prone to form α-helices [28] , raising the question , What nucleates this process ? We investigated whether CT may convert to amyloid-like fibrils at low pH by measuring Thioflavin T ( ThT ) fluorescence of CT over time at different pH values . When ThT binds to β-sheet polymers in amyloid-like fibrils , it gives an increased fluorescence [29] . At pH 5 . 5 and below , CT converted to a ThT-positive state , which was not observed at higher pH , or for NT at any pH tested ( Figure 11A ) . Analysis of the ThT-positive aggregates by transmission EM showed typical amyloid-like fibrils , 5–10 nm thick , elongated and nonbranched ( Figure 11B ) . Only samples of CT incubated at pH 5 . 5 showed the presence of amyloid-like fibrils . Furthermore , the CT fibrils were positive for Congo red staining and showed green birefringence under polarized light ( Figure 11C ) , another hallmark of an amyloid-like fiber [30] . The spidroins' terminal domains are highly conserved , both between species and between different types of silks [31] , which suggest that they play important roles in spider silk formation rather than for the silks' mechanical properties . Further supporting the hypothesis of general polymerization mechanisms between different types of silks , CA is found in the distal parts of several different spider silk glands and occur at the same location as the observed structural changes of NT and CT will take place provided that their behavior in vitro is recapitulated in vivo . NT and CT are unique to spidroins and there are no known homologues . The lock ( accomplished by NT ) and trigger ( accomplished by CT ) mechanism proposed herein is therefore likely unique for spider silk formation , in contrast to the previously identified shear-induced polymerization mechanism that also apply to , for example , silk worm silk formation [32] . A detailed understanding of the natural spinning process will be vital for the development of a spinning process capable of generating truly biomimetic spider silk fibers and may provide novel insights into Nature's way of confining amyloid fibril formation to a specific location . In summary , the spidroin N- and C-terminal domains show synchronous and opposite structural changes in response to the physiological conditions of the spinning duct . CT unfolds into β-sheet nuclei that can trigger rapid polymerization of the spidroins , whereas gradually locked NT dimers alleviate the need for rapid diffusion [11] , [33] , firmly interconnect the spidroins , and allow for propagation of pulling forces along the peptide chains . These events are driven by CO2 and proton gradients that ensure temporal and spatial confinement of the divergent structural changes of CT and NT . This novel lock and trigger mechanism elegantly explains how silk formation can occur at a very high speed , more than 1 m/s [34] , and at the same time be confined to the very distal part of the spinning duct . Concentric ISMs [12] were used to measure the concentrations of hydrogen , carbonate , sodium , potassium , and chloride ions . Thin-walled borosilicate glass capillaries of two different diameters were used for construction of concentric ISMs . The capillary forming the outer barrel ( outer diameter 2 . 0 mm , inner diameter 1 . 5 mm , A-M Systems 6185 ) was pulled to a tip diameter of 2–4 µm using a Flaming/Brown micropipette puller ( Sutter Instrument Co . US , Model P87 ) . The tip of the outer barrel was silanized by back-filling with N , N-dimethyltrimethylsilylamine ( Fluka 41716 ) , after which the barrel was mounted on a micromanipulator and heated using a hot air gun giving temperatures of 200–300°C for 60 s . Ion-selective cocktails for H+ ( Fluka 95291 ) , CO32− ( described by Chesler et al . ) [35] , Na+ , K+ , and Cl− were sucked into the tip to form a 100 to 200 µm long column , and a backfilling solution ( pH electrode , 150 mM NaCl pH 7 . 4; CO32− electrode , 10 mM NaHCO3 , 150 mM NaCl ) was added in the middle of the outer barrel . The inner barrel ( outer diameter of 1 . 2 mm and inner diameter of 0 . 9 mm , A-M Systems 6160 ) was pulled to a tip diameter of 1 µm and filled with 3 M KCl pH 7 . 4 . The inner barrel was then inserted into and secured in the outer barrel , the inner glass tip being positioned 4–10 µm away from the outer barrel tip . A silver wire was inserted into the inner barrel and connected to an amplifier . The ISMs were calibrated using pH 6 . 87 and pH 7 . 42 buffers , 50 , 100 , 200 , and 400 mM Na+ or Cl− , or using 1 , 2 , 4 , and 8 mM K+ , respectively . Carbonate electrodes were calibrated as described [35] . Adult female N . clavipes collected in Florida from September to November were kept in individual containers and fed water . Spiders were anaesthetized with CO2 gas before severing at the pedicle . Dissection of the major ampullate glands was carried out in a modified spider Ringer [36] ( with 2 mM MgCl2 , 2 mM CaCl2 , 3 mM KCl , and 10 mM glucose ) buffered with 26 mM bicarbonate and 5% CO2 , yielding a pH of 7 . 4 . Major ampullate glands were mounted in a submersion-style incubation chamber and superfused with HCO3− and CO2-buffered modified spider Ringer at room temperature . ISM measurements were performed in triplicates in different parts of the gland . The difference in potential between the bath and the inside of the gland was recorded on a chart recorder ( Zipp and Konnen ) and later translated into change in concentration of the ion of interest using the Nernst equation ( H+ , Na+ , K+ , Cl− ) or a modified Henderson–Hasselbalch equation [35] ( CO32− ) to get the concentration of HCO3− . Determined pH values and HCO3− concentrations were used to calculate pCO2 according to the Henderson–Hasselbalch equation , assuming equilibrium . To study the influence of CA activity on the pH gradient , some glands were incubated for 1 h in 0 . 1 mM methazolamide ( M4156 , Fluka ) , a membrane-permeable CA inhibitor , prior to pH measurements , after which the methazolamide was washed away for 30 min and pH measurements repeated . Some glands were subjected to CO2 permeability studies . Glands were dissected , mounted , and superfused with HCO3− and CO2-buffered spider Ringer at room temperature as described above . A pH electrode was inserted into the gland , after which the surrounding Ringer solution was buffered by 26 mM bicarbonate and 100% CO2 . pH measurements were continued up to 1 h to see if intraluminal pH changed in response to the elevated pCO2 surrounding the gland . The Ringer solution was then changed again , being buffered by 26 mM bicarbonate and 5% CO2 , yielding a pH of 7 . 4 , after which the pH electrode was removed from the gland and put in the Ringer and pH was recorded . This was made to ensure that the electrode had not been drifting . Spiders ( A . diadematus , N . clavipes , E . australis , and Tegenaria sp . ) were anesthetized and sacrificed as described above . Dissection was carried out in 67 mM sodium phosphate buffer at pH 7 . 2 or in a modified Spider Ringer ( see above ) . Some opisthosomas were fixed and embedded directly after removal of the exoskeleton , whereas others were dissected so that the major and minor ampullate glands could be isolated before fixation . Tissues for histochemical localization of CA activity were immersion fixed in 2 . 5% ( v/v ) glutaraldehyde in 67 mM phosphate buffer , pH 7 . 2 , for 24 h at 4°C and subsequently rinsed with phosphate buffer , pH 7 . 2 . After fixation , tissues were dehydrated using increasing concentrations of ethanol , infiltrated and embedded in a water-soluble glycol methacrylate ( Leica Historesin embedding kit ) . Historesin embedded major and minor ampullate glands and opisthosomas were sectioned at 2 µm in a microtome ( Leica RM 2165 ) and stained for CA activity using a histochemical method [14] . The method involves incubation of sections in a medium containing NaHCO3 , CoSO4 , H2SO4 , and KH2PO4 , whereby carbon dioxide leaves , pH increases , and a cobalt–phosphate–carbonate complex is formed at sites with CA activity . This complex is then converted into a black cobalt–sulphide precipitate . The sections were counterstained with Azure blue . For control of unspecific staining , the CA inhibitor acetazolamide was included in the incubation medium . A . ventricosus MiSp NT and CT coding gene fragments corresponding to ( NT: GSGNSQPIWT NPNAAMTMTN NLVQCASRSG VLTADQMDDM GMMADSVNSQ MQKMGPNPPQ HRLRAMNTAM AAEVAEVVAT SPPQSYSAVL NTIGACLRES MMQATGSVDN AFTNEVMQLV KMLSADSANE VST ) and ( CT: GSGNSTVAAY GGAGGVATSS SSATASGSRI VTSGGYGYGT SAAAGAGVAA GSYAGAVNRL SSAEAASRVS SNIAAIASGG ASALPSVISN IYSGVVASGV SSNEALIQAL LELLSALVHV LSSASIGNVS SVGVDSTLNV VQDSVGQYVG ) were amplified by PCR with the full-length MiSp gene as template [37] , cloned into a modified pET vector ( resulting in the target proteins being fused to His tag–Thioredoxin–His tag followed by a thrombin cleavage site ) and transformed into BL21 ( DE3 ) Escherichia coli . The E . coli were grown at 37°C in LB medium containing 70 mg/l kanamycin until OD600 was about 0 . 9 . The temperature was lowered to 30°C , IPTG was added to a final concentration of 0 . 3 mM , and the cells were incubated for about 4 h . The E . coli were then harvested by centrifugation at 6 , 400×g for 20 min at 4°C ( Sorvall RC 3BP+ , 500 ml flasks ) , after which the pellet was resuspended in 20 mM Tris pH 8 . 0 , 1 mg/ml lysozyme was added , and the solution was incubated on ice for 30 min . Next , DNase and MgCl2 were added and the mixture was kept on ice for 30 min . The cell lysate was centrifuged ( 27 , 000×g ) at 4°C for 20 min ( centrifuged as above , 50 ml tubes ) . For purification of CT , the supernatant was loaded on a Ni-NTA column and the fusion protein was eluted with 300 mM imidazole . For purification of NT , which is mainly found in the pellet after lysis , pellets were resuspended in 20 mM Tris pH 8 . 0 containing 2 M urea , sonicated for 2 min , and the supernatant was treated as for CT . The fusion proteins were then dialyzed against 20 mM Tris pH 8 . 0 overnight at 4°C , cleaved by 1/1 , 000 ( w/w ) thrombin , and run over a Ni-NTA column to remove the fusion tag . This resulted in essentially pure NT or CT ( >90% purity as determined by SDS PAGE gel electrophoresis and Coomassie staining ) . For NMR structure determination , we initially expressed a 150-amino-acid-residue-long C-terminal part of A . ventricosus MiSp ( full-length sequence above ) . The expressed protein was labeled with 15N , and the NMR spectrum showed that the first 25 residues adopt random coil fold . Therefore , A . ventricosus MiSp CT was truncated and residues 31–150 ( marked in bold in the sequence above ) were expressed in minimal medium and labeled by 15N and 13C/15N . The NMR sample was prepared by adding 8% ( v/v ) D2O and 0 . 02% ( w/v ) NaN3 to a 1 mM solution of uniformly 13C/15N-labelled protein in 20 mM sodium phosphate buffer ( pH 6 . 8 ) with 20 mM NaCl . All NMR experiments were carried out at 298 K on a 600-MHz Varian Unity Inova spectrometer equipped with an HCN triple-resonance pulsed-field-gradient cold probe . The following 2D and 3D spectra were acquired for backbone resonance assignment ( number of complex points given in parentheses ) : [15N-1H]-HSQC ( 1024×128 ) , HNCA ( 1024×24×40 ) , CBCA ( CO ) NH ( 2048×48×40 ) , HNCO ( 1024×24×40 ) , HN ( CA ) CO ( 1024×24×40 ) , and for side-chain assignment and NOE restraint collection ( mixing time given in parentheses ) : 15N-resolved NOESY-HSQC ( 1024×38×150 , 60 ms ) , 13C ( aliphatic ) -resolved NOESY-HSQC ( 768×52×150 , 60 ms ) , and 13C ( aromatic ) -resolved NOESY-HSQC ( 1024×16×150 , 60 ms ) . Additionally , in order to identify intermolecular NOEs , a 13C/15N-filtered 13C ( aliphatic ) -resolved NOESY-HSQC spectrum ( 768×34×70 , 60 ms ) was recorded on a sample containing 50% 13C/15N-labelled and 50% unlabelled proteins [38] that was prepared by mixing equal amounts of labeled and unlabelled proteins in 8 M urea followed by dialysis against the NMR sample buffer . The same sample was afterwards used to probe interactions with CS2 . Aliquots of 20% CS2 in DMSO were added in a stepwise manner to the NMR sample of CT , yielding CS2 concentrations of 50 mM , 100 mM , and 200 mM , and a 2D [15N-1H]-HSQC spectrum was recorded each time . To account for perturbations due to DMSO , a reference experiment was performed by adding DMSO only in the same amounts . CS2-induced chemical shift perturbations were calculated by comparing the spectrum at 200 mM CS2 with the spectrum at the end of the reference titration with DMSO , and using the formula ( ) [39] . All spectra were processed with Bruker TopSpin 3 . 1 and analyzed using CARA [40] . The assigned chemical shifts have been deposited in BioMagResBank ( accession number 19579 ) . To probe interactions between NT and CS2 , NT from MaSp1 from E . australis was expressed and purified as previously described [10] . Chemical shift perturbations of MaSp NT backbone amides were determined at pH 7 . 2 and 200 mM NaCl and at pH 5 . 5 upon addition of CS2 ( 0 to 200 mM ) as described for CT . For 2D [15N-1H]-HSQC NMR spectra of MiSp CT , samples at pH 5 . 0 , 5 . 3 , and 5 . 5 were prepared by diluting 50 µl of a concentrated stock solution of MiSp CT in 20 mM sodium phosphate buffer , 20 mM NaCl , 0 . 03% NaN3 , pH 6 . 8 with 200 µl of 100 mM CD3COOD/CD3COONa , 20 mM NaCl , 0 . 03% NaN3 buffer , and adding 20 µl of D2O . Automated peak picking of the three NOESY spectra was performed using UNIO-ATNOS/CANDID 2 . 0 . 2 [41] . Distance restraints were obtained from these peak lists using the internal NOE calibration procedure of CYANA 2 . 1 [42] . Intermolecular contacts were identified by analysis of the 13C , 15N-filtered NOESY spectrum and used as distance restraints with an upper limit of 5 Å . No explicit torsion-angle restraints were used in the input . Structure calculations were performed using CYANA 2 . 1 [42] and involved seven iterations of automated NOE assignment with the routine CANDID [41] followed by a simulated annealing procedure starting in the first cycle from a homology model generated based on the MiSp CT structure from N . antipodiana [19] ( PDB accession code 2M0M ) that was annealed in 15 , 000 steps of torsion-angle dynamics . This approach was used to reduce the assignment ambiguity during the first cycles of the automated NOE assignment and resulted in significantly more unambiguous distance restraints in the final cycle of the calculation concomitantly with a lower target function value . The 20 conformers with the lowest residual restraint violations were energy minimized in a water shell using the program CNS 1 . 2 [43] , [44] , and their coordinates were deposited in PDB ( accession code 2MFZ ) . Table S2 shows an overview of the restraints used and structural statistics . Ramachandran statistics for structured part ( residues 20–120 ) are 94 . 2% most favored , 5 . 8% additionally allowed regions; for all residues including the unstructured N-terminal tail , 88 . 3% most favored , 11 . 1% additionally allowed , 0 . 3% generously allowed , and 0 . 3% disallowed regions . For analysis of amyloid fibril formation , 10 µM of A . ventricosus MiSp NT and CT were incubated under quiescent conditions at 25°C with 10 µM ThT in 20 mM sodium phosphate or 50 mM sodium acetate buffer with or without 154 mM NaCl at different pH values between 5 . 0 and 7 . 5 . ThT fluorescence was recorded on a BMG FLUOstar Galaxy plate reader using bottom optics in 96-well polyethylene glycol-coated black polystyrene plates with a clear bottom ( Corning Glass , 3881 ) using a 440-nm excitation filter and a 490-nm emission filter . For analysis of amyloid fibrils , 10 µM of A . ventricosus MiSp NT and CT were incubated overnight ( 12–16 h ) under quiescent conditions at 25°C in 20 mM sodium phosphate buffers with or without 154 mM NaCl at pH 7 . 5 , 6 . 5 , and 5 . 5 , respectively . Samples were incubated overnight and 2 µl were adsorbed on copper grids , stained with 2 . 5% uranyl acetate in 50% ethanol for about 20 s , and examined and photographed with a Hitachi H7100 electron microscope at 75 kV . Ten µM A . ventricosus MiSp CT was incubated at 37°C with shaking ( 250 rpm ) for 2 . 5 h at pH 5–7 in 20 mM sodium phosphate and 50 mM sodium acetate buffers , respectively . Samples were centrifuged , supernatant removed , washed with dH2O , and then centrifuged again . The supernatant was removed and 10 µl dH2O was added , the sample was vortexed , and droplets ( 0 . 8 µl ) were applied to microscopical slides , air dried , and stained with Congo red B [45] . After mounting under coverslips , the materials were examined in a polarization microscope for Congophilia and green birefringence . A . ventricosus MiSp NT and CT stability between pH 5 . 0 and 7 . 5 with and without 154 mM NaCl was determined by urea denaturation . Like in previous denaturation studies of MiSp CT from N . antipodiana [19] , we used a two-state model for analyzing our denaturation data . Although a two-state transition is supported by a CD isodichroic point at 203 nm [46] for NT at low pH ( Figure 4A ) , this is not the case for CT at any pH , or NT at pH 7 . 5 ( Figure 4B ) . To emphasize that we assumed a two-state transition for both NT and CT , the urea concentrations derived from fitting the data to a two-state unfolding model are referred to as apparent half-denaturation ( [den]50% ) . Notably , the main conclusion from these experiments—that NT and CT respond in completely opposite ways to lowered pH—is not dependent on whether a two-state transition applies or not . For NT , urea-induced denaturation was performed by diluting the protein to 5 µM in 20 mM HEPES/20 mM MES with 0–7 M urea in 0 . 25 M steps . Tryptophan fluorescence emission spectra were measured on a spectrofluorometer ( Tecan Safire 2 ) using Costar black polystyrene assay plates with 96 flat bottom wells . The samples were excited at 280 nm using a 5 nm bandwidth , and emission spectra were recorded in 1 nm steps between 300 and 400 nm using a 10 nm bandwidth . Spectra were recorded at constant pH values ranging from 5 . 0 to 7 . 5 with 0 . 2–0 . 5 unit steps . For CT , CD spectroscopy at 222 nm was used to determine [den]50% as a function of pH . The CT samples were diluted to 7 . 5 µM in 20 mM sodium phosphate buffer and ran with 0–7 M urea in 0 . 25 M steps . At each pH , the average 222 nm CD ellipticity from three temperature scans for different urea concentrations were obtained with the settings described below ( see CD spectroscopy ) . The ellipticities for each measured pH values ranging from 5 . 0 to 7 . 5 with 0 . 2–0 . 5 unit steps were plotted against the urea concentration and fitted to a two-state unfolding model in order to determine the [den]50% by KaleidaGraph . CD spectra were recorded from 260 to 190 nm at 25°C in 0 . 1 mm path length quartz cuvettes using an Aviv 410 Spectrometer . The wavelength step was 0 . 5 nm , averaging time 0 . 3 s , scan speed 20 nm/min , time constant 100 ms , and bandwidth 1 nm . The spectra shown are subtracted for background and averaged over three consecutive scans . The HT voltages were always below 600 V during the entire scans . Spectra of 7 . 5 µM A . ventricosus MiSp NT ( 110 µg/ml ) or CT ( 90 µg/ml ) in 20 mM sodium phosphate buffer at pH 7 . 5 , pH 6 . 5 , or pH 5 . 5 were recorded at 25 , 45 , 65 , 85 , and 95°C and at 25°C again after cooling . For temperature melting curves , the CD at 222 nm was monitored between 25 and 95°C with 1°C/min increase . Deuteration buffers were prepared by freeze-drying 200 µl of 20 mM sodium phosphate buffer , pH 5 . 5 or 6 . 5 , followed by reconstitution in 200 µl D2O ( Cambridge Isotopes ) . A . ventricosus MiSp CT was diluted from 555 µM stock solution , pH 6 . 5 , to 55 . 5 µM in deuterated phosphate buffer , pH 6 . 5 or 5 . 5 . We removed 19 . 5 µl aliquots after 400 s , 50 min , 100 min , 200 min , or 300 min ( pH 5 . 5 ) or after 40 s , 5 min , 10 min , 20 min , or 30 min ( pH 6 . 5 ) . Aliquots were placed in prechilled 500 µl Eppendorf tubes containing 0 . 5 µl 5% TFA ( Merck ) , vortexed , and immediately frozen in liquid nitrogen . For a fully deuterated control , CT was incubated in deuterated phosphate buffer , pH 6 . 5 , for 24 h at 25°C . Samples were stored at −80°C until ESI MS analysis . Samples were thawed and immediately injected into an HPLC system using a chilled 25 µl Hamilton syringe . CT protein was digested in a Porozyme pepsin cartridge ( Applied Biosystems ) , and peptides were trapped and desalted in a Waters Symmetry C18 trap column ( Waters ) . Two 140D solvent delivery systems ( Applied Biosystems ) were employed , operating at 20 µl/min ( for washing with 0 . 05% TFA ) or at 15 µl/min ( for elution with 70% acetonitrile , 0 . 2% formic acid ) . Digestion and desalting were carried out in a single step for 10 min , and peptides were then eluted in a single step and delivered to the mass spectrometer via a TaperTip emitter ( Proxeon ) . The entire flow system was submerged in an ice bath . ESI spectra were acquired in positive-ion mode with a Waters Ultima API mass spectrometer ( Waters ) equipped with a Z-spray source . The source temperature was 80°C , the capillary voltage was 2 . 5 kV , and the cone and radiofrequency lens 1 potentials were 100 and 38 V , respectively . The mass spectrometer was operated in single-reflector mode to achieve a resolution of 10 , 000 ( full width at half maximum ) . The mass scale was calibrated using [Glu1]fibrinopeptide B . Peptic peptides were identified based on a map of pepsin-digested undeuterated CT using automated liquid chromatography–tandem mass spectrometry ( LC-MS/MS ) analysis with a Waters NanoAcquity system ( Waters ) . Peptide sequences were identified by individual analysis of collision-induced dissociation ( CID ) spectra using the Waters MassLynx and ProteinLynx software packages ( Waters ) . We analyzed 100 µl of 1 mg/ml A . ventricosus MiSp CT equilibrated 10 min in 20 mM HEPES/MES pH 7 . 5 or 5 . 5 using Sephacryl S-100 ( GE Healthcare ) in the same buffers and at a flow rate of 0 . 5 ml/min . Molecular mass standards aprotenin ( 6 . 5 kDa ) , ribonuclease ( 13 . 7 kDa ) , CA ( 29 kDa ) , ovalbumin ( 43 kDa ) , and conalbumin ( 75 kDa ) were used for calibration .	The spinning process of spider silk is crucial for making webs or other complex constructions to catch spider's prey . The main components of the silk are spidroins , which are large and repetitive proteins that have conserved nonrepetitive terminal domains ( NT and CT ) . Spiders manage both to store the highly aggregation-prone spidroins in solution at extreme concentrations in the silk glands and then to rapidly convert these spidroins into a solid fiber within fractions of a second as they spin fibres . This process has been extensively studied and is thought to involve a pH gradient , but how this pH gradient is generated and maintained was not resolved . Here , we measured the pH at locations along the ampullate gland and determined that the pH decreases to 5 . 7 in the middle of the spinning duct . We also observed that the carbon dioxide pressure is simultaneously increased and that its accumulation may affect the stability of CT . We find that active carbonic anhydrase ( CA ) is crucial to maintain the pH gradient along the gland . Detailed molecular studies of NT and CT under the disparate conditions present along the gland revealed a lock and trigger mechanism whereby in more neutral pH conditions , precocious spidroin aggregation is prevented , and when in more acidic pH conditions , NT dimers firmly interconnect the spidroins and the CT unfolds into β-sheet nuclei that can trigger rapid polymerization of the spidroins . We conclude that this mechanism enables temporal and spatial control of silk formation and may be harnessed in attempts to produce artificial silk replicas .	[ "Abstract", "Introduction", "Results", "and", "Discussion", "Materials", "and", "Methods" ]	[ "biochemistry", "physiological", "processes", "proteins", "protein", "folding", "physiology", "protein", "structure", "biology", "and", "life", "sciences", "biophysics" ]	2014	Carbonic Anhydrase Generates CO2 and H+ That Drive Spider Silk Formation Via Opposite Effects on the Terminal Domains
Retrograde axonal transport requires an intricate interaction between the dynein motor and its cargo . What mediates this interaction is largely unknown . Using forward genetics and a novel in vivo imaging approach , we identified JNK-interacting protein 3 ( Jip3 ) as a direct mediator of dynein-based retrograde transport of activated ( phosphorylated ) c-Jun N-terminal Kinase ( JNK ) and lysosomes . Zebrafish jip3 mutants ( jip3nl7 ) displayed large axon terminal swellings that contained high levels of activated JNK and lysosomes , but not other retrograde cargos such as late endosomes and autophagosomes . Using in vivo analysis of axonal transport , we demonstrated that the terminal accumulations of activated JNK and lysosomes were due to a decreased frequency of retrograde movement of these cargos in jip3nl7 , whereas anterograde transport was largely unaffected . Through rescue experiments with Jip3 engineered to lack the JNK binding domain and exogenous expression of constitutively active JNK , we further showed that loss of Jip3–JNK interaction underlies deficits in pJNK retrograde transport , which subsequently caused axon terminal swellings but not lysosome accumulation . Lysosome accumulation , rather , resulted from loss of lysosome association with dynein light intermediate chain ( dynein accessory protein ) in jip3nl7 , as demonstrated by our co-transport analyses . Thus , our results demonstrate that Jip3 is necessary for the retrograde transport of two distinct cargos , active JNK and lysosomes . Furthermore , our data provide strong evidence that Jip3 in fact serves as an adapter protein linking these cargos to dynein . Active transport of proteins and organelles between the neuronal cell body and axon terminals is necessary for the formation and maintenance of functional neural circuits . Anterograde ( to axon terminals ) and retrograde ( to the cell body ) transport rely on motor proteins of the Kinesin and Dynein families respectively . These motors use the energy of ATP hydrolysis to walk along microtubule tracks , carrying cargo to its proper destination . Though 15 kinesin families exist in mammals [1] , only 1 retrograde microtubule based motor protein , cytoplasmic dynein , is responsible for the majority of retrograde cargo transport in axons [2]–[4] , leading to intriguing questions about the nature of dynein-cargo interaction specificity which have been largely unexplored [5] . The core cytoplasmic dynein motor is composed of an array of proteins that includes two motor domain-containing heavy chains , two intermediate chains , two light intermediate chains , and four light chains which bind the intermediate chains [6] . Though recombinant dynein heavy chain can function in microtubule sliding assays in vitro [7] , dynein complex interacting proteins have been shown to be essential for the initiation of retrograde cargo movement in vivo . Dynactin , a large dynein-interacting protein complex , and Lis1 have been separately shown to be co-factors that are necessary for the initiation of retrograde transport [8]–[10] . Loss of either of these factors leads to decreased retrograde transport frequency of some cargo and can lead to the accumulation of dynein components as well as cargo in axon terminals [9] . Retrograde cargo is thought to either bind directly to the core dynein complex proteins or , alternatively , to additional adapter proteins . It is tempting to speculate that the use of distinct adapter proteins may confer specificity to motor-cargo interactions in the dynein motor system . Despite their importance for the understanding of dynein-based cargo transport , the identity of specific dynein cargo adapters is dramatically lacking [5] . We used the advantages of the zebrafish system , including its amenity to forward genetics and live imaging , to identify Jip3 ( JNK-interacting protein 3 ) as a cargo-specific adapter for dynein-based axonal transport . Through a forward genetic screen , we isolated a mutant strain ( jip3nl7 ) that exhibited swellings in axon terminals of long sensory axons , a potential sign of interrupted retrograde transport . jip3nl7 carried a mutation in Jip3 , a scaffold protein shown previously to serve as an adapter and facilitator of synaptic cargo anterograde transport through its interaction with Kinesin-1 [11]–[13] . In addition to anterograde transport machinery , Jip3 interacts with components of the dynein motor complex and c-Jun N-terminal Kinase ( JNK ) . Indeed , Jip3 was first identified as a scaffold protein that links JNK to its upstream activating kinases , facilitating JNK activation [14] . Interestingly , Cavalli and colleagues demonstrated that Jip3 and activated JNK ( pJNK ) colocalized with p150glued ( dynactin complex protein ) distal to sciatic nerve injury . Based on this data , they postulated that Jip3-JNK-dynein interaction may be important during retrograde damage signaling [15] . Furthermore , in this and other studies , Jip3 has been shown to biochemically interact with components of the retrograde motor complex , specifically p150glued [15] and dynein light intermediate chain ( DLIC; [13] ) . Thus , an intriguing possibility is that Jip3 could serve as an adapter for dynein-mediated retrograde transport of JNK and other cargo; however , neither this hypothesis nor the possibility that Jip3 is required for retrograde transport of any cargo , has been directly addressed to date . Our work reveals discrete and direct roles for Jip3 in the retrograde transport of two cargos , pJNK and lysosomes . Using an in vivo imaging technique we developed for use in the zebrafish , we found specific retrograde transport defects in jip3nl7: frequencies of lysosome and pJNK retrograde transport were decreased causing accumulation of both cargos in axon terminals . Further analyses showed that direct Jip3-JNK interaction was necessary for retrograde clearance of pJNK from axon terminals and provided evidence that increased levels of pJNK were directly responsible for axon terminal swellings . Surprisingly , JNK activity and Jip3-JNK interaction had no impact on lysosome localization . Rather , co-transport analysis of lysosomes with both Jip3 and DLIC provided strong evidence that DLIC-lysosome interaction during retrograde transport relies on Jip3 . Thus , based on our data we posit that Jip3 serves as an adapter protein for the retrograde transport of two distinct cargos , pJNK and lysosomes , and that failed retrograde clearance of pJNK contributes to the dysmorphic axon terminals in jip3nl7 mutants . jip3nl7 was isolated in a forward genetics screen for which we utilized the TgBAC ( neurod:EGFP ) nl1 transgenic zebrafish ( hereafter referred to as neurod:EGFP; [16] ) . This transgenic strain expresses an EGFP reporter in the central and peripheral nervous systems , including the posterior lateral line ( pLL ) ganglion and the long sensory axons emanating from it ( Figure 1A , 1B; for screen details consult the Materials and Methods ) . We focused our screen on the long sensory axons of the pLL because of their planar character and superficial localization . These axons originate from the pLL ganglion , located just posterior to the ear , and extend along the trunk , branching to innervate mechanosensory hair cells that reside within surface sensory organs called neuromasts ( NMs; axon terminals innervating NM3 and terminal NMs are shown in Figure 1B′ and 1B″ respectively ) . Initial pLL nerve extension and NM formation is complete by 2 dpf ( days post-fertilization ) , and by 5 dpf a functional neural circuit has developed between NM hair cells and afferent pLL axons [17] . The recessive jip3nl7 mutant ( Figure 1C ) was isolated because it displayed truncation of pLL axons ( incomplete penetrance; Figure 1C″ ) and swollen axon terminals innervating all trunk NMs ( penetrance 100%; NM3 in Figure 1C′ and data not shown ) . To determine if long central nervous system axons were also affected by loss of Jip3 , we analyzed axons of the reticulospinal tract as well as the efferent axons that project from the CNS to innervate the pLL NMs by crossing the jip3nl7 mutation into the TgBAC ( phox2b:EGFP ) w37 transgenic line [18] . Similar to pLL afferents , both reticulospinal tract and pLL efferent axons were truncated in jip3nl7 mutants ( Figure 1D , 1E ) . jip3nl7 mutants were homozygous viable and the pLL axonal phenotype did not have a maternal component , as progeny derived from homozygous crosses displayed identical phenotypes to that of progeny derived from heterozygous crosses ( data not shown ) . We used a positional cloning approach to isolate the genomic locus containing the jip3nl7 gene mutation . Zebrafish Jip3 , which mapped to this locus , is similar to its mammalian orthologs and contains two coiled coil domains , one leucine zipper deemed integral for Kinesin Light Chain ( KLC ) and dynactin binding [19] , [20] , and a JNK binding domain ( Figure 1F ) . Sequencing of jip3 from jip3nl7 mutants revealed a mutation at nucleotide 552 which created a premature stop codon , truncating the Jip3 protein at amino acid 184 ( Figure 1F ) . In situ hybridization analysis showed that , similar to mouse [21] , jip3 was expressed in the central and peripheral nervous systems of the zebrafish embryo ( Figure 1G ) . jip3 expression was lost in jip3nl7 , perhaps due to nonsense-mediated mRNA decay ( Figure 1H ) . Consequently , jip3nl7 is likely a Jip3 null . Initial investigations revealed the pLL nerve phenotypes were not due to impaired pLL patterning , neuronal cell death , abnormal glial support/myelination , or gross cytoskeletal abnormalities ( Figure S1 ) . As Jip3 has been shown to interact with members of the anterograde and retrograde motor complexes [11]–[13] , [22] , [23] and interruptions in transport have been associated with axon swellings like those observed in jip3nl7 [24] , [25] , we next focused our investigations on the potential function of Jip3 in axonal transport . To study the function of Jip3 in axonal transport , we developed methods to visualize microtubule-based axonal transport in the pLL system in vivo , in intact zebrafish embryos and larvae ( Figure 1I ) . Zebrafish are ideal for such a preparation as they are transparent through early embryonic and larval development , facilitating in vivo live imaging , and transient transgenesis can be used reliably to express tagged cargos of interest mosaically . Using these advantages , we developed a protocol that requires no surgical or invasive techniques to visualize protein or organelle transport in the long and planar axons of the pLL . To image axonal transport in zebrafish pLL axons , zygotes are injected with DNA encoding a cargo of interest tagged with a fluorescent reporter . Expression of these constructs is controlled by a neuron-specific 5 kilobase portion of the neurod promoter ( 5kbneurod; [26] ) . This results in mosaic expression of the desired cargo in the pLL ganglion , which , in ideal preparations , labels 1 to 2 neurons . Neurons expressing cargo are then monitored for full axon extension , innervation of NMs , and the absence of cargo accumulation in neuronal cell bodies and axons to assess optimal concentrations of DNA for injection . Using this approach , cargo transport can be visualized in individual pLL axons during axon extension ( 1–2 dpf ) , post-extension ( after 2 dpf ) , and after functional synaptic connections are established ( 5 dpf ) . We first utilized this technique to observe the localization and transport of a Jip3-mCherry fusion in pLL neurons and their axons . During axon extension ( 30 hpf ) , Jip3-mCherry localized to the neuronal cell body and axon growth cones ( Figure 1J , 1K ) , similar to Jip3 localization in cultured neurons [27] . We then visualized Jip3 transport at 2 dpf , just after pLL nerve extension completes , and analyzed transport parameters using kymograph analysis ( Figure 1L , 1M and Video S1 ) . Jip3-containing cargo traveled at average velocities of 1 . 60 µm/sec in the anterograde direction and 1 . 35 µm/sec when moving in the retrograde direction ( N = 7 larvae ) ; these parameters are consistent with fast anterograde and retrograde transport [1] . Next , we assayed the localization and transport of ssNPY-mCherry [28] , a marker of Golgi-derived vesicles , to determine if loss of Jip3 affects the axonal transport of this generalized cargo . At 5 dpf , we observed large accumulations of mCherry positive puncta in axon terminals of jip3nl7 mutants but not in wildtype siblings ( Figure S2A , S2B; for this and other experiments , mutants were identified using the genotyping protocol described in the Materials and Methods , except where otherwise indicated ) . In vivo imaging and kymograph analysis demonstrated bidirectional movement of mCherry-positive puncta in wildtype and jip3nl7 mutants ( Figure S2C–S2F; Videos S2 and S3 ) with decreased frequency of anterograde and retrograde transport of this cargo in jip3nl7 at 2 dpf with a tendency toward a decrease at 5 dpf ( Figure S2G ) . Neither distance nor velocity of cargo movement were altered ( Figure S2H , S2I ) , potentially implicating Jip3 in cargo-motor attachment , rather than modulation of motor activity . Next , we set out to determine the identity of the mCherry labeled retrograde cargo ( s ) by looking for accumulation of commonly transported retrograde cargos in jip3nl7 axon terminals using immunofluorescence [29] , [30] . Neither late endosomes ( Rab7-positive ) nor autophagosomes ( LC3-positive ) accumulated in jip3nl7 axon terminals ( Figure S3A–S3D ) . Consistent with a previous study on Jip3's role in anterograde transport of TrkB [13] , TrkB levels were decreased in jip3nl7 axon terminals , as assayed by TrkB antibody labeling ( Figure S3E , S3F ) . In contrast , the axon terminal swellings in jip3nl7 were rich in lysosomes that were visualized using two separate markers , Lamp1 ( detected by immunofluorescence; Figure 2A , 2B ) and Lysotracker red ( vital dye; Figure S3G , S3H ) . We then asked whether abnormalities in lysosomal transport caused lysosome accumulations in axon terminals by employing our in vivo imaging approach , using a Lamp1-mTangerine fusion [31] to mark lysosomes in pLL axons ( Figure 2C–2F; see Videos S4 and S5 ) . The ability of a Lamp1-EGFP fusion construct to label lysosomes was confirmed by double labeling with the vital dye Lysotracker red ( Figure 2G ) . Similar to our immunolabeling results , Lamp1-mTangerine accumulated in the axon terminals of jip3nl7 mutants but not wildtype controls ( Figure 2E , 2F ) . Live imaging analysis demonstrated that , though Lamp1-mTangerine transport parameters were not altered at 2 dpf , the number of lysosomes moving in the retrograde direction was significantly decreased at 3 dpf in jip3nl7 axons ( Figure 2H–2J; WT = 15 . 08±2 . 71 vs . jip3nl7 = 5 . 14±2 . 71 particles/100 µmmin , p≤0 . 002; Wilcoxon rank-sum ) . A similarly reduced frequency of lysosome retrograde transport was also observed at 5 dpf , while distance and velocity of movement were largely unaffected at all stages ( Figure 2K , 2L ) . These data show that retrograde lysosome transport relies on Jip3 . Jip3 has been shown to interact with components of the Kinesin-1 motor to regulate anterograde transport [11]–[13] , but a role for Jip3 in retrograde transport has not been described previously . Therefore , we next sought to address how Jip3 functioned to regulate retrograde axonal transport . Jip3 was originally identified as a JNK-interacting protein and has been shown to facilitate JNK activation in vitro [14] . Thus , we would predict that loss of Jip3 would lead to decreased JNK activation . As JNK activity can impact numerous intracellular processes that could potentially affect axonal transport machinery [32] , [33] , we assayed levels and localization of active JNK ( pJNK ) using pan-pJNK immunolabeling . Surprisingly , instead of a decrease , we found elevated levels of pJNK in the mutant axon terminals innervating all NMs from 2 dpf onward ( Figure 3A–3I and data not shown; see Materials and Methods for an explanation of fluorescent intensity measurement ) . In contrast , total JNK ( tJNK ) levels in jip3nl7 were comparable to controls ( Figure 3J and Figure S4A–S4D ) . Western blot analysis of whole embryo extracts revealed no increase in overall tJNK or pJNK levels in jip3nl7 ( Figure S4E , S4F ) , pointing to a change in localization of pJNK rather than overall JNK expression or activity . Given the ability of Jip3 to bind components of the retrograde motor and pJNK [14] , [15] , we reasoned that Jip3 might directly mediate pJNK retrograde transport/clearance from axon terminals by attaching this active kinase to the dynein motor complex . To determine if Jip3 has a specific role in pJNK transport , we used two complimentary approaches . First , we developed an axon injury model for use in the zebrafish pLL nerve to indirectly assay pJNK transport , similar to a protocol previously used in mouse sciatic nerve ( Figure 4A; see Materials and Methods for procedure details; [15] ) . Following injury , cargos that are transported in the anterograde direction will accumulate proximal to the injury site , whereas retrograde cargos will accumulate distal to the injury site . Severing the pLL nerve between NM2 and NM3 at 5 dpf resulted in accumulation of pJNK in the pLL nerve proximal and distal to the site of injury in wildtype larvae by 3 hours post-injury . In contrast , pJNK failed to accumulate distal to the site of injury in jip3nl7 mutants ( Figure 4B–4E , 4J ) , indicating failed retrograde pJNK transport in mutant axons . Total JNK levels were not significantly different proximal or distal to injury site in jip3nl7 mutants ( Figure 4F–4I , 4K ) , though there was a strong trend towards decreased levels of the tJNK anterograde pool ( proximal to the injury site ) in jip3nl7 mutants . This data supports the hypothesis that loss of Jip3 inhibits pJNK retrograde transport , which would lead to accumulations of this kinase in axon terminals . Next , we asked whether dynein motor components were normally transported to axon terminals in jip3nl7 mutants , as the perturbation of this transport could indirectly affect retrograde cargo movement . Using immunolabeling for two components of the dynein complex ( Dynein heavy chain and p150glued ) , we demonstrated proper localization of these core dynein motor proteins to jip3nl7 mutants , confirming that the retrograde motor can reach axon terminals in jip3nl7 mutants ( Figure S5A–S5G ) . From this data , we can also infer that even in the absence of Jip3 , the initiation of dynactin-mediated , dynein movement was intact since these retrograde motor components did not accumulate in axon terminals [9] , [10] . Finally , we used our in vivo live imaging to concretely determine if retrograde JNK transport was impaired in jip3nl7 mutant pLL axons using transient expression of JNK3 tagged with mEos . We chose to use JNK3 for our in vivo analysis because Jip3 has been shown to bind most strongly to the JNK3 homolog [14] , and jnk3 is strongly expressed in the zebrafish nervous system ( Figure S6A , S6B ) . Phospho-JNK immunolabeling of embryos expressing JNK3-mEos driven by the 5kbneurod promoter in pLL axons demonstrated that a large portion of JNK3-mEos positive vesicles carried the active form of this kinase ( Figure 5A ) . Live imaging experiments revealed JNK3-mEos positive puncta traveled bidirectionally in wildtype and jip3nl7 mutants at 2 dpf ( Figure 5B , 5C; Videos S6 and S7 ) . Using kymograph analysis ( Figure 5D , 5E ) , we found a decrease in the number of JNK3-mEos positive puncta moving in the retrograde direction at 2 dpf in jip3nl7 mutants ( Figure 5F; wildtype:2 . 99±0 . 48 vs . jip3nl7:1 . 15±0 . 58 particles/100 µmmin , p≤0 . 05; Wilcoxon rank-sum ) while retrograde movement distance and velocity were largely unchanged ( Figure 5G , 5H ) . Taken together with the results from our injury model , these data confirmed that the frequency of retrograde pJNK transport was hindered in jip3nl7 mutants . Based on our data and previous work showing that Jip3 can bind components of the dynein motor complex [15] , we hypothesized that direct Jip3-JNK interaction was necessary for the retrograde transport of pJNK . To address this , we first asked whether Jip3 and JNK3 were transported together in pLL axons using a dual cargo transport assay . We co-injected Jip3-mCherry and JNK3-mEos plasmids and identified embryos in which both constructs were expressed in the same pLL neuron . Notably , co-injection of these and other cargos used for dual transport analysis ( see below ) resulted in almost 100% co-expression . Sequential imaging of Jip3 and JNK3 positive vesicles at 2 dpf revealed a high degree of co-transport , primarily in the retrograde direction ( Video S8 ) . While only 16% of vesicles in the anterograde pool were positive for both Jip3 and JNK3 , 87% of vesicles in the retrograde pool carried both proteins ( N = 5 embryos ) . This data supported a role for Jip3 in the retrograde transport of activated JNK . Importantly , since mEos is a green to red photoconvertable molecule , we used extreme caution during these dual imaging experiments to prevent accidental photoconversion and noted no green to red shift in the vesicles imaged during these sessions ( data not shown ) . Next , we addressed whether the direct interaction between Jip3 and JNK was necessary for retrograde pJNK transport by asking whether the pJNK accumulation in jip3nl7 could be rescued with a Jip3 variant that lacked the JNK binding domain ( Jip3ΔJNK: amino acids 202–214; [32] ) . DNA constructs were injected into zygotes to mosaically express Jip3-mCherry or Jip3ΔJNK-mCherry in individual pLL ganglion neurons . At 4 dpf , axon terminals expressing the respective fusions were imaged live and scored for axon morphology before larvae were individually immunolabeled for pJNK and the same axon terminals were re-imaged . As each NM is innervated by 2 axons and this innervation is segregated in space [34] , we could use the non-expressing half of the NM to identify which larvae were jip3nl7 mutants as well as utilize it as a normalizing factor for the quantification of pJNK immunofluorescence . Though full-length Jip3 rescued axon terminal swellings and the accumulation of pJNK , Jip3ΔJNK was unable to rescue either phenotype ( Figure 6A–6E ) . Importantly , expression of Jip3ΔJNK by mRNA injection rescued axon length , providing evidence that deletion of this region did not result in protein instability or failed processing , and pointing to a JNK-independent mechanism for Jip3's role in axon outgrowth ( Figure S7 ) . In summary , these data show that direct interaction between Jip3 and JNK is necessary for pJNK retrograde transport and also revealed a correlation between the accumulation of pJNK due to loss of Jip3-JNK interaction and the generation of axon terminal swellings . To determine if high levels of pJNK in axon terminals were sufficient to cause axon terminal swellings , we conditionally and mosaically expressed a constitutively active form of JNK3 ( caJNK3; [35] , [36] ) fused to EGFP under the control of a heat shock promoter in pLL neurons of wildtype larvae . Fifteen hours after activation at 4 dpf , we identified larvae that were expressing this construct in pLL axon terminals . Subsequently , these larvae were individually immunolabeled using anti-pJNK and anti-GFP antibodies to determine if caJNK3 could alter axonal morphology and additionally determine if axonal swellings correlated with elevated pJNK levels . Using this assay , we found that increased pJNK levels by expression of caJNK3 correlated with the presence of axon terminal swellings ( Figure 6F ) . Interestingly , expression of caJNK3 did not always elevate pJNK levels ( 8 out of 17 larvae ) and axon terminals were not swollen in these instances ( data not shown ) . To test if axon terminal swellings were a result of JNK activity , we mutated the site phosphorylated by the upstream activating MAPKK to render caJNK3 inactive ( caJNK3-IA; [37] ) . To assay the efficacy of the caJNK3 and caJNK3-IA constructs , we expressed both individually using RNA-mediated whole embryo expression and assayed phospho-cJun levels , a direct downstream JNK target , by Western blot analysis . As predicted , caJNK3 elevated levels of p-cJun ( Figure 6H ) while caJNK3-IA did not ( Figure 6I ) . Induction of caJNK3-IA using a protocol identical to that used of caJNK3 did not cause axonal swellings in any of the 16 larvae we imaged ( Figure 6G ) , confirming that JNK activity was indeed required for the generation of axon terminal swellings . These experiments demonstrated that high JNK activity is sufficient to induce axonal swellings and provided strong evidence that the axon terminal swellings in jip3nl7 mutants are due to increased pJNK levels at axon terminals . Our data demonstrated that lysosomes accumulate in jip3nl7 mutant axon terminals ( see Figure 2 ) and elevated pJNK levels cause axon terminal swellings ( see Figure 6 ) . Next , we asked whether elevated pJNK could cause lysosomal accumulation . To test this , we used the approach described above to conditionally expressed caJNK3 at 4 dpf in wildtype larvae . Larvae expressing caJNK3 in pLL neurons were immunolabeled with an anti-Lamp1 antibody and axon terminals were imaged . This analysis demonstrated that elevation of pJNK levels did not increase Lamp1 levels above controls ( Figure 7A , 7B ) . Importantly , lysosome number and dynamics appeared normal in the presence of activated JNK , as Lysotracker red vital dye labeling was similar between caJNK3 expressing axons and non-expressing neighboring axons ( Figure 7C , 7D ) . Based on genetic work in Drosophila , JNK has been postulated to act as a “switch” , controlling anterograde vs . retrograde motor activity for cargo transport [38] . Thus , we asked whether Jip3-JNK interaction could be a potential regulator of directional lysosome transport . First , we used sequential imaging to determine if JNK3 and lysosomes were co-transported by co-expressing JNK3-mEos and Lamp1-mTangerine in pLL axons and imaging their transport at 2 dpf ( N = 6; Video S9 ) . This analysis demonstrated that only ∼19% of Lamp1-positive vesicles moving in the anterograde or retrograde direction were co-labeled with JNK3-mEos . Interestingly , 72% of JNK3 positive retrograde vesicles label with Lamp1-mTangerine , suggesting that , though lysosomes do not rely on JNK3 for their movement , JNK3 was transported with lysosomes towards the cell body . Finally , we tested whether Jip3-JNK interaction had any function in lysosome transport , which , if disrupted , could lead to lysosome accumulation in axon terminals in the absence of Jip3 . To address this , we assayed whether lysosome accumulation in jip3nl7 mutants could be rescued by expressing Jip3ΔJNK ( and Jip3 as a control ) by RNA injection . For this assay , RNA was co-injected with the Lamp1-mTangerine DNA construct to visualize lysosomes in individual axons ( see Figure 2E , 2F ) . Rescue score was determined as the average of the scores recorded by 2 blind , independent raters and was based on the ratio of punctate lysosomes ( similar to wildtype in Figure 2E ) vs . aggregates ( as in mutants in Figure 2F ) . This analysis determined that Jip3ΔJNK was as effective as full-length Jip3 at suppressing lysosome accumulation in jip3nl7 mutants ( Figure 7E ) . We did not , however , observe complete rescue , potentially due to RNA degradation by 3 dpf . To complement this analysis , we implemented a DNA-based expression strategy that would allow expression of the rescue constructs at later stages . We expressed Jip3-mCherry and Jip3ΔJNK-mCherry in pLL axons using the 5kbneurod promoter and assayed larvae for lysosome accumulation using Lamp1 immunolabeling at 4 dpf . Larvae were imaged live at 4 dpf to identify the axon terminals expressing these constructs and to identify mutant and wildtype siblings based on axonal phenotype of mCherry negative axons . Subsequently , larvae were individually immunolabeled for pJNK and Lamp1 and the same axon terminals were reimaged . Consistent with our previous results ( see Figure 6 ) , Jip3ΔJNK failed to rescue axon terminal swellings or pJNK accumulation in jip3nl7 mutants but was capable of suppressing the elevation of Lamp1 levels similar to full-length Jip3 ( Figure 7F–7I and data not shown; N = 5 out of 8 jip3nl7 mutants injected with Jip3ΔJNK showed full rescue ) . Together , these data argue that Jip3-JNK interaction is not necessary for retrograde lysosome transport and supports a JNK-independent role for Jip3 in lysosome clearance from axon terminals . In cultured cells , DLIC , a dynein accessory protein , functions in dynein-dependent lysosome transport [30] . As Jip3 has been shown to interact with DLIC [22] , we hypothesized that Jip3 might serve as an adapter for lysosome-DLIC attachment during retrograde lysosome transport in axons . To ascertain whether Jip3 co-localized with moving lysosomes and could function in such a direct role , we performed sequential imaging of axons expressing both Jip3-mCherry and Lamp1-EGFP cargos at 2 and 3 dpf . Co-transport analysis revealed that Jip3 is present on lysosomes moving in the retrograde direction at both time-points ( Figure 8A–8E; Video S10 ) . Interestingly , the percentage of lysosomes that were transported in the retrograde direction labeled with Jip3 was higher at 3 dpf than at 2 dpf ( 2 dpf: 15%±3 . 8% , N = 5 vs . 3 dpf: 37%±4 . 2% , N = 4 ) . This may indicate a differential reliance on Jip3 for the transport of this organelle beyond 2 dpf , leading to the decrease in lysosome retrograde transport frequency only after 2 dpf in jip3nl7 ( see Figure 2 ) . Finally , we co-expressed DLIC tagged with mTangerine ( mTangerine-DLIC ) and Lamp1-EGFP to characterize DLIC localization and co-transport with lysosomes and determine if this association is lost in jip3nl7 mutants . At 3 dpf , mTangerine-DLIC localized to discrete puncta along the axon and in axon terminals in wildtype larvae ( Figure 8F ) . In contrast , in jip3nl7 mutants , DLIC accumulated in axon terminals , similar to lysosomes and pJNK ( Figure 8G ) . Co-transport analysis of mTangerine-DLIC and Lamp1-EGFP cargos revealed a decrease in the ratio of DLIC-positive lysosomes moving in the retrograde direction in jip3nl7 mutants ( Figure 8H–8M; Video S11 ) . This observation points to a failure of lysosome-dynein interaction during transport with loss of Jip3 . Interestingly , there was a slight decrease in DLIC-Lamp1 vesicle co-transport in the anterograde direction as well in jip3nl7 mutants suggesting that this complex may move bidirectionally . In summary , our data supports a model where the independent interaction of Jip3 with pJNK and lysosomes is required for the attachment of these cargoes to the dynein motor for clearance from axon terminals ( Figure 9 ) . Our results revealed a novel role for Jip3 in retrograde axonal transport . We provided evidence that loss of Jip3 led to a decreased frequency of retrograde transport of an active kinase ( pJNK ) and lysosomes but not other components of the endosomal or autophagocytic system . We demonstrated that direct interaction of Jip3 and JNK was necessary to prevent pJNK accumulation and the axon terminal swellings characteristic of the jip3nl7 mutant but had no effect on lysosome accumulation . Additionally , exogenous expression of activated JNK phenocopied the jip3nl7 mutant axon terminal swellings but did not cause lysosome accumulation , providing evidence that high levels of active JNK cause this phenotype in a lysosome-independent manner . Finally , our co-transport analysis suggested that Jip3 directly facilitated lysosome interaction with the dynein motor through binding to the accessory protein DLIC . Given the decrease in frequency of cargo movement , the normal distribution of dynein components in jip3nl7 mutant axon terminals , and the high rate of Jip3-lysosome and Jip3-JNK3 co-transport , we posit that Jip3 likely serves as an adapter protein that mediates attachment of these cargos to the dynein motor ( Figure 9 ) . Jip3 has been implicated in anterograde axonal transport in several studies through its interaction with both Kinesin light chain and Kinesin heavy chain components of the Kinesin-1 motor [12] , [13] , [23] . We became interested specifically in Jip3's function in retrograde transport as jip3nl7 demonstrated the unusual quality of extreme swellings in axon terminals , the end of the line for anterograde transport . A function for Jip3 in retrograde transport has indeed been posited by Cavalli et al . as they demonstrated that Jip3 co-localized with pJNK distal to nerve ligation and co-purified from similar membrane fractions as dynein components [15]; however , our study is the first to provide conclusive evidence that Jip3 is required for retrograde transport of pJNK , as pJNK accumulates in axon terminals in jip3nl7 mutants , Jip3 and JNK3 are co-transported , and direct Jip3-JNK interaction is functionally required for pJNK retrograde transport . Thus , our work identifies pJNK as a Jip3-dependent retrograde cargo . In addition , through the implementation of our in vivo imaging approach , we found that the frequency of retrograde JNK3 transport was decreased with loss of Jip3 , but the processivity of the motor ( reflected by run length ) and velocity of movement were unchanged . This data , in combination with previous biochemical studies of Jip3-JNK and Jip3-dynein interaction [15] , provide strong evidence that Jip3 functions as an adapter for pJNK , linking it to the dynein complex for transport , while not affecting motor movement itself . Using a combination of immunolabeling and in vivo imaging techniques , we further show that Jip3 is necessary for retrograde transport of lysosomes through interaction with the dynein accessory protein DLIC . DLIC has been shown to be an important mediator of dynein-based lysosome movement in culture systems [30] and was shown to biochemically interact with Jip3 in another system [22] . Thus , Jip3 could provide a link between lysosomes and dynein through its interaction with DLIC . In support of this , Jip3 is co-transported with lysosomes , the retrograde transport velocities for Jip3 alone were highly similar to those observed for lysosomes , and DLIC-lysosome co-transport was significantly decreased in jip3nl7 mutants . Together , these data provides strong evidence that Jip3 serves as an important adapter protein for lysosome-DLIC interaction and subsequent retrograde lysosome transport . Notably , Jip3 was implicated in the anterograde transport of DLIC to axon terminals in C . elegans [22] . However , instead of a decrease , we observed increased levels of DLIC in jip3nl7 axon terminals , arguing that this Jip3 function may not be conserved in vertebrates or is compensated for by another member of the Jip family [39] . Elevated levels of activated JNK , lysosome accumulation and axonal dysmorphology have been co-associated with neurodegenerative disorders [40] . Interestingly , though our studies indicated that Jip3-JNK interaction was not required for lysosome retrograde transport , JNK3 was frequently present on lysosomes moving in the retrograde direction , suggesting that Jip3 could serve to attach both cargos to the dynein motor simultaneously . Furthermore , our results point to a lysosome-independent etiology of axon terminal swellings in jip3nl7 mutants . Evidence to support a lysosome-independent mechanism includes: 1 ) the ability to induce axonal swellings without lysosome accumulation by exogenous expression of constitutively active JNK; 2 ) the absence of axon morphological changes following expression of an inactivated form of the constitutively active JNK; and 3 ) rescue of lysosome accumulation , but not pJNK levels or axonal swellings , in jip3nl7 mutant axon terminals by Jip3ΔJNK expression . Thus , our work provides evidence that axonal swellings can occur downstream of this active kinase without causing concomitant accumulation of organelles in the autolysosomal pathway . The exact etiology of axonal swellings in jip3nl7 mutants due to elevated levels of activated JNK remains to be determined . Importantly , jip3nl7 mutants did not exhibit a global disruption of retrograde axonal transport , which would indirectly lead to cargo accumulations . Evidence supporting the specificity of transport disruptions includes: 1 ) absence of the accumulation of other cargo ( late endosomes , autophagosomes , and signaling endosomes ) in jip3nl7 axon terminals; and 2 ) normal localization of dynein heavy chain and p150glued in jip3nl7 axon terminals , indicating that dynactin-based initiation of dynein transport is not hindered [9] , [10] . Thus , our data supports a direct role for Jip3 as an adapter for the transport of two specific retrograde cargos , pJNK and lysosomes . In summary , our data demonstrate novel and separate roles for Jip3 in the retrograde axonal transport of activated JNK and lysosomes . It is tempting to speculate that Jip3-dependent retrograde clearance of activated JNK may be a novel and crucial strategy for the removal of this active kinase from axon terminals , bypassing traditional phosphatase pathways . Furthermore , we show that enhanced JNK activity can indeed cause axon terminal swellings , similar to those observed in the jip3nl7 mutant , in the absence of lysosome accumulation . Thus , we have shown that there can be an independent etiology for these tightly coupled events observed in disease models . The similarities between the axonal swellings , high levels of pJNK , and accumulation of lysosomes in jip3nl7 and neurodegenerative diseases such as Alzheimer's Disease points to an intricate relationship between these phenotypes during pathogenesis . Our studies begin to unravel how Jip3-dependent regulation of retrograde axonal transport may underlie or modulate such disease states . Adult AB and WIK zebrafish and AB/WIK hybrids were maintained at 28 . 5°C and staged as described [41] . Embryos were derived from natural matings or in vitro fertilization , raised in embryo media , and developmentally staged using previously established methods [42] . Strains utilized included TgBAC ( neurod:EGFP ) nl1 [16] , TgBAC ( phox2b:EGFP ) w37 [18] , TgBAC ( neurog:DsRed ) nl6 , TgBAC ( foxd3:EGFP ) nl5 transgenics and mitfaw2 [43] , and mapk8ip3nl7 ( jip3nl7 ) mutants . We used Escherichia coli-based homologous recombination to modify a neurog1- and foxd3-containing bacterial artificial chromosome ( BAC ) clones [44] . The neurog1 BAC clone zK171N3 contains 63 . 8 kb of upstream and 106 . 1 kb of downstream sequence of neurog1 , while the foxd3 BAC clone zC137J12 contains 66 . 2 kb of upstream and 122 . 1 kb of downstream sequence of foxd3 ( http://www . sanger . ac . uk/Projects/D_rerio/mapping . shtml ) . After recombination , the modified BAC clones contained DSRedExpress-1 and EGFP positioned at the endogenous start site of neurog1 or foxd3 , respectively . The accuracy of recombination was evaluated by PCR , sequencing , and analysis of transient expression . To obtain germline transgenics , we microinjected 20–80 pg of BAC DNA ( linearized with Srf I for neurog1 BAC and supercoiled for foxd3 BAC ) into zebrafish zygotes , raised injected fish to adulthood , and screened their progeny for reporter gene expression . The germline transmission rate was 2 . 3% for neurog1 BAC and 1 . 4% for the foxd3 BAC . The TgBAC ( neurog1:DSRed ) nl6 and TgBAC ( foxd3:EGFP ) nl5 strains have been outcrossed for multiple generations and transfmitted the transgenes in a Mendelian manner . The jip3nl7 mutant was identified in a standard three-generation N-ethyl-N-nitrosourea ( ENU ) mutagenesis screen [45] , [46] . For this screen , TgBAC ( neurod:EGFP ) nl1-positive larvae were screened at 4 dpf for axon truncation and the presence of axonal swellings under epifluorescence . For genetic mapping , heterozygous carriers of jip3nl7 on a polymorphic AB/WIK background were incrossed to produce homozygous , heterozygous and wildtype progeny . Initial chromosome assignment was done by bulk segregate analysis of DNA pools from 20 wildtype and 20 mutant individuals using microsatellite markers ( http://zfin . org/ZFIN ) . Flanking regions were identified using individual wildtype and mutant larva and markers z15457 , z21697 , and a designed marker , CA50 ( forward: 5′-TTACACACTTTCAGCCTGTC , reverse: 5′-CCTTTATGCCACGGTCACA ) . Genomic DNA was isolated from larvae by incubating it overnight at 55°C in PCR Extraction Buffer ( 10 mM Tris pH = 8 , 2 mM EDTA , 0 . 2% Triton X-100 , 200 µg/ml Proteinase K ) . Total RNA was isolated from larvae using Trizol according to the manufacture's protocol ( Invitrogen ) and cDNA was generated using Superscript II reverse transcriptase and oligo dT primers ( Invitrogen ) . The full mapk8ip3 ( jip3 ) cDNA was amplified using following primers ( forward: 5′-CGTTAAACGAGCTTCGGACA , reverse: 5′-GCGTGTCACTTTGAGTTTGG ) based on the predicted sequence and subsequently entered into GenBank ( KC170712 ) . Full-length jnk3 was amplified using primers ( forward: 5′-ATGAACAGACGTTTCTTATATAACTGC , reverse: 5′-CACGGCTGCACCTGCGCTG ) designed against the annotated sequence ( NM 001037701 ) . Full-length dynein light intermediate chain was amplified using primers ( forward: 5′-TGTCACTCAAGCCTGCGAAG , reverse: 5′-GGATTTGTCGTTTTCAGCAG ) designed against the annotated sequence ( NM 001017669 ) . To genotype jip3nl7 carriers , a 385 bp region around the mutation was amplified from genomic DNA by PCR using annealing T = 55°C and the following primers ( forward: 5′-TTTGTCTGTTGAAATTGCT , reverse: 5′-ACGGTCCATACCCATGATT ) . PCR products were then digested with SpeI , as the single nucleotide change generates this restriction site in the jip3nl7 allele , producing two bands , 243 and 142 bp . RNA in situ hybridization was performed as described [47] . Digoxygenin-labeled antisense RNA probes were generated for jip3 and jnk3 using the full-length cDNA cloned . Whole mount immunohistochemistry was performed following established protocols [48] . The following antibodies were used: anti-GFP ( 1∶1000; Invitrogen #A11122 ) , anti-pJNK ( 1∶100; Cell Signaling #9251S ) , anti-tJNK ( 1∶100; Cell Signaling #9252 ) , anti-p150glued ( 1∶100; Signal Transduction Labs #610473 ) , anti-dynein heavy chain ( 1∶100; gift of R . Vallee; [49] ) , anti-Rab7 ( 1∶100; Sigma #R8779 ) , anti-Lamp1 ( 1∶100; Developmental Studies Hybridoma Bank ) , anti-LC3 ( 1∶100; Novus #NB100-2331 ) , anti-TrkB ( 1∶100; Santa Cruz Biotechnology #sc-12 ) and Alexa-488/568/647 ( 1∶750; Invitrogen ) . Antibodies not used previously in zebrafish were validated by Western blot analysis ( see below: Figure S3I–S3L; Rab7–24 kD , LC3–14 . 5 kD , TrkB-69 kD and 18 kD , Lamp1–27 kD ) . For TUNEL labeling , embryos were processed as previously described [50] with minor modifications according to the manufacturer's instructions ( In situ cell death kit , Roche ) . For Lysotracker red vital dye staining , 4–5 dpf larvae were incubated in Lysotracker red ( 1∶10 , 000; Invitrogen ) for 15 minutes in embryo media , washed briefly , embedded in 1 . 2% low-melt agarose , and imaged . All fluorescently labeled embryos were imaged using a FV1000 laser scanning confocal system ( Olympus ) . Brightfield or Nomarski microscopy images were collected using a Zeiss Imager Z1 system . Images were processed using ImageJ software [51] . Brightness and contrast were adjusted in Adobe Photoshop and figures were compiled in Adobe Illustrator . For western blot analysis , protein was isolated from wildtype and jip3nl7 3 dpf larvae by homogenizing individuals in extraction buffer ( 150 mM NaCl , 50 mM Tris pH = 7 . 4 , 5 mM EDTA , 0 . 05% NP40 , 25 mM NaF , 10 mM Na3VO4 , 1 mM DTT , 10 µL/mL protease inhibitor ) at a ratio of 4 µL buffer per embryo . The equivalent of 4 embryos was run in each lane on a 12% SDS-PAGE gel and transferred onto a PVDF membrane ( Millipore ) . Primary antibodies were applied overnight: anti-pJNK ( 1∶1000; Cell Signaling #9251S ) , anti-tJNK ( 1∶1000; Cell Signaling #9252 ) , anti-p150glued ( 1∶1000; Signal Transduction Labs #610473 ) , anti-dynein heavy chain ( 1∶1000; gift of R . Vallee; [49] ) , anti-Rab7 ( 1∶2000; Sigma #R8779 ) , anti-Lamp1 ( 1∶4000; Developmental Studies Hybridoma Bank ) , anti-LC3 ( 1∶500; Novus #NB100-2331 ) , anti-TrkB ( 1∶100; Santa Cruz Biotechnology #sc-12 ) , and anti-p-cJun ( 1∶1000; Cell Signaling 9164S ) . After washing , an anti-rabbitHPR , anti-mouseHRP , or anti-ratHRP secondary ( 1∶10 , 000; Jackson Immuno ) was applied for 90 minutes . Protein was visualized using SuperSignal West Pico Chemiluminescent Substrate according to the manufacture's specification ( Thermo Scientific ) . If necessary , the blot was then stripped with 25 mM glycine ( pH = 2 . 5 ) and re-probed with rabbit anti-α-actin ( 1∶10 , 000; Sigma ) . To generate constitutively active JNK3 that could be activated in a temporally specific manner , we fused MKK7 to JNK3 and placed this fusion behind a heat-shock inducible promoter ( hsp70:mkk7-JNK3-egfp , referred to as caJNK3 in the text ) . To generate an inactive form of the same construct ( referred to as caJNK3-IA in the text ) , two amino acids were mutated ( T221A and Y223F ) to render JNK3 not able to be phosphorylated , which is required for its activity [37] . For induction of transcription of both constructs , 4 dpf larvae injected with 10 pg of the caJNK3 or caJNK3-IA constructs were heat-shocked at 38°C for 1 hour . Larvae were then transferred to 28 . 5°C prior to analysis . Zygotes were injected with plasmid DNA encoding fluorescently tagged cargos of interest with expression driven by the 5kbneurod promoter [26] . At 30 hpf , 2 dpf , or 5 dpf , embryos or larvae were sorted under epifluorescence to identify individuals with tagged cargo expression in a few cells of the pLL ganglion . For imaging , embryos were mounted in 1 . 2% low melting point agarose on a glass coverslip , submerged in embryo media containing 0 . 02% tricaine and imaged using a 60X/NA = 1 . 2 water objective on an upright Fluoview1000 confocal microscope ( Olympus ) . For each embryo , a region of interest ( 30–200 µm ) was selected in the pLL nerve in which a long stretch of axon was observable in a single plane . Scans were taken at the fastest possible speed ( 3–5 frames per second ) for 600 to 2500 frames . Embryos were subsequently released from agarose and processed for genotyping . For co-transport , embryos expressing both constructs in a single cell were selected and imaged as described above using sequential imaging of the 488 and 568 nm excitation channels . 600 frames were collected at 2–3 frames per second . Transport parameters were analyzed using kymograph analysis in the MetaMorph software package ( Molecular Devices , Inc . ) . Kymographs were generated from each imaging session and used to determine distance moved in individual bouts of movement ( uninterrupted straight lines ) and velocity of movement ( slope of uninterrupted straight lines ) . Typically , 10–50 traces were analyzed in each kymograph and these were averaged within individual embryos for statistical analysis . The number of particles moving in each direction was estimated based on traces on the kymographs and then normalized to length of axonal segment and total imaging time . Five day old zebrafish larva ( neurod:EGFP carriers ) were anesthetized in 0 . 02% tricaine ( MS-222; Sigma ) and embedded in 3% methylcellulose on a slide . Pulled thick-walled glass capillaries were used to sever the nerve between NMs 2 and 3 . Slides were immersed in Ringer's solution ( 116 mM NaCl , 2 . 9 mM KCl , 1 . 8 mM CaCl2 , 5 mM HEPES pH = 7 . 2 , 1% Pen/Strep ) and incubated at 28 . 5°C for 3 hours . Larva were then collected and immunolabeled for pJNK or tJNK and EGFP . Details of image and statistical analyses are described below . For analysis of pJNK and tJNK intensity in axon terminals and after nerve injury , individuals were immunolabeled as described above . For consistency of labeling , larvae that were directly compared were processed in the same batch . Confocal Z-stacks ( 0 . 5 µm between planes ) were taken of the area of interest using a 40X/NA = 1 . 3 oil objective with identical settings . Images were analyzed using ImageJ [51] . For fluorescent intensity measurements of pJNK or tJNK in wildtype and mutant axon terminals , summed projections of the regions of interest were generated only through regions that contained the neurod:EGFP signal and converted to 8 bit in ImageJ . In the pLL nerve injury analysis , a 30 µm , neurod:EGFP-positive region encompassing the proximal or distal edge of the severed axon was selected and summed projections through only this segment were compiled for analysis . By restricting our analysis to the neurod:EGFP axons we eliminated a majority of the fluorescent signal from the surrounding tissue . Prior to statistical comparison , the mean background fluorescent intensity , measured in a region adjacent to the NM axon terminal or injury site , was subtracted from the values generated . For analysis of pJNK levels in the DNA rescue experiment , axon terminals expressing Jip3-mCherry or Jip3ΔJNK-mCherry ( typically innervating half a NM ) and control terminals not expressing these constructs ( the alternate half of the NM ) were outlined in similar summed confocal projections and the mean fluorescent intensity was measured . The ratio of pJNK fluorescence in the axons expressing the rescue construct to those not expressing the rescue construct were compared for statistical analysis . Statistical analysis was performed using the JMP software package . Data suitable for parametric analysis were analyzed using ANOVA , with Tukey-Kramer highly significant difference post-hoc contrasts for more than two variables . Data not suitable for parametric analysis were analyzed using Wilcoxon rank-sum analysis . Ordinal data was analyzed using Chi Square test . In all cases , data from individual embryos was averaged prior to analysis making each N equivalent to an embryo . All animal work was approved by and conducted according to guidelines of the Oregon Health & Science University IACUC .	To form and maintain connections , neurons require the active transport of proteins and organelles between the neuronal cell body and axon terminals . Inhibition of this “axonal” transport has been linked to neurodegenerative diseases . Despite the importance of this process , to date there was no vertebrate model system where axonal transport could be studied in an intact animal . Our study introduces zebrafish as such a model and demonstrates its power for the analysis of axonal transport . We used this system to 1 ) initiate a genetic screen to find novel mediators of axonal transport; 2 ) develop in vivo imaging strategies to visualize axonal transport in real time in the intact animal; and 3 ) discover , using these methods , that JNK interacting protein 3 ( Jip3 ) is required for the transport of two cargos , a kinase and lysosomes , from axon terminals to the cell body ( retrograde transport ) . In the absence of Jip3 , these cargos accumulate and axon terminals become dysmorphic , though the retrograde transport of other cargos is normal . Interestingly , abnormal localization of these cargos has been linked to axonal disease states , but our work is the first to identify a specific adapter protein necessary for their transport from axon terminals .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "animal", "models", "developmental", "biology", "zebrafish", "developmental", "neuroscience", "model", "organisms", "biology", "neuroscience" ]	2013	JNK-Interacting Protein 3 Mediates the Retrograde Transport of Activated c-Jun N-Terminal Kinase and Lysosomes
Rift Valley fever ( RVF ) outbreaks are recurrent , occurring at irregular intervals of up to 15 years at least in East Africa . Between outbreaks disease inter-epidemic activities exist and occur at low levels and are maintained by female Aedes mcintoshi mosquitoes which transmit the virus to their eggs leading to disease persistence during unfavourable seasons . Here we formulate and analyse a full stochastic host-vector model with two routes of transmission: vertical and horizontal . By applying branching process theory we establish novel relationships between the basic reproduction number , R0 , vertical transmission and the invasion and extinction probabilities . Optimum climatic conditions and presence of mosquitoes have not fully explained the irregular oscillatory behaviour of RVF outbreaks . Using our model without seasonality and applying van Kampen system-size expansion techniques , we provide an analytical expression for the spectrum of stochastic fluctuations , revealing how outbreaks multi-year periodicity varies with the vertical transmission . Our theory predicts complex fluctuations with a dominant period of 1 to 10 years which essentially depends on the efficiency of vertical transmission . Our predictions are then compared to temporal patterns of disease outbreaks in Tanzania , Kenya and South Africa . Our analyses show that interaction between nonlinearity , stochasticity and vertical transmission provides a simple but plausible explanation for the irregular oscillatory nature of RVF outbreaks . Therefore , we argue that while rainfall might be the major determinant for the onset and switch-off of an outbreak , the occurrence of a particular outbreak is also a result of a build up phenomena that is correlated to vertical transmission efficiency . Rift Valley fever ( RVF ) is an emerging zoonotic disease with pronounced health and economic impacts , particularly to vulnerable African communities with low resilience to economic and environmental challenges [1–3] . Studies have shown that the disease has two distinct cycles: the epizootic/epidemic and the enzootic/inter-epidemic or endemic [4] . During the inter-epidemic cycle , disease transmission occurs at low levels in nature during periods of average rainfall . The virus is thought to be maintained through transovarial transmission from the female Aedes mosquito to her eggs and by occasional amplification cycles in nearby livestock [5] . The epidemic activities have been found to be highly correlated to heavy rainfall and flooding ( in particular in eastern and southern regions of Africa ) that stimulate hatching of Aedes mosquito eggs , resulting in a massive emergence of both uninfected and infected Aedes mosquitoes [4 , 6] . The infected ones if feeding on nearby vulnerable ruminants/livestock , would then trigger virus amplification , leading to an epizootic . An epizootic is mainly driven by the subsequent elevation of various Culex mosquito populations , which serve as excellent secondary vectors if immature mosquito habitats remain flooded for a long enough period [6 , 7] . These disease epidemic activities occur at very irregular intervals of up to 15 years in the southern and eastern regions of Africa as well as in the horn of Africa [1 , 3] . This characteristic temporal pattern of disease outbreaks adds an additional complication towards efforts for understanding and predicting occurrence of outbreaks . Findings from a pioneering empirical study in Tanzania on the subject of disease temporal and spatial patterns [7] suggest that continuous endemicity of Rift Valley fever virus ( RVFV ) may lead to periodic disease outbreaks . Similar observations have also been reported in Kenya [3] and South Africa [8 , 9] . Although correlation between RVF outbreaks and the warm phase of El Niño/Southern Oscillation ( ENSO ) phenomena which lead to abnormal rainfall has been reported [10] , there have been instances where no outbreaks were recorded following seasons of exceptionally above normal rainfall [7] . Moreover , in some Sub-Saharan regions , such as West Africa RVF outbreaks are not known to be correlated with above average rainfall [2] . In Senegal , it is reported that disease outbreaks have occurred during drought and normal rainy seasons [11 , 12] , and have been attributed to human-induced movement of livestock and trade and loss of herd immunity over time [11 , 13] . However , a common situation could be the mechanism that lead to virus endemicity during dry season which is also suspected to be through transovarian transmission in Aedes vexans female mosquitoes [11] . The inter-epidemic period in Senegal is estimated to be 5-7 years , a time length thought to closely correspond to the time it takes for renewal of a domestic herd of ruminants [12] . This suggests that while rainfall might be the major determinant factor for the onset and switch-off of an outbreak , it is likely to not be the only factor responsible for this temporal characteristic pattern of disease outbreaks . Although consensus is yet to be achieved , studies have suggested that causal association between local environment factors , livestock density and movement , and encroachment of mosquitoes into new geographical area might be responsible for modifying temporal patterns of RVF outbreaks [14–16] . Findings by recent studies [7 , 17] suggested that once RVFV had been introduced to a new geographical area , it becomes endemic and also pointed out that these newly established endemic areas constitute a source for future outbreaks once favourable environmental conditions are satisfied . Clearly , here the role of transovarial transmission cannot be neglected because it is essential for infection reactivation and scale of virus transmission in response to climatic conditions . This leads to a suspicion that apart from environmental conditions and other factors disease outbreak may be a result of a build up phenomena that depend on the efficiency of vertical transmission . Therefore , the present research study aims to investigate factors underlying the characteristic temporal patterns of RVF outbreaks and explore possibilities of predicting these outbreak patterns based on disease inter-epidemic activities . Over the past decades mathematical models have been used to translate assumptions concerning transmission and spread of RVF at population level . From the pioneer RVF models by Favier et al . [18] and Gaff et al . [19] , several models have been formulated and analysed using deterministic compartmental modelling approach [20–28] . Although these models have potential for examining factors underlying dynamics of the disease , they fail to capture observed fluctuations on the occurrence of RVF outbreaks . Nevertheless , extending these models to include seasonality yielded rich dynamics including chaotic behaviour [28] . Chitnis et al . [24] suggested that seasonality combined with mosquito vertical transmission and/or introduction of new infected individuals after immunity wanes was necessary for the survival of RVF and inter-epidemic persistence . On the other hand a study in [29] used a seasonally forced deterministic model to explore different scenarios of infection persistence including vertical transmission and alternate wildlife hosts , and concluded that RVF persistence is a delicate balance between numerous species of susceptible hosts , mosquito species , vertical transmission and environmental stochasticity . In these situations such dynamics are attributed to climatic variations disregarding the fact that interaction between the deterministic dynamics and demographic stochasticity is central for explaining realistic disease patterns [30] . Deterministic models are typically assumed to be reasonable approximations for infinitely large homogeneous populations , and arise from the analysis of mean field stochastic models , such that if one considers finite populations which is the case of livestock , stochastic interactions even within a well-mixed system may introduce new phenomena [31] . Therefore , it is more likely that these disease characteristic temporal patterns could be captured by fully stochastic models [31 , 32] , which are known to show large oscillations caused by the stochasticity exciting the system’s natural frequency [33 , 34] . Stochastic effects are known to show major impacts whenever the prevalence of infection in either the host or vector population , or both are low and can be highly significant during the period immediately after the introduction of infection into a population [35] . In this study we formulate a full host-vector stochastic model which takes into account mechanisms of vertical transmission on the vector population . Our aim is to examine the impact of stochastic effects and virus endemicity on the invasion and persistence of the disease . Stochastic effects can also lead to disease extinction during endemic settings [36] . To investigate these situations we employ branching process theory [37–39] , which has been successfully applied in vector-borne epidemic models ( for more details see [35 , 40] ) . Here we extend the analysis presented in [35] to include vertical transmission while implementing infection rates that depend on the sizes of both host and vector populations . Our objective is to examine the impacts of mosquito biting behaviour and host efforts to avoid the biting on the invasion and persistence of the disease in the presence of vertical transmission . Although stochasticity can cause large departures from equilibrium , potentially allowing the number of infectives to fall to low levels [35] , it could act passively to kick the system between different deterministic states [41] , as well as interacting with the non-linearity to excite the transients [32] , leading to either periodic or non-periodic oscillations . Using power spectra analysis we investigate the periodicity of fluctuations of RVF outbreaks as was undertaken for avian influenza in [31] . This is accomplished by formulating the model as a master equation which is then studied using van Kampen’s system size expansion [42] , to provide a prediction for the dominant period of disease oscillations . Since the macroscopic dynamics can then be viewed as a sum of a deterministic and a stochastic part , this approach provides a unique opportunity to investigate the effects of stochasticity on disease endemicity and outbreaks . The approach has been successfully applied while investigating the effects of stochastic amplification [34 , 43] and seasonal forcing [32 , 44 , 45] on disease outbreaks in particular in childhood diseases and more recently on avian influenza [31] . Our objective here is to test ideas about whether the oscillatory patterns of disease outbreaks can be predicted by simply looking at disease inter-epidemic activities . Based on historical data of occurrence of disease outbreaks in particular in Kenya , Tanzania and South Africa , we suspect vertical transmission and chance events to influence the observed characteristic pattern of disease outbreaks . This analysis provides prediction of the dominant period of disease fluctuations depending on the efficiency of vertical transmission . The results highlight the role of continuous RVFV endemicity driven by vertical transmission on mosquitoes , on the periodicity of disease outbreaks which agree with findings from empirical studies [3 , 7 , 9] . Therefore , it is reasonable to argue that it could be possible to reduce the frequency and intensity of RVF outbreaks by controlling transovarial transmission efficiency . To analytically investigate temporal dynamics of a RVF model by means of stochastic processes we formulate a simple but realistic stochastic host-vector model that captures all important features of RVF dynamics . The present study does not use primary data ( medical records or public records ) , rather during model development we calibrate the model towards temporal characteristic patterns of RVF epidemic and inter-epidemic activities observed in East Africa and Southern Africa . In particular , the data used reflect patterns observed in Kenya , Tanzania and South Africa ( see [3 , 7 , 9 , 46 , 47] and references therein ) . A description of all model parameters and their respective values , ranges and sources is given in Table 1 . We investigate both disease epidemic and inter-epidemic activities in a livestock population where the transmission of the infection is intermediated by Aedes mosquitoes only . Thus , neglecting the presence of Culex species which are known to be the secondary vectors of the disease as in [24] . Aedes mosquitoes are responsible for both initial spread and persistence of the disease since the female can transmit the virus transovarially to her eggs [2 , 48] . The mosquito sub-model is an SI type model , that is , with only two compartments: susceptible and infectious . This way we ignore the exposed class and mosquitoes once infected remain infected for life . The livestock sub-model is an SIR type model , that is , susceptible , infectious and recovered . Animal hosts enter the susceptible class through birth at a constant rate , μ2 . When an infectious Aedes mosquito bites a susceptible animal , there is a finite probability , β21 that the animal becomes infected . Once an animal host is successfully infected by an infected vector , it moves from susceptible class S2 to infectious class I2 . After some time , the infectious animal host either recovers at rate ϵ2 and moves to recovered class , R2 or dies naturally at per capita rate of μ2 . Female Aedes mosquitoes ( we do not include male mosquitoes in our model because only female mosquitoes bite animals for blood meals ) enter the susceptible class through birth at rate , b1 . The term birth for mosquitoes accounts for and is proportional to the egg-laying rate; and survival of larvae [24] . Since most density-dependent survival of mosquitoes occurs in the larvae stage , we assume a constant emergence rate that is not affected by the number of eggs laid; that is , all emergence of new adult mosquitoes is limited by the availability of breeding sites [24] . Susceptible vectors , S1 are infected when they bite an infected animal with probability β12 and depending on the ambient temperature and humidity [49] the mosquitoes move from S1 to the infectious class , I1 . To reflect the vertical transmission in Aedes mosquitoes a proportion of infected , q1 newly hatched mosquitoes joins class I1 . Mosquitoes leave the population through a per capita natural death rate , μ1 . Although births and deaths are intrinsically distinct events , we assume , for simplicity , that the vector birth and death rates have the same values , which means that the total population size N1 = S1 + I1 is kept constant . A key feature of the model is that the rate at which new infections occur in both host and vector is proportional to both host and vector population . That is , the total number of bites varies with both the host and vector population sizes . This allows more realistic modelling of situations where there is a high ratio of mosquitoes to livestock and where livestock availability to mosquitoes is reduced through control intervention as well as the efforts a host takes to prevent mosquito bites ( such as swishing its tail ) [24 , 28] . Thus , the force of new infections in livestock is λ 21 = α 1 α 2 β 21 I 1 α 1 N 1 + α 2 N 2 and the force of new infections in mosquitoes is λ 12 = α 1 α 2 β 12 I 2 α 1 N 1 + α 2 N 2 , where α1 is the number of times one Aedes mosquito would want to bite a host per day , if livestock were freely available ( for details on their derivation see supplementary material section A ) . This is a function of the mosquitoes gonotrophic cycle ( the amount of time a mosquito requires to produce eggs ) and its preference for livestock blood . α2 is the maximum number of mosquito bites a host can sustain per day . This is a function of the hosts exposed surface area , the efforts it takes to prevent mosquito bites ( such as swishing its tail ) , and any vector control interventions in place to kill mosquitoes encountering hosts or preventing bites [24] . This formalism allow us to evaluate how mosquito biting behaviour and vertical transmission in Aedes female mosquitoes impact both the probabilities of disease invasion and extinction and disease fluctuations . The former is accomplished by employing branching process theory which is central for determining critical epidemic behavioural thresholds [35] , and for the later we used system-size expansion technique [57] and Fourier analysis . However , a standard incidence function used in mosquito transmitted diseases usually assumes that mosquitoes bite a particular host at a constant rate irrespective of the number of available hosts . Therefore , for very large N2 the above forces of infection can be approximated by the following standard incidence functions λ 21 ′ = β 21 α m 0 I 1 N 1 and λ 12 ′ = β 12 α I 2 N 2 as the model forces of infections . In this case α is the mosquito biting rate , such that α/N2 is the rate at which a particular host is bitten by a particular mosquito , m0 = N1/N2 is the ratio female mosquitoes to hosts and β21 and β12 are the probabilities of successful transmission per bite [58 , 59] . All the transitions of the host and the vector associated with their corresponding rates are illustrated graphically in Fig 1 . Setting the livestock population size to remain constant , we can omit the equation containing R2 , since it can be obtained when S2 and I2 are known . Therefore , the basic ingredients of our new model framework are susceptible livestock S2 , infected livestock I2 and infected Aedes mosquitoes I1 . Unlike in deterministic models the numbers in these classes are no longer treated as continuous varying quantities [35] , but instead as integers since individual-based stochastic models consider movements of individuals between classes to be discrete [60] . To be precise , these transitions are assumed to take place in a small time interval ( t , t + Δt ) with inflows and outflows of magnitude unity . If we denote the numbers in each class as s2 , i2 and i1 respectively , the general state of the system is then written as σ = ( s2 , i2 , i1 ) . Thus , T ( σ′\|σ ) represents the transition probability per unit time from state σ to the state σ′ . Note that we characterize the events taking place in the system into three distinct groups: Infection T ( s 2 - 1 , i 2 + 1 , i 1 \| s 2 , i 2 , i 1 ) = β 21 α ′ m 0 i 1 N 1 s 2 , T ( s 2 , i 2 , i 1 + 1 \| s 2 , i 2 , i 1 ) = β 12 α ′ i 2 N 2 ( N 1 - i 1 ) , ( 1 ) Birth/Death T ( s 2 + 1 , i 2 , i 1 \| s 2 , i 2 , i 1 ) = μ 2 N 2 , T ( s 2 - 1 , i 2 , i 1 \| s 2 , i 2 , i 1 ) = μ 2 s 2 , T ( s 2 , i 2 , i 1 - 1 \| s 2 , i 2 , i 1 ) = μ 1 i 1 . ( 2 ) Recovery T ( s 2 , i 2 - 1 , i 1 \| s 2 , i 2 , i 1 ) = ( ϵ 2 + μ 2 ) i 2 . ( 3 ) where α ′ = α 1 α 2 α 1 m 0 + α 2 for general forces of infections λ21 and λ12 , and α′ = α for standard forces of infections λ 21 ′ and λ 12 ′ . For better illustration we summarize all of the processes taking place in the system and their corresponding rates and probabilities of occurrence in Table 2 . Note that these rates are the conditional instantaneous stochastic rates of individuals entering or leaving each compartment at time t and also depend on the sizes of each compartment . Using the probabilities in Table 2 , we can now construct the master equation in its general form [34 , 42 , 61] , describing temporal evolution of the probability distribution of determining the system in state σ at time t . d P ( σ ; t ) d t = ∑ σ ′ ≠ σ T ( σ \| σ ′ ) P ( σ ′ ; t ) - ∑ σ ′ ≠ σ T ( σ ′ \| σ ) P ( σ ; t ) , ( 4 ) where σ = ( s2 , i2 , i1 ) represents the state of the system , P ( σ , t ) is the probability of the system in the state σ at time t . This can also be referred to as the forward Fokker-Planck ( or forward Kolmogorov ) equation , which is a differential equation for the probability density function P ( σ , t ) of determining the system in σ at time t and it cannot be solved exactly . An alternative analytical approach can be the derivation of the moments of the distribution of the state σ . However , for the purpose of our study we analyse the master equation using van Kampen’s system-size expansion [42] , see Section C . 2 of S1 Methods . In the following sections we determine both the probabilities of a major outbreak and extinction after introduction of a single or few infectives into a population that is otherwise susceptible . In any disease model , a question of fundamental interest is to determine conditions under which a disease if introduced into a community with no immunity will develop into a large outbreak , and if it does , conditions under which the disease may become endemic . For this purpose , a key threshold parameter called the basic reproduction number , R0 is derived and analysed usually in deterministic epidemic models . In this context it is defined as the average number of secondary cases produced by a single infected individual during his or her entire infectious period , in a population which is entirely susceptible . In this regard , it is soon clear that when R0 < 1 each infected individual will produce less than one infected case and the probable result is that the disease will die out . On the contrary , if R0 > 1 each individual will produce more than one case and eventually the infection will invade the population . However , in the stochastic models , invasion of an infection into a susceptible population is not guaranteed by having R0 > 1: stochastic extinction can occur during the period immediately following introduction , when there are few infective individuals [35] . Thus , rather than the major outbreak that would be expected based on the behaviour of the deterministic model , only a minor outbreak might occur . During this early stage after the introduction of the pathogen , little depletion of susceptibles will have occurred and so probabilities of major outbreaks can be derived using the linear model that arises by assuming that the populations are entirely susceptible [62–64] . Thus , in the resulting model , the number of infectives can be approximated through a multi-type linear birth-death process [62] . In a multi-type branching process , individuals in the population are categorised into a finite number of types and each individual behaves independently [35] . An individual of a given type can produce offspring of possibly all types and individuals of the same type have the same offspring distribution [65 , 66] . In our model the disease is spread via two modes of infection transmission: vertical and horizontal . Thus , an infectious mosquito produces an infected animal , and a proportion q1 of infectious mosquitoes produce infectives of the same type while an infected animal produces an infected mosquito . Therefore , by assuming that secondary infections arise independently and at a constant rate over the infectious period of each infective , then the distribution of secondary infections follow geometric distributions [35] , with means R 0 11 , R 0 21 and R 0 12 for mosquito-to-mosquito , mosquito-to-animal and animal-to-mosquito transmission respectively ( for more details see subsection B . 2 of S1 Methods ) . In this settings , for horizontal transmission the probability generating functions ( PGF ) for the joint distribution of the dynamic variables when a single infected mosquito was introduced at time 0 can be obtained and it is given by G i ( s ) = E [ ∏ j = 1 2 s j X i j ] . ( 5 ) For vertical transmission the PGF is simply G 1 2 [67] . Note that {Xij , i , j = 1 , 2} is the number of infectives of type j produced by an infective of type i . G ( s ) is the probability generating function of the distribution of secondary infections and Eq ( 5 ) can be solved to find the extinction probability if there is initially one infective individual present . Extinction in the linear model is most likely to occur early in the process , so this corresponds to the occurrence of minor outbreaks in the nonlinear model , whereas non-extinction in the linear model corresponds to a major outbreak in the nonlinear model [35] . Eq ( 5 ) can be expanded to obtain the following formula [35] , G i ( s 1 , s 2 ) = ∑ k 1 , k 2 s 1 k 1 s 2 k 2 P ( X i 1 = k 1 , X i 2 = k 2 ) = 1 1 + ∑ j = 1 2 R j i ( 1 - s j ) ( 6 ) where i is equal to 1 or 2 . An infective animal only directly give rise to secondary infections in the vector population . Thus , we have that P ( X21 = j , X22 = k ) is equal to P ( X21 = j ) when k = 0 and zero otherwise . Consequently the generating function G2 ( s1 , s2 ) is a function of s1 alone , G 2 ( s 1 , s 2 ) = 1 1 + R 12 ( 1 - s 1 ) . ( 7 ) However , when effects of vertical transmission are included , infective mosquitoes not only give rise to secondary infections in the animal population but also to secondary infection in the mosquito population through transmission from mother to eggs . Therefore , the generating function G1 ( s1 , s2 ) is a function of s1 and s2 , G 1 ( s 1 , s 2 ) = 1 1 + R 11 ( 1 - s 1 ) + R 21 ( 1 - s 2 ) . ( 8 ) Extinction probabilities can be calculated by solving the pair of equations , G 1 ( G 2 ( s 1 , s 2 ) ) = s 1 and G 2 ( G 1 ( s 1 , s 2 ) ) = s 2 , ( 9 ) resulting from composition of functions in Eqs ( 7 ) and ( 8 ) . The pair ( s1 , s2 ) = ( 1 , 1 ) is always a solution . If R0 ≤ 1 it is the only solution , whereas for R0 > 1 there is another solution with both s1 and s2 being less than unity [38] , where R 0 = q 1 2 + 1 2 q 1 2 + 4 R 12 R 21 with R 12 = 1 ϵ 2 + m 2 + μ 2 α 1 α 2 β 12 α 1 N 1 + α 2 N 2 S 1 0 being the number of new infections in Aedes mosquitoes generated by single infected animal and R 21 = α 1 α 2 β 21 α 1 N 1 + α 2 N 2 S 2 0 1 μ 1 the number of new infections in animals generated by single infected Aedes mosquito . So far we have formulated a fully stochastic host-vector model with both horizontal and vertical transmission , under well-mixed conditions and constructed the master Eq ( 4 ) . To analyse the model we apply two methods: one is to simulate the system using the Gillespie algorithm [68] , which gives the exact realization of temporal disease evolution . The other is analytical and consists of performing van Kampen’s system-size expansion [34 , 42] of the master equation , which allows for quantitative prediction of the power spectrum of the time fluctuations of each of the system variables , and , therefore , of the dominant period of disease outbreaks [31] . Full details of van Kampen’s system size expansion are discussed in Section C of S1 Methods . This method allows us to derive analytical approximate solutions which involves making the following substitutions , s 2 = N 2 ϕ 1 + N 2 x 1 , i 2 = N 2 ϕ 2 + N 2 x 2 , i 1 = N 1 ψ + N 1 x 3 , where ϕ1 , ϕ2 , ψ are fractions of the susceptible livestock , the infected livestock and infected Aedes mosquitoes respectively , with xl ( l = 1 , 2 , 3 ) describing the stochastic corrections to the variables s2 , i2 , i1 . This expands the master equation in powers of N 1 - 1 / 2 and N 2 - 1 / 2 , such that the probability distribution P ( s2 , i2 , i1;t ) can be written in terms of the new variables x1 , x2 , x3 . Then , in comparison to the leading order , yield the following deterministic system in terms of fractions as follows: d ϕ 1 d t = - β 21 α ′ m 0 ψ ϕ 1 + μ 2 ( 1 - ϕ 1 ) , d ϕ 2 d t = β 21 α ′ m 0 ψ ϕ 1 - ( ϵ 2 + μ 2 ) ϕ 2 , d ψ d t = β 12 α ′ ϕ 2 ( 1 - ψ ) + μ 1 q 1 ψ - μ 1 ψ . ( 10 ) When integrating the above deterministic Eq ( 10 ) with respect to t we obtain trajectories of the mean behaviour which show damped oscillations tending to a fixed point see Fig 2 . This is eventually the expected long-term behaviour for realistic parameter values for host-vector models . This further confirm the results of system stability analysis . The stability of the steady state of this system is tractable , and can be obtained by deriving the deterministic limit ( see subsection D of S1 Methods ) . It is easy to verify that these equations have a trivial fixed point , named the disease-free equilibrium E0: ϕ 1 0 , ϕ 2 0 , ψ 0 ; and a unique non-trivial fixed point named the endemic equilibrium E: ϕ 1 = a + μ 2 R 0 ( a + μ 2 ) R 0 , ϕ 2 * = μ 1 μ 2 ( 1 - q 1 ) ( R 0 - 1 ) b ( a + μ 2 ) , ψ * = μ 1 μ 2 g ( 1 - q 1 ) ( R 0 - 1 ) a ( b μ 2 + μ 1 g ( 1 - q 1 ) ) , where a = β21α′ m0 , b = β12α′ , g = ϵ2 + μ2 and R 0 = 1 1 - q 1 β 21 α ′ m 0 μ 1 β 12 α ′ ϵ 2 + μ 2 is the basic reproductive number . From the stability’s analysis in Section D of S1 Methods , we know that when R0 < 1 , the disease-free equilibrium point E0 is stable while when R0 > 1 , the endemic equilibrium point E* exists and is stable . A fundamental question is whether the existence of a stable fixed point in the deterministic system generates oscillations and multi-year periodicity in the corresponding stochastic system [34] . In order to investigate this and describe the stochastic fluctuations of the system by an analytical method , we introduce step operators which allow us to express the master Eq ( 4 ) in a more compact form which further facilitate the expansion of the system . Details are given in Section C . 2 of S1 Methods , where it is shown that the resulting master equation can be written in a power series of N 1 - 1 / 2 and N 2 - 1 / 2 and the step operators in terms of the fluctuation variables x1 , x2 and x3 . Then , at next-to-leading order of the newly formed master equation ( ? ? ) we obtain a linear Fokker–Planck equation for the fluctuation variables xl ( l = 1 , 2 , 3 ) , ∂ Π ∂ t = - ∑ k , l = 1 3 A k l ∂ ( x l Π ) ∂ x k + 1 2 ∑ k , l = 1 3 B k l ∂ 2 Π ∂ x k ∂ x l . ( 11 ) This is equivalent to a set of Langevin equations [42] for the stochastic corrections to the deterministic Eq ( 10 ) having the form d x k d t = ∑ l = 1 3 A k l x l + ξ k ( t ) , ( k , l = 1 , 2 , 3 ) , ( 12 ) where ξk ( t ) ( k = 1 , 2 , 3 ) are Gaussian white noises with zero mean and a cross-correlation function given by 〈 ξ k ( t ) ξ l ( t ′ ) 〉 = B k l δ ( t - t ′ ) . Note that system Eq ( 12 ) combines both the deterministic and stochastic contributions . Given that we are interested in evaluating fluctuations of the system trajectories around the non-trivial fixed point of the deterministic system , we evaluate the entries of the Jacobian matrix Akl and Bkl of the noise covariance matrix at this stable fixed point . Explicit expressions for these two matrices are given in subsection C . 2 of S1 Methods . The Langevin Eq ( 12 ) describe temporal evolution of the normalized fluctuations of variables around the equilibrium state . By Fourier transformation of these equations , we are able to analytically calculate the power spectral densities ( PSD ) that correspond to the normalized fluctuations , independent of community sizes N1 and N2 . By taking the Fourier transform of Eq ( 12 ) , we transform them into a linear system of algebraic equations , which can be solved . After taking averages , in the three expected power spectra of the fluctuations of susceptible livestock , infected livestock and infected Aedes mosquitoes around the deterministic stationary values we obtain: P S 2 ( ω ) = ⟨ \| x ˜ 1 ( ω ) \| 2 ⟩ = B 11 ω 4 + Γ S 2 ω 2 + χ S 2 \| D ( ω ) \| 2 , P I 2 ( ω ) = ⟨ \| x ˜ 2 ( ω ) \| 2 ⟩ = B 22 ω 4 + Γ I 2 ω 2 + χ I 2 \| D ( ω ) \| 2 , P I 1 ( ω ) = ⟨ \| x ˜ 3 ( ω ) \| 2 ⟩ = B 33 ω 4 + Γ I 1 ω 2 + χ I 1 \| D ( ω ) \| 2 , ( 13 ) The complete derivation of these PSDs and detailed descriptions about the way the functions χi , Bkl , Γk and D ( ω ) depend on model parameters are discussed in subsection C . 3 of S1 Methods . In the absence of vertical transmission , that is , R11 = 0 the solutions of the equations G1 ( s1 , s2 ) = s1 and G2 ( s1 , s2 ) = s2 are provided in [35] and for the case of introduction of a single infectious vector , it is reproduced here as follows: To obtain the extinction probability requires determining the smallest non-negative root of s 1 = 1 1 + R 21 [ 1 - 1 1 + R 12 ( 1 - s 1 ) ] , ( 14 ) which is obviously given by 1 + R 12 R 12 ( R 21 + 1 ) . ( 15 ) Note that this is smaller than 1 if and only if the product R12R21 = R0 , H is greater than 1 . Consequently , when R0 , H ≤ 1 , the relevant solution is 1 and so a major outbreak can never happen [35 , 63] . For R0 , H > 1 , both the probability of extinction and of a major outbreak , are found by swapping the roles of R12 and R21 in the preceding elaboration . An interesting observation in host-vector systems is that R0 , H can be greater that one even if either R12 or R21 is less than unity . This leads to an asymmetry relationships between either with the probability of extinction or invasion and the reproductive numbers which may stem from the disparity between the sizes of the host and vector populations [35] . To further investigate this phenomenon we compute the probability of extinction and invasion while varying the biting ability of the vector when host ability to avoid a mosquito bite is taken into account . This is accomplished by varying the parameters α1 ( number of bites that a mosquito would like to bite a host ) and α2 ( number of bites a host would sustain ) when plotting the extinction and invasion probabilities . This is possible since in our approach we generalized the mosquito biting rates so that they can be applied to wider ranges of population sizes . Instead of letting the total number of mosquito bites on livestock depend on the number of mosquitoes as in [35] , we set the total number of bites to vary with both the livestock and mosquito population sizes . Results from Fig 3 ( c ) and 3 ( d ) further rephrase the roots of the observed asymmetry highlighting that although the high ratio of mosquitoes to livestock is a major factor , any form of intervention to reduce livestock availability to mosquitoes can lead to such disparity . And disease extinction is only possible if the ratio mosquitoes to livestock is kept at a very low level resulting in values of α1 less than 0 . 1 see Fig 3 ( c ) . This explains why when environmental conditions are satisfied , that is , during rainy seasons disease outbreaks are expected as a result of the presence of massive numbers of potential vectors , implying large values of α1 . From Fig 3 ( d ) we see that for α1 around 0 . 5 invasion probabilities are close to 0 . 8 . Hence , if mosquito biting activities are much more frequent disease invasion is expected but it is dependent on the availability of hosts . An interesting feature is that for α1 ≤ 1 invasion probability is zero regardless of the availability of hosts . This indicates that any intervention aimed at reducing the appetite of mosquitoes to bite might be a viable control strategy . Note also that the above observation may imply that infection does not die out merely because there are few susceptible hosts but because the number of infective vectors have reduced . Moreover , without virus reservoirs in either host or vector population or virus introduction from the outside even in the presence of optimal climatic conditions , disease activities are almost impossible . Therefore , in the following section we examine the relationships of disease persistence , extinction and spread when effects of vertical transmission efficiency are taken into consideration . In the presence of vertical transmission , determining the probability of extinction requires solving one of the equations in Eq ( 9 ) when R11 ≠ 0 . In this regard , the extinction probability following the introduction of a single infectious mosquito is given by the smallest non-negative root [63] of s 1 = 1 1 + R 11 ( 1 - s 1 ) + R 21 [ 1 - 1 1 + R 12 ( 1 - s 1 ) ] . ( 16 ) After rearranging the above equation we obtain R 11 R 12 ( 1 - s 1 ) 2 s 1 + ( R 11 + R 12 + R 12 R 21 ) ( 1 - s 1 ) s 1 - R 12 ( 1 - s 1 ) + s 1 - 1 = 0 , ( 17 ) which is a cubic polynomial in s1 . Note that for R11 = 0 this equation reduces to quadratic Eq ( 14 ) . It is evident that s1 = 1 is a solution to Eq ( 17 ) and the remaining solutions are found by solving the quadratic equation R 11 R 12 s 1 2 - ( R 11 + R 12 + R 11 R 12 + R 12 R 21 ) s 1 + R 12 + 1 = 0 . ( 18 ) Denoting A = R11R12 , B = R11 + R12 + R11R12 + R12R21 and C = R12 + 1 , there exist a unique feasible solution to Eq ( 18 ) given by s 1 = B - B 2 - 4 A C 2 A , for more details see section B . 3 of S1 Methods . Studies have shown that in the absence of vertical transmission in mosquitoes RVFV dies out when R0 < 1 and becomes endemic when R0 > 1 . However , in the presence of vertical transmission the disease may persist even for R0 < 1 [24 , 27 , 28] . This situation stems from the fact that in host-vector systems , R0 results from a complete cycle of host-vector-host or vector-host-vector transmission and does not reflect the average number of secondary infections of a specific population type [69] . For instance , R0 = 0 . 75 may result from a product of host reproductive number R12 = 5 and vector reproductive number R21 = 0 . 15 . Nevertheless , in each generation , the number of host infections is proportional to the number of infected mosquitoes , and decreases proportionally to the vertical infection efficiency . However , if the host reproductive number is high it is likely to boost up new vector infections in future generations . Fig 4 shows the dependency of probability of disease invasion on R12 , R21 and vertical transmission efficiency R11 . The invasion probability increases linearly with increments on vertical transmission efficiency with significant impact when vertical infection efficiency exceeds 20% . Other studies have found that it is only from such levels of vertical transmission efficiency that time of viral persistence is observed [69 , 70] . Another interesting relationship is that as the invasion probability increases with vertical infection efficiency the horizontal transmission R0 , H = R12 × R21 tends to decrease highlighting an asymmetric relationship with R12 and R21 as highlighted in the previous section . Since one of the main confounding factors to such asymmetric relationship is the ratio female mosquitoes to hosts , we further investigate this phenomena by examining how both vertical transmission efficiency and ratio mosquitoes to hosts impact both the invasion and extinction probabilities . This is depicted in Fig 5 where we also provide a plot for both numerical and analytical solution of the extinction probability Eq ( 18 ) when varying vertical transmission efficiency . The results show that the invasion probability increases exponentially with respect to the ratio mosquitoes to hosts but increases linearly with respect to vertical transmission efficiency , Fig 5 ( a ) . However , it saturates when the ratio mosquitoes to hosts is close to α2 , the number of bites a host would sustain , see Fig 5 ( b ) . This indicates that any adequate intervention aimed at preventing ruminants from being bitten is a viable control strategy regardless of the ratio mosquitoes to hosts . Since , Eq ( 18 ) is a polynomial of degree two its numerical and analytical solutions overlap and the extinction probability decreases quasi-linearly with respect to vertical infection , with the invasion lying above 0 . 5 Fig 5 ( c ) . This stems from the fact that the horizontal basic reproductive number , R0 , H is greater than unity , meaning that there are sustained host-to-vector and vice versa transmission cycles regardless of the efficiency of vertical transmission . A clear effect of the ratio mosquitoes to hosts is observed in Fig 5 ( b ) where for very low vertical transmission efficiency and m0 = 1 . 0 the extinction probability is almost certain . This suggests that in the absence of vertical transmission , if every mosquito is for only one ruminant then there is a high probability that the disease will die out . This result from the fact that in such settings the chance of a ruminant being bitten twice in quick succession ( once to catch the infection and once to pass it before recovery ) is very small [59] . This is also depicted in ( a ) where for m0 ≤ 1 the invasion probability is almost null regardless of the efficiency of vertical transmission , but for q 1 ⋙ 0 . 8 invasion would be possible . More interestingly is the fact that for high ratios of female mosquitoes to hosts the level of vertical infection necessary for invasion decreases substantially Fig 5 ( b ) . RVF is known to be endemic in Sub-Saharan Africa [14] with some differences in temporal patterns . In general it is emphasized that outbreaks occur at irregular intervals of up to 15 years in eastern and southern regions of the continent [7] . However , a closer look at temporal patterns of disease outbreaks in Tanzania and Kenya ( East Africa ) and South Africa ( Southern Africa ) shows existence of some possible differences in the temporal characteristic patterns of disease outbreaks . Fig 6 depicts temporal characteristic patterns of disease outbreaks from 1930 to 2007 in Tanzania [7] , from 1951 to 2007 in Kenya [3] and from 1950 to 2011 in South Africa [9] . The prevalence shown for Kenya and South Africa is artificial , it is only for representation purposes since real information regarding prevalence of the disease at each year is not available . Although data regarding reported cases for each outbreak during the recent years may exist , it is not complete [2 , 7] . For instance , in Tanzania , data for the years 1960 , 1963 and 1968 is missing . The plots in Fig 6 are based on data reported in [3] for Kenya , in [9] for South Africa and in [7] for Tanzania . According to Pienaar and Thompson [9] during this period South Africa experienced only three major outbreaks ( 1950-1951 , 1974-1976 and 2010-2011 ) and the remaining are considered smaller or isolated outbreaks . Interestingly the 1974 outbreak lasted for 3 consecutive years , a situation which can be compared to the 1960 outbreak that occurred in Kenya which continued until 1964 [3] . From the time series Fig 6 ( b ) we observe that after each major outbreak including the outbreak in 1985-1986 in South Africa there are subsequent outbreaks occurring nearly each year . According to findings by Murithi et al . [3] during the period 1950-2007 only 11 large scale outbreaks were recorded in Kenya with an average inter-epizootic period of 3 . 6 years ( range 1-7 years ) . However , for Tanzania an average inter-epizootic period of 7 . 9 years ( range 3-17 years ) is reported [7] . These disease post-epidemic activities in ruminants are known to occur without clinical cases and can only be detected where active surveillance is carried out [47 , 71] . Could it be that these differences in temporal patterns are results of a deficit of surveillance system to cover all remote regions that are vulnerable to the disease or are due to differences in the ecology of the vector ? This question takes us to another question which is the driving force of this study . Could it be possible that smaller or sporadic RVF outbreaks occur every year after major outbreaks without noticeable outbreaks or clinical cases due lack of active surveillance ? Could the prevalence of these outbreaks show multi-year periodicity ? If disease prevalence data could be available we would apply techniques of wavelet analysis which performs a time-scale decomposition of a time signal to estimate spectral characteristics of the signal as a function of time [31 , 72] . This would allow us to predict the dominant period of outbreak fluctuations when varying some model parameters in particular , vertical transmission which is known to be the driving force behind the continuous disease endemicity in these regions [7] . Since reliable information is not available , in the following section we theoretically estimate the power spectra of disease oscillations taking into account effects of demographic stochasticity and vertical transmission . Fig 7 ( first row ) depicts the power spectrum density ( PSD ) for fluctuations of the total number of susceptible livestock , infected livestock and infected mosquitoes as derived in Eq ( 13 ) , when using the standard or simplified version of the forces of infection . Our derivation of exact expressions for the power spectrum of the stochastic variables around the endemic equilibrium , see ( Eq ( 13 ) ) gives additional benefits . Using the expression for the power spectrum density ( PSD ) for variable I2 we examine how changes in female Aedes vertical transmission efficiency affects the periodicity of RVF outbreaks . In Fig 8 ( a ) we observe that an increase in vertical transmission efficiency causes a significant increase in the frequency of disease outbreaks . To better illustrate this phenomenon , we show that for vertical transmission of q1 = 0 . 05 the dominant period of disease outbreaks is about 10 years while for q1 = 0 . 5 the dominant period is about 1 year . These results suggest that with low efficiency of vertical transmission there is a high probability of disease extinction after a major outbreak , followed by a long period without outbreaks . This stems from the fact that the mosquito life cycle is relatively short and vertically acquired infections are multiplicatively diluted with every generation such that the virus is rapidly lost unless there is regular amplification in the host population . This could be only possible if renewal of susceptible livestock would happen with high frequency . Since the PSD Formula ( 13 ) describes components of the deterministic model we can examine effects of the nature of the basic reproduction number R0 on outbreaks periodicity . If R0 is less than or equal to unity , with a high probability the disease outbreak is relatively small . This is the reason why most studies would rather concentrate on the complementary case . However , our analysis ( see Fig 8 ( b ) and 8 ( c ) ) shows that the most important and interesting case is where R0 is near unity . We see that as R0 moves away from unity the PSD surface becomes flatter , indicating that more frequencies are involved in the stochastic fluctuations . This simply means that when increasing R0 , the dominant period decreases ( the dominant frequency increases ) , however for larger values ( R0 > 2 ) the PSD becomes totally flat . In this region ‘coherence resonance’ , that is , a phenomenon in which random fluctuations sustain nearly periodic oscillations around the deterministic endemic equilibrium is lost and becomes white noise . Furthermore , we examine the PSD surface for nearly extreme values of vertical transmission efficiency q1 = 0 . 05 and q1 = 0 . 5 . For larger values of vertical transmission the frequency of system fluctuation tends to increase , resulting in continuous endemicity of the disease as has been observed in some of the endemic regions [7] . While for small values of vertical infections the frequency of outbreaks is significantly reduced . We have explored the use of analytical tools to measure and examine effects of demographic stochasticity in host-vector models with two routes of transmissions . Host-vector models are designed to explain the dynamics of diseases in which transmission of the pathogen is mediated by a vector . For our study case which is Rift Valley fever ( RVF ) , the vector is a mosquito of genus Aedes with special ability of transmitting the virus to its offspring transovarially . In disease dynamics , this leads to two modes of transmission: horizontal and vertical . The later is of great epidemiological significance for it allows for investigating the contribution of this mode of transmission to disease spread and endemicity . The analytical tools applied are: branching process theory to examine the impact of stochastic effects on the invasion and persistence of RVF infection when vertical transmission is taken into account and the van Kampen method to investigate effects of mosquito vertical transmission on the characteristic temporal patterns of multi-year periodic disease outbreaks . Using branching process theory we have determined novel relationships among vertical infection , host-to-vector and vector-to-host reproductive numbers with both the invasion and extinction probabilities . These horizontal basic reproductive numbers are found to exhibit an asymmetric relationship with the probabilities of a major outbreak and extinction . Previous studies on host-vector models , using this technique highlighted that the existing asymmetry relationship between the disease transmission potentials from hosts to vectors and from vectors to hosts could stem from the fact that the disease invasion probability starting from a single infective host and the invasion probability starting from a single infective vector can differ significantly , even though the overall basic reproductive number of the infection is the same in both cases [35] . This asymmetry can lead to a situation where the overall basic reproduction number is greater than unity while either the vector or host reproductive number is less than unity , resulting in dramatic implications for disease control efforts . Unlike in previous models , we set the forces of infections to vary according to the sizes of both the host and vector populations . In this settings we further investigated the implications of this asymmetry relationships to disease control strategies by computing the invasion and extinction probabilities when varying the mosquito biting ability α1 and the host ability to avoid mosquito bites α2 . Our model predictions suggest that although the ratio of mosquitoes to livestock is a major factor , any form of intervention to reduce livestock availability to mosquitoes can lead to such disparity . Previous studies have shown that in the absence of vertical transmission in mosquitoes Rift Valley fever virus ( RVFV ) dies out when R0 < 1 and becomes endemic when R0 > 1 . However , in the presence of vertical transmission the disease may persist even for R0 < 1 [24 , 27 , 28] . To further investigate the role played by this mode of transmission , for the first time using branching process theory we derive both the invasion and extinction probabilities on a host-vector model that includes vertical transmission . It has been shown for host-vector models without vertical transmission that in regard to invasion probability the two transmission potentials can show complex relationships , causing the invasion probability to remain almost constant as a given model parameter is varied . However , it is not the case of our model which has two routes of infection transmission . Our results suggested that invasion probability increases linearly with increments on vertical transmission efficiency with significant impact when vertical infection efficiency exceeded 20% as found in other studies of vector-borne diseases [69 , 70] . Adams and Boots [69] found that vertical infection could only be important in dengue ecology , if the efficiency in nature is substantially greater than that found in empirical studies . On the contrary , vertically acquired infections are multiplicatively diluted at every mosquito life-cycle generation , such that , the virus is rapidly lost unless there is regular amplification in the host population . However , regular amplification of the virus in the host population is not certain for several factors . Recovered ruminants from RVF infection are immune for several days if not months [73] , and vaccinated animals may produce a high level of neutralizing antibodies , making them protected against subsequent RVF viral infections [74] . However , how long do these neutralizing antibodies persist and other immune responses such as innate , humoral and cell mediated are not known with good degrees of certainty and require further investigation [2] . Another interesting factor is livestock renewal either through birth or migration , and the livestock viraemic phase whose intensity and duration may vary according to the inoculated dose , the virus strain and the degree of natural susceptibility of the infected ruminant [2] . Also , a factor that could serve as a constraint to regular amplification of the disease during the inter-epidemic period is the ratio mosquitoes to hosts ( m0 ) . For the first time we derived an explicit solution translating both the probability of major outbreak or extinction in a stochastic host-vector model with both horizontal and vertical transmissions . Our results showed that for m0 ≤ 1 the invasion probability is almost zero indicating that if mosquitoes are fewer compared to livestock , it is almost impossible for the infection to invade the community because sustained transmission may be impossible . An interesting pattern was observed when vertical transmission efficiency was in the range q1 ≫ 0 . 8 , the disease could invade even for m0 ≤ 1 . This finding suggest that the interplay between the two is also a determinant factor for disease spread and if not persistence . This interplay was more paramount for m0 > 1 where the levels of vertical transmission efficiency decreased substantially . This is another interesting finding in this paper , which highlights how interaction between the ratio mosquitoes to hosts and vertical infection efficiency influence both the invasion and extinction probabilities . In the case m0 > 1 there is a clear indication that during outbreak situation effects of vertical infection are easily diluted at every generation and this mode of transmission becomes more significant mostly at early stage of the epidemic . However , the invasion probability saturated for m0 close to α2 ( host availability ) . Highlighting that if preventing measures targeting the host population are in place , the spread of infection will eventually saturate even for m0 ≫ 1 and higher level of vertical infection . Results from experimental studies have indicated that depending on the host’s innate susceptibility or resistance the infection may be classified as: severe acute lethal infection , delayed onset of complications or mild to asymptomatic infection [75–77] . Low level asymptomatic circulation and host re-introduction from external reservoir populations are also likely to be important factors [24 , 28 , 69] . Chamchod et al . [27] concluded that re-introduction of susceptible animals from external sources ( either through movement or buying ) may lead to a certain probability of some subsequent outbreaks if the renewal takes place every year . Certainly in such a situation if vertical transmission is very low we are likely to observe long intervals with no outbreaks just like the situation in Tanzania ( see Fig 6 ( a ) ) ; while for high values of vertical transmission we are likely to observe frequent waves of disease outbreaks as compared to the situation in South Africa Fig 6 ( b ) . Our results in Fig 4 further indicated that although invasion probability increases with vertical infection efficiency , the horizontal transmission reproductive number tends to decrease , highlighting an asymmetric relationship between the host and vector reproductive numbers . This further highlights the role of vertical transmission efficiency in inducing complex behaviours in the dynamics of RVF outbreaks . Such complex dynamics may partially be explained from the fact that effects of vertical infection are further compounded by effects of the diapause phenomena in Aedes mosquitoes [69] , and the ratio female mosquitoes to livestock . In summary , our analysis reveals that higher values of vertical transmission or vertical infection efficiency increase the frequency of disease outbreaks and highlights the importance of the interplay between horizontal and vertical transmission [19 , 24 , 27 , 28] in the spread and persistence of the disease . Previous RVF modelling studies [24 , 27 , 28] have relied on the use of seasonal type functions in order to explain periodicity or subsequent waves of RVF outbreaks in endemic regions as well as characterizing the nature of the resulting oscillations when mosquito population varies according to seasons or climatic conditions [24 , 27 , 28] . This is the standard paradigm in the framework of deterministic models [31] , where seasonal and/ or climatic extrinsic forcing and intrinsic host-pathogen dynamics are both used in order to understand the nature of different types of disease oscillations and system’s attractor structures [78] . However , more recently , it has become clear that the interaction between the deterministic dynamics and demographic stochasticity is fundamental to understand realistic patterns of disease outbreaks [30] . To the best of our knowledge this is the first time a non seasonal full stochastic host-vector model is used to explain the temporal characteristic patterns of disease multi-year periodicity depending on vertical transmission efficiency . This was accomplished by performing van Kampen [57] system size expansion , which allows us to derive an approximate analytical solution of the model . This method enables us to further view the population-level dynamics as being composed of a deterministic part and a stochastic part , where the spectrum of stochastic fluctuations is intimately related to the stability of the deterministic level dynamics [32] . Through power spectra analysis we were able to calculate the power spectrum of the stochastic fluctuations analytically and by comparison with simulations we can gain general insights into mechanisms underlying the peaks . Our analysis predicts complex fluctuations with a dominant period of 1 to 10 years for acceptable parameter values , which essentially depends on the efficiency of vertical transmission . Moreover , this dominant period was found to be significantly sensitive to the ratio mosquitoes to hosts and mosquitoes lifespan . These findings are in good agreement with observations , which indicate that in endemic areas RVFV is known to circulate continuously and outbreaks occur at irregular intervals of up to 15 years [3 , 79] , or 10-15 or even 3-7 years [3 , 80] . Note however , that these periods of disease outbreaks are not known with exact details due to lack of appropriate infrastructure and active disease surveillance . Although , we do not reproduce the exact known patterns of RVF outbreaks fluctuations in every country or region , we provide a plausible explanation , showing that the interplay between the stochastic component and vertical transmission is central to our understanding of the erratic patterns of disease outbreaks characterized by a dominant period of 1 to 10 years . Our results indicated that an increase in the vertical transmission efficiency increases the frequency of disease outbreaks , hence reducing the periodicity of outbreaks to nearly a dominant period of one year . This further confirms our findings through branching process theory as discussed above . When vertical infection efficiency is higher RVFV is likely to circulate every year with virus amplification at every rainfall season leading to yearly sporadic cases of disease outbreaks . This situation can be compared with the observation of disease outbreaks in South Africa as shown in Fig 6 ( b ) . According to a review by Pienaar and Thompson [9] since the first outbreak in 1950 , South Africa has experienced only three major outbreaks ( 1950-1951 , 1974-1976 and 2010-2011 ) , with sporadic or isolated outbreaks in between . Two interesting temporal patterns can be discussed: ( 1 ) the post-epidemic disease activities or disease activities between two major outbreaks are of one year cycle; ( 2 ) the second major outbreak lasted for three consecutive years . Could it be that the efficiency of vertical transmission in South Africa is relatively higher , sustaining continuous endemicity patterns ? Our analysis provides a simple but one of the most relevant explanations for this situation . An increase in vertical transmission efficiency leads to low frequency of disease outbreaks of nearly one year cycle which is in good agreement with findings from empirical studies [8 , 9] . The epidemic continued through the winter , spilling over into the next rainfall season . It is believed that such spillover was possible due to warm temperatures and wet conditions during winter , which are conductive for reproduction of mosquitoes maintaining infection through winter . However , other dynamical factors such as susceptible livestock recruitment ( or movement ) , mosquito seasonal abundance and livestock immune responses could play a role on fluctuations of RVF outbreaks [24 , 27 , 28] . Perhaps a combination of these factors was responsible for the 1974-1976 and 1960-1964 outbreaks in South Africa and Kenya respectively , which lasted for at least three consecutive years [3 , 9] . Such ‘long-lasting’ consecutive outbreaks are not common and their underlying factors are not yet fully understood . On the other hand , our model predicts that for low levels of vertical transmission the frequency of outbreaks becomes very low resulting in a dominant period of disease outbreaks of 10 years and above . These findings suggest that when efficiency of vertical transmission is very low the virus may require a long period of time to build up and eventually trigger an initial phase of the outbreak . This is a reasonable explanation for why there have been instances with no records of outbreaks following seasons of exceptionally above normal rainfall . This is likely to be the situation in East Africa , for example Tanzania ( see Fig 6 ( a ) ) . In this part of the continent outbreaks occur at irregular intervals followed by long periods ( inter-epidemic period ) without records of disease outbreaks , however , RVFV activities have been detected but with no clinical signs in the mammalian host [46 , 47 , 71] . During this inter-epidemic period ( IEP ) the virus exists but it fails to further amplify within the host during every wet season . Our explanation is that since the mosquito life cycle is very short , in the absence of regular amplification of the virus in the mammalian host population , vertically acquired infections can be rapidly lost . Low virus activities result in lower immunity in the host population and create conditions for large outbreaks whenever the virus may have sufficiently built up . In summary , for low vertical infection efficiency we expect long intervals without outbreaks . This is another contribution of this paper highlighting how our understanding of RVF ecology and epidemiology has been advanced by the work undertaken . For a long time entomological studies have highlighted the relationship between abnormal rainfall and RVF outbreaks [3 , 4 , 10 , 81] . However , optimum climatic conditions and the presence of mosquitoes have not completely explained the epidemiology of RVF outbreaks [82] . For instance , abundant rainfall , which normally correlates with increased number of mosquitoes in East Africa , was not often associated with RVF outbreaks in West Africa [2] , and even in East Africa there have been instances where no outbreaks were recorded following seasons of exceptionally above normal rainfall [7] . These observations suggest that while rainfall might be the major determinant factor for the onset and switch-off of an outbreak [7] , it is likely to not be the only factor responsible for the characteristic pattern of disease outbreaks . Other factors such as causal association between local environmental factors , livestock density and movement , encroachment of mosquitoes into new areas and livestock immune responses could be responsible for the observed characteristic pattern of disease outbreaks [7] . However , in this study we maintain the focus on the role of vertical transmission , ratio female mosquitoes to livestock and chance event on the oscillation of disease outbreaks and endemicity as we expect our results to be valid even when the above factors have been taken into account . Nevertheless , effects of livestock immune responses and livestock re-introduction or movement deserve their own further investigation .	Rift Valley fever ( RVF ) is a relatively novel vector-borne zoonotic disease , with long and irregular periods between outbreaks . Although outbreaks are highly correlated with occurrence of abnormal rainfall and flooding regimes , there have been instances with above normal rainfall where no disease outbreaks were recorded . This suggests that while rainfall might be the major determinant for the onset and switch-off of an outbreak , it may not be the only factor responsible for the temporal characteristic pattern of RVF outbreaks . We suspect that apart from influence of environmental conditions , disease outbreaks result from build up phenomena that depend on vertical transmission efficiency . Therefore , in this study we focus on formulating a stochastic host-vector model with two routes of transmission ( horizontal and vertical ) without seasonality to investigate the role of vertical transmission in the dynamics of the disease . First we apply branching process theory to characterize the relationships between vertical transmission and invasion probability and between vertical transmission and extinction probability . Finally , we apply an analytic approach , van Kampen’s system size expansion method to characterize disease complex fluctuations that are simply a result of interaction between demographic stochasticity , non-linearity and vertical transmission efficiency . Our results show that it is possible to reduce the frequency and intensity of RVF outbreaks by simply reducing the efficiency of vertical transmission .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "invertebrates", "livestock", "medicine", "and", "health", "sciences", "rift", "valley", "fever", "virus", "pathology", "and", "laboratory", "medicine", "infectious", "disease", "epidemiology", "pathogens", "vector-borne", "diseases", "microbiology", "tropical", "diseases...	2016	Predicting Rift Valley Fever Inter-epidemic Activities and Outbreak Patterns: Insights from a Stochastic Host-Vector Model
Although the malaria burden in the Lao PDR has gradually decreased , the elimination of malaria by 2030 presents many challenges . Microscopy and malaria rapid diagnostic tests ( RDTs ) are used to diagnose malaria in the Lao PDR; however , some studies have reported the prevalence of sub-microscopic Plasmodium infections or asymptomatic Plasmodium carriers in endemic areas . Thus , highly sensitive detection methods are needed to understand the precise malaria situation in these areas . A cross-sectional malaria field survey was conducted in 3 highly endemic malaria districts ( Xaysetha , Sanamxay , Phouvong ) in Attapeu province , Lao PDR in 2015 , to investigate the precise malaria endemicity in the area; 719 volunteers from these villages participated in the survey . Microscopy , RDTs and a real-time nested PCR were used to detect Plasmodium infections and their results were compared . A questionnaire survey of all participants was also conducted to estimate risk factors of Plasmodium infection . Numbers of infections detected by the three methods were microscopy: P . falciparum ( n = 1 ) , P . vivax ( n = 2 ) ; RDTs: P . falciparum ( n = 2 ) , P . vivax ( n = 3 ) ; PCR: Plasmodium ( n = 47; P . falciparum [n = 4] , P . vivax [n = 41] , mixed infection [n = 2]; 6 . 5% , 47/719 ) . Using PCR as a reference , the sensitivity and specificity of microscopy were 33 . 3% and 100 . 0% , respectively , for detecting P . falciparum infection , and 7 . 0% and 100 . 0% , for detecting P . vivax infection . Among the 47 participants with parasitemia , only one had a fever ( ≥37 . 5°C ) and 31 ( 66 . 0% ) were adult males . Risk factors of Plasmodium infection were males and soldiers , whereas a risk factor of asymptomatic Plasmodium infection was a history of ≥3 malaria episodes . There were many asymptomatic Plasmodium carriers in the study areas of Attapeu province in 2015 . Adult males , probably soldiers , were at high risk for malaria infection . P . vivax , the dominant species , accounted for 87 . 2% of the Plasmodium infections among the participants . To achieve malaria elimination in the Lao PDR , highly sensitive diagnostic tests , including PCR-based diagnostic methods should be used , and plans targeting high-risk populations and elimination of P . vivax should be designed and implemented . The malaria burden in the Lao People’s Democratic Republic ( PDR ) has gradually decreased thanks to the efforts of the Lao government and the support of partners such as the World Health Organization ( WHO ) , and the Global Fund to Fight AIDS , Tuberculosis and Malaria [1] . In 2015 , the malaria-associated mortality rate ( number of deaths/100 , 000 population ) was 0 . 03 , which was lower than the target in Millennium Development Goal 6 ( MDG 6 ) ( <0 . 20 ) [2] . However , in the same year , the malaria-associated morbidity rate ( annual parasite incidence [API]: number of cases/1 , 000 population ) was 4 . 9 , which was higher than the target in MDG 6 ( <0 . 6 ) . Now , the Ministry of Health ( MOH ) of the Lao PDR and the WHO have adopted the goal of eliminating malaria by the year 2030 [3 , 4] . In central , provincial and district hospitals in the Lao PDR , malaria is typically diagnosed by microscopy , whereas rapid diagnostic tests ( RDTs ) are used as a sub-standard diagnostic method at locations in rural areas , such as health centers and selected villages with high malaria endemicity ( API ≥10 ) . Recently , highly sensitive methods , such as polymerase chain reaction ( PCR ) [5] , ultra-sensitive PCR [6–8] and loop-mediated isothermal amplification ( LAMP ) [9–11] are becoming available and being used to detect low-level malaria infections in research bases in endemic areas . Such methods can detect sub-microscopic malaria infections and asymptomatic Plasmodium carriers in the endemic areas . We conducted a malaria field survey of 10 villages in Xepon district , Savannakhet province in the Lao PDR , where malaria was highly endemic from August to September in 2013 ( the rainy season ) [5] . A nested PCR using dried blood samples from healthy villagers revealed that there were many asymptomatic Plasmodium falciparum carriers who could be considered to be parasite reservoirs or a cryptic malaria group . In most cases , the parasite densities among the villagers were below the microscopic threshold , i . e . , sub-microscopic malaria cases . Interestingly , these parasite carriers were likely to be grouped within a family [5] . An Oxford research team also reported similar results but for many P . vivax carriers from the Thapangthong and Nong districts , which are also located in Savannakhet province , from March to July in 2015 ( dry season to the beginning of the rainy season ) [8] . Recently , there has been a gradual decrease in the number of reported cases of malaria in Savannakhet province thanks to the extensive efforts of the provincial health office and the support of several partners . In contrast , the number of reported cases in southern provinces , such as Salavan , Sekong , Attapeu and Champasak , remain high . According to the national data collected by the Center of Malariology , Parasitology and Entomology ( CMPE ) of the MOH in 2013 , the API of Attapeu province was 37 . 8 , which was the highest among all provinces in the Lao PDR . Therefore , in the present study , a cross-sectional field survey was conducted in Attapeu province , which borders Vietnam and Cambodia in the southern part of the Lao PDR ( Fig 1 ) , to investigate the prevalence of asymptomatic Plasmodium infections among the people in malaria-endemic villages , the distribution of the species of malaria parasites , and the risk factors of Plasmodium infections . We also evaluated the performance of our PCR technique in comparison to microscopy and RDTs . The research proposal was reviewed and approved by the National Ethics Committee for Health Research , Ministry of Health , Lao PDR ( No . 049 NIOPH/NECHR ) in 2014 . Written informed consent was obtained from all of the participants prior to the interview and the collection of blood for the diagnosis of malaria . The guardians of child participants ( <18 years old ) consented to their participation . In May 2015 , which was the beginning of rainy season , a cross-sectional field survey was conducted in the three districts that showed the greatest malaria endemicity in Attapeu province , Lao PDR: Xaysetha , Sanamxay and Phouvong ( Fig 1 ) . In 2013 , the APIs of Xaysetha , Sanamxay and Phouvong were 31 . 2 , 59 . 9 and 103 . 4 , respectively . The populations of Xaysetha , Sanamxay and Phouvong were 32 , 888 , 30 , 551 and 12 , 432 , respectively , in 2014 . The sample size of each district was calculated based on the malaria prevalences ( API ) in 2013 and the populations in 2014 , with a confidence level of 95% and a confidence interval ( CI ) of 5% . The minimum sample sizes of Xaysetha , Sanamxay and Phouvong were 47 , 87 and 143 , respectively . Two malaria high endemic villages ( strata 3; API ≥10 ) that were accessible by car were randomly selected in each district using village lists in district health offices . Village leader or village health volunteer informed the villagers to join the malaria survey and asked not to go to their fields , forests or fishing on the day of the survey , if possible . The survey team was based at a temple , meeting place or health center in the villages and conducted an interview with and malaria blood test for the voluntary participants who came to us ( convenience sampling ) . We accepted all voluntary participants who came to us regardless of their age , gender , occupation , ethnicity or health conditions . Information on demographics ( age , gender , ethnicity , occupation , marital status , educational level ) , tympanic temperature , any current symptoms and signs ( time first noticed ) , malaria treatment , number of previous malaria episodes ( malaria history ) and use of insecticide-treated bed nets was collected by using a questionnaire form . An approximately 150–200 μL blood sample was collected from each participant by finger prick using a lancet . Three diagnostic methods were applied to detect Plasmodium infections in blood samples: microscopy ( thin and thick blood smears ) , RDTs ( SD Bioline Malaria Ag Pf/Pv , Standard Diagnostics , Inc . , Gyeonggi-do , Republic of Korea ) and PCR . The malaria RDTs were conducted on-site , whereas microscopy and the PCR were conducted at IPL in Vientiane after the survey . When Plasmodium infections were detected by the RDTs , an antimalarial medicine ( Coartem; artemether + lumefantrine ) was prescribed free of charge on-site by the staff members from the district hospital or health center that was conducting the survey in the village . For the PCR , blood samples were collected on filter papers ( Whatman FTA Classic Cards , GE Healthcare Life Science , UK ) in accordance with the manufacturer’s instructions . DNA was extracted from the dried blood spots on the filter papers with a QIAamp DNA Mini Kit ( Qiagen , Hilden , Germany ) in accordance with the manufacturer’s instructions . Six punched-out circles ( 3 . 175 mm [1/8 inch] ) from the dried blood spot on the filter paper were used for DNA extraction , which was equivalent to 30–40 μL of whole blood . The extracted DNA was eluted with 50 μL of elution buffer in the kit and preserved until use at -30°C . To identify malaria parasite infection , a real-time nested PCR was conducted using a primer set that was reported in a previous study ( S1 Table ) [12 , 13] . In the primary real-time PCR , a universal primer set for amplifying the partial cytochrome b gene on the mitochondrial genome of all human malaria parasites was used . In the secondary real-time PCR , primer sets that were specific for P . falciparum and P . vivax were used to detect the two species . The real-time PCR was performed using SsoAdvanced Universal SYBR Green Supermix ( Bio-Rad Laboratory , Inc . , USA ) using 2 μL of the extracted DNA as a template , which was equivalent to 1 . 2–1 . 6 μL of whole blood . The primary PCR product was diluted 25 times with PCR-grade water , and 2 μL of the diluted primary PCR product was used as a template for the secondary real-time PCR . Serial diluted recombinant plasmid DNAs containing the cyt b region of P . falciparum or P . vivax were used as the positive control for each assay . PCR-grade water was used as the negative control for each assay . A sample was considered negative if there was no increase in the SYBR Green ( fluorescent ) signal after 35 cycles . The PCRs were performed independently , in triplicate . The sample was considered positive for Plasmodium DNA when positive results were obtained at least twice . As an external quality control for the PCR , some of the dried blood samples on the filter papers were sent to a laboratory in the National Center for Global Health and Medicine , Japan and examined by experienced researchers . The sensitivity and specificity of microscopy and the malaria RDTs for the diagnosis of P . falciparum and P . vivax were calculated using the results of the real-time nested PCR as a reference . For bivariate analyses , the Chi-square test and Fisher’s exact test were used to evaluate an association between variables and Plasmodium infection . A P value less than 0 . 05 was considered statistically significant . Multivariate logistic regression analyses that adjusted for the effects of other variables were conducted to estimate the association between variables and Plasmodium infection and asymptomatic Plasmodium infection using SPSS version 18 . 0 ( SPSS INC . , Chicago , IL , USA ) . All variables with a P-value of 0 . 20 from univariate analysis were entered into a multivariate logistic regression analysis . Multicollinearity among all independent variables was tested before logistic regression . Asymptomatic Plasmodium infection was defined by the following criteria: Plasmodium DNA was positive by PCR , tympanic temperature was less than 37 . 5°C ( no fever ) at the time of the survey and no history of any subjective symptoms and signs at the time of the survey and in the preceding 14 days . A total of 719 villagers ( male , n = 336; female , n = 383 ) participated in this survey ( Table 1 ) . The socio-demographic data of the participants are summarized in Table 2 and S2 Table . Most of the adult participants ( n = 472; 65 . 6% ) were farmers , and 160 of the participants ( 22 . 3% ) were students . The study population included two major ethnic groups: Lao Loum ( n = 189; 26 . 3% ) and Lao Therng ( n = 529; 73 . 6% ) . More than half of the participants ( n = 410; 57 . 0% ) had never been to primary school or had dropped out of primary school before graduation . Overall , Plasmodium DNA was detected in 47 ( 6 . 5% ) participants by the PCR ( S3 Table ) . Three of them were also detected by microscopy: P . falciparum ( n = 1; 0 . 1% ) and P . vivax ( n = 2; 0 . 3% ) , and 5 of them were also detected by the RDTs: P . falciparum ( n = 2; 0 . 3% ) and P . vivax ( n = 3; 0 . 4% ) including the individuals who were microscopy-positive . Including all diagnostic tests , 5 of the 47 participants were P . falciparum , 41 of the 47 participants were P . vivax and 1 of the 47 participants was mixed infection with P . falciparum and P . vivax . Using the results of the real-time PCR as a reference , the sensitivity of microscopy for detecting P . falciparum and P . vivax was calculated as 16 . 7% ( 95% CI -13 . 2–46 . 5 ) and 4 . 7% ( 95% CI -1 . 6–10 . 9 ) , respectively ( Table 3 ) . The sensitivity of the RDTs for P . falciparum and P . vivax was 33 . 3% ( 95% CI -4 . 4–71 . 1 ) and 7 . 0% ( 95% CI -0 . 6–14 . 6 ) , respectively . The specificity of microscopy and the RDTs was 100 . 0% for both P . falciparum and P . vivax . With regard to gender , 41 of the 336 male participants were Plasmodium-positive , whereas only 6 of the 383 female participants were Plasmodium-positive ( S4 Table ) . The positive rate of the male participants ( 12 . 2% ) was significantly higher than that of the female participants ( 1 . 6% ) ( P < 0 . 001 ) . The positive rate of soldiers ( 50% ) was the highest among the occupations ( S4 Table ) . The prevalence of Plasmodium infections was heterogeneous among the districts: Xaysettha 0 . 40 ( 95% CI -0 . 38–1 . 18 ) , Phouvong 7 . 11 ( 95% CI 3 . 61–10 . 60 ) and Sanamxay 12 . 11 ( 95% CI 8 . 09–16 . 13 ) ( Table 4 ) . A flowchart summary of malaria screening based on the clinical and PCR data is shown in Fig 2 . The average body temperature of the participants was 36 . 8°C ( 34 . 8°C–39 . 6°C ) , and only 14 participants ( 14/719: 1 . 9% ) had a fever ( >37 . 5°C ) at the time of the survey . The average body temperature of the 47 PCR-positive participants was 36 . 7°C ( 36 . 0°C–37 . 6°C ) . A summary of the socio-demographic and clinical data of the 47 PCR-positive participants is shown in S3 Table . Only one of the 47 PCR-positive participants had a fever ( >37 . 5°C ) at the time of the survey . This febrile participant ( ID: PT-015 , 8 years of age , female ) showed negative results for the blood smear and RDTs but was found to be infected with P . vivax using the real-time nested PCR . The fever started 2 days before the survey . This girl had used an insecticide-treated bet net while sleeping . The average age of all participants was 26 . 4 years ( 0–90 years , median age: 24 years ) , whereas that of the 47 PCR-positive participants was 29 . 0 years ( 6–75 years , median age: 25 . 0 years ) . Fifteen of the 47 PCR-positive participants answered that they had some symptoms or signs of health problems , such as fever , shivering and nausea , but only one participant had a fever ( >37 . 5°C ) at the time of the survey ( S3 Table ) . Eleven of these 15 participants answered that the symptoms or signs started more than 3 days before the survey . In contrast , 32 of the 47 PCR-positive participants had no symptoms or signs of health problems at the time of and the preceding 14 days before the survey ( S3 Table ) . Results of bivariate analyses to estimate an association between variables and Plasmodium infections are shown in S4 Table . Risk factors associated with Plasmodium infections were estimated by multiple logistic regression analyses . Three variables were independently associated with Plasmodium infections: being male ( AOR: 6 . 04 , 95% CI: 1 . 35–27 . 10 ) , Soldier ( AOR: 28 . 58 , 95% CI: 2 . 89–282 . 60 ) , and being Lao Loum ( AOR: 2 . 71 95% CI: 1 . 12–6 . 59 ) ( Table 5 ) . The risk factor associated with asymptomatic ( cryptic ) Plasmodium infections was a history of 3 or more malaria episodes ( multiple malaria infections ) ( Model 1 , AOR: 12 . 66 , 95% CI: 1 . 21–132 . 00 ) ( S5 Table ) . The present PCR examination revealed 47 asymptomatic Plasmodium infections ( 6 . 5% ) among the 719 voluntary participants from malaria-endemic villages in Attapeu province in May 2015 . Most of them ( 66 . 0%; 31/47 ) were adult males ( ≥18 years old ) , and the dominant parasite species was P . vivax . Males and soldiers were associated with Plasmodium infections whereas a history of 3 or more malaria episodes was associated with cryptic Plasmodium infections . Of the 32 asymptomatic Plasmodium carriers , 30 ( 93 . 8% ) had infections that were below the detection threshold of microscopy , and 28 ( 87 . 5% ) had infections that were below the detection threshold of the RDTs . To achieve the goal of malaria elimination in the Lao PDR by 2030 , sensitive diagnostic methods—such as a PCR- or LAMP-based method—should be utilized , and plans targeting high-risk populations ( males and soldiers ) and the elimination especially of P . vivax should be designed and implemented .	The Lao People’s Democratic Republic ( Laos ) is a country in the Greater Mekong Subregion . In Laos , the numbers of reported cases of malaria and deaths due to malaria have been gradually decreasing . Recently , the Lao government adopted a goal of eliminating malaria by 2030 . To achieve this goal , we must understand the precise situation in each endemic area . With this background , we conducted a field survey in Attapeu , one highly endemic province , in 2015 . We collected blood samples from 719 villagers , and most ( 98 . 1% , 705/719 ) had no fever ( <37 . 5°C ) at the time of survey . The samples were examined using three diagnostic methods: microscopy , rapid diagnostic test ( RDT ) and polymerase chain reaction ( PCR ) . PCR revealed parasitemia in 6 . 5% of the villagers ( 47/719 ) , and only one showed a fever ( ≥37 . 5°C ) at the time of survey . The dominant species was Plasmodium vivax ( 87 . 2% , 41/47 ) . Males and soldiers tended to be infected with malaria parasites whereas those with a history of multiple malaria episodes ( ≥3 times ) tended to be asymptomatic malaria carriers . These findings suggest the need to implement targeted elimination plans for vivax malaria that include measures to protect males and soldiers , who are at high risk of malaria infection .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "parasite", "groups", "medicine", "and", "health", "sciences", "lao", "people", "pathology", "and", "laboratory", "medicine", "plasmodium", "tropical", "diseases", "plasmodium", "falciparum", "parasitic", "diseases", "parasitic", "protozoans", "parasitology", "ethnicities...	2017	The detection of cryptic Plasmodium infection among villagers in Attapeu province, Lao PDR
Host defence peptides ( HDPs ) are expressed throughout the animal and plant kingdoms . They have multifunctional roles in the defence against infectious agents of mammals , possessing both bactericidal and immune-modulatory activities . We have identified a novel family of molecules secreted by helminth parasites ( helminth defence molecules; HDMs ) that exhibit similar structural and biochemical characteristics to the HDPs . Here , we have analyzed the functional activities of four HDMs derived from Schistosoma mansoni and Fasciola hepatica and compared them to human , mouse , bovine and sheep HDPs . Unlike the mammalian HDPs the helminth-derived HDMs show no antimicrobial activity and are non-cytotoxic to mammalian cells ( macrophages and red blood cells ) . However , both the mammalian- and helminth-derived peptides suppress the activation of macrophages by microbial stimuli and alter the response of B cells to cytokine stimulation . Therefore , we hypothesise that HDMs represent a novel family of HDPs that evolved to regulate the immune responses of their mammalian hosts by retaining potent immune modulatory properties without causing deleterious cytotoxic effects . Host defence peptides ( HDPs ) are found in all living organisms and play a pivotal role as effector components of the innate immune system [1] , [2] . They act as the first line of defence against pathogenic assaults from bacteria , fungi , eukaryotic parasites and viruses [3]–[5] . A range of HDPs with varied sequence lengths , structures and activities have been characterized [6] and since sequence identity between them is often very poor , their classification is based largely on homologous secondary structures . The two predominant HDP groups found in nature are the cathelicidins , characterized by α-helical secondary structure , and the defensins , which contain β-sheets stabilized by intra-molecular disulfide bridges [7]–[9] . Despite the diversity in their sequences and structures , HDPs are typically small amphipathic peptides ( 12–50 amino acids ) with a net positive charge ( +2 to +9 ) and consist of at least 50% hydrophobic amino acids [10] . These biochemical properties are central to the HDPs antimicrobial function by allowing their interaction with , and disruption of , negatively charged bacterial membranes [10] . The contribution of mammalian HDPs to the innate immune response extends beyond direct bacterial killing . The elevated expression of HDPs in response to damage ( injury or infection ) has led to the suggestion that mammals utilize these peptides as ‘alarmins’ to activate the mobilization of a comprehensive immune response [11] . Besides their antimicrobial activity , HDPs function as potent immune regulators , selectively altering host gene expression , inducing chemokine production , inhibiting bacterial- or hyaluronan-induced pro-inflammatory cytokine production , promoting wound healing and modulating T and B cell function [reviewed in [12]–[14] . The net result of these activities is a balance between pro- and anti-inflammatory immune responses which prevents an exacerbated inflammatory response while concurrently stimulating the resolution of infection and repair of damaged epithelia . The immune response elicited by helminth ( worm ) parasites is akin to the innate immune response to tissue injury and wound healing [15] , [16] . Typically , this consists of a suppression of classical pro-inflammatory responses and the induction of anti-inflammatory regulatory Th2 type immune responses . While classical Th1-associated inflammatory mediators can provide protection from helminths [17] , there is a substantial cost in collateral damage to host tissue [15] , [18] . In addition , due to their migration and feeding activities , helminth parasites cause considerable local tissue damage . Therefore , it has been proposed that on exposure to helminths , the most beneficial outcome is to shut down a destructive Th1-type response in favour of a Th2 response that rapidly and effectively heals tissue [15] , [17] , [18] . Ultimately this means that the parasite is tolerated by the host , remaining in situ for many years and thus successfully completes its lifecycle . Some advances have been made in identifying the signalling molecules that initiate helminth-associated Th2 responses . Many of these ( such as IL-33 and Thymic stromal lymphopoietin ( TSLP ) ) are thought to be released by epithelial cells damaged by migrating parasites [15] , [19] . However , a number of helminth-derived products have also been shown to modulate the function of innate immune cells and thus are potentially instrumental in the initiation of Th2 immune responses [19] , [20] . We have previously shown that a cysteine protease secreted by the trematode helminth Fasciola hepatica prevented the induction of pro-inflammatory macrophages and dendritic cells [21] , [22] . In addition , peroxiredoxin , also secreted by F . hepatica , promoted the development of Th2 host immune responses [23] , [24] . Importantly , homologues of these proteins are found in other medically-important trematode parasites which we have suggested reveals a common mechanism of immune-modulation employed by this class of pathogen . As part of our on-going analysis of the secretome of F . hepatica we recently discovered a novel 8 kDa protein . On analysis , this protein shared structural and biochemical similarities to mammalian cathelicidins and was therefore termed F . hepatica Helminth Defence Molecule 1 ( FhHDM-1 ) [25] . Like the human cathelicidin LL-37 precursor CAP-18 , FhHDM-1 is proteolytically processed ( by a parasitic endopeptidase , cathepsin L1 ) to release a 34-residue C-terminal peptide previously named FhHDM-1 p2 [25] . This peptide adopts an amphipathic helix structure and , like LL-37 , can bind to Escherichia coli lipopolysaccharide ( LPS ) to prevent its interaction with the toll-like receptor ( TLR ) 4/MD2/CD14 complex on macrophages . Hence we proposed that F . hepatica utilized FhHDM-1 as a molecular mimic of mammalian cathelicidin-like HDPs as a means of controlling host innate immune responses [25] . Phylogenetic analysis showed that FhHDM-1 is a member of a family of HDMs conserved throughout several major animal and human trematodes such as Schistosoma , Fasciola , Opisthorchis , Clonorchis and Paragonimus species [25] . Importantly , all HDM molecules in this family have preserved the C-terminal amphipathic helix . Here , we have performed a comparative functional study between several anti-microbial HDPs derived from well-characterized mammalian cathelicidins and parasite-derived peptides . For the helminth-derived peptides we selected FhHDM-1p2 and two homologs derived from Schistosoma mansoni which we term S . mansoni HDM-1 ( SmHDM-1p146 ) and HDM-2 ( SmHDM-2p58 ) . In addition , we included a peptide derived from a previously characterized secretory molecule of S . mansoni , termed Sm16-p73 [26] , which our phylogenetic studies suggest is a divergent member of the HDM superfamily [25] . We show that , in contrast to the mammalian HDPs ( LL-37 , CRAMP , BMAP-28 and SMAP-29 ) , the trematode-derived HDMs are not cytotoxic or bactericidal . However , like the mammalian HDPs , the trematode HDMs suppress the activation of macrophages by microbial stimuli and alter the isotype of immunoglobulin secreted by B cells . We propose that HDMs represent a novel family of HDPs that have undergone specific adaptation to retain potent immune modulatory properties in the absence of deleterious cytotoxic effects and are exploited by helminth pathogens to regulate the immune responses of their mammalian hosts . Four synthetic cathelicidin-derived peptides from diverse mammalian species were used: SMAP-29 from sheep [27] , [28] , CRAMP from mice [29] , LL-37 from human [30] , and BMAP-28 from cattle [31] . The 34-residue C-terminal FhHDM-1 peptide , termed FhHDM-1p2 , has been previously described [25] . Sm16-p73 peptide is 35 residues in length , corresponding to residues 73–107 of the full-length protein S . mansoni Sm16 ( GenBank accession number: AAD26122 . 1 ) . SmHDM-1p146 is 35 residues in length and corresponds to residues 146–180 of the full length protein ( GenBank accession number XP_002580563 . 1 ) . Finally , SmHDM-2p58 is a 32 residue peptide that corresponds to residues 58–98 from the full length protein ( GenBank accession number: XP_002576627 . 1 ) . All mammalian and trematode peptides were synthesized by GenScript ( NJ , USA ) and supplied endotoxin-free . The single letter code sequence of each peptide is shown in Table 1 . The biochemical characteristics for each of the HDMs and HDPs were calculated using tools available from the Antimicrobial Peptide Database ( http://aps . unmc . edu/AP/main . php ) [32] and are presented in Table 1 . Predicted properties were total net charge , Boman index , hydrophobic ratio and total hydrophobic residues on the same hydrophobic surface of the alpha helix . We have previously shown , using circular dichroism ( CD ) spectroscopy , that FhHDM-1 has the propensity to adopt alpha-helical structure in solution [25] . To assess whether HDMs from the related trematode parasite S . mansoni also form alpha helices , secondary structure prediction was performed using JPred 3 ( [33]; http://www . compbio . dundee . ac . uk/www-jpred/ ) . Specific regions predicted to form alpha helices were then subjected to helical wheel analysis using Heliquest ( [34]; http://heliquest . ipmc . cnrs . fr/cgi-bin/ComputParams . py ) to identify those with distinct hydrophobic faces; i . e . are amphipathic . The atomic structures of the vertebrate HDPs LL-37 ( PDB ID: 2K6O ) and BMAP-28 ( PDB ID: 2KET ) were visualised for comparison using the PyMOL Molecular Graphics System , Version 1 . 5 . 0 . 4 Schrödinger , LLC . ( http://pymol . org/ ) . Lipopolysaccharide ( LPS ) binding was performed using a quantitative chromogenic Limulus amoebocyte assay ( Chromo-LAL assay; Associates of Cape Cod Incorporated ) following manufacturer's recommendations . Assays were performed in flat-bottom endotoxin- and glucan-free 96-well plates ( Associates of Cape Cod Incorporated ) . Stock solutions of each peptide were prepared in endotoxin-free water ( 80 µg/ml ) and diluted to a final concentration of 250 pmol/ml . In the first step , 25 µl of peptide solution was mixed with 25 µl of a solution containing 1 endotoxin U/ml of Escherichia coli O113:H10 LPS and incubated for 30 min at 37°C to allow peptide and LPS binding to occur . The second step involved the addition of 50 µl of the chromo-LAL reagent . The liberation of ρ-nitroaniline was monitored every 60 sec at 405 nm with a Synergy H1 hybrid reader ( Biotek ) while the temperature was maintained at 37°C . Each peptide concentration was also incubated with 25 µl of LPS-free water as a control to determine if the peptide itself could activate the Chromo-LAL assay . The experiment was conducted twice in triplicate . Standard deviation was calculated from these six replicates . The minimal inhibitory concentration ( MIC ) of each peptide against various bacteria was determined using a standardized dilution method according to NCSLA guidelines [35] . Overnight colonies of E . coli , Pseudomonas aeruginosa , Salmonella typhimurium , Staphylococcus epidermis and Staphylococcus aureus , were suspended to a turbidity of 0 . 5 OD units and further diluted in Mueller-Hinton broth ( MHB ) . For determination of MIC , peptides were prepared in an acetic acid/BSA solution and used in graded concentrations ( 0 , 1 , 2 , 4 , 8 , 16 , 32 , 64 , and 128 µg/ml ) from a stock solution . Ten microliters of each concentration was added to each corresponding well of a 96-well polypropylene microtiter plate and 1×105 bacteria in the volume of 90 µL . The plate was incubated at 37°C for 16 h and then read at 600 nm with a Synergy H1 hybrid reader ( Biotek ) . The peptide's haemolytic activities were determined using human red blood cells ( RBCs; Research Blood Components , LLC ) in 96-well polypropylene microtiter plates . One hundred µl of 0 . 5% RBC suspension was added to an equal volume of a peptide ( 8–256 µg/ml ) . After 1 h at 37°C , plates were centrifuged at 1420×g for 5 min and the optical density of the supernatant was measured at 414 nm with a Synergy H1 hybrid reader ( Biotek ) . Values for 0% and 100% lysis were obtained by adding PBS or Triton X-100 ( 1%; final concentration ) to RBCs , respectively . All assays were performed in triplicate and the values of percent lysis were within a 1% standard deviation range . RAW 264 . 7 murine macrophages ( 5×105 cells ) were incubated in the presence of the fluorescent dye TO-PRO ( Life Technologies ) for 60 sec . After the peptides ( 50 µM ) were added to the culture media the uptake of dye was measured for 360 sec by flow cytometry . RAW 264 . 7 murine macrophages ( 1×106cells ) were incubated with a range of concentrations ( 2 . 5–50 µM ) of peptides for 1 h at 37°C . The culture supernatants were collected and assayed for LDH activity with the CytoTox LDH release kit ( Promega ) according to the manufacturer's instructions . The amount of LDH released is expressed as a percentage of the total amount of LDH released from cells treated with lysis buffer ( regarded as 100% cytotoxicity ) . Oocysts of the Iowa C . parvum isolate [36] were propagated in experimentally infected newborn Cryptosporidium–free Holstein bull calves to obtain parasite material for study as previously described [37] , [38] . Oocysts were isolated by sucrose density gradient centrifugation , stored in 2 . 5% ( W/V ) potassium dichromate ( 4°C ) and used within 6 weeks of isolation [39] . Oocysts of the TU502 C . hominis isolate [40] , [41] were propagated in gnotobiotic piglets and isolated from feces at Tufts University [42] and used within 4 weeks of isolation . Prior to excystation , oocysts were treated with hypochlorite [37] . In vitro excystation ( 37°C , 0 . 15% [W/V] taurocholate , 2 h ) of oocysts used for all experiments was ≥90% . Sporozoites were isolated from excysted oocyst preparations by passage through a polycarbonate filter ( 2 . 0 µm pore size; Poretics , Livermore , California ) and used immediately . Sporozoite viability after incubation with peptides was assessed using fluorescein diacetate ( FDA ) and propidium iodide ( PI ) with modification [43] . In brief , freshly excysted sporozoites were incubated ( 15 min , 37°C ) in minimal essential medium ( MEM ) containing individual peptides ( 2 . 5 , 0 . 25 , 0 . 025 µM ) or in MEM alone ( n = 3 ) . Peptide concentrations were selected based largely on studies by our group and others evaluating the effects of various antimicrobial peptides on C . parvum viability [44]–[47] . Heat-killed ( 20 sec , 100°C ) sporozoites were used as a control . FDA ( 8 mg/ml final concentration ) and PI ( 3 mg/ml final concentration ) were then added to the sporozoite preparations and incubated further ( 5 min , 21°C ) . A minimum of 100 sporozoites were then examined by epifluorescence microscopy for each preparation , and the percent viability was determined . Percent reduction of viability was calculated as ( [MEM-treated sporozoite viability−peptide-treated sporozoite viability]÷MEM treated sporozoite viability ) ×100 . The mean values for test and control preparations were examined for significant differences using Student's t-test . CD11b+F4/80+ macrophages were derived ( 99% purity ) from the bone marrow of BALB/c mice by culturing with M-CSF ( ebioscience ) over 6 days and then seeded in RPMI ( with 10% FBS v/v ) at a concentration of 1×105 cells/ml . These cells were incubated with peptides ( 0 . 5 µM–5 µM ) for 1 h at 37°C , in the absence of mCSF . Following two washes with ice-cold PBS , cells were incubated with a combination of E . coli LPS ( 10 ng/ml; Sigma ) and IFNγ ( 10 ng/ml; BD Pharmingen ) overnight . Supernatants were then collected and the amount of TNF measured by ELISA according to the manufacturer's instructions ( BD Pharmingen ) . B cells were isolated from the spleens of BALB/c mice by negative selection using a B cell isolation kit containing biotin-conjugated mAbs to CD43 , CD4 , and Ter-119 ( Miltenyi Biotec ) and then seeded at a concentration of 1×106 cells/ml in RPMI ( with 10% FBS v/v ) . The cells were treated with a range of concentrations of peptides ( 0 . 5 µM–5 µM ) for 1 h at 37°C . After washing , the B cells were incubated with a combination of either E . coli LPS ( 10 µg/ml; Sigma ) and IL-4 ( 10 ng/ml; BD Pharmingen ) or E . coli LPS ( 10 µg/ml ) and IFNγ ( 200 ng/ml; BD Pharmingen ) for six days . Supernatants were then collected and the amount of IgG1 or IgG2a measured by ELISA ( Sigma ) . Statistical comparisons were performed with Prism 4 . 0 Software ( Graph- Pad ) , using two-tailed Student's t test for comparisons of two data sets , and ANOVA for multiple comparisons . Statistically significant differences were determined by a p value of <0 . 05 , <0 . 01 , **<0 . 001 . The biochemical properties for each of the HDMs and HDPs are presented in Table 1 . The cathelicidins are known to be highly cationic peptides . Except for the FhHDM-1p2 peptide , which has a net charge of 0 , all the peptides in the present study have a net positive charge ( +3 to +9 ) with a percentage of hydrophobic residues ranging from 34 to 44 . The Boman index is an estimated potential of peptides to bind to other proteins . For this index , a low value ( ≤1 ) suggests that a peptide has more antibacterial activity , whereas values ranging from 2 . 5–3 . 0 indicate that a peptide is multifunctional with hormone-like activities [48] . While BMAP-28 has a low Boman index ( 0 . 81 ) which correlates to its high antimicrobial activity , the Boman indices for the other peptides range from 1 . 34–3 . 11 . Overall , there is no striking difference between the biochemical properties of HDMs and HDPs . The atomic structures of the vertebrate HDPs LL-37 , CRAMP , BMAP-28 and SMAP-29 have been previously solved and determined that each form an amphipathic helix [31] , [49]–[51] ( figure 1 ) . The structures of LL-37 and BMAP-28 are shown in figure 1A as representative for this group of peptides . Secondary structure prediction of the parasite-derived peptides FhHDM-1p2 , SmHDM-1p146 , SmHDM-2p58 and Sm16-p73 revealed that all possessed regions likely to form alpha helices . Furthermore , helical wheel analysis showed that each molecule contained an alpha helix ( 32–35 amino acids ) toward the C-terminal that was distinctly amphipathic . The number of residues forming the hydrophobic face of the parasite molecules ranged from 6–9 ( figure 1B ) . Thus , like their vertebrate HDP counterparts , the secreted helminth parasite molecules also form distinct amphipathic helices . Mammalian HDPs have the ability to bind to and thus neutralize the bacterial endotoxin LPS [52]–[57] . The capacity of different mammalian and helminth peptides to bind LPS from E . coli O113:H10 was compared using the chromogenic Limulus amoebocyte assay ( Chromo-LAL ) . The Limulus amoebocyte lysate contains enzymes that are activated in a cascade of reactions in the presence of LPS [58]–[60] . The final enzyme in the series splits the chromophore , ρ-nitroaniline ( ρNA ) , from the chromogenic substrate , generating a yellow color . The amount of ρNA released is proportional to the amount of free LPS present in the system . Consistent with that published in the literature , mammalian HDP LL-37 , BMAP-28 , SMAP-29 and CRAMP inhibited the activation of the Chromo-LAL assay at a concentration of 250 pmol/ml , indicating that they interact with LPS and thus prevent the activation of the enzymatic cascade [52] , [53] , [55] , [61] ( figure 2 ) . BMAP-28 was the most potent , preventing the activation of the enzyme cascade with a Vmax of 3 . 1 and SMAP-29 was less effective with a Vmax of 26 . 5 . Of the helminth-derived HDPs , both Sm16-p73 and SmHDM-1p146 displayed no LPS-binding capacity with a Vmax of 40 . 4 and 31 . 1 , respectively , which were above the control reaction with no peptide ( Vmax of 28 . 5 ) . By contrast , the helminth peptide SmHDM-2p58 was the second best inhibitor of the series with a Vmax of 12 . 5 , just after BMAP-28 . We found that FhHDM-1p2 was capable of directly activating the Chromo-LAL assay itself ( data not shown ) and , therefore , its binding capacity could not be evaluated using this test . However , using a plate-binding assay we have previously demonstrated that FhHDM-1p2 does indeed bind to LPS [25] . The bactericidal properties of mammalian HDPs are mediated by direct antimicrobial activities , and are therefore easily evaluated as the minimal concentration capable of inhibiting visible microbial growth ( MIC ) against a panel of bacterial species . Consistent with previous reports [28] , [31] , [62] , [63] , we found that SMAP-29 and BMAP-28 were effective against a broad group of gram-negative bacteria , including E . coli , P . aeruginosa , and S . typhimurium and two gram-positive bacteria , S . aureus and S . epidermis , with MIC values of <0 . 25–8 µg/ml ( Table 2 ) . LL-37 and its mouse counterpart , CRAMP , showed bactericidal activity against gram-negative bacteria with MIC ranging from 2 to 8 µg/ml , but were ineffective against the gram-positive bacteria tested ( MIC<128 µg/ml ) . The inactivity of LL-37 against S . aureus and S . epidermis is consistent with other studies [64] . However , the anti-microbial effect of LL-37 on Staphylococcus could be strain dependent as several studies have reported an effect of this peptide on both S epidermis [65] and S . aureus [66] , [67] . Despite structural similarities with the mammalian peptides tested , none of the HDMs demonstrated bactericidal activity against any species of bacteria at the concentrations tested ( <0 . 25 to 128 µg/ml ) . We have recently shown that cationic peptides , including the cathelicidin LL-37 , are highly parasiticidal against the apicomplexan parasite C . parvum in vitro [47] . Using the same methodology , we compared the anti-parasite activity of the mammalian HDPs to that of the four helminth-derived peptides . In keeping with our previous data [35] , LL-37 exhibited parasiticidal activity at a concentration of 2 . 5 µM against C . parvum and the related species C . hominis ( figure 3 ) . The other mammalian cathelicidins showed greater parasiticidal activity , reducing the viability of protozoans at lower concentrations of 0 . 025 and 0 . 25 µM . Of particular note , BMAP-28 demonstrated the highest potency against both species of protozoan at 2 . 5 µM ( P<0 . 01 ) . In stark contrast to these results , was the relative absence of parasiticidal activity of the helminth-derived peptides , even at the highest concentration of 2 . 5 µM . The predominant mechanism of HDP bactericidal activity is the formation of pores in the membrane lipid bilayer , destroying its integrity and causing cell death [68] . However , this effect is not specific to bacterial cells and HDPs have also been shown to be cytolytic to eukaryotic cells , particularly at high concentrations [69] . TO-PRO is a membrane impermeant dye and therefore its detection within cells is indicative of pore formation . Using this dye , we demonstrated that , as expected , all mammalian peptides at a concentration of 50 µM ( equivalent to the highest concentration tested in the bactericidal assays ) induced the formation of pores in a murine macrophage cell line ( figure 4 ) . However , at the same concentration none of the helminth peptides exhibited this effect . To more completely assess the cytotoxicity of the peptides , we first examined their haemolytic activity against human RBCs at various concentrations ( 8–256 µg/ml ) . After one hour of co-incubation , all of the mammalian peptides induced concentration dependent hemolysis ( Table 3 ) . BMAP-28 was the most potent of all the peptides , with 50% of RBCs lysed at the lowest concentration of peptide tested ( 8 µg/ml ) and 70% at the highest concentration of 256 µg/ml . In comparison , at this highest concentration , the other mammalian peptides were much less cytotoxic , lysing only 14 . 5–29 . 3% of RBCs . Notably , under the same experimental conditions , the S . mansoni-derived peptides did not lyse the cells at any concentration tested . The F . hepatica-derived FhHDM-1p2 showed some low-level cytolytic activity , with 11 . 4% of cells lysed at the highest concentration of 256 µg/ml . Lactate Dehydrogenase ( LDH ) is a soluble cytosolic enzyme that is released into culture media following loss of membrane integrity resulting from either apoptosis or necrosis . Therefore LDH release is an indicator of cell membrane integrity and acts as a measure to assess cytotoxicity . Consistent with the demonstration of hemolysis , higher concentrations ( >10 µM ) of mammalian peptides also resulted in the death of murine macrophages , with the highest concentration tested ( 50 µM ) resulting in 100% cytotoxicity , compared to the effect of a lysis buffer ( figure 5 ) . By contrast , none of the helminth peptides induced cell death at any concentration tested . This lack of LDH detection was not due to enzyme inhibition by the helminth peptides; when the peptides were added directly to culture supernatant from lysed cells the LDH activity was unchanged . Consistent with these data , the cells treated with helminth peptides ( 10–50 µM ) looked morphologically normal by light microscopy ( data not shown ) . Activation of macrophages by microbial stimuli is central to the induction of innate immune responses . IFNγ is one of the key cytokines in the innate immune response to intracellular pathogens , and augments cellular responses to TLR ligands such as bacterial LPS [70] , [71] . To prevent excessive inflammation potentially leading to sepsis , HDPs have been shown to inhibit the response of macrophages to these inflammatory mediators using mechanisms that are independent of direct binding to LPS [72] . Significantly , macrophages isolated from animals and humans infected with helminth parasites are also hyporesponsive to stimulation with LPS and IFNγ [73] , [74] . Therefore , here we investigated whether helminth-derived peptides , like mammalian HDPs , could suppress the activation of inflammatory macrophages . At concentrations below cytotoxic levels ( <5 µM ) , all mammalian peptides significantly inhibited TNF production in response to the combined stimulation with LPS and IFNγ ( figure 6 ) . Titration of the peptide concentration showed that even at concentrations as low as 0 . 5 µM , the inhibitory activity was preserved . In contrast , at the lowest concentration tested , the helminth-derived peptides had no effect on the activation of macrophages . However , as the concentration was increased , the helminth peptides significantly suppressed the inflammatory response of macrophages and in most cases more effectively than the mammalian peptides ( figure 6 ) . It is worth noting that the helminth-derived peptides could be tested at concentrations up to 50 µM as they are non-toxic to cells , whereas due to their cytotoxicity the HDPs were not tested at concentrations above 10 µM ( figures 4 , 5 ) . At these higher concentrations ( 10 , 25 and 50 µM ) the helminth-derived peptides significantly reduced the production of TNF from activated macrophages in a concentration dependent manner ( data not shown ) . In addition to directly inhibiting inflammatory innate immune responses , there is evidence that mammalian HDPs have an additional role in regulating the magnitude of the adaptive antibody responses . For example , it has been shown that CRAMP functions to positively regulate the level of IgG1 produced by B cells [75] , and LL-37 reportedly decreased the production of IgG2a from mouse splenic B cells activated with LPS and IFNγ [76] . Consistent with these reports , our analyses showed that with the exception of BMAP-28 , all the mammalian HDPs significantly increased the production of IgG1 in response to a Th2 biased environment ( LPS and IL-4 ) ( figure 7A ) . The apparent reduction in IgG1 production recorded for the higher concentration of BMAP-28 , likely reflects some level of cell death rather than a reduction in antibody production . Due to this cytotoxicity the HDPs were not tested at concentrations above 5 µM . While there was greater variability between the HDMs , peptides from F . hepatica and S . mansoni significantly increased the production of IgG1 in response to LPS and IL-4 even at low concentrations . Conversely , mammalian peptides reduced the production of IgG2a in a Th1 ( LPS and IFNγ ) biased environment ( figure 7B ) . For the helminth peptides , only concentrations above 5 µM had the same effect on B cells , significantly ( p<0 . 001 ) inhibiting IgG2a secretion ( data not shown ) , suggesting a lower potency than mammalian-derived peptides . Parasitic helminths secrete molecules that modulate host immune responses to establish an environment that facilitates their survival and a prolonged reproductive phase [20] , [77] , [78] . Co-evolution of helminths with their hosts means that these parasites are well adapted to the host's immune system , making use of endogenous regulation mechanisms to manipulate the immune response to their benefit . In this study , we compared the biological activities of a series of helminth-derived cathelicidin-like peptides to that of their mammalian homologues and suggest how their production by helminths can facilitate a successful parasitic life cycle . In vitro , most mammalian HDPs are effective antimicrobial agents against a range of organisms including gram-negative and gram-positive bacteria , protozoa , viruses and fungi [10] , [79] , [80] . In general , the expression of HDPs is increased at the onset of an infection and therefore the anti-pathogenic activity was thought to be one of the most important immediate responses that the mammalian host evolved to deal with invading pathogens . It has been proposed that the specificity of HDPs for particular microbes is subjected to significant variation and is particularly influenced by the types of microbial biotas to which each HDP species is exposed [81] . Consistent with this theory , we showed some variation in the anti-microbial capabilities of the mammalian HDPs examined . While sheep and bovine derived peptides were effective against a broad range of both gram-positive and gram-negative bacteria , the mouse and human HDPs were largely ineffective against the gram-positive species tested . However , we found that none of the helminth-derived peptides displayed gram positive or negative bactericidal activity , even at the highest concentration tested , implying that their specialised function is not anti-microbial . However , we cannot exclude the possibility of the peptides having anti-microbial activity on other , untested pathogenic bacteria . Despite lacking bactericidal activity , we showed previously that FhHDM1-p2 , like the mammalian peptides , interacts with LPS , thus effectively neutralising the ability of infecting bacteria to induce an inflammatory response [25] . We suggested that this may be a mechanism used by the parasite to prevent excessive activation of innate cells in response to the translocation of microbes into circulation occurring as a result of damage to the skin and/or gut epithelium during migration of the parasite [25] . As the two major factors mediating interaction between LPS and HDPs are hydrophobicity and cationicity [82] , inspection of the sequence of the other HDMs would predict a universal ability to bind to LPS . However , while SmHDM-1p146 appeared to bind LPS as efficiently as the mammalian peptides , neither SmHDM-2p58 or Sm16-p73 were particularly potent , indicating that the neutralization of LPS may not be a common function of the helminth-derived peptides . A number of studies have demonstrated the ability of amphibian and mammalian HDPs to kill protozoan parasites in vitro [83] , [84] . For example , BMAP-28 possesses potent activity against the agent of human leishmaniasis , Leishmania major [83] , and we have shown that LL-37 can reduce the viability and infectivity of sporozoites of C . parvum , an intestinal infection of humans and agricultural animals [47] . In the present study we confirm the activity of LL-37 against C . parvum and the related parasite C . hominis and show that the other mammalian cathelicidins tested also have anti-protozoan activity; BMAP-28 exhibited the most potent in vitro activity . By contrast , and consistent with their lack of antibacterial activity , the helminth-derived HDMs did not kill either parasite in vitro . Clearly , the particular physico-chemical properties of HDMs do not confer an ability to penetrate and disrupt the surface membrane of these parasitic organisms . Although widely defined as antimicrobial , in fact , at the concentrations normally found at human mucosal surfaces and in physiological salt conditions , the mammalian cathelicidin-like peptides do not display bactericidal activity . However , at these same concentrations and under the same conditions , the peptides exhibit a variety of immune modulatory functions [12] , [85] . This has led to the suggestion that the cathelicidins are principally immune modulators rather than antimicrobials , and like the mammalian defensins , have traded their bactericidal capacities to acquire the ability to broadly regulate the immune response [12] , [86] . The mammalian cathelicidins have a diverse effect on the cellular immune response , but in particular , the peptides have a crucial role in regulating TLR-dependent innate inflammatory responses . This means that they function to maintain homeostasis in response to natural shedding of microflora-TLR agonists as well as controlling the systemic inflammatory response to infection or tissue damage [12] , [86] , [87] . We have shown here that the helminth-derived cathelicidin-like peptides also regulated the innate immune response to TLR stimulation by inhibiting TNF release from macrophages stimulated with bacterial LPS . While the helminth HDMs were not as effective as their mammalian homologues at the lowest concentration tested , they were correspondingly more potent as their concentration was increased towards quantities that are likely secreted during infection . It is probable that the function of these secreted HDMs is similar to the predicted role for the mammalian HDPs , i . e . prevention of an excessive inflammatory response , which acts to prevent the expulsion of the worm and to protect the host from exacerbated tissue damage . The role of HDPs in regulating the adaptive immune response has been less extensively studied . Recent studies have shown that LL-37 decreased the production of IgG2a from murine B cells stimulated with LPS and IFNγ [76] , and that CRAMP increased the amount of IgG1 in response to IL-4 [75] . These results are consistent with the suggestion that HDPs are engaged in the process of infection resolution and wound healing , as autoreactive IgG1 antibody production is central to tissue repair processes [88] , while IgG2a are associated with IFNγ/Th1-mediated inflammatory responses . Consistent with these reports , the helminth-derived peptides performed in a similar manner to their mammalian counterparts , successfully enhancing the production of IgG1 . At high concentrations , the helminth-derived peptides were far more effective than mammalian HDMs at suppressing the release of IgG2a in response to IFNγ , supporting their role as potent anti-inflammatory agents . Due to their immune modulatory activities , there is considerable interest in developing HDPs as therapeutics such as anti-inflammatory agents , adjuvants and wound healing agents . The therapeutic potential of HDPs has been demonstrated and a number of peptides are being developed as anti-inflammatory agents [89] . However , the clinical use of these peptides as injectable therapeutics has been hampered by indications of toxic side-effects on mammalian cells and their ability to lyse eukaryotic cells [90] , [91] . This had led to intense research into understanding how HDPs function in terms of their physico-chemical properties . Among the factors that appear to influence specificity between their activity against prokaryotic and eukaryotic cells are the ability to form an amphipathic α-helical structure , hydrophobicity , overall charge distribution , and minimal peptide length [92] . At first sight , the biochemical characteristics of the helminth-derived HDMs would predict an inherent cytotoxic activity: they form amphipathic helices ( figure 1 ) , they have a comparable proportion of hydrophobic amino acids to mammalian HDPs and most of them are cationic ( Table 1 ) . However , based on the assays employed in this study , we found no correlation between the level of hydrophobicity and cytotoxic activity . The majority of mammalian HDPs have an overall net charge ranging from +4 to +6 . 37 [93] , implying an optimal range for biological activity . HDPs with a net positive charge of <+4 are found to be inactive , whereas increasing the net charge from +4 to +8 confers antimicrobial activity and some haemolytic activity [79] . Three of the HDMs used in this study , FhHDM-1p2 , SmHDM-1p146 and SmHDM-2p58 , all had a net charge <+4 which is consistent with a non-bactericidal , non-haemolytic peptide . However , despite possessing a net charge of +5 , Sm16-p73 displayed neither antimicrobial nor cytotoxic activities . Recent studies have proposed that rather than a simple correlation between net charge and haemolysis it is the localisation of the positively charged amino acids within the peptide that also dictates membrane interaction and selectivity . By increasing the charge of an amphipathic HDP analog from +8 to +9 , by the addition of one positive charge on the polar face , the haemolytic activity of the peptide was enhanced 32-fold [79] . Using this parameter , we calculated that Sm16 has four positively charged amino acids on its polar face compared to six for LL-37 , CRAMP and BMAP-28 and seven for SMAP-29 , which may provide some clue as to the difference in cytotoxicity between these peptides . Likewise , all the helminth HDMs used in this study have less positive charges on their polar face compared to LL-37 , which was the least cytotoxic of all the mammalian peptides examined in this study . It is generally accepted that the cytoplasmic membrane is the main target of many mammalian HDPs , whereby peptide accumulation in the membrane causes increased permeability and a loss of barrier function , resulting in the leakage of cytoplasmic components and cell death [94] . However , the cytotoxic concentrations of HDPs are higher than the concentrations required for the destruction of microbes , which , some authors suggest , reveals a cell-selective killing mechanism [85] , [94] . The physiological concentration of HDPs at mucosal sites is typically less than 2 µg/ml [12] , [80] , well below the concentration that is cytotoxic to mammalian cells in vitro . We have found that helminth HDMs are abundant molecules within the secretions of helminth parasites , which during a multi-parasite infection would likely be at relatively high concentrations in circulation . Therefore , it is essential for the success of the parasite that these peptides do not possess cytotoxic activity , whilst at the same time retain the beneficial immune modulatory properties . The complete absence of antimicrobial activity by helminth-derived peptides is likely linked to this need to prevent host cell death during infection . The extraordinary capacity of helminths to regulate the immune response is central to their longevity in the mammalian host and thus underpins their success as parasitic organisms [77] , [78] . Therefore , it is perhaps unsurprising that helminth secretory products contain homologues of components of the host immune system that target the same mammalian pathways . In addition to the HDMs identified here , helminth parasites express highly conserved cytokine gene families that , like their mammalian counterparts , ligate specific receptors on immune cells . Brugia malayi and Ancylostoma ceylanicum express homologues of the mammalian cytokine macrophage migration inhibitory factor ( MIF ) [95] , and in a Th2 environment , such as that activated by helminth infection , Brugia MIF synergises with IL-4 to induce the development of regulatory M2 macrophages [96] . Helminths also express members of the Tumour Growth Factor- ( TGF ) β and TGF-β receptor superfamilies [97]–[99] , and similar to the mammalian cytokine , Heligmosomoides polygyrus TGF-β homologue has been shown to directly induce the differentiation of regulatory T cells , demonstrating a key role in parasite immune regulation [100] . The cysteine protease inhibitors , cystatins , are an ancient and conserved family of peptides in the animal and plant kingdoms [101] . The cystatins of parasitic worms differ substantially from those produced by free-living nematodes with regard to their immune modulatory properties [102] . In particular , the acquisition of an asparaginyl endopeptidase site , similar to that of vertebrate cystatin C , confers an ability to reduce the activation of host T cell responses by directly inhibiting the presentation of antigen by dendritic cells [103] , [104] , suggesting a specific adaption to regulate host immune responses [102] , [105] . Similar to the cystatins , HDPs are conserved in all organisms , including plants , animals and humans [106] . However , we show here that while the helminth-derived HDMs are effective immune modulators , they display no bactericidal activity . These observations would suggest that like the cystatins , HDMs have become specifically adapted to support a parasitic lifestyle , losing the more ancient property of direct antimicrobial killing but acquiring the ability to regulate immune responses in order to promote their survival within the mammalian host . It is possible that HDM immune modulation arose in trematodes following their divergence from the chordate lineage ( as acoelomates ) and their subsequent specialisation to an endoparasitic lifestyle , distinct from the free-living acoelomate turbellarian flatworms from which HDM homologues have yet to be identified ( unpublished observation ) . The immune-modulatory properties of mammalian HDPs , and in particular their ability to prevent excessive inflammatory induced pathology associated with bacterial sepsis , has attracted interest in exploiting these as anti-infectives . However , their cytotoxicity , as also shown in the present study , has presented a major drawback for their in vivo use . Accordingly , the absence of cytotoxicity and retention of immune-modulatory activity observed for the helminth-derived HDMs offer an opportunity to design novel immunotherapeutics to combat microbial pathogens and immune-related disorders .	In mammals , secreted host defence peptides ( HDPs ) protect against a wide range of infectious pathogens . They also perform a range of immune modulatory functions which regulate the immune response to pathogens , ensuring that the protective inflammatory response is not exacerbated and that post-infection repair mechanisms are initiated . We identified a novel family of molecules secreted by medically-important helminth pathogens ( termed helminth defence molecules; HDMs ) that exhibit striking structural and biochemical similarities to the HDPs . To further investigate the extent of this similarity , we have performed a comparative functional study between several well characterized , anti-microbial , mammalian HDPs and a series of parasite-derived peptides . The parasite HDMs displayed immune modulatory properties that were similar to their HDP homologs in mammals , but possessed no antimicrobial or cytotoxic activity . We propose that HDMs of these helminth pathogens underwent specific adaptation , losing their anti-microbial activity but retaining their ability to regulate the immune responses of their mammalian hosts . This absence of cytotoxicity and retention of immune-modulatory activity offers an opportunity to design novel immunotherapeutics derived from the HDMs which could be used to combat destructive inflammatory responses associated with microbial infection and immune-related disorders .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "infectious", "diseases", "immunology", "biology" ]	2013	Cathelicidin-like Helminth Defence Molecules (HDMs): Absence of Cytotoxic, Anti-microbial and Anti-protozoan Activities Imply a Specific Adaptation to Immune Modulation
Rapid pathogen identification during an acute febrile illness is a critical first step for providing appropriate clinical care and patient isolation . Primary screening using sensitive and specific assays , such as real-time PCR and ELISAs , can rapidly test for known circulating infectious diseases . If the initial testing is negative , potentially due to a lack of developed diagnostic assays or an incomplete understanding of the pathogens circulating within a geographic region , additional testing would be required including highly multiplexed assays and metagenomic next generation sequencing . To bridge the gap between rapid point of care diagnostics and sequencing , we developed a highly multiplexed assay designed to detect 164 different viruses , bacteria , and parasites using the NanoString nCounter platform . Included in this assay were high consequence pathogens such as Ebola virus , highly endemic organisms including several Plasmodium species , and a large number of less prevalent pathogens to ensure a broad coverage of potential human pathogens . Evaluation of this panel resulted in positive detection of 113 ( encompassing 98 different human pathogen types ) of the 126 organisms available to us including the medically important Ebola virus , Lassa virus , dengue virus serotypes 1–4 , Chikungunya virus , yellow fever virus , and Plasmodium falciparum . Overall , this assay could improve infectious disease diagnostics and biosurveillance efforts as a quick , highly multiplexed , and easy to use pathogen screening tool . Appropriate diagnostic assay selection for infectious diseases depends on multiple parameters including clinical presentation and endemic pathogens known to circulate within a specific geographic region . Rapid point-of-care PCR [1 , 2] and lateral flow immunoassays [3 , 4] as well as more complex PCR [5–7] and laboratory based antigen capture ELISAs [8 , 9] can generate a clinically actionable diagnosis in patients presenting with an acute febrile illness . These assays are sensitive , rapid , and relatively inexpensive , making this testing approach ideal for initial diagnostic testing . If these assays are negative , however , additional testing including increasingly multiplexed assays and agnostic next-generation sequencing can be utilized . Multiplexed assays such as the MAGPIX [10 , 11] or multiplexed real-time PCR [12–15] can increase the number of targets being tested . For example , Munro and colleagues described a multiplexed PCR assay with detection on the MAGPIX or Luminex instruments capable of detecting multiple influenza viruses with performance similar to real-time PCR [11] . Similarly , a multiplexed real-time RT-PCR assay , developed by Santiago and colleagues and approved by the FDA as an in vitro diagnostic device , detects all four dengue virus serotypes in a single tube reaction [15] . In cases where the initial testing methods do not result in positive pathogen identification , next-generation sequencing ( NGS ) is another alternative for clinically actionable infectious disease diagnostics [16] . However , metagenomic sequencing can be challenging due to a large host background , necessitating high sequencing depth to generate sufficient on target reads for pathogen detection . Targeted NGS , in which a specific signature is amplified [17 , 18] or enriched from a complex sample using hybridization [19] , can increase pathogen specific reads sufficiently to allow detection on desktop sequencers such as the Ion Torrent or the MiSeq . Using these approaches , however , adds time-to-answer due to library preparation , sequencing , and analysis . A potential solution described here is the use of the NanoString nCounter platform for highly multiplexed pathogen detection . This system utilizes direct hybridization and detection of a nucleic acid target and can be highly multiplexed ( up to 800 different targets ) . Since this technology has been successfully implemented for quantitative gene expression studies [20–22] , we investigated whether this platform could be used for broad , targeted pathogen detection in a situation where rapid testing ( ex . real-time PCR ) was negative . In this context , we developed and evaluated a panel containing 195 different assay targets against 164 different viruses , bacteria and parasites . Overall , this panel was not as sensitive as real-time PCR; however , this assay successfully identified multiple pathogens quickly , demonstrating utility as a pathogen screening assay . All organisms used in this study ( listed in S1 File ) are maintained at United States Army Medical Research Institute of Infectious Diseases ( USAMRIID ) or were provided by the Unified Culture Collection ( UCC ) or the American Type Culture Collection ( ATCC , Manassas , VA ) . Samples included bacterial , parasite DNA , cell culture supernatant from virus-infected cells treated with TRIzol LS ( ThermoFisher Scientific , Waltham , MA ) or gamma irradiation . Total nucleic acid from each unpurified sample was extracted using the EZ1 Virus Mini Kit v2 . 0 ( Qiagen , Valencia , CA ) with the EZ1 robot ( Qiagen ) according to the manufacturer’s instructions . Total nucleic acid was eluted in 90 μl elution buffer . Due to a limited supply , Coxiella burnetii DNA was amplified using the REPLI-g Whole Genome Amplification Kit ( Qiagen ) according to the manufacturer’s instructions . The number of C . burnetii genome equivalents ( GE ) was approximated using the genome of C . burnetii RSA493 ( GenBank# NC_002971 ) and the C+G ( 42 . 7% ) and A+T ( 57 . 3% ) genome percentages . Based on these calculations , 1 GE is approximately 2 . 05 fg . The approximate number of GE for Plasmodium falciparum 3D7 DNA ( ATCC ) was similarly determined to be approximately 23 . 89 fg . A custom Broad Pathogen Detection Assay ( BPDA ) targeting a broad panel of medically important viruses , bacteria , and parasites was designed and acquired from NanoString Technologies ( Seattle , WA ) . Using 195 different capture and reporter probes , this assay targeted 164 different pathogens of concern for human health ( S1 File ) . After initial testing showed lower than desired assay sensitivity , nested primers targets were designed by NanoString using Primer3 software [23–25]; see S1 File for the sequences . These multiplexed primers were used in a multiplexed target enrichment ( MTE ) reaction to amplify the capture/reporter target prior to detection . Primer pairs for 4 probe targets could not initially be designed and were redesigned for incorporation into a subsequent MTE iteration . Individual MTE primer pairs for all available pathogens were evaluated for amplicon generation using SuperScript One-Step with Platinum taq ( Thermo Fisher Scientific ) with the following cycling conditions: 50°C for 15 minutes , 95°C for 5 minutes , 40 cycles of 95°C for 30 seconds , 60°C for 1 minute and 72°C for 1 minute . The final reagent concentrations per 20 μL reaction were: 1X Reaction Mix , 4 mM MgSO4 , 0 . 25 mg/mL BSA , 50 μM primers , 0 . 4 units of Platinum Taq . Amplicon generation was visualized on the 2100 Bioanalyzer ( Agilent Technologies , Santa Clara , CA ) using the DNA 1000 Kit ( Agilent Technologies ) . After confirming successful amplification using selected individual primer pairs , all primers were combined into a single 500 nM primer mixture for the MTE reaction . Sample cDNA was generated by adding 4 μL purified total nucleic acid to 1μL of SuperScript VILO MasterMix ( Thermo Fisher Scientific ) and incubating at 25°C for 10 minutes , 42°C for 60 minutes , and 5 minutes at 85°C . The sample was then added to 5 μL of TaqMan PreAmp Master Mix ( Thermo Fisher Scientific ) and 1 μL of the 500 nM primer mixture . MTE used the following cycling conditions: 94°C for 10 minutes , then 18 cycles of 94°C for 15 seconds , 60°C for 4 minutes , and a 4°C hold . The entire enriched sample ( 11 μL ) was used for detection on the NanoString nCounter platform . Pathogen nucleic acid ( total nucleic acid and MTE amplified nucleic acid ) was initially denatured at 95°C for 5 minutes and immediately transferred to ice for 2 minutes . Next , 5 μl of each denatured sample or the entire MTE reaction ( 11 μl ) was added to a 20 μl NanoString master mix containing the BPDA Reporter Codeset plus 130 μL hybridization buffer followed by 5 μl of the BPDA Capture Codeset . Reactions were immediately placed on a thermocycler for an overnight incubation at 65°C ( for ~16 hours ) , loaded onto a sample cartridge using the nCounter Prep Station , and scanned using the NanoString nCounter Digital Analyzer . A sample was called positive for a specific target if the number of counts was greater than the average of the internal negative controls for that target plus three times the standard deviation of the negative controls . The impact of the MTE reaction on sensitivity was assessed using SYBR Green real-time PCR assays using primers internal to the MTE primers [See S1 File for Bundibugyo virus ( BDBV ) , Marburg virus ( MARV ) , Ebola virus ( EBOV ) , influenza B virus , and P . falciparum assay information] . Total nucleic acid for each organism was serially diluted and amplified by MTE . The levels of enrichment were measured by real-time RT-PCR using Superscript RT-PCR reagents ( Thermo Fisher Scientific ) and SYBR Green . The cycling conditions for the SYBR Green RT-PCR assays were: 50°C for 15 minutes , 95°C for 5 minutes , 45 cycles of 95°C for 5 seconds and 60°C for 20 seconds , and a final melt step from 60°C—95°C at a rate of 0 . 2°C per second . The final reagent concentrations per 20 μL reaction were as follows: 1X Reaction Mix , 3mM MgSO4 , 0 . 25 mg/mL BSA , 1 μM primers , 2X SYBR Green , and 1 unit of Platinum Taq . Following optimization and characterization of the MTE reaction , assay performance was determined utilizing the MTE reaction and detection on the NanoString platform . A preliminary limit of detection ( LOD ) for BDBV , MARV , EBOV , influenza B virus , and P . falciparum was determined with and without MTE by serially diluting organism and testing for positive detection . The preliminary LOD was defined as the lowest concentration of organism having all three replicates testing positive . Testing at the preliminary LOD was repeated ( ten replicates ) without MTE for BDBV and MARV Angola and with MTE for influenza B virus , BDBV , and MARV Angola to show assay reproducibility . Similarly , LODs were conducted using existing real-time RT-PCR assays as previously described [26 , 27] . Mock clinical samples were prepared in order to evaluate the ability of the BPDA to detect samples that had been extracted from whole human blood . Dengue virus serotype 3 ( DENV-3 ) in TRIzol LS and gamma-irradiated EBOV were diluted in human whole blood treated with EDTA ( Bioreclamation , Westbury , NY ) in 200 μl samples . Samples were extracted using the Qiagen EZ1 XL Advanced with the EZ1 Virus Mini kit 2 . 0 by adding an equal volume of ATL buffer ( Qiagen ) to each sample prior to being placed on the automated instrument for extraction . Each sample was run in triplicate with and without MTE amplification prior to being tested with the pathogen panel . Since the BPDA assay could be a useful diagnostic tool , de-identified human clinical samples were tested . These samples were acquired through USAMRIID’s Special Pathogens Laboratory . All samples were de-identified prior to use , and all studies were conducted in compliance with United States Department of Defense , federal , and state statutes and regulations relating to the protection of human subjects , and adheres to principles identified in the Belmont Report . All data and human subjects research were gathered and conducted for this publication under an Institutional Review Board approved determination FY17-31 as defined by 32 CFR 219 . 102 ( f ) . This sample set comprised of Chikungunya virus positive and negative samples , as determined by real-time PCR [28] . Total nucleic acid was extracted from each sample using the Qiagen EZ1 XL Advanced with the EZ1 Virus Mini kit 2 . 0 and tested using the BPDA . We developed a Broad Pathogen Detection Assay ( BPDA ) for use with the NanoString nCounter platform in order to quickly screen a sample for multiple pathogens in a single tube reaction . This assay consisted of 195 detection probes targeting 164 different viral , bacterial , and parasitic pathogens of concern for human health . Initial testing using the highest concentration of organism available showed positive detection of multiple pathogens including EBOV , MARV , P . falciparum , and C . burnetii ( see Table 1 for a selected list and the S1 File for a full detection list ) . However , some pathogens such as Crimean-Congo hemorrhagic fever virus ( CCHFV ) were not detected even at this high concentration ( Table 1 ) . A multiplexed target enrichment ( MTE ) step utilizing a complex PCR to amplify the pathogen-specific probe hybridization site was used for use prior to detection with the BPDA to mitigate this issue . Incorporating this upfront target enrichment step increased the assay sensitivity as shown by the now positive detection of CCHFV and increased read counts for almost all pathogens tested ( Table 1 ) . Having shown the effectiveness of the MTE step for increasing assay sensitivity , we wanted to further assess the target enrichment capability of this method by comparing the amount of target amplicon present with and without MTE . Comparing real-time PCR results for the MTE amplified and non-amplified reactions , MTE enrichment showed a decrease in Cq values , indicating an increase in the target amplicon ( Fig 1 ) . Statistical analysis ( two-way ANOVA with Bonferroni correction ) identified that all points for each virus were significantly different with the exception of the highest influenza B virus concentration . Assay limit of detection ( LOD ) studies were conducted with serially diluted organism in order to define assay performance for medical relevant concentrations of the tested organism . Incorporating the MTE enrichment improved detection and lowered LODs ( Fig 2 ) . This improvement was most notable for influenza B virus which was undetectable without enrichment but tested positive following MTE ( Fig 2B ) . Similarly , the preliminary LOD for MARV improved from 2 . 2 x 106 PFU/ml without MTE to 3 . 75 x 104 PFU/ml after enrichment ( Fig 2D ) . Generally , the assays were highly specific for the targeted organism . For example , BPDA showed positive results for only P . falciparum while other Plasmodium species including knowlesi , malariae , ovale , and vivax were called as true negatives . To confirm assay reproducibility at the preliminary LOD , ten replicates of BDBV , influenza B virus , and MARV were tested with and without MTE ( Table 2 ) . For influenza B virus , no replicates tested positive at the highest concentration used; however , MTE incorporation resulted in repeated detection of all viruses . In addition , comparison of BPDA performance to real-time PCR , the current gold standard for molecular based pathogen detection , showed real-time RT-PCR was the more sensitive technique ( Table 2 ) . Testing EBOV and DENV-3 spiked into whole blood at three different concentrations showed the clinical applicability of this assay . Application of the MTE did not impact the overall testing results , but MTE increased the number of EBOV-specific counts for all dilutions and replicates ( Table 3 ) . DENV-3 tested positive with all replicates and dilutions; however , MTE did not result in increased DENV-3 counts ( Table 3 ) . Interestingly , the pre-amplification of DENV-3 signatures resulted in a lower number of counts as compared to the same samples without MTE , potentially suggesting suboptimal amplification for that DENV-3 isolate . Further characterization of the clinical utility of the BPDA showed positive detection across 14 de-identified , human clinical samples with potential Chikungunya virus ( CHIKV ) infections ( Table 4 ) . Both the BPDA and the MTE-BPDA assays correctly identified the 12 real-time RT-PCR positive samples and the two negative samples ( Table 4 ) . Incorporating the MTE component increased the number of CHIKV counts by 1–2 log for all of the positive samples tested . Positively identifying the etiologic agent for an acute febrile illness can be critical for ensuring appropriate administration of treatment and supportive care; however , proper identification can be challenging . Highlighting the importance of broad pathogen screening and appropriately fielded diagnostics , a recent study by Schoepp and colleagues found ~70% of the suspected Lassa fever patients admitted to the Lassa Fever Ward in Kenema , Sierra Leone , were negative for both Lassa virus and the malaria parasite , both hyperendemic pathogens in the region [29] . This study found serological evidence of filovirus infection ( EBOV and MARV ) in the years prior to the explosive Ebola virus disease outbreak in West Africa [29] , and it is likely that previous infections with EBOV as well as MARV were misidentified as severe malaria or Lassa fever . Accurate testing of these acute febrile patients with multiplexed assays could have identified the risk of an EBOV outbreak earlier . Here , we developed and evaluated a highly multiplexed , broad pathogen detection assay for use following negative detection using singleplex assays ( ex . real-time PCR ) . This assay targets 164 different human pathogens of public health concern and includes viruses , bacteria , and parasites . We included multiple organisms with overlapping clinical presentation , such as Plasmodium , Lassa virus , and Ebola virus . We also included a large number of less common organisms in order to maximize the diagnostic potential of the assay . While the assay performed well without target enrichment , applying MTE prior to running on the NanoString platform greatly improved assay sensitivity . While extensive primer optimization was not conducted in this study , such an optimization would likely improve the overall assay sensitivity and the detection variation we observed . As we were unable to do this optimization , all testing was conducted with and without MTE . Overall , 98 of the 164 pathogens on the panel that we had available for testing were positively identified including endemic pathogens to West Africa such as EBOV , Lassa virus , dengue virus , and the malaria parasite P . falciparum . Assay run time , from start to finish , is approximately 27 hours with approximately one hour ( or 30 minutes without the MTE ) of hands-on time . Future efforts include the acquisition and testing of the remaining pathogens on the panel . Characterization of the BPDA using mock clinical and clinical samples showed the efficacy of this assay for detecting pathogen in patient samples . Specifically , testing of spiked human serum showed positive detection of EBOV and DENV-3 at clinically relevant concentrations . In addition to mock clinical samples , the assays correctly identified all CHIKV human clinical samples , demonstrating the ability to correctly identify pathogens from natural infections . These studies were in agreement with real-time RT-PCR testing establishing preliminary 2-by-2 testing that would be required for regulatory use . There are a variety of easy to use multiplexed assays described in the literature [6 , 7 , 30 , 31]; however , there are inherent limitations in multiplexability within a single assay . Other technologies offer higher levels of multiplexing through microarray [32 , 33] and next-generation sequencing [16–19 , 34] . However , these assays are highly technical , and the large number of targets makes full validation of each signature highly challenging in both cost and time . Furthermore , clinical validation of these assays would be further complicated by the regulatory requirement to validate each potential organism the assay could detect [35 , 36] . Preliminary limit of detection ( 5 dilutions in triplicate for 164 organisms ) and confirmation of the preliminary limit of detection ( twenty replicates of 164 organisms ) testing alone would require 2 , 460 reactions at approximately $100 USD/reaction ( ~$575 , 000 USD ) . Mock clinical testing would further expand the testing numbers and cost . Ideally , comparison of primer/primer interactions and a direct comparison of each target to a gold standard ( ex . real-time PCR ) would be performed for diagnostic applications . Within current regulatory paradigm , this type of validation also remains cost prohibitive similar to other highly multiplexed assays ( ex . microarray and next-generation sequencing assays ) as this would require independent testing of each target on the panel in a statistically robust manner . However , results presented here provide proof of concept testing results and the framework for such a validation by demonstrating the proof of concept utilization of this technology for infectious disease diagnostics .	Identifying the causative agent in an acute febrile illness can be challenging diagnostically , especially when organisms in a particular region have overlapping clinical presentation or when that pathogen’s presence is unexpected . Ebola virus , for example , was not considered in an acute febrile illness differential diagnosis in West Africa until the explosive outbreak in 2013 presented the risk of infection . Besides the cost and time of screening a single patient sample for a large number of pathogens , limited sample volumes place further restrictions on what assays can be applied . Here , we developed a broad pathogen screening assay targeting 164 different human pathogens and show positive detection of over 100 of the organisms on the panel including Ebola virus , Plasmodium falciparum , and a large number of rare pathogens . The hands on time and sample volume requirement is minimal . The assay performed well in mock clinical and human clinical samples , demonstrating the clinical utility of this assay in cases where the initial diagnostic testing results in negative results . Our results provide a framework for further validation studies that would be required for formal clinical diagnostic applications .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "dengue", "virus", "medicine", "and", "health", "sciences", "parasite", "groups", "pathology", "and", "laboratory", "medicine", "togaviruses", "plasmodium", "pathogens", "microbiology", "orthomyxoviruses", "alphaviruses", "parasitology", "viruses", "organisms", "apicomplexa...	2018	A highly multiplexed broad pathogen detection assay for infectious disease diagnostics
Alignments of orthologous protein sequences convey a complex picture . Some positions are utterly conserved whilst others have diverged to variable degrees . Amongst the latter , many are non-exchangeable between extant sequences . How do functionally critical and highly conserved residues diverge ? Why and how did these exchanges become incompatible within contemporary sequences ? Our model is phosphoglycerate kinase ( PGK ) , where lysine 219 is an essential active-site residue completely conserved throughout Eukaryota and Bacteria , and serine is found only in archaeal PGKs . Contemporary sequences tested exhibited complete loss of function upon exchanges at 219 . However , a directed evolution experiment revealed that two mutations were sufficient for human PGK to become functional with serine at position 219 . These two mutations made position 219 permissive not only for serine and lysine , but also to a range of other amino acids seen in archaeal PGKs . The identified trajectories that enabled exchanges at 219 show marked sign epistasis - a relatively small loss of function with respect to one amino acid ( lysine ) versus a large gain with another ( serine , and other amino acids ) . Our findings support the view that , as theoretically described , the trajectories underlining the divergence of critical positions are dominated by sign epistatic interactions . Such trajectories are an outcome of rare mutational combinations . Nonetheless , as suggested by the laboratory enabled K219S exchange , given enough time and variability in selection levels , even utterly conserved and functionally essential residues may change . Universally-spread proteins suggest a common ancestor dating back to the last universal common ancestor ( LUCA ) , ∼3 . 5 billion years ago [1]–[3] . They indicate that most positions have drifted from their original state , but many , typically active-site positions , remained unchanged . However , many [4] , if not most [5] , exchanges between orthologous sequences are incompatible — namely , amino acids that are functional in one sequence can be deleterious in related ones . Thus , as protein sequences evolve , they navigate a rugged fitness landscape formed by a complex network of epistatic interactions [4] , [6]–[11] . Computational analysis suggests that this ruggedness distinctly relates to sign epistasis , and not to any other mechanism that could limit divergence ( e . g . a large fraction of mutations that are intolerable ) [12] . Sign epistasis relates to a qualitative rather than a quantitative effect , namely a mutation that is beneficial within one genetic background but deleterious at another . This form of epistasis results in a very slow yet constant divergence rate , suggesting that protein sequences have not yet reached the limits of their divergence potential [6] . Experimentally , protein sequence divergence and epistasis have been studied [4] , [10] , [13] , [14] . It remains unclear , however , how sequence space is traversed at crucial active-site positions , particularly in positions that have a key functional role and are highly conserved . Similarly , the mechanism ( s ) leading to incompatibility — i . e . , an originally tolerated exchange becoming unacceptable in orthologous sequences , are also unclear . Our case study regards an ancient substitution that occurred within the active site of the highly ubiquitous phosphoglycerate kinase ( PGK ) . PGK is an essential glycolytic enzyme that catalyzes the interconversion of 1 , 3-bisphosphoglycerate and ADP to 3-phosphoglycerate and ATP . Its enzymatic activity is directly related to growth and thereby to organismal fitness [15] . PGK can be traced back to LUCA [1] , [2] , [16] and its evolutionary rate is extremely slow . It is categorized in the 10th percentile of slowly evolving proteins [17] . Indeed , the sequence identity between the human and the archaeon Methanococcus mazei PGKs is 32% , and nearly all active-site positions are entirely conserved in all known PGKs . However , Position 219 comprises an exception . It is lysine in all known bacterial and eukaryotic PGKs , and this lysine was shown to play a critical role in ATP binding and phosphate transfer [18] . The ancestor of all PGKs is also predicted to have had a lysine at position 219 . Archaeal PGKs , however , possesses mostly serine or threonine at this position , but never a lysine . Although a transition from lysine to serine and subsequently to other amino acids had occurred at an early point in PGK's history , this event was singular , as no successive exchanges seem to have occurred within the lysine clade . Indeed , as described below , the K219S exchange renders human and Escherichia coli PGKs non-functional , and the reciprocal S219K replacement in the archaeal M . mazei PGK has the same effect . We aimed at disentangling the incompatibility at PGK's position 219 and thus address the following conundrum . On the one hand , exchanges at numerous other positions erased the trails that had historically enabled to traverse sequence space at this position - the closest archaeal sequences ( the serine/threonine clade ) and bacterial/eukaryotic sequences ( the lysine clade ) are separated by 250 amino acid exchanges . Further , the traversing trails are presumed to be intricate and scarce [6] . On the other hand , it is assumed that “functional proteins must form a continuous network which can be traversed by unit mutational steps without passing through nonfunctional intermediates” [19] . Unraveling this network is fundamental to our understanding of molecular evolution . As previously shown [20] , the phylogenetic tree of 620 diverse PGK sequences largely follows the species tree with a clear separation between the three kingdoms of life ( Figure 1 ) . Overall , within kingdoms , conservation is high . For instance , the amino acid identity of human relative to chicken PGK is ∼90% , and ∼60% relative to yeast ( Saccharomyces cerevisiae ) . However , between kingdoms , the average sequence identity drops to approximately 35% . The sequence identity between human and E . coli PGKs , for example , is 38% . The archaeal kingdom is the most diverged one with , on average , ∼30% identity to bacterial or eukaryotic PGKs . Nonetheless , the active-site residues known to play a role in PGK's enzymatic function are conserved throughout , with only two exceptions: positions 219 and 336 ( numbering is according to human PGK; Figure S1 ) . All bacteria and eukaryotes have lysine at position 219 , as does the inferred ancestor of all PGKs ( lysine predicted with 81% posterior probability; data not shown ) . In contrast , no archaeal PGK has a lysine at position 219 , with serine being most abundant ( Figure 1 ) . The phylogenetic pattern of position 336 is less consistent . Nonetheless , exchanges in both 219 and 336 are incompatible ( see below ) . Given its clear phylogenetic pattern , and the well-defined catalytic role of lysine 219 in human PGK , we focused on the exchange at position 219 . We chose human PGK ( thereafter marked as hPGK ) , rather than that of E . coli , because of the large available body of biochemical and structural data [18] , [21]–[24] . Specifically , lysine 219 of hPGK was shown to be directly involved in substrate binding and domain closure upon catalysis [18] . Lysine 219 also contributes to formation of the active site transition state electrostatics [21] ( Figure S2 ) . An in-vivo selection system for PGK was developed by inactivating E . coli's pgk gene . An E . coli-Δpgk strain is publicly available . This strain was obtained via transposon mutagenesis and screening for loss of ability to grow on glucose as the sole carbon source [25] . However , although this strain exhibited no PGK activity , our sequencing of the pgk open reading frame , and of 200 bps of its flanking regions , revealed no mutations . Previous attempts to obtain a pgk chromosomal deletion in E . coli had failed [26] . We could , however , knock out the pgk gene when the transformed cells were plated on minimal growth media supplemented only with glycerol and succinate . Indeed , a bidirectional pathway such as glycolysis can be circumvented by supplying the metabolites from both ends of the pathway [25] . The resultant Δpgk strain was not only unable to utilize glucose as a carbon source ( as the previously reported Δpgk strain [25] ) , but its growth on glycerol-succinate was completely inhibited when glucose was added at any level beyond 1 µM . The previously reported E . coli-Δpgk therefore confers only a partial phenotype , and the previous failure to obtain a pgk knockout [26] may relate to plating on rich medium . The human pgk gene was placed within a plasmid dubbed pZA-hPGK under the tight regulation of a tetracycline-inducible promoter . At ≥8 ng/ml of the inducer AHT ( anhydrous tetracycline ) , E . coli-Δpgk cells harboring this plasmid ( ∼40 copies per cell [27] ) grew at the same rate as the parental strain with the endogenous , chromosomal E . coli PGK in glycerol-succinate media with 5 mM glucose . At the range of 0–8 ng/ml AHT , cellular PGK activity levels and growth rates were concomitantly dependent on the inducer's concentration ( Figure 2 ) . When lysine 219 of the plasmid-encoded hPGK was mutated to serine , no growth on glucose containing media was observed ( Figure 2 ) . In agreement , a 400-fold decrease in kcat/KM value was measured for the K219S mutant ( Table 1 , Figure S3 ) . The K219S exchange in E . coli's PGK resulted in a similar effect , namely no growth on media containing ≥1 µM glucose with the mutant expressed from the pZA plasmid ( data not shown ) . The reciprocal replacement , S219K , had a similar effect in the archaeal , M . mazei PGK . Expressed from the pZA plasmid , wild-type M . mazei PGK supported growth on glucose similar to E . coli and human PGK . However , its S219K mutant showed no growth in glucose containing media ( data not shown ) . We sought to identify mutations that could rescue the enzymatic function of hPGK-serine 219 ( second-site suppressor mutations ) . To this end , seven iterative rounds of random mutagenesis and selection were performed to gradually increase the ability of K219S mutated hPGK genes to confer growth on glucose . The gene library was generated by error-prone PCR of the hPGK-serine 219 gene , giving rise to an average of 2 . 5 mutations per gene . To minimize the chances of serine 219 being reverted to lysine , the serine codon at position 219 was designed to be TCT , whereby all three bases must mutate simultaneously to give lysine . Nonetheless , mutations to lysine and wild-type contaminants still dominated the first and second rounds ( appearing in 10 to 20 percent of the surviving clones ) . Thus , during all rounds , a special cloning protocol was applied to prevent exchanges of the serine at position 219 ( see Text S1 ) . Because hPGK-serine 219 gave no growth on glucose when expressed from the pZA plasmid , we began by over-expressing it from the high copy plasmid pZUC ( ∼900 copies per cell [28] ) , thus enabling some growth on glucose supplemented plates ( ∼1 mm colonies after 72 hours ) . The selection pressure was gradually strengthened by decreasing the inducer concentration ( thus lowering the expression level ) , decreasing incubation time , and subsequently , by re-cloning the libraries to plasmids of lower copy numbers ( Methods , Tables 2 and S1 ) . In each round , 10–20 randomly chosen surviving colonies were isolated and the harbored hPGK genes were sequenced . All surviving colonies were subsequently pooled . The plasmid DNA was isolated and the pgk gene was subjected to further mutagenesis , re-cloning and selection . By round 5 , the sequences of randomly sampled variants indicated near complete convergence with 4–7 substitutions within and around hPGK's active site ( the mean being 5 . 2±1 . 1 substitutions per gene; the sequence composition of the evolved variants is discussed below ) . Indeed , no further improvement in growth , or substantial sequence changes , could be observed in the subsequent round . In the seventh round , with the aim of isolating the most active hPGK K219S variants , we conducted random mutagenesis by error-prone PCR with the selected gene pool of round 6 as template , followed by 3 cycles of selection on glucose plates ( extracting the plasmid pool , retransforming and plating ) . Eight of the largest colonies of round 7 were examined . They exhibited growth rates identical to wild-type hPGK , but similar catalytic efficiencies ( kcat/KM values; Table 1 and S2 ) relative to the best variants from round 6 . The likely explanation for having improved growth yet identical catalytic efficiency is that surface mutations from basic to acidic residues ( lysine to glutamate ) were highly abundant in round 7 variants ( Figure S4 ) . These led to a reduction in PGK's pI value , thus meeting the conditions of the E . coli cytoplasm ( the pI of hPGK is 8 . 75 [29] whereas the pH of the E . coli cytoplasm is 7 . 5 [30] ) . As such , these surface mutations are possibly resulting in improved stability and elevated amounts of soluble protein [31] , therefore increasing enzyme dose rather than specifically compensating for the serine-lysine exchange . Such compensatory mutations that increase stability and/or solubility and therefore enzyme doses are commonly observed , e . g . in the clinical emergence of drug resistance [32] . Variants that represented the sequence trends in each round along the described evolutionary trajectory were characterized ( the choice of variants is described in Methods ) . Since these variants were originally selected in various copy number plasmids , they were all re-cloned to the low copy number pZA plasmid and re-sequenced to confirm their mutational compositions . Cell cultures ( Δpgk ) harboring the evolved hPGKs were grown in glycerol-succinate media supplemented with 5 mM glucose and 8 ng/ml AHT , such that the growth rates were most sensitive to the enzymatic efficiency of these variants relative to wild-type hPGK ( Figure 2 ) . The evolved variants exhibited increasingly higher growth rates that matched the order of their appearance along the trajectory ( Figure 3a ) . Next , serine 219 was exchanged back to lysine by site-directed mutagenesis of the evolved hPGK variants . The series of evolved hPGK variants exhibited an opposite trend to the one observed with the original serine 219 variants - growth rates gradually decreased with sequential rounds of evolution . The loss of activity with lysine at position 219 , however , was mild relative to the gain of activity with serine ( Figure 3a ) . Thus , a zone of relative neutrality with respect to position 219 has been revealed in hPGK's sequence space . These sequences could theoretically facilitate serine-lysine transitions . For example , variants G2-v15 and G4-v4 ( G stands for the round number from which the variant was isolated , and v is the variant's sequential number within this round ) can accommodate either lysine or serine at position 219 yet exhibit growth rates that are not very far from wild-type . The growth rates of these transition variants in media containing only glucose were lower than in medium containing glucose plus succinate and glycerol ( Figure S5 ) . Nonetheless , growth was clearly observed in these variants , with either lysine or serine at 219 . In contrast , the K219S mutant of hPGK conferred no growth in any glucose containing media , even when succinate and glycerol comprised the primary carbon source ( Figure 3a ) . To validate the observed trend in growth rates , these representative variants were purified and assayed in vitro . The kcat/KM values and growth rates were generally correlated ( Figure 3b ) . Most variations related to lower kcat/KM values relative to growth rates ( e . g . variant G7-v12 ) , as discussed above , are attributed to surface mutations . Two mutations appeared in nearly all the library selected variants from round 4 to round 7: methionine at position 239 exchanged to isoleucine , and glutamate at position 403 to aspartate ( Table 1 ) . Indicatively , these mutations appear in nearly all the transition variants - i . e . , variants in which both serine and lysine are accommodated with little effect on growth rates ( Table S2 ) . Although located away from the active site and ∼9 Å from lysine 219 , glutamate 403 plays a role in hPGK's catalytic cycle , and specifically in the inter-domain motions that relate to substrate binding and phosphoryl transfer [21] . Phylogenetically , positions 219 and 403 seem correlated: >98% of sequences that have a lysine at position 219 ( 535/547 ) have glutamate at position 403 . None of the 72 sequences in the archaeal clade ( non-lysine 219 ) have glutamate at position 403 , with alanine being the most frequent amino acid ( Figure S6a ) . Methionine 239 is in close proximity to lysine 219 ( 3 . 5 Å ) . However , phylogenetically , it is weakly correlated to position 403 . Amongst the sequences that have a lysine at position 219 , methionine is most frequent ( 76%; as in human PGK ) , and the remaining sequences have mostly isoleucine or leucine at position 239 ( 117 and 14 , respectively ) . Amongst archaeal PGKs , none have a methionine whilst most sequences have a valine ( 60/73 ) and the remaining have isoleucine or leucine ( Figure S6b ) . Sequence wise , the outlier was variant G6-v9 that also exhibited the highest kcat/KM value . Instead of the E403D mutation present in all other transition sequences , it contained the A397V mutation that appeared in 35% of the selected variants of this round ( Table 1 ) . However , A397V was purged by the 7th round that was completely dominated by the aforementioned pair of mutations ( M239I plus E403D ) . Indeed , E403D and A397V never appeared together , suggesting they might have a beneficial effect individually but interact with negative epistasis ( see below ) . Archaeal PGKs have not only diverged away from the absolute conservation of lysine at position 219 seen along the bacterial and eukaryote clades , but have also drifted considerably at this position . Archaeal PGKs with serine , but also with threonine , valine , arginine and leucine are found within relatively close phylogenetic proximity ( Figure S7 ) . Since the laboratory-evolved transition sequences accepted serine , and were also functional with lysine , we tested whether they might accept additional amino acids at position 219 . Thus , a variable NNS codon ( N , any nucleotide; S , G or C ) , that can give rise to any one of the 20 possible amino acids ( but only one stop codon ) was introduced at position 219 , of both the evolved transition variant G4-v4 and wild-type hPGK . The resulting gene libraries were cloned into the pZA plasmid , transformed into Δpgk cells and selected for survival on glucose . The fraction of glucose viable transformants was 26% in the G4-v4 based library and 4% in the wild-type library . Sequencing indicated that all 19 clones from the wild-type library carried lysine at 219 . Variant G4-v4 , however , accepted serine , lysine and five additional amino acids ( Figure 4a ) . The observed diversity in G4-v4 showed near-complete overlap with the natural diversity at position 219 . The only exceptions were glycine that appeared viable in our selection ( with the slowest growth rate; Figure 4b ) but does not appear in known natural sequences , and leucine that appears in 5 out of the 73 archaeal PGKs included in the tree , but not in the G4-v4 sequenced clones . Indeed , when artificially introduced at position 219 of variant G4-v4 , leucine did not confer any growth with glucose ( data not shown ) . Can we deduce the order in which mutations that may enable exchanges at key positions such as 219 might occur ? Two key mutations underlined the experimental , directed evolution towards the release of position 219 from serine incompatibility: M239I and E403D . To unravel a minimal trajectory that can lead to serine compatibility , the individual and cumulative effects of M239I and E403D were tested at the background of both lysine and serine at 219 . The double mutant ( M239I and E403D ) , either with lysine or serine at position 219 , exhibited highly similar growth rates , and these were only slightly lower than those exhibited by the evolved transition variant G4-v4 ( Figure 5a ) . The eight possible sequence permutations , including wild-type with either serine or lysine 219 , were therefore constructed and tested for their growth rates . Analysis of the growth rates of the eight permutations indicates which trajectories are more likely to occur – these trajectories comprise ‘ridges’ in the genotype-phenotype space , i . e . , the three mutational steps that connect the two fitness peaks , serine-219 and lysine-219 PGKs , with a minimal loss of fitness . These hypothetical trajectories indicate three trends ( Figure 5b–c ) . Firstly , the trajectory unraveled by our directed evolution experiment is underlined by strong sign epistasis . This trajectory can be referred to as a ‘reverse’ trajectory , namely from an extant lysine 219-restricted sequence ( e . g . human PGK ) to a 219-permissive sequence ( the putative ancestral state ) . As predicted [6] , both M239I and E403D , alone and in combination , exhibit a negative fitness effect in the context of lysine 219 , but a positive effect when combined with serine 219 . Not only does the sign revert , but the magnitude of fitness effects differs as well . However , the fitness gain at the background of serine 219 is ‘infinitely’ higher ( from no growth to 0 . 013 min−1 ) relative to the loss at a lysine 219 background . Secondly , the hypothetical ‘forward’ trajectory , i . e . , a trajectory that might mimic the divergence of a 219-permissive sequence ( the putative ancestral state ) into the extant lysine-restricted sequences ( e . g . human PGK ) , shows the same trend of sign epistasis . The mutations at 239 and 403 ( to methionine 239 and glutamate 403 , in the presumed forward direction ) provide a small yet consistent fitness gain in the lysine 219 context ( 1 . 4-fold ) yet result in a complete fitness loss on the background of serine 219 . Indeed , this asymmetric effect is the underlining reason for the serine incompatibility of the extant lysine-219 PGK sequences . Thirdly , although both mutations , I239M or D403E , contribute individually to serine-incompatibility ( the hypothetical ‘forward’ direction ) , they interact with strong negative epistasis . The combination of both mutations completes the specialization towards lysine 219 and also makes hPGK incompatible to the serine exchange ( no growth on glucose is observed with hPGK carrying serine 219 ) . The above eight permutations were also examined for growth in presence of the A397V mutation that appeared in round 6 and was subsequently removed in round 7 ( see ‘sequence composition’ above ) . This analysis indicated that A397V is inferior to the E403D mutation in increasing the fitness of serine 219 ( Figure S8 ) . The combination of A397V and E403D showed very slow growth , that was much slower than that of E403D on its own . Thus , A397A and E403D , that never appeared together in a single variant , interact with negative epistasis . A397V therefore represents an alternative trajectory , albeit a trajectory that does not coincide with the optimal path and enforces a severe loss of fitness upon an exchange of lysine 219 to serine [7] . An ancient divergence event at the active site of an essential metabolic enzyme , PGK , provided us with a model for studying protein sequence divergence and incompatibility . The results of our laboratory evolution experiment indicate that as few as two mutations were sufficient to enable an exchange at a critical active site position that seems to have diverged with the archaeal lineage , presumably >3 billion years ago [33] . There are clear caveats associated with this experiment , and with the derived conclusions regarding the ‘fitness’ of individual PGK sequences and mutants . The laboratory conditions do not reproduce the range of environments and challenges that organisms face in the wild . Thus , the fact that under our experimental conditions , the growth rates with wild-type hPGK and its various mutants were similar to the parental strain with the endogenous , E . coli PGK , does not obviously imply the same fitness . Nonetheless , the experiment reveals a dramatic shift , from no growth on glucose in hPGK carrying serine 219 , to near-normal laboratory growth upon acquiring the two rescuing mutations . The shift , however , occurred under two constraints . Firstly , the trajectories enabling this exchange exhibit clear marks of sign epistasis , or indeed , of ‘reciprocal sign epistasis’ [7]- a large gain of function in the exchanged context ( serine 219 ) is necessarily accompanied by a certain loss of fitness with respect to the original one ( lysine 219 ) . In other words , the fitness landscape is rugged and comprises two discrete peaks ( serine or lysine 219 ) , and all crossings between these peaks are via valleys of lower fitness ( Figure 5b ) [6]–[12] . However , our results also indicate that the valleys can be readily crossed . The exchanges at positions 219 and 403 are likely to have had a key contribution to the historical crossing , but substitutions at other positions that were not unraveled by our very limited exploration of PGK's sequence space may have further facilitated this crossing . An equally plausible explanation is that the historical crossing was enabled by a temporal relaxation in the selection pressure acting on PGK ( Figure 5c ) . Relaxation may be afforded by an elevation in protein dose . Whereas the transition variant exhibited lower fitness than wild-type under limiting conditions ( Figure 5a ) , this gap disappeared at ≥2-fold inducer level ( AHT ≥15 ng/ml; data not shown ) . In nature , higher protein doses are readily obtained by promoter mutations [10] or by gene duplication [34] . Another enabling factor is a change in the environment , and specifically , alternative carbon sources . PGK is absolutely essential for growth in the presence of glucose , but growth rates of the evolved transition variants are far less compromised with additional carbon sources such as succinate and glycerol ( Figure S5 ) . Indeed , even highly essential metabolic processes can be bypassed by evolution [35] . Secondly , the exchange trajectory is highly specific in its sequence composition . Our selections were not bottlenecked and we sequenced a large variety of clones ( 110 in total ) . Nonetheless , two mutations , M239I and E403D , appeared in nearly all evolved variants . The mutation A397V comprises an alternative yet a transient solution , as it interacts with strong negative epistasis with the more optimal mutation E403D . There are also a number of other , global suppressor mutations that increase the activity upon exchange to serine , but their effects are relatively minor . Overall , the picture is consistent with the theory presented by Kondrashov et al . [6] , and with previous findings regarding adaptive trajectories [10] , [14] . That is to say , traversing sequence space at key , highly conserved positions can occur . However , the rate is painstakingly slow , primarily for two reasons: ( i ) The mutations that enable these key exchanges are highly specific , and thus the likelihood of traversing is low . ( ii ) Sign epistasis makes the accumulation of these enabling mutations dependent on a relaxation in the selection pressure , i . e . , environmental changes or protein dose changes . As we did not select for gain of function with any other amino acid than serine , the appearance of sequences that permit variability at position 219 was spontaneous , with respect not only to lysine , but to other amino acids as well . It seems that the two transition mutations , M239I and E403D , largely alleviated the enzyme's dependency on lysine , rather than having serine take over its role . Further , there is near-complete overlap between the amino acids tolerated by the evolved variant G4-v4 and those seen in archaeal PGKs ( Figure 4 ) . The sequence compositions isolated here , and particularly isoleucine at 239 and glutamate at 403 , may therefore reflect the ancestral state with respect to position 219 at the node that separated the lysine-clade from the archaeal one . This node may represent a sub-optimal PGK yet with a unique capability to tolerate a wide range of amino acids at position 219 . Our results also show how specialization towards lysine ( and probably towards any other amino acid seen in extant PGKs ) erased the ancestral state with respect to an exchange in position 219 . Thus , while hPGK may have been reverted to an ancestral state with respect to position 219 , this reversion depended on a rare combination of at least two mutations in other sites [36] . The challenge of reverting relates to the fact that the mutations that augment G4-v4's activity with lysine severely compromised activity with serine . This challenge is also manifested in the need to implement a special protocol that blocks the reversion to lysine in the laboratory evolution experiment . Needless to say that such blocks are impossible in natural evolution . Foremost , a strong negative epistatic effect seems to have led to the incompatibility seen in the extant human PGK ( Figure 5 ) . Thus , incompatibility may not be the outcome of drift , as in the classical Dobzansky-Muller model for hybrid incompatibility [37] , [38] . Rather , incompatibility is the outcome of the evolution towards higher performance with lysine-219 , and the highly deleterious effect of the mutations that improve the activity with lysine-219 at the background of serine-219 . From a sequence space point of view , this work reinforces the view that “functional proteins must form a continuous network which can be traversed by unit mutational steps without passing through nonfunctional intermediates” [19] . The experimental approach taken here provides crucial insights in this respect . For example , a lysine-serine exchange cannot occur via a single nucleotide substitution . Thus , an intermediary amino acid must have existed at some stage . Ancestral inference does not reveal possible candidates , as the predicted nodes carry at position 219 either lysine ( the LUCA's PGK ) or serine ( the ancestor of archaeal PGKs ) . However , the screen for tolerable amino acids at position 219 of the transition variant G4-v4's revealed that the growth rate with arginine exceeded all other amino acids including lysine ( Figure 4b ) . Indeed , arginine does comprise a bridging codon between serine and lysine ( Figure S9 ) . Thus , the experiment reveals that arginine is the most probable transition between lysine and serine in the corresponding PGK clades . From a protein structure-function point of view , our results support a somewhat unorthodox view: some sites are conserved not because a given amino acid is functionally irreplaceable ( e . g . lysine 219 of human PGK ) , but rather because “there has not been enough time to create the right combination of amino acids at other sites to allow them to evolve” [6] . The structural and functional features that underline the exchange at position 219 are unknown at this stage , but the evolutionary basis of this particular exchange seems to have been clarified . PGK sequences were collected by combining three BLAST searches using human , E . coli and M . mazei PGK sequences as queries . Each of the three BLAST searches contained sequences from all three kingdoms of life , suggesting that the searches were exhaustive . Glycerate kinase from Neisseria meningitides was structurally aligned with the PGK sequences to serve as an outgroup , in agreement with the protein fold classification [39] . The sequences were aligned using MUSCLE [40] , and ≥98% redundancy was removed using CDHIT [41] . A phylogenetic tree was built using PHYML [42] and the JTT substitution matrix . The tree was visualized with FIGTREE . As observed in a tree of 131 PGK sequences [20] , the three kingdoms formed distinct clades . However , the inter-kingdom topology deviates from the species tree [43] whereby the eukaryotic branch diverges from within the bacterial branch with the spirochaetes being the outgroup . The tree and the alignment were further processed by PAML [44] for ancestral sequence prediction using default parameters . The pgk knockout strain was constructed using the technique of Datsenko and Wanner [45] with primers PGK_KO_F and PGK_KO_R . A set of plasmids with decreasing copy numbers were constructed by modifying pZA11MCS ( EXPRESSYS ) as described in Text S1 ( Figure S12 ) . To generate the library for the first round of directed evolution , the K219S mutant of hPGK was amplified by error-prone PCR ( GeneMorph II Random Mutagenesis Kit; Agilent ) at an average of 2 . 5±2 mutations per gene . The resulting gene library was re-cloned to pZUC ( the K219S mutant did not grow on media containing glucose when expressed from the low copy pZA plasmid; Figure 2 ) . The growth upon copy number manipulation correlated with very high protein expression levels ( Figure S10 ) . The resulting plasmid library was transformed into Δpgk cells and plated on minimal media agar plates supplemented with 20 mM glucose , kanamycin and ampicillin ( selection plates ) . An aliquot of the transformation reaction was plated on glycerol-succinate plates with both antibiotics to measure the transformation efficiency . The colonies that grew on the selection plates were pooled and the plasmid DNA was isolated . Prior to pooling , 10 to 20 variants were sequenced from randomly picked clones . The hPGK gene was amplified from the plasmid pool by using the special cloning protocol designed to eliminate mutations at position 219 ( Figure S11 ) . The PCR product was digested by Nco1 and Not1 , purified from gel and subsequently cloned to the appropriate plasmid for the next round of selection . In the 2nd round , gene shuffling was applied to obtain favorable combinations of mutations that accumulated in the 1st round ( see Text S1 ) . In later rounds , mutagenesis was achieved via the use of 60 cycles of non-proofreading TAQ polymerase . The subsequent rounds ( R2 to R7 ) were conducted with certain modifications to allow an increase in selection pressure: ( a ) Use of lower copy plasmids; ( b ) lower inducer ( AHT ) concentration in the selection plates; ( c ) shorter time of incubation of the selection plates . A summary of the selection conditions for all rounds is provided as Table 2 . Representative variants from the selected libraries were chosen for further analysis . The variants were chosen so they reflect the mutational trend within their round , and order in which mutations fixed in course of the progressing generations from G1 to G7 . For instance , the E403D mutation already appeared in 3 out of 9 clones that were sequenced from the first round library , and this mutation was enriched in consecutive rounds . Thus , the chosen variant G1-v5 carried this mutation ( Table 1 ) . Conversely , the M239I mutation became dominant ( ≥30% of the clones ) only by the third round , and the combination of both mutations in a single clone represents the general trend in the fourth round library , as represented in variant G4-v4 . In round 6 , the A397V mutation became enriched ( found in 35% of the sequences variants ) and appeared to be mutually exclusive from the E403D mutation . Thus , variant G6-v9 was chosen as a representative sequence . We also assumed that mutations that appeared only once in the sequenced variants from all rounds , did not contribute to the enhancement of hPGK- serine 219's activity . Thus , another criterion for choosing variants for analysis was a minimal number of ‘hitchhiking’ mutations . Finally , the growth rates , kinetic parameters and sequence compositions of 12 additional variants are given as Table S2 – these coincide with the variants presented in the main text . The chosen variants were re-cloned to pZA . For each clone , a lysine 219 counterpart was prepared by all-around PCR with primers S219K_F and S219K_R . Mutations M239I and E403D were similarly incorporated ( primers listed in Table S3 ) . Plasmids encoding individual hPGK variants were transformed into Δpgk cells and the sequences were revalidated . Single colonies were used to inoculate 2 ml of glycerol-succinate media with kanamycin and ampicillin . The OD600 nm were measured , and fresh glycerol-succinate media supplemented with 5 mM glucose and 8 ng/ml AHT were inoculated with these overnight cultures at the same initial bacterial density ( OD600 nm = 0 . 02 ) . The inoculated media ( 200 µl ) were distributed in random fashion in a 96 well-plate . The plate was incubated at 37°C , and 95% relative humidity , in a rotating incubator ( Liconic Instruments , Woburn , MA ) . Using an automated robotic platform ( Evoware II , Tecan ) and a multi-well reader ( Infinite M200-pro , Tecan ) the OD600 nm was measured at 20 min intervals . The growth rate was calculated by plotting the log2 of the OD600 nm against time of incubation , and calculating linear regression values over a window size of 2 hours . The 3 highest rates ( ΔOD/Δt values ) along the entire growth curve were extracted and averaged . Doubling time was calculated as the reciprocal of the average growth rate [46] . Error bars correspond to standard errors from 8 independent measurements of each variant .	Orthologs are proteins in different species sharing the same function and structure . However , the mechanisms that underline the divergence of different sequences from a single ancestor remain unclear , particularly because many amino acid exchanges between orthologs result in loss of function ( incompatibility ) . We aimed at disentangling an ancient divergence event within the active-site of a universally spread enzyme that mediates ATP synthesis . Using laboratory evolution experiments , we found that an exchange in a functionally critical active-site residue that is incompatible within contemporary orthologs is enabled by few mutations . These mutations lead to transition sequences in which , unlike the extant sequences , a wide range of amino acids is tolerated . Our experiment reveals the properties of these transition sequences that may resemble the historical ancestral states that underlined this divergence event , and the mechanisms that led to incompatibility within the contemporary orthologs . Our results support theoretical predictions and reshape our understanding of protein structure-function . That a given position is entirely conserved and essential for function does not indicate that it will never exchange , but rather , that the exchange may depend on changes in many other positions .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "systems", "biology", "biochemistry", "genetics", "biology", "genomics", "evolutionary", "biology" ]	2013	Mechanisms of Protein Sequence Divergence and Incompatibility
MRP4 ( multidrug resistance-associated protein 4 ) is a member of the MRP/ABCC subfamily of ATP-binding cassette ( ABC ) transporters that are essential for many cellular processes requiring the transport of substrates across cell membranes . Although MRP4 has been implicated as a detoxification protein by transport of structurally diverse endogenous and xenobiotic compounds , including antivirus and anticancer drugs , that usually induce oxidative stress in cells , its in vivo biological function remains unknown . In this study , we investigate the biological functions of a Drosophila homolog of human MRP4 , dMRP4 . We show that dMRP4 expression is elevated in response to oxidative stress ( paraquat , hydrogen peroxide and hyperoxia ) in Drosophila . Flies lacking dMRP4 have a shortened lifespan under both oxidative and normal conditions . Overexpression of dMRP4 , on the other hand , is sufficient to increase oxidative stress resistance and extend lifespan . By genetic manipulations , we demonstrate that dMRP4 is required for JNK ( c-Jun NH2-terminal kinase ) activation during paraquat challenge and for basal transcription of some JNK target genes under normal condition . We show that impaired JNK signaling is an important cause for major defects associated with dMRP4 mutations , suggesting that dMRP4 regulates lifespan by modulating the expression of a set of genes related to both oxidative resistance and aging , at least in part , through JNK signaling . In Drosophila , one important feature of the aging process appears to be the similarity between the changes in gene expression that occur during aging and oxidative stress response [1] , [2] , [3] . For instance , the up-regulation of genes encoding for some chaperones and/or detoxification agents in response to oxidative stress has been found to highly correlate with the aging process [1] , [2] , [3] . Hsp proteins may promote longevity by facilitating the clearance of damaged proteins that accumulate during aging [4] . Another example is the JNK signaling pathway which can be triggered by a variety of insults , including oxidative stress , and has been shown to be a genetic determinant of aging in Drosophila [5] . Mutations in the JNK cascade increase stress sensitivity and lead to shortened lifespan . Conversely , flies with increased JNK activity can sustain oxidative stress and live longer [6] . Although genome-wide surveys [1] are powerful and have linked a set of genes between stress response and aging , the majority of them have not been tested experimentally for lifespan; some genes involved in both processes may still be missing by genome-wide surveys . Here we report that a new gene , namely dMRP4 , which has not been reported on the survey list [1] , clearly plays a role in both aging process and oxidative stress . The multidrug resistance-associated protein 4 ( MRP4 ) belongs to the subfamily C ( also known as ABCC ) of the ATP-binding cassette ( ABC ) transporter protein family . It has been classified as a detoxification protein that is implicated in transport of structurally diverse endogenous and xenobiotic compounds , including antivirus and anticancer drugs that usually induce oxidative stress in cells and lead to toxicity [7] , [8] , [9] . MRP4 mRNA and protein are widely expressed in many tissues of mammals including humans [10] , suggesting that this transporter may be involved in different physiological processes . However , several recent studies have shown that mammalian MRP4 is not essential for development , since MRP4-knockout mice are viable and do not reveal any abnormalities [11] , [12] , [13] , [14] . Therefore , the biological function of MRP4 remains largely unknown . MRP-associated drug resistance has represented an important clinical problem in the treatment of cancers . Some cancer cells seem to adopt a survival strategy to protect against chemotherapy-induced oxidative stress by increasing transport of chemotherapeutics out of cells , as a result of induction of MRP , including MRP4 [15] , [16] , [17] , [18] , [19] . Indeed , up-regulation of MRP4 expression has been linked to a variety of human cancers [20] , [21] , [22] , [23] , [24] . The induction of hepatic MRP4 by oxidative stress has also been observed in mammalian liver injury after chemical treatments and this response appears to be regulated primarily at a transcriptional level [25] , [26] . However , oxidative stress-inducing agents do not always induce MRP4 [27] , [28] , [29] , [30] , raising the possibility that the induction of MRP4 expression during oxidative stress may be agent-dependent and/or cell type-specific . Furthermore , no study has attempted to address whether MRP4 is required for general oxidative stress resistance at a whole organismal level . We have previously identified the Drosophila homolog of mammalian MRP4 , called dMRP4 , during an unbiased screen for genes whose overexpression causes an abnormal response to hypoxia in adult flies [31] . dMRP4 encodes a protein sharing 43% overall amino acid identity and 63% similarity with the human MRP4 [32] , [33] . In this study , we have investigated the possible involvement of dMRP4 in resistance to oxidative stress . By genetic manipulation , we present evidence that dMRP4 is associated with changes in lifespan under both oxidative stress and normal conditions , likely through a mechanism that is linked to JNK signaling in Drosophila . To test our hypothesis that the expression of dMRP4 may be regulated by oxidative stress in Drosophila , we first analyzed dMRP4 transcriptional activity in response to oxidative stimuli by feeding flies with paraquat , which generates superoxide in mitochondria [34] and has been widely used as an oxidative stress inducer in vivo . The expression of dMRP4 was strongly induced in wild-type flies fed with 10 mM paraquat for 12 hours ( Fig . 1A ) . Similar induction patterns were observed in parallel with two known oxidative stress-responsive genes [3] , [6] , [35] , puc ( puckered ) and gstD1 ( glutathione s transferase D1 ) . To test whether dMRP4 responds to other oxidative stressors , we analyzed its transcriptional changes in flies treated with hydrogen peroxide as well as hyperoxia . Up-regualtion of dMRP4 was clearly observed after hydrogen peroxide or hyperoxia treatment , in parallel with two known up-regulated markers , gstD1 and hsp22 , under these conditions [1] ( Fig . 1B–C ) . These results indicate that Drosophila dMRP4 is a bona fide oxidative stress-responsive gene . To test whether dMRP4 indeed might play a role in oxidative stress resistance , we generated two mutations by excision of two independent EP elements near the dMRP4 gene ( Fig . 2A ) . Analysis of the dMRP4 expression by RT-PCR indicated that dMRP4 RNA was undetectable in these mutants ( Fig . 2B ) . However , the more sensitive assay with qt-PCR revealed about 8% dMRP4 mRNA retaining in both homozygous mutations ( Fig . 2C ) . Currently it is not clear if this transcript residual was resulted from splice forms of the predicted full length mRNA or from an alternative transcription start site of the remaining dMRP4 transcript after the truncation . Nevertheless , these results indicate that the two dMRP4 alleles represent strong loss-of-function mutations . In addition , flies homozygous for both mutations were viable and fertile , suggesting that dMRP4 may not be an essential gene for development . However , it cannot be ruled out that the remaining residual in these mutations might still retain some vital function during development . To address whether induction of dMRP4 is required for defense against oxidative stress , we monitored the survival of adult flies treated with three most commonly used oxidative stressors: paraquat , hydrogen peroxide , or hyperoxia . In each condition the two dMRP4 alleles or their transheterozygous combination displayed similar and reproducible phenotypes: flies lacking dMRP4 reduced profoundly their viability under oxidative stress relative to controls ( Fig . 2D–F , Log-rank test , p<0 . 001 ) . These results demonstrate that wild-type dMRP4 is required for oxidative stress resistance in Drosophila . Oxidative stress is known to activate a protective program involving induction of a number of stress-responsive genes in cells [3] , [6] , [35] , [36] . JNK signaling is activated in response to oxidative stress and is a major genetic factor in control of oxidative stress tolerance and aging process [3] , [6] , [35] , [36] , [37] . Since puc ( a phosphatase inhibitor of JNK ) is often used as a marker for activation of the JNK pathway [3] , [6] , [35] , [36] , [38] , we tested whether there were any differential expression changes of JNK signaling by examining puc induction in dMRP4 mutant flies fed with paraquat . Compared to the pattern in wild-type flies , puc expression was completely diminished in dMRP4 mutant flies under oxidative stress ( Fig . 3A ) . To further evaluate whether dMRP4 might play a general role in JNK signaling , induction of other JNK-mediated marker genes , such as gstD1 [6] , hsp68 and Jafrac1 , was also examined . Although expression of all these marker genes was induced in wild-type flies after paraquat feeding , their induction , with exception for gstD1 , was significantly reduced in the dMRP4 mutant flies ( Fig . 3C–D ) , indicating that activation of JNK signaling by oxidative stress requires a wild-type dMRP4 function . Because flies deficient for JNK signaling become more susceptible to stress [6] , a phenotype resembling what we have observed with flies deficient for dMRP4 , impairment of JNK signaling in dMRP4 mutants may be an important cause for increased lethality when animals face oxidative insults . There was also a possibility that dMRP4 itself may be a component of the JNK pathway . To test whether dMRP4 might be a component of the JNK pathway , we examined dMRP4 response in flies with reduced activities of JNK signaling by the expression of a dominant negative form of Bsk ( BskDN ) ( Basket , a Drosophila homolog of JNK ) . BskDN can mimic bsk mutant phenotypes in flies and cells [39] . In this experiment , BskDN expression was induced in adult flies by actin-GeneSwitch-Gal4 ( actGS-Gal4 ) , a RU486-mediated system [40] that drives ubiquitous expression in whole fly . In the presence of drug RU486 , BskDN expression was activated from the UAS driven transgene . The relative mRNA levels from RU486-fed flies were compared to control flies carrying the same induction system ( actGS>dMRP4 ) without drug feeding . Inhibition of JNK activity by BskDN , as shown by puc expression , did not repress dMRP4 induction in response to paraquat ( Fig . 3E ) , indicating that JNK signaling is not required for dMRP4 induction under this stress . Next we asked whether stimulation of JNK signaling might influence dMRP4 induction . This was achieved by conditionally expressing an activated version of Hep ( HepAct ) ( hemipterus , a Drosophila homolog of JNKK ) . HepAct has been shown to be a JNK gain-of-function mutant [39] . Constitutive activation of JNK signaling by HepAct did not change dMRP4 expression in paraquat-fed flies relative to controls ( Fig . 3F ) . These results indicate that unlike those direct targets of JNK , dMRP4 induction by paraquat is independent of JNK activity , and therefore dMRP4 is not a direct component , but instead acts in parallel on a signaling that perhaps only regulates expression of some downstream effectors , of the JNK pathway . If dMRP4 is essential for oxidative resistance in Drosophila , an increased dMRP4 expression may increase oxidative resistance in wild-type flies . To test this hypothesis , we used the RU486-system to test the role of dMRP4 overexpressing in paraquat resistance . Adult flies carrying tub5GS>dMRP4 , after being fed with RU486 for dMRP4 induction ( Fig . 4I ) , significantly improved survival rates following acute treatment with paraquat ( 30 mM ) compared to control flies ( Fig . 4A ) . Importantly , RU486 feeding itself had no effect on survival under the same condition ( Fig . 4B ) . These experiments underline the protective role of dMRP4 from paraquat challenge . It also implies that this protection does not need dMRP4 to be elevated before reaching adulthood . Because mammalian MRP4 has been implicated in protecting the liver from oxidative stress [25] , [26] , we sought to investigate whether it was also the case in Drosophila . Drosophila fat body is an analogous tissue to mammalian liver and white adipose tissue [41] , [42] . yolk-Gal4 is expressed specifically in the female fat body [43] . We tested whether overexpression of dMRP4 in the fat body could provide overall protection against oxidative damage to the whole fly . Induction of dMRP4 in female fat body by yolk-Gal4 led 4-fold increase in the dMRP4 transcript ( Fig . 4H ) and rendered flies much more tolerant to paraquat treatment as compared to controls ( yolk-Gal4/+ or dMRP4/+ ) ( Fig . 4C , Log-rank test , p<0 . 01 ) . Similarly , overexpression of dMRP4 by S106-Gal4 , an inducible driver expressed predominantly in adult fat body [40] , [44] , [45] , significantly increased survival of paraquat-fed flies in the presence of RU486 ( Fig . 4D ) . Again , RU486 treatment showed dose-dependent induction of dMRP4 expression ( Fig . 4J ) but played no role in mortality under the same condition ( Fig . 4E ) . Thus , the Drosophila fat body appears to be an important tissue for dMRP4 to sustain paraquat-induced oxidative stress . Furthermore , the protective role of dMRP4 under paraquat challenge is applicable for both sexes . The anti-oxidative effect of dMRP4 on lifespan was further tested by exposing flies to hyperoxia . Flies overexpressing dMRP4 by RU484 induction clearly lived longer under 90% oxygen environment compared to controls ( Fig . 4F , Log-rank test , p<0 . 001 ) . We conclude that wild-type dMRP4 function is to promote resistance to oxidative stress in Drosophila . Aging shares many features with oxidative stress [1] . The free radical theory has proposed a link between aging and oxidative stress [46] , [47] . Recent studies from genetic manipulation of many genes in Drosophila have presented evidence that resistance to oxidative stress genetic often correlate with increased lifespan [6] , [48] , [49] , [50] . Since manipulation of dMRP4 can influence lifespan under oxidative stress , it would be important to examine whether dMRP4 regulates lifespan under non-stress conditions . We observed that mutations in dMRP4 dramatically caused a shortened normal adult lifespan ( Fig . 5A , Log-rank test , p<0 . 0001 ) . In particular , dMRP4M2/M2 flies had a mean lifespan ( as measured by 50% survival ) of 45 days and a maximum lifespan ( as measured by the 90 percent survival ) of 60 days . Compared to wild-type controls , dMRP4M2/M2 flies had a major reduction in the mean lifespan of about 47% and a decrease in maximum lifespan of 24% ( Fig . 5A ) . Similar results were observed with dMRP4M1/M1 flies ( Fig . 5A ) . The overall mortality rates of these groups were compared using Partial Slopes Rank-Sum Test [51] over the linear portion of the increase in mortality . Despite an apparent initiation of early mortality before day 30 in survival of dMRP4 mutants , there was no significant difference in slopes between the mutants and wild type ( Fig . 5B ) , indicating that loss of dMRP4 decreased lifespan by lowing the whole mortality trajectory , but not the rate of increase in mortality with age . Thus , although dMRP4 is not required for normal development , it is required for normal lifespan under non-stress conditions . Since flies overexpressing dMRP4 were more resistant to oxidative stress , we tested whether overexpressing dMRP4 would be sufficient to extend lifespan . RU486-mediated overexpression was used to minimize the influence of genetic background on lifespan assays . RU486-fed tub5GS>dMRP4 flies lived significantly longer than their siblings without RU486 feeding ( Fig . 5C , Log-rank test , p<0 . 0001 ) . The lifespan extension by tub5GS>dMRP4 expression appeared to be correlated with the dose of RU 486 . In one case , the mean lifespan was extended to 16% and the maximum lifespan to 8% ( Fig . 5C , RU486 100 ug/ml ) . In the other case , when flies were fed with 20 ug/ml RU486 , this group of flies showed only about 9% of increase in the mean lifespan and 5% of increase in the maximum lifespan , even though their overall lifespan appeared to significantly increase ( Fig . 5C , Log-rank test , p<0 . 0001 ) . Increased lifespan was not due to chronic RU486 treatment because no significant difference in lifespan was seen between treated or untreated tub5GS-Gal4 groups ( Fig . 5D , Log-rank test , p = 0 . 3 ) . We conclude that another dMRP4 function is to promote normal lifespan in Drosophila . In these experiments the lifespan extension clearly correlated with increased expression of dMRP4 , but it remained unclear whether tissue-specific dMRP4 overexpression was sufficient to extend lifespan and whether the overall levels and/or timing of such expression would be critical . Interestingly , S106>dMRP4 flies treated with RU486 did not live longer ( Fig . 5E , Log-rank test , p = 0 . 37 ) even though the fat body-specific expression of dMRP4 did show resistance to paraquat , suggesting that there might be different requirements between resistance to oxidative stress and lifespan extension . Again , RU486 treatment showed no difference between parallel controls ( Fig . 5F , Log-rank test , p = 0 . 09 ) . Moreover , high levels of ubiquitous dMRP4 expression by da-Gal4 throughout development were not beneficial and instead , there was a negative correlation with lifespan ( Fig . 5G , Log-rank test , p<0 . 0001 ) . These observations suggest that in order for dMRP4 overexpression to be beneficial for lifespan extension , the spatial and temporal such expression with proper levels have to be tightly controlled . In order to learn the molecular mechanism by which dMRP4 regulates lifespan , we selectively studied transcription profiling of several genes whose expression changes have been linked to both aging and stress [1] . Among five hsp ( heat shock protein ) genes examined , expression of three genes , hsp68 , hsp70 and l ( 2 ) efl ( lethal ( 2 ) essential for life , a small hsp gene ) was severely down-regulated in dMRP4 mutant flies ( Fig . 6A ) , while they were significantly up-regulated when dMRP4 was overexpressed ( Fig . 6B ) . Overexpression of dMRP4 was also sufficient to increase expression of other two hsp genes , hsp22 and hsp83 ( Fig . 6B ) . Since l ( 2 ) efl is a known target of dFOXO ( Drosophila forkhead transcription factor ) in lifespan regulation [52] , it raised the possibility that dMRP4 might regulate expression of other dFOXO-dependent genes . Indeed , expression of the dFOXO target gene thor , which encodes 4E-BP ( eIF4E binding protein ) , was also greatly enhanced when dMRP4 was overexpressed . Since both thor and hsp68 are target genes of JNK signaling [6] , [52] , we further examined expression patterns of several other JNK targets ( Fig . 3A–D ) . Like hsp68 , basal expression of puc and gstD1 was down-regulated in dMRP4 mutant flies and was up-regulated with dMRP4 overexpression ( Fig . 6A–B ) . Furthermore , basal expression of Jafrac1 was increased when dMRP4 was overexpressed , even though its expression was not affected by dMRP4 mutation under normal condition . Thus , in addition to regulating the JNK-dependent gene expression under oxidative stress , dMRP4 also regulates the basal transcription of such genes under normal conditions . Increased expression of hsp22 [53] , hsp68 [6] , [54] , hsp70 [55] , l ( 2 ) efl [52] , Jafrac1 [54] , [56] , has been reported to increase Drosophila lifespan . We hence suggest that increased expression of these genes by elevated dMRP4 expression may account for , at least in part , the dMRP4-mediated lifespan extension . Increasing age is accompanied with physiological decline . The locomotor decline is one of prominent physiological changes as they grow older . The climbing ability , measured by negative geotaxis , of adult fly reflects a function of age in Drosophila [57] , [58] . To determine whether the onset of aging associated with dMRP4 , we performed a negative geotaxis test for flies with different ages . Although there was no difference in negative geotaxis behavior between 5-days old dMRP4 mutant and wild-type adults , the age-associated functional decline became visible in dMRP4 mutant flies already at day 10 of adulthood , at a time when no mortality was seen regardless of mutant or wild-type controls ( Fig . 7C ) . By age 40 days , although there was a progressive functional decline in the control group , it was clearly worse in dMRP4 mutant groups ( w1118 vs dMRP4M2/M1 , Fig . 7C ) . Thus , the functional decline as they aged was faster in dMRP4 mutants than in controls . Activation of JNK signaling can increase stress resistance and extend lifespan in both Drosophila [6] , [52] , [59] , and C . elegans [60] . Our observations ( Fig . 3A–D , Fig . 6A–B ) suggest that the deficiency in basal transcription and stress response of JNK signaling may be an important cause for loss of stress tolerance and normal lifespan with dMRP4 mutant flies . If this were the case , increasing JNK signaling might be expected to correct dMRP4 deficiency . We tested this hypothesis by recombination of a pucE69 chromosome into the dMRP4 mutant background . pucE69/+ flies were more resistant to paraquat and lived longer under normal conditions [6] ( Fig . 6A and B ) . When dMRP4 mutant flies also heterozygous for pucE69 were challenged with paraquat , they behaved like pucE69/+ flies alone: they lived significantly longer not only than dMRP4 mutant flies , but also longer than wild-type controls ( Fig . 7A , p<0 . 01 ) . Consistent with a previous report [6] , pucE69/+ flies extended normal lifespan ( 27% mean lifespan and 24% maximum lifespan ) of control flies ( dMRP4M2/+ ) under non-stress conditions ( Fig . 7B , p<0 . 0001 ) . More strikingly , the puc , dMRP4 double mutant flies remarkably extended the mutant mean lifespan by 61% ( dMRP4M2/M1 vs pucE69/+ , dMRP4M2/M1 ) and maximum lifespan by 42% ( Fig . 7B , p<0 . 0001 ) . These results demonstrate that dMRP4 deficiency in stress resistance and lifespan regulation is correlated with a defect in JNK signaling . These results also place puc genetically in epistatic interaction with dMRP4 in both stress resistance and lifespan regulation . We tested whether the functional decline with age might also be associated with JNK activity by comparing the climbing ability between wild-type and pucE69/+ flies . Increased JNK signaling did not appear to benefit wild-type flies before 30 days of age , as climbing tests did not reveal a significant difference in locomotor function between wild-type and pucE69/+ flies ( Fig . 7C ) . However , after 40 days of age , increased JNK activity indeed improved climbing ability , and therefore functional aging in wild-type flies ( w1118 vs pucE69/+ in the 40 d group , Fig . 7C ) , suggesting that JNK activity is required for fitness of older flies . We then tested whether the age-associated functional decline of dMRP4 mutants could be caused by impaired JNK signaling as well . The climbing ability of puc , dMRP4 double mutant flies was restored to the level comparable to that of wild-type flies in the first 30 days of age . Therefore , early functional decline of dMRP4 mutants is possibly associated with a decline of JNK signaling ( Fig . 7C ) . Furthermore , by age of 40 days , puc , dMRP4 double mutant flies behaved like pucE69/+ flies , showing better climbing performance even over wild-type flies ( Fig . 7C ) . Thus , the JNK activity can seemingly rescue all defects that are associated with dMRP4 phenotypes . We conclude that dMRP4 plays a critical role in regulation of JNK-mediated oxidative resistance and aging process . The MRP4 subfamily and its homologs have not been reported in any lifespan-related studies including genome-wide surveys . In this study , we have investigated the physiological function of dMRP4 gene in Drosophila . A main finding from our work is that dMRP4 regulates lifespan under both normal conditions and oxidative stress , concomitantly with changes of JNK activity in vivo . Our main finding is based on several observations: First , dMRP4 is required for induction of some JNK-dependent genes in response to paraquat-induced oxidative stress . Second , elevated dMRP4 expression stimulates basal transcription of some JNK-dependent genes downstream of JNK signaling . Third , increased JNK activity in dMRP4 mutant background can rescue dMRP4-related phenotypes identified in this work , supporting our hypothesis that dMRP4 may regulate oxidative resistance and lifespan , at least in part , through JNK signaling . The finding that dMRP4 has a role in lifespan is particularly intriguing because we are able to show for the first time that a drug transporter like MRP4 is involved in lifespan regulation . Like Drosophila dMRP4 , MRP4 KO mice show no visible phenotype [11] , [12] , [13] , [14] , and mrp-4 knockdown in C . elegans with RNAi results in no observed phenotype either [61] , [62] . These observations together suggest that MRP4 and its homologs across species do not contribute to normal development in the animal world . However , unlike in other species , we found that the Drosophila dMRP4 is required for adult lifespan . Flies deficient for dMRP4 live significantly shorter , under both stressful and normal conditions . Subsequently , our work reveals that dMRP4 acts as a modulator of a network of gene expression since loss- or gain-of dMRP4 function leads to major changes in the transcriptional profiling of a number of genes that may contribute to lifespan regulation . Therefore we suggest that gene expression changes mediated by dMRP4 may represent a molecular mechanism by which dMRP4 regulates lifespan . For instance , hsp genes have been implicated in regulation of both stress resistance and lifespan extension [4] , [63] , and are among the best-known biomarkers of aging in C . elegans [63] , [64] , in Drosophila [1] , [65] , and perhaps even in humans [66] . Given the fact that the expression of hsp reporters in young individual flies has been observed to be partially predictive of remaining lifespan [65] , down-regulation of several hsp gene expression ( i . e . hsp68 , hsp70 , l ( 2 ) efl ) in dMRP4 mutant background could explain the shorter lifespan of these flies , while their up-regulation ( i . e . hsp22 , hsp68 , hsp70 , hsp83 , l ( 2 ) efl ) at a young age by dMRP4 overexpression may help protect against oxidative stress and extend lifespan of wild-type flies . This scenario is consistent with previous notions that genes are involved in stress responses generally share similar involvement with aging [1] . In addition to hsp genes , the interaction of dMRP4 with JNK signaling may provide an alternative mechanism to explain dMRP4 functions . Because the JNK pathway is known to be crucial in stress resistance and aging , impairment of JNK signaling in dMRP4 mutant flies , indicated by transcriptional down-regulation of several known JNK-related effecters , could result in dMRP4-associated phenotypes . The acute phenotype is seen particularly when the animal faces stressors such as paraquat-induced oxidative stress , which recapitulates the phenotype shown by mutations in the JNK pathway [6] . The effect of the JNK pathway on lifespan has also been observed during aging under normal conditions . Flies with reduced JNK activity have a shorter lifespan [6] , a phenotype similar to that seen in dMRP4 mutant flies . Furthermore , some downstream effectors in the JNK pathway also exhibit phenotypes that are reminiscent of dMRP4 . For instance , loss of Jafrac1 function leads to an exaggerated sensitivity to paraquat-induced oxidative stress and a shortened lifespan , while overexpression of Jafrac1 increases oxidative resistance and extends lifespan [56] . Interestingly , expression of Jafrac1 transcription is down-regulated in the dMRP4 mutant in response to oxidative stress ( Fig . 2D ) and is up-regulated by dMRP4 overexpression ( Fig . 6B ) . How dMRP4 regulates Jafrac1 remains to be investigated . One possible scenario is that dMRP4 executes its functions through interacting with JNK signaling to modulate the expression of downstream effectors such as Jafrac1 especially that the expression of Jafrac1 itself is regulated by JNK signaling [56] . After all , the most compelling evidence for the relationship between dMRP4 and JNK signaling comes from our genetic epistatic assays . When JNK signaling is enhanced in dMRP4 mutant background , all dMRP4-related defects are restored , and puc , dMRP4 double mutant flies now phenocopy pucE69/+ flies , clearly proving that JNK signaling plays a central role in realizing dMRP4 functions . Our work also suggests that promoting lifespan by increasing JNK signaling may be a result of its ability to antagonize oxidation on macromolecules , thereby postponing aging . Compared to JNK signaling , the effect of increased dMRP4 expression on lifespan extension seems less dramatic . Yet this phenotype , together with the results showing that loss- or gain-of JNK function does not alter dMRP4 expression , indicates that dMRP4 functions as a modulator of , but not a component within , JNK signaling . Furthermore , if dMRP4 is one of upstream modulators of JNK/Puc signaling , it is conceivable that its overexpression cannot entirely recapitulate the effect of JNK/Puc activation and consequently , it may not be as effective as a direct manipulation of JNK/Puc signaling with respect to lifespan . Together our results , we propose a working model to summarize how dMRP4 executes its functions in conjunction with JNK signaling ( Fig . 7D ) . Future work needs to explore how a transmembrane protein such as dMRP4 could integrate its signal into the JNK pathway under both stress and normal conditions . Although in human and mammalian models of cholestasis , MRP4 has been implicated in providing protection against oxidative stress , the genetic basis for this resistance has not yet been addressed . Therefore , the connection between tissue oxidative stress , survival of the animal , and the physiological function of MRP4 , has been lacking . In this work we show that overexpression of dMRP4 in Drosophila fat body , the equivalent tissue of mammalian liver and white adipose tissue , can confer oxidative resistance to the whole animal , suggesting a functional importance of dMRP4 in the fat body in the protection of Drosophila against oxidative stress . Drosophila fat body has recently been reported as a primary site of lipid oxidative damage after paraquat treatment [67] . dFOXO , whose expression is predominately restricted to the fat body , appears to regulate sensitivity of paraquat-induced oxidative damage and age-associated degeneration of behavioral rhythms through this tissue [68] . Furthermore , overexpression of dFOXO in the adult fat body can increase stress resistance and retard aging process [44] , [69] , supporting the physiological role of fat body in stress defense for the whole organism . Strikingly , we show in this work that expression of two targets of dFOXO , l ( ) efl and thor , are greatly induced when dMRP4 is overexpressed , raising the intriguing possibility that dMRP4 may promote stress resistance and lifespan extension by activation of dFOXO , for instance through JNK signaling [52] . However , unlike the finding that global induction of dMRP4 can promote lifespan , we have not observed a significant lifespan extension when dMRP4 overexpression is restricted in fat body . This observation suggests that the ability of stress resistance may not be an absolute factor associated with longevity in a particular tissue . It is also possible that in order for dMRP4 to benefit for longer life , more tissues with its elevated expression need to be involved . Our studies in fact have not ruled out the roles of dMRP4 in tissues other than the fat body to survival even under oxidative stress . The main function of MRP4 family is known for their ability to transport a variety of diverse endogenous and xenobiotic compounds . An interesting speculation could be raised as to whether dMRP4 might function simply as a transport in paraquat resistance . In this scenario , flies deficient in dMRP4 might not be able to efficiently exclude paraquat out of cells , thereby leading to substrate-related toxic effects . However , this assumption would hardly explain why flies deficient in dMRP4 lose their resistance to hydrogen peroxide and hyperoxia . In addition , there is no report for paraquat as a potential substrate of any MRP4 members thus far . The deteriorate influence by da>dMRP4 overexpression is notable because this phenotype has not been seen in overexpression studies of mammalian MRP4 . Although use of the whole animal in this study clearly differs from use of cultured cells in mammalian researches , it is more likely that high levels of dMRP4 expression may interfere with normal development , resulting in a pleotropic impact on later assays . An early report did observed that overexpression of two EP lines , which all targeted dMRP4 , in larvae caused neuromuscular phenotypes [70] . Given the considerable conservation of pathways between Drosophila and mammals , it will be interesting to test if manipulating MRP4 in mammalian liver cells could confer resistance to the liver , or even to the whole animal subjected to chemotherapy-induced oxidative stress . Finally , our proposed mechanism that interactions between dMRP4 and JNK signaling may shed new light on the clinic problems for long-lived cancer cells with drug resistance due to elevated expression of MRP including MRP4 proteins . EP3177 and EP3655 were described previously [31] . Other stocks: w1118 , w; TM3 , Sb , Ser/TM6B , Tb , w; Sco/CyO; MKRS/TM6B , Tb , daughterless ( da ) -Gal4 , S106-Gal4 , pucE69 , UAS-BskDN and UAS-HepAct strains were obtained from Bloomington stock center . These strains have been backcrossed to w1118 for 8–10 times before experiments . yolk-Gal4 [43] was kindly provided by Norbert Perrimon and was backcrossed into w1118 background for 8 times . Actin-GeneSwitch-Gal4 ( actGS-Gal4 , [71] , [72] ) was a gift from Dirk Bohmann . tublin5-GeneSwitch-Gal4 ( tub5GS-Gal4 , [73] ) was a gift from Scott Pletcher . These Gal4 strains have been backcrossed into w1118 background for 6 times before use . Flies were raised on standard Drosophila food ( per liter: 17 . 3 g of yeast , 73 . 1 g of cornmeal , 10 g of soy flour , 77 ml of light corn syrup , 4 . 8 ml of propionic acid , and 5 . 7 g of agar ) . To generate dMRP4 mutant flies , two independent EP lines , EP3177 and EP3655 were first backcrossed into w1118 background for 8 times . EP males were crossing to w1118; Δ2-3 Sb/TM3 females that provides with transposase . Males with mosaic color eyes were excised and subsequently balanced with w1118; TM3 , Sb , Ser/TM6B , Tb strain . The balanced excisions were then repeatedly backcrossed via the balancer strain for 8 times to establish excision stocks . They were identified by loss of the expression of the mini-white gene . The genomic deletions were determined by sequencing with specific primers spanning the EP insertion region . Two deletions obtained had truncated the 5′-end of putative dMRP4 transcript , which was designated as dMRP4 mutation 1 ( w1118; dMRP4M1 ) and dMRP4 mutation 2 ( w1118; dMRP4M2 ) ( Fig . 2A ) . dMRP4M1 was excised from EP3655 , which inserted at 47 bp from the transcription start site of the predicted gene CG14709 , resulting a 2 . 7 kb deletion that removed 1179 bp upstream of dMRP4 transcript and a 1521 bp region including 585 bp of the entire exon 1 encoding the first 25 amino acids of the protein , as well as 936 bp of the intron 1 . dMRP4M2 was resulted from an excision of EP3177 , which inserted at 88 bp from the transcription start site of the predicted gene CG14709 . This led to a 3 kb deletion that has removed 2117 bp upstream of dMRP4 transcript and an 883 bp region spanning the entire exon 1 and part of intron 1 . The pucE69 , dMRPM2 recombination strain was generated by recombination of pucE69 and dMRPM2 onto the same 3rd chromosome . Both the balanced pucE69 and dMRPM2 were repeatedly backcrossed via w1118; TM3 , Sb , Ser/TM6B , Tb for 8 times before the recombination experiments . The presence of both mutations after meiotic recombination was verified by genetic cross and by PCR with specific primers . Resultant pucE69/+ , dMRPM2/M2 double mutants were then continuously backcrossed via w1118; TM3 , Sb , Ser/TM6B , Tb for more than 10 times and were kept with the balancer as a parent stock . To induce dMRP4 overexpression , adult flies carrying different Gal4 drives were crossed to homozygous EP3177 lines . For RU486 induction , a 25 mg/ml RU486 ( mifepristone , Sigma ) stock solution made in 100% ethanol was diluted with water for desired concentrations . 250 ul of diluted RU486 solution was added onto the surface of standard fly food . This “on food” method has been shown to be simple and effective over other RU486 supply methods [74] . The vials were allowed to dry for 24 hours before use . The same solution without RU486 was added to fly food for control experiments . In most experiments , three to four day-old males , grouped with 20 flies per vial , were fed on a 3 mm Whatmann paper soaked with 10 mM paraquat ( N , N′-dimethyl-4 , 4′-bipyridinium dichloride , Sigma ) in 5% sucrose/PBS . Flies of different genotypes were also fed only with 5% sucrose/PBS as experimental controls . Under this condition all flies can live up for 10 days perfectly . Scores were done every 12 hours for the number of dead flies . Fresh paraquat was added daily . All tests were performed at 25°C . Flies were not starved before adding paraquat in this test to avoid unnecessary stress . Survival comparisons were analyzed by Kaplan–Meier Log-rank Test using Graph Pad Prism4 . p<0 . 05 was considered statistically significant . In RU486-induced experiments , 20 adult males ( 2–4 days old ) per vial were fed with different concentrations of RU486 for 4–6 days . They were then transferred on a 3 mm Whatmann paper soaked with 30 mM paraquat in 5% sucrose for acute survival test , or with 10 mM paraquat in 5% sucrose for mRNA induction at 24 hours . Control flies were from the same collection and were treated in parallel . For RNA , all samples were collected at the end of treatments and were immediately frozen in dry ice for RNA preparations . Eight day-old males were fed with different concentrations of hydrogen peroxide ( v/v , Sigma ) in 5% sucrose/PBS . Control flies were fed with 5% sucrose/PBS only . RNA for qt-PCR was extracted from these flies after 24 hours treatment . For survival tests , ten day-old males with different genotypes were fed with 3% hydrogen peroxide . Fresh hydrogen peroxide was added every day . Scoring and analysis were done essentially as described in paraquat treatment . Eight day-old males , grouped with 20 flies per vial on regular food , were exposed to a steady flow of 95% or 90% oxygen bubbled through water in a sealed chamber . RNA for qt-PCR was extracted from these flies after indicated time points . For survival tests , twelve day-old males with different genotypes were treated with 90% oxygen as above . For RU486 induction , flies from the same breeding were divided into two groups , one group fed on food containing RU486 ( 150 ug/ml ) and the other on normal food through the experiments . Flies were transferred to fresh vials every 2–3 days . Scoring was done every day . Flies were collected within 24 hours of eclosion and grouped into 20 males per vial . Tests were performed at 25°C . For each experiment , at least 200 flies of each genotype were tested . For GeneSwitch experiments , males of genotypes w1118; tub5GS-Gal4 , w1118; actGS-Gal4 , or w1118; S106-Gal4 were crossed to w1118; EP3177 or w1118 females , respectively . Male progeny from these crosses were aged for 3 days after eclosion , and then were divided into 20 flies per vial , with or without indicated concentrations of RU486 in food . Flies were transferred to fresh vials with or without RU486 every other day and dead flies were scored at the time of transfer . All experiments were conducted at least two times from independent biological breeding . The maximum lifespan was the mean lifespan of last 10% of survival animals in each cohort . 10–20 male flies , ages from 5–40 days at 25°C , were transferred to a clean plastic vial , rested for 3 min , and then measured for bang-induced vertical climbing distance at room temperature ( 20–21°C ) . The performance was scored as percentage of flies crossing 7 cm within 10 seconds in a single vial , which was expressed as average of 5 repeated tests for a single vial . 80–100 flies were tested for each genotype at each time point . Total RNA was isolated from whole flies using RNeasy Mini Kit ( Qiagen , Maryland , USA ) according to the manufacturer's instructions . cDNA synthesis was performed with oligo-dT and random primers using SuperScript III first-strand synthesis system ( Invitrogen , Carlsbad , CA ) . Semiquantitative PCR was performed as described [31] . Real-time PCR was performed in duplicate using SYBR Green on an ABI 7900HT Real-Time PCR system ( Applied Biosystems ) according to the manufacture's protocol . All samples were analyzed from at least 3 independent of experiments . Data was normalized first to the level of the rp49 mRNA prior to quantifying the relative levels of mRNA between controls and experimentally treated samples . All detailed primers are available upon request . All survival data were analyzed by Kaplan–Meier Log-rank Test for overall survival and by the Student's t-test for mean and maximum lifespan using Graph Pad Prism4 . The log mortality was determined by OASIS program [51] . Treated data were then plotted using Graph Pad Prism4 . Other comparisons were determined either by Student's t-test or One way ANOVA followed by post hoc t-test . p<0 . 05 was considered statistically significant .	The drug transporters are often known for their ability to transport different physiological-related compounds across cell membranes . Although the abnormal up-regulation of some these transporters is believed to be the common cause of the clinic problem called drug resistance , the biological functions of these transporters remain largely unknown . Here we show that a Drosophila homolog of the mammalian drug transporter plays a role in lifespan regulation . Mutations of this gene increase the sensitivity to oxidative stress and reduce lifespan , while overexpression of this gene increases resistance to oxidative stress and extends lifespan . By molecular and genetic analyses , we have linked functions of this gene to a key signaling transduction pathway that has been known to be important in lifespan regulation .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "medicine", "and", "health", "sciences", "biology", "and", "life", "sciences" ]	2014	A Drosophila ABC Transporter Regulates Lifespan
Glossina fuscipes fuscipes is the main vector of human and animal trypanosomiasis in Africa , particularly in Uganda . Attempts to control/eradicate this species using biological methods require knowledge of its reproductive biology . An important aspect is the number of times a female mates in the wild as this influences the effective population size and may constitute a critical factor in determining the success of control methods . To date , polyandry in G . f . fuscipes has not been investigated in the laboratory or in the wild . Interest in assessing the presence of remating in Ugandan populations is driven by the fact that eradication of this species is at the planning stage in this country . Two well established populations , Kabukanga in the West and Buvuma Island in Lake Victoria , were sampled to assess the presence and frequency of female remating . Six informative microsatellite loci were used to estimate the number of matings per female by genotyping sperm preserved in the female spermathecae . The direct count of the minimum number of males that transferred sperm to the spermathecae was compared to Maximum Likelihood and Bayesian probability estimates . The three estimates provided evidence that remating is common in the populations but the frequency is substantially different: 57% in Kabukanga and 33% in Buvuma . The presence of remating , with females maintaining sperm from different mates , may constitute a critical factor in cases of re-infestation of cleared areas and/or of residual populations . Remating may enhance the reproductive potential of re-invading propagules in terms of their effective population size . We suggest that population age structure may influence remating frequency . Considering the seasonal demographic changes that this fly undergoes during the dry and wet seasons , control programmes based on SIT should release large numbers of sterile males , even in residual surviving target populations , in the dry season . Tsetse flies ( Diptera: Glossinidae ) are the sole vectors of pathogenic trypanosomes in tropical Africa , where they cause Human African Trypanosomiasis ( HAT ) , or sleeping sickness , one of the most seriously neglected tropical diseases . HAT is a zoonosis caused by the flagellate protozoa Trypanosoma brucei rhodesiense in East and Southern Africa and by T . b . gambiense in West and Central Africa [1] . The only country with known infection foci of both parasites is Uganda [2] . The World Health Organization ( WHO ) has estimated that there are around 10 , 000 cases of HAT as the recent epidemics are beginning to decline , but 60 million people continue to live at risk in 37 countries covering about 40% of Africa [3] . In addition to HAT , trypanosomes transmitted by tsetse cause a fatal disease in livestock , called Nagana , which represents a major impediment to agricultural development in Africa . No vaccines exist to prevent the disease and drugs currently available to treat HAT are expensive , can cause severe side-effects , and are difficult to administer in remote villages . As a consequence , an effective alternative for controlling the disease is to target the tsetse vector [1] , [4] . In 2001 , the African Union launched the Pan African Tsetse and Eradication Campaign ( PATTEC ) to increase efforts to manage this plague , which is considered one of the root causes of hunger and poverty in most sub-Saharian African countries [5] . Glossina fuscipes fuscipes , a member of the palpalis complex , is one of the most important vectors of human and animal trypanosomiasis in Africa . It is a riverine species confined to forested patches along rivers and lacustrine environments [6] . Its range extends across the central part of the African continent from Sudan , Democratic Republic of Congo to Uganda . As a trypanosome vector , G . f . fuscipes is exposed to a large reservoir of parasites , as it feeds on both domestic and wild animals in addition to humans . Attempts to control/eradicate tsetse require in-depth information about their population characteristics such as dispersal rates , distribution , densities and reproductive biology . The riverine nature of G . f . fuscipes has resulted in a patchy distribution of its populations and as a consequence of drift , populations arising from historical colonization events show a considerable population structure [7] . Nevertheless , Beadell et al . [8] inferred a high dispersal capacity for G . f . fuscipes , demonstrating ongoing gene flow among apparently isolated populations , with an equilibrium between drift and gene flow in western and south-eastern Uganda . Since populations undergo seasonal contractions during the year due to changes in water availability , Krafsur [9] suggests that high levels of genetic drift during the dry season could be masking effects due to gene flow . The capacity of G . f . fuscipes to disperse and colonize may also depend on the number of times a female mates in the wild and whether the matings are with the same or different males . This specific mating behaviour influences the effective population size , and may constitute a critical factor in determining the success of control methods [10] , [11] . Some aspects of mating behaviour , such as the effect of age on mating competitiveness , have been studied in laboratory colonies [12] , but to date , the polyandrous behaviour of G . f . fuscipes has not investigated in the laboratory or in the wild . Data on the proportion of tsetse females that mate more than once can be obtained in two ways: through the number of fathers ( male genotype ) represented in her offspring [13] , [14] or through genotyping stored sperm in the spermatheca of the female . In the first case , the genotyping of offspring can reveal the minimum number of males that sire a brood , but not necessarily the number of males with which a female had mated , as females may bias paternity towards one or a few of their mates , resulting in an underestimation of the actual level of polyandry [15] . In the second case a more accurate estimate of the number of mates can be obtained , through the genotyping of the female's stored sperm supply [16] , [17] . Using microsatellite markers to genotype sperm , we ascertained the minimum number of males that were able to transfer sperm to a female's spermatheca in two Uganda populations . The interest in Uganda is based on the fact that eradication efforts by PATTEC are at the planning stages in this country . The results obtained in two sites , which are eco-geographically differentiated , are of particular interest , as in both populations a large proportion of females were found to have mated more than once . The remating frequencies , validated with probability values obtained with two inference statistical models , are relevant for interpreting the reproductive biology of the species but may also have an immediate impact on the strategy to be employed for eradication success . Natural populations of G . f . fuscipes were sampled from two localities in Uganda: Kabunkanga ( KB , Western territory , 0°58′37 . 88″N , 30°32′47 . 40″E ) and Buvuma Island in Lake Victoria ( BV , Southern zone , 0°15′23 . 15″N , 33°12′22 . 86″E ) ( Figure 1 ) . Both sites are favourable for this riverine species and harbour well established populations . Males and females were collected using biconical traps located 500 m apart at both sites . The traps were checked daily and the average daily fly catch per trap was recorded . The collections from Kabunkanga were made in November 2008 , at the end of the dry season from four traps with an average of 15 flies/day/trap . The collections from Buvuma Island were made at the beginning of April 2008 , during the wet season , from five traps with an average of 58 flies/day/trap . Individuals of each sex were removed from the traps and placed in tubes containing 95% ethanol . The Kabunkanga ( KB ) sample was composed of 20 males and 29 females , while for Buvuma Island ( BV ) 20 males and 40 females were analyzed . The number of males and females in each sample mirrored the sex-ratio observed in the collections . The age and the reproductive history of the sampled flies were unknown , but all the 29 Kabukanga females and the 40 females collected in Buvuma had mated as their spermathecae contained sperm . More precise information about the age structure of the flies collected in each sample could have been obtained from ovarian inspection and/or wing fray analysis [18]; however the extent of damage observed in the wings due to trapping and EtOH preservation , did not permit wing fray analysis . Ovarian age was not assessed . For each site , all of the collected flies were considered to compute allele frequencies and variability estimates . For the remating analysis , the 29 females from KB and 30 females , randomly chosen from the BV collection , were examined . For sperm isolation , the ethanol preserved female body was rehydrated in physiological solution ( 0 . 9% NaCl ) for 24–48 h before dissection . The spermathecae were easily isolated from the abdomen , stored in 70% ethanol to permit the sperm to coagulate in a “sperm bundle” [19] and then dissected in a drop of 1× PBS ( Phosphate buffered saline ) . The sperm bundle was isolated and DNA extraction was performed using QIAamp DNA Micro Kit ( Qiagen , Valencia , CA ) . DNA extraction from the legs was performed using the protocol described in Baruffi et al . [20] . The DNA extracted from legs and sperm was used as PCR template for the amplification of microsatellite markers ( SSRs ) . Nineteen SSR loci were previously isolated from a G . f . fuscipes SSR enriched library [7] . For eight of these loci ( A06 , A09 , A112 , B05 , C7 , C107 , D06 , and D109 ) the described primer sequences were adopted [7] , [8] , [21] . For the remaining 11 loci ( A03 , B03 , B06 , B11 , B109 , C104 , D3 , D05 , D12 , D101 and D103 ) primer sequences and amplification conditions were determined using DNA extracted from Kabunkanga flies as PCR template . Amplification reactions were performed in 15 µl volumes containing 1 µl of genomic DNA , 1× reaction buffer , 1 . 5 mM MgCl2 , 25 µM dNTP , 1 U Taq polymerase ( Invitrogen , Carlsbad , CA ) and 10 pmol of each primer . Reactions were performed with an Eppendorf MasterCycler Gradient thermocycler . After an initial denaturing step of 10 min at 96°C , the PCR consisted of 40 cycles of 1 min at 96°C , 1 min at optimal annealing temperature , and 1 min at 72°C , followed by a final extension step of 15 min at 72°C . Microsatellite loci were analyzed using an ABI PRISM 310 Genetic Analyzer and the GeneScan program ( Applied Biosystems ) . An individual was declared null ( non-amplifying allele ) after at least two amplification failures . Mitotic chromosome spreads were obtained from freshly deposited larvae obtained from the Slovakia laboratory strain . Briefly , brain tissues were incubated in 1% sodium citrate for 10 min at room temperature and transferred to methanol-acetic acid 3∶1 solution for 4 min . The material was disrupted in 100 µl 60% acetic acid and dropped onto clean slides and dried . Pre-hybridization was performed according to Willhoeft [22] . In situ hybridization was performed using the following protocol: the probe DNA was labelled using the Biotin High Prime kit ( Roche , Basel , Switzerland ) and detection of hybridization signals was performed using the Vectastain ABC elite kit ( Vector Laboratories , Burlingame , CA , USA ) and Alexa Fluor 594 Tyramide ( Invitrogen ) . Chromosomes were DAPI stained and the slides were mounted using the VECTASHIELD mounting medium ( Vector Laboratories , Burlingame , CA , USA ) . Chromosomes were screened under an epiflorescence Zeiss Axyoplan microscope; images were captured using an Olympus DP70 digital camera . For the chromosomal location of SSRs on mitotic chromosomes the karyotype description in Willhoeft [22] has been adopted . The polymorphic information content ( PIC ) of each of the 19 SSR loci was determined using the program Cervus 3 . 0 [23] . For each locus and population , the number of alleles ( Na ) , frequency range , observed heterozygosity ( HO ) and expected heterozygosity ( HE ) were estimated using the program Genepop version 4 [24] . The same software was also used to test for linkage disequilibrium between pairs of loci in each population ( 100 batches , 1000 interactions per batch ) and for deviations from Hardy-Weinberg ( HW ) equilibrium , at each locus/population combination , using Fisher's exact test . The Bonferroni correction was used for all tests involving multiple comparisons [25] . The average exclusion probability ( Excl . ) , i . e . the probability of excluding a single unrelated candidate parent from the parentage of a given offspring , knowing the genotype of the second parent , was estimated using the program Cervus 3 . 0 . For each locus and population , the frequency of null alleles was calculated using the Brookfield estimation [26] in Micro-Checker 2 . 2 . 3 [27] . For the X-linked loci the number of alleles and the frequency range were evaluated using the data from both males and females , whereas heterozygosity , exclusion tests and frequency of null alleles , were calculated using the data obtained from only the females . Microsatellite Analyser ( MSA ) software , version 4 . 05 [28] was applied to determine the degree of genetic differentiation between Kabunkanga and Buvuma in terms of Fst [29] . There are three potential sources of errors associated with the genotyping of the sperm stored in the spermathecae [30] , [31]; Two different approaches were used to determine the minimum number of mates per female . The first is a simple descriptive method , based on direct count , which does not involve any probabilistic model . The second approach , which incorporates information derived from the allele frequency in each population using the Hardy-Weinberg principle , provides expected values of multiple matings . This information would be lost if one followed only the first approach . It is worth noting that the expected values of multiple matings also take into account cases in which both males and females , in the population , share the same alleles for each locus . These cases are not recognizable as rematings in the direct count . For the second approach two different viewpoints were adopted: ( a ) the maximum likelihood technique and ( b ) the Bayesian analysis . For elementary explanations of these methods see [35]–[37] . The characteristics of the 19 identified SSR loci , in terms of primer sequence , amplification conditions and PIC values , are summarized in Table 1 . The characterization was performed on DNA from single flies ( 29 females and 20 males ) collected in KB . Eleven of these loci are X-linked while the remaining eight are spread along the L1 and L2 autosomes , as assessed by chromosomal in situ hybridization analyses ( Figure 2 ) . Out of these 19 loci , 4 autosomal ( A03 , B11 , C7 , D101 ) and 2 X-linked ( C107 and D3 ) loci appear to be good candidates for sperm genotyping in remating studies , as they display high PIC values and are easy to score . The variability estimates describing the suitability of the six loci: A03 , B11 , C7 , D101 , C107 and D3 , for remating analysis in KB and BV , are shown in Table 2 . The number of alleles per locus ranged from 6 to 12 with a mean of 8 . 83 in the KB population , and from 3 to 11 with a mean of 7 . 00 in the BV population . After Bonferroni correction [25] for multiple comparisons , Fisher's exact test revealed that the six loci are in Hardy-Weinberg equilibrium in both populations . No significant genotypic linkage was detected between the six loci ( Fisher's exact test , Genepop ) and therefore they can be considered as independent loci . Analyses performed with Micro-Checker [27] indicated that the average frequency of null alleles is low , 0 . 02 in KB and 0 . 01 in BV . The accuracy of these six loci for assessing remating is measured by their combined probability of excluding ( Excl ) an unrelated candidate parent from parentage when the genotype of the mother is known . The combined exclusion value is 0 . 99 in KB and 0 . 93 in BV . The different levels of variability between KB and BV populations is accompanied by a significant level of differentiation [38] , as the estimate of FST is equal to 0 . 174 between the two populations . The six microsatellite loci were successfully amplified from sperm DNA isolated from the spermathecae of 29 KB and 30 BV females . Our deductions are based on molecular data , which provide information on the number of males that were able to transfer sperm in a PCR-detectable quantity to a female's spermathecae . Consequently , a conservative ( minimum ) estimate of the number of males with which a female had mated , was determined in the Kabunkanga and Buvuma wild populations . Although our conditions were able to detect the presence of a second male sperm at a ratio as low as 1∶10 , an undetected sperm contribution cannot be excluded . Furthermore cases of failure of sperm transfer , apparently after normal copulations , have been reported [39] , [40] . Our results provide the first direct evidence that remating is a common event in the wild and what is more , females of G . f . fuscipes may store sperm from different males . These are biologically relevant data for interpreting the reproductive biology of this tsetse species , as it appears that many females preserve sperm from different mates , that could potentially be used for insemination . It is also known that this fly is able to maintain the sperm alive for long time [41] . The simultaneous presence of sperm derived from each mating suggests that one of the potential mechanisms of cryptic female choice , such as sperm displacement , [42]–[45] is not operating in this species . On the other hand , the storage of sperm from more than one male generates the opportunity for sperm competition for fertilization . Whether post-copulatory specific events/mechanisms are operating in the female storage organs to control or drive sperm use , is an important open question , which may clarify how the copulations are translated into fertilization in this fly . It is noteworthy that in G . austeni twice mated females utilize sperm from both matings for fertilization of oocytes [10] . If this is the case also for G . f . fuscipes , considering the high frequency of remating , this sperm use by polyandrous females may have a strong impact on the effective population size of the population . Both direct count estimates of remating and probability estimates , obtained with the two inference methods , are significantly lower in Buvuma than in Kabunkanga: more than fifty per cent ( 57% ) of females mated more than once in Kabunkanga while a smaller proportion ( 33% ) remated in Buvuma . Various factors , which may be interrelated , could be responsible for the observed difference . First , the lower genetic variability in Buvuma , with respect to Kabunkanga , diminishes the discriminatory power of the six SSR loci in this island population , as revealed by the lower combined exclusion probability estimate ( Excl 0 . 93 versus 0 . 99 ) . Probably this observation is not related to the choice of loci , as Beadell et al . [8] demonstrated that in Uganda there is a significant decline of microsatellite allelic richness from West to East: Kabunkanga and Buvuma are located at a great geographic distance in the West and East , respectively , of the predicted range of the species ( Figure 1 ) . Thus , considering that the Excl estimate is related to the level of genetic variability , with an Excl value of 1 . 00 , we would have increased our remating estimates , obtaining an expected value of 0 . 58 for Kabukanga and 0 . 36 for Buvuma . Since there is still a difference in the remating frequencies between the two populations , other interrelated eco-geographic and demographic factors must account for the difference . The average age structure may have played a role . In Buvuma Island , flies were caught in April , at the beginning of the rainy season when the population was expanding as also confirmed by the high fly density in the traps , which is about four times greater than the density in Kabunkanga . The Kabunkanga flies were collected in November , at the end of the cooler dry season , when the population undergoes seasonal demographic contractions with a high level of mortality particularly among the young teneral flies while the remaining flies concentrate in moist refugia . In the absence of objective observations regarding the age , such as ovarian measurements and wing-fray analysis [18] , we can speculate that in an expanding population , such as Buvuma , the proportion of young flies may be greater than that in a residual population after a seasonal bottleneck , such is the case of the Kabunkanga sample [9] , [46]–[51] . It is a reasonable assumption that the surviving flies collected in Kabunkanga at the end of the dry season , had more time and opportunity to remate , than those from Buvuma . In addition , according to Abila et al . [12] , male mating competitiveness increases with age , i . e . older males copulate significantly more frequently than younger flies and the peak of female receptivity is between the 8th–13th day after emergence [52] . It has been also reported that Glossina females tend to mate more than once with no apparent difference in receptivity and the number of matings appears to be directly related to the amount of semen in the spermathecae: young females contain less semen than older ones [53] . On the basis of these observations , it can be speculated that a demographic parameter such as age could be the cause of the observed difference in remating frequency between Kabunkanga and Buvuma . However , this hypothesis must be confirmed by appropriate analyses . Finally , as the two study sites , Kabunkanga and Buvuma island harbour well established populations which show a significant level of genetic differentiation ( Fst = 0 . 174 ) , we cannot exclude that the distinct genetic backgrounds of the two populations had an effect on the extent of the observed remating estimates . Several considerations concern the applied aspects of the present findings . As the Sterile Insect Technique ( SIT ) is being entertained for tsetse population control , the presence of remating and the fact that females maintain sperm from different mates , potentially available for insemination , may constitute a critical factor for the success of eradication programmes . Although specific experiments would be necessary to assess the sperm use and the possible presence of paternity skew in populations , multiple mating may potentially help maintain genetic variability and increase the effective population size . Thus polyandry may affect the long-term stability and effective size of G . f . fuscipes populations . In cases of eradication programmes , re-infestation of cleared areas and/or in cases of residual populations , the occurrence of remating may , unfortunately , enhance the reproductive potential of the re-invading propagules in terms of their effective population size . The comparison of two populations highlights another important factor , which , if confirmed , influences the remating frequency , i . e . the population age structure . Consequently , any vector control programme for G . f . fuscipes , according to the present results , must address the greater dimension of the young expanding population in the early wet season , and the increased rate of remating of the fewer , remaining adults after the bottleneck in the dry season . For instance in the case of SIT , a large number of sterile males should be released , also in a population with a reduced number of individuals because of the high rate of remating . These considerations agree with the recommendation to release aged , more competitive , sterile males in all cases [12] . Finally , analyses have identified the presence of parasitic Wolbachia infections in some individuals of natural populations of G . f . fuscipes , including those from Uganda described here . As it has been suggested that Wolbachia-associated incompatibilities may promote polyandry [54] , future studies can now investigate the potential influence of Wolbachia in the remating phenomenon described here . As Wolbachia infections are entertained as a tool to drive genetically desirable phenotypes into natural populations [55] , female mate choice and remating may also have an impact on strategies of population replacement .	Glossina fuscipes fuscipes is the most common tsetse species in Uganda where it is responsible for transmitting Trypanosoma brucei rhodensiense and Trypanosoma brucei gambiense parasites causing sleeping sickness in humans in addition to related trypanosomes that cause Nagana in cattle . An understanding of the reproductive biology of this vector is essential for the application of sustainable control/eradication methods such as Sterile Insect Technique ( SIT ) . We have analysed the number of times a female mates in the wild as this aspect of the reproductive behaviour may affect the stability and size of populations . We provide evidence that remating is a common event in the wild and females store sperm from multiple males , which may potentially be used for insemination . In vector eradication programmes , re-infestation of cleared areas and/or in cases of residual populations , the occurrence of remating may unfortunately enhance the reproductive potential of the re-invading propagules . We suggest that population age structure may influence remating frequency . Considering the seasonal demographic changes that this fly undergoes during the dry and wet seasons , control programmes based on SIT should release large numbers of sterile males , even in residual surviving target populations , in the dry season .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "statistics", "population", "genetics", "gene", "pool", "mathematical", "computing", "mathematics", "effective", "population", "size", "population", "modeling", "theoretical", "ecology", "population", "biology", "biostatistics", "genetic", "polymorphism", "biology", "popula...	2011	Polyandry Is a Common Event in Wild Populations of the Tsetse Fly Glossina fuscipes fuscipes and May Impact Population Reduction Measures
Encounters between human neutrophils and zymosan elicit an initially protrusive cell response that is distinct from the thin lamella embracing antibody-coated targets . Recent experiments have led us to hypothesize that this behavior has its mechanistic roots in the modulation of interactions between membrane and cytoskeleton . To test and refine this hypothesis , we confront our experimental results with predictions of a computer model of leukocyte mechanical behavior , and establish the minimum set of mechanistic variations of this computational framework that reproduces the differences between zymosan and antibody phagocytosis . We confirm that the structural linkages between the cytoskeleton and the membrane patch adherent to a target form the “switchboard” that controls the target specificity of a neutrophil's mechanical response . These linkages are presumably actin-binding protein complexes associating with the cytoplasmic domains of cell-surface receptors that are engaged in adhesion to zymosan and Fc-domains . Our recent quantitative comparison of the physical responses of human neutrophils to zymosan ( an insoluble , particulate fraction from yeast cell walls and prominent model system in the study of fungal infection [1] , [2] ) and to antibody-coated targets has exposed differences between these two forms of phagocytosis [3] . Zymosan phagocytosis typically commences as a chemotactic-like , pseudopodial protrusion that collides with , and eventually overflows , its target . In contrast , antibody-mediated phagocytosis represents more of an enveloping process in which a thin cellular lamella advances along all sides of a firmly held target to achieve engulfment [3] . The quantitative measure that most clearly illustrates this difference is the distance over which a neutrophil initially pushes an adherent target outwards ( Fig . 1A ) . This push-out distance is 1 . 03±0 . 3 ( SD ) µm in the case of zymosan , versus 0 . 12±0 . 14 ( SD ) µm for antibody-coated targets . The “detour” in cell deformation at the onset of zymosan phagocytosis proves time-consuming: zymosan engulfment takes ∼2 . 5 times longer ( 167±73 ( SD ) s ) than the uptake of antibody-coated particles of similar size ( 66±19 ( SD ) s ) . The experiments also provided the maximum cortical tension of neutrophils and the fastest speed of target inward motion during the two forms of phagocytosis . A ∼two-fold difference in cortical tension ( 0 . 3 mN/m versus 0 . 14 mN/m , with the higher value observed during zymosan engulfment ) contrasts with an indistinguishable target speed ( ∼33 nm/s ) in the two cases . Since the cortical tension is the dominant driver of target inward motion , a conserved target-internalization speed implies that the cytoplasmic viscosity rises concurrently with the cortical tension during phagocytosis , in agreement with a previously reported tight balance between the cortical tension and cytoplasmic viscosity of leukocytes [4] . Such large variations in phagocytic behavior call for a mechanistic explanation . An important clue was provided by our drug-inhibition experiments [3] , which exposed a dichotomous mechanical role of actin: in addition to driving the formation of the protrusive pseudopod that pushes zymosan outward at first , actin also participates in the suppression of protrusion during the efficient uptake of antibody-coated targets . We therefore hypothesized that structural interactions between the cytoskeleton and the membrane patch adherent to a target play a pivotal role in the target-specific mechanoregulation of phagocytosis . Although this is a plausible hypothesis , there is no guarantee that it is physically realistic or will explain the observed behavior in detail , be it in terms of cell morphology , target trajectory , or other measurable parameters . It is to verify , adapt , and refine this hypothesis that we present a quantitative analysis of its mechanical implications through the use of a previously validated computational model of neutrophil phagocytosis . Mechanistic scenarios of target engulfment are “hard-wired” into adaptations of this computational model by translating a set of “cause-effect rules” ( or physical/biochemical mechanisms deemed biologically plausible ) into mathematical equations , as explained in the Methods section . Our modeling strategy is to implement these equations in a self-consistent manner in a physically realistic computational framework , and solve the resulting problem numerically to predict the cellular behavior . A given model then is iteratively optimized by improving initial guesses of adjustable parameters until the predictions of this framework satisfactorily match experimental observations . In earlier work , we have developed and extensively validated a finite-element model of the human neutrophil , and applied it successfully to chemotactic and phagocytic mechanics [5] , [6] . It is based on the “reactive interpenetrating flow” formalism [5] , [7] . Briefly , it conceptualizes the cell interior as a mixture of two materials , the cytoskeleton and the cytosol , enclosed by an envelope , the cell membrane . The cytoskeleton plays the crucial role in controlling the dynamics of cell deformation , whereas the cytosol is a “filler” material whose relocation is driven by pressure gradients . The two phases can convert into each other , reflecting , for instance , the polymerization/depolymerization of G- and F-actin . A movable and deformable boundary endowed with surface tension represents the cortical membrane . Mass and momentum conservation equations determine the evolution of this continuum mechanical model in a self-consistent manner . To model phagocytic behavior , we have identified a number of necessary prescriptions to account for adhesion , basic signaling , and the generation of mechanical forces [6] . First , we implement an adhesive interaction between the cell surface and the target . Once cell contact with a patch of target is established , detachment is proscribed ( as observed ) . The leading edge of the adherent membrane region is assumed to stimulate transient production of a generic signaling “messenger” that locally triggers conversion of the low-viscosity ( cytosolic ) phase into the high-viscosity ( cytoskeletal ) phase , akin to cytoskeletal polymerization ( Fig . 1B ) . The degree of polymerization in turn determines the magnitude of a repulsive ( or “disjoining” ) force between the membrane and the cytoskeleton that then leads to local protrusion . ( This continuum model of protrusion encompasses , but is not limited to , the Brownian ratchet mechanism [8] . ) Finally , our previous work on neutrophil phagocytosis of antibody-coated beads exposed the necessity to invoke an attractive force between the membrane adherent to the bead and the cytoskeleton that essentially acts to “flatten” the neutrophil onto the bead [6] . Mathematical equations representing the physics of the above mechanistic concepts have been published previously [6] . We solve these model equations numerically through a Galerkin finite-element method using a mesh of quadrilaterals as described in Dembo [9] and Herant et al . [5] . Briefly , the calculation is advanced over a time step Δt ( determined by the Courant condition or other fast time scale of the dynamics ) by means of five sequential operations: This computational cycle is repeated until the simulation is complete . Cylindrical symmetry allows the use of a two-dimensional mesh to solve the axisymmetric version of the model equations . Numerical convergence is confirmed by checking that the results are not sensitive to variations of the tolerance of the different iterations performed by the code as well as to variations of the spatial resolution . Calculations are conducted on PC workstations and typically take a few hours per run . Two distinct yet closely related questions arise from the experimental comparison of the phagocytosis of zymosan and antibody-coated beads [3]: What are the differences in cell-signaling cascades that are triggered by the different receptors recognizing these targets ? What distinguishes the mechanical processes that govern the different forms of particle internalization ? Biological phagocytosis research primarily addresses the former ( e . g . , [12] , [13] ) ; in contrast , this paper focuses on the latter , i . e . , on physical mechanisms that are fundamental to our understanding of not only phagocytosis but all processes involving eukaryotic cell motility . We have previously established an optimal computational model describing the phagocytic uptake of antibody-coated beads by human neutrophils [6] . Our strategy here is to take this successful computational model of Fcγ-mediated phagocytosis as a baseline , and determine what changes must be made to recover the behavior observed in the phagocytosis of zymosan . The “virtual” phagocytic target in all simulations is a rigid spherical particle with a diameter of 3 . 2 µm . We prescribe the time course of the neutrophil cortical tension as measured , i . e . , increasing from a resting value of 0 . 025 mN/m to 0 . 3 mN/m during the engulfment of zymosan , and to 0 . 15 mN/m in the case of antibody-coated beads . We previously presented a more elaborate model of the behavior of the cortical tension [14]; however , using observed rather than modeled cortical tension values reduces the complexity of the comparative analysis of phagocytic mechanics that is the primary focus of this paper . Fig . 2 shows the results of our adapted model of zymosan phagocytosis , as well as simulations of Fcγ-mediated phagocytosis using the previously established model [6] ( see also Supporting Videos S1 and S2 ) . Differences between the two models are summarized in Table 1 . Both simulations are in excellent agreement with corresponding observations . To achieve such agreement in the case of zymosan phagocytosis , the following changes had to be incorporated into our original model of Fcγ-mediated phagocytosis: We discuss each of these aspects in turn . We have shown previously that a combination of attractive force and lack of protrusion at the membrane-bead interface accounts for two distinctive features of Fcγ-mediated phagocytosis: the absence of significant outward motion of the bead at the onset of engulfment , and the thin lamellar pseudopod embracing the bead [6] . In contrast , zymosan particles do exhibit significant outward motion before engulfment , and the pseudopods surrounding zymosan are thick ( Fig . 2; [3] ) . Hence we find that the attractive force is superfluous when modeling this case . Yet even after its removal , the modeled outward motion of the target remains much less than typically observed in zymosan phagocytosis ( Fig . 3 , blue line ) . To reproduce the measured initial push-out distances of zymosan particles , it is necessary to postulate that in this case , the protrusive force driving out free ( non-adherent ) membrane also pushes to some extent against the membrane in contact with the target – much as if the contact patch acted as a chemoattractant rather than a locus of contraction-inducing adhesion . On the other hand , when implementing this polymerization-driven protrusion force at the cell-target interface and ascribing to it the same “full” strength as at the free membrane , the resulting outward motion of the target far overshoots the observed distances ( Fig . 3 , red line ) . Only by choosing a middle ground—i . e . , setting the strength of the protrusive force at the cell-target interface to 50% of the value acting at the free membrane—are we able to recover the correct behavior ( Fig . 2 ) . How does one interpret these differences in local protrusion ? Consider a cell-surface patch in adhesive contact with an external rigid object . If the engaged adhesion receptors are strongly anchored in the cytoskeleton , they tightly couple the latter to the external object . Then , any disjoining force ( such as due to de novo actin polymerization ) must act against the tensile stress in the molecular structures linking the object to the cytoskeleton . A protrusive deformation will ensue only if the disjoining force exceeds the strength of this link . On the other hand , if an external object adheres only via membrane binding without further internal structural linkage , local disjoining forces will drive protrusion as if the contact patch was a region of free membrane . Based on this reasoning , we conclude that the adhesion of human neutrophils to antibody-coated targets exhibits strong cytoskeletal coupling , whereas in phagocytic adhesion to zymosan the cytoskeletal linkage is weak ( but not non-existent ) . We also see that the completion of engulfment takes longer for zymosan particles than for antibody-coated beads of the same size . This difference in engulfment duration can be accounted for in the simulations of zymosan phagocytosis by a 25% weaker stimulus of polymerization and protrusive force . Finally , the combination of high cortical tension and large initial target-push-out distance should lead to a much faster inward motion of zymosan particles than actually observed . The only remaining parameter that can be adjusted in the simulations to slow down the inward motion of zymosan is the interior viscosity of the cell body . Whereas this viscosity remains constant throughout the simulation of Fcγ-mediated phagocytosis , our model of zymosan phagocytosis implements a five-fold increase of the viscosity that occurs concurrently with the rise of the cortical tension . This change in cytoplasmic viscosity is assumed to take place throughout the cell interior because otherwise , the intracellular region with the lowest viscosity would determine the rate of cell rounding . To match the measurements , this lowest viscosity would have to have the value currently used for the whole cell interior , and the viscosity in the remainder of the cell would be even higher . Physically , the viscosity increase corresponds to a high degree of polymerization and/or cross-linking of cytoskeletal components throughout the cell . This computational study examines the mechanistic underpinnings of distinct physical responses of human neutrophils to zymosan and antibody-coated targets . A direct quantitative comparison of a finite-element model of the neutrophil with recent single-cell/single-target experiments [3] allows us to corroborate or discard mechanistic hypotheses about the mechanoregulation of phagocytosis . Key to the success of this comparison has been a suitable experimental design , i . e . , an essentially axisymmetric configuration that isolated the cell-target interactions of interest from potential interference by other cellular processes ( such as cell-substrate interactions of adherently kept immune cells ) . Our computational framework integrates the reactive interpenetrating flow formalism [7] with cell adhesion , basic signaling , and the autonomous generation of forces [6] . We use this framework to establish the variations in the modeled interplay of mechanical forces that most closely reproduce the observed differences in cell behavior . This enquiry complements previous studies of zymosan- and antibody-mediated phagocytosis that have focused on differences in receptor-mediated recognition and biochemical signaling [12] , [13] , [15] , [16] . By considering generic mechanistic principles , our computational approach is able to cover mechanical outcomes from underlying processes involving a range of cellular receptors and associated signaling reactions . Our overall strategy has been to vet mechanistic scenarios of phagocytic target uptake by postulating and testing biologically plausible cause-effect relationships . Additional assumptions implemented in our simulations include the irreversibility of cell-target adhesion and the time dependence of the cortical tension ( as measured ) . We do not impose any particular aspects of the cell morphology; instead , the time courses of both the shape ( including surface area ) of our “virtual immune cell” as well as the cytoskeletal density distribution ( as seen in Figs . 2 , B and E , and in the Supporting Videos S1 and S2 ) are outcomes of the simulation . Summarizing our findings , a simplified “mechanistic timeline” of zymosan phagocytosis encompasses the following stages: The primary difference between this mechanistic sequence and Fcγ-mediated phagocytosis is the following . Neutrophil contact with an antibody-coated target suppresses cell protrusion directly underneath the cell-target contact region , presumably due to a stronger structural association of the adherent membrane with the cytoskeleton . On a molecular scale , we speculate that these linkages are actin-binding protein complexes that also associate with the cytoplasmic domains of Fcγ-receptors engaged in adhesion . ( The postulated difference in cytoskeletal coupling may be due either to distinct strengths of individual linkages between receptors and actin , or to a difference in densities of the engaged receptors . ) As a result , protrusion is limited to the cell surface not in contact with the target , leading to a pseudopodial lamella that envelops the target . Worth highlighting in this physical perspective of phagocytosis is the common role of the cortical tension as primary driver of inward target motion . Note that this contrasts with the notion that an actual inward pulling force , presumably generated by molecular motors , should mainly be responsible for drawing the target into a cell . Neither our experiments ( e . g . , in the presence of myosin-II inhibitors ) [3] nor our modeling work found evidence for a significant participation of such a contractile force in inward target movement . Instead , the cortical-tension-driven tendency of a cell to round up , in conjunction with strong adhesion between the cell membrane and target , appears to be the dominant mechanical cause of target motion into the cell , as also supported by the synchronous onsets of cortical tension rise and target inward movement seen in Figs . 2 , C and D . In closing , this study not only illuminates fundamental mechanisms driving the target-specific physical immune responses of human neutrophils , it also reinforces that the present computational framework represents a biologically plausible and physically realistic model of “virtual immune cells” . In addition to correctly reproducing distinct cell morphologies observed in a range of experiments , this model also matches the dynamics of cell deformation , such as the overall engulfment times , or the time-dependent target trajectories measured in phagocytosis experiments . As an early model for calculation of a spectrum of autonomous cellular motions at this level of complexity , it is a step toward one of the key goals of computational biology , i . e . , achieving true predictive power .	Recent micropipette experiments have provided a unique live view of “one-on-one” interactions between human neutrophils and their phagocytic targets . Our results revealed surprising differences between two prominent immunological pathways: the response to fungal targets ( mimicked using zymosan particles ) , and antibody-mediated phagocytosis . Whereas antibody-coated targets were “pulled” into the cell in a straightforward manner , zymosan particles were internalized only after an initial outward “push” . We hypothesized that structural interactions between the cytoskeleton and the membrane patch adherent to a target play a pivotal role in the control of this target specificity . To verify and refine this hypothesis , we here compare our experimental results with predictions of suitable adaptations of a previously validated computational model of neutrophil mechanical behavior . By optimizing the model to best match our experiments , we corroborate that the primary mechanistic origin of the target-specific cell behavior indeed lies in the strength of cytoskeletal membrane anchors .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "biophysics/theory", "and", "simulation", "immunology/immune", "response", "immunology/innate", "immunity", "computational", "biology", "physics/interdisciplinary", "physics" ]	2011	Protrusive Push versus Enveloping Embrace: Computational Model of Phagocytosis Predicts Key Regulatory Role of Cytoskeletal Membrane Anchors
Many chronic infections , including malaria and HIV , are associated with a large expansion of CD21−CD27− ‘atypical’ memory B cells ( MBCs ) that exhibit reduced B cell receptor ( BCR ) signaling and effector functions . Little is known about the conditions or transcriptional regulators driving atypical MBC differentiation . Here we show that atypical MBCs in malaria-exposed individuals highly express the transcription factor T-bet , and that T-bet expression correlates inversely with BCR signaling and skews toward IgG3 class switching . Moreover , a longitudinal analysis of a subset of children suggested a correlation between the incidence of febrile malaria and the expansion of T-bethi B cells . The Th1-cytokine containing supernatants of malaria-stimulated PBMCs plus BCR cross linking induced T-bet expression in naïve B cells that was abrogated by neutralizing IFN-γ or blocking the IFN-γ receptor on B cells . Accordingly , recombinant IFN-γ plus BCR cross-linking drove T-bet expression in peripheral and tonsillar B cells . Consistent with this , Th1-polarized Tfh ( Tfh-1 ) cells more efficiently induced T-bet expression in naïve B cells . These data provide new insight into the mechanisms underlying atypical MBC differentiation . Over the past decade it has become evident that many chronic infections and autoimmune diseases are associated with fundamental differences in the composition and functionality of B cell memory pools . [1] For example , the chronic infections HIV , malaria , and tuberculosis—that together cause more than 5 million deaths annually and continue to elude conventional vaccine development [2]—are all associated with an expansion of somatically hypermutated CD21−CD27−CXCR3+CD11c+ B cells that upregulate inhibitory receptors and exhibit decreased effector function [3–6] , a subset of memory B cells ( MBCs ) that has been referred to as ‘atypical’ or ‘exhausted’ . It has been postulated that atypical MBCs contribute to the humoral deficiencies associated with HIV and malaria , and may pose a challenge for the development of effective vaccines for these and other chronic infections . Despite the potential clinical relevance of atypical MBCs to the prevention and treatment of chronic infections and autoimmune diseases , little is known about the cellular and molecular mechanisms that drive their differentiation . In the case of malaria , children generally mount short-lived antibody responses to Plasmodium falciparum infection , leaving them susceptible to repeated bouts of malaria [7] . Similarly , the only malaria vaccine candidate that has been tested in phase 3 clinical trials to date induces short-lived antibody responses [8 , 9] and confers only partial , short-term protection against malaria in African children [10] . It is now well-established that long-lived humoral immunity depends on the activation of highly functional T follicular helper ( Tfh ) cells that support the differentiation of naive B cells into long-lived plasma cells ( LLPCs ) and MBCs in the germinal center ( GC ) reaction [11] . Although several Tfh subsets have been described in humans , data in healthy U . S . adults indicates that Th2-polarized , CXCR3-Tfh cells provide superior B cell help [12] . Consistent with the observation that malaria induces short-lived antibody responses , we recently observed that acute febrile malaria in children preferentially activates Th1-polarized PD-1+CXCR3+ Tfh ( Tfh-1 ) cells that exhibit reduced B cell helper function [13] , which is in line with several recent studies in mice showing that excessive IFN-γ suppresses germinal center B cell responses and anti-Plasmodium humoral immunity [14–17] . Taken together , these observations suggest that Th1 cytokines and Tfh-1 cells may play a role in the differentiation of atypical MBCs . Here we conducted ex vivo analyses of immune cells of P . falciparum-exposed children in Africa , as well as in vitro modeling of malaria infection , to gain insights into the molecular and cellular conditions that are associated with the expansion of atypical MBCs . To gain insight into the regulators of atypical MBC differentiation we analyzed publicly available genome-wide expression data we generated from naïve B cells ( CD19+CD21+CD27− ) , classical MBCs ( CD19+CD21+CD27+ ) and atypical MBCs ( CD19+CD21−CD27− ) isolated from the peripheral blood of adults with lifelong malaria exposure ( S1 Table ) [5] . A principal components analysis ( Fig 1A ) of genes selected for involvement in lymphocyte differentiation and germinal center regulation ( listed in Fig 1B ) showed that atypical MBCs are transcriptionally distinct from classical MBCs . We found that relative to classical MBCs , atypical MBCs upregulate TBX21 [fold change ( FC ) 2 . 7 ( range 1 . 3–5 . 5 ) , false discovery rate ( FDR ) adjusted p value = 1 . 008 E-10] and AICDA ( FC 2 . 2 , FDR p = 0 . 048 ) , and downregulate BACH2 ( FC -2 . 1 , FDR p = 2 . 733 E-07 ) and MYC ( FC -2 . 5 , FDR p = 1 . 549 E-15 ) ( Fig 1B ) . TBX21 encodes the Th1-lineage defining transcription factor T-bet , which we found is upregulated in B cells of malaria-exposed children ( n = 15; S2 Table ) relative to healthy U . S adults ( n = 10 ) in a bi-modal distribution with approximately 18% of CD19+ B cells expressing intermediate levels of T-bet ( T-betint ) and 8% expressing high levels of T-bet ( T-bethi ) ( Fig 2A ) . On average , atypical MBCs as a percentage of total B cells were 12 . 0% and 2 . 5% for Malian children and U . S . subjects , respectively . Among T-bethi B cells , 83 . 5% were atypical MBCs ( 95% CI: 80 . 6–86 . 3 ) and 12 . 0% were activated MBCs ( 95% CI: 9 . 3–14 . 6 ) ( Fig 2B ) . Conversely , 79 . 8% of atypical MBCs ( 95% CI: 74 . 1–85 . 5 ) were T-bet+ and of these 63 . 3% were T-bethi ( 95% CI: 56 . 2–70 . 4 ) . Moreover , in an independent experiment ( n = 10 Malian children ) T-bethi B cells of malaria-exposed children expressed markers that are known to be associated with atypical MBCs , with higher surface expression of FCRL5 , CD11c , CXCR3 and CD95 , and decreased expression of CD35 , CD40 , CXCR5 and CCR7 [5 , 18] ( Fig 3 ) . Additionally , FCGR2B , a receptor known to reduce antibody production in B cells , was also upregulated in T-bethi B cells in an independent set of samples ( n = 7 Malian children ) ( Fig 4 ) . Consistent with this , T-bethi B cells exhibited lower phosphorylation of B cell receptor ( BCR ) signaling molecules following BCR cross-linking ( Fig 5A ) —a functional feature of atypical MBCs described previously . [5] Moreover , within CD21-CD27- atypical MBCs , T-bet expression correlated inversely with phosphorylation of BCR signaling molecules ( Fig 5B ) . In mice , T-bet is known to be a selective inducer of IFN-γ-mediated IgG2a class switching in B cells . [19 , 20] Therefore , we investigated the relationship between T-bet expression and isotype switching in B cells of Malian children . Overall , the percentage of IgG-expressing mature B cells was higher in Malian children compared to U . S . adults ( S1 Fig ) . We found that B cells expressing intermediate or high levels of T-bet were more likely to surface express IgG3 , compared to T-bet negative B cells which skewed toward IgG1 surface expression ( Fig 6 ) . Accordingly , T-bet expression ( MFI ) correlated with the percentage of IgG3+ B cells but not IgG1+ B cells ( S2 Fig ) . In addition , total serum IgG3 had the greatest fold increase among IgG subclasses during acute malaria ( Fig 7A ) and ( S3 Fig ) , which correlated with serum IFN-γ levels during the same malaria episode ( Fig 7B ) . Therefore , in malaria-exposed children , high T-bet expression appears to be a useful marker for atypical MBCs , as its correlates with phenotypic and BCR signaling properties of atypical MBCs . Moreover , T-bet expression in B cells of these children is associated with skewing toward IgG3 class switching . In a longitudinal analysis of serum collected from children at their healthy baseline before the malaria season , during acute febrile malaria , and 7 days post anti-malarial treatment , we confirmed in the present study population that acute febrile malaria tends to induce Th1-skewed cytokine production , with no IL-5 or IL-13 detected at any time point ( Fig 8 ) —a response that has been associated with the activation of Tfh-1 cells that exhibit impaired B cell helper function in children [13] as well as germinal center dysfunction in Plasmodium-infected mice . [14–17] Because Th1 cytokines are known to drive T-bet expression in murine B cells [20 , 21] , we hypothesized that frequent febrile malaria episodes in children would be associated with an increase in T-bethi B cells . To test this hypothesis , we compared the change in T-bet expression in B cells in a subset of age-matched children who either had ≤1 or ≥5 febrile malaria episodes documented during two consecutive years of intensive clinical surveillance , and who had PBMCs available before and after the two-year surveillance period . Malaria episodes were defined as parasitemia of ≥2500 parasites/μL of blood , an axillary temperature of ≥37 . 5°C and no other cause of fever discernible by the study physician . We found that only children with ≥5 febrile malaria episodes had a significant increase in the percentage of T-bethi B cells from before to after the two-year period ( Fig 9A–9C ) . Together these data suggest that malaria-induced Th1 cytokines drive T-bet expression in B cells , and thus play a role in the differentiation of T-bethi atypical MBCs . To test the hypothesis that P . falciparum-induced cytokines drive T-bet expression in B cells we simulated malaria infection in vitro by co-culturing PBMCs of healthy U . S . donors with P . falciparum-infected red blood cells ( iRBCs ) . The resulting supernatant induced T-bet expression in purified naïve B cells but only in the presence of BCR cross-linking by anti-IgM antibodies ( Fig 10A ) , whereas exposure of purified naïve B cells to iRBC lysate alone in the presence of BCR cross-linking did not induce T-bet expression ( Fig 10A ) . A similar pattern was observed for the activation markers CD69 and CD86 on naïve B cells ( S4A and S4B Fig ) . Similar fold increases in T-bethi B cells were observed in both naïve and memory B cells ( Fig 10B and 10C ) . Using cytokine multiplex assays , we measured 20 cytokines in the supernatants of U . S . PBMCs stimulated with uninfected RBCs ( uRBCs ) or iRBCs . Of the cytokines detected , ten ( IFN-γ , IL-2 , IL-5 , IL-9 , IL-13 , IL-17A , IL-17F , IL-21 , IL-22 and TNF ) were significantly higher in the supernatants of iRBC-stimulated PBMCs ( S5 Fig and S6 Fig ) . However , only the concentration of IFN-γ in the supernatant of PBMCs stimulated with iRBC lysate correlated with T-bet expression in naïve B cells treated with these supernatants plus anti-IgM ( Fig 11 ) . Accordingly , neutralization of IFN-γ in the supernatant or blocking IFN-γ Ra on B cells reduced T-bet expression in naïve B cells , but did not diminish expression of the activation marker CD69 ( Fig 12A and 12B ) . Of note , combining antibodies that neutralize IFN-γ or block IFN-γ Ra in the same experiment further reduced the percentage of T-bet-expressing B cells to ~4% ( Fig 12C ) . Consistent with this , BCR cross-linking plus recombinant human IFN-γ ( rhIFN-γ ) induced T-bet expression in a dose dependent manner and at a concentration of IFN-γ that was similar to that found in supernatants of iRBC-stimulated PBMCs ( Fig 13A ) . While BCR stimulation alone drove intermediate T-bet expression in B cells , high T-bet expression was only observed with the addition of supernatants of iRBC-stimulated PBMCs or rhIFN-γ ( Fig 13B ) . Taken together , these data show that P . falciparum-induced IFN-γ drives T-bet expression in B cells through the IFN-γ receptor , consistent with the observation that repeated febrile malaria episodes are associated with increased T-bethi B cells . Next , we examined the effect of IFN-γ plus BCR cross-linking on T-bet expression in tonsillar B cell subsets of U . S . children . Consistent with our findings in peripheral B cells , we found that rhIFN-γ plus BCR cross-linking induced T-bet expression in naïve B cells ( CD10- , IgD+ ) , MBCs ( CD10- , IgD- ) and light zone GC B cells ( CD10+IgD-CXCR4- ) ( Fig 14A–14C ) , whereas rhIFN-γ or BCR cross-linking alone did not significantly induce T-bet expression . Because febrile malaria in children has been shown to preferentially activate a subset of circulating Th1-polarized Tfh ( Tfh-1 ) cells [13] , we tested the hypothesis that Tfh-1 cells drive T-bet expression in B cells . We FACS-sorted PBMCs from healthy U . S . adults into naïve B cells ( CD19+CD21+IgD+ ) , Th-2 polarized circulating Tfh ( cTfh ) cells ( CD4+PD-1+CXCR5+CXCR3- ) , Tfh-1 cells ( CD4+PD-1+CXCR5+CXCR3+ ) and Th-1 cells ( CD4+PD-1+CXCR5-CXCR3+ ) ( gating strategy shown in S7 Fig ) . Autologous naïve B cells were cultured with each T cell subset in the presence of the super antigen staphylococcal enterotoxin B ( SEB ) , or with SEB alone . We observed no difference in the percentage of viable B cells among the three co-culture conditions at day 2 ( S8A Fig ) . In each condition after 2 days in culture , naïve B cells significantly upregulated the activation marker CD69 ( S8B Fig ) and Blimp-1 ( S8C Fig ) as well as AID to varying degrees ( S8D Fig ) . As expected , the secreted cytokine profile in supernatants after 2 days was Th1-skewed in the presence of Th1 and Tfh-1 cells , and Th2/Th17-skewed in the presence of cTfh cells ( Fig 15 ) . Accordingly , a higher percentage of T-bethi B cells were detected when cultured with Tfh-1 and Th1 cells , compared to B cells cultured with cTfh cells or SEB alone ( Fig 16A and 16B ) . The addition of rhIFN-γ to the naïve B cell/T cell subset culture increased the percentage of T-bet expressing B cells ( S9 Fig ) . As expected , Tfh-1 cells , which more efficiently drove T-bet expression in B cells , were less efficient in driving autologous naïve B cells to class switch ( S10A Fig ) or produce antibodies ( S10B Fig ) . Several chronic infections are associated with an expansion of ‘atypical’ or ‘exhausted’ MBCs that upregulate inhibitory receptors and exhibit decreased effector function [3–5] . Here we conducted ex vivo analyses of immune cells of malaria-exposed children in Africa as well as in vitro studies to gain insight into the cellular and molecular conditions associated with the differentiation of atypical MBCs . In malaria-exposed children we found that T-bethi B cells express markers that are known to be associated with atypical MBCs ( FCRL5 , FCGR2B , CD11c , CXCR3 and CD95 ) [5 , 18] . Consistent with this , T-bethi B cells exhibited lower phosphorylation of BCR signaling molecules following BCR cross-linking—a functional feature of atypical MBCs described previously . [5] Moreover , within CD21-CD27- atypical MBCs , T-bet expression correlated inversely with phosphorylation of BCR signaling molecules . Together these data suggest that atypical MBC differentiation occurs along a spectrum in which altered expression of key transcription factors drives the hierarchal upregulation and co-expression of inhibitory receptors ( e . g . FCRL5 , FCGR2B ) , which in turn leads to progressive loss of BCR signaling and effector functions—potentially analogous to what has been described for T cell exhaustion . [22] We found that B cells that express T-bet at intermediate or high levels were more likely to surface-express IgG3 , compared to T-bet negative B cells which skewed toward IgG1 expression . Moreover , total serum IgG3 had the greatest fold increase among IgG subclasses during acute malaria , which correlated with serum IFN-γ levels . Although differences in IgG subclasses between mice and humans make direct comparisons difficult , in mice , T-bet is a selective inducer of IFN-γ-mediated class switching to IgG2a [19 , 20 , 23] , which is functionally similar to human IgG1 and IgG3 in terms of FcR binding and complement fixation capacity . Interestingly , a recent study showed that HIV infection drives the expansion and maintenance of T-bet+ B cells ( discussed further below ) that correlate with an overrepresentation of surface-expressed and soluble IgG1 and IgG3 . [24] Therefore , there may be a consistent theme in mice and humans: that IFN-γ drives T-bet expression in B cells , which promotes class switching to IgG subclasses that are potent triggers of effector mechanisms . Further human studies are needed to determine the generalizability of these findings in other settings , and the mechanisms by which this occurs . However , based on these data we speculate that in the context of pediatric malaria , intermediate T-bet expression contributes to IgG3 class switching , while T-bet ‘overexpression’ may play a role in atypical MBC differentiation . It will also be of interest to investigate the potential role of intermediate T-bet expression in B cell activation given that the expression of CXCR3 and CD95 appears similar in T-bet intermediate and T-bet high B cells ( Fig 3 ) . A preliminary longitudinal analysis of a subset of children suggested a correlation between the incidence of febrile malaria and the expansion of T-bethi B cells . Consistent with the Th1-skewed cytokine response that febrile malaria induces , we found that Th1-cytokine-containing supernatants of iRBC-stimulated PBMCs plus BCR crosslinking induced T-bet expression in naïve B cells , a response that was abrogated by neutralizing IFN-γ or blocking the IFN-γ receptor . Accordingly , recombinant human IFN-γ plus BCR cross-linking drove T-bet expression in peripheral B cells . Importantly , the same conditions drove T-bet expression in B cells derived from secondary lymphoid tissue ( tonsils ) where germinal center reactions occur . Because Tfh cells are known to play a critical role in the activation and differentiation of naïve B cells in secondary lymphoid tissue [25] , we also examined T-bet expression in naïve B cells following co-culture with various T cell subsets . Our previous studies confirmed that functionally distinct memory Tfh cell subsets can be detected in the circulation of malaria-exposed children [13 , 26] . Additionally , we found previously that Th-1 polarized Tfh cells ( Tfh-1 ) ( T-bet+ , IFN-γ-producing ) that exhibit impaired B cell helper function are preferentially activated during acute febrile malaria in children [13] . In the present study , we observed higher T-bet expression in naïve B cells co-cultured with autologous Tfh-1 and Th1 cells , which produced high levels of IFN-γ and IL-18 , consistent with the observation that IFN-γ drives T-bet expression in naïve B cells . Whereas Tfh-1 and Th1 cells primarily induce T-bethi B cells , intriguingly , Th-2 polarized cTfh cells , which produce much less IFN-γ , appear capable of inducing both intermediate and high T-bet expression in naïve B cells , albeit at lower levels , suggesting additional mechanism by which cTfh cells may induce T-bet in B cells . Because malaria can induce IFN-γ production in multiple cell types , it remains possible that IFN-γ from sources other than Tfh cells could drive T-bet expression in B cells in vivo . However , because of the proximity of Tfh cells and B cells in secondary lymphoid tissue , it seems plausible that Tfh-derived IFN-γ may play a greater role in driving T-bet expression in B cells . Together these data support the hypothesis that malaria-induced activation of Tfh-1 cells contributes to the expansion of T-bethi atypical MBCs in malaria-exposed children . In general , little is known about T-bet+ B cells in humans [27] . In healthy adults T-bet has been detected in memory B cells and plasmablasts , but at lower levels than other T-bet+ lymphocytes . [28] T-bet expression in circulating CD21-CD27- ‘tissue-like’ MBCs has been described in healthy adults , in whom CD21-CD27- B cells are a relatively rare population [24 , 29] , but whether CD21-CD27- B cells in healthy adults represent the same population of CD21-CD27- atypical MBCs that are expanded in disease settings remains unclear . Interestingly , a recent study in humans showed that yellow fever and vaccinia vaccination stimulates an acute T-bet+ B cell response and that the T-bethiCD85jhi population may function as an early responder during acute viral infections . [24] Of note , the same study reported that HIV infection maintains an expanded T-bet+ B cell population that was primarily comprised of T-bethiCD85jhi B cells . [24] Several recent studies have described T-bet expression in B cells of individuals with autoimmune diseases . For example , transcriptome analysis of CD21-/lo versus CD21+ mature naïve B cells from subjects with rheumatoid arthritis or common variable immunodeficiency found that TBX21 expression was upregulated in CD21-/lo B cells [30] . Similarly , transcriptome analysis of CD19+ B cells isolated from individuals with systemic lupus erythematosus revealed increased TBX21 expression compared to CD19+ B cells of healthy controls . [31] Importantly , HIV and malaria-associated atypical MBCs exhibit markedly reduced cytokine and antibody production capacity [4 , 5 , 32] , whereas T-bet+ CD19+ B cells in individuals with autoimmune diseases can produce proinflammatory cytokines and autoreactive antibodies [33–35] . Therefore , T-bet+ B cells that arise in humans in the context of chronic infections versus autoimmunity may differ phenotypically and functionally , although further studies are needed to determine if this is a consistent pattern . That IFN-γ drives T-bet expression in activated human B cells is consistent with prior studies in mouse models [20 , 21 , 36] . T-bet expressing B cells termed age-associated B cells ( ABCs ) appear in mice with age , autoimmunity and viral infections [37][38 , 39] . ABCs are generated through the interplay of IL-4 , IL-21 , and IFN-γ in concert with TLR engagement [40] , and have been shown to play a role in the pathogenesis of lupus-like autoimmunity [39] and anti-viral immunity [41 , 42] . Although murine ABCs are similar to human atypical MBCs in that they upregulate T-bet and CD11c , and downregulate CD21 , unlike atypical MBCs [5] , murine ABCs proliferate in response to TLR agonists , produce IL-10 and IFN-γ and differentiate into ASCs—distinct functional profiles that call into question the relatedness of murine ABCs and human atypical MBCs that are associated with chronic infections . Instead , the available evidence suggests that murine ABCs more closely resemble the phenotype and function of T-bet+ B cells in humans with autoimmune diseases [31 , 33 , 35 , 43] . It will be of interest in future studies to employ methods such as siRNA gene silencing and ChIP-seq to determine whether T-bet plays a causal role in atypical MBC differentiation , and if so , how it directly affects B cell programming , and what other transcription factors may be involved [44] . For example , we found that expression of the gene encoding the transcription factor Bach2—which predisposes GC B cells to enter the memory pool [45 , 46]—was downregulated in atypical MBCs ( Fig 1B ) . A high priority should also be placed on ascertaining the ‘plasticity’ [47] of atypical MBCs , and whether and how their apparent loss of function can be reversed . In this regard , Kardava et al showed that HIV-associated human B cell exhaustion could be attenuated by siRNA downregulation of inhibitory receptors , particularly Fc receptor-like-4 ( FCRL4 ) and sialic acid-binding Ig-like lectin 6 ( Siglec-6 ) [32] . However , emerging data suggests that the array of inhibitory receptors expressed by atypical MBCs varies by disease; for example , malaria-associated atypical MBCs upregulate the expression of FCRL3 and FCRL5 rather than FCRL4 [5] . In summary , we show that atypical MBCs in malaria-exposed individuals highly express T-bet , and that exposure to malaria-induced Th1 cytokines and Tfh-1 cells correlates with the expansion of T-bethi B cells . These data provide insight into the mechanisms underlying atypical MBC differentiation and open the possibility of preventing or reversing the expansion of atypical MBCs in various disease states . The Ethics Committee of the Faculty of Medicine , Pharmacy and Dentistry at the University of Sciences , Technique and Technology of Bamako , and the Institutional Review Board of the National Institute of Allergy and Infectious Diseases , National Institutes of Health approved this study . Written informed consent was obtained from participants or parents or guardians of participating children prior to inclusion in the Mali study . Human pediatric tonsil tissue was obtained from the pathology department at the Children's National Medical Center in Washington , DC , and written informed consent was obtained from the parents or legal guardians of all donors . All collected tonsil samples were anonymized . The field study was conducted in the rural village of Kalifabougou , Mali where intense P . falciparum transmission occurs from June through December each year . The cohort study has been described in detail elsewhere [48] . Briefly , 695 healthy children and adults aged 3 months to 25 years were enrolled in an ongoing cohort study in May 2011 . Exclusion criteria at enrollment included a hemoglobin level <7 g/dL , axillary temperature ≥37 . 5°C , acute systemic illness , underlying chronic disease , or use of antimalarial or immunosuppressive medications in the past 30 days . The present study focused on 74 children aged 3–12 years who had venous blood samples collected at their healthy uninfected baseline before the malaria season , as well as during and 7 days after treatment of their first acute malaria episode of the ensuing 6-month malaria season . Clinical malaria episodes were detected through active and passive surveillance and were defined as ≥ 2 , 500 asexual parasites/μL , an axillary temperature of ≥37 . 5°C and no other cause of fever discernible by physical exam . All individuals with signs and symptoms of malaria and any level of parasitemia detected by microscopy were treated according to the Malian National Malaria Control Program guidelines . Peripheral blood samples from healthy U . S . adult donors enrolled in NIH protocol # 99-CC-0168 ) were also analyzed . Fresh human tonsils were obtained from the pathology department of the Children's National Medical Center in Washington , DC . All tonsils were from children . Demographic and travel history data were not available from these anonymous donors , but prior P . falciparum exposure is unlikely . Thick blood smears were stained with Giemsa and counted against 300 leukocytes . Parasite densities were recorded as the number of asexual parasites per microliter of blood based on a mean leukocyte count of 7500 cells/μL . Mali blood samples ( 8 ml ) were drawn by venipuncture into sodium citrate-containing cell preparation tubes ( BD , Vacutainer CPT Tubes ) and transported 45 km to the laboratory where PBMCs and plasma were isolated and frozen within three hours according to the manufacturer's instructions . Plasma was frozen at −80°C . PBMCs were frozen in fetal bovine serum ( FBS ) ( Gibco , Grand Island , NY ) containing 7 . 5% dimethyl sulfoxide ( DMSO; Sigma-Aldrich ) . U . S . blood samples were drawn into heparinized tubes ( BD ) and PBMCs were isolated from whole blood by Ficoll-Hypaque density gradient centrifugation ( GE Healthcare , Uppsala , Sweden ) according to the manufacturer's instructions , and frozen under the same conditions as the Malian samples . For all assays , PBMCs were rapidly thawed in a 37°C water bath , washed in PBS with 10% heat-inactivated FBS and then in complete RPMI ( RPMI 1640 with L-glutamine supplemented with 10% heat-inactivated FBS , and penicillin/streptomycin 10 , 000 μg/ml , [all from GIBCO , Invitrogen] ) . Tonsils were homogenized using wire mesh and passed through a cell strainer to make a single cell suspension . B cells were then negatively selected using a human B cell enrichment kit ( STEMCELL Technologies ) . Previously published gene expression microarray data [5] was re-analyzed to examine the expression of genes involved in lymphocyte differentiation and germinal center regulation in naïve , classical MBCs and atypical MBCs using R . Expression data were imported from . CEL files using the read . celfiles ( ) command from the oligo package library [49] and normalized using robust multi-array average ( RMA ) [50] . Samples were checked for quality using density and principal components analysis ( PCA ) plots , which confirmed the presence of a previously identified outlier sample was removed from the analysis . The data were then median normalized and filtered to remove any low information probes that were not on the list of selected genes or any genes highlighted in the previous publication [5] . Probes were considered “low information” if they had mean log2 expression less than 5 or log2 expression standard deviation below 1 . After normalization and filtering , a 3-dimensional PCA plot was created using the default settings of the prcomp ( ) analysis in R . Differential expression of naïve B cells ( N ) , classical MBCs ( C ) and atypical MBCs ( A ) was tested using limma [51] , then heatmaps were generated using heatmap . 2 ( ) from the gplots package library[52] . For surface staining and sorting , PBMCs were washed in PBS with 4% heat-inactivated FCS and cells were incubated with live/dead fixable stain ( Invitrogen ) and the indicated fluorescently labeled antibodies . Cells stained for sorting were kept on ice until sorted on a FACS Aria ( BD Biosciences ) . For intracellular transcription factor staining , PBMCs were treated with Transcriptional Factor Fixation/ Permeabilization kit ( ebioscience ) . FACS analyses were performed on a BD LSR II flow cytometer ( BD Biosciences ) and analyzed using FlowJo software ( Tree Star , Inc ) . Antibody details in S3 Table . PBMCs were FACS-sorted into PD-1+CXCR3+CXCR5+CD4+ ( Tfh-1 cells ) , PD-1+CXCR3-CXCR5+CD4+ ( cTfh cells ) , PD-1+CXCR3+CXCR5-CD4+ ( Th1 cells ) , CD4-CD19+CD21+CD27- ( naïve B cells ) and CD4-CD19+CD21+CD27+ ( MBCs ) . Each T cell subset ( 1 . 5 × 104 to 5 × 104 ) was co-cultured with naïve B cells or MBCs at ratio of 1:1 for 2 , 4 , 6 , 8 or 12 days in complete medium with staphylococcal enterotoxin B ( SEB ) ( 1 . 5μg/ml; Sigma-Aldrich ) in 96 round U-bottomed plates at 37°C . After co-culture , B cell number , phenotype , cytokine and Ig levels in supernatants were assessed using multiplex cytokine or isotyping assays and staining cells with anti-human monoclonal antibodies . Antibody details in S3 Table . PBMCs were cultured with uninfected red blood cells ( uRBCs ) or P . falciparum-infected red blood cells ( iRBCs ) . PBMCs were cultured in complete RPMI ( RPMI 1640 plus 10% fetal calf serum , 1% penicillin/streptomycin ) in flat-bottom 96 well plates , at 37°C in a 5% CO2 atmosphere . PBMCs were stimulated with iRBC or uRBC lystate in a ratio of 3 RBCs per PBMC . At day three of co-culture , cells were centrifuged and supernatants were recovered for cytokine analysis or frozen at -80 degrees . Plasma or supernatants were thawed and immediately analyzed with either ProcartaPlex human cytokine assays ( affymetrix , ebioscience ) or LEGENDplex ( Biolegend ) as recommended by the manufacturer . Briefly , 100 μL of plasma at 1∶2 dilution or 100 μL of supernatant were incubated with anti-cytokine antibody-coupled magnetic beads at room temperature shaking at 300 RPM in the dark . After several washes , the beads were then incubated with a biotinylated detector antibody at room temperature before incubation with streptavidin-phycoerythrin . Finally , the complexes were resuspended in 125 μL of detection buffer and 100 beads were counted with a Luminex 200 device ( BioRad Laboratories , Inc . ) . Final concentrations were calculated from the mean fluorescence intensity and expressed in pg/mL using standard curves with known concentrations of each cytokine . Total B cells or naïve B cells from PBMC were negatively selected using EasySep Human B Cell Enrichment Kit ( STEMCELL Technologies ) as recommended by the manufacturer . Using the appropriate antibodies , the final purities of the start and isolated fractions were assessed . The purity of the final fraction was approximately 98% . For PBMC supernatant-B cell stimulation , B cell receptors were cross-linked with 1–2μg/mL of anti-IgM ( for naïve B cells only ) or 2μg/mL each of anti-λ/κ light chains ( for total B cells ) in the presence of PBMC , PBMC/uRBC , PBMC/iRBC or iRBC lysate only supernatant . For cytokine blocking experiments , anti-cytokine neutralizing or receptor blocking antibodies were added either before or during the addition of supernatant . B cell receptors were also cross-linked in the presence of different concentrations of recombinant cytokines . For tonsillar B cell studies , total B cells negatively isolated from tonsils were stimulated with 2μg/mL each of anti-λ/κ light chains in the presence or absence of recombinant human IFN-γ . Staining with anti-CD10 , IgD and CXCR4 , identified the different B cell subsets . Whole PBMCs were thawed and plated on a 96 well plate with 2 × 106 cells of each donor per well , and stained for CD10 , CD19 , CD20 CD21 , CD27 and FCRL5 ( 509f6 ) ( Biolegend ) at 4°C in 4% PBS-FBS for 20 min . Cells were washed in 0 . 5% PBS-BSA and incubated at 37°C for 30 min before adding F ( ab' ) 2 anti-IgM and anti-IgG ( Southern Biotech , Birmingham , AL and Jackson ImmunoResearch , respectively ) at a final concentration of 10 μg/ml and incubating at 37°C for 5 min . For the detection of phospho-proteins by flow cytometry , cells were fixed and permeabilized according to the manufacturer's protocol using the FoxP3 Staining Buffer Set ( eBioscience ) . Cells were then stained with antibodies specific for phospho-Syk ( Y352 ) ( pSyk ) and phospho-PI3Kinase p85 ( Y458 ) /p55 ( Y199 ) ( pPI3K ) and T-bet . Antibody details in S3 Table . Continuous data were compared using the paired or unpaired Student's T-test and ANOVA . Bonferroni adjustments ( T-tests ) and Sidak adjustments ( ANOVA ) were applied to correct for multiple comparisons where appropriate . Correlations were calculated with Pearson correlation coefficient and their significance was determined using Fisher’s Z-test . All analyses were performed in Prism 6 . 0e ( GraphPad Software ) or R 3 . 1 . 2 [53] .	Antibodies are proteins in blood that help kill microbes such as viruses , bacteria and parasites . Antibodies are produced by B cells with the help of T follicular helper ( Tfh ) cells . Some microbes for which we have no effective vaccines , such as HIV and malaria , establish chronic infections that are not cleared by the immune system . These chronic infections are associated with ‘atypical’ B cells that are less able to produce antibodies . We studied blood samples of malaria-exposed children to understand why normal B cells become atypical B cells . We found that atypical B cells express high levels of T-bet—a protein that is important for determining the fate of other types of immune cells . Children who frequently got malaria had more T-bet expressing B cells than children who rarely got malaria . We also found that malaria parasites cause immune cells to secrete inflammatory substances that cause normal B cells to express T-bet . Similarly , the inflammation-prone Tfh cells that malaria activates , which are relatively poor B cell helpers , also caused normal B cells to express T-bet . This study helps us understand why atypical B cells arise during chronic infections—information that could lead to strategies to improve antibody responses through vaccination .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "blood", "cells", "chemical", "bonding", "innate", "immune", "system", "medicine", "and", "health", "sciences", "children", "immune", "cells", "immune", "physiology", "cytokines", "immunology", "tropical", "diseases", "parasitic", "diseases", "cell", "differentiation", ...	2017	Malaria-induced interferon-γ drives the expansion of Tbethi atypical memory B cells
In 2008 a nosocomial outbreak of five cases of viral hemorrhagic fever due to a novel arenavirus , Lujo virus , occurred in Johannesburg , South Africa . Lujo virus is only the second pathogenic arenavirus , after Lassa virus , to be recognized in Africa and the first in over 40 years . Because of the remote , resource-poor , and often politically unstable regions where Lassa fever and other viral hemorrhagic fevers typically occur , there have been few opportunities to undertake in-depth study of their clinical manifestations , transmission dynamics , pathogenesis , or response to treatment options typically available in industrialized countries . We describe the clinical features of five cases of Lujo hemorrhagic fever and summarize their clinical management , as well as providing additional epidemiologic detail regarding the 2008 outbreak . Illness typically began with the abrupt onset of fever , malaise , headache , and myalgias followed successively by sore throat , chest pain , gastrointestinal symptoms , rash , minor hemorrhage , subconjunctival injection , and neck and facial swelling over the first week of illness . No major hemorrhage was noted . Neurological signs were sometimes seen in the late stages . Shock and multi-organ system failure , often with evidence of disseminated intravascular coagulopathy , ensued in the second week , with death in four of the five cases . Distinctive treatment components of the one surviving patient included rapid commencement of the antiviral drug ribavirin and administration of HMG-CoA reductase inhibitors ( statins ) , N-acetylcysteine , and recombinant factor VIIa . Lujo virus causes a clinical syndrome remarkably similar to Lassa fever . Considering the high case-fatality and significant logistical impediments to controlled treatment efficacy trials for viral hemorrhagic fever , it is both logical and ethical to explore the use of the various compounds used in the treatment of the surviving case reported here in future outbreaks . Clinical observations should be systematically recorded to facilitate objective evaluation of treatment efficacy . Due to the risk of secondary transmission , viral hemorrhagic fever precautions should be implemented for all cases of Lujo virus infection , with specialized precautions to protect against aerosols when performing enhanced-risk procedures such as endotracheal intubation . Viral hemorrhagic fever ( VHF ) is an acute systemic illness classically involving fever , a constellation of initially nonspecific signs and symptoms , and a propensity for bleeding and shock . VHF may be caused by more than 25 different viruses from four taxonomic families: Arenaviridae , Filoviridae , Bunyaviridae , and Flaviviridae . Transmission of hemorrhagic fever viruses is through direct contact with blood and bodily fluids during the acute illness . Although patient isolation and specific VHF precautions ( consisting of surgical mask , double gloves , gown , protective apron , face shield , and shoe covers ) are advised for added security , experience has shown that routine universal and contact precautions are protective in most cases [1] . Aerosol precautions , such as the use of N95 particulate filters , are only recommended when performing specific potentially aerosol-generating procedures , such as endotracheal intubation . South Africa has often played a role of “sentinel” for VHF in countries further to the north through the travel and admission of undiagnosed patients to South African hospitals , often with subsequent nosocomial transmission to healthcare workers . For example , cases of Marburg and Ebola hemorrhagic fevers have been reported in Johannesburg in persons initiating travel in Zimbabwe [2] and Gabon [3] , respectively . In 2008 a nosocomial outbreak of five cases of VHF occurred in Johannesburg [4] , [5] ( figure 1 ) . The primary patient was a tour operator who was evacuated from Lusaka , Zambia . The etiologic agent was determined to be a novel arenavirus and the name “Lujo virus” was proposed . The source of the patient's infection is unknown , but assumed to be a rodent , as with all other pathogenic arenaviruses . Recent field studies of small mammals in Zambia did not result in isolation of Lujo virus , although another novel arenavirus was discovered [6] . Arenaviruses are divided into two groups: the New World ( or Tacaribe ) complex , and the Old World ( or Lymphocytic Choriomeningitis/Lassa ) complex , with various members of both groups causing VHF in South America and Africa , respectively [7] Lassa virus , the distribution of which is confined to West Africa , is the only other Old World arenavirus associated with VHF [8] . Lujo virus is only the second pathogenic arenavirus to be recognized in Africa and the first in over 40 years . Some arenavirus infections , especially Lassa fever , have shown benefit with the use of the nucleoside analogue ribavirin [9] . Because of the remote and resource poor locations where Lassa fever typically occurs , as well as the history of civil unrest in West Africa in recent decades , there have been few opportunities to undertake in-depth study of the clinical manifestations or pathogenesis of Lassa fever or other VHFs , or the response of these infections to treatment options typically available in industrialized countries . We describe the clinical features of the five recognized cases of Lujo hemorrhagic fever ( LHF ) in the 2008 outbreak in South Africa and summarize their clinical management , as well as providing additional epidemiologic detail , with a focus on the risks for secondary transmission . The initial description of the outbreak [4] was published primarily under the auspices of the South African National Institute for Communicable Diseases , which had a blanket ethics approval for use of all the patients' data . The same data set has been used for this publication , with ethics committee approval , with the exception of further data collated on the one survivor , who provided written consent for use of data and images related to her illness . The five patients' ages ranged from 33 to 47 years . There were two white females , two black females , and one white male . The incubation periods of the 3 secondary and 1 tertiary cases ranged from 9-13 days . Four of the five patients died ( CFR 80% ) . Based on the five cases of LHF recognized to date , the clinical disease associated with LHF is remarkably similar to Lassa fever [7] . Surprisingly , the two viruses are genetically quite distinct ( up to 38 . 1% on the nucleotide level ) , with Lujo virus grouping much closer genetically to Old World arenaviruses not associated with VHF [5] Lassa fever classically begins with non-specific signs and symptoms including fever , general malaise , headache , myalgia , chest or retrosternal pain , and sore throat with progressive diarrhea and other gastrointestinal involvement [7] , [9] . Severe cases may progress to a capillary leak syndrome with septic shock , rash , facial and neck swelling , and multi-organ system failure . The facial and neck swelling seen in both LHF and Lassa fever appear to be specific to Old World arenavirus infection and may help differentiate it from other African VHFs . Like in Lassa fever ( and despite the slight misnomer “VHF” ) , major bleeding was not a prominent feature in the patients with LHF , although minor bleeding was common . The AST and ALT are typically elevated in Lassa fever , with AST much greater than ALT and high levels of AST associated with a poor prognosis [7] . This same pattern was seen in all five patients with LHF , with the only survivor manifesting the lowest peak AST and AST: ALT ratio . Some distinctive features of LHF relative to typical Lassa fever were the abrupt disease onset ( typically indolent in Lassa fever ) and the presence of DIC , which is generally not considered to be part of the pathogenesis of Lassa fever , although the matter has not been extensively studied [9] . Although rash is consistently seen in light-skinned persons with Lassa fever , for unknown reasons it is almost never seen in blacks . All of the white patients and one of the two black patients with LHF manifested a very prominent rash . Interestingly , the black patient without rash was HIV infected , suggesting that the rash of LHF may be immune mediated . Patient 5 also had relative bradycardia , an interesting finding given reports of depressed cardiac function in an animal model of arenavirus infection [12] . The CFR associated with this outbreak of LHF was 80% . The CFR of hospitalized patients with Lassa fever is typically in the 20–30% range , ranging up to 50% in some nosocomial outbreaks [13] . However , mild and asymptomatic Lassa virus infection is thought to be common , with mortality rates less than 5% when infection in the community is considered [7] , [14] . No antibody survey of case contacts or community members in the region of origin of the index case in Zambia has been conducted to determine if mild or asymptomatic infection with Lujo virus occurs . The four nosocomial infections of Lujo virus illustrate the risk to healthcare workers . Although no specific exposures were reported and some degree of personal protective equipment was worn by all four secondary or tertiary cases , it appears that strict barrier nursing practices were not always maintained and full VHF precautions were often implemented late in the course of treatment , if at all . Furthermore , the four infected healthcare workers generally had very close and sometimes prolonged contact with the patient , including in closed settings , such as the medical evacuation flight of Patient 1 , augmenting the possibility of exposure to blood and bodily fluids . They also performed procedures that are often considered to be high risk , such as endotracheal intubation , insertion of indwelling intravascular catheters , and dialysis . The transmissibility of other emerging viruses such as SARS and MERS coronaviruses has similarly been enhanced when such procedures have been performed [15] . In addition to the 4 secondary/tertiary cases , another 94 persons were identified as contacts and monitored , including support staff ( kitchen , laundry , cleaning ) , laboratory and radiography technicians , and nursing staff . We did not categorize contacts in terms of risk at the time , but now estimate that at least 30 of these would be reasonably categorized as high risk . Nevertheless , no suspected cases of LHF were noted in this group . We suspect that the degree of transmissibility of Lujo virus is likely analogous to that of Lassa virus , for which , although reliable reproduction numbers and secondary attack rates are difficult to ascertain , they are generally thought to be low . Nevertheless , occasional outbreaks with secondary and tertiary cases are sometimes seen , especially when barrier nursing practices are not maintained [16] , [17] . Until the matter can be studied more thoroughly , VHF precautions should certainly be implemented for all suspected and confirmed cases of LHF , with specialized precautions to protect against aerosols when performing endotracheal intubation [1] . Despite the high prevalence of HIV infection in many areas of sub-Saharan Africa , including some areas where VHF is common , data are scarce on HIV and hemorrhagic fever virus co-infection , such as was the case with our Patient 4 . She was also infected with hepatitis B virus . A 68 year old Sierra Leonean man with a history of HIV infection and chronic progressive neurological deterioration was infected with Lassa virus in 2006 [18] The patient survived despite severe disease requiring intubation and mechanical ventilation . In the 2000–2001 outbreak of Ebola virus in Uganda , the CFR was not statistically different between those who were HIV positive and negative [19] . The samples were anonymously tested and no clinical data were reported . Although the clinical data on Patient 4 are also sparse , there were no obvious differences in the clinical manifestations of LHF in this patient compared to the others , with the exception of the aforementioned absence of rash . It is also interesting to note that her peak fever ( 38 . 5°C ) and leukocyte count ( 14×109/L ) were not particularly high , consistent with her compromised immune system . There have been very few controlled studies on the management of VHF . Most recommendations represent the informal consensus of experienced clinicians and investigators . Supportive therapy is the mainstay [20] . The pathogenesis of severe cases of VHF is thought to be similar to severe sepsis , with a severe inflammatory response syndrome mediated in part by various soluble cytokines and chemokines and nitric oxide [21] . Therefore , the basic management principles of shock are also recommended for VHF [20] , [22] However , since most VHFs occur in resource-poor areas with little access to advanced ICU medicine , opportunities to use and make observations on the efficacy of these or other advanced treatment options are rare . Although obviously not a controlled trial , we were nevertheless able to make some detailed observations on the management of five patients with LHF , who were often treated in more advanced healthcare settings . The most detailed data are from Patient 5 , who was the only patient for whom a specific diagnosis of VHF was considered and confirmed early in the course of disease . Despite receiving ribavirin at disease onset , Patient 5's clinical status deteriorated and her illness was severe and prolonged . Although these results could be interpreted as lack of efficacy of ribavirin against Lujo virus , this is unlikely considering the drug's proven efficacy in other arenavirus infections [8] , [23]–[25] Of greater importance was probably the fact that ribavirin was administered orally for the first 6 days of treatment . Efficacy of oral ribavirin for arenavirus infection has not been definitively shown and , in light of the significant first-pass hepatic metabolism resulting in an oral bioavailability of only ∼50% , it is unlikely that oral administration reliably reaches the minimum inhibitory concentration for arenaviruses in serum [26] Serum levels are undoubtedly further diminished by decreased gut absorption , vomiting , and diarrhea in these severely ill patients . Various adjunctive therapies with demonstrated or theoretical efficacy in severe sepsis were administered to Patient 5 and a few of the other patients , including HMG-CoA reductase inhibitors ( statins ) , N-acetylcysteine [27] , [28] , recombinant factor VIIa , [29] , [30] , [31] mechanical ventilation , plasmapheresis , and hemodialysis . Animal models of sepsis have suggested that statin drugs may improve outcomes in septic shock [32] , [10] . Furthermore , a large , population-based cohort analysis in Canada showed reduced risk of sepsis in patients with cardiovascular disease who were treated with statins [11] . Patient enrolment is currently ongoing for prospective trials of statin therapy after the development of sepsis . N-acetylcysteine is an antioxidant and free radical scavenger that resulted in decreased nuclear factor-κB and interleukin-8 in patients with sepsis , suggesting a blunting of the inflammatory response [28] . Recombinant factor VIIa is a prohaemostatic agent thought to act at the local site of tissue injury and vascular wall disruption by binding to exposed tissue factor to promote generation of thrombin and platelet activation . [29] . The drug has been used in hemophilia and other coagulation disorders , as well as in liver disease , reversal of anticoagulant therapy , and for episodes of excessive or life threatening bleeding related to surgery or trauma [30] , [31] . Other therapies being explored for sepsis and , in some cases specifically for VHF , such as the recombinant inhibitor of the tissue factor/factor VIIa coagulation pathway , rNAPc2 , and activated protein C , were not used in this outbreak due to lack of availability and/or risk of bleeding . The seemingly counterintuitive use of anticoagulants like rNAPc2 stemmed from work with an Ebola virus animal model to ameliorate the effects of tissue factor resulting in DIC [21] . It is difficult to assess the contribution of the various therapies to the patient outcomes . Although hemofiltration has been suggested in patients with refractory hemodynamic septic shock , with a significant decrease in ICU mortality in responders [33] , and plasmapheresis appeared to have a brief positive effect in Patient 2 , we are reluctant to advocate treatments or procedures that potentially increase healthcare worker exposure to blood . In fact , one explanation for the high secondary attack rate associated with this outbreak could be that such high-risk procedures were frequently undertaken . Many of the drugs employed in the management of Patient 5 are already clinically approved . Investigation of many of these compounds in animal models of VHF is warranted , including in LHF model using strain 13/N guinea pigs [34] . Ideally , controlled clinical trials in humans would also be undertaken , although the feasibility of this is dubious for most VHFs , with the possible exception of Lassa fever , for which many infections occur across West Africa , or perhaps through a “multicenter” approach through advanced planning with Ministries of Health and other partners in endemic areas for VHFs [35] , [36] , [21] . Until controlled efficacy data are available , and considering the high CFR often associated with VHF , we feel that it is both logical and ethical to explore the use of these approved compounds in treatment of patients with VHF when possible . Treating clinicians should make a concerted effort to collect and publish detailed , repeated , and systematic clinical observations to facilitate objective evaluation of their efficacy . The pace of discovery of arenaviruses has increased considerably in recent years , with over ten new viruses being isolated since 2000 . Pathogenic arenaviruses will almost certainly continue to be discovered . Furthermore , rapid population growth , especially in Africa , and incursion for both economic and leisure activities into natural habitats harboring rodents will likely put humans at risk . The clinical findings and management experience reported here will be of use to clinicians faced with patients with arenavirus infections and as well as other VHFs .	Viral hemorrhagic fever is a syndrome often associated with high fatality and risk of secondary transmission . In 2008 , an outbreak of a novel hemorrhagic fever virus called Lujo occurred in Johannesburg , South Africa , with secondary transmission from the index patient to four healthcare workers . Four of the five patients died . Lujo belongs to the arenavirus family and is only the second pathogenic arenavirus , after Lassa virus , to be recognized in Africa and the first in over 40 years . Because most viral hemorrhagic fevers occur in remote , resource-poor settings , few in-depth controlled studies of their clinical manifestations , transmission dynamics , pathogenesis , or response to treatment options are possible . We describe the clinical features of the five cases in this outbreak and summarize the clinical management , as well as providing additional epidemiologic detail . Lujo virus causes a clinical syndrome remarkably similar to Lassa fever . The treatment options used in these five cases are discussed as well as the recommended precautions to prevent secondary transmission .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "viral", "diseases", "infectious", "diseases", "medicine", "and", "health", "sciences" ]	2014	Clinical Features and Patient Management of Lujo Hemorrhagic Fever
Salmonella Typhi and Salmonella Paratyphi A are the agents of enteric ( typhoid ) fever; both can establish chronic carriage in the gallbladder . Chronic Salmonella carriers are typically asymptomatic , intermittently shedding bacteria in the feces , and contributing to disease transmission . Detecting chronic carriers is of public health relevance in areas where enteric fever is endemic , but there are no routinely used methods for prospectively identifying those carrying Salmonella in their gallbladder . Here we aimed to identify biomarkers of Salmonella carriage using metabolite profiling . We performed metabolite profiling on plasma from Nepali patients undergoing cholecystectomy with confirmed S . Typhi or S . Paratyphi A gallbladder carriage ( and non-carriage controls ) using two-dimensional gas chromatography coupled with time-of-flight mass spectrometry ( GCxGC-TOFMS ) and supervised pattern recognition modeling . We were able to significantly discriminate Salmonella carriage samples from non-carriage control samples . We were also able to detect differential signatures between S . Typhi and S . Paratyphi A carriers . We additionally compared carriage metabolite profiles with profiles generated during acute infection; these data revealed substantial heterogeneity between metabolites associated with acute enteric fever and chronic carriage . Lastly , we found that Salmonella carriers could be significantly distinguished from non-carriage controls using only five metabolites , indicating the potential of these metabolites as diagnostic markers for detecting chronic Salmonella carriers . Our novel approach has highlighted the potential of using metabolomics to search for diagnostic markers of chronic Salmonella carriage . We suggest further epidemiological investigations of these potential biomarkers in alternative endemic enteric fever settings . Enteric fever is a systemic infection caused primarily by Salmonella enterica serovars Typhi ( S . Typhi ) and Paratyphi A ( S . Paratyphi A ) ; the disease has the highest incidences in low- and middle-income countries with limited access to clean water and poor hygiene standards [1] . The causative agents are transmitted via the fecal-oral route and , after entering the gastrointestinal tract , they translocate through gut mucosa and spread systemically , reaching the liver , spleen , bone marrow , and gallbladder [2] . There are limited data on the precise modes of pathogenesis and interactions between invasive Salmonella and the human host , but the organisms are thought to employ several strategies for avoiding immune defenses [3] . One of the principal covert mechanisms of S . Typhi and S . Paratyphi A is the ability to colonize the gallbladder , resist the antimicrobial activity of bile [4] , and induce a chronic carriage [5 , 6] . Colonization of the gallbladder is thought to be facilitated by the formation of biofilms [7] , which may be associated with an inability to effectively treat carriage with antimicrobials [5] . It is estimated that approximately 2–5% of individuals living in endemic areas that have experienced an episode of enteric fever will become a chronic carrier [8] . Typically , chronic carriers , such as Typhoid Mary , are asymptomatic [9] and many have no history of an acute disease episode [10] , making the prospective detection of S . Typhi or S . Paratyphi A carriers a major challenge . Nonetheless , intermittent fecal shedding of bacteria is thought to be a major contributing factor for disease maintenance in endemic areas and the detection of these individuals is considered to be a public health priority for reducing the disease burden [11 , 12] . Current methods for the prospective detection of Salmonella chronic carriers include serial fecal culture[13] , the detection of the bacteria in bile or on gallstones after cholecystectomy[5] , and elevated antibody responses against the Vi polysaccharide antigen[14] . However , all these methods have various limitations ( e . g . logistics , invasiveness , and sensitivity ) and are seldom performed [5 , 13 , 15] . New approaches for detecting Salmonella carriers are warranted , and we have previously used metabolomics for identifying biomarkers associated with acute enteric fever [16 , 17] . Here , we aimed to identify metabolite biomarkers specific for S . Typhi or S . Paratyphi A carriage . Using plasma samples from patients undergoing cholecystectomy with confirmed S . Typhi or S . Paratyphi A gallbladder carriage and appropriate controls we performed two-dimensional gas chromatography coupled with time-of-flight mass spectrometry ( GCxGC-TOFMS ) . After generating metabolite profiles , our primary strategy was to use chemometric bioinformatics to investigate if carriers could be differentiated from non-carriers , thus generating a biomarker pattern ( latent biomarker ) of carriage that may be developed into a diagnostic methodology . As opposed to focusing on single metabolites as diagnostic markers , we hypothesize that a combination of co-varying metabolites could potentially provide a means for more sensitive and specific biomarkers of typhoid carriage . Additionally , as secondary aims we sought to distinguish between S . Typhi and S . Paratyphi A carriers and to investigate if the identified metabolite profiles were unique to carriage with respect to those identified during acute infection [16] . This study was conducted according to the principles expressed in the Declaration of Helsinki and was approved by the institutional ethical review boards of Patan Hospital , The Nepal Health Research Council , and the Oxford Tropical Research Ethics Committee ( OXTREC , Reference number: 2108 ) . All enrollees were required to provide written informed consent for the collection , use and storage of tissue and blood collected surgery . The study was conducted at Patan Hospital , a 318-bed government hospital located in the Lalitpur Sub-Metropolitan City in the Kathmandu valley , Nepal . Non-specific febrile disease is common at Patan Hospital and S . Typhi and S . Paratyphi A are the most common bacteria cultured from blood of febrile patients in this location [18] . The study generating the plasma samples has for this investigation been previously described [19] . Briefly , EDTA blood was collected from a subset of patients undergoing cholecystectomy at Patan Hospital from June 2007 to October 2010 . A questionnaire related to the patient’s health and demographics was administered prior to surgery along with a stool sample for microbiological culture . Surgeons collected bile samples and gallbladder tissue during the procedure . After recruiting 1 , 377 cholecystectomy patients over three years and culturing their bile we identified 24 and 22 individuals with S . Typhi and S . Paratyphi A inside their gallbladder , respectively; 35/46 ( 76% ) were female and the median age was 34 . 5 years ( range; 20–67 ) . Plasma samples were stored at -80°C until analysis . The samples constituted a subset of those enrolled in the study and were comprised of plasma samples from patients with confirmed S . Typhi ( n = 12 ) and S . Paratyphi A ( n = 5 ) gallbladder carriage i . e . individuals from whom S . Typhi or S . Paratyphi A was isolated from their bile after cholecystectomy . We additionally analyzed plasma from individuals who underwent surgery but had sterile bile ( n = 20 ) i . e . a surgical control population without exhibiting growth of bacteria in their bile . Additional patient information and patient group metadata can be found in S1 File and S1 Table . For extraction and derivatization of the samples prior to analysis with GCxGC-TOFMS we used the plasma protocol for metabolomics analysis at the Swedish Metabolomics Centre ( SMC ) with a 50μl starting volume [20] ( and described in detail in [16] ) . Quality control ( QC ) samples were prepared by pooling 100μl aliquots of seven samples from each class . The extracted and derivatized plasma samples were analyzed on a GCxGC-TOFMS as described previously [16] but with minor modifications . The MS transfer line was set at 325°C and the detector voltage at 1 , 700V . The analytical run order was constructed by randomizing the three sample groups ( S . Typhi carriers , S . Paratyphi A carriers , and non-carriage controls ) . The run also included QC samples incorporated at the beginning and end of the run and after every sixth sample . In addition , blank samples ( milli-Q water and extraction mix ) and n-alkane series ( C8-C40 ) for retention index calculation were analyzed . The GCxGC-TOFMS data was processed as NetCDF files using an in-house Matlab script ( MATLAB R2014b , Mathworks , Natick , MA , USA ) applying hierarchical multivariate curve resolution ( HMCR ) [21] on GCxGC-TOFMS data . The processing resulted in resolved chromatographic peaks with semi-quantitative metabolite concentrations and corresponding mass spectra . The mass spectra were subjected to library search in NIST MS Search 2 . 0 to give the peaks a putative annotation . In-house libraries from SMC and publicly available libraries ( from US National Institute of Science and Technology ( NIST ) and the Max Planck Institute in Golm ( http://gmd . mpimp-golm . mpg . de/ ) ) together with libraries of all sample peaks ( to detect split peaks ) and peaks from the acute enteric fever study were used in the identification . Split peaks and metabolites with a different number of TMS groups were investigated by comparing retention indexes , mass spectral matches , raw data profiles and loading positions in a PCA model and for peaks with comparable values/profiles only one peak was included in further analysis . Other criteria for peak exclusion were: low quality peak/mass spectrum , column bleed artifacts , internal standard , high analytical run order correlation ( Pearson correlation coefficient >❘0 . 5❘ ) , and highly deviating QC samples ( RSD>0 . 5 , peaks with RSD between 0 . 3 and 0 . 5 were manually investigated ) . Multivariate projection methods provide a statistical tool for deciphering multivariate or latent biomarkers based on combinations of variables ( e . g . metabolites , proteins etc . ) . Initially unsupervised modeling with principal component analysis ( PCA ) [22] was used for peak investigation during the identification process , and for data overview with metabolites selected for inclusion . From PCA modeling the general trends in the metabolite data can be obtained with possible outliers . The variable raw data was investigated and highly deviating samples were detected in a few metabolites . Metabolites with one sample having a centered and UV scaled value >4 in a PCA model with all samples were investigated . PCA models with and without missing value replacement were obtained to investigate the effect . In general , metabolites with a value >5 were selected for missing value replacement but also metabolites where the deviating sample was affecting the significance of the metabolite ( either by reinforcing a weak trend or creating an opposite trend compared to the majority of the samples in the same class ) . All the following models were calculated with missing value replaced data . Unsupervised modeling was followed by supervised modeling using orthogonal partial least squares-discriminant analysis ( OPLS-DA ) [23] . OPLS-DA models were obtained to investigate the metabolite profiles related to chronic carriage for i ) a three-class model with S . Typhi carriers , S . Paratyphi A carriers , and non-carriage controls , ii ) a combination of S . Typhi and S . Paratyphi A carriage samples compared to non-carriage controls , iii ) separate models for S . Typhi and S . Paratyphi A carriage samples compared to non-carriage controls and iv ) for S . Typhi carriage samples compared to S . Paratyphi A carriage samples . All data was centered and scaled to unit variance and all OPLS-DA models were validated with a seven-fold cross-validation [24] . The multivariate significance criterion was based on the latent significance concept recently developed in our research group ( Jonsson et al . , submitted ) . The latent significance concept takes advantage of the unique features of OPLS modeling to highlight significant metabolites . In OPLS , the variation in the measured variables ( here metabolite concentrations ) can be divided into one part related to the response of interest ( here class information regarding carriage ) and into one part unrelated to the response . In the latent significance concept the orthogonal variation is subtracted , creating latent model covariance loadings , wlatent , to be able to highlight metabolites with a significant alteration only focusing on the response of interest . Univariate p-values were calculated using the Mann-Whitney U-test and metabolites with p≤0 . 05 were considered univariate significant . The selection of significant metabolites was based on multivariate significance only . To investigate how well metabolite patterns and the individual metabolites could distinguish between the sample groups receiver operating characteristic ( ROC ) curves were constructed . For the metabolite panels the cross-validated scores from the OPLS-DA models were used to construct the ROC curves and for the individual metabolites the relative metabolite concentrations were used . Area under the curve ( AUC ) values were calculated from the ROC curves with values ranging from 0 . 5 to 1 ( where 0 . 5 represents a random classifier and 1 represents a perfect classifier ) . The bootstrap percentile resampling method ( using 1 , 000 bootstrappings ) was used to calculate 95% confidence intervals for the AUC values[25] . Modeling was performed in SIMCA ( version 14 , Umetrics , Umeå , Sweden ) , ROC curve analysis was performed in Matlab ( R2014b , Mathworks , Natick , MA , USA ) and figures were created in GraphPad Prism ( 5 . 04; GraphPad Software Inc . , La Jolla , CA , USA ) . In order to compare the profiles during acute enteric fever and chronic carriage the metabolites found to be significant for separating carriers from non-carriers in the current study were compared to significant metabolites separating acute enteric fever samples from afebrile controls in a previous study [16] . Initially , the mass spectra for the carriage samples were compared to the acute enteric fever samples during the identification process to find matches between both putatively identified and unidentified metabolites . The metabolites were then compared by significance and direction of change to highlight both similarities and differences between acute enteric fever and chronic carriage . Plasma samples from patients with confirmed S . Typhi ( n = 12 ) and S . Paratyphi A ( n = 5 ) gallbladder carriage together with control samples without gallbladder carriage of any bacteria ( n = 20 ) were analyzed by GCxGC-TOFMS ( S1 File and S1 Table ) . This resulted in 691 detected peaks and after further investigation 195 putative metabolites were selected for downstream analysis . Exclusion was mainly associated with split peaks , low quality peaks or mass spectra , peaks with a high correlation with run order , and deviation in quality control samples . Of the selected peaks 69/195 ( 35 . 4% ) had a putative annotation , 8/195 ( 4 . 1% ) had an assigned metabolite class , 18/195 ( 9 . 2% ) were of uncertain identity , and 100/195 ( 51 . 3% ) were of unknown identity ( S2 Table ) . Examining the raw processed data by principal component analysis ( PCA ) further revealed ten putative metabolites that had extremely high concentration levels in single samples , these metabolite concentrations were replaced by missing values to avoid misclassification . To identify metabolite profiles that may be associated with Salmonella carriage we used unsupervised multivariate modeling and generated a primary PCA model of the 195 metabolites in the 37 plasma samples . The model indicated a direction of separation of the non-carriage controls from the S . Typhi and S . Paratyphi A carriage samples; this separation was chiefly along the second principal component ( S1A Fig and Table 1 ) . We next applied supervised modeling through OPLS-DA to further investigate the strength and details of the trend observed in the PCA model . A three-class OPLS-DA model of S . Typhi carriers , S . Paratyphi A carriers , and non-carriage controls was fitted to generate an overview of the relation between the sample classes . The plasma samples from the Salmonella carriers were clearly segregated from the non-carriage samples along the first component whilst the S . Paratyphi A carriage samples were separated from the S . Typhi carriage samples along the second component ( p = 0 . 0031 ) ( S1B Fig and Table 1 ) . For a more detailed investigation of the metabolite profiles separating the carriage samples from the non-carriage samples a two-class OPLS-DA model was generated with S . Typhi and S . Paratyphi A carriage samples combined in one class and non-carriage controls in the other class . The metabolite profiles from the Salmonella carriage samples were significantly divergent from the profiles of the non-carriage samples ( p = 2 . 810−6 ) ( Fig 1A and Table 1 ) . We next segregated metabolite profiles generated in the Salmonella carriage plasma samples by infecting organism and performed further pairwise OPLS-DA comparisons with the non-carriage controls . The models revealed that the S . Typhi carriage samples and the S . Paratyphi A carriage samples were clearly segregated from the non-carriage controls ( p = 3 . 110−5 and p = 3 . 710−4 , respectively ) ( Fig 1C and 1D and Table 1 ) , as indicated in the combined S . Typhi and S . Paratyphi A model . Further investigation of the metabolite distributions ( using the model covariance loadings for the first OPLS-DA component w[1] ) highlighted a shift of metabolites towards the non-carriage control group in the S . Typhi model ( Fig 1E ) ( i . e . more metabolites with a higher relative concentration in the non-carriage group than in the S . Typhi group ) . Notably , this pattern was not observed in the S . Paratyphi A carriage model , in which the metabolites were more evenly distributed , although the S . Paratyphi A group was comprised of few samples ( Fig 1F ) . The three-class OPLS-DA model , together with the different appearance of the model covariance loadings of the S . Typhi and S . Paratyphi A models , indicated potential differences between metabolite profiles in the S . Typhi carriage samples and S . Paratyphi A carriage samples . Therefore , we generated an OPLS-DA model comparing the metabolite profiles between the S . Typhi and S . Paratyphi A carriers . This analysis resulted in a weaker model for the identification of differences in metabolite profiles between the two Salmonella serovars during carriage ( p = 0 . 07 ) ( Fig 1B and Table 1 ) . To further investigate the diagnostic potential of the metabolite patterns and the individual metabolites ROC curves were constructed . ROC curves for the metabolite patterns based on the cross-validated scores from the pairwise OPLS-DA models are shown in Fig 2 . For all pairwise OPLS-DA models comparing Salmonella carriage samples and non-carriage controls ( with S . Typhi and S . Paratyphi A carriage samples combined or separated ) the ROC curves indicated strong diagnostic potential with AUC values ≥ 0 . 95 ( Fig 2A–2C ) . The weaker OPLS-DA model obtained when comparing S . Typhi and S . Paratyphi A carriage samples resulted in a ROC curve with an AUC value of 0 . 833 ( Fig 2D ) . Focusing only on the comparison between the combined Salmonella carriage samples and the non-carriage controls ROC curves were constructed for each of the 195 metabolites ( included in the OPLS-DA model ) based on the relative metabolite concentrations . AUC values from these ROC curves are listed in S2 Table . None of the individual metabolites had an AUC value as high as or higher than the AUC value for the metabolite pattern ( Fig 2A ) . In a prior study we investigated metabolites associated with acute enteric fever in patients with culture-confirmed acute S . Typhi infections , S . Paratyphi A infections , and afebrile controls [16] . Using these existing data , we compared the metabolite profiles related to acute enteric fever and chronic carriage; both conducted in the same setting in Nepal . As our aim was to segregate Salmonella carriers from non-carriers , we used models where S . Typhi and S . Paratyphi A infections were combined and compared to controls ( non-carriage controls from the current investigation and afebrile controls from the former investigation ) to identify significant metabolites differentiating between the sample groups . The resulting metabolite comparison is summarized in Fig 3A ( S3 Table ) . Investigating comparable metabolites between acute enteric fever and chronic carriage highlighted three metabolites that were increased in the S . Typhi/S . Paratyphi A group; seven were increased in the control group . However , there were more metabolites with different directions of change in acute infection and carriage with 16 metabolites increased in the S . Typhi/S . Paratyphi A group in acute infection . Notably , we observed a substantial number of metabolites that were significantly different in only acute infection or carriage respectively . Of particular interest were five metabolites ( glutaric acid , hexanoic acid , and three metabolites with unknown identity ) that were elevated in the S . Typhi/S . Paratyphi A group in the carriage samples only ( Fig 3B ) . These metabolites were potentially differential and may have utility for prospectively identifying Salmonella carriers . Using these five metabolites to calculate an OPLS-DA model for Salmonella carriage in comparison to the non-carriage samples resulted in a significant separation between the groups ( p = 1 . 4*10−7 ) ( Fig 3C and Table 1 ) . A ROC curve constructed using the cross-validated scores from this five metabolite OPLS-DA model indicated that the strong diagnostic potential for distinguishing between Salmonella carriers and non-carriage controls was maintained using only these five metabolites ( Fig 3D ) . This ROC curve had an AUC value ( 0 . 935 ) that was higher than any of the AUC values from ROC curves of the five individual metabolites ( S2 Fig and S2 Table ) . Typhoid carriage remains one of greatest enigmas in infectious disease research . Due to the imperceptible nature of carriers the mechanisms and precise epidemiological role of S . Typhi carriage in humans are poorly defined . However , carriers are thought to be essential for disease maintenance and may be important for generating new genotypes . Therefore , prospectively detecting carriers is a key objective in regional enteric fever elimination . Here we employed metabolomics and chemometric bioinformatics to analyze plasma samples from patients in Nepal undergoing cholecystectomy that were confirmed carriers of S . Typhi or S . Paratyphi A . Using this patient population in an endemic enteric fever area we were able to generate metabolomic biomarker profiles indicative of Salmonella carriage . The descriptive nature of the study makes it difficult to reveal what the detected metabolite profiles physiologically reflect . However , notable patterns with diagnostic potential were identified and some of these aspects will be discussed . By exploiting specific metabolite profiles we were able to separate Salmonella carriers from non-carriers in a multivariate model . Further , segregating the carriage group into S . Typhi carriers and S . Paratyphi A carriers and comparing them independently to non-carriage controls generated an equivalent separation of the two carriage groups from non-carriage controls . Notably , metabolite profiles , as appose to individual metabolites , showed greater diagnostic potential ( in terms of AUC values for ROC curves ) . In addition , differences in the distribution of the metabolites between the two models were highlighted . We observed that the majority of the metabolites had a lower relative concentration in the S . Typhi carriage samples than the non-carriage controls . We are uncertain of the precise reason for a relative reduction in metabolite concentrations in the carriers compared to the controls , but speculate that this is evidence of S . Typhi controlling the inflammatory response . S . Typhi is a covert pathogen that has the ability to hide from the immune system and regulate inflammation during infection [26] . The Vi capsule polysaccharide , which protects outer membrane proteins from immunological interactions , likely controls much of this immune regulation [27 , 28] . During chronic carriage S . Typhi is within a protected , immune-privileged environment inside the gallbladder , and provoking an inflammatory response is incompatible with long-term persistence . Our data are consistent with S . Typhi manipulating systemic inflammatory responses during carriage , indicating that S . Typhi is perfectly adapted for long-term persistence in the gallbladder . Furthermore , S . Paratyphi A does not express the Vi polysaccharide and does not , therefore , regulate the inflammatory response in the same manner; the trend of increased metabolite concentrations in the non-carriage controls was not observed in the S . Paratyphi A model . Our data additionally indicated differences in metabolite profiles between S . Typhi and S . Paratyphi A carriers , but these differences were weaker than those observed between carriers and non-carriage controls . Although the key aim of this study was to differentiate carriers from non-carriers , distinguishing S . Typhi carriers from S . Paratyphi A carriers may be relevant in disease epidemiology and for studying the carriage mechanisms of these two organisms . There is a need for improved diagnostic methods for acute enteric fever and chronic carriage; therefore we additionally compared the metabolite profiles between chronic carriers and acute enteric fever cases . We identified a greater number of differences than similarities in metabolite profiles between the acute infection and chronic carriage . Among the differences there were metabolites with a different direction of change in both conditions and also metabolites found significant in one of the two disease stages only . In identifying diagnostic biomarkers with the potential for prospectively detecting S . Typhi and S . Paratyphi A carriers , the most suitable candidates are metabolites with a higher concentration in carriers than non-carriers ( more appropriate for point-of-care test development ) , and not relevant during acute infection . We identified five metabolites , including glutaric acid and hexanoic acid , which fulfilled these criteria . The profile of these five combined metabolites exhibited a stronger diagnostic potential in comparison to the five individual metabolites , highlighting the utility of using a combination of metabolite markers . We are uncertain of the relevance of these carboxylic acids , but these chemicals have antibacterial properties [29] and may be the result of altered gut microbiome substrate fermentation . Short-chain fatty acids , including hexanoic acid , are major fermentation products of the gut microbiome and concentrations can fluctuate by a shift in the composition of the gut microbiome [30–32] . Similarly , the same phenomenon can induce an increased catabolism of amino acids , resulting in an increase in the concentration of glutaric acid . It is speculative , but Salmonella spp . have been shown to alter the gut microbiome in murine models of infection [33 , 34] . The increase in hexanoic acid and glutaric acid observed during Salmonella carriage was not observed during acute enteric fever , suggesting that the release of organisms from the gallbladder into the duodenum may impact on the gastrointestinal microbiota . However , whether this discrepancy is a result of Salmonella associated alterations of the gut microbiome arising during the two disease stages or due to other metabolic processes remains unclear . The major limitation of this study was the small sample size , which was difficult to overcome given the ethical , surgical , and diagnostic issues of identifying and confirming invasive Salmonella carriers . We suggest that resulting metabolite profiles should be validated in an additional , independent cohort . However , this was not possible in this setting due to the factors stated above . Furthermore , as all patients included in this study presented with gallbladder conditions we cannot rule out that the primary cause for the cholecystectomy may impact on the metabolite signatures . However , the surgical conditions were comparable for both carriers and non-carriers and , as there is a known association between Salmonella carriage and cholecystitis [35] there is significant value in investigating carriage in this patient cohort . The non-carrier controls included in this study were culture-negative , meaning that there was no growth of any bacteria in their bile . It would be of additional interest to include controls with gallbladder carriage of other bacteria in this comparison to be able to demonstrate that these metabolite are truly Salmonella-specific . Further studies in alternative endemic enteric fever populations are required to investigate the metabolite markers associated with S . Typhi and S . Paratyphi A carriage . If a set of metabolite markers for Salmonella carriage passes a further rigorous validation the next challenge is to convert this diagnostic panel into an accurate and inexpensive test suitable for use in resource-limited areas . An important feature of such a test is simultaneous detection of multiple biomarkers . A recent review summarizes current advantages in the field of multiplexed point-of-care testing[36] . In general , differing microfluidic techniques have great potential for such multiplexing , although many challenges remain . One promising example shows the measurement of three metabolites in human serum using microfluidic paper-based analytical devices[37] . In conclusion , our novel approach highlights the potential of using metabolomics to search for diagnostic markers of chronic Salmonella carriage . We identified metabolite patterns signifying carriage of S . Typhi and S . Paratyphi A in the gallbladder among a cohort of patients with cholelithiasis in Nepal . These findings are encouraging in the search for a diagnostic assay that may be able to access the reservoirs of S . Typhi and S . Paratyphi A carried asymptomatically within human populations .	Enteric fever , caused by typhoidal Salmonella serovars , remains a substantial public health problem in many low- and middle-income countries . The human-restricted nature of these organisms combined with the development of new vaccines suggests that regional elimination of enteric fever should be possible . However , individuals that chronically carry Salmonella in their gallbladder , such as the notorious Typhoid Mary , complicates enteric fever transmission and maintain circulation of the organisms . The prospective detection of chronic Salmonella carriers is therefore a critical step for regional enteric fever elimination . However , there are currently no diagnostic methods routinely in use for this purpose . Here , we used a novel method for identifying chronic Salmonella carriers by comparing metabolite patterns in plasma samples from patients with chronic Salmonella carriage against non-carriage controls . We could significantly distinguish Salmonella carriers from non-carriers based on a large set of metabolites . Five metabolites were then highlighted , after comparing metabolite patterns obtained during chronic Salmonella carriage and acute enteric fever respectively , which could significantly distinguish Salmonella carriers from non-carriers . These potential biomarkers require further evaluation in epidemiological investigations of enteric fever in alternative endemic settings but this study provides a first step towards improved detection of Salmonella carriers .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "biliary", "system", "medicine", "and", "health", "sciences", "liver", "pathology", "and", "laboratory", "medicine", "gallbladder", "pathogens", "microbiology", "biomarkers", "salmonella", "typhi", "surgical", "and", "invasive", "medical", "procedures", "multivariate", ...	2018	Diagnostic metabolite biomarkers of chronic typhoid carriage
The main constraint to the fight against container-breeding mosquito vectors of human arboviruses is the difficulty in targeting the multiplicity of larval sources , mostly represented by small man-made water containers . The aim of this work is to assess the feasibility of the “auto-dissemination” approach , already tested for Aedes aegypti , as a possible alternative to traditional , inefficient control tools , against Ae . albopictus in urban areas . The approach is based on the possibility that wild adult females , exposed to artificial resting sites contaminated with pyriproxyfen , can disseminate this juvenile hormone analogue to larval habitats , thus interfering with adult emergence . We carried out four field experiments in two areas of Rome that are typically highly infested with Ae . albopictus , i . e . the main cemetery and a small green area within a highly urbanised neighbourhood . In each area we used 10 pyriproxyfen “dissemination” stations , 10 “sentinel” sites and 10 covered , control sites . The sentinel and control sites each contained 25 Ae . albopictus larvae . These were monitored for development and adult emergence . When a 5% pyriproxyfen powder was used to contaminate the dissemination sites , we observed significantly higher mortality at the pupal stage in the sentinel sites ( 50–70% ) than in the controls ( <2% ) , showing that pyriproxyfen was transferred by mosquitoes into sentinel sites and that it had a lethal effect . The results support the potential feasibility of the auto-dissemination approach to control Ae . albopictus in urban areas . Further studies will be carried out to optimize the method and provide an effective tool to reduce the biting nuisance caused by this aggressive species and the transmission risk of diseases such as Dengue and Chikungunya . These arboviruses pose an increasing threat in Europe as Ae . albopictus expands its range . Aedes albopictus ( Skuse , 1894 ) is native to Southeast Asia . In recent decades this mosquito has invaded and efficiently colonized temperate areas of the US and Europe , thanks to passive transportation of eggs in used tires and the ability to produce diapausing eggs [1] , [2] . In its tropical range the species is a secondary vector of human arboviruses , such as Dengue , while in temperate regions its impact on public health is mostly due to its aggressive and diurnal biting behavior . However , in recent years , Ae . albopictus has been the sole vector of large epidemics of Chikungunya virus in La Reunion , France [3] , [4] and Kerala , India [5] . In August 2007 the species was responsible for a Chikungunya virus outbreak in the Province of Ravenna in north-eastern Italy where more than 250 human cases were confirmed [6] and where the mosquito has repeatedly been identified as infected by Chikungunya [7] , [6] , [8] and Usutu viruses [9] . In 2010 Ae . albopictus was responsible for 2 and 31 cases of endemic transmission of Dengue in France and China , [10] , [11] . The report of a human blood index ≥80% in Rome raises particular concern for the potential of this species to vector pathogens in urban areas where humans represent the major blood-meal source [12] . In Italy , Aedes albopictus was detected for the first time in Genoa in 1990 and the next year in Padua [13] , [14] . In the following years it spread to 19 out of 20 Italian regions and over 82 out of 107 Italian provinces [15] . It was first detected in Rome in 1997 [16] and since then it has colonized the whole urban area through three phases “a first massive spread , a following maintenance of infestation , and the colonization of alternative winter breeding sites with favourable climatic conditions” [17] . The successful invasion of Rome and of other urban areas in Italy and worldwide is driven by the ability of this mosquito to exploit a large variety of water containers as larval breeding sites , including the catch basins of storm drains , used tires , domestic containers , vases , etc . The difficulty in identifying and treating all these sources makes control extremely difficult . A survey carried out in Rome's zoological gardens , which estimated the larval density of mosquitoes breeding in catch basins , showed that a large number of productive alternative larval biotopes exist [18] . Despite this , the most commonly utilized strategy to reduce Ae . albopictus densities in Italy is the treatment of catch basins with larvicidal compounds [19] . This only targets an unknown percentage of the overall aquatic habitat . Indoor insecticidal treatments and house screening are not suitable due to the largely exophilic behavior of the species , and outdoor adulticidal treatments are recommended only in case of emergencies ( e . g . virus outbreaks or very high nuisance in sensible locations such as outdoor recreational areas ) due to their environmental impact and their low cost-efficiency ratio [19] . The aim of this work is to assess the feasibility of a new approach for the control of Ae . albopictus in urban areas , inspired by results obtained on the tropical Dengue vector , Ae . aegypti , in Thailand [20] and in Peru [21] . This approach , named “auto-dissemination” , is based on the possibility that wild adult mosquitoes exposed to artificial resting sites contaminated with pyriproxyfen ( PPF , a juvenile hormone analogue ) , can disseminate insecticide to larval breeding sites , thus preventing adult emergence . This strategy is facilitated by the oviposition behaviour of both Ae . aegypti and Ae . albopictus , that typically scatter the eggs from a single gonothrophic cycle among several temporary sites . This increases the probability of at least some larvae reaching the adult stage [22] . Second , extraordinarily low doses of PPF ( Ae . aegypti: LC50 = 0 . 011 p . p . b . , [23] , 0 . 0039 p . p . b . [24]; Ae . albopictus: LC50 = 0 . 11 p . p . b . [25] , are needed to interfere with the metamorphosis of juvenile stages [26] , and/or to cause morphological and functional aberrations in emerging adults , such as decreased fertility in males and females [27] , [28] . Third , evidence from laboratory experiments shows that females either forced to walk on PPF-treated paper or topically contaminated can contaminate larval sites and significantly inhibit adult emergence [29] , [30] . Thus , the “auto-dissemination” approach can be proposed as a ‘pull’ ( i . e attraction of wild mosquitoes to PPF-contaminated sites for contamination ) and ‘push’ ( i . e . dispersal of contaminated mosquitoes and dissemination of PPF to larval habitats ) control strategy with the potential to target the myriad of cryptic larval breeding sites that cannot be reached by traditional larvicidal applications . We here present the results of four “auto-dissemination” experiments carried out in two areas of Rome with high infestations of Ae . albopictus: a cemetery and an enclosed garden . Dissemination stations ( DS , Figure 1 ) were adapted from modified sticky traps ( ST ) , previously developed by our group [31] . The four sticky surfaces were replaced by four 12×8 cm black cotton cloths ( Figure 1b ) . A thick net was placed over the water to prevent mosquitoes from ovipositing ( Figure 1c ) . Before each experiment , each DS was filled with 700 ml of tap water and each cloth was dusted with 1 g of powdered PPF . This was obtained by manually grinding 0 . 5% or 5% PPF tablets ( Proxilar , I . N . D . I . A . Industrie Chimiche S . p . A . ) to an granule average size of 40–80 micron . Larval/pupal mortality in sentinel sites ( SS ) that were potentially contaminated with PPF by wild mosquitoes was compared to mortality in uncontaminated control sites ( CS ) . Sentinel sites were assembled by inserting a 600 ml Pyrex beaker into a standard ovitrap ( i . e . a black vase ) . This was done to allow easy decontamination after the experiments ( PPF adheres to plastics , but not to glass ) . The border of the beaker was covered with black tape to avoid reflections . Control sites were similar to SS but closed with white nets to prevent mosquitoes from entering the beakers and transferring PPF . Each SS and CS contained 200 ml of tap water , 0 . 07 g of cat biscuits and 25 third-instar Ae albopictus larvae . These were obtained from eggs collected from the “Sapienza” University campus by ovitraps during the weeks before the experiments . After hatching , the larvae were reared outdoors on the terrace of the Department of Public Health and Infectious Diseases of “Sapienza” University at a density of about 1 larva/ml . The experiments were carried out in the following two sites located in a highly urbanized area of Rome ( Italy ) adjacentto the University “Sapienza” campus . During the experiments , SS and CS were monitored every two days . For each larval cohort , we derived cumulative totals of dead larvae , dead pupae and emerged adults , as follows: i ) live larvae were left to develop further; ii ) dead larvae and pupae were counted and removed ( Figure 2b ) ; iii ) live pupae were counted and transferred by pipette to a separate disposable cup containing uncontaminated water; these cups were covered with netting and maintained under semi-natural conditions until adult emergence or pupal death; iv ) the water level in the SS and CS was maintained at 200 ml volume; v ) temperature was recorded using one data logger for each corridor in Site 1 , and by individual data loggers located close to each SS in Site 2 . Monitoring continued until all the original 25 larvae were either dead or emerged ( 7–12 days ) . Mortality in SS and CS was analysed by nonparametric alternatives to the t-test , since the distribution was not normal . We used the Kolmogorov-Smirnov ( K-S ) two-sample test ( two-tailed ) , which is powerful even when the distributions differ in terms of dispersion We used Wilcoxon signed-rank tests ( two-tailed ) to compare the mortality in SS between experiments within same field site: this is a non-parametric paired difference test that compares two related samples or repeated measures on a single sample . Two mixed-effect logistic regression models ( one per study area ) were applied to quantify PPF-related mortality . Each model includes , as independent variables , the number of replicates ( “replicate effect” ) , the treatment ( i . e . potential contamination in SS by wild mosquitoes vs . no contamination in CS ) and the interaction between the replicates and the treatments , which allows an evaluation of the effect on the mortality in the sentinel sites in each replicate . The code identifying each SS and each CS was included as random effect to correct for the repeated measures within sites ( “site-effect” ) . Table 1 shows the overall mortality observed in SS and CS in the four experiments . Overall mortality in SS was ≥50% in the 3 experiments carried out with 5% PPF concentration , and 20 . 8% in the first experiment in Site 1 , with 0 . 5% concentration . In all experiments , >90% of the mortality recorded in SS occurred at the pupal stage . In Site 2 however , about 13% of deaths occurred during adult emergence . This has been frequently noted in insects treated with PPF [32] , but never in mosquitoes . The overall mortality in CS was ≤2 . 5% . In Site 1 , a single experiment ( 1 . 1 ) utilised the 0 . 5% PPF formulation and a subsequent experiment utilised the 5% formulation ( 1 . 2 ) . No significant differences in mortality in SS vs . CS were observed using the 0 . 5% concentration ( K-S , p = 0 . 164; Figure 2 ) . However , at 5% concentration mortality in SS was higher than that observed in CS ( K-S , p<<0 . 001; Figure 2 ) . Mortality in SS was lower in Exp . 1 . 1 ( median = 2 , Q1 = 1 , Q3 = 8 ) than in Exp 1 . 2 ( median = 18 , Q1 = 13 , Q3 = 23 . 75 ) ( significant difference shown by Wilcoxon test , p = 0 . 004 ) . In Site 2 , both experiments ( 2 . 1 and 2 . 2 ) utilised the 5% concentration . The difference in mortality in SS vs . CS was significant only in Exp . 2 . 2 ( K-S , p = 0 . 015; Figure 3 ) , although the trend was confirmed in Exp . 2 . 1 ( K-S , p = 0 . 055; Figure 3 ) . No difference in mortality in SS was observed between the two replicates ( Wilcoxon test , n . s . ; Exp 2 . 1: median = 13 . 5 , Q1 = 0 . 25 , Q3 = 23 . 75; Exp . 2 . 2: median = 13 , Q1 = 2 . 25 , Q3 = 24 . 5 ) . The results of the mixed-effect logistic regression model showed that mortality was always significantly higher in SS than in CS . In Site 1 , mortality was 9- and 66 . 5-fold higher in SS than in CS- in Exp . 1 . 1 and in Exp 1 . 2 , respectively , although mortality observed in SS was approximately 3 times lower in Exp 1 . 1 than in Exp 1 . 2 ( Table 2 ) . In Site 2 , mortality was 49- and 37-fold higher in SS than in CS in Exp . 2 . 1 and Exp 2 . 2 , respectively ( Table 3 ) . Mortality was similar in the two replicates and no significant differences were found in the interaction between “replicate” and “treatment” . Figure 4 shows mortality in SS in the four experiments . In Exp 1 . 1 mortality was concentrated in 3 out of 10 SS , while in Exp 1 . 2 it ranged between 40 and 100% . In both Site 2 experiments mortality >76% was observed in 5 out of 10 SS , 4 of which were located in the same position in both replicates . A number of experiments carried out under laboratory conditions have shown that adult Ae . albopictus topically contaminated with PPF can transfer enough material to water containers to exert a significant lethal effect on pre-imaginal stages developing therein [23] , [20] , [29] , [30] . For the first time we tested the hypothesis that wild Ae . albopictus adults can act as “auto-disseminators” of PPF and inhibit adult emergence from sentinel sites . Although it was not possible to measure PPF concentration in those sites ( the low doses of the product are beneath the limits of detection for any published method ) , the evidence supports the working hypothesis and suggests that auto-dissemination could represent a valid , novel approach for reducing Ae . albopictus densities in urban temperate areas . We observed significantly higher mortality in our sentinel sites than in the controls . This shows that PPF was transferred by mosquitoes into sentinel sites and elicited a lethal effect . Mortality was almost exclusively limited to the pupal stage , i . e . the stage on which PPF is known to have its major effect [26] . Note that control sites were exposed to exactly the same experimental conditions as sentinel sites , except for the fact that contact with potentially contaminated mosquitoes was prevented by a net cover in the controls . Mortality was not uniformly distributed among sentinel sites . This strongly suggests that PPF contamination occurred at some sites , but not others - presumably as a result of differences in the frequency of visits made by contaminated mosquitoes . This is particularly evident in Site 2 , where mortality was observed in 5 out of 10 SSs , 4 of which were located at the same position in both replicates ( Figures 3 and 4 ) , leading to hypothesize that these sites were more attractive for the mosquitoes than the remaining ones . The same applies for the first replicate in Site 1 . In the second replicate , however , mortality ranging from 40 to 100% was observed in all SSs suggesting that these were all visited and contaminated but with differing frequency . In the cemetery experiments , a 10-fold increase in PPF concentration resulted in a 3-fold increase in mortality ( Table 2 ) . Median mortality in sentinel-sites , in the three experiments carried out with a 5% PPF concentration , was higher in Site 1 than in Site 2 ( 18 and 13 deaths/25 larvae , respectively ) . In the mixed-effect logistic regression model , the variance due to the “site effect” was 0 . 12 ( Standard deviation = 0 . 35 ) and 1 . 74 ( Standard Deviation = 1 . 32 ) , respectively . This is the likely consequence of the fact that , as mentioned above , in the Site 2 most mortality was observed only in 50% of the sentinel sites , while in Site 1 40–100% mortality was observed in every sentinel site . This presumably reflects the different ecology of the two experimental areas . Site 1 , being an underground corridor , is ecologically very homogeneous and it is reasonable to hypothesize that all sentinel sites were equally attractive to the flying wild mosquitoes . The enclosed garden in Site 2 was far more heterogeneous and sentinel sites were located outdoors in a wider area compared to Site 1 . It is reasonable to hypothesise that under these conditions , sentinel sites vary in attractiveness depending on contrast against background , shade , humidity , etc . However , even under such heterogeneous conditions the overall mortality in sentinel sites was about 37–49 fold higher than in controls . It is also important to emphasize that the experiments were not designed with the aim of identifying environmental factors that might optimise impact , although the effect of sun exposure has already been hypothesized in auto-dissemination experiments carried out against Ae . aegypti in Peru [21] . In fact , it is relevant to highlight that the major/minor attractiveness of sentinel sites to mosquitoes is likely to affect the results of the experiments , but has a relatively low practical relevance . In fact , if the approach is applied to reduce Ae . albopictus adult densities , the mosquitoes themselves will disseminate PPF to the most attractive ( i . e . most productive ) natural breeding sites . Further studies are required to assess the ideal conditions for the location of dissemination stations , as well as to evaluate the persistence of the lethal effect under different environmental ( e . g . temperatures , sun exposures ) and ecological ( e . g . air flow , presence of animals , of humans , other disturbances/attractions ) variables . In the only other auto-dissemination experiment ever carried out in the field outdoors ( i . e . in Peru with a 0 . 5% PPF concentration [21] ) , the overall reductions in Ae . aegypti adult emergence was 49–84% , as opposed to a 7–8% mortality in controls . In the single experiment carried out in Rome with the same PPF concentration , the overall reduction in Ae . albopictus adult emergence was 20 . 8% , as opposed to a 2 . 4% mortality in controls . This is quite encouraging: in fact a lower overall effect could have been expected in our experiments , since PPF LC50 reported for Ae . albopictus ( 0 , 11 ppb [25] ) is about 10 times higher than that reported for Ae . aegypti ( 0 . 011 ppb [23]; 0 . 0039 ppb [24] ) . It is also worth noting that in Rome control sites were monitored concurrently to sentinel sites and were located at the same distance from dissemination stations , while in Peru they were separated in time . Thus , our results rule out the hypothesis that PPF could be passively transported by wind . If this had occurred , we would have expected higher mortality in at least some control sites . Overall , the results from these small-scale experiments carried out in Rome strongly encourage further studies to evaluate the feasibility of the exploitation of the auto-dissemination approach to control Ae . albopictus densities in urban areas . In fact , the experimental design probably underestimates the overall impact of the approach , as other known effects of PPF on mosquitoes – such as sterilizing effects on adult females [28] and a decrease in male spermiogenesis [27] - were not taken into consideration . Moreover , the effect of auto-disseminated PPF was only monitored on third instar larvae ( or later stages ) , while in natural breeding sites younger larvae will also be contaminated , presumably increasing the overall impact . The auto-dissemination approach has the potential to effectively counter the main challenge to conventional larviciding approaches by effectively targeting the myriad of cryptic breeding sites that these mosquitoes utilise . Based on results from mark-release-recapture experiments carried out in the campus of Sapienza University in Rome [33] , it is possible to hypothesize that gravid , PPF-contaminated Ae . albopictus females could contaminate breeding sites in a 200-m radius area around a dissemination station . Other relevant advantages of the auto-dissemination approach are: i ) a higher residual activity of PPF ( 4 months in water [34] ) compared to that of other compounds commonly used for larval control and ii ) no risk for human health , due to the low-toxicity of the product for vertebrates and the high sensitivity of mosquito larvae/pupae [35] . With regard to this latter point , although PPF is effective against many insects , the proposed approach targets container-breeding species with such tiny amounts of compound , disseminated exclusively to their breeding sites , that impacts on non-target species are likely to be minimal . Finally , the auto-dissemination approach could be a very cost-effective control tool . Once deployed , the dissemination station does not require frequent maintenance , nor frequent toxicant applications , thanks to PPF's outstanding persistence [28] . It is also of note that other Culicidae species ( e . g . Culex pipiens , a common nuisance species in urban Italy , sharing some Ae . albopictus breeding sites ) may contribute to the dissemination of the product [36] and be affected by it themselves . Indeed , larger-scale field experiments are needed to evaluate more precisely the worldwide feasibility of the approach and to promote its use against Ae . albopictus . With appropriate modifications to the dissemination stations , auto-dissemination tools may be simple enough to be deployed by members of the public . In the experiments carried out in Rome , PPF dissemination stations were adapted from existing sticky-traps for collecting Ae . albopictus and Ae . aegypti females [31] , [37] . The sticky surfaces were replaced with cloth surfaces dusted with pulverized PPF . In Peru , dissemination stations consisted of 1-liter plastic disposable pots lined with damp black cloth , These dissemination tools could be optimized to increase their attractiveness for mosquitoes and their overall practicability . The fabric of the cloths could be optimized for the release of PPF to mosquito legs , and the method of application of the compound into the cloth could be optimized and standardized . Other methods to attract and contaminate adult mosquitoes could be developed ( see , for instance , [30] ) . Finally , the auto-dissemination approach could be exploited to spread other mosquito toxic compounds , such as other juvenile hormones or fungi , which might have an even greater impact than PPF [38] , [39] , [40] , [41] .	Aedes albopictus ( the Asian Tiger mosquito ) is one of the most invasive and aggressive disease vectors in the world . It is a serious public nuisance and a public health risk , due to its ability to transmit pathogens to humans . The control of this mosquito is complicated by the difficulty in targeting either juveniles ( due to their ability to colonise myriad and often cryptic domestic and natural water containers ) or adults ( due to their diurnal and outdoor resting and biting patterns ) . The aim of our work was to assess the feasibility of a novel approach to control Ae . albopictus in urban areas . This was based on the possibility that wild adult females , exposed to artificial resting sites contaminated with an insect growth regulator , could subsequently contaminate their breeding sites and kill both the larvae and pupae developing therein . The results obtained from field experiments carried out in highly infested areas in Rome demonstrate that this approach has potential as a valid alternative to traditional ( and ineffective ) larval control efforts in urban areas . We therefore introduce a new perspective in the fight against the Asian Tiger Mosquito .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "medicine", "infectious", "diseases", "public", "health", "and", "epidemiology", "ecology", "biology", "public", "health" ]	2012	The “Auto-Dissemination” Approach: A Novel Concept to Fight Aedes albopictus in Urban Areas
Effective T cell responses can decisively influence the outcome of retroviral infection . However , what constitutes protective T cell responses or determines the ability of the host to mount such responses is incompletely understood . Here we studied the requirements for development and induction of CD4+ T cells that were essential for immunity to Friend virus ( FV ) infection of mice , according to their TCR avidity for an FV-derived epitope . We showed that a self peptide , encoded by an endogenous retrovirus , negatively selected a significant fraction of polyclonal FV-specific CD4+ T cells and diminished the response to FV infection . Surprisingly , however , CD4+ T cell-mediated antiviral activity was fully preserved . Detailed repertoire analysis revealed that clones with low avidity for FV-derived peptides were more cross-reactive with self peptides and were consequently preferentially deleted . Negative selection of low-avidity FV-reactive CD4+ T cells was responsible for the dominance of high-avidity clones in the response to FV infection , suggesting that protection against the primary infecting virus was mediated exclusively by high-avidity CD4+ T cells . Thus , although negative selection reduced the size and cross-reactivity of the available FV-reactive naïve CD4+ T cell repertoire , it increased the overall avidity of the repertoire that responded to infection . These findings demonstrate that self proteins expressed by replication-defective endogenous retroviruses can heavily influence the formation of the TCR repertoire reactive with exogenous retroviruses and determine the avidity of the response to retroviral infection . Given the overabundance of endogenous retroviruses in the human genome , these findings also suggest that endogenous retroviral proteins , presented by products of highly polymorphic HLA alleles , may shape the human TCR repertoire that reacts with exogenous retroviruses or other infecting pathogens , leading to interindividual heterogeneity . Adaptive immunity to viral infection relies on appropriate activation of T cells by complexes of viral peptides with MHC molecules . The host MHC haplotype , the nature of the viral peptide targeted and the T cell receptor ( TCR ) repertoire of responding T cells are three interconnected parameters that play a decisive role in the outcome of infection . Indeed , the MHC is the predominant genetic factor affecting susceptibility to many infectious diseases [1]–[4] . For example , the HLA locus shows the strongest genetic association with control of HIV infection , with certain HLA alleles having been consistently found to confer a protective advantage [3] , [5] , [6] . Although the precise underlying mechanism is not completely understood , T cell responses restricted by protective HLA/MHC alleles often comprise narrower TCR repertoires , dominated by public TCR sequences , and exhibit higher magnitude , avidity or depth , and thus greater contribution to HIV or SIV control , than those restricted by non-protective HLA/MHC alleles [7]–[9] . Bias in the use of certain TCRVα ( Vα ) or TCRVβ ( Vβ ) chains has been frequently observed in the T cell response to several antigenic epitopes , and public T cell responses with identical or similar TCRs have been found to dominate the response of different individuals to a given epitope . Skewed TCR usage has often correlated with higher functional avidity to a given antigenic epitope , and , in diverse systems , also translated into more efficient protection against infection [10]–[12] . Despite the potential importance in cellular immunity to infection , however , the mechanisms by which TCR biases ( and particularly high-avidity T cell responses to viral infections ) are generated and maintained remains incompletely understood . The mechanisms leading to bias in the T cell response will vary considerably depending on the antigenic peptide and MHC combination . TCR repertoire bias can be generated during thymic selection , leaving only certain Vα or Vβ chains able to respond to a given antigen [10] . It can also be generated at the initiation of the immune response , where clones using particular Vα or Vβ chains may have a recruitment or proliferative advantage and can quickly dominate the response [10] . Lastly , bias can also be generated during chronic viral infection either due to preferential maintenance of certain T cell clones or differential margins for cross-reactivity with viral escape mutations [10] or by prior or concurrent infection with heterologous viruses , sharing cross-reactive epitopes [13] . We have previously described the TCRβ-transgenic strain EF4 . 1 , which generates increased frequencies of CD4+ T cells reactive with the H2-Ab-restricted env122-141 epitope of Friend murine leukemia virus ( F-MLV ) [14] . Virus-specific EF4 . 1 CD4+ T cells show bias in the use of endogenous Vα2 chains in their response to infection with Friend virus ( FV ) , a retroviral complex of F-MLV and spleen focus-forming virus ( SFFV ) [14] , [15] . Use of Vα2 chains by virus-specific CD4+ T cells creates higher avidity for the same epitope than use of other Vα chains , and although they represent a minority in the naïve repertoire , high-avidity Vα2 T cells become the dominant subset at the peak of the response [15] . Here we have examined the potential mechanisms underlying the formation of TCR repertoire diversity in this system , which might be responsible for the high-avidity response to FV infection . We have found that a thymic selection event was necessary for the dominance of Vα2 virus-specific CD4+ T cells during the response to FV infection . Selection of virus-specific CD4+ T cells was mediated by a self peptide encoded by an endogenous retrovirus with substantial similarity to F-MLV . Unexpectedly , despite deleting a sizeable fraction of virus-specific CD4+ T cells , negative selection by this endogenous retrovirus was necessary for a predominantly high-avidity response to FV infection . On average , 4% of EF4 . 1 CD4+ T cells in virus-naïve mice react with the env122-141 peptide , of which approximately 25% stain positive with the anti-Vα2 monoclonal antibody B20 . 1 [14] , [15] . Vα2 env-specific CD4+ T cells were previously [14] , [15] found to be >30-fold more sensitive than non-Vα2 T cells to stimulation with a 20-mer env122-141 peptide spanning the core env128-134 epitope [16] . This higher avidity of Vα2 CD4+ T cells was not due to recognition of the core epitope-flanking residues by this family of Vα chains , as has been described for other TCR – epitope combinations [17] , since it was maintained against a series of N-terminal truncated peptide epitopes ( Figure S1A ) . Thus , Vα2 CD4+ T cells would recognize with higher avidity all the nested peptides of variable lengths likely to be generated during in vivo processing of env [18] . To examine whether the polyclonal Vα2 CD4+ T cell population displayed higher affinity for F-MLV env-derived epitopes even at the clonal level , we generated hybridoma cell lines from primary EF4 . 1 CD4+ T cells stimulated in vitro with either a low ( 10−7 M ) or a high ( 10−5 M ) peptide dose . In agreement with our previous findings [14] , [15] , 71% ( 20/28 ) and 30% ( 9/30 ) of hybridoma cell lines derived from CD4+ T cells stimulated with the low or high peptide dose , respectively , were Vα2+ . Similarly to primary EF4 . 1 CD4+ T cells , randomly selected Vα2 T cell hybridomas were more sensitive to stimulation with all the peptides tested than non-Vα2 ones , irrespective of whether a high or low peptide dose was used for their generation ( Figure S1B ) . Thus , the higher avidity of Vα2 CD4+ T cells for F-MLV env-derived epitopes was also observed at the level of individual clones . To assess whether low-avidity F-MLV env-reactive CD4+ T cells were characterized by expression of any particular family of endogenous Vα chains , we screened env122-141-specific non-Vα2 CD4+ T cells for expression of Trav transcripts encoding different Vα families . Although this analysis indicated enrichment for Trav9 expression ( encoding Vα3 ) , only a small percentage of env122-141-reactive non-Vα2 CD4+ T cells stained positive with the anti-Vα3 . 2 monoclonal antibody RR3-16 ( unpublished data ) , and only 2 out of 4 F-MLV env-reactive non-Vα2 T cell hybridomas were positive for Vα3 . 2 ( Table S1 ) . However , Vα3 . 2 is used preferentially in CD8+ T cells in B6 mice , whereas the other three of the four expressed Vα3 family members are preferentially expressed in CD4+ T cells [19] . It was therefore possible that env122-141-reactive non-Vα2 CD4+ T cells that did not stain positive with the RR3-16 antibody were also expressing Vα3 . Indeed , cloning and sequencing of expressed endogenous Trav genes from theses hybridomas revealed that they were all members of the Trav9 family ( Table S1 ) . Thus , similarly to selective usage of Vα2 chains in high-avidity cells , low-avidity F-MLV env122-141-reactive CD4+ T cells selectively used Vα3 chains . However , in the absence of a Vα3-specific antibody that can detect all Vα3 family members , these cells were referred to here as non-Vα2 cells . Lastly , we tested whether biased use of Vα2 chains also characterized the response of non-transgenic CD4+ T cells to F-MLV env . CD4+ T cells from wild-type ( wt ) C57BL/6 ( B6 ) mice 7 days post FV infection were stained with an env123-141-presenting tetramer ( Ab-env123-141 ) . In comparison with a control tetramer , staining with Ab-env123-141 tetramer identified a measurable population of env122-141-specific CD4+ T cells in all infected mice ( Figure 1A ) , in agreement with published data [20] , [21] . FV infection had no impact on the frequency of Vα2 cells in naïve ( CD44lo ) and total memory ( CD44hi ) CD4+ T cells ( 15% and 12% , respectively ) , with minimal variation between individual mice ( Figure 1B ) . In contrast , the frequency of Vα2 cells in Ab-env123-141 tetramer+ CD4+ T cells varied considerably between 4% and 23% . These results revealed substantial deviation in Vα2 usage in Ab-env123-141 tetramer+ CD4+ T cells from the same usage in total CD4+ T cells , but also indicated substantial heterogeneity . However , this particular tetramer is known to bind only some env122-141-specific T cell clones but not others [14] , [22] . Furthermore , at the peak of their response , env122-141-specific CD4+ T cell reversibly downregulate up to 70% of their surface TCR [15] , [23] , which could prevent tetramer binding . Indeed , combining adoptive transfer of EF4 . 1 CD4+ T cells and tetramer staining revealed that Ab-env123-141 tetramer staining was restricted to env122-141-reactive CD4+ T cells with above-average TCR levels , independently of Vα usage , and TCR re-expression improved tetramer staining ( Figure 1C–F ) . Thus , detection of env122-141-reactive CD4+ T cells by Ab-env123-141 tetramer staining was eclipsed by in vivo antigen-induced TCR downregulation . Collectively , these results both validated and necessitated the use of env122-141-reactive EF4 . 1 CD4+ T cells that can be indelibly identified , independently of Ab-env123-141 tetramer binding , to study the requirements for induction of a high-avidity CD4+ T cell response to F-MLV env . F-MLV env122-141-reactive CD4+ T cells in EF4 . 1 mice have a naïve phenotype [14] , and it was therefore likely that the F-MLV env122-141-reactive TCR repertoire and associated avidity differences were the result of thymic selection events . We searched the mouse proteome for the presence of self-derived epitopes with homology to F-MLV env122-141 . This approach identified the single-copy endogenous ecotropic MLV at the Emv2 locus [24] . Emv2 shares 80% homology with F-MLV at the DNA sequence level , and although it represents a full-length provirus , it is unable to produce infectious particles due to a single inactivating point-mutation in the pol gene [24] . Nevertheless , Emv2 has full potential for env expression , and , importantly , the env122-141 epitope is almost identical between Emv2 and F-MLV , with the exception of a Y instead of an L at position 128 ( Figure S2A ) . For this reason , Emv2 and F-MLV env-derived epitopes were referred to here as env122-141Y and env122-141L , respectively . Position 128 , together with 129 and 133 , have been previously mapped as important contact residues for the SB14-31 TCR ( Figure S2B ) , which was the donor of the TCRβ chain transgene used in EF4 . 1 mice [16] . We next investigated whether or not Emv2 could be involved in T cell selection . In vitro stimulation with the env124-138Y epitope activated a fraction of EF4 . 1 CD4+ T cells , which was however smaller than the fraction activated by the env124-138L epitope ( Figure 2A ) . As EF4 . 1 mice generate a polyclonal TCR repertoire , it was unclear whether the same CD4+ T cells could respond to both epitopes . However , analysis of env124-138L-reactive T cell hybridomas revealed the same TCR could be activated by both env124-138L and env124-138Y epitopes , albeit less potently by the latter peptide ( Figure 2B ) . Thus , F-MLV env124-138-reactive TCRs have the potential to recognize Emv2 env . This analysis also revealed that Emv2 was not causing complete tolerance of either env124-138L or env124-138Y epitopes . We then confirmed that Emv2 was expressed in primary and secondary lymphoid organs . Using primers specific to the spliced env mRNA that could distinguish between genuine transcripts and contaminating genomic DNA , Emv2 was found to be expressed at low levels in the majority of mice tested ( Figure 2C ) . This low level of expression was further confirmed by comparison with a newly-generated B6 congenic strain lacking Emv2 ( Figure S3 ) . To evaluate the extent of Emv2-mediated deletion of env-reactive CD4+ T cells more directly , we generated B6-Emv2−/− EF4 . 1 mice and compared them with Emv2-expressing EF4 . 1 mice . Emv2-deficient EF4 . 1 mice contained a significantly higher frequency of env124-138L-reactive CD4+ T cells than Emv2-sufficient EF4 . 1 mice , with Emv2 , when present , being responsible for the deletion of approximately 35% of these cells ( Figure 2D ) . Thus , albeit low , expression of Emv2 in B6 mice significantly impacted on the frequency of env124-138L-reactive EF4 . 1 CD4+ T cells . Emv2-mediated deletion of a proportion of env124-138L-reactive EF4 . 1 CD4+ T cells suggested that Emv2 may impinge on resistance to FV infection . We therefore examined the possible effect of Emv2 expression on FV control . Firstly , we infected non-transgenic B6 and Emv2-deficient B6 mice and measured the levels of infected cells in the spleen . B6 mice are relatively resistant to FV infection due to genetic resistance at the Fv2 locus and due to mounting a strong adaptive immune response [1] , [25] , resulting in control of the infection by the third week . Percentages of FV-infected ( glyco-Gag+ ) erythroid precursor ( nucleated Ter119+ ) cells were significantly lower in B6-Emv2−/− mice than in wt counterparts at day 7 of infection ( Figure 3A ) . Nevertheless , wt B6 mice effectively controlled FV infection to levels comparable with those in B6-Emv2−/− mice by the second week of infection ( Figure 3A ) . Thus , Emv2 deficiency did not extensively increase the natural resistance of B6 mice to FV infection . The modest increase in resistance to FV infection in B6-Emv2−/− mice suggested that this low Emv2 expression was immunologically relevant , but did not indicate if any arm of the adaptive immune response was affected . We thus measured the FV-specific CD4+ T cell , CD8+ T cell and antibody responses in these mice . In contrast to the MHC class II-restricted env122-141L epitope , the FV-derived MHC class I-restricted epitopes that have been described do not share extensive homology or cross-reactivity with those derived from Emv2 [26]–[28] . We examined the CD8+ T cell response to FV by measuring numbers of activated CD44hiCD43+CD8+ T cells , irrespective of antigen specificity , in the spleens of B6 and B6-Emv2−/− mice 7 days post FV infection ( Figure S4A ) . The two types of host showed comparable expansion of CD44hiCD43+CD8+ T cells , suggesting they mounted a CD8+ T cell response of similar magnitude ( Figure S4A ) . We further measured the CD8+ T cell response to the immunodominant Db-restricted epitope from the leader sequence gPr80gag85-93 encoded by the F-MuLV gag gene [28] . CD8+ T cells specific to the gPr80gag85-93 epitope display strong bias for the use of Vα3 . 2 and Vβ5 . 2 in combination , which allows their identification by flow cytometry [29] . Expectedly , FV infection led to an increase in the percentage of Vα3 . 2+Vβ5 . 2+ cells in antigen-experienced ( CD44hi ) , but not naïve ( CD44lo ) CD8+ T cells ( Figure S4B ) . However , this expansion of Vα3 . 2+Vβ5 . 2+CD44hiCD8+ T cells was comparable in B6 and B6-Emv2−/− mice 7 days post FV infection ( Figure S4B ) . We next examined the FV-specific antibody response of B6 and B6-Emv2−/− hosts . As FV-neutralizing antibodies are not readily detected in B6 mice on day 7 post FV infection [25] , we measured titers of antibodies that were able to bind F-MLV-infected cells ( Figure S4C ) . On day 7 post FV infection all mice produced both IgG and IgM F-MLV-infected cell-binding antibodies that could be measured by flow cytometry ( unpublished data ) . However , titers of these antibodies were low at this early time-point and in a proportion of FV-infected mice they were below 50 , a value that we set as the detection limit ( Figure S4C ) . Importantly , serum titers of both IgG and IgM F-MLV-infected cell-binding antibodies were similar between B6 and B6-Emv2−/− hosts ( Figure S4C ) . Lastly , the frequency of Ab-env123-141 tetramer+ CD4+ T cells as well as the frequency of Vα2 cells within this population was highly variable between individual mice and as a result not statistically different between groups of B6 and B6-Emv2−/− hosts on day 7 post FV infection ( Figure S4D ) . However , staining with the Ab-env123-141 tetramer may have underestimated the frequency of env122-141L-reactive CD4+ T cells ( Figure 1 ) and it was also possible that env122-141L-reactive CD4+ T cells selected in the presence or absence of Emv2 expressed TCRs with distinct Ab-env123-141 tetramer-binding properties . Furthermore , virus-specific CD4+ T cells can mediate both direct and indirect protection against FV infection [15] , [23] , [30] , and env122-141-specific CD4+ T cells have been shown to mediate direct cytotoxic activity [31] . It was thus uncertain whether weakened immunity in Emv2-expressing mice was directly linked to a potentially less effective CD4+ T cell response . We therefore examined the effect of Emv2 expression on the FV-specific CD4+ T cell response functionally and directly . To this end , equal numbers of Emv2-selected or -nonselected EF4 . 1 CD4+ T cells were transferred into the same type of host . This approach ensured that only donor EF4 . 1 CD4+ T cells differed with respect to exposure to Emv2 . Surprisingly , the two types of donor CD4+ T cells provided comparable and almost complete protection of wild-type B6 hosts , at the peak of FV replication on day 7 post infection ( Figure 3B , C ) . To rule out that differences in antiviral activity between the two types of donor CD4+ T cells were not missed because this activity was already maximal , we have additionally used B6 . A-Fv2s hosts , expressing the susceptibility allele at the Fv2 locus , which confers susceptibility to FV infection by enhancing proliferation of infected erythroid precursors [25] . The two types of donor CD4+ T cells provided significant , suboptimal and , importantly , comparable protection in B6 . A-Fv2s hosts , at the peak of FV replication and expansion of infected erythroid precursors on day 7 post infection in this strain ( Figure 3C ) . Thus , Emv2-mediated selection did not impair the antiviral activity of CD4+ T cells exerted in wt hosts . To further examine direct CD4+ T cell-mediated protection we transferred equal numbers of Emv2-selected or -nonselected EF4 . 1 CD4+ T cells into T and B cell-deficient Rag1−/− Fv2s hosts . Both types of donor CD4+ T cells were similarly protective against severe FV-induced splenomegaly ( Figure 3D ) that otherwise develops in these hosts [15] . In addition , the two types of donor CD4+ T cells caused comparable levels of anemia in these T and B cell -deficient hosts ( Figure 3D ) , which results from bone marrow pathology [14] . Lastly , FV-neutralizing antibodies were similarly and efficiently induced in T cell-deficient Tcra−/− hosts by transfer of either type of donor CD4+ T cells , although they were slightly , but not significantly higher in hosts of Emv2-nonselected CD4+ T cells on day 7 post infection ( Figure 3E ) . Nevertheless , at this time-point , the two types of donor CD4+ T cells induced comparable titers of IgG or IgM antibodies that were able to bind F-MLV-infected cells , which also included antibodies potentially mediating antibody-dependent cell-mediated cytotoxicity ( Figure 3E ) . Collectively , these results demonstrated that despite selecting against a significant fraction of env124-138L-reactive CD4+ T cells , Emv2 expression did not compromise CD4+ T cell function against FV infection . Retention of full CD4+ T cell-mediated antiviral activity , despite deletion of over a third of env124-138L-reactive CD4+ T cells in Emv2-expressing mice , suggested that the deleted cells were not contributing to immunity against FV infection . We therefore assessed the impact of Emv2 expression on both the magnitude and composition of the CD4+ T cell response to FV . Equal numbers of EF4 . 1 CD4+ T cells from either Emv2-sufficient or -deficient donor B6 mice , positive for CD45 . 2 ( encoded by the Ptprc2 allele ) , were adoptively transferred into Ptprc1/2 syngeneic B6 recipients that were positive for both CD45 . 1 and CD45 . 2 . Recipient mice were infected with FV on the day of T cell transfer and FV-responding donor CD4+ T cells were identified as CD44hiCD45 . 2+CD45 . 1− cells ( Figure S5 ) . Consistent with increased precursor frequency in Emv2-deficient donor mice , significantly higher numbers of total responding CD4+ T cells could be recovered at the peak of the response from secondary recipients that received Emv2-nonselected than those that received Emv2-selected donor CD4+ T cells ( Figure 4A ) . Notably , the two types of donor CD4+ T cells generated comparable numbers of high-avidity responding CD4+ T cells , and the numerical increase in total numbers of responding CD4+ T cells from Emv2-nonselected donors was due to significantly higher expansion of low-avidity non-Vα2 responding CD4+ T cells from these donors in comparison with the expansion of non-Vα2 responding CD4+ T cells from Emv2-selected donors ( Figure 4A ) . As a result , peak expansion of Emv2-selected CD4+ T cells was dominated by high-avidity Vα2 CD4+ T cells , whereas that of Emv2-nonselected CD4+ T cells was dominated by low-avidity non-Vα2 CD4+ T cells ( Figure 4B , C ) . Thus , Emv2 expression converted a predominantly low-avidity response to FV to a predominantly high-avidity response . The shift from a predominantly Vα2 response of Emv2-selected CD4+ T cells to a predominantly non-Vα2 response of Emv2-nonselected CD4+ T cells could be the result of Emv2-induced modulation of either the relative frequency in the naïve repertoire of the two subsets of env124-138L-reactive CD4+ T cells , or their relative avidity for env124-138L ( or both ) . We first measured the overall functional avidity to env124-138L of EF4 . 1 CD4+ T cells selected with or without Emv2 as an indicator of potential avidity repertoire changes . Surprisingly , although the presence of Emv2 reduced the precursor frequency of env124-138L-reactive CD4+ T cells , it had no effect on the avidity with which they responded to env124-138L stimulation ( Figure 5A ) . This result suggested that Emv2-mediated selection either affected high- and low-avidity cells similarly , or that potential loss of higher-avidity T cells was compensated by an increase in average avidity of the remaining T cells . To examine whether Emv2 preferentially selected against high-avidity env124-138L-reactive cells , we measured their frequency separately in either Vα2 or non-Vα2 CD4+ T cells from EF4 . 1 mice . Notably , Emv2 expression significantly reduced the frequency of non-Vα2 , but not Vα2 env124-138L-reactive cells in EF4 . 1 CD4+ T cells ( Figure 5B ) , indicating that it only selected against non-Vα2 CD4+ T cells . Correspondingly , the avidity of Vα2 CD4+ T cells to env124-138L was not altered by Emv2 expression , whereas the avidity of non-Vα2 CD4+ T cells was 3-fold higher in the absence than in the presence of Emv2 ( Figure 5C ) . Nevertheless , non-Vα2 CD4+ T cells from Emv2-deficienct mice still displayed lower avidity than Vα2 CD4+ T cells from either Emv2-deficienct or -sufficient mice ( Figure 5C ) . To test whether Emv2-mediated changes in the frequency and avidity for env124-138L of non-Vα2 CD4+ T cells could account for the dominance of this subset in the in vivo response to FV of Emv2-nonselected CD4+ T cells , we examined the in vitro response of Emv2-selected or -nonselected primary naïve EF4 . 1 CD4+ T cells to env124-138L stimulation . As a result of differences in initial frequency and functional avidity between virus-naïve Vα2 and non-Vα2 env122-141-specific cells , the composition of the responding population varied according to the amount of env122-141 presentation [14] and Vα2 T cells dominated the response at doses lower than 10−7 M ( Figure 5D ) . Importantly , this percentage of Vα2 cells was significantly lower at all peptide doses in CD4+ T cells selected in the absence than in the presence of Emv2 ( Figure 5D ) , demonstrating that selection by this single provirus heavily influenced the clonal composition of env124-138L-reactive CD4+ T cells , in favor of high-avidity cells . We further confirmed that this effect of Emv2 expression of reducing the overall frequency of env124-138L-reactive cells , but significantly increasing the percentage of high-avidity Vα2 cells in the env124-138L-reactive population was already evident in CD4+CD8− thymocytes ( Figure S6 ) , consistent with a thymic , rather than peripheral event . Preferential deletion by Emv2 of non-Vα2 CD4+ T cells , which had low avidity for F-MLV env124-138L raised the possibility that these cells may have been cross-reactive with Emv2-encoded env124-138Y . Indeed , lack of Emv2 expression in EF4 . 1 mice had a small , non-significant effect on env124-138Y-reactive Vα2 CD4+ T cells , but caused a significant 3 . 5-fold increase in the frequency of env124-138Y-reactive non-Vα2 CD4+ T cells ( Figure 5E ) . Furthermore , Vα2 CD4+ T cells from either Emv2-deficient or -sufficient EF4 . 1 mice could only react with env124-138Y at the highest dose of 10−5 M , whereas non-Vα2 CD4+ T cells from Emv2- deficient mice were markedly more sensitive to env124-138Y than those from Emv2-sufficient mice ( and as sensitive as Vα2 CD4+ T cells to env124-138L ) ( Figure 5F ) . Together , these findings indicated that Emv2 expression was not affecting env124-138L-reactive Vα2 CD4+ T cells because they displayed low avidity for env124-138Y , but was deleting a significant proportion of non-Vα2 CD4+ T cells that could react with either env124-138L or env124-138Y . Although EF4 . 1 CD4+ T cells selected by Emv2 mounted high-avidity responses to the index env124-138L sequence in vitro , and to FV infection in vivo , and retained full antiviral activity , counter-selection of env124-138Y-reactive clones indicated that this repertoire would be less able to respond to viral escape mutations , and especially to an L128Y mutation . To extend these findings , we used another variant of env , which differed from F-MLV env in two of the three putative TCR-binding residues . This variant has Y and S in positions 128 and 129 , respectively ( referred to as env124-138YS ) and is a naturally-occurring functional form of ecotropic env , encoded by the Fv4 locus in certain strains and species of mouse , other than the B6 strain [32] , [33] . Again , a very small fraction of EF4 . 1 Vα2 CD4+ T cells could react to env124-138YS , regardless of the presence or absence of Emv2 ( Figure 6A ) . In contrast , lack of Emv2 led to a 7-fold increase in the frequency of env124-138YS-reactive EF4 . 1 non-Vα2 CD4+ T cells , which now made a sizable fraction ( Figure 6A ) . Thus , non-Vα2 CD4+ T cells from Emv2-deficient EF4 . 1 mice could react with the index sequence and the two env variants and with high avidity to env124-138Y , suggesting that Emv2-mediated selection significantly reduced the ability of CD4+ T cells , at the population level , to recognize these env variants . This analysis of polyclonal cells from Emv2-deficient EF4 . 1 mice did not reveal whether the same T cell could react to all three env variants or if env124-138L- , env124-138Y- and env124-138YS-reactive non-Vα2 CD4+ T cells were distinct . We therefore tested the reactivity of hybridoma cell lines generated from env124-138L-reactive EF4 . 1 CD4+ T cells that developed either in the presence or the absence of Emv2 expression to other env variants . Similarly to non-Vα2 CD4+ T cell hybridomas from Emv2-sufficient donors , all 4 non-Vα2 CD4+ T cell hybridomas tested from Emv2-deficient donors used members of the TCRVα3 family ( encoded by the Trav9 gene family; Table S2 ) . Notably , neither Emv2-selected nor -nonselected env124-138L-reactive non-Vα2 CD4+ T cell hybridomas responded to env124-138Y more potently than Vα2 CD4+ T cell hybridomas from the same donor strain , and only 1 out of 4 had a measureable response to env124-138YS ( Figure 6B , C ) . These findings suggested that the env124-138L-reactive non-Vα2 CD4+ T cells that developed in Emv2-deficient EF4 . 1 mice were largely distinct from env124-138Y- and env124-138YS-reactive T cells in the same mice . They also indicated that env124-138L-reactive non-Vα2 CD4+ T cells were not inherently more cross-reactive than env124-138L-reactive Vα2 CD4+ T cells at the clonal level . To gain a more detailed view of the depth , defined here as the ability to tolerate epitope mutations , of env124-138L-reactive Vα2 or non-Vα2 TCRs , we screened Emv2-selected or -nonselected env124-138L-reactive T cell hybridomas for reactivity against a library of env126-138 peptide variants carrying all possible single mutations in each of the amino acid residues in positions 128 , 129 and 133 ( Figure S7 ) . Amino acids that elicited at least 40% of the maximal response were listed in the order they were preferred by the individual TCRs ( Figure 7 ) . All Vα2 T cell hybridomas displayed strong preference for L at position 128 and also recognized similar amino acids with hydrophobic side chains , namely F , I , M and V , but not the less hydrophobic Y ( Figure 7A ) . Vα2 T cell hybridomas also showed strong preference and specificity for the amino acid residues of the index sequence against which they were derived , T or highly similar S at position 129 , and N at position 133 ( Figure 7B , C ) . Overall , the depth of Vα2 T cell hybridomas was highly homogeneous and unaffected by Emv2 expression . In contrast to Vα2 T cell hybridomas , and as expected by their low avidity for the index env124-138L sequence , none of the non-Vα2 T cell hybridomas derived from Emv2-deficient mice displayed strong preference for L at position 128 ( Figure 7A ) . The latter hybridomas did , however , respond strongly to env variants with a different amino acid residue at this position , most frequently V or I , or in the case of clone E2H10 the unrelated S ( Figure 7A ) . Non-Vα2 T cell hybridomas selected by Emv2 were also heterogeneous , with two clones showing similar preference and specificity for V or I , and two other clones showing much wider reactivity to at least 10 different amino acid residues , including L ( Figure 7A ) . Furthermore , the low reactivity to the index env124-138L sequence of two of the four non-Vα2 T cell hybridomas derived from Emv2-deficient mice , but not those derived from Emv2-sufficient mice , could be enhanced by substitutions at another position ( C for clone E2H10 , and S or T for clone E2L18 , instead of N at position 133 ) ( Figure 7C ) . Non-Vα2 T cell hybridomas that could recognize L at position 128 also preferred the amino acid residue of the index env124-138L sequence at the two other positions ( T and N for positions 129 and 133 , respectively ) ( Figure 7B , C ) . Collectively , these results confirmed the differential preference for L at position 128 between Vα2 and non-Vα2 T cell hybridomas and further suggested that selection by Emv2 enriched the non-Vα2 repertoire for clones with relative indifference for this position . Analysis of the env-reactive CD4+ T cell repertoire in B6 mice revealed a clear effect of Emv2-mediated selection . However , in addition to Emv2 , the presence of numerous other endogenous retroviruses could affect the formation of the env124-138L-reactive CD4+ T cell repertoire , even if their primary amino acid sequence is not closely homologous with that of F-MLV env . Furthermore , the functional avidity of env124-138L-reactive CD4+ T cells could also be affected by additional genetic determinants other than endogenous retroviruses . To address this question we generated congenic EF4 . 1 mice on the 129S8 background . 129S8 mice share the same MHC class II allele with B6 mice ( H2-Ab ) , thus allowing restriction of env124-138L-specific EF4 . 1 CD4+ T cells . However , they do differ substantially with respect to the composition of endogenous retroviruses and , importantly , 129S8 mice are naturally devoid of endogenous ecotropic MLVs [34] , [35] . Similar frequency of env124-138L-reactive Vα2 CD4+ T cells developed in B6 and 129S8 EF4 . 1 mice ( Figure 8A ) . In contrast , the frequency of env124-138L-reactive non-Vα2 CD4+ T cells was significantly higher on the 129S8 than on the B6 background ( Figure 8A ) , and was comparable with that on the Emv2-deficient B6 background ( Figure 5B ) , as was their functional avidity ( Figure 8B ) . This finding indicated that deletion of env124-138L-reactive non-Vα2 CD4+ T cells in B6 , but not in B6-Emv2−/− or 129S8 mice was mediated primarily by Emv2 . Surprisingly , however , the functional avidity of env124-138L-reactive Vα2 CD4+ T cells in 129S8 mice was very much reduced in comparison with that of Vα2 CD4+ T cells in B6 mice ( Figure 8B ) , and was as low as that of low-avidity non-Vα2 CD4+ T cells . As , a result of differences in frequency and functional avidity , the env124-138L-specific response of 129S8 mice was dominated by non-Vα2 CD4+ T cells at all peptide doses , in contrast to that of B6 mice , which was dominated by Vα2 CD4+ T cells at low peptide doses ( Figure 8C ) . To further explore the origin of high-avidity env124-138L-reactive Vα2 CD4+ T cells in B6 , but not in 129S8 mice , we tested the response of a series of B6×129S8-EF4 . 1 F1 mice . In comparison with B6×129S8-EF4 . 1 F1 mice , which inherited Emv2 from the B6 parent , B6-Emv2−/−×129S8-EF4 . 1 F1 mice , which lacked ecotropic MLVs , had elevated frequencies of env124-138L-reactive non-Vα2 CD4+ T cells , whereas frequencies of env124-138L-reactive Vα2 CD4+ T cells were similar ( Figure 8D ) . These results confirmed that elevated frequencies of env124-138L-reactive non-Vα2 CD4+ T cells in 129S8 mice were indeed due to lack of Emv2-mediated selection . Interestingly , both B6×129S8-EF4 . 1 and B6-Emv2−/−×129S8-EF4 . 1 F1 mice generated env124-138L-reactive Vα2 CD4+ T cells with higher avidity than those of 129S8 mice ( Figure 8E ) , suggesting that a genetic contribution of the B6 parent , other than Emv2 , was necessary for the development of high-avidity env124-138L-reactive Vα2 CD4+ T cells . To assess whether this genetic contribution arose from polymorphisms in the Trav locus itself , we tested B6-Tcra−/−×129S8-EF4 . 1 F1 mice , which inherited Emv2 from the B6 parent , but could generate endogenous Vα chains only from the locus inherited from the 129S8 parent . The presence of Emv2 in B6-Tcra−/−×129S8-EF4 . 1 F1 mice had the predicable effect on the frequency of env124-138L-reactive non-Vα2 CD4+ T cells ( Figure 8D ) , which displayed comparably low avidity in all three F1 strains tested ( Figure 8E ) . Surprisingly , however , env124-138L-reactive Vα2 CD4+ T cells that had developed in B6-Tcra−/−×129S8-EF4 . 1 F1 mice were also low-avidity , which was comparable with that of Vα2 CD4+ T cells in 129S8 mice ( Figure 8E ) , suggesting that the ability of B6 mice to generate high-avidity env124-138L-reactive Vα2 CD4+ T cells was germline-encoded . Consequently , the env124-138L-specific response of B6×129S8-EF4 . 1 F1 mice , but not of isogenic mice lacking either Emv2 or the B6-origin Trav , was dominated by Vα2 CD4+ T cells at low peptide doses ( Figure 8F ) . The peak percentage of Vα2 CD4+ T cells in the env124-138L-reactive population was lower in B6×129S8-EF4 . 1 F1 mice than in B6 mice , as the former were expressing endogenous Vα chains from both parental Trav loci . Thus , the combined effect of Emv2 on the frequency of non-Vα2 T cells and of Trav on the avidity of Vα2 T cells was necessary for the dominance of high-avidity Vα2 CD4+ T cells in the response to env124-138L . As a result of the combinatorial process that creates TCRs , their specificity is random and has to undergo selection . Thymic positive and negative selection of developing T cells ensures that mature T cells in the periphery have a functional TCR and minimal reactivity to self proteins , respectively [36] . Negative selection is thought to decrease the frequency , avidity and cross-reactivity of the developing TCR repertoire specific to foreign epitopes that may be similar to self-derived epitopes presented in the thymus [36] and promote peptide specificity [37] . Here we used a well-characterized molecular system to show that negative selection by a defined self peptide from Emv2 env indeed decreased the frequency in the naïve CD4+ T cell repertoire of clones specific to a range of foreign env epitopes , thus reducing the magnitude of the CD4+ T cell response to all env epitope variants . However , negative selection counter-intuitively also promoted the avidity of the CD4+ T cell response to F-MLV env by shifting the clonal composition of responding CD4+ T cells in favor of high-avidity cells . CD4+ T cells play a central coordinating role in the orchestration of adaptive immunity to infection , and may also mediate direct antiviral activity . Recent studies in diverse systems have indicated an essential role for the CD4+ T cell response in the control of retroviral infection [15] , [38]–[42] . We have previously shown that protection of wt mice against acute FV infection is proportional to the frequency of virus-specific CD4+ T cells [23] . Surprisingly , we found that although negative selection significantly reduced both the precursor frequency and peak expansion of F-MLV env-specific CD4+ T cells , it did not compromise CD4+ T cell-mediated antiviral activity . This finding suggested that not all virus-specific CD4+ T cells were equal in their ability to mediate antiviral functions . Indeed , negative selection by Emv2 env affected CD4+ T cells with low avidity for F-MLV env , but not those with high avidity for the same epitope . Preservation of full antiviral activity in the Emv2-selected CD4+ T cell repertoire therefore indicated that this activity is primarily , if not exclusively , exerted by high-avidity CD4+ T cells . High-avidity virus-specific CD4+ T cells may be superior in certain direct antiviral or indirect helper functions than low-avidity ones , but there may also be important exceptions . High-avidity CD4+ T cells responding to FV infection have been reported to show enhanced ex vivo production of IFN-γ and IL-21 cytokines and reduced expression of PD-1 inhibitory receptor [15] than low-avidity counterparts , properties that may contribute to superior antiviral activity . However , T follicular helper ( Tfh ) differentiation and function were previously found to be similar between high- and low-avidity virus-specific CD4+ T cells [15] , suggesting that provision of T cell help for the production of virus-neutralizing antibodies may be more sensitive to the frequency of virus-specific CD4+ T cells , rather than their avidity . However , in addition to the frequency of virus-specific CD4+ T cells , the virus-specific antibody response is also proportional to the frequency of rare antigen-specific B cells . Thus , when availability of T cell help is abundant , the virus-specific antibody response may be limited by the frequency of antigen-specific B cells and additional T cell help would not be expected to enhance antibody production . Consistent with this idea , adoptive transfer of virus-specific EF4 . 1 CD4+ T cells into wt B6 mice did not accelerate the virus-neutralizing antibody response [23] . In addition to an effect of Emv2 on the availability of T cell help for the FV-specific antibody response , Emv2 could in principle also directly affect the development of virus-specific B cells [43] . Although we observed comparably low FV-specific antibody responses between B6 and B6-Emv2−/− mice at the peak of FV infection , our results did not exclude a potential direct effect of Emv2 on FV-specific B cell and antibody responses at later time-points , when these responses are fully induced . Indeed , Emv2-encoded env shares 79% amino acid identity with F-MLV env and it is therefore possible that Emv2 expression , especially when upregulated , might affect the FV-specific antibody response . As previously shown , high-avidity F-MLV env122-141L-specific Vα2 CD4+ T cells are a minority subset in the naïve repertoire and only dominate the immune response to FV as a result of their preferential expansion during infection [15] . We have now found that for this ability of high-avidity F-MLV env122-141L-specific Vα2 CD4+ T cells to dominate the peak response , negative selection by Emv2 of at least some of the competitor low-avidity F-MLV env122-141L-specific non-Vα2 CD4+ T cells is necessary . These findings indicate that even subtle thymic events can have profound effects on the induction of an effective T cell response to retroviral infection . Recently , a comprehensive theoretical study has indicated that HLA class I alleles that associated with control of HIV infection , such as HLA-B5701 , sample far fewer self peptides than other HLA alleles [5] . As a result of less stringent negative selection , a higher frequency of CD8+ T cells restricted by these protective alleles were predicted to recognize viral peptide epitopes and to cross-react with variants of the targeted epitopes [5] . Our results with a single self peptide provide further experimental confirmation of negative selection reducing both the precursor frequency and cross-reactivity of env-specific CD4+ T cells , although in this case the effect on cross-reactivity was more pronounced at the population , rather than the single-cell level . These results also suggest that from the thousands of self peptides that can mediate thymic selection of retrovirus-specific T cells , the main effects may be mediated by only a few self peptides . Moreover , self peptides with such strong influence may also be polymorphic between different individuals , which might contribute to the partial association of HLA polymorphisms with virus control [3] , [5] , [6] . In addition to polymorphisms at the MHC/HLA locus or of self peptides mediating thymic selection , the Trav/TRAV and Trbv/TRBV loci may also display allelic sequence variation . A polymorphism in the TRBV9 gene has been shown to affect TCR affinity for and functional recognition of an HLA-B3501-restricted epitope from the EBNA-1 protein of Epstein-Barr virus ( EBV ) , leading to a public T cell response dominated by the high-affinity variant [44] . Similarly , we found that the ability of Vα2 chains to confer high avidity for env122-141L in EF4 . 1 mice seems to be germline-encoded , as only Vα2 chains encoded by the B6 , but not the 129 Trav locus had this ability . It is tempting to speculate that amino acid residues unique to the B6-germline Trav14-encoded Vα2 chains participate in recognition of the strongly interacting L ( or a limited set of amino acids with similar properties ) at env position 128 . Notably , the CD8+ T cell response to an HLA-B8-restricted epitope from the latent antigen EBNA 3A of EBV uses almost exclusively identical Vα and Vβ , as well as other TCR-region sequences , and comprehensive structural studies have shown that a unique amino acid residue in the germline-encoded complementarity-determining region 2 ( CDR2 ) of the preferred Vα chain , encoded by TRAV26-2 , is critically required for binding to a residue from the peptide epitope [45] . Despite the vast number of somatically-generated random TCRs that can arise during T cell development , these studies highlight the potential for germline-encoded residues to provide exquisite specificity and competitive advantage to the TCRs that carry them . In addition to likely representing the best-fit for recognition of Ab-restricted env122-141L , the dominance of Vα2 EF4 . 1 CD4+ T cells could also result from preferential pairing of the transgenic TCRβ chain with Vα2 chains in general . This is unlikely to be the case as the usage of Vα2 cells was not increased in either total or env122-141L-reactive EF4 . 1 CD4+ T cells , and indeed in the env122-141L-reactive preimmune repertoire clones using other Vα chains were at least 3 times more frequent than those using Vα2 . However , although non-Vα2 env122-141L-reactive CD4+ T cells were still the majority in Emv2-expressing mice , their ability to participate in the response to FV and compete with env122-141L-reactive Vα2 CD4+ T cells was severely compromised by Emv2 . Thus , the dominance of Vα2 CD4+ T cells in the response to FV infection can be seen as a combination of germline-encoded advantage in Ab-restricted env122-141L recognition conferred to Vα2 CD4+ T cells and of Emv2-mediated self-tolerance of other non-Vα2 CD4+ T cells capable of recognizing Ab-restricted env122-141L . One important novel insight of the current study is the proof of principle that negative selection is not necessarily always impairing high-avidity T cell responses . By counter-selecting some cross-reactive CD4+ T cells , negatively selecting self peptides have the ability to significantly enhance the avidity for the response to at least some epitope variants . Higher precursor frequency and cross-reactivity with emerging epitope variants seem to be the best correlates for an effective cytotoxic CD8+ T cell response [5] . Whether higher avidity for the primary infecting epitope , rather than cross-reactivity with epitope variants better describes an effective CD4+ T cell response to retroviral infection needs to be further addressed . It should be noted that differences in avidity for antigen in this system were defined functionally . Indeed , Vα2 env122-141L-specific primary CD4+ T cells or hybridomas reacted to much lower concentrations of env122-141L peptide stimulation in vitro than their non-Vα2 counterparts . Furthermore , this higher sensitivity translated to higher in vivo expansion and increased potential for cytokine production [15] . It is currently unclear whether differences in functional avidity between Vα2 and non-Vα2 env122-141L-specific CD4+ T cells resulted from overall higher affinity of individual TCRs of these polyclonal populations for the peptide-MHC class II complex . Although dissociation kinetics between TCRs and peptide-MHC class II tetramers are often informative with respect to the biochemical affinity of these TCRs , they may not be universally useful . For example , the available env123-141-Ab tetramer ( Ab-env ) is known to bind only some env124-138L-specific CD4+ T cell clones but not others , irrespective of their functional avidity or Vα usage [14] , [22] . Therefore , this reagent could not be used to access the biochemical affinity of all env124-138L-specific CD4+ T cells in the polyclonal repertoire . Furthermore , identification of antigen-specific cells using a sensitive two-dimensional binding assay has recently demonstrated that the affinity of many CD4+ T cells that participate in the response to two separate antigens is below detection with peptide-MHC class II tetramers [46] . Thus , peptide-MHC class II tetramers may generally only detect some but not all antigen-specific CD4+ T cells . In addition , such detection is conditional on expression of sufficient TCR levels . Indeed , we have found that the extensive , antigen-induced downregulation of their TCR in vivo , eclipses detection with the Ab-env123-141 tetramer of even the env122-141L-reactive CD4+ T cells that could otherwise bind this reagent . Similar observations have been recently made with peptide-MHC class I tetramer staining of virus-specific effector CD8+ T cells [47] , suggesting that the inability of peptide-MHC multimers to identify antigen-specific effector T cells that have downregulated their TCRs may be a general problem for T cells restricted by both classes of MHC molecules . Negative selection ensures minimal reactivity of developing thymocytes to self proteins . However , endogenous retroviruses are a large constituent of mammalian genomes and thus represent a potentially large pool of self proteins able to mediate selection , both positive and negative . Self peptides encoded by endogenous MLVs have been shown to mediate positive selection of CD4+ T cells with specificity for an unrelated H2-Ek-restricted moth cytochrome C peptide , and to enhance the response of mature CD4+ T cells with this specificity in the periphery [48] . We found that Emv2 was expressed at very low levels in the thymus of B6 mice , in agreement with a previous report [49] , and was undetectable by qRT-PCR in some of the mice . It should be noted , however , that the qRT-PCR method employed was specific only for the spliced env mRNA that is transcribed by Emv2 . This was chosen to eliminate the possibility of detecting contaminating genomic DNA or viral genomic RNA , but may underestimate the total amount of spliced and unspliced mRNA that leads to the production of other viral proteins . Nevertheless , as demonstrated by its effect on thymic development , this low level of Emv2 expression was clearly functional . Endogenous retroviruses have been known for many years to cause a range of different diseases in mice , including cancer , immunodeficiency and autoimmunity , although a similar causal effect in humans has been questioned [50] . Immune reactivity to endogenous retroviruses has been amply demonstrated in mice where is has been strongly associated with the development of spontaneous autoimmune conditions [51] , [52] . Interestingly , immune reactivity to endogenous retroviruses has also been frequently observed in humans during infection , inflammation , autoimmunity and cancer [50] , [53]–[56] . Expression of human endogenous retroviruses , as well as CD8+ T cell responses against their antigens , have been documented in HIV infection [57] , [58] . Furthermore , a whole-genome association study has suggested that part of the effect of the protective HLA-B*5701 allele during the asymptomatic period of HIV infection may be mediated by a linked human endogenous retrovirus at the same locus [59] . Human endogenous retroviral antigens have also been reported to serve as targets for CD8+ T cell-mediated rejection of cancer cells [60] . It might be evident from the studies in humans and the results of the current study that peptide epitopes encoded by endogenous retroviruses have a strong influence on T cell thymic selection and may also participate in the shaping of the peripheral T cell response . It is also clear that endogenous retroviruses do not always cause immunological tolerance , and although their activation in infected or transformed cells may provide a non-mutable target for immune attack , activation of endogenous retroviruses may also trigger inflammatory or autoimmune phenomena frequently associated with infection and cancer . Further study of endogenous retrovirus regulation during infection , autoimmunity or cancer , and of the immune responsiveness to them should shed more light into their pathogenic potential . All animal experiments were approved by the ethical committee of the NIMR , and conducted according to local guidelines and UK Home Office regulations under the Animals Scientific Procedures Act 1986 ( ASPA ) . Inbred C57BL/6J ( B6 ) , A/J and B6 . SJL-Ptprca Pep3b/BoyJ ( CD45 . 1+ B6 ) mice were originally obtained from The Jackson Laboratory ( Bar Harbor , Maine , USA ) and were subsequently maintained at NIMR animal facilities . Inbred 129S8/SvEvNimrJ ( 129S8 ) mice were developed from an 129/Sv substrain , maintained at NIMR animal facilities , and were subsequently deposited at The Jackson Laboratory . The B6 TCRβ-transgenic strain EF4 . 1 , expressing a transgenic TCRβ chain from a T cell clone specific to F-MuLV env122-141 presented by H2-Ab , has been described [14] . 129S8-congenic EF4 . 1 mice were generated by serial backcrossing of B6-EF4 . 1 mice for 10 nuclear generations onto the 129S8 genetic background . B6-backcrossed Rag1-deficient ( Rag1−/− ) mice [61] and T cell receptor α-deficient ( Tcra−/− ) mice [62] were also maintained at NIMR animal facilities . Fv2s-congenic B6 ( Fv2s ) and Rag1−/− ( Fv2s Rag1−/− ) mice have been previously described [25] . Emv2-deficient ( Emv2−/− ) B6 mice were created by introducing the Emv2 integration site of chromosome 8 from the A/J strain , which lacks this proviral integration , by serial backcrossing for at least 12 nuclear generations onto the B6 genetic background . Lack of Emv2 was validated by PCR for both the D8Mit49 microsatellite marker close to the locus that detects polymorphisms in A/J ( Emv2− ) and B6 ( Emv2+ ) strains of mice ( D8Mit49 forward 5′-TCTGTGCATGGCTGTGTATG-3′ and D8Mit49 reverse 5′-TGGTGTGCTGCTGATGCT-3′ ) , and also for the actual integration site using three primers , two of which were flanking the integration site ( forward 5′-ACCCACTAAGTAACCCAGGCTGCCTCAGCT-3′ and reverse 5′-GACCAGAATAGAAAGACGTTCAAGTGAGCT-3′ ) and one located in the Emv2 LTR ( 5′-ATCAGCTCGCTTCTCGCTTCTGTACCCGCG-3′ ) ( Figure S3 ) . Spleen or lymph node single-cell suspensions were prepared from EF4 . 1 mice and 5×105 cells per well were stimulated in 96-well plates with the indicated amount of env peptide variants . The frequency of env-reactive cells in stimulated CD4+ T cells was defined as the frequency of cells that responded to 18-hr stimulation , before cell division or death had occurred , by upregulating CD69 expression . Correct identification of env-reactive CD4+ T cells by CD69 upregulation was confirmed in control experiments by co-staining for CD154 ( CD40L ) expression in stimulated T cells . Both antibodies were obtained from eBiosciences . For assessment of T cell activation on day 3 , cells were labeled with CFSE before stimulation and responding cells were identified by CFSE dilution . Single-cell suspensions were prepared from spleens and lymph nodes from Emv2-sufficient or -deficient EF4 . 1 mice and stimulated in vitro with 10−7 M or 10−5 M env122-141L peptide and 4 ng/ml recombinant human IL-2 for 4 days . CD4+ T cells were subsequently purified from stimulated cultures using immunomagnetic positive selection ( StemCell Technologies , Vancouver , BC , Canada ) and fused to TCRαβ-negative BW5147 thymoma cells to produce hybridoma cell lines . Established hybridoma cell lines were stimulated with a range of env peptide variants presented by dendritic cells . Dendritic cells were obtained from cultures of bone marrow cells isolated from B6 mice and supplemented with granulocyte macrophage colony-stimulating factor ( GM-CSF ) . GM-CSF was obtained from culture supernatant of ×63 cells transfected with mouse Csf2 and was used at 1∶10 dilution . Bone marrow cells were culture in these conditions for 7 days , at which point they consisted of 50–70% dendritic cells . These cells were then used to stimulate hybridoma cells at a ratio of 5×104 dendritic cells to 1×105 hybridoma cells , for 18 hrs , in the presence or absence of env peptide variants . Dendritic cell-hybridoma cell co-cultures were plated in flat-bottom 96-well plates in 200 µl final volume . The concentration of peptides used is indicated in individual figures and figure legends . In additional experiments peritoneal macrophages were also used as antigen-presenting cells with results comparable to the use of dendritic cells . Macrophages were isolated from B6 mice following plating of the peritoneal cavity exudate cells for 1 hr and washing off the non-adherent fraction . Env-specific responses were assessed by measuring the amount of IL-2 secreted in co-culture supernatants using an AlamarBlue ( Invitrogen , Carlsbad , CA , USA ) -based CTLL-2 assay . Trav and Traj usage by T cell hybridomas was probed by staining with an anti-Vα2 ( clone B20 . 1 ) or anti-Vα3 . 2 ( clone RR3-16 ) monoclonal antibodies , and by reverse transcription ( RT ) -PCR amplification and sequencing of expressed Trav genes , using previously described primers [63] . Trav and Traj segment identification and alignment , and confirmation of productive rearrangements were performed on the International Immunogenetics Information System website ( http://www . imgt . org ) . The FV used in this study was a retroviral complex of a replication-competent B-tropic F-MuLV and a replication-defective polycythemia-inducing spleen focus-forming virus ( SFFVp ) . Stocks were propagated in vivo and prepared as 10% w/v homogenate from the spleen of 12-day infected BALB/c mice . Mice received an inoculum of ∼1 , 000 spleen focus-forming units of FV . All viral stocks were free of Sendai virus , Murine hepatitis virus , Parvoviruses 1 and 2 , Reovirus 3 , Theiler's murine encephalomyelitis virus , Murine rotavirus , Ectromelia virus , Murine cytomegalovirus , K virus , Polyomavirus , Hantaan virus , Murine norovirus , Lymphocytic choriomeningitis virus , Murine adenoviruses FL and K87 , and Lactate dehydrogenase-elevating virus . Virus inocula were injected via the tail vein in 0 . 1 ml of phosphate-buffered saline . FV-infected cells were detected by flow cytometry using surface staining for the glycosylated product of the viral gag gene ( glyco-Gag ) , using the matrix ( MA ) -specific monoclonal antibody 34 ( mouse IgG2b ) , followed by an anti-mouse IgG2b-FITC secondary reagent ( BD , San Jose , CA , USA ) . For the assessment of anemia , mice were bled by a small incision of the tail vein and blood was collected into heparinized capillary tubes . Complete blood counts were measured on a VetScan HMII hematology analyzer ( Abaxis , CA , USA ) , following the manufacturer's instructions . RBC counts of uninfected mice were ∼9 . 95×106 per mm3 of blood . FV-induced splenomegaly in infected mice was expressed as spleen index , which is the ratio of the weight of the spleen ( in mg ) to the weight of the rest of the body ( in g ) . Serum titers of FV-neutralizing antibodies were measured as previously described [25] . The dilution of serum which resulted in 75% neutralization was taken as the neutralizing titer . Serum titers of F-MLV-infected cell-binding antibodies were determined by flow cytometry following primary staining of F-MLV-infected Mus dunni cells with serial dilutions of serum samples and secondary staining with fluorescently labeled anti-mouse IgG1 ( clone A85-1 ) , anti-mouse IgG2a/c ( clone R19-15 ) , anti-mouse IgG2b ( clone R12-3 ) or anti-mouse IgM ( clone R6-60 . 2 ) antibodies ( BD ) . B6 mice express the IgG2c isotype , which may not be efficiently detected by anti-IgG2a reagents [64] . Although the R19-15 monoclonal antibody has higher affinity for IgG2a , it can be effectively used for detection of IgG2c . This was confirmed by staining of F-MLV-infected Mus dunni cells that were first incubated with serum from FV-infected mice , with the anti-IgG1 or anti-IgG2a/c or anti-IgG2b reagents separately ( Figure S8A ) . The three reagents used separately resulted in comparable staining intensity , which allowed us to use all three IgG subclass-specific antibodies in combination . For IgG titers , F-MLV-infected Mus dunni cells were first incubated with serum samples and then with anti-IgG1 , anti-IgG2a/c and anti-IgG2b antibodies mixed together . Serum samples were 2-fold serially diluted , starting from an initial dilution of 1∶50 . The last positive serum dilution resulting in staining intensity at least twice the background level was taken as the binding titer ( Figure S8B ) . Single-cell suspensions were prepared from the spleens and lymph nodes of donor CD45 . 2+ EF4 . 1 mice by mechanical disruption . Spleen suspensions were treated with ammonium chloride for erythrocyte lysis . CD4+ T cells were enriched using immunomagnetic positive selection ( StemCell Technologies ) according to the manufacturer's instructions . Purity of the isolated CD4+ T-cell population was routinely higher than 92% . A total of approximately 1×106 EF4 . 1 CD4+ T cells were injected in B6-congenic CD45 . 1+CD45 . 2+ recipients via the tail vein in 0 . 1 ml of air-buffered Iscove's Modified Dulbecco's Media . When adoptive transfer of CD4+ T cells was combined with FV infection , purified CD4+ T cells and virus stocks were injected separately into recipient mice within a 24 hour-period . Spleen-cell suspensions were stained with directly-conjugated antibodies to surface markers , obtained from eBiosciences ( San Diego , CA , USA ) , CALTAG/Invitrogen , BD Biosciences ( San Jose , CA , USA ) or BioLegend ( San Diego , CA , USA ) . Donor-type env-specific CD4+ T cells were identified as CD44hiCD45 . 2+CD45 . 1−CD4+ cells . Four- and 8-color cytometry were performed on FACSCalibur ( BD Biosciences ) and CyAn ( Dako , Fort Collins , CO ) flow cytometers , respectively , and analyzed with FlowJo v8 . 7 ( Tree Star Inc . , Ashland , OR , USA ) or Summit v4 . 3 ( Dako ) analysis software , respectively . Total RNA was extracted from whole organs using TRI-reagent ( Sigma-Aldrich , St . Louis , US ) according to the manufacturer's instructions , precipitated with isopropanol and washed in 75% ethanol before being dissolved in water . DNase digestion and cleanup was performed with the RNeasy Mini Kit ( Qiagen , Hilden , Germany ) and cDNA produced with the high capacity reverse transcription kit ( Applied Biosystems , Carlsbad , US ) with an added RNase inhibitor ( Promega Biosciences , Madison , US ) . A final clean-up was performed with the QIAquick PCR purification kit ( Qiagen ) . Level of expression of Emv2 RNA was determined by qRT-PCR using DNA Master SYBR Green I kit ( Roche , Mannheim , Germany ) and the ABI Prism 7000 or 7900HT Detection System ( TaqMan , Applied Biosystems , Foster City , CA ) . The following primers were used for the amplification of target transcripts: Hprt: forward 5′-TTGTATACCTAATCATTATGCCGAG-3′ and reverse 5′- CATCTCGAGCAAGTCTTTCA-3′; Emv2: forward 5′-CCAGGGACCACCGACCCACCGT-3′ and reverse 5′-TAGTCGGTCCCGGTAGGCCTCG-3′ . Emv2-specific primers amplified only the spliced form of env mRNA , thus eliminating the possibility of residual genomic DNA or RNA contamination contributing to Emv2 signal . The housekeeping gene Hprt was used to normalize the Critical Threshold ( CT ) values for Emv2 . Analysis was conducted with the ΔCT method [65] and Emv2 expression corresponding to an Emv2 CT value of 40 ( the total number of amplification cycles used ) was set at 1 arbitrary unit . A theoretical detection limit of 2 arbitrary units was also used , which represents the detectable Emv2 signal in the penultimate cycle of amplification . Statistical comparisons were made using SigmaPlot 12 . 0 ( Systat Software Inc . , Germany ) . Parametric comparisons of normally-distributed values that satisfied the variance criteria were made by unpaired Student's t-tests . Linear percentages of FV-infected cells , spleen indices and nAb titers , which did not pass the variance test , were compared with non-parametric two-tailed Mann-Whitney Rank Sum or Wilcoxon Signed Rank tests .	Our immune systems defend against viral infection . However , the immune response to a virus often differs substantially between individuals , as does the outcome of infection . The antiviral immune response relies on recognition of viral proteins by T lymphocytes using T cell antigen receptors ( TCRs ) . TCRs are created randomly in an individual and each T lymphocyte has a different TCR . T lymphocytes with TCRs that recognize our own ( self ) proteins are removed during development , to prevent autoimmunity . Our cells can also make proteins that belong to endogenous retroviruses ( ERVs ) , relics of ancestral retroviral infection that accumulated during evolution to constitute a large proportion of our genomes . The impact of ERVs on lymphocyte development and the subsequent influence on antiviral immunity is incompletely understood . Here , we use a mouse model to investigate this link and show that the T lymphocyte response to exogenous retrovirus infection is heavily influenced by an ERV . Surprisingly , we find that ERVs do not always have a negative impact on immunity , and in our model they improve the sensitivity with which T lymphocytes react to retroviral infection . This work may thus provide a basis for the understanding of the heterogeneity in immunity to retroviral infections in genetically diverse individuals .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "adaptive", "immunity", "immune", "cells", "immunity", "t", "cells", "immunity", "to", "infections", "immunology", "biology", "immune", "response" ]	2012	Negative Selection by an Endogenous Retrovirus Promotes a Higher-Avidity CD4+ T Cell Response to Retroviral Infection
Buruli ulcer , an emerging tropical disease caused by Mycobacterium ulcerans ( MU ) , is characterized by disfiguring skin necrosis and high morbidity . Relatively little is understood about the mode of transmission , pathogenesis , or host immune responses to MU infection . Due to significant reduction in quality of life for patients with extensive tissue scarring , and that a disproportionately high percentage of those affected are disadvantaged children , a Buruli ulcer vaccine would be greatly beneficial to the worldwide community . Previous studies have shown that mice inoculated with either M . bovis bacille Calmette–Guérin ( BCG ) or a DNA vaccine encoding the M . ulcerans mycolyl transferase , Ag85A ( MU-Ag85A ) , are transiently protected against pathology caused by intradermal challenge with MU . Building upon this principle , we have generated quality-controlled , live-recombinant strains of BCG and M . smegmatis which express the immunodominant MU Ag85A . Priming with rBCG MU-Ag85A followed by an M . smegmatis MU-Ag85A boost strongly induced murine antigen-specific CD4+ T cells and elicited functional IFNγ-producing splenocytes which recognized MU-Ag85A peptide and whole M . ulcerans better than a BCG prime-boost vaccination . Strikingly , mice vaccinated with a single subcutaneous dose of BCG MU-Ag85A or prime-boost displayed significantly enhanced survival , reduced tissue pathology , and lower bacterial load compared to mice vaccinated with BCG . Importantly , this level of superior protection against experimental Buruli ulcer compared to BCG has not previously been achieved . These results suggest that use of BCG as a recombinant vehicle expressing MU antigens represents an effective Buruli ulcer vaccine strategy and warrants further antigen discovery to improve vaccine efficacy . Buruli ulcer ( BU ) , the clinical manifestation of subcutaneous infection by Mycobacterium ulcerans ( MU ) , is a highly disfiguring flesh-eating skin disease with significant associated morbidity [1–5] . Much of the pathology related to M . ulcerans infection is due to secretion of the potent cytotoxin , mycolactone , which not only leads to tissue necrosis but also possess immunosuppressive and analgesic properties [6–8] . Although infections are present worldwide , endemicity occurs in impoverished areas with below-average access to appropriate diagnosis and medical treatment . Unfortunately this condition disproportionately afflicts children , and over 50% of those affected are less than 15 years of age [9] . Not only is Buruli ulcer capable of causing body-wide scarring , loss of limbs , damage to eyes , osteomyelitis , and secondary colonization of necrotic tissues , but infection is often affiliated with severely negative social stigma and ostracism [10 , 11] . If diagnosed early enough , an eight-week course of multiple anti-mycobacterial drugs are required to eliminate infection [12] . However , due to the typically indolent and misleading nature of initial disease symptoms , skin ulceration is often allowed to progress to a point at which antibiotic treatment is ineffective in preventing further necrosis . At this stage , surgical removal of infected tissue and skin grafting is required . The mode of transmission , pathogenesis , and host immune responses to M . ulcerans infection are poorly understood , and there is currently no effective prophylactic vaccine available [13] . Mycobacterium bovis bacille Calmette-Géurin ( BCG ) is the most widely administered vaccine in the world and is currently the only WHO-approved vaccine for tuberculosis [14–16] . Although BCG confers highly variable protection against pulmonary tuberculosis ( TB ) , defense against disseminated forms of tuberculosis in children has been more consistently observed [17] . Previous retrospective studies have attempted to assess if vaccination with BCG offers protection against Buruli ulcer . Prior vaccination with BCG was observed to offer short term protection , mainly in the form of a delay in onset of tissue ulceration , although up to 45% protection was observed against MU-associated osteomyelitis [18–21] . Mouse models of Buruli ulcer infection have also examined the protective qualities of BCG vaccination and have yielded similar results , whereby the time to onset of footpad swelling in BCG-primed animals is delayed compared to unprimed controls [22] . Alternative anti-Buruli ulcer vaccination strategies have involved use of DNA-based vaccines which direct the expression of MU antigens within host cells . Roupie et al . utilized a plasmid-based vaccine to encode several mycolactone polyketide synthase domains , where DNA-primed mice were boosted with recombinant homologous proteins and were subsequently infected with virulent M . ulcerans 1615 [23] . Despite the evidence for antigen-specific antibody and T cell responses to this vaccination , little protection was observed and none of the vaccines yielded protection matching the level conferred by BCG priming . However in other studies , Tanghe et al . repeatedly observed that upon priming with two doses of plasmid DNA encoding MU mycolyl transferase Ag85A ( MU-Ag85A ) and subsequently boosting with purified recombinant protein , the onset of mouse footpad swelling was delayed to a level approaching that of BCG priming alone [24 , 25] . Based on the combined observations that both BCG inoculation and exposure to certain MU antigens could impart protection in the mouse model of M . ulcerans infection , we examined whether enhanced protection could be conferred by a prime of recombinant BCG ( rBCG ) strain overexpressing MU-Ag85A . Due to the fact that BCG was previously shown to be ineffective at boosting its own prime responses , we also examined the use of recombinant M . smegmatis expressing MU-Ag85A as a boosting agent [26] . Previous studies have highlighted the importance of Th1-based CD4+T cell responses during mycobacterial infection , and the skewing of adaptive immunity away from such responses have been observed to coincide with ulceration in Buruli ulcer patients [27–33] . In addition to the strong proliferation of CD4+ Ag85A-specific T cells and increase in functional IFNγ-producing splenocytes stimulated by our recombinant vaccine , MU bacterial load in the mouse footpad was significantly reduced . Most importantly , either a single administration of BCG-MUAg85A or the prime-boost model using recombinant mycobacteria significantly increased the time to onset of footpad swelling in MU-infected mice , demonstrating the first Buruli ulcer vaccine to achieve enhanced protection over BCG vaccination . Female C57BL/6 mice were obtained from Jackson Laboratories . Mice were 6–8 weeks old at time of vaccination and 16–18 weeks old by time of challenge . Mycobacterium bovis BCG-Danish ( BCGD ) was cultured in liquid Difco Middlebrook 7H9 media or on solid Difco Middlebrook 7H10 agar supplemented with 0 . 5% glycerol , OADC , and 0 . 05% tyloxapol . Selection of BCG transformants was accomplished by supplementing liquid or solid media with 25 or 40 μg/ml kanamycin , respectively . Liquid cultures less than 50 ml were grown at 37°C and shaken at 120 rpm . Liquid cultures greater than 50 ml were expanded to no more than 250 ml in one liter roller bottles and rotated at 6 rpm . Large-volume accession lots of vaccine strains were frozen in 1 ml aliquots at a concentration of OD600 1 ( ~108 CFU/ml ) . Mycobacterium ulcerans strain 1615 was kindly provided by Dr . Pamela Small ( University of Tennessee ) and was grown in Middlebrook 7H9 or 7H10 as above at 32°C . For sequencing and manipulation of plasmid DNA , DH5α Escherichia coli ( E . coli ) was grown in lysogeny broth ( LB ) or on agar plates supplemented with appropriate antibiotics . The M . ulcerans Ag85A open reading frame with its endogenous signal sequence was amplified from genomic DNA using the following primers engineered with NdeI and EcoRV restriction sites: forward , 5’-GAGACATATGAAGCTTGTTGACAGGTTTCGTGGC-3’ , reverse , 5’-GAGAGATATCGGCGCCCTGGG TGTCACC-3 . ’ MU-Ag85A amplicons were cloned into the pMV261 backbone , where antigen expression was under the control of the constitutive mycobacterial hsp60 promoter and plasmid selection was mediated by kanamycin resistance ( henceforth , pSL401 ) . The hemagglutinin ( HA ) epitope was fused to the C-terminus of MU-Ag85A . Electrocompetent BCG and M . smegmatis cells were prepared by pelleting log phase liquid cultures ( OD600 0 . 6–0 . 8 ) at 3000 rpm for 10 minutes and washing three times in 10% glycerol with 0 . 05% tyloxopol . Mycobacteria were electroporated using 0 . 5 μg plasmid DNA and recovered in Middlebrook 7H9 media at 37°C overnight . Accession lot of vaccine strains underwent quality control to assess antigen expression , absence of contamination , and retention of plasmid DNA . For immunoblotting , bacterial lysates were prepared by pelleting 10 ml of log-phase culture at 3000 rpm for five minutes . Pellets were washed by repeated centrifugation and resuspension in 10 ml PBST . The final pellet was resuspended in 200 μl lysis buffer with glass beads and vortexed for three minutes . SDS PAGE gels were loaded with a mixture of 15 μl clarified lysate with Laemmli buffer and run under 130V for one hour . Protein was subsequently transferred to PVDF membranes by electrophoresis under 30V for one hour . Membranes were blocked by shaking in 5% milk dissolved in TBS with 0 . 1% tween ( TBST ) at room temperature for one hour . HRP-conjugated mouse anti-HA ( clone 3F10 , Roche ) antibodies were diluted 1:1000 in 5% milk-TBST and incubated with membranes for one hour at room temperature . After washing , proteins were detected via chemiluminescence ( Lumi-light , Roche ) and exposure of X-ray film . For plasmid sequencing , plasmid isolation was performed by incubating mycobacterial pellets with Qiagen Miniprep buffer P1 and lysozyme at 60°C for one hour . This was followed by the manufacturer’s protocol . Eluted plasmid DNA was then heat shocked into chemically competent E . coli DH5α and plasmid DNA was isolated from the resulting transformant colonies . The presence of the correct plasmid insert was assessed by gel electrophoresis of NdeI/EcoRV restriction digests . Plasmid inserts were also sequenced and analyzed using Clone Manager ( Sci-Ed ) software . For the assessment of contamination 100 μl of thawed accession lot material was spread plated on chocolate agar and incubated at 37°C for up to two weeks . For priming , C57BL/6 mice were vaccinated subcutaneously in the scruff of the neck with 100 μl ( ~107 cells ) from a thawed vaccine accession lot vial of BCG-empty vector or recombinant BCG-MU-Ag85A . Eight weeks post-prime , mice were boosted retro-orbitally with 100 μl ( ~107 cells ) of M . smegmatis-MU-Ag85A and were then challenged intradermally via the footpad with 105 M . ulcerans 1615 at two weeks post-boost . Vials of MU1615 challenge material were consistently pulled from the same accession lot and were fully virulent by testing pathology in mouse footpad models . At two to three week intervals , the height and width of footpads from infected mice were measured with digital calipers for signs of swelling . To comply with IACUC protocol , mice were euthanized once height of swelling exceeded 4 . 5 mm ( prior to visible ulceration ) , in order to prevent animal suffering . Levels of Ag85-specific CD4+ T cells were assessed by tetramer staining and flow cytometric analysis . At weekly time points after BCG inoculation , blood samples were collected retro-orbitally and peripheral blood mononuclear cells ( PBMCs ) were isolated by gradient centrifugation through 1 ml of Lympholyte M ( Cedarlane ) . The resulting PBMC layers were removed , washed in 10 ml PBS , and were resuspended in 2 ml of ACK lysis buffer for three minutes to eliminate erythrocyte contamination . PBMC pellets were then washed and resuspended in APC-conjugated M . tuberculosis Ag85B-MHCII tetramer ( 1:500 , NIH Tetramer Core Facility ) diluted in flow buffer ( 2% FBS in PBS ) . This tetramer recognizes a 15 amino acid epitope ( FQDAYNAAGGHNAVF ) with high sequence homology to M . ulcerans Ag85A . The staining proceeded for 30 minutes at 37°C followed by addition of 1:500 FITC-conjugated anti-mouse CD4 ( clone GK-1 . 5 , Biolegend ) and 1:200 PE-Cy5 anti-mouse CD8 ( clone 53–6 . 7 , Biolegend ) for 30 minutes on ice . The PBMCs were subsequently washed in 3 ml of flow buffer , followed by resuspension in 4% paraformaldehyde . Samples were fixed for 30 minutes before flow cytometric analysis using a BD LSRII and FlowJo analysis software ( Tree Star Inc . ) . For central and effector memory cell staining , the above tetramer protocol was combined with additional antibody incubations , including 30 minutes on ice using APC-Cy7 anti-mouse CD4 ( clone GK-1 . 5 ) , PE-Cy7 anti-mouse CD44 ( clone IM7 , BD Pharmingen ) , and FITC anti-mouseCD62L ( clone MEL-14 , BD Pharmingen ) . At weeks 5 and 12 post-challenge , mice were euthanized and infected footpads were removed for disinfection and homogenization . Footpads were disinfected by submerging in 70% ethanol for five minutes and subsequently washing three times with PBST . Footpads were then were finely minced using surgical scissors and placed in mortar and pestles for homogenization . These homogenates were then subjected to N-acetyl-L-cysteine ( NALC ) /NaOH disinfection by combining with a 50:50 mixture of 4% NaOH and 2 . 9% sodium citrate containing 1% NALC . Disinfection was allowed to proceed for 20 minutes at room temperature . Homogenates were then pelleted at 3000 rpm for 5 minutes , washed three times in PBST , and were filtered through a 40 μm mesh . Glass microscope slides were marked with a 0 . 79 cm2 circle and 5 μl of filtrate was spread evenly within this area . After drying the smear , slides were heat-fixed by passing briefly over a Bunsen burner flame and then were allowed to cool . Smears were stained using auramine-rhodamine ( BD Bioscienes ) for three minutes , followed by a rinsing in water , destaining using acid alcohol , counterstaining with potassium permanganate , and rinsing again . Dried slides were viewed at 100x oil immersion under a Nikon X . Acid-fast bacilli ( AFB ) were enumerated within four random fields of view ( FOV ) per animal ( 16 images total per vaccinated group ) and total AFB were calculated by multiplying counts by the numbers of 0 . 038 mm2 FOVs under the 100x lens per marked smear area per microliter of filtrate applied . 96-well PVDF plates were equilibrated with 70% ethanol , washed with PBS , and coated with 1 μg/ml capture anti-mouse IFNγ antibody ( clone AN18 , Mabtech ) overnight . Mice that had received the full vaccination scheme described above were euthanized two weeks after boost to isolate splenocytes in RPMI complete ( RPMI with L-glutamine and 10% fetal bovine serum ) . After blocking plates with RPMI media , 6 x 105 splenocytes were added to each well along with various agonists: ConA positive control ( 100 μg/ml ) , MU-Ag85A peptide ( 100 μg/ml ) , or heat killed M . ulcerans ( 1 mg/ml ) . Following a 16 hour stimulation at 37°C , the plates were washed with PBS + 0 . 05% tween 20 and 1:1000 secondary anti- mouse IFNγ antibody ( clone R46A2 , Mabtech ) was added for two hours at 37°C . The plates were washed again before addition of VectaStain avidin peroxidase complex ( Vector Labs ) for one hour at room temperature . 3-amino-9-ethylcarbazole substrate in acetate buffer was added for five minutes until the reaction was stopped by submerging plates in deionized water . Plates were dried overnight and spots were enumerated using a CTL Immunospot plate reader . This study was reviewed and approved by the Duke University Institutional Animal Care and Use Committee ( IACUCU protocol A065-13-03 ) . Duke IACUC protocols adhere to the USDA , AAALAC , Animal Welfare Act , Guide for Care and Use of Laboratory Animals and Public Health Service Policy on Humane Care and Use of Laboratory Animals . In order to express the immunodominant Ag85A from M . ulcerans ( MU-Ag85A ) in recombinant mycobacterial vectors , M . bovis BCG ( BCG ) and M . smegmatis ( Msmeg ) , electrocompetent cells were transformed with pSL401 ( Fig 1A ) . This replicating plasmid drives transcription of MU-Ag85A containing a C-terminal fusion to hemagglutinin ( HA ) using the strong , constitutive hsp60 promoter . Mycobacterial transformants were selected by plasmid-encoded resistance to kanamycin and replication of plasmids was regulated by oriM in mycobacteria , and by oriE in E . coli shuttle strains . Large-volume vaccine accession lots of transformants were frozen and a quality control protocol was used to assess antigen expression , vaccine lot purity , and plasmid sequence integrity . First , three random vials of each lot were thawed for Western analysis of MU-Ag85A expression using anti-HA antibody . Fig 1B displays the expected 35 kDa molecular weight band for MU-Ag85A expressed in BCG ( BCG MU-Ag85A ) . Expression of the antigen in Msmeg appears to generate an additional larger band , which was not a result of plasmid mutation upon sequencing . Accession lots of BCG and Msmeg transformed with empty-vector DNA , BCG pHA and Msmeg pHA , were also generated to be used as negative controls in subsequent studies . As expected , lysates from these lots did not yield anti-HA reactive bands on the same immunoblot . Due to potential instability of recombinant inserts when expressed in mycobacteria , an analysis of plasmid integrity was performed on both the BCG MU-Ag85A and Msmeg MU-Ag85A lots . Re-isolated pSL401 plasmid DNA from three thawed accession lots vials was examined for appropriately sized MU-Ag85A inserts and sequence . Ten out of ten plasmid transformants from each vaccine lot contained the correct MU-Ag85A insert band , suggesting a high proportion of recombinant mycobacteria retained the plasmid ( Fig 1C and 1D ) . Sanger-sequencing of these re-isolated plasmids was performed to ensure all transformants also contained the correct MU-Ag85A sequence devoid of extraneous mutation . A final aspect of the quality control procedure was to identify any contaminating microorganisms present in the accession lots which could confound subsequent studies . No contamination was observed after incubating thawed accession lot material on both chocolate agar and mycobacterial growth medium for several weeks . Upon establishing the integrity of stable antigen expression and both plasmid and culture purity , BCG MU-Ag85A and Msmeg MU-Ag85A vaccine lots were ready for further immunological and protection studies . The production of CD4+ T cell-mediated responses is vital for anti-mycobacterial immunity [31 , 32] . To determine if vaccination with BCG MU-Ag85A could generate an antigen-specific adaptive immune response , C57BL/6 mice were primed with 107 bacilli by intravenous injection . At intervals ranging from 1–8 weeks post-prime , peripheral blood samples were collected for isolation of circulating lymphocytes . Flow cytometric analysis of MHCII tetramer staining was used to identify the percentage of CD4+ T cells which recognized Ag85A . As seen in Fig 2A , BCG MU-Ag85A induced significantly larger populations of Ag85A-specific helper T cells compared to BCG-pHA vaccinated or unprimed mice . Background tetramer-positive T cell populations were most likely observed during BCG pHA priming due to endogenous Ag85A expressed by BCG . Over the course of initial BCG prime , populations of Ag85A-specific T cells peaked at 2 weeks and waned to pre-vaccination levels by week 8 . Previous reports have determined that multiple doses of BCG are not effective at boosting primary responses in humans [26] . In order to determine if a heterologous mycobacterium could be used to boost initial T cell proliferation , an intravenous dose of 107 Msmeg MU-Ag85A was administered at 8 weeks post-prime . One week post-boost , circulating Ag85A-specific T cells increased over 3-fold and remained significantly higher compared to BCG-primed mice . Additionally , maximal levels of Ag85A-specific T cells were maintained for 3-fold longer following the boost than after priming . Establishment of memory CD4+ T cell reservoirs has been shown to be important for development and downstream efficacy of anti-mycobacterial vaccines [34 , 35] . To examine the level of antigen-specific memory cells induced by vaccination with BCG MU-Ag85A , C57BL/6 mice were primed with 107 bacilli retro-orbitally . Three weeks later , peripheral lymphocytes were stained for CD4 and the T cell memory markers , CD62L and CD44 . Priming alone with either BCG pHA or BCG MU-Ag85A halved the naïve ( CD62LloCD44lo ) population of cells ( Fig 2B ) while increasing the total population of CD4+ T cells by 20% ( Fig 2C ) . As predicted by the Ag85 MHCII tetramer experiment in Fig 2A , the levels of antigen-specific memory cells were significantly higher in BCG MU-Ag85A vaccinated mice , with both effector and central CD4+ memory populations increasing by 1 . 7 and 2-fold respectively . Together these data suggested that utilization of BCG MU-Ag85A as a vaccine prime or Msmeg MU-Ag85A as a boost is an effective regimen for producing high levels of CD4+ helper T cells and memory populations capable of recognizing an immunogenic MU protein . Many studies have highlighted the requirement of Th1 IFNγ-producing responses for inducing successful anti-mycobacterial immunity ( 27–33 ) . In order to determine if the antigen-specific T cells produced during vaccination could generate these types of responses , C57BL/6 mice were intravenously primed with recombinant BCG MU-Ag85A for 8 weeks , then boosted with Msmeg MU-Ag85A . As a control , a group of mice was also primed with Msmeg MU-Ag85A , and then homologously boosted with the same strain . Two weeks post-boost , the mice were euthanized and splenocytes were harvested for in vitro stimulation with either MU-Ag85A peptide or heat-killed M . ulcerans 1615 ( HKMU ) . Enzyme-linked immunospot ( ELISPOT ) was then used to quantify the number of functional Th1 splenocytes capable of producing IFNγ in response to MU antigens . Fig 3 displays the number of IFNγ+ spot-forming units ( SFU ) counted 24 hours after agonist stimulation . Both BCG pHA and Msmeg pHA vaccinations were capable of generating responses , however , priming with BCG MU-Ag85A yielded significantly greater numbers of IFNγ+ splenocytes . The strongest responses were observed when splenocytes were stimulated with HKMU , against which priming with BCG MU-Ag85A increased SFU over BCG pHA by 2 . 5-fold . Similar trends were observed using Msmeg MU-Ag85A as a prime , although overall responses were much lower than those observed using BCG strains . These data suggested that expression of MU-Ag85A in the BCG background increases responsiveness of functional Th1 cells not only to Ag85A , but also to other antigens contained within M . ulcerans bacteria . Previous studies have shown that BCG vaccination in mice delays the onset of MU-induced pathology in the footpad model of Buruli ulcer [22 , 36] . To determine if vaccination with BCG MU-Ag85A could deliver improved control of MU infection compared to standard BCG vaccination , we carried out a direct comparison in the footpad challenge model . To recapitulate a more physiologically relevant vaccination route as used in humans , these groups of mice were subcutaneously primed . A 107 subcutaneous does of recombinant mycobacteria was chosen because it was found to be superior over lower does in protection against MU challenge ( S1 Fig ) . Following intradermal challenge with 105 virulent MU1615 , the height and width of challenged footpads were measured using digital calipers over the course of infection . After footpad vertical swelling reached 4 . 5 mm , mice were euthanized to eliminate suffering . In C57BL/6 mice footpad swelling reaches its peak after an average of 6–8 weeks post-challenge , although swelling can readily be observed by week 3 post-challenge ( Fig 4A ) . Upon priming with BCG-pHA , the average peak swelling was reached at 12–15 weeks post-challenge , however , the use of BCG MU-Ag85A as a prime extended the time required for manifestation of peak swelling to 25–30 weeks post-challenge . In contrast , priming with Msmeg pHA afforded no protection against footpad swelling and the use of Msmeg MU-Ag85A only showed a slight trend toward delay in footpad swelling ( S2 Fig ) . Mycolactone production is known to contribute to the histological observations of necrosis within MU-infected tissues [37] . MU-infected footpads were collected for histopathological analysis to assess internal tissue damage in unprimed or vaccinated animals at 12 week post-challenge . Fig 4B displays representative images from H&E stained tissue sections from an unprimed , BCGD pHA-primed , or a BCG MUAg85A-primed mouse . In unprimed mice , footpads consistently showed ulcerative loss of epidermis with micro-hemorrhaging , as well as large areas of internal necrosis and general inflammatory infiltrate . These features were rare amongst BCGD MUAg85A-primed footpads , which was consistent with the reduced swelling observed at this time point . BCG pHA-primed footpads exhibited an intermediate phenotype , where prominent edema replaced necrotic features and was concordant with the level of swelling observed at week 12 post-challenge . To visualize and assess organization of M . ulcerans in vivo , Ziehl-Neelson staining was also performed on tissue sections from infected footpads of vaccinated or unvaccinated mice . Granulomatous lesions and large masses of extracellular acid fast bacilli ( AFB ) were observed in unprimed footpads , while groups of extracellular AFB were detected in BCG-primed mice ( Fig 4C ) . However , AFB could not be detected in the BCG MU-Ag85A-primed footpad sections available for ZN staining suggesting a comparatively lower level of overall bacterial burden in these tissues . The delay in footpad swelling and reduction in bacterial burden induced by BCG vaccination has previously been shown to temporarily protect mice against the need for euthanasia after MU challenge [22 , 36] . To determine if BCG MU-Ag85A could enhance the protective phenotype observed over BCG vaccination , C57BL/6 mice were subcutaneously primed with BCG or recombinant BCG alone , or BCG- primed and heterologously boosted with Msmeg MU-Ag85A as before . Ten weeks later , mice were challenged with an intradermal footpad injection of high dose ( 105 bacilli ) virulent MU 1615 . Once infection induced vertical footpad swelling surpassing 4 . 5 mm , mice were euthanized . Fig 5A displays survival curves ( time-to-euthanasia ) for mice left unprimed or primed with either BCG pHA or BCG MU-Ag85A . While priming with BCG pHA could significantly increase protection over unprimed mice ( p = 0 . 002 ) , strikingly , a single subcutaneous vaccination with BCG MU-Ag85A further significantly increased survival in mice challenged with virulent MU 1615 ( p<0 . 0001 ) . Importantly , compared to the BCG prime alone , use of BCG MU-Ag85A was significantly more efficacious against experimental Buruli ulcer in mice ( p<0 . 002 ) . Furthermore , boosting BCG MU-Ag85A with an injection of Msmeg-MUAg85A at week 8 post-prime appeared to enhance the protective efficacy of this vaccine , even when compared to unprimed mice or BCG pHA-primed mice that also received a boost ( Fig 5B ) ( p<0 . 03 ) . Interestingly , unprimed mice which received a dose of Msmeg MU-Ag85A did significantly increase protection over unboosted , unprimed mice to a small degree ( p<0 . 05 ) , however , administration of the Msmeg MU-Ag85A boost did not statistically significantly increase protection in BCG pHA or BCG MU-Ag85A-primed mice compared to primes which did not receive a boost ( p<0 . 07 and p<0 . 1 , respectively ) . Significant protection was not observed when the Msmeg background itself was used as a prime , regardless of recombinant insert or boosting ( S2 Fig ) . Overall these data suggest that the anti-mycobacterial immune response generated by increased proliferation of functional and antigen-specific CD4+ T cells can contribute to decreased swelling , reduced bacterial burdens , and an overall greater lifespan for MU 1615 infected mice . Previous studies of experimental Buruli ulcer in mice have shown a correlation between the degree of footpad swelling and the M . ulcerans bacterial load in infected tissues [38] . To determine if the observed reduction in swelling and enhanced survival associated with BCG MU-Ag85A vaccination correlated with lower bacterial burdens , mice which had received vaccinations and challenges were euthanized for isolation of persistent MU 1615 in the challenged footpads . Infected footpads were removed from groups at 5 and 12 weeks post-challenge and acid-fast bacilli present in smears from filtered footpad homogenates were stained with auramine-rhodamine . Importantly , microscopic evaluation of bacterial load was assessed to ensure similar bacterial counts were achieved when compared to plating of colony forming units ( CFU ) . As seen in S3 Fig , comparable trends were observed between acid fast staining of footpad homogenates and CFU plating . Fig 6A shows the average MU burdens for mice left unprimed or primed with BCG pHA or BCG MU-Ag85A during two time points post-challenge . Representative images of re-isolated MU can be seen in Fig 6B . At both 5 and 12 weeks post-challenge , BCG MU-Ag85A subcutaneous vaccination significantly reduced the footpad bacterial burden compared to unprimed mice by 3 . 1-log and 2 . 3-log , respectively . Importantly however , priming with BCG MU-Ag85A at both time points yielded significantly better protection against bacterial replication in the footpads compared to BCG pHA-primed mice , displaying a 0 . 5-log and 1 . 2-log reduction at 5 and 12 weeks respectively . This reduction in bacterial burden correlated well with the differential ability of each vaccine to reduce footpad swelling at similar time points post-infection ( Fig 4A ) and represents a potential mechanism for BCG MU-Ag85A-mediated protection against MU pathology . Buruli ulcer is an insidious disease whose persistence and high morbidity is complicated by both social stigma of those infected and relatively poor access to diagnosis and healthcare in the most afflicted areas . The current standard of care requires lengthy adherence to rifampin and daily intramuscular streptomycin , drugs which are associated with side effects including nephrotoxicity and hearing loss [39] . This , added to the disproportionately high infection rates in children , makes it apparent that a Buruli ulcer vaccine is greatly needed . Previous studies in both humans and animal models have demonstrated that Mycobacterium bovis BCG vaccination affords some level of protection against the pathology observed during MU infection , although total protection has not been achieved . Interestingly , although exposure to MU antigens in the form of DNA vaccination or homologous protein boosting can readily induce antigen-specific T cell and antibody responses in mice , no vaccine has been able to achieve a level of protection better than BCG inoculation . Comparable to tuberculosis , the importance of T cells and Th1 responses to the containment of MU during infection has been established , but the necessary sub-populations , antigens , and cytokine milieu which correlate with protection against Buruli ulcer have not been fully recognized . However , the immune responses and protection developed by exposure to the immunodominant M . tuberculosis mycolyl transferase , Ag85A , in mice and humans have been well characterized [40–42] . Previous studies have shown that DNA-based or rBCG vaccines encoding TB-Ag85A can generate responses which induce antigen-specific IFNγ+ T cell populations , circulating titers of anti-Ag85A IgG , and which reduce virulent M . tuberculosis burdens in murine lungs and spleens [43–46] . In a human clinical trial , use of a recombinant viral vector expressing M . tuberculosis Ag85A was well tolerated as a boosting agent to BCG and induced potent and durable Th1-type responses , although without evidence for increased efficacy [47–49] . Although DNA-based Ag85A primes have been shown to induce protection in experimental tuberculosis , vaccination with rBCG expressing Ag85A appears to have greater efficacy in animal models [50] . Worth noting is the fact that most tuberculosis vaccines currently in clinical trials are designed as boosting agents for BCG , suggesting that BCG is itself a viable and relevant priming agent . This supports the potential for development of improved rBCG priming vaccines not only for tuberculosis but also other important yet neglected mycobacterial diseases . Encouragingly , use of the MU-Ag85A antigen has been demonstrated to develop protection approaching that of BCG in experimental mouse models of Buruli ulcer , and served as a starting point for our recombinant BCG MU-Ag85A approach . BCG’s extensively documented safety in immunocompetent individuals , well-established infrastructure for vaccine administration , relatively low production cost , and previously demonstrated protection make it an ideal candidate for development as an anti-Buruli ulcer vaccine vehicle . Given the prior observations that exposure to MU-Ag85A could afford some level of protection , we generated a quality-controlled , live-recombinant BCG based Buruli ulcer vaccine strain which expressed MU-Ag85A antigen . Using this strain as a prime , a subsequent administration of M . smegmatis expressing MU-Ag85A was used to examine the potential to boost primary vaccination . Two aspects of immunogenicity were examined with this vaccine: induction of antigen-specific CD4+ T cell proliferation and production of IFNγ-secreting cells capable of responding to MU stimulation . Over the course of 8 weeks , BCG MU-Ag85A vaccination significantly increased levels of circulating helper T cells which recognized MU Ag85A over those levels produced by BCG . Upon analysis of MHCII tetramer staining , this proliferative effect could also be boosted for longer than the prime responses , after administration of Msmeg MU-Ag85A . Furthermore , by 3 weeks post-prime , significantly increased levels of antigen-specific effector and central memory CD4+ T cells were observed in BCG MU-Ag85A-vacinated mice versus BCG vaccination alone . Using ELISPOT analysis , mice which were primed with BCG MU-Ag85A produced significantly more IFNγ-secreting splenocytes after stimulation with MU-Ag85A peptide . Importantly , ELISPOT analysis also revealed a significant increase in the splenocyte response to heat-killed MU , suggesting that vaccination could not only increase antigen-specific recognition but also enhance Th1 immune responses to whole M . ulcerans . Vaccine efficacy of BCG MU-Ag85A was also determined by investigating the potential to affect footpad swelling , bacterial burden , and overall survival of MU-challenged mice . Upon subcutaneous priming with BCG MU-Ag85A and intravenous boosting with Msmeg MU-Ag85A , increased time-to-onset of swelling , significantly reduced numbers of footpad acid fast bacilli , and significantly increased time of survival for MU challenged mice was observed . Previous studies have provided evidence for diminished effects of MU infection among BCG vaccinated mice , however , our data reveal that BCG overexpressing MU-Ag85A displayed enhanced protection over BCG at multiple levels of MU pathology . Importantly , a single administration of BCG-MUAg85A alone was sufficient to significantly increase the survival of MU-challenged mice , whereas DNA-based MU-Ag85A vaccines required two prime injections and a homologous protein boost to achieve a similar level of protection observed using standard BCG [25] . Since multiple doses of BCG at 10 or 18 weeks apart were not shown to increase efficacy against experimental BU [26] , we decided to use the relatively innocuous M . smegmatis to generate a boosting agent . Indeed , heterologously boosting a BCG MU-Ag85A prime with M . smegmatis overexpressing MU-Ag85A served to enhance vaccine efficacy to an even greater degree . It is interesting to note that although ELISPOT responses to the M . smegmatis were lower overall in comparison to the BCG background , boosting with Msmeg-MU-Ag85A did increase circulating levels of antigen-specific T cell populations . These elevated helper T cell populations remained in circulation for longer than the BCG prime alone , which may account for the enhancement of protection . Of note was the relative inhibition of bacterial replication in the footpad when comparing infected BCG or BCG MU-Ag85A-primed mice . Interestingly , a more significant 14-fold difference in bacterial load was observed 12 weeks post-infection versus 5-fold at 5 weeks , an effect which appears to synchronize with disparity in survival between those groups at similar time points . Overall , the significant protection against swelling afforded by BCG MU-Ag85A is hypothesized to be due in large part to the reduction in bacterial load , especially in light of an over 1000-fold and 100-fold reduction comparing unprimed to BCG-MUAg85A-vaccinated mice at 5 and 12 weeks post-challenge , respectively . To our knowledge , this is the first study that has identified a Buruli ulcer vaccine candidate which has performed significantly better than standard BCG vaccination alone . As seen with numerous tuberculosis vaccine design studies however , full protection against MU-mediated pathology was not achieved by this regimen . Several factors could explain the lack of full protection , including the insufficiency of available antigens in BCG to elicit appropriate T cell populations necessary for MU containment . An alternative explanation may be associated with the large amount of evidence detailing the immunosuppressive properties of mycolactone produced by MU , whereby reduced cytokine secretion , interference with T cell signaling , and inhibition of inflammatory cell chemotaxis have all been demonstrated as functions of the mycolactone toxin [51–55] . The immunosuppressive properties of mycolactone present in the extracellular milieu may be potent enough to counteract the vaccine-induced cell-mediated responses that would have otherwise been more effective . Use of additional animal models for vaccine efficacy testing could be more suitable in determining mechanisms of protection . The guinea pig has been well established as a model of cutaneous infections and has been used successfully in Buruli ulcer studies [37 , 56] . Guinea pig skin possesses more immunological and structural similarities with human skin compared to the mouse model , and may prove to be an advantageous model for future MU vaccine studies . An alternative approach to vaccine design could be to utilize mycolactone deficient stains of MU as recombinant vaccine vehicles , which may encode antigens capable of inducing robust anti-MU immunity while lacking the virulence of wildtype MU . Such an approach has been investigated using mycolactone-negative MU5114 with limited success , but potentially more immunogenic strains or mutants of mycolactone-deficient MU could provide better protection [36] . In addition , other mycobacterial backgrounds , such as M . marinum , could present an essential antigenic profile unmet by BCG , stemming from substantial sequence homology shared with M . ulcerans [57] . Both of these approaches would undoubtedly require extensive safety profiling for use in humans , a standard already very well achieved by BCG . Despite the lack of full protection by BCG MU-Ag85A , our results lend credence to the possibility of expressing alternative MU antigens , or combinations of antigens , which may increase the protective efficacy of BCG . Others have shown that by expressing combinations of immunodominant M . tuberculosis antigens in rBCG , such as Ag85A and Ag85B together or with ESAT6 and TB10 . 4 , bacterial burdens were further ablated over expression of single antigens [58–62] . Discovery of protective antigens specific to MU in addition to the cross-reactive antigens responsible for the baseline protection conferred by BCG will be essential to making BCG a safe and efficacious Buruli ulcer vaccine .	Buruli ulcer , caused by subcutaneous infection with Mycobacterium ulcerans , is a highly disfiguring flesh-eating skin disease with significant morbidity . Besides surgical intervention , 8-week combination antibiotics is the standard of care . However , problems with resistance and toxicity warrant their replacement with efficacious vaccines . Several attempts to generate a vaccine have met with limited success and , to date , BCG remains the only vaccine capable of conferring transient protection . Here we demonstrate that a recombinant BCG-based vaccine expressing the immunodominant M . ulcerans Ag85A is capable of significantly enhancing protection in experimental Buruli ulcer compared to standard BCG , with a decrease in bacterial burden , pathology , and increase in survival . These results support further Buruli ulcer vaccine development using the highly safe and well-established BCG vehicle .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[]	2015	Recombinant BCG Expressing Mycobacterium ulcerans Ag85A Imparts Enhanced Protection against Experimental Buruli ulcer
Eastern equine encephalitis ( EEE ) virus ( Togaviridae , Alphavirus ) is a highly pathogenic mosquito-borne zoonosis that is responsible for occasional outbreaks of severe disease in humans and equines , resulting in high mortality and neurological impairment in most survivors . In the past , human disease outbreaks in the northeastern U . S . have occurred intermittently with no apparent pattern; however , during the last decade we have witnessed recurring annual emergence where EEE virus activity had been historically rare , and expansion into northern New England where the virus had been previously unknown . In the northeastern U . S . , EEE virus is maintained in an enzootic cycle involving the ornithophagic mosquito , Culiseta melanura , and wild passerine ( perching ) birds in freshwater hardwood swamps . However , the identity of key avian species that serve as principal virus reservoir and amplification hosts has not been established . The efficiency with which pathogen transmission occurs within an avian community is largely determined by the relative reservoir competence of each species and by ecological factors that influence contact rates between these avian hosts and mosquito vectors . Contacts between vector mosquitoes and potential avian hosts may be directly quantified by analyzing the blood meal contents of field-collected specimens . We used PCR-based molecular methods and direct sequencing of the mitochondrial cytochrome b gene for profiling of blood meals in Cs . melanura , in an effort to quantify its feeding behavior on specific vertebrate hosts , and to infer epidemiologic implications in four historic EEE virus foci in the northeastern U . S . Avian point count surveys were conducted to determine spatiotemporal host community composition . Of 1 , 127 blood meals successfully identified to species level , >99% of blood meals were from 65 avian hosts in 27 families and 11 orders , and only seven were from mammalian hosts representing three species . We developed an empirically informed mathematical model for EEE virus transmission using Cs . melanura abundance and preferred and non-preferred avian hosts . To our knowledge this is the first mathematical model for EEE virus , a pathogen with many potential hosts , in the northeastern U . S . We measured strong feeding preferences for a number of avian species based on the proportion of mosquito blood meals identified from these bird species in relation to their observed frequencies . These included: American Robin , Tufted Titmouse , Common Grackle , Wood Thrush , Chipping Sparrow , Black-capped Chickadee , Northern Cardinal , and Warbling Vireo . We found that these bird species , most notably Wood Thrush , play a dominant role in supporting EEE virus amplification . It is also noteworthy that the competence of some of the aforementioned avian species for EEE virus has not been established . Our findings indicate that heterogeneity induced by mosquito host preference , is a key mediator of the epizootic transmission of vector-borne pathogens . Detailed knowledge of the vector-host interactions of mosquito populations in nature is essential for evaluating their vectorial capacity and for assessing the role of individual vertebrates as reservoir hosts involved in the maintenance and amplification of zoonotic agents of human diseases . Our study clarifies the host associations of Cs . melanura in four EEE virus foci in the northeastern U . S . , identifies vector host preferences as the most important transmission parameter , and quantifies the contribution of preference-induced contact heterogeneity to enzootic transmission . Our study identifies Wood Thrush , American Robin and a few avian species that may serve as superspreaders of EEE virus . Our study elucidates spatiotemporal host species utilization by Cs . melanura in relation to avian host community . This research provides a basis to better understand the involvement of Cs . melanura and avian hosts in the transmission and ecology of EEE virus and the risk of human infection in virus foci . Eastern equine encephalitis ( EEE ) virus ( Togaviridae , Alphavirus ) is responsible for outbreaks of severe disease in humans and equines , causing high mortality and neurological sequelae in most survivors [1 , 2] . EEE virus is maintained in an enzootic transmission cycle involving ornithophagic mosquitoes , specifically Culiseta melanura ( Coquillett ) ( Diptera: Culicidae ) , and passerine birds in freshwater swamp foci [1–6] . In the past , disease outbreaks have occurred intermittently with no apparent pattern . Since 2003 , however , the northeastern U . S . and southeastern Canada have experienced a resurgence of EEE virus activity with expansion into new regions [7] . These outbreaks occur when ecological factors and environmental conditions favor virus amplification followed by overflow into human and equine populations . It is widely acknowledged that Cs . melanura feeds predominately on birds; however , the identity of key bird species that may serve as superspreaders of EEE virus has not been established in various virus foci [8–11] . Regional differences exist in the proportion of blood meals by Cs . melanura from various avian species that may be due to the availability and abundance of these birds among other ecological and physiological factors . Vector-host interaction studies conducted in EEE virus foci in the northeastern U . S . have identified >50 bird species as hosts for Cs . melanura , among which Wood Thrush and American Robin were most common [12–14] . Serological surveys also indicate that many of these bird species were frequently exposed to EEE virus [2 , 15] . The percentage of viral antibody was the highest for Wood Thrush , followed by American Robin , Ovenbird , and Swamp Sparrow in studies conducted in New Jersey and Massachusetts [2 , 16] . Successful transmission of EEE virus in avian host communities is governed by the abilities of host species to maintain , amplify , and transmit the virus to mosquito vectors ( mainly Cs . melanura ) , and by ecological factors that influence contact rates between competent avian hosts and the mosquito vector . Earlier studies have identified a number of avian species as hosts for Cs . melanura . However , the potential for vector host preference to affect transmission of EEE virus in multiple foci has not been fully explored [12–14] . In this study , we investigated vector-host contact rates between Cs . melanura and avian hosts by identifying host species from blood meals , and the potential for heterogeneity in vector host preference to influence EEE virus transmission dynamics . The main objectives of this study were to 1 ) quantify vector host-feeding preferences in EEE virus foci , 2 ) identify key bird species that serve as frequent hosts for Cs . melanura , and as reservoir hosts for the virus , and 3 ) determine the extent to which these preferences shape the virus transmission dynamics . To achieve these objectives , we used PCR-based molecular methods and direct sequencing of the mitochondrial cytochrome b gene for profiling of blood meals in Cs . melanura to quantify its contact with vertebrate hosts , and to infer epidemiologic implications of its feeding behavior in four historic EEE virus foci in the northeastern U . S . We conducted avian point count surveys to determine spatiotemporal host community composition experienced by Cs . melanura and EEE virus . Finally , we developed a novel empirically informed mathematical model to describe enzootic transmission of EEE virus in a community of multiple avian hosts and the mosquito vector , Cs . melanura . Field studies were conducted in four historic EEE virus foci , Chester , Killingworth , Madison , and North Stonington , CT ( Fig 1 ) . These four locations were considered to be virus foci because from 1996 to 2014 , EEE virus was detected in 115 pools of mosquitoes , including 74 ( 63% ) pools of Cs . melanura , the enzootic vector for EEE virus . North Stonington had the greatest number of positive pools ( n = 47 , 40 . 9% ) , followed by Chester ( n = 42 , 36 . 5% ) , Killingworth ( n = 13 , 11 . 3% ) , and Madison ( n = 13 , 11 . 3% ) . Majority of the positive pools were identified during 2009 ( n = 56 , 48 . 7% ) , which contributed to the rationales for initiating the present study during 2010–2011 ( Table A in S1 Text ) . In 2010 , four mosquito pools tested positive in mosquitoes from North Stonington but none from the other three sites . Interestingly , positive mosquito pools were identified in North Stonington in 8 years of the nearly two decades during which active mosquito surveillance has been conducted for EEE and other arboviruses in Connecticut . A total of 6 , 234 female Cs . melanura were collected from the four EEE virus foci using 120 resting boxes ( or 11”x11” stackable fiber nursery pots ) placed on dry forested uplands within sight of red maple/Atlantic white cedar swamp habitats , and along the edges of these swamps in 8 sites , Chester 3 , Killingworth 1 , Madison 3 , and North Stonington 1 , during May through October , 2010–2011 , and according to the established protocol [17] ( Table 1 ) . Greater numbers of Cs . melanura were collected during 2011 ( total n = 4390; Chester 1936 , Killingworth 777 , Madison 1129 , and North Stonington 548 ) than in 2010 ( total n = 1844; Chester 548 , Killingworth 545 , Madison 577 , and North Stonington 174 ) . Multiple collection peaks were observed during the trapping season , which suggested 2–3 generations of Cs . melanura each year ( Fig A in S1 Text ) . Resting boxes were examined daily , and a battery-powered handheld aspirator was used to collect engorged mosquitoes . Specimens were transported in coolers containing dry ice to the laboratory . Mosquitoes were then identified to species using a dissecting microscope and an identification key [18] . Specimens with visible blood meals were transferred to 1 . 5 mL microtubes , labeled with a unique number , and stored in an ultra-low temperature freezer . Mosquito abdomens were removed with the aid of a dissecting microscope and disposable razor blades for blood meal analysis . DNA was extracted from the abdominal content of engorged mosquitoes individually by using DNAzol BD ( Molecular Research Center , Cincinnati , OH , USA ) according to the manufacturer’s recommendation with some modifications as described elsewhere [13 , 19] . Extracted DNA from the mosquito blood meals served as DNA templates in subsequent polymerase chain reaction assays with primers based on vertebrate mitochondrial cytochrome b sequences according to published protocols [13 , 19] . Sequencing of both DNA strands was carried out on 3730xL DNA Analyzers , using Big Dye chemistries ( Applied Biosystems Inc . , Grand Island , NY ) at the Keck Sequencing Facility , Yale University , New Haven , CT . Sequences were analyzed and annotated using ChromasPro version 1 . 7 . 5 ( Technelysium Pty Ltd . , Tewantin , Australia ) , and identified by comparison to the GenBank DNA sequence database utilizing the BLAST search ( BLASTN ) of the National Center for Biotechnology Information [20] . Avian point count surveys were conducted in the study sites weekly from April through October , 2010–2011 in order to assess species composition and relative spatial and temporal abundance of bird species , according to the previously described protocols [21 , 22] ( Table 1 ) . A skilled observer knowledgeable of the identification and vocalizations ( i . e . songs and call notes ) of the local bird fauna conducted the point count surveys . Three 100 m-diameter point count circles , located 50–200 m apart , were established at each resting box site with the aid of a Garmin eTerx 10 GPS unit ( Garmin International , Inc . , Olathe , KS ) and a rangefinder ( Bushnell Outdoor Products , Overland Park , KS ) . To estimate bird distances within the point count circles , flagging tape was affixed to trees or shrubs at distances of 15 , 30 , and 50 m . Landmarks within the point count circles were also used to estimate distances . Counts began shortly after sunrise under favorable weather conditions , when bird activity and vocalization are highest . The observer approached the point count with as little disturbance as possible and began counting birds upon arrival at the site . With the aid of binoculars and a stopwatch , the observer recorded the distance and number of bird species seen and heard during a 15-minute period . Status and habitat codes , including observations of nestlings , fledglings , and juvenile birds , and individuals detected outside or flying over the point count circles , were recorded . Temperature and wind speed was measured with a hand-held anemometer ( La Crosse Technology , La Crosse , WI ) and recorded on the form along with the start and finish times ( Fig B in S1 Text ) . Bird nomenclature followed the 7th edition of the “American Ornithologists’ Union” Checklist [23] . The modeling methods are briefly described here , and more detailed information is provided in the “Supporting Information” . The data for each of the four EEE virus foci was pooled to increase sample size and to illustrate general trends . Across the four locations , eight bird species were selected based on their high abundance in bird counts or high prevalence as the source of blood meals , including Wood Thrush ( Hylocichla mustelina ) , American Robin ( Turdus migratorius ) , Tufted Titmouse ( Baeolophus bicolor ) , Common Grackle ( Quiscalus quiscula ) , Chipping Sparrow ( Spizella passerina ) , Black-capped Chickadee ( Poecile atricapillus ) , Northern Cardinal ( Cardinalis cardinalis ) , and Warbling Vireo ( Vireo gilvus ) . The remaining bird species at each location were combined into a ninth category of other birds . Our model simulates the dynamics of EEE virus in each of these eight bird species , the remaining birds and Cs . melanura , over a season of 180 days . We chose this time interval to simulate a typical mosquito activity season . The feeding index was used to model the preference of Cs . melanura for the different bird species [24 , 25] . The feeding index assesses the proportion of blood meals from a host species relative to the abundance of that species in the host community , or the relative likelihood of a blood meal on a given bird species per bird of that species . Thus a feeding index shows the relative preference for one bird species compared to the other species examined . As a result , we chose the ninth category , other birds , as the frame of reference for the other species , assigning its feeding index the value of 1 . We calculated the feeding index for the eight bird species from the bird count and blood meal data collected in this study . The force of infection is defined as the per-capita rate , or hazard , at which susceptible hosts become infected , with a separate force of infection for each bird species and one for the mosquitoes . The number of female mosquitoes , the feeding index and bird abundances determine the relative number of mosquito bites on each bird species . The proportion of infectious mosquitoes and the vector-to-host transmission rate provide the force of infection for the bird species , while the proportions of each infectious bird and the host-to-vector transmission rate give the mosquito force of infection . We assumed that host-to-vector and vector-to-host transmission rates were constant across all bird species , due to a lack of comprehensive information on species-specific transmission rates . Statistical sampling models were used to account for sampling error due to small numbers of bird counts and blood meals for some species . Utilizing Markov chain Monte Carlo methods , 1000 samples were selected for both the counts and blood meals . For each of these samples , the feeding index of the selected bird species were calculated , and the median and inner 95% quantiles were reported . For each of the selected bird species , we used a standard SIR sub-model to simulate the EEE virus transmission dynamics , with the population of each species divided into susceptible , infectious , and recovered compartments . We assumed that vectors do not recover from infection , using a SI sub-model . Each sub-model includes births and deaths , with birth balancing deaths leaving population sizes constant . The number of susceptible hosts increases according to their birth rates , and decreases due to infection and death . For simplicity , we assumed that there was no death due to infection in birds and mosquitoes , so that death is only due to background mortality , and that the population size of each bird species and mosquitoes remained constant throughout the simulation period . Mosquitoes and each bird species transition from susceptible to infectious according to their forces of infection; birds transition from infectious to recovered at a recovery rate of 1 per day , i . e . 1 day mean duration of infectious viremia . The model then simulated for a period of 180 days , starting with all bird species completely susceptible and one tenth of one percent of mosquitoes infectious . Studies suggest that vector-to-host transmission is guaranteed when an infectious vector feeds from a host [26] , and thus the vector-to-host transmission rate was set to 1 . Limited data exists regarding the value of the host-to-vector transmission rate among the various host species . In order to focus the analysis on the effect of variable biting rates on the model's output , the host-to-vector transmission rate is assumed to be constant across all of the host species . The all-species host-to-vector transmission rate was determined by fitting the model proportion of birds that became infected over the model period to the observed proportion of seropositives amongst the various bird species from a previous study [15] . For bird species that were not present in the previous study , the mean proportion of seropositives from that study was used . A total of 1 , 798 Cs . melanura with visible blood meals were collected from the four virus foci , and blood meal sources were identified in 1 , 127 ( 62 . 7% ) specimens by DNA sequencing . These included Chester 348 of 581 ( 59 . 9% ) , Killingworth 249 of 364 ( 68 . 4% ) , Madison 361 of 598 ( 60 . 4% ) , and North Stonington 169 of 255 ( 66 . 3% ) . The remaining blood-fed Cs . melanura either did not produce visible amplification products or the sequencing results were insufficiently conclusive to assign a host species . In addition to Cs . melanura , 372 engorged specimens of 12 species in the genera of Aedes , Anopheles , Coquillettidia , Culex , Culiseta , and Ochlerotatus were collected , and blood meal analyses conducted ( Table B in S1 Text ) . However , because the focus of the present study was on Cs . melanura , the principal mosquito vector of EEE virus , results of these analyses are not presented here . Of the 1 , 127 engorged Cs . melanura blood meals that were successfully linked to a vertebrate host at the species level , 99 . 4% were from avian hosts , comprising 65 species from 27 families and 11 orders . Passeriformes constituted the most numerous hosts representing 97 . 5% of avian blood meals . Comparatively few Cuculiformes ( 1 . 0% ) , Columbiformes ( 0 . 6% ) , Accipitriformes ( 0 . 3% ) , Strigiformes ( 0 . 2% ) , and six other avian orders were additionally identified . Among taxonomic families , Turdidae ( thrushes ) served as the most frequent hosts ( 33 . 4% ) , followed by Paridae ( chickadees and titmice , 18 . 6% ) , Cardinalidae ( cardinals and tanagers , 10 . 9% ) , Icteridae ( blackbirds , 9 . 5% ) , Vireonidae ( vireos , 8 . 0% ) , and 22 other avian families ( Table 2 ) . Four mammalian species belonging to the Cervidae ( White-tailed Deer ) , Bovidae ( Domestic Cow and Sheep ) and Sciuridae ( Eastern Gray Squirrel ) were also identified in individual or mixed blood meals . A total of 37 avian families in 14 orders were encountered at the study sites during the point count survey with differing frequencies calculated over 7 months , April to October . The most frequent avian order was Passeriformes ( 86 . 2% of all birds ) with 20 families including Paridae ( chickadees and titmice , 17 . 7% ) , Icteridae ( blackbirds , 9 . 7% ) , Turdidae ( thrushes , 8 . 8% ) , Parulidae ( wood warblers , 8 . 1% ) , and Emberizidae ( New World Sparrow , 6 . 7% ) . Other relatively frequent avian orders included Piciformes ( [Picidae ( woodpeckers ) ] , 5 . 2% ) , Anseriformes ( [Anatidae ( ducks , geese , and swans ) ] , 3 . 3% ) , and Columbiformes ( [Columbidae ( doves and pigeons ) ] , 1 . 6% ) ( Table G in S1 Text ) . We compared percentage of avian-derived blood meals for Cs . melanura with average avian frequencies in the four EEE virus foci . The eight selected bird species ( Wood Thrush , American Robin , Tufted Titmouse , Common Grackle , Chipping Sparrow , Black-capped Chickadee , Northern Cardinal , and Warbling Vireo ) were found to have a larger feeding index ( Table L in S1 Text ) than the remaining birds , indicating these species were fed upon more frequently by Cs . melanura ( Tables 3–6 and C-F in S1 Text ) . The feeding index was highest for Wood Thrush followed by Warbling Vireo . As a result , both Wood Thrush and Warbling Vireo see early peaks in infections ( Fig 4 ) . The infections in Wood Thrush in turn increase the prevalence of infection in Cs . melanura ( Fig C in S1 Text ) . As a result , the infection rate for some of the remaining bird species , which are less preferred as blood meal hosts , can be seen to increase slightly , allowing EEE virus to persist through the entire season . In order to understand the role that vector feeding preference plays in the amplification of EEE virus , and given the limitations of the collected data , several assumptions were made . Data to estimate the transmission and recovery rates for the host species is limited . As a result , we chose these rates to be the same across the host species . With the assumption that transmission rates do not vary by species , limited sensitivity analysis has shown that species that act as amplifying hosts are insensitive to changes in transmission and recovery rates . If data were available to estimate these parameters separately for each host , then a wide range of substantially different dynamic patterns of infection between host species would be possible . Pooling the samples over time leads to a model that does not include changes in the mosquito or bird populations over the 180-day model period . If the observed seasonal changes in bird species abundance were incorporated into the model , we expect that overall infection rates amongst the bird species would remain about the same , but the time when infections peak in each species could shift depending on the abundance over time . Further investigations are underway to better understand the potential influence of these factors . We have also assumed that the feeding index values remain constant across the 180-day model period . Wood Thrush may be more susceptible to being bitten during their molting period from late July through August; incorporating this shift in biting preference over time into our model would lead to Wood Thrush becoming infected later in the season than in our current results [30] . Several bird species had comparatively few observed bird counts or identified blood meals . In particular , some bird species in some sample periods had 0 observations in the bird count but a positive number of blood meals , which for our simple estimate would give an infinite feeding index in these periods . As a result , these feeding index values are extremely sensitive to small perturbations in the data . This sensitivity can be seen in the confidence intervals for the feeding index for both Wood Thrush and Warbling Vireo , as both species had relatively small observed bird counts , which led to high coefficients of variation in our Poisson sampling model for bird counts . In order to further examine previously excluded host species , more sophisticated sampling models would be needed . Despite these limitations , the model provides valuable insight into the role that vector feeding preferences play in the transmission of EEE virus . In particular we notice that observed bird species with relatively small abundances could drive a majority of the infections across all bird species , due to the amplification of the virus in those species early in the season . In conclusion , we found that Cs . melanura was exposed to diverse avian communities but preferentially focused feeding on Wood Thrush . The model suggests that this species may play a vital role in supporting EEE virus amplification , subject to confirmation that Wood Thrush is a competent host . Culiseta melanura had fed frequently on several other bird species , including American Robin , Tufted Titmouse , Common Grackle , Chipping Sparrow , Black-capped Chickadee , Northern Cardinal , and Warbling Vireo , that were shown to play a less important role in maintaining EEE virus transmission later in the season .	Eastern equine encephalitis ( EEE ) is a highly pathogenic mosquito-borne virus responsible for outbreaks of severe disease in humans and equines , causing high mortality and neurological impairment in most survivors . In the past , human disease outbreaks in the northeastern U . S . occurred sporadically with no apparent pattern; however , during the last decade , this region has experienced changes in the frequency of EEE virus activity with expansion into new localities . We studied vector-host interaction using molecular methods in order to: 1 ) quantify vector host-feeding preferences of Culiseta melanura , 2 ) identify key bird species that serve as frequent hosts for Cs . melanura and reservoir hosts for the virus , and 3 ) determine the extent to which these preferences shape virus transmission dynamics in four historic foci of EEE virus activity in Connecticut . We also examined avian population density and dynamics to determine spatiotemporal host community composition experienced by Cs . melanura and EEE virus . We developed an empirically informed mathematical model to describe enzootic transmission of EEE virus in a community of multiple avian hosts and the primary mosquito vector . To our knowledge this is the first mathematical model for EEE virus , a pathogen with many potential hosts , in the northeastern U . S . Our study clarifies the host associations of Cs . melanura in four foci of virus activity in the northeastern U . S . , identifies vector host preferences as the most important transmission parameter , and quantifies the contribution of preference-induced contact heterogeneity to enzootic transmission . Our study also elucidates spatiotemporal host species utilization by Cs . melanura in relation to avian host abundance . We identified eight bird species ( Wood Thrush , American Robin , Tufted Titmouse , Common Grackle , Chipping Sparrow , Black-capped Chickadee , Northern Cardinal , and Warbling Vireo ) with a larger feeding index than the remaining birds , indicating these species were fed upon by Cs . melanura more frequently , and therefore are important in maintenance of EEE virus at these four foci . This research provides a basis for better understanding the involvement of Cs . melanura in the transmission and ecology of EEE virus , and the risk of human infection in virus foci .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[]	2016	Dynamics of Vector-Host Interactions in Avian Communities in Four Eastern Equine Encephalitis Virus Foci in the Northeastern U.S.
Hybrid incompatibilities play a critical role in the evolution and maintenance of species . We have discovered a simple genetic incompatibility that causes lethality in hybrids between two closely related species of yellow monkeyflower ( Mimulus guttatus and M . nasutus ) . This hybrid incompatibility , which causes one sixteenth of F2 hybrid seedlings to lack chlorophyll and die shortly after germination , occurs between sympatric populations that are connected by ongoing interspecific gene flow . Using complimentary genetic mapping and gene expression analyses , we show that lethality occurs in hybrids that lack a functional copy of the critical photosynthetic gene pTAC14 . In M . guttatus , this gene was duplicated , but the ancestral copy is no longer expressed . In M . nasutus , the duplication is missing altogether . As a result , hybrids die when they are homozygous for the nonfunctional M . guttatus copy and missing the duplicate from M . nasutus , apparently due to misregulated transcription of key photosynthetic genes . Our study indicates that neutral evolutionary processes may play an important role in the evolution of hybrid incompatibilities and opens the door to direct investigations of their contribution to reproductive isolation among naturally hybridizing species . Across diverse taxa , hybrid incompatibilities arise as a byproduct of genetic divergence among incipient species . The basic genetic underpinnings of this process are well understood: two or more mutational differences between species interact epistatically to cause hybrid inviability or sterility [1–3] . However , what is less clear , and often very challenging to uncover , is the nature of the molecular changes and evolutionary forces that lead to hybrid incompatibilities . What sort of mutations are perfectly functional within species but cause reproductive failure or death in hybrids ? Do such mutations accumulate within species by neutral processes or are they positively selected , perhaps providing an ecological advantage or resolving an intragenomic conflict ? Addressing the first of these questions is most straightforward in systems with well-developed genetic tools that facilitate positional cloning , which explains why most progress has been made in traditional models like Drosophila , Arabidopsis , and rice . However , insight into the evolutionary forces acting on hybrid incompatibilities during the speciation process requires a focus on young species pairs with natural populations . Over the past two decades , genetic dissection of diverse incompatibilities has provided some hints about their evolutionary origins ( reviewed in [4–7] ) . Often , hybrid incompatibility genes show molecular signatures of positive selection [8–13] , and there is suggestive evidence that incompatibility alleles can arise through ecological adaptation [14–16] or recurrent bouts of intragenomic conflict [11 , 17–19] . On the other hand , there is also evidence from a handful of cases , all involving gene duplicates , that the evolution of hybrid dysfunction need not involve natural selection [20–22] . The idea that gene duplication might play a key role in hybrid incompatibilities was initially proposed by Muller as a variant of his original model [3] . He explained how gene duplication , followed by degenerative mutations and divergent copy loss , could lead to a difference in gene position between species with missing ( or inactive ) copies acting as recessive incompatibility alleles [3] . This same scenario was emphasized later as an explanation for defects in pollen development between subspecies of Asian cultivated rice , Oryza sativa [23] , and more recently , as a general mechanism of hybrid breakdown via neutral processes [24 , 25] . There have now been three empirical demonstrations of gene duplication/transposition giving rise to interspecific hybrid male sterility; one case involves Drosophila melanogaster-D . simulans hybrids [26] and the other two arise from crosses between O . sativa and wild species [21 , 22] . Gene duplication also causes lethal and sterile combinations that segregate within Arabidopsis thaliana [27 , 28] and O . sativa [20] . However , it is not yet clear whether divergent resolution of gene duplicates contributes to hybrid incompatibilities between wild species in the early stages of divergence . Only by identifying examples in young species pairs , particularly those with sympatric populations and still connected by some degree of gene flow , will it be possible to evaluate the contribution of such loci to speciation . In this study we investigate the molecular genetic basis of hybrid seedling lethality between two closely related sister species of yellow monkeyflower , Mimulus guttatus and M . nasutus . These recently diverged species ( 200-500kya; [29] ) co-occur throughout much of their shared range in western North America , where reproductive isolation between sympatric populations occurs through a number of prezygotic [30–33] and postzygotic barriers [34–40] . Despite substantial reproductive isolation , patterns of shared variation across their genomes indicate historical and ongoing gene flow between the two species [29 , 31 , 41] . Here we focus on sympatric populations of M . guttatus and M . nasutus located at Don Pedro Reservoir ( DPR ) in central California , where both species coexist within centimeters of one another . Species at DPR are strongly isolated by divergence in flowering time and mating system [33]; nevertheless , studies have shown low levels of hybridization [33] and a clear signal of introgression [29] . Using high-resolution genetic mapping and genome-wide expression analyses we identify a duplicate gene pair as the cause of Mimulus hybrid lethality . As the first case of hybrid incompatibility genes identified between naturally hybridizing species , this study opens the door to direct investigations of their evolutionary dynamics and contribution to reproductive isolation . Hybrid lethality occurs in the hybrid progeny of M . guttatus and M . nasutus from the sympatric DPR site and is easily characterized by seedlings that completely lack chlorophyll ( white seedlings , see S1 Fig ) . As a first step toward investigating the genetic basis of this phenotype , we self-fertilized and intercrossed the inbred lines DPR102-gutt ( M . guttatus ) and DPR104-nas ( M . nasutus ) . We then examined phenotypic ratios of white and green seedlings among their selfed progeny and reciprocal F1 and F2 hybrids ( S1 Fig , S1 Table ) . Although we never observed white seedlings in the selfed progeny of parental lines or in F1 hybrids , we discovered that roughly 1/16 of F2 hybrid seedlings were white ( maternal parent listed first: DPR102-gutt x DPR104-nas , N = 516 , 7 . 36% white seedlings; DPR104-nas x DPR102-gutt , N = 661 , 5 . 75% white seedlings ) . Segregation of white seedlings in reciprocal F2 hybrids suggests a nuclear , rather than cyto-nuclear , genetic incompatibility . Chi-squared tests rejected several genetic models that could potentially explain the observed phenotypic ratios , but could not reject a two-locus model involving only recessive alleles in either F2 population , or when their ratios were combined ( S1 Table ) . These results suggest that hybrid lethality between sympatric M . guttatus and M . nasutus is caused by a two-locus , recessive-recessive hybrid incompatibility . To genetically map Mimulus hybrid lethality , we performed two rounds of bulked segregant analysis ( BSA ) . In the first round , we pooled DNA from green and white F2 seedlings into eight separate tubes ( six individuals per pool , four replicates each for green and white ) . Because incompatibility alleles act recessively , our expectation was that pooled white seedlings should be homozygous ( for either DPR102-gutt or DPR104-nas alleles ) at markers linked to hybrid lethality loci , whereas green seedlings should segregate 1:2:1 ( for DPR102-gutt homozygotes: heterozygotes: DPR104-nas homozygotes ) . Of the 126 size-polymorphic markers ( spanning much of the Mimulus genome ) that we used for genotyping , four showed an association with seedling phenotype: the four tubes with white seedlings carried only ( or mostly ) DPR104-nas alleles , whereas green seedlings carried both parental alleles . All four markers map to a region of roughly 40 cM on linkage group 14 ( inferred by marker position in [35] ) , which we named hybrid lethal 14 ( hl14 ) . To identify the partner locus , we performed a second round of BSA controlling for genotype at hl14 . We generated 60 F3 families by self-fertilizing green F2 hybrids that were homozygous for DPR104-nas alleles at hl14 ( determined by genotyping flanking markers ) ; these F3 families segregated green and white seedlings in ratios of either 3:1 or 1:0 . We reasoned that if hybrid lethality is caused by hl14 and a single interacting locus , white F3 hybrids should be homozygous for DPR102-gutt alleles at the partner , whereas green F3 families that do not segregate white seedlings should be homozygous for DPR104-nas alleles . Based on this logic , we formed two separate pools of DNA from F3 hybrids: one with 34 white seedlings and one with 26 green seedlings from non-segregating families . Note that each F3 seedling was derived from a different family ( i . e . , from a unique F2 maternal parent ) so that at markers unlinked to hybrid lethality , both pools should carry each of the two parental alleles at ~50% frequency . For the two pools , we performed whole genome sequencing , generated a genome-wide SNP dataset , and calculated average allele frequency difference in 200-SNP sliding windows ( 100-SNP overlap between windows ) . Using this approach , we discovered that the top 5% most divergent windows were located in contiguous windows along the distal end of chromosome 13 ( S2 Fig ) , which we named hybrid lethal 13 ( hl13 ) . To fine-map hl13 and hl14 , we generated a large DPR104-nas x DPR102-gutt F2 mapping population , oversampling white seedlings to roughly equalize frequencies of the two phenotypes ( white = 44% , green = 56% , N = 2 , 652 ) . Each F2 individual was genotyped with two pairs of markers targeted just outside of the hybrid lethality loci: M208 and M236 at hl13 , and M241 and M132 hl14 . As expected , nearly all white seedlings were homozygous for DPR102-gutt alleles at the hl13-linked markers and homozygous for DPR104-nas alleles at the hl14 markers ( 92% , N = 1174 ) , whereas green seedlings never carried this genotype ( N = 1478 ) ( Fig 1 ) . Because we later discovered that both M208 and M236 are proximal to hl13 , an additional 2 , 182 F2 hybrids were screened with one of two more distal markers ( M263 or M255 ) that allowed us to flank the locus . We genotyped informative recombinants at additional size-polymorphic and SNP-based markers designed in each interval . Although white seedlings must be destructively sampled for DNA , green seedlings were allowed to grow into adult plants so that informative recombinants could be self-fertilized and phenotyped via progeny testing . In this way , we determined if green F2 hybrids were heterozygous for hl13 and/or hl14 ( versus homozygous for compatible alleles ) . Using this strategy , we mapped the hl13 locus to a 72 . 2 kb-region at the distal end of chromosome 13 that contains 24 genes ( Fig 2A ) . At the same time , we mapped the hl14 locus to a 51 . 6 kb-region of chromosome 14 that contains six genes , as well as a gap of unknown size in the M . guttatus IM62 reference genome ( Fig 2B ) . For each hl13 and hl14 candidate gene , we identified its top blast hit ( s ) in Arabidopsis thaliana , gene ontology terms , known mutant phenotypes , and predicted functions ( S2 Table ) . In the hl13 interval , we discovered several strong functional candidates for hybrid lethality including Migut . M02023 , a homolog of pTAC14 ( PLASTID TRANSCRIPTIONALLY ACTIVE CHROMOSOME 14 ) . In A . thaliana , pTAC14 is essential for proper chloroplast development and mutants show a chlorotic lethal phenotype [42] that appears identical to DPR102-gutt x DPR104-nas F2 hybrid lethality . In addition to Migut . M02023 on chromosome 13 , we also identified a highly similar and slightly truncated protein homolog of pTAC14 ( 99 . 1% amino acid similarity along length of truncated homolog ) , Migut . O00467 , located on an unmapped scaffold of the IM62 M . guttatus reference genome ( v2 . 0 scaffold_193 ) . To investigate the possibility that this additional copy of Mg . pTAC14 resides on chromosome 14 , we turned to several large-insert IM62 genomic libraries ( six fosmid and two BAC libraries ) that were generated and end-sequenced as part of the reference genome assembly effort [43] . Among these libraries , only a single end-sequence of one fosmid blasts to v2 . 0 scaffold_193 . Intriguingly , the other end-sequence of this same fosmid blasts to the first exon of Migut . N01489 , a gene within the mapped interval of hl14 . This finding provides evidence that a second copy of Mg . pTAC14 is located on chromosome 14 in IM62 , despite it being absent from the current genome assembly . Using PCR , we confirmed that the DPR102-gutt genome also contains two copies of pTAC14 ( S3 Fig ) . However , despite exhaustive PCR and cloning efforts ( using many different primer combinations ) , we recovered only one copy of pTAC14 from DPR104-nas genomic DNA . To determine if Mimulus pTAC14 duplicates genetically map to hybrid lethality loci , we obtained a set of 96 DPR102-gutt x DPR104-nas F2 hybrids carrying each of the nine possible two-locus genotypes at hl13 and hl14 ( 10 replicates for each green genotype and 16 replicates of the white seedling genotype , see Fig 3 ) . Using a set of conserved primers spanning exons 6–8 , we PCR-amplified and sequenced Mimulus pTAC14 from each of these F2 hybrids . Across this region , 10 SNPs define three distinct haplotypes of pTAC14: “G1” and “G2” from DPR102-gutt and “N1” DPR104-nas ( Fig 3 ) . Remarkably , we discovered a perfect association between pTAC14 haplotype and hl13/hl14 genotype: G1 is present in all individuals with DPR102-gutt alleles at hl13 , G2 is in all individuals with DPR102-gutt alleles at hl14 , and N1 is in all individuals with DPR104-nas alleles at hl13 . From this pattern , we infer that both DPR102-gutt and DPR104-nas carry copies of pTAC14 at hl13 ( hereafter referred to as Mg . pTAC14_1 and Mn . pTAC14_1 , respectively ) , but that only DPR102-gutt carries a copy at hl14 ( referred to as Mg . pTAC14_2 ) . To examine sequence similarity among Mimulus pTAC14 genes , we obtained full-length genomic sequences from both DPR parents and generated a neighbor-joining tree ( Fig 4 , S3 Fig ) . As expected , chromosome 13 copies of pTAC14 from DPR102-gutt and IM62 cluster together ( Mg . pTAC14_1 and Migut . M02023 ) . Likewise , chromosome 14 copies of pTAC14 from DPR102-gutt and IM62 cluster together ( Mg . pTAC14_2 and Migut . O00467 ) . However , somewhat counterintuitively , the DPR104-nas copy ( Mn . pTAC14_1 ) , which is located on chromosome 13 , clusters more closely with M . guttatus copies on chromosome 14 than with copies on chromosome 13 ( Fig 4B ) . Consistent with a causal role for Mimulus pTAC14 duplicates in hybrid lethality , we discovered several lines of evidence that suggest only one of the two copies is functional in DPR102-gutt . First , Mg . pTAC14_1 ( at hl13 ) contains a frameshift mutation in exon 7 , which results in the production of numerous premature stop codons in downstream sequence ( Fig 4A , S3 Fig ) . Second , from each inbred line , we PCR-amplified only a single copy of the gene from leaf cDNA: Mg . pTAC14_2 in DPR102-gutt and Mn . pTAC14_1 in DPR104-nas . Third , pTAC14 expression is nearly absent in white F2 hybrid seedlings , which inherit hl13 from DPR102-gutt ( containing Mg . pTAC14_1 ) and hl14 from DPR104-nas ( containing no copy of pTAC14 ) . Using qPCR and primers that amplify both Mimulus pTAC14 duplicates , we found strong expression in green parental and F2 seedlings , but not in white F2 seedlings ( qPCR on eight additional functional candidates in the hl13 and hl14 intervals showed no association between expression and seedling phenotype; S2 Table , S4 Fig ) . Additionally , we performed RNAseq on DPR102-gutt , DPR104-nas , green F2 , and white F2 seedlings . Consensus sequences generated from de novo assemblies of DPR102-gutt reads that align to Migut . M02023 and/or Migut . O00467 ( high sequence similarity between pTAC14 duplicates means that reads align equally well to both copies ) correspond to Mg . pTAC14_2; consensus sequences generated in the same manner from DPR104-nas reads correspond to Mn . pTAC14_1 . Moreover , RNAseq SNP variation in green F2 seedlings suggests they express only Mg . pTAC14_2 from DPR102-gutt and/or Mn . pTAC14_1 from DPR104-nas . In contrast , read coverage of pTAC14 transcripts in white F2 seedlings is exceptionally low: of the 1 , 092 genes that are significantly differentially expressed between white F2 seedlings and green seedlings ( DPR102-gutt , DPR104-nas , green F2 hybrids ) , the duplicate copies of pTAC14 are the two most underexpressed ( Fig 5 ) . Taken together , these results provide strong evidence that Mimulus hybrid lethality is caused by nonfunctional pTAC14 duplicates: white hybrid seedlings carry unexpressed Mg . pTAC14_1 alleles at hl13 and are missing pTAC14 alleles altogether at hl14 . Comparison of genome-wide RNAseq patterns among DPR102-gutt , DPR104-nas , green F2 , and white F2 seedlings provides additional support for disrupted pTAC14 function as a cause of hybrid lethality . White F2 seedlings show a strong signature of genome-wide misexpression: of 27 , 948 annotated genes , 1 , 092 ( 3% ) are significantly misexpressed in all three pairwise comparisons between white and green seedlings ( Fig 5 ) . Among transcripts that are underexpressed in white seedlings ( N = 209 ) , we found a significant enrichment of genes involved in photosynthesis and/or located within the thylakoid and photosynthetic membranes . Among overexpressed transcripts ( N = 883 ) , we observed an enrichment of heat shock proteins and glutathione peroxidase proteins ( S3 Table ) . Furthermore , consistent with disrupted pTAC14 function , we discovered evidence for severe misexpression of chloroplast-encoded genes in white seedlings ( Fig 6 , S5 Fig ) . In A . thaliana , knockouts of pTAC14 disable the PEP ( plastid-encoded bacterial type ) RNA polymerase , which leads to reduced transcription of some chloroplast-encoded genes , particularly those involved in photosynthesis ( e . g . , photosystem I , photosystem II , and cytochrome b6f ) , and increased transcription of others such as the rpo genes [42] . Of 52 putative Mimulus chloroplast genes , those involved in photosynthesis , and thus likely to be transcribed by PEP RNA polymerase , were often significantly underexpressed in white F2 seedlings . In contrast , homologs of A . thaliana rpo genes were significantly overexpressed . Several additional putative chloroplast genes ( e . g . , ATP synthase , NADH Dehydrogenase , ribosomal proteins ) that are likely transcribed by both PEP and the nuclear-encoded phage-type ( NEP ) RNA polymerase [44–50] were also significantly misexpressed ( both up- and downregulated ) in white seedlings . Taken together , these patterns of gene misexpression in Mimulus F2 white seedlings , which show a remarkable similarity to patterns observed in A . thaliana pTAC14 knockouts [42] , provide additional support for a causal role of pTAC14 duplicates in Mimulus hybrid lethality . Identifying the molecular genetic basis of hybrid incompatibilities between recently diverged , wild species is a critical first step toward understanding their evolutionary origins and role in speciation . We have shown that duplicate copies of Mimulus pTAC14 , a gene critical for chloroplast development in A . thaliana [42] , causes hybrid lethality between sympatric M . guttatus and M . nasutus at the DPR site . We fine-mapped hybrid lethality to hl13 and hl14 , two small nuclear genomic regions on chromosomes 13 and 14 . In DPR102-gutt ( M . guttatus ) , pTAC14 is present in each of these genomic intervals , but only the hl14 copy is expressed . In DPR104-nas ( M . nasutus ) , pTAC14 is present only in the hl13 interval , consistent with either of two possibilities: the hl13 copy is ancestral and this line lacks the duplication , or a large deletion has removed all trace of the gene from hl14 . As a consequence of divergent resolution of these duplicate genes , F2 hybrids that are homozygous for DPR102-gutt alleles at hl13 and homozygous for DPRG102-nas alleles at hl14 contain no functional copy of Mimulus pTAC14 . These hybrids fail to produce chlorophyll and die in the cotyledon stage of development , remarkably similar to what is observed in pTAC14 knockouts in A . thaliana [42] . To our knowledge , this is the first pair of hybrid incompatibility genes identified between naturally hybridizing species . Using complementary genetic mapping and functional genomics approaches , our study provides strong evidence that nonfunctional Mimulus pTAC14 is the cause of hybrid lethality between DPR M . guttatus and M . nasutus . But what causes the lack of pTAC14 expression in white hybrid seedlings ? In M . nasutus ( DPR104-nas ) , because pTAC14 is missing entirely from the hl14 interval , the hl13 copy ( Mn . pTAC14_1 ) is the only one expressed . In M . guttatus ( DPR102-gutt ) , the situation is less clear . Although both copies ( Mg . pTAC14_1 at hl13 and Mg . pTAC14_2 at hl14 ) are present and highly similar in exons ( S3 Fig ) , our qPCR and RNAseq experiments demonstrate that only one of them–Mg . pTAC14_2 –is expressed . Further work will be required to determine the molecular nature of this change in gene expression . The most obvious possibility is that non-sense mediated decay has efficiently targeted Mg . pTAC14_1 , which carries a series of premature stop codons . Another possibility , is that a cis-regulatory mutation disrupts Mg . pTAC14_1 transcription in DPR102-gutt . Alternatively , expression might be prevented by the epigenetic silencing of one duplicate by the other , as was recently shown for sterile and lethal combinations segregating within A . thaliana [28 , 51] . Whatever its cause , disrupted expression is not the only problem with DPR102-gutt Mg . pTAC14_1; this gene copy also carries a 1-bp insertion that , if transcribed , would result in a truncated , and potentially nonfunctional , protein . We do not yet know which of these two functional changes to DPR102-gutt Mg . pTAC14_1 arose first . The evolution of hybrid lethality in this system thus appears entirely consistent with a scenario of duplication and neutral non-functionalization within M . guttatus . Given the ubiquity of gene duplications in plant and animal genomes , divergent resolution of paralogs due to degenerative mutation and genetic drift has been proposed as a major source of hybrid incompatibilities [24 , 25 , 52] . Although initially redundant duplicate genes might sometimes evolve new or partial functions favored by selection [53] , our study and others suggest that duplicates involved in hybrid incompatibilities are more often subject to mutations that disable function in one copy . Within A . thaliana and between closely related Oryza species , divergent resolution of duplicates has occurred through nonsense mutations [20 , 22] and disruptions to expression [21 , 27 , 51] . In a more distantly related species pair of Drosophila , hybrid sterility is caused by a gene transposition , with degenerative mutations having presumably removed any remnant of the duplication that likely preceded its evolution [26] . Remarkably , then , to explain the evolution of hybrid dysfunction in Mimulus and several other diverse systems , there is no need to invoke processes beyond mutation and genetic drift . In addition to showing that Mimulus hybrid lethality is due to nonfunctional pTAC14 , our analyses have begun to provide some insight into the duplication history of this gene . As might be expected , within M . guttatus ( DPR102-gutt and IM62 ) , pTAC14 copies on chromosome 13 are most related and pTAC14 copies on chromosome 14 are most related ( Fig 4 ) . However , somewhat counterintuitively , pTAC14 from DPR104-nas , which is located on chromosome 13 , is most closely related to the M . guttatus copies on chromosome 14 . We interpret this finding , along with the fact that we find no trace of pTAC14_2 at hl14 in DPR104-nas , as evidence that Mimulus pTAC14_1 on chromosome 13 is the ancestral copy . Under this scenario , both the duplicate copy on chromosome 14 ( Mg . pTAC14_2 ) and the M . nasutus copy on chromosome 13 ( Mn . pTAC14_1 ) would have arisen from a similar genetic variant ( Fig 7 ) . Standing genetic variation within and between populations of M . guttatus is high [29 , 54–57] , so it is likely this ancestral populations carried multiple variants of pTAC14_1 . Both the duplicated copy in M . guttatus ( Mg . pTAC14_2 ) and the ancestral copy in the selfing M . nasutus ( Mn . pTAC14_1 ) would be expected to carry only a small subset of ancestral variation . Unfortunately , we have not yet been able to assess molecular patterns of Mimulus pTAC14 variation in a wider sample of M . guttatus and M . nasutus . Although whole genome resequence data are available from a number of lines [29 , 54 , 57] , short-read sequences of Mimulus pTAC14 align equally well to both annotated copies in the IM62 reference genome ( Migut . M02023 and Migut . O00467 ) . Once Mimulus pTAC14 is sequenced from a broader sample of individuals , we speculate that Mn . pTAC14_1 from M . nasutus and Mg . pTAC14_2 from M . guttatus will be included in distinct monophyletic groups nested within the greater diversity of sequences present at the ancestral Mg . pTAC14_1 from M . guttatus . Interestingly , white seedlings are often observed segregating at low frequencies within M . guttatus populations , which manifest as epistatic inbreeding depression [58 , 59] that may be due to divergent resolution of duplicate genes similar to the ones characterized here . However , future studies investigating natural variation for functional and non-functional pTAC14 alleles within species will be required to determine whether the hl13-hl14 incompatibility plays a role in such phenotypes outside of the DPR population . Our study provides the first detailed study of hybrid incompatibility genes from naturally hybridizing species and contributes to a growing body of literature that suggests hybrid incompatibilities might result from neutral evolutionary change within species . Going forward , it will be important to address whether these barriers can persist in the face of ongoing gene flow . Theoretical treatments of this question have consistently concluded that the maintenance of hybrid incompatibility alleles between hybridizing populations relies heavily on a selective advantage within species [60–64] . If so , neutrally evolving hybrid incompatibility alleles might be precluded from affecting reproductive isolation in any more than a transient fashion , with gene flow temporarily constrained until the hybrid incompatibility degrades with time . Nevertheless , other factors such as strong linkage to selected alleles ( e . g . , [16] ) and constraints on gene dosage ( e . g . , [65] ) may play an important role in such incompatibilities . By showing that the duplication of pTAC14 underlies hybrid lethality among sympatric Mimulus species , we now have a natural system in place to test broader questions regarding the evolutionary significance of hybrid incompatibilities in speciation . In this study , we used inbred lines of M . guttatus and M . nasutus derived from wild individuals collected from Don Pedro Reservoir ( DPR ) in central California [33] . Wild-collected seed was sown on moist Fafard 3-B potting soil in 2 . 5” pots , cold-stratified in the dark at 4C for two weeks , and moved to the UGA greenhouses to germinate under 16 hour days at 23°C ( growth conditions were constant across all experiments with the exception of plants used in RNAseq , which were grown in a growth chamber ) . Because M . nasutus is a selfer , the DPR104-nas line used in this study was naturally inbred . To generate the DPR102-gutt inbred line , we self-fertilized M . guttatus for three generations , which is expected to result in a genome that is 87 . 5% homozygous . The DPR102-gutt and DPR104-nas inbred lines were intercrossed to generate reciprocal F1 and F2 hybrids ( maternal parent always listed first in crosses ) . To ensure that both inbred lines were homozygous at hl13 and hl14 , we analyzed white:green seedling ratios in 32 F2 families ( generated from distinct F1 hybrids , 16 from each cross direction ) ; indeed , all F2 families segregated white seedlings at a frequency of one sixteenth , confirming homozygosity at both hybrid lethality loci . DNA was extracted from seedlings and adult leaf tissue using a standard CTAB-chloroform protocol [66] modified for use in a 96-well format . Genotyping was performed using a combination of exon-primed intron-spanning size polymorphic markers containing 5’ fluorescent tags ( 6-FAM or HEX ) and SNP-containing gene fragments that were analyzed through Sanger sequencing . We designed size-polymorphic and SNP markers from polymorphisms observed in whole genome re-sequence data of multiple lines of M . guttatus and M . nasutus [29 , 67 , 68] and confirmed polymorphisms by genotyping parental lines used in our study . A standard touchdown PCR protocol was used in all amplifications and Sanger sequencing reactions were prepared using BigDye v3 . 1 mastermix ( Applied Biosystems , Foster City , USA ) . Genotyping and Sanger sequencing reactions were run on an ABI3730XL automated DNA sequencer at the Georgia Genomics Facility and analyzed using GENEMARKER [69] and Sequencher ( Gene Codes Corporation , Ann Arbor , USA ) software , respectively . For whole genome sequencing of bulked segregants , we generated equimolar amounts of DNA from green ( N = 26 ) and white hybrid seedlings ( N = 34 ) . Green and white DNA was pooled separately and sent to the Duke Center for Genomic and Computational Biology , where Illumina libraries with unique barcodes were prepared and sequenced using the Illumina Hi-seq platform ( 100bp single-end reads ) . Reads from both pools were aligned to the M . guttatus ( IM62 ) reference genome ( https://phytozome . jgi . doe . gov ) , along with previously generated whole genome re-sequence data for DPR104-nas [29] . Reads were aligned using Burrows-Wheeler Aligner ( bwa , [70] ) with a minimum alignment quality threshold of Q29 ( filtering done with samtools , [71] ) . We identified 235 , 922 SNPs that differentiated the IM62 reference genome from DPR104-nas using the samtools mpileup function , which provided a list of SNPs that differentiate these two lineages . We used the samtools mpileup function to estimate the frequency of each SNP ( ‘alternate allele frequency’ ) within white and green BSA pools . Since SNPs were not based on differences between DPR104-nas and DPR102-gutt , our analysis assumes that M . guttatus lines IM62 and DPR102-gutt ( which is not sequenced ) share a common set of SNPs . To examine the potential for transcriptional misexpression associated with hybrid lethality , we performed quantitative PCR on a subset of strong candidate genes within hl13 and hl14 ( 9 genes total , S2 Table ) . We compared gene expression patterns in seedlings from DPR102-gutt , DPR104-nas , green F2s , and white F2s . We extracted RNA from pools of 10 seedlings for each genotypic class using a Zymo MicroRNA Kit ( Zymo Research , Irvine , USA ) followed by cDNA synthesis with GOscript Reverse Transcriptase ( Promega , Madison , USA ) . We designed exon-specific primers to amplify fragments of each gene ( S4 Table ) , amplified fragments using standard touchdown PCR , and visualized gene fragments on a 1% agarose gel . We performed an RNAseq experiment to compare genome-wide expression profiles between white and green seedlings . We used lines of DPR102-gutt and DPR104-nas that had been inbred for 5 generations and their green and white F2 progeny , which resulted in three classes of green seedlings ( DPR102-gutt , DPR104-nas , and green F2s ) and a single class of white seedlings ( white F2s ) . Seedlings with fully expanded cotyledons began to emerge within 3 days and continued to emerge for a week thereafter . We collected pools of 10 seedlings from each biological class directly into 2mL Eppendorf tubes filled with liquid nitrogen . We then extracted RNA from these pools using the Zymo Quick-RNA microprep kit ( Zymo Research , Irvine , USA ) and estimated RNA concentration using a qubit fluorometer ( Life Technologies , Paisley , UK ) . High quality RNA was subsequently submitted to the Duke Center for Genomic and Computational Biology , where Kapa Stranded mRNA-Seq libraries ( Kapa Biosystems , Wilmington , USA ) were prepared and samples were sequenced across a single lane of Illumina Hiseq 4000 with single-end 50 bp reads . In total , our analysis involved three replicates each of DPR102-gutt and DPR104 green seedlings , five replicates of green F2 seedlings , and six replicates of white F2 seedlings , where each replicate was a pool of 10 seedlings . We utilized the cufflinks pipeline [72] to assess patterns of differential expression among the four genotypic classes ( DPR102-gutt , DPR104-nas , green F2s , and white F2s ) . We aligned trimmed and filtered reads ( Q>20 ) to the M . guttatus IM62 reference genome in TopHat2 [73] , which resulted in an average of 19 million reads aligned per biological replicate . We then assembled transcriptomes in cufflinks using default settings , the reference IM62 transcriptome to guide the assembly ( -g ) , and allowing for both fragment bias ( -b ) and multi-read ( -u ) corrections . We used ‘cuffnorm’ to normalize transcript abundance for each genotypic class and ‘cuffdiff’ to calculate differential expression for all pairwise comparisons using default parameters for both program . For data management and sorting , we used Microsoft excel , the R statistical package [74] , and the R package CummeRbund [72] . Gene ontology ( GO ) enrichment analyses were carried out for particular subsets of data that exhibited patterns of differential expression between white and green seedlings . To perform these analyses , we used GOstat [75] and GO::TermFinder [76] implemented in the Phytomine user interface ( https://phytozome . jgi . doe . gov ) . For GO term analyses , we used all annotated genes in the v2 . 0 IM62 M . guttatus reference assembly to serve as the background population and used a Bonferroni cutoff value of 0 . 05 to test for significant GO term enrichment in our subset of differentially expressed genes . For our analysis of chloroplast-encoded genes , we generated a list of putative chloroplast genes , since no chloroplast genome assembly is currently available for M . guttatus . We generated this set by first downloading a list of 135 genes present in the chloroplast genome in A . thaliana from the TAIR database ( www . arabidopsis . org ) . We used this list to identify M . guttatus homologs in the Phytomine database ( https://phytozome . jgi . doe . gov ) . This approach yielded a set of 52 putative Mimulus chloroplast genes that are currently included ( and , presumably , misassembled ) in the nuclear genome ( no homologs were identified for the other 83 genes used in our search ) . A substantial fraction of these genes ( 69% ) occur along chromosome 4 from positions 6 , 719 , 000–7 , 985 , 375 , which contains 183 genes total . To obtain full-length pTAC14 sequences from DPR102-gutt and DPR104-nas , we amplified both genomic and cDNA using primers designed within conserved exonic sequence . PCR fragments were amplified using Phusion High-Fidelity DNA Polymerase ( New England Biolabs , Ipswich , USA ) and either directly sequenced or sequenced after cloning into the TOPO TA Vector ( Thermo Fisher , Carlsbad , USA ) . Additionally , we extracted full-length transcript sequences from RNAseq data by performing de novo assemblies on reads that mapped to candidate genes using the Geneious Assembler ( Biomatters , Newark , USA ) . When reads mapped to duplicated genes , they were combined into a single de novo assembly and 95% confidence consensus sequences were constructed . Using PHYML [77] , we constructed a neighbor-joining tree for pTAC14 using 4 , 319 bp of genomic sequence ( excluding 5’ and 3’ UTRs and insertions/deletions coded as single variants ) with branch support determined with 1000 bootstraps . For the tree presented in Fig 4B , we used the general time-reversible model with four substitution rate categories and allowed the program to estimate the proportion of variable sites and the gamma distribution parameter ( varying these parameters produced identical consensus trees ) .	The evolution of hybrid incompatibilities ( gene interactions that cause hybrids to be sterile or inviable ) is a common outcome of genomic divergence between lineages . However , evaluating the importance of hybrid incompatibilities for speciation requires that we identify the causal genes and evolutionary forces in recently diverged , wild species . We discovered that hybrid seedling lethality between two closely related sister species of yellow monkeyflower is caused by duplicate copies of a gene critical for chloroplast development . Because each lineage carries its one functional gene copy in a distinct genomic location , some hybrids inherit only inactive ( or missing ) alleles . We infer that hybrid lethality in this young species pair has arisen through divergent resolution of gene duplicates by degenerative mutations and ( likely ) genetic drift . These findings are an important step toward understanding the evolutionary dynamics of hybrid incompatibility genes in nature , as well as the role of such genes in species divergence .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "plant", "cell", "biology", "brassica", "chloroplasts", "alleles", "plant", "science", "model", "organisms", "genome", "analysis", "experimental", "organism", "systems", "seedlings", "plants", "cellular", "structures", "and", "organelles", "arabidopsis", "thaliana", "re...	2018	Gene duplicates cause hybrid lethality between sympatric species of Mimulus
Dimensionality is a fundamental component that can have profound implications on the characteristics of physical systems . In cell biology , however , the majority of studies on cell physical properties , from rheology to force generation to migration , have been performed on 2D substrates , and it is not clear how a more realistic 3D environment influences cell properties . Here , we develop an integrated approach and demonstrate the combination of mitochondria-tracking microrheology , microfluidics , and Brownian dynamics simulations to explore the impact of dimensionality on intracellular mechanics and on the effects of intracellular disruption . Additionally , we consider both passive thermal and active motor-driven processes within the cell and demonstrate through modeling how active internal fluctuations are modulated via dimensionality . Our results demonstrate that metastatic breast cancer cells ( MDA-MB-231 ) exhibit more solid-like internal motions in 3D compared to 2D , and actin network disruption via Cytochalasin D has a more pronounced effect on internal cell fluctuations in 2D . Our computational results and modeling show that motor-induced active stress fluctuations are enhanced in 2D , leading to increased local intracellular particle fluctuations and apparent fluid-like behavior . Mechanical properties of cells have important implications in many areas of biology and medicine , from cancer metastasis to blood-borne diseases to cardiovascular functions [1]–[7] . For instance , recent studies have shown that highly metastatic cancer cells tend to generate higher traction forces [1] , [8] and are more deformable [9]–[13] . Understanding intrinsic intracellular mechanical properties , such as internal fluctuations and viscoelasticity , can provide insights toward fundamental functional capabilities of cells , including the abilities to migrate , change shape , and exert and respond to force . There are a number of techniques that have been developed that enables cell mechanical properties to be investigated , including micropipette aspiration [14]–[16] , atomic force microscopy ( AFM ) [17] , traction force microscopy [18] , [19] , optical [9] or hydrodynamic force-based cell stretching [10] , and various forms of particle tracking microrheology [20]–[25] . These techniques have revealed important insights towards the mechanical states of cells . However , recent studies have shown that the cell microenvironment plays a critical role in regulating cell properties and behavior . Effects such as dimensionality , shear flow , interstitial flow , chemokine gradients , co-culture conditions , and matrix and substrate mechanics have all been demonstrated to alter cell migratory behavior , mechanical properties , and signaling [2] , [26]–[40] . While cell mechanics is of considerable importance , it is currently difficult to measure mechanical properties in physiologically realistic , 3D environments . Many techniques can only be applied to cells in 2D or suspended cells . Passive particle tracking microrheology is the most practical technique for this task , as no additional constructs and instrumentation are required besides the ability to visualize intracellular particles . The microrheology technique refers to the tracking of tracer particles and assessing mechanical properties based on the particle motions . In the “passive” case , no external forces are applied and the particle motions are intrinsic to the material [41] , [42] . For cells , those forces result from thermal activity and molecular motors . This is in contrast with “active” microrheology , in which probe particles are externally forced , such as with laser tweezers or magnetic tweezers , and their motions in response to the applied force are tracked [43]–[45] . Additionally , in order to have the ability to control the microenvironment , it is advantageous to perform microrheology in a microfluidic device with easily and precisely tunable inputs , such as interstitial flow and co-culture conditions . In this study , we demonstrate an integrated approach that applies mitochondria-tracking microrheology in a compartmentalized microfluidic device . Tracking the fluctuations of intracellular organelles has traditionally been performed in 2D . Our approach provides flexibility and practicability for studies analyzing the effects of environmental factors , especially in 3D , on intracellular mechanics . We describe the key steps to enable this to be practiced , and we look into important practical considerations , specifically temperature effects and a comparison between ballistically injected nanoparticle tracking and mitochondria tracking microrheology . We then focus on a key environmental factor that modulates cell behavior – dimensionality – and demonstrate its impact on intracellular mechanics and drug-induced effects , specifically cytoskeletal disruption via Cytochalasin D . Notably , we primarily use mitochondria as the tracer particles of interest because they are endogenous and exist in high abundance throughout the cell , enabling a spatial distribution of intracellular mechanical properties to be computed for each cell . This is advantageous over ballistic particle injection microrheology , since particle injection efficiency may be low for some cell types thus resulting in only few traceable particles per cell . The number of traceable particles per cell is reduced further over longer term cultures as cells divide . Additionally , it is unclear if the ballistic injection protocol has adverse or transformative effects on targeted cells , as cells are transiently placed under stressful conditions , vacuumed and pressurized , without media . Finally , mitochondria are important multifunctional organelles that play critical roles in cell energetics , behavior , and apoptosis [46] . Thus , the very fluctuations of these tracers could provide direct insights toward mitochondrial transport and cell bioenergetics [47] , [48] . MDA-MB-231 metastatic breast adenocarcinoma cells were cultured in DMEM ( Dulbecco's Modified Eagle Medium , Life Technologies ) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37°C and 5% CO2 . These cells expressed GFP-actin via Lifeact-GFP ( obtained from Dr . Keiko Kawauchi's lab at the Mechanobiology Institute ( MBI ) , Singapore ) . The microfluidic device platform used in these experiments is shown in Figs . 1 a and b . An SU8 master of the device can be microfabricated as described previously [49] , [50] , and PDMS soft lithography can be performed to generate microfluidic devices . The channels are treated with PDL ( poly-D-lysine hydrobromide; 1 mg/mL , Sigma-Aldrich ) for at least 4 hours and then rinsed twice with sterile water prior to experiments to promote collagen matrix adhesion [49] , [50] . To seed the device ( Figs . 1a-d ) , MDA-MB-231 cells are resuspended into 50 µL of complete growth media . 200 µL of type I rat tail collagen solution ( Becton Dickinson ) ( at a final collagen concentration of 2 . 5 mg/mL and a pH of about 7 . 4 ) with about 5 million MDA-MB-231 cells per mL is prepared on ice . A 200 µL pipettor is used to directly inject about 60 µL of this solution into the central channel of the device . Surface tension and microposts help keep the gel solution localized in the central channel ( Fig . 1b ) . After loading , the device with the loaded cells and gel is placed inside a 37°C incubator for 30 minutes to allow for crosslinking and gelation of the collagen . The device is flipped upside down after 1 minute of incubation , followed by 3 more flips 3–5 minutes apart in order to prevent cell sedimentation to the top or bottom surfaces before the collagen fully gels . After the collagen is fully gelled , complete growth media is pipetted directly into the two side reservoirs and the device is incubated for at least 1 day before experimentation . 2D experiments are performed in tissue culture glass-bottom ( for high resolution imaging ) well-plates without any additional substrate coating . Cells are seeded and cultured in serum containing growth media . MDA-MB-231 cells have been shown to proliferate well on substrates including glass and PDMS in the presence of serum without any other substrate coating [51] , as serum contains adhesion proteins [52] . Fifteen minutes prior to experimentation , the media at the reservoir regions is replaced with new complete growth media with 500 nM added Mitotracker Red solution ( Life Technologies ) . In about 15 minutes , the mitochondria of cells in the device are fluorescently labeled and traceable . The device is placed in an environmental chamber set to ideal culture conditions ( 37°C , 5% CO2 ) . Cells near the middle of the gel are located and fluorescent images are acquired at high temporal resolution ( 50 ms per frame ) for 100 s . IMARIS ( Bitplane , St . Paul , MN ) or other auto-tracking software is used to trace the displacement vs . time curve for each tracer and the mean-squared displacements ( MSDs ) are calculated . Net displacements are subtracted for each tracer to account for drift and persistent motions . All experiments performed in this study were done in an environmental chamber set to 37°C and 5% CO2 unless otherwise specified . At least 9 individual cells are studied in mitochondria tracking experiments in each case for 2D and 3D cultures with and without Cytochalasin treatment over multiple plates or devices , with at least 68 mitochondria tracked in each case . Once the MSDs are computed , post-processing following previous studies to fit MSDs to a locally weighted polynomial and applying an algebraic approximation of the Generalized Stokes-Einstein Relation ( GSER ) can be performed to compute the shear modulus [20] , [41] , [42] , [53]–[56]:where is the shear modulus , s is the Laplace frequency , kB is the Boltzmann constant , T is the temperature , <Δr2 ( t ) > is the MSD ( in 3D assuming an isotropic medium ) , t = 1/s is the time interval , a is the radius of the tracer particle , and Γ is the gamma function . s can be substituted by iω to obtain the complex shear modulus G* ( ω ) = G′+iG″ , where G′ and G″ are the elastic and loss moduli , respectively . Note that due to active , non-thermal , motions inside living cells , the GSER becomes unreliable at low frequencies ( below 10 Hz ) [21] , [45] , [57] , which we consider and discuss in Results . To use nanoparticles as tracers , a ballistic injection step [58] , [59] must be performed prior to cell loading into devices . Briefly , a ballistic particle delivery system ( Bio-Rad Laboratories , Carlsbad , CA ) is used to propel fluorescent 500 nm carboxylated polystyrene particles into near-confluent cells on a 10 cm2 dish . A spinning disk confocal microscope was used with dual lasers for exciting GFP and mCherry . Images were acquired using a 63×1 . 4NA oil immersion objective and an electron multiplying CCD camera ( Hamamatsu Photonics , Hamamatsu , Japan ) . The image plane is focused inside the cell both in 2D and 3D . In 3D , cells in the middle of the gel were chosen to avoid boundary effects . Over the course of image acquisition , which was only 100 s per video , no significant focal drifts were observed . Baseline MSD due to system noise of beads stuck on the surface of a cover glass has been measured to be ∼10−4–10−3 µm2 ( N = 30 ) , which is below our average MSDs for mitochondria motion in our results . This baseline measurement is shown in Fig . S1 in Supporting Information S1 . Cytochalasin D ( Sigma-Aldrich ) disrupts the cytoskeletal network by capping actin filaments [60] . Cells are treated with 5 µM Cytochalasin D , dissolved in dimethylsulfoxide ( DMSO ) and mixed in complete growth media , for at least 1 hour prior to the experiments on cytoskeletal disruption . Error bars are standard error of the mean ( s . e . m . ) and statistical significance is computed from One-way ANOVA tests for p<0 . 05 unless stated otherwise . Simulations are performed as previously described [61]–[63] . Briefly , a 3D domain is generated with fixed or periodic boundary conditions . A fixed boundary is hard , such that particles cannot penetrate through it . A periodic boundary is one where a particle crossing it comes out on the opposing boundary . Actin monomers are polymerized into filaments in the presence of actin binding proteins ( motors and cross-linkers ) . Motors have two arms , each binding to a different filament , and walk towards the barbed end of the filaments , generating tension , and have mechanochemical rates in accordance with the literature [64] . Cross-linkers are mechano-sensitive and have binding and unbinding rates that respond to force . Simulations are run for several hundred seconds and the system is allowed to reach steady-state . We tracked the fluctuations of fluorescently labeled mitochondria in highly metastatic breast adenocarcinoma cells ( MDA-MB-231 ) in the 2D imaging plane ( Fig . 2a ) and showed that on average there is local anisotropic motion ( Fig . 2b ) . This motion is likely due to locally anisotropic mechanical properties within the cell [27] , although probe asymmetries could also lead to this effect . The anisotropy is also apparent in nanoparticle tracking data ( Fig . S2 in Supporting Information S1 ) , suggesting that there is likely anisotropic mechanical properties in the cytoskeleton . To maximize the appreciation of this effect , we rotated the coordinate axes for each probe in each cell such that the variance of the displacements in the two principal orthogonal directions have maximal ratio . As shown in Fig . 2b , the average anisotropy appears to persist for over 10 s , as the 1D MSDs in the two principal orthogonal directions do not overlap . It can also been seen in Fig . 2b that there is substantial heterogeneity in the MSD data of even a single cell . The distribution and average for a single cell , however , does appear comparable to the distribution and average of the bulk data ( of 9 cells ) shown in Fig . S2a in Supporting Information S1 , which suggests that intracellular heterogeneity is a key factor driving heterogeneity in the total data . Thus , considering individual mitochondria fluctuations is important in capturing the heterogeneity in the system . We then applied the Generalized Stokes-Einstein Relation ( GSER ) to compute the shear moduli based on 1D displacement data along the principal directions , as shown in Figs . 2 c and d . The average diameter of the mitochondria was calculated to be 140 nm by assuming that particle MSDs scale as the inverse of particle diameter [65] , and comparing with average MSDs from 500 nm nanoparticle tracking data ( Fig . S2 in Supporting Information S1 ) . Because of the abundance of mitochondria in each cell , spatial distributions of the shear modulus can be mapped , as shown in Figs . 2 e and f ( at 10 Hz ) . The directions of maximum and minimum fluctuations are superposed at each position indicating the local principal directions of anisotropy . Globally , because there were no imposed external forces , there does not appear to be any correlated anisotropy in this cell , as shown in Fig . S3 in Supporting Information S1 . At higher frequencies , the elastic modulus is larger than the loss modulus and the material of the cell behaves more solid-like . At lower frequencies , however , GSER becomes unreliable due to non-thermal effects , primarily active fluctuations induced by molecular motors [21] , [45] , [57] , [66] . It is likely that the magnitude of the shear modulus is underestimated at low frequencies because of active stress fluctuations . We considered this in more detail in the modeling section , and for the remaining experimental results we primarily analyzed properties of the intracellular displacement fluctuations ( MSDs ) rather than computed shear moduli . Next we investigated the impact of dimensionality on mitochondrial fluctuations in the cell . Traditional studies of cell mechanics , including migration and viscoelasticity , are performed on 2D substrates . However , it has emerged from recent studies that cells behave differently when embedded in a 3D microenvironment [67] , which is physiologically more relevant . Here we investigated the impact of dimensionality on intracellular mechanical properties . Our results demonstrate that intracellular fluctuations are different in 2D compared to 3D . Specifically , cells in 2D appear to have larger and more fluid-like fluctuations at longer time scales , as shown in Figs . 3 a and b . At a time interval of 10s , the 2D MSD is over three times larger than the 3D MSD . We then investigated the power-law dependence of the MSDs , as shown in Figs . 3 c and d . The power-law exponent β is time dependent and is larger in 2D compared to 3D . At a time interval of 1 s , β about doubles in 2D from 3D . A power-law exponent of 0 corresponds to solid-like materials and a power-law exponent of 1 corresponds to fluid-like materials [41] , [42] . Thus , 2D cells exhibit more fluid-like internal fluctuations . These results suggest that dimensionality plays an important role in modulating intracellular mechanics . We note here that for 3D experiments , because our microfluidic device is only 200 µm high , cells are always within 100 µm from a rigid boundary ( glass or PDMS ) , which may impact cell behavior . The effects of boundary proximity on intracellular properties will require further studies . In our experiments , we do however select for cells that are surrounded by collagen in all 3 dimensions . In these experiments , in 2D cells were grown on glass and in 3D cells were embedded in collagen I . The difference in substrate composition could potentially alter cell stiffness by modulating cell adhesion or substrate stiffness . Note , however , the cells used here have been shown to adhere and proliferate well on glass ( and other substrates such as PDMS ) without additional coating beyond growth media containing serum [51] because of adhesion proteins in serum [52] . Previous studies have shown that increasing substrate stiffness could lead to an increase in cell stiffness [1] , [68] , [69] . Additionally , inhibiting cell adhesion via integrin blocking may reduce cell stiffness [39] , likely by reducing internal prestress . These effects , if present , will induce a vertical shift in the log-log plot of the MSDs , particularly at short time-scales , for which the GSER is valid . In our results , we found that for short times , the magnitudes of the MSDs are comparable between cells in 2D and 3D , suggesting that the stiffness of the cells in 2D and 3D in our experimental conditions are comparable . The difference emerges at long times , where active ( motor-induced ) motions tend to dominate . This effect has not been previously elucidated based on changing matrix composition alone . These results are also consistent with recent work that demonstrated that inhibiting motor activity via blebbistatin treatment or ATP depletion [45] can alter exclusively the long time characteristics of the MSDs . This suggests that the effect we observed is likely due to a modulation in motor activity , which we further explore in the modeling section . It is noteworthy that directed active transport , defined as highly persistent ( constant velocity ) motions that are not random over long times , of mitochondria is likely negligible . These motions would result in MSD = v2t2 , where v is velocity and t is time , so β would equal 2 . This does not appear to occur based on recent work [45] . We nevertheless subtracted net displacements from our data , as mentioned in the methods , so our analysis should not depend on directed active transport . However , active fluctuations , meaning persistent but random motions due to motors , are likely not negligible [21] , [45] , [57] . If active fluctuations are negligible , we expect the log-log plot of the MSDs to appear flat with a weak power-law dependence on time interval ( small constant positive slope ) , consistent with active microrheology methods [43] , [45] . Active fluctuations manifest in the long time regime in the MSDs , as shown by the increase in β . This increase is consistent with non-thermally driven motions , as demonstrated in other studies [45] , [57] by comparing active and passive microrheology measurements of cytoskeletal networks containing motors . A key goal in understanding cell mechanics in diseases such as cancer is to develop platforms and methods that can screen for therapeutics that modulate cell mechanical properties . In order for effective screening to be performed , however , it is necessary to be reasonably confident that cell responses in the screening assays can reproduce cell responses in physiological environments . Here , our interest is in the effect of the dimensionality of the cellular microenvironment . We have already shown that the mechanical properties of cells , as inferred from MSDs , are different between 2D and 3D environments , which suggests that 2D cell-based assays for cell mechanics may not properly recapitulate 3D physiological cell responses . We next tested whether cytoskeletal disruption would impact cell mechanical properties in different ways in a dimensionality-dependent manner . We disrupted the cytoskeletal actin network with Cytochalasin D , which destabilizes the network by capping actin and inhibiting polymerization [60] , and measured intracellular fluctuations in both 2D and 3D ( Fig . 4 ) . 5 µM Cytochalasin D treatment for 30 min . has been shown to be sufficient in substantially disrupting intracellular mechanics to a steady level [70] , [71] . Our results show that the mechanical response of cells to this treatment appears to be significantly dependent on dimensionality . Cytochalasin D reduces intracellular fluctuations at short time intervals , which suggests an increase in network stiffness , and increases β ( at least in 2D ) at longer time intervals , which suggests an increase in fluidity ( Figs . 4 e and f ) . The net impact of Cytochalasin D appears to be more pronounced at 2D , at least with respect to changes in β . We then considered the morphological impact of Cytochalasin D treatment in 2D and 3D and showed that in both cases , the cytoskeleton is clearly disrupted , as demonstrated by punctate actin aggregation ( Fig . 5 ) . We speculate that a decrease in fluctuations at low time intervals may be due to inclusion of the probes within a region of increased local actin concentration due to actin aggregation . The fluid-like behavior at longer time intervals is likely due to decreased actin network connectivity , since aggregates are formed and the initial homogeneous distribution in the cytosol is collapsed . Additionally , our results show that while the morphological impact of Cytochalasin D is apparent in both 2D and 3D , the measured mechanical effects are more pronounced in 2D . In order to gain insights into the physical mechanisms that may contribute towards dimensional modulation of intracellular behavior , we performed Brownian dynamics simulations of actin networks in 3D and under conditions mimicking “2D” . 3D simulations are conducted in a 3×3×3 µm3 cubical domain with periodic boundary conditions . 2D simulations are conducted in a 3×3×1 µm3 domain with periodic boundaries in the x and y directions but with fixed boundaries in the 1 µm-thick z-direction . Super-resolution imaging studies have shown that the height of the cytoskeleton in 2D is submicron [72] . Typical confocal images of MDA-MB-231 cells in 2D and 3D are shown in Fig . S4 in Supporting Information S1 , which elucidates dimensional effects on cytoskeletal morphology . Our simulation results demonstrate that 2D networks cause the actin filaments to align in the x-y plane and the stress fluctuations ( from mean stress ) in those directions increase; this coincides with a decrease in stress fluctuations in the z-direction , as shown in Fig . 6 . The stress fluctuation distributions are shown in Fig . S5 in Supporting Information S1 , illustrating that the variance of the data over time is larger in the x-y plane in 2D . In our simulations , we specifically considered stress fluctuations because they are the source of motions in the cytoskeleton . When motors are inactive , as in the first 100 s of the simulations shown in Figs . 6 c-e , stress fluctuations are low , since thermal collisions have extremely low persistence . When motors are active , t>100 s , the average stress level is increased , due to motors walking along filaments and generating tension ( prestress ) in the network . Additionally , the stress fluctuates over time as tension is generated and released due to motor walking and crosslinker binding and unbinding , causing internal movements . Displacement fluctuations ( MSDs ) are proportional to the ratio of the internal force fluctuations to the shear modulus of the material , in accordance with GSER ( for the passive case with inactive motors ) and modifications to the GSER that incorporate active sources of internal force [21] , [73] . We also derive a model that relates internal energetics to internal motions later . What we essentially show with our simulation results is that the magnitude of the force fluctuations is different in 2D compared to 3D when all other parameters are the same , i . e . concentration , mechanics , and kinetics of actin , motors , and crosslinkers . Knowing that there is a difference in the source ( force ) that drives internal motions now allows us to explore the implications of this property on the behavior and qualitative features of intracellular movements . Therefore , we next aimed to quantitatively describe how alterations in active internal stress fluctuations , such as from dimensional modulation , can lead to distinct changes in the characteristics of MSDs of intracellular particles . We developed a simple effective temperature model , which we derived in Text S1 in Supporting Information S1 . In this model , non-thermal , motor-induced active stress fluctuations manifest in the following form in the Laplace frequency domain:where Teff is the effective temperature including non-thermal effects , T is the actual temperature , A corresponds to the amount of motor activity per unit time that leads to active stress fluctuations ( and we show in Text S1 in Supporting Information S1 that A is proportional to the added kinetic energy in the system induced by motors ) , s is the Laplace frequency , and s0 is a characteristic frequency below which the effect of motors simply leads to a constant , plateaued scaling of the actual temperature . Essentially , this model states that at relatively high frequencies ( s>>s0 ) , there is an extra frequency dependent term in the GSER . At relatively low frequencies ( s<<s0 ) , the impact of motor activity approaches a plateau . This form factor is consistent with trends observed in previous experimental studies , which also suggest that the experimentally relevant regime is s>>s0 [21] , [66] . For systems that exhibit larger stress fluctuations , A is larger . Using this model , the GSER then becomes:and we can now consider two limiting cases: elastic solids and viscous fluids . For a solid ( or more precisely in this case harmonically bound Brownian particles [74]–[76] , is constant , now denoted as G , and the GSER states:In the limit that s>>s0 , the Laplace transform pair of the MSD in the time domain is:Similarly for a purely viscous Newtonian fluid , , where η is the viscosity , a constant . In this case: which , again for s>>s0 , in the time domain is:We speculate that for a power-law viscoelastic material , the Laplace transform pair of the MSD for s>>s0 may take the forms: ( 9 ) where C and α ( between 0 and 1 ) are constants and Γ is the gamma function . These relations predict the two regimes seen in our experimental data for MSD's and the trend for β's and suggest an explanation for the impact of dimensionality on intracellular mechanics and dynamics . The planarization of the cytoskeletal network in 2D leads to an enhancement in motor-induced stress fluctuations in that plane and thus an increase in the factor A in the equations . In the time domain , an increase in A leads to an earlier onset of the second power-law regime and a corresponding increase in β during those time intervals . The predicted trends from this model for 2D ( larger A ) and 3D ( smaller A ) are shown in Fig . 7 and are in agreement with the experimental data shown in Fig . 3 . Fig . 7c shows the simulated frequency spectrum of the MSD based on the effective temperature model for different levels of motor activity . These trends are comparable to previously observed experimental microrheology studies for various cell types [21] , [66] . In addition to giving mechanistic insights towards the impact of dimensionality on intracellular motions , the Brownian dynamics simulation results shown in Fig . 6 also reveal potential mechanisms driving local mechanical anisotropy in cytoskeletal networks . Anisotropy in viscoelasticity may arise from anisotropy in the alignment of filaments in the cytoskeleton . This is observed in the “2D” simulations . Due to alignment of filaments in the x-y plane in the 2D simulations , the magnitude of the average motor-induced stress is much lower in the z-direction , as shown in Figs . 6 d and e . The amount of internal stress is proportional to the stiffness of the cell , as demonstrated by prior work [77] , [78] measuring both cell traction and rheological properties . Thus , our simulations suggest that anisotropy could arise due to filament alignment and that the direction perpendicular to the plane of filament alignment will exhibit reduced stress . The shear modulus in the direction of alignment should be decreased , assuming that the Young's and shear moduli are proportional and according to the tensegrity model [77] . This is consistent with previous experimental results that demonstrated that intracellular MSDs are larger along the direction of induced alignment of endothelial cells under shear flow [27] . Taken together , our simulation and experimental results suggest that there may be local intracellular network alignment leading to local anisotropy in mechanical properties . We tested for potential differences in mechanical properties of cells in 3D inside an incubation chamber ( 37°C , 5% CO2 ) and at ambient conditions . Our results show that the trends and magnitudes of the MSDs are comparable ( Figs . S2 a and b ) . Additionally , we performed microrheology experiments in the microfluidic device in 3D using ballistically injected nanoparticles ( 500 nm diameter ) . The motion of intracellular particles should scale as the inverse of their diameters for particles larger than the mesh size of the cytoskeleton , which is around 50 nm [45] , [79] , [80] . The trends of the MSDs are comparable between mitochondria-tracking and particle-tracking microrheology ( Fig . S2 in Supporting Information S1 ) , suggesting that mitochondria are effective probes for measuring intracellular mechanics . Differences are likely attributed to probe geometries or alterations in cell mechanics due to bead injection . There may also be mitochondria-specific motors that induce further distinctions and enhance mitochondria fluctuations , although the overall features of the motions of endogenous and ballistically-injected exogenous particles are similar [45] . Mitochondria and other intracellular granules have been previously used in a number of 2D cell microrheology studies [21] , [23] , [27] . Based on our simulations and experimental results , dimensionality and cytoskeletal disruption both have distinct effects on intracellular mechanics . The more sold-like motion of intracellular particles in 3D ( lower β ) in Fig . 3 is likely due to the geometry of the cytoskeleton . In 2D , cells and their cytoskeleton are more planar [72] , whereas in 3D , the cytosolic cytoskeletal network is more isotropic since there is less geometric constraint . Based on our simulations , the result of this geometric difference is that motor activity is enhanced in the imaging ( x-y ) plane in 2D but not in 3D . Enhanced motor activity leads to increased stress fluctuations ( from mean stress ) in the cell , giving rise to increased active internal motions ( MSDs at long time scales ) . Therefore , based on the effective temperature model , the enhancement in stress fluctuations in 2D leads to an earlier emergence of the second power-law ( larger β ) regime in the MSDs . Disrupting the cytoskeleton via Cytochalasin D demonstrated visually the aggregation of actin ( Fig . 5 ) , with localized regions of increased actin concentration . Increased local concentration around the probes will lead to increased local stiffness , which would suppress the MSDs . This is consistent with our experimental results at short time intervals ( Fig . 4 ) . For longer time intervals ( 1 s ) , there is an increase in β for Cytochalasin D treated cells . This may be due to a loss of connectivity in the cytoskeleton due to actin aggregation , such that the network inside the cell is no longer well percolated . This loss of percolation leads to a decrease in global stress inside the cell . This is consistent with previous experiments demonstrating that Cytochalasin D treatment abolishes cell traction forces [71] . A decrease in internal tension and network connectivity leads to networks that have decreased global stiffness and more fluid-like behavior [63] , [77] , [78] , [81]–[84] , since the network becomes more like disconnected aggregates diffusing in the cytoplasmic fluid . While dimensionality is a fundamental feature of all physical systems , its impact on cell behavior is not well understood . This is especially important to consider , since many cell biology and mechanics experiments have been performed in 2D environments , which is not physiologically accurate . Towards that end , we have demonstrated the ability to integrate 3D cell culture in an environmentally tunable microfluidic platform with intracellular particle tracking microrheology , thus illustrating and enabling a practical means to study intracellular mechanics on-chip in 3D . Key advantages include lower volumes of reagents required and a more tunable microenvironment . We then demonstrated the importance of dimensionality in altering intracellular mechanics and when testing for effects in response to cytoskeletal disruption . We found that cells in 3D exhibit mechanical characteristics distinct from 2D , and their mechanical response to cytoskeletal disrupting drugs are different from cells in 2D . Cells in 2D appear to have more fluid-like intracellular fluctuations and their response to Cytochalasin D treatment is more pronounced than cells in 3D . Finally , through Brownian dynamics simulations of active actin networks and an effective temperature model , we showed that dimensionality can impact the magnitude of the non-thermal motor-induced stress fluctuations inside the cell , leading to differences in intracellular dynamics and particle motion . Cytoskeletal remodeling due to actin disrupting drugs such as Cytochalasin D is likely impacted by motor activity which is prominent throughout the cell , and enhanced activity in 2D may be responsible for the more pronounced effects seen in our experiments . Future studies exploring the relation between polymerization dynamics , which are altered by Cytochalasin and intracellular actin regulatory proteins such as Mena [85] , and cytoskeletal tension and morphology in the presence of motor activity , can provide new insights toward the fundamental mechanical state of individual cells as well as disease states in cancer cells . Furthermore , during the metastatic process , cancer cells must undergo many instances in which dimensional modulation plays a role . For instance , as individual cells invade through small pores of the tumor ECM ( especially non-proteolytically ) , intra- and extravasate across tight endothelial layers , and obstruct small microvessels during circulation , they exhibit substantial deformations that can significantly squeeze even the nucleus [34] , [51] , [86] , [87] . These events , which induce geometric confinement , can effectively cause the cell to behave mechanically , at least locally , as though it were constrained in a 2D or even 1D environment , thus altering intracellular fluctuations , transport , and motor behavior . The intracellular space is also crowded and compartmentalized , and diffusion of important molecules such as adenosine phosphates ( ATP and ADP ) is slow ( compared to free diffusion in water ) [88]–[90] . Motor activity and active fluctuations may play an important role in transporting sources of ATP by actively redistributing mitochondria in a fluid-like manner over long time scales . Our results show that dimensional modulation from 3D to 2D shortens the time interval before the onset of fluid-like mitochondrial motions , which may alter ATP-dependent and bioenergetic processes as well as the transport of macromolecules in the cell . Dimensionality , a prominent feature in the biological landscape , has characteristic effects on intracellular mechanics , with implications on internal force generation , transport , and mitochondrial dispersion . Properly recreating the dimensionality of physiological environments in in vitro systems may elicit relevant behavior in pathological processes .	Biomechanical properties at the cellular and subcellular levels are important in providing proper biological functions , from cell migratory capabilities to intracellular transport . Deregulation in these properties can lead to disease states such as cancer metastasis . We develop and demonstrate an integrated experimental and computational approach to study intracellular mechanics . We demonstrate that a key environmental factor , dimensionality , plays a significant role in modulating intracellular mechanical behavior . This is important as typical cell biology and mechanics experiments are performed on 2D substrates , which do not capture the physiological features of 3D matrices and may not induce physiologically accurate cell properties . We further develop an effective temperature model to describe how dimensionality changes intracellular particle motion by altering the activity of molecular motors .	[ "Abstract", "Introduction", "Methods", "Results", "and", "Discussion" ]	[ "biotechnology", "bioengineering", "biology", "and", "life", "sciences", "cell", "mechanics", "biophysics", "biomechanics" ]	2014	Impact of Dimensionality and Network Disruption on Microrheology of Cancer Cells in 3D Environments
Current therapies for cutaneous leishmaniasis are limited by poor efficacy , long-term course of treatment , and the development of resistance . We evaluated if pentavalent antimony ( an anti-parasitic drug ) combined with imiquimod ( an immunomodulator ) was more effective than pentavalent antimony alone in patients who had not previously been treated . A randomized double-blind clinical trial involving 80 cutaneous leishmaniasis patients was conducted in Peru . The study subjects were recruited in Lima and Cusco ( 20 experimental and 20 control subjects at each site ) . Experimental arm: Standard dose of pentavalent antimony plus 5% imiquimod cream applied to each lesion three times per week for 20 days . Control arm: Standard dose of pentavalent antimony plus placebo ( vehicle cream ) applied as above . The primary outcome was cure defined as complete re-epithelization with no inflammation assessed during the 12 months post-treatment period . Of the 80 subjects enrolled , 75 completed the study . The overall cure rate at the 12-month follow-up for the intention-to-treat analysis was 75% ( 30/40 ) in the experimental arm and 58% ( 23/40 ) in the control arm ( p = 0 . 098 ) . Subgroup analyses suggested that combination treatment benefits were most often observed at the Cusco site , where L . braziliensis is the prevalent species . Over the study period , only one adverse event ( rash ) was recorded , in the experimental arm . The combination treatment of imiquimod plus pentavalent antimony performed better than placebo plus pentavalent antimony , but the difference was not statistically significant . Clinical Trials . gov NCT00257530 Leishmaniasis includes a spectrum of diseases occurring throughout Asia , Africa , and the Americas which are caused by infection with Leishmania parasites transmitted by the bite of infected sandflies [1] . Disease manifestations are determined predominantly by the host's immune response and the parasite species [2] . In Peru , the predominant species include the Leishmania ( Viannia ) complexes of L . braziliensis , L . peruviana and L . guyanensis that are all associated with cutaneous leishmaniasis . Mucocutaneous leishmaniasis is caused predominantly by L . braziliensis infection [3] , [4] . There is no vaccine for leishmaniasis and current therapies are limited by poor efficacy , the requirement for prolonged treatment , and increasing development of clinical resistance . The drugs most commonly used include pentavalent antimonials , various amphotericin B lipid formulations and a variety of other drugs used to a lesser extent , including pentamidine , miltefosine , and paromomycin [5] . In Peru , the most commonly used first-line treatment for cutaneous and mucocutaneous leishmaniasis is pentavalent antimony ( meglumine antimoniate or sodium stibogluconate ) with a success rate varying between 60% and 80% [6] . Amphotercin B is typically used in those patients who do not respond to pentavalent antimony . The current standard treatment regimes for cutaneous leishmaniasis all involve monotherapy . The use of combination therapy may improve efficacy , and if toxic drugs can be used at lower levels , improve tolerance . Host immune mechanisms play an important role in the efficacy of anti-Leishmania chemotherapy [reviewed in 7] . An essential component of cell-mediated immunity against Leishmania is the development of a Th1 type response that activates macrophages via IFN-γ to either inhibit or kill the parasite [2] . Activation of the innate immune response is essential for the subsequent development of the Th1 type cell-mediated immune response . Imiquimod is a small molecule that activates Toll-like receptors 7 and 8 ( TLR 7/8 ) on antigen-presenting cells and mediates the production of a variety of cytokines including IFN-α , IFN-γ , TNF-α , IL-1 and IL-12 leading to the induction of enhanced Th1 immune responses [8] , [9] . In addition , it has been demonstrated that imiquimod can directly activate macrophage killing of Leishmania amastigotes in the absence of a T-cell-mediated response [10] . Enhancing the local immune response at the site of cutaneous infection , therefore , may be a logical approach to enhance parasite clearance . We previously reported that combination therapy with imiquimod plus parental pentavalent antimony was more effective than pentavalent antimony alone in patients who had previously failed treatment with pentavalent antimony [11] . The objective of this clinical trial was to determine whether the combination therapy can also be beneficial for cutaneous leishmaniasis patients who have not previously been treated . The study design was a randomized controlled trial where 80 subjects were recruited at two clinics associated with the leishmaniasis clinic at the Instituto de Medicina Tropical ‘Alexander von Humbolt’ - Universidad Peruana Cayetano Heredia ( UPCH ) – Hospital Nacional Cayetano Heredia ( in Lima and in Cusco , Peru ) . Subjects recruited in Lima are typically infected with L . peruviana , L . guyanensis , or L . braziliensis while those recruited at the Cusco site are infected predominantly with L . braziliensis [12] . We therefore planned to analyze the data overall and for the Lima and Cusco sites individually . Each clinic recruited 20 experimental and 20 control subjects and assigned treatment based on a 1∶1 randomization list generated by 3M Pharmaceuticals Inc . Subjects assigned to the control arm received the standard pentavalent antimony treatment plus an application of placebo vehicle cream applied to each lesion 3 times per week . Subjects assigned to the experimental arm received pentavalent antimony plus 5% imiquimod cream identically applied . Study I . D . numbers and corresponding treatment packages were prepared so that both subjects and study investigators were blind to treatment allocation throughout the study . Although diagnostic and clinical procedures were performed in either Lima or Cusco , most subjects were recruited from smaller communities of within a one day's drive ( e . g . Ancash , Churin , Yumpe , Cajatambo , Yauyos , Sicuani , Madre de Dios and surrounding areas ) . Study teams at each clinic comprised doctors , nurses , and technologists , all of whom were supervised by Dr . Llanos-Cuentas . Following the completion of therapy , follow-up visits were scheduled at 1 , 2 , 3 , 6 , 9 and 12 months . The primary outcome was cure , defined as complete re-epithelization with no inflammation . The following inclusion and exclusion criteria were assessed using a pre-screening protocol to determine patients' eligibility . Inclusion criteria: Exclusion criteria: Before enrollment , subjects also underwent an electrocardiogram and blood sampling for limited biochemical analysis ( eg . ALT , AST , total bilirubin , alkaline phosphatase , pancreatic amylase , glucose and creatinine ) . When a bacterial superinfection was suspected at the site of Leishmania infection , subjects received oral or systemic antibiotic treatment ( dicloxacillin or clindamycin ) before entry into the trial ( during the screening period ) . Subjects who met all inclusion/exclusion criteria underwent a limited physical exam to ensure that no mucosal lesion had developed . Borders around each lesion were drawn using a plastic sheet placed on the lesion to accurately document size . Lesions were photographed with a digital camera ( at daylight setting ( no flash ) at a distance of 15 cm , with patient ID code and date clearly indicated beside the millimeter ruler . The photos were stored in a computer at UPCH in Lima . A punch biopsy specimen was taken from the border of the lesion prior to the first treatment and a section placed into media to culture the parasite . The primary outcome was cure defined as complete re-epithelialization with no inflammation assessed at between 1 and 12 months post-treatment . The clinical response was evaluated using standard criteria for assessing the size and severity of the cutaneous lesion including size of lesion , reduced ulceration , re-epithelialization , and inflammation reduction . The criteria used to evaluate the response to treatment have been used for the past several years by the investigators [7] , [8] as follows . At each of the follow-up time points ( ie . at 1 , 2 , 3 , 6 , 9 , 12 months ) , the overall response to therapy was classified as: Improvement , Cure or Failure based on the characteristics listed below: Improvement: Significant reduction in the size of the lesion at the time of evaluation ( Stages M1–M3 ) compared to baseline . Cure: Complete re-epithelialization of the lesion without inflammation ( Stage M4 ) . Failure: Any one , or combination , of the following options: Local side effects of the topical treatment at the application site ( pain , pruritus , erythema and swelling ) were graded as follows: Grade 0 = none; Grade 1 = mild ( easily tolerated ) ; Grade 2 = moderate ( sufficiently discomforting to interfere with daily activities ) ; Grade 3 = severe ( prevents normal daily activity ) . At each follow-up visit after the preceding treatment , a lesion evolution score was assigned based on the least improved lesion in the case of patients with more than one lesion and was recorded in the clinical report form ( CRF ) . This evaluation included measuring the size of the lesion ( s ) and taking a standardized photograph of each lesion . A limited physical examination was also performed , recording all adverse events experienced since the last follow-up visit . The clinical diagnosis of cutaneous leishmaniasis was confirmed in all patients by directly identifying the parasite by smear ( Giemsa staining ) , by culture ( in Novy-Macneal Nicolle media ) , and/or by PCR of the Leishmania minicircle DNA prior to randomization . Leishmania species determination was possible in those cases where the parasite was successfully cultured from the lesion . To differentiate between infections with L . braziliensis , L . peruviana , and L . guyanensis , the mannose phosphatase isomerase isoenzyme ( MPI ) gene was sequenced from the cultured parasites . The MPI enzyme represents the only isoenzyme identified by Multi Locus Enzyme Electrophoresis ( MLEE ) to differentiate these species [12] and it has recently been established that the sequence of the MPI gene could differentiate between infections with these different species [13] , [14] . All parasite species determinations were performed before the treatment blind was broken . Sample size was estimated such that the log-rank test for equality of survival curves would have 80% power to detect a statistically significant difference in proportions cured at three months of at least 32% ( hazard ratio of 2 . 6 ) ( estimates of proportions cured at 3 months were based on previously published data ) [11] . This test assumes a constant hazard ratio over time . Descriptive statistics ( means±standard deviations ( SD ) and proportions ) were calculated to summarize socio-demographic , clinical and epidemiological characteristics of patients at baseline for experimental and control arms . These characteristics included age , sex , study site , region of infection acquisition , occupation and lesion-specific data ( number , location , size , type , duration , presence of adenopathy and presence of bacterial superinfection ) . Intention-to-treat and efficacy analyses were conducted . The intention-to-treat analysis was the primary analysis . In this analysis , missing values at day 20 were treated as failures and any subsequent missing value was assigned the outcome status of the patient's most recent visit . The effect of this decision in dealing with missing values , was examined in a secondary intent-to-treat analysis where all non-cured categories ( missing and not improved/worse ) were considered failures . Additional secondary analyses were performed . This included an efficacy ( per protocol ) approach and a comparison of the proportions of subjects cured at each timepoint using chi-square tests . The log rank test was used to compare the Kaplan-Meier survival curves . The experimental protocol for this study was designed in accordance with the general ethical principles outlined in the Declaration of Helsinki , 2000 and ICH ( International Committee for Harmonization ) guidelines for Good Clinical Practice . Approvals from the ethics review boards of Universidad Peruana Cayetano Heredia UPCH , McGill University , and the National Institute of Health in Peru ( INS-Peru ) were obtained prior to the initiation of the study . The UPCH and INS review committees approved a Spanish version of the consent form . The study investigators explained the nature of the investigation and the risks involved to each patient , or parent of a patient below 18 years of age , prior to recruitment . Written informed consent was obtained from each patient enrolled in the study and the patient/parent was informed that he/she was free to voluntarily withdraw from the study at any time . Program directors from DNDi and a study monitor employed by DNDi periodically reviewed study documentation ( ie . all CRFs and SOPs and corresponding source documents for each patient ) and implementation procedures . The monitoring visits provided DNDi with the opportunity to evaluate the progress of the study , verify the accuracy and completeness of CRFs , resolve any inconsistencies in the study records , as well as to ensure that all protocol requirements , applicable regulations , and investigators' obligations were fulfilled . There was no conflict of interest in this study . 3M Pharmaceuticals Inc . provided the randomized allocation of imiquimod and placebo creams at no cost; however , they did not provide financial support for this study , nor were they otherwise involved in study design , data analysis , interpretation of the data or in the writing of the manuscript . Patient recruitment took place between December 2005 and June 2006; follow-up continued to June 2007 . From the 157 patients screened , 80 patients met the inclusion/exclusion criteria , 78 ( 97 . 5% ) completed the 20-day course of treatment , and 75 ( 93 . 8% ) were followed for the entire 12 months following the end of therapy ( Figure 1 ) . The demographic characteristics of the patients are presented in Table 1 . There were no significant differences between the intervention groups . The principal characteristics of the lesions are presented in Table 2 . On average , each patient had 2 lesions . The mean duration and sizes of the lesions were not judged to be different between the experimental: antimony/imiquimod and control: antimony/placebo groups . There were more lesions on the torso in the control: antimony/placebo group ( n = 14 lesions in 5 subjects ) than in the experimental: antimony/imiquimod group ( n = 0 ) . There were more lesions with bacterial super-infections in the experimental: antimony/imiquimod group ( n = 9 lesions in 4 subjects ) than in the control: antimony/placebo group ( n = 1 ) . There is no evidence to suggest that these differences could have influenced the outcome of the trial . There were no significant differences between the intervention groups with respect to adverse events at the site of application ( Table 3 ) . One patient in the experimental antimony/imiquimod group however had a serious adverse event and was unable to complete the full course of treatment . This was a 39-year-old male who developed a generalized rash on day 9 of treatment , at which time treatment was suspended and he was treated with anti-histamines . After the rash subsided , treatment resumed on day 13 but was discontinued the next day when the rash returned and again treated with anti-histamines to resolve the rash . He was then given only imiquimod cream for the remainder of the 20 days without recurrence of the rash . The rash was considered to be due to the pentavalent antimony . At the 2-month follow-up period , this patient was considered a treatment failure and was given amphotericin B , which resulted in a complete cure with no rash . He was considered to be a treatment failure in the final analysis . The evolution of the lesions by intervention group is detailed in Table 4 . Most missing values occurred at month 2 ( n = 28 ) , but at the 12-month evaluation , outcome data were obtained from 75 of the 80 patients . In the primary intention-to-treat analysis ( Table 5 ) , the overall cure rate was 75% ( 30/40 ) in the experimental: antimony/imiquimod group compared to 58% ( 23/40 ) in the control: antimony/placebo group ( p = 0 . 098 ) at 12 months . In the secondary intention-to-treat analysis ( Table 4 ) , where all non-cured categories ( missing , not improved/worse ) were considered failures , a statistically significant difference was found in cure rates between the two groups at 12 months . We further performed an efficacy analysis where the patients who did not complete the treatment ( one from each group ) and those who were lost during the follow-up ( three from the placebo group ) were not considered in the analysis . In this analysis , the cure rate at the final 12-month follow-up was 77% ( 30/39 ) in the experimental antimony/imiquimod group and 58% ( 21/36 ) in the control antimony/placebo group ( p = 0 . 085 ) . Figure 2 displays the Kaplan-Meier ( KM ) curves depicting time to cure for each group within the 12-month period following enrolment into the study . Relapses ( n = 3 ) were considered failures throughout . Missings were considered as described for Table 5 . The log rank test comparing the KM curves of the two groups revealed a non- statistically significant result ( p = 0 . 174 ) , although the combined therapy group demonstrated a higher probability of cure compared to the placebo group . Patients who did not respond to therapy after 2–3 months follow-up were treated with amphotericin B , and all infected lesions were cured . All of these subjects were considered as treatment failures in the intention-to-treat analyses . Analysis by treatment center ( intention-to-treat , primary analysis ) revealed that , at 12 months , the Lima cohort had a 75% cure rate in the experimental: antimony/imiquimod group compared to 65% in the control: antimony/placebo group ( p = 0 . 490 ) ( Table 6 ) . In comparison , the Cusco cohort had a 75% cumulative cure rate in the experimental: antimony/imiquimod group compared to a 50% cure rate in the control: antimony/placebo group ( p = 0 . 102 ) . Efficacy analysis for the Cusco cohort , removing patients who did not complete therapy or were lost to follow-up , showed a cure rate in the experimental: antimony/imiquimod group of 79% ( 15/19 ) compared to 47% ( 8/17 ) in the control: antimony/placebo group ( p = 0 . 046 ) . The major difference between treatment center appeared to be the lower cure rate in the Cusco control: antimony/placebo group . Baseline characteristics of the lesions from the patients from Lima and Cusco are shown in Table 7 and show that the lesions were , on average , larger in the Cusco cohort . L . braziliensis infections are generally more difficult to cure and have a higher relapse rate than L . peruviana and L . guyanensis infections and could account for the lower cure rate in the Cusco control: antimony/placebo group ( 6 , 15 ) . Species determination , performed as previously detailed [14] , [15] on 28 cultures established from patient lesions revealed that L . braziliensis infections were present in 11 of 12 subjects from Cusco and 5 of 16 from Lima ( Figure 3 ) . The distribution of infecting species was consistent with prior studies showing that patients in Cusco are predominantly infected with L . braziliensis [12] , [16] and this could also explain why the lesions from the Cusco cohort were on average larger than the lesions from the Lima cohort . This study assessed whether the combination of topical imiquimod plus systemic pentavalent antimony was superior to monotherapy with pentavalent antimony as a first-line treatment for cutaneous leishmaniasis in Peru . The final outcome did not reach statistical significance in the intention-to-treat analysis , although the combination therapy did have a higher cure rate . In the planned subset analysis by site , there was a similar suggestion that the combination therapy was superior to monotherapy in the Cusco cohort which was predominantly infected with L . braziliensis , a species usually more difficult to treat than L . peruviana and L . guyanensis [6] . The results from this trial are consistent with a previous open-label pilot trial involving this combination therapy as a first-line therapy for cutaneous leishmaniasis in Peru [17] as well as trials on patients who had previously failed treatment [11] , [18] . An important consideration of this study is that activation of a toll like receptor ( TLR ) is therapeutically beneficial in the treatment of this parasitic infection and this could have implications beyond leishmaniasis . Activation of TLR 7/8 with imiquimod could work in at least two ways to enhance resolution of Leishmania infections . First , it could directly stimulate macrophages to synthesize nitric oxide resulting in direct killing of the parasite , as previously detailed in Leishmania-infected cultured macrophages [10] . Secondly , imiquimod could mediate a better anti-Leishmania Th1 immune response that would result in the production of IFN-γ and macrophage activation resulting in enhanced parasite killing . This would be consistent with the observation that topical imiquimod , used as a vaccine adjuvant with a killed Leishmania preparation , has been shown to enhance the Th1 immune response against a subsequent live infection with L . major in BALB/c mice [19] . Since the immune status of leishmaniasis patients is known to affect drug efficacy [7] , the parasite killing by pentavalent antimony plus the enhanced innate and/or acquired immune response against Leishmania could be expected to act synergistically to increase the cure rate and/or reduce the relapse rate . We have previously observed that treatment with imiquimod alone had a transient beneficial effect as determined by a reduction in lesion size in the first few days of treatment but did not ultimately clear the infection [17] . This has also been observed for Old World cutaneous leishmaniasis [20] . This could suggest that imiquimod can activate an initial innate immune response as previously observed in infected and non-infected macrophages in vitro [10] , [21] . Future studies should include analyses of relevant immunological parameters in the infected lesions treated with topical imiquimod to provide a better understanding of how imiquimod enhances cure and possibly define more effective therapy regimes . An interesting observation was that there was a higher failure rate in the control: antimony/placebo group in the Cusco cohort compared to the Lima cohort . This could be related to the predominance of L . braziliensis in Cusco as observed in this and previous studies [12] , [16] . As there is a higher failure and/or relapse rate with L . braziliensis infections than with L . peruviana or L . guyanensis infections [6] , [15] , it is possible that the combination therapy may have reduced the relapse rate associated with L . braziliensis infections in the Cusco cohort . However , in addition to the different species , there may be other differences between the two sites that necessitate a larger sample size . In any event , if the combination therapy were to be considered for L . braziliensis infections , it would be necessary to implement a pre-treatment species determination . Two previous studies using the combination of imiquimod with pentavalent antimony have been carried out in Iran in areas endemic for L . tropica . One study reported that there was no beneficial effect [22] and the other study reported a beneficial effect of the combination therapy [23] . It is difficult to compare our Peruvian study with these results as there were differences in: definitions for cure , methods of administration of the pentavalent antimony , length of follow-up , and prevalent Leishmania species . It is noteworthy that , in Iran , L . tropica is resistant to pentavalent antimony treatment [24] and this could further complicate the interpretation of the outcome . It would , however , be of interest to carry out an imiquimod combination immunotherapy trial in the Old World involving L . major infections that are responsive to pentavalent antimony or in combination with other anti-Leishmania treatments . There was one adverse event that resulted in deviation from the treatment protocol when a generalized skin rash developed in one patient in the experimental ( antimony/imiquimod ) group . The rash resolved when the pentavalent antimony treatment was stopped and the patient continued to be treated with topical imiquimod alone . This patient was eventually considered a treatment failure at 2 months follow-up and was counted as a failure in the experimental group even though he had not received the complete course of combination therapy . Future studies should include combining topical imiquimod with other therapies for cutaneous leishmaniasis such as , for example , with topical paromomycin or oral miltefosine . If the cure rate with a combined imiquimod/paromomycin or imiquimod/miltefosine therapy reaches a cure rate of 80% at 3 months post-treatment , then this can be used to replace pentavalent antimony as the first-line treatment in Peru . Pentavalent antimony could then be used only in those cases that do not respond to the combination topical therapy .	Neglected tropical diseases ( NTDs ) are a group of tropical infections including trypanosomiasis , filariasis , schistosomiasis , onchocerciasis , leishmaniasis and other such diseases of poverty . Of the classic neglected diseases , leishmaniasis has among the highest level of morbidity and mortality . Infection with Leishmania parasites causes severe disease in humans , including fatal visceral leishmaniasis and cutaneous leishmaniasis resulting in severe scarring , often in the face . This is a difficult infection to treat because the current therapies are generally poorly effective . The present study carried out a placebo-controlled , double-blinded study to investigated whether a combined therapy with imiquimod plus pentavalent antimony was superior to the standard therapy of pentavalent antimony alone as a first-line treatment for cutaneous leishmaniasis in Peru . A higher cure rate with the combination therapy was observed , but could not be conclusively proven .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "microbiology/immunity", "to", "infections" ]	2009	First-Line Therapy for Human Cutaneous Leishmaniasis in Peru Using the TLR7 Agonist Imiquimod in Combination with Pentavalent Antimony
The molecular events leading to the development of the bat wing remain largely unknown , and are thought to be caused , in part , by changes in gene expression during limb development . These expression changes could be instigated by variations in gene regulatory enhancers . Here , we used a comparative genomics approach to identify regions that evolved rapidly in the bat ancestor , but are highly conserved in other vertebrates . We discovered 166 bat accelerated regions ( BARs ) that overlap H3K27ac and p300 ChIP-seq peaks in developing mouse limbs . Using a mouse enhancer assay , we show that five Myotis lucifugus BARs drive gene expression in the developing mouse limb , with the majority showing differential enhancer activity compared to the mouse orthologous BAR sequences . These include BAR116 , which is located telomeric to the HoxD cluster and had robust forelimb expression for the M . lucifugus sequence and no activity for the mouse sequence at embryonic day 12 . 5 . Developing limb expression analysis of Hoxd10-Hoxd13 in Miniopterus natalensis bats showed a high-forelimb weak-hindlimb expression for Hoxd10-Hoxd11 , similar to the expression trend observed for M . lucifugus BAR116 in mice , suggesting that it could be involved in the regulation of the bat HoxD complex . Combined , our results highlight novel regulatory regions that could be instrumental for the morphological differences leading to the development of the bat wing . Vertebrate limbs show a wide range of morphological variety ranging from fins to limbs . The developing tetrapod limb is made up of three skeletal elements: the stylopod ( humerus/femur ) , zeugopod ( ulna/tibia , radius/fibula ) , and autopod ( carpals/tarsals; metacarpals/metatarsals; phalanges ) [1 , 2] . Autopods are highly specialized , composed of different numbers and lengths of digits , and exhibit varying degrees of interdigital soft tissue ( webbing ) . Autopods are a hallmark of tetrapod diversity and are essential for adaptation to life on land , in the sea and in the air . Bats are an extreme example of this . To form a wing , bat forelimbs have gone through three major changes: elongation of digits II-V , retention of membranous tissue forming the inter-digital patagia ( chiropatagium ) and a relative reduction in the diameter of the ulna [3–5] . These morphological innovations are clearly apparent in bat fossils from 52 . 5 million years ago [6 , 7] . The genetic changes that led to the development of these specialized limb structures and mammalian flight are likely to have occurred prior to the radiation of the Chiroptera , one of the most diverse mammalian orders . Enhancers can regulate spatiotemporal gene expression during vertebrate development [8] and nucleotide changes within them can lead to phenotypic differences , such as limb malformations [9] . For example , regulatory regions in the 5’Hoxd locus have been implicated in digit specification during mammalian autopod development and loss of interactions with these regions can result in limb phenotypes , similar to Hoxd10-Hoxd13 deletions [10] . Nucleotide changes in enhancers have also been linked to morphological differences between species [11] . One such example is the Prx1 limb enhancer . The replacement of the mouse sequence of this enhancer with the homologous bat Prx1 sequence resulted in mice with longer forelimbs [12] . The recent availability of several bat genomes ( Myotis lucifugus , Myotis davidii , Pteropus vampyrus , and Pteropus alecto ) [13–16] now make it possible to identify specific nucleotide changes in the bat lineage , as compared to other mammals , that could have a role in the development of the unique limb morphology of the bat . Various computational approaches have been used to identify regulatory elements that could be involved in species-specific morphological changes [17–21] . These include human accelerated regions ( HARs ) and human accelerated conserved noncoding sequences ( HACNSs ) , which are highly conserved sequences that have acquired a disproportionate number of nucleotide substitutions since humans diverged from our common ancestor with chimpanzees [20 , 22 , 23] . Based on epigenetic marks , Capra and colleagues predicted that at least 30% of these noncoding HARs are developmental enhancers [24] . So far , 62 out of 92 tested HARs have shown enhancer activity in mouse transgenic assays , and 7 out of 26 HARs , where the activity of the human and chimp sequences were compared , showed differential enhancer activity [25] . These include the limb enhancer sequences HAR2/HACNS1 , which showed no limb specific activity for the non-human homologous sequence [22] , and 2xHAR . 114 , which displayed restricted limb activity for the human sequence compared to the chimpanzee sequence [24] . These findings indicate that the identification of accelerated regions could serve to detect sequences that function as gene regulatory elements and could possibly give rise to characteristic phenotypes among species . Here , we set out to identify enhancers whose alteration could have contributed to bat wing development . We utilized mouse limb-specific ChIP-seq datasets and available bat genomes to identify bat accelerated regions ( BARs ) . We identified 166 BARs with potential enhancer activity in developing limbs and functionally analyzed five of them in mouse , finding all five M . lucifugus cloned BARs ( BAR2 , BAR4 , BAR61 , BAR97 , BAR116 ) to be functional limb enhancers . Comparison of the enhancer activity of mouse and M . lucifugus orthologous BAR sequences revealed expression differences for three of the four tested sequences ( BAR4 , BAR97 , and BAR116 ) , suggesting that these sequences could be accelerated in bats due to functional differences . Amongst them , M . lucifugus BAR116 , which resides in a gene desert on the telomeric side of the HoxD locus , showed robust forelimb and weak hindlimb expression , a trend similar to bat Hoxd10 and Hoxd11 gene expression as we determined using whole-mount in situ hybridization on bat and mouse embryos . We sought to identify specific sequences that could be responsible for bat wing development . To generate a high-confidence list of candidate enhancers , we implemented a comparative genomics approach ( Fig 1 ) that pinpoints bat accelerated regions ( BARs ) , which are genomic sequences that are evolving very slowly in vertebrates , but experienced rapid sequence changes in the common ancestor of extant bats . We analyzed multiple sequence alignments of 58 vertebrates , excluding bat genomes ( see Materials and Methods; Fig 1 ) , to generate 2 . 7 million vertebrate conserved sequences using PhastCons [26] . To focus our analysis on potential limb developmental enhancers , we constrained our search to conserved sequences that overlap with 39 , 260 ChIP-seq peaks for H3K27ac and p300 from embryonic day ( E ) 10 . 5 and E11 . 5 mouse limbs . These include two previously reported datasets [18 , 27] and an H3K27ac E11 . 5 developing mouse limb autopod dataset generated for this project ( see Materials and Methods , Fig 1 ) . We then tested these candidates for statistically significant numbers of substitutions in the ancestor of four bats with sequenced genomes , compared to the set of vertebrate conserved sequences , using PhyloP [28] ( see Materials and Methods ) . Using four bats and numerous vertebrate genomes in our analysis assisted in reducing false positives that can result from sequencing , assembly , and alignment errors . We identified 166 BARs that overlap genomic marks of active enhancers in developing mouse limbs and show significant evidence of accelerated substitution rates in the genome of the common ancestor of extant bats ( false discovery rate ( FDR ) < 5%; S1 Table ) . Like many known developmental enhancers , the average BAR is 1 , 542 base pairs ( bp ) in length and does not overlap gene transcription start sites ( TSS ) . We found that 73% of BARs are more than 20 kilobases ( kb ) from the closest TSS , 45% are more than 100 kb from a TSS , and five are in gene deserts that are greater than 1 megabase ( Mb ) across . Thirty-eight BARs are adjacent to transcription factors ( TFs ) involved in limb development ( see Materials and Methods; S1 Table ) and overall we observed an enrichment for limb TFs near BARs ( OR = 2 . 88 , p-value < 0 . 0001; S2 Table ) . BARs were also found to cluster around each other more densely than expected ( p<0 . 001; permutation test compared to the set of candidate limb enhancers from the various ChIP-seq datasets ) , with several clusters being adjacent to developmental genes . For example , we found five BARs ( BAR4 , 18 , 22 , 71 and 148 ) clustered near sprouty homolog 1 ( Spry1 ) , a gene associated with skeletal myogenesis [29 , 30] . In addition , for 4 out of 5 BARs ( BAR4 , BAR18 , BAR22 , BAR71 , BAR148 ) near Spry1 , we found more pleiomorphic adenoma gene 1 ( Plag1 ) motifs in the inferred sequence of the bat ancestor than in the orthologous mouse sequence ( all p-values<0 . 1; see Materials and Methods ) . Plag1 is a zinc finger protein whose loss in mice results in retarded growth [31] and it has been associated with stature in bovines [32] , limb bone length in pigs [33] and height in humans [34] . We next set out to identify transcription factor binding site ( TFBS ) changes in each of the 166 BARs by estimating the sequence of the common ancestor of the four bat genomes ( M . lucifugus , P . vampyrus , M . davidii and P . alecto; see Materials and Methods ) and comparing this ancestral bat sequence to the orthologous mouse sequence . We predicted TFBS in the mouse and ancestral bat sequences of each BAR and tested for significant loss or gain of TFBS of 745 TFs expressed in the developing limb using motifDiverge [35] . Most TFs only had significant changes in TFBS for a single BAR , but several showed consistent patterns of loss or gain across multiple BARs . When all BARs are analyzed collectively as a single sequence , 34 TFs have significantly more TFBS in the bat ancestor compared to mouse , and 146 TFs have significantly fewer TFBS ( FDR<0 . 05 , S2 Table ) . The most striking TFBS changes in the ancestral bat BAR sequences were gains of sites for Nr2c2 , Sp4 , Zfp281 , and Zfp740 each of which is enriched in twelve or more BARs . Nr2c2 , also known as the testicular nuclear receptor 4 ( Tr4 ) , is involved in osteoblast maintenance and differentiation [36 , 37] . Mice lacking Tr4 do not have apparent skeletal abnormalities , however , they display a reduction in bone mineral density and long bone volume , showing premature aging , spinal curvature [38] , and osteoporosis [36] . Zfp281 and Zfp740 are expressed in the developing limb [39] but have yet to be characterized for their limb function . Two additional TFBS gains are worth noting , Egr1 and Zic2/3 . The Egr genes are C2H2-type zinc finger proteins that function as transcriptional regulators with an important role in mitogenesis and differentiation . Specifically , Egr1 is involved in mouse wound repair , endochondral bone repair and data suggests that EGR1 is upregulated during skeletal muscle wound healing [40 , 41] . Zic2 and Zic3 belong to the C2H2-family of Zinc fingers , are known to be involved in morphogenesis and patterning during development and are associated with muscle and skeletal defects [42–45] . We also observed a significant depletion for specific TFBS when comparing the ancestral bat sequences to mice collectively over all BARs ( S2 Table ) . By rank , the most depleted and fourth most depleted TFs were OSR2 and OSR1 respectively . Odd-skipped related genes , Osr1 and Osr2 , belong to the C2H2 Zinc finger family [46 , 47] and are expressed in the embryonic limb mesenchyme [48 , 49] . Both Osr1 and Osr2 are associated with osteoblast regulation , chondrogenesis [50 , 51] , synovial joint formation , and their removal in mice leads to fusion of these joints [52] . Also worth mentioning are Tgif1 and Meis1 . Tgif1 , the Thymine/Guanine interacting factor 1 , is a repressor of TGF-β/Smad signaling , and is expressed in the developing limb mesenchyme [53] . Meis1 , a TALE homeobox TF , is a marker of the stylopod region and its overexpression abolishes distal limb structures during development [54] . Combined , our results identify TFBS gains and losses in BARs that might have a functional role . To determine whether BARs are functional limb enhancers , we selected five BARs ( BAR2 , BAR4 , BAR61 , BAR97 and BAR116 ) and tested them for enhancer activity using a mouse transgenic assay . The BAR candidates were chosen based on their location , residing within 1Mb of a known limb developmental genes whose alteration leads to a skeletal or limb phenotype ( Table 1; Fig 2 ) . BAR2 resides near Twist2 , a bHLH transcription factor which has been shown to terminate the Shh/Grem1/Fgf autoregulatory loop by repression of Grem1 expression in early limb morphogenesis [55] . BAR4 is in close proximity to Spry1 , a known antagonist of FGF signaling [56] , that along with other Sprouty proteins ( Spry2 , Spry4 ) , is expressed in skeletal muscle stem cells , chondrocytes , limb buds , muscles and tendons during development [29 , 57 , 58] . BAR61 overlapped with the zone of polarizing activity ( ZPA ) regulatory sequence ( ZRS ) , a previously characterized limb enhancer of Shh that when mutated leads to limb malformations in humans , mice , dogs and cats [59 , 60] ( Table 1 ) . BAR97 is in close proximity to Spg20 , a gene that is expressed in the limb , face and brain during morphogenesis and is an inhibitor of BMP signaling that is linked to short stature and spastic paraplegias [61] . BAR116 is located on the telomeric side of the HoxD cluster ( Fig 2 ) which is known to be an important regulator of skeletal development [62] . Regions spanning each of the five BAR candidate enhancers ( Table 1; S1 Table ) were amplified from M . lucifugus , cloned into the Hsp68-LacZ vector that contains an Hsp68 minimal promoter followed by the LacZ reporter gene [63] , and injected into single-cell mouse embryos . Transgenic embryos were harvested at E12 . 5 . This stage was chosen since it is equivalent to CS16E in Carollia perspicallata and Miniopterus natalensis bat embryos , a stage when digits are identifiable and forelimbs ( FL ) lose their symmetry in the anterior to posterior ( AP ) axis compared to hindlimbs ( HL ) [64–66] . All assayed M . lucifugus BAR sequences showed limb enhancer activity in our transgenic mouse assay ( Fig 3 ) . M . lucifugus BAR2 was active in the limb autopod ( 3/5 embryos ) but also demonstrated enhancer expression in the brain and nonspecifically throughout the whole embryo ( 4/5 embryos; S1 Fig ) . M . lucifugus BAR4 was positive for enhancer activity in the proximal limb ( 3/4 embryos; Figs 3 and 4 ) . M . lucifugus BAR61 demonstrated strong activity in the posterior-half of the autopod in FL and HL tissues ( 3/4 embryos ) ( Figs 3 and 4 ) . BAR97 displayed weak and diffuse enhancer expression in the posterior-half of the autopod ( 3/4 embryos; Figs 3 and 4 ) . M . lucifugus BAR116 showed strong enhancer activity throughout the proximal and distal FL , covering the entire autopod and zeugopod regions . A weaker enhancer activity in the proximal portion of the HL was also observed for BAR116 ( 3/5 embryos; Figs 3 and 4 ) . In total , all five examined M . lucifugus BAR sequences displayed enhancer activity in the developing forelimb or hindlimb . To compare the species-specific enhancer activity of our predicted BARs , we set out to analyze the orthologous mouse sequences of four BARs ( BAR4 , BAR61 , BAR97 , BAR116; S1 Table ) . Due to the nonspecific expression pattern of M . lucifugus BAR2 , the orthologous mouse sequence was not analyzed . Regions covering each of the mouse BAR sequences were cloned into the Hsp68-LacZ vector and tested for enhancer activity at E12 . 5 . Mouse BAR61/ZRS ( Shh ) , was active in the posterior-half of the developing autopod ( 2/4 embryos , Fig 4; S2 Fig ) , similar to the corresponding M . lucifugus sequence ( BAR61 ) . However , the three other tested mouse BAR sequences ( BAR4 , BAR97 , BAR116 ) showed differential enhancer activity . Mouse BAR116 ( HoxD ) and mouse BAR4 ( Spry1 ) sequences were both negative for enhancer activity ( Fig 4; S2 Fig ) . For mouse BAR4 it is worth noting that only one of the six positive embryos showed weak staining that was similar to M . lucifugus BAR4 and for mouse BAR116 none of the embryos showed similar enhancer activity , with the majority ( 3/4 ) being completely negative for LacZ ( S2 Fig ) . Mouse BAR97 ( Spg20 ) showed differential enhancer activity compared to M . lucifugus BAR97 , being active in the midbrain , the zeugopod ( 2/4 embryos ) and the forebrain ( 1/4 embryos , Fig 4; S2 Fig ) . Combined , our results suggest that the accelerated sequence changes observed in BARs could lead to differences in limb enhancer expression . Our injected BAR sequences included the PhastCons conserved sequences that defined the BAR element along with the flanking sequence under the ChIP-seq peak ( S1 Table ) . We next wanted to determine whether the M . lucifugus BAR116 PhastCons sequence , rather than sequence differences in the flanking regions , is essential for the observed E12 . 5 limb enhancer activity . We generated a synthetic construct that has the mouse BAR116 PhastCons sequence along with the flanking M . lucifugus sequence . This bat-mouse BAR116 composite sequence was analyzed using a similar mouse transgenic enhancer assay at E12 . 5 and displayed inconsistent limb expression in two out of the four LacZ positive transgenic embryos ( Fig 5 , S2 Fig ) . The loss of limb enhancer activity that we observed for the bat-mouse BAR116 composite construct suggests that the M . lucifugus BAR116 PhastCons element itself , and not the flanking sequence within the ChIP-seq peak , is essential for limb enhancer activity at this time point . The most robust difference in enhancer activity observed was for M . lucifugus BAR116 , which showed strong FL expression for the bat sequence but was negative for the mouse sequence . We wanted to analyze whether the M . lucifugus BAR116 enhancer expression pattern recapitulates that of the bat HoxD gene expression . We carried out whole-mount in situ hybridization on Hoxd10 , Hoxd11 , Hoxd12 and Hoxd13 , important developmental genes expressed during both early limb bud outgrowth and later autopod development [67] , in developing bat ( M . natalensis ) and mouse embryos . At CS15 , we observed FL expression of Hoxd10 and Hoxd11 in both proximal ( zeugopod ) and distal ( autopod ) domains , while Hoxd12 and Hoxd13 expression were mainly limited to a large distal domain within the autopod ( Fig 6A–6D ) . The expression of all of these genes was strongest in a distal domain encompassing digits II-IV . The HL patterning at CS15 had both proximal and distal domains of expression for Hoxd10 and Hoxd11 , while Hoxd12 and Hoxd13 were found in the autopod region ( Fig 6A’–6D’ ) . In the HL , the expression of the HoxD genes we analyzed appeared uniform and fairly symmetrical across the distal edge of the autopod . At CS16 , the matching stage for E12 . 5 in mice , the bat autopod expands , becoming highly asymmetrical and digit rays become apparent . In the bat FL the distal expression of Hoxd10 and Hoxd11 are indistinguishable from one another , with strong expression occurring in a ‘triangular’ domain between digits II–IV ( Fig 6E and 6F ) . Hoxd12 has a more expansive autopod expression , extending from the posterior region of digit II and throughout the posterior portion of the autopod , with expression being focused in the proximal half of the developing digit rays ( Fig 6G ) . Hoxd13 was found throughout the autopod , and was strongest in the regions around the developing digits and in the interdigital region between digits III-V ( Fig 6H ) . Interestingly , at CS16 , we observed that HoxD expression in the HL is lost in the distal portion for Hoxd10 and Hoxd11 , while being maintained in the region where the calcar develops ( Fig 6E’ & 6F ) . This differential expression in the FL and HL was similar for M . lucifugus BAR116 , whereby we observed a robust FL expression within the entire limb but reduced HL expression in transgenic mice at E12 . 5 . Expression of Hoxd12 and Hoxd13 at CS16 was maintained in the bat HL ( Fig 6G’ & 6H’ ) . In CS17 FLs , when the digit rays extend , expression of all the HoxD genes examined becomes progressively restricted to the regions surrounding the digits and is reduced in the distal interdigital tissue . In the HL , expression of Hoxd10 and Hoxd11 is still absent , and is reduced in the calcar region ( Fig 6I’ & 6 J’ ) . Hoxd12 expression appears reduced and is only found surrounding the digit rays while Hoxd13 is found throughout the autopod region ( Fig 6K’ & 6L’ ) . In summary , although the observed expression pattern for Hoxd10 and Hoxd11 did not fully recapitulate the M . lucifugus BAR116 limb enhancer expression pattern , we did observe lower HL expression for both these genes at an equivalent stage that matches the decreased activity of the M . lucifugus BAR116 enhancer . Using comparative genomics , developing limb ChIP-seq datasets and mouse enhancer assays , we characterized genomic regions in bat genomes that could play a role in mediating gene expression changes underlying the unique morphological development of the bat wing . We identified 166 BARs which showed a global enrichment for Nr2c2 , Zfp281 , Zfp740 , Zic2/3 and Egr1 , and depletion for Osr1 , Osr2 , Tgif1 and Meis1 TFBS when comparing their mouse sequences to the inferred ancestral bat sequences ( S2 Table ) . Analysis of five M . lucifugus BARs using a mouse transgenic assay showed all of them to have enhancer activity in the developing limb at E12 . 5 . Examination of the mouse orthologous sequences for four of these BARs showed three to be differentially expressed compared to the M . lucifugus sequence , including BAR116 that showed strong FL versus HL expression , matching the differential expression pattern observed for Hoxd10 and Hoxd11 in developing bat limbs . The M . lucifugus BAR4 sequence showed enhancer expression in the proximal FL [68] , while the mouse orthologous sequence BAR4 was negative for enhancer activity at this time point ( S1 and S2 Figs ) . BAR4 is near Spry1 , a gene involved in skeletal and muscle development in mice that when overexpressed leads to chondrodysplasia , a skeletal disorder leading to arrested development [57 , 69] . Interestingly , 5 of our 166 identified BARs reside in the Spry1 locus ( Fig 2 , S1 Table ) , suggesting that altered regulation of this gene could have been important for bat wing development . It is also worth noting that all five Spry1 BARs are located in the same topological associating domain [TAD; [70]] ( Fig 2 ) . The presence of multiple BARs in this region could implicate a co-operative and tightly coordinated regulation of Spry1 during bat limb myogenesis and digit elongation via Fgf signaling [29 , 57] . The transcription factor Plag1 could be differentially binding these enhancers and affecting the regulation of Spry1 or other nearby genes . Interestingly BAR61 , the well-characterized Shh limb enhancer ( ZRS ) , did not show differential enhancer activity between M . lucifugus and mice at E12 . 5 . Both BAR61 sequences had similar enhancer activity in the ZPA of the autopod ( S1 and S2 Figs ) , however , both also seemed to have an expanded domain compared to the previously characterized E11 . 5 expression pattern [59] . These analyses do not exclude differences in expression pattern that may occur at later stages of development , or quantitative differences , which cannot be picked up through these transient transgenic assays . BAR116 is of extreme interest , due to its telomeric location relative to the HoxD locus ( Fig 2 ) . The HoxD locus is conserved across vertebrates and has a critical spatiotemporal role in skeletal development [67] . Hoxd10 and Hoxd13 in particular are known to directly interact with Shh during limb outgrowth [71 , 72] and mutations in these genes lead to various limb malformations including synpolydactyly , split hand and foot , and distally located skeletal elements [67 , 73] . Several cis regulatory elements for the HoxD cluster have been shown to drive limb development and are conserved among vertebrates [10 , 74] . The M . lucifugus BAR116 portrayed enhancer activity in the FL mesenchyme but had reduced activity in HL ( 5/5 embryos ) . Its orthologous mouse sequence was negative for enhancer activity , despite showing evolutionary conservation in vertebrates . It is worth noting that a 700 bp partially overlapping region to mouse BAR116 , termed CNS9 found in the HoxD telomeric region ( Fig 2 ) , was previously examined for enhancer activity in a lentiviral-mediated mouse transgenesis system and was also negative [75] . M . lucifugus BAR116 showed strong enhancer activity throughout the developing FL and weak expression in the proximal portion of the HL . Using whole-mount in situ hybridizations , we analyzed the expression patterns of Hoxd10-13 at CS15-CS17 for M . natalensis and mice at matching developmental time points . Our results showed an overall similarity to previously characterized Hoxd10-13 bat and mouse limb expression patterns [15 , 76] . The robust FL expression and weaker HL expression of BAR116 showed a similar trend to that of Hoxd10 and Hoxd11 ( Fig 6 ) . Interestingly , Hoxd9 also showed a reduction in HL expression at CS16 [15] , similar to the one we observed for Hoxd10 and Hoxd11 , suggesting that BAR116 could possibly be regulating these and other HoxD genes that are located 5’ to these genes . Additionally , by using 4C combined with HoxD telomeric deletion assays in mice , Hoxd9-Hoxd11 were suggested to interact , during early phase of HoxD expression , with telomeric enhancers promoting forearm/arm development [75] . BAR116 ( CNS9 ) lies within this telomeric region and is 850kb from the telomeric boundary of this topological domain suggesting that it could regulate HoxD genes . There are also two functional limb enhancers on both sides of it , CNS39 ( 60kb centromeric to CNS9; Fig 2 ) and CNS65 ( 229kb telomeric to CNS9; Fig 2 ) , that are thought to interact with HoxD genes [75] . The differential enhancer activity we observed for M . lucifugus BAR116 compared to the negative enhancer activity of its mouse sequence and bat-mouse BAR116 composite sequence at E12 . 5 , could imply that bats have acquired a novel enhancer function or a temporal specific chromatin conformation essential for forelimb morphology during autopod development in this locus . Our data suggests that accelerated regions could be used to identify species-specific developmental enhancers that serve as critical determinants during morphoevolution as has been demonstrated previously [20 , 22 , 24] . In bats , an examination of the evolution of echolocation identified Foxp2 as a major constituent of vocal and orofacial development pathways , finding conserved noncoding elements in close proximity of Foxp2 that changed significantly in echolocating bats when compared to non-echolocating species [77] . Similarly , our study utilizes different species to test orthologous sequences , but focuses specifically on regions that are predicted to be developmental limb enhancers through ChIP-seq . There have been previous reports that analyzed bat-specific limb enhancers [12 , 76] . However , our study is the first to examine , in a genome-wide manner , putative limb enhancers in bats . Interestingly , the Prx1 known bat limb enhancer whose replacement in the mouse led to longer forelimbs [12] , was not identified in our analyses as a BAR element . It is worth noting that our study also had many caveats . The mouse transgenic enhancer assay that we used is not quantitative and can be inconsistent due to differences in integration sites and transgene copy number . We also could not test sequences in bat embryos and so were limited to observing expression changes only in mice . For our in situs , due the scarcity of these embryos , we were only able to use M . natalensis embryos whereas in our mouse transgenic assays we used sequences from M . lucifugus , which could also lead to differences in expression patterns . Moreover , we only examined 5 of 166 BARs , which does not represent the majority of the BARs found in our pipeline . For our global TFBS analysis , we analyzed what we determined to be the ancestral bat sequence that could lead to us missing several TFBS changes . In addition , the TFBS matrixes we used were mainly human and mouse based and could differ in bats . In M . lucifugus BAR sequences we also noticed repetitive regions that were not present in the mouse BARs , which could explain the enrichment for specific TFBSs during our motifDiverge analysis . Despite these caveats , our study shows that the use of tissue specific ChIP-seq datasets combined with sequence acceleration can be an efficient means to identify sequences that are important in determining morphological changes between species . Mouse work was approved by the UCSF Institutional Animal Care and Use Committee ( protocol number AN100466 ) and was conducted in accordance with AALAC and NIH guidelines and also by the University of Cape Town Faculty of Health Sciences animal ethics committee application number FHS AEC 012/052 . Ethical approval to collect bats was given by the University of Cape Town , Faculty of Science Animal Experimentation Committee ( 2006/V4/DJ , 2008/V16/DJ and 2012/V39/NI ) with permission to sample granted by the Western Cape Nature Conservation Board ( AAA004-00030-0035 ( 2006 ) , AAA007-00041-0056 ( 2012–2014 ) . To identify BARs , we employed a statistical phylogenetic test for accelerated nucleotide evolution in the common ancestor of all extant bats . This is an extension of a previously proposed likelihood ratio test for acceleration in a single species or clade [28] . This new ancestral lineage version of the likelihood ratio test is implemented in the PhyloP function ( option–-branch ) in the open source software package PHAST [78] . The input to PhyloP is a multiple sequence alignment for each genomic region to be tested for acceleration , plus a phylogenetic tree of the species in the alignment that is estimated from genome-wide data ( in this case , four-fold degenerate sites ) . To apply this statistical test to bat limb development , we first identified a collection of candidate enhancers for limb development genes by intersecting evolutionarily conserved elements with enhancer-associated histone modifications and transcription factor binding events measured in the developing mouse limbs ( Fig 1 ) . Specifically , we took the union of all peaks from two previously published ChIP-seq experiments targeting H3K27ac or p300 [18 , 27] and an H3K27ac dataset generated for this project . Next , we generated a set of vertebrate conserved elements that were agnostic to the rate of nucleotide substitutions in bats . We started with 60-way vertebrate multiple sequence alignments with mouse as the reference species ( UCSC Genome Browser , mm10 assembly ) . We dropped the two bat genomes ( M . lucifigus and P . vampyrus ) from the alignments to ensure that high rates of nucleotide differences between the bats and other vertebrates would not prevent us from identifying conservation in other species . Finally , we ran the PhastCons program with default settings [26] on the resulting genome-wide alignments . This analysis identified 4 , 384 , 943 conserved elements , many of which were less than 100 bp long and , thus , too short for statistical tests for acceleration [28] . However , we observed that many short elements frequently clustered together on the chromosome and that known functional elements ( e . g . , coding exons ) were often tiled with multiple conserved elements separated by short gaps . Hence , we iteratively merged adjacent elements until the ratio of the distance between the elements merged over the total length of the region was less than or equal to 0 . 1 . This merging algorithm was the result of empirical experiments aimed at producing one or a small number of merged elements per exon . We also experimented with adjusting the parameters of PhastCons to produce longer elements , but found that post-processing , by merging , recapitulated exons more effectively . Next , we intersected all merged regions greater than 100 bp with the ChIP-seq peaks and unmasked the M . lucifigus and P . vampyrus sequences from the multiple alignments . Regions with more than 50% missing sequence from either bat or more than 25% of nucleotides overlapping a coding exon were dropped to produce a collection of 20 , 057 candidate limb enhancers . Prior to PhyloP analysis , we integrated sequences from two additional bat genomes into the candidate enhancer alignments . We obtained assembled contigs for two bats , M . davidii and P . alecto , that were sequenced to high coverage ( 100x ) [13] . We used the BLAST algorithm to identify alignments of the mouse sequence from each candidate enhancer to contigs from M . davidii and P . alecto [79] . The single best hit with an e-value less than or equal to 0 . 01 was then blasted back to the mouse genome . If this produced a reciprocal best hit ( i . e . , the top scoring alignment to the mouse genome overlapped the original candidate enhancer sequence ) , we added the M . davidii or P . alecto sequence to the 60-way multiple alignment for that candidate enhancer . This produced alignments with between two and four bats present per enhancer . The two additional bat species were added to the phylogenetic tree corresponding to the 60-way alignments ( UCSC Genome Browser ) and their branch lengths were adjusted using their relationship to M . lucifigus and P . vampyrus . We then restricted our analysis to regions containing at least one bat . Finally , we used PhyloP to test each candidate enhancer for accelerated nucleotide substitutions along the ancestral bat lineage . The resulting p-values were adjusted for multiple testing using a false discovery rate ( FDR ) controlling procedure [80 , 81] . We call all candidate enhancers with FDR < 5% Bat Accelerated Regions ( BARs ) ( S1 Table ) . Their genomic distribution and sequence composition were analyzed using custom Python scripts . Significant associations with functions and phenotypes of nearby genes were identified using GREAT after lifting BARs over to mm9 coordinates [82] . We curated a list of limb-associated genes by exhaustively looking through the literature for evidence found in mouse or human and used resampling tests to assess associations between BARs and these genes compared to random sets of PhastCons elements . To look for TFBS differences , we manually curated a list of limb-associated TFs ( S2 Table ) . BARs were analyzed for loss and gain of binding sites for each TF using motifDiverge [35] . We first compared the ancestral bat sequence to mouse . We used prequel to computationally infer the sequence of the common ancestor of extant bats using our multiple alignments [78] . We created the corresponding aligned mouse sequence from these alignments . We then called a TFBS a hit if its FDR exceeded a threshold of 0 . 01 . We then used motifDiverge [35] to test if the total number of TFBS in the bat ancestor was significantly different than the number of TFBS in mouse for each TF in each individual BAR . We repeated these tests collectively over all BARs . ChIP was performed using the LowCell# ChIP kit ( Diagenode ) as previously described [83] . About 70 , 000 cells were pooled per IP and sonicated with a Covaris sonicator ( S220 Focused-ultrasonicator , Covaris ) . Of the sheared chromatin , 30ul was used for each ChIP experiment with the antibody anti-acetyl histone H3 ( Lys27 ) clone CMA309 ( Milipore 05–1334 ) . Following the manufacturer’s directions , each library was constructed using the Rubicon ThruPLEX library construction kit . Each library included 10 ul of ChIP material for a total of 14 cycles of amplification . Sequencing was carried out using an Illumina HiSeq and FASTQ files were aligned to the Mus Musculus genome ( mm9 ) using Bowtie 0 . 12 . 8 [84] . A single base pair mismatch was permitted and reads with multiple alignments were discarded . The ChIP-seq library was sequenced to a depth of 168M total reads with 137M aligning uniquely . The input sample was sequenced to a depth of 111M reads total and 81M aligning uniquely . In each case , approximately 18% of sequences failed to align . We sorted and indexed the alignments using SAMtools 0 . 1 . 18 [85] and then converted to BED files with the bam2bed utility , a part of bedtools 2 . 17 . 0 [86] . To identify enriched H3K27ac islands in the limb samples , the peak-finding tool SICER 1 . 1 [87] was used . PCR was carried out either on M . lucifugus or M . musculus DNA using primers that were designed to amplify candidate enhancer peak sequences with additional 100–500 bp outside of predicted regions ( S1 Table ) . The bat-mouse BAR116 composite sequence was synthesized ( Biomatik ) and sequence validated . PCR products and the synthetic sequence were cloned into the Hsp68-LacZ vector [63] and sequence verified . All transgenic mice were generated by Cyagen Biosciences using standard procedures [88] , and harvested and stained for LacZ expression at E12 . 5 as previously described [89] . Pictures were obtained using an M165FC stereo microscope and a DFC500 12-megapixel camera ( Leica ) . To be designated as an enhancer , we required consistent spatial expression patterns present in at least two embryos . ChIP-seq data has been made publically available through NCBI ( ChIP-seq BioProject ID: PRJNA252737 as experiment ID SRX793524 ) .	The limb is a classic example of vertebrate homology and is represented by a large range of morphological structures such as fins , legs and wings . The evolution of these structures could be driven by alterations in gene regulatory elements that have critical roles during development . To identify elements that may contribute to bat wing development , we characterized sequences that are conserved between vertebrates , but changed significantly in the bat lineage . We then overlapped these sequences with predicted developing limb enhancers as determined by ChIP-seq , finding 166 bat accelerated sequences ( BARs ) . Five BARs that were tested for enhancer activity in mice all drove expression in the limb . Testing the mouse orthologous sequence showed that three had differences in their limb enhancer activity as compared to the bat sequence . Of these , BAR116 was of particular interest as it is located near the HoxD locus , an essential gene complex required for proper spatiotemporal patterning of the developing limb . The bat BAR116 sequence drove robust forelimb expression but the mouse BAR116 sequence did not show enhancer activity . These experiments correspond to analyses of HoxD gene expressions in developing bat limbs , which had strong forelimb versus weak hindlimb expression for Hoxd10-11 . Combined , our studies highlight specific genomic regions that could be important in shaping the morphological differences that led to the development of the bat wing .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "sequencing", "techniques", "split-decomposition", "method", "vertebrates", "animals", "mammals", "multiple", "alignment", "calculation", "developmental", "biology", "sequence", "motif", "analysis", "molecular", "biology", "techniques", "mammalian", "genomics", "embryos", "...	2016	Bat Accelerated Regions Identify a Bat Forelimb Specific Enhancer in the HoxD Locus
Neural Mass Models provide a compact description of the dynamical activity of cell populations in neocortical regions . Moreover , models of regional activity can be connected together into networks , and inferences made about the strength of connections , using M/EEG data and Bayesian inference . To date , however , Bayesian methods have been largely restricted to the Variational Laplace ( VL ) algorithm which assumes that the posterior distribution is Gaussian and finds model parameters that are only locally optimal . This paper explores the use of Annealed Importance Sampling ( AIS ) to address these restrictions . We implement AIS using proposals derived from Langevin Monte Carlo ( LMC ) which uses local gradient and curvature information for efficient exploration of parameter space . In terms of the estimation of Bayes factors , VL and AIS agree about which model is best but report different degrees of belief . Additionally , AIS finds better model parameters and we find evidence of non-Gaussianity in their posterior distribution . Dynamical systems models instantiated using differential equations are a mainstay of modern neuroscience and provide mathematical descriptions of neuronal activity over multiple spatial and temporal scales [1 , 2] . In imaging neuroscience a widely adopted framework , called Dynamic Causal Modelling ( DCM ) , has been developed for fitting such models to brain imaging data using a Bayesian approach [3] . This allows inferences to be made about changes in parameters ( eg . effective connectivity ) in the human brain using noninvasive imaging data . There is now a library of DCMs which differ according to their level of biological realism and the data features they explain . DCM can be applied to fMRI [3] , EEG and MEG [4] and invasive electrophysiological data [5] . The Bayesian approach to model fitting in DCM is based on the Variational Laplace ( VL ) algorithm [6] . One of its core assumptions , the ‘Laplace Assumption’ , is that the posterior distribution is Gaussian . This assumption is typically instantiated by finding the maximum posterior parameter vector , using numerical optimisation , and making a Taylor expansion around this value and retaining terms up to second order [7] . It has been found to be more robust than higher-order moment expansions on empirical data [8] . In VL , the posterior is assumed to factorise into a product of probability distributions , one over latent variables controlling noise variances and one over model parameters . Each distribution is multivariate Gaussian with mean and covariance that are iteratively updated to maximise an approximation to the model evidence [6] . The Laplace approximation is attractive because it provides a computationally simple method for both quantifying posterior uncertainty in model parameters and approximating the model evidence for Bayesian model comparison . A theoretical motivation for the the Laplace approximation is that the posterior will tend to a Gaussian in the limit where the number of data points goes to infinity [9] . But as previously noted in the context of DCM [10] , it is questionable as to whether posteriors are Gaussian for datasets that are encountered in practice which naturally have a finite number of data points . The VL algorithm has two potential weaknesses ( i ) as with any local optimisation method working in a non-convex domain [11] it may fall into a local maxima and ( ii ) the distribution around the maxima may be non-Gaussian . In this paper we compare VL to Monte Carlo methods in the challenging context of identifying Neural Mass Models ( NMMs ) [12] . The advantage of Monte Carlo methods is that , provided the sampling process runs for a sufficiently long time , the samples converge in distribution to the exact posterior . This obviates the need for Gaussian assumptions but at the cost of potentially very long sampling times . To address these issues this paper uses the Annealed Importance Sampling ( AIS ) algorithm [13] with proposals made using a Langevin Monte Carlo ( LMC ) procedure [14] . The use of AIS has two benefits ( i ) it can accomodate multiple local maxima and ( ii ) it provides an estimate of the Bayesian model evidence . The use of LMC improves convergence properties because proposals are made using local gradient and curvature information [14 , 15] . Previously , the Metropolis-Hastings ( MH ) algorithm has been used to validate VL in the context of DCM for fMRI [16] . Whilst these findings are largely consistent with the Laplace assumption this study is incomplete in a number of respects ( i ) only results from a single Markov chain were reported thus raising the possibility that a local maxima was found , ( ii ) no sample-based estimate of the model evidence was provided , and ( iii ) the neurodynamical models used in fMRI are based on linear dynamical systems , so this finding may not hold for the nonlinear dynamical models [17] underlying other DCMs such as those for M/EEG data . This paper assesses how well the two Bayesian estimation algorithms ( AIS-LMC and VL ) perform inference for NMMs . These models have been chosen as they are highly nonlinear and underlie the first proposed DCM for M/EEG data [17] . In order to validate our software implementation and fine tune parameters of the AIS algorithm , we additionally evaluate these algorithms in the simpler context of linear and nonlinear regression models . Annealed Importance Sampling ( AIS ) [13] provides samples from a posterior density using a sequence of densities at a series of monotonically increasing inverse temperatures βj with j = 0 . . J , β0 = 0 and βJ = 1 . For the jth temperature the algorithm produces a sample from the unnormalised density f j ( w ) = p ( y \| w ) β j p ( w ) ( 1 ) An independent sample w ( i ) from the posterior density is produced by generating a sequence of points w1 , w2 , … wJ as follows Generate w1 from p ( w ) Generate w2 from w1 using T1 ( w2\|w1 ) … Generate wj from wj−1 using Tj−1 ( wj\|wj−1 ) … Generate wJ from wJ−1 using TJ−1 ( wJ\|wJ−1 ) and then let w ( i ) = wJ . We refer to the process of producing a single independent sample as a ‘trajectory’ . The transition densities Tj can be chosen in any of the usual ways for constructing Markov chains [18] and may themselves involve several steps . The only requirement is that Tj is chosen to leave fj as the invariant distribution . For example , for a simple density estimation problem , Neal [13] specified each Tj to be a sequence of Metropolis moves each defined using an isotropic Gaussian proposal with increasing width . For a linear regression problem with non-Gaussian priors he employed a Hamiltonian Monte-Carlo ( HMC ) approach [19] . In this paper we will use Langevin Monte Carlo ( LMC ) , as recent work shows this to provide higher effective sample size per unit of computation time as compared to HMC [15] . The above process is repeated i = 1 . . I times to produce I independent samples from the posterior density . Because the samples are produced independently , without interaction among trajectories , the AIS algorithm is amenable to ‘embarrassing parallelization’ [20] . Specifically , trajectories can be assigned to individual computer processors or processor cores thus greatly speeding up the implementation . Each sample is also accompanied by an importance weight v ( i ) = f 1 ( w 1 ) f 0 ( w 1 ) f 2 ( w 2 ) f 1 ( w 2 ) f 3 ( w 3 ) f 2 ( w 3 ) … f J ( w J ) f J - 1 ( w J ) ( 2 ) which can be evaluated as log v ( i ) = ∑ j = 1 J β j - β j - 1 log p ( y \| w j ) ( 3 ) To avoid numerical overflow we first create adjusted weights ui v m a x = max ( log v ) u i = exp ( log v ( i ) - v m a x ) ( 4 ) and let u ¯ be the mean adjusted weight . The normalised importance weights are q i = u i ∑ i u i ( 5 ) A derivation of the formula for the importance weights is provided in [13] and included in S1 Text . The variance of the importance weights is an indicator of the quality of the approximation to the posterior density [13] . In this paper the transition densities Tj in AIS are implemented using a Langevin Monte Carlo ( LMC ) sampler , which leads to proposals being accepted with high probability even for nonlinear and high dimensional inference problems , as it uses information about the gradient and curvature of the unnormalised density , fj . The use of LMC follows from the definition of the log joint and its gradient as a function of w L ( w ) = log p ( y \| w ) + log p ( w \| μ , Λ ) g ( w ) = d L ( w ) d w ( 13 ) A proposal is drawn as w s * ∼ p ( w s * \| w s ) p ( w s * \| w s ) = N ( w s * ; m , C ) m = w s + 1 2 C g ( w s ) C = h 2 ( Λ + F − 1 ( 14 ) where Λ is the prior precision , ws is the sth sample , and h is a step size parameter ( fixed at 0 . 5 for all applications in this paper ) . The quantity F is the Fisher Information matrix ( see S2 Text ) and quantifies the precision of the parameters conferred by the data . This has analytic forms for many probabilistic models such as logistic regression [14] and is readily computed for differential equation models using an approach based on forward sensitivity analysis [27 , 28] . The Metropolis-Hastings ( MH ) criterion is then applied to accept proposals with probability r = p w ( w s * ) p w ( w s ) p ( w s \| w s * ) p ( w s * \| w s ) ( 15 ) where pw ( ws ) = exp[L ( ws ) ] . The proposal is always accepted if r > 1 . We set w s + 1 = w s * if the sample is accepted and ws+1 = ws if it is rejected . The above proposal ( Eq 14 ) has the same functional form as the Simplified Manifold MALA algorithm as applied to ODEs [14 , 27] . Here the ‘manifold’ is defined by C and m and its computation has been ‘simplified’ as the curvature has been assumed to be locally constant . For Gaussian likelihoods , this same local linearity assumption is also the basis of the Gauss-Newton optimization algorithm [29] . In the usual application of LMC [14 , 15] , Eqs 14 and 15 , are repeatedly applied until one obtains samples from the posterior density . However , in this paper we use LMC to provide a single sample at each temperature in an AIS trajectory . Specifically , the transition kernel , Tj−1 ( wj\|wj−1 ) , starts at ws = wj−1 and produces w j = w s * using the modifield log joint Lj−1 . This modification requires multiplication of the likelihood , gradient and Fisher information by βj−1 . The LMC updates are otherwise identical . Because LMC is used to produce only a single sample at each temperature the total number of LMC steps is equal to the number of temperatures . We now briefly comment on the computational scalability of the combined AIS-LMC algorithm . Because AIS is based on importance sampling its accuracy is proportional to the number of annealing runs ( “trajectories” ) [13] . As trajectories are independent , and can be assigned to cores on multiple core computer architectures , the accuracy will therefore scale with the number of cores ( at almost no increase in computer time ) . For a fixed number of cores computer time scales linearly with the number of trajectories . The computational bottleneck within each AIS trajectory is the evaluation of the gradient of the log joint and the Fisher information , required for each LMC step . These quantities can be efficiently computed for ODE models using forward sensitivity or adjoint methods [27 , 28] . The computation time of these methods scales linearly with the length of time series being modelled , and adjoint methods are typically more efficient than forward sensitivity methods if the number of parameters is much larger than the number of dynamical states . In multiple linear regression an [N × 1] data vector y is generated as y = X β + e ( 16 ) where X is an [N × p] design matrix , β is a [p × 1] vector of regression coefficients , and e is an [N × 1] zero-mean IID Gaussian noise vector with entries having variance σ2 . To provide a simple nonlinear model with multiple maxima , we consider a regression model where the parameters of interest are nonlinearly related to the regression coefficients y = ∑ i x i β i + e β i = w i 2 ( 17 ) This model will have multiple maxima over the various combinations of positive and negative values of wi . We also consider an exponential approach-to-limit or ‘approach’ model where y ( t ) = - 60 + V a 1 - exp ( - t / τ ) + e ( t ) ( 18 ) with parameters w1 = log τ and w2 = log Va . This models the ramping up of a voltage from −60 to −60 + Va with a time constant τ , and has the same mathematical form as Biochemical Oxygen Demand ( BOD ) models [30] previously used to evaluate Bayesian inference methods [31] . As the VL algorithm assumes that the posterior distribution is Gaussian it will be interesting to see if this is indeed the case . We use Royston’s test for multivariate normality [34] using a Matlab implementation by Trujillo-Ortiz et al [35] . This is a multivariate extension of the Shapiro-Wilks test and we apply it to Monte Carlo samples from the posterior densities produced by AIS . As these samples are independent there is no need for ‘thinning’ or assessments of Effective Sample Size [36] . The algorithms on which this research is based have been implemented in Matlab in the ‘Monte Carlo Inference ( MCI ) ’ toolbox and will be distributed as part of a forthcoming release of the Statistical Parametric Mapping ( SPM ) package . AIS and LMC , for example , are implemented in the spm_mci_ais . m and spm_mci_lgv . m functions available in the subdirectory /toolbox/mci/inference/ . The Variational Laplace ( VL ) algorithm is instantiated in the SPM software [33] ( in the function spm_nlsi_GN . m ) and described elsewhere [6 , 37] . We also include a brief mathematical description in S4 Text . In VL , the posterior is assumed to factorise into a product of probability distributions , one over latent variables controlling noise variances and one over model parameters . Each distribution is multivariate Gaussian with mean and covariance that are iteratively updated to maximise an approximation to the model evidence [6] . Importantly , the multivariate nature of each Gaussian allows parameter dependencies to be accommodated . This optimiser is the standard approach used for the majority of DCM applications in neuroimaging . Known noise variances ( see below ) are implemented for the VL algorithm by setting the prior over the log noise precision to have a mean corresponding to the true ( known ) value , and a variance of 10−8 ( i . e . very tight ) . By default , the implementation of VL in SPM initialises parameters at the prior mean . A simple way of potentially handling optimisation problems with multiple maxima , however , is to run the VL algorithm multiple times where each run is initialised using a different sample from the prior . We will refer to this procedure as Multistart VL . We first provide results on a multiple linear regression model , as there are analytic formulae for the posterior distribution and model evidence [7] , and the Laplace approximation is exact . This comprised p = 7 regressors chosen from a discrete cosine basis set over N = 20 ‘time points’ , with additive noise of standard deviation σ = 0 . 2 . The prior variances , Λ p p - 1 were set to 10 for each regressor and the prior means , μp to zero . The regression coefficients were drawn from the prior . The AIS algorithm was applied to this data using J = 512 temperatures and I = 32 independent samples . We fitted the true model ( with 7 regressors ) and a reduced model to the same data but this time using only the first 6 regressors . Using the 32 samples produced by AIS , we could not reject the hypothesis that the posterior was Gaussian using Royston’s test for the full ( p = 0 . 67 ) and reduced ( p = 0 . 68 ) models . This is of course to be expected as the posterior distribution is indeed Gaussian for linear regression models [7] . For the full model , the normalised importance weights had high entropy , H = 4 . 07 , and many trajectories had significant weight , Iq = 21 . The AIS estimates of the log model evidences for the full , log p ( y\|m = f ) , and reduced models , log p ( y\|m = r ) and the corresponding log Bayes factor , and computation times , are provided in Table 1 . The estimates very closely match the analytic values . Note that the VL estimates correspond to the analytic values for the case of linear regression [7] . We then re-estimated the evidences using different numbers of AIS samples and temperatures , with results plotted in Fig 2 . Theses results show good agreement with analytic values for J = 128 and above . The error bars on AIS model evidence estimates were computed using bootstrapping ( over trajectories ) as described in the section on ‘Annealed Importance Sampling’ , in the subsection on ‘Model Evidence’ . To produce the following results the differential equations underlying the neural mass models ( see S3 Text ) were integrated using implicit backward-differentiation formulas ( BDFs ) and the resulting nonlinear equations solved using Newton’s method as implemented in the CVODES software [40] . With a relative tolerance of 10−2 and an absolute tolerance 10−4 this algorithm took an average of 75ms ( averaged over ten runs ) to produce the time series for the two-region model . This was lower than the 229ms for Matlab’s ODE15s integrator and the 90ms for SPM’s ( implemented in the function spm_int_L . m ) . Both VL and AIS model estimation approaches therefore used the CVODES implementation . For the LMC algorithm used in AIS , gradients were computed using a forward sensitivity method as implemented in CVODES . For VL , gradients and curvatures were computed using central differences as implemented in the SPM function spm_nlsi_GN . m . The simulations that follow make use of the two-region neural mass model depicted in Fig 1 and described above . We generated data from a model with strong forward and backward connections . This is specified using the parameter values w1 = w2 = 1 which set the connections a21 and a12 according to S1 Table . The other parameters were set to zero . Data was then generated from the model as described above using zero mean additive Gaussian noise having standard deviation σs = 0 . 01 . The resulting time series are shown in black in Fig 5 . The priors over model parameters for Bayesian model fitting are as described at the end of the above subsection ‘Two-Region Model’ in the section on ‘Neural Mass Models’ . We then fitted two models to the data using AIS , a ‘full’ model , which has the same structure as the model from which the data were generated , and a ‘reduced’ model which did not have the backward connection . We used I = 32 , J = 512 and model estimation took 5290s and 4610s for the full and reduced models . The estimated log model evidences were 1563 . 6 for the full model and 1293 . 4 for the reduced model , corresponding to a Log Bayes Factor of 270 . 2 in favour of the full model . Using the 32 samples produced by AIS , we could not reject the hypothesis that the posterior was Gaussian using Royston’s test for the full ( p = 0 . 32 ) and reduced ( p = 0 . 15 ) models . The AIS acceptance rates , aj , averaged over the I = 32 trajectories , showed a gradual decrease with βj . Averaging aj over the high temperatures ( βj < 0 . 5 ) gave a value of ahigh = 0 . 43 and over the low temperatures of alow = 0 . 19 . These acceptance rates show that the cost function is being sufficiently explored and are in line with other Bayesian annealing methods [38] . The normalised importance weights had lower entropy than for the previous models above , H = 2 . 59 , and fewer trajectories with significant weight , Iq = 12 . We also fitted the full and reduced models using VL , which took 22s and 24s ( using 19 and 22 VL iterations ) respectively . The estimated log model evidences were 1524 . 1 for the full model and 1288 . 4 for the reduced model , corresponding to a Log Bayes Factor of 235 . 74 in favour of the full model . Thus , the VL and AIS estimates agree reasonably well for the reduced model ( within 0 . 4 per cent ) but not for the full model ( within only 2 . 5 per cent ) . Which are we to believe ? As described in S1 Text , it is also possible to use the VL posterior as a proposal density to provide an importance sampling estimate of the model evidence , without using any annealing . We refer to this procedure as ISVL and used it to generate 1000 samples . ISVL is highly computationally efficient , requiring only 90s of compute time . The estimate of the log evidence was 1562 . 8 for the full model which agrees very well with the AIS estimate ( within 0 . 05 per cent ) . Fig 6 plots the log evidences and log Bayes factor as a function of the number of temperatures J . These indicate that a fine-grained temperature resolution J is required to obtain good results . We also note that the log joint probability , L ( see Eq 13 ) , of the posterior mean AIS solution increases with J , with values of L = 1583 , 1584 , 1588 , 1589 for J = 64 , 128 , 256 , 512 . The log joint probability of the true parameters is L = 1589 , whereas the log joint of the VL posterior mean is only L = 1157 . Fig 7 plots the posterior densities from fitting the full model for VL and the AIS solution with J = 512 . The estimates are generally in agreement but the AIS posterior means are closer to the true parameter values ( w1 = w2 = 1 , w3 to w10 equal to 0 ) for eight out of ten parameters . This is reflected in the higher joint probability mentioned above . Given that we know the true parameters we can also compute the Root Mean Squared Error ( RMSE ) between true and posterior mean parameters . For VL this is 0 . 21 and for AIS it is 0 . 11 . Annealed Importance Sampling has a number of appealing properties . It can provide accurate estimates of the posterior parameter distribution and of the model evidence by avoiding local maxima and without making assumptions of Gaussianity . Samples from AIS converge in distribution to the true posterior density . Sub-optimal model evidence approximations [46] based on the Prior Arithmetic Mean ( PAM ) or Posterior Harmonic Mean ( PHM ) emerge as special cases of AIS with only two temperatures . Unlike Markov chain Monte Carlo methods , the samples produced are not serially correlated thus making any corrections involving effective sample size unnecessary . We have described an implementation of AIS using a transition kernel based on an LMC sampler . The use of LMC here is critical as it allows proposals to be made based on local gradient and curvature information . Our empirical results show that the resulting proposals are accepted with probabilities in a desirable range ( similar to the target of 20 to 40% in Zhou et al . [38] ) even for nonlinear dynamical systems models at low temperature . We have compared AIS to inferences based on the VL approximation in the context of neural mass models . In terms of the estimation of Bayes factors , the two methods agree as to which model is best but report different degrees of belief , especially at high signal to noise ratio . AIS tends to produce higher model evidence estimates both for optimal and suboptimal models . AIS finds better parameter estimates than does VL , as quantified by the joint log probability , especially in data regimes with high signal to noise ratio . A possible explanation as to the dependence on SNR could be that there are more or deeper local minima at high SNR . Moreover , a multistart VL procedure with computer time matched to AIS does not find better solutions . Additionally , we found evidence of non-Gaussianity in the AIS posteriors . Thus it appears that AIS is useful due to its ability to avoid local maxima , and its ability to characterise non-Gaussian parameter posteriors . We have also used an Importance Sampling procedure to estimate the model evidence . This method , which we’ve referred to as ISVL , is highly computationally efficient as it uses the posterior from VL as a proposal density , but it proved unreliable . Similarly , other more standard approaches such as AMC worked well on linear and nonlinear regression problems but it was not possible to derive good AMC-based model evidence estimates for neural mass models . In order to apply AIS one must decide upon an annealing schedule and in this paper we used a 5th-order geometric schedule , discretised using 512 temperatures and explored using 32 trajectories . This proved sufficient over a range of statistical models from linear and nonlinear regression to nonlinear differential equation models . Our empirical work has shown that the required number of temperatures and trajectories did not show a strong dependence on the number of model parameters or model nonlinearity . However , the need to specify the parametric form of the schedule , number of temperatures and trajectories is clearly a weakness of the AIS approach and is an area of ongoing research . Previous work in this direction has focussed on the Sequential Monte Carlo ( SMC ) method which can be viewed as a generalisation of AIS . SMC represents probability densities using particles , as in the particle filter , but is applied at a sequence of temperatures rather than to a sequence of temporally ordered data . In particular Zhou et al . [38] have shown how SMC can be used for model comparison . Automatic annealing schedules can be derived by resampling at every temperature so as to maximise the effective sample size of the particle ensemble . An alternative approach grounded in statistical physics is based on the notion of contact flows and thermodynamic processes [47] . A potential drawback of SMC as compared to AIS , however , is that because particles interact during optimisation , SMC is not amenable to embarrasing parallelisation . Additionally , an application of SMC to nonlinear differential equations [38] used a similar number of temperatures as we do ( 500 as compared to our 512 ) but used many more trajectories ( 1000 as compared to our 32 ) . This suggests that SMC may be more computationally demanding . Another development in this direction is Langevin Importance Sampling [48] which does not require specification of an annealing schedule as temperatures are sampled using Langevin dynamics . This flexibility again comes at the cost of interaction among trajectories ( or particles ) and therefore also compromises parallelisation . Beal [23] has also suggested interesting ways of improving AIS . First , automatic annealing schedules could be produced by introducing finer graining of temperatures in regions of the path for which forward and reverse estimates are inconsistent . Second , Eq 11 suggest that better model evidence estimates could be produced by generating more samples at each temperature . This algorithim would then become more similar to thermodynamic integration [46] which , however , is naturally more computationally demanding than AIS [24] . Whilst our model fitting using AIS was parallelised over multiple cores , alternative efforts can be made to speed up implementation . For example , Wang et al . [49] have shown how the integration of neural mass models can be implemented on Graphical Processing Units ( GPUs ) , resulting in a reduction of computing time by a factor of approximately seven . Additionally , Aponte et al . [50] have pursued a similar GPU approach for DCM for fMRI and shown how it can be used in the context of model evidence computation using thermodynamic integration . This GPU approach has been used to estimate parameters of DCM for fMRI models using an Adaptive Monte Carlo algorithm , again resulting in an order of magnitude reduction in computation time [41] . See also [51] for generic methods for parallelisation of single Markov chains . Dynamical models have also been fitted to neuroimaging data using a range of global optimisation methods . For example , mean field models have been fitted to EEG using particle swarm optimisation [52] and stochastic nonlinear oscillator models have been fitted to EEG using a multi-start algorithm [53] . Additionally , DCMs have been fitted to fMRI data using a method that combines local search with Gaussian process approximation [41] . This method provides better parameter estimates than VL with only a modest increase in computational cost ( much less than AIS ) . However , like the other global optimisation methods ( see also [54] ) , it does not produce an estimate of the posterior distribution or model evidence . This paper has compared the ability of VL and AIS to make inferences about two-region neural mass models based on simulated data . These simulations are a caricature of the DCM for ERP approach [17] as they are simplified in a number of respects ( i ) we have fixed parameters such as time delays between regions , synaptic time constants and synaptic response magnitudes , to known true values , ( ii ) we have not estimated observation noise , ( iii ) we have used only two brain regions whereas most practical applications use upwards of four [55–57] , ( iv ) we have assumed that the electrical activities of brain regions are directly observed , rather than being filtered through a lead field matrix to produce observations in M/EEG sensor space , ( v ) we have used simulated rather than empirical M/EEG data . Further work will be needed to establish whether the findings from our caricature follow over to DCM for ERP . This paper has used an independent model optimisation approach to compute Bayes factors , in which the evidence is computed separately for each model of interest . But in the context of AIS one can traverse a path from the posterior of one model to the posterior of another , with the resulting importance weights providing a direct approximation of the corresponding Bayes factor [13] . Direct computation of Bayes factors in this way is also possible in the context of SMC and a transdimensional AIS algorithm [58] . If one has a nested model , as in the empirical NMM examples in this paper in which the reduced model is nested within the full model model , Savage-Dickey approximations can also be used [59] . It would be interesting to compare Savage-Dickey against the direct path integral methods based on AIS . This paper has explored one method for combining VL and sampling methods , ISVL , in which the VL posterior is used as a proposal density for importance sampling . However , this method did not provide good estimates of the model evidence . Other proposals for combining sampling with variational methods view the sequence of samples produced by a Markov chain as auxiliary variables in a variational inference problem [60] . An alternative approach , proposed in [13] would be to use AIS to traverse a path from the VL posterior to the true posterior at a series of intermediate temperatures , another interesting avenue for future work .	The activity of populations of neurons in the human brain can be described using a set of differential equations known as a neural mass model . These models can then be connected to describe activity in multiple brain regions and , by fitting them to human brain imaging data , statistical inferences can be made about changes in macroscopic connectivity among brain regions . For example , the strength of a connection from one region to another may be more strongly engaged in a particular patient population or during a specific cognitive task . Current statistical inference approaches use a Bayesian algorithm based on principles of local optimization and the assumption that uncertainty about model parameters ( e . g . connectivity ) , having seen the data , follows a Gaussian distribution . This paper evaluates current methods against a global Bayesian optimization algorithm and finds that the two approaches ( local/global ) agree about which model is best , but finds that the global approach produces better parameter estimates .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "engineering", "and", "technology", "applied", "mathematics", "signal", "processing", "neuroscience", "simulation", "and", "modeling", "algorithms", "regression", "analysis", "optimization", "mathematics", "statistics", "(mathematics)", "neuroimaging", "research", "and", "a...	2016	Annealed Importance Sampling for Neural Mass Models
The replication terminus region ( Ter ) of the unique chromosome of most bacteria locates at mid-cell at the time of cell division . In several species , this localization participates in the necessary coordination between chromosome segregation and cell division , notably for the selection of the division site , the licensing of the division machinery assembly and the correct alignment of chromosome dimer resolution sites . The genome of Vibrio cholerae , the agent of the deadly human disease cholera , is divided into two chromosomes , chrI and chrII . Previous fluorescent microscopy observations suggested that although the Ter regions of chrI and chrII replicate at the same time , chrII sister termini separated before cell division whereas chrI sister termini were maintained together at mid-cell , which raised questions on the management of the two chromosomes during cell division . Here , we simultaneously visualized the location of the dimer resolution locus of each of the two chromosomes . Our results confirm the late and early separation of chrI and chrII Ter sisters , respectively . They further suggest that the MatP/matS macrodomain organization system specifically delays chrI Ter sister separation . However , TerI loci remain in the vicinity of the cell centre in the absence of MatP and a genetic assay specifically designed to monitor the relative frequency of sister chromatid contacts during constriction suggest that they keep colliding together until the very end of cell division . In contrast , we found that even though it is not able to impede the separation of chrII Ter sisters before septation , the MatP/matS macrodomain organization system restricts their movement within the cell and permits their frequent interaction during septum constriction . Most bacteria harbour a single chromosome and , in the rare case in which the genetic material is divided on several chromosomes , the extra-numerous ones appear to have derived from horizontally acquired mega-plasmids that subsequently gained essential genes [1] . This is notably the case for Vibrio cholerae , the agent of the deadly human diarrheal disease cholera , whose genome is divided between a 2 . 961 Mbp ancestral chromosome , chrI , and a 1 . 072 Mbp plasmid-derived chromosome , chrII [2] . The preferential transcription of chrII genes during colon colonization compared to in vitro growth under aerobic conditions suggests that this genomic organization is important for rapid adaptation to different environments [3] . Likewise , other bacteria harbouring multipartite genomes can adopt several different life cycles [4] , [5] , [6] , [7]: the rhyzobium , the burkholderia and the vibrio , can alternatively spread freely in the environment or interact as symbionts or pathogens with eukaryotic cells; the borrelia are obligate parasites that need to infect several different eukaryotic organisms in the course of their life cycle . Thus , multipartite genomes seem to offer a selective advantage for the adaptation to very different environmental conditions . However , the necessary coordination between replication , chromosome segregation and cell scission raises questions on the management of the different chromosomes of such bacteria . Bacterial chromosomes harbour a single origin of bidirectional replication and are generally circular . Replication ends in a region opposite of the origin of replication , the terminus region , in which is usually found a specific recombination site dedicated to the resolution of chromosome dimers , dif [8] . Fluorescent microscopic observation of chromosome segregation in mono-chromosomal bacteria revealed that it is concurrent with replication and starts with the active positioning of sister copies of the origin region into opposite cell halves [9] , [10] , [11] . As replication progresses along the left and right chromosomal arms , newly replicated loci are progressively segregated towards their future daughter cell positions . However , the mean time during which sister loci remain together before separation is variable [12] . In particular , sister copies of the terminus region co-localize at mid-cell until the initiation of cell division in E . coli and P . aeruginosa [9] , [10] , [13] , [14] . This mode of segregation can participate in the coordination between chromosome segregation and cell division . Indeed , nucleoid occlusion factors impede the assembly of the cell division machinery until a time when the only genomic DNA left at mid-cell consist of the sister copies of the terminus region in Escherichia coli and Bacillus subtilis [15] , [16]; the long co-localization of sister termini at mid-cell is at least in part dictated by the MatP/matS macrodomain organisation system in E . coli [17] , [18]; a DNA translocase , FtsK , which is recruited to mid-cell as part of the divisome and which pumps chromosomal DNA in the orientation dictated by repeated polar motifs that point towards dif , the KOPS , promotes the orderly segregation of the DNA within the terminus region of E . coli chromosome [13] , [19] , [20] . One of the functions of FtsK is to control the resolution of chromosome dimers , which result from homologous recombination events between circular sister chromatids , by the addition of a cross-over between sister dif sites at the time of constriction [21] . FtsK is also thought to participate in sister chromatid decatenation [22] , [23] and to create a checkpoint to delay constriction until sister terminus regions have been fully segregated [19] , [24] , [25] . V . cholerae chrI and chrII are circular and harbour a single dif site in the region opposite of their origin of replication , dif1 and dif2 , respectively ( Figure 1A , [26] ) . Segregation of the two chromosomes is concurrent with replication and both chromosomes adopt a longitudinal organization within the cell [27] . However , chrII is replicated late in the C period of the cell cycle , when most of chrI has been replicated , and the initiation of its segregation is consequently delayed [28] . In addition , the origin region of chrI , OriI , locates to the old pole of newborn cells and one OriI sister migrates to the other pole after replication ( Figure 1A , [27] ) . The origin region of chrII , OriII , locates to mid-cell in newborn cells and the two OriII sisters migrate towards the ¼ and ¾ positions after replication ( Figure 1A , [27] ) . This is at least in part dictated by the presence of a partition machinery of a chromosomal type on chrI , parABS1 , and a partition system that groups with plasmid and phage machineries on chrII , parABS2 ( Figure 1A , [27] , [29] , [30] ) . The last chromosomal regions to be segregated are the terminus regions of chrI and chrII , TerI and TerII , respectively ( Figure 1A , [27] ) . Both TerI and TerII locate at or close to the new pole in newborn cells ( Figure 1A , [27] ) . Replication termination of the two V . cholerae chromosomes is synchronous [28] and unreplicated TerI and TerII are recruited to mid-cell at approximately the same time ( Figure 1A , [27] ) . V . cholerae is closely related to E . coli in the phylogenetic tree of bacteria and its genome harbour the same dam co-occurring DNA maintenance machineries as E . coli [31] . This includes a unique E . coli MatP ortholog and the presence of cognate matS sites in both TerI and TerII . In addition , a common pair of tyrosine recombinases , XerC and XerD , serves to resolve dimers of each of the two V . cholerae chromosomes despite the sequence divergence of dif1 and dif2 [26] . Dimer resolution is controlled by a unique E . coli FtsK ortholog , whose translocation activity is oriented by KOPS motifs that point towards the dimer resolution site of each of the two chromosomes [26] . By analogy to E . coli , MatP is thought to maintain sister copies of TerI and TerII at mid-cell and FtsK to promote the orderly segregation of the DNA within TerI and TerII . Correspondingly , the separation of sister copies of a locus situated at 40 kbp from dif1 seemed coordinated with cell division ( Figure 1A , [27] , [32] ) . However , sister copies of a locus situated at 49 kbp from dif2 separated before cell division , which questioned the role of FtsK and MatP on TerII segregation ( Figure 1A , [32] ) . The aim of this work was to identify the contribution of MatP to the segregation dynamics of TerI and TerII . We show by replication profiling that dif1 and dif2 are located next to the replication terminus of chrI and chrII , respectively . Simultaneous visualization of the positions of dif1 and dif2 within the cell then allowed us to confirm the late and early separation of TerI and TerII , respectively . However , we show that TerII sisters keep colliding with each other at mid-cell during constriction by genetically probing the relative frequency of sister chromatid contacts occurring at mid-cell at the time of cell division along the two chromosomes and by time-lapse fluorescent microscopy . We further show that the frequency of these collisions depends on the MatP/matS macrodomain organization system , possibly because it restricts the movements of TerII within the cell . We also show that MatP promotes the late mid-cell co-localization of TerI sisters . However , TerI loci remain in the vicinity of the cell centre and sister chromatid contacts remain frequent in its absence . Replication profiling of V . cholerae cells by deep sequencing indicated that termination most frequently occurred at a distance of ∼90 kbp and ∼70 kbp from the reference loci that had been used by Srivastava et al . for the simultaneous visualization of the positions of TerI and TerII ( Figure S1A , [32] ) . It was therefore possible that the behaviour of these loci did not fully reflect TerI and TerII segregation dynamics . To confirm the segregation pattern of the terminus regions of chrI and chrII , we simultaneously visualized the intracellular location of dif1 and dif2 in cells that were exponentially growing in minimal media . We used the lacO/LacI-mCherry system to label the dif1 locus and the pMT1 parS/yGFP-ParB system to label the dif2 locus . Cells were classified according to their length in bins of 0 . 25 µm . They had a median length of 3 . 2 µm ( Figure S2A ) . The smallest cells , i . e . the youngest cells , contained a single dif1 spot at one of the two cell poles ( Figure 1B ) . This pole , which results from the previous division event , is hereafter referred to as the new pole . The preferential localization of dif1 towards the new pole was used to orientate the cells . A single dif2 spot was also observed in the youngest cells ( Figure 1B ) . This spot was located in the younger cell half , at an intermediate position between the dif1 spot and the middle of the cell ( Figure 1B ) . The polarity of the dif1 and dif2 spots decreased as a function of cell elongation and the median position of each spot reached mid-cell in cells of an intermediate length ( Figure 1B ) . The majority of the longest cells , i . e . the closest to cell division , displayed a single dif1 spot , which was located at mid-cell and was flanked by two dif2 spots ( Figure 1B ) . Indeed , <15% of the cells from the 4 . 25 µm–4 . 5 µm bin displayed two dif1 spots whereas >80% of them displayed two dif2 spots ( Figure 1C ) . In addition , the proportion of cells containing two dif2 spots reached 100% in the cells that were longer than 4 . 5 µm whereas only 50% of these cells displayed two dif1 spots ( Figure 1C , grey points ) . Marker frequency analysis indicated that the earlier timing of appearance of cells with two dif2 foci was not due to an earlier timing of replication of dif2 compared to dif1 ( Figure S1A ) . The same pattern of segregation was observed when the dif1 and dif2 labelling systems were switched , excluding any possible artefact linked to the visualization strategy ( Figure S3 ) . Finally , dif2 sisters were found to segregate further away from each other and from mid-cell than dif1 sisters ( Figure 1G ) . Taken together , these results suggest that in the vast majority of cases TerII sisters separated before cell division whereas TerI sister separation was delayed until the end of cell division . We next investigated the influence of the MatP/matS macrodomain organization system on TerI and TerII segregation . V . cholerae cells in which MatP was disrupted were slightly longer than wild-type cells . In minimal medium , they had a median length of 3 . 77 µm ( Figure S2B ) . Nevertheless , growth competition indicated that they lost less than 0 . 23% of fitness per generation ( Figure S4 ) . The smallest cells had a single dif1 and a single dif2 spot , which were both positioned closer to mid-cell than in wild-type cells ( Figure 1D ) . This was accompanied by an increase in position variability ( Figure 1D ) . As a consequence , mid-cell recruitment was no longer directly observable in cells of intermediate lengths ( Figure 1D ) . In addition , the timing of separation of dif1 spots was now very similar to the timing of separation of dif2 spots ( Figure 1E ) . Marker frequency analysis indicated that this was not due to a change in the relative replication timing of dif1 and dif2 ( Figure S1B ) . Many cells of intermediate length now displayed two dif1 and two dif2 spots and most of the cells of the following bins had two dif1 and two dif2 spots ( Figure 1E ) . This was directly reflected in the proportion of cells displaying a single dif1 spot and a single dif2 spot and the proportion of cells with two dif1 and two dif2 spots in the entire population ( Figure 1F , number of spots ) . The separation of dif2 sisters remained slightly ahead of the separation of dif1 sisters ( Figure 1E ) , which was reflected in the higher proportion of cells harbouring a single dif1 spot and two dif2 spots than cells harbouring a single dif2 spot and two dif1 spots ( Figure 1F , 3 spots disposition ) . However , the disposition of spots became more random and many cells now displayed dif2 spots more centrally located than dif1 spots ( Figure 1F , 3 spots disposition and 4 spots disposition ) . Finally , sister dif sites migrated to opposite cell halves after their separation ( Figure 1D ) and the distances between the sisters of both sites were similar ( Figure 1G ) . Taken together , these results suggested that MatP contributed to the precise positioning of TerI before and after replication and that it delayed the separation of TerI sisters to the time of cell division . MatP also contributed to the precise positioning of TerII . However , it was unable to impede TerII sisters from separating before septum constriction . As the densities of matS sites in TerI and TerII are very similar , we were intrigued by the apparent inability of MatP to block TerII sister separation . Lesterlin et al . designed an assay based on the interruption of the lacZ reporter gene by two copies of loxP to detect sister chromatid contacts ( SCC ) behind replication forks [33] . The assay was based on the proximity of the loxP sites: the cleavage points of the Cre recombinases on each strand of the tandem sites were separated by only 55 bp to prevent intra-molecular recombination . As a result , a functional lacZ ORF could only be reconstructed via intermolecular recombination events ( Figure 2A ) . As dif-recombination is under the control of FtsK in V . cholerae [26] , which was expected to restrict it to mid-cell and to the time of septum constriction [21] , we reasoned that 55 bp dif-cassettes could be used to monitor the proximity of TerI and TerII sisters to the cell division machinery at the time of constriction ( Figure 2B ) . We engineered a strain in which XerC production was under the control of the arabinose promoter to permit the stable inheritance of dif-cassettes . To help repress any leaky XerC production , we inserted the E . coli lacZ promoter and the E . coli lacI repressor gene in anti-orientation at the end of the xerC ORF . We also replaced the ATG translation initiation codon by the less favourable TTG codon and removed the ribosomal binding site ( Figure 2B ) . The dif sites harboured by the first and second chromosomes of the El Tor N16961 strain , dif1 and dif2 , possess divergent overlap regions ( Figure 2C , [26] , [34] ) . To compare the excision of 55 bp dif1- and dif2-cassettes ( lac2dif1 and lac2dif2 ) , we inserted them at the same genomic position , in place of the dif locus of chromosome II , and monitored the frequency of full blue colonies that were obtained three hours after the induction of XerC production ( Figure 2C ) . Recombination worked well for both dif sites ( Figure 2D ) . In both cases , blue colony formation strictly depended on XerC production and on the presence of a fully functional ftsK allele ( Figure 2D ) . Little or no recombination can occur between dif1 and dif2 thanks to their sequence divergence ( Figure 2C ) . The use of lac2dif2 on chrI and lac2dif1 on chrII thus prevented any risk of Xer-mediated intrachromosomal rearrangements due to recombination between the dif sites of the cassette and the dimer resolution site of the chromosome during the course of the experiment ( Figure 2E ) . Therefore , the dimer resolution site of the chromosome could be left , which avoided any artefact in the measured excision frequencies linked to the formation of chromosome dimers by recombination between sister copies of the cassettes ( Figure S5 ) . The dif sites of the cassettes used on each of the two V . cholerae chromosomes are identical to the dimer resolution site of the other chromosome . However , this site did not influence the proportion of blue colonies that were formed ( Figure 2E and Figure S6 ) . Both intramolecular and intermolecular recombination events can generate single dif site products . In contrast , three dif site products can only be generated via intermolecular recombination . Such products are transient because they can be converted to single dif products by subsequent intramolecular recombination ( Figure 2A ) . Nevertheless , we could detect their appearance with 55 bp cassettes , demonstrating that recombination occurred via SCC ( Figure 3A ) . As a point of comparison , we engineered 1 kbp dif-cassettes , a distance sufficient for intramolecular recombination . With such cassettes , we did not observe any intermolecular recombination intermediates , suggesting that 1 kbp cassette excision mainly resulted from intramolecular recombination events on separate chromatids ( Figure 3A ) . FtsK-YFP localized to mid-cell in long cells ( Figure 3B , white arrow ) and at one of the two poles in short cells ( Figure 3B , white arrow head ) . This was reminiscent of the pattern of localization of the cell division machinery of Caulobacter crescentus , which assembles at mid-cell but remains bound to the new pole after cell scission [35] . Time-lapse observations confirmed that such a scenario applied to V . cholerae FtsK , demonstrating that it assembled at mid-cell as part of the cell division machinery ( Figure S7A ) . In addition , treating cells with cephalexin , which blocks septum constriction , led to a dramatic reduction in the level of dif-recombination without affecting the recruitment of FtsK to the cell division apparatus ( Figure 3C ) . No loss of cell viability was observed during the course of the cephalexin treatment ( Figure S7B ) . We conclude that dif-recombination occurs during or shortly after septum constriction in V . cholerae . Finally , deletion of recA did not affect the proportion of excision events that could be detected using 55 bp- and 1 kbp-cassettes , indicating that activation of dif-recombination was independent from chromosome dimer formation in V . cholerae ( Figure 3D ) . This result is strikingly different from what is observed using dif-cassettes in E . coli [36] , [37] . The reasons for this difference are the subject of another study ( Gally , Midonet , Demarre and Barre , unpublished results ) . Taken together , these results demonstrate that the proportion of blue colonies formed following lac2dif1 and lac2dif2 recombination events can be used as a relative measure of the respective frequency of contacts between monomeric sister chromatids that occur at mid-cell at the time of septum constriction in V . cholerae . Cells in which lac2dif2 were inserted in the immediate vicinity of dif1 yielded a high level ( ∼60% ) of blue colonies , demonstrating dif1 SCC during constriction ( Figure 4A ) , in agreement with the co-localization of dif1 sisters ( Figure 1 ) . However , interchromatid recombination dropped rapidly when lac2dif2 was not in the immediate vicinity of the dif1 locus ( Figure 4A ) . The frequency of blue colony formation did not diminish in cells in which recA was deleted , confirming that 55 bp cassette recombination on chrI was not restricted to chromosome dimers ( Figure S8A ) . Strikingly , we obtained a very high proportion of blue colonies ( ∼90% ) when lac2dif1 was inserted at dif2 ( Figure 4B ) despite the apparent early separation of dif2 sisters ( Figure 1 ) . In addition , blue colony formation remained high ( ∼45% ) within a 160 kbp region surrounding dif2 , from a position at 9 kb on the left of the dif locus to 152 kb on the right of it ( Figure 4B ) . The same results were obtained after recA deletion , confirming that TerII SCCs were unlikely due to chromosome dimers ( Figure S8A ) . Taken together those results suggested that dif2 sisters contacted each other at mid-cell at the time of cell division as frequently as dif1 sisters , despite their apparent early separation . On chrII , the extent of the region displaying a high frequency of SCC at the time of septum constriction corresponded to the putative MatP domain ( Figure 4B ) . The only notable exception was next to a matS site that is isolated from the rest of the matS region by the V . cholerae superintegron ( Figure 4B ) . Correspondingly , we observed more than a 4-fold reduction in blue colony formation within TerII upon matP disruption ( Figure 4C ) . Indeed , dif2 was the only locus where cassette excision remained above the background level ( Figure 4C ) . Cassette excision remained independent from chromosome dimer formation ( Figure S8B ) . In contrast , the disruption of matP only had a very modest , albeit significant , effect on SCCs within TerI ( Figure 4C ) . The remaining SCCs were still independent from homologous recombination ( Figure S8B ) . Correspondingly , SCCs occurred in a much smaller region than the putative MatP domain on chrI ( Figure 4B ) . Taken together , these results suggested that MatP was the main contributor to TerII SCC occurring at mid-cell at the time of cell division . The high frequency of SCCs detected at dif2 with our genetic assay suggested that dif2 sisters frequently collided at mid-cell during septum constriction despite their early separation . To directly demonstrate that such collisions occurred , we followed the segregation dynamics of dif2 sisters by time-lapse fluorescence microscopy . We expected collisions to be transient because two dif2 spots were observed in almost all of the wild type cells longer than 4 . 5 µm ( Figure 1 and Figure S9A ) . Therefore , we reasoned that short time intervals had to be used between each image acquisition . However , a balance had to be achieved between the detection of the supposedly transient dif2 collisions and the fraction of the cell cycle during which dif2 spots could be tracked in any given cell due to photobleaching . With 30 s time intervals , dif2 foci could be observed for 100 min . A total of 74 wild-type cells were followed , out of which 44 showed a complete cell division event . In 42 of these cells , i . e . in ∼95% of the observed cell division events , dif2 sisters separated before septum invagination , in agreement with our snapshot analysis . However , dif2 sister collisions were frequent ( Figure 5A and Movie S1 ) . As a result , dif2 sisters were found to co-localize at mid-cell at some stage of the cell constriction process in 70% of the cells , which fits with the high frequency of dif2 SCCs observed with the genetic assay ( Figure 5A and Movie S1 ) . On average , 3 . 2 collisions were observed after the initial separation of the dif2 sisters and before cell fission . In the majority of cases , re-joining of the dif2 sisters was transient , i . e . co-localization was only observed during 2 consecutive frames . In some instances , however , dif2 sisters remained co-localized for several minutes . We also followed 131 matP− cells , out of which 30 displayed a complete analysable cell division event . In all of these cells , dif2 sisters separated before septum invagination ( Figure 5A and Movie S2 ) . The positions of the two dif2 sisters were no longer restricted to the ¼–¾ cell region and , in several cases , one of the two dif2 spots located near the old pole at the time of division ( Figure 5A and Movie S2 ) . Indeed , only 0 . 6 collisions were observed on average in each cell after the initial separation of the dif2 sisters and before cell fission . These events lasted for a single frame in the vast majority of cases . Finally , co-localization of the dif2 sisters during septum constriction was only observed once , which fits with the loss of dif2 SCCs monitored with the genetic assay . Taken together , these results suggested that MatP allowed FtsK to process dif2 sisters during cell division by restricting the range of their movements to the ¼–¾ cell region and that other factors played a similar role for dif1 sisters in its absence . Possibly the most striking observation of our study was that TerII sisters kept colliding against each other at mid-cell after their initial separation in the cell cycle , up to and after the initiation of the constriction process ( Figure 4 and 5 ) . During the three hours of our genetic assays , cells underwent ∼8 divisions , as judged by the number of colony forming units at the beginning and at the end of the experiments . Therefore , the ∼90% frequency of blue colony formation that we observed with a recombination inserted at dif2 corresponded to a rate of 25% of β-galactosidase+ cell formation per generation . As only one out of the two possible intermolecular recombination events could yield β-galactosidase+ cells ( Figure 2A ) , this result suggested that >50% of SCC occurred between TerII sisters during each cell division event ( Figure 4 ) . Moreover , we observed the same frequency of blue colony formation with the lacZdif2 probe when it was inserted at the dif2 locus on chrII ( Figure 2D , lac2dif2 ) and when it was inserted at the dif1 locus on chrI ( Figure 4B , dif1 locus ) , suggesting that SCCs at cell division were as frequent within TerII as within TerI . Accordingly , frequent collisions of dif2 sisters were observed at the time of cell division when following the growth of individual cells by fluorescence microscopy with 30 s time intervals ( Figure 5 ) . Interchromatid recombination events during constriction were observed in a specific 160 kb region of chrII , which corresponded to the putative MatP domain of the chromosome ( Figure 4 ) . The relative frequency of interchromatid recombination curve consisted of a plateau with a central peak at the dif2 locus ( Figure 4 ) . Our results suggested the plateau was due to the action of the MatP/matS system ( Figure 4 ) . Our snapshot analysis of the positioning of dif1 in wild type and matP− cells indicated that MatP was a major contributor to the organization and management of TerI at the time of cell division , as observed in E . coli ( Figure 1 ) . However , the relative frequency of interchromatid recombination curve on chrI simply consisted of a sharp peak centred on dif1 with no plateau in the MatP region ( Figure 4 ) . In addition , the relative frequency of SCCs was not dramatically affected in matP− cells ( Figure 4 ) . This is in sharp contrast to what we could have expected based on the role of MatP in the formation of a FtsK loading region in E . coli [13] . Taken together , these observations suggest that other factors than MatP contribute to the management of dif1 sisters at the time of cell division , which partially masked its action in our genetic assay . We are currently investigating the relative contribution of likely candidates for TerI mid-cell localization using the power of our SCC assay . We think that these factors might be common to other bacteria in which sister copies of the terminus regions remain at mid-cell for a long period during cell division , such as P . aeruginosa and C . crescentus . However , they could not , or might not yet , be adapted to the management of the recently acquired chrII of V . cholerae . As a result , the MatP/matS system was left as the sole contributor for TerII SCCs during cell division , which helped reveal its action . The disruption of matP had a profound impact on the subcellular localization of dif1 and dif2 ( Figure 1 ) . In particular , MatP seemed to impede the separation of dif1 sisters until cell division ( Figure 1 ) . MatP is able to create bridges between two matS sites [38] . However , we do not think that sister chromatids are tethered together by such bridges since MatP did not impede the separation of dif2 sisters ( Figure 1 ) . Careful analysis of the location of dif1 and dif2 spots in wild type and matP− cells rather suggested that MatP helped create a molecular leash that confined Ter regions in the ¼–¾ portion of the cell: even though the median positions of dif2 sisters in the cell population indicated their separation before cell division , they did not migrate very far apart from each other and away from mid-cell ( Figure 1 ) . In particular , ∼90% of dif2 spots were located at a distance of less than a quarter of the cell length in cells longer than 4 . 5 µm ( Figure S9A ) . Results from our genetic assay suggested that the movements of such sister loci around the median position probably allowed for their frequent collision at mid-cell at the time of cell division . Even though their medians were equivalent , the distributions of the distances between dif2 sisters in wild type and matP− cells were markedly different ( Figure 1G ) . Indeed , in matP− cells longer than 4 . 5 µm , only ∼57% of the dif2 spots remained in the ¼–¾ portion of the cell ( Figure S9B ) . This might be sufficient to explain a large drop in sister collisions . In contrast , ∼83% of dif1 spots remained in the ¼–¾ portion of the cell , which might explain the low impact of the matP disruption on the frequency of SCC ( Figure S9C and S9D ) . Further work will be required to investigate the molecular nature of the MatP leash . An attractive possibility would be that MatP restrains the movement of catenation loops between the two circular chromatid sisters by binding together the matS sites of each sister chromatid . Our results suggest that multiple redundant factors , including MatP in the enterobacteriaceae and the Vibrios , ensure that sister copies of the terminus region of bacterial chromosomes remain sufficiently close to mid-cell to be processed by FtsK . In this regard , it is remarkable to observe that , even though initiation of chrII replication responds to the same global cell cycle regulatory networks than chrI initiation [39] , it occurs at a later time point in the cell cycle [28] , which results in synchronous chrI and chrII replication termination ( Figure S1 , [28] ) . This is likely to participate in delaying TerII sister separation until the time of cell division . We observed that matP− cells were longer than wild type cells in agreement with the notion that coordination of cell division and chromosome segregation is a key feature of the bacterial cell cycle ( Figure S1 ) . What is the functional role of this coordination ? The late segregation of the terminus region might facilitate the action of FtsK on unresolved catenation links or chromosome dimers . Under laboratory conditions , we did not observe any significant chromosome dimer resolution defect ( Figure S4 ) . However , these results have to be interpreted with caution since the disorganization induced by the absence of MatP should only slightly delay the time required for FtsK to bring together sister dif sites . Genetic engineering methods are described in Text S1 . Bacterial strains and plasmids used in this study are listed in Tables S1 and S2 , respectively . All V . cholerae strains were derivatives of the El Tor N16961 strain . A lacO array was inserted adjacent to dif1 and a PMT1 parS was inserted adjacent to dif2 . LacIE . coli-mCherry and yGFP-ParBpMT1 were produced via the leaky expression of a synthetic operon under the E . coli lacZ promoter that was inserted at the V . cholerae lacZ locus . A C-terminal fusion between FtsK and a yellow fluorescent protein , FtsK-YFP , was inserted in place of the endogenous V . cholerae ftsK allele to visualize its localisation . Protocols for Microscopy are detailed in Text S1 . The snapshot images were analysed using the Matlab-based sofware MicrobeTracker [40] , [41] . Details for the analysis are described in [27] . For bright field ( BF ) and fluorescence microscopy 2 µl of an exponentially growing culture sample were placed on a microscope slide coated with a thin agarose layer ( 1% ) made using the growth medium . The slide was incubated at 30°C during the images acquisition . The images were acquired with an Evolve 512 EMCCD camera attached to an Axio Observe spinning disk from Zeiss and recorded every 30 seconds with step size of 0 . 4 µm in the Z-axis ( 3 images were acquired for each channel ) . The BF image 3 is subtracted from the BF image 1 to obtain the phase image . Blue colony formation assay: 0 . 2 mM IPTG were used to repress xerC transcription . 0 . 1% arabinose was used to produce XerC . Freshly grown cultures were diluted in 5 mL of LB supplemented with arabinose to reach 0 . 02 of optical density at 600 nm . They were incubated for 180 mn at 37°C with shaking . Serial dilutions of the cells were plated on LB agar plates supplemented with X-gal and IPTG before and after the induction of recombination . Southern blot assay: Cephalexin was added at the final concentration of 10 µg/ml at the same time as the arabinose . Cells were collected at the beginning of the incubation and after 40 , 80 and 120 mn for genomic DNA extraction . Recombination products were analysed an EcoRV/HphI digest and 1 kbp fragment corresponding to the lacZ promoter as a probe . Signals were detected using a Typhoon instrument and quantified using the IQT 7 . 0 software ( GE Healthcare ) .	The genome of Vibrio cholerae is divided into two circular chromosomes , chrI and chrII . ChrII is derived from a horizontally acquired mega-plasmid , which raised questions on the necessary coordination of the processes that ensure its segregation with the cell division cycle . Here , we show that the MatP/matS macrodomain organization system impedes the separation of sister copies of the terminus region of chrI before the initiation of septum constriction . In its absence , however , chrI sister termini remain sufficiently close to mid-cell to be processed by the FtsK cell division translocase . In contrast , we show that MatP cannot impede the separation of chrII sister termini before constriction . However , it restricts their movements within the cell , which allows for their processing by FtsK at the time of cell division . These results suggest that multiple redundant factors , including MatP in the enterobacteriaceae and the Vibrios , ensure that sister copies of the terminus region of bacterial chromosomes remain sufficiently close to mid-cell to be processed by FtsK .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "cell", "biology", "genetics", "biology", "and", "life", "sciences", "microbiology", "molecular", "cell", "biology" ]	2014	Differential Management of the Replication Terminus Regions of the Two Vibrio cholerae Chromosomes during Cell Division
Coalescent theory is routinely used to estimate past population dynamics and demographic parameters from genealogies . While early work in coalescent theory only considered simple demographic models , advances in theory have allowed for increasingly complex demographic scenarios to be considered . The success of this approach has lead to coalescent-based inference methods being applied to populations with rapidly changing population dynamics , including pathogens like RNA viruses . However , fitting epidemiological models to genealogies via coalescent models remains a challenging task , because pathogen populations often exhibit complex , nonlinear dynamics and are structured by multiple factors . Moreover , it often becomes necessary to consider stochastic variation in population dynamics when fitting such complex models to real data . Using recently developed structured coalescent models that accommodate complex population dynamics and population structure , we develop a statistical framework for fitting stochastic epidemiological models to genealogies . By combining particle filtering methods with Bayesian Markov chain Monte Carlo methods , we are able to fit a wide class of stochastic , nonlinear epidemiological models with different forms of population structure to genealogies . We demonstrate our framework using two structured epidemiological models: a model with disease progression between multiple stages of infection and a two-population model reflecting spatial structure . We apply the multi-stage model to HIV genealogies and show that the proposed method can be used to estimate the stage-specific transmission rates and prevalence of HIV . Finally , using the two-population model we explore how much information about population structure is contained in genealogies and what sample sizes are necessary to reliably infer parameters like migration rates . Genealogies can provide valuable information about the demographic history of a population because the demography of a population can dramatically shape the structure of a genealogy [1] , [2] . For example , fluctuations in population size will shift the distribution of branching events , or coalescent times , over a genealogy relative to what would be expected for a population with a constant size [3] . Other aspects of a population's demographic history can also leave behind distinctive genealogical patterns . For example , the structuring of a population into different subpopulations can influence the topology of genealogies , which is often seen as clustering among individuals sampled from the same subpopulation [4] . These observations have led to great interest in statistical methods for inferring demographic trends and parameters from genealogies and given rise to the new field of phylodynamic inference [2] , [5]–[8] . Most statistical methods for reconstructing the demographic history of a population from genealogies have been motivated by coalescent theory , which provides a probabilistic framework for relating the demographic history of a population to a genealogy of individuals sampled from that population [9] , [10] . Critically , coalescent models provide a way to compute the probability of a given genealogy under a given demographic model . It is therefore possible to estimate parameters of a demographic model , such as population size , from a genealogy using likelihood-based inference methods . Extensions of this basic idea have been used to estimate changes in population size over time , for example by the Bayesian skyline methods available in the BEAST phylogenetic software package [11] , [12] . Coalescent theory has also been extended to consider different forms of population structure , giving rise to structured coalescent models [13] , [14] . Statistical methods that allow fitting of structured coalescent models to genealogies have the ability to estimate parameters relating to population structure , including migration rates between populations [7] , [15] . Recent developments in phylodynamics have focused on developing models and statistical methods for more complex demographic scenarios , which have been largely motivated by the application of coalescent methods to pathogens like RNA viruses with rapidly changing population sizes . For example , coalescent models have been developed for populations where birth ( i . e . transmission ) rates vary over time [16] , [17] . Importantly , the framework of Volz et al . [16] also considers the coalescent process in populations where transmission rates change over time in a nonlinear manner , as is often the case for epidemiological models like the well-known Susceptible-Infected-Recovered ( SIR ) model [18] . Coalescent models have also been developed for common epidemiological scenarios with population structure that alters the rate of coalescence in the population [19] , but these models are limited to populations at equilibrium . Finally , Volz [20] presented a framework that brings together both complex population dynamics and population structure . This approach has great appeal as it generalizes coalescent models to allow both birth and migration rates to change over time as a function of the underlying population dynamics , which may be nonlinear and far from equilibrium . Although recent advances with structured coalescent models have enabled the analysis of more complex epidemiological models , the statistical challenge remains of efficiently fitting stochastic population dynamic models to genealogies . These models can be extremely high-dimensional due to a large number of latent state variables for which we have no direct observations . In Rasmussen et al . [21] , a particle filtering approach was used to marginalize out these latent variables by forward simulating population dynamic trajectories from the epidemiological model and then averaging over these trajectories to a compute a marginal likelihood . For unstructured models , adapting particle filtering methods to coalescent-based inference is relatively straightforward as the likelihood of a genealogy is simply a function of the simulated population dynamic trajectories . However , for structured models the likelihood also depends on the internal states of lineages in the genealogy , which may change over time as lineages move between populations [20] . The probable state of a lineage can only be calculated retrospectively conditional on the population's demographic history and the state of the lineage at the time of sampling . As we show below , these backward-time dependencies prevent the direct application of forward-time particle filtering methods to structured models . We therefore present a new statistical approach for fitting stochastic population dynamics models to genealogies using the structured coalescent approach presented in Volz [20] using a modified particle filtering algorithm . This modified algorithm allows for efficient particle filtering under structured coalescent models where the probability that a lineage is in a certain population may depend on both the past dynamics of the population as well as future sampling of lineages . Using this algorithm , we can fit stochastic , nonlinear epidemiological models with essentially any form of population structure to genealogies as long as the model is Markovian . Because population structure arises naturally in many epidemiological models , we define population structure in a very broad sense and consider any model where the population of infected hosts is structured into different nonequivalent states and therefore lineages in different infected hosts do not necessarily have an equal probability of coalescing . This includes models with spatial structure , multiple stages of infection and models of vector-borne and other multi-host pathogens . The paper has the following structure . First , we present the forward-time epidemiological models that we use as examples throughout the paper . Next , we review the framework first developed in Volz [20] for how coalescent models can be derived for a corresponding forward-time population dynamic model . We then describe how we can fit structured epidemiological models to genealogies given the corresponding structured coalescent model . The statistical method we describe combines MCMC methods with our particle filtering algorithm , and is a variation of the particle MCMC algorithm of Andrieu et al . [22] . Using simulated genealogies , we show that this algorithm can accurately reconstruct population dynamics in structured populations and obtain reliable estimates of epidemiological parameters such as transmission rates . We then apply our approach to the HIV epidemic in Detroit , Michigan in order to estimate stage-specific transmission rates and infer how prevalence and incidence have changed over the course of the epidemic . Finally , we explore under what conditions parameters relating to population structure can be inferred from genealogies and how factors such as sample size affect uncertainty in our estimates . In this paper , we use epidemiological models to demonstrate how mechanistic population dynamic models can be fit to genealogies . More specifically , we will consider the type of Susceptible-Infected-Recovered ( SIR ) models widely used to study the transmission dynamics of infectious diseases [18] , [23] . In SIR-type models , the host population is divided into different compartments depending on the host's state ( e . g . susceptible or infected ) . For generality , we let be the vector that holds the number of hosts in each compartment at time , for example for the standard SIR model . For stochastic models , the state variables in are treated as random variables . We consider an epidemiological model to be structured if there is more than one class of infected host . In this section , we consider formulating structured coalescent models for the type of structured epidemiological models just presented . As shown in Volz [20] , thinking about population dynamic models as simple birth-death processes can be useful when deriving coalescent models that correspond to a given forward-time model . If we randomly sample individuals from a population and trace their ancestry back in time , then coalescent events in the genealogy will correspond to birth events in the population when both the parent and child lineages are ancestral to sampled individuals . While deaths may affect the overall population size , deaths can be ignored along lineages ancestral to sampled individuals because we know that a lineage could not have died out at an earlier time if it persisted to be sampled at some later time . For a structured population , we also must consider individuals transitioning between different subpopulations through migration events that occur independently of birth events , although for the type of models we will consider here a lineage can also transition between populations by being born into a different population than its parent . The same birth-death-migration framework can be applied to pathogens if we assume that each infected host corresponds to a single individual in the pathogen population . In this case , births in the pathogen population occur at transmission events between hosts . Deaths in the population will correspond to recovery or mortality of infected hosts . If each infected host is represented by a single pathogen lineage , coalescent events in the genealogy will correspond to transmission events if both the infected host and the infector are sampled or give rise to descendent infections that are sampled . For structured epidemiological models , we also must consider a pathogen lineage transitioning among populations , or compartments in SIR-type models , independent of transmission events . For example , in the three-stage model , pathogen lineages can transition between different stages of infection . Here , we will refer to all transitions between states that occur independently of transmission as migration for generality . This allows many epidemiological models with some form of population structure to be thought of as a birth-death-migration process . To formalize the birth process , we adopt the notation of Volz [20] and let be a matrix that specifies the birth rate of new lineages in the population at time , where , meaning that can be a function of the epidemiological parameters and the population state variables . Lineages may be in any one of states . The rate at which lineages currently in state give birth to lineages in state is given by the element . The rate at which migration , or transitions between states independent of birth events , occurs is given by another matrix . The rate at which lineages currently in state migrate to state is given by the element . We treat birth and migration as distinct processes because , as we will see , they affect the coalescent process in different ways since coalescent events can only occur at birth events but migration events can affect the probability of a particular lineage coalescing with another lineage . The total number of lineages in each state is given by a vector , such that gives the total number of individuals in the population in state at time . From here in , we drop the time indices and just refer to the matrices and or the vector , but emphasize that the rates in and and the population sizes in can be time-dependent . We illustrate the and matrix notation by decomposing the three-stage and two-population SIR models presented above into their component birth and migration processes . For the three-stage model , we have ( 4 ) ( 5 ) In the matrix , births occur through transmission of the pathogen from any of the three stages of infection to susceptible individuals . Because all new infections begin in the early stage , only the leftmost column of the matrix has nonzero elements . The nonzero elements in the matrix correspond to migration between stages through disease progression from early to chronic and from chronic to AIDS . For the two-population model , we have ( 6 ) ( 7 ) Because transmission events can move the pathogen within and between the two populations in either direction , all entries in the matrix are nonzero . The matrix has all zero entries because there is no migration between populations independent of transmission . Before moving on , we note that for an infectious pathogen our coalescent models make the implicit assumption that coalescent events in the genealogy correspond to transmission events between hosts . In essence then , we are ignoring the within-host coalescent process and assuming that all infected hosts are represented by a single lineage . This implies that lineages immediately coalesce once in the same infected host , which may not be true for certain pathogens where multiple lineages can persist within a host for long periods of time . Nevertheless , in general our assumption that each infected host is represented by a single pathogen lineage will be valid as long as super-infection is rare and there is a strong bottleneck in the pathogen population at transmission events so that it is unlikely that more than one lineage is transmitted between hosts . To fit a structured coalescent model to a genealogy , we need to compute the likelihood of the coalescent model given the genealogy . To compute this likelihood , we can partition the genealogy into any number of discrete time intervals . We label the time partitioned genealogy , where is the time of the first event in the genealogy and is the final event time going forwards in time ( usually the terminal-most sampling event ) . Time points are chosen to correspond to the times at which events in the genealogy occur such as coalescent and sampling events . We can then further subdivide the genealogy into smaller intervals that correspond to the time steps used to simulate from the epidemiological model so that at any time point we have the state variables corresponding to that time . With the time partitioned genealogy , we can compute the likelihood over each interval in the genealogy , , and then take the product over all intervals to compute the total likelihood of the model given . Computing the likelihood over a time interval requires us to first compute the probabilities that the lineages present in the genealogy did or did not coalesce within that time interval . The probability of a coalescent event in turn depends on the expected rate of coalescence under the model . This expected rate can be computed for a coalescent model with any arbitrary population structure using the formalism summarized above for the rates of birth in . As shown in Volz [20] , the rate of coalescence for two lineages and is ( 8 ) where , for example , is the probability that lineage is in state . How these lineage state probabilities are computed is explained below . We can make intuitive sense of the coalescent rate in ( 8 ) by noting that is the total rate at which lineages in state give birth to lineages in state in the population and that is the probability that lineages and are the two lineages involved in a particular birth event . However , since we do not know the true states of and we must sum over all possible combinations of states for these two lineages . The total rate of coalescence for all lineages present in the genealogy over an interval of time is then ( 9 ) Given the rates of coalescence , we can then compute the likelihood over a time interval under the coalescent model . If the time interval does not end in a coalescent event , we have ( 10 ) Alternatively , if the interval does end in a coalescent event between two lineages and , we have ( 11 ) As alluded to above , computing the coalescent rates requires us to compute the probability of each lineage in the genealogy being in each possible state . At the time of sampling , we may know the state of a lineage from information gathered from the infected host from which the sample was obtained . Alternatively , if we do not know the state of the host at the time of sampling exactly , we can assign prior probabilities to the lineage being in each state under a multinomial distribution . Either way , given the initial state or state probabilities at the time of sampling , we need to be able to compute the probability of the lineage being in each state at any point in the past . Going backwards in time , the lineages transition between states at the rates given in the and matrices , which in turn depends on the population states and the parameters . Given these transition rates , we have a continuous time Markov process on a discrete state space along each branch . We can therefore use master equations to track how the lineage state probabilities change going backwards through time . In other words , we can write down differential equations for how the probability mass assigned to each state flows between states as we move into the past . As shown in Volz [20] , the general form that these master equations take for any lineage and state is ( 12 ) where ; that is is the expected number of lineages in state in the genealogy at a given point in time . Further details on how the lineage state probabilities are computed and get updated at coalescent events are given in Text S1 . For convenience , we introduce the notation to denote the lineage state probabilities for all lineages in the genealogy at time and to denote the complete mapping of lineage state probabilities onto the genealogy over the entire time partitioned genealogy . The goal of phylodynamic inference for the type of models presented above will generally be to infer the parameters of interest from the genealogy along with the latent population state variables , such as the number of infected or susceptible hosts over time . In a Bayesian context then , we would like to infer the joint posterior density of the model parameters and the latent state variables . Up to a normalizing constant , this posterior density is given by ( 13 ) From ( 13 ) , we see that this joint density can be factored into three parts: the coalescent likelihood which we outlined how to compute above; the prior density on the population state variables as defined by the epidemiological process model; and the prior density on the parameters . Although we may be able to compute each component individually and thereby the posterior probability of a given set of parameters and population states , the posterior density is not analytically tractable in general and we must resort to sampling from the posterior using MCMC methods . However , it may be difficult or impossible to sample from complex , high-dimensional densities such as using standard MCMC methods . We could , for example , use a Gibbs sampler to iteratively sample from the conditional posterior densities of and any component of , but this strategy can be extremely inefficient owing to strong correlations among the parameters and the state variables , leading to slow MCMC mixing [28] . In Rasmussen et al . [21] , a particle MCMC approach known as the particle marginal Metropolis-Hastings ( PMMH ) algorithm was therefore used to sample from the joint posterior density of and . The main motivation behind using the PMMH algorithm is that we can jointly update and together [22] . Each MCMC iteration , we first propose new parameter values and then run a particle filtering algorithm to get a numerical approximation of the posterior density of the latent state variables , which we refer to as . Particle filtering , also known as sequential Monte Carlo , provides a computational means of approximating high dimensional densities by providing samples ( i . e the particles ) distributed according to the desired density , and are often used in the context of nonlinear and non-Gaussian state space models [29]–[31] . We review how particle filters can be used to fit epidemiological models to genealogies in Text S1 . After running the particle filtering step in the PMMH algorithm , we can then sample a particle from to get a proposal for the latent state variables that is adapted to the parameters in . We can also use the particle filter to compute the marginal likelihood of by marginalizing out the state variables . Because we jointly accept and based on the marginal likelihood , we do not have to independently update , leading to a much more efficient MCMC sampler . Despite marginalizing out the latent state variables , the remarkable feature of the PMMH algorithm is it provides an exact ( i . e . unbiased ) approximation to the density of interest , . The PMMH algorithm is summarized in pseudo-code below . We simulated mock genealogies under each model to test the performance of the PMMH algorithm before applying the method to real data . Mock genealogies were obtained by first forward simulating from the population dynamic model while tracking all infected hosts in the population and the parent-offspring relationships at transmission events . From the forward simulations , we could then trace the lineages of infected individuals backwards through time to obtain the true genealogy for a fraction of sampled lineages . All population dynamic simulations were performed using the tau-leaping algorithm so that the epidemiological dynamics included demographic noise [32] . The three-stage model was parameterized to reflect the natural history of HIV because we planned to apply our method to real HIV genealogies ( see Table 1 ) . We set the disease progression and AIDS death rate to values that give an average time between infection and death of about 10 years , consistent with observed patterns . The incidence scaling parameter was set to zero so that in the simulations there was a linear scaling between incidence and prevalence . The epidemic simulations were seeded with one early-stage infection at time zero and run for 37 years to reflect the timespan of the HIV epidemic in the U . S . To obtain mock genealogies from the complete infection trees , we sampled 200 individuals in the last six years of the epidemic to reflect the fact that most HIV sequences have been sampled in the recent past . For all parameters , we chose to use uniform priors over a wide range of biologically plausible values so that the choice of prior would have minimal influence on our estimates . For the two-population model , we added seasonality to the model by seasonally forcing the base transmission rate using a sinusoidal forcing function , where ( 17 ) The strength of seasonality was the same in both populations but we allowed to differ between the two populations to get asynchronous dynamics between populations . The values of all fixed parameters in the model are also shown in Table 1 . For the genealogies , 120 infected hosts were randomly sampled over time with sampling effort proportional to disease prevalence in each population . For the two-population model , we fixed the initial conditions for the number of susceptible and infected hosts in each population . For the simulation experiments , we wished to compare estimates obtained by fitting stochastic models using the PMMH algorithm against estimates obtained by fitting deterministic models . To fit deterministic models , we used a Metropolis-Hastings sampler where , whenever new parameters are proposed , the likelihood of the genealogy under the new parameters is computed by conditioning on a deterministic trajectory of the state variables simulated from the model using these new parameters . We applied our method to a set of HIV-1 partial pol sequences collected from men who have sex with men ( MSM ) in the metropolitan area of Detroit , Michigan . The dataset contained 437 HIV-1 subtype B sequences which were originally collected for drug resistance testing between 2004 and 2011 . More information about this dataset can be found in Volz et al . [26] . Data were anonymized by staff at the Michigan Department of Community Health before being provided to investigators . Because this research falls under the original mandate for HIV surveillance and was de-identified , it was classified as human subjects research but was exempt from further IRB review . We reconstructed time-scaled genealogies from the HIV sequences in BEAST using a relaxed molecular clock [33] . All sequences identified as likely recombinants were removed from the alignment prior to the analysis . Tips in the genealogy corresponding to sampled infected individuals were assigned prior probabilities of being in each infection stage based on the time since infection estimated from CD4 cell counts and genetic diversity within the host [34] . From the HIV genealogies , we estimated the transmission rates , and as well as the incidence scaling parameter . All other parameters were fixed at the values given in Table 1 . Rather than estimate initial conditions , the time of the initial introduction of HIV into Detroit was estimated , at which point the epidemic was seeded with one early-stage infection in a completely susceptible population . All priors on the parameters were uniform . For the time of initial introduction the prior was truncated at 1973 as a lower bound and the root time of each tree as an upper bound . To ensure our phylodynamic estimates of HIV incidence were reasonable , we compared our estimates against incidence back-calculated from Michigan Department of Community Health surveillance data using the method of Yan et al . [35] . For all results shown in this paper , the PMMH algorithm was run for at least 100 , 000 iterations or until the MCMC fully converged . For the Metropolis-Hastings step , we chose a multivariate normal proposal density for , which can take into account the correlations among different parameters by optimizing the covariance parameters that specify the density . For the particle filter , we found that using a small number of particles ( ) was sufficient . Running the particle filter with a small number of particles tends to increase the error , or variance , in the marginal likelihood estimates . However , this error will not affect inference as long as the marginal likelihood estimates are not systematically biased because the error in the estimates will get averaged out in the encompassing MCMC algorithm . Nevertheless , with too few particles we run the risk of the MCMC getting stuck at erroneously high values of the likelihood . Our choice of was therefore a compromise between minimizing the error in the marginal likelihood estimates and the time taken to run the particle filter . Resampling within the particle filter was done by multinomial sampling with replacement . Resampling times were chosen to minimize the variance in the marginal likelihood estimates and were usually placed around coalescent events , as most of the variation in particle weights arises at coalescent times . The PMMH algorithm was implemented in the software package PHYLter and Java source code is freely available at http://code . google . com/p/phylter/ . Running the PMMH algorithm for 100 , 000 iterations using the simulated HIV genealogies took approximately 10 hours ( 0 . 36 s per iteration ) on a 3 . 4 GHz Intel i7 processor without any parallelization across cores . The most computationally intensive component of the algorithm is computing the lineage state probabilities , which involves numerically solving the master equations for each lineage in the genealogy and has a time complexity of , where is the number of possible lineage states . On the other hand , run times scale linearly with the number of particles and lineages in the genealogy . Thus , the efficiency of the algorithm is mainly limited by the number of states in the model . Before applying the PMMH algorithm to genealogies reconstructed from real data , we ran extensive simulations to ensure that we could accurately recover epidemiological parameters and population dynamics from mock genealogies . We simulated 100 stochastic realizations of an epidemic from the three-stage model , keeping track of the underlying infection tree so that we could obtain the true genealogy for a fraction of sampled lineages . From the simulated epidemic dynamics , we can see that demographic stochasticity generates considerable variation in when the epidemic begins and peaks ( Figure S1 ) . Even with this variability , we accurately inferred stage-specific prevalence and transmission rates from the mock genealogies using the PMMH algorithm ( Figure 1 ) . The 95% credible intervals generally contained the true prevalence for all three stages of infection ( Figure 1A ) . We were also able to estimate the stage-specific transmission rates associated with each stage of infection ( Figure 1B–D ) , even though there were strong correlations among the different transmission rates as seen in the pairwise joint posterior densities ( Figure 1E–G ) . Overall , out of all 100 simulations , the 95% credible intervals contained all three transmission rates 94 times , while the posterior coverage was greater than 95% for each parameter individually . In contrast , when we fit deterministic models to the same set of genealogies , the credible intervals contained the true parameters only 79% of the time . The PMMH algorithm therefore appears to give reliable estimates of parameters and epidemiological dynamics and outperforms deterministic methods when stochasticity plays a role in the epidemic dynamics . Given that we were able to reliably estimate transmission parameters and prevalence in our simulation study , we next applied the method to HIV genealogies reconstructed from sequences collected in Detroit , Michigan . A critical question in HIV epidemiology is to what extent transmission during the early stages of infection contributes to overall HIV incidence . Transmission during early infection may influence the effectiveness of interventions based on antiretroviral treatment in limiting the epidemic [36] , [37] . If most new cases of HIV result from recently infected individuals , then prevention strategies that rely on treating diagnosed individuals , who are likely in later stages of infection , will directly prevent few transmissions . Thus , the transmission rate from early HIV infections ( EHI ) is a key parameter of great interest , although difficult to measure directly from traditional surveillance data . Phylogenetic studies of HIV have used the high degree of clustering and short branch times within these clusters to argue for a high EHI transmission rate [4] , [38] . However , clustering alone cannot be taken as definitive evidence for high EHI transmission as similar patterns can arise simply from epidemic transmission dynamics [26] . In this section , we demonstrate that our inference framework can be used to estimate the EHI transmission rate and the number of new HIV infections attributable to EHI from HIV genealogies using models that explicitly consider HIV's transmission dynamics , as well as the stochastic nature of the epidemic dynamics . Time-scaled genealogies were reconstructed using BEAST from HIV-1 partial pol sequences isolated from men who have sex with men ( MSM ) in the metropolitan area of Detroit . A representative genealogy randomly sampled from the BEAST posterior is shown in Figure S2 . We then fit our three-stage SIR model to 10 genealogies sampled from the BEAST posterior to take into account uncertainty in the genealogy . From these genealogies , we estimated the transmission rate for each stage , including the EHI transmission rate , along with the stage-specific dynamics of prevalence and the incidence ( i . e number of new cases ) attributable to each stage over the course of the epidemic . Parameters estimated from the representative HIV genealogy are shown in Figure 2 and estimates from all 10 genealogies are given in Table 2 . We estimated that transmission rates are higher during the early and AIDS stages than during the chronic stage , as expected from previous studies [39]–[41] . The transmission rate from EHI is about 20 times higher than during the chronic stage and about five times higher than during the AIDS stage ( Figure 2A–C ) . We also found evidence for a nonlinear dependence of incidence on prevalence , quantified through the incidence scaling parameter . Although estimated values of are small , the posterior density is clearly centered away from zero , indicating that incidence scales nonlinearly with prevalence ( Figure 2D ) . Overall , parameter estimates were largely consistent across genealogies , although there was considerable variation in the time of initial introduction of HIV into Detroit estimated from different trees . This is likely attributable to the large amount of variation in the root times inferred for different trees , as we inferred earlier times of introduction from trees with earlier root times . Stage-specific HIV prevalence inferred from the genealogies shows a predictable transition from most infections being in the early stage at the beginning of the epidemic to most infections being in the chronic or AIDS stages later in the epidemic ( Figure 3A ) . This is expected given the longer duration of the chronic and AIDS stages . In general , our phylodynamic estimates of the epidemic dynamics closely track HIV incidence imputed from surveillance data from the beginning of the epidemic through the peak ( Figure 3B ) . While our phylodynamic estimates do not capture the fluctuations in incidence that occur after 1990 , there was nothing in our model that would allow us to reproduce this pattern , which likely results from complex changes in HIV treatment and behavioral changes [34] . Although there was also considerable variability in the population dynamics inferred from different genealogies , this variation occurs primarily during the early stages of the epidemic ( Figure 3C ) . Again , this appears to be associated with uncertainty in the root times of trees; dynamics inferred from trees with earlier root times show an earlier rise and peak in incidence . After the epidemic peaks , the incidence estimated from different trees seems to converge on similar values . Estimates of incidence attributable to each stage show that EHI contributed to most new infections at the beginning of epidemic when EHI prevalence was high ( Figure 3B ) . After the epidemic peak , infections arising from EHI remains high proportional to EHI prevalence , consistent with the higher transmission rate we estimated for EHI . In the late 2000's , we estimated that between 40 to 50% of all new infections arise from EHI , indicating that early stage infections still play a major role in driving HIV transmission . These large estimates for number of new infections arising from EHI are consistent with the phylodynamic estimates of Volz et al . [34] , who fit a more complex but deterministic epidemiological model to the same set of HIV sequences . While our results for the three-stage model suggest that the PMMH algorithm works effectively and can be used to estimate key epidemiological parameters like HIV transmission rates , we were also interested in how much information genealogies contain about the structure of populations in general . To explore this question , we used the two-population model presented in ( 3 ) , for which we can tune the strength of population structure by altering the mixing rate between populations . Mock genealogies were simulated under three values of : low ( 0 . 01 ) , medium ( 0 . 05 ) and high ( 0 . 2 ) . At , for example , about one in every one hundred transmission events occurs between populations . For all three values of , we were able to accurately infer the epidemiological parameters of interest and the population dynamics from the simulated genealogies ( Figure 4 & Figure S3 ) . While we can easily estimate under all three demographic scenarios , the posterior densities become skewed towards increasingly high values of as mixing increases between the populations ( Figure 4A–C ) . This indicates that it may be very difficult to obtain precise estimates of or other parameters pertaining to population structure when populations are only weakly structured . We can visually explore how much information a genealogy contains about population structure and pathogen movement by comparing the true lineage states to the computed lineage state probabilities . In Figure 5A–C , the true state of each lineage over time is mapped onto the genealogies . For ease of viewing , we only display a representative subtree of each genealogy . As expected , under low mixing lineages change states very slowly leading to a high degree of clustering among lineages sampled from the same population , whereas under high mixing lineages move rapidly between states and there is little clustering . We can then compare the true lineage states with the state probabilities computed under the median posterior values of the estimated parameters ( Figure 5D–F ) . When is low , the state of the lineages at the time of sampling is highly informative about the state of the lineage going into the past . However , when we increase to 0 . 05 , the state of the sampled lineages is less informative about the past states and we can see that the lineage state probabilities fluctuate seasonally according to the asynchronous dynamics between populations . When is high , the lineages move between states so rapidly that there is high uncertainty in the lineage states over the entire tree . This loss of information regarding the lineage states is readily observed by considering how the entropy , or uncertainty , in the lineage states changes going backwards in time ( Figure 5G–H ) . Visualizing the flow of information along the lineages in the trees shows how uncertainty in parameters like depends on how rapidly information about the lineage states decays . When is low , lineages remain in the same state long enough that once a coalescent event is reached , information about the probable state of the lineages is still present . In this case , the probable states of the coalescing lineages provides additional information about the transmission event with respect to whether the transmission event occurred within or between populations . By combining information from coalescent events across the entire tree , we can then estimate the rates at which transmission occurs within and between populations . However , if all information about the past lineage states is lost before lineages coalesce , the observed coalescent events will no longer be informative about whether transmission occurred within or between populations and therefore parameters like will be difficult to precisely estimate . The preceding observations about uncertainty in lineage states suggest that it may be possible to estimate more precisely if we increase the number of sampled lineages . Increasing the sampling fraction will also increase the coalescent rate among lineages , thereby increasing the probability of lineages coalescing before all information about their probable state is lost . To test this idea , we simulated genealogies under the same three values of but varied the sample size . With a sample size of 100 , the same as used above , we see that the likelihood is peaked around the true value of when mixing is low but the likelihood profile is fairly flat when mixing is high ( Figure 6A–C ) . Increasing the sample size to 500 resulted in more curved likelihood profiles but the likelihood remains relatively flat with high mixing ( Figure 6D–F ) . Doubling the sample size again to 1 , 000 , the likelihood profiles show significant curvature for all values of ( Figure 6G–I ) . This suggests that while the sample size does play a significant role in determining whether parameters like can be precisely estimated from genealogies , extremely large sample sizes may be required to estimate parameters pertaining to population structure when the population is only weakly structured . The approach outlined in this paper allows for structured , stochastic epidemiological and other population dynamic models to be fit to genealogies in order to jointly infer past population dynamics and model parameters . We believe this to be an important step forward in the field of phylodynamics because many populations are structured in ways that could bias estimates of demographic parameters when using coalescent-based methods if population structure is not properly taken into account . Furthermore , unlike earlier methods for fitting structured coalescent models to genealogies ( e . g . [7] , [15] ) , our framework can accommodate non-equilibrium and nonlinear population dynamics and allows birth and migration rates to vary over time . We can also include stochasticity in our models when fitting them to data obtained from real populations , which may behave very differently than what would be expected under deterministic models . We can therefore fit the type of mechanistic population dynamic models typically used by epidemiologists and ecologists , which often include population structure , to genealogies . As we have shown , fitting stochastic population dynamic models to genealogies through a structured coalescent model poses some challenges to statistical inference not normally dealt with in the statistical literature on fitting generic state space models to observational data . Under our structured coalescent models , the probability of a genealogy depends conditionally on both the population state variables as well as the states of individual lineages over time . However , going backwards in time , the probability that a lineage is in a certain state can strongly depend on the state that the lineage was sampled in at some future point in time . Particle filtering methods , which are widely used to fit state space models to other sources of data , can perform very poorly under these circumstances because the state of the system , in this case the lineage states , can depend strongly on the future states of the system . One strategy we initially tried was therefore to use a Gibbs sampling approach to iteratively sample from the conditional posterior densities of the population state and lineage state variables in independent steps to avoid the problem of having both forward and backward time dependencies in the model . Unfortunately , we found that such a Gibbs sampling strategy can be very inefficient and suffer from extremely poor MCMC mixing when there are strong correlations among the parameters and the lineage states . For example , in our two-population model , the mixing parameter controls how rapidly lineages move between states and is thus highly correlated with the lineage states . If we update conditional on our current lineage states , the proposed value of will need to be very close to the current value in order for the proposal to have high enough probability to be accepted conditional on the current lineage states . We therefore explore a potentially very large parameter space taking only small steps at a time . Given these issues , we decided to use a modified version of the PMMH algorithm originally proposed by Andrieu et al . [22] . In this approach , we simply propose new parameter values each MCMC iteration and then run the particle filter to numerically integrate over the population state variables . To make the particle filtering algorithm as efficient as possible within each MCMC step , we allow for resampling by first weighting the particles according to the expected lineage state probabilities . Once we have run the particle filter forwards in time , we can then compute the true lineage state probabilities backwards in time and apply an additional round of importance sampling to correct for any bias introduced by using the expected lineage state probabilities . With the true lineage state probabilities of each particle , we can compute the coalescent likelihood of the genealogy while summing over all possible lineage states . We can therefore integrate over both the unobserved population state variables and the lineage state variables when computing the marginal likelihood of the parameter proposal . We thus have an efficient MCMC algorithm for sampling from the posterior density of the parameters without having to design independent proposals for the population states or the lineage states . The PMMH sampler therefore has a major practical advantage over other MCMC approaches that can be easily quantified . For the models considered in this paper , the PMMH algorithm typically converged in less than 100 , 000 iterations whereas for the Gibbs sampler we could run millions of MCMC iterations and still not converge . The efficiency of this approach will hopefully make it possible to also consider phylogenetic uncertainty in the future by sampling genealogies in addition to epidemiological parameters in the MCMC algorithm . Whether or not the type of coalescent models considered here are appropriate for a particular pathogen is another important issue . The coalescent models assume that each infected host corresponds to a single pathogen lineage . If this were indeed always the case then coalescent events in the genealogy would always correspond to transmission events in the population . In reality , coalescent events will not occur instantaneously at transmission events but at some time before the actual transmission event because there will be a waiting time between when a lineage is transmitted and when it coalesces with another sampled lineage in the host . How closely the actual transmission event corresponds in time with the coalescent event will likely depend on the within-host dynamics of the pathogen [42] . For chronic viral infections like HIV where multiple lineages can persist within a given host for months or years , this may result in a large discrepancy in the timing of transmission and coalescent events . Nevertheless , a simulation study using a realistic distribution of within-host coalescent times for HIV found that the difference in timing between coalescent and transmission events was not sufficient to bias estimates of epidemiological parameters [34] . This may be due to the fact that a large fraction of HIV transmissions are due to recently infected individuals , in which case the within-host coalescent event cannot have occurred very long before the actual transmission event . A more principled approach to pursue in the future may be to impute the actual times of transmission conditional on the time of the coalescent events using information about within-host population dynamics . For example , additional information about pathogen population sizes over the course of a typical infection could provide an informative prior on waiting times between transmission events and coalescent events within hosts . Another possible violation of the coalescent model occurs if sampled individuals have descendants that are themselves sampled , which can occur when samples are collected serially over time . The coalescent model implicitly assumes that when a new lineage is sampled , that lineage is sampled from a different host than any other lineage already in the genealogy . However , if a lineage is sampled from a host that has other sampled descendant lineages in the genealogy , then this results in a coalescent event in the tree that does not correspond to a transmission event in the population . A similar problem would arise if we unwittingly sampled more than one lineage from a single infected host . However this is likely to occur only if sampling is dense relative to prevalence over time . For example , if sampling is dense at the beginning and the end of an epidemic , then with a high probability hosts sampled at the beginning of the epidemic will likely have sampled descendants at the end of the epidemic . We acknowledge that the coalescent models used in this paper cannot adequately handle these types of situations , although for the HIV analysis it is unlikely that this is a serious problem seeing as all sequences were sampled in the recent past when prevalence was high . In cases where this is likely to be a serious problem , it may be worth developing metapopulation coalescent models , such as those introduced by Dearlove and Wilson [43] , that allow hosts to be infected by more than a single lineage . As our application to HIV showed , the PMMH algorithm allowed us to infer key epidemiological parameters like stage-specific transmission rates directly from genealogies . However , in the case of HIV , individuals stay in the same stage of infection for long periods of time relative to the timescale of the epidemic . The stage of infection of sampled individuals is therefore highly informative about the state of the lineage going into the past . Our experience with HIV may therefore not be representative of our general ability to infer parameters pertaining to pathogen transmission or movement in structured populations . In fact , our simple two-population SIR model revealed certain conditions under which it may be inherently difficult to estimate parameters relating to population structure . When lineages move between states rapidly due to transmission or migration any particular lineage is likely to have changed states multiple times before a coalescent event is reached , leading to high uncertainty about the state of lineage over the majority of the genealogy . This is somewhat analogous to the problem of site saturation in phylogenetic inference , where multiple transitions at a particular site along branches can render that site phylogenetically uninformative [44] . In the case of rapid transition rates among population states , observing the state of lineages at the time of sampling offers little or no information about the structure of the population because all information about the state of the lineage is quickly lost . Under these circumstances , it will be difficult to precisely estimate migration rates or other parameters relating to population structure from genealogies as we saw from the likelihood profiles of the mixing parameter in the two-population model , although it may be possible with many samples or a large sample fraction . This echoes earlier work on inference with structured coalescent models , where researchers have found it difficult to estimate migration rates from genealogies even without the complication of complex population dynamics [7] , [45] . Although it may not always be possible to precisely estimate parameters relating to population structure from genealogies , we can imagine several cases in which the ability to fit mechanistic epidemiological models to genealogies that include population structure may be extremely useful . For example , our methods could be used to fit spatially structured models to genealogies of samples collected in different locations and could potentially complement recently developed phylogeographic methods that consider spatial structure but do not generally take into account local population dynamics at any particular location [46] , [47] . For instance , incorporating both spatial and temporal dynamics could be important when the structure of a population is not static but changes over time due to changes in migration rates , which themselves may vary due to non-stationary population dynamics across locations . Our approach can also be applied in cases where sampling effort is distributed unevenly among populations so that the assumption of random sampling in unstructured coalescent models has obviously been violated . In this case , structured coalescent models can be used to control for non-random sampling as long as sampling is random within the subpopulations defined in the coalescent model . Finally , our methods can be applied to multi-host or vectored pathogens where lineages can move among different host or vector species . As shown in Rasmussen et al . [48] for the case of dengue , including the dynamics of both the host and vector populations in coalescent models may be necessary in order for population dynamics inferred from genealogies of vector-borne pathogens to be accurate . We end by noting that the methods presented here can be used to fit epidemiological models to genealogies as well as other sources of data simultaneously . For example , we previously showed how unstructured epidemiological models can be fit to a genealogy and a time series of case reports simultaneously and it would be straightforward to extend the methods presented here to include time series or other observational data [21] . This could be especially helpful when certain parameters or aspects of the dynamics are difficult to infer from one data source but for which an alternative data source could be highly informative . For example , case report data may be aggregated over different subpopulations obscuring some of the heterogeneity present in the population but could be revealed by also considering information present in a genealogy . Consolidating data sources in this way will likely play an important role in epidemiological modeling in the future , especially as molecular sequence data become increasingly available and phylodynamic methods become integrated into modern epidemiology .	Mathematical models play an important role in our understanding of what processes drive the complex population dynamics of infectious pathogens . Yet developing statistical methods for fitting models to epidemiological data is difficult . Epidemiological data is often noisy , incomplete , aggregated across different scales and generally provides only a partial picture of the underlying disease dynamics . Using nontraditional sources of data , like molecular sequences of pathogens , can provide additional information about epidemiological dynamics . But current “phylodynamic” inference methods for fitting models to genealogies reconstructed from sequence data have a number of major limitations . We present a statistical framework that builds upon earlier work to address two of these limitations: population structure and stochasticity . By incorporating population structure , our framework can be applied in cases where the host population is divided into different subpopulations , such as by spatial isolation . Our framework also takes into consideration stochastic noise and can therefore capture the inherent variability of epidemiological dynamics . These advances allow for a much wider class of epidemiological models to be fit to genealogies in order to estimate key epidemiological parameters and to reconstruct past disease dynamics .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "and", "health", "sciences", "mathematics", "statistics", "(mathematics)", "epidemiology", "population", "dynamics", "biology", "and", "life", "sciences", "population", "genetics", "population", "biology", "physical", "sciences", "evolutionary", "biology", "st...	2014	Phylodynamic Inference for Structured Epidemiological Models
The vast majority of meiotic recombination events ( crossovers ( COs ) and non-crossovers ( NCOs ) ) cluster in narrow hotspots surrounded by large regions devoid of recombinational activity . Here , using a new molecular approach in plants , called “pollen-typing” , we detected and characterized hundreds of CO and NCO molecules in two different hotspot regions in Arabidopsis thaliana . This analysis revealed that COs are concentrated in regions of a few kilobases where their rates reach up to 50 times the genome average . The hotspots themselves tend to cluster in regions less than 8 kilobases in size with overlapping CO distribution . Non-crossover ( NCO ) events also occurred in the two hotspots but at very different levels ( local CO/NCO ratios of 1/1 and 30/1 ) and their track lengths were quite small ( a few hundred base pairs ) . We also showed that the ZMM protein MSH4 plays a role in CO formation and somewhat unexpectedly we also found that it is involved in the generation of NCOs but with a different level of effect . Finally , factors acting in cis and in trans appear to shape the rate and distribution of COs at meiotic recombination hotspots . Meiosis reduces the level of ploidy by half . To fulfill this goal , homologous chromosomes ( homologs ) are segregated at the first meiotic division . In most eukaryotes , accurate segregation is ensured by the formation of at least one reciprocal recombination event or crossover ( CO ) between the chromatids of homologs [1] . In addition to this crucial mechanical role , COs increase genetic diversity by reshuffling alleles along the genome . In all eukaryotes , CO distribution along chromosomes is not homogeneous . COs tend to be clustered in narrow regions ( two to three kilobases wide ) called hotspots where CO frequencies are greatly enhanced compared to large adjacent regions almost devoid of any recombinational activity [2] . For example , 80% of all recombination occurs in 10 to 20% of the human genome [3] . The molecular organization of hotspots has been deciphered in the two yeasts Saccharomyces cerevisiae and Schizosacchamomyces pombe . As most of the proteins involved in the meiotic recombination process are evolutionary conserved , it is thought that their basic features are similar in all eukaryotes: meiotic recombination is initiated by DNA double-strand breaks ( DSBs ) formed early in meiotic prophase at the leptotene stage by the Spo11 protein [4] , [5] . The initiating DSBs are repaired preferentially by interactions with a non-sister chromatid . After completion of the DSB repair process , both COs and non-reciprocal recombination events , also called non-crossovers ( NCOs ) can be recovered [6] ( Figure 1 ) . COs and NCOs cluster around the DSBs sites in hotspot regions . CO rates peak at the center of the hotspots and then decrease on either side of this region [2] , [7] . Most studies of meiotic recombination hotspots characterized COs whereas because NCOs are difficult to detect these have rarely been analyzed . COs per se are the major determinant of linkage disequilibrium ( non-random association of genetic markers ) breakdown . In addition , the gene conversion tracks contained in both COs and NCOs shape the haplotype landscape . Indeed , CO associated gene conversion events soften the boundaries between haplotype blocks while NCOs create holes within blocks [8] . Thus it is important to appreciate both phenomena as they have implications for genetic association analyses . The ratio of COs to NCOs varies from one hotspot to another , from 14∶0 to 0∶7 in S . cerevisiae with a very low CO to NCO ratio next to telomeres ( excess of NCOs ) and repression of both CO and NCOs close to centromeres [9] . The CO to NCO ratio is also extremely variable in human and mice from more than 12∶1 to 1∶10 [10] . In plants , NCOs have been detected at only a few loci and mainly in maize: at the bronze loci , the CO to NCO ratio varies from 30∶1 to 1∶6 depending on the presence or not of large indels in the region [11] . In Arabidopsis , to date little information is available . Using antibodies directed against the meiotic DSB repair protein DMC1 , DSB sites have been estimated to be between 100 to 200 per meiosis in male meiosis of various accessions [12]–[14] whereas COs vary between 7 to 11 depending on the accessions studied [15] . If these breaks are mainly repaired as NCOs , there should be a large excess ( at least 20 times ) of NCO events compared to COs but the relative ratio of DSBs repaired on homologous versus sister chromosomes is totally unknown in Arabidopsis as in other higher eukaryotes . Several recent studies have tried to tackle the question of the meiotic NCOs rate in A . thaliana . Genome-wide studies using Next Generation Sequencing ( NGS ) gave contradictory results with NCOs found to be either a rare meiotic event [16] or , in striking contrast , several hundred times more frequent than COs [17] . Another study analyzed gene conversion rates at several loci but found only one NCO event at one loci among more than 106 tetrads , the others being associated with COs [18] . Thus , NCO features are poorly understood in plants . Meiotic hotspots are under the control of a series of genes that channel DSB repair toward different pathways . In most organisms , two CO pathways coexist . One is dependent on a group of proteins called ZMM [19] . When one of these proteins is absent , there is a dramatic reduction in COs and the remaining COs do not exhibit interference ( a phenomena described by H . J . Muller in 1916 [20] , where , on the same molecule , multiple COs are more widely spaced than expected if they were placed randomly [21] ) . A second pathway is controlled , at least partially , by the Mus81 complex and COs in this pathway are interference-free . The ratio of interfering to non-interfering COs varies considerably from one species to another . Caenorabditis elegans has only interfering COs while all COs in S . pombe are interference-free . Both pathways appear to contribute equally in S . cerevisiae . However , in most higher eukaryotes , it seems that a vast majority of COs belong to the interfering pathway: 90 to 95% in mice and Humans , 85% in A . thaliana [21] . Up to now , none of the ZMM proteins have been shown to play a role in the formation or processing of NCOs . Our current understanding of the organization of hotspots benefits from analyses in a few species , essentially fungi and mammals . The findings of these studies suggest that there are similarities but also differences in the formation and control of these hotspots . We characterized hotspot regions in a very different model , the plant A . thaliana . We set up a “pollen-typing” molecular approach ( see Results ) , based on the “sperm typing” technique developed previously in mammals [22] , that allowed us to detect and characterize hundreds of CO and NCO molecules at different hotspots . We obtained evidence for the existence of factors acting in cis and in trans that appear to influence CO rate and distribution and our data also suggest a role for the ZMM protein MSH4 in NCO formation . A previous study of CO distribution over the entire A . thaliana chromosome 4 , in large populations of hybrids between “Columbia-0” ( Col ) and “Landsberg erecta-4” ( Ler ) , identified the regions 14a and 130× as good candidates for true CO hotspots [23] . In these two regions of a few kilobases only , CO rates were found to be 20 to 30 times higher than the chromosomal average ( 4 . 8 cM/Mb ) . However , classical genetic techniques could not be used to further investigate these regions , as several tens of thousands of plants would have been needed to obtain enough COs to characterize these regions . Thus we set up a “pollen-typing” technique ( see Materials and Methods; Figure 1; [24] ) , which parallels the “sperm-typing” technique used for hotspot studies in mice and humans [25]–[27] . Briefly ( see Material and Methods for more details ) , genomic DNA ( gDNA ) was extracted from millions of pollen grains collected from a series of F1 ColxLer hybrids and precisely quantified by PCR ( see Material and Methods ) . Taking into account the CO frequency estimated by our genetic map , the gDNA was then diluted to obtain less than one putative recombinant molecule per PCR reaction . Recombinant products were detected by two rounds of allelic PCRs ( Figure 1 ) . When an appropriate dilution was reached , a large series of PCR reactions was performed to detect the presence or absence of at least one template molecule in each reaction . CO rates could then be estimated using the Bayesian inference approach described in the Material and Methods . The meiotic origin of these molecules was assessed with control reactions carried out in parallel with pollen and leaf ( somatic ) DNA . Using the same amounts of gDNA , CO molecules could be detected in pollen DNA , but never in leaf DNA ( Figure 2 ) . Indeed , when the few positive PCR products amplified from a large input of leaf genomes ( at least 15 times more genomes than in the PCR reaction used to detect recombinant molecules on pollen DNA ) were sequenced , they were found to result from non-specific amplification . They did not correspond to a single locus of the Arabidopsis genome but rather to a complex mixture of loci from different chromosomes ( data not shown ) in contrast to products obtained from pollen DNA ( see below ) . We were able to amplify and characterize hundreds of recombinant CO molecules at both regions in gDNA extracted from the pollen of ColxLer hybrid plants ( 167 and 104 COs at 130× and 14a respectively ) . The CO rate in pollen gDNA was 0 . 55% ( Confidence Intervals ( CI ) : 0 . 29–0 . 95 ) at 14a , and 0 . 53% ( CI:0 . 34–0 . 78 ) at 130× . The recombinant molecules were confirmed by sequencing and their exchange point mapped precisely . All CO molecules characterized contained a single transition between the parental haplotypes . Two distinct CO peaks were observed in the 14a region ( subsequently referred to as 14a1 and 14a2 ) , which both fit a Gaussian distribution ( Figure 3A ) . The width of the hotspots within which 95% of COs occurred ( determined by best fit normal distributions , see Material and Methods ) was 1 , 475 bp and 3 , 775 bp for 14a1 and 14a2 , respectively . Their respective medians are 3 , 047 bp apart , with a “valley” in between where just a few COs were detected . The CO frequency was null on either side of this region ( Figure 3A ) . At 14a1 and 14a2 , CO rates peak at 261 and 127 cM/Mb , respectively ( 54 and 26 times the chromosomal average ) . To investigate the relationship between CO frequency at 14a and chromatin we plotted published low nucleosome density ( LND ) data over the same region ( Figure 3C ) , where a high signal represents an absence of nucleosomal DNA [28] . Regions of LND are typically observed at the 5′ of genes coincident with transcriptional start sites ( TSS ) . Consistent with this observation , the LND peaks were located upstream of the two genes within 14a . Strikingly we observed an overlap of the 14a CO frequency peaks and the LND peaks ( Figure 3C ) which suggest that DNA accessibility promotes COs at the 14a hotspot . At 130× , the CO rate was also null on both sides of the region and reached a maximum ( close to the center ) at 167 cM/Mb , which is 35 times the chromosomal average ( Figure 3B ) . The distribution of COs differed from that observed in 14a: it was broad ( more than 7 kb ) and irregular with alternating “peaks” and “valleys” ( Figure 3B ) which does not fit well with a unique Gaussian curve . Interestingly , there was little correlation between CO peaks and LND at 130× ( Figure 3D ) , suggesting that hotspots exist , which have different relationships to nucleosome density . Altogether , both the CO rate and distribution at 14a and 130× clearly indicate the existence of hotspots in A . thaliana . We then looked at the distribution of exchange points in the recombinant molecules in each orientation at both loci ( Figure 4 ) . At the 14a hotspots ( 14a1+14a2 ) , discrepancies between CO distribution in the reciprocal orientations “Col to Ler” and “Ler to Col” ( i . e . ‘CtoL’ and ‘LtoC’ ) were observed: ‘LtoC’ exchanges were shifted to the left of ‘CtoL’ exchanges ( Figure 4A ) . A comparison of cumulative CO distribution patterns showed that this leads to an excess of the Col allele at the center of both hotspots ( Figure 4A ) . At 14a1 , the Col allele was over-transmitted by 68% and this difference was highly significant ( p-value = 0 . 00111 ) , while it was only barely significant ( p-value = 0 . 0734 ) for the 14a2 hotspot , probably due to the lower CO number . These patterns are consistent with the hypothesis that the Ler allele has a stronger initiation activity than the Col allele at these hotspots [29] . The mean position of the two reciprocal distributions ‘CtoL’ and ‘LtoC’ was separated on average by 213 bp and 483 bp for the 14a1 and 14a2 hotspots respectively . In contrast , at the 130× hotspots both alleles appeared equally proficient at initiating recombination ( Figure 4B ) . At meiotic recombination hotspots , DSBs are repaired as either COs or NCOs . In plants , very few meiotic NCOs have been characterized because of the difficulty in detecting molecular events unless they are linked to a phenotypic change . We characterized NCO events at both the 14a1 and 130× hotspots , with different molecular approaches adapted to the polymorphisms available at each hotspot ( Figure 1; see Material and Methods; [30] ) . For 14a1 , the polymorphisms at the center of the hotspot were not suitable for a pollen typing strategy . Thus we used a cloning strategy based on a method described in [31] ( Figure 1; see Material and Methods , [30] ) : after two rounds of allele-specific PCR performed on pollen DNA , the fragment corresponding to the 14a1 hotspot region was cloned . 3000 clones were individually genotyped at three SNPs: two ( #35 and #37 ) located on opposite sides of the center of the hotspot and one ( #33 ) on the left border ( Figure 5 ) . For the control reaction , a similar series of PCRs , cloning and genotyping was performed with DNA extracted from F1 ColxLer leaves . Positives clones obtained with pollen DNA were sequenced ( see Material and Methods; Figure 1 ) . Among 3 , 000 molecules tested , 8 and 7 NCO events were detected at polymorphism #35 and #37 respectively , and none at the most external SNP #33 ( Figure 5 ) . No positive clones were obtained in DNA extracted from leaves ( 0/2850 ) at SNP#35 and #37 demonstrating that NCOs were specific to pollen DNA . The cumulative NCO frequency for both SNPs ( #35 and #37 ) ( 1/203 , 0 . 50% ( CI: 0 . 30–0 . 82 ) ) was similar to the overall CO frequency estimated with the pollen typing approach suggesting that this hotspot is equally prone to produce NCOs and COs ( 0 . 55% ) . NCO events at both sites were all restricted to a single polymorphism , either #35 or #37 , i . e . without co-conversion of left and/or right flanking markers , which are located 111 bp and 482 bp away for #35 and 166 bp and 340 bp for #37 ( Figure 5 ) . Thus , the mean minimal tract was 1 bp if only the polymorphism converted was considered and , the mean maximal tract was 552 bp if the tract was extended to either side just before the next non-converted polymorphism ( 276 bp when the minimal and maximal mean are averaged ) . We recovered unequal numbers of NCOs in both directions: two ‘LtoCtoL’ and six ‘CtoLtoC’ were detected at polymorphism #35 while one ‘LtoCtoL’ and six ‘CtoLtoC’ at polymorphism #37 ( Figure 5 ) . When all NCOs were pooled , the difference between NCO rates in reciprocal orientations ( ‘LtoCtoL’ versus ‘CtoLtoC’ ) was significant ( p-val = 0 . 018 ) . This result strengthens the hypothesis that initiation occurs preferentially on the Ler allele at the 14a1 hotspot ( see above ) . At 130× , NCO molecules were characterized using a PCR-based “pollen-typing” strategy ( see Materials and Methods; Figure 1; [30] ) . Allele-specific PCR was performed using either Col specific or Ler specific primers on 96 samples , each containing 4 , 145 F1 pollen genomes or 4 , 800 F1 leaf genomes . Then , to specifically detect NCO molecules , allele-specific PCR was carried out at three different SNPs ( Figure 1; see Material and Methods; [30] ) : the three polymorphic sites #21 , #44 and #52 were 2339 and 2052 bp away , respectively ( see green dots in Figure 3B ) . SNP#44 is next to the center of the hotspot where the CO frequency is maximal , #21 is located in the left section of 130× where the CO rate is low and #52 is to the right where CO rates were average . For the control , DNA extracted from leaves ( almost 468 , 000 genomes ) , no PCR product was amplified at SNP #44 . In DNA extracted from pollen ( almost 398 , 000 genomes ) , 29 NCO events were found at SNP#44 ( Figure 6A ) demonstrating that NCOs were specific to pollen DNA . Thirty and four NCO events were found at SNP#21 and #52 , respectively ( Figure 6B , 6C ) . The observed NCO frequency was approximately 0 . 007% ( CI: 0 . 005–0 . 010 ) , 0 . 008% ( CI: 0 . 005–0 . 011 ) and 0 . 001% ( CI: 0 . 0004–0 . 0033 ) at #44 , #21 and #52 , respectively . When the results obtained at the three SNPs were pooled , the NCO rate is 0 . 017% , which is roughly thirty times less than the overall CO frequency ( 0 . 53% ) . At SNP #44 , 23/29 NCO tracts extended to the right toward the neighboring polymorphism ( 89 bp away ) , while polymorphisms to the left were co-converted in only five tracts , but over a greater distance ( up to 791 bp ) ( Figure 6B ) ; similarly , at #21 , 20/30 NCOs included the first two polymorphisms on the left ( 275 bp ) whereas only five extended to the right but again over a greater distance ( up to 1028 bp ) ( Figure 6A ) . This apparent non-symmetrical distribution of the NCO tracks reflects the asymmetrical scattering of the SNPs on either side of #44 and #21 . At both sites , numerous SNPs are present only on one side , leading to an accurate analysis of the breakpoints whereas on the other side only distant SNPs were available . The minimal track length means were comparable for both SNPs ( 160 bp at #21 and 278 bp at #44 ) whereas the maximal mean track length was more than three times longer at #21 ( 1798 bp ) when compared to #44 ( 492 bp ) . The longest NCO track was found at #52: six SNPs were co-converted along a tract of 1882 bp that could extend up to 3045 bp . Interestingly , none of the NCO tracks covered either #21 and #44 or #44 and #52 . At #21 three NCO tracts were chimeric ( with more than two exchange points ) : two ‘CtoLtoCtoLtoC’ and one ‘LtoCtoLtoCtoL’ ( Figure 6A ) . We also noticed that as observed for 14a1 , an excess of ‘LtoCtoL’ ( 36 ) NCOs compared to ‘CtoLtoC’ ( 23 ) NCOs were detected at both #21 and #44 . At #52 , only ‘LtoCtoL’ NCOs were obtained . In the latter case , it was not possible to determine whether there was an absence of ‘CtoLtoC’ NCOs in our starting pool of almost 398000 genomes or if these were missed by pollen-typing . However , when the results were pooled for #21 and #44 , the difference was not significant ( ‘LtoCtoL’: 0 . 009% ( CI: 0 . 007–0 . 013 ) ; ‘CtoLtoC’: 0 . 006% ( CI: 0 . 004–0 . 009 ) ) . In A . thaliana , on the basis of chiasma counts in mutant backgrounds , it is assumed that 85% of COs belong to the interference dependent pathway ( class I ) , while the remaining 15% are interference-free ( class II ) [32] . To test the contribution of both CO pathways at the 130× and 14a hotspots , we analyzed CO rate and distribution in an atmsh4 mutant background in which interfering COs are absent . Crosses were made between hemizygous Col and Ler lines containing a T-DNA insertion in the AtMSH4 gene ( see Material and Methods ) . Meiosis appeared regular in both AtMSH4+/− Col and Ler parents and the F1s AtMSH4+/+ or AtMSH4+/− . Meiosis , however , was disturbed in the F1 Atmsh4−/− with a dramatic reduction in chiasma number as described in ( Higgins et al . 2004; data not shown ) . We set up pools of Atmsh4−/− or AtMSH4+/− or AtMSH4+/+ F1 plants , extracted gDNA from their pollen and performed pollen-typing PCR to detect CO molecules . CO rates were not statistically different in AtMSH4+/+ and AtMSH4+/− at either 14a or 130× ( data not shown ) . Thus pollen gDNA from AtMSH4+/+ and AtMSH4+/− was pooled and is referred to as “MSH4” in the following experiments . As expected , in the Atmsh4−/− pollen , when we conducted the experiment at the 14a locus , we detected a dramatic decrease ( 12 fold ) in CO frequency compared to the “MSH4” CO rate ( Table 1 ) . However , this frequency is likely to be slightly over-represented because the proportion of viable pollen grains depends on the number of bivalents ( i . e . pairs of homologous chromosomes containing a CO ) . We then analyzed CO distribution ( Figure 7A ) . Surprisingly , the two hotspots , 14a1 and 14a2 , were affected differently by the mutation . In “MSH4” , the majority of COs ( 61% ) occurred in 14a1 ( ratio 14a1/14a2: 1 . 6 ) . At contrario in Atmsh4−/− , the proportion of COs between 14a1 and 14a2 was inversed ( ratio 14a1/14a2: 0 . 5; chi2 p-value = 8 . 7 10−5 ) ( Figure 7A; Table 2 ) . At 130× , we performed two series of overlapping PCRs to cover the whole area ( see Material and Methods; Figure 7B ) . We also obtained a lower rate of CO frequency in Atmsh4−/− , but at a different level in the left ( 13 times lower ) , and right ( 78 times lower ) sections of the loci ( Figure 7B; Table 1 ) . Next , we tested the NCO frequency in the Atmsh4 mutant background and “MSH4” at both loci . At the SNP#35 in the center of 14a1 , 41 NCOs were detected and confirmed among 11 , 896 colonies in pollen DNA extracted from Atmsh4−/− and 20 among 3 , 600 colonies for “MSH4” pollen DNA . Thus , at this marker the NCO rate was comparable in both genetic backgrounds ( Table 3 ) . At 130× , in the Atmsh4−/− pollen DNA , among 83 , 656 molecules , two NCOs were detected at #44 , which gave a NCO frequency of 0 . 0024% ( CI: 0 . 00074–0 . 0086 ) . In “MSH4” , the NCO frequency ( 84 events/545844 genomes ) was significantly higher at the same SNP 0 . 015% ( 0 . 012–0 . 019 ) ( p-value = 0 . 0035 ) ( Table 3; Figure S3 ) . Therefore , the NCO frequency at #44 decreased considerably ( six fold ) in the absence of MSH4 . We selected two other Arabidopsis accessions for which we could use the same allele-specific primers to perform pollen typing but which have different levels of polymorphisms within the DNA sequence at the 14a hotspot: Pyl-1 ( 8AV ) and Ws-4 ( 530AV ) , ( Material and Methods ) . Between Col and the three accessions Ler , Pyl1 and Ws-4 , there are 0 . 43% , 0 . 53% and 0 . 63% of polymorphisms distributed along the 5 kb of the 14a hotspot ( Figure S1 ) . We also included the “MSH4” data in this study because it is another ColxLer F1with exactly the same sequence at both the 14a and 130× loci . We observed considerable variation in CO rates at the 14a loci . Strikingly , the 14a hotspots almost disappeared in ColxPyl-1 . There were 100 times less COs than in “MSH4” , 60 times less than in ColxLer and 23 times less than in ColxWs and even 12 times less than in the mutant Atmsh4−/− background ( Table 4; Figure 8A ) . In ColxWs , the CO rate ( 0 . 21% ) was in the same range as in ColxLer ( 0 . 55% ) but significantly less than in “MSH4” ( 1 . 27%; Table 4 ) . Surprisingly , we observed that the CO and NCO rates and CO distribution at 130× differed significantly between the two ColxLer F1s used in this study ( ColxLer and “MSH4” ) , whereas no significant variation was obtained at 14a ( Table 4; Figure 8 ) . We also observed another difference , with ‘LtoC’ and ‘CtoL’ exchanges peaking in the same interval in ColxWs and in “MSH4” but distant in ColxLer ( see above ) ( Figure S2 ) . Moreover in “MSH4” we recovered a comparable number of ‘LtoCtoC’ and ‘CtoLtoC’ NCOs ( Figure S3 ) . Thus the bias in recombination appears to only exist in ColxLer . We believe that the differences in CO rates and distribution between these two F1s are robust , as they were observed in several different experiments ( data not shown ) but as mentioned above the hotspot sequences are identical in these two lines . Here , we characterized several meiotic recombination hotspots on A . thaliana chromosome 4 . Both the rate and distribution of COs and the occurrence of NCOs across these regions confirm that they are indeed true meiotic recombination hotspots . At 14a , CO distribution patterns fit with the existence of two independent hotspots located very close to each other . 14a1 has a very high peak rate of COs and is narrow ( 1 , 475 bp ) while 14a2 is broader ( 3 , 775 bp ) and peaks at less than half the rate of 14a1 . At 130× , the CO landscape is more complex: the CO distribution is broad and does not conform to a single Gaussian curve . Instead , it is irregular with alternating peaks and valleys . In this region , COs may originate from a single initiation zone: the irregularities observed in the distribution pattern could be explained by the presence of several insertions/deletions - which are 20 , 13 , 12 , 70 , 10 , 7 , 11 and , 3 bp wide respectively - along the region ( Figure 3B; Figure 6 ) . Such heterologies could either block branch migration of double Holliday junctions or channel recombination intermediates towards NCOs or exchanges between sister chromatids , as suggested previously for the mouse HS22 hotspot [33] . However , even if at 130× , heterologies could result in a slight drop in CO rate , there is not a dearth of COs as described in several mammalian hotspots [6] , [10] , [34] , [35] . In fact , the RuvAB branch migration helicase has been shown to bypass 1000 bp heterologies in vitro [36] . Alternatively , 130× could be , like 14a , a cluster of several close hotspots ( three or more ) , each which derived from a discrete initiation zone and the resulting CO distributions could overlap extensively . This last hypothesis is strengthened by the fact that NCOs are initiated independently in at least three regions around SNPs #21 , #44 and #52 and the conversion tracks do not overlap between these three regions ( Figure 6 ) . Interestingly , in maize , there are two regions where the fine-scale distribution of COs has been characterized ( the a1 [37] , [38] and bronze ( bz ) regions [39] , [40] ) . At a1 the pattern is similar to that of 130× with COs distributed throughout a wide region ( 10 kb ) with peaks and valleys but with no region devoid of COs between two peaks . At bz , there are at least three sections where hotspots have been detected in a 99 kb region but based on the data provided whether each section is itself a cluster of hotspots or a unique hotspot could not be determined [39] , [40] . The occurrence of hotspot clusters ( two or more ) within less than 12 kb was also described in at least four regions in human: DNA1-DNA2-DNA3 , DMB1-DMB2 [41] , NID2a-NID2b , MSTM1a-MSTM1b [34] . We can therefore hypothesize that in Arabidopsis there are regions of tens of kb that are permissive for recombination and that within these regions , recombination hotspots arise at sites where particular sequence motifs and/or chromatin modifications target the activity of Spo11 [42]–[46] . The center of all three of the hotspots described here lies close to gene promoters , which are active in A . thaliana meiocytes in both Col and Ler ( three to four times higher than the average transcription level , [47] ) . In the other Arabidopsis hotspot described recently , 3a , the apparent CO peak lies in a short intergenic region where transcription terminates for both genes [48] . In maize , the majority of characterized hotspots are localized in genes [40] , [49] but there was not sufficient resolution to determine if the CO peak lies within the promoter region . Most DSB hotspots in S . cerevisiae [50]–[52] coincide with promoter regions while in S . pombe , hotspots lie preferentially in large intergenic regions [53] . In human and mice recombination activity occurs near genes but away from transcription start sites ( TSS ) [54] . However , even if the localization of meiotic recombination hotspots seems drastically different in S . cerevisiae and in mice , the underlying mechanisms may not be so divergent . In S . cerevisiae , meiotic DNA DSBs are formed in nucleosome depleted regions enriched in histone H3 trimethylated on lysine 4 ( H3K4me3 ) [55] . In mice , an enrichment of H3K4me3 is also detected at DSBs sites [56] . Recently , a genome-wide correlation between recombination sites , the insertion sites of the transposable element Mu and the chromatin modification H3K4me3 was also reported in maize . As Mu insertion sites also correlate strongly with recombination sites , it was suggested that the local chromatin structure could play a key role in both mechanisms [57] . Others chromatin modifications have been shown to be associated with meiotic recombination hotspots in others species . Acetylation of lysine 9 of histone H3 in S . pombe [46] and levels of H3K9m3 and H2AK5ac in C . elegans modulate meiotic DSB formation [58] , [59] . In A . thaliana , the 3a and 14a hotspots but not 130× lie in a low nucleosome density region [48] ( this study ) . It is important to note that the nucleosome data we are comparing to was generated from somatic ( seedling ) tissues and we therefore cannot rule out that nucleosome occupancy may differ during meiosis . However , nucleosome occupancy in yeast and mammals is similar between meiotic and mitotic cells [60]–[62] . Thus it could be that several different chromatin states act on the localization of meiotic DSBs . The potential similarity between the location of S . cerevisiae and Arabidopsis hotspots in promoter regions could be due to the resemblance of their genomic structures . Both are compact , have a high density of genes along chromosomes and small intergenic regions [63] , [64] . We measured the CO rate and distribution at 14a during meiosis of four different F1s . In ColxWs , the CO rate was significantly different to “MSH4” but even more strikingly , the hotspot almost disappeared in ColxPyl1 . There were even less COs than in the Atmsh4−/− background . Both Ws and Pyl1 exhibit differences at the DNA sequence level compared to Col ( 0 . 53% and 0 . 63% respectively; Figure S1 ) that could explain this variation . A correlation between a decrease in the rate of meiotic recombination and polymorphism level has been reported in various species including plants and it has been shown that the mismatch repair machinery is involved in this drop in meiotic recombination rate [65] . Transposon insertion has also been shown to modify meiotic recombination at a1 [37] and bz [40] in maize . Alternatively , the disappearance of the hotspot in the ColxPyl1 cross could be due to a modification of a sequence inside 14a , crucial for the initiation of recombination . Similar results have been reported in mice and human where a mutation in the recognition site of PRDM9 can dramatically influence the hotness of a hotspot [66] , [67] . Thus a combination of factors acting in “cis” or “trans” could influence the behavior of the 14a loci during meiosis . Although the 8 . 8 kb sequence of the 14a locus in the two other F1s , ColxLer and “MSH4” , is identical , significant differences were observed at their hotspots . In ColxLer a bias in CO distribution and directionality of NCOs suggests that preferential initiation on the Ler chromosome occurred but this was not observed in “MSH4” . At 130× , which is also identical in both F1s , we also observed significant differences in CO rate and distribution between ColxLer and “MSH4” but the NCOs were the same . The accessions used to obtained these two F1s are clearly related to each other ( http://arabidopsis . info/protocols/ler . html ) but nevertheless they have evolved since they were isolated and it could be that some key mutations were selected that changed in “trans the local behavior of the region . Alternatively , the epigenetic status of the two F1s may have changed leading to differences in the activity of hotspots . We isolated and characterize dozens of meiotic NCOs at two loci for the first time in Arabidopsis thaliana . Three recent studies addressed the detection and rate of meiotic NCOs in A . thaliana . One recorded gene conversion events associated or not with COs at seven loci but only one NCO was detected at one locus [18] . The two other studies used NGS for genome-wide detection of COs and NCOs . These two studies gave contradictory results on the rate of NCOs . One estimated NCOs to be rare meiotic events ( on average two per meiosis ) [16] whereas the other one predicted up to 3000 NCOs per meiosis [17] . However , in this latter study , up to 30 to 40 COs per meiosis were also predicted which is three to four times higher and not consistent with numerous genetics or cytological studies performed with wild type crosses , suggesting a large over estimation of NCO rates in this study . Our data clearly show that NCO rates are highly unhomogeneous between hotspots . At 130× , NCOs were detected at three polymorphic sites distributed along the hotspot with an overall rate of 0 . 016% , thus 30 times less than that of COs . At 14a , the observed NCO rate was similar to that of COs ( 0 . 5% ) . In A . thaliana , meiotic DSBs sites have been estimated at between 100 and 200 per meiotic cell based on the number of RAD51 or DMC1 foci ( two DSB repair proteins ) at mid-prophase [12]–[14] . The proportion of these DSBs repaired on the sister or homologous chromatids is unknown , but if only half of them are repaired as NCOs , there should be five to ten times more NCOs than COs . However , neither the NGS genome-wide data [16] or our data at two hotspots support this . We propose that NCOs are very small and in most cases are not detectable because they do not convert a SNP . Indeed , most of the NCO tracks detected in this study were single SNPs . We also detected three NCO events with a discontinuous conversion pattern: two ‘CtoLtoCtoLtoC’ and one ‘LtoCtoLtoCtoL’ ( Figure 6A ) . In all cases , the discontinuity was related to a single SNP ( A/T or C/A ) . Similar complex conversion events have been detected in other species [68]–[70] . It was suggested that chimeras could result from template switches between non-sister and sister-chromatids during DSB repair [68] , [70] . Alternatively , at this particular locus , the mismatch repair machinery may fail to convert all mismatches contained in the heteroduplex generated by the homologous recombination machinery . According to the current view of meiotic DSB repair ( reviewed in [71] ) , in most species ( including S . cerevisiae , mammals and plants ) most if not all NCO events arise through a “Synthesis Dependant Strand Annealing” ( SDSA ) mechanism whereas COs are formed by two distinct pathways , which generate either interfering COs ( class I ) or non-interfering COs ( class II ) . MSH4 belongs to a group of highly conserved proteins , called ZMM , that are essential for the “class I” CO pathway [19] . In S . cerevisiae ( reviewed in [19] ) , and Sordaria macrospora [72] , the mutation of MSH4 leads to a pronounced decrease in CO number . Therefore , in Arabidopis Atmsh4 , the decrease in COs that we observed at 130× and 14a was expected but the change in CO distribution was more surprising . Thus , 130× and 14a are very likely to be clusters of hotspots and the proportion of MSH4 dependent COs appears to vary markedly between hotspots within the cluster and between hotspots . Surprisingly , we also found that in the Atmsh4 mutant at 130× there is a six fold decrease in the frequency of NCO events . In A . thaliana , mice and S . macrospora , the MSH4 protein is localized in numerous foci along chromosome axes as early as mid-late leptotene and then the number of foci decreases to zero at the end of pachytene [32] , [72] , [73] . In these three species , however , the maximum number of MSH4 foci far exceeds the number of COs . Furthermore , in S . macrospora MSH4 is present at virtually all sites of interaction between homologs ( COs and NCOs ) at the onset of zygotene , where it appears to play a role in the orderly progression of the pairing and synapsis processes [72] . We now propose that in all eukaryotes , beyond its role in class I CO formation , MSH4 is involved in the formation or stabilization of at least some of the recombination intermediates leading to NCOs . In conclusion , we have formally demonstrated that true meiotic recombination hotspots exist in the plant Arabidopsis thaliana . We have also established that COs and NCOs occur at very high rates at these hotspots , as observed in yeast and mammals . However , we have shown that the pattern of COs and NCOs differs from that described in other species . There is therefore a need for the analysis of more hotspots in a diverse range of species in order to understand the underlying mechanisms that control them . The Arabidopsis thaliana accessions “Columbia-0” ( Col , 186AV ) , “Landsberg erecta” ( Ler , 213AV ) , Pyla-1 ( Pyl-1; 8AV ) and Wassilewskija-4 ( Ws-4; 530AV ) were obtained from the “Centre de Ressources Biologiques” at the “Institut Jean Pierre Bourgin” , Versailles , France . The MSH4 mutant lines SALK_136296 ( Col-0 ) [74] and CSHL_GT14269 ( Ler ) [75] were provided by NASC ( http://nasc . nott . ac . uk/ ) . All plants were grown in the greenhouse under standard conditions . After crossing , we used PCR screening to select homozygous AtMSH4+/+ , +/− and −/− plants among the F1 . The oligonucleotides used were: ( i ) N636296U CTTCTTGCAGGTTfGTGTTTG - N636296L GCCAGCTGTTTTTGTTGTC and N636296L - LbSalk2 TCCCGCTCAGAAGAACTC to genotype the wild type and the mutant allele respectively in Col ( ii ) GT14269U CCGTTCAAATGTTTGCCATAC - GT14269L TTTCACCTTCCTAACGGTGC and CSHLds5-4 TACGATAACGGTCGGTACGG - GT14269L to genotype the wild type and the mutant allele respectively in Ler . Characterization of male meiosis by cytology was carried out as described in [76] . Chiasma counts , performed on ColxLer and “MSH4” as described in [15] , showed similar numbers of chiasmata in the two F1s ( Figure S4 ) . Genomic DNA from pollen was extracted as described in [24] . Briefly , whole inflorescences from hybrid plants were harvested in 10% saccharose , and crushed in a “Waring Blender” ( two 4 sec pulses at full speed ) . The homogenate , containing intact microspores and pollen grains , was then filtered and stored at −20°C until DNA extraction . Pollen grains and microspores were resuspended and incubated with proteinase K at 65°C for three hours with gentle shaking . Then , pollen grains were disrupted by mixing with glass beads with a vortex at full speed for 1 to 3 minutes . One volume liquid phenol was added and tubes were rocked for 30 min at 4°C . After centrifugation , the supernatant was recovered and nucleic acids were precipitated with sodium acetate and ethanol . Genomic DNA was dissolved in ( 10 mM Tris-Cl pH8 , 1 mM EDTA , 100 µg/ml RNAseA ) and incubated at room temperature for 15 min . Four volumes of freshly made ( 5 M guanidine isothiocyanate , 50 mM Tris-Cl pH 8 ) were then added , and DNA was purified with DNeasy minicolumns ( Qiagen ref . 69106 ) . Genomic DNA was extracted from young leaves as described in [77] . Four volumes of freshly made ( 5 M guanidine isothiocyanate , 50 mM Tris-Cl pH 8 ) were added to the extract , and DNA was purified with DNeasy minicolumns ( Qiagen ref . 69106 ) . Quantification of gDNAs was performed as described in [24] . Briefly , PCR reactions were performed in 20 µl of buffer [78] with 1 U Taq DNA polymerase and 0 . 1 U Pfu DNA polymerase . Whenever less than 100 pg/µl of genomic DNA was used , herring sperm DNA ( Clontech ) was added into the reactions ( 1 ng/µl ) . Primers ( sequence and genomic coordinates ) are listed in Table S1 . Pairs of oligonucleotides used are listed in Table S2 . Products amplified from the gDNA extracts were quantified in a series of dilutions through two rounds of PCR , using the nested allele-specific oligonucleotides ( ASOs ) listed in Table S1 and Table S2 . The product of the first PCR was diluted 1/1000 in the second reaction . The thermal cycling profile of the reactions was: ( ( ( 92°C;2 min ) ( ( 92°C;20 sec ) ( Tm;30 sec ) ( 68°C;30 sec + 45 sec/kb ) ) ) ×30 ( 68°C;90 sec/kb ) ( 4°C;∞ ) ) . After the second PCR , the proportion of negative wells among a set of aliquot reactions was approximated by e−m , where ‘m’ is the mean number of DNA molecules per well in the first reaction . Parental molecules were thus quantified using ASOs all specific to either Col or Ler DNA . CO molecules were amplified with primers specific to either Col or Ler allele on one side , and Ler or Col respectively on the other side ( Figure 1 ) . Primers ( sequence and genomic coordinates ) are listed in Table S1 . Pairs of oligonucleotides are listed in Table S2 . For mapping CO exchange points , a series of aliquot reactions was carried out , which was predicted to contain an average of less than 0 . 2 CO molecules , so that more than 90% of positive reactions issued from a single CO molecule . PCR products were then sequenced in order to locate exchange points from single CO molecules . To detect COs in “MSH4” and “Atmsh4” , two series of overlapping PCRs were performed . The two PCRs overlapped on each side of the SNP#44 . The first PCR used the primers 130×0LeL1 and 130×52CoR1 or 130×0CoL1 and 130×52LeR2 . Then primers 130×7LeL5 and 130×47CoR2 or 130×7CoL4-130×47LeR4 were used for the left distribution whereas for the right distribution we used 130×44LeL4 and 130×72CoR2 or 130×44CoL4 and 130×72LeR4 ( Table S1 ) . At 14a , the Ler primers were used to detect COs in Ws and Pyl1 . NCO molecules at polymorphisms #44 , #21 and #52 in the 130× hotspot were detected using a PCR-based strategy adapted from [25] . The outline of this approach is described in Figure 1 [30] . An allele-specific PCR was performed using either Col specific primers or Ler specific primers on 96 samples each containing 4 , 145 F1 pollen genomes for ColxLer , 5 , 685 for “MSH4” or 4 , 800 F1 ColxLer leaf genomes . Then , two sets of allele-specific PCRs were carried out in parallel at one SNP to specifically detect NCO molecules . When both left and right PCR reactions were positive at one SNP , recombinant PCRs molecules were fully sequenced to ( i ) confirm the NCO event and ( ii ) map the recombinant point ( Figure 1 ) . Pairs of primers for these experiments are listed in Table S3 . For 14a1 , we followed the procedure described ( [30]; see Figure 1 ) . Twenty four PCRs were carried out on DNA extracted from F1 ColxLer pollen corresponding to 48 , 000 F1 genomes , half with the primers specific from the Col parental allele and located outside the hotspot region and the other half with the Ler primers ( Table S1 and Table S2 ) . Then after a second role of allele-specific PCR on each pool of DNA , the PCR products were pooled . The PCR products were then digested with BglII and XbaI and ligated into pCRIITOPOblunt ( Invitrogen ) between the BamHI and XbaI unique sites , using standard procedures . The ligation products were then used to transform DH10B E . coli strain by electroporation . Transformed cells were spread onto LB agar plates containing 100 µg/ml carbenicillin , 0 . 2 mM IPTG and 40 µg/m X-gal . Following blue/white screening , individual colonies were transferred to 200 µl of LB medium containing 100 µg/ml carbenicillin in 1 ml MASTERBLOCK microplates ( Greiner Bio-One ref . 780215 ) and grown with gentle shaking at 37°C for 16 hours . Then , 100 µl of each cell culture was transferred to 96 well V-bottom microplates and spun down at 3200×g for 10 min . Cell pellets were resuspended in 100 µl of sterile water . Bacterial clones were then genotyped using the “Chemicon Amplifluor SNPs Genotyping System” . For this purpose , oligonucleotides either specific to each parent at polymorphisms #33 , #35 and #37 , or non-specific primers , were designed using the Amplifluor AssayArchitect software ( Table S4 ) . Genotyping was then performed as described in [31] . Plasmids which appeared to contain a NCO event by genotyping were fully sequenced to precisely map the gene conversion event . The differences between CO distribution in reciprocal orientations at a given hotspot were tested as follows: ( i ) CO breakpoints located on each side of the median position were grouped separately for the two reciprocal orientations ( ‘CtoL’ CL and ‘LtoC’ LC ) , thus providing four numbers COCLleft , COCLright , COLCleft , COLCright; ( ii ) these numbers were grouped in a contingency table for testing the association between left/right and CL/LC classification using the two-tailed Fisher's exact test . The difference between ‘CtoLtoC’ and ‘LtoCtoL’ NCO rates at polymorphisms #35 and #37 in the 14a1 hotspot was tested using the one tailed Fisher's exact test . The parameters of the best fitting Gaussian distributions were calculated using an R script ( Supplemental File S1 ) , which computes the least sum of squared differences between observed and theoretical ( Gaussian ) integrated distributions over every interval between successive SNPs . Estimated CO frequencies and associated confidence intervals were analyzed as follows: Repeated PCR experiments were performed on highly diluted pollen DNA samples collected on a F1 plant obtained from a cross between two homozygous parents carrying different alleles at two marker loci . Two primer pairs were used for PCR amplification . The first pair was specific for molecules carrying the alleles of the first parent at both loci , and could thus amplify half of the non-recombinant molecules . The second primer pair was specific for molecules carrying the first parental allele at the first locus and the second parental allele at the second locus , and could then amplify half of the recombinant molecules . A PCR reaction was considered positive if the template contained at least one molecule corresponding to the primer pair used . The same initial pollen DNA sample ( unknown concentration C ) was used as the template for all experiments , but at different dilutions . An Si series of experiments ( indexed with i ) was performed with the first primer pair , and an Sj series ( indexed with j ) with the second primer pair . For the kth series , a total of Nk PCR reactions was carried out using as template the initial DNA sample diluted at the rate Dk . Let us note yk as the number of reactions that did not produce a product . A Bayesian inference approach was used to infer the recombination rate between the two marker loci , as well as its 95% confidence intervals . For the first primer pair , amplifying non-recombinant molecules , the probability of no amplification in a given well follows a Poisson law and is: . The likelihood of the observed PCR results obtained with this primer pair follows a binomial law: With the second primer pair , amplifying recombinant molecules , the probability of no amplification is where r is the recombination rate between the two marker loci . The likelihood of the observed results is then: The joint likelihood of all observations ( both primer pairs ) is: Because C and r are independent , the a posteriori probability of parameters C and r knowing the observations is: where fC and fr are the prior density functions of C and r respectively . The prior distributions for r and C were taken uniformly distributed in [0;0 . 1] and respectively , which supposes that r<10% , and that at least 10% positive reactions are expected with the least concentrated sample . The a posteriori distributions of r and C were numerically computed by a two-dimensional scan of the parameter space ( R script available upon request , and the 95% confidence intervals on r and C were determined from these distributions . 95% confidence intervals on CO and NCO frequencies were computed based on the binomial law , by numerically adjusting the frequencies corresponding to distribution function values equal to 0 . 025 and 0 . 0975 . When the value of a parameter was estimated from the data , the associated 95% confidence interval is defined so there was less than 5% chance that the true value of the parameter lies outside this interval .	During meiosis , genomes are reshuffled by recombination between homologous chromosomes . Reciprocal recombination events called crossovers are clustered in several kilobase-wide regions called hotspots , where their frequency is greatly enhanced compared to adjacent regions . Our understanding of hotspot organization is based on analyses performed in only a few species and rules differ between species . For the first time , hundreds of recombination events were analyzed in Arabidopsis thaliana revealing several new features: ( i ) crossovers are concentrated in hotspots where their rate reaches up to 50 times the genome average; ( ii ) non-crossovers events , ( also called gene conversions not associated with crossovers ) also occur in hotspots but at very different levels; and ( iii ) in the absence of the recombination protein MSH4 , the crossover rate is dramatically reduced ( 70 times less than the wild-type level ) and the crossover distribution within a hotspot is also largely modified; unexpectedly , the non-crossover rate was also altered ( 15% of the wild-type level at a hotspot ) . Finally we showed that factors acting in cis and in trans may influence the level and distribution of crossovers at and between hotspots .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2013	Contrasted Patterns of Crossover and Non-crossover at Arabidopsis thaliana Meiotic Recombination Hotspots
Seasonal influenza virus infections can cause significant morbidity and mortality , but the threat from the emergence of a new pandemic influenza strain might have potentially even more devastating consequences . As such , there is intense interest in isolating and characterizing potent neutralizing antibodies that target the hemagglutinin ( HA ) viral surface glycoprotein . Here , we use cryo-electron microscopy ( cryoEM ) to decipher the mechanism of action of a potent HA head-directed monoclonal antibody ( mAb ) bound to an influenza H7 HA . The epitope of the antibody is not solvent accessible in the compact , prefusion conformation that typifies all HA structures to date . Instead , the antibody binds between HA head protomers to an epitope that must be partly or transiently exposed in the prefusion conformation . The “breathing” of the HA protomers is implied by the exposure of this epitope , which is consistent with metastability of class I fusion proteins . This structure likely therefore represents an early structural intermediate in the viral fusion process . Understanding the extent of transient exposure of conserved neutralizing epitopes also may lead to new opportunities to combat influenza that have not been appreciated previously . Influenza displays three glycoproteins that embroider the viral surface: hemagglutinin ( HA ) , neuraminidase ( NA ) , and Matrix-2 ion channel . All of these proteins are necessary for the viral replication cycle . Among the surface glycoproteins , HA is the principal target for neutralizing antibodies . HA is a class I viral fusion protein that facilitates viral entry by interacting with sialic acid receptors on the host cell and then fusing the viral and cell membranes in acidic endosomal compartments . HA is expressed as a precursor form termed HA0 , which is then cleaved by host-cell proteases into HA1 and HA2 domains , resulting in a trimer of heterodimers . HA1 contains the membrane-distal sialic acid receptor-binding site ( RBS ) , while HA2 includes the fusion machinery , proximal to the membrane . Proteolytic cleavage at the HA1–HA2 junction liberates the hydrophobic fusion peptide , which then becomes buried in the center of the trimer . The HA1 and HA2 domains remain covalently linked by a disulfide bond after cleavage . This cleaved , prefusion conformation is metastable and poised to undergo pH-induced conformational changes but must not do so prematurely . After influenza virus binds to the host cell , it is endocytosed and trafficked into endosomal compartments , in which the lumen is acidified . Near pH 5 . 5 , HA undergoes large conformational rearrangements , which lead to insertion of its fusion peptide into the host membrane . This process drives fusion of the host and viral membranes and release of the viral RNA genome into the cytoplasm [1] . Influenza evolution , particularly in HA , occurs rapidly , with antigenic drift sometimes conferred by single amino acid changes near the RBS [2] . The RBS , which is the most conserved region of the HA head , forms a shallow pocket to which a series of antibodies has been shown to bind and neutralize [3] . Although amino acid mutations in HA are used to evade the host immune system , there are conserved areas of HA that are vital to viral fitness . Head-binding antibodies have been shown to be very effective in neutralizing the virus but usually only in a strain-specific manner , although some broadly neutralizing antibodies are known to target the RBS [3 , 4] . Monoclonal antibody ( mAb ) 5J8 , a prototypic head antibody with breadth , targets the RBS via a long complementary determining region heavy chain 3 ( CDRH3 ) that mimics the sialoglycan receptor and neutralizes H1 strains from 2009 to the pandemic of 1918 [5] . RBS antibodies typically have a broader neutralization profile due to sequence conservation of this site , while other regions on the HA1 head are hot spots for strain-specific antibodies [6] . Originating as avian influenza , H7 strains have crossed over intermittently to infect humans [7] . Laboratory studies have shown only three amino acids are required to completely change receptor specificity from avian to human , allowing human-to-human transmission and increasing the chance of a new pandemic [8] . In fact , an outbreak of H7N9 virus in China in 2013 has been linked to close contact of humans with chickens and ducks , which were identified as a reservoir for the virus [7 , 9] . Since then , H7N9 has continued to circulate in poultry reservoirs , causing a spike in human infections in recent years [7] . Several H7-specific mAbs have been described recently , including H7 . 137 , H7 . 167 , and H7 . 169 , which target highly conserved regions of the HA head adjacent to the RBS [10] . Another antibody isolated in the same study , mAb H7 . 5 , was shown to recognize an epitope that overlaps a portion of the H7 . 167 epitope and neutralize similar strains of human H7 virus , including those from mallards and chickens , but which binds at a different angle [10] . The H7 . 5 mAb potently neutralizes human H7 strains of influenza virus , including strains isolated in outbreaks from 2003–2013 in Shanghai , the Netherlands , and New York [10] . In the course of our epitope-mapping studies of H7-specific mAbs using negative-stain electron microscopy ( nsEM ) , we noted a striking effect of mAb H7 . 5 binding to HA , in which the soluble H7 ectodomain trimer falls apart . Using biochemical and structural approaches described here , we delineate the phenotypic effect of H7 . 5 on HA trimers . We also report several cryo-electron microscopy ( cryoEM ) structures of H7 . 5 fragment antigen binding ( Fab ) bound to cleaved or uncleaved H7 HA trimers . Together , these data reveal a previously unappreciated antigenic determinant on HA that , while somewhat inaccessible , may be exploited as a new target for vaccine design , given its relative conservation and structural importance . These studies also indicate that the HA trimer is likely sampling subtly different prefusion conformations that may provide clues about the early aspects of the conformational changes that accompany the fusion process . Antibody H7 . 5 was described in our previous study that identified several broadly neutralizing antibodies isolated from the peripheral blood B cells of donors who participated in a vaccination trial with monovalent , inactivated influenza ( H7N9 ) , A/Shanghai/02/2013 vaccine candidate ( DMID13-0033 ) [10] . mAb H7 . 5 was shown to neutralize H7 viruses and exhibit strong inhibition of hemagglutination activity against a variety of H7 HAs . Preliminary analysis revealed that H7 . 5 recognizes multiple strains of H7 HAs with Kd ( dissociation constant ) of less than 0 . 1 nM , even in the monovalent Fab form , when measured using biolayer interferometry ( S1 Fig ) . The breadth and high affinity for the emerging H7N9 viruses made H7 . 5 an interesting target for in-depth structural studies . Hence , we first determined the structure of unliganded H7 . 5 Fab by X-ray crystallography at 2 . 0 Å resolution ( Fig 1A , Table 1 ) , but its complex with H7 HA could not be obtained , despite extensive screening of crystallization conditions . In parallel to the crystallization efforts , we used nsEM to characterize H7 . 5 Fab in complex with HA protein from different H7 strains . The difficulty in obtaining crystals of the complex became immediately apparent upon inspection of the complexes in nsEM ( Fig 1B and 1C ) . Reference-free 2D classification of cleaved H7 NY HA ( A/New York/107/2003 ) in complex with H7 . 5 Fab showed substantial heterogeneity relative to previously observed prefusion complexes of HA bound to other head-binding antibodies [10] ( Fig 1B and 1C ) . In nsEM 2D classes , we observed several different phenomena , including variable stoichiometries of bound Fabs and heterogeneous species that were difficult to interpret ( Fig 1B and 1C ) . The effect of H7 . 5 was observed consistently with both cleaved and uncleaved forms of H7 HA from multiple strains including H7 NY HA ( A/New York/107/2003 ) , H7 Sh2 HA ( A/Shanghai/2/2013 ) , and H7 NL HA ( A/Netherlands/219/2003 ) ( S2 Fig ) . We therefore attempted to capture intermediate conformations of the complex by incubating the cleaved trimer with H7 . 5 for shorter periods of time . Indeed , a 5-minute incubation resulted in enough reliable observations to deduce that H7 . 5 bound to the HA1 head but induced an unfamiliar structural phenotype in our 2D classes ( Fig 1B ) . Within a single data set , all combinations of the H7 . 5/H7 HA complex stoichiometry were observed ( from zero to three Fabs per trimer and individual protomers bound to H7 . 5 Fab ) that probed the conformational landscape of HA ( Fig 1B and 1C ) . Relative to other prefusion HA structures , the HA1 heads appeared separated , with H7 . 5 apparently prying them apart or stabilizing a more open form of the HA trimer ( Fig 1D ) . Unlike cleaved H7 HA trimers that rapidly fell apart into protomers , uncleaved H7 HA ( HA0 ) remained in a trimeric conformation even after overnight incubation with H7 . 5 ( Fig 1B ) , although the separation of the HA heads make it distinct from the closed , prefusion conformation observed in a large number of crystal structures . This relatively stable complex enabled us to image a larger number of intact particles . In comparisons of 2D classes of cleaved and uncleaved trimeric complexes of H7 HA with H7 . 5 , there appears to be no large difference whether zero , one , two , or three Fabs are bound ( Fig 1B ) . Hence , the uncleaved H7 HA , which is stable in the presence of H7 . 5 , is likely to adopt an overall similar structure to the cleaved H7 HA prior to H7 . 5-induced degradation . We do note that our higher resolution analysis discussed below does indicate that there are likely subtle differences in the molecular interactions between protomers of cleaved and uncleaved HA , depending on the stoichiometry of H7 . 5 binding . Our EM data demonstrated that H7 . 5 had a substantial disruptive effect on the prefusion structure of the HA trimer . To further validate the influence of H7 . 5 on the HA trimer structure , we investigated the susceptibility of H7 HA to trypsin protease digestion with or without H7 . 5 bound . Indeed , this experiment showed that addition of H7 . 5 Fab to H7 Sh2 HA in the presence of trypsin resulted in degradation of HA into many peptidic fragments ( S3 Fig ) . When H7 . 5 Fab was added to H7 Sh2 HA , 0 . 1% trypsin was enough to induce the cleavage of H7 Sh2 HA , and 2% was sufficient to completely degrade the H7 HA trimer . Without H7 . 5 Fab , the H7 Sh2 HA trimers remained resistant to protease cleavage up to a trypsin concentration of 2% . These data indicate that H7 . 5 may prematurely trigger structural changes or fluctuations of H7 HA trimers that acquire protease-sensitive conformations , which is reminiscent of the typical behavior of HA trimers in low-pH environments [11 , 12] . To interrogate the structure of H7 . 5 bound to H7 HA at a higher resolution , we employed single particle cryoEM of the uncleaved H7 NY HA in complex with H7 . 5 Fab . Reference-free 2D classification resulted in many different views of the complex , and secondary structure features were clearly visible ( Fig 2A ) . The cryoEM 2D classes of the complex in vitreous ice were similar to those observed in nsEM , although the majority of the classes appeared to have three Fabs bound to the HA trimers . Projection images corresponding to meaningful 2D classes were subjected to iterative angular reconstitution and reconstruction . The resulting density map exhibited significant heterogeneity in the Fab densities attached to the head as well as in the membrane-proximal part of the stem . 3D classification was then performed , and among the resulting classes , one class was characterized by having well-resolved Fabs bound to the head domain . Data in this class were subjected to further refinement resulting in a 3-fold symmetric map with a global resolution of approximately 3 . 5 Å ( EMDB-9139 , PDB 6MLM ) that is well resolved in all but the membrane-proximal stem region ( Fig 2B , Table 2 ) . An initial homology model was created using an X-ray structure of the H7 HA1/HA2 protomer ( A/New York/30732-1/2005 , 3M5G ) [13] , combined with our X-ray structure of the H7 . 5 Fab , and then individually docked into our cryoEM density map exhibiting a nice fit ( Fig 2B , 2C and 2D ) . Next , iterative rebuilding and refinement was carried out in Rosetta [14] , yielding a high-resolution atomic model . Since the membrane-proximal part of the map was not well resolved , a protocol for density-subtracted refinement of this region alone was pursued , resulting in a local density map at approximately 3 . 9 Å resolution ( S4 Fig ) . The membrane-proximal part of the model was refined under constraints of this density map , and an atomic model was created from combining the two builds . To confirm that local ( Brownian ) motion was responsible for the disorder in the membrane-proximal region—as opposed to the two maps displaying H7 HA in different conformations—coordinates assigned in the density-subtracted local refinement of cryoEM data were applied to the original data , and a reconstruction was performed . Indeed , density corresponding to HA1 and to the parts of HA2 not included in local refinement were clearly visible , albeit at an expected lower resolution ( S5 Fig ) , thereby justifying combining the two builds into a meaningful global conformation supported by our experimental observation . The epitope on the H7 HA trimer recognized by H7 . 5 Fab was delineated based on the cryoEM model ( Fig 3A ) . Intriguingly , the H7 . 5 antibody was found to simultaneously interact with two separated surfaces on two adjacent HA protomers ( Figs 2D and 3B ) . Moreover , our model illustrated that the epitope recognized by H7 . 5 is not completely accessible when all three HA1 heads are close together , as observed in the apo cleaved H7 trimer crystal structure . Interestingly , the heavy-chain framework region 3 ( H-FR3 ) of H7 . 5 was juxtaposed to the RBS of an adjacent HA protomer ( S6 Fig ) , and in a closed-trimer model , CDRH3 of H7 . 5 would clash with residues 189 to 194 in the 190 helix of the adjacent protomer ( Fig 3D ) . Compared to the apo cleaved H7 trimer model , the buried interprotomer surfaces in the H7 . 5 epitope were separated substantially from each other in the H7 . 5-bound H7 NY HA model ( Fig 3B ) . The recessed nature of this epitope suggests that one of the surfaces was likely the primary binding interface recognized by H7 . 5 antibodies , while the other surface allosterically alters upon antibody binding . Indeed , further interface analysis revealed that the HA surface interacting with CDRH2 and CDRL3 of H7 . 5 ( surface 1 ) generates a total buried surface area of 720 Å2 , with substantial favorable energy estimated to be −5 kcal/mol and an extensive hydrogen bond network between H7 . 5 and the HA in surface 1 ( Fig 3B and 3C ) . In contrast , the surface between the adjacent HA protomer and H-FR3 of H7 . 5 ( surface 2 ) is significantly smaller ( only 240 Å2 ) and appears to be energetically unfavorable with a ΔG ( change in Gibbs free energy ) of +1 . 2 kcal/mol ( Fig 3B and S6 Fig ) . To confirm the structural observations , we performed mutagenesis studies of residues in the antibody paratope by individually reverting the residues to the corresponding germline residues and measuring the affinity changes of the H7 . 5 mutants to H7 Sh2 HA . Consistent with the interface analysis , mutations in CDRH2 and CDRL3 , particularly H-S58W , H-T57A , L-T94Q , and L-Y96A of H7 . 5 , drastically reduced binding to H7 , while mutations in H-FR3 of H7 . 5 , namely H-M73R and H-S74K , which are proximal to surface 2 , did not impact the binding affinity ( Fig 3E and S6 Fig ) . These results strongly support the hypothesis that surface 1 is the primary binding surface driven by favorable energy changes , while surface 2 is a collateral result of the antibody-binding event . Although mutations in H-FR3 or CDRH3 of the H7 . 5 antibody did not affect binding of the antibody itself , Fab binding requires the adjacent protomer to be displaced from the tight 3-fold axis at the apex of the trimer to reveal the full epitope ( Fig 3D and S6 Fig ) . The H7 . 5 epitope therefore includes residues in the inter-HA head contact region ( Fig 3E and S6 Fig ) . Notably , these residues are conserved across all human and zoonotic H7 strains from 1996–2017 ( Fig 3F and 3G and S7 Fig ) but not amongst other strains , with the exception of proline at position 62 . ( S7 Fig ) . Binding of all three Fabs ultimately results in a conformation of HA with the heads splayed apart . To further probe the conformational dynamics of the H7 HA when it interacts with H7 . 5 Fab , we performed hydrogen–deuterium exchange mass spectrometry ( HDX-MS ) studies on cleaved or uncleaved H7 NL HA in the presence or absence of H7 . 5 Fab . The differential HDX-MS results of H7 NL HA and H7 NL HA/H7 . 5 complex were mapped onto the H7 NL HA crystal structure ( PDB 4FQV ) ( Fig 4A ) . The HDX-MS data revealed two protected regions ( colored in blue ) in HA1 of H7 HA upon H7 . 5 binding . These two discontinuous regions ( residues E150–K166 and residues S197–L201 ) form the potential binding epitope , which is consistent with the cryoEM results ( Fig 3C ) . Moreover , HDX-MS experiments performed on cleaved H7 NL HA alone indicated a low level of exchange in one of the regions ( residues E150–K166 ) , suggesting less solvent accessibility in the binding epitope region . On the other hand , H7 . 5 binding led to deprotection of regions ( colored in red ) in HA1 and HA2 , which indicated that HA became more conformationally flexible to accommodate the binding of multiple H7 . 5 Fabs . The structural dynamic differences between cleaved and uncleaved H7 NL HA were also compared ( Fig 4B ) . The HDX-MS results indicated that the cleaved H7 NL HA showed more protection ( colored in blue ) in the HA2 subunit compared to uncleaved H7 NL HA , which is more prone to protease cleavage . This finding is consistent with the helical domains in cleaved HA being more stable and is likely related to the maturation of HA from uncleaved HA0 to HA1 and HA2 . Detailed results are shown in S8 Fig . To trap additional structural intermediates of HA , we next attempted cryoEM reconstructions of H7 . 5 Fab bound to cleaved H7 Sh2 HA and H7 NL HA , which differ greatly in season and global location of isolation . Guided by our nsEM experiments ( Fig 1B ) , we succeeded in capturing an intermediate structure of cleaved H7 HA using a short incubation time with H7 . 5 Fab immediately prior to sample vitrification for EM experiments . The predominant classes that we observed had three H7 . 5 Fabs bound to H7 HA and were able to generate asymmetric 3D reconstructions at approximately 7 . 4 and approximately 9 . 2 Å resolution ( EMDB-9142 , EMDB-9143 ) ( Fig 5C and 5D , S9 Fig and Table 2 ) . Similar to the high-resolution structure of H7 . 5 bound to uncleaved H7 HA , we still observed movement of the HA1 heads away from the apical 3-fold axis in the cleaved H7 HAs . Further , in all reconstructions of the cleaved H7 HAs , one of the HA1 heads appeared to be splayed out further than the other two ( Fig 5C and 5D ) . In the case of cleaved H7 NL HA , we also refined a class with only two Fabs bound ( S9C Fig ) . We next compared our high-resolution cryoEM-derived model of the H7 . 5 bound to H7 NY HA , with separated heads to a crystal structure of HA in the canonical closed prefusion conformation ( PDB 3M5G ) . Within a single protomer , our hybrid docking revealed that the HA1 movement was coupled to a subtle change in its position relative to HA2 ( Fig 5 ) . At the trimer level , head separation was accompanied by movement in the HA2 central helices ( Fig 5A ) . Hence , the HA head motions appear to be communicated downward to the stem , resulting in a small separation of the stem protomers . We hypothesize that this looser packing may trigger events that ultimately liberate the fusion peptide or otherwise destabilize the HA trimer and would lead to initiation of the downhill fusion process , a phenotype that would explain our initial nsEM results with the cleaved HA showing dissociation of the trimer into protomers after incubation with H7 . 5 . Additionally , the head separation correlates with HA2 residues G57–K58 adopting an α-helical conformation ( Fig 5B ) . These two amino acids were observed previously as a random coil following the C-terminus of HA2 helix A [13] . The segment ( residues G57–I73 ) connects helix A with the HA2 central helix; in the stable , trimeric HA2 postfusion state , this segment was observed to be α-helical , forming a continuous extended helix with helix A and the central helix [15] . The α-helical organization of residues G57–K58 also correlates with further separation of the membrane-proximal H7 stem . Thus , this structure likely represents an intermediate conformational state of the class I fusion protein of influenza virus . Fig 6 summarizes the results and the events leading up to the trimer dissociation when H7 . 5 is added . With the ever-present risk of new avian influenza outbreaks looming , it is important to explore novel ways to target influenza . The recently identified H7 . 5 recognizes a new conserved epitope on the HA head that can be targeted by neutralizing antibodies . The H7 . 5 epitope , conserved throughout H7 HA , resides at the HA head protomer–protomer interface and is only transiently accessible . Neutralizing antibodies that bind the HA1 head often target the RBS , blocking HA from interacting with sialic acid on host cells [16] . Although H7 . 5 does not bind directly to the RBS , it does however block sialic acid binding to HA [10] and potently blocks intact HA from approaching the cell surface . The prefusion structure of HA has been described extensively , with HA1 and HA2 tightly interacting in a compact arrangement in both cleaved and uncleaved structures [10 , 17 , 18] . Portions of the postfusion HA structure have also been crystallized and described [19 , 20] . Interestingly , mAb H7 . 5 induces premature dissociation without exposure to low pH , and the somewhat open conformation of the receptor-binding domains may be reminiscent of early steps in the influenza virus HA-mediated membrane-fusion process . After H7 . 5 binds , it likely induces progressive opening of the trimer , and as more antibodies bind , the trimer is pushed over the activation barrier required for transition into a postfusion conformation [1] . In fact , there are other recently reported HA head–directed antibodies that have indirectly been shown to induce a similar phenotype [21 , 22] . Notably , these antibodies do not bind to epitopes that overlap with H7 . 5 ( S10 Fig ) . Antibody-induced decay of HA , via different head-binding antibodies , likely represents a new mechanism of influenza neutralization inhibition . To date , HA has been shown to adopt a stable , closed prefusion conformation [17 , 23] . However , despite significant effort , intermediates in the fusion process have eluded high-resolution characterization . It is likely that these intermediates are highly unstable and rapidly transition to the postfusion form [24] . Altogether , our cryoEM studies reveal that HA appears to be somewhat dynamic in its prefusion state , wherein protomers are able to undergo structural fluctuations . Indeed , we observed movement throughout the HA molecule , including fluctuations in the HA2 stem . The inferred movements of the HA heads to enable antibody H7 . 5 binding suggest that the HA1 head is “breathing , ” and this motion is reminiscent of conformational masking described for the HIV envelope glycoprotein [25–27] . Indeed , a very recent study using single-molecule FRET ( Förster resonance energy transfer ) observed that H5 HA undergoes reversible conformational changes [28] . The structural fluctuations observed in our study , as well as the smFRET study , indicate that movement of the HA1 heads alters interactions between the HA1/HA2 stem portion and the fusion peptide , which reside in the trimer core . HDX-MS studies of H3 HA liberated from whole virions demonstrated a stepwise progression of acid-induced fusion peptide liberation followed by HA head separation [29] . Whether these movements are similarly recapitulated amongst all strains of HA remains to be investigated , but there are likely to be conserved and coordinated motions , given the critical role that the HA-fusion machine plays in viral entry . Thus , there may be ways to exploit these structural fluctuations and access conserved epitopes that have not previously been appreciated relative to the more canonical RBS and stem antibodies . H7 . 5 Fab for crystallization was prepared as previously described [30] . In brief , the heavy and light chains of H7 . 5 were cloned independently into the phCMV3 vector and fused with an N-terminal IgK secretion signal . A His6 tag was added to the C-terminus of the Fab heavy chain . Recombinant cDNAs encoding the Fab heavy and light chains were purified and cotransfected into 293F cells by 293fectin ( Invitrogen ) . After 6–7 days of expression at 37 °C , the Fabs were purified from the supernatant by Ni-NTA Superflow ( Qiagen ) and monoS chromatography ( GE Healthcare ) . H7 . 5 Fab was concentrated to 10 . 0 mg/mL in the buffer of 50 mM NaOAc ( pH 5 . 5 ) for crystallization screening on our high-throughput robotic CrystalMation system ( Rigaku ) at The Scripps Research Institute ( TSRI ) using sitting-drop vapor diffusion . The best crystals grew in the well with 0 . 1 M MES ( pH 6 . 5 , 0 . 01 ) , M cobalt chloride , and 1 . 8 M ammonium sulfate as mother liquor at 4 °C . Crystals were cryoprotected with mother liquor supplemented with 15% ethylene glycol . X-ray diffraction data were collected to 2 . 00 Å resolution on beamline 23-ID-D at the Advanced Photon Source ( APS ) . The diffraction data were processed with HKL2000 [31] , and the structure was determined by molecular replacement with initial models of PDB 3N9G and PDB 4KMT in Phaser [32] . Refinements were carried out in PHENIX [33] , and model rebuilding was performed manually in Coot [34] and the model validated by MolProbity [35] . Hydrogen bond and buried molecular surface analyses were calculated using the PDBePISA server of EMBL-EBL . Structure figures were generated by MacPyMOL ( DeLano Scientific , LLC ) and UCSF Chimera package . Chimera is developed by the Resource for Biocomputing , Visualization , and Informatics at the University of California , San Francisco ( supported by NIGMS P41-GM103311 ) [36] . Baculovirus-expressed HA was prepared for the study as previously described [3 , 37 , 38] . In brief , the HA ectodomain sequence was cloned into the pFastBac vector with an N-terminal gp67 secretion signal peptide , a C-terminal BirA biotinylation site , thrombin cleavage site , foldon trimerization domain , and His6 tag . Recombinant bacmid DNA was generated via the Bac-to-Bac system ( Invitrogen ) , and Baculovirus was generated by transfecting purified bacmid DNA into Sf9 cells . HAs were expressed by infecting the High Five cells with the recombinant virus , shaking at 110 rpm for 72 hours at 28 °C . The secreted HAs were purified from the supernatant via Ni-NTA Superflow ( Qiagen ) and gel filtration ( to collect only the HA trimer ) . The HA trimer was concentrated in the buffer of 20 mM Tris-HCl ( pH 8 . 0 ) and 150 mM NaCl and aliquoted for storage . For the protease susceptibility assay , the H7 HA in the experimental group was firstly mixed with H7 . 5 Fab with molar ratio of 1:3 and diluted to 2 mg/ml ( for H7 only ) in the buffer of 20 mM Tris-HCl ( pH 8 . 0 ) and 150 mM NaCl . The control group H7 was diluted with the same buffer to the sample concentration . Both groups of H7 samples were then aloquated and mixed with 0% , 0 . 1% , 0 . 2% , 1% , and 2% of trypsin ( dissolved in the same buffer above ) , and each reaction was incubated at room temperature overnight . The samples were then submitted for sodium dodecyl suflate polyachrylamide gel electrophoresis ( SDS-PAGE ) analysis to determine the stability of the H7 protein . An Octet RED instrument ( FortéBio , Inc . ) was used to determine Kd of the antibody–antigen interactions by biolayer interferometry . To determine the binding affinity of H7 . 5 Fab to its ligands , HAs ( biotinylated as previously described [3 , 37] ) were immobilized onto streptavidin-coated biosensors ( FortéBio , Inc . ) and incubated with H7 . 5 Fab at 62 . 5–500 nM for 120 seconds for association and then incubated in PBS buffer with 0 . 2% BSA for dissociate for 120 seconds . The signals for each binding event were measured in real time , and Kd values were calculated by fitting to a 1:1 bivalent analyte model . In this case , the Kd values were estimated to be less than 10−3 nM , as no dissociation was observed . H7 . 5 IgG was digested with 4% papain w/w for 4 hours before adding iodoacetamide . The Fab was purified using a protein A column and then added to H7 NY HA with a 3x molar excess of Fab and incubated for 30 minutes at room temperature . The complex was purified using an S200i column ( GE Healthcare , Amersham , United Kingdom ) and immediately added to 400 mesh copper grids with 2% uranyl formate . Micrographs were taken on a Tecnai Spirit with a TemCam F416 camera using Leginon [39] . Particles were selected using DoG Picker [40] and then stacked and aligned into 2D classes in Appion [41] with MRA/MSA [42] . Particles not representing HA trimers or the complex were removed , and a clean stack was processed in RELION [43] . Sample preparation for uncleaved H7 HA was similar to negative stain . Directly after size exclusion chromatography ( SEC ) purification , the complex was added at 1 mg/mL to 2/2 gold Quantifoil grids with amphipol and frozen in liquid ethane using a Vitrobot . Cryo sample preparation for cleaved H7 HA with H7 . 5 Fab does not include a column purification step . Complexes were formed , incubated for one minute at room temperature , and immediately frozen . 1 , 324 micrographs were recorded of the uncleaved H7 NY HA/H7 . 5 Fab complex on a Gatan K2 summit detector mounted on a Titan Krios operating at 300 kV . Data were collected in counting mode at a nominal magnification of 29 , 000x . Dose rate was approximately 10 electrons/ ( camera_pixel*s ) , and frame exposure time was 200 ms . Total exposure time was 10 seconds , with a total dose of 67 electrons/Å2 . Movie micrograph frames were aligned using MotionCor2 [44] , dose weighted , and integrated . Contrast transfer function ( CTF ) models were determined in GCTF [45] . Projection image candidates of H7 NY HA/H7 . 5 were identified using a difference-of-Gaussians picker [40] . The resulting set of candidate projection images were subjected to 2D class averaging implemented in RELION 2 . 1b1 [43] . 227 , 202 projection images that generated well-formed class averages were selected for further data processing . Iterative angular reconstitution and reconstruction was attempted . Due to on-symmetry axis preferred orientations , reconstruction artefacts were insurmountable for asymmetric refinement . C3 symmetry was imposed in a second attempt , and a stable , albeit somewhat stretched , 3D reconstruction of the data set was obtained . The data set was then subjected to 3D classification , and a subclass of the data ( 30 , 032 projection images ) was identified; its reconstruction was characterized by persistent Fab densities—indicating full stoichiometric occupancy—and no obvious reconstruction artifacts . This subclass of data was then refined under C3 symmetry constraint to a final resolution of 3 . 5 Å . Fab Fv regions were well resolved , as were HA1 and part of HA2 . The membrane-proximal stem of HA2 was largely disordered . Attempting to recover a structure of the membrane-proximal stem , a globular mask was placed encompassing the disordered part of the map . Inverting the mask , the ordered part of the map could be isolated from which a mask was created to isolate the constant part of the original map . This constant part map was then used to project onto R2 along projection directions deduced from Euler and X , Y coordinate assignments obtained during the earlier-referenced refinement , followed by subtraction from the original projection images . This density-subtracted data set was subjected to further iterative angular reconstitution and reconstruction , resulting in a density map of the membrane-proximal stem resolved to 3 . 9 Å . To investigate if the two maps could be combined and interpreted as one instance of multiple only locally diverging—breathing—conformations , the Euler and X , Y coordinates determined in the stem refinement was applied to the corresponding original projection images and the coinciding 3D object reconstructed . This density map exhibited , albeit noisy , density for the constant part of the structure , suggesting that the membrane-proximal part of the stem is breathing locally around the position determined in the local map . Treatment of our cleaved H7 NL HA/H7 . 5 and H7 Sh2 HA/H7 . 5 data sets was performed in RELION 2 . 1b1 and followed a standard procedure , as outlined above , prior to local refinement procedures and as outlined in the supplemental information ( S5 Fig ) . FSC 0 . 143 between independently refined half sets were 7 . 4 Å ( H7 Sh2 HA/3 H7 . 5 Fabs ) , 9 . 2 Å ( H7 NL HA/3 H7 . 5 Fabs ) , and 10 . 2 Å ( H7 NL HA/2 H7 . 5 Fabs ) . A homology model was created ( Modeller; [46] ) based on the X-ray structure of HA7 HA1/H2 ( A/New York/30732-1/2005; PDB 3M5G ) [13] . The model was created by combining with the H7 . 5 Fab X-ray structure and rigid-body fitted to the cryoEM density map . A fragment library consisting of 200 7mers for each amino acid position in the model was compiled from a nonredundant protein structure database . Iterations of manual and Rosetta fragment-library-based centroid rebuilding-and-refinement was then performed [14] . The resulting model was all-atom-refined under constraints of the density map . Glycans were manually built in COOT [34] and final rounds of real-space refinement performed in PHENIX 1 . 12 [47] . The membrane-proximal part of HA2 was built similarly in the local map referenced above . The builds were then combined into the final structure . Evaluation of builds were performed in MolProbity [35] , EMRinger [48] , and Privateer [49] and by the PDB validation server . Antigen–antibody complexes were prepared by mixing HA H7 and H7 . 5 antibodies at 1:1 . 1 stoichiometric ratio and incubating at room temperature for 30 minutes and then kept at 0 °C . For control experiments , free antigens were also diluted with the same volume of protein buffer and treated the same way as complexes . To initiate HDX reactions , 2 μl of prechilled protein stock solution ( free uncleaved HA NL H7 , 1 . 8 mg/ml; H7 . 5-uncleaved HA NL H7 , 4 . 5 mg/ml; cleaved HA NL H7 , 1 . 6 mg/ml; or H7 . 5-cleaved HA NL H7 , 4 . 5 mg/ml ) was diluted into 4 μl D2O buffer ( 8 . 3 mM Tris , 150 mM NaCl , in D2O , pDREAD 7 . 2 ) at 0 °C . At the indicated times of 10 seconds , 100 seconds , 1 , 000 seconds , 10 , 000 seconds , and 100 , 000 seconds , the exchange reaction was quenched by the addition of 9 μl of optimized quench solution ( 6 . 4 M GuHCl , 1 M TCEP , 0 . 8% formic acid [pH 2 . 4] ) at 0 °C . After incubating on ice for 5 minutes , the quenched sample was diluted 5-fold with 0 . 8% formic acid containing 16 . 6% glycerol , immediately frozen on dry ice , and stored at −80 °C . In addition , undeuterated samples and equilibrium-deuterated control samples were also prepared . All samples were then loaded onto our in-house DXMS apparatus for online digestion and separation . The resulting peptides were directed into an OrbiTrap Elite Mass Spectrometer ( Thermo Fisher Scientific , San Jose , California ) for DXMS analysis . Instrument settings have been optimized for HDX analysis . The data acquisition was carried out in a data-dependent mode and the 5 or 10 most abundant ions were selected for MS/MS analysis . Proteome Discoverer software was used for peptide identification . The centroids of each peptide was calculated with HDExaminer and then converted to corresponding deuteration levels with corrections for back exchange . The deuteration levels of the peptides were further sublocalized using overlapping peptides by MATLab program . Nonduplicate HA sequences were fetched from the databases at www . fludb . org in FASTA format . Strain A/Aichi/2/1968 HA gene ( Genbank Accession M55059 ) was then defined as a reference for residue numbering . We then used an in-house script to build a database of every HA sequence’s residues at every position relative to the reference . First , we performed pairwise Clustal alignment of every sequence to the reference sequence [50] . The results were parsed into a database keyed on positions relative to the reference , with gaps notated as subpositions following the last identical residue . We are then able to interrogate the database by providing position numbers and desired residues at the position , providing an output of sequences that meet the search criteria as well as breakdowns of residues at the positions . Analysis of natural HA variants was done , as previously described in Wu and colleagues , 2018 [51] . Briefly , a total of 6 , 984 HA sequences made up of a subset of up to 20 sequences per year per subtype were analyzed from the Global Initiative for Sharing Avian Influenza Data ( GISAID; https://gisaid . org ) . Sequences were aligned using MAFFT version 7 . 157b [52] . Sequence logos were generated by WebLogo [53] .	Influenza viruses cause severe respiratory infections on a global scale annually . Vaccine efforts are hampered by the virus’s naturally high mutation rate , which results in wide variation between influenza strains of the antigens that are produced and recognized by antibodies , particularly in the surface glycoprotein hemagglutinin ( HA ) . However , broadly neutralizing antibodies ( bnAbs ) are a class of antibodies that develop during natural infections that are capable of inhibiting infection across multiple strains . In this study , we structurally characterized one such bnAb , H7 . 5 , which targets a unique semioccluded yet highly conserved region on the HA head . We showed , using both negative-stain and high-resolution cryo-electron microscopy ( cryoEM ) , that after a short incubation , H7 . 5 fragment antigen binding ( Fab ) induces HA to fall apart , effectively preventing infection . We found that H7 . 5 binds to an epitope only accessible through transient “breathing” of the HA head , and this observation provides insight into the conformational transitions necessary for viral fusion as well as key information about a unique vaccine target .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "medicine", "and", "health", "sciences", "chemical", "characterization", "immune", "physiology", "crystal", "structure", "pathology", "and", "laboratory", "medicine", "influenza", "pathogens", "immunology", "condensed", "matter", "physics", "microbiology", "orthomyxoviruses...	2019	Potent anti-influenza H7 human monoclonal antibody induces separation of hemagglutinin receptor-binding head domains
A drug exerts its effects typically through a signal transduction cascade , which is non-linear and involves intertwined networks of multiple signaling pathways . Construction of such a signaling pathway network ( SPNetwork ) can enable identification of novel drug targets and deep understanding of drug action . However , it is challenging to synopsize critical components of these interwoven pathways into one network . To tackle this issue , we developed a novel computational framework , the Drug-specific Signaling Pathway Network ( DSPathNet ) . The DSPathNet amalgamates the prior drug knowledge and drug-induced gene expression via random walk algorithms . Using the drug metformin , we illustrated this framework and obtained one metformin-specific SPNetwork containing 477 nodes and 1 , 366 edges . To evaluate this network , we performed the gene set enrichment analysis using the disease genes of type 2 diabetes ( T2D ) and cancer , one T2D genome-wide association study ( GWAS ) dataset , three cancer GWAS datasets , and one GWAS dataset of cancer patients with T2D on metformin . The results showed that the metformin network was significantly enriched with disease genes for both T2D and cancer , and that the network also included genes that may be associated with metformin-associated cancer survival . Furthermore , from the metformin SPNetwork and common genes to T2D and cancer , we generated a subnetwork to highlight the molecule crosstalk between T2D and cancer . The follow-up network analyses and literature mining revealed that seven genes ( CDKN1A , ESR1 , MAX , MYC , PPARGC1A , SP1 , and STK11 ) and one novel MYC-centered pathway with CDKN1A , SP1 , and STK11 might play important roles in metformin’s antidiabetic and anticancer effects . Some results are supported by previous studies . In summary , our study 1 ) develops a novel framework to construct drug-specific signal transduction networks; 2 ) provides insights into the molecular mode of metformin; 3 ) serves a model for exploring signaling pathways to facilitate understanding of drug action , disease pathogenesis , and identification of drug targets . Most drugs exert their therapeutic actions through interactions with specific protein targets . These target proteins are dominated by two categories: enzymes that catalyze reactions essential for the functioning of organisms , and receptors that transmit signals by interacting with messenger molecules [1 , 2] . The interactions of drugs and their targets initiate the signal transduction cascade that is usually propagated by the involved proteins and multiple pathways . These proteins and pathways act in the mode of crosstalk networks [3] . The process of such signaling transduction converts the chemical signals to a specific cellular response such as gene expression , cell division , and inhibition of cell death and apoptosis [4] . The signaling cascade usually ends at the recipients of chemical signals such as transcription factors ( TFs ) , which have specific binding sites on DNA and play critical roles in the gene expression regulation [5] . In complex diseases such as cancer [6 , 7] , neuropsychiatric disorders [8] , and diabetes [9] , these molecules involved in the signal transduction cascade that are altered and , thus , become attractive targets for disease treatment [10 , 11] . Therefore , targeting signaling pathways has become an important approach to discovering new drugs through traditional experimental methods [12 , 13] and to predicting drug repositioning through systematic approaches [14] . However , the primary challenge for utilizing signal transduction pathways for drug discovery is to synopsize the drug signaling pathways into one comprehensive system , including the major causal genetic factors for pathology of the complex disease and the most elemental components in the drug action . Recent high-throughput technologies such as array-based mRNA and microRNA expression , genome-wide association studies ( GWAS ) , and next-generation sequencing ( NGS ) have provided massive amounts of data , enabling investigation of drug effect through pharmacogenomic network approaches . For example , the Connectivity Map ( CMap , build 02 ) studied the effect of 1 , 309 small chemicals on gene expression in four cultured human cells [15] . Furthermore , multiple reliable drug-centered databases such as DrugBank [16] , KEGG ( Kyoto Encyclopedia of Genes and Genomes ) DRUG [17] , PharmGKB ( The Pharmacogenomics Knowledge Base ) [18] , and STITCH ( Search Tool for Interactions Chemicals ) [19] , provide comprehensive and detailed drug information for computational discovery and/or drug design . Therefore , it is possible to integrate known drug targets , genes involved in drug pharmacokinetics ( PK ) and pharmacodynamics ( PD ) processes , drug-induced gene expression data , and disease-gene associations . Additionally , network-assisted approaches have become powerful tools to explore disease-gene , gene-gene , as well as drug-target associations in pharmacology and human disease [20–23] . Therefore , we hypothesized that the construction of a signaling pathway network to connect the upstream components and downstream signal recipients for an individual drug would increase power to identify genes that play critical roles in drug action or disease development . In this study , we develop a computational framework , called DSPathNet , to construct one signaling pathway network ( SPNetwork ) for a particular drug via amalgamating drug knowledge with drug-induced gene expression data . The main purposes are to capture the principal components in the drug signal transduction process and to provide an alternative approach to identifying critical elements and modules ( subnetworks ) relevant to drug action . We illustrate the utility of DSPathNet using the metformin , one of the most widely prescribed anti-diabetic drugs in the world which has been recently shown to be useful for cancer treatment and prevention in people at higher risk [24–26] . We started with the collection of known drug-related genes and inference of TFs from metformin-induced gene expression data . Considering that most of the known drug-related genes participate in PK and PD processes and are located in the upstream of the signaling cascade based on their function , we defined them as “metformin upstream genes . ” Likewise , we defined the TFs that receive and transmit the chemical signals at the end of the signaling cascade as “metformin downstream genes . ” After overlaying the two sets of genes onto human SPNetwork , we employed random walk algorithms to construct a metformin-specific SPNetwork . The random walk-based methodology aims to identify the pathways that are closet to the known disease genes compared to other methods [27] and offers the best predictive performance [28] . The network is expected to enrich with signaling genes involved in metformin signal transduction . We performed the comprehensive gene enrichment analyses of the network using the disease genes of type 2 diabetes ( T2D ) from GWAS catalog [29] , cancer genes from Cancer Gene Census [30] , one T2D GWAS [31] , three cancer GWAS [32 , 33] , and one novel GWAS of cancer patients with T2D using metformin from BioVU [34] . The enrichment analysis results showed that the network contained a significant number of T2D and cancer disease genes and genes related to metformin action , indicating that the framework is promising as a method to identify critical genes involved in disease pathology and drug action . Additionally , the metformin-specific SPNetwork generated here provides potential metformin targets and molecular insights for further delineating the mechanism of metformin action . In this study , we develop a novel computational framework to build a Drug-specific Signaling Pathway Network , namely DSPathNet , for constructing a signaling pathway network ( SPNetwork ) for an individual drug of interest . The drug-specific SPNetwork is expected to contain critical components in the drug’s signal transduction cascade . These components are genes that harbor genetic variations contributing to the pathology of the drug indication or drug response . Thus , the drug-specific SPNetwork would facilitate our understanding of the molecular mechanisms of drug action , disease pathogenesis , and identification of novel drug targets . To prove the principle , we utilized the drug metformin as an example to evaluate the framework . Fig 1 outlines the framework to build the metformin-specific SPNetwork and S1 Table summarizes the data sources , software and evaluation data used in the study . Briefly , we first collected metformin upstream genes from multiple sources and inferred metformin downstream genes from metformin-induced gene expression data . We compiled a human SPNetwork from two databases , Pathway Commons [35] and TRANSFAC [36] , as a background pathway system for all signal transduction processes in humans . To weight the association of each node with metformin action , we assigned a functional similarity score to each node based on their Gene Ontology ( GO ) annotations and metformin upstream genes . The human SPNetwork included 37 , 881 edges and 4 , 367 nodes . Then , we utilized the metformin upstream and downstream genes as seeds to produce the metformin-specific SPNetwork from the human SPNetwork via random walk approaches . In this process , we applied a crossing network strategy to generate the drug-specific SPNetwork from background human SPNetwork by longitudinal and lateral movements . Finally , we computationally evaluated the metformin-specific SPNetwork by examining the enrichment of genes in the network using two types of data . The first includes the disease genes of type 2 diabetes ( T2D ) and cancer , the two diseases in which metformin has been actively studied . The second contains the individual genotyping data from five GWAS datasets: one T2D GWAS dataset , three cancer GWAS datasets , and one GWAS dataset of cancer patients with T2D treated by metformin . Our evaluation results indicated that the metformin-specific SPNetwork was significantly enriched with genes with mutations that could contribute to the pathology of T2D and cancer , and genes that may be associated with metformin-associated cancer survival ( Table 1 ) . To further investigate the molecular mechanisms underlying metformin action , we built a crosstalk subnetwork based on common genes to T2D and cancer , network topology , and functional analyses . We revealed several critical components , modules , and pathways that might be involved in metformin action . In order to generate a complete and reliable SPNetwork , we extensively collected the metformin related genes , rigorously selected the expressed genes induced by metformin , and comprehensively compared the performance using T2D GWAS data after the SPNetwork generation . For each step , we provide the detailed information as below . The final metformin-specific SPNetwork generated above comprised 477 nodes and 1 , 366 edges ( Fig 3B , S5 Table ) . Among the 477 nodes , 215 belonged to metformin upstream network , while 303 belonged to metformin downstream network . There were 41 bridge nodes between them . Thus , 174 genes were unique to the metformin upstream network , and 262 genes were unique to the metformin downstream network . From here , we refer to the three gene sets as upstream genes ( number of genes: 174 ) , downstream genes ( 262 ) , and bridge genes ( 41 ) for follow-up network topological and functional analyses . To explore the topological properties of this SPNetwork , we calculated node degrees ( connectivity ) and their distribution [41] . In this network , degree values of nodes ranged from 1 to 79 and the average degree was 5 . 73 . The degree distribution was strongly right-skewed , indicating that most nodes had a low degree and only a small portion of the nodes had a high degree ( Fig 3C ) . The nodes with a high degree act as hubs in the network and hold the whole network together [41] . In biological networks , hubs are more likely to be essential genes [42] and disease genes [43–45] . Using the hub defining method proposed by Yu et al . [46] , we determined 38 hubs whose degrees were larger than 14 . Among them , one gene ( PPARG ) belonged to both metformin upstream and downstream gene sets , two genes ( TP53 and SREBF1 ) were metformin upstream genes , 13 belonged to the metformin downstream gene set only , and 22 were novel genes . After extracting these hubs from metformin-specific SPNetwork , we generated a hub-centered subnetwork ( Fig 3D ) . Among the 38 hubs , 19 ( 50 . 00% ) are included in ‘pathway in cancer’ and 9 ( 23 . 68% ) in ‘MAPK signaling pathway’ according to KEGG pathway annotation . The MAPK signaling pathway plays important roles in the pathology of both cancer [47] and diabetes [48] . Thus , the 477 genes had two genes belonging to metformin upstream and downstream genes , 33 to the metformin upstream genes , 58 to the metformin downstream genes , and 384 novel genes ( S6 Table ) . The novel genes may provide a valuable resource for further investigation of the pathology of T2D and cancer , and the metformin action . We further examined pathway enrichment in these 477 nodes based on KEGG pathway annotation using the online tool WebGestalt [49] . We identified 69 significant pathways ( adjusted P-value < 1 . 00 × 10–4 ) ( S7 Table ) . According to the KEGG pathway first-level category annotation ( Materials and Methods ) , 12 pathways belonged to ‘environmental information processes , ’ nine to ‘cellular processes , ’ 18 to ‘organismal systems , ’ and 29 to ‘human disease . ’ Among these 12 environmental information processes pathways , eight were signal transduction pathways , of which the top three pathways were ‘MAPK signaling pathway’ ( 32genes , adjusted P-value: 3 . 39 × 10–22 ) , ‘mTOR signaling pathway’ ( 13 genes , adjusted P-value: 6 . 39 × 10–14 ) and ‘ErbB signaling pathway’ ( 15 genes , adjusted P-value: 1 . 89 × 10–13 ) . Among the 18 pathways related to organismal systems , five belonged to the endocrine system , of which the top three pathways were ‘adipocytokine signaling pathway’ ( 22 genes , adjusted P-value: 3 . 19 × 10–25 ) , ‘PPAR signaling pathway’ ( 22 genes , adjusted P-value: 5 . 36 × 10–25 ) , and ‘insulin signaling pathway’ ( 23 genes , adjusted P-value: 1 . 91 × 10–19 ) . Among the 29 pathways related to human disease , 15 were directly related to cancer . Importantly , the pathway ‘type II diabetes’ ( 10 genes , adjusted P-value: 1 . 12 × 10–10 ) and the ‘maturity onset diabetes of the young’ ( 8 genes , adjusted P-value: 1 . 94 × 10–10 ) were among the enriched pathways . Together , the evidence indicates that the metformin-specific SPNetwork involves both diabetes and cancer at the pathway level . In the metformin-specific SPNetwork , there were 41 genes ( bridge genes ) common to both the metformin upstream and downstream networks . As mentioned above , most of them ( 31 , 75 . 6% ) were novel linkers ( S6 Table ) . To interrogate their roles , we compared them with upstream genes ( 174 ) and downstream genes ( 262 ) via network topological and functional analyses , as described below . Since metformin is a well-studied drug for T2D treatment , the metformin-specific SPNetwork was expected to contain genes that have genetic association with T2D . To examine this expectation , we comprehensively performed enrichment analysis using two sets of genes . The first one contained 131 genes collected from 66 T2D GWAS studies curated by the NHGRI GWAS Catalog database ( April 1 , 2014 ) [29] . Those genes have been reported to be significantly associated with T2D based on GWA studies . Here , we selected these genes having at least one SNP with P-value less than 1 . 0 × 10–8 as T2D associated genes . The second set included the T2D-related genes from the WTCCC T2D study [31] as mentioned above . Among the 477 nodes in the metformin-specific SPNetwork , 11 genes were found in the first set of 131 genes . Compared to the human protein-coding genes ( 20 , 716 ) , the network was significantly enriched for T2D associated genes ( Hypergeometric test , P-value: 1 . 36 × 10–4 ) . Similarly , among the 131 T2D disease genes , 43 existed in the human SPNetwork ( 4 , 367 ) . Thus , compared to all nodes in the human SPNetwork , the metformin-specific SPNetwork was significantly enriched for T2D associated genes too ( P-value: 3 . 62 ×10–3 ) . These 11 genes were CDKN2B , HNF1A , HNF4A , IRS1 , ITGB6 , KCNJ11 , LEP , PPARD , PPARG , SND1 , and TCF7L2 . Among them , KCNJ11 , PPARG , and TCF7L2 have the strongest genetic association among genes that appear in the T2D GWAS studies based on a comprehensive review [59] . Among the 477 genes in metformin-specific SPNetwork , 445 had genotyping data from WTCCC T2D GWAS dataset . Among them , 169 genes belonged to T2D-related genes . Compared with all genes with genotyping data in the GWAS , the network was significantly enriched with T2D-related genes ( Hypergeometric test P-value: 3 . 08 ×10–5 ) . We further compared the 169 genes with the genes having genotyping data in the human SPNetwork . Among the 4 , 367 nodes in the human SPNetwork , 3 , 446 genes had genotyping data , in which 1 , 048 genes were T2D-related genes . Thus , the metformin-specific SPNetwork was significantly enriched for the T2D-related genes as compared to the whole human SPNetwork in this study ( P-value: 7 . 47 ×10–3 ) . Fig 5A shows the comparison of P-value distributions of genes in whole GWAS data ( T2D GWAS ) , human SPNetwork , and metformin-specific SPNetwork . These comparisons indicate that the network is enriched with genes that might be involved in the pathology of T2D . We further generated a subnetwork for 169 nominally significant genes with T2D ( Fig 5B ) by their direct links . Among the 169 genes , 50 genes had SNPs whose P-values were less than 0 . 01 in the WTCCC T2D GWAS . In addition , there were six genes observed in both the 131 GWAS Catalog genes and the 169 genes; they are CDKN2B , ITGB6 , KCNJ11 , PPARD , PPARG , and TCF7L2 . Among them , the SNP rs4506565 in gene TCF7L2 has the strongest significance ( P = 5 . 68 ×10–13 ) . TCF7L2 encodes a transcription factor that regulates the transcription of several genes . It is a key element in the WNT signaling pathway , which has been reported to contribute to T2D risk significantly [59] . Above pathway analysis indicated that the metformin-specific SPNetwork was significantly associated with cancer-related pathways . Here , we further examined if the SPNetwork is enriched with cancer genes from four data sets . The first one included 509 cancer genes downloaded from the Cancer Gene Census ( December 11 , 2013 , http://cancer . sanger . ac . uk/cosmic/census ) . Among them , 64 genes were included in the metformin-specific SPNetwork . Compared to all human genes or the protein-coding genes in the human SPNetwork , the network was significantly enriched with cancer genes ( Hypergeometric test , P-value: 1 . 64 × 10–29 and 6 . 48 × 10–8 , respectively ) . Interestingly , 3 of the 64 genes ( HNF1A , PPARG , and TCF7L2 ) were in the T2D GWAS Catalog gene list , and 21 genes belonged to 169 T2D-related genes ( see above ) . This observation strongly indicates that metformin may affect the shared genetic risk factors between T2D and cancer . Such information provides clues for how metformin acts in T2D and cancer treatments . This observation also provides evidence for epidemiological studies of metformin in both T2D and cancer [50] . Additionally , we performed the GSEA of the metformin-specific SPNetwork using three cancer GWAS datasets from the Cancer Genetic Markers of Susceptibility ( CGEMS ) projects ( breast cancer [32] , pancreatic cancer [33] , and prostate cancer [32] ) . Table 1 summarizes the corresponding gene numbers in each GWAS dataset . Compared with all genes with genotyping in each GWAS dataset , the metformin SPNetwork was slightly significantly enriched in nominally significantly associated genes ( Hypergeometric test P-values: 0 . 0144 , 0 . 0120 , and 0 . 0053 for breast , pancreatic , and prostate cancer , respectively ) . Though the results of these statistical tests are not as robust as that of the genotyping data from the T2D GWAS study , the results confirm that the metformin-specific SPNetwork was enriched with genetic factors associated with cancer development . From above analyses , the metformin-specific SPNetwork is enriched with genes associated with T2D and cancer . Several studies over the last few years have demonstrated that patients using metformin have reduced cancer risk and improved cancer survival in T2D patients [24 , 26 , 60 , 61] . Thus , we evaluated whether metformin-specific network enrich genes associated with cancer survival among cancer patients with T2D using metformin . In this study , we took advantage of GWAS data of cancer subjects with T2D treated with metformin from BioVU [34 , 62] ( Materials and Methods ) . Hereafter , this dataset is referred as “metformin GWAS . ” Among the 477 nodes in the metformin-specific SPNetwork , 458 genes had genotyping and 177 genes were nominally significantly ( P-value < 0 . 05 ) associated with T2D with better survival . Compared with all genes with genotyping data in the metformin GWAS data , the metformin-specific SPNetwork was enriched with nominally significant genes too ( Hypergeometric test , P-value: 0 . 0181 ) . We further compared the P-value distribution of metformin GWAS data for three gene sets: the metformin-specific SPNetwork , human SPNetwork , and all genes in metformin GWAS data set ( S6 Fig ) . The genes in the metformin SPNetwork had the highest proportion of P-values ( P-value < 0 . 05 ) in metformin GWAS data at the gene level . Among the 177 genes , 81 genes were included in the 169 genes whose smallest P-values were less than 0 . 05 in T2D GWAS data . While most of them did not link to each other ( S7 Fig ) , these 81 genes directly linked to other 175 genes to form a subnetwork that included 256 nodes and 910 edges . This feature indicated that the 81 genes and their direct interactors dominated the metformin-specific SPNetwork . For example , the 256 nodes accounted for 53 . 7% of all nodes and the 910 edges accounted for 66 . 6% of all edges in the metformin-specific SPNetwork . Additionally , among the 81 genes , 17 belonged to ‘pathway in cancer’: COL4A1 , COL4A2 , ERBB2 , GLI3 , ITGB1 , MECOM , MMP1 , PLD1 , PRKCA , RARB , RXRG , SMAD3 , TCF7L1 , TCF7L2 , TGFA , TGFB2 , and ZBTB16 . Collectively , the above observations indicate that the network was enriched in genes that might contribute to overall survival among cancer patients with metformin therapy . From above analyses , we observed that the metformin-specific SPNetwork was enriched with genes associated with T2D and cancer , and genes associated with metformin-associated cancer survival . To gain more insights into how metformin act in T2D and cancer treatment , we generated a subnetwork to synopsis the crosstalk between T2D and cancer based on the common genes with nominal significance ( P-value < 0 . 05 ) among the four GWAS data sets ( T2D , CGEMS breast cancer , pancreatic cancer , and prostate cancer ) . There were 25 genes common to all the four gene sets ( Fig 6A ) , and there were only five edges in the metformin-specific SPNetwork ( S8 Fig ) . By further examining degree distributions of the common 25 genes and their direct interactors ( 71 genes ) , we found that their interactors had significantly more interactions than the 25 genes as well as all the genes in the metformin-specific SPNetwork ( Wilcoxon’s test P-value: 2 . 1 × 10–4 and 2 . 4 × 10–9 , respectively ) ( Fig 6B ) . The 25 genes included one hub ( PPARG ) while the 71 genes included 21 of the 38 hub nodes in the metformin-specific SPNetwork . Similarly , the 25 genes contained three bridge nodes while the 71 genes contained 15 of the 41 bridge nodes between metformin upstream and downstream network . These observations indicate that the interactors of the 25 common nodes were more likely to play important roles for signal transduction . Starting with the 25 genes and their 71 interactors , we assembled a subnetwork by their direct links among 96 nodes . The subnetwork comprised 96 nodes and 269 edges ( S9 Fig ) . To further explore the metformin treatment mechanisms in T2D and cancer through the protein modules , we utilized software CFinder to perform network cluster and community analysis [63] . We required each node in one module participate at least one 3-vertex clique . Accordingly , we obtained three modules , which contained 6 , 9 , and 51 genes , respectively ( S10 Fig ) . We found no gene shared between the first and second modules , but one gene ( STK11 ) common to the first and third modules , or five genes ( EIF4E , PPARGC1A , PRKCA , RPS6KB1 , and SREBF1 ) common to the second and third modules . All the genes of the first and second modules belonged to metformin upstream network while most of the genes in the third module belonged to metformin downstream network . We merged them to form a network , which included 60 nodes and 210 edges ( Fig 6C ) . Since this subnetwork was generated from common genes to T2D and cancer genotyping data , we defined it as a crosstalk subnetwork of metformin action in T2D and cancer . We realized that , if we removed the nodes ( CDKN1A , ESR1 , MAX , MYC , PPARGC1A , STK11 , and SP1 ) , the connections among three modules would be lost ( S11 Fig ) . Among them , three ( MAX , MYC , and SP1 ) were both the bridge nodes and hub nodes . Therefore , these seven nodes might be functionally critical in the metformin signal transduction cascade . To further explore how the three modules and the seven key nodes might be related to metformin treatment in term of biological function meaning , we performed the KEGG pathway enrichment analysis on each module . Table 3 summarizes the enriched pathways for each module ( adjusted P-value < 1 . 0 × 10–4 ) . We labeled the enriched KEGG pathways ( adjusted P-value < 1 . 0 × 10–9 ) for each module in Fig 6C . In the first and second modules , there were two common pathways: adipocytokine signaling pathway and insulin signaling pathway . Adipocytokine signaling pathway was the top pathway in the first module ( adjusted P-value: 2 . 01 × 10–13 ) . The adipocytokine is a group of cytokines secreted by adipose tissue , which contributes to the development of insulin resistance , T2D , and cardiovascular disease [64 , 65] . The insulin signaling pathway , the top pathway in the second module , plays important roles in many complex diseases such as diabetes , obesity [66] , and neurological disorders [67] . In addition , the mTOR signaling pathway and ErbB signaling pathway were also enriched in the second module . There were 28 pathways enriched in the third community . According to KEGG pathway annotation at the second level , 15 of these 28 pathways belonged to human disease , six to signal transduction , and three belonged to the endocrine system , one to cell communication , one to cell growth and death , one to development , and one to environmental adaptation . Among the 15 human disease related pathways , 11 were for specific types of cancer . Therefore , the three modules reflected different biological processes involved in T2D and cancer . Additionally , the pathway analyses highlighted the seven nodes that are not only topological linkers but also functional linkers in the crosstalk SPNetwork of metformin action in T2D and cancer . Starting from above crosstalk subnetwork and the seven key nodes , we manually checked their publications and integrated the experimental evidence for further understanding their roles in the metformin actions . Through careful review , we summarized their function and action together and found that a novel MYC-centered pathway was hidden under the crosstalk subnetwork , which may play important roles in metformin action in T2D and cancer ( Fig 7 ) . The Myc-centered pathway included AMPK , STK11 , MYC , SP1 , and CDKN1A , which formed two small motifs: AMPK-STK11-MYC and MYC-SP1-CDKN1A . It is well known that metformin exerts anti-diabetes and anti-cancer effects via mitochondrial complex I inhibition [68 , 69] . Mitochondrial complex I inhibition increases AMP/ATP ratio , which activates AMP-activated protein kinases ( AMPKs ) [70] to cause human disease [71] . In the crosstalk subnetwork , the first module contained core members of AMPK signaling pathways ( PRKAA2 , PRKAB2 , and PRKAG2 ) , which is linked to the second and third modules through the STK11-MYC interaction . The gene LKB1encodes a key upstream activator of AMPK [51] and is known to be inactivated through mutations during lung carcinogenesis [72] . Furthermore , the metformin induces activation of LKB1 [73] . For the MYC and LKB1 , several lines of evidence show they are in opposite action in tumor . For example , LKB1 is overexpressed partly by degradation of MYC protein to inhibit lung carcinoma cell proliferation [74] . Nevertheless , their direct relationship is not clear . Recent studies have shown that metformin has an ability to reduce MYC protein level in vivo and in vitro in several types of cancer , including lung cancer [75] and prostate cancer [76] . Based on the integrative network and function analyses with experimental evidence , we suggested a feed-forward loop ( AMPK-STK11-MYC ) exists in metformin action . This network motif may act cohesively to strengthen the inhibition of MYC expression . In addition , in the crosstalk subnetwork , three nodes ( CDKN1A , MYC , and SP1 ) formed a 3-node clique . The network small motif bridges the three modules together . The SP1 is a TF that binds to the GC-rich motif of numerous genes’ promoters and is involved in many cellular processes , including cell differentiation , cell growth , apoptosis , immune responses , response to DNA damage , and chromatin remodeling . It has been reported that SP1 could cooperate with MYC to activate transcription of the human telomerase reverse transcriptase gene ( TERT ) , which is responsible for maintenance of the length of telomeres and its defects may lead to diseases including cancer [77] . During the process of carcinogenesis , expression of MYC and SP1 is known to be up-regulated [78] . It has been reported that metformin has an ability to down-regulate MYC [75 , 76] and SP1 [61] . Additionally , MYC [79 , 80] and SP1 [81 , 82] are also the key transcription factors involved in the regulation of insulin and insulin regulated gene transcription . MYC could directly induce both impaired insulin secretion and loss of β-cell mass [83] . SP1 could regulate the upstream target STK11 expression [84 , 85] . MYC could activate AMPK in multiple cell lines [86] . AMPK activation could reduce SP1 translocate from cytoplasm to nucleus [87] . The CDKN1A , a cyclin-dependent kinase inhibitor p21 , inhibits proliferation both in vitro and in vivo . After metformin treatment , the expression of CDKN1A is upregulated in hepatocellular carcinoma [88] and bladder cancer cells [89] . Additionally , multiple lines of evidence have demonstrated that MYC can suppress the expression of CDKN1A in cancer like colorectal cancer [90] . Therefore , taken all evidence together with the crosstalk network , we propose a new biological pathway for metformin action focused on four key nodes ( CKDN1A , MYC , SP1 , and STK11 ) ( Fig 7 ) . The pathway highlights several new questions , which may have been missed by previous studies . Specifically , we speculate that MYC and its networks are the key downstream targets of metformin . Further investigations are needed to illustrate this mechanism . In this study , we developed a computational framework ( DSPathNet ) to construct a signaling pathway network for a given drug , specifically , metformin . The framework first collected metformin upstream genes from different data sources and inferred chemical signaling receptor TFs based on metformin-induced gene expression data . Then , a metformin-specific SPNetwork was produced using the random walk-based algorithms by applying longitudinal and lateral movements starting from metformin upstream genes and downstream TFs . By examining the enrichment of disease genes in the network , the metformin-specific SPNetwork proved to be enriched with genes that could contribute to the pathology of T2D and cancer , or reducing cancer risk in T2D patients undergoing metformin treatment . Starting from the genes common to T2D and cancer GWAS data , we further produced a crosstalk subnetwork of metformin action in T2D and cancer . Through comprehensive network and functional analyses and literature mining , we identified seven critical genes ( CDKN1A , ESR1 , MAX , MYC , PPARGC1A , STK11 , and SP1 ) , some of which have been implicated in previous studies . Furthermore , the MYC and its motifs were suggested to play important roles in metformin action . In summary , this study has the following major results: 1 ) we developed a computational framework for building drug-specific signaling pathway networks; 2 ) we generated a metformin-specific signaling pathway network that is significantly enriched with genes associated with T2D , cancer , or metformin-associated cancer survival , and 3 ) we pinpointed the MYC-centered pathway that may play important roles in metformin action . These results demonstrate that the computational framework effectively integrates various types of data , such as prior drug knowledge and drug-induced gene expression to identify critical genetic factors responsible for drug indications and drug response . This framework is a novel approach that provided a broader and deeper understanding of metformin actions in both T2D and cancer . This computational approach can be applied to other drugs as well . This framework applies a new network generation strategy that focused on a drug of interest . In our framework , we utilized the gene expression data to infer the drug related gene expression regulators TFs , which is different from the methods that have been developed to infer signaling pathway networks directly from gene expression data [91] . As we know , the gene expression represents the transcriptional changes in the downstream genes of a pathway and provides an indirect view of pathway structure and gene activity after modulation of the system . Thus , the gene expression cannot directly represent the activity state of many signaling components that mediated the cellular response [92] . It is well known that the signal transduction network is not linear; rather it is quite complex [3] . During the development of this framework , we observed only two genes overlapped between metformin upstream genes and downstream genes . This small overlap presents us with a big challenge: how to fill the gap to rebuild a complete cascade for drug action ? To tackle this challenge , we proposed a novel strategy from background human SPNetwork through both longitudinal and lateral movements . For the longitudinal movement , we employed the software NetWalker that implemented the random walk with a starting probability . For the lateral movement , we took advantage of K-Walk algorithm that simulates random walks in the network using a Markov Chain to build the most relevant subnetwork . In this study , we combined them together to achieve our goal . Table 2 summarizes the number of genes in each step and the hypergeometric tests based on the number of genes with smallest P-value less than 0 . 05 in the corresponding network compared to all genotyping data in T2D GWAS data . The evaluation results indicated that the process is promising since it recruited more informative genes; the significance of the association between the network and disease-related mutation signals became stronger . However , the major concern regarding the framework is to rebuild a complete and reliable human SPNetwork and to control false positives from both public data and prediction results caused by the computational tools . To balance these two factors , we rigorously compiled the information involved in the signaling pathways , extensively collected the drug related data from multiple data sources , applied rigorous parameters during the use of computational approaches , and performed comprehensive evaluations for metformin-specific SPNetwork . To increase the accuracy of results , we only included the protein-protein pairs with experimental evidence and excluded the pairs only involved in the protein complexes . Thus , the coverage of the human SPNetwork was lower than a typical protein-protein interaction network; it contained only 37 , 881 edges and 4 , 367 proteins . With the rapid development of human experimental technologies , we believe more data with higher coverage and accuracy will become available , which will enable the construction of a more comprehensive signaling pathway network with high quality . To collect as many metformin-related genes as possible , in addition to the public databases DrugBank and PharmGBK , we further performed literature mining from PubMed abstracts , which provided an additional 19 genes . To ensure the accuracy of TF inference , we only utilized gene expression data from the four treatments of metformin that showed significant consistency with each other . To comprehensively evaluate if the metformin-specific SPNetwork was enriched with mutation signals of T2D and cancer , we not only took advantage of the well-studied disease genes but also individual genotyping data from GWAS data sets . Thus , our framework has the ability to recruit more key components in the drug signal transduction process . It could be potentially applied to other drugs for the purpose of deciphering their signaling pathway networks and identifying critical genes . Another limitation of this framework is the absence of a control network representing the normal state . The signaling network at the normal state may provide additional insights into drug action . However , it is very difficult and challenging to construct a normal-state signaling transduction network for drug action . Though some pathway data sources such as KEGG provide the relevant signaling networks in the normal state , most of them only provide a limited view by focusing on one or two related pathways . Compared to these individual pathway networks , the metformin-specific SPNetwork provides a comprehensive view by including many well-known metformin-related pathways , T2D-related pathways , and cancer-related pathways ( Results ) . This computational framework is strongly dependent on the available literature about the investigated drugs . Thus , it is not suitable for these drugs or chemicals that do not have many basic research reports . However , it is known that , during the drug development , most of them cannot be approved by FDA even after entering the clinical trials [93] . Furthermore , as the time and costs for developing novel drugs dramatically increased recently , many drug developers prefer to find new uses for existing drugs including the approved and non-approved drugs . As more large-scale data become publicly available , researchers could utilize the framework to build a SPNetwork for each drug of interest , and then examine the relationship between the network and disease genes , or calculate network similarities with the known drugs for a certain indication . These relationship or network similarities may provide more clues for drug repurposing at the network level . Therefore , the framework will be promising for identification of drugs that may be used to treat secondary indications by constructing and comparing the drug-specific SPNetworks . Moreover , since the drug-specific SPNetwork contains comprehensive information regarding the drug action of the components , we speculated that some off-targets might be included in the network . Thus , our network approach can be extended to evaluate the association between drugs and their potential side effects . However , it is challenging to identify large-scale side effect data associated with genes or their proteins . So far , several studies have used the available biochemical data to determine candidate targets for specific side effects [94–96] . Such data is limited and likely with a high false positive rate . When more relevant data becomes available in future , our approach will be applied to assess drugs’ side effects . An important output of this study is the metformin-specific SPNetwork consisted of metformin related genes , metformin related TFs , and many novel genes . The network provides a valuable gene pool for further investigation of metformin action . Metformin has been used to treat diabetic disorders for many years because of its ability to lower glucose levels and improve insulin sensitivity [97] . Recently , several findings from epidemiological studies have shown that metformin can reduce cancer risk and improve cancer survival in the T2D patients [60 , 98 , 99] , including a recent electronic health record ( EHR ) study we participated in that demonstrated the effect was seen for many cancer types [26] . However , the molecular mechanisms underlying metformin action are complex and remain unclear , especially for its ability of decreased cancer risk [100 , 101] . In this study , we first constructed a complex metformin-specific SPNetwork and then produced a crosstalk subnetwork from the metformin-specific SPNetwork . This subnetwork contained three modules highlighting different pathways ( Fig 6C ) . The first and second modules were enriched with genes from the insulin signaling pathway and adipocytokine signaling pathway , and the third module was enriched with genes involved in cancer related pathways . The adipocytokine signaling pathway contains the major components of AMPK signaling pathway according to KEGG annotation . Through seven nodes , the first and second modules were linked to the third module . These observations suggest that the metformin possibly affects the AMPK signaling pathway and the insulin signaling pathway directly , which subsequently decrease the chance of cancer development . This outlook is consistent with a previous review [102] . The seven nodes act as bridges linking the first and second modules to the third module . We predicted they might play critical roles in the metformin signaling transduction process ( Fig 6C ) . Among them , two genes ( PPARGC1A and STK11 ) belonged to metformin upstream genes; one ( ESR1 ) to metformin downstream genes; four genes ( CDKN1A , MAX , MYC , and SP1 ) were both hubs and bridge nodes . It is well known that gene STK11 , also known as LKB1 , encodes a member of the serine/threonine kinase family that regulates cell polarity and functions as a tumor suppressor [103] . Additionally , previous studies have shown that mutations in the STK11 gene influence insulin sensitivity and metformin efficacy [104 , 105] . The MYC gene encodes a protein that plays a role in cell cycle progression , apoptosis , and cellular transformation [106] . It has been shown that MYC gene plays important roles in the anticancer metabolic effects of metformin [75 , 76] . The PPARGC1A gene encodes a transcriptional coactivator that regulates the genes involved in energy metabolism . Its variant rs2970852 has been reported to modify the effects of metformin on triacylglycerol levels [107] . Recent studies have shown that gene regulation induced by metformin involves the transcription factor SP1 in cancers [61 , 108] . Moreover , the expression of CDKN1A ( also known as P21 ) is upregulated in hepatocellular carcinoma [88] and bladder cancer cells [89] after metformin treatment . The evidence from these studies suggests that our approach is effective for identifying the key components in the signaling pathway . To further investigate detailed information for these genes , more experimental validations are needed . To our knowledge , there is no any positive evidence for the association of the genes ESR1 and MAX of the seven critical genes with metformin action . Thus , they are two novel genes for further experimental validation . In addition to the DSPathNet framework to effectively recruit critical components in the mode of drug action , there are other ways to expand this approach . First , integrating multiple layers of data involving the signal cascade beyond gene expression data into a comprehensive method might improve our ability to identify the association between the genetic changes and their response to drugs . Second , although we have shown the utility of two sources for compiling the human SPNetwork , there are other data worth exploring such as those involved in the metabolism , protein phosphorylation , and protein kinase and phosphatase interactions . While this study focused on one medication , metformin , the computational framework is broadly applicable to any drug for which induced gene expression data is available . Moreover , several experimental data sources are available for further data integration and mining such as the Connectivity Map project [15] , Genomics of Drug Sensitivity in Cancer [109] , Cancer Cell Line Encyclopedia ( CCLE ) [110] , and anticancer compounds in breast cancer [111] . Finally , analyzing the crosstalk among different types of diseases in the context of networks will offer an intriguing opportunity to explore the underlying molecular mechanisms of drug action , which will provide an alternative approach for drug repurposing . Before generating the metformin-specific SPNetwork , we need a global signal transduction network for humans as the background network . We therefore integrated signaling transduction related associations from Pathway Commons with experimental evidence [25] , and TF-TF/target pairs from TRANSFAC [26] . The Pathway Commons database collected publicly available pathways from multiple organisms with over 1 , 400 pathways and 687 , 000 interactions . We first downloaded the edge data specific for humans from the Pathway Commons ( release 2011 . 10 ) . Since the interactions that occur within the protein complexes do not reveal the flow of signaling information [3] , we excluded the edges that came from the same complex . This process resulted in 33 , 614 pairs among 3 , 502 proteins . Additionally , we obtained 1 , 325 pairs among 487 TFs , and 2 , 723 pairs between 428 TFs and 1 , 315 targets downloaded from TRANSFAC database ( release 2011 . 4 ) . The TRANSFAC database manually collects eukaryotic TFs , their genomic binding sites , and DNA binding profiles with experimental evidence [112] . After merging the two data sets and removing the redundancies , we obtained a network with 37 , 881 edges and 4 , 367 nodes . This network was used to represent global signaling pathways in humans . To further weight the association of each node in human SPNetwork with metformin action , we assigned a functional similarity score by calculating its functional similarity to the metformin upstream genes using the R package GoSemSim based on GO annotations [113] . GO annotations have three functional domains ( k ) : molecular function ( MF ) , biological process ( BP ) , and cellular component ( CC ) . First , for a given node i in each domain ( k ) , we calculated its score as Scorei = ∑j = 1nScorei , j/n , where n is the number of existing scores between node i and metformin upstream gene j . Second , for the given node i in all domains , we calculated a final score as S^ = ∑k = 1NScorek/N , where N is the number of the domains having scores for the node . Gene expression profiles of cancer cells following drug treatment are useful for better understanding cellular changes reflective of drug treatment [114] . In this study , we integrated the known TF-target association and drug-induced gene expression data to infer the metformin downstreams . We first comprehensively collected the TF-target associations , then calculated the up- or down-regulated genes from drug-induced gene expression data , and finally performed the hypergeometric test to evaluate the over-representation of the up- or down-regulated genes in multiple TF target gene datasets . To compile a target gene set for each TF comprehensively , we downloaded data from two sources: TRANSFAC Professional ( release 2011 . 4 ) and MSigDB database [38] . From the TRANSFAC database , we extracted known TFs and their targets in human . From the MSigDB , we downloaded the gene sets that share one TF binding site . The gene sets were derived from a comparative analysis of human , mouse , rat , and dog genomes and were organized by TF binding motifs . Genes associated with different binding motifs that correspond to a common transcription factor were combined into one gene set . After merging the two data sets , we obtained 666 human TFs and 8 , 502 human targets . To calculate the potential differentially expressed genes induced by metformin , we downloaded ten gene expression datasets from Connectivity Map website ( version 2 . 0 ) . The gene expression datasets were generated from metformin treated cell lines . We calculated the ranked probes by using the method described in Lamb et al . [15] and selected the top 100 and bottom 100 probes in each treatment to represent the differentially expressed probes [115] . We examined the expression consistency among them using the software GSEA . We noticed that , among ten metformin treatment data sets , four had the highest consistency based on GSEA analysis [38] . Therefore , we utilized these four treatment gene expression data to perform the GSEA leading edge analysis to detect differentially expressed probes . Then , by mapping the differently expressed probes to genes using Ingenuity Pathway Analysis Tool ( http://www . ingenuity . com/ ) , we obtained the up-regulated genes and down-regulated genes . Finally , we performed the hypergeometric test to evaluate the over-representation of the up- or down-regulated genes in the different TF gene sets . The TFs with P-value less than 0 . 05 were identified as significant TFs related to metformin action and their genes as metformin downstream genes . Considering that the signal transduction cascade is not linear , we adopted a two-step strategy to construct the metformin-specific SPNetwork from the human SPNetwork . More specifically , in the first step , we utilized the software NetWalker to expand metformin upstream genes and downstream genes for longitudinal conduction [116] . The NetWalker implements the random walk with a starting probability . In this study , we gave equal starting probability of 0 . 5 to each gene in the metformin upstream genes and downstream genes and required those nodes with both local P-value < 0 . 05 and global P-value < 0 . 05 as the expanded genes . In the second step , we expanded the nodes from in the first step by lateral movement by applying the K-Walk method implemented in the Python package GenRev [117] . The K-Walk algorithm simulates random walks in the network using a Markov Chain to build the most relevant subnetwork , connecting seed nodes by walk a fixed length L or up to a maximal length Lmax in a large network . A subnetwork is obtained by keeping only edges that are above a minimal relevance threshold . The threshold is automatically fixed after the subnetwork has the maximum score . As such , the limited K-Walk algorithm computes edge and node relevance from random walks connecting the seed nodes [118] . We used one T2D GWAS data set , three cancer GWAS data sets , and one GWAS data set for T2D patients with metformin treatment . The T2D GWAS data was individual-level genotype data generated from the WTCCC [31] . The three cancer GWAS datasets were generated by the Cancer Genetic Markers of Susceptibility ( CGEMS ) project: breast cancer [32] , pancreatic cancer [33] , and prostate cancer [32] . We downloaded the genotype data from the National Center for Biotechnology Information ( NCBI ) dbGaP with approved access for the CGEMS project . For these four GWAS datasets , we first removed individuals with genotyping rate < 95% and SNPs with missing rate >5% . A single SNP associated test was conducted using the Armitage trend test for SNPs with a minor allele frequency ( MAF ) > 0 . 05 . S10 Table summarizes the data . T2D cancer patients from Vanderbilt University Medical Center ( VUMC ) were identified using the Synthetic Derivative ( SD ) , a de-identified copy of the electronic health records from VUMC . Eligible subjects were individuals who 1 ) had a cancer diagnosis ( excluding non-melanoma skin cancers ) between January 1 , 1995 and December 31 , 2010 identified through the Vanderbilt tumor registry , and 2 ) were older than 18 years at the time of cancer diagnosis . Using a previously developed algorithm [119 , 120] , we identified T2D subjects having at least two pieces of clinical information in their medical record: 1 ) ICD9 code for type 2 diabetes , 2 ) medications for type 2 diabetes , or 3 ) clinical labs suggestive of T2D ( random glucose >200 mg/dl or hemoglobin A1c > 6 . 5% ) . Individuals without at least two of the above types of information were excluded . At least two mentions of metformin use ( mono-therapy or combined therapeutic ) and one mention of metformin use within 5 years after cancer diagnosis were required for study inclusion . Individuals on other T2D medications were excluded from analysis . Subjects were followed for overall mortality that was determined through linkage with the Vanderbilt tumor registry . Physician-reported European descent individuals with an available DNA sample in the Vanderbilt biobank ( BioVU ) [121] were genotyped on either the Illumina HumanOmni1-Quad or the Illumina HumanOmni5-Quad . Only the consensus single nucleotide polymorphisms ( SNPs ) between the two genotyping platforms were used . Standard quality control ( QC ) procedures were applied to remove individuals and autosomal SNPs not meeting standard QC criteria ( i . e . related individuals , discordant sex , sample efficiency < 98% , genotyping efficiency < 98% , deviations from Hardy-Weinberg equilibrium ( p < 1×10–6 ) , and MAF < 5% ) . Palindromic SNPs were also removed . After QC , 461 individuals and 551 , 745 SNPs remained . Principal components were estimated using EIGENSTRAT [122] . The association between each SNP , assuming an additive genetic model , and overall survival was examined using Cox proportional hazards models , adjusted for age , sex and one principal component , using the GenABLE package of R [123] . The GWAS analysis of this set is ongoing and will be reported in a separate publication . In this study , we defined the genes having at least one SNP with nominal P-value less than 0 . 05 as disease or drug related genes . The SNP is located in the gene’s region or its 20kb up- or down-stream sequence based on the gene annotation and human reference genome build 36 for T2D GWAS study and cancer GWAS studies and build 37 for metformin GWAS study . To identify pathways overrepresented in gene sets , we performed KEGG pathway enrichment analyses using WebGestalt [49] ( version 1/30/2013 ) . Given a list of genes , a hypergeometric test was performed for the enrichment of these genes , which was implemented in the WebGestalt tool . To control the error rate in the analysis results , WebGestalt also provides a corrected P-value based on the Benjamini-Hochberg method [124] . To summarize the enriched pathways , we took advantage of KEGG pathway category annotation , which included the two-level categories and represent the relative abundance of the pathways [125] . These pathways are grouped into seven categories at the first level of KEGG annotation and 43 categories at the second level of KEGG annotation . At the second-level category , we further calculated a Z-score for each category to represent the KEGG pathway relative abundance: Z-score = x-uσ , where x is the number of pathways in one category in the first or second level , u is the mean of the pathway number in the first or second category , σ is the standard deviation of the pathway number in the first or second category . The pathway categories were selected for further analysis if their Z-scores were higher than zero . In this study , we adopted the statistical design for gene set enrichment analysis [126] to compare a gene set ( A ) in the drug-specific network to a reference gene set ( B ) . The design has been commonly used to conduct the gene annotation enrichment analysis [127] . Suppose that the gene set ( A ) has n genes , of which most genes ( n’ ) belong to the reference gene set ( m ) . Among n’ gene , k genes belong to a given category ( C ) . And the reference gene set has j genes belong to the same category ( C ) . Based on the definition of the hypergeometric test , we performed the hypergeometric test to get a P-value to evaluate the significance of enrichment for category C in the gene set A . For network property analysis , we calculated degree of each node and degree distribution of all nodes , which are the most basic measures of biological networks [41] . The node degree ( connectivity ) is the number of links of a node in the network . If degree distribution of one network follows a power law , the network would have only a small portion of nodes with a large number of links ( i . e . , hubs ) [41] . To determine the hubs in metformin-specific SPNetwork , we adopted the method utilized by Yu et al . [46] , as we did in a previous study . We first drew a degree distribution for the whole network to define a specific degree value as a cut-off point ( S12 Fig ) . If a node has the degree greater than the cut-off value , then the node is a hub . To identify the modules , we performed the cluster and community analysis using the software CFinder ( version 2 . 0 . 5 ) [63] . CFinder is a fast program to locate and visualize overlapping , densely interconnected groups of nodes in undirected network . We required each node in the module being involved in at least one 3-vertex clique . We visualized the networks using Cytoscape ( version 3 . 2 ) [128] .	A deep understanding of a drug’s mechanisms of actions is essential not only in the discovery of new treatments but also in minimizing adverse effects . Here , we develop a computational framework , the Drug-specific Signaling Pathway Network ( DSPathNet ) , to reconstruct a comprehensive signaling pathway network ( SPNetwork ) impacted by a particular drug . To illustrate this computational approach , we used metformin , an anti-diabetic drug , as an example . Starting from collecting the metformin-related upstream genes and inferring the metformin-related downstream genes , we built one metformin-specific SPNetwork via random walk based algorithms . Our evaluation of the metformin-specific SPNetwork by using disease genes and genotyping data from genome-wide association studies showed that our DSPathNet approach was efficient to synopsize drug’s key components and their relationship involved in the type 2 diabetes and cancer , even the metformin anticancer activity . This work presents a novel computational framework for constructing individual drug-specific signal transduction networks . Furthermore , its successful application to the drug metformin provides some valuable insights into the mode of metformin action , which will facilitate our understanding of the molecular mechanisms underlying drug treatments , disease pathogenesis , and identification of novel drug targets and repurposed drugs .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	Deciphering Signaling Pathway Networks to Understand the Molecular Mechanisms of Metformin Action
Enterovirus 71 ( EV71 ) causes hand , foot and mouth disease , a mild and self-limited illness that is sometimes associated with severe neurological complications . EV71 neurotropic determinants remain ill-defined to date . We previously identified a mutation in the VP1 capsid protein ( L97R ) that was acquired over the course of a disseminated infection in an immunocompromised host . The mutation was absent in the respiratory tract but was present in the gut ( as a mixed population ) and in blood and cerebrospinal fluid ( as a dominant species ) . In this study , we demonstrated that this mutation does not alter the dependence of EV71 on the human scavenger receptor class B2 ( SCARB2 ) , while it enables the virus to bind to the heparan sulfate ( HS ) attachment receptor and modifies viral tropism in cell lines and in respiratory , intestinal and neural tissues . Variants with VP197L or VP197R were able to replicate to high levels in intestinal and neural tissues and , to a lesser extent , in respiratory tissues , but their preferred entry site ( from the luminal or basal tissue side ) differed in respiratory and intestinal tissues and correlated with HS expression levels . These data account for the viral populations sequenced from the patient’s respiratory and intestinal samples and suggest that improved dissemination , resulting from an acquired ability to bind HS , rather than specific neurotropism determinants , enabled the virus to reach and infect the central nervous system . Finally , we showed that iota-carrageenan , a highly sulfated polysaccharide , efficiently blocks the replication of HS-dependent variants in cells and 2D neural cultures . Overall , the results of this study emphasize the importance of HS binding in EV71 pathogenesis and open new avenues for the development of antiviral molecules that may prevent this virus’s dissemination . Enterovirus 71 ( EV71 ) is one of a limited number of enterovirus genotypes that have the ability to infect the central nervous system ( CNS ) [1–3] and has emerged as a major health-care threat across the Asia-Pacific region [4–7] . Although this virus is typically associated with mild hand , foot and mouth disease ( HFMD ) epidemics , EV71 has been increasingly associated with neurological disorders that range from aseptic meningitis with or without pulmonary edema to brain stem encephalitis and poliomyelitis-like acute flaccid paralysis , particularly among children less than 6 years old [8–10] . EV71 transmission depends on hygiene , water quality , and the extent of crowding [11] . EV71 is typically transmitted through fecal-oral routes , although transmission via respiratory secretions also frequently occurs , particularly in countries with high standard sanitation [12 , 13] . Following infection , EV71 predominantly replicates in the intestinal mucosa and , to a lesser extent , in the respiratory mucosa [5 , 14] . Subsequently , the virus can enter the bloodstream and disseminate to a variety of organs , including the central nervous system ( CNS ) [15] . EV71 dissemination to the CNS remains a rare event and the viral neurotropic determinants remain ill-defined , despite extensive epidemiological studies and experimentation in animal models . EV71 tissue tropism may be modulated by its ability to bind to a variety of receptors [16] . The human scavenger receptor class B2 ( SCARB2 ) , a major transmembrane lysosomal protein , is ubiquitously expressed and is the best studied receptor [17] . In the human CNS , SCARB2 is expressed on neurons and glial cells [18] , and a transgenic mouse model expressing human SCARB2 has exhibited susceptibility to EV71 infection and the development of ataxia , paralysis and death [19] . The second EV71 entry receptor , P-selectin glycoprotein ligand-1 ( PSGL-1 ) , is expressed exclusively on leukocytes and is only bound by certain EV71 strains [20 , 21] . Finally , attachment molecules , such as heparan sulfate glycosaminoglycan ( HS ) [22] , sialic acids [23] , nucleolin [24] , vimentin [25] and annexin II [26] also enhance EV71 infectivity and may contribute to viral dissemination and neurotropism . Residues that are crucial for the binding of EV71 to certain of these receptors have been mapped on the VP1 viral capsid protein . For instance , 172Q and adjacent amino acids in the EF loop are required for binding to human SCARB2 and for efficient infection of cells expressing this receptor [27] while K98E , E145A and L169F substitutions confer binding ability to murine SCARB2 [28] . Similarly , EV71 strains with a glycine or a glutamine at position 145 of VP1 are able to bind PSGL-1 , while strains with a glutamate at this position are not [21] . Binding of EV71 to annexin II has been mapped to the BC loop region of VP1 ( residues 40–100 ) [26] . Finally , positively charged residues at the 5-fold axis of EV71 capsids ( VP1 162K , 242K , and 244K ) appear to be essential for attachment to HS . When these residues are mutated , compensatory mutations ( E98A , T100K and Q145R-T237N ) can restore binding . Similarly , strains with 98E and 145E are poor HS binders , while E98K restores binding [29] . These observations are based on in vitro adapted EV71 isolates and do not address the in vivo significance of dissemination and neurotropism . In addition , the implication of immune escape mutations may add a level of complexity . VP1 residues 98 and 145 have been shown to be under positive selection and may be targeted by neutralizing antibodies [30 , 31] . Finally , EV71 virulence determinants that have been identified in animal models contradict epidemiological studies in humans . VP1 residue 145E was shown to be a virulence factor for both cynomolgus monkeys [31 , 32] and mice [33 , 34] , while isolates with 145G/Q are also frequently associated with severe neurological disease in humans [35–38] . Mutations in other genomic regions may explain these apparent contradictions . We previously isolated EV71 from an immunocompromised patient with disseminated disease . Genomic analysis revealed that this clinical strain clusters with EV71 in the human EV-A species and is related to the subgenogroup C1 . Comparison of five full-length genomes sequenced directly from respiratory , gastrointestinal , nervous system and blood specimens highlighted a critical non-synonymous single amino acid change within the EV71 VP1 BC loop ( L97R ) that could potentially lead to dissemination in natural infections . EV71-VP197R , which was absent in the respiratory sample , was present as a dominant population in blood and cerebrospinal fluid samples and as a mixed population ( EV71-VP197R/L ) in the stool sample [39] . However , the mechanisms by which the mutant virus could disseminate are not clear . In this study , we investigated the capacity of EV71-VP197L and EV71-VP197R to bind and mediate infection in different cell lines and in 3D reconstituted respiratory , intestinal and neural tissues . We demonstrated that both derivatives are dependent on SCARB2 for effective infection , while their specific tropism is linked to a different ability to bind HS and further correlates with the expression level of this attachment receptor in the tested tissues . This study provides new insight regarding HS binding as a critical determinant of EV71 dissemination in humans and opens the door to antiviral strategies aimed at preventing EV71 in-host adaptation and dissemination . We previously showed that the VP1 L97R mutation conferred a binding advantage in RD and Vero cells . In addition , we showed in vitro that this mutation was associated with a second non-conservative mutation ( E167G/A ) in the VP1 EF loop [39] . To define the role of these mutations more precisely , we generated 4 infectious clones with distinct combinations of each of the mutations ( Table 1 ) . Viral stocks were prepared in monkey kidney cells ( Vero ) and titrated in Vero , colorectal carcinoma ( Caco-2 ) , rhabdomyosarcoma ( RD ) and neuroblastoma cells ( SH-SY5Y ) . After 4 days of infection , viruses were sequenced from each cell line . EV71-VP197L167G and EV71-VP197R167E viral stocks were unstable independent on the cell line ( Table 1 ) and thus could not be amplified and purified in sufficient quantities to be used in functional assays . EV71-VP197L167E ( originally isolated from the lower respiratory tract sample of the patient with a disseminated infection ) and EV71-VP197R167G ( in vitro adapted from EV71-VP197R167E , which was originally isolated from the stool , cerebrospinal fluid and blood of the patient ) were thus used for all further investigations . Titration showed that EV71-VP197L167E and EV71-VP197R167G exhibited differential cell tropism , with significantly enhanced replication being observed for EV71-VP197R167G in RD and SH-SY5Y cells ( Fig 1 ) . Notably , the RNA load measured in the stocks prepared in Vero cells was comparable for the two viruses ( 12 . 38 log RNA copies/ml for EV71-VP197R167G and 12 . 30 log RNA copies/ml for EV71-VP197L167E ) . As SCARB2 plays a critical role for EV71 infection [17] , we investigated whether the differential cell tropism of the two variants is linked to a different dependence on this receptor . We used the CRISPR/Cas9 system to knockout SCARB2 in Caco-2 and RD cells , two cell lines presenting different susceptibilities to the two variants ( Fig 1 ) . A striking reduction in viral titers was observed in SCARB2 knockout RD and Caco-2 cells for EV71-VP197L167E and EV71-VP197R167G ( Fig 2 ) suggesting that both variants depend on this receptor for effective infection regardless of the cell line . This was also shown by transfection of human SCARB2 in mouse L929 cells that are non-permissive for infection by EV71 [17] . Exogenous expression of human SCARB2 conferred susceptibility to both EV71-VP197L167E and EV71-VP197R167G ( S1 Fig ) confirming that the two viruses are dependent on SCARB2 . We then investigated the implication of another ubiquitously expressed receptor , the HS attachment receptor , in this observed differential cell tropism . We measured the amount of EV71-VP197R167G and EV71-VP197L167E viruses bound to untreated cells or cells pretreated with heparinase III ( Fig 3A ) . EV71-VP197R167G exhibited a binding advantage over EV71-VP197L167E in RD , Vero and SH-SY5Y cells but not in Caco-2 cells . This advantage was suppressed after cleavage of HS from the cell surface ( Fig 3A , left panel ) . In contrast , EV71-VP197L167E did not show a significant difference in any cell line ( Fig 3A , right panel ) . To confirm this observation , competition experiments were performed with a viral stock ( EV71-VP197R/L-167G/E ) containing equivalent amounts of each derivative . The competition was run under 3 conditions: using untreated cells , using cells having surface HS digested with heparinase , or using a viral population that was pretreated with heparin , a soluble HS analogue . Both binding ( 1 hour post-infection [hpi] ) and replication ( 24 hpi ) efficiencies were monitored by real time-quantitative polymerase chain reaction ( RT-qPCR ) , and the dominant species was characterized by Sanger sequencing for each binding ( Fig 3B ) and replication condition ( Fig 3C ) . The overall binding was lower for the mixed population than for the EV71-VP197R167G stock ( Fig 3B versus 3A ) . Accordingly , the pretreatment of cells with heparinase reduced viral binding less significantly , while viral replication was affected in SH-SY5Y , Vero and Caco-2 cells . Sanger sequencing of the entire VP1 region after binding showed that in Vero , SH-SY5Y and RD cells , EV71-VP197R167G ( with R at position 97 encoded by CGC and G at position 167 encoded by GGG ) was the dominant variant in non-treated conditions , which was not the case in Caco-2 cells , where EV71-VP197L167E ( with L at position 97 encoded by CTC and E at position 167 encoded by GAG ) and EV71-VP197L167G were dominant ( Fig 3C ) . Upon heparinase treatment , the dominant consensus sequence shifted towards EV71-VP197R/L167G/E , EV71-VP197L167G/E or EV71-VP197L167E in binding and replication assays . Notably , viral species with a positive charge at VP1 position 98 ( K encoded by GAA instead of E encoded by AAA ) emerged after treatment of SH-SY5Y , RD and Caco-2 cells with heparinase ( Fig 3C and 3D ) . Pre-incubation of viruses with heparin also resulted , although in a less stringent manner , in a shift of the dominant population from EV71-VP197R167G to a mixed EV71-VP197R/L167G/E population , except after replication in SH-SY5Y and RD cells . To determine if the improved binding and replication of EV71-VP197R167G in cells correlated with HS expression levels , quantification of HS expression was assessed by immunofluorescence ( Fig 3D and 3E ) . SH-SY5Y , RD and Vero cell lines strongly express HS , while Caco-2 showed low HS expression . A recent publication showed that variants with VP1 E167G do not bind heparin [29] , and the differential binding and replication of EV71-VP197R167G confirmed our previous observations with EV71-VP197R167E in SH-SY5Y and Vero cells [39] , indicating that the acquired HS binding ability is due only to the L97R substitution . Finally , we confirmed the different affinity of EV71-VP197R167G and EV71-VP197L167E for HS using a heparin sepharose binding assay . After incubation of a mixed viral stock ( EV71-VP197R/L-167G/E ) composed of 50% of each derivative with sepharose beads , we sequenced the viral populations present in the non-heparin-binding population ( flow-through ) or in the heparin-binding population ( eluate ) ( Fig 3E ) . As expected , EV71-VP197L167E was highlighted in the non-heparin-binding population , while EV71-VP197R167G was detected in the heparin-binding population . To better replicate the viral tropism observed in the patient , infection assays were performed with EV71-VP197R167G , EV71-VP197L167E and EV71-VP197R/L167G/E in complex tissue models , in vitro reconstituted from human primary cells , that mimic the upper respiratory tract , small intestine and neural tissues composed of neurons and glial cells . Because of the acquired ability of EV71-VP197R167G to bind HS , its sensitivity to iota-carrageenan ( ı-carrageenan ) , a highly sulfated polysaccharide , was investigated and compared to the sensitivity of EV71-VP197L167E in Vero and 2D neural cultures . As expected , EV71-VP197R167G infection was strongly inhibited by ı-carrageenan in Vero cells ( IC50 = 17 . 8 μg/ml ) , whereas EV71-VP197L167E infection was only slightly affected by the treatment ( IC50>200 μg/ml ) ( Fig 7A ) . Similar results were obtained in neural cultures , where the infection of EV71-VP197R167G and EV71-VP197R/L167G/E was inhibited up to 90% in a dose-dependent manner . EV71-VP197L167E infectivity was also inhibited but to a lesser extent ( up to 50% ) . Treatment of the mixed viral population altered the dominant sequence from EV71-VP197R167G at 1 dpi towards EV71-VP197R/L167G/E ( Fig 7B ) . In 2012 , we reported an EV71 genogroup C1 disseminated infection in an immunocompromised patient treated with rituximab who was hospitalized with respiratory and neurological symptoms [39] . Although EV71 genogroup C1 infections have not been associated with severe diseases in the past , several complicated cases have been recently reported in Europe , underscoring the importance of defining EV71 virulence determinants [43 , 44] . By sequencing the full genome of the virus from different samples , we identified the acquisition of a non-conservative mutation in VP1 ( L97R ) , which was absent at the beginning of the infection in the respiratory tract but was present in a mixed population in the gastrointestinal tract and became dominant in the blood and CSF at later time points . We showed that this mutation conferred a binding advantage to the virus in certain cell lines . In this study , we extended these findings and showed that the VP1 L97R substitution , as well as an associated mutation ( VP1 E167G ) observed after in vitro culturing , confers an ability of the virus to use HS as an attachment receptor while it does not alter the dependence of the virus on the ubiquitously expressed SCARB2 entry receptor . Tan and colleagues recently showed that variants with VP1 E167G do not bind heparin [29] . In addition , in our previous publication , we observed the same binding advantage of the VP197R mutant in neuroblastoma and Vero cells that was observed in this study with the VP197R167G mutant [39] . This strongly suggests that the VP1 L97R mutation confers an HS binding ability , while the VP1 E167G mutation has a stabilizing function as proposed previously based on VP1 3D structure [39] . Interestingly , after competition between EV71-VP197L167E and EV71-VP197R167G for binding and replication in cells pretreated with heparinase , EV71-VP197R was excluded from the viral population , while a new variant ( EV71-VP198K ) which is also known to promote HS binding [29] , emerged . Thus , it is very likely that after cleavage with heparinase at the 1→4 linkages between hexosamine and glucuronic acid residues , HS is still bound by EV71-VP198K but not by EV71-VP197R , suggesting that these two variants present different HS binding specificities . To assess EV71 replication in vivo in the gastrointestinal and respiratory tracts and in the CNS , we analyzed the replication of HS-dependent ( EV71-VP197R167G ) and independent ( EV71-VP197L167E ) variants in relevant tissue culture models that mimicked the upper and lower respiratory airway epithelia , intestinal and neural tissues . To reproduce natural infection routes , we inoculated intestinal and respiratory tissue cultures that grow at an air-liquid interface from the apical ( corresponding to the lumen ) or basal ( corresponding to the blood-borne ) tissue sides . A comparison of viral growth of HS-dependent and HS-independent variants and competition experiments highlighted that HS-dependent variants had improved access to respiratory and intestinal tissues from the basal tissue side , where HS is highly expressed , while HS-independent variants show improved replication after infection from the air-exposed side of respiratory tissues , where HS is only slightly present , in line with previous publications [45 , 46] . Infection from the luminal side of intestinal tissues sustained efficient replication of both variants and correlated with intermediate expression of HS at this side . We also highlighted high expression of SCARB2 at the apical side of respiratory and intestinal tissues and low expression at the basal side . HS may thus be necessary in tissues with low SCARB2 expression to concentrate the virus at the cell surface and promote subsequent interaction with SCARB2 . This would explain why HS-dependent variants infect the basal side of respiratory and intestinal tissues better than HS-independent variants . Our findings in respiratory and intestinal tissues correlate with the viral populations isolated from the different specimens of the patient . We only observed an HS-independent variant in the bronchoalveolar lavage of the patient , a sample collected from the luminal side of the respiratory tissue . Importantly , sequencing of EV71 that was detected in bronchoalveolar lavages from two other immunosuppressed patients who presented neurological complications in Switzerland also identified viruses from the same genogroup and with the same VP1 sequence ( GenBank accession numbers MH256663 and MH256664 ) . Of note in these 2 cases as in our case , the infecting virus had a glutamate at position 145 , a residue known to confer poor HS binding ability [29] . This suggests that HS-independent variants may be preferentially transmitted via the respiratory route . We observed a mixed viral population in the stool of the patient , a sample that reflects viral population isolated from the luminal side of the gastrointestinal tract . A mixed viral population was also obtained after apical inoculation of intestinal tissue cultures with a viral population composed of EV71-VP197L167E and EV71-VP197R167G . Since both ex vivo and in vivo intestinal tissues are the favored site of EV71 replication , and since these tissues express HS at the basal and apical surface , it is very likely that the emergence of an HS-dependent virus arose at this site . Whether this site was reached after replication in the respiratory tract and propagation via shedding in oral secretions and swallowing or whether infection occurred simultaneously in the two sites remains an open question . Concerning neurotropism , the viral population sequenced from the CSF of the patient was HS-dependent while results obtained in 2D and 3D neural cultures were more intricate . In single infections , HS-dependent and HS-independent variants revealed similar neurotropic potential as both showed high replication levels in primary neural cells without significant difference 5 dpi . However , in competition experiments , HS-independent variants outcompeted HS-dependent variants 5 dpi while the opposite occurred at earlier time points ( 1 hpi and 1 dpi in 2D and 4 hpi in 3D cultures ) . The higher amount of cell-associated HS-dependent viruses shortly after inoculation is easily explicable by the high expression of HS in neural tissues . However , the fact that the improved binding is not reflected by an improved replicative fitness 5 dpi is less obvious . A similar situation was observed in Vero cells where the binding advantage of EV71-VP197R167G did not result in improved replication relative to EV71-VP197L167E . Whether another step of the viral life is affected by the VP1 L97R mutation in Vero , 2D and 3D neural cultures is currently under investigation . Of note in this study and in our previous publication [39] , the HS-dependent variant was fitter in SH-SY5Y neuroblastoma cells at any time post infection . This difference between primary neural cultures and immortalized cells may rely on the nature of these culture models ( cancer neuron precursors for neuroblastoma cells versus mature neurons and glial cells for both 2D and 3D neural cultures ) and emphasizes the need to use relevant primary tissue culture models to study viral pathogenesis . Two explanations may account for the detection of HS-dependent virus only in the patient’s CSF and our experimental observations that the HS-independent virus outcompetes the HS-dependent virus in neural tissues 5 dpi ( thought both variants replicate to high level in independent infections ) . First , the patient’s CSF may have been sampled shortly after neural invasion , at a time where the HS-dependent variant was also fitter in neural cultures . However , this would imply that both variants were present together in the CNS where the HS-dependent variants outcompeted the HS-independent variants . The fact that only HS-dependent variants were found in the patient’s blood does not support this hypothesis . More probably , the ability to bind HS promoted EV71-VP197R167G replication in HS-enriched tissues ( such as the basal layers of the gastrointestinal mucosa , muscle or endothelial cells ) , resulting in sustained viremia and subsequent dissemination to neural tissues . In this context , EV71 neurotropism may not rely on an improved growth in neural tissues but rather on an improved dissemination ability selected upstream of CNS contamination over the course of an infection . The implication of HS binding in dissemination and neurotropism has been investigated for several viruses and remains controversial . Studies on aphtoviruses [47] , flaviviruses [48] , togaviruses [49] , alphaviruses [50] and even enteroviruses such as coxsackievirus B3 [51] suggest that binding to HS may lower viremia due to virus trapping in tissues expressing high HS levels , while it may improve neurotropism and ability to cross the blood-brain barrier [52] . Of note , two recent papers reported that infection by EV71 with VP1 E145Q substitution was associated with an attenuated phenotype in mice and cynomolgus monkeys and that this attenuation was linked to the ability of VP1 145Q to bind HS [31 , 34] . The authors conclude that HS binding and in vivo virulence are negatively correlated due to trapping and abortive infection of HS-dependent viruses in tissues expressing high levels of HS but low levels of SCARB2 . In the clinical case investigated here , the HS binding variant was the only variant sequenced from the blood of the patient suggesting that HS binding did not prevent viremia and dissemination but rather promoted it . Several hypotheses could explain these different observations: first , mice and monkeys may not faithfully reproduce EV71 infection in human as suggested by studies showing that isolates with VP1 145G/Q are frequently associated with severe neurological disease in humans [35–38]; second , the inoculation site may modulate the infection outcome . Huang and colleagues [53] showed that broad mutant spectra with divergent mutations are observed at the initial infection sites ( the respiratory and digestive systems ) and that a selection bottleneck occurs afterwards with subsequent enrichment of advantageous mutations in the viral population . Intraperitoneal or intravenous injections may thus prevent the natural evolution of the viral population observed after fecal-oral or respiratory transmission; third , the binding affinity for HS may differ according to the number and/or location of positive charges on the viral capsid , and different mutations ( VP1 E145Q versus VP1 L97R ) may result in different binding intensities and different trapping strengths; finally , the host immunity certainly plays a key role . In the clinical case presented here , the patient underwent immunosuppressive therapy with rituximab , an anti-CD20 antibody that depletes the peripheral B‐cell pool . Seroneutralisation experiments performed with the patients’ serum at a time when the VP1 97R mutation was already present in the plasma failed to highlight the presence of an antibody-mediated selective immune pressure against HS-dependent and HS-independent variants [39] . This absence of neutralizing antibodies was correlated with a negative complement fixation assay confirming a poor antibody response against enterovirus linked to the immunosuppressed state of the patient . Fujii et al showed that viruses with VP1 145E are more resistant to neutralizing antibodies than viruses with VP1 145G [31] . The absence of immune pressure may thus be a prerequisite for the emergence of HS-dependent variants in vivo , since critical residues , such as VP1 98 and 145 , seem to play an important role both in immune escape [30 , 31] and HS binding [29 , 34 , 54 , 55] . The emergence of HS-dependent mutants and their dissemination could thus be favored in the presence of a poor antibody response . Additional clinical investigations are needed to further assess the implication of HS binding in dissemination and disease severity in humans . The recent emergence of severe cases associated with EV71 genogroup C1 infections and the lack of an efficient vaccine or antiviral to fight EV71 infections [56 , 57] highlights an important need for the development of effective antiviral strategies [58] . In this study , we demonstrated the key role of HS attachment receptor binding in EV71 dissemination in an immunocompromised host , and our preliminary results in cells and neural cultures highlighted the inhibition of HS-dependent variants by soluble HS mimetic compounds . Soluble heparan sulfate analogs may thus be a useful tool to limit enterovirus replication and dissemination . In conclusion , our data may help to unravel the key determinants of EV71 dissemination and neurotropism and bridge the gap between EV71 neurological diseases and the lack of a therapeutic approach . Respiratory ( http://www . epithelix . com/products/mucilair ) and EpiIntestinal ( https://www . mattek . com/products/epiintestinal/ tissues ) were ordered from Epithelix and MatTek biotechnology companies . There , the tissues are developed from anonymized samples and after Ethical approval . The study was conducted according to the Declaration of Helsinki on biomedical research ( Hong Kong amendment , 1989 ) , and the research protocol was approved by our local ethics committee . RD ( human rhabdomyosarcoma , ATCC CCL-136 ) and Vero ( monkey kidney , ATCC CRL-1587 ) cells were cultured in Dulbecco's Modified Eagle Medium ( DMEM ) and GlutaMAX ( 31966021 , Gibco , Thermo Fisher Scientific , Switzerland ) supplemented with 10% ( v/v ) fetal bovine serum ( FBS ) ( P40-37500 , Pan Biotech , Chemie Brunschwig , Switzerland ) and 100 μg/ ml of penicillin and streptomycin ( 15140–122 , Gibco , Thermo Fisher Scientific , Switzerland ) . Caco-2 ( human colorectal adenocarcinoma , ATCC HTB 37 ) and SH-SY5Y cells ( human neuroblastoma , ATCC CRL- 2266 ) were grown in Eagle’s minimum essential medium ( EMEM ) ( BE12-125F , Lonza , Switzerland ) and DMEM/F12 medium ( 31331093 , Gibco , Thermo Fisher Scientific , Switzerland ) , respectively , supplemented with 2 mM L-glutamine ( 25030–024 , Gibco , Thermo Fisher Scientific , Switzerland ) , 1% non-essential amino acids ( 11140–035 , Gibco , Thermo Fisher Scientific , Switzerland ) and 10% ( v/v ) FBS . HEK 293T/17 ( human kidney , ATC C#CRL-11268 ) cells were grown in DMEM ( 41965039 , Gibco , Thermo Fisher Scientific , Switzerland ) supplemented with 10% ( v/v ) fetal bovine serum ( FBS ) , 100 μg/ ml of penicillin and streptomycin , 2 mM l-glutamine , 1 mM sodium pyruvate ( 11360–039 , Gibco , Thermo Fisher Scientific , Switzerland ) , and 1% non-essential amino acids . Cultures were maintained at 37°C in a 5% CO2 atmosphere . Infection medium was similar but with reduced FBS concentrations ( 2 . 5% for Vero and RD cells and 5% for Caco-2 and SH-SY5Y cells ) . Confluent Vero , Caco-2 , RD or SH-SY5Y cells pre-plated in a 96 well plate , were washed twice with heparinase III digestion buffer ( 0 . 1 M sodium acetate pH 7 . 0 , 1 mM calcium acetate and 0 . 2% BSA ) and incubated for 1 h at 37°C with 3 . 5 mIU/ml of heparinase III ( AMS . HEP_ENZ III_S , Amsbio , Switzerland ) ( 50 μl/well ) . Control samples were incubated with digestion buffer alone . Heparinase III efficient cleavage was assessed by immunofluorescence with a mouse anti-Δ-heparan sulfate F69-3G10 antibody ( 370260-S , Amsbio , Switzerland , diluted 1:500 ) specific for a heparan sulfate neo-epitope generated after digestion . EV71-VP197R/L167G/E mixed viral population was pre-incubated with 300 μg/ml of soluble heparin sodium salt from porcine intestinal mucosa ( H4784 , Sigma-Aldrich , Merck , Switzerland ) diluted in culture medium or with culture medium alone as control for 1 h at 37°C . Confluent monolayers of Vero , Caco-2 , RD or SH-SY5Y cells were challenged with 50 μl of the pre-incubated virus-heparin or virus-alone mixture and virus binding and replication assays were performed . Confluent cells were first seeded in 96-well plates and binding and replication assays were performed 24h later as previously described [39] . Briefly , the medium was removed , cells were washed with cold binding buffer ( Hanks' Balanced Salt Solution [HBSS , 14175053 , Gibco , Thermo Fisher Scientific , Switzerland] containing 1% BSA [wt/vol] and 0 . 1% sodium azide [wt/vol] ) and incubated for 1h on ice with 50 μl of viral suspension containing 5108 viral RNA copies . After removal of unbound virus with 3 washes with 200 μl of cold binding buffer , cells were either lysed directly in 200 μl of easyMAG lysis buffer to quantify bound viruses by RT-qPCR or overlaid with fresh culture medium and left an additional day at 37°C before lysis and viral load quantification by RT-qPCR . To determine the dominant population in binding and replication experiment run with EV71-VP197R/L167G/E , the VP1 region was PCR-amplified and sequenced . Aliquots of 200 μl heparin sepharose beads ( ab193268 , Abcam , UK ) were immobilized to Pierce cellulose acetate filter spin cups ( 69702 , Thermo Fisher Scientific , Switzerland ) by centrifugation at 1500 rpm for 5 min . The beads were then equilibrated with 600 μl of binding buffer ( 0 . 02M Tris-HCl , 0 . 14M NaCl , pH 7 . 4 ) and centrifuged for 5 min at 1500 rpm . Subsequently , 100 , 200 or 600 μl of viral suspension of EV71-VP197R/L167G/E containing respectively 109 , 2109 and 6109 viral RNA copies were added and incubated with the heparin sepharose beads for 1h at 4°C with gentle mixing . The flow-through fraction was collected by centrifugation ( 5 min at 1500 rpm ) and the heparin sepharose beads were washed 5 times with 200 μl of binding buffer . Bound viral particles were then eluted twice by sequential incubation of 5 min with 200 μl of elution buffer ( 0 . 02M Tri-HCl , 2M NaCl , pH 7 . 4 ) . The collected fractions ( input , flow-through and eluate ) were analyzed by RT-qPCR and VP1 was PCR amplified and sequenced . RNA was extracted using the NucliSens easyMAG magnetic beads system ( BioMérieux , France ) according to the manufacturer’s instructions . RT-qPCR was performed using the quantitative Entero/Ge/08 assay as previously described [61] in a one-step format using the QuantiTect Probe RT-PCR Kit ( Qiagen , Switzerland ) according to the manufacturer's instructions in a StepOne Applied Biosystems thermocycler . As a quantitative reference standard for each run , 10-fold dilution series ( from 2 . 5108 to 2 . 5105 copies/ml ) of the in vitro transcribed full-length pBMH-EV-D68 was used . The RNAse P housekeeping gene ( 4316831 , Thermo Fisher Scientific , Switzerland ) was quantified by qPCR in binding and replication experiments to confirm homogeneous cell number . Retro-transcription was performed using Superscript II ( Invitrogen ) and either random hexamer primers ( Roche ) or a specific primer ( 2A-3408-R: 5’CTGGGTTTTGAAAAGCTGACC3’ ) as previously described [39] . The VP1 region was amplified by PCR ( Fwd 5’TGCTCGAGATGGAGTATTCG3’ and Rev 5’CTGGGTTTTGAAAAGCTGACC3’ ) and nested PCR ( Fwd 5’CGACTACTACACTACAGGCTTGGTTAG3’ and Rev 5’CATTGGGCGAGGTATCCAC3’ ) with platinum Taq DNA polymerase ( P/N 10966026 , Invitrogen , Thermo Fisher Scientific , Switzerland ) . PCR products were purified as previously described [39] and sequenced ( Fasteris-DNA sequencing service , Switzerland ) . Two SCARB2 targeting guide RNAs ( sgRNA1 5’-CACCGCGATGCTGCTTCTACACGGC-3’ & 3’-CGCTACGACGAAGATGTGCCGCAAA-5’ , and sgRNA2 5’-CACCGCCGGCATTGTCTGACGTAT-3’ & 3’-CGGCCGTAACAGACTGCATACAAA-5’ for the sgRNA2 ) were subcloned in the pLentiCRISPRV2 vector ( Addgene # 52961 ) . Recombinant lentiviruses ( plentiCRISPRV2 , pLentiCRISPRV2-SCARB2-sg1 or pLentiCRISPRV2-SCARB2-sg2 ) were produced by transient transfection of HEK 293T/17 cells as previously described [62] . Caco-2 and RD cells ( 70–80% confluency in 6-well plates ) were transduced with 1 ml of viral supernatants . Five days post transduction the cells were trypsinised and subjected to puromycin selection ( 58-58-2 , Invivogen , LABFORCE , Switzerland; 10 μg/ml for Caco-2 cells and 6 μg/ml for RD cells ) . Cells were maintained in the selection medium and the expression of SCARB2 was assessed by Western Blot . Human upper ( MucilAir ) and lower ( SmallAir ) airway epithelia ( Epithelix , Geneva , Switzerland ) were developed from isolated nasal polyp and distal lung epithelial cells originating from surgeries as previously described [40 , 41] . Human small intestine tissues ( EpiIntestinal ) were purchased from MatTek ( Ashland , USA ) and are reconstituted from primary human small intestinal epithelial cells isolated from the ileum of healthy adult donors . After full differentiation , the EpiIntestinal tissues faithfully reproduce the pseudo-stratified architecture of the human small intestine composed of enterocytes , paneth cells , M cells , tuft cells and intestinal stem cells [https://www . mattek . com/products/epiintestinal/] . The tissues , cultured at an air-liquid interface ( at 33°C for MucilAir and at 37°C for the other tissues ) , were infected apically or basally as previously described [63] . Briefly , for apical infection , 108 viral RNA copies ( in 100 μl of culture medium ) culture medium were applied at the tissue/air interface while for basal infection , 6108 viral RNA copies ( in 600 μl of culture medium ) were applied at the tissue/liquid interface . After 4 hours , residual viruses were removed by extensive apical or basal washes and then , every day , 200 μl of culture medium was applied apically for 20 min to collect apically released viruses . Basal medium was also collected daily and replaced with 500 μl of fresh medium . RNA was extracted from apical samples collected 5 dpi . 2D and 3D neural tissues were engineered from human induced pluripotent neural stem cells ( GSC-4311 , MTI-GlobalStem , Thermo Fisher Scientific , Switzerland ) and contained a mixed population of mature neurons and glial cells [64] . 2D cultures were inoculated with 107 viral RNA copies ( in 100 μl of culture medium ) . Residual inoculum was removed after 1 h and tissues were lysed 1 hpi and 5 dpi for RNA extraction . 3D cultures were inoculated with 4*107 viral RNA copies ( in 40 μl of medium ) and tissues were lysed 4 hpi or 5 dpi for RNA extraction . For all tissues , replication was quantified by RT-qPCR and VP1 was PCR amplified and sequenced in both single infections or competition assays . Proteins were loaded on a 10% SDS-PAGE gel , transferred on a PVDF membrane ( 162–0177 , BIO-RAD , Switzerland ) that was hybridized with the primary Ab overnight at 4°C . The membranes were washed twice with TTBS 0 . 05% [TBS buffer ( 10 mM Tris HCl , pH 7 . 5 , 500 mM NaCl ) with 0 . 05% Tween] for 10 min and incubated for 1 h at 37°C with the secondary Ab . The membranes were washed twice with TTBS for 10 min and developed with the ECL system ( 34080 , Thermo Fisher Scientific , Switzerland ) according to the manufacturer’s protocol . Images were acquired with the Fujifilm LAS 4000 luminescence imager . Images were acquired on a Zeiss LSM700 confocal microscope with a 63x/1 . 4 oil objective leading to the calibration of 0 . 396 microns per pixel . HS expression was measured with MetaMorph software version 7 . 7 . 6 ( Molecular Devices , Sunnyvale , CA ) . I-carrageenan ( C1138 , Sigma-Aldrich , Merck , Switzerland ) was serially diluted and added for 1h at 37°C to EV71-VP197R167G , EV71-VP197L167E or EV71-VP197R/L167E/G ( multiplicity of infection [MOI] , 0 . 01 PFU/cell ) . The mixture was then added to confluent 2D neural cells or Vero cells grown in a 96-well plate at a density of 13 × 104/well . After 1 h at 37°C , monolayers were washed and overlaid with fresh medium . One dpi , viral replication was monitored by immunocytochemistry ( ICC ) in Vero cells and by RT-qPCR in neural cells . For ICC , cells were washed with PBS , fixed with cold methanol:acetone ( 1:1 ) for 1 min and permeabilised with PBS-Triton 0 . 1% for 5 min on ice . Cells were then incubated 1h at 37°C with mouse anti-EV71 VP2 monoclonal antibody ( MAB979 , Millipore , Merck , Switzerland; diluted 1:500 in PBS-BSA 1% ) . After 3 washes with PBS-Triton 0 . 1% , cells were incubated 1 h at 37°C with the anti-mouse HRP-labelled secondary antibody ( 7076 , Cell Signaling Technology; diluted 1:1000 in PBS-BSA 1% ) . After extensive washing , DAB substrate solution ( D4293-50SET , Sigma-Aldrich , Merck , Switzerland ) was added for 15 min before final washing with PBS . More than 400 cells per well were scored under a light microscope . The percent of inhibition of virus infectivity was determined by comparing the percentage of infected cells in presence of increasing concentrations of ı-carrageenan relative to the percent of infected cells in absence of ı-carrageenan ( set as reference at 100% ) . All data were generated from duplicate wells in three independent experiments . For quantification by RT-qPCR , neural cultures were lysed in 200 μl of easyMAG lysis buffer . RNA was extracted , viral replication was quantified by RT-qPCR and VP1 was PCR amplified and sequenced . The percent of inhibition of virus infectivity was determined by comparing the viral load measured in neural cells in presence of increasing concentrations of ı-carrageenan relative to the viral load measured in absence of ı-carrageenan ( set as reference at 100% ) . Values are expressed as mean ( ± SEM ) . Experiments were performed at least in biological duplicates , with N in the figure legend indicating the number of replicates . Two-way ANOVA , Fisher’s exact tests and determination of the fifty percent inhibitory concentration ( IC50 ) was determined using GraphPad Prism 7 . 02 software .	Enterovirus 71 ( EV71 ) has been the cause of major hand-foot-and-mouth disease outbreaks , particularly in the Asia-Pacific region . However , the recent emergence of severe neurological cases associated with EV71 infection in Europe and the lack of an efficient vaccine or antiviral agent to fight EV71 infections highlight two critical needs: ( A ) the identification of ill-defined viral factors that contribute to viral dissemination and pathogenesis in humans and ( B ) the development of effective antiviral strategies . Herein , based on clinical observation in an immunocompromised host , we have demonstrated that heparan sulfate attachment receptor played a critical role in EV71 virulence and that “in host” EV71 adaptation to a heparan sulfate-dependent virus was likely responsible for its dissemination . To our knowledge , this is the first study highlighting the key determinants of EV71 dissemination based on a clinical case and proposing a new therapeutic approach against EV71 neurological diseases .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "cell", "binding", "cell", "physiology", "medicine", "and", "health", "sciences", "chemical", "characterization", "respiratory", "infections", "biological", "cultures", "microbiology", "neuroscience", "pulmonology", "caco-2", "cells", "digestive", "system", "research", "a...	2018	A VP1 mutation acquired during an enterovirus 71 disseminated infection confers heparan sulfate binding ability and modulates ex vivo tropism
Ciliopathies are genetic disorders arising from dysfunction of microtubule-based cellular appendages called cilia . Different cilia types possess distinct stereotypic microtubule doublet arrangements with non-motile or ‘primary’ cilia having a 9+0 and motile cilia have a 9+2 array of microtubule doublets . Primary cilia are critical sensory and signaling centers needed for normal mammalian development . Defects in their structure/function result in a spectrum of clinical and developmental pathologies including abnormal neural tube and limb patterning . Altered patterning phenotypes in the limb and neural tube are due to perturbations in the hedgehog ( Hh ) signaling pathway . Motile cilia are important in fluid movement and defects in motility result in chronic respiratory infections , altered left-right asymmetry , and infertility . These features are the hallmarks of Primary Ciliary Dyskinesia ( PCD , OMIM 244400 ) . While mutations in several genes are associated with PCD in patients and animal models , the genetic lesion in many cases is unknown . We assessed the in vivo functions of Growth Arrest Specific 8 ( GAS8 ) . GAS8 shares strong sequence similarity with the Chlamydomonas Nexin-Dynein Regulatory Complex ( NDRC ) protein 4 ( DRC4 ) where it is needed for proper flagella motility . In mammalian cells , the GAS8 protein localizes not only to the microtubule axoneme of motile cilia , but also to the base of non-motile cilia . Gas8 was recently implicated in the Hh signaling pathway as a regulator of Smoothened trafficking into the cilium . Here , we generate the first mouse with a Gas8 mutation and show that it causes severe PCD phenotypes; however , there were no overt Hh pathway phenotypes . In addition , we identified two human patients with missense variants in Gas8 . Rescue experiments in Chlamydomonas revealed a subtle defect in swim velocity compared to controls . Further experiments using CRISPR/Cas9 homology driven repair ( HDR ) to generate one of these human missense variants in mice demonstrated that this allele is likely pathogenic . Primary cilia are solitary and immotile cellular appendages that serve as signaling hubs for pathways such as Hedgehog ( Hh ) during development [1] . Motile cilia initiate and maintain fluid flow and are critical in the brain for cerebral spinal fluid flow and are necessary for mucus transport in the lungs [2] . During development , motile cilia are responsible for initiating flow at the embryonic node which is critical for setting up left-right asymmetry in the mammalian body [3–5] . While all cilia have common core components such as tubulin and intraflagellar transport proteins , motile cilia possess several accessory structures such as inner dynein arms ( IDAs ) , outer dynein arms ( ODAs ) , radial spokes , and the nexin-dynein regulatory complex ( N-DRC ) . In Chlamydomonas reinhardtii , data indicate that the N-DRC functions to link the A microtubule of one doublet with the B microtubule of the adjacent doublet . It coordinates the activities of the outer and inner dynein arms to regulate flagellar beat frequency and waveform [6 , 7] . Studies in Chlamydomonas have led to the identification of several N-DRC proteins many of which appear to be conserved in mammals [8 , 9] . As in Chlamydomonas , mutations in putative mammalian N-DRC proteins CCDC164 ( DRC1 ) , CCDC65 ( DRC2 ) , and most recently , GAS8 ( DRC4 ) are correlated with defects in ciliary motility [10–13] . The human homolog of Gas8 was originally identified in human breast cancer and referred to as Growth Arrest Specific 11 ( GAS11 ) [14] . This gene shares 56% protein identity to an N-DRC component in Chlamydomonas known as DRC4 , the protein product of the paralyzed flagella 2 ( PF2 ) gene [15] . Loss of PF2 ( DRC4 ) in Chlamydomonas leads to loss of IDAs and the majority of the N-DRC ( N-DRC proteins DRC3-7 ) visible by transmission electron microscopy ( TEM ) and results in a slower forward swimming velocity and defective waveform [6 , 16 , 17] . The function of Gas8 in the mammalian N-DRC remains poorly understood . Gas8 localizes to the axoneme of motile cilia and also to the base of primary cilia in vertebrate cells [18] . This led us to question if Gas8 serves as an N-DRC component in motile cilia and whether it has a separate role in non-motile primary cilia . This possibility is supported by data from knockdown studies of Gas8 in NIH3T3 cells showing defects in Hh pathway responses . Expression of truncated versions of Gas8 , after knockdown of endogenous Gas8 , revealed that the C-terminal region of Gas8 bound to and facilitated the transport of Smoothened into the cilium in response to Hh pathway activation using the Smoothened agonist ( SAG ) [18 , 19] . In mammals , cilia are essential for normal regulation of Hh signaling activity with many of the Hh signaling components such as Smoothened , Patched and Gli transcription factors dynamically localizing in primary cilia [20–22] . Primary Ciliary Dyskinesia ( PCD , OMIM #244400 ) is a human disease characterized by abnormal motile cilia . PCD patients exhibit bronchiectasis , infertility , and chronic respiratory infections , and in some cases can present with hydrocephalus . A subset of PCD patients will also have a reversal of their left-right body axis that includes situs inversus totalis which is referred to as Kartagener syndrome [23] . PCD patients often have changes in cilia axonemal ultrastructure that include defects in the inner or outer dynein arms , central complex , radial spokes , and the N-DRC [24–28] . These structural defects alter ciliary beat frequency ( CBF ) , ciliary waveform , and cilia orientation . Recent data indicate that mutations in the putative mammalian N-DRC components CCDC164 , CCDC65 , and Gas8 correlate with the clinical presentation of PCD . Mutations in these genes lead to dyskinetic cilia with subtle changes in cilia ultrastructure pointing to an importance for these components in ciliary motility . Other proteins such as CCDC39 and CCDC40 are responsible for the assembly and attachment of the IDAs and N-DRC in motile cilia . The absence of these proteins results in severe motility defects [10 , 11 , 29–31] . In this study , we investigate a role for Gas8 in both primary and motile cilia in vivo . For this we generated a Gas8 genetrap mutant mouse . Gas8 mutants present with severe hydrocephalus and cilia motility defects on both the ependyma and trachea , as well as a situs inversus phenotype . Given the role for Gas8 in cilia motility and recent data suggesting it is a PCD causing allele , we screened human PCD patients for GAS8 mutations and identified two independent missense variants . The potential pathogenicity of these alleles was tested by rescue experiments in Chlamydomonas PF2 mutants and by generating a mouse model for one of the variants . In contrast to the PCD phenotypes , we did not observe Hh associated defects in any of the mutant mice or cell lines derived from them or other phenotypes typically associated with defects in primary cilia function . These results suggest that GAS8 plays a highly conserved role in ciliary motility and mutations in Gas8 are associated with human disease through their impact on motile cilia . A β-geo cassette containing the β-galactosidase enzyme , a neomycin resistance cassette , an N-terminal splice acceptor and poly-A tail was inserted in intron 7 of the Gas8 mouse allele ( Fig 1A , herein referred to as Gas8GT ) . RT-PCR analysis using primers located before the genetrap insertion indicates that the 5’ end of the transcript is generated ( Fig 1B , left ) . In contrast , primers located 3’ to the insertion failed to detect any Gas8 mRNA ( Fig 1B , right ) . Western blot analysis shows a product of expected size ( 57kDa ) in wildtype and heterozygous Gas8 mice . This product is absent in homozygous mutants ( Fig 1C ) . Additionally , the Gas8::β-geo fusion protein is detected ( approx . 230kDa ) in heterozygous and homozygous mutants indicating that the genetrap allele is being transcribed and translated . Loss of Gas8 led to lethality at approximately postnatal day 14 ( P14 ) with few living to P21 . All mutants presented with severe hydrocephalus ( Fig 1D ) . Gas8GT mutant mice also presented with situs inversus at a rate of 36% ( 6 of 16 mutants ) in live births based on position of the heart and stomach ( Fig 1E and 1F ) . Both the hydrocephalic and situs inversus phenotypes suggested a defect in the function of motile cilia . Based on a previous study reporting Gas8 as a positive effector of Hh signaling in mammals , we anticipated Gas8GT mice would present with phenotypes related to Hh signaling defects , especially since this mutation would lack the putative Smo binding domain ( amino acids 386–478 ) . However , we did not observe any hedgehog-associated phenotypes in limb patterning or neural tube formation . To further test a role for Gas8 in the Hh pathway , we isolated Mouse Embryonic Fibroblasts ( MEFs ) from Gas8WT and Gas8GT mutant mice and treated them with 150nM Smoothened agonist ( SAG ) . The MEFs were then immunolabeled for Smo and acetylated tubulin to analyze differences in Smo trafficking into the cilium ( Fig 2A ) . In contrast to the outcome of the knockdown studies , there was no difference between the amounts of Smo present in the cilia of Gas8GT mutants when compared to Gas8WT cilia ( Fig 2B ) . These data indicate that a least in the Gas8GT mutants , Gas8 is not an essential factor involved in regulating Smo cilia trafficking . Similarly , none of the Gas8GT mutants exhibited defects in dorsal ventral patterning of the neural tube typical of altered Hh signaling ( Fig 2C ) . To investigate the hydrocephalus phenotype , cilia morphology , ultrastructure and motility on ependymal and tracheal cells was assessed . DIC analysis and immunofluorescence staining of trachea indicate motile cilia are present on the epithelium , but the Gas8GT protein fails to localize to these cilia ( Fig 3 ) . We counted cilia from trachea TEMs for broken doublet rings and found that about 9% of cilia from Gas8GT mutants showed disorganization of the arrangement of the microtubule doublets ( Fig 4A arrowhead and 4E ) . To analyze ultrastructure within the doublets , we averaged 202 doublets of both genotypes to reduce variability due to random sectioning of the 96nm repeat of the microtubule doublet . We did not observe any major structural differences in the inner or outer dynein arms ( Fig 4B ) . However , high speed video and Fourier transformation analysis revealed that cilia are largely static with only a few moving ( Fig 4C and 4D , S1 Movie and S2 Movie ) . Those cilia that did moved were dyskinetic , resulting in an inability of cilia to propel fluid as seen by tracking of fluorescent beads added to either brain ventricle or trachea preparations ( Fig 4F and 4G ) . Beat frequency of cilia that remained motile in Gas8GT mutants was modestly decreased from 17 . 0Hz in Gas8WT to 12 . 7Hz in Gas8GT ( Fig 4H ) . Cilia length is also affected in Gas8GT mice , with Gas8GT motile cilia measuring 0 . 9μm shorter than Gas8WT motile cilia ( Gas8WT 5 . 3μm and Gas8GT 4 . 4μm ) ( Fig 4I ) . Cilia orientation in Gas8GT tracheas is also more randomized than in Gas8WT controls ( Fig 4J ) . These phenotypes observed in the motile cilia of Gas8GT mutant mice are similar to those observed in PCD patients and animal models . The phenotypes in the Gas8GT mutants led us to evaluate whether mutations in GAS8 are associated with PCD in humans . We identified two independent missense variants , c . 595G>A E199K and c . 1172C>T A391V , in human patients through a previously published screen ( Fig 5A ) [32] . The E119K patient is of Latino decent and presented with heterotaxy . Unaffected parents of the patient are heterozygotes , and an unaffected female sibling is a homozygote . This allele appears at a frequency of 11% in Latino populations ( 87 homozygotes and 1279 heterozygotes in a total of 11564 alleles sequenced according to ExAC ) . The prevalence of this allele in the Latino community makes it unlikely to be associated with disease . The A391V patient met the diagnostic criteria for PCD . This allele is infrequent , occurring only 3 times heterozygously and 0 times homozygously in 84864 alleles sequenced according to ExAC . Both variants affect highly conserved regions across multiple species ( Fig 5B ) . We utilized the PolyPhen-2 program to predict the pathogenicity of these alleles . The A391V allele had a PolyPhen-2 score of 0 . 762 suggesting that it is a potentially damaging mutation while the E199K allele had a score of only 0 . 082 , suggesting that this is a benign mutation . Given the low allele frequency , PolyPhen-2 score , and the confirmation of the PCD diagnosis in the patient carrying the A391V variant , we chose to test potential pathogenicity of this allele in mice . To further assess the potential pathogenicity of the human allele , we created a mouse harboring the A391V mutation via homology driven repair with CRISPR/Cas9 technology . Sequencing confirmed the presence of the c . 1172 C>T mutation resulting in an A391V amino acid change ( Fig 6A ) . We crossed Gas8AV mice onto the Gas8GT background to create compound heterozygous ( Gas8GT/AV ) mice . To determine the impact on motile cilia and test possible cause of the hydrocephalus , we took brains from 6 week old mice and analyzed cilia beat and the ability of motile cilia to move fluid . While there were no differences in beat frequency , bead flow analysis shows a modest decrease in the ability of Gas8GT/AV cilia to move fluid compared to Gas8GT/WT cilia ( Fig 6B and 6C ) . Compound heterozygotes develop mild hydrocephalus at approximately 10 weeks of age ( Fig 6D ) but there were no evident laterality defects . While all the Gas8GT/AV mice analyzed ( n = 6 ) display hydrocephalus at this age , the severity ranged from mild ( arrowhead ) to moderate ( arrow ) . The phenotype in the Gas8GT/AV mice is not as severe as in the Gas8GT mice and hydrocephalus was not present in any ( n = 2 ) of the Gas8GT/WT or ( n = 2 ) of the Gas8WT/AV mice analyzed . We chose to generate the A391V mouse model because of the PCD symptoms of the patient but given the lack of full PCD symptoms in the E199K patient , we decided to test first whether or not the E199K is pathogenic in Chlamydomonas before proceeding to a potential mammalian model . Alignment of GAS8 and the Chlamydomonas orthologue DRC4 revealed that E199 in GAS8 aligns with D198 in DRC4 ( Fig 5B ) . To better understand the mechanisms underlying these defects , we generated strains expressing the Chlamydomonas equivalent ( D198K ) of the human E199K alleles in a null mutant background ( pf2 ) . Interestingly , transformation with DRC4-DK-GFP rescued the severe motility defects seen in the pf2 null mutant , but measurements of forward swimming velocities revealed a subtle defect in the swimming phenotype of the rescued strains ( Fig 6E ) . Furthermore , the DRC4-DK-GFP protein is assembled at wild-type levels in the flagellar axonemes of Chlamydomonas , as assayed by western blot ( S1 Fig ) . These observations show that the D198K DRC4 mutant protein is properly localized in the axoneme and may not correspond to a pathogenic allele . Defects involving cilia motility cause severe phenotypes in humans including infertility , hydrocephalus , respiratory defects , and reversal of left-right asymmetry . Much of our understanding about cilia motility has come from studies in organisms such as Chlamydomonas . These studies and how defects in cilia motility cause disease are now being extended into mammalian systems . Recently GAS8 was implicated as a cause for Primary Ciliary Dyskinesia ( PCD ) as well as a positive effector of Smoothened transport into cilia during Hh pathway activation [12 , 13 , 19] . To further evaluate the connection between Gas8 and PCD in mammals , we generated a mouse with a β-geo cassette inserted in intron 7 of the Gas8 gene . Insertion of the β-geo genetrap cassette effectively eliminated the presence of wildtype transcript and protein in mutants as verified by RT-PCR and western blot analysis . Though the Gas8GT mutant allele is translated into a large fusion protein between the N-terminal portion of Gas8 and β-geo , it does not localize to motile cilia . Gas8GT mutant mice present with hydrocephalus starting at postnatal day 5 ( P5 ) that becomes more pronounced as the mice mature and eventually leads to mortality between P14-P21 . Development of hydrocephalus is associated with severe impairment of cilia motility on ependymal cells lining the ventricles of the brain . Previous studies using image average procedures to analyze flagella ultrastructure in Chlamydomonas showed that strains with mutations in PF2/DRC4 , the Gas8 homolog , were associated with the loss of the majority of the N-DRC complex along with a subset of the IDAs [6 , 16 , 17 , 33] . In contrast to the Chlamydomonas results , the N-DRC and IDA do not appear to be overtly affected in Gas8 mutant mice based on standard thin-section TEM analysis . However , loss of Gas8 does effect microtubule organization , as indicated by a higher percentage of cilia with disorganized microtubule doublets in Gas8GT mutant mice when compared to Gas8WT mice . Altered cilia microtubules were recently also recently observed in human Gas8 patients [12 , 13] . Together these data suggest that defects in the mammalian N-DRC may not always be detectable using traditional TEM averaging of cross-sections . The inability to observe ultrastructural defects in human PCD patients could be attributed to having only one N-DRC per 96nm repeat . Future studies using better imaging approaches such as cryo-electron tomography and image averaging of longitudinal sections to assess the human N-DRC will likely continue to reveal structural and functional differences similar to those described for the radial spokes by Lin , et al 2014 . Most Gas8GT mutant cilia failed to move , however those that were observed moving displayed a modest decrease in beat frequency . The most distinguishable phenotype observed in the cilia that moved was a very rigid and short wave pattern . This pattern has also been observed in other cilia motility mutants thought to affect the NDRC [10–12] . These changes in waveform and the lack of overall motility result in the defective fluid flow observed in these mice . Previous data show a complex relationship between planar cell polarity ( PCP ) and fluid flow in establishing motile cilia orientation [34 , 35] . Gas8GT cilia show a more random distribution of cilia orientation than their Gas8WT counterparts supporting the necessity of proper fluid flow in establishing cilia orientation . Variants in Gas8 were recently identified in human PCD patients . These mutations resulted in a similar , albeit not significant , decrease in beat frequency along with an abnormally rigid ciliary waveform [12 , 13] . This motility phenotype is similar to our observation in the mutant mice . Here we identified an additional independent missense mutation , c . 1172C>T A391V , in PCD patients as well as a variant c . 595G>A E199K that appears to have minimal effect on cilia motility . The A391V mutations lies in close proximity to the other published mutants , C309* , A334* , and G357* suggesting that this area is critical for GAS8 function . Similarly , the genetrap cassette in the Gas8GT allele was inserted in close proximity ( K337 ) to the A334* mutation . The E199K mutation also affects a highly conserved region within Gas8 that is proposed to be a Microtubule Association Domain ( GMAD ) [36] . To test pathogenicity of the A391V allele , we used CRISPR/CAS9 homology driven repair ( HDR ) to generate a mouse line mimicking the human mutation . Mice compound heterozygous for the Gas8GT and Gas8AV mutations develop age dependent , mild hydrocephalus , but did not present with situs defects ( n = 6 Gas8GT/AV mice ) . The phenotype was associated with a reduced ability of ependymal cilia to move fluid . Interestingly , beat frequency was not significantly altered from that of controls , suggesting that the defect lies within a subtle waveform difference or in cilia orientation . These data suggest that the A391V allele is pathogenic though more in-depth analysis of ciliary defects will be necessary to determine the precise mechanism . Data from the Chlamydomonas rescue experiments suggest that the E199K allele may have very subtle effects on motility . The D198K rescued strain in Chlamydomonas showed a small but statistically significant reduction in forward swim velocity of approximately 10 percent . While statistically significant , additional work is needed to determine whether such small changes might impact ciliary motility and have pathogenic consequences in different tissues and different organisms . As this variant is commonly found in Latino populations , it seems more likely that this variant is a benign polymorphism . Gas8 was previously implicated as a modulator of the Hh pathway . In vitro data indicated that the C-terminal region of Gas8 binds to Smoothened ( Smo ) and acts at the base of primary cilia as a regulator of Smo entry into the cilium following Hh pathway activation [19] . These data showed that in the absence of Gas8 , Smo accumulation in the cilium is abrogated and that it cannot activate the Gli transcription factors and turn on downstream genes . Based on these in vitro findings , we expected to see Hh defects in our mutant mice . However , the Gas8 mutants survive to birth and have normal digit number and patterning as well as normal neural tube dorsal ventral patterning . Furthermore , there were no significant differences in Smo accumulation in cilia between Gas8WT and Gas8GT MEFs after SAG stimulation , suggesting that in this mutant model , Gas8 does not act as a regulator for Smo entry . The role that Gas8 plays at the base of primary cilia remains uncertain; however , we do not see any other pathologies that would suggest there is a defect in primary cilia such as cystic kidney disease . The data presented here solidify GAS8 as a disease causing gene in humans and elucidate the mechanisms by which loss of Gas8 causes disease . We identified new independent , homozygous missense mutations and used model systems to test the pathogenicity of the alleles . Importantly , these results suggest the A391V allele is pathogenic while the E199K variant is not . Our results demonstrate the importance of testing the potential pathogenicity of human alleles in easily amenable model systems such as Chlamydomonas and further reveal the ease with which CRISPR/Cas9 has now made it possible to conduct similar tests in mouse models . The Gas8 mutant mouse line was generated using embryonic stem cell line CH0760 ( BayGenomics ) in which a β-galactosidase neomycin resistance fusion cassette was inserted into intron 7 of Gas8 . The insertion site was confirmed by genomic PCR and sequence analysis . PCR primers for genotyping were designed based on the insertion site and are as follows 5’-GGGACAAGCAGATTCTGGTC-3’ , 5’-CAGGGTTACACACAGAGAAACC-3’ , and 5’-CCGCAAACTCCTATTTCTG-3’ . The Gas8GT embryonic stem cells were from the 129P2/OlaHsd genetic background and were injected into C57BL/6 blastocysts using standard procedures . Chimeras were bred with albino C57BL/6 females and germline transmission was confirmed by coat color and subsequent PCR genotyping . All experimental procedures were approved by the Institutional Animal Care and Use Committee ( IACUC ) regulations at the University of Alabama at Birmingham under the animal protocol number ( 130208061 ) . RNA was isolated from Gas8WT , Gas8GT/WT , and Gas8GT mouse embryonic fibroblasts using Trizol reagent according to the manufacturer’s protocol ( cat# 15596–026 , Thermo-Fisher Scientific ) . cDNA was generated using Verso cDNA kit ( cat# AB-1453/B , Thermo-Fisher Scientific ) . 5’ Gas8 RT-PCR was performed using the following primers spanning exons 3 and 4: 5’-GAATCGAAGAATACCACCATC-3’ and 5’-CTGAGAAGATGGCTATGTAG-3’ . 3’ Gas8 RT-PCR was performed with primers spanning exons 9 and 10: 5’-CTGGACCCCACAGCATTAAC-3’ and 5’-CTTGATGGTGGTATTCTTCG-3’ . Actin control primers: 5’-ATGGGTCAGAAGGACTCCTA-3’ and 5’-GGTGTAAAACGCAGCTCA-3’ were used in all samples . Animals were anesthetized by a 0 . 1 ml per 10 g of body weight intraperitoneal injection of 2 . 5% tribromoethanol ( cat# T48402 , Sigma–Aldrich ) , killed by cardiac puncture , and perfused with PBS followed by 4% paraformaldehyde ( cat# 19943 , Thermo-Fisher Scientific ) . The brains were further fixed in 4% paraformaldehyde 1h at room temperature followed by successive dehydration through 1 hour alcohol incubations at 30% and 50% and placed finally in 70% overnight . Tissues were further dehydrated through 1 hour alcohol incubations at 80% , 95% , and finally 100% . Tissues were placed in xylenes for 1 hour and then placed in a 50/50 xylenes/paraffin mix for 1 hour at 60°C under vacuum followed by a final paraffin penetration in paraffin at 60°C under vacuum for 1 hour and then paraffin embedded . The brains were sectioned at 10μm and stained with Cresyl Violet stain as previously described [37] . Fresh tracheas were extracted from p21 Gas8WT , Gas8GT/WT , and Gas8GT mice . Samples were submerged in ice cold RIPA ( 10mM Tris pH7 . 5 , 150mM NaCl , 1%NP-40 , 1% sodium deoxycholate , 0 . 1% SDS ) mixed with one cOmplete Protease Inhibitor tablet ( cat# 11 836 170 001 , Roche Diagnostics ) per 10mL at 300uL per 5mg of tissue . Tissues were sonicated 3x for 10 seconds each . After sonication , tissues were placed on a rotary agitator for 2 hours at 4°C and then spun for 20 minutes at 12 , 000rpm at 4°C . Supernatant was removed and protein levels were assayed using a BioRad DC protein assay kit ( cat# 5000111 , Bio-Rad ) . Approximately 20μg per sample was used for SDS-PAGE with a 12% Tris-Glycine gel ( cat# 00252562 , Thermo—Fisher Scientific ) . Proteins were transferred overnight to nitrocellulose . The membrane was blocked for 45 minutes in 5% milk in PBS and incubated with primary antibody in 5% milk in PBS with 0 . 02% Tween-20 overnight at 4°C . Primary rabbit anti-Gas8/DRC4 antibody was used at 1:20000 [6] . Primary mouse anti-GAPDH was used as a loading control at 1:1000 ( cat# ab8245 , Abcam , Cambridge UK ) . Blots were washed 5x for 5 minutes each in 0 . 02% PBS-Tween-20 . Secondary antibody in 5% milk in 0 . 02% PBS-Tween-20 was added and the blot was incubated for 1 hour at room temperature with the following secondary antibodies: anti-mouse IRDye 800CW ( cat# 827–08364 , LI-COR , Lincoln NE USA ) and anti-rabbit IRDye 680RD ( cat# 926–68071 , LI-COR ) . Blots were washed 5x for 5 min each in 0 . 02% PBS-Tween-20 and then dried . Images were taken on a LI-COR Odyssey CLx imaging system ( LI-COR ) . Mouse embryonic fibroblasts ( MEFs ) were generated from E14 . 5 embryos and cultured in DMEM growth medium with High Glucose , 0 . 05mg/ml Penicillin/Streptomycin , 2mM L-Glutamine , 0 . 2mM β-mercaptoethanol , and 20% Fetal Bovine Serum ( FBS ) . Prior to immunolabeling , MEFs were cultured in reduced serum medium containing 0 . 5% FBS for 48 hours to induce primary cilia formation as previously described [38] . Cells were fixed in 4% paraformaldehyde and permeabilized with 0 . 1% Triton X-100 in PBS with 2% donkey serum , 0 . 02% sodium azide and 10 mg/ml bovine serum albumin ( BSA ) . Cells were labeled with anti-acetylated α-tubulin , 1:1000 ( cat# T-6793 , Sigma-Aldrich ) , anti-SmoN , 1:1000 ( gift from Dr . Matthew Scott , Stanford University ) . Sections from E10 . 5 neural tubes were immunolabeled with the following antibodies from Developmental Studies Hybridoma Bank ( University of Iowa , Iowa City , IA ) : anti-ShhN 1:1000 ( 5E1 ) , anti-FoxA2 1:1000 ( 74 . 5a5 ) , anti-Mnr2 1:1000 ( 81 . 5C10 ) , anti-Pax7 1:1000 ( Pax7 ) , and anti-Msx1+2 1:1000 ( 4G1 ) as previously described [38] . Trachea sections were labeled with anti-Gas8/DRC4 , 1:2000 [6] . All incubations and washes were carried out in PBS with 2% normal donkey serum , 0 . 02% sodium azide and 1% BSA . Primary antibody incubations were performed for 16–24 hours at 4°C and secondary antibody incubations were performed for 1 hour at room temperature . Secondary antibodies all from Thermo-Fisher Scientific include the following: Alexa Fluor-594 donkey anti-mouse ( cat# A21203 ) , Alexa Fluor-488 donkey anti-mouse ( cat# A21202 ) , Alexa Fluor-594 donkey anti-rabbit ( cat# A21207 ) , and Alexa Fluor-488 donkey anti-rabbit ( cat# A21206 ) . Nuclei were visualized by Hoechst nuclear stain . Coverslips were mounted using Immu-Mount ( cat# 9990402 , Thermo-Fisher Scientific ) . Fluorescence imaging was performed using a Nikon TE-2000U inverted microscope ( Melville , KY ) outfitted with a PerkinElmer UltraVIEW ERS 6FE-US spinning disk laser apparatus ( Shelton , CT ) and a Hamamatsu C9100 . DIC images of p14 trachea prepared for IF were used for length analysis . Images were captured with a 40x objective ( Plan-Fluor , 1 . 3NA ) . Length was measured manually by drawing a line from the tip of the cilium to the base using Volocity v6 . 3 . Gas8WT and Gas8GT MEFs were grown to confluency on 0 . 17mm coverslips and serum starved for 48 hours to induce ciliation . Cells were treated with 150nM Smoothened agonist ( SAG ) ( cat#566660 , CALBIOCHEM ) for 2 hours in low serum media to induce Hedgehog pathway activation and Smoothened translocation . Cells were fixed and stained as described in the immunofluorescence section and imaged by spinning disk confocal . Amount of Smoothened per cilia volume was measured using Volocity v6 . 3 software . Postnatal day 14 ( P14 mice were anesthetized and perfused with PBS followed by a perfusion of 2% glutaraldehyde in 0 . 1M cacodylate buffer pH 7 . 4 . Tracheas were extracted and fixed overnight at 4°C in 2% glutaraldehyde in 0 . 1M cacodylate buffer pH 7 . 4 . Samples were then washed thoroughly four times for 15 minutes each in 0 . 1M cacodylate Buffer pH 7 . 4 . A post fix in 1% OsO4 in 0 . 1M cacodylate buffer pH 7 . 4 was performed . Samples were washed two times for 10 minutes each in 0 . 1 M cacodylate pH 7 . 0 . Samples were then prepped in 1% tannic acid in 0 . 1M cacodylate Buffer pH 7 . 0; 30 minutes followed by 1% NaSO4 in 0 . 1M cacodylate Buffer pH 7 . 0; 5 minutes . Dehydrate the samples in 50% , 75% , and 95% at 4°C for 20 minutes each and finally 100% EtOH for 20 minutes; warm to RT° . Dehydrate samples totally with four washes of 100% EtOH 15 minutes each . Infiltrate the sample with Propylene Oxide for 30 minutes . Mix the EMbed 812 according to instructions from EMS and Infiltrate with 25% Embed in propylene oxide for 30 minutes , 50% for 40 minutes , 75% overnight , 100% for four hours , 100% for 1 hour and harden at 60°C . Samples were sliced at 90nm and imaged on a Phillips CM110 Electron Microscope . TEM averaging of doublets was performed by isolating individual doublets from cilia and importing the doublets into Photoshop CS5 . Individual doublets were aligned to a single template doublet and then averaged and flattened . TEMs were used to determine cilia orientation . Cilia orientation was determined by measuring the angle of central pairs by drawing a line across the central doublets and measuring the angle relative to the image . Each angle was normalized to the average ( or most common angle ) after setting the average angle to 0° . The frequency of angles in each image was measured and plotted . Brains of experimental mice were extracted , sliced in half to expose the ependymal of the lateral ventricles and placed in pre-warmed , pre-oxygenated artificial cerebrospinal fluid ( 125mM NaCl , 2 . 5mM KCl , 1 . 25mM NaH2PO4 , 2mM CaCl2 , 1mM MgCl2 , 25mM NaHCO3 , 25mM Glucose , pH 7 . 35 ) . Brains were placed on a Zeiss Axioskop microscope and imaged with a 5x objective ( Plan-Neofluor , 0 . 15NA ) and a 10x objective ( Fluor , 0 . 5NA ) using a Photometrics CoolSnap HQ CCD camera at 30fps . Red fluorescent latex beads ( cat# L3530-1mL , Sigma-Aldrich ) were diluted 1:100 from stock and 10μL of diluted beads was added to the ventricles . Bead tracking analysis was performed using the MTrack2 plugin in FIJI . Mice trachea were dissected out into fresh PBS and cut lengthwise into strips . Trachea were kept in warm media ( DMEM F/12 , 20% FBS , and Pen-Strep and allowed to adapt for 20 minutes in an environmental chamber ( 37°C , 45% relative humidity , and 5% CO2 ) before imaging with Differential Interference Contrast ( DIC ) . All high speed video was captured at 240fps using a modified Casio Exilim EX-ZR100 attached to a Nikon TE-200 using a 60x water objective ( Plan-Apo WI NA = 1 . 2 ) . Videos saved as quick time files were then extracted into individual frames using VirtualDub 1 . 10 . 4 software and all analysis was performed in ImageJ . Kymographs were created using Metamorph v6 . 1 . A line was drawn from the tip of the cilium to the base and kymographs were made from the results . To make the D198K mutation in Chlamydomonas , the DRC4-GFP plasmid [6] was used as template for PCR with the primers 5’-CAGTGCTGTGAGCCTGACG and 5’-AAACCAAAGCACCTTGAGCG to generate a 1483bp product that contains the restriction sites BclI and ClaI flanking the desired mutation site . The PCR product was cloned into pGEM-T-Easy ( cat# A1360 , Promega Corp ) to generate the plasmid pf2-Y1-A . This plasmid was further digested with KpnI and SpeI to removed repetitive DNA and subcloned into pBlueScript to generate the plasmid pf2-Y1-B . The D198K mutation was introduced into pf2-Y1-B using the primers 5’-GAAGATGCTGCGAGACaAaATGGAGCTGCGGAGAAAG-3’ and 5’-CTTCTACGACGCTCTGtTtTACCTCGACGCCTCTTTC-3’ and the QuickChange II kit ( Agilent Technologies ) to generate the plasmid pf2-Y1-C . After sequence verification by Genewiz , the pf2-Y1-C plasmid was digested with KpnI and SpeI and subcloned back into pf2-Y1-A to generate the plasmid pf2-Y1-D . The pf2-Y1-D plasmid was digested with BclI and ClaI to release the 1483 bp fragment now carrying the D198K mutation . This fragment was subcloned back into the original DRC4-GFP plasmid by Genewiz . The completed plasmid , DRC4-D198K-GFP , was linearized with EcoRI for transformation into the pf2-4 strain [6] . Transformants were screened as described above . RT-PCR confirmed that the D198K mutation was expressed in the rescued strains without any other sequence modification . Forward swimming velocity was recorded and measured as previously described [6] . For transformations with the control DRC4-GFP plasmid , rescued colonies were recovered at a frequency of 5–15% CRISPR/sgRNA target sequences were queried using the MIT CRISPR server . Three sites most proximal to the desired SNP change were selected to test nuclease efficiency . CRISPR1: 5’- CTTCTCCACAGCAGCGTTCA GGG-3’ ( reverse strand ) , CRISPR2: 5’-GGTGCTGGCCGCCTCCAACC TGG-3’ ( forward strand ) , CRISPR3: 5’-GACACAAGCGTTAATGCTGTGGG-3’ ( reverse strand ) . Pronuclear injections were performed with Cas9 mRNA ( 100 ng/ul ) , CRISPR3/sgRNA ( 50 ng/ul ) and ssODN ( 200 ng/ul ) . Efficiency of nuclease activity was assessed using a blastocyst assay . In brief , injected zygotes were cultured to the blastocyst stage and lysed to obtain genomic DNA . Genomic DNA was used in PCR and the amplicons ( 215 bp ) were resolved by heteroduplex mobility assay ( HMA ) . CRISPR3 was found to be most efficient and was used to generate the SNP edited mouse ( C57Bl/6 background ) . Injected zygotes were cultured to 2-cell stage in KSOM mixed with the NHEJ inhibitor SCR7 at a final concentration of 10 mM . The 2-cell stage embryos were transferred to psuedopregnant recipient female mice , which gave birth to 13 pups . The SNP was introduced with the help of a 154 nt single stranded oligo DNA ( ssODN ) HDR template . Since the PAM sequence ( CCC>Pro ) could not be modified without changing the amino acid , multiple silent changes were made in the protospacer ( sgRNA binding ) sequence ( indicated by small letters in the sequence below ) . These changes were made to eliminate the chances of the sgRNA binding to the repaired allele . The SNP change introduced a restriction enzyme recognition enzyme site ( BsmBI/Esp3I ) and the silent changes introduced two new restriction enzyme recognition sites ( BtgI and HaeII ) . HaeII sites were used to distinguish the wildtype and the modified alleles . Specific primers were also designed that can preferentially amplify the modified allele . HDR template ( ssODN ) 5’-GGCCCTGAACGCTGCTGTGGAGAAGAGAGAGGTTCAGTTCAATGAGGTGCTGGCCGTCTCCAACCTGGACCCCACgGCgcTgACGtTgGTGTCCCGCAAACTTGAGGTAGGTGCCCTCCTGTCCTGTGCTGTGGTACGCCTTCTTGGGTGGCAC-3’ . After the initial characterization of the F0 litter by PCR , the 215 bp amplicons were cloned , and selected individual clones were subjected to Sanger sequencing . Sequence analysis of the 13 pups revealed that 2 had complete knock-in of the edited/repair sequence , 1 pup had incorporated the silent changes but did not have the desired SNP change , and 1 pup had indels . F0 animals were bred with wildtype C57Bl/6 mice to test germline transmission of the desired alleles . All alleles were successfully transmitted through the germline , and the positive F1 animals were used to create homozygous and compound heterozygous F2 animals . Cilia length analysis , bead flow tracking , cilia orientation , and cilia beat frequency were tested with Student’s t-test and graphed in Microsoft Excel . Smoothened trafficking assay and Chlamydomonas swim speeds were tested by ANOVA followed by Student’s t-test with a Bonferroni correction and graphed in Microsoft Excel . All error is represented in Standard Error of Means ( SEM ) .	Growth-Arrest Specific 8 ( Gas8 ) is implicated in dual roles at both the primary cilium to regulate hedgehog signaling and in motile cilia to coordinate cilia movement . To investigate these roles in vivo , we created a Gas8 genetrap mutant mouse . Though no overt primary cilia phenotypes were evident in the Gas8 genetrap mutant mice , there were severe motility defects and the mice presented with Primary Ciliary Dyskinesia ( PCD ) like symptoms including situs inversus and hydrocephalus . We also identified two potential disease causing GAS8 missense variants ( A391V and E199K ) in humans . Utilizing CRISPR/Cas9 we generated a mouse to mimic the A391V allele . When we crossed the Gas8AV mutants with the Gas8GT mutant , the compound Gas8GT/AV heterozygous animals developed mild hydrocephalus . Rescue experiments using Chlamydomonas with mutations in the Gas8 homolog revealed only a modest decrease in swim velocity raising the possibility that the E199K allele is not pathogenic .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "microtubules", "pathogens", "alleles", "animal", "models", "respiratory", "system", "model", "organisms", "trachea", "cellular", "structures", "and", "organelles", "cytoskeleton", "r...	2016	Mutation of Growth Arrest Specific 8 Reveals a Role in Motile Cilia Function and Human Disease
Coronavirus replication takes place in the host cell cytoplasm and triggers inflammatory gene expression by poorly characterized mechanisms . To obtain more insight into the signals and molecular events that coordinate global host responses in the nucleus of coronavirus-infected cells , first , transcriptome dynamics was studied in human coronavirus 229E ( HCoV-229E ) -infected A549 and HuH7 cells , respectively , revealing a core signature of upregulated genes in these cells . Compared to treatment with the prototypical inflammatory cytokine interleukin ( IL ) -1 , HCoV-229E replication was found to attenuate the inducible activity of the transcription factor ( TF ) NF-κB and to restrict the nuclear concentration of NF-κB subunits by ( i ) an unusual mechanism involving partial degradation of IKKβ , NEMO and IκBα and ( ii ) upregulation of TNFAIP3 ( A20 ) , although constitutive IKK activity and basal TNFAIP3 expression levels were shown to be required for efficient virus replication . Second , we characterized actively transcribed genomic regions and enhancers in HCoV-229E-infected cells and systematically correlated the genome-wide gene expression changes with the recruitment of Ser5-phosphorylated RNA polymerase II and prototypical histone modifications ( H3K9ac , H3K36ac , H4K5ac , H3K27ac , H3K4me1 ) . The data revealed that , in HCoV-infected ( but not IL-1-treated ) cells , an extensive set of genes was activated without inducible p65 NF-κB being recruited . Furthermore , both HCoV-229E replication and IL-1 were shown to upregulate a small set of genes encoding immunomodulatory factors that bind p65 at promoters and require IKKβ activity and p65 for expression . Also , HCoV-229E and IL-1 activated a common set of 440 p65-bound enhancers that differed from another 992 HCoV-229E-specific enhancer regions by distinct TF-binding motif combinations . Taken together , the study shows that cytoplasmic RNA viruses fine-tune NF-κB signaling at multiple levels and profoundly reprogram the host cellular chromatin landscape , thereby orchestrating the timely coordinated expression of genes involved in multiple signaling , immunoregulatory and metabolic processes . CoV infect a wide range of host species and cell types , but only six CoV are known to cause disease in humans [1] . Four human coronaviruses , HCoV‑229E , HCoV‑OC43 , HCoV‑NL63 and HCoV‑HKU1 , are mainly associated with upper respiratory infections , while severe acute respiratory syndrome ( SARS ) -CoV and middle east respiratory syndrome ( MERS ) -CoV may cause serious pathology in the lower respiratory tract , with acute lung injury , respiratory failure and death occurring in a significant number of infected individuals [1] . Lung biopsies obtained from SARS patients or infected primates consistently showed increased expression levels of ( i ) proinflammatory cytokines , such as interleukin ( IL ) -1 , TNF and IL-6 , ( ii ) chemokines , such as IL-8 ( CXCL8 ) , IP-10 ( CXCL10 ) and MCP-1 ( CCL2 ) and ( iii ) several other NF-κB target genes , suggesting that CoV infection triggers prototypical innate immune reactions involving the upregulation of inflammatory genes [2–4] . Other reports revealed a more diverse and incoherent pattern of host cell gene expression in response to SARS-CoV , MERS-CoV and HCoV-229E , respectively [5–7] . To date , it remains unclear if these differences are a consequence of different CoV strains/isolates and cell types being used in the respective studies or result ( at least in part ) from additional confounding factors , such as paracrine effects of cytokines or chemokines secreted during infection . Studies that define ( i ) the gene sets that are expressed specifically in CoV-infected but not bystander cells and ( ii ) the signaling pathways triggered in response to viral replication in these infected cells are still lacking [8] . Also , it remains to be studied if changes in host cellular mRNA levels in infected cells result from differential transcription rates of specific genes , altered mRNA stability , or both mechanisms . Transcriptional gene regulation in response to pathogens involves the activation of a cascade of signal-regulated events that trigger de novo recruitment of RNA polymerase II or the release of paused RNA pol II at transcriptional start sites ( TSS ) [9] . Ongoing transcription requires phosphorylation of RNA pol II at the C-terminal domain ( CTD ) heptapeptide repeats by CTD kinases . This modification marks the transition from the preinitiation to the initiation complex and is also characteristic for pausing of catalytically active pol II 40–60 nucleotides downstream from the TSS [10] . Accordingly , P ( S5 ) -pol II levels are high at the 5’ end of transcribed genes and decrease in a gene-specific manner towards the 3’ end during productive elongation within each cycle of pol II-mediated mRNA synthesis [11] . A potent signal-dependent TF that regulates transcription in innate immunity is NF-κB , a ( hetero ) dimer composed of the subunits p65 , p105/p50 or p100/p52 , c-Rel or RelB [12] . These subunits are retained in the cytoplasm by inhibitor of NF-κB ( IκB ) proteins . In the canonical NF-κB pathway , phosphorylation by IκB kinases ( IKKs ) and subsequent ubiquitylation and proteasomal degradation of IκBs liberates p65 subunits . In the non-canonical pathways , p100 or p105 are proteolytically processed to generate free p52 or p50 , respectively [13] . Liberated NF-κB subunits then translocate to the nucleus , where their DNA binding and activity is modulated in a cell type- and stimulus-specific manner , a complex process that is mainly controlled by the chromatin accessibility of NF-κB motifs , post-translational modifications and cooperative interaction ( s ) with other TFs or coregulators such as histone acetyltransferases ( HATs ) [14] . Transcriptional processes often depend on a highly abundant array of gene-regulatory elements in the human genome , called enhancers [15] . Within enhancer regions , histone H3K4me1 and H3K27ac are strongly associated with combinations of cis-elements that primarily encode binding motifs for lineage-determining TFs , but some enhancers also recruit stimulus-specific inducible TFs [16] . Active enhancers can be distinguished by the presence of regulated histone acetylation at H3K27 . Information on possible cooperative activities of TFs in enhancer-dependent gene regulatory networks may be inferred from the motif architecture of the underlying cis-elements [15 , 17] . While enhancers are known to represent binding platforms for multiple TFs , the specific relationships between cooperative TF binding , enhancer selection and subsequent patterns of gene regulation in response to external cues including viral infections are still not very well understood [14 , 16] . In the current study , we investigated , in a systematic manner , the dynamic gene expression profile in cells infected with HCoV-229E and compared these changes to those induced by the major inflammatory cytokine IL-1 . We also assessed the chromatin status of HCoV-229E-infected cells at a genome-wide level . Our data suggest that cytoplasmic RNA viruses may exert profound effects on cellular chromatin structure and gene expression , involving attenuated nuclear regulation of the NF-κB pathway and activation of an extensive set of genomic enhancers . To analyze HCoV-229E-induced host cell gene responses in a systematic manner , we first conducted a series of transcriptome-wide studies using the A549 lung epithelial carcinoma cell model . To detect transcriptional changes caused by viral replication , the cells were either infected at low multiplicity with ( replication-competent ) HCoV-229E ( MOI 0 . 001 ) or inoculated with replication-incompetent ( heat- or UV-inactivated ) HCoV-229E . At 16 and 48 h post infection ( p . i . ) , increasing amounts of HCoV-229E RNA ( as assessed by RT-PCR using primer pairs specific for the nsp8 and spike protein coding regions , respectively ) were detectable in HCoV-229E-infected cells , but not in control cells inoculated with heat- or UV-inactivated HCoV-229E ( S1A Fig ) . To determine the impact of HCoV-229E infection on the entire host cell transcriptome , we used Agilent microarrays containing 60 , 000 probes covering annotated genes and non-coding RNAs . Infection with HCoV-229E was found to cause differential expression of 108 genes at 16 h p . i . and 239 genes at 48 h p . i . ( Fig 1A , upper panels and S1 Table ) . A nearly identical set of differentially expressed genes was identified in control cells inoculated with heat-inactivated virus , that is , in the absence of ( detectable ) viral replication ( Fig 1A , lower left and middle panels ) , suggesting that the observed differential gene expression was caused to a significant extent by soluble factors present in the viral stock or interactions between cells and non-infectious virus particles . However , 2 days after infection we were also able to identify a specific set of 26 upregulated and 11 downregulated genes that were differentially regulated only in cells infected with replication-competent HCoV-229E but not in cells inoculated with inactivated virus ( Fig 1A , lower right panel ) . In a next set of experiments , we used immunofluorescence analysis of HCoV-229E nucleocapsid ( N ) protein expression to ( i ) monitor HCoV-229E infection and spread in A549 cells ( S1B Fig ) and ( ii ) characterize at the single-cell level the gene expression changes in infected as opposed to adjacent non-infected cells present in the same or separate culture ( s ) . Laser microdissection coupled with immunofluorescence was used to excise ( i ) HCoV-229E-infected cells , ( ii ) cells immediately adjacent to infected cells , ( iii ) cells located at least 150 μm away from infected cells ( "distant cells" ) and ( iv ) cells from a separate , non-infected cell culture ( for details , see Fig 1B and S1B Fig ) . Two independent series of microarray studies revealed almost no changes of gene expression in adjacent , distant and uninfected cell populations , while infected cells showed a total of 99 up- and 53 down-regulated genes , respectively ( Fig 1C and S2 Table ) . Overrepresentation ( ORA ) and gene set enrichment analyses ( GSEA ) for the genes identified in the time-course experiment ( Fig 1A ) and the laser-microdissected samples ( Fig 1B and 1C ) across all pair-wise comparisons identified the KEGG pathways 04141 ( protein processing in the endoplasmic reticulum ) and 03013 ( RNA transport ) as the pathways that were most significantly affected by HCoV-229E infection ( S2A Fig ) . Visualization of the mRNA expression values in KEGG 04141 and 03013 pathway maps illustrates that many mRNAs are induced or suppressed at several crucial nodes within these pathways ( S2B Fig ) . The initially identified set of 37 genes regulated by replicating virus only ( Fig 1A ) contained several strongly upregulated genes of these pathways and included transcription factors ( EGR1 , DDIT3 ( also called CHOP , GADD153 or C/EBPzeta ) , ATF3 , KLF6 , ZNF165 , BHLHE41 , ZBTB43 ) , cytokines and stress signaling proteins ( IL6 , GADD45A/DDIT1 , TNFAIP3/A20 ) , phosphatases ( DUSP1 ) , and further factors involved in ER stress ( DNAJB9/ERDJ4 , NEURL3 , HERPUD1 , HSPA5/GRP78/BIP ) or regulators of translation ( PPP1R15A/GADD34 , HSPA1A/HSP701A , HSPA1B/HSP701B ) . Fig 1D summarizes the observed changes in the expression levels of these genes and shows that the differential expression is specific for ( and restricted to ) infected cell populations . The upregulation of selected genes was confirmed by independent RT-qPCR experiments ( Fig 1E ) . Taken together , these data suggest that HCoV-229E replication regulates complex sets of genes involved in transcription , stress responses , cytokine signaling and nucleotide metabolism . While A549 cells are a suitable model for studying pathology caused by HCoV-229E in lung epithelial cells , their low infection efficiency with HCoV-229E precluded in-depth molecular analyses requiring very large numbers of infected cells . We , therefore , decided to use HuH7 cells , a widely used human hepatoma cell line that is known to support efficient HCoV-229E replication and infectious particle formation , for subsequent high-resolution analyses at the genome-wide level [18] . As a control , we treated HuH7 cells with IL-1 , which allowed us to compare HCoV-229E-linked gene signatures with those of a well characterized cytokine that regulates a broad range of inflammatory genes through stress kinase or NF-κB pathways . HuH7 cells were either infected with HCoV-229E ( MOI = 1 ) or were treated with IL-1 for 1 h and subsequently used for transcriptome analyses . We first asked if the HCoV-229E-linked gene signature observed in A549 cells could also be confirmed in HuH7 cells . We found that 19 out of 26 genes upregulated by HCoV-229E in A549 cells were also upregulated in HuH7 cells , defining a cell type-independent set of virus-induced host cell target genes ( Fig 2A ) . The genes downregulated in A549 cells were not expressed in HuH7 cells ( Fig 2A ) . The kinetic regulation of seven of these genes and their specific induction in cells productively infected with HCoV-229E but not in cells incubated with heat-inactivated virus was validated by RT-qPCR ( S3A Fig ) . IL-1 induced six of these genes ( ATF3 , DUSP1 , TNFAIP3 , KLF6 , PPP1R15A , and EGR1 ) indicating that HCoV-229E shares pathways with this cytokine but activates additional specific sets of genes ( Fig 2A ) . At the genome-wide level , HCoV-229E was found to upregulate 1 , 073 genes and to downregulate 717 genes , whereas IL-1 induced 492 genes and downregulated 95 genes by at least 2-fold on average in repeated microarray studies ( Fig 2B , Venn diagrams ) . The left heatmap in Fig 2B shows all 61 jointly regulated genes which includes well-characterized inflammatory genes ( e . g . IL8 , CXCL1 , CXCL2 , CXCL5 , CXCL6 , ICAM1 ) as well as signaling proteins regulated during inflammatory cell activation ( e . g . DUSP1 , DUSP8 , DUSP10 , EGR1 , c-JUN , NFKBIA , NFKBIAZ , TNFAIP3 , MAP3K8 ( TPL-2/COT ) , ZC3H12A ( MCPIP1 ) ) . Despite these similarities we observed a substantial number of genes which are upregulated by HCoV-229E , but not by IL-1 . The right heatmap in Fig 2B shows the top 50 genes that are specifically upregulated by HCoV-229E . These genes were consistently induced by HCoV-229E in all four individual microarray experiments and were not regulated by heat-inactivated virus ( S3B Fig ) . The differential regulation patterns in response to IL-1 or HCoV-229E were further corroborated by analyzing mRNA expression of representative genes by RT-qPCR ( S4A Fig ) . These data clearly show the major impact of HCoV-229E replication on inflammatory pathways but also reveal much broader virus-induced effects on cellular gene expression . To determine if HCoV-229E-induced genes are enriched for specific biological pathways , the complete microarray data sets were analyzed by GSEA . Fig 2C shows the top ten pathways with significant enrichment of regulated genes and the overall direction of regulation of the entire gene sets . The complete list of expression values and the components for all KEGG pathways of Fig 2C are shown in S3 Table . Fig 2D highlights the similarities and differences of HCoV-229E- versus IL-1-regulated changes in mRNA expression of individual genes in six of these pathways , all of which representing intracellular processes . In four KEGG pathways ( 04060 , 04141 , 04630 , 04010 ) , HCoV-229E replication mainly caused an upregulation of genes ( e . g . CXCL2 , HERPUD1 , DUSP1 , FUT1 ) , while in pathway 00190 the virus infection was mainly associated with a downregulation of genes . In comparison , and consistent with its role as a MAPK pathway-activating cytokine , IL-1 was found to primarily upregulate genes in the pathways 04060 ( cytokine-cytokine receptor interaction ) and 04010 ( MAPK signaling pathway ) . Together , the data shown in Fig 1 and Fig 2 define a core set of genes that are consistently induced by HCoV-229E in both A549 and HuH7 cells . Most of these genes are specifically regulated by HCoV-229E , while another subset of genes is shared with IL-1 . To investigate if and to what extent HCoV-229E affects host cell mRNA expression at the transcriptional level , we analyzed the Ser5 phosphorylated form of RNA polymerase II ( P ( S5 ) -pol II ) that accumulates during formation of the transcription initiation complex near the transcriptional start site ( TSS ) [11] . In order to investigate the recruitment of P ( S5 ) -pol II to the promoter of the HCoV-inducible IL8 gene , we performed ChIP experiments which showed a time-dependent recruitment with a maximum at 24 h p . i . of RNA polymerase II and its phosphorylated form to the promoter/TSS region of the IL8 gene ( S4B Fig ) . This condition was chosen for ChIP-seq experiments using the P ( S5 ) -pol II antibody . The data were then used to assemble averaged profiles of P ( S5 ) -pol II recruitment for groups of up- and downregulated genes that belong to the KEGG pathways and which were affected by HCoV-229E replication and/or IL-1 treatment . Upregulated genes belonging to the cytokine pathway KEGG_04060 showed increased recruitment of P ( S5 ) -pol II by both , HCoV-229E and IL-1 , and a prototypical decline of P ( S5 ) -pol II towards the 3’ ends of the genes ( Fig 3A ) . In this group of genes , IL-1-mediated P ( S5 ) -pol II signals were stronger than the CoV-mediated effects ( Fig 3A ) , in line with the stronger induction of mRNA expression of genes such as IL8 or CXCL2 ( Fig 2D ) . Upregulated genes belonging to the ER stress pathway KEGG_04141 showed no regulation by IL-1 , but strong upregulation of P ( S5 ) -pol II signals by HCoV-229E ( Fig 3A ) , again correlating with mRNA expression ( Fig 2D ) . For downregulated genes of KEGG pathways 04060 or 04141 , we found only low levels of constitutive and unregulated P ( S5 ) -pol II occupancy ( Fig 3A ) . This observation suggests that altered mRNA levels of these genes may be regulated by post-transcriptional mechanisms independently from a low level of basal transcription . A meta-gene analysis of P ( S5 ) -pol II recruitment across the entire transcriptome showed a good correlation of P ( S5 ) -pol II recruitment with mRNA expression . For 194 genes shown to be strongly ( ≥ 4-fold ) upregulated by HCoV-229E and for 59 genes shown to be regulated by both IL-1 and HCoV-229E , an inducible P ( S5 ) -pol II recruitment could be confirmed ( Fig 3B ) . Collectively , these data indicate that HCoV-229E induces a genome-wide transcriptional response to upregulate a broad range of cellular genes . To examine histone modifications at genomic regions with increased recruitment of RNA polymerase II in HCoV-229E-infected cells , we analyzed the acetylation patterns of H3K9 , H3K36 and H4K5 . All three modifications are highly enriched at promoters , TSS and gene bodies , where they function to open nucleosomal DNA and additionally provide binding sites for multiple coactivators [19] . Fig 3C shows the histone acetylation pattern for 8 genes whose transcription is specifically induced by HCoV-229E as indicated by the P ( S5 ) -pol II recruitment pattern . At these genes , basal H3K9ac was found to be high and only regulated for two genes ( ANKRD1 , FICD ) . At all genes , the basal H3K36ac level was increased by HCoV-229E but not by IL-1 . H4K5ac levels were low in uninfected and IL-1 treated cells , but were consistently increased by HCoV-229E . These profiles show that H4K5ac and , for most genes , H3K36ac are characteristic inducible histone marks of HCoV-229E-induced genes . Fig 3D shows examples of read count distribution for eight genes , which are induced by both HCoV-229E and IL-1 . These genes show high levels of non-regulated H3K9ac at their promoters and five of them show inducible H3K36ac and H4K5ac ( CXCL2 , CXCL1 , TNFAIP3 , EGR1 , IL8 ) . As shown in Fig 2B ( left heatmap ) , some genes are stronger induced at the mRNA level by IL-1 ( e . g . CXCL2 , CXCL1 , EGR1 , IL8 ) , correlating well with the observed stronger P ( S5 ) -pol II recruitment to these genes . Quantification of read counts confirmed similar inducible changes in histone modifications and P ( S5 ) -pol II recruitment for the upregulated genes in KEGG pathway 04060 and KEGG pathway 04141 , but not for the downregulated genes ( S4C Fig ) . Taken together , the data show that HCoV-229E activates specific sets of host cell genes at the transcriptional level by increasing phosphorylation of RNA pol II and acetylation of H3K36 and H4K5 at the promoter and TSS regions . The gene set that was transcriptionally induced by both HCoV-229E and IL-1 contained several well-characterized NF-κB target genes , such as IL8 , CXCL2 and TNFAIP3 . This raised the question of whether HCoV-229E elicits non-canonical NF-κB signaling or activates the canonical NF-κB pathway , the latter involving IκBα degradation [20] . To address this question , cells were infected with HCoV-229E for different periods and signaling events were analyzed by Western blot experiments in comparison to cells that were stimulated for 1 h with IL-1 . This cytokine caused strong phosphorylation ( mean fold 7 . 4 +/- 4 . 6 s . d . ) and nearly complete degradation of IκBα ( mean fold 0 . 10 +/- 0 . 073 s . d . ) ( Fig 4A and 4B ) . IκBα protein levels were also reduced at 24 h after infection with HCoV-229E but to a significantly lesser extent ( mean fold 0 . 34 +/- 0 . 12 s . d . ) and without a concomitant virus-mediated increase in IκBα phosphorylation at this time point ( Fig 4A and 4B ) . A transient early induction of Ser32 phosphorylation of IκBα by HCoV-229E after 3 h of infection was only seen inconsistently in two out of six experiments ( Fig 4A and 4B ) . IL-1 did not affect the expression of the core IKK complex catalytic subunits IKKα and IKKβ or of the regulatory subunit NEMO , but the cytokine strongly activated the phosphorylation of IKKβ ( mean fold 12 +/- 6 . 3 s . d . ) and moderately that of IKKα ( mean fold 2 . 1 +/- 0 . 75 s . d . ) ( Fig 4A and 4B ) . In contrast , HCoV-229E infection led to decreased levels of P- IKKβ ( mean fold 0 . 52 +/- 0 . 26 s . d . ) , IKKβ ( mean fold 0 . 53 +/- 0 . 13 s . d . ) and NEMO ( mean fold 0 . 48 +/- 0 . 14 s . d . ) but remained without effects on P-IKKα / IKKα ( Fig 4A and 4B ) . Additionally , HCoV-229E was found to induce the expression of TNFAIP3 ( A20 ) ( mean fold 1 . 6 +/- 0 . 60 s . d . ) , a potent cytosolic inhibitor of the IKK complex ( Fig 4A and 4B ) [21] . Another IKK substrate is S536 of p65 NF-κB [22] . This modification was reduced upon HCoV-229E infection ( mean fold 0 . 54 +/- 0 . 33 s . d . ) but triggered by IL-1 ( mean fold 3 . 2 +/- 1 . 0 s . d . ) ( Fig 4A and 4B ) . Together , these data suggest that HCoV-229E replication on the one hand inefficiently triggers NF-κB activity by an unusual mechanism involving partial and selective IκBα degradation , but on the other hand dampens NF-κB activity by leading to reduced expression of IKK core complex subunits . This negative regulation occurred at the translational level , as the mRNAs encoding the three IKK subunits and IκBα were weakly induced by virus ( S5A Fig ) . Besides induction of TNFAIP3 , these effects may involve ( negative ) regulatory steps at the level of translation , as HCoV-229E upregulated the phosphorylation of the eukaryotic translation initiation factor subunit eIF2α at S51 ( mean 2 . 0 +/- 0 . 62 s . d . ) , corroborating data obtained previously for the avian infectious bronchitis virus ( IBV ) ( Fig 4A and 4B ) [23] . Next , we analyzed the amounts of NF-κB subunits in soluble ( N1 ) and chromatin ( N2 ) nuclear fractions in response to HCoV-229E infection or IL-1 . Consistent with previously published data , IL-1 strongly increased the concentrations of p50 , p52 , p65 and c-Rel NF-κB subunits by 5 to 10-fold in the soluble N1 fraction , whereas p65 was the most prominently increased subunit stably associated with the N2 chromatin fraction ( Fig 5A and 5B , S6A Fig ) [24 , 25] . In HCoV-229E-infected cells , a modest 1 . 5 to 3-fold increase of p65 was observed in the N1 fractions , but also of p50 and p52 which are characteristic for the activation of the non-canonical pathway ( Fig 5A and 5B ) . This experiment also showed that the p65 protein was the only subunit that was detectably increased in the N2 chromatin fraction after virus infection . These small changes were significantly different from basal nuclear levels of NF-κB subunits and were not seen when inactive virus was used ( Fig 5A and 5B ) . We also detected the HCoV-229E N protein in both nuclear compartments , confirming earlier studies on a partial nuclear localization of CoV N proteins ( Fig 5A ) ( reviewed in [26] ) . Collectively , the data shown in Fig 4 , Fig 5A and 5B and S6A Fig demonstrate that HCoV-229E infection causes an attenuated nuclear NF-κB response involving both the non-canonical and canonical NF-κB pathways . To answer the question of whether HCoV-229E selectively regulates nuclear functions of the N2 fraction-associated p65 NF-κB , we analyzed the chromatin recruitment of p65 NF-κB to distinct genomic loci by ChIP-seq experiments , focusing on the promoters of sixteen HCoV-229E- or IL-1-regulated genes for which virus-mediated changes in P ( S5 ) -pol II occupancy and histone acetylation pattern had been identified in the experiments reported above ( see Fig 3C and 3D ) . For genes that were activated by HCoV-229E but not by IL-1 , we only observed relatively low and unregulated p65 signals ( Fig 5C ) . In contrast , IL-1 strongly stimulated the recruitment of p65 to 5 of the 8 genomic regions that contained genes shown to be regulated at the mRNA and transcriptional level by both HCoV-229E and IL-1 ( see above and Fig 5D ) whereas , for HCoV-229E-infected cells , only a moderate increase of p65 binding to the CXCL2 , CXCL1 , TNFAIP3 and IL8 loci could be detected ( Fig 5D ) . The latter pattern of p65-dependent transcriptional regulation was also seen for the NFKBIA gene which encodes the highly regulated IκBα protein ( S5A–S5C Fig ) . The functional relevance of the NF-κB system for the transcriptional regulation of virus-induced ( CHAC1 , ANKRD1 ) versus virus- and cytokine-induced genes ( IL8 , CXCL2 ) was tested using the IKKβ inhibitor PHA-408 [27] . This compound partially blocked IL-1- and HCoV-229E-induced IκBα degradation in HuH7 cells ( Fig 5E ) , suggesting that IKK activation also contributes to HCoV-229E-mediated NF-κB activation . Accordingly , PHA-408 also inhibited the virus-inducible mRNA expression of all 4 genes as well as the upregulation of IL8 and CXCL2 mRNAs by IL-1 ( Fig 5F ) . ChIP-PCR experiments confirmed the virus- versus IL-1-specific p65 and P ( S5 ) -pol II recruitment patterns presented above ( Fig 5G ) . PHA-408 suppressed P ( S5 ) -pol II recruitment to the IL8 , CXCL2 , CHAC1 and ANKRD1 loci and virus- or IL-1-inducible p65 recruitment to the IL8 and CXCL2 promoters ( Fig 5G ) . The IKKβ inhibitor also partially suppressed viral replication by about 30% as assessed by measuring N protein levels ( Fig 5E ) . This result is in line with a positive role of IKKs for viral replication that may proceed through substrates unrelated to NF-κB as shown for a few other RNA viruses [20] . However , the inhibitory effects of PHA-408 at the mRNA and chromatin levels were stronger than the suppression of viral protein synthesis , suggesting that the IKK-NF-κB pathway may also exert pro-viral functions ( Fig 5E–5G ) . To address the function of NF-κB by a complementary experimental approach , p65 expression was downregulated by RNAi in HeLa cells , which unlike HuH7 cells are suitable for both efficient HCoV-229E infection ( shown in S6B and S6C Fig ) and transfection of shRNAs . Almost complete reduction of p65 suppressed steady-state protein levels of the p65 target gene IκBα but did not affect the IL-1- or virus-mediated IκBα degradation or viral N protein synthesis ( Fig 6A ) . Knockdown of p65 suppressed virus- or IL-1-induced mRNA expression as well as P ( S5 ) -pol II and p65 recruitment to the NF-κB target genes IL8 and CXCL2 ( Fig 6B and 6C ) . Moreover , decreased viral titers in these cells suggested that p65 is required for a step subsequent to viral protein synthesis ( Fig 6D ) . Further evidence for a fundamental role of the IKK-NF-κB system in the HCoV-229E host response was derived from TNFAIP3 knockdown experiments performed under identical conditions . Reduction of TNFAIP3 protein resulted in a 70% inhibition of the infection rate as assessed by N protein levels ( Fig 6E ) . At the same time , basal mRNA expression levels of IL8 and ANKRD1 were strongly increased ( Fig 6F ) . As a result , the fold regulation of these genes by HCoV-229E or IL-1 was almost completely lost ( Fig 6F ) . Basal CXCL2 and CHAC1 levels were less affected by TNFAIP3 knockdown , however , signal-mediated regulation by HCoV-229E or IL-1 was also inhibited ( Fig 6F ) . Collectively , the data shown in Figs 5 and 6 suggest important functional roles for IKKβ , TNFAIP3 and p65 in virus-mediated changes of host cell gene expression and viral replication and reveal that HCoV-229E infection also restricts the maximal activity of the IKK complex . This occurs by inversely balancing the levels of positive regulators of the NF-κB pathway ( IKKβ , NEMO ) versus negative cytosolic regulators of the IKK complex ( TNFAIP3 , IκBα ) . This fine-tuning allows for some basal NF-κB activity that supports viral replication but limits high nuclear concentration of NF-κB subunits avoiding excessive transcription of inflammatory and other potentially antiviral host cell genes . The ChIP-seq measurements also allowed us to address the genome-wide distributions of HCoV-229E- versus IL-1-regulated chromatin changes . In all instances , we detected significant overlaps between P ( S5 ) -pol II recruitment and histone modifications after virus infection or cytokine treatment ( S7 Fig ) . With respect to p65 NF-κB , 6 , 977 peaks were detected , of which 353 were regulated by IL-1 , 82 were regulated by both IL-1 and HCoV-229E , and 50 by HCoV-229E only ( Fig 7A ) . Quantification of p65 binding revealed much stronger signals for genome regions that were regulated specifically by IL-1 alone or regulated consistently by both IL-1 and HCoV-229E , whereas p65 peaks induced by HCoV-229E but not IL-1 had comparably lower signal intensities ( Fig 7B ) . The majority of constitutive p65 peaks were localized close to promoter-TSS regions , whereas most of the regulated p65 peaks were located at more distant positions ( > 20 kb away from the next TSS ) , suggesting that p65 is involved in occupying distant enhancer structures in these cases ( Fig 7C ) . This conclusion is also supported by an increased acetylation of histones H3 and H4 in genome regions with increased p65 binding ( Fig 7D ) . An example of such an IL-1- and HCoV-229E-regulated intergenic region , supported by inducible p65 recruitment to p65 DNA motifs combined with histone modifications in the same genomic region , is provided in Fig 7E . Taken together , the data lead us to conclude that , compared to the potent NF-κB activator IL-1 , HCoV-229E causes a weak activation of the p65 NF-κB pathway at the chromatin level and regulates recruitment of this transcription factor to a small but specific group of mainly non-coding genomic regions . These results raised the question of whether HCoV-229E infection leads to a more general activation , possibly affecting the enhancer repertoire across the entire HuH7 genome . To address this possibility , we determined the total number of active enhancers as assessed by the co-ocurrence of H3K4me1 and H3K27ac [28] . We found 36 , 107 regions carrying both histone marks ( Fig 8A , S8A Fig ) . In total , 2 , 736 ( 7 . 6% ) of all enhancers were identified to be regulated by IL-1 and/or HCoV-229E using a 2-fold change in H3K27ac as a threshold level ( Fig 8A , sum of Venn diagrams ) . 1 , 432 and 1 , 744 enhancers , respectively , were specifically regulated by either HCoV-229E or IL-1 , whereas 440 enhancers ( 16% ) were found to be activated by both stimuli ( Fig 8A ) . Meta-profiling showed that all regulated enhancers had low levels of H3K27ac prior to the activation and responded strongly with a symmetrical broadening of the H3K27ac peak following activation by IL-1 or HCoV-229E ( Fig 8B ) . Compared to IL-1-specific enhancers , the HCoV-229E-specific enhancers were found to cover smaller regions of the genomic DNA , suggesting that these regions are occupied by a smaller number of DNA-binding factors and cofactors ( Fig 8B ) . Quantitative assessment of histone modifications corroborated the specific increases of H3K27ac for all three groups of enhancers and little regulation of H3K4me1 ( Fig 8C ) . It should be noted that all enhancer regions also showed additional regulation of H3K36ac , H4K5ac and H3K9ac . As these lysines are modified by different HATs such as CBP/p300 or GCN5 [29–31] , this observation suggests that different HAT enzymes are recruited to and modify these structures ( Fig 8C ) . Fig 8D provides 3 examples for each of the different enhancer groups . Notably , there was no virus-regulated p65 binding to the 992 HCoV-229E-specific or any of the IL-1-specific enhancers , suggesting that the vast majority of HCoV-229E enhancers is controlled by transcription factor combinations that differ from those employed by IL-1-regulated enhancers which always showed inducible p65 recruitment ( Fig 8C and 8D ) . In line with this hypothesis , de novo motif searches revealed a complex assembly of cis-elements that were characteristic for each of the 3 regulated groups of enhancers ( Fig 8E ) . These structures matched to a large number of known transcription factor-binding sites ( Fig 8E ) . While the HCoV-229E-specific enhancers are predicted to bind various combinations of AP-1 subunits ( Jun , JunD , FOSL2 , ATF4 ) and C/EBPs , the IL-1 regulated enhancers are enriched for NF-κB and alternative combinations of AP-1 sites ( Fig 8E ) . Additionally , a variety of other TFs , such as FOXC2 , FOXL1 , FOXB1 , DDIT3 and others , can bind to these sites in different combinations ( Fig 8E ) . Thus , the three groups of enhancers provide composite binding sites for stimulus-specific combinations of TFs . Accordingly , GO classifications of all annotated genes next to the HCoV-229E-specific enhancers revealed an enrichment for catabolic and ER stress pathways , whereas genes next to enhancers regulated by both IL-1 and HCoV-229E or by IL-1 alone are highly enriched for multiple immunoregulatory pathways ( S8B Fig ) . ChIP-PCR experiments using PHA-408 confirmed that the two enhancers on Chr . 1 and on Chr . 10 shown in Fig 8D differ in their virus- or IL-1-mediated p65 recruitment and sensitivity to IKKβ inhibition ( S9 Fig ) . In conclusion , these data reveal that HCoV-229E infection activates specific histone modification patterns in a large number of genomic regions that are likely to orchestrate the genome wide gene response to this virus . Apart from a number of shared enhancer regions , this pattern is fundamentally different from the prototypical NF-κB p65-driven IL-1 enhancer structures and provides an explanation for the large divergent sets of virus-specific genes ( summarized in Fig 9A ) . Emerging or re-emerging RNA viruses cause significant morbidity and mortality in humans and animals [1 , 32 , 33] . With few exceptions , these RNA viruses complete their viral life cycle in the cytoplasm but need to reprogram and/or adjust specific biosynthetic and other pathways required for viral replication and production of infectious virus progeny according to their specific requirements [34 , 35] . Simultaneously , intrinsic host cell defense systems are activated and , in many cases , counteracted by a multitude of strategies that viruses developed during viral evolution [32] . Despite significant scientific progress in the past few years , the complete spectrum of interactions between plus-strand RNA viruses and their hosts is far from being understood , in particular if it comes to understanding complex regulatory networks at a genome-wide level [8 , 36] . Here , we provide a high-resolution view of both transcriptome and chromatin changes in cells infected with the alphacoronavirus HCoV-229E , a virus that causes mild upper respiratory tract disease and is efficiently transmitted in the human population , indicating excellent adaption to its human host . The sophisticated fine-tuning of IKK-NF-κB activity and the identification of thousands of coronavirus-specific enhancers reported in this study provides important insight into the enormous complexity by which an RNA virus resets the host cell chromatin response to change expression levels of many hundreds of host cell genes and , at the same time , stimulates future studies into the molecular interplay between corona- and other RNA viruses and their hosts . There is conflicting evidence regarding the activation or importance of the NF-κB pathway for human CoV [4 , 37–40] . In our view , this may be attributed to the methods used to study NF-κB activation or the ectopic expression of viral proteins rather than infection with intact virus but may also be linked to biological differences between CoV that belong to different virus species/genera and are known to infect different hosts . Here , we provide the first in-depth and high-resolution analysis of endogenous NF-κB activation and p65 DNA binding in virus-infected cells . In addition , we compared the HCoV-229E-triggered responses directly to those elicited by IL-1 , a potent inducer of the classical NF-κB pathway . Our data suggest two different consequences of HCoV-229E infection for NF-κB activity: On the one hand the viral infection leads to NF-κB activation , as seen by IκBα degradation , p65 chromatin recruitment and the inducible transcription of NF-κB target genes . This elevated NF-κB activity has a pro-viral function , as evidenced by impaired virus replication after IKK inhibition or p65 knockdown . Mechanistically , elevated NF-κB activity also allows for synthesis of the A20 protein which is required for efficient virus replication . These findings are in line with the recent observation that deletion of the A20-encoding gene protects mice from influenza A virus infection [41 , 42] . On the other hand , we also observed that HCoV-229E has acquired specific mechanisms to dampen NF-κB activity , as evidenced by lower levels of p65 chromatin recruitment and comparably low induction of gene expression . HCoV-229E infection failed to induce IKK phosphorylation and only caused insufficient phosphorylation and incomplete degradation of IκBα . In addition , the steady state levels of NEMO and IKKβ decreased during infection which could be due to several mechanisms including phosphorylation of eIF2α to shut off protein de novo synthesis [43 , 44] . The simultaneous induction and restriction of the NF-κB response is typically seen for many different RNA viruses [20] . This mechanism is visualized in Fig 9B and presumably serves to ensure a corridor of elevated NF-κB activity allowing to support viral replication , while also preventing full activity of the transcription factor that counteracts virus infection by enabling the synthesis of anti-viral mediators such as IFN or other cytokines . This could also explain the pro-viral function of A20 , as it prevents exaggerated anti-viral gene expression and NF-κB activity . At the level of gene expression we found that the HCoV-229E-regulated gene pattern is overlapping with the genes affected by the zoonotic virus SARS-CoV including IL-6 and several chemokines [4 , 6] . To our knowledge , none of these earlier studies addressed the question of whether the observed changes of host cell gene expression were caused by the replicating virus itself or ( partly or largely ) by factors potentially present in the virus preparation used to infect/inoculate the cells to be used for subsequent analyses . Using appropriate controls and two different human cell lines , we were able to identify a common set of genes that is specifically induced in productively infected cells with ongoing HCoV-229E genome replication and expression . This gene set contains factors involved in the ER stress response , consistent with an earlier study showing that some genes from this group ( PERK , ATF3/4 , ERO1α ) are induced at the protein level in IBV-infected avian cells [23 , 45 , 46] . Our mRNA expression and bioinformatics analysis identified a large number of additional genes of the ER stress response that are induced by HCoV-229E , such as CHAC1 , an enzyme that degrades glutathione [47 , 48] and others ( FICD , HERPUD1 , DNAJB9/ERDJ4 ) functioning as regulators of AMPylation , ubiquitylation or co-chaperones , respectively [49–51] . Moreover , HCoV-229E was found to upregulate transcription factors ( TFs ) with known functions such as EGR1 and c-JUN , but also poorly characterized TFs such as zinc finger ( ZNF ) 165 , basic helix-loop-helix-type transcription factors ( BHLHE40/41 ) or zinc finger and BTB domain-containing 43 ( ZBTB43 ) [52 , 53] . In an attempt to dissect virus-specific effects from well-characterized inflammatory stimuli , we performed a detailed comparison of HCoV-229E-and IL-1-regulated genes and were able to identify a common set of genes that appears to be regulated by both HCoV-229E and IL-1 . This small group of genes mainly involves factors involved in innate immune regulation or cytokine functions ( e . g . CXCL2 , CXCL1 , CXCL5 , IL8 , GDF15 ) or MAPK pathway signaling ( e . g . DUSP1 , 8 , 10 , 16 , c-JUN , JUN D ) , and helps explain the inflammatory phenotype that , in many cases , is associated with coronavirus infections . Altogether , our comprehensive analysis of the HCoV-229E-regulated transcriptome provides numerous candidate genes that warrant further functional investigations to better understand their potential role in the viral life cycle , host defense mechanisms or other pathways . A detailed analysis of the recruitment pattern of RNA pol II and phosphorylated pol II revealed that about 200 HCoV-229E-induced genes are strongly upregulated at the transcriptional level . 49 of these genes were shared with IL-1 , a stimulus that very potently activates P ( S5 ) -pol II binding to its target genes [24] . Furthermore , the data support the idea that , at these loci , HCoV-229E infection reorganizes the chromatin structure at or near promoter regions , TSS and gene bodies , for example by increasing acetylation of H3K36 and H4K5 , suggesting that virus-induced host cell responses strictly depend on nuclear transcriptional mechanisms . The restricted chromatin targeting of p65 by HCoV-229E may also involve specific cooperative interactions with other eukaryotic TFs , such as MAPK- or ER stress-activated AP-1 proteins , interactions with upregulated corepressors such as ANKRD1 , or interactions with the nuclear pool of N proteins . In line with this idea , ANKRD1 was recently described to suppress NF-κB activity , whereas viral N proteins have consistently been reported to translocate to the nucleus , although to date there is no evidence for an involvement of the N protein in chromatin processes [26 , 54 , 55] . Additional support for HCoV-229E-specific TF complexes comes from experiments conducted to characterize the enhancer repertoire of virus-infected cells . A striking finding of our study is the identification of more than 1 , 000 enhancers that are activated specifically by HCoV-229E . The lack of p65 recruitment and the discovery of specifically enriched motifs for specific groups of TFs distinguish these structures from the IL-1-regulated enhancers . These motifs contain binding sites for a number of AP-1 proteins ( FOSL2 , JUND , c-JUN ) that are downstream of the ERK or JNK pathways known to be triggered by SARS-CoV or IBV infection [23 , 56] . Another abundantly found motif is bound by various members of the forkhead superfamily of TFs ( FOXC1/2 , FOXL1 , FOXB1 ) . These multifunctional proteins regulate transcriptional programs in cancer and innate immunity but their roles in RNA virus infection are unknown and await further investigation [57 , 58] . Additionally , the motif TGATGXAA is found in 108 virus-specific enhancers ( Fig 8E ) . This sequence matches the consensus sequence for the C/EBP-ATP response element CARE ( TGATGXAAX ) [59] . The CARE element is crucial for the ATF4-dependent activation of a specific set of genes that are ( up ) regulated during amino acid starvation [60] . ATF4 and its heterodimerization partners of the C/EBP or JUN families of TFs control metabolic gene expression programs but also the initiation of cell death upon activation of PERK-dependent phosphorylation of eIF2α during the ER stress response [61–63] . Thus , the large enhancer repertoire identified in this study likely coordinates a global gene response to cope with increased translational demand and the elevated load of misfolded proteins in the ER during viral replication . Altogether , the identification in non-coding genomic regions of three types of enhancers that respond to virus , to IL-1 , or to both against a background of 90% of all the other enhancer regions that did not change in response to these stimuli , provides an impressive example of the versatile usage of a small portion of the enhancer repertoire of a human cell to trigger stimulus-specific gene expression [14–16] . In conclusion , our results provide comprehensive insight into host cell transcriptome changes induced by HCoV-229E infection and link this information to the underlying chromatin changes as summarized in Fig 9 . We also show novel mechanisms ensuring the induction of a well-balanced and self-limiting NF-κB response to support viral propagation . A549 human alveolar basal epithelial cells and HeLa cells ( both ATCC ) and HuH7 human hepatoma cells ( Japanese Collection of Research Bioresources ( JCRB ) cell bank ) [64] were maintained in Dulbecco’s modified Eagle’s medium ( DMEM ) , complemented with 10% fetal calf serum ( FCS ) , 2 mM L-glutamine , 100 U/ml penicillin and 100 μg/ml streptomycin . The genome sequence of the human coronavirus ( strain 229E ) used in this study is available from GenBank ( accession number 304460 ) . Infections of A549 , HeLa and HuH7 cells were performed at 33°C using the indicated multiplicities of infection ( MOI ) . Virus titers ( TCID50/ml ) were determined using HuH7 cells . IL-1 treatment was done at 33°C using identical conditions and cell culture medium . Quantitative analysis of mRNA expression of individual A549 or HuH7 or viral genes was performed as described [24] . Transcriptomes were determined using Agilent 60k microarrays followed by KEGG pathway analysis of regulated gene sets determined by overrepresentation or gene set enrichment analyses . Cells expressing viral N protein were excised with a Leica LMD6000 system . Then , total RNA was extracted and subjected to RT-qPCR or microarray analysis . Indirect immunofluorescence analyses were performed on Leica DMIRE2 or DMi8 fluorescence microscopes as described [65] . Expression , phosphorylation and subcellular distribution of cellular proteins were analyzed as described [24] . Protein:DNA complexes were cross-linked in vivo by formaldehyde treatment , immunoprecipitated from denatured cell extracts and enriched DNA fragments were purified and quantified by ChIP-PCR or deep DNA sequencing . ChIP-seq reads were mapped to the human genome ( built HG19 ) , binding events were normalized to input samples and differential binding of phosphorylated RNA polymerase or p65 or changes in histone acetylation events were quantified and further analyzed by bioinformatics as described [24] . Statistics ( Mann-Whitney Rank , Wilcoxon signed rank or t-tests ) were calculated using R , SigmaPlot11 , GraphPadPrism6 . 0 and MS EXCEL2010 . Complete experimental procedures including reagents , buffers , nucleotide sequences , additional methods and software used are described in detail in S1 Supporting Experimental Procedures . Microarray ( GSE89167 ) and ChIP-seq ( GSE89212 ) data have been deposited at geo@ncbi . nlm . nih . gov .	Coronaviruses are major human and animal pathogens . They belong to a family of plus-strand RNA viruses that have extremely large genomes and encode a variety of proteins involved in virus-host interactions . The four common coronaviruses ( HCoV-229E , NL63 , OC43 , HKU1 ) cause mainly upper respiratory tract infections , while zoonotic coronaviruses ( SARS-CoV and MERS-CoV ) cause severe lung disease , including acute respiratory distress syndrome ( ARDS ) . The molecular basis for this fundamentally different pathology is incompletely understood . Our study provides a genome-wide investigation of epigenetic changes occurring in response to HCoV-229E . We identify at high resolution a large number of regulatory regions in the genome of infected cells that coordinate de novo gene transcription . Many of these genes have immunomodulatory functions and , most likely , contribute to limiting viral replication , while other factors may promote viral replication . The study provides an intriguing example of a virus that completes its entire life cycle in the cytoplasm while sending multiple signals to the nuclear chromatin compartment to adjust the host cell repertoire of transcribed genes . The approach taken in this study is expected to provide a suitable framework for future studies aimed at dissecting and comparing host responses to representative coronaviruses with different pathogenic potential in humans .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "phosphorylation", "viral", "transmission", "and", "infection", "gene", "regulation", "microbiology", "dna", "transcription", "epigenetics", "microbial", "genetics", "chromatin", "chromosome", "biology", "proteins", "gene", "expression", "viral", "replication", "viral", "...	2017	The NF-κB-dependent and -independent transcriptome and chromatin landscapes of human coronavirus 229E-infected cells
To maintain a particular cell fate , a unique set of genes should be expressed while another set is repressed . One way to repress gene expression is through Polycomb group ( PcG ) proteins that compact chromatin into a silent configuration . In addition to cell fate maintenance , PcG proteins also maintain normal cell physiology , for example cell cycle . In the absence of PcG , ectopic activation of the PcG-repressed genes leads to developmental defects and malignant tumors . Little is known about the molecular nature of ectopic gene expression; especially what differentiates expression of a given gene in the orthotopic tissue ( orthotopic expression ) and the ectopic expression of the same gene due to PcG mutations . Here we present that ectopic gene expression in PcG mutant cells specifically requires dBRWD3 , a negative regulator of HIRA/Yemanuclein ( YEM ) -mediated histone variant H3 . 3 deposition . dBRWD3 mutations suppress both the ectopic gene expression and aberrant tissue overgrowth in PcG mutants through a YEM-dependent mechanism . Our findings identified dBRWD3 as a critical regulator that is uniquely required for ectopic gene expression and aberrant tissue overgrowth caused by PcG mutations . The eukaryotic genome is packaged in a macromolecular complex termed chromatin . Chromatin is composed of DNA , RNA , histones , and non-histone proteins . The nucleosome , the basic unit of chromatin , consists of a histone octamer containing two copies of histones ( H3 , H2A , H2B , and H4 ) and 147 base pairs of DNA wrapped around the octamer [1] . Variants of H2A , H2B , and H3 differ from the canonical histones by a few amino acids [2] . Moreover , canonical histones are encoded by multiple repeated sequence arrays and expressed during the S-phase , while the variants are encoded by single-copy genes and expressed in the interphase [2 , 3] . Once the histone variants are incorporated into nucleosomes , they confer distinct physical and biochemical properties to DNA templates and thus regulate DNA replication , repair and gene transcription [3 , 4] . The deposition of histone variants is mediated by specific chaperone complexes . For example , H3 . 3 deposition , which often occurs in actively transcribed regions , is mediated by a histone chaperone named histone repressor A ( HIRA ) and its associated chaperone Yemanuclein ( YEM ) [5–10] . We previously showed that HIRA/YEM activity is negatively regulated by dBRWD3 ( Bromodomain and WD repeat-containing protein 3 ) [11] , which adds a second layer of complex regulation to H3 . 3 deposition . Dendritic arborization of peripheral neurons and photoreceptor development are disrupted in dBRWD3 mutants . These phenotypes are effectively suppressed by mutations in yem or H3 . 3 , indicating that dBRWD3 functions largely through restricting YEM-dependent H3 . 3 deposition [11] . However , it remains unknown where in the genome this regulation of H3 . 3 deposition takes place and how it affects transcription . Distinct patterns of transcriptional activation and inactivation of the genome contribute to the diversity of cell types in multicellular organisms . To inactivate transcription , Polycomb group ( PcG ) proteins bind to specific genomic regions and modify histones posttranslationally [12 , 13] . PcG proteins are grouped into two evolutionarily conserved complexes , PRC1 and PRC2 . In Drosophila , the PRC1 complex consists of Polycomb , Posterior sex combs , Sex combs extra ( Sce , the Drosophila homolog of human RING1 ) , Polyhomeotic proximal , Polyhomeotic distal with an accessory molecule , and Sex comb on midleg ( Scm ) [14] . The PRC2 complex is composed of Enhancer of zeste ( E ( z ) ) , Suppressor of zeste 12 , and extra sex combs [13] . Functionally , PRC1 adds a monoubiquitin moiety onto histone H2AK119 ( H2AK118 in Drosophila ) , whereas PRC2 catalyzes the trimethylation of H3K27 ( H3K27me3 ) . The combined activities of PRC1 and PRC2 repress transcription by compacting chromatin [15] . PcG proteins may also silence gene expression through a compaction-independent mechanism , such as by blocking transcription initiation [16–18] . Misregulation of transcription within typically inactive genomic regions leads to the disorganization of tissues and organisms [19] . For instance , loss of PcG function causes the ectopic expression of Homeotic ( Hox ) genes specific to posterior segments and thus disrupts the anterior-to-posterior body plan in embryos [20 , 21] . In Drosophila , loss of PRC1 leads to ectopic expression of Unpaired 1–3 , driving aberrant cell proliferation and tissue overgrowth by activating the JAK-STAT pathway [22] . In humans , loss of PRC1 function has been shown to promote tumorigenesis [23 , 24] . Reduced expression of the PRC1 subunit CBX7 has been implicated in bladder , breast , colon , glioma , lung , pancreatic , and thyroid carcinomas [25–31] , and CBX7 knockout mice develop lung and liver carcinoma [29] . Loss of PRC2 function also causes tumor formation . For example , the tumor-driving H3 . 3K27M mutation in pediatric diffuse intrinsic pontine gliomas ( DIPGs ) results in the inactivation of the PRC2 complex , causing ectopic expression of LIN28B , PLAG1 , and PLAGL1 , and leading to the de-differentiation and hyperproliferation of tumor cells [32–34] . A second mutation in the PRC2 complex genes in patients with neurofibromatosis increases the likelihood of developing malignant peripheral nerve sheath tumors [35 , 36] . Currently , no therapeutic strategies have been developed for tumorigenesis caused by ectopic gene expression . This is mainly because little is known about how ectopic gene expression is initiated within de-repressed genomic regions , and how it differs from conventional transcription initiation . Here we show that dBRWD3 is specifically required for ectopic gene expression and tissue overgrowth caused by PcG mutations . dBRWD3 sustains PcG mutation-induced ectopic gene transcription by regulating H3 . 3 deposition , which in turn affects the way RNA polymerase II occupies transcription start sites . Thus , our results suggest that human BRWD3 could be a potential therapeutic target for PcG mutation-driven tumors . In the process of investigating how dBRWD3 might affect gene expression , we unexpectedly found that the dBRWD3 mutations suppress the lethality of Scm mutants . Similar to other PcG mosaic mutants [22] , ScmD1 mosaic mutant flies died in the pupal stage . Interestingly , a significant portion of the mosaic ScmD1 , dBRWD3s5349 double mutants survived to the adult stage , so did the mosaic ScmD1 , dBRWD3PX2 double mutants ( Table 1 ) . To explore the relationship between dBRWD3 and PcG genes , we examined the genetic interaction between dBRWD3 and Posterior sex comb ( Psc ) , another PcG gene . We found that knockdown of Psc was semi-lethal , whereas simultaneous knockdown of Psc and dBRWD3 was fully viable ( Table 2 ) . Taken together , these results suggest a role for dBRWD3 as a suppressor of PcG genes . Since ectopic gene expression underlies many phenotypes of PcG mutations , we then investigated whether the dBRWD3 mutations also suppresses ectopic gene expression . While the second thoracic segment-specific Hox gene , antennapedia ( Antp ) , was repressed in wild-type eye clones ( Fig 1A and 1B , arrow ) , it was ectopically expressed in Scm mutant eye clones located in the posterior region ( Fig 1C , arrow ) . This ectopic Antp expression was dramatically reduced in ScmD1 , dBRWD3s5349 or ScmD1 , dBRWD3PX2 double-mutant eye clones ( Figs 1D , S1A , S1B and S1C ) . Similarly , Antp is ectopically expressed in Sce1 mutant eye clones ( Fig 1E , arrows ) but not in dBRWD3s5349 , Sce1 double-mutant clones ( Figs 1F and S1D ) . Interestingly , dBRWD3 is dispensable for the orthotopic expression of Antp in wings ( Figs 1G , 1H , 1I and S1E ) . Overall , these observations reveal that dBRWD3 is involved in the ectopic expression of Antp caused by PcG mutations . To determine whether dBRWD3 suppresses ectopic gene expression other than Antp in the eyes , we knocked down Pc in the central nervous system by Elav-GAL4 , reducing the level of Pc mRNA to 5% ( S2A Fig ) . Ubx is ectopically expressed in the Pc-depleted brains but not in the control ( Fig 2A , 2B , 2C , 2D , 2F and 2O ) . In addition , the Pc depleted ventral nerve cord was thinner and more elongated compared to the control ( Fig 2E , bracket ) . We found that both ectopic expression of Ubx and elongation of ventral nerve cords were suppressed by depletion of dBRWD3 ( Fig 2G and 2H ) . On the other hand , orthotopic expression of Ubx in the ventral nerve cord was not affected in the dBRWD3 , Pc double knockdown ( Fig 2G ) or in dBRWD3 knockdown animals ( Figs 2I , S2B , S2C and S2D , arrowhead ) . Thus , ectopic expression of Ubx depends on dBRWD3 whereas orthotopic expression of Ubx does not . Depleting E ( z ) , which encodes the H3K27 methyltransferase in PRC2 , caused ectopic expression of Ubx in brains ( Figs 2J , 2L and S2E ) and condensation failure in ventral nerve cords ( Fig 2K , bracket ) . The ectopic expression of Ubx and condensation failure of ventral nerve cords were also suppressed by knockdown of dBRWD3 ( Fig 2M , 2N and 2P ) . By contrast , orthotopic expression of Ubx was not affected in dBRWD3 , E ( z ) -doubly depleted ventral nerve cords ( S2F , S2G and S2H Fig ) . Taken together , our data indicates that ectopic Hox gene expression depends on dBRWD3 whereas orthotopic Hox gene expression does not . In addition to ectopic expression of Hox genes , loss of Ph , Psc , or Pc induces ectopic expression of unpaired ( upd ) 1–3 , and therefore activation of the JAK-STAT pathway that leads to overgrowth of tumor-like tissues [22 , 37–39] . By RT-qPCR , we detected mild increases of upd1 and upd2 mRNAs ( Fig 3A and 3B ) and a strong induction of upd3 mRNA ( Fig 3C ) in the mosaic ScmD1 mutant brain-eye complex . This upregulation of upd1-3 was prevented or significantly weakened in the mosaic ScmD1 , dBRWD3s5349 double mutants compared with mosaic ScmD1 mutants ( Fig 3A , 3B and 3C ) . Consistently , our immunofluorescent micrographs showed that Upd3 accumulated in ScmD1 mutant clones adjacent to the morphogenetic furrow ( Figs 3D , 3E and S3A , arrows ) , but not in ScmD1 , dBRWD3s5349 double-mutant clones ( Figs 3F and S3A ) or wild-type clones ( Fig 3G ) . Similarly , we detected accumulation of Upd3 in Sce1 ( Figs 3H and S3B , arrows ) and SceKO ( S4 Fig ) mutant clones , but not in the dBRWD3s5349 , Sce1 double-mutant clones ( Figs 3I and S3B , arrow ) . On the other hand , we found that orthotopic upd3 expression in the posterior end of the 2nd instar eye disc was not altered in dBRWD3s5349 mutant clones ( Fig 3J and 3K , arrows ) , indicating that the regulation of upd3 by dBRWD3 is specific to ectopic expression . We also used the 10XSTAT-GFP reporter to determine whether the JAK-STAT pathway , which is activated by Upd1-3 , is affected by dBRWD3 in PcG mutant cells [40 , 41] . In contrast to the weak and uniform expression observed in wild-type antennal discs ( Fig 3L and 3M , arrows ) , the GFP signal was much higher in ScmD1 mutant clones ( Fig 3N ) , likely due to up-regulation of upd1-3 . It remained unchanged in ScmD1 , dBRWD3s5349 double mutants ( Fig 3O , arrow ) . The STAT activity in the antennal disc was not affected in dBRWD3s5349 single mutants ( Fig 3P , arrow ) . Given these results , we propose that dBRWD3s5349 suppresses ectopic activation of the JAK-STAT pathway caused by ScmD1 or Sce1 mutations . By ectopically expressing Upd1 , Upd2 , and Upd3 , mutations in PRC1 and PRC2 complexes cause cell autonomous and non-autonomous proliferations [22 , 37] . Consistently , we found that ScmD1 ( Fig 4A ) or Sce1 ( Fig 4B ) mutant clones were larger than wild-type clones ( Fig 4C , 4D and 4E ) . Since Upd3 is a diffusible ligand stimulating non-cell-autonomous proliferation , we found the non-clonal area of mosaic ScmD1 ( Fig 4D ) or Sce1 ( Fig 4E ) eye antennal discs were also larger . Overall , mosaic ScmD1 or Sce1 mutant discs were 1 . 5- ( Fig 4D ) or 1 . 8-fold in size ( Fig 4E ) compared to wild-type respectively . This tissue overgrowth could be suppressed by dBRWD3s5349 ( Fig 4D , 4E , 4F and 4G ) . As a control , the mosaic dBRWD3s5349 alone did not reduce the disc size ( Fig 4H and 4I ) . Quantitatively , the numbers of clonal , non-clonal , and total mitotic cells marked by phosphorylation of H3S10 ( H3S10ph ) were increased in mosaic ScmD1 mutant eye-antennal disc ( Fig 4J ) . It was reduced to a wild-type level in mosaic ScmD1 , dBRWD3s5349 mutant eye-antennal discs ( Fig 4K , 4L and 4M ) . A similar suppression of proliferation was found in mosaic dBRWD3s5349 , Sce1 mutant eye-antennal discs ( Fig 4N ) as opposes to Sce1 mutants ( Fig 4O and 4P ) . The elongated Pc and E ( z ) depleted ventral nerve cords and control ventral nerve cord had comparable mitotic indices and undetectable expression of upd1 , upd2 and upd3 , indicating that the elongation of the ventral nerve cord was not caused by excessive proliferation . To determine whether the dBRWD3 mutation also suppresses other types of oncogenic tissue overgrowth , we sampled tissue overgrowth caused by the warts ( wts ) mutation that activates the hippo pathway [42 , 43] . We found that dBRWD3s5349 did not suppress the expression of the hippo pathway target gene , expanded ( ex ) ( S5A and S5B Fig ) and tissue overgrowth ( Fig 4Q and 4R ) . Thus , the dBRWD3 mutation appears to suppress the oncogenic tissue overgrowth specifically related to PcG mutations . To examine the growth-inhibition effect of the dBRWD3 mutation beyond the developmental stage , we generated overgrown eyes and surrounding tissues by knockdown of ph-p ( Fig 4S and 4T ) . In dBRWD3 and ph-p double-knockdown eyes , the tissue overgrowth phenotype was suppressed ( Fig 4U and 4V ) . From these data , we infer that the tissue overgrowth induced by PcG gene depletion requires dBRWD3 . dBRWD3 contains bromodomain I and II ( BRDI and BRDII ) that were predicted to be acetylated histone-binding domains ( S6A Fig ) [44] . To investigate the function of these bromodomains , we complemented ScmD1 , dBRWD3s5349 double-mutant cells with wild-type dBRWD3 , dBRWD3-N1287A , and dBRWD3-N1451A , in which the conserved asparagines in the BC loops of BRDI and BRDII were mutated into alanines . The wild-type dBRWD3-RFP , dBRWD3-N1287A-RFP , and dBRWD3-N1451A-RFP could restore the ectopic upd3 expression in ScmD1 , dBRWD3s5349 double-mutant cells ( Figs 5A , S6B and S6C ) . However , when both the BRDI and BRDII were disrupted , the dBRWD3-N1287A , N1451A-RFP ( designated as the dBRWD3-2BC-RFP ) could not restore the ectopic expression of upd3 ( Fig 5B ) . This failure to complement is not related to expression levels because dBRWD3-2BC-RFP was expressed more than wild-type dBRWD3-RFP ( S7 Fig ) . Therefore , the BRDI and BRDII of dBRWD3 are functionally redundant in supporting ectopic gene expression . We also complemented ScmD1 , dBRWD3s5349 double-mutant cells with a HLH motif-deleted , ΔN-dBRWD3 , which no longer interacts with the DNA damage binding protein 1 ( DDB1 ) and cannot be recruited to cullin4/DDB1 organized E3 ligase [11] . The ΔN-dBRWD3-RFP also failed to restore the ectopic upd3 expression in ScmD1 , dBRWD3s5349 double-mutant cells ( Fig 5C ) . Together , these results suggest that the activities of dBRWD3 binding to acetylated histones and cullin 4/DDB1 organized E3 ligase are essential for maintaining ectopic gene expression . Previously , we demonstrated that dBRWD3 limits HIRA/YEM-mediated H3 . 3 deposition [11] . However , it is not clear whether the BRDI , BRDII , and HLH motif of dBRWD3 are important for dBRWD3 regulation of H3 . 3 . When we complemented dBRWD3s5349 mutant cells with ΔN-dBRWD3-RFP or dBRWD3-2BC-RFP , the H3 . 3 levels in the dBRWD3s5349 mutant cells remained higher than those in wild-type cells ( S8A and S8B Fig , arrows ) . By contrast , dBRWD3-N1287A and dBRWD3-N1451A reduced the H3 . 3-dendra2 to a normal level ( S8C and S8D Fig , arrows ) , indicating a negative correlation between accumulation of H3 . 3 and ectopic gene expression . Indeed , the negative correlation was also observed in the dBRWD3 knockdown brains , where the endogenous H3 . 3 levels were higher than in control brains ( S8E Fig ) . When we co-immuno-stained the mosaic discs with ant-Antp antibody , the ectopic anti-Antp signals were strongly reduced along with accumulated H3 . 3 in ScmD1 , dBRWD3s5349 double-mutant clones ( S9B Fig , arrows ) . We next examined whether the increased H3 . 3 deposition suppresses ectopic gene expression . To this end , we introduced a yem mutation to reduce the dBRWD3 mutation-induced H3 . 3 deposition ( S9C Fig ) [11] . In ScmD1 , dBRWD3s5349 , yemGS21861 triple-mutant clones , the ectopic expression of Antp was restored and coincided with the reduction of H3 . 3 ( Figs 5D , S9C and S10A ) , suggesting that the dBRWD3 mutation suppresses the Scm mutation through a YEM-dependent mechanism . Moreover , upd3 was ectopically expressed in this triple mutant clone ( Figs 5E and S10B ) . Consistently , the size of the ScmD1 , dBRWD3s5349 , yemGS21861 triple-mutant eye-antennal disc was larger than that in the ScmD1 , dBRWD3s5349 double-mutant ( Fig 5F ) . To further substantiate the role for H3 . 3 in ectopic gene expression , we investigated whether ectopic gene expression could be suppressed by YEM-induced H3 . 3 deposition ( S11A , S11B and S11C Fig ) without any mutation in dBRWD3 . YEM over-expression effectively suppressed the ectopic expression of Antp ( Figs 5G and S10C ) . Taken together , these data indicate that dBRWD3 supports ectopic gene expression and tissue overgrowth mediated by PcG mutations by limiting HIRA/YEM-mediated H3 . 3 deposition . To understand how the dBRWD3 mutation suppresses ectopic gene expression , we investigated whether dBRWD3 is required for the removal of pre-existing H3K27me3 and H2A118ub at Antp and Ubx loci upon depletion of E ( z ) and Pc . The ChIP-qPCR analysis revealed a reduction of H3K27me3 in the distal region of Antp and at Ubx when E ( z ) was depleted ( Fig 6A and 6B ) . Similarly , H2A118ub levels at Antp and Ubx were also decreased in the Pc depleted brains ( Fig 6C and 6D ) . When dBRWD3 was depleted by RNAi together with E ( z ) or Pc , the H3K27me3 or H2A118ub levels at Antp and Ubx loci remained low or became lower ( Fig 6A , 6B , 6C and 6D ) , indicating that knockdown of dBRWD3 promotes or does not affect the removal of the pre-existing H3K27me3 and H2A118ub . Moreover , dBRWD3 is not required for the removal of pre-existing H3K27me3 in the E ( z ) depleted wings at the global level , as revealed by the equally reduced H3K27me3 immunostaining signals in the E ( z ) knockdown , and E ( z ) , dBRWD3 double-knockdown wings ( S12A , S12B and S12C Fig ) . H3K27me3 levels were not changed in the Pc knockdown , and Pc , dBRWD3 double-knockdown wings compared with the control ( S12D , S12E , S12F Fig ) . Similarly , dBRWD3 was not required for the removal of pre-existing H2AK118ub in the Sce mutant clones at the global level ( S12G and S12H Fig ) . trithorax ( trx ) encodes an H3K4 monomethyltransferase [45] and antagonizes PcG activity by binding to PRE sites , the enhancer cis-elements targeted by PcG proteins . In different cellular contexts , ectopic gene expression might or might not depend on trx [46 , 47] . To investigate the requirement of trx in ectopic expression of Antp in eyes , we generated ScmD1 , trxE2 double-mutant eye clones and found that the ectopic expression of Antp was suppressed ( S13 Fig ) , indicating that trx , like dBRWD3 , is required for ectopic Antp expression . We next investigated whether dBRWD3 and trx function in a linear pathway or in parallel . We found that over-expression of trx in the eye disc proper , peripodial epithelium of the eye disc , and wing disc was sufficient to induce ectopic expression of Antp or Abd-B ( Figs 6E , S14A , S14B and S14D ) , but not in a dBRWD3 knockdown background ( Figs 6F , S14C and S14E ) . In addition , Trx-induced Ubx ectopic expression in wing discs was strongly suppressed by knockdown of dBRWD3 , albeit incompletely ( S14F and S14G Fig ) . It seems that H3 . 3 deposition underlies the suppression of this ectopic gene expression , since Trx-induced Abd-B ectopic expression was also completely suppressed by YEM over-expression ( Fig 6G ) . To understand how dBRWD3 affects Trx-induced ectopic gene expression , we used ChIP to determine the levels of H3K4me1 and PolII over Ubx and Abd-B loci . Compared to the control , Trx increased H3K4me1 levels at the enhancer regions of Ubx and Abd-B irrespective of dBRWD3 depletion ( Fig 6H ) , which was later found to have no effects on Trx-induced monomethylation of H3K4 on a global scale ( S15A , S15B , S15C , S15D , and S15E Fig ) . These data indicate that dBRWD3 is epistatic to trx with respect to ectopic gene expression . By contrast , Trx-induced PolII levels at the transcription start sites and 5' ends of Ubx and Abd-B were significantly reduced when dBRWD3 was depleted by RNAi ( Fig 7A and 7B ) , while the PolII levels for orthotopically expressing Antp were not affected ( Fig 7C ) . When trx is overexpressed in the wing imaginal discs , RNA PolII increased on the Antp promoter , which is likely contributed by the purely orthotopic Antp expression and the Trx-induced ectopic Antp expression . The additional knockdown of dBRWD3 restored the PolII occupancy to a level similar to the orthotopic Antp expression control ( S16 Fig ) , indicating that knockdown of dBRWD3 suppressed only the Trx-induced increase of PolII occupancy but not the PolII occupancy of orthotopically expressing Antp . Similarly , PolII phospho-CTD Ser5 levels around the transcription start sites of Ubx and Abd-B were reduced upon knockdown of dBRWD3 ( Fig 7D and 7E ) . We also detected higher levels of H3K4me3 , a marker for active chromatins , at the transcription start sites and 5' ends of Ubx and Abd-B upon trx over-expression in a dBRWD3-dependent manner ( Fig 7F and 7G ) . Nevertheless , the levels of H3K4me3 at the orthotopically expressed Antp were not sensitive to dBRWD3 depletion ( Fig 7H ) . These observations suggest that , in ectopic gene expression , dBRWD3 is required for the activation of chromatin specifically at transcription start sites but not in enhancer regions . To examine whether the reduction of PolII and H3K4me3 are directly caused by H3 . 3 deposition , we examined dBRWD3 and H3 . 3 levels across Ubx and Abd-B loci . ChIP revealed that dBRWD3 was present at these regions and Trx over-expression moderately reduced the occupancy of dBRWD3 ( Fig 8A and 8B ) . Although dBRWD3 was recruited to the promoters , 5' and 3' regions of these ectopically expressed loci , it predominantly reduced H3 . 3 levels at the transcription start sites ( Fig 8C and 8D ) . These results imply that dBRWD3 maintains PolII levels of ectopically expressed genes by limiting H3 . 3 deposition at the transcription start sites . Moreover , dBRWD3 was also present at Antp locus ( Fig 8E ) and limited H3 . 3 levels at the promoters and the transcription start site of Antp ( Fig 8F ) . Nevertheless , PolII and H3K4me3 levels at Antp were not affected by knockdown of dBRWD3 ( Fig 7C and 7H ) , indicating that the sensitivity toward H3 . 3 but not H3 . 3 levels at the promoter and transcription start site per se distinguishes ectopic gene expression from orthotopic gene expression . Since in ectopic gene expression PolII occupancy appears to be more sensitive to H3 . 3 deposition at transcription start sites , we speculated that the initiation of ectopic transcription is more vulnerable to perturbation . To test this hypothesis , we reduced the activities of the general transcriptional factor TFII-D by knocking down various TATA box-binding protein ( TBP ) -associated factors ( TAF ) . Although TAFs are essential factors , animals with 65% reduction of Taf5 or 40% reduction of Taf7 in the central nervous system can grow to the adult stage without discernible defects ( S17A and S17B Fig ) . Ubx-expressing neurons in Pc , Taf5 or Pc , Taf7 double-knockdown brains exhibited 88% or 90% reduction in number relative to Pc depleted brains , respectively ( Fig 9A ) . Next , we investigated whether ectopic expression is more sensitive to general transcriptional factor TFII-H subunits , Cdk7 and CycH , which phosphorylate PolII CTD serine 5 . Similarly , Pc , Cdk7 or Pc , CycH double depletion by RNAi significantly reduced the number of neurons ectopically expressing Ubx in brains ( Figs 9A , S17C and S17D ) . By contrast , orthotopic expression of Ubx in ventral nerve cords was not affected by partial depletion of Taf5 , Taf7 , Cdk7 or CycH ( Fig 9B and 9C ) . The Trx-induced ectopic expression of Abd-B was also sensitive to knockdown of CycH ( Figs 9D and S18A ) , whereas the orthotopic expression of Antp was not ( Figs 9E , 9F and S18B ) . Interestingly , the ectopic expression domain of Abd-B extended to the ventral compartment by over-expressing CycH ( Figs 9G and S18A ) . As a control , over-expression of CycH alone did not induce ectopic expression of Abd-B ( Figs 9H and S18A ) or affect the orthotopic express of Antp ( Figs 9I and S18C ) . Collectively , we propose that ectopic gene expression involves more sensitive coordination between H3 . 3 deposition , TFII-D , and TFII-H activities than is required for orthotopic gene expression . A genome-wide H3 . 3 ChIP study revealed that H3 . 3 is enriched in enhancers , promoters , and gene bodies of actively transcribed genes [48] . This correlation suggests that H3 . 3 deposition may promote gene transcription , a concept that has been supported by the fact that H3 . 3s are more likely to possess marks associated with active gene expression , including trimethylation at lysine 4 ( H3 . 3K4me3 ) and acetylation at several lysine residues [49 , 50] . However , an elegant study demonstrated that gene transcription remains normal when H3 . 3 is replaced by H3 . 3K4A , casting doubt regarding the importance of H3 . 3K4me3 [51] . Moreover , the extent to which H3 . 3 deposition truly promotes gene transcription is difficult to determine from genetic studies because knockout of H3 . 3 concurrently leads to up-regulation of one set of genes and down-regulation of another[52] . In the case of trans-retinoid acid induced expression of Cyp26A1 in embryonic stem cells , H3 . 3 is actively deposited to the enhancer before induction . Upon induction , H3 . 3 is depleted from the enhancer but deposited into the promoter . Knockdown of H3 . 3 reduces the binding of RAR and Tip60 to the enhancer region , indicating that deposition of H3 . 3 at enhancer regions facilitates the activation of inducible genes [53] . However , the role of H3 . 3 at promoter and gene body in transcription remain unclear . It is even less clear how H3 . 3 affects ectopic gene expression , a pathological condition frequently associated with various cancers in humans [54 , 55] . In this study , we provide several lines of evidence that show that the regulation of H3 . 3 by dRBWD3 is required for the ectopic gene expression observed in PcG mutants or upon over-expression of trx . Firstly , only dBRWD3 transgenes that are able to reduce H3 . 3 levels in dBRWD3 mutant cells are capable of restoring ectopic expression of upd3 in Scm , dBRWD3 double-mutant cells . Secondly , the loss of yem , which prevents the aberrant incorporation of H3 . 3 , also restores the ectopic expression of upd3 in Scm , dBRWD3 , yem triple-mutant clones . Conversely , the over-deposited H3 . 3 induced by YEM is sufficient to suppress the ectopic Antp expression and Trx-induced ectopic Abd-B expression . In ectopic gene expression , H3 . 3 deposition at the enhancers is variably regulated by dBRWD3 . Nevertheless , through a not entirely clear mechanism , the H3K4me1 induced by Trx at enhancers is insensitive to dBRWD3 . By contrast , the dBRWD3 depletion-enhanced H3 . 3 deposition at transcription start sites interferes with H3K4me3 and PolII enrichment at the same regions as well as the 5' ends of the gene bodies . Taken together , these observations suggest that H3 . 3 deposition at these regions disrupts transcription by interfering with trimethylation of H3K4 as well as PolII engagement or activation . This notion is supported by a recent finding that ectopic gene expression persists longer in Hira mutants , in which H3 . 3 deposition is reduced at the promoter and 5' regions [56] . Different from the known role of H3K4me3 on the promoters , our data shows that H3K4me3 is enriched on the gene bodies of ectopically or orthotopically expressed genes rather than the promoters . This is most likely due to the bias associated with the selected PCR amplicons . dBRWD3 regulates the deposition of H3 . 3 more prominently at the promoters and transcription start sites in both ectopic and orthotopic gene expression . However , dBRWD3 regulates the PolII occupancy and H3K4me3 levels only in ectopic gene expression . Due to a not-yet-defined “robustness” of transcription , orthotopic gene expression is rendered insensitive to the increase of H3 . 3 . Our data suggest that the same robustness of orthotopic gene expression closely cooperates with TFII-D and TFII-H since orthotopic gene expression remains intact under the suboptimal TFII-D or TFII-H activities . Further studies are needed to understand molecular nature of the robustness . In addition to their negative effects on ectopic gene expression , H3 . 3 deposition and dBRWD3 may also interact with PcG in different contexts . For example , it has been reported that H3 . 3 deposition directs PRC2 to bivalent promoters in ES cells [57] . In addition , loss of dBRWD3 up-regulates Pc , pho , and tna but down-regulates phol , Jarid2 and the trithorax group genes ash1 and Iswi [11] . The regulation of ash1 , which encodes H3K4 monomethylase , is particularly interesting because an independent transcriptome analysis of adult heads in dBRWD3 hypomorphic mutants also confirmed a lower level of ash1 mRNA . However , functional studies revealed that ash1 depletion by RNAi rescues rather than exacerbates the rough eye phenotype caused by dBRWD3 depletion . In addition , the H3K4me1 levels in dBRWD3 mutant cells are similar to those in wild-type cells . Further studies are needed to determine the significance of dBRWD3 regulation of ash1 mRNA . Finally , knockdown of dBRWD3 causes further reduction of H3K27me3 or H2AK118ub in the E ( z ) or Pc depleted brains , suggesting that H3 . 3 deposition may accelerate the removal of these repressive marks . Further investigations are warranted to understand whether the accelerated removal of repressive marks is mainly contributed by the nucleosome turnover associated with H3 . 3 deposition or involves activation of demethylases and deubiquitylase . The H3K4me3 levels at the promoter and 5' ends of genes correlate well with active transcription . In fact , it is both a cause and a consequence of active transcription . As a cause , H3K4me3 recruits the TFII-D subunit TAF3 , a general transcription factor involved in PolII engagement and transcription initiation [58] . Based on such a scenario , we propose that dBRWD3 increases H3K4me3 levels to the extent required for promoters to recruit TFII-D in ectopic gene expression . Consistently , partial knockdown of Taf5 or Taf7 affects ectopic but not orthotopic gene expression . In other words , ectopic and orthotopic gene expression may require different levels of TFII-D activities . During active transcription , H3K4me3 is established by hSet1A/B , which is recruited to actively transcribed gene regions by CTD Ser5-phosphorylated PolII [59] . The phosphorylation of PolII’s CTD Ser5 is mediated by CDK7/CycH and is preferentially required for ectopic gene expression . Consistent with this idea , we demonstrated that partial knockdown of Cdk7 or CycH indeed affected ectopic gene expression without a discernible effect on orthotopic gene expression , perhaps because it was more dependent on the phosphorylation of PolII CTD Ser5 . A complement study in RING1A , RING1B double knockout embryonic stem cells revealed that de-repressed loci displaying higher levels of PolII phospho-CTD Ser5 are ectopically expressed at higher levels [60] , supporting our findings that phosphorylation of PolII CTD Ser5 plays an unique role in ectopic gene expression that is not shared with orthotopic gene expression . Based in part on these findings , we propose that dBRWD3 could play a preferential role in ectopic gene expression by facilitating the phosphorylation of PolII CTD Ser5 . PcG mutations and reduced expression of PcG proteins contribute to tumorigenesis in several human malignancies [25–31 , 34 , 35 , 61–64] . Hence , understanding the regulation of ectopic gene expression will have important medical implications . It has been shown that the ectopic expression of upd1 , upd2 , and upd3 also underlies tissue overgrowth in Drosophila PcG mutants , suggesting an evolutionarily conserved role for PcG in tumor suppression from insects to humans . Based on our results , we speculate that inhibition of the BRWD3 , TFII-D , and TFII-H complex , for example by the CDK7 inhibitor THZ1 [65–67] , might preferentially suppress a broad spectrum of tumors driven by PcG mutations . In summary , we found that ectopic gene expression differs from orthotopic gene expression in their sensitivities to dBRWD3 . Inactivation of dBRWD3 selectively suppresses ectopic gene expression and tissue overgrowth induced by loss of PcG function . p-ENTR-dBRWD3-N1287A-3XFlag , p-ENTR-dBRWD3-N1451A-3XFlag , p-ENTR-dBRWD3-N1287A , N1451A-3XFlag ( p-ENTR-dBRWD3-2BC-3XFlag ) were generated with the Thermo Scientific Phusion Site-Directed Mutagenesis kit using the previously described p-ENTR-dBRWD3-3XFlag as a template . p-ENTR-HA-yem was generated by PCR from the cDNA clone RE33235 , Drosophila Genetic Resource Center . p-ENTR-dBRWD3-N1287A-3XFlag , p-ENTR-dBRWD3-N1451A-3XFlag , and pENTR- dBRWD3-2BC-3XFlag were recombined into the pUWR vector ( DGRC Gateway collection ) to generate pUWR-dBRWD3-N1287A-3XFlag-RFP , pUWR-dBRWD3-N1451A-3XFlag-RFP , pUWR-dBRWD3-2BC-3XFlag-RFP . p-ENTR-HA-yem was recombined into the pTWF vector ( DGRC Gateway collection ) to generate pTWF-HA-yem . Flies were raised in standard conditions at 25°C except as otherwise mentioned . The dBRWD3s5349 , and yemGS21861 were described earlier [11 , 68] . SceKO and ScmD1 , trxE2 were kindly provided by Dr . Muller [39 , 47] . hs-H3 . 3-GFP was a gift from Dr . Kami Ahmad [69] . GMR-GAL4 ( stock number 9146 ) , Elav-GAL4 ( stock number 458 ) , ScmD1 ( stock number 24158 ) , Sce1 ( stock number 24618 ) , wtsx1 ( stock number 44251 ) , ex-LacZ ( stock number 11067 ) , UAS-mCD8-GFP ( stock number 5146 ) , UAS-Taf5-shRNA ( stock number 35367 ) , and UAS-Taf7-shRNA ( stock number 55216 ) were obtained from the Bloomington stock center . UAS-trx ( stock number 12194 ) and OK107-GAL4 , were obtained from the Drosophila Genetic Resource Center , Kyoto . The ScmD1 , dBRWD3s5349 double mutant , ScmD1 , dBRWD3PX2 double mutant , dBRWD3s5349 , Sce1 double mutant , ScmD1 , dBRWD3s5349 , yemGS21861 triple mutant , and wtsx1 , dBRWD3s5349 double mutant were generated by recombination . UAS-Psc-dsRNA ( NIG3886R-4 ) UAS-ph-p-dsRNA ( NIG18412R-1 ) , UAS-E ( z ) -dsRNA ( NIG6502R-3 ) , UAS-Pc-dsRNA ( NIG32443R-1 ) , UAS-CycH-dsRNA ( NIG7405R-1 ) , and UAS-Cdk7-dsRNA ( NIG3319R-1 ) were obtained from the fly stocks of the National Institute of Genetics , Kyoto , Japan ( NIG-FLY ) . UAS-dBRWD3-dsRNA ( VDRC40209 ) was obtained from the Vienna Drosophila RNAi Center ( VDRC ) . The transgenic flies ubi-dBRWD3-N1287A-3XFlag-RFP , ubi-dBRWD3-N1451A-3XFlag-RFP , ubi-dBRWD3-N1287A , N1451A-3XFlag-RFP ( ubi-dBRWD3-2BC-3XFlag-RFP ) , and pTWF-HA-yem were generated by microinjection for germ-line transformation . The transgenic flies ubi-H3 . 3-dendra2 , ubi-dBRWD3-3XFlag-RFP , ubi-delta-N-dBRWD3-3XFlag-RFP , and 10XSTAT-nlsGFP were described previously [11 , 41] . Genotypes for mosaic mutant clones in eyes: Total RNA was isolated from instar larval mosaic eye brain complexes using TRIzol reagent ( Invitrogen ) . Following the manufacturer’s protocol , cDNA was synthesized using oligo ( dT ) and SuperScript reverse transcriptase ( Invitrogen ) . OmicsGreen qPCR 5X Master Mix ( Omics Bio ) was used for real-time quantitative PCR on a CFX96 connect Real-Time PCR System ( Bio-Rad ) . RPL32 was used as an endogenous loading control . 3rd instar larval eye imaginal discs were dissected in PBS and fixed for 17 minutes in 4% formaldehyde , followed by three 10-min washes in PBS supplemented with 0 . 3% Triton-X-100 ( PBT ) and 30-min blocking in PBT containing 5% normal donkey serum ( NDS ) . After blocking , discs were incubated with primary antibody either overnight at 4°C or 2 hours at room temperature in PBT containing 5% NDS . After incubation with primary antibody , discs were washed three times in PBT before incubating with secondary antibody in PBT containing 5% NDS for one hour at room temperature . After three subsequent washes , discs were mounted with glycerol . Primary antibodies used in this study include mouse anti-H2AK118ub ( 1:100 , Millipore , E6C5 ) , rabbit anti-H3K27me3 ( 1:100 , Millipore ) , rabbit anti-H3K4me1 ( 1:100 , Active Motif ) , rabbit anti-H3S10ph ( 1:500 , Millipore ) , mouse anti-Ubx ( 1:20 , DSHB , Ubx ) , mouse anti-Antp ( 1:20 , DSHB , 8C11 ) , rabbit anti-upd3 ( 1:750 ) , and mouse anti-β-Galactosidase ( 1:1000 , Sigma , GAL-50 ) . Secondary antibodies include goat anti-mouse Cy3 ( 1:1000 , Jackson ImmunoResearch ) , goat anti-mouse Cy5 ( 1:1000 , Jackson ImmunoResearch ) , and goat anti-rabbit Cy3 ( 1:1000 , Jackson ImmunoResearch ) . Chromatin immunoprecipitation was done with a ChIP-IT High Sensitivity ( HS ) Kit ( Active Motif ) following the instructions provided by the manufacturer . Briefly , 300 pairs of brain lobes ( leaving out the attached ventral nerve cords ) or wing imaginal discs of 3rd instar larvae were collected . The collected tissues were fixed with complete tissue fixation solution ( 28μl 37% formaldehyde in 970μl PBS ) at room temperature for 15 minutes . Fixation was stopped with the stop solution at room temperature for 5 minutes . The fixed tissues were washed with ice-old PBS wash buffer and then immersed in the tissues with the chromatin prep buffer . The fixed tissues were sonicated using the UP50H Ultrasonic Processor ( Hielscher-Ultrasound Technology ) , with 30% amplitude and 20 pulse cycles of 30 seconds on followed by 30 seconds off . 6 μg sheared chromatin was incubated with 1μg antibodies overnight at 4°C . 30 μl Protein G agarose beads were added to each IP reaction . The mixture was rotated at 4°C for 3 hours . The ChIP reactions were loaded into columns and washed . The ChIP DNA was obtained by eluting the columns with elution buffer AM4 . The elute was treated with Protease K at 55°C for 30 minutes , 80°C for two hours , followed by column clean-up . OmicsGreen qPCR 5X Master Mix ( Omics Bio ) was used for real-time quantitative PCR on a CFX96 connect Real-Time PCR System ( Bio-Rad ) to measure the amount of ChIP DNA and input DNA containing indicated sequences from enhancers , promoters , and 5' transcription regions of Antp , Ubx , and Abd-B . Primary antibodies used include rabbit anti-H2AK118ub ( Cell signaling , D27C4 ) , mouse anti-H3K27me3 ( Abcam , mAbcam 6002 ) , rabbit anti-H3K4me1 ( Abcam , ab8895 ) , rabbit anti-H3K4me3 ( Abcam , ab8580 ) , mouse anti-PolII ( Abcam , 4H8 ) , and rabbit anti-PolII phospho-CTD Ser5 ( Abcam , ab5131 ) , rabbit anti-HA ( Cell Signaling , C29F4 ) and mouse anti-Flag ( Sigma , M2 ) All confocal images were obtained by LSM 700 laser scanning confocal microscope ( Carl Zeiss ) . For quantitative analysis of protein levels , the antibody staining conditions , laser power , and pinhole sizes were kept identical among groups . Pixel number , pixel intensity , and area were provided by the built-in software in LSM 700 . The areas of clones ( marked by the absence of GFP ) and non-clones ( marked by GFP ) were calculated by the total GFP positive and GFP negative areas respectively . Antp-positive regions in ventral nerve cords were manually marked . The Antp expression areas and lengths were calculated by the built-in software according to the marked regions . H3S10ph-positive mitotic cells and Antp-positive brain cells were manually counted .	Genetic information is stored in our genomic DNA , and different cells retrieve distinct sets of information from our genome . While it is important to activate genomic regions encoding proteins that are essential for a given cell type , it is equally important to silence genomic regions encoding proteins that are potentially harmful to this type of cells . One of the gene silencing mechanisms frequently used during and after development is mediated by the Polycomb group ( PcG ) proteins . If this guardian function does not perform correctly due to PcG mutations , genes that are normally silenced—such as oncogenes—are expressed aberrantly . Due to the activation of oncogenes and the loss of other PcG functions , PcG mutant cells often begin to display hallmarks of cancer , such as proliferating beyond control , acquiring stem-cell-like properties , and migrating to distant sites . If the transcriptional mechanisms underlying aberrant gene expression in PcG-mutant cancer cells differ from gene expression in normal cells , we may be able to selectively inhibit the growth of cancer cells without affecting their normal counterparts . Here we show that the difference between these two types of gene expression resides in their sensitivity to dBRWD3 , a negative regulator of the deposition of histone H3 variant H3 . 3 . Our results indicate that the inactivation of dBRWD3 or promotion of H3 . 3 deposition may selectively suppress ectopic gene expression and tumorigenesis driven by mutations in PcG .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "medicine", "and", "health", "sciences", "gene", "regulation", "dna-binding", "proteins", "cloning", "dna", "transcription", "mutation", "developmental", "biology", "molecular", "biology", "techniques", "eyes", "morphogenesis", "research", "and", "analysis", "methods", ...	2016	dBRWD3 Regulates Tissue Overgrowth and Ectopic Gene Expression Caused by Polycomb Group Mutations
Enteropathogenic and enterohemorrhagic Escherichia coli ( EPEC/EHEC ) are human intestinal pathogens responsible for diarrhea in both developing and industrialized countries . In research laboratories , EPEC and EHEC are defined on the basis of their pathogenic features; nevertheless , their identification in routine laboratories is expensive and laborious . Therefore , the aim of the present work was to develop a rapid and simple assay for EPEC/EHEC detection . Accordingly , the EPEC/EHEC-secreted proteins EspA and EspB were chosen as target antigens . First , we investigated the ideal conditions for EspA/EspB production/secretion by ELISA in a collection of EPEC/EHEC strains after cultivating bacterial isolates in Dulbecco’s modified Eagle’s medium ( DMEM ) or DMEM containing 1% tryptone or HEp-2 cells-preconditioned DMEM , employing either anti-EspA/anti-EspB polyclonal or monoclonal antibodies developed and characterized herein . Subsequently , a rapid agglutination latex test ( RALT ) was developed and tested with the same collection of bacterial isolates . EspB was defined as a biomarker and its corresponding monoclonal antibody as the tool for EPEC/EHEC diagnosis; the production of EspB was better in DMEM medium . RALT assay has the sensitivity and specificity required for high-impact diagnosis of neglected diseases in the developing world . RALT assay described herein can be considered an alternative assay for diarrhea diagnosis in low-income countries since it achieved 97% sensitivity , 98% specificity and 97% efficiency . Annually , nearly five million cases of diarrhea are reported around the world leading to 800 thousand deaths per year in under-fives [1] , [2] , and Escherichia coli is the etiological agent responsible for most of them [3] . The E . coli isolates associated with diarrhea are classified into pathotypes on the basis of specific virulence factors , pathogenesis or clinical manifestation [4] . Among them , enteropathogenic E . coli ( EPEC ) and enterohemorrhagic E . coli ( EHEC ) continue to represent a threat to human health worldwide [5] . Both pathotypes can induce the attaching and effacing ( A/E ) lesion on the intestinal mucosa , characterized by intimate bacterial adhesion , destruction of microvilli , and accumulation of polymerized actin in pedestals beneath intimately attached bacteria [6] . The A/E lesion formation is caused by effector proteins that are secreted into the enterocytes by the type III secretion system [4] . All genes necessary for the A/E lesion formation are located in a pathogenicity island called locus of enterocyte effacement ( LEE ) . After the establishment of initial contact via EspA containing filaments , two further effector proteins , EspB and EspD , are translocated into the host cell membrane where they form a pore structure [7] , [8] , which allows the translocation of effector proteins . The delivery of the translocated intimin receptor ( Tir ) into the host cell membrane is followed by dissolution of EspA filaments and intimate bacterial attachment via binding of Tir to the bacterial adhesin intimin [9] , [10] . EHEC but not EPEC produces the Shiga toxins , which are associated with the development of severe complications of infection , namely hemorrhagic colitis ( HC ) and the hemolytic uremic syndrome ( HUS ) [11] . Moreover , some EPEC strains may carry a large plasmid known as the EPEC adherence factor plasmid ( pEAF ) [12] , [13] , which encodes the bundle-forming pilus ( BFP ) [14] , [15] . Since pEAF is not present and BFP is not produced by all isolates , this pathotype has been divided in the subgroups typical EPEC ( tEPEC ) and atypical EPEC ( aEPEC ) , where BFP is produced only by tEPEC [14] , [16]–[18] . Epidemiologically , EHEC is more common as a food or water-borne pathogen in industrialized countries , and EPEC remains a significant cause of diarrhea in low-income countries , responsible for high rates of infant morbidity and mortality [15] , [19] , [20] , but it is worth to mention that aEPEC has been now considered an emerging pathogen in both industrialized and developing countries [21]–[27] . EPEC and EHEC have been defined on the basis of their pathogenic properties; however , this detection in routine laboratories is expensive and laborious for developing countries . Therefore , in these settings they are defined only with a serogroup agglutination-based test [28] . As LEE-encoded virulence factors are common to EPEC and EHEC strains , intimin has been considered the first target for diagnosis [29] , mainly its conserved region ( Int388–667 ) [30] , [31] . Essentially , intimin detection in EPEC and EHEC isolates is appropriate by immunofluorescence and/or immunoblotting , i . e . , after bacterial permeabilization , allowing anti-intimin antibody accessibility [32]–[34] . Alternative targets for EPEC/EHEC diagnosis are the LEE-secreted proteins , including EspA and EspB . For production and delivery of EspA and EspB , special culture conditions are required [7] , [9] , [35]–[38] . Only a few developed antibodies against EspA or EspB have been used either in the characterization of EPEC or EHEC [7] , [39] , [40] , and only anti-EspB polyclonal antibodies have been evaluated for diagnosis [41] , [42] . Therefore the goal of the present work was to develop a rapid and simple assay for EPEC/EHEC detection , especially for EPEC , a pathotype that lacks an internationally recognized standard diagnostic test [43] . Accordingly , we first investigated the ideal conditions for EspA/EspB production/secretion in a collection of EPEC/EHEC isolates , employing either anti-EspA/anti-EspB polyclonal or monoclonal antibodies developed and characterized herein . Subsequently , we defined EspB as a target antigen and EspB monoclonal antibodies as a tool for the rapid agglutination latex test ( RALT ) to be considered an alternative assay for diarrhea diagnosis in developing countries . We analyzed in this study a collection of 71 aEPEC [17] , 31 tEPEC [18] , [32] and 23 EHEC [44] , belonging to different serotypes characterized as LEE-positive isolates . We also included for ELISA cut-off definition and specificity of the RALT , 20 LEE-negative diarrheagenic E . coli ( DEC/LEE− ) , 20 fecal E . coli negative for DEC virulence factors ( NVF E . coli ) isolates and 20 Enterobacteriaceae isolates ( Aeromonas hydrophila , Edwardsiella tarda , Enterobacter cloacae , Enterococcus faecalis , Klebsiella pneumoniae , Morganella morganii , Pseudomonas aeruginosa , Proteus mirabilis , Providencia spp . , Salmonella spp . , Serratia marcescens , Shigella boydii and Shigella flexneri ) from our laboratory collection . The prototype tEPEC E2348/69 [45] was included in the assays as a positive control for EspA/EspB-producing strain . These experiments were conducted in agreement with the Ethical Principles in Animal Research , adopted by the Brazilian College of Animal Experimentation , and they were approved by the Ethical Committee for Animal Research of Butantan Institute ( Protocol 469/08 ) . EspA and EspB antibodies: development and characterization EspA and EspB recombinant proteins were obtained from E . coli BL21 clones containing the pET28a-EspA or pET28a-EspB plasmid . Protein induction , production and purification were done as described elsewhere [7] . These proteins were employed for raising the rabbit polyclonal ( PAb ) [31] and the monoclonal ( MAb ) antibodies [44] , [46] . Bacterial isolates were cultivated in Luria Bertani ( LB ) broth at 37°C for 18 h . Each culture was then inoculated at a 1∶50 dilution at 37°C for 6 h ( 250 rpm ) into Dulbecco’s modified Eagle’s medium ( DMEM ) , DMEM containing 1% tryptone ( DMEM-T ) or preconditioned DMEM ( DMEM-PC ) , which was prepared by incubation of DMEM without antibiotics or fetal bovine serum with monolayers of HEp-2 for 24–48 h . The supernatant referred to as “preconditioned medium” was collected , adjusted to pH 7 . 4 , and filtered through a 0 . 2 mm membrane [47] . After growth of the bacteria , the cultures were centrifuged at 13 , 000×g for 10 min and the supernatants were stored at 4°C for 16–18 h . A 100-µL aliquot of supernatants was used to coat the microplates in indirect ELISA assays . The microplates ( MaxiSorp microplates , Nunc , Rochester , NY , USA ) were then kept at 37°C for 2 h . After blocking with 1% bovine serum albumin ( BSA ) at 37°C for 30 min , the microplates were incubated with anti-EspA MAb ( 5 µg/mL ) or MAb anti-EspB ( 10 µg/mL ) or with 30 µg/mL anti-EspA PAb or anti-EspB PAb at 37°C for 1 h . Antigen-antibody binding was detected by the addition of either peroxidase-conjugated goat anti-mouse IgG ( 1∶5 , 000 ) or peroxidase-conjugated goat anti-rabbit IgG ( 1∶5 , 000 ) and OPD ( 0 . 5 mg/mL ) and H2O2 as enzyme substrates . The peroxidase reaction was stopped by the addition of 1 N HCl . The absorbance was measured at 492 nm in a Multiskan EX ELISA reader ( Labsystems , Milford , MA , USA ) . The absorbance values from duplicates of three independent experiments from LEE-positive and LEE-negative isolates after reaction with anti-EspA or anti-EspB antibodies were analyzed by GraphPrism 5 . 01 , using Student’s t-test and two-away ANOVA . The differences were considered statistically significant when p≤0 . 05 . The receiver operating characteristic ( ROC ) curve was employed for determining the cut-off value using the ELISA data , considering the highest sensitivity and specificity . Prior to testing the isolates by rapid agglutination latex test ( RALT ) , the beads were coupled with anti-EspB MAb . Briefly , beads in a 2 . 5% aqueous suspension ( 1 µm diameter – Polyscience , Warrington , PA , USA ) were washed three times with PBS and incubated with 8% glutaraldehyde in PBS at room temperature for 4 h . Next , 200 µg anti-EspB MAb were added and the mixture incubated at room temperature for 16–18 h for coupling , followed by further incubation in the presence of 0 . 2 M ethanolamine and BSA . Both incubations were with gentle mixing at room temperature for 30 min . Between incubations , the coated beads were washed and centrifuged ( 7 , 200×g ) for 6 min . After the last washing procedure , the pellet was ressuspended in the storage buffer ( Polyscience , Warrington , PA , USA ) and kept at 4°C for 7 days . For RALT , bacterial lysate was prepared using 20 mg of isolates grown on DMEM-agar at 37°C for 16–18 h and suspended in 80 µL of lysis buffer [Bacterial Protein Extraction Reagent ( B-PER ) , Thermo Scientific , Rockford , IL , USA] , followed by incubation at room temperature for 15 min . The assay was performed on a slide glass using 20 µL of bacterial lysate and 20 µL of latex beads coupled to anti-EspB MAb , and checking for agglutination after 5 min of gentle mixing . EspA and EspB are noteworthy antigens , demonstrated by the anti-EspA and anti-EspB polyclonal antibodies IgG titers ( 1∶10 , 240 and 1∶40 , 960; respectively ) and detection limit of 78 and 156 ng/mL , respectively . Secretory hybridomas of antibodies against EspA and EspB were obtained and subcloned by limiting dilution . Anti-EspA and anti-EspB MAbs produced by the selected clones ( 3C12 and 4D9 , respectively ) , were classified as IgG2a and showed a dissociation constant of 1 . 66×10−10 and 2×10−9 M , with detection limit of 19 and 17 ng/mL , respectively . The reactivity of all antibodies , as well as the efficiency of different culture media ( DMEM , DMEM-T and DMEM-PC ) were determined in the collection of tEPEC , aEPEC and EHEC isolates by indirect ELISA . Using either anti-EspA PAb or anti-EspA MAb , the production of EspA by tEPEC and EHEC isolates was the same regardless the culture medium ( Figure 1 A and B ) . Considering either anti-EspB PAb or MAb , production of EspB by EHEC isolates was also medium independent . On the other hand , when LEE-positive isolates were evaluated as a group , the production of EspB was higher in DMEM compared to DMEM-T ( p<0 . 0001 ) or to DMEM-PC ( p = 0 . 003 ) , and no difference was observed between DMEM-PC and DMEM-T ( p = 0 . 129 ) ( Figure 2 ) . Therefore , comparing the production of EspB in 125 LEE-positive and 60 LEE-negative isolates using both anti-EspB antibodies , we observed by ROC curves that regardless the medium employed the sensitivity ( Se ) and specificity ( Sp ) were higher with MAb ( Se = 90 . 4% , confidence intervals of 83 . 8 to 96 . 4% and Sp = 96 . 4% , confidence intervals of 87 . 7 to 99 . 6% ) ( Figure 2B ) than PAb ( Se = 82 . 4% , confidence intervals of 74 . 6 to 88 . 6% and Sp = 92 . 9% , confidence intervals of 82 . 7 to 98% ) ( Figure 2A ) . And the ELISA cut-off value was lower for MAb than for PAb ( 0 . 027 and 0 . 152 , respectively ) . For RALT , bacterial isolates were grown on DMEM-agar and the test was done using latex sensitized with anti-EspB MAb . Figure 3 presents typical negative and semi-quantitative positive reactions . From the total of the positive reactions by RALT , + correspond to 44 . 8%; ++ to 26 . 4%; +++ to 11 . 2% and ++++ to 14 . 4% of the isolates . By this assay only four LEE-positive isolates ( one aEPEC and three tEPEC ) did not react with anti-EspB MAb and one false positive occurred ( Proteus mirabilis ) ( Table 1 ) . Considering the LEE-positive and -negative isolates , the test exhibited 97% sensitivity , 98% specificity and 97% efficiency . A fast and inexpensive diagnosis for EPEC/EHEC infections is highly required considering their global prevalence , the severity of the diseases associated with them , and the fact that the use of antibiotics to treat EHEC infections can be harmful . One appropriate approach for their rapid detection may utilize the secreted proteins EspA and/or EspB , since the espA and espB genes are present in LEE positive isolates and they are the major secreted proteins by both pathogens [4] . Thus , the aim of the present study was to develop and define sensitivity and specificity of EspA and EspB antibodies , determine the ideal target antigen , and design a simple and rapid test for the diagnosis of both emerging pathogens worldwide . Production and secretion of virulence factors in pathogenic bacteria are tightly and coordinately regulated . Growth phase and environmental conditions characteristic of the host , including temperature and partial O2 pressure , are the stimulus for virulence factor expression in various gram-negative pathogens [48] , [49] . Additionally , in our experience , the production of virulence factors is a critical point for diarrheagenic E . coli diagnosis [50]–[52] . Thus , initially , one group of isolates ( including tEPEC , aEPEC and EHEC ) was cultivated in different media: LB broth , DMEM , E . coli broth and Minimum medium . Besides , other culture conditions were tested , including pH ( 7 . 2 and 5 . 5 ) , CO2 presence , and growth time period ( 6 , 18 and 24 h ) . Our results showed that in general DMEM favored the production of secreted proteins after 6-h growth culture , but with individual variation ( data not shown ) . Some reports describe that the use of preconditioned DMEM ( DMEM-PC ) provides signals from epithelial cells affecting virulence factors expression [47] . Also the secretion of plasmid-encoded toxin ( Pet ) by enteroaggregative E . coli is dependent on the addition of tryptone to DMEM ( DMEM-T ) [53] . Considering this , the bacterial isolates from our collection were cultivated in DMEM , DMEM-T and DMEM-PC , but EspB production and secretion was enhanced when bacterial isolates were cultivated in DMEM without enrichment . Another important point of the present work was the evaluation of the four antibodies raised . We expected that EspA would be a biomarker for diagnosis and anti-EspA antibodies a detecting tool , since this protein is the major component of a transiently expressed surface organelle , which forms a direct link between the bacterium and the host cell [7] . However , our data pointed out EspB as the target antigen , and MAb anti-EspB the best antibody for defining LEE-positive isolates . Nakasone et al . [42] , [54] also defined EspB as the target antigen for identifying LEE-positive strains . In fact , EspA filaments exhibit antigenic polymorphisms [55] . The indirect ELISA using anti-EspB MAb showed 90 . 4% and 96 . 4% , sensitivity and specificity , respectively , indicating its possible use in routine diagnostic laboratories . However , this methodology requires specific laboratory instrumentation , making it difficult to be performed in low-income country settings . Therefore , we standardized here a rapid agglutination test using latex beads coated with anti-EspB MAb ( RALT ) , which has the sensitivity and specificity required for high impact diagnosis of neglected diseases in the developing world [56] . Two other assays have been described for LEE-positive isolates based on EspB detection; the 16–18 h reversed passive latex agglutination test ( RPLA ) [41] and a 10 min immunochromatographic test ( IC ) [42] . Although more time consuming , the RPLA test was more sensitive than the IC test [42] . Serotyping-based diagnosis is the only methodology available in limited-resources settings , employing either commercial or in-house antisera [28] . The standardized RALT for detection of EPEC and EHEC will have a remarkable impact in the diagnosis of these pathotypes , demonstrated by 97% sensitivity , 98% specificity and 97% efficiency in EspB detection . Also , no cross-reaction was observed with other DEC pathotypes and E . coli negative for DEC virulence factors . Among the enterobacteria species only one Proteus mirabilis was recognized by MAb anti-EspB . However , P mirabilis can be easily differentiated from EPEC/EHEC by biochemical methods employed for species identification [57] , a step necessary prior to the performance of our RALT . Thus the established agglutination latex in the present study is a simple , rapid ( 5 min ) and easy to perform test , which can be employed in less equipped laboratories in low-income countries .	A rapid and low-cost diagnosis for EPEC/EHEC infections is extremely required considering their global prevalence , the severity of the diseases associated with them , and the fact that the use of antibiotics to treat EHEC infections can be harmful . For EHEC , the detection of Stx toxins has already been developed , but for EPEC , an internationally recognized standard diagnostic test is lacking . Thus , the approach for their rapid detection in this study was the use of the secreted proteins EspA and/or EspB , since they are the major secreted proteins in both pathogens . EspB was defined as a biomarker and its corresponding monoclonal antibody as the tool for EPEC/EHEC diagnosis using a latex agglutination assay , which can be employed in less equipped laboratories in developing countries .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "bacterial", "diseases", "infectious", "diseases", "medicine", "and", "health", "sciences", "diagnostic", "medicine", "antigen-antibody", "reaction", "analysis", "emerging", "infectious", "diseases", "antibody", "production", "serodiagnosis", "immunoassays", "antibody", "is...	2014	Development of a Rapid Agglutination Latex Test for Diagnosis of Enteropathogenic and Enterohemorrhagic Escherichia coli Infection in Developing World: Defining the Biomarker, Antibody and Method
Mutated NLRP3 assembles a hyperactive inflammasome , which causes excessive secretion of interleukin ( IL ) -1β and IL-18 and , ultimately , a spectrum of autoinflammatory disorders known as cryopyrinopathies of which neonatal-onset multisystem inflammatory disease ( NOMID ) is the most severe phenotype . NOMID mice phenocopy several features of the human disease as they develop severe systemic inflammation driven by IL-1β and IL-18 overproduction associated with damage to multiple organs , including spleen , skin , liver , and skeleton . Secretion of IL-1β and IL-18 requires gasdermin D ( GSDMD ) , which—upon activation by the inflammasomes—translocates to the plasma membrane where it forms pores through which these cytokines are released . However , excessive pore formation resulting from sustained activation of GSDMD compromises membrane integrity and ultimately causes a pro-inflammatory form of cell death , termed pyroptosis . In this study , we first established a strong correlation between NLRP3 inflammasome activation and GSDMD processing and pyroptosis in vitro . Next , we used NOMID mice to determine the extent to which GSDMD-driven pyroptosis influences the pathogenesis of this disorder . Remarkably , all NOMID-associated inflammatory symptoms are prevented upon ablation of GSDMD . Thus , GSDMD-dependent actions are required for the pathogenesis of NOMID in mice . NLRP3 , also called cryopyrin , assembles an inflammasome complex upon sensing danger signals triggered by structurally different exogenous and endogenous molecular entities [1–3] . Failure to clear the insults or restore homeostasis leads to chronic activation of this inflammasome , a response that underlies various inflammatory and metabolic diseases , including gout , diabetes , and atherosclerosis [4] . Activating mutations in the NLRP3 gene also cause constitutive activation of the NLRP3 inflammasome in patients with a spectrum of autoinflammatory disorders known as cryopyrinopathies or cryopyrin-associated periodic syndromes ( CAPS ) , which include neonatal-onset multisystem inflammatory disease ( NOMID ) , Muckle-Wells syndrome ( MWS ) , and familial cold autoinflammatory syndrome ( FCAS ) [5 , 6] . CAPS are monogenic disorders with some degree of genotype-phenotype correlation , with NOMID exhibiting the most severe manifestations [5 , 6] . Each of the CAPS phenotypes displays multiple symptoms , including systemic inflammation , recurrent or chronic fever , and urticaria-like rash [5 , 6] . Consistent with the NLRP3 inflammasome role in interleukin ( IL ) -1β and IL-18 maturation , cryopyrinopathies are associated with excessive production of these cytokines . Accordingly , IL-1-blocking drugs are widely used in the management of these disorders . However , it appears that some CAPS patients only partially respond to IL-1 biologics [7–9] . In addition , skeletal lesions , the hallmark of NOMID , are refractory to IL-1 blockade [10–13] . These clinical observations underscore the complexity of cryopyrinopathies by suggesting that other actions of the inflammasomes beyond maturation of cytokines also contribute to the pathogenesis of these disorders . Indeed , the NLRP3 inflammasome also processes gasdermin D ( GSDMD ) into GSDMD-N ( N-terminal domain ) and GSDMD-C ( C-terminal domain ) [14–16] . GSDMD-N translocates to the plasma membrane , where it binds phospholipids and forms pores at the plasma membrane through which IL-1β and IL-18 are secreted by living cells [17–19] . Sustained activity of the inflammasomes causes excessive maturation of GSDMD and pore formation; this leads to membrane perforation and , ultimately , pyroptosis [17 , 20–23] . This form of cell death provokes the uncontrolled release of not only IL-1β and IL-18 but also cytoplasmic contents , resulting in the recruitment of immune cells and propagation of inflammation [17 , 24] . Thus , pyroptosis is not a silent endpoint , but the extent to which this pathologic process influences the pathogenesis of cryopyrinopathies is unknown . Knockin mice harboring specific mutations found in CAPS patients were engineered in an attempt to generate preclinical disease-relevant models for genotype-phenotype relationship studies [25–28] . These models recapitulate some clinical features though disease manifestations are , in general , more severe in mice than in humans . Nonetheless , these seminal studies revealed that pyroptosis may be responsible for the persistent inflammatory responses in mice with impaired IL-1β and IL-18 signaling [8 , 29] . Here , we used NOMID mice to determine the role that GSDMD and pyroptosis play in this disease model . NOMID mice exhibited systemic inflammation , stunt growth , and damage to multiple organs . These anomalies were absent in NOMID mice lacking GSDMD , which were indistinguishable from wild-type ( WT ) littermates . These results reveal a nonredundant function of GSDMD in the onset and progression of NOMID in mice . The NLRP3 inflammasome complex—which comprises NLRP3 itself , the adapter protein , apoptosis-associated speck-like protein containing a CARD ( ASC ) , and caspase-1—processes pro-IL-1β and pro-IL-18 into IL-1β and IL-18 , respectively [1] . This inflammasome also cleaves GSDMD into GSDMD-N and GSDMD-C [14–16] . GSDMD-N forms pores at the plasma membranes through which IL-1β and IL-18 are secreted by living cells; excessive pore formation causes pyroptosis , a response that can be assessed in vitro by quantifying the release of lactate dehydrogenase ( LDH ) [18 , 19] . Consistent with the literature , GSDMD was cleaved upon stimulation of WT mouse bone marrow–derived macrophages ( BMMs ) with lipopolysaccharide ( LPS ) and nigericin ( Fig 1A ) . Two cleaved GSDMD fragments were detected; whether the larger fragment was further processed to generate the smaller fragment is not known . GSDMD maturation correlated with the release of not only IL-1β ( Fig 1B and S1 Data ) but also LDH ( Fig 1C and S1 Data ) , indicating that BMMs undergo NLRP3 inflammasome-dependent pyroptosis under these experimental conditions . To reinforce this conclusion , GSDMD processing , cytokine production , and pyroptosis were determined using cells isolated from mice lacking GSDMD or components of either the NLRP3 canonical inflammasome ( e . g . , NLRP3 or caspase-1 ) or noncanonical inflammasome ( e . g . , caspase-11 ) . Maturation of IL-1β and GSDMD was impaired in BMMs lacking any component of the classical NLRP3 inflammasome but was unaffected in Casp11 null cells ( Fig 1A–1C ) , as expected . These results strengthen the view that GSDMD is a key effector of the NLRP3 inflammasome pathway . The identification of more than 100 NLRP3 sequence variants underscores the challenges of genotype–phenotype relationship studies for CAPS [30] . In efforts to fill this gap , several preclinical CAPS-relevant models were developed [25–28] . They included knockin mice , which harbored a D301N NLRP3 mutation , the mouse ortholog of the human D303N mutation found in NOMID patients [25] . Mating of Nlrp3fl ( D301N ) /+ mice with lysozyme M-Cre−/+ ( LysM-Cre−/+ ) mice yielded control and Nlrp3fl ( D301N ) /+;LysM-Cre−/+ mice , in which the autosomal dominant mutation in Nlrp3 was induced in myeloid cells; these mice are referred to as NOMID mice . We previously reported that the phenotype of NOMID mice with myeloid-restricted activation of NLRP3 , which included systemic inflammation and skeletal anomalies , resembled that of mice broadly expressing the mutated protein [25 , 31 , 32] . This mouse model provided the opportunity to determine the impact of GSDMD deficiency in the pathogenesis of NOMID . Consistently , GSDMD cleavage in WT BMMs required priming and secondary signals triggered by LPS and nigericin , respectively ( Fig 1D ) . By contrast , GSDMD proteolysis in NOMID BMMs was induced by LPS alone though the response was maximal in the presence of the ionophore . Likewise , secretion of IL-1β and LDH by WT cells necessitated the combined actions of LPS and nigericin , whereas these responses were significantly induced by the endotoxin alone in NOMID cells ( Fig 1E and 1F; and S1 Data ) . Notably , secretion of IL-1β and LDH was abolished in cells lacking GSDMD . Thus , mature IL-1β is constitutively produced in NOMID cells , but its release requires GSDMD . NOMID mice are runted , and they usually die by 2 to 3 weeks of age [25 , 31] , whereas Gsdmd null mice are apparently normal [16] . Consistent with these reports , NOMID pups were indistinguishable from WT and Gsdmd null littermates at birth but exhibited growth retardation and significantly lower body weight by 12 days of age ( Fig 2A and 2B; S1A Fig and S1 Data ) . Additional macroscopic aberrations in NOMID mice included the presence of skin lesions ( Fig 2A ) and splenomegaly ( Fig 2C and 2D; S1 Data ) . Skin and spleen abnormalities and the small body size phenotype of NOMID were all normalized in mutant mice lacking GSDMD ( Fig 2A–2D ) . Growth delay , systemic inflammation , perinatal lethality , and spleen and skin abnormalities have been reported for other models of CAPS [26 , 27 , 29] . Deletion of Il-1 receptor completely abolished these outcomes in NOMID mice but not in FCAS and MWS mice [8 , 29] , findings that are consistent with the view that , in contrast to humans , FCAS and MWS are unexpectedly more severe than NOMID in mice . The release of not only IL-1β and IL-18 but also other pro-inflammatory factors during pyroptosis may be responsible for the persistent residual inflammatory responses in FCAS and MWS models . Thus , it will be informative to determine the effects of GSDMD deficiency on disease progression in other preclinical models of CAPS . While we were wrapping up this work , a report indicated that lack of GSDMD in mice prevented the onset and progression of Familial Mediterranean Fever , a disease in which aberrant pyrin inflammasome activities caused IL-1β oversecretion and pyroptosis [33] . Deficiency in GSDMD also protected mice against endotoxic shock , consistent with activation of this protein by intracellular LPS [14 , 16] . A recent paper suggested an interplay between caspase-8 and caspase-11-GSDMD axis in the execution of endotoxic shock [34] . Collectively , these findings indicate that inactivation of GSDMD arrests pathogenic signals induced by various inflammasomes . IL-1β propagates inflammation through various mechanisms , including perturbation of chemokine and cytokine signaling networks , responses that lead to the expansion and recruitment of neutrophils to several organs . This cytokine also promotes anemia owing to its negative effects on erythroid progenitors and erythropoietin signaling as well as alteration of the expression of ferritin and ferroportin [35–38] . Accordingly , NOMID mice produced higher levels of IL-1β in bone marrow compared to WT counterparts ( Fig 3A and S1 Data ) , a response that correlated with excessive GSDMD processing in vivo in bone marrow , though the cleaved fragment was barely detected in this compartment ( S1B Fig ) . This observation was not unexpected considering that excessive generation of GSDMD-N caused cytolysis; as a result , the cleaved fragment may have been lost during the sampling process . NOMID mice also exhibited peripheral leukocytosis ( Fig 3B and S1 Data ) driven by neutrophilia ( Fig 3C and S1 Data ) and anemia ( Fig 3D; S1C Fig and S1 Data ) , as we previously reported [25 , 32] . The identity of myeloid cell subpopulations , which are prone to pyroptosis and propagate inflammation , in this model is unknown , a knowledge gap that future studies should address . In any case , while Gsdmd ablation had no effect on the number of blood cells compared to WT mice , it abrogated or attenuated the onset of leukocytosis and anemia in NOMID mice ( Fig 3A–3D ) . Accordingly , the bone marrow compartment of NOMID mice contained abnormally high levels of Gr1+/CD11b+ cells and low levels of Ter119+ cells ( Fig 3E and 3F; S1 Data ) , responses that were normalized upon Gsdmd deletion . Histological analyses showed massive neutrophilic infiltration in the liver , dermal and hypodermal layers of the skin , and the spleen of NOMID mice compared to WT or Gsdmd−/− counterparts ( Fig 4 ) . Inflammation in the spleen was characterized by disorganized structures of white and red pulps . Because skeletal complications—including low bone mass—are hallmarks of NOMID , we investigated these outcomes in NOMID mice . Histological examinations of skeletal elements showed disorganized columns of chondrocytes with profoundly altered morphology . The epiphysis was hypocellular ( Fig 4 ) , a phenotype that was previously reported to be caused by massive chondrocyte death [25 , 32] and reminiscent of the human disease [39] . The number of osteoclasts , cells responsible for bone resorption , was markedly increased in NOMID mice relative to control mice . Remarkably , all organs that were analyzed in NOMID;Gsdmd−/− mice were all spared from inflammation-induced damage ( Fig 4 ) . Thus , deletion of GSDMD abolishes inflammatory responses and organ demise in NOMID mice . Blockade of IL-1 activity has been the main strategy for neutralizing pathogenic signals induced by this cytokine in CAPS and other autoinflammatory disorders . However , these drugs have shortcomings , including high cost and the requirement for parenteral administration . Thus , there is still a medical need for the development of safe and affordable drugs for the treatment of autoinflammatory diseases . Breakthrough research demonstrating that GSDMD-mediated pyroptosis releases cytoplasmic contents , including IL-1β and IL-18 , offers a novel node for therapeutic intervention . Stemming from its mechanisms of action , blockade of GSDMD and subsequently pyroptosis should , in theory , provide superior efficacy compared with targeted blockade of IL-1β . The compelling evidence indicating that inactivation of GSDMD blocks inflammatory responses induced by the NLRP3 inflammasome lends support to discovery efforts aimed at identifying selective inhibitors of GSDMD actions . All procedures were approved by the Institutional Animal Care and Use Committee ( IACUC ) of Washington University School of Medicine in St . Louis , Missouri . All experiments were performed in accordance with the relevant guidelines and regulations described in the IACUC-approved protocol number 20160245 . Gsdmd−/− mice [16] were kindly provided by Dr . V . M . Dixit ( Genentech , South San Francisco , CA ) . Nlrp3−/− , Casp1−/− , and Nlrp3fl ( D301N ) /+ mice and lysozyme M ( LysM ) -Cre mice have previously been described [25 , 31 , 40 , 41] . Casp11−/− mice were purchased from The Jackson Laboratory . All mice were on the C57BL6J background , and mouse genotyping was performed by PCR . Complete blood counts were performed by the Washington University School of Medicine DCM Diagnostic Laboratory as previously described [31] . Bone marrow cells were flushed out as previously described , and photographed [31] . BMMs were obtained by culturing mouse bone marrow cells in culture media containing a 1:25 dilution of supernatant from the fibroblastic cell line CMG 14–12 as a source of M-CSF , a mitogenic factor for BMMs , for approximately 5 days in a 10-cm dish following the procedures that we published [42] . Nonadherent cells were removed by vigorous washes with PBS , and adherent BMMs were detached with trypsin-EDTA and were cultured in culture media containing a 1:50 dilution of CMG at 4 × 104/well in a 96-wells plate ( for the analysis of IL-1β and LDH ) or 1 . 2 × 106/well in a 6-wells plate ( for Western blot analysis ) . BMMs were treated with 100 ng/mL LPS for 3 hours , then with 15 μM nigericin for 30 minutes . Extracts from BMMs or bone marrow cells were prepared by lysing cells or cell pellets , respectively , with RIPA buffer ( 50 mM Tris , 150 mM NaCl , 1 mM EDTA , 0 . 5% NaDOAc , 0 . 1% SDS , and 1 . 0% NP-40 ) plus phosphatase inhibitors and Complete Protease Inhibitor Cocktail ( Roche , Brighton , MA ) . Protein concentrations were determined by the Bio-Rad method , and equal amounts of proteins were subjected to SDS-PAGE gels ( 12% ) . Proteins were transferred onto nitrocellulose membranes and incubated with GSDMD antibody ( 1:1 , 000 , ab209845 , Abcam , Cambridge , MA ) or β-actin ( 1:5 , 000 , sc-47778 , Santa Cruz Biotechnology , Dallas , Texas ) overnight at 4°C , followed by a 1-hour incubation with secondary goat anti-mouse IgG ( 1:5 , 000 , A21058 , Thermo Fisher Scientific , Grand Island , NY ) or goat anti-rabbit IgG ( 1:5 , 000 , A21109 , Thermo Fisher Scientific ) , respectively . The results were visualized using Li-Cor Odyssey Infrared Imaging System ( LI-COR Biosciences , Lincoln , Nebraska ) . Mouse bone marrow cells were flushed from femur and tibia . For flow cytometry analysis of the leukocytes , red blood cells ( RBCs ) were depleted with RBC lysis buffer ( Roche , Brighton , MA ) . Cells ( 0 . 5–1 × 106 ) were incubated with Fc block ( anti-mouse CD16/32 , BioLegend , San Diego , CA ) to block nonspecific Fc binding , stained with isotype control or APC-anti-mouse Ter119 ( BioLegend , San Diego , CA ) , FITC-anti-mouse CD11b ( eBioscience , Grand Island , NY ) , and PE-anti-mouse Ly-6G/Ly-6C ( Gr1 ) antibody ( BioLegend , San Diego , CA ) according to the supplier’s instructions . Flow cytometry was performed using BD LSRFortessa or BD FACSCanto II Flow Cytometer system , followed by analysis with FlowJo software ( Tree Star , Ashland , Oregon ) . All tissues were harvested and fixed in 10% formalin . Long bones were decalcified in 14% ( w/v ) EDTA for 5 days at room temperature . All tissues were embedded in paraffin , sectioned at 5 μm thickness , and mounted on glass slides . Sections were stained with hematoxylin–eosin ( HE ) or TRAP as previously described [42] . BMMs were treated with 100 ng/mL LPS for 3 hours , then with 15 μM nigericin for 30 minutes . Cell death was assessed by the release of LDH using LDH Cytotoxicity Detection Kit ( TaKaRa , Mountain View , CA ) . IL-1β levels in conditioned media were measured by ELISA ( eBiosciences , Grand Island , NY ) . For IL-1β measurements in bone marrow , flushed bone marrow was centrifuged , and the supernatants were collected as described previously [42] . IL-1β levels were quantified using the eBioscience ELISA kit . Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test or two-way ANOVA with Tukey's multiple comparisons test in GraphPad Prism 7 .	The NLRP3 inflammasome plays an important role in the maturation of interleukin ( IL ) -1β and IL-18 . Accordingly , NLRP3 gain-of-function mutations , which cause a spectrum of autoinflammatory disorders known as cryopyrin-associated periodic syndromes ( CAPS ) , are associated with excessive IL-1β and IL-18 production . Although CAPS-associated inflammatory symptoms are treated with IL-1-blocking agents , emerging evidence indicates that some CAPS patients only partially respond to these drugs . Persistent inflammatory responses have also been reported in CAPS mice deficient in IL-1β and IL-18 signaling and may be the consequences of the pro-inflammatory cell death , pyroptosis , which is induced by gasdermin D ( GSDMD ) , the other effector of the inflammasomes . Consistent with this view , we found that damage to multiple organs that manifested in a mouse model of CAPS was prevented by ablation of GSDMD .	[ "Abstract", "Introduction", "Results", "and", "discussion", "Materials", "and", "methods" ]	[ "flow", "cytometry", "inflammatory", "diseases", "innate", "immune", "system", "medicine", "and", "health", "sciences", "immune", "physiology", "cytokines", "pathology", "and", "laboratory", "medicine", "spleen", "immunology", "animal", "models", "bone", "marrow", "ce...	2018	Gasdermin D mediates the pathogenesis of neonatal-onset multisystem inflammatory disease in mice
Why are mitochondria almost always inherited from one parent during sexual reproduction ? Current explanations for this evolutionary mystery include conflict avoidance between the nuclear and mitochondrial genomes , clearing of deleterious mutations , and optimization of mitochondrial-nuclear coadaptation . Mathematical models , however , fail to show that uniparental inheritance can replace biparental inheritance under any existing hypothesis . Recent empirical evidence indicates that mixing two different but normal mitochondrial haplotypes within a cell ( heteroplasmy ) can cause cell and organism dysfunction . Using a mathematical model , we test if selection against heteroplasmy can lead to the evolution of uniparental inheritance . When we assume selection against heteroplasmy and mutations are neither advantageous nor deleterious ( neutral mutations ) , uniparental inheritance replaces biparental inheritance for all tested parameter values . When heteroplasmy involves mutations that are advantageous or deleterious ( non-neutral mutations ) , uniparental inheritance can still replace biparental inheritance . We show that uniparental inheritance can evolve with or without pre-existing mating types . Finally , we show that selection against heteroplasmy can explain why some organisms deviate from strict uniparental inheritance . Thus , we suggest that selection against heteroplasmy explains the evolution of uniparental inheritance . Our model is based on an idealized life cycle of a single-cell diploid eukaryotic organism , such as the algae Chlamydomonas reinhardtii . Diploid cells contain n mitochondria and haploid cells have n/2 mitochondria . All mitochondria are initially wild type but mitochondria can mutate from wild type to mutant ( and vice versa ) . The starting population contains haploid gametes with a nuclear allele regulating biparental inheritance ( B ) . Gametes are evenly split between two nuclear self-incompatible mating types ( B1 and B2 ) . In the basic model , we assume no recombination between the mitochondrial inheritance and mating type loci because these are tightly linked in many isogamous organisms [9] ( later we explore recombination and no mating types ) . Cell types are characterized by the proportion of wild type and mutant mitochondria that they carry and their nuclear allele ( haploid ) or genotype ( diploid ) . Our life cycle has four discrete stages and is similar to the life cycles used in previous models [1 , 5 , 8] . Since we begin with a population of gametes , the first stage is random mating . Here , gametes randomly mate with the opposite mating type to produce diploid cells . Matings are controlled by the nuclear allele in gametes . In biparental inheritance ( between B1 and B2 gametes ) , both gametes contribute mitochondria to the B1B2 diploid cells ( see later for uniparental inheritance ) . The second stage is mutation . Each mitochondrion can mutate to the other haplotype with probability μ . The third stage is selection . Here , diploid cells have a relative fitness based on the proportion of each haplotype in the cell . We assume that fitness decreases as the level of heteroplasmy increases . The fourth stage is meiosis , where diploid cells produce gametes that contain a single nuclear allele and n/2 mitochondria . As mitochondria are stochastically partitioned into gametes [9] , diploid heteroplasmic cells produce gametes with varying degrees of heteroplasmy . First , we let the population of B1 and B2 gametes reach mutation-selection equilibrium . We then simulate a mutation leading to uniparental inheritance of mitochondria by converting a small proportion ( 10−2 ) of B1 gametes to U1 gametes . We assume no further mutations between B and U alleles . Matings between U1 and B2 gametes result in uniparental inheritance , in which the U1B2 cell inherits mitochondria from U1 alone . ( Matings between U1 and B1 are not possible as they are the same mating type . ) The population now consists of three alleles ( U1 , B1 and B2 ) and two genotypes ( U1B2 and B1B2 ) . The model tracks the proportion of each cell type at each stage of the life cycle . U1 spreads at the expense of B1 when uniparental inheritance is more advantageous than biparental inheritance ( the frequency of B2 always remains at 0 . 5 ) , and the simulation ends when the alleles reach equilibrium ( see Model and S1–S6 Model for details of the model ) . To explore whether a cost to heteroplasmy could have led to the evolution of uniparental inheritance , we study several scenarios . We first examine the simplest case , where mutations in mitochondria are neither advantageous nor disadvantageous ( neutral mutations ) , but heteroplasmic cells incur a fitness cost proportional to the degree of heteroplasmy . Because no empirical data relate fitness to the degree of heteroplasmy , we consider three forms of fitness function to describe selection against heteroplasmy: concave , linear and convex ( Fig 1A ) . For each fitness function , we vary the cost of heteroplasmy ( ch ) , given by ch = 1 − h where h is the fitness of the most heteroplasmic cell in the population , to see how this affects the spread of U1 . We generate the concave fitness function by w ( i ) ={1−ch ( in/2 ) 2for0≤i<n/2 , 1−ch ( n−in/2 ) 2forn/2≤i≤n , the linear function by w ( i ) ={1−ch ( in/2 ) for0≤i<n/2 , 1−ch ( n−in/2 ) forn/2≤i≤n , and the convex function by w ( i ) ={1−chin/2for0≤i<n/2 , 1−chn−in/2forn/2≤i≤n . We also vary μ ( mutation rate ) and n ( number of mitochondria ) to ensure that our findings are robust . Second , we explore the effect of advantageous or deleterious mutations ( non-neutral mutations ) on the spread of U1 . Third , we relax the assumption of tight linkage between mating type and inheritance loci by exploring two cases: recombination between mating types and the absence of mating types altogether . Finally , we examine whether selection against heteroplasmy can explain the rare , but nevertheless important , exceptions to uniparental inheritance . To ensure that our results generalize to more than two mitochondrial types , we developed a second model that considers three mitochondrial types ( S6 Model ) . We find that U1 always replaces B1 , resulting in complete uniparental inheritance in the population ( Fig 1B ) . These findings are independent of the number of mitochondria per cell ( Fig 1C ) , mutation rate ( Fig 1D ) , fitness function ( Fig 1E ) , and cost of heteroplasmy ( Fig 1F ) ( see S1–S10 Tables for more parameter combinations ) . We find the same results when we generalize the model to three mitochondrial haplotypes ( S1 Fig ) . In our model , heteroplasmic cells are generated by mutation . During meiosis , heteroplasmic cells produce gametes with varying levels of heteroplasmy , including homoplasmic gametes . Uniparental inheritance maintains this variation created by meiosis , which leads to homoplasmic U1B2 cells ( Fig 2A–2B and S2A–S2B Fig ) . Mutants that arise in U1B2 cells quickly segregate into U1 gametes that carry mutant haplotypes only ( Fig 3A–3B and S3A–S3B Fig ) , which leads to U1B2 cells that are homoplasmic for mutant mitochondria ( Fig 2B and S2B Fig ) . Since we assume that mutations are neutral , cells homoplasmic for mutant mitochondria suffer no fitness costs . U1B2 cells carrying mutant mitochondria produce B2 gametes that also carry mutant mitochondria ( Fig 3D and S3D Fig ) . When these B2 gametes mate with B1 gametes carrying wild type mitochondria , the resulting B1B2 cells are highly heteroplasmic ( Fig 2C–2E and S2C Fig ) . As U1 spreads , matings between U1 and B2 become more likely , increasing the level of heteroplasmy in both B1B2 cells and in B1 and B2 gametes ( Figs . 2C–2E and 3C–3F and S2C and S3C–S3D Figs . ) . Increased levels of heteroplasmy reduce the fitness of both B1 and B2 gametes ( w¯B1 , w¯B2 in Fig 3A and S3A Fig ) and B1B2 cells ( w¯B1B2 in Fig 2A and S2A Fig ) . The difference in fitness between B1 and B2 becomes stronger ( Fig 3A and S3A Fig ) as more B2 gametes that carry mutant mitochondria are produced ( Fig 3D and S3D Fig ) . As a result U1 spreads at the expense of B1 . In the above description ( Figs . 2 and 3 ) , the mutation from B1 to U1 occurred in gametes homoplasmic for wild type mitochondria . When U1 is introduced into heteroplasmic gametes , it takes fewer generations to reach equilibrium because B2 gametes homoplasmic for mutant mitochondria are produced more quickly ( S4 Fig ) . Our results are robust to changes in the frequency at which U1 gametes are introduced ( S5 Fig ) . For more detailed model dynamics , see S1 Text and S1–S2 Videos . U1 spreads more slowly when mutation rate ( μ ) is lower ( Fig 1D ) and number of mitochondria ( n ) is higher ( Fig 1C ) . Reducing μ slows the spread of U1 because mutant mitochondria are produced more slowly , slowing the generation of B2 gametes that only carry the mutant haplotype . Increasing n has the same effect . While varying the cost of heteroplasmy does not change the qualitative behavior of the model , it does affect the number of generations required for U1 to replace B1 ( Fig 1F ) . In general , U1 spreads more quickly when the cost of heteroplasmy is low for all three fitness functions ( Fig 1F ) . Strong selection against heteroplasmy ( e . g . ch = 1 ) slows the production of B2 gametes homoplasmic for the mutant haplotype because a transition via heteroplasmy is needed to lead to U1B2 cells homoplasmic for mutant mitochondria . Heteroplasmy levels thus remain low in B1B2 cells , and U1 takes longer to replace B1 ( S6A and S6D Fig ) . At lower costs of heteroplasmy ( e . g . ch = 0 . 2 ) , more B2 gametes that are homoplasmic for the mutant haplotype are produced and levels of heteroplasmy in B1B2 cells increase , leading to a faster spread of U1 ( S6B and S6E Fig ) . Although levels of heteroplasmy in B1B2 cells increase even further as the cost of heteroplasmy approaches 0 ( e . g . ch = 0 . 01 ) , selection against heteroplasmy is now very weak , which slows the spread of U1 compared with ch = 0 . 2 ( S6C and S6F Fig ) . When the number of mitochondria is higher , U1 spreads more quickly when the cost of heteroplasmy is low . This is because B2 gametes homoplasmic for mutant mitochondria are produced more slowly at higher values of n and strong selection against heteroplasmy compounds this problem ( S7 Fig ) . A similar logic can be applied to understand the differences between the three fitness functions . Since heteroplasmic cells are under weaker selection when fitness is concave ( followed by linear and convex respectively ) ( Fig 1A ) , the level of heteroplasmy is highest using a concave function ( S8 Fig ) . Thus , U1 spreads more quickly using a concave function ( followed by linear and convex respectively ) when the cost of heteroplasmy is high because it is easier to generate heteroplasmic cells , and thus easier to generate B2 gametes homoplasmic for mutant mitochondria , when selection against heteroplasmic cells is weaker ( Fig 1F and S8 Fig ) . As the cost of heteroplasmy decreases , the number of generations for U1 to spread under the three fitness functions converges because it becomes easier to generate B2 gametes homoplasmic for mutant mitochondria ( Fig 1F ) . We next investigate how the U1 allele spreads when mutations are non-neutral , as is the case for most mtDNA mutations [13] . We start by assuming that mutations are deleterious so that cells carrying mutant mitochondria are more strongly selected against than cells that carry wild type mitochondria . We assume that a mutation from wild type to mutant haplotype is more common than the reverse [14] . We let the probability of a mutation from mutant to wild type haplotype be μb = μ/100 . We vary the selection coefficient of the mutant haplotype to see how this affects the spread of the U1 allele ( the fitness of a cell homoplasmic for the mutant haplotype is 1 − sd , where sd is the selection coefficient of the mutant haplotype ) . Essentially there are now two fitness functions: one governing the effect of mitochondria on cell fitness ( where the selection coefficient determines the magnitude of the effect ) and one governing the cost of heteroplasmy . For deleterious mutations , we assume that fitness decreases as a concave function of the number of mutants , as this relationship is experimentally established [15] . We examine both concave and convex fitness functions for selection against heteroplasmy ( yielding two combinations ) . Again , U1 replaces B1 unless the fitness of heteroplasmic cells and the fitness of deleterious mutants are governed by a concave function and the selection coefficient is sufficiently large ( S9 Fig and S11–S12 Tables ) . U1 generally spreads more slowly as sd increases and it always spreads more slowly compared to when mutations are neutral ( S11–S12 Tables ) . Stronger selection against mutant haplotypes leads to fewer B2 gametes homoplasmic for mutant mitochondria , which slows the spread of U1 ( S10 Fig ) . Next we explore the effect of advantageous mutations on the spread of U1 . In this case , cells that carry mutant haplotypes have an advantage over those carrying wild type haplotypes ( the fitness of a cell homoplasmic for the wild type haplotype is 1 − sa , where sa is the selection coefficient of the mutant haplotype ) . We account for the rarity of advantageous mutations by setting μb = 100μ . Because it is unknown how fitness relates to the accumulation of advantageous mtDNA mutations , we model this relationship with both a concave and convex function . As in the deleterious case , we model selection against heteroplasmy by testing both concave and convex fitness functions ( giving four combinations ) . U1 always replaces B1 unless mutations are highly advantageous ( sa = 0 . 1 ) and both the fitness of heteroplasmic cells and the fitness of advantageous mutants are governed by a concave function ( S9 Fig and S13–S14 Tables ) . U1 spreads more quickly when sa = 0 . 001 and sa = 0 . 01 because B2 gametes homoplasmic for mutant haplotypes now have a fitness advantage and are produced more quickly ( S10 Fig ) . In contrast , U1 spreads more slowly when sa = 0 . 1 because the mutant haplotype quickly replaces the wild type as the dominant haplotype before U1 has replaced B1 . Once B1 gametes carry mostly mutant haplotypes , B1 × B2 matings are less costly because they predominantly involve mutant haplotypes . We find the same patterns for non-neutral mutations when we generalize our model to three mitochondrial types ( S15 Table ) . Previously , U × U matings were not possible because we assumed tight linkage between mating type and inheritance loci . But if we allow recombination to occur between these loci , U1 × U2 matings become possible . In this scenario , the number of gametes increases to four ( B1 , B2 , U1 and U2 ) , as does the number of genotypes ( B1B2 , U1B2 , U1U2 and U2B1 ) . There are three main ways in which mitochondrial inheritance could be regulated in U1 × U2 matings . ( 1 ) One U allele is dominant to the other , leading to uniparental inheritance; ( 2 ) each U allele ensures inheritance of its mitochondria , resulting in biparental inheritance; or ( 3 ) inheritance is more or less random so that some matings result in uniparental inheritance and some in biparental inheritance . We model all three cases . When U1 × U2 matings lead to uniparental inheritance , the U1U2 genotype always spreads until it is fixed in the population , leading to complete uniparental inheritance ( Fig 4A and S16–S18 Tables ) . When U1 × U2 matings lead to biparental inheritance , however , uniparental inheritance does not become fixed and the population reaches a polymorphic equilibrium ( Fig 4B–4C ) . Under these conditions , the frequency of uniparental inheritance at equilibrium is ≤ 0 . 5 ( S19–S21 Tables ) . Uniparental inheritance cannot exceed 0 . 5 because increasing the frequency of U1 or U2 simply increases the proportion of biparental U1 × U2 matings . The frequency of uniparental inheritance remains very low when we assume a concave fitness function ( Fig 4B ) , but reaches its maximum ( 0 . 5 ) when we assume a linear or convex fitness function ( Fig 4C ) ( see S12–S13 Figs . for an explanation ) . When the probability of recombination ( Pr ) is sufficiently high ( 10−4 ≤ Pr ≤ 0 . 5 in S11 Fig ) , the U1B2 and U2B1 genotypes have the same frequency at equilibrium ( S11B–S11D Fig ) . Now uniparental inheritance is no longer associated with a single mating type but is evenly split between the two mating types ( S19–S21 Tables ) . When Pr is sufficiently small ( Pr = 10−5 in S11 Fig ) , the recombination rate is so low that the mating type and inheritance loci are essentially linked and the U1B2 genotype becomes fixed ( as in the general model ) ( S11A Fig ) . When we assume a mixture of uniparental inheritance and biparental inheritance , we let U1 × U2 matings lead to biparental inheritance with probability Pb and to uniparental inheritance with probability 1 − Pb . Lowering Pb increases the frequency of uniparental inheritance , and uniparental inheritance becomes fixed when Pb = 0 ( Fig 4A and 4E ) . Under linear and convex fitness functions , the equilibrium always maximizes the level of uniparental inheritance ( Tables S22–S23 ) . Under concave fitness , however , uniparental inheritance is only maximized for particular values of Pb ( roughly Pb ≤ 0 . 2 for the parameter values we considered ) ( S22 Table; rows 2–3 ) . ( See S5 Model for how we determine when uniparental inheritance is maximized . ) We also find that uniparental inheritance can evolve in the complete absence of mating types . The no mating types scenario differs from the recombination case in that UB equals the sum of U1B2 and U2B1 at equilibrium ( Fig 4A and 4F ) ( see S2 Text for more details ) . In this section , we explore whether relaxing some of the assumptions in our general model can lead to mitochondrial inheritance patterns that resemble some of the known exceptions to uniparental inheritance . Exceptions to uniparental inheritance fall in three main categories: organisms that ( 1 ) regularly inherit mitochondria from both parents; ( 2 ) normally inherit mitochondria from one of the two parents but on occasion inherit mitochondria from both; and ( 3 ) inherit mitochondria from either or both parents . Baker’s yeast , Saccharomyces cerevisiae , regularly inherits mitochondria from both parents ( though uniparental inheritance also occurs ) , but heteroplasmy is transient because the diploid cell has only a few mitochondria [16] and divides repeatedly , which separates heteroplasmic cells into cells homoplasmic for either mitochondrial type ( vegetative segregation ) [9 , 10] . Vegetative segregation is usually completed within twenty generations , but up to 50% of zygotes may be homoplasmic after the first division ( [10] and references therein ) . Thus , Saccharomyces may restore homoplasmy as quickly as organisms that actively destroy one mitochondrial lineage [17] . Similarly , the geranium Pelargonium zonale often inherits cytoplasmic organelles from both parents ( chloroplasts in this case ) . As with Saccharomyces , heteroplasmy is transient in Pelargonium because of rapid vegetative segregation of heteroplasmic cells shortly after syngamy [9] . We added mitotic divisions to our model to test whether vegetative segregation could maintain biparental inheritance under selection against heteroplasmy . When we include mitosis before selection ( which assumes that vegetative segregation occurs swiftly , before selection has time to act ) , uniparental inheritance does not spread , provided that the number of mitochondria is low ( n = 4 ) and the number of divisions is high ( S24 Table; rows 7–8 ) . Under these conditions , biparental inheritance is stable because heteroplasmic cells resulting from biparental inheritance segregate into homoplasmic cells before selection acts . If there are insufficient mitotic divisions , or if selection acts before vegetative segregation is complete , then uniparental inheritance replaces biparental inheritance , although it spreads much more slowly than when there are no mitotic divisions ( S24 ( rows 3–6 ) and S25 Tables ) . When there are more mitochondria per cell ( e . g . n = 8 ) , biparental inheritance is only stable if the number of cell divisions increases to compensate ( S24 Tables; rows 9–10 ) . Thus , biparental inheritance can be stable under selection against heteroplasmy but only under a narrow set of conditions , explaining why this form of inheritance is rare . In other isogamous organisms , including the acellular slime molds Physarum polycephalum and Didymium iridis and the algae Chlamydomonas reinhardtii , mitochondria from both gametes mix before one mitochondrial lineage is destroyed post-fertilization , often by nucleases [18–20] . This mechanism is not perfect and these organisms sometimes deviate from strict uniparental inheritance [9 , 18–20] . While uniparental inheritance is the norm in the slime mold P . polycephalum , sometimes both mitochondrial lineages survive , leading to varying degrees of biparental inheritance [18] . Could uniparental inheritance still spread under such conditions ? Since mating types and inheritance loci are tightly linked in Physarum [18] , we explore this question using our general model that assumes linkage . Now , U1 × B2 matings lead to biparental inheritance with probability Pb and to uniparental inheritance with probability 1 − Pb . For the parameter values that we examined , the U1B2 genotype always goes to fixation when Pb < 1 and the fitness function is linear or convex ( S26 Table ) . ( When fitness is concave , Pb must be roughly <0 . 05 for the U1B2 genotype to become fixed . ) Under these conditions , the frequency of biparental inheritance at equilibrium is equal to Pb ( S26 Table ) . In this scenario , the level of biparental inheritance in the population simply reflects the likelihood that an individual mating results in biparental inheritance . Chlamydomonas reinhardtii and Didymium iridis can inherit mitochondria from either or both parents [19 , 20] . Chlamydomonas normally inherits mitochondria from the mt − parent and chloroplasts from the mt + parent , but under some circumstances it can inherit mitochondria from mt + and chloroplasts from mt − or mitochondria and chloroplasts from both [20] . Didymium iridis has random , biased , or dominant patterns of uniparental inheritance . Under random uniparental inheritance , either parental strain is equally likely to be the mitochondrial donor while , under biased inheritance , one strain is more likely to be the mitochondrial donor [19] . Under dominant inheritance , one strain is always the donor . Didymium also has low levels of biparental inheritance [19] . In this scenario , we test whether selection against heteroplasmy could lead to the evolution of a system with a mixture of uniparental inheritance ( from either parent ) and biparental inheritance . We assume that mating types can recombine and that U1 × U2 matings can lead to mitochondria being inherited from U1 , U2 or both . Mitochondria are inherited from U1 with probability PU1 , from U2 with probability PU2 and from both parents with probability Pb ( where PU1+PU2+Pb=1 ) . Now , uniparental inheritance comes from U1 × B2 matings , U2 × B1 matings and those U1×U2 matings with uniparental inheritance . Irrespective of the values of PU1 and PU2 , we find the same results as with our earlier model in which U1×U2 matings led to a mixture of uniparental and biparental inheritance ( S22–S23 Tables ) . This is because equilibrium depends only on the value of Pb . ( Since uniparental inheritance quickly eliminates most heteroplasmic cells , U1U2 cells are almost entirely homoplasmic regardless of which gamete donates mitochondria . ) Consequently , different probabilities of inheriting mitochondria biparentally ( Pb ) , from mating type 1 ( PU1 ) or from mating type 2 ( PU2 ) lead to a range of inheritance patterns that include uniparental inheritance ( from both parents ) and biparental inheritance ( see S27 Table for some examples ) . Lastly , selection against heteroplasmy provides an explanation for the cases in which mitochondria are inherited from one parent while chloroplasts are inherited from the other ( e . g . in Chlamydomonas and pines [20 , 21] ) . If uniparental inheritance simply evolved to maintain homoplasmy in cells , it should not matter which parent donates mitochondria or chloroplasts . Selection against heteroplasmy has implications for the evolution of the mitochondrial genome . Because of a smaller effective population size , which is more strongly affected by genetic drift , and higher mutation rates , mtDNA should be less conserved than the nuclear genome [35 , 36] . Indeed , mitochondrial transfer RNAs and synonymous sites mutate 5–50 times more frequently than comparable elements in the nuclear genome [35 , 37] . Because the mitochondrial genome is effectively asexual , any deleterious mutations in the fittest haplotype cannot be rescued ( except by unlikely back mutations ) . This effect , known as Muller's Ratchet , should eventually lead to irreparable genome meltdown [38 , 39] . In stark contrast to theoretical predictions , however , mitochondrial coding genes are more conserved than analogous nuclear oxidative phosphorylation genes [36] . When mtDNA mutates , only one of the many mtDNA molecules in the cell is affected , leading to a heteroplasmic cell . Selection against heteroplasmy should reduce the probability that mtDNA mutations spread throughout the cell , which , in turn , should oppose changes to mtDNA . Thus , selection against heteroplasmy may not only explain the evolution of uniparental inheritance but also why mitochondrial coding genes have thus far managed to resist the effects of Muller's Ratchet . The starting population is evenly split between B1 and B2 gametes , and all gametes contain type wild type mitochondria ( i . e . P ( M0 , τ1= ( 0 , B1 ) ) =0 . 5 , P ( M0 , τ1= ( 0 , B2 ) ) =0 . 5 and P ( M0 , τ1= ( p , g ) ) =0 , ∀p>0andg=U1 ) . We first allow this population to reach equilibrium , which we define as the point at which the proportion of cell types change by less than 10−12 ( except when the probability that a mitochondrion mutates into another mitochondrion is 10−10 ( μ = 10−10 ) , in which case we define equilibrium to be a change of less than 10−13 ) . We then introduce U1 gametes that are homoplasmic for wild type mitochondria by setting P ( Mge1 , τ1= ( 0 , U1 ) ) =0 . 01 , where ge1 is the number of generations taken to reach the first equilibrium . To maintain the total proportion of the population at 1 , we remove the corresponding proportion of cells from the B1 population ( i . e . P ( Mge1 , τ1= ( 0 , U1 ) ) =P ( Mge1 , τ1= ( 0 , U1 ) ) −0 . 01 ) . In two instances , we alter the way in which U1 is introduced . In S4 Fig , we introduce U1 into the most heteroplasmic gamete with a frequency greater than 0 . 01 , and in S5 Fig we vary the introductory frequency of U1 . Our life cycle is very similar to the life cycle used by Hadjivasiliou and colleagues [1] , which examined the genomic conflict , mutational clearance , and mitochondrial-nuclear coadaptation hypotheses . Gametes with n / 2 mitochondria randomly mate with the opposite mating type to produce diploid cells containing n mitochondria . In effect , this is random mating in which all matings between the same mating type ( i . e . U1U1 , B1B1 , B2B2 and U1B1 ) are lethal , and the only viable genotypes are U1B2 and B1B2 . Consider a biparental mating involving a gamete in state Mt , τ1= ( p , B1 ) , where τ1 is the gamete stage of the life cycle . For this gamete to produce a diploid cell with type Mt , τ2= ( i , B1B2 ) , where τ2 is the diploid cell stage of the life cycle that precedes mutation , it must mate with a gamete of type Mt , τ1= ( i−p , B2 ) . The probability of this mating is 2P ( Mt , τ1= ( p , B1 ) ) P ( Mt , τ1= ( i−p , B2 ) ) , where the factor of 2 accounts for the two ways in which we can choose B1 and B2 ( B1 then B2 or B2 then B1 ) . We restrict the values of p and i – p to biologically valid combinations . First , 0 ≤ p ≤ n / 2 , as the B1 gamete cannot carry negative numbers of mutant mitochondria nor can it contain more mutant mitochondria than the total number of mitochondria in the gamete . Likewise , 0 ≤ i – p ≤ n / 2 for the B2 gamete , which , when rearranged , gives i – ( n / 2 ) ≤ p ≤ i . Valid values for p lie in the range of intersection of these two inequalities , giving max ( 0 , i – ( n / 2 ) ) ≤ p ≤ min ( n / 2 , i ) . We can thus obtain the probability of forming any given diploid cell type after random mating with the sum , P ( Mt , τ2= ( i , B1B2 ) ) =2 ( ∑p=max ( 0 , i−n/2 ) min ( n/2 , i ) P ( Mt , τ1= ( p , B1 ) ) P ( Mt , τ1= ( i−p , B2 ) ) ) . Because uniparental matings between U1 and B2 gametes contain mitochondria from U1 alone , U1B2 cells initially have n / 2 mitochondria . To restore the total complement of n mitochondria , we sample n / 2 mitochondria with replacement from the n / 2 mitochondria in the U1B2 cell and add the n / 2 sampled mitochondria to the original set of mitochondria to form a cell with n mitochondria . For a gamete with identity Mt , τ1= ( p , U1 ) to produce a diploid cell with identity Mt , τ2= ( i , U1B2 ) , it must sample n / 2 mitochondria containing i – p mutant mitochondria and n / 2 – ( i – p ) wild type mitochondria . The mitochondrial state of the B2 gamete is irrelevant because its mitochondria are discarded and we will refer to this cell as Mt , τ1= ( r , B2 ) . Sampling of the n / 2 mitochondria follows a binomial distribution , which we denote T ( i – p;n / 2 , ( 2p ) / n ) , where i – p refers to the number of mutant mitochondria that need to be sampled , n / 2 refers to the number of mitochondria being sampled , and ( 2p ) / n is the probability of drawing a single mutant mitochondrion from a U1B2 cell with p ( out of n / 2 ) mutant mitochondria ( where ( 2p ) / n is obtained by rearranging p / ( n / 2 ) ) . The probability of sampling i – p mutant mitochondria ( and ( n / 2 ) – ( i – p ) wild type mitochondria ) is given by T ( i−p;n2 , 2pn ) = ( n/2i−p ) ( 2pn ) i−p ( 1−2pn ) n2−i−p . The restrictions on p and i – p are the same as those in biparental mating . Since U1 will form the same initial U1B2 cell regardless of the B2 gamete with which it mates , the probability of producing each type of U1 gamete is multiplied by the probability of selecting any B2 gamete . The probability of forming a given U1B2 cell after random mating is determined by P ( Mt , τ2= ( i , U1B2 ) ) =∑p=max ( 0 , i−n2 ) min ( n2 , i ) ( 2P ( Mt , τ1= ( p , U1 ) ) T ( i−p;n2 , 2pn ) ∑r=0n2P ( Mt , τ1= ( r , B2 ) ) ) . We denote the post-mutation states of cells as Mt , τ3= ( i , G ) , ( where τ3 indicates the post-mutation life cycle stage ) . If we define the number of wild type mitochondria that mutate to mutant mitochondria to be a and the number of mutant mitochondria that mutate to wild type mitochondria as b , a post-mutation cell in state Mt , τ3= ( i , G ) must be derived from a pre-mutation cell in state Mt , τ2= ( i−a+b , G ) ( because the pre-mutation cell gains a mutant mitochondria and loses b mutant mitochondria to form the post-mutation cell ) . Similarly , if the post-mutation cell has j wild type mitochondria , then the pre-mutation cell must have j + a – b wild type mitochondria , where j = n – i . First , we must work out the probability that a cell mutates a of its wild type mitochondria to mutant mitochondria . We define Y ( a;n – i + a – b , μ ) as the probability that a pre-mutation cell has a mutations in its n – i + a – b wild type mitochondria given that each mitochondrion mutates with probability μ . The accumulation of mutations is binomially distributed such that Y ( a;n−i+a−b , μ ) = ( n−i+a−ba ) μa ( 1−μ ) n−i−b . Similarly , we define Y ( b;i – a + b , μb ) to be the probability that a pre-mutation cell acquires b mutations in its i – a + b mutant mitochondria given that each mitochondrion mutates with probability μb . This probability is given by Y ( b;i−a+b , μb ) = ( i−a+bb ) μbb ( 1−μb ) i−a . For any combination of values for a , b and i , multiplying Y ( a;n – i + a – b , μ ) by Y ( b;i – a + b , μb ) gives the probability of a particular transition from a pre-mutation cell with identity Mt , τ3= ( i−a+b , G ) to a post-mutation cell with identity Mt , τ3= ( i , G ) . To get the overall probability that such a transition occurs , we multiply the probability of the transition by the proportion of pre-mutation cells in the population . To produce the post-mutation population , we sum all possible transitions between pre-mutation and post-mutation cells . All valid transitions must satisfy 0 ≤ a ≤ i ( because the post-mutation cell cannot receive more than i mutant mitochondria ) and 0 ≤ b ≤ n – i ( because the post-mutation cell cannot receive more than n – i wild type mitochondria ) . Thus , we can determine the post-mutation population by P ( Mt , τ3= ( i , G ) ) =∑a=0i∑b=0n−iY ( a;i−a+b , μ ) Y ( b;n−i+a−b , μb ) P ( Mt , τ2= ( i−a+b , G ) ) . In the neutral scenario , μ = μb ( i . e . the rate of mutation from wild type to mutant is equal to the rate of mutation from mutant to wild type ) . The relative fitness of a cell , w ( i ) , is a measure of how likely a cell type is to survive and reproduce , and we assume that cells carrying multiple mitochondrial types have lower fitness . For the first fitness function , the relative fitness of a cell with i mutant mitochondria is determined according to the following piecewise concave function: w ( i ) ={1−ch ( in/2 ) 2for0≤i<n/2 , 1−ch ( n−in/2 ) 2forn/2≤i≤n , ( 1 ) for even values of n and 0 ≤ ch ≤ 1 , where ch is the cost of heteroplasmy . In this function , a cell containing n / 2 mutant and n / 2 wild type mitochondria has minimum relative fitness . The post-selection population of each cell type is then given by: P ( Mt , τ4= ( i , G ) ) =w ( i ) P ( Mt , τ3= ( i , G ) ) . We also make use of two alternative fitness functions . The first of these is the piecewise linear function: w ( i ) ={1−ch ( in/2 ) for0≤i<n/2 , 1−ch ( n−in/2 ) forn/2≤i≤n . ( 2 ) The third fitness function is the piecewise convex function: w ( i ) ={1−chin/2for0≤i<n/2 , 1−chn−in/2forn/2≤i≤n . ( 3 ) We normalize the post-selection population by P ( Mt , τ5= ( i , G ) ) =P ( Mt , τ4= ( i , G ) ) σ , where σ=∑i=0nP ( Mt , τ4= ( i , U1B2 ) ) +P ( Mt , τ4= ( i , B1B2 ) ) , so that the sum of the proportions of the population equals 1 . The cell must first duplicate its chromosomes and double its mitochondrial complement ( from n to 2n ) . This cell with 2n mitochondria then produces gametes with n / 2 mitochondria . Meiosis occurs in two steps . First , we sample n mitochondria with replacement from a cell containing n mitochondria and add the set of sampled mitochondria to the original set of mitochondria to form a cell containing 2n mitochondria ( this is the same process that occurs in uniparental mating only with n mitochondria rather than n / 2 mitochondria ) . We let Mt , τ6= ( l , 2G ) represent the cell with doubled mitochondria and nuclear genotype , where l takes values in {0 , 1…2n} and 2G takes values in {U1U1B2B2 , B1B1B2B2} . For a cell to contain l mutant mitochondria after duplication of its mitochondria , it must sample l – i mutant mitochondria . We denote the probability of sampling l – i mutant mitochondria from Mt , τ5= ( i , G ) as F ( l – i;n , i / n ) . Sampling follows a binomial distribution such that F ( l−i;n , in ) = ( nl−i ) ( in ) l−i ( 1−in ) n−l+i . We obtain Mt , τ6= ( l , 2G ) by P ( Mt , τ6= ( l , 2G ) ) =∑i=max ( 0 , l−n ) min ( l , n ) F ( l−i;n , in ) P ( Mt , τ5= ( i , G ) ) . During the second step of meiosis , the cells with 2n mitochondria produce gametes with n / 2 mitochondria . Biologically , this occurs in two steps . In meiosis 1 , the homologous chromosomes are pulled apart to produce two haploid cells that contain two identical nuclear alleles ( sister chromatids ) and n mitochondria . In meiosis 2 , the two cells divide to produce four gametes , each with a single nuclear allele and n / 2 mitochondria . Since mitochondria segregate independently of nuclear alleles during cell partitioning , we model this as a single step . We define S ( p;2n , l , n / 2 ) to be the probability of obtaining p mutant mitochondria in n / 2 draws from a cell in state Mt , τ6= ( l , m , 2G ) . Here , sampling is without replacement and follows a hypergeometric distribution , giving S ( p;2n , l , n2 ) = ( lp ) ( 2n−ln2−p ) ( 2nn2 ) . Gametes produced by meiosis are represented by Mt+1 , τ1= ( p , g ) . We determine the probability of obtaining a particular gamete using P ( Mt+1 , τ1= ( p , U1 ) ) =12 ( ∑l=02nS ( p;2n , l , n2 ) P ( Mt , τ6= ( l , U1U1B2B2 ) ) ) , P ( Mt+1 , τ1= ( p , B1 ) ) =12 ( ∑l=02nS ( p;2n , l , n2 ) P ( Mt , τ6= ( l , B1B1B2B2 ) ) ) , and P ( Mt+1 , τ1= ( p , B2 ) ) =12 ( ∑l=02nS ( p;2n , l , n2 ) P ( Mt , τ6= ( l , U1U1B2B2 ) ) ) +12 ( ∑l=02nS ( p;2n , l , n2 ) P ( Mt , τ6= ( l , B1B1B2B2 ) ) ) . Factors of 1 / 2 in the above three equations take into account that half of the gametes produced from parent cells with nuclear genotype U1B2 will carry the U1 allele and the other half will carry the B2 allele ( with similar logic applied for gametes produced from parent cells with nuclear genotype B1B2 ) . Meiosis completes a single generation of the life cycle . The relative fitness of U1B2 cells is given by w¯U1B2=∑i=0nP ( Mt , τ3= ( i , U1B2 ) ) w ( i ) ∑i=0nP ( Mt , τ3= ( i , U1B2 ) ) , while the relative fitness of B1B2 cells is w¯B1B2=∑i=0nP ( Mt , τ3= ( i , B1B2 ) ) w ( i ) ∑i=0nP ( Mt , τ3= ( i , B1B2 ) ) . Although gametes are not subject to selection in our model , and thus do not technically have fitness values , it is informative to track gamete relative fitness throughout the simulation . We define a gamete’s relative fitness as the fitness that a diploid cell would have if it had the same mitochondrial composition as the gamete . Since gametes contain n / 2 mitochondria , they will have minimum fitness when they carry n / 4 wild type and n / 4 mutant mitochondria . To rescale the fitness function , we substitute n / 2 for n in the diploid cell fitness functions . For example , Equation ( 1 ) becomes wg ( i ) ={1−ch ( in/4 ) 2for0≤i<n/4 , 1−ch ( ( n/2 ) −in/4 ) 2forn/4≤i≤n/2 . Once the fitness function is scaled to gametes , we can determine the relative fitness of the three gametes by w ¯ U 1 = ∑ i=0 n/2 P ( M t , τ 1 = ( i , U 1 ) ) w g ( i ) ∑ i=0 n/2 P ( M t , τ 1 = ( i , U 1 ) ) , w ¯ B 1 = ∑ i=0 n/2 P ( M t , τ 1 = ( i , B 1 ) ) w g ( i ) ∑ i=0 n/2 P ( M t , τ 1 = ( i , B 1 ) ) , and w¯B2=∑i=0n/2P ( Mt , τ1= ( i , B2 ) ) wg ( i ) ∑i=0n/2P ( Mt , τ1= ( i , B2 ) ) . See S1–S6 Model for details of the other models .	Mitochondria contain genes that encode the machinery needed to power cells . Unlike the nuclear genome , the mitochondrial genome is typically inherited from one parent only ( uniparental inheritance ) . The most common explanation for uniparental inheritance is the genomic conflict theory , which states that uniparental inheritance evolved to prevent the spread of ‘selfish’ mitochondria that replicate quickly but produce energy inefficiently . Current explanations have a major problem: when using realistic parameters , mathematical models cannot show that uniparental inheritance can replace biparental inheritance . Clearly , we need a new explanation that fits with standard population-genetic theory . Recent evidence suggests cells may incur a cost when they carry multiple types of mitochondria . Here we show mathematically that uniparental inheritance could have evolved to avoid the costs of maintaining multiple mitochondrial lineages within a cell . Our results explain the long-standing evolutionary mystery of uniparental inheritance and provide insight into the evolution of mating types and binary sexes . Selection against heteroplasmy also has implications for the evolution of the mitochondrial genome because new mitochondrial haplotypes always lead to heteroplasmy before becoming fixed in the population . Thus , selection against heteroplasmy may explain why mtDNA coding-genes have slower substitution rates than analogous genes within the nucleus .	[ "Abstract", "Introduction", "Results", "Discussion", "Model" ]	[]	2015	Selection against Heteroplasmy Explains the Evolution of Uniparental Inheritance of Mitochondria
Mosquitoes are incriminated as vectors for many crippling diseases , including malaria , West Nile fever , Dengue fever , and other neglected tropical diseases ( NTDs ) . microRNAs ( miRNAs ) can interact with multiple target genes to elicit biological functions in the mosquitoes . However , characterization and function of individual miRNAs and their potential targets have not been fully determined to date . We conducted a systematic review of published literature following PRISMA guidelines . We summarize the information about miRNAs in mosquitoes to better understand their metabolism , development , and responses to microorganisms . Depending on the study , we found that miRNAs were dysregulated in a species- , sex- , stage- , and tissue/organ-specific manner . Aberrant miRNA expressions were observed in development , metabolism , host-pathogen interactions , and insecticide resistance . Of note , many miRNAs were down-regulated upon pathogen infection . The experimental studies have expanded the identification of miRNA target from the 3′ untranslated regions ( UTRs ) of mRNAs of mosquitoes to the 5′ UTRs of mRNAs of the virus . In addition , we discuss current trends in mosquito miRNA research and offer suggestions for future studies . miRNAs , which are ~22 nt long non-coding RNAs derived from larger hairpin RNA precursors , are involved in the post-transcriptional regulation of target genes in many physiological and pathological processes; therefore , they are of interest as therapeutic targets for treating various diseases [1] . In Drosophila , miRNAs control developmental processes , and once they are activated , more than 50 target genes can be regulated temporally and spatially [2] . In mammals , miRNAs may control the activity of ~30% of all protein-coding genes and participate in the regulation of most cellular processes [3] . The mosquito is a vector for numerous crippling diseases , including malaria , West Nile fever , Dengue fever , and other parasitic infections [4] . Studies suggest the worldwide distribution of medically important mosquito species that transmit infectious agents that cause millions of deaths annually . In addition , environmental changes , such as global warming , frequent international travel , and drug resistance , have contributed to keeping mosquito-borne diseases a public health concern [5] . Therefore , we must understand the molecular mechanisms underlying vector biology and host-pathogen interactions to develop novel vector control strategies to reduce disease . Recently , the genome sequences of several important vector mosquitoes have enabled studies of the molecular basis of mosquito feeding , immune function , and development . Studies suggest that miRNA expression in Anopheles gambiae , Anopheles stephensi , Aedes aegypti , Culex quinquefasciatus , and Aedes albopictus [6–12] have roles in ovary development , blood digestion , and immunity against infections . Furthermore , big data platforms generated from studies of the mosquito genome , transcriptome , and proteome would add in the understanding of vector biology and host-pathogen interactions [13–16] . Studies of miRNAs in mosquito species may provide clues to elucidate their effects on biological functions , invasions of parasites , and induction of immune protection for preventing disease . In this context , we reviewed current literature on mosquito miRNA repertoires and outlined physiological and pathological roles assigned to miRNAs to offer a foundation for future work . A systematic search of the published research for medical subject headings ( MeSH ) “mosquito” and “microRNA , ” “miRNA* , ” or “miRNA” was conducted using the electronic online database PubMed . This was supplemented by searches of Google Scholar and Web of Knowledge using the same MeSH terms as well as iterative reviews of reference lists of relevant published papers . After duplicate publications were deleted , all of the records were screened and the abstracts were reviewed if they contained relevant data on mosquito miRNA . Highly relevant papers were selected for full-text reviews . Two reviewers independently extracted and categorized data about the authors , the publication year , country in which the study was performed , and the samples . The characteristics of the study , laboratory methods , miRNAs found , and referred functions were also analyzed . Finally , the reviewers resolved the discrepancies through discussion and consensus . The search process and studies included are depicted in Fig 1 . Overall , the search strategy yielded 129 entries . Following the removal of 40 duplicates , 87 titles and abstracts were assessed , and 42 articles appeared to be potentially relevant for inclusion in the review . After eight articles were excluded based on the exclusion criteria , there were 34 articles that fulfilled the eligibility criteria , and these were included in the analysis . We located 1 , 540 , 1 , 893 and 383 putative miRNAs from the Anopheles 16 genomes project ( mostly through computational methods ) , VectorBase , and miRBase databases , respectively ( Table 1 ) . Table 2 depicts 29 experimental studies that mention miRNA profiles in the tissues or organs of many mosquito species . These selected experimental studies were stratified into seven categories according to mosquito species: Aedes aegypti ( 10 ) , Aedes albopictus ( 8 ) , Anopheles gambiae ( 4 ) , Anopheles stephensi ( 4 ) , Culex quinquefasciatus ( 1 ) , Culex pipiens ( 2 ) , and Anopheles anthropophagus ( 1 ) as shown in Table 2 . Overall , over 800 distinct mosquito miRNA sequences have been identified by experimental studies in Aedes , Anopheles , and Culex subgenera ( Table 2 ) . Fig 2 depicts differentially expressed miRNAs that are sex-specific in the mosquito life cycle from egg to adult . Mosquito body elements , such as the head , thorax , gut , and ovary , may have distinct expression profiles and these data appear in Fig 3 . miRNAs with roles in interactions between mosquitoes and pathogens appear in Fig 4 . The predicted miRNA targets in mosquitoes and the targets of regulated miRNAs identified at this time appear in Table 3 . Approximately 3 , 000 mosquito species from 34 genera exist and some are disease vectors . Previous studies revealed genetic determinants that affect the ability of different strains of mosquitoes to transmit pathogens , such as gene profiles [50] and transcriptomes [51] . miRNAs are key to the regulation of gene expression at transcriptional and post-transcriptional events and advances in miRNA have illuminated a role for these small RNAs in development and vector-pathogen interactions . miRNA types and amounts ( overall number of distinct miRNAs ) vary across mosquito species ( as shown in Table 1 ) . Direct sequencing revealed variations in miRNA profiles among different mosquito species . By analyzing the miRNA profiles of three mosquito species using the miRBse database , we found 52 miRNAs shared among three mosquito lineages ( Aedes aegypti , Culex quinquefasciatus , and Anopheles gambiae ) , and others were specific to certain mosquitoes , for example , aae-miR-2940 , aae-miR-2943 , and aae-miR-2945 were only observed in Aedes aegypti . Of 111 known miRNAs expressed across developmental stages An . stephensi [7] , 103 were identified in in Ae . albopictus [52] . Hu’s group identified 7 and 19 miRNAs unique to Ae . aegypti and An . stephensi , respectively [35] . Interestingly , although Ae . aegypti and Ae . albopictus are related , aae-miR-1174 was not found in Ae . albopictus developmental stage libraries [29] . Mosquitoes appear to have retained highly conserved miRNAs during their evolution . Homologous miRNAs identified in mosquito species indicate evolutionary pressure for miRNA sequence conservation and potentially critical functions of these miRNAs . Among conserved miRNAs , some were shared among many species , such as mir-281 , mir-184 , mir-989 , and mir-278 , which are generally expressed in An . gambiae [13] , Ae . Aegypti , and Cu . quinquefasciatus [37] . The expression patterns of conserved miR-14 , miR−184 , miR-210 , miR-970 , and miR-998 in Ae . aegypti are similar to the patterns found in An . stephensi [9] . However , novel miRNAs , usually identified as subsets of differentially expressed miRNAs , had distinct characteristics; for example , aae-mir-2946 , which could only be found in Ae . aegypti [9] , and cqu-mir-2951 and cqu-mir-2952 , which are only found in Cu . quinquefasciatus [30] . Sequence analysis of novel miRNAs indicates that they often lacked orthologs found in other mosquito species . Novel miRNAs are potentially restricted to certain species but they are less abundant than known conserved miRNAs [37] . Although Anopheles , Aedes , and Culex genera may have shared a common ancestor approximately million years ( MYr ) ago [53] , each lineage has specific miRNAs , indicating a loss or gain of these miRNAs in species to achieve and control different functions ( S1 Table ) . An investigation of stage-specific miRNAs may provide an understanding of mosquito biology and provide mosquito-specific targets for disease control . Significant stage-specific expression was observed for miRNAs in various species ( Fig 1 ) . In Anopheles , aan-miR-2943 and afu-miR-980 were only expressed in the egg stage in An . anthropophagus [32] and An . funestus , [54] respectively . ast-miR-2943 and ast-miR-2945 were highly expressed in An . stephensi embryos , and ast-miR-1890 had a peak expression in An . stephensi pupae [35] . Jain and colleagues [7] reported that 36 miRNAs were differentially expressed among various developmental stages of An . stephensi , including larval male and female , pupal male and female , and adult male and female . Among them , ast-miR-1891 , ast-miR-190-3p , ast-miR-285 , ast-miR-988-3p , and ast-miR-989 were absent in the larval stage , but ast-miR-8-3p was the most abundant in the male and female larval stages . ast-bantam-3p was the most abundant in the male and female pupal stages of development , and ast-miR-281-5p and ast-miR-bantam-3p were the most abundant in adult males and females , respectively . ast-miR-14 had a relatively strong signal from the late embryonic to adult stages [36] . The consistent expression of ast-miR-14 suggests that it may be essential throughout development , from embryos to aged adults . Expression analysis of miRNAs revealed distinct patterns from early embryo to adult stages in Aedes . In Ae . albopictus , aal-mir-M1 was only expressed in embryos , and aal-mir-9a was mainly expressed in embryo and larval stages . aal-let-7 was only expressed in pupal and adult stages , and aal-miR-1175 was widely expressed in all of the life stages , except for embryos [31] . There was aal-miR-286b accumulation in the embryo and aal-miR-2942 was the most expressed in larvae although it was normally expressed at the egg , pupae , and adult stages . aal-miR-1891 was more expressed in adult females than adult males , suggesting a possible regulatory role in blood feeding and egg development [12] . In addition , the increased expression of aal-miR-2941 , aal-miR-2943 , and aal-miR-2946 occurred in embryos , [29] which is consistent with the results for Ae . aegypti [9] . In Ae . aegypti , aae-miR-275 was required for egg maturation , but aae-miR-bantam , aae-miR-275 , and aae-miR-8 were highly expressed during the pupal period . However , aae-miR-275 was prominent at the beginning of the pupal stage , and aae-bantam and aae-miR-8 peaked at the mid-pupal stage [23] . The expression of the same miRNAs may differ across stages . Li and colleagues [9] found that aae-miR-989 had 2 read counts in the embryo stage , but 33 read counts in sugar-fed Ae . aegypti female adults . Conserved miRNAs are likely to be involved in important functions in mosquito lineages . In Ae . aegypti and An . stephensi , miR-2943 and miR-2945 were highly expressed in embryos . Bantam and miR-1890 were highly expressed during the pupal developmental period , and miR-1891 was most abundantly expressed in adult males [36] [35] . Therefore , the expression of some stage-specific miRNAs may be conserved in most lineages and stage-specific miRNAs may be involved in the regulation of growth , differentiation , and reproduction during a specific developmental stage . Understanding how sex-specific miRNA expression occurs in mosquitoes ( Fig 2 ) has great significance towards its role in blood feeding and disease transmission . For instance , 29 miRNAs ( based on read count of miRNAs ) had sex-biased expression in An . anthropophagus [32]; of these , 9 miRNAs were up-regulated in females and 20 miRNAs had decreased or no expression . Among them , aan-miR-989 was highly expressed in female mosquitoes , but not in males—similar to patterns in An . gambiae [6] . In other studies , miR-989 was up-regulated in adult female An . stephensi [7] and Ae . aegypti [36] compared to adult males , suggesting functional conservation among mosquitoes . In An . gambiae , the expression of aga-miR-34 was more pronounced in the midguts of females , while aga-miR-277 was highly expressed in the midguts of males [6] . miR-1891 was most abundantly expressed in Ae . aegypti and An . stephensi adult males [35] . In addition , Northern blot and sequencing counts indicated that the expression of miR-184 and miR-1000 in male adults was higher than in female adults in Ae . Aegypti [9] and An . anthropophagus [32] . Moreover , sex-specific miRNA expression diverged during larval , pupal , and adult mosquito stages [7] . Fewer miRNA differences were identified during An . stephensi immature stages , and two miRNAs ( ast-miR-184b and ast-miR-1175-5p ) were up-regulated in male larvae; one miRNA ( ast-miR-285 ) was down-regulated in the female pupal stage . Maximal differences in miRNA expression between sexes were observed during the adult stages , except for ast-miR-989 , and several miRNAs were down-regulated in female An . stephensi , including miR-7 , which was also reported in An . anthropophagus [32] . The expression of miR-989 was restricted to adult females and predominantly in the ovaries of Anopheles and Aedes . Mead’s group observed reduced miR-989 in post-blood-meal ( PBM ) females ( 72 h ) [36] . To investigate the role of sex-specific miRNAs in mosquito reproduction , Jain’s group [7] injected miR-989-specific antagomirs in female mosquitoes and their expression affected multiple functions in ovaries after blood-feeding . Thus , miR-989 may be associated with female reproduction , and its function may be conserved among divergent mosquitoes . Different body parts of the mosquito , such as the head , thorax , gut , and ovary , have distinct expression profiles ( Fig 3 ) . aga-miR-317 was more expressed in the head compared to the thorax , leftover ( carcass ) , and midgut [6] . The preferential expression of aga-miR-34 , aga-miR-277 , aga-miR-12 , and aga-miR-283 occurred in the thorax of both males and females , and twice as much in heads . Midgut-specific miRNAs have been identified in An . gambiae , Ae . Albopictus , and other mosquito species . For instance , aga-miR-12 and aga-miR-283 were predominantly expressed in the midgut . aga-miR-1175 , aga-miR-1174 , and aga-miR-281 were expressed only in the midgut . In addition , the miR-1174/miR-1175 miRNA cluster was highly expressed in An . gambiae gut PBM [6] . In Ae . albopictus , the midgut-specific aal-miR-281 was the most abundant miRNA . A high expression of aae-miR-1890 was observed in the midgut of female Ae . aegypti , and mature aae-miR-1890 peaked at 24 h PBM and declined sharply by 36 h PBM in the female mosquito midgut [17] . In addition , miR-281 , miR-1174 , and miR-1175 were also only found to be expressed in the midgut of adults in Ae . aegypti , Ae . albopictus , and Cu . quinquefasciatus [9] [30] . In An . stephensi and Ae . aegypti , the expression of miR-989 was predominantly in the ovaries [36] . aae-miR-8 , aae-miR14 , and aae-miR-275 were highly expressed in the vitellogenic fat body [23] , and aae-miR-8 was substantially increased PBM in female Ae . aegypti fat bodies [28] [23] . In addition , 41 miRNAs were differentially expressed in the testes and pre-vitellogenic ovaries . Among them , aga-mir-2944a-2 and aga-mir-286b were up-regulated in the testes and during oogenesis , suggesting a role in gametogenesis [34] . Then , 103 extracellular miRNAs were identified from Ae . aegypti and Ae . albopictus saliva; of these , 31 miRNAs were previously unidentified and designated as novel . aae-mir-281-2-5p , aae-mir-281 , aae-mir-2940 , aae-mir-bantam , aae-mir-125 , and aae-mir-263a were highly expressed in uninfected and infected Ae . aegypti saliva , while aal-mir-8 and aal-mir-125 were equally expressed in uninfected and infected Ae . albopictus saliva [24] . Therefore , different tissues/organs possess different miRNA expression profiles , and tissue/organ-specific miRNAs may be of more value than some ubiquitously-expressed miRNAs in investigating and explaining specific physiological functions , or as specific indicators to distinguish infections . The role of miRNAs in the post-transcriptional regulation of gene expression has been recognized to contribute to physiological and immune pathways that affect development , metabolism , host-pathogen interactions , and insecticide resistance . The stage-specific expression of miRNAs in the four developmental stages ( eggs , larvae , pupae , and adults ) has been confirmed using high-throughput sequencing followed by Northern blot analysis and quantitative polymerase chain reaction ( PCR ) [7 , 12 , 37] . To understand the role of regulated miRNAs in mosquito development , the knock-in and knock-down of specifically and temporally expressed miRNAs were conducted in Ae . albopictus by microinjection . The knock-down of aal-miR-286b and aal-miR-2942 decreased the hatching of embryos and eclosion rate of larvae , respectively , when compared with the knock-in groups . Reduced longevity and fecundity ( aal-miR-1891 ) in adults was observed in the miR-1891 knock-down groups compared to the knock-in and control groups [12] . Female mosquitoes require sugar for energy metabolism and a blood meal for egg development , and recent studies have indicated that blood feeding leads to the differential expression of many genes , proteins , and miRNAs [55–57] . miRNA abundance differs under sugar-fed and blood-fed conditions , and ast-miR-2796-5p was observed exclusively in sugar feeding An . stephensi with extremely low read counts [8] . aae-miR-375 was only found in blood feeding Ae . aegypti mosquitoes [22] . Most miRNAs ( 107 ) were found in a blood-fed library of An . stephensi compared with sugar-fed and Plasmodium-infected libraries . ast-miR-286b , ast-miR-2944a-3p , and ast-miR-309 were significantly expressed in blood-feeding ( BF ) 42 h with no reads present in sugar feeding , indicating that the expression of these miRNAs may be induced by a blood meal [8] . Expressions of 4 miRNAs ( aga-miR-7 , aga-miR-92a , aga-miR-317 , and aga-miR-N3 ) were significantly changed in blood-fed An . gambiae [13] . Expression changes occurred in aga-miR-34 and aga-miR-989 in leftovers and midguts [6] . Moreover , variations in miRNA expression are temporally regulated . aae-miR-275 , which is required for blood digestion in Ae . aegypt , was elevated 7 . 2-fold from 0 to 12 h PBM [23] . The depletion of aae-miR-275 in Ae . aegypti females by injection of its specific antagomir led to severe defects in blood digestion , fluid excretion , and egg development . aae-miR-1890 is induced after blood feeding and peaks at 24 PMB , and systemic depletion of aae-miR-1890 resulted in decreased egg development and deposition , suggesting that miR-1890 may be key to mosquito blood digestion [17] . In contrast with up-regulated miRNAs after blood feeding , some miRNAs were down-regulated . For example , reduced ast-miR-989 was observed 72 h after a blood meal [36] . aga-let7 was decreased in the midguts and other parts/leftovers [6] , but most miRNAs were increased after blood feeding [9] . The malarial vector Anopheles initiates strong immune responses by inducing the expression of key anti-Plasmodium effectors upon the invasion of Plasmodium parasites , which are largely regulated by 3 immune signaling pathways , namely , the Toll , Jak/Stat , and immune deficiency ( IMD ) pathways [58 , 59] . miRNAs may fine-tune immune responses and other physiological processes . The expression of aga-miR-34 , aga-miR-1174 , and aga-miR-1175 decreased in the midgut after P . falciparum infection , while aga-miR-989 and aga-miR-305 were elevated in infected midguts . A functional study showed that aga-miR-305 increased susceptibility to P . falciparum infection and proliferated midgut microbiota [33] . Infection of An . stephensi and An . gambiae with the rodent malarial parasite P . vinckei petteri and P . berghei caused the differential expression of multiple miRNAs [8 , 13] . For instance , 6 miRNAs were significantly up-regulated after P . berghei infection[13]; of these , aga-miR-317 and aga-miR-2940 were more than 5- and 3-fold unregulated . Then , 4 miRNAs were markedly up-regulated in infectious blood feeding 42 h ( ast-miR-124 , ast-miR-137 , ast-miR-1000 , and ast-miR-932 ) and 5 d ( ast-miR-1175-3p , ast-miR-1174 , ast-miR-281-3p , and ast-miR-281-5p ) of infectious blood-feeding . Meanwhile , 10 miRNAs ( ast-miR-285 , ast-miR-2944a-5p , ast-miR-309 , ast-miR-210-3p , ast-miR-1891 , ast-miR-981 , ast-miR-315-5p , ast-miR-932 , ast-miR-124 , and ast-miR-7 ) were significantly down-regulated in the infectious blood feeding 5 d group compared with the 42 h group after P . vinckei petteri infection [8] . In addition , Dicer1 , Dicer2 , Drosha , and Ago1 are involved in miRNA biogenesis and increased polysome loading after infection in mosquitoes . The knock-down of Dicer1 and Ago1 changed mosquito susceptibility to the Plasmodium parasite [6] [60] . Thus , mosquito miRNAs may participate in reactions against Plasmodium invasion . Flavivirus genus viruses are spread by mosquitoes and cause diseases , including Dengue and West Nile fever . To determine whether flavivirus infection could alter miRNA expression , Skalsky’s group infected female Cu . quinquefasciatus mosquitoes with WNV ( West Nile virus ) , and cuq-miR-92 and cuq-miR-989 had altered expressions [30] . Slonchak’s group found aae-miR-2940 was selectively down-regulated in Aedes albopictus cells in response to WNV infection to restrict viral replication [38] . Campbell and co-workers observed that the expressions of 35 mosquito miRNAs were modulated upon DENV ( Dengue virus ) infection in Aedes aegyptis , [20] and Liu’s group noted that 66 miRNAs of Ae . albopictus were differentially expressed after DENV-2 infection [52] . Therefore , aal-miR-34-5p and aal-miR-87 may contribute to anti-pathogen and immune responses during DENV-2 infection [52] . aae-miR-375 is the key to DENV replication , which may enhance DENV-2 infection in an Ae . aegypti cell line [22] . aae-miR-252 was induced more than three-fold after DENV-2 infection in an Ae . albopictus C6/36 cell line , which inhibited DENV replication by suppressing the expression of the DENV envelope protein [11] . aal-miR-281 , an abundant midgut-specific miRNA , facilitates DENV-2 replication in Ae . albopictus [27] . Chikungunya virus ( CHIKV ) is a alphavirus transmitted predominantly by Aedes aegypti and Aedes albopticus , and it causes severe symptoms , including the risk of death [24] . Shrinet’s group evaluated the role of host miRNAs upon CHIKV infection in Ae . albopictus and they observed an altered expression of 8 miRNAs [61] . Maharaj and co-workers reported 59 and 30 miRNAs upregulated in Ae . aegypti and Ae . albopictus CHIKV-infected saliva , respectively , indicating the importance of saliva miRNAs in regulating CHIKV infection in mammals [24] . Wolbachia are widespread in invertebrates and can manipulate reproduction , reduce the host life span , and inhibit pathogen infections , such as DENV , filarial nematodes , and malarial parasites [21 , 62 , 63] . In 2011 , a microarray analysis of miRNAs revealed that ~13 miRNAs were differentially expressed in Wolbachia-infected female Ae . aegypti mosquitoes [39] . Also , aae-miR-12 was differentially expressed in Ae . aegypti infected with Wolbachia . Then , Osei-Amo found that the inhibition of aae-miR-12 reduced Wolbachia density in Wolbachia-infected Aag2 mosquito cell lines [41] . Decreased aae-miR-2940 and aae-miR-184 was observed in AGO2-silenced and Wolbachia-infected cells [22] . Then , aae-miR-2940 was induced and exclusively found in Wolbachia-infected mosquitoes [39] . Wolbachia uses host aae-miR-2940 to regulate a methyltransferase gene to block DENV replication [40] . aae-miR-989 , aae-miR-306-5p , and aae-miR-1889 were down-regulated in Wolbachia-infected Ae . aegypti , while aae-miR-2765 and aae-bantam-5 were up-regulated [25] . Therefore , Wolbachia influences miRNA expression and alters natural miRNA profiles in the mosquito . Pyrethroid resistance due to excessive and improper usage of pyrethroids is an impediment to combating mosquito-borne diseases . To validate whether miRNAs have a role correlated with insecticide resistance , Lei’s group measured miRNA expression in pyrethroid-resistant and susceptible strains of lab populations and confirmed the dysregulated miRNAs . Of these , miR-278-3p was up-regulated in the susceptible Culex pipiens pallens strain [10] . In another study , cpi-miR-71 was significantly down-regulated in female adults from a deltamethrin-resistant strain , indicating that cpi-miR-71 may play a contributing role in deltamethrin resistance [37] . Then , the overexpression of cpi-miR-71 in female mosquitoes had reduced resistance to deltamethrin . Differentially expressed miRNAs in these studies provide a basis for the investigation of pyrethroid resistance in the future . To understand the role of regulated miRNAs in development , sugar feeding , blood feeding , and pathogen invasion , we must identify relevant targets . Studies show that bioinformatic analysis and in vivo assays can be used to identify the targets of regulated miRNAs . Targets were predicted by identifying miRNA seed-binding sites on the 3' UTR of genes using RNAhybrid [64] , miRanda [65] , TargetScan [66] , PicTar [67] , and other in-house pipelines . Dual-luciferase reporter assays to assess target identification were used , and degradome sequencing has recently been used to identify cleaved targets of regulated miRNAs by sequencing degraded mRNA [7] . This may allow researchers to overcome the limits of bioinformatic predictions and locate target genes for miRNAs . Studies to predict miRNA targets in mosquitoes and targets of regulated miRNAs have been identified ( Table 3 ) . Ovary-specific aae-miR-309 was found to target SIX4 and contribute to Ae . aegypti mosquito reproduction [42] . The ortholog of SIX4 in D . melanogaster is required for gonadogenesis [42] . Blood-feeding in mosquitoes is a major metabolic challenge , and aae-miR-1890 was shown to bind the 3' UTR of JHA15 mRNA ( with presumed role in blood digestion ) and control JHA15 mRNA stability in a stage- and tissue-specific manner to regulate blood digestion [17] . In addition , secreted wingless-interacting molecule ( swim ) , an important gene that could interrupt Wg signaling activity in Drosophila , was regulated in female mosquito fat body [18] . Pathogenic agents can alter host-derived miRNAs , which then modulate the host gene expression to cause translational inhibition and mRNA decay . Cactus and REL1 genes were targets of blood-induced aae-miR-375 , and the injection of an miRNA mimic into mosquitoes led to fold-changes in immune gene transcripts , suggesting that aae-miR-375 enhanced DENV-2 infection [22] . Furthermore , three targets of aae-miR-2940-5p have been validated . Metalloprotease ftsh ( MetP ) was found to be the first target of miRNA and it was important for the replication/maintenance of Wolbachia [39] . The second target of aae-miR-2940 was DNA methyltransferase ( Dnmt2 ) . The overexpression of Dnmt2 increased DENV-2 replication and reduced Wolbachia density [21] . In addition , arginine methyltransferase 3 was found to be a target of aae-miR-2940 [40] , which was positively regulated and beneficial for Wolbachia replication . Other miRNAs significantly regulated during development or in the presence of pathogens have not been explored . More work is needed to identify their potential roles in metabolic processes , phagocytosis , and immune defense . In addition , miRNAs have been predicted to have multiple gene targets , suggesting the importance of these molecules in regulatory networks . A comparative analysis of miRNA profiles of different mosquito species revealed that nearly half of known miRNAs are conserved . Conserved miR-184 and miR-998 were identified in An . gambiae , Ae . aegypti , An . stephensi , Ae . albopictus , and other mosquitoes [9] [24] , which indicates evolutionary pressure for miRNA conservation and potentially critical functions of these miRNAs in various species [7] . Species-specific miRNAs are thought to be novel and potentially specific to mosquitoes in low read counts , indicating a loss/gain or rapid change of miRNAs during evolution to achieve and control species-specific functions . However , some highly conserved miRNAs , such as miR-282 and miR-927 , found in Ae . aegypti and An . gambiae were not confirmed in Cu . quinquefasciatus [29] , indicating that known and novel miRNAs can exhibit species-specific patterns . Species-specific miRNAs may contribute to the susceptibility of different mosquitoes to unique pathogens and mosquito-specific targets for disease control and prevention . Many miRNAs have spatio-temporal patterns of expression essential for regulating complex physiological activity of mosquitoes . Research has shown a significant reduction in miR-989 72 h after a blood meal in An . stephensi and Ae . aegypti , predominantly in adult female ovaries [36] . Some regulated miRNAs were differentially expressed during larval to pupal stages and during pupal to adult metamorphosis [7] . Recent research has indicated possible regulatory effects of aae-miR-8 in reproduction as it is highly expressed in female mosquito fat body PBM [23] . Of note , certain miRNAs ( aga-miR-996 , aga-miR-279 , aga-miR-306 , aga-miR-79 , aga-miR-9b , and aga-miR-275 ) were expressed evenly and ubiquitously throughout the An . gambiae body [6] , and most highly expressed miRNAs ( miR-1 , miR-184 , and miR-263 ) were expressed in most developmental stages in many mosquito species [35] . For example , the consistent expression of ast-miR-14 from the late embryonic to the adult stage indicates that it likely plays an important role across all life stages [36] . Several miRNAs are sexually dimorphic , such as miR-989 with an expression that is restricted to adult An . stephensi and Ae . aegypti females[36] . In addition , aga-miR-277 was highly expressed in An . gambiae [6] , and miR-1891 was abundantly expressed in Ae . aegypti and An . stephensi adult males [35] . These differences may be a result of unique reproduction strategies and disease transmission . Moreover , several sex-specific miRNAs were observed in pupal , larva , and adult mosquitoes , but embryogenesis has not been investigated and how these are regulated in a sex-specific manner is unclear . Early studies have identified extracellular miRNAs in saliva and serum in humans and mammals , and these have roles in intercellular communication , coinciding with the transfer of functional and intact proteins , lipids , and nucleic acids between cells . Recent studies identified extracellular miRNAs as being dysregulated in mosquitoes . Maharaj’s group [24] isolated saliva containing extracellular miRNAs from mosquito salivary glands and these were key to pathogen transmission from mosquito to vertebrates . They found 103 mature miRNAs in Ae . aegypti and Ae . albopictus saliva . Subsequent experiments confirmed that saliva miRNAs can regulate CHIKV infection . Therefore , extracellular miRNAs may have concomitant changes with intracellular miRNAs and synergistically modulate viral replication . Aberrant miRNA expression in mammals may be used as a biomarker for disease and some miRNAs can influence the onset and courses of cancer [68] and vascular/heart diseases [69] [70] . miR-21 was misexpressed in diseased hearts [70] , and Let7-f , miR-27b , and mir-130a had a proangiogenic role [69] . Studies have shown that infection can alter the expression of mosquito miRNAs . In Ae . aegypti and Ae . albopictus , aberrantly expressed miRNA profiles were noted after CHIKV , WNV , and Wolbachia infections [23–25] . In An . gambiae and An . stephensi , Plasmodium infection changed miRNAs expression [11] . While drafting this paper , Saldana’s group reported that the Zika virus modulated 17 host miRNAs in Ae . aegypti mosquitoes at all post-infection points [71] . Also , many miRNAs are reported to be down-regulated by Plasmodium , Dengue , and CHIKV , although responses differed from infections . These data will increase our understanding of pathogen-vector interactions and provide potential avenues to investigate and develop miRNA-based strategies . In particular , Tsetsarkin’s group [72] developed an miRNA-targeted approach by introducing mosquito-specific mir-184 and mir-275 miRNAs to selectively restrict the replication of Dengue type 4 virus ( DEN4 ) in Ae . albopictus and Ae . aegypti , which may mitigate the risk of introduction and dissemination . miRNA profiles have been identified using cloning , microarray or direct sequencing , and influencers of miRNA expression have been reported . However , there is limited research about miRNA target analysis and the gene regulatory networks of miRNA families . There are only a few existing studies regarding specific functions , which limits our understanding of miRNA repertoires and functions in the mosquito . Attention should focus on discovering veiled miRNAs in vector mosquitoes , clarifying the occurrence and spatial-temporal regulation of specific miRNAs , especially miRNA signatures rather than a single miRNA alone . This would allow setting of criteria for target prediction using machine-learning algorithms and exploration of miRNA:mRNA networks in post-transcriptional level . These ideas may spur research to allow miRNAs to be used as vector control tools .	Millions of human infections are caused by mosquito-borne diseases , so understanding the molecular and genetic mechanisms that determine variability in transmission efficiency and insect susceptibility may assist with novel vector control strategies . Recently , microRNAs ( miRNAs ) have been studied as post-transcriptional regulators of gene expression in vertebrates , invertebrates , and viruses . Because interactions between vectors and parasites are the keys to malarial transmission dynamics , identifying miRNAs involved in the developmental cycles of the mosquito and interactions between the mosquito and pathogens that allow the survival , proliferation , and disease transmission may help with disease control . In this review , we summarize the research on mosquito miRNAs regarding their expression and function in various physiological and pathological processes , and explore future research in this area .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "body", "fluids", "chikungunya", "infection", "gene", "regulation", "tropical", "diseases", "animals", "wolbachia", "micrornas", "neglected", "tropical", "diseases", "infectious", "disease", "control", "insect", ...	2018	microRNA profiles and functions in mosquitoes
Myeloid dendritic cells ( DCs ) can capture HIV-1 via the receptor CD169/Siglec-1 that binds to the ganglioside , GM3 , in the virus particle membrane . In turn , HIV-1 particles captured by CD169 , an I-type lectin , whose expression on DCs is enhanced upon maturation with LPS , are protected from degradation in CD169+ virus-containing compartments ( VCCs ) and disseminated to CD4+ T cells , a mechanism of DC-mediated HIV-1 trans-infection . In this study , we describe the mechanism of VCC formation and its role in immune evasion mechanisms of HIV-1 . We find HIV-1-induced formation of VCCs is restricted to myeloid cells , and that the cytoplasmic tail of CD169 is dispensable for HIV-1 trafficking and retention within VCCs and subsequent trans-infection to CD4+ T cells . Interestingly , introduction of a di-aromatic endocytic motif in the cytoplasmic tail of CD169 that results in endocytosis of HIV-1 particles , suppressed CD169-mediated HIV-1 trans-infection . Furthermore , super-resolution microscopy revealed close association of CD169 and HIV-1 particles in surface-accessible but deep plasma membrane invaginations . Intriguingly , HIV-1 particles in deep VCCs were inefficiently accessed by anti-gp120 broadly neutralizing antibodies , VRC01 and NIH45-46 G54W , and thus were less susceptible to neutralization . Our study suggests that HIV-1 capture by CD169 can provide virus evasion from both innate ( phagocytosis ) and adaptive immune responses . Myeloid dendritic cells ( DCs ) are professional antigen presenting cells that play sentinel roles in sensing pathogens and priming adaptive immunity [1] . HIV has , however , exploited DCs to spread to CD4+ T cells and thus DCs have been suggested to play a role in systemic HIV dissemination from peripheral mucosa to secondary lymphoid tissues [2 , 3] . While DCs are infected with HIV and DC-derived progeny viruses can infect CD4+ T cells [4–7] , productive infection of DCs is limiting for several reasons including low receptor/co-receptor density , presence of cell-intrinsic restriction factors and innate sensing mechanisms eliciting anti-virus immune responses such as type I interferon secretion [8–11] . In contrast , DCs can capture HIV-1 particles and transmit captured virus to CD4+ T cells without establishing productive infection in DCs via a tight cell-to-cell junction called virological synapse [12] , a mechanism of DC-mediated HIV-1 trans-infection , that might have evolved to bypass DC-intrinsic anti-viral responses . Recently , our group and others have identified CD169 , also known as Siglec-1 , as a predominant receptor for mature DC-mediated capture of HIV-1 and subsequent virus transmission to T cells [13 , 14] . CD169 , a type I transmembrane protein , is the largest member of the sialic-acid-binding immunoglobulin-like lectin ( Siglec ) family , containing 17 extracellular repeats of immunoglobulin like domain including a N-terminal V-set domain that recognizes α2–3 linked sialic acid residues , a single transmembrane domain and a short cytoplasmic tail ( CT ) [15] . Upon HIV-1 binding to CD169 on mature DCs , HIV-1 particles accumulate in CD81 tetraspanin+ compartments [13 , 14] . These compartments are , however , only weakly or poorly stained with endosome/lysosome markers such as CD63 and Lamp1 [16 , 17] . Whether or not these HIV-1+ compartments are connected to cell surface has been matter of intense debate [reviewed in [18]] . While early studies suggested that endocytosis of HIV-1 particles was important for efficient trans-infection of T cells [19–21] , recent studies , however , have called these findings into question and have suggested that surface bound HIV-1 particles present in plasma membrane invaginations were the major source of viruses contributing to efficient DC-mediated HIV-1 trans-infection of T cells [22 , 23] . Interestingly , the CT of human CD169 contains 44 amino acids , and there are no defined signaling motifs or phosphorylation sites that could contribute to potential virus particle trafficking and internalization upon ligand binding . Therefore , how CD169-bound HIV-1 particles are accumulated and viral infectivity preserved in these compartments remains unclear . In this study , we have investigated the role of CD169 in trafficking of HIV-1 in mature DCs and facilitating HIV-1 trans-infection of T cells . We found that CD169-mediated HIV-1 trafficking to non-endocytic plasma membrane invaginations is cell-type specific , and that trans-infection could be achieved even in the absence of the CT . Trans-infection efficacy was correlated with the ability of CD169 to retain HIV-1 particles on the cell surface . Interestingly , a single amino acid substitution ( Ala to Tyr at position 1683 ) in the CT of CD169 resulted in the endocytosis of CD169-bound HIV-1 and the mutant CD169 was unable to support trans-infection of T cells , suggesting surface retention by CD169 is critical for HIV-1 to gain access to the trans-infection pathway . Furthermore , using super resolution microscopy , we observed that CD169 and HIV-1 particles were closely associated in LPS-matured DCs in compartments at the cell periphery , approximately 800 nm to 1 μm in depth from the cell surface . These peripheral virus-containing plasma membrane invaginations were not observed in DCs matured by exposure to IFN-α alone , suggesting a requirement for a LPS-inducible host co-factor for formation of the CD169+ HIV-1 containing plasma membrane invaginations . Intriguingly , HIV-1 particles localized within plasma membrane invaginations in LPS-matured DCs were inefficiently accessed by and hence less susceptible to α-gp120 broadly neutralizing antibodies compared to cell free viruses . Our study here , therefore , demonstrates that CD169-mediated capture and trafficking of HIV-1 within DCs can not only provide virus evasion from endocytic mechanisms that can lead to virus particle degradation in lysosomal compartments but also protect HIV-1 from neutralizing antibodies via formation of virus-containing surface-exposed plasma membrane invaginations in LPS-matured DCs . Previous studies have reported that upon virus capture by mature DCs , HIV-1 particles accumulate in compartments at the cell periphery [23 , 24] . Furthermore , formation of DC–T cell conjugates results in polarized release of captured virus particles towards T cells for establishment of optimal CD4+ T cell infection [25] . We , as well as others , have recently reported that HIV-1 particles in these compartments are strongly colocalized with CD169 [13 , 14] . Since CD169 was also colocalized with HIV-1 at the DC–T cell virological synapse [13 , 14] , we wanted to determine the mechanism by which CD169 mediates trafficking of HIV-1 particles in mature DCs . First , we sought to establish a cell line which could recapitulate the formation of peripheral virus-containing compartments ( VCCs ) that are observed upon HIV-1 capture by CD169 in mature DCs [13 , 14] . CD169 was stably transduced into a monocytic cell line THP-1 , Raji B cell line and HeLa cells . Cell-surface CD169 expression was tested by flow cytometry and found to be comparable to or higher than that observed on mature DCs ( S1 Fig A ) . Furthermore , induced expression of CD169 on primary cells ( LPS treatment of DCs , mature DCs ) or engineered expression of CD169 on cell lines ( THP-1 , Raji or HeLa ) resulted in a dramatic enhancement in virus capture ( Fig . 1A ) . Next , we examined if any of the cell lines were able to recapitulate the formation of CD169+ VCCs found in mature DC . Mature DCs , THP-1/CD169 , Raji/CD169 and HeLa/CD169 cells were incubated with HIV Gag-mCherry VLPs and stained for total CD169 following membrane permeabilization with TritonX-100 ( +Tx100 ) or without membrane permeabilization to visualize cell surface CD169 ( Surface ) expression ( Fig . 1B ) . In all the cells tested , VLPs were strongly colocalized with CD169 when stained after membrane permeabilization , as reported previously [14] . In mature DCs , VLPs were often found within compartments at the cell periphery some of which were stained with CD169 without membrane permeabilization ( Fig . 1B ) . In THP-1/CD169 cells , VLPs were strongly colocalized with CD169 in compartments similar to those found in mature DCs ( Fig . 1B ) . Interestingly , similar to mature DCs , CD169+ VLP+ compartments in THP-1/CD169 cells were also partially accessible to surface applied anti-CD169 antibodies . While VLPs captured by Raji/CD169 cells were strongly colocalized with CD169 , VLPs remained at the surface in the absence of formation of VCCs . In contrast , captured VLPs were found in intracellular CD169+ compartments in HeLa/CD169 cells , since anti-CD169 antibody was unable to stain VCCs without membrane permeabilization ( Fig . 1B ) . We next determined if differential localization of HIV-1 particles upon CD169 capture in cell lines could affect CD169-mediated trans-infection . While HIV-1 particles captured by mature DCs , THP-1/CD169 cells or Raji/CD169 cells were transmitted to CD4+ T cells , resulting in robust infection of T cells ( Fig . 1C; trans-infection was enhanced more than 10-fold in CD169+ cells compared to CD169low immature DCs or empty vector transduced control cell lines ) , HeLa/CD169 cells failed to transmit HIV-1 to T cells ( Fig . 1C ) . These findings suggest that retention of HIV-1 particles at the cell surface upon CD169-mediated capture ( Fig . 1B ) is necessary for virus access to the trans-infection pathway . A corollary of these findings is that endocytosed HIV-1 particles are incompetent for CD169-mediated trans-infection . CD169 has been reported as a phagocytic receptor on porcine macrophages that can mediate endocytosis of PRSSV [26] . However , to date , no previously defined endocytosis signaling motifs have been described in the CT of human CD169 . Since CD169 was trafficked to and colocalized with HIV-1 in surface-accessible compartments in myeloid cells ( Fig . 1 ) , we postulated that there was an unidentified trafficking motif in the CT that contributed to colocalization of CD169 and HIV-1 in VCCs . Two CD169 CT truncation mutants were constructed ( Fig . 2A ) , one of which has a stop codon right after the transmembrane domain of CD169 ( CD169ΔCT ) [15] . Since previous studies have demonstrated severe reduction in cell surface expression of plasma membrane targeted proteins upon deletion of cytoplasmic tails [27 , 28] , we constructed a second CD169 CT mutant that expressed the first four amino acids of CT ( CD169ΔCT4R ) . These CD169 CT mutants were transduced into THP-1 cells and the ability of these stably transduced cell lines expressing CD169 mutants to capture HIV and form VCCs was compared to that observed with THP-1 cells expressing wild type CD169 ( THP-1/CD169 ) ( Fig . 1B ) . Deletion of the cytoplasmic tail ( CD169ΔCT ) resulted in decreased expression of CD169 in in THP-1 cells ( Fig . 2B and S1 Fig B ) . Furthermore , cell surface expression of CD169ΔCT was further reduced ( Fig . 2C and D ) and resulted in severe attenuation of HIV-1 capture ( Fig . 2E ) . Interestingly , inclusion of the membrane proximal 4 arginine residues in the cytoplasmic tail resulted in higher expression of CD169 in cells and partial rescue of cell surface expression of CD169 ( Fig . 2C , 2D and S1 Fig B ) , and importantly , capture of HIV-1 particles ( Fig . 2E ) . The efficiency of virus capture by THP-1/CD169ΔCT4R cells was much lower than that exhibited by wt THP-1/CD169 cells ( Fig . 2E ) , in correlation with CD169 expression level on the cell surface ( Fig . 2C and D ) . We next co-cultured CD4+ T cells with THP-1 cells expressing CD169 CT mutants to investigate the role of CD169 CT in mediating HIV-1 trans-infection . Interestingly , THP-1/CD169ΔCT4R but not THP-1/CD169ΔCT cells could transmit HIV-1 to CD4+ T cells ( Fig . 2F ) . Furthermore , there was no significant difference in the efficiency of trans-infection ( T cell infection per amount of virus captured by THP-1 cells ) mediated by THP-1/CD169 and THP-1/CD169ΔCT4R cells ( Fig . 2G ) . Finally , CD169+ VCCs were also observed in THP-1/CD169ΔCT4R cells ( Fig . 2H ) , suggesting that the CD169 CT sequences downstream of the four arginine residues were dispensable for the formation of VCCs and CD169-mediated HIV-1 trans-infection . Whether endocytosed HIV-1 particles in DCs remain competent for trans-infection has been a matter of significant debate [6 , 13 , 16 , 18 , 21–24 , 29] . Since CT sequences proved dispensable for CD169 mediated trans-infection and HIV-1 particles captured by CD169 remained within surface-accessible VCCs ( Fig . 2 ) we hypothesized that HIV-1 has exploited CD169-dependent trafficking to evade host phagocytic responses that target captured pathogens to degradative compartments . To test this hypothesis , we introduced a single point mutation in the CT of CD169 that introduces a di-aromatic motif ( Ala to Tyr at position 1683 ) such as one known to be essential for mannose receptor-mediated phagocytosis of bacterial pathogens bearing terminal mannosylated proteins in their cell wall [30 , 31] ( Fig . 3A ) . This mutant CD169 , designated as CD169YF , was constitutively expressed in THP-1 cells via retroviral transduction . CD169YF expression was confirmed both by western blotting ( Fig . 3B ) and flow cytometry ( Fig . 3C and S1 Fig B ) , and was expressed at similar levels at the cell surface as wild type CD169 ( Fig . 3D ) . Interestingly , kinetics of anti-CD169 antibody internalization were enhanced in THP-1/ CD169YF compared to THP-1 cells expressing wild type CD169 , suggesting the single amino acid substitution functioned as an internalization signal motif ( S2 Fig A ) . We next investigated the localization of HIV Gag-mCherry VLPs upon capture by THP-1/CD169YF cells . Both wt CD169 and CD169YF expressing THP-1 cells were challenged with VLPs and stained for CD81 , a tetraspanin protein that colocalizes with HIV-1 in VCCs in mature DCs [16 , 17] , or CD63 and Lamp1 ( late endosomal compartment markers ) . In THP-1/CD169 cells , VLPs were colocalized with CD81 , but not with CD63 or Lamp1 ( Fig . 3E ) , which is consistent with previous reports on HIV-1 localization in mature DCs [16 , 17 , 23 , 24] . In contrast , colocalization of HIV Gag-mCherry VLPs in THP-1/CD169YF cells was reduced within CD81+ compartments but enhanced within CD63+ or Lamp1+ compartments ( Fig . 3E ) . In addition , VCCs in THP-1/CD169YF were inefficiently accessed by surface-applied antibodies ( S2 Fig B and C ) , suggesting that CD169YF internalized VLPs to late endosomes or lysosomes . These differences in intracellular localization of HIV Gag-mCherry VLPs between THP-1/CD169 and THP-1/CD169YF cells were statistically significant ( Fig . 3F and S2 Fig B ) . While THP-1/CD169YF cells captured HIV-1 particles as efficiently as THP-1/CD169 cells ( Fig . 3G ) , HIV-1 trans-infection of CD4+ T cells by THP-1/CD169YF cells was completely abrogated ( Fig . 3H and I ) . Collectively , these results suggest that endocytosed HIV-1 particles are incompetent for accessing the CD169-dependent HIV-1 trans-infection pathway . We next sought to characterize the architecture in greater detail of CD169+ VCCs in mature DCs . CD169 expression in DCs is induced upon treatment with TLR ligands such as LPS and polyI:C [14] . As opposed to TLRs , exposure to IFN-α that results in partial maturation of DCs [32 , 33] can also upregulate CD169 expression [14] , though putative differences in IFN-α and TLR-induced maturation phenotypes might alter virus trafficking in differentially matured DCs [34] . Therefore , DCs differentially matured with LPS or IFN-α ( referred as LPS-DC or IFN-α-DC , respectively ) , were used for determining HIV-1 localization in phenotypically divergent CD169-expressing primary cells . While CD169 was highly upregulated on both LPS-DCs and IFN-α-DCs ( Fig . 4A ) , IFN-α-DCs displayed a partial maturation phenotype expressing low levels of the activation antigens , CD86 and HLA-DR consistent with previously published findings [32 , 33] . HIV-1 capture by both LPS-DCs and IFN-α-DCs and subsequent trans-infection of CD4+ T cells were similarly enhanced over that observed with immature DCs ( Fig . 4B and C ) . We next investigated the nature of the CD169+ VCCs formed in IFN-α-DCs and LPS-DCs by conventional deconvolution microscopy , electron microscopy and super-resolution microscopy . While HIV-1 particles were strongly colocalized with CD169 at the cell periphery in both cell types , virus-containing pocket-like compartments were only found in LPS-DCs but not in IFN-α-DCs ( Fig . 4D ) . Electron microscopy also revealed virus-containing pocket-like compartments in LPS-DCs as previously reported ( Fig . 4E , S3 Fig A to C and [19 , 20] ) . In contrast most of HIV-1 particles were found at the surface in IFN-α-DCs ( Fig . 4E and S3 Fig D to F ) in valleys between dendritic extensions or present in structures presumably formed upon collapse of the dendrites in IFN- α-DCs ( Fig . 4E and S3 Fig D to F ) . This divergent localization of HIV-1 in IFN-α- and LPS-DCs was further investigated by super resolution microscopy . We used a fluorescence photoactivation localization microscopy ( FPALM ) with bi-plane capture technique [35–37] which allowed us to visualize CD169+ VCCs in mature DCs at 20–40 nm ( X-Y ) and 50–80 nm ( Z ) resolution . In agreement with conventional deconvolution and electron microscopy ( Fig . 4D and E ) , HIV-1 and CD169 were accumulated in pocket-like compartments in LPS-DCs , while HIV-1 was found mostly at the cellular edge in IFN-α-DCs ( Fig . 4F , top panels , S4 Fig D to F and S1 Movie ) . Focusing at the cross section of these cells ( along the line between a and b in the top panels ) , the depth of the compartments harboring HIV-1 particles in LPS-DCs was measured at 800 nm-1 μm ( Fig . 4F , middle panels ) . In contrast , HIV-1 particles ( p24gag ) and CD169 clustered in a long “valley-like” structure that appeared to be on the surface of IFN-α-DCs ( Fig . 4F , middle panels ) . In both cell types , p24gag molecules ( green ) were closely associated with CD169 ( red ) ( Fig . 4F , bottom panels , S4 Fig and S2 Movie and S3 Movie ) , implying an important role of CD169 in the formation of VCCs in DCs . All together , these results suggested that LPS and IFN-α treatment of DCs resulted in divergent CD169+ VCCs and that formation of CD169+ HIV-1 containing pocket-like structures in DCs requires a LPS-induced host factor . We next sought to determine if HIV-1 particles in CD169+ VCCs in LPS-DCs and IFN-α-DCs are exposed to the extracellular milieu . Uninfected or virus-exposed CD169+ LPS-DCs were subjected to extensive proteolytic digestion with either trypsin or pronase ( Fig . 5A ) . Cell-surface CD169 expression and amount of HIV-1 particles that remained associated with LPS-DCs and IFN-α-DCs following protease treatment was determined by FACS and p24gag ELISA , respectively ( Fig . 5A ) . In the absence of HIV-1 binding , CD169 was mostly present at the LPS-DC and IFN-α-DC surface , and remained sensitive to cleavage by pronase but not trypsin ( Fig . 5B and C , No virus ) , suggesting that extracellular domain of CD169 lacks trypsin-recognition sequences . Interestingly , when HIV-1 containing compartments were formed prior to pronase treatment , CD169 was still sensitive to pronase-digestion ( Fig . 5C , + Virus ) . We next investigated if HIV-1 particles associated with CD169 in LPS-DCs and IFN-α-DCs were sensitive to protease digestion . LPS-DC- or IFN-α-DC-associated HIV-1 particles were insensitive to trypsin exposure ( Fig . 5D ) , consistent with the findings that CD169 was trypsin resistant ( Fig . 5B and C ) and CD169—HIV-1 interaction is a protein ( CD169 ) —lipid ( GM3 ) interaction [13 , 14 , 38 , 39] . In contrast , consistent with the ability of pronase to effectively cleave cell-surface exposed CD169 ( Fig . 5B ) , pronase treatment decreased LPS-DC- or IFN-α-DC-associated HIV-1 content by ~60% ( Fig . 5D ) , suggesting VCCs were accessible to surface-applied pronase . The pronase-resistant cell-associated HIV-1 fraction might be attributed to those virus particles that either remain bound to residual CD169 ( ~20% of the CD169 molecules remained cell-associated even after pronase treatment; Fig . 5C ) , or p24gag in the cytoplasm after virus fusion with mature DCs . All together , these results suggest that the majority of CD169+ VCCs in LPS-DCs and IFN-α-DCs remain accessible from the cell surface and thus sensitive to surface-applied pronase digestion . We next sought to determine if CD169-bound HIV-1 particles in CD169+ VCCs in LPS-DCs or IFN-α-DCs were accessible to surface-applied large molecular probes , such as anti-gp120 broadly neutralizing antibodies ( bNAbs ) or anti-CD169 mAbs . LPS-DCs or IFN-α-DCs were pulsed with fluorescent HIV-1 Lai-iGFP particles and stained for either HIV-1 gp120 or CD169 prior to fixation and permeabilization such that antigens accessible to surface-applied antibodies would only be visualized . As a comparison , staining for total HIV-1 gp120 or CD169 was performed in parallel after fixation and permeabilization ( + Tx100 ) . Most of the HIV-1 particles and CD169 in IFN-α-DCs were found at the cell surface and could be visualized with surface-applied anti-gp120 ( Fig . 6A ) or anti-CD169 antibodies ( Fig . 6B ) . To quantify accessibility of captured HIV-1 particles to surface-applied antibodies , the fraction of fluorescent HIV-1 particles overlapping with antibody staining was calculated using Manders' coefficients . Quantification revealed no significant differences between surface-exposed and total molecule staining for both HIV-1 gp120 and CD169 in IFN-α-DCs ( Fig . 6C and D ) . Though some of the Lai-iGFP+ VCCs in LPS-DCs were stained by surface-applied anti-gp120 or anti-CD169 antibodies ( Fig . 6A and B ) , some of the virus particles present at the “bottom” of the pocket-like structures ( arrowheads , Fig . 6A and B , LPS ) were inaccessible to both anti-gp120 and anti-CD169 antibody suggesting that CD169+ VCCs in LPS-DCs were either closed structures or inaccessible to surface-applied probes due to steric hindrance . Furthermore , we observed statistically significant differences in Manders’ coefficients amongst cells ( LPS-DCs ) stained by the two distinct staining techniques ( Fig . 6C and D ) . The differences in surface accessibility of antibodies to CD169+ VCCs between LPS-DCs and IFN-α-DCs prompted us to hypothesize that HIV-1 particles localized within VCCs in LPS-DCs might remain competent for mature DC-mediated trans-infection even in the presence of anti-gp120 bNAbs . To test this hypothesis , neutralization assays were performed using anti-gp120 bNAbs ( VRC01 and NIH45–46 G54W ) and two-domain sCD4 ( sCD4–183 ) . Either LPS-DC- or IFN-α-DC-associated HIV-1 was incubated with increasing concentrations of VRC01 , NIH45–46 G54W or sCD4–183 prior to co-culture with CD4+ T cells . In parallel , cell free HIV-1 infection of CD4+ T cells was performed in the presence or absence of VRC01 , NIH45–46 G54W or sCD4 . While VRC01 and NIH45–46 G54W inhibited cell free CCR5-tropic HIV-1 ( pseudotyped with Bal Env ) infection of CD4+ T cells efficiently [IC50 ( VRC01 ) = 0 . 035 ± 0 . 005 μg/ml ( Fig . 6E and H ) , IC50 ( NIH45–46 G54W ) = 0 . 012 ± 0 . 004 μg/ml ( Fig . 6F and I ) ] , transfer of LPS-DC-associated HIV-1 particles was inefficiently neutralized [IC50 ( VRC01 ) = 1 . 152 ± 0 . 308 μg/ml ( Fig . 6E and H ) and IC50 ( NIH45–46 G54W ) = 0 . 223 ± 0 . 062 μg/ml ( Fig . 6F and I ) ] . Interestingly , transfer of HIV-1 particles from IFN-α-DCs to T cells was more susceptible to neutralization by VRC01 and NIH45–46 G54W [IC50 = 0 . 508 ± 0 . 155 μg/ml ( Fig . 6E and H ) and 0 . 069 ± 0 . 016 μg/ml ( Fig . 6F and I ) , respectively] than that mediated by LPS-DCs , though efficiency of IFN-α-DC mediated transfer in the presence of VRC01 and NIH45–46 G54W was still greater than that observed for cell-free infection of T cells . In contrast , sCD4–183 , a small gp120-neutralizing reagent ( 26kD ) was able to equally inhibit all three modes of infection , namely cell free , IFN-α-DC-mediated and LPS-DC-mediated HIV-1 infection of CD4+ T cells ( Fig . 6G ) . The IC50 values for cell free , IFN-α-DC or LPS-DC-associated HIV-1 infection of CD4+ T cells were 0 . 171 ± 0 . 047 , 0 . 307 ± 0 . 039 and 0 . 467 ± 0 . 117 μg/ml , respectively ( Fig . 6J ) . These results suggest that HIV-1 association with CD169 in VCCs within mature DCs protects viruses from detection by anti-gp120 bNAbs and might provide virus evasion from adaptive immune responses in vivo . In this study , we have characterized CD169+ VCCs in mature DCs and found that captured HIV-1 particles in LPS-matured DCs were localized within surface-connected plasma membrane invaginations at depths of ~800nm—1μm ( Fig . 4 ) . We hypothesize that multivalent association of CD169 and HIV-1 particles or clustering of multiple CD169 molecules ( induced upon virus particle binding ) might enhance localized concentration of receptor-ligand complexes that are retained at the cell surface because of the inability of CD169 to mediate endocytosis . Recruitment of a LPS-induced myeloid cell-specific co-factor ( s ) upon virus capture to the localized membrane microdomain might place additional strain and stress on the membrane that is relieved by formation of membrane invaginations ( Fig . 7 ) , though the mechanisms that inhibit membrane closure and endosome formation remain to be identified . Interestingly , localization of HIV-1 within VCCs in LPS-matured DCs reduced the accessibility of anti-gp120 bNAb , VRC01 , to virus particles and hence , reduced the neutralization efficiency of anti-gp120 bNAbs , VRC01 and NIH45–46 G54W ( Fig . 6 ) . Localization of HIV-1 and CD169 in a lattice-like structure in the VCCs might provide steric hindrance only to large molecules such as neutralizing antibodies . HIV-1 particles captured by IFN-α-DC-associated HIV-1 were also less susceptible to VRC01 and NIH45–46 G54W compared to cell free HIV-1 , though localization of HIV-1 particles in IFN-α-DCs was in compartments that lacked comparable depth to that observed in LPS-DCs ( Fig . 4 ) , presumably because of the lack of recruitment of the co-factor ( s ) in IFN-α-DCs necessary for formation of membrane invaginations . Since HIV-1 particles were found at the bottom of dendrites and/or surrounded by dendrites ( Fig . 4E ) forming clusters of CD169 and HIV-1 in a "valley-like" structure ( Fig . 4F ) , this unique localization of HIV-1 in the IFN-DCs might also hinder access of VRC01 and NIH45–46 G54W to HIV-1 . Acute infection of HIV-1 in vivo induces various pro-inflammatory cytokines including type-I interferon [40] . Such inflammatory conditions can differentiate monocytes at the site of infection into inflammatory DCs [41–43] . We have reported previously that inflammatory DCs generated in vitro are CD169+ and efficiently disseminate HIV-1 to T cells [14] . While triggering type I IFN responses induces the expression of number of interferon-stimulated genes , some of which are anti-viral , and restrict virus replication , induction of CD169 might offset ISG-mediated restrictions to virus replication in the peripheral mucosa . Thus , in acute phase of infection , type I IFN-induced CD169 on inflammatory DCs might support establishment of infection in mucosal CD4+ T cells . In addition to IFNs , previous studies have demonstrated increases in serum LPS levels over the course of HIV-1 infection , primarily due to the compromised integrity of gut epithelium [44] . Mucosal damage-associated-translocation of LPS might lead to systemic activation of DCs and upregulation of CD169 that not only enhances virus spread to CD4+ T cells , but also might provide evasion from humoral responses that develop but fail to neutralize cell-to-cell transmission in the mucosal tissues . A great deal of effort currently supports the design of viral vector-based immunoprophylactic regimens that express anti-gp120 bNAbs to induce protection in vivo [45 , 46] . Since DC-mediated trans-infection of CD4+ T cells has been suggested as an important pathway of HIV-1 dissemination in vivo [2 , 3] , significantly increased antibody titers might be necessary in vivo to achieve neutralization of IFN-α-DC or LPS-DC-mediated HIV-1 dissemination . CD169 is expressed exclusively on myeloid cells in vivo [15 , 47] , and interestingly formation of CD169+ VCCs upon HIV-1 capture was only recapitulated in DCs and THP-1 monocytoid cell line , but not Raji B cells or HeLa cells , constitutively expressing CD169 ( Fig . 1B ) . While HIV-1 particle associated with CD169 remained at the cell surface in Raji/CD169 cells , virus particles accumulated in intracellular , surface-inaccessible compartments in HeLa/CD169 cells . These results suggest the formation of surface-connected VCCs might require a cofactor specific to myeloid cells . Interestingly , CT sequences of CD169 proved dispensable for VCC formation , since truncation of CT downstream of the four membrane-proximal arginine residues ( THP-1/CD169ΔCT4R ) preserved cell surface expression of CD169 and importantly resulted in VCC formation in THP-1 cells upon HIV-1 capture ( Fig . 2 ) . Previous results from our laboratory and others have also demonstrated that HIV-1 capture by DCs is also reduced upon treatment with β-methyl-cyclodextrin , a cholesterol sequestering reagent [17 , 48] . Since VCC formation occurred even in the absence of CT , it is possible that lateral association and clustering of CD169 is driven by interaction of transmembrane domain of CD169 with a protein and/or lipid molecule in such cholesterol-rich plasma membrane microdomains . It is of note that expression of such cofactor ( s ) is regulated by LPS stimulation of DCs but not upon treatment with IFN-α alone ( Fig . 4D ) . Future studies will be needed to identify the nature of this myeloid-cell specific co-factor by comparing TLR4-mediated ( TRIF or MyD88-dependent ) and IFNAR-mediated ( JAK-STAT dependent ) signaling pathways in myeloid cells . CD169 is a pattern recognition receptor that captures diverse bacterial and viral pathogens by recognizing α2 , 3-sialylated glycoconjugates on the pathogen surface [47] . In addition to HIV-1 , capture of other enveloped viruses such as murine leukemia virus , nipah and hendra hemorrhagic fever viruses by CD169 is also dependent on binding α2 , 3-sialylated GSLs incorporated in the virus particle membranes [14 , 49] . Though some studies have implicated CD169 as an endocytic receptor that mediates internalization of pathogens into early endosomes [26] , CD169 , unlike other members of the Siglec protein family , CD169 has no defined endocytic motifs in its CT [15 , 47] . Furthermore , studies described in this report suggest that HIV-1 particles captured by CD169 are not targeted for endocytosis but rather retained on the myeloid cell-surface in plasma membrane invaginations . Interestingly , exogenous introduction of a di-aromatic endocytic motif in the CT of CD169 resulted in HIV-1 internalization and dramatic attenuation of CD169-mediated HIV-1 trans-infection ( Fig . 3 ) . Collectively , these results strongly suggest a requirement for HIV-1 retention at the cell surface for accessing the mature DC/CD169-mediated trans-infection pathway . We hypothesize that this unique trafficking pattern is beneficial to HIV-1 since it provides virus particles evasion from endocytic pathways in DCs that can result in degradation of virions and/or antigen presentation to T cells to elicit robust adaptive immune responses [50] . It is interesting to speculate that HIV-1 might have evolved to assemble and exit from GM3-enriched plasma membrane microdomains [49] such that GM3-dependent interactions of HIV-1 with CD169 provide virus sanctuary from both myeloid cell-intrinsic phagocytic mechanisms of virus degradation and antibody-dependent detection and neutralization of virus infectivity . Furthermore , as DCs have been proposed to be the first cells to encounter HIV-1 particles in the genital mucosa [2] , topical administration of such reagents might prevent sexual transmission of HIV-1 Therefore , development of agents that target HIV-1–CD169 interaction might be an attractive potential anti-viral therapeutic to curtail the HIV-1 pandemic . This research has been determined to be exempt by the Institutional Review Board of the Boston University Medical Center since it does not meet the definition of human subjects research , since all human samples were collected in an anonymous fashion and no identifiable private information was collected . Human CD169 was cloned into a retroviral expression vector , LNCX ( LNC-CD169 ) and has been described previously [14] . Truncations in cytoplasmic tail of human CD169 were introduced by PCR using the following primer sets: for CD169/ΔCT , CD169–4188-sense ( ATCAGGGACAGGCCATGTCC ) and CD169-ΔCT-antisense ( TTTTTATCGATCACCAGGTGTAGCAGGCCC CCAGG ) ; for CD169/ΔCT4R , CD169–4188-sense and CD169-ΔCT4R antisense ( TTTTTATCGATTAACGCCTCCTTCTCCAGGTGTAGCAGGC ) . Point mutation in the cytoplasmic tail of CD169 ( A1683Y ) was introduced by PCR-based site-directed mutagenesis ( QuikChange; Agilent Technologies ) using the following primers: CD169-YF-sense ( CGAGAATTCGGTGGAGATGTATTTTCAGAAAGAGACCACGC ) and CD169-YF-antisense ( GCGTGGTCTCTTTCTGAAAATACATCTCCACCGAATTCTCG ) . A SbfI-ClaI fragment containing truncations or mutations in the CT of CD169 was replaced into the corresponding portion of LNC-CD169 . All clones were verified by sequencing . Stable expression of CD169 CT mutants in THP-1 monocytic cell line , HeLa cell line and Raji B cell line was accomplished by transduction with VSV-G pseudotyped LNC-CD169 mutant retroviral vectors followed by G418 selection as previously described [14] . CD169 positive cells were further purified either by MACS ( Miltenyi Biotec ) or FACS ( BD AriaIII ) . Protein expression was confirmed by western blot analysis and flow cytometry ( BD Calibur ) as described below . The expression plasmid , pGag-EGFP , that expresses a HIV-1 Gag-eGFP fusion protein , was obtained from the NIAID AIDS Reference and Reagent Program . HIV-1 Gag-mCherry expression plasmid that expresses a red fluorescent Gag-mCherry fusion protein has been described previously [24] . HIV-1 LaiΔenv-luc ( Env deficient HIV-1 Lai containing a luciferase reporter gene in place of the nef orf ) , Lai/Balenv-luc ( a CCR5-tropic infectious proviral construct encoding luciferase ) and a CCR5-tropic infectious proviral plasmid Lai/YU-2env have been described previously [51–53] . Lai-imCherry , a proviral construct producing red fluorescent infectious virus particles , was derived from Lai-iGFP [14] by replacing the GFP-encoding fragment with that of mCherry . The CCR5-tropic HIV gp160 ( Bal env ) expression vector was generated from a CXCR4-tropic HIV gp160 ( Lai env ) expression vector [54] by replacing the entire Lai env gene with the corresponding region of Bal env . Human dendritic cells ( DCs ) were derived from CD14+ peripheral blood monocytes , as described previously [14] . DCs were matured with ultrapure E . coli K12 LPS ( 100 ng/ml; Invivogen ) or IFN-α ( 1000 U/ml; PBL Interferon Source ) for 2 days prior to use in the assays . Primary human CD4+ T cells were positively isolated from CD14-depleted PBMCs , using CD4-conjugated magnetic beads and LS MACS cell separation columns ( Miltenyi Biotech ) . Positively isolated CD4+ T cells were activated with 2% PHA ( Invitrogen ) for 2 days , washed and cultured in IL-2 ( 50 U/ml ) containing RPMI supplemented with 10% FBS . HEK293T ( human kidney epithelial cell line ) , Raji ( human B cell line , obtained from the NIH AIDS Research and Reference Reagent Program ) , THP-1 ( human monocytic cell line , clone ATCC , obtained from the NIH AIDS Research and Reference Reagent Program ) , and HeLa cells have been described previously [39] . Replication competent viruses , Lai/Balenv-luc , Lai/YU-2env and Lai-imCherry , were derived from HEK293T cells via calcium phosphate transfection as described previously [54] . Fluorescent HIV Gag derived virus-like particles ( VLPs ) were generated via transient transfections of HEK293T cells with HIV Gag-eGFP or HIV Gag-mCherry expression plasmids . HIV-1 vectors pseudotyped with Bal Env were generated from HEK293T cells via co-transfection of HIV-1 LaiΔenv-luc with HIV-1 Bal Env expression plasmid . Viruses or VLP-containing cell supernatants were harvested 2 days post-transfection , cleared of cell debris by centrifugation ( 300 x g , 5 min ) , passed through 0 . 45 μm filters , and stored at—80°C until further use . For some experiments , viruses in the supernatants were concentrated by ultracentrifugation on a 20% sucrose cushion [24 , 000 rpm and 4°C for 2 hr with a SW32Ti rotor ( Beckman Coulter ) ] . The virus pellets were resuspended in PBS , aliquoted and stored at -80°C . The capsid content of infectious HIV-1 particles or VLPs was determined by a p24gag ELISA [54] . VSV-G pseudotyped LNC-CD169 mutant retroviral vectors were prepared as described elsewhere [14] . Mature DCs ( 1x105; see above ) , THP-1 cells ( 1x105 ) , Raji cells ( 1x105 ) or HeLa cells ( 5x104 ) were incubated with virus ( 10–20 ng p24gag ) for 2 hr at 37°C in complete RPMI media , washed 4 times with PBS and analyzed for capture using either p24gag ELISA . Virus capture was quantified by measuring p24gag associated with lysed cells using an in-house p24gag ELISA described previously [54] . For transfer of Lai/Balenv-luc infectious viruses , 1x105 of mature DCs , THP-1 cells , Raji cells or 5x104 HeLa cells were incubated with virus ( 10–20 ng p24gag ) for 2 hr at 37°C in complete RPMI media , washed 4x with PBS and co-cultured with autologous or heterologous CD4+ T cells at a 1:1 or 1:2 cell ratio in complete RPMI media with IL-2 . The cells were lysed at 48 hours post infection and luciferase activity in the cell lysates was measured using Bright-Glo ( Promega ) . All assays were performed with cells derived from a minimum of three independent donors and each experiment was performed in triplicate . To assess expression of CD169 CT mutants in THP-1 cells , cell lysates were resolved with SDS-PAGE and transferred onto PVDF membranes . Membranes were probed with mouse anti-CD169 antibody ( 7D2 , Novus Biologicals ) or rabbit anti-actin antibody ( SIGMA ) . To measure CD169 expression on the cell surface , cells were stained with Alexa488-conjugated mouse anti-CD169 ( AbD Serotec ) and analyzed with a FACS Calibur ( BD ) , as detailed in supporting methods ( see S1 Text ) . To determine the extent of cell-surface exposure of CD169 bound HIV-1 particles on mature DC surface , cells incubated in the presence or absence of 10 ng ( p24gag ) of HIV-1 for 2 hours were washed extensively with cold-PBS and chilled at 4°C for 30 min , prior to incubation with pronase ( 4 mg/ml , in Ca2+ containing PBS , Roche ) for 30 min at 4°C . Alternatively , virus-exposed cells were incubated with 0 . 25% trypsin ( Invitrogen ) for 5 min at 37°C . After the treatment , cells were washed extensively with cold-PBS . The amount of cell-associated HIV-1 particles was measured by p24 ELISA ( described above ) and CD169 expression was measured by FACS as described above . The values were normalized to those of untreated samples . To investigate sensitivity of DC-associated HIV-1 to gp120-targeting neutralizing reagents , VRC01 , NIH45–46 G54W and sCD4–183 ( obtained from the NIH AIDS Reagent Program ) , 5x104 mature DCs were incubated with HIV-1 Bal Env pseudotyped LaiΔenv-luc particles ( 10 ng p24gag ) for 2 hours at 37°C , washed 4 times with PBS and chilled at 4°C for 15 min . Serially diluted VRC01 , NIH45–46 G54W or sCD4–183 starting at 10 or 20 μg/ml in final was added to HIV-1 pulsed DCs or cell free virus ( 50 ng of p24gag ) and incubated for 1 hour at 4°C . Cells were washed twice with cold-PBS and autologous or heterologous CD4+ T cells were added at 1:2 ratio to monitor trans-infection of CD4+ T cells as described above . Cell free HIV-1 was added directly to CD4+ T cells . These experiments were performed in triplicates with DCs from at least nine independent donors . Cell free infections of CD4+ T cells were performed with cells derived from at least five independent donors in triplicates . Nonlinear regression was used to estimate a fitted curve and IC50 values were calculated in GraphPad Prism 5 . To investigate structure of HIV-1 containing CD169+ compartments in DCs , super resolution FPALM ( fluorescence photoactivated localization microscopy ) was used . LPS- or IFN-α-stimulated DCs ( 1x106 cells ) were incubated with 2 . 5 μg p24gag Lai/YU-2env for 2 hours at 37°C and washed extensively to remove unbound viruses . Cells were fixed with 4% PFA , permeabilized , blocked with normal donkey serum and stained for HIV-1 with mouse monoclonal anti-p24 antibody ( AG3 . 0 , obtained from the NIH AIDS Reagent Program ) , followed by secondary donkey anti-mouse IgG-Cy3B . Cells were blocked with 20% normal mouse serum , and CD169 expression was visualized with Alexa647-conjugated anti-CD169 mAb ( AbD Serotec ) . Cells were attached onto a grass coverslip and subjected to microscopy analysis . Images were recorded with a Vutara 200 super-resolution microscope ( Bruker Nano Surfaces , Salt Lake City , UT ) based on the Biplane FPALM approach [36] . Samples were imaged using a 647 nm and 488 nm excitation lasers , respectively , and 405 nm activation laser in photoswitching buffer comprising of 20 mM cysteamine , 1% betamercaptoethanol and oxygen scavengers ( glucose oxidase and catalase ) in 50mM Tris buffer at pH 8 . 0 . Images were recorded using a 60x/1 . 2 NA Olympus water immersion objective and Photometrics Evolve 512 EMCCD camera with gain set at 50 , frame rate at 50 Hz and maximal powers of 647 nm , 488 nm and 405 lasers set at 8 , 10 , and 0 . 05 kW/cm2 respectively . Data was analyzed by the Vutara SRX software ( version 4 . 09 ) . Briefly , particles were identified by their brightness from the combined images taken in both planes and two color channels simultaneously . If a particle was identified in multiple subsequent camera frames , data from these frames was combined for the specific identified particle . Identified particles were then localized in three dimensions by fitting the raw data in a customizable region of interest ( typically 16x16 pixels ) centered around each particle in each plane with a 3D model function which was obtained from recorded bead data sets . The four-recorded fields were aligned automatically by computing the affine transformation between each pair of planes . Sample drift was corrected by cross-correlation of the determined localized particles [55] or tracking of fiduciary markers . Fit results were stored as data lists for further analyses . Structure of HIV-1 containing CD169+ compartments in DCs was visualized by electron microscopy . 4 . 5 x106 LPS- or IFN-α-stimulated DCs were incubated with 9 μg p24gag Lai/YU-2env for 2 hours at 37°C and washed extensively to remove unbound viruses . Cells were fixed with 4% PFA and 1% glutaraldehyde in 0 . 1 M PHEM buffer ( 60 mM PIPES , 25 mM HEPES , 2 mM MgCl2 and 10 mM EGTA ) . Cells were further fixed with 2% osmium tetroxide , dehydrated in ethanol and embedded in epoxy resin as previously reported [29] . Ultra-thin sections ( 60–80 nm ) of embedded cells were stained with 3% uranyl acetate and 1% lead citrate and subjected to imaging with a Philips CM-12 electron microscope at 100kV . To determine if VLP containing compartments remained connected to the cell surface , HeLa/CD169 cells ( seeded on coverslips in a 24-well tissue culture plate on the day before ) , THP-1/CD169 cells , Raji/CD169 cells or mature DCs ( 2x105 cells ) were incubated with 10 ng p24gag of VLP Gag-mCherry for 2 hours at 37°C , washed extensively to remove unbound VLPs , chilled to 4°C and stained with Alexa488-conjugated mouse anti-CD169 mAb ( AbD Serotec ) on ice for 1 hour , prior to fixation with 4% PFA . For total CD169 staining , virus-exposed cells were fixed , permeabilized and stained with Alexa488-conjugated mouse anti-CD169 for 1 hour at RT . To determine intracellular localization of VLPs , THP-1/CD169 ( 2x105 ) cells were incubated with 10 ng p24gag Gag-mCherry VLPs for 2 hours at 37°C , washed and fixed with 4% PFA . Virus-containing compartments were visualized by staining with anti-human CD81 ( BD ) , anti-human CD63 ( Santa Cruz ) or anti-human Lamp1 ( Santa Cruz ) at 10 μg/ml followed by secondary Alexa594-conjugated goat anti-mouse IgG ( Invitrogen ) at 10 μg/ml . To visualize HIV-1 containing compartments in DCs , LPS- or IFN-α-stimulated DCs ( 4x105 cells ) were incubated with 1 μg Lai-iGFP for 2 hours at 37°C and washed extensively to remove unbound viruses . For staining of surface-exposed HIV-1 particles or CD169 , cells were chilled and stained with anti-gp120 antibody ( VRC01 ) or anti-CD169 mAb ( 7D2 , Novus Biologicals ) , respectively , at 10 μg/ml on ice for 1 hour , prior to fixation with 4% PFA . Alternatively , virus-exposed cells were fixed with 4% PFA , permeabilized with TritonX-100 and then stained with anti-gp120 antibody ( VRC01 ) or anti-CD169 mAb ( 7D2 , Novus Biologicals ) . Cells were then stained with Alexa594-conjugated goat anti-human IgG ( for visualizing gp120 staining , Invitrogen ) or Alexa594-conjugated goat anti-mouse IgG ( for CD169 staining , Invitrogen ) . Nuclear staining was visualized with DAPI ( Sigma ) and cells were mounted on a glass slide with Fluoromount G ( Southern Biotech ) . Images were acquired using a Olympus IX70 microscope equipped for DeltaVision deconvolution ( Applied Precision ) . Images were deconvolved using the SoftWoRx software ( Applied Precision ) , processed with ImageJ and pseudocolored for data presentation . For the colocalization study on THP-1 cells , images were acquired for at least 20 cells , deconvoluted , flattened for maximum intensity in order to avoid selection bias inherent in analysis of single focal plane images and analyzed for Pearson’s coefficient of correlation ( R ) with ImageJ . For the quantification of accessibility of antibodies to HIV-1 particles in CD169+ VCCs in differentially stimulated mature DCs , images were acquired on 10–15 cells , deconvolved , and flattened for maximum intensity . To specifically quantify the fraction of fluorescent HIV-1 particle overlapping with antibody signals ( acquisition of red ( antibody ) on green ( HIV-1 ) ) , Manders’ coefficients were calculated using ImageJ . Thresholds were set as the mean ± standard deviation of intensity at each channel .	Dendritic cells ( DCs ) are professional antigen presenting cells , and their sentinel roles are important to elicit a potent antiviral immunity . However , HIV-1 has exploited DCs to spread infection by several mechanisms . One such mechanism is the DC-mediated trans-infection pathway , whereby DCs transmit captured virus to CD4+ T cells . We have recently identified the type I interferon ( IFN-I ) inducible protein , CD169 , as a receptor on DCs which mediates HIV-1 capture and trans-infection . We have also demonstrated extensive co-localization of HIV-1 with CD169 within peripheral non-lysosomal compartments in DCs , although the mechanism and biological importance of the compartment formation remain unclear . Here in this study , we report that a myeloid cell specific co-factor interacts with CD169 following virus capture leading to compartment formation . This co-factor is induced in DCs by an IFN-I-inducing TLR ligand LPS , but not by IFN-I itself . Though the CD169+ HIV-1 containing compartments are surface-accessible , these compartments have considerable depth and are connected to the surface , such that captured virus particles localized within these unique structures are protected from detection by anti-gp120 broadly neutralizing antibodies . Our study suggests that CD169–HIV-1 interaction provides an evasion mechanism from degradation by phagocytosis and neutralization by anti-viral humoral responses .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	CD169-Mediated Trafficking of HIV to Plasma Membrane Invaginations in Dendritic Cells Attenuates Efficacy of Anti-gp120 Broadly Neutralizing Antibodies
It is often supposed that a protein's rate of evolution and its amino acid content are determined by the function and anatomy of the protein . Here we examine an alternative possibility , namely that the requirement to specify in the unprocessed RNA , in the vicinity of intron–exon boundaries , information necessary for removal of introns ( e . g . , exonic splice enhancers ) affects both amino acid usage and rates of protein evolution . We find that the majority of amino acids show skewed usage near intron–exon boundaries , and that differences in the trends for the 2-fold and 4-fold blocks of both arginine and leucine show this to be owing to effects mediated at the nucleotide level . More specifically , there is a robust relationship between the extent to which an amino acid is preferred/avoided near boundaries and its enrichment/paucity in splice enhancers . As might then be expected , the rate of evolution is lowest near intron–exon boundaries , at least in part owing to splice enhancers , such that domains flanking intron–exon junctions evolve on average at under half the rate of exon centres from the same gene . In contrast , the rate of evolution of intronless retrogenes is highest near the domains where intron–exon junctions previously resided . The proportion of sequence near intron–exon boundaries is one of the stronger predictors of a protein's rate of evolution in mammals yet described . We conclude that after intron insertion selection favours modification of amino acid content near intron–exon junctions , so as to enable efficient intron removal , these changes then being subject to strong purifying selection even if nonoptimal for protein function . Thus there exists a strong force operating on protein evolution in mammals that is not explained directly in terms of the biology of the protein . Why do some parts of proteins evolve more slowly than others ? Why , in turn , do some proteins evolve more slowly than others ? Intragenic conserved regions are typically considered to reflect domains of functional importance to the protein [1] . Similarly , proteins with a high density of important functional sites should evolve slowly . There are , however , potentially multiple other correlates to rates of protein evolution [1] . The expression parameters of a gene ( rate of expression , protein abundance , and number of tissues in which a gene is expressed ) are consistently reported to be important predictors [2–5] . This may in part reflect selection to resist mistranslation [6] . Other possible covariates include essentiality and the number of protein interactions , but the issues here are more contentious , not least because of covariance with expression parameters [7–17] . Here we test the hypothesis that selection acting to ensure that introns are correctly removed skews amino acid content in predictable ways and imposes constraints on rates of protein evolution . In mammalian genes , which are rich in introns [18] , correct removal of introns often requires the presence , in the flanking exons , of splice-enhancer domains , these being short ( six nucleotide ) blocks required for binding of serine/arginine-rich proteins [19] . The need for splice enhancers can impact the use of synonymous codons in the domains flanking intron–exon junctions , such that when a synonymous codon is used commonly in splice enhancers it is preferred over its less commonly used synonym [20 , 21] . Moreover , selection to preserve splice enhancers affects both the synonymous single nucleotide polymorphism profile [22 , 23] and the rate of evolution at synonymous sites of splice-enhancer-associated domains [24] . Might the same forces also act to cause skews in amino acid usage in the vicinity of intron–exon junctions ? In a preliminary analysis , we showed that there is a tendency for enrichment near boundaries of an amino acid whose codons are common in splice enhancers: lysine is coded by AAA and AAG , both of which are common in splice enhancers , and at both 5′ and 3′ ends of exons , lysine's proportional usage increases [24] . Is it more generally the case that an amino acid's usage increases near intron–exon junctions if it commonly features in splice enhancers ? Conversely , are some amino acids avoided near such boundaries if they are rare in splice-enhancer domains ? To address these issues , we derive patterns of amino acid preference in the vicinity of intron–exon boundaries and compare these patterns with a metric of enrichment of amino acids in splice enhancers relative to rates of usage in the genome . In turn , we ask whether selective constraints are stronger near intron–exon boundaries , and whether such constraints explain much of the variation between proteins in their rate of evolution . For 178 , 382 human exons we considered the trends in amino acid composition as one approaches the intron–exon boundary , as assayed by the rank correlation , rho , between distance from the boundary and proportional usage of the amino acid . Considering the 2-fold and 4-fold blocks of the 6-fold degenerate amino acids as different groupings , we found that of 46 independent comparisons ( 23 amino acid groups 5′ and 3′ prime ) , 34 showed significant trends for enrichment or avoidance near intron–exon boundaries ( Table 1 ) . After Bonferroni correction 26 remained significant ( with 46 comparisons p < 0 . 001 indicates significance ) . For all plots for individual amino acids see Figure S1 . We repeated the analysis for 115 , 466 exons from 14 , 005 mouse genes and found that patterns of preference are strikingly similar between the two species ( Table 1 ) . In mice , 34 amino acids again showed significant trends , and the correlation of rho values for 46 comparisons in mice versus human was extremely high ( Pearson product moment correlation , r = 0 . 96 , p < 0 . 0001 ) . Do these effects necessarily relate to the nucleotide content of the codons , as the splice-regulation model requires ? One might conjecture instead that these effects reflect some coincidence of exon boundaries with protein substructures having unusual amino acid contents . Several facts strongly support the hypothesis that the trends seen are at least in part driven by effects at the nucleotide level . Notably , while the 2-fold block of arginine ( amino acid r in Table 1 ) was strongly preferred near boundaries at both 3′ and 5′ ends , the 4-fold redundant block ( amino acid R ) showed the reverse pattern . A comparable difference was seen for the 2-fold ( amino acid l ) and 4-fold ( amino acid L ) blocks for leucine . The same pattern was seen in mouse genes . A preference for certain amino acids , regardless of the nucleotide content of their codons , would not have predicted this . If splice-enhancer domains impact amino acid usage near intron–exon boundaries , we expect that those amino acids preferred in splice enhancers should be preferred near junctions ( i . e . , rho < 0 ) . To test this we developed a metric of involvement of codons in splice-enhancer hexamers , which we term the hexamer preference index ( HPI ) . Using hexamers found both in mouse and human to define the HPI ( and ignoring 3′ and 5′ differences ) , we found a striking predictability of patterns of preference near boundaries ( Spearman rank correlation between HPI and rho for preference/avoidance near boundaries , rho = −0 . 54 , p < 0 . 0001 , n = 46 ) . As an alternative to rho , we can employ the slope of the best-fit regression line between proportional usage of an amino acid and distance from intron–exon junctions . A negative slope , like a negative rho , indicates preferential usage near junctions . Using this slope on the best-fit regression line revealed , as expected , the same trend ( Spearman rank correlation , slope versus HPI = −0 . 57 , p < 0 . 0001; Figure S2 ) . The trend for preference of high HPI amino acids near boundaries was also seen in mice ( e . g . , using mouse–human overlap set of hexamers , correlation of rho with HPI = −0 . 49 , p = 0 . 0005; correlation of slope with HPI = −0 . 52 , p = 0 . 0002 ) . These results are not greatly affected by considering 5′ and 3′ ends separately ( Spearman rank correlation between rho 5′ and HPI 5′ using human 5′-specific hexamers = −0 . 59 , p = 0 . 003 , n = 23 , Figure 1A; between rho 3′ and HPI 3′ using human 3′-specific hexamers = −0 . 57 , p = 0 . 004 , n = 23 , Figure 1B ) . This is reflected in the fact that trends in usage ( rho ) and patterns of HPI are similar 5′ and 3′ ( Pearson correlation , r , between rho 5′ and rho 3′ for the 23 amino acid classes = 0 . 80 , p < 0 . 0001; Pearson correlation between HPI 5′ and HPI 3′ for the 23 amino acid classes = 0 . 95 , p < 0 . 0001 ) . One might suppose that our measure of HPI might be biased by incomplete knowledge of enhancers . We can control for this , in part , by recognizing that splice enhancers tend to be adenine rich and cytosine poor . Consider then the composite measure AC bias = frequency of adenine in synonymous codon set – frequency of cytosine . For example , in the 4-fold degenerate set for alanine ( GCN ) , of the 12 bases in four possible synonymous codons , adenine and thymine both featured 1/12 of the time , and guanine and cytosine both featured 5/12 of the time . So AC bias for alanine is 1/12 − 5/12 = −1/3 . This AC bias was a robust predictor of preference/avoidance near boundaries ( Spearman rank correlation , AC bias versus rho = −0 . 67 , p < 0 . 0001 ) ( Figure 2 ) . Avoidance of cytosine in the synonymous codons appeared to be a somewhat stronger predictor of patterns of avoidance or preference of amino acids than was preference for adenine ( Spearman rank correlation , cytosine content of codons versus rho = 0 . 67 , p < 0 . 0001; adenine content of codons versus rho = −0 . 37 , p = 0 . 01 ) . Neither thymine nor guanine content showed any trends ( p >> 0 . 05 ) . These results suggest that the general profile of enhancers and the specifics employed to define HPI are about equally good predictors of patterns of preference/avoidance . The above results suggest that selection acts to prefer nucleotides that permit efficient intron removal . Does this in turn affect rates of protein evolution ? Were there such an effect , we should expect that smaller exons should evolve more slowly , as a higher proportion of sequence is near ( e . g . , within 70 bp ) boundaries . Indeed , from a set of 36 , 683 mouse–human aligned exons , we found that small exons do tend to have low rates of evolution ( Spearman rank correlation between the number of nonsynonymous substitutions per nonsynonymous site [KA] and exon length , rho = 0 . 15 , p < 0 . 0001 ) . This might , however , be owing to a trend for genes with small exons to be disproportionately in functional classes of protein that have intrinsically low rates of evolution . To control for this we considered , for all genes with more than two internal exons , the Spearman rank correlation between exon KA and the size of the exon . As each correlation coefficient is derived from a given gene , between-gene variation in KA ( and indeed the number of synonymous substitutions per synonymous site [KS] ) is controlled for in any such analysis . If splice control impacts rates of exon evolution we expect that on the average this correlation should be positive , while the null hypothesis , that small exons have low rates of evolution because they derive from classes of genes with intrinsically low KA , predicts a mean rho of zero . The distribution of rho was very strongly skewed to positive values ( median rho = +0 . 14 , Wilcoxon rank test , p < 0 . 0001 , n = 3 , 629 ) . Restricting analysis to genes with ten or more exons only strengthened this conclusion ( median rho = +0 . 16 , p < 0 . 0001 , n = 1 , 286 ) . Is there also a trend for lower rates of evolution near boundaries ? Using all exons , asking about the proportion of all sites a given distance from a boundary ( 5′ or 3′ ) in which we see a nonsynonymous change , we observed the predicted low rate of amino acid evolution near boundaries ( Spearman rank correlation , proportion of aligned sites showing nonsynonymous change versus distance from boundary , rho = 0 . 955 , p < 0 . 0001 ) ( Figure 3 , circles ) . Might this result simply be an artefact of the possibility that small exons might both come disproportionately from a class of slow-evolving genes and contribute more data to the estimate of divergence near the exon–intron junctions than they do to the more distant sites ? To control for this , we again considered divergence rates within 40 codons of boundaries ( 5′ and 3′ ) but considered only the 1 , 836 exons that are at least 80 codons long . This way all exons contribute approximately the same amount of data at all distances from the junction . We found that the lower rate of evolution near the boundary remained highly robust ( rho = 0 . 7685 , p < 0 . 0001 ) ( Figure 3 , squares ) . Note , however , that absolute rates of evolution , at any given distance from the boundary , were higher in this long exon set . This is consistent either with reduced density of splice-control elements near boundaries in long exons or with a splice-unrelated force acting more profoundly on long exons . There is good evidence for the former . When we examined the density of putative exonic splice enhancers ( ESEs ) in the exon span within 100 bp of a boundary at either end ( or all of the exon in the case of exons shorter than 200 bp ) , we found a robust negative correlation between enhancer density and exon size ( rho = −0 . 18 , p < 0 . 0001 ) . Comparably , when we considered exons longer than 200 bp to be long exons and those shorter than this to be short exons , we found that ESEs occupy a median of 31% of the short exons , but only 21% of the 200 bp near the boundaries ( 100 bp 5′ and 100 bp 3′ ) of the long exons . This is consistent with the idea that there is less space in short exons to pack in the information necessary to enable proper splicing . As expected , KA was lower in ESEs than in nonenhancers ( Figure 4 ) ( see also [24] ) . This was also true if we restricted analysis to exons longer than 200 bp ( paired test , p < 0 . 0001 ) ( Figure S3 ) . These results tally with the finding that genes with long introns tend to have low rates of evolution [12] , as exons flanked by long introns tend to be richest in ESEs [25] . As expected from the above results , genes with a high proportion of sequence within , for example , 70 bp of an intron–exon junction showed lower KA ( Table 2; Figure 5 ) . Using alternative bounds ( 50 or 100 bp ) did not qualitatively affect conclusions ( Table 2 ) . The difference between a gene with all sequence within 70 bp of exon boundaries and one with very little ( <10% ) was striking ( mean KA = 0 . 032 for those with small exons and 0 . 083 for those with less than 10% of sequence near junctions ) . Were all things equal , this result suggests that the rate of evolution of an intron-rich gene is on average approximately under 40% that of an intron-poor gene . It is , however , unlikely that all things are equal . To allow for this , we performed a paired test . For each gene we concatenated all sequences in the alignment flanking ( within 72 nucleotides ) intron–exon boundaries , both 5′ and 3′ , and concatenated all of the middle sections of exons ( defined as anything beyond 72 nucleotides ) . As before we considered only internal exons . We then calculated KA for the concatenated flanks and the concatenated middle sections and considered the gene-specific ratio of the two . We then considered the mean of the gene-specific ratio for all genes . By necessity we had to eliminate all genes with no exon larger than 144 bp , leaving 3 , 058 genes . Moreover , as accurate estimation of KA probably requires a minimum of 100 codons , we restricted analysis to those genes with at least 300 bp in the concatenated flanks and in the concatenated middle of exons . We found that the mean ratio of the rate of evolution ( KA ) of the middle part of exons to the flanks within the same gene was 1 . 93 ( Wilcoxon signed rank test , p < 0 . 0001 , n = 666 ) . Requiring at least 600 bp in both flanks and middle sections , the middle was estimated to evolve 2 . 3 times faster than the flanks . When we considered the exon flanks to be 102 bp , the mean ratio of middle to flank was 2 . 5 when requiring a minimum of 300 bp in each class ( n = 368 ) . Requiring a minimum of 600 bp , the middle parts of exons evolved on average 2 . 7 times faster than the exon flanks from the same genes ( n = 167 ) . Overall , then , it seems safe to conclude that exon centres evolve at about 2 . 3 times the rate of exon flanks from the same gene , the precise estimate depending on parameter choices . These results demonstrate that exon flanks evolve more slowly than exon centres , regardless of the functional class of the protein . The mean KA of flanking domains was around 0 . 04 in the above samples . A gene with short exons should then have approximately a KA of 0 . 04 , controlling for between-gene heterogeneity . By contrast one with 90% of sequence not near boundaries should have a KA of on average around 0 . 086 , assuming exon centres of such long exons evolve 2 . 3 times faster than flanks ( 0 . 04 × 2 . 3 × 0 . 9 + 0 . 04 × 0 . 1 = 0 . 086 ) . Controlling then for functional class , we estimated that a gene with all sequence near intron–exon boundaries should evolve at about 46% ( 0 . 04/0 . 086 ) the rate of one with proportionally little sequence near boundaries . This estimate can be downwardly adjusted if we consider that some of the genes with long exons have more than 90% of sequence near boundaries: at the limit intronless genes should evolve with KA ≅ 0 . 092 , i . e . , at 2 . 3 times the rate of small exon genes . Likewise , if our estimate of the ratio of rates of evolution is higher , then the discrepancy between intron-poor and intron-rich genes will be greater . Using the 2 . 7 ratio , for example , intron-rich genes evolve at 37% of the rate of intronless genes , controlling for protein function . Equally , the estimate can be upwardly adjusted if we presume a more modest ratio of rates of evolution of internal parts of exons to flanks . Overall , it seems fair to suppose that constraints imposed in the proximity of intron–exon boundaries can reduce the rate of evolution of a gene by a half or more , if the gene is full of small exons rather than lacking introns . That this is similar to the prior estimate , not controlling for between-gene heterogeneity , suggests that selection on exon flanks is a major determinant of rates of evolution . How does the effect of selection in the vicinity of intron–exon junctions compare with and covary with other strong predictors of rates of protein evolution ? In principle any relationship between rate of protein evolution and proportion of sequence near a boundary might in part be because genes with many introns tend to be housekeeping genes [26] , and housekeeping genes ( those expressed in many tissues ) tend to have low rates of evolution [4 , 27 , 28] . The two parameters ( expression breadth and proportion of sequence near boundaries ) both appear , however , to be good predictors when controlling one for the other ( Table 2 ) . Use of alternative metrics of gene expression ( mean rate and peak rate ) ( see Table 2 ) make no qualitative difference to the conclusion that , before and after control for covariates , the proportion of sequence near intron–exon junctions is at least as strong a predictor of rates of evolution as expression parameters , if not stronger . After expression parameters , the dispensability of a protein may , in mammals , also be a good predictor [12] . From a sample of 1 , 198 mouse genes for which knockout experiments have resolved whether they are essential or not , and for which we have orthologues , we can ask whether essential and nonessential genes ( a ) differ in their proportion of sequence near intron–exon junctions and ( b ) differ in their rate of evolution . Confirming the prior report [12] , we found that essential proteins evolve at about two-thirds the rate of nonessential ones ( mean KA for nonessential proteins , 0 . 07; for essential proteins , 0 . 049; p < 0 . 0001 , Mann-Whitney U test ) . However , the two classes are no different as regards the proportion of sequence near intron–exon boundaries ( mean proportion of sequence near boundaries for nonessential proteins , 0 . 618; for essential proteins , 0 . 607; p = 0 . 67 , Mann-Whitney U test ) . There is , therefore , no reason to suppose that the lower rate of evolution of genes with much sequence near intron–exon boundaries is owing to their being more likely to be essential . Equally , there is no reason to suppose that the lower rate of evolution of essential genes is owing to their having more sequence near intron–exon boundaries . Note too that the difference in evolutionary rate between essentials and nonessentials is more modest than that between genes with high and low proportion of sequence near intron–exon junctions . The majority of our sample is of unknown dispensability . These genes have a mean KA of 0 . 059 , more or less as expected , given the means for the essential and nonessential genes and assuming that 30% of mouse genes are essential [12] . Let us now consider two models for what might happen after a new intron has been inserted . In the first , a new intron might be favoured only if enough splice-enhancer domains in adequate proximity are already present to enable efficient removal of the intron ( model 1 ) . An alternative model ( model 2 ) might suppose that immediately after introduction of a new intron , proper excision , owing to a dearth of local splice enhancers , is not always possible . If , then , some transcripts preserve the original mRNA by proper excision , but others fail to so do , the new intron would effectively reduce the rate of protein production for a given transcription rate . Such a mutation might be weakly deleterious such that fixation through drift is still possible . Selection may then favour shifts in amino acid usage to enable more efficient splicing . The second model is especially interesting as it suggests that intra-protein amino acid usage is not dictated simply by protein requirements alone . Both models predict that should enhancer domains be employed , they may then be under selection to preserve functionality . Both also predict that amino acids that feature commonly in the hexameric sequences describing splice enhancers should be more common near intron–exon junctions , as observed . How they differ is in the prediction of subsequent evolution following gain/loss of introns . Model 1 supposes that if an intron inserts but is not successfully removed owing to a dearth of splice-enhancer domains in the near vicinity , the insertion may simply be too deleterious to be tolerated and is hence lost from the population . By contrast , model 2 considers the possibility that compensatory nonsynonymous changes can further occur that permit more efficient intron removal . To discriminate these two classes , one needs a sufficiently sized dataset of intron losses or gains in humans . Unfortunately , intron gain appears to be vanishingly rare in humans and mammals more generally . However , functional retroposed genes do provide a means to ask about the consequences of intron loss . Is it then the case that , after retroposition , the residues that , in the original parental copy of the gene , flanked intron–exon junctions are more prone to change ? We examined a set of 49 old functional retroposed genes for which , in all cases , there existed mouse and human parent and retroposed sequences . For all sites in the alignment that specified an amino acid in all four lineages , we considered the proportion of retrogene-specific changes ( see Materials and Methods ) . We then considered how this varied as a function of the distance from what was , in the parental gene , the intron–exon boundary . Merging figures for 3′ and 5′ ends , we found that the rate of evolution in retrogenes is higher close to what was the boundary ( Spearman rank correlation , proportion of sites subject to change in retrogenes versus distance from ancient boundary , rho = −0 . 48 , p = 0 . 019 ) ( Figure 3 ) . Moreover , retrogenes that are derived from genes in which a high proportion of the sequence was near exon boundaries ( genes with predominantly small exons ) tended to have higher overall rates of evolution ( proportion of parent sequence 70 bp from boundary versus number of retrogene-specific changes per base pair , rho = +0 . 38 , p = 0 . 008 , n = 49 ) . The difference in behaviour between genes that have lost their introns and intron-containing genes ( Figure 3 ) suggests that constraints that exist near intron–exon boundaries have been released in the retrogenes , and , hence , that these sites are now free to change . This evidence , therefore , lends some support to the converse possibility , namely that , after intron insertion , exonic domains flanking the new boundary changed , probably to permit better splicing . The result does not specifically show that all the change involved the evolution of new splice enhancers; however , with the data showing that the HPI predicts trends in amino acid usage near junctions and low nonsynonymous rates in ESEs ( Figure 4 ) [24] , this is likely to explain much of the effect . We have found that , in both mouse and human , most amino acids show skewed usage in the vicinity of intron–exon junctions . These patterns appear owing to preference at the nucleotide level , as evidenced by the different behaviours of the 2-fold and 4-fold blocks of leucine and arginine . To a first approximation , the patterns are well explained by the abundance of the relevant codons , relative to levels in the genome , in splice enhancers . The preferences are also reflected in reduced rates of evolution near intron–exon boundaries and in intron-rich genes more generally . Indeed , the proportion of sequence near intron–exon boundaries is , to the best of our knowledge , one of the strongest predictors to date of rates of protein evolution ( for analysis of alternatives see [12] ) . That in retrogenes the domains that used to be near intron–exon junctions show increased rates of evolution supports the view that intron–exon junctions are domains on which constraint operates . Were it the case that new introns are only tolerated if the full repertoire of splice-control elements is already in place , we would not expect that , on loss of introns , these domains would show unusually high rates of evolution . Although by necessity our sample size of retrogenes is small , we suggest that model 2 , evoking evolution to modify amino acid content after intron insertion , is more parsimonious . Whether the elements being preferred are necessarily and exclusively splice enhancers remains uncertain . First , as can be seen in Figure 4 , sequence putatively not in enhancers is more highly constrained near boundaries , at least at the 3′ end . This suggests the possibility of constraint imposed near boundaries independent of splice enhancers and/or inaccuracy in the definition of enhancers . Further , there are a few strong outliers in the distribution of HPI versus preference near boundaries ( Table 1 ) . In human sequences , of 46 comparisons , 14 fail to match with the expectation that if HPI is negative , rho should be positive and vice versa , of which nine are significant and six significant after Bonferroni correction: I5′ , l5′ , Q5′ , F3′ , I3′ , and l3′ ( Table 1 ) . Glutamine ( CAA and CAG ) is unique in being preferred in splice enhancers and avoided both 3′ and 5′ at boundaries . Three amino acids are strongly preferred near boundaries ( rho << 0 ) but disfavoured in splice enhancers ( HPI < 0 ) , these being the 2-fold degenerate codons of leucine ( TTA and TTG ) , isoleucine ( ATC , ATA , and ATT ) , and phenylalanine ( TTC and TTT ) . Tyrosine ( TAC and TAT ) may be a weaker outlier ( rho < 0 both 5′ and 3′ , HPI < 0 ) . The same outliers are seen in mouse genes ( Table 1 ) . Are these apparent exceptions instructive of some other force driving amino acid choice near boundaries , or might they reflect limitations in our understanding of splice-enhancer hexamers ? Were the latter the case we might expect that a surrogate measure of involvement in splice enhancers might reveal these exceptions to simply have poorly described roles in splice enhancers . As noted above adenine and cytosine content of the synonymous codon blocks of each amino acid well predicts HPI ( Figure 2 ) . Fitting the best-fit regression of AC bias to rho ( using both 5′ and 3′ data ) , we indeed find from inspection of the standardised residuals ( Figure S4 ) that , both 3′ and 5′ , isoleucine and leucine usage now sit within the 95% confidence intervals , as does phenylalanine usage 5′ . However , phenylalanine usage 3′ is a little outside the line , as is glutamine 5′ usage . Another possibility is that the presence of exonic splice suppressors may impact amino acid usage . Wang et al . [29] have identified 131 decamers that function as splice suppressors . We therefore adapted our method to calculate a decamer preference index ( DPI ) to correspond with these splice suppressors ( Table 1 ) . DPI and HPI are not themselves correlated ( for mouse–human set for HPI , Spearman rank correlation between HPI and DPI = −0 . 05 , p = 0 . 7 ) . Relating DPI scores to either the slope or the rho values for amino acid preference , we find only a marginal tendency for DPI to explain amino acid preferences ( Spearman rank correlation , rho versus DPI , −0 . 27 , p = 0 . 07; slope versus DPI , −0 . 26 , p = 0 . 07 ) . Splice suppressors hence appear to have less impact on amino acid usage than do splice enhancers . Taking a combined measure , the mean of DPI and HPI , marginally improves the fit between amino acid preference and involvement in splice regulation ( Spearman rank correlation between mean of DPI and HPI and rho , −0 . 61 , p < 0 . 0001; for HPI alone , −0 . 54 , p < 0 . 0001 ) . AC bias remains a better predictor . Involvement in splice suppressors may , however , explain some of our apparent exceptions . Notably , phenyalanine and the 2-fold block of leucine , while having a negative HPI , have a strongly positive DPI ( 9 . 8 and 14 . 9 , respectively ) . Similarly , glutamine , while having a positive HPI , has a strongly negative DPI ( −6 . 1 ) . The converse roles of these amino acids in splice enhancers and splice suppressors may hence explain their apparently aberrant behaviour . Indeed , on a plot of the mean of DPI and HPI these amino acids no longer appear as outliers ( Figure S5 ) . Isoleucine remains an exception , being negative for both HPI and DPI but preferred near boundaries . The only other model for selection near intron–exon junctions , the so-called cryptic splice-site avoidance model [21 , 30] , does not predict any tendency for cytosine avoidance near boundaries . The relevance of this model is unclear as both AG[A\|G] ( arginine ) and AG[C\|T] ( serine ) appear to have patterns of usage near boundaries at both 5′ and 3′ ends as expected given their HPI scores , whereas the cryptic splice-site avoidance model would predict avoidance at 5′ ends . This model cannot also obviously explain why 3′ usage of phenyalanine might be discordant . One further striking peculiarity is notable . The profile of usage of glycine ( GGN ) shows a curious pattern at both 3′ and 5′ ends ( also seen in mouse , data not shown ) ( Figure 6 ) : at every third codon the usage is much higher than at the intervening distances from the boundary . With the sample sizes in question ( ∼10 , 000 glycines at these positions ) , this is not a sample-size artefact . The effect is highly repeatable , being found regardless of the phase of the exon ( Figure S6 ) . At both the 3′ and 5′ ends , it is found for all of the four ( GGN ) codons when analysed separately , although it may be most pronounced for GGA ( data not shown ) . This appears to reflect a pattern at the protein level , at least in part owing to collagens , whereby glycines are very commonly three apart ( see Figure S7 ) . Given that introns tend to prefer G\|G insertion sites , codons starting GG may well be hot spots for insertion , potentially at all positions . This together with the apparent periodicity in the occurrence of glycine might explain the observations . We leave this to future analysis . Whatever the cause , it points to a limitation of our method , which assumes that trends towards boundaries are monotonic and consistent . For the most part ( see Figure S1 ) these assumptions appear relatively sound , although 5′ usage of proline suggests a U-shaped function . The hypothesis that the domains under constraint are uniquely splice enhancers might also predict that amino acids not having a role in splice enhancers tend to be gained in retrogenes in boundary proximal domains . Unfortunately , from a sample of 803 gains/losses for retrogenes and 229 in parental genes in regions near intron–exon junctions ( <30 codons ) , we find no amino acid showing statistically significant differences between parental and retrogenes . However , the top three most discordant amino acids ( judged by the chi-squared value ) all show net gain in the retrogenes and net loss in the parental genes , and , as might be predicted , are all avoided in splice enhancers . These are the 4-fold block of leucine ( 49 gains to 39 losses in retrogenes; nine losses to 18 gains in parental genes; chi-squared = 4 . 13 ) , histidine ( 20 gains to 13 losses in retrogenes; three gains to seven losses in parental genes; chi-squared = 2 . 89 ) , and the 4-fold block of serine ( 48 gains to 29 losses in retrogenes; 14 gains to 17 losses in parental genes , chi-squared = 2 . 66; N . B . , for three degrees of freedom p < 0 . 05 occurs at chi-squared > 7 ) . It would be unwise to read too much into this observation , not least because there are several other amino acids with strong avoidance in splice enhancers that show no evidence of switching substitutional profile ( notably alanine , cysteine , phenylalanine , and valine ) . No amino acids show any good evidence for being gainers in the parental gene but losers in the retrogene . Firmer conclusions regarding the patterns of amino acid loss and gain will require larger sample sizes . Given the outliers being possibly explained by splice-suppressor roles and the strange behaviour of glycine , we do not wish to suggest that the need for splice enhancers determines all amino acid bias , nor all constraint , seen near intron–exon boundaries . Constraints operating near intron–exon boundaries not explained by splice enhancers may nonetheless reflect selection on splice regulation of some form ( e . g . , exonic splice suppressors ) . These caveats aside , it is notable that constraints in the vicinity of intron–exon boundaries appear to be one of the stronger , if not the strongest , predictors of rates of protein evolution in mammals . Naturally , for intron-poor genomes the same will not apply . We established a dataset of 178 , 382 human exons derived from the RefSeq track at the University of California Santa Cruz genome browser ( http://genome . cse . ucsc . edu/cgi-bin/hgTables ) , March 2006 release . We obtained a set of 21 , 990 RefSeq files with the exon structure of the CDS specified . All files were checked to ensure that the coding sequence started with ATG , finished with a stop codon , had no internal stop codons , had no codons of uncertain translation , and was a multiple of three . This resolved to a dataset of 19 , 384 RefSeq files . We eliminated all first and last exons , leaving a sample of 178 , 382 exons . We trimmed all exons so that the first base was the first base of the first complete codon , and the last base the last of the final complete codon . As , to ensure correct splicing , first and last codons are by necessity highly skewed in usage , these too were eliminated . For each codon and in turn each amino acid , we considered proportional usage of that amino acid at a given distance from the junction both 3′ and 5′ . All exons were divided in two , so a given codon never featured in both 3′ and 5′ calculations . This sample was not purged for duplicates . However , we repeated the analysis on a more stringently defined set of over 2 , 000 genes and 14 , 000 exons , previously purged for duplicates [21] . We confirmed that all qualitative trends are identical ( data not shown ) . We then considered the trend in usage of each amino acid as a function of the distance from the boundary . This we did by calculating Spearman rank correlations ( rho ) between the distance from the boundary ( 5′ or 3′ ) and proportional usage of the amino acid ( i . e . , in proportion to the number of residues at that given distance ) . Note that a negative rho implies an amino acid that is preferred near boundaries , and a positive rho implies a tendency to be avoided . To simplify numbering on the plots , we refer to amino acid positions by reference to the number of full codons between the given position and the relevant end of the trimmed exon . We split the three 6-fold degenerate amino acids into a block of four and a block of two . The block of two is specified by the usage of the lowercase letter ( i . e . , “S” implies TCA , TCC , TCG , and TCT , while “s” implies AGC and AGT ) . In relevant circumstances , the 2-fold and 4-fold blocks were treated as separate amino acids . Changes between the 2- and 4-fold blocks were not , however , treated as nonsynonymous changes . As with the derivation of the human exon set , we obtained a set of mouse exons via the RefSeq track at the University of California Santa Cruz genome browser . For analysis of trends in amino acid preference near junctions , these exons were handled as described above . For analysis of orthologous exons , we obtained the human–mouse orthologue list from Mouse Genome Informatics ( ftp://ftp . informatics . jax . org/pub/reports/index . html ) . We identified all pairs for which both mouse and human sequence had a RefSeq entry . As before , we eliminated all full coding sequences that were not well translated ( more than one stop , ambiguous codons , etc . ) . We further eliminated those in which the number of exons differed between the orthologues . We then compared the phases of the putatively orthologous exons . Gene pairs in which any orthologous exon did not have the same phase in mouse and human were eliminated , leaving 7 , 767 genes . Any genes in which any orthologous exon differed by more than 5% in size were also eliminated , leaving 5 , 057 genes . First and last exons were removed , and all remaining orthologous exons were trimmed to start at the first full codon and end at the end of the last complete codon . They were then aligned at the peptide level using muscle v3 . 6 [31] . This left 36 , 683 aligned orthologous internal exons . Burge and colleagues have characterised numerous hexameric sequences that function as splice enhancers [22 , 25 , 32 , 33] . For each hexamer we can then define a series of full codons that could potentially be present in the hexamer . If we consider a series of six nucleotides , n1n2n3n4n5n6 , then codons n1n2n3 , n2n3n4 , n3n4n5 , and n4n5n6 are specified in their entirety . We sum all such possible codons for all specified splice-enhancer hexamers . This provides a measure of ESE hexameric involvement of all possible codons , within any given hexamer dataset . The three stop codons were removed , and the proportions renormalised . To provide a metric of involvement of an amino acid in ESEs , we compared rates of involvement of codons in the hexamers with those in the genome as a whole . To this end , we normalised ( after stop codon removal ) the relative abundances of all codons as specified in the appropriate codon usage database ( http://www . kazusa . or . jp/codon ) . We then generated 10 , 000 sets of random hexamers , each set being the same size as the input hexamer list . Hexamers were generated by joining two codons selected at random in proportion to their frequency in the appropriate genome . We parsed each random hexamer in the same manner as we parsed the input list , extracting all non-stop codons . For each amino acid , given the frequencies of the relevant synonymous codons , we then determined the mean and standard deviation in relative abundance in the 10 , 000 random sets . The difference between the observed frequency of an amino acid in the real hexamer set and in the randomised sets , normalised by the standard deviation in the randomised sets , then is our HPI ( i . e . , a Z score ) . A high HPI value indicates that a given amino acid is enriched in ESEs compared with what is expected given its content in the genome , and given the underlying variance expected based on the number of hexamers used as input . Source code to calculate HPI is freely available from L . D . H . In principle , the HPI score for an amino acid will change as a function of both input codon frequencies and with the input set of known ESE hexamers . In practice , we find that employing mouse rather than human codon frequencies makes little or no difference ( data not shown ) . In this analysis we thus employed human codon frequencies to assemble random hexamers . As regards the input list for hexamers , we considered three sets: two sets specific to human 5′ and 3′ exonic ends ( 95 5′ enhancers and 177 3′ enhancers ) and a set of 175 hexamers found both in mouse and human at either exonic end . We found that scores for 5′ and 3′ ends were very similar to each other . Unless otherwise stated , we employed the mouse–human conserved set . Use of this latter set is advantageous as it is most probably enriched for strong enhancers . The splice-enhancing hexamers in all datasets have two striking features , notably an abundance of adenine and a dearth of cytosine , relative to their usage in the human genome . In the human genome , cytosine constitutes 26 . 0% of all nucleotides in coding sequences ( derived from table of codon usage as noted above ) but only 12 . 5% in splice enhancers , while adenine is 25 . 6% of all nucleotides in coding sequences but is 49 . 0% of the nucleotides in splice enhancers . Guanine is used in approximately the same amount in hexamers and in the genome ( 26 . 4% in genome and 25 . 7% in hexamers ) . Thymine is , like cytosine , underused in hexamers ( 12 . 4% ) , but its usage in the genome is just 22 . 0% , so its relative reduction in hexamers is less dramatic than that of cytosine . As expected , amino acids with few cytosine nucleotides in their codon set and many adenine residues tend to have positive HPI values ( Spearman rank correlation , HPI versus cytosine content of codons , rho = −0 . 63 , p = 0 . 0012; HPI versus adenine content of codons , rho = +0 . 71 , p = 0 . 0002 , n = 23 ) . A composite measure of adenine and cytosine bias of codons ( frequency of adenine in synonymous codon set minus frequency of cytosine ) is a good predictor of HPI ( Spearman rank correlation = 0 . 85 , p < 0 . 0001 , n = 23 ) . For the DPI pertinent to splice suppressors we extracted the 131 decamers provided by Wang et al . [29] from http://www . cell . com/cgi/content/full/119/6/831/DC1 . The protocol to define DPI scores was identical to that to calculate HPI , except that random decamers were made by random selection of four codons and trimming off of the final two bases . The eight full codons in the decamers were employed to define expected frequencies . Mouse retroposed gene copies were identified using the procedure described in Vinckenbosch et al . [34] . For humans , we used a previously established retrocopy dataset [34] . To identify orthologous retrocopies shared between humans and mouse , we used human–mouse chained alignments available from the University of California Santa Cruz ( hg17 versus Mm6 ) . Similar to our previous procedure [34] , we first extracted the best alignments that overlapped with the genomic location of human retrocopies and that were >15 kb ( this length ensures that the alignment also covers surrounding , nonretrocopy-derived sequences in the two species ) . If no such alignments could be identified , presence/absence in mouse was not determined . We then scanned the chained alignments for an aligned block ( putative orthologous sequence in the chain ) that overlapped with the human retrocopy . If such a block was found , its corresponding mouse coordinates were compared to the mouse retrocopy set . Mouse retrocopies overlapping with these coordinates were considered orthologues of human retrocopies . In total , we identified 56 orthologous retrocopy pairs , of which 49 showed intact open reading frames in both species . The fact that these retrocopies emerged in the common ancestor of humans and mice ( at least approximately 75–90 million years ago ) and possess intact open reading frames strongly suggests that they have been selectively preserved by natural selection . Thus , they likely represent functional retroposed gene copies ( retrogenes ) . Functionality of these human–mouse retrocopies is further supported by their generally higher transcription levels and lower KA/KS values relative to younger , lineage-specific retrocopies [34] . To infer retrogene-specific changes , the sets of four sequences were aligned at the protein level using Muscle [31] . The sequences were then cut into individual exons by reference to the human annotation of parental genes . Exons were trimmed so as to contain only complete codons . The 5′ end of the first exon and the 3′ end of the last exon were ignored . All sites in the amino acid alignment that specified the same amino acid in three of the four sequences but a different amino acid in the third were considered , by parsimony , to be informative . That is , if the two human sequences specify amino acid X , as does the mouse parent gene at a given position , while the mouse retrogene is amino acid Y , then an X→Y change is inferred to have occurred in the mouse retrogene . The total number of retrogene changes is simply the sum of those in the mouse and those in the human retrogene , employing this strict 3:1 criterion . Gene expression estimates were obtained from Su and colleagues [35] , employing the March 2006 annotation ( http://wombat . gnf . org/index . html ) . Mas5 files with Affymetrix present/absent calls were used . Human gene expression data were obtained by merging U133A and GNF1h chip datasets . In both mouse and human , average expression was obtained from samples of the same tissues . Probes matching to more than one gene were eliminated from further analyses . Indexes of gene activity were obtained only from samples obtained from normal adult tissues . Levels and breadth of expression were calculated . Three indexes for expression levels were obtained: peak , average , and median expression . The peak level was the highest score across all analysed tissues . Breadth of expression was calculated from present/absent calls . For the analysis of mean/median levels , for each gene we considered only those tissues in which a gene was expressed ( judged by present/absent call ) . When multiple probes matched the same gene we considered a gene to be expressed in a given tissue if half or more of the probes indicated presence .	Most of the DNA in our genes is actually not involved in the specification of proteins . Rather , the bits with the protein-coding information ( exons ) are separated from each other by noncoding bits , introns . Before a gene can be translated into protein these introns are removed and the exons are spliced back together to be translated into protein . While information about which DNA to remove is largely in the introns themselves , parts of the exons near the intron–exon boundary can , for example , function as splice enhancer elements . In principle , then , these parts of exons have two functions: to specify the amino acids of the resulting protein and to enable the correct removal of introns . What impact might this have on a gene's evolution ? We show that near intron–exon boundaries , amino acid usage is biased towards nucleotides involved in splice control . Moreover , these parts of genes evolve especially slowly . Indeed , we estimate that a gene with many exons would evolve at under half the rate of the same gene with no introns , simply owing to the need to specify where to remove introns . Likewise , genes that have lost their introns evolve especially fast near the former intron's location . Thus , human proteins may not be as optimised as they could be , as their sequence is serving two conflicting roles .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "none", "computational", "biology", "evolutionary", "biology", "genetics", "and", "genomics" ]	2007	Splicing and the Evolution of Proteins in Mammals
Long-term memories are likely stored in the synaptic weights of neuronal networks in the brain . The storage capacity of such networks depends on the degree of plasticity of their synapses . Highly plastic synapses allow for strong memories , but these are quickly overwritten . On the other hand , less labile synapses result in long-lasting but weak memories . Here we show that the trade-off between memory strength and memory lifetime can be overcome by partitioning the memory system into multiple regions characterized by different levels of synaptic plasticity and transferring memory information from the more to less plastic region . The improvement in memory lifetime is proportional to the number of memory regions , and the initial memory strength can be orders of magnitude larger than in a non-partitioned memory system . This model provides a fundamental computational reason for memory consolidation processes at the systems level . Memories are stored and retained through a series of complex , highly coupled processes that operate on different timescales . In particular , it is widely believed that after the initial encoding of a sensory-motor experience , a series of molecular , cellular , and system-level alterations lead to the stabilization of an initial memory representation ( memory consolidation ) . Some of these alterations occur at the level of local synapses , while others involve the reorganization and consolidation of different types of memories in different brain areas . Studies of patient HM revealed that medial temporal lobe lesions severely impair the ability to consolidate new memories , whereas temporally remote memories remain intact [1] . These results and more recent work ( see e . g . [2] ) suggest that there may be distinct memory systems , and that memories , or some of their components , are temporarily stored in the medial temporal lobe and then transferred to other areas of the cortex . Is there any fundamental computational reason for transferring memories from one area to another ? Here we consider memory models consisting of several stages , with each stage representing a region of cortex characterized by a particular level of synaptic plasticity . Memories are continuously transferred from regions with more labile synapses to regions with reduced but longer-lasting synaptic modifications . Here we refer to each region as a stage in the memory transfer process . We find that such a multi-stage memory model significantly outperforms single-stage models , both in terms of the memory lifetimes and the strength of the stored memory . In particular , memory lifetimes are extended by a factor that is proportional to the number of memory stages . In a memory system that is continually receiving and storing new information , synaptic strengths representing old memories must be protected from being overwritten during the storage of new information . Failure to provide such protection results in memory lifetimes that are catastrophically low , scaling only logarithmically with the number of synapses [3]–[5] . On the other hand , protecting old memories too rigidly causes memory traces of new information to be extremely weak , being represented by a small number of synapses . This is one of the aspects of the classic plasticity-rigidity dilemma ( see also [6]–[8] ) . Synapses that are highly plastic are good at storing new memories but poor at retaining old ones . Less plastic synapses are good at preserving memories , but poor at storing new ones . A possible solution to this dilemma is to introduce complexity into synaptic modification in the form of metaplasticity , by which the degree of plasticity at a single synapse changes depending on the history of previous synaptic modifications . Such complex synapses are endowed with mechanisms operating on many timescales , leading to a power-law decay of the memory traces , as is widely observed in experiments on forgetting [9] , [10] . Furthermore , complex synapses can vastly outperform previous models due to an efficient interaction between these mechanisms [11] . We now show that allowing for a diversity of timescales can also greatly enhance memory performance at the systems level , even if individual synapses themselves are not complex . We do this by considering memory systems that are partitioned into different regions , the stages mentioned above , characterized by different degrees of synaptic plasticity . In other words , we extend the previous idea of considering multiple timescales at single synapses to multiple timescales of plasticity across different cortical areas . To determine how best to partition such a memory system , we take the point of view of an engineer who is given a large population of synapses , each characterized by a specific degree of plasticity . Because we want to focus on mechanisms of memory consolidation at the systems level , we use a simple binary model in which synaptic efficacies take two possible values , weak and strong . Previous work has shown that binary synapses are representative of a much wider class of more realistic synaptic models [5] . It seems likely that the mechanisms for storing new memories exploit structural aspects and similarities with previously stored information ( see e . g . semantic memories ) . In our work , we are interested in different mechanisms responsible for storing new information that has already been preprocessed in this way and is thus incompressible . For this reason , we restrict consideration to memories that are unstructured ( random ) and do not have any correlation with previously stored information ( uncorrelated ) . After constructing multi-state models , we estimate and compare their memory performance both in terms of memory lifetime and the overall strength of their memory traces . We first analyzed a homogeneous model ( single partition ) , in which all the synapses have the same learning rate ( see Fig . 1 ) . We consider a situation in which new uncorrelated memories are stored at a constant rate . Synapses are assumed to be stable in the absence of any overwriting due to the learning of new memories . Each memory is stored by modifying a randomly selected subset of synapses . As the synapses are assumed to be bistable , we reduce all the complex processes leading to long term modifications to the probability that a synapse makes a transition to a different state . As memories are random and uncorrelated , the synaptic transitions induced by different memories will be stochastic and independent . To track a particular memory we take the point of view of an ideal observer who has access to the values of the strengths of all the synapses relevant to a particular memory trace ( see also [11] ) . Of course in the brain the readout is implemented by complex neural circuitry , and the estimates of the strength of the memory trace based on the ideal observer approach provide us with an upper bound of the memory performance . However , given the remarkable memory capacity of biological systems , it is not unreasonable to assume that specialized circuits exist which can perform a nearly optimal readout , and we will describe later a neural circuit that replicates the performance of an ideal observer . More quantitatively , to track a memory , we observe the state of an ensemble of synapses and calculate the memory signal , defined as the correlation between the state of the ensemble at a time and the pattern of synaptic modifications induced by the event of interest at time . Specifically , we can formalize this model description by assigning the value to a potentiated synapse and to a depressed one . Similarly , a plasticity event is assigned a value if it is potentiating and if depressing . We then define a vector of length , where is the state of synapse at time . Similarly , the memories are also vectors of length , , where is the plasticity event to which synapse is subjected at time . If we choose to track the memory presented at time , then we define the memory trace as the signal at time , which is just the dot product of two vectors , . The signal itself is a stochastic variable , since the updating of the synaptic states is stochastic . This means that if one runs several simulations presenting exactly the same memories , the signal will be different each time , see right hand side of Fig . 1a . The mean signal , understood as the signal averaged over many realizations of the Markov process , can be computed analytically . For the homogeneous model , a continuous-time approximation to the mean signal takes the simple form of an exponential , , where is the total number of synapses and is the learning rate , see Methods and Text S1 for details . We must compare this mean signal to the size of fluctuations in the model , i . e the noise . The memory noise is given by the size of fluctuations in the overlap between uncorrelated patterns , which here is approximately , see Text S1 for details . Therefore , the signal-to-noise ratio . One can track a particular memory only until it has grown so weak it cannot be discerned from any other random memory . Memory lifetime , which is one measure of the memory performance , is then simply defined as the maximum time over which a memory can be detected . More quantitatively it is the maximum time over which the SNR is larger than 1 . The scaling properties of the memory performance that we will derive do not depend on the specific critical SNR value that is chosen . Moreover , it is known that the scaling properties derived from the SNR are conserved in more realistic models of memory storage and memory retrieval with integrate-and-fire neurons and spike driven synaptic dynamics ( see e . g . [12] ) . As we mentioned , the dynamics of the Markov model we consider are stochastic . Therefore , throughout the paper , we will discuss results from stochastic models for which we have derived corresponding mean-field descriptions . Fig . 1b shows the mean-field result for two extreme cases when all synapses have the same degree of plasticity . If the synapses are fast and the transition probability is high ( ) , then the memory is very vivid immediately after it is stored and the amount of information stored per memory is large , as indicated by the large initial SNR ( ) . However the memory is quickly overwritten as new memories are stored . In particular , the memory lifetime scales as which is extremely inefficient: doubling the lifetime requires squaring the number of synapses . It is possible to extend lifetimes by reducing the learning rate , and in particular by letting the learning rate scale with the number of synapses . For the smallest that still allows one to store sufficient information per memory ( i . e . that allows for an initial SNR∼1 ) , , the memory lifetimes are extended by a factor that is proportional to . This trade-off between memory lifetime and initial SNR ( i . e . the amount of information stored per memory ) cannot be circumvented through the addition of a large number of synaptic states without fine-tuning the balance of potentiation and depression [5] . These shortcomings can be partially overcome by allowing for heterogeneity in the transition probabilities within an ensemble of synapses . Specifically , if there are equally sized groups of synapses , each with a different transition probability ( ) , then the most plastic ones will provide a strong initial SNR while the least plastic ones will ensure long lifetimes . Intermediate time-scales are needed to bridge the gap between the extreme values . In Fig . 1c we plot the SNR as a function of time . Transition probabilities are taken to be of the form , where is the fastest learning rate , is the slowest learning rate and . Time is expressed in terms of the number of uncorrelated memories on the lower axis , and we choose an arbitrary rate of new uncorrelated memories ( one per hour ) to give an idea of the different orders of magnitudes of the timescales that are at play ( from hours to years ) . This model , which we call the heterogeneous model is already an interesting compromise in terms of memory performance: as we increase the number of synapses , if the slowest learning rate is scaled as , then both the initial SNR and the memory lifetime scale advantageously with the number of synapses ( ) . Moreover , the model has the desirable property that the memory decay is a power law over a wide range of timescales , as observed in several experiments on forgetting [13] . In the heterogeneous model , the synapses operate on different timescales independently from each other . We now show that the performance can be significantly improved by introducing a feed-forward structure of interactions from the most plastic group to the least plastic group of synapses . How is this possible ? While the least plastic synapses can retain memories for long times , their memory trace is weak . However , this memory trace can be boosted through periodic rewriting of already-stored memories . If a memory is still present in one of the groups of synapses ( called hereafter a ‘memory stage’ ) , the stored information can be used to rewrite the memory in the downstream stages , even long after the occurrence of the event that created the memory . It is important to notice that not all types of rewriting can significantly improve all the aspects of the memory performance . For example , if all memories are simply reactivated the same number of times , then the overall learning rate changes , so that the initial memory trace becomes stronger , but the memory lifetimes are reduced by the same factor . Rather , an alternative strategy is to reactivate and rewrite a combination of multiple memories , one which has a stronger correlation with recent memories and a weaker correlation with the remote ones . We have built a model , which we will call the memory transfer model , that implements this idea . We consider synapses divided into interacting stages . We assume that all the stages have the same size and that synapse in stage can influence a counterpart synapse in stage . In particular , synapses in the first stage undergo stochastic event-driven transitions as before ( Fig . 2a ) . They therefore encode each new memory as it is presented . On the other hand , synapses in downstream stages update their state stochastically after each memory is encoded in the first stage . Specifically , at time , a memory of length consisting of a random pattern of potentiating ( ) and depressing ( ) events is presented to the synapses in stage one , which have synaptic state . Synapse is subjected either to a potentiating ( ) or to a depressing ( ) event with probability 1/2 , and is updated with a probability as in the previous models . Therefore , the updating for synapses in stage 1 is identical to that for ensemble 1 in the synaptic model with heterogeneous transition probabilities which we discussed previously . Now , however , we assume that a synapse in stage 2 is influenced by the state of synapse in stage 1 in the following way . If synapse in stage 1 is in a potentiated ( depressed ) state at time ( or respectively ) , then synapse in stage 2 will potentiate ( depress ) at time with probability . The update rule for synapses in stage 3 proceeds analogously , but depends now on the state of synapses in stage 2 , and so on . In other words , after each memory is stored , a random portion of the synaptic matrix of each stage is copied to the downstream stages with a probability that progressively decreases . We will show later that this process of “synaptic copying” can actually be mediated by neuronal activity which resembles the observed replay activity [14]–[19] . Transition probabilities of the different memory stages are the same as in the heterogeneous model: . We will follow the SNR for a particular memory by measuring the correlation of the synaptic states in each stage with the event of interest . Once again , we can derive a mean-field description of the stochastic dynamics . The upshot is that the mean signal in stage obeys the differential equationwhich expresses clearly how the signal in stage is driven by that in stage . This is precisely the mechanism behind the improvement of memory performance compared to the heterogenous model without interactions . The memory trace in the first stage decays exponentially as new memories are encoded , as in the homogeneous case ( see Fig . 2a ) . Memory traces in downstream stages start from zero , increase as the synaptic states are propagated , and eventually decay once again to zero . Information about all the stored memories is transferred between stages because the synapses that are “copied” are correlated to all the memories that are still represented at the particular memory stage . The most plastic stages retain the memories for a limited time , but during this time they transfer them to less plastic stages . This explains why the memory traces of downstream stages are non-monotonic functions of time: at stage , the memory trace keeps increasing as long as the information about the tracked memory is still retained in stage . The memory trace in the second stage is already greater than that of an equivalent heterogeneous model with independent synaptic groups ( Fig . 2a ) . This effect is enhanced as more stages are added . The memory trace takes the form of a localized pulse that propagates at an exponentially decreasing rate ( Fig . 2b ) . It begins as a sharply peaked function in the fast learning stages but slowly spreads outward as it propagates toward the slow learning stages . This indicates that although the memory is initially encoded only in the first stage ( presumably located in the medial temporal lobe ) , at later times it is distributed across multiple stages . Nonetheless , it has a well defined peak , meaning that at intermediate times the memory is most strongly represented in the synaptic structure of intermediate networks . An analytical formula for the pulse can be derived , see Methods and Text S1 , which allows us to calculate the SNR and memory lifetimes ( Fig . 3 ) . Now , when reading out the signal from several stages of the memory transfer model , we must take into account the fact that the noise will be correlated . This was not the case for the heterogeneous model without interactions . In fact , if we consider a naive readout which includes all stages , the noise will increase weakly with the number of stages . On the other hand , if we only read out the combination of stages which maximizes the SNR , one can show that the noise is independent of and very close to the uncorrelated case . In fact , this readout is equivalent to reading out only those groups whose SNR exceeds a fixed threshold , which could be learned , see Text S1 for more details . Fig . 3a shows the SNR for memories in the heterogeneous model ( dashed lines ) and the memory transfer model ( solid lines ) for a fixed number of synapses and different numbers of groups . The curves are computed using the optimal readout described above , for which noise correlations are negligible . Both the SNR for intermediate times and the overall lifetime of memories increase with increasing in the memory transfer model . The increase in SNR is proportional to , see Fig . 3b , while the lifetime is approximately linear in for large enough , see Fig . 3c . While the initial SNR is reduced compared to the heterogeneous model ( by a factor proportional to ) , it overtakes the SNR of the heterogeneous model already at very short times ( inset of Fig . 3a ) . Importantly , the memory transfer model also maintains the propitious scaling seen in the heterogeneous model of the SNR and memory lifetime with the number of synapses . Specifically , if the slowest learning rate is scaled as , then the very initial SNR scales as ( but almost immediately after the memory storage it scales as ) and the lifetime as . Hence the lifetime is extended by a factor that is approximately with respect to the memory lifetime of both the heterogeneous model and the cascade synaptic model [11] in which the memory consolidation process occurs entirely at the level of individual complex synapses . The improvement looks modest on a logarithmic scale , as in Fig . 3a , however it becomes clear that it is a significant amelioration when the actual timescales are considered . In the example of Fig . 3a the memory lifetime extends from three years for the heterogeneous model , to more than thirty years for the memory transfer model . As the memory lifetime extends , the initial signal to noise ratio decreases compared to the heterogeneous model ( but not compared to the cascade model , for which it decreases as , where is the number of levels of the cascade , or in other words , the complexity of the synapse ) . However , the reduction is small , and after a few memories the memory transfer model already outperforms the heterogeneous model . In the example of Fig . 3 the heterogeneous model has a larger SNR only for times of the order of hours . This time interval should be compared to the memory lifetime which is of the order of decades . The consolidation model we have described involves effective interactions between synapses that must be mediated by neuronal activity . We now show that it is possible to build a neuronal model that implements these interactions . We consider a model of identical stages , each one consisting of recurrently connected McCulloch-Pitts neurons ( the total number of plastic synapses is ) . Neurons in each stage are connected by non-plastic synapses to the corresponding neurons in the next stage ( feed-forward connections ) . See Fig . 4a for a scheme of the network architecture . The model operates in two different modes: encoding and transfer . Importantly , we must now be more careful concerning our definition of time . The unit of time we have used up until now was simply that of the encoding of a memory , i . e . one time step equals one memory . Now we have two different time scales: the encoding time scale and the neuronal time scale . The encoding time scale is just the same as before , i . e . it is the time between learning new memories . The neuronal time scale is much faster . Specifically , in the neuronal model we encode a new memory and then stimulate the neurons to drive the transfer of patterns of synaptic weights . The time-step used in the Hebbian learning process when a memory is encoded , as well as the time-step used during this transfer process is a neuronal time scale , perhaps from milliseconds to hundreds of milliseconds . The time between memory encodings , on the other hand , might be on the order of minutes or hours , for example . During encoding , a subset of neurons in the first stage is activated by the event that creates the memory and the recurrent synapses are updated according to a Hebbian rule , see Fig . 4b , c . Specifically , one half of the neurons are randomly chosen to be activated ( ) , while the remaining neurons are inactive ( ) , where is the state of the neuron in stage . A synapse is then potentiated ( ) with a probability if and is depressed ( ) with probability if , where is a binary synapse from neuron to neuron in stage . Consistent with the previous analysis , we assume that the neuronal patterns of activity representing the memories are random and uncorrelated . No plasticity occurs in the synapses of neurons in downstream stages during encoding . During transfer , a random fraction of neurons in each stage is activated at one time step , and the network response then occurs on the following time-step due to recurrent excitatory inputs . Specifically , at time , for all neurons which have been activated in stage , and otherwise . At time the recurrent input to a neuron in stage due to this activation is . If then and otherwise , where is a threshold . At time all neurons are silenced , i . e . and then the process is repeated times . The initially activated neurons at time are completely random and in general they will not be correlated with the neuronal representations of the stored memories . However , the neuronal response at time will be greatly affected by the recurrent synaptic connections . For this reason , the activity during the response will be partially correlated with the memories stored in the upstream stages , similar to what happens in observed replay activity ( see e . g . [14]–[19] ) . During transfer , the activated neurons project to counterpart neurons in the downstream stage . Crucially , we assume here that the long-range connections from the upstream stage to the downstream one are up-regulated relative to the recurrent connections in the downstream stage . In this way , the downstream state is “taught” by the upstream one . In the brain this may occur due to various mechanisms which include neuromodulatory effects and other gating mechanisms that modulate the effective couplings between brain regions . Cholinergic tone , in particular , has been shown to selectively modulate hippocampal and some recurrent cortical synapses ( see [20] ) as well as thalamocortical synapses [21] . Recent studies have also shown that the interactions between cortical and subcortical networks could be regulated by changing the degree of synchronization between the rhythmic activity of different brain areas ( see e . g . [22] ) . In our model we assumed that , due to strong feedforward connections , whenever we have . The pattern of activation in stage therefore follows that of stage during the transfer process . Importantly plasticity only occurs in the recurrent synapses of the downstream stage , i . e . stage is ‘teaching’ stage . For illustration we first consider a simple learning rule which can perfectly copy synapses from stage to stage , but only for the special case of , i . e . single-neuron stimulation . Following this , we will consider a learning rule which provides for accurate but not perfect copying of synapses but which is valid for any . Fig . 5 shows a schematic of the transfer process when . In this simplest case , only one presynaptic synapse per neuron is activated . To successfully transfer this synapse to the downstream stage a simple rule can be applied . First , the threshold is set so that . If there is a presynaptic spike ( ) followed by a postsynaptic spike ( ) , then potentiate ( ) with a probability equal to the intrinsic learning rate of the synapses , . If there is no postsynaptic spike ( ) then the corresponding synapse should be depressed ( ) . This leads to perfect transfer . In general and therefore it is not possible to perfectly separate inputs with a single threshold . Nevertheless , a learning rule which can accurately copy the synapses in this general case is the following . Consider two thresholds , which are ‘low’ and ‘high’ respectively . On any given transfer ( there are of them per stage ) is set to one of these two thresholds with probability . If then if and , then set with a probability . In words , this says that if despite the high threshold , the presynaptic activity succeeded in eliciting postsynaptic activity , then the synapses in stage must have been strong , therefore one should potentiate the corresponding synapses in stage . Similarly if then if and , then set with a probability . In words , this says that if despite the low threshold , the presynaptic activity did not succeed in eliciting postsynaptic activity , then the synapses in stage must have been weak , therefore one should depress the corresponding synapses in stage . For this learning rule to work , both stages and must be privy to the value of the threshold . Therefore , there must be some global ( at least common to these two stages ) signal available . This could be achieved via a dynamical brain state with long-range spatial correlations . For example , globally synchronous up-state and down-state transitions [23] , which are known to occur during so-called slow-wave sleep would be ideally suited to shift neuronal thresholds . Alternatively , theta oscillations have been shown to be coherent between hippocampus and prefrontal cortex in awake behaving rodents during working memory [24] and learning tasks [25] and would also be suited to serve as a global signal for synaptic plasticity . We have stated that this second learning rule involving two thresholds can lead to accurate learning in the general case . Concretely , we can completely characterize the transfer process between any two stages via two quantities: the transfer rate , which is the fraction of synapses transferred after replays of the transfer process , and the accuracy of transfer which is the fraction of transferred synapses which were correctly transferred . Both of these quantities depend on the stimulation fraction and the threshold and can be calculated analytically , see Methods . In short , the stimulation of neurons during the transfer process leads to a unimodal input distribution which is approximately Gaussian for . The transfer rate is proportional to the area in the tails of this distribution above the high threshold and below the low threshold , while the accuracy is the fraction of this area which is due only to strong synapses ( above the high threshold ) or to weak synapses ( below the low threshold ) . It is easy to see that as the thresholds are moved away from the mean into the tails the transfer rate will decrease while the accuracy will increase . There is therefore a speed-accuracy tradeoff in the transfer process . Additionally , the transfer process can be implemented even if we relax the assumption of strong one-to-one feedforward connections and allow for random feedforward projections , see Text S1 . In this case a two-threshold rule is still needed to obtain performance above chance level , although an analytical description is no longer straightforward . The neuronal implementation of the transfer process reveals an important fact: the probability of correctly updating a synapse does not depend solely on its intrinsic learning rate , but rather on the details of the transfer process itself . In our simple model , the transfer rate is where is a factor which depends on the threshold of the learning process relative to the distribution of inputs and is the intrinsic learning rate of the synapses in the downstream stage . Additionally , since the likelihood of a correct transfer is , the rate of correct transfers is , while there is also a “corruption” rate equal to which is the probability of an incorrect transfer . Obviously , if a given fraction of synapses is to be transferred correctly , the best strategy is to make as close to one as possible and increase accordingly . In the limit the neuronal model is exactly equivalent to the mean-field model we studied earlier with the transfer rate playing the role of the learning rate . For a modified mean-field model with a “corruption” term can be derived , see Text S1 for details . Fig . 6 illustrates that the neuronal implementation quantitatively reproduces the behavior of the synaptic mean-field model . Specifically , the transfer rate can be modified by changing the number of transfers , as shown in Fig . 6a . In this case , although the intrinsic synaptic properties have not changed at all , learning and forgetting occur twice as fast if is doubled . The combined SNR of ten stages with 1000 all-to-all connected neurons each averaged over ten realizations ( symbols ) is compared to the mean-field model ( line ) in Fig . 6 . In this case , the parameters of the neuronal model have been chosen such that the transfer rates are equal to , and . In conclusion , we showed that there is a clear computational advantage in partitioning a memory system into distinct stages , and in transferring memories from fast to slow stages . Memory lifetimes are extended by a factor that is proportional to the number of stages , without sacrificing the amount of information stored per memory . For the same memory lifetimes , the initial memory strength can be orders of magnitude larger than in non-partitioned homogeneous memory systems . In the Results we focused on the differences between the heterogeneous and the memory system model . In Fig . S15 in Text S1 we show that the SNR of the memory transfer model ( multistage model ) is always larger than the SNR of homogeneous model for any learning rate . This is true also when one considers that homogeneous models can potentially store more information than the memory transfer model . Indeed , in the homogeneous model all synapses can be modified at the time of memory storage , not only the synapses of the first stage . However , the main limitation of homogeneous models with extended memory lifetimes comes from the tiny initial SNR . If one reduces the amount of information stored per memory to match the information stored in the memory transfer model , it is possible to extend an already long memory lifetime but the initial SNR reduces even further ( see Text S1 for more details ) . Our result complements previous studies ( see e . g . [8] , [26] , [27] ) on memory consolidation that show the importance of partitioning memory systems when new semantic memories are inserted into a body of knowledge . Two-stage memory models were shown to be fundamentally important to avoid catastrophic forgetting . These studies focused mostly on “memory reorganization” , as they presuppose that the memories are highly organized and correlated . We have solved a different class of problems that plague realistic memory models even when all the problems related to memory reorganization were solved . The problems are related to the storage of the memory component that contains only incompressible information , as in the case of random and uncorrelated memories . These problems are not related to the structure of the memories and to their similarity with previously stored information , but rather they arise from the assumption that synaptic efficacies vary in a limited range . We showed here that this problem , discovered two decades ago [3] and partially solved by metaplasticity [11] , can also be solved efficiently at the systems level by transferring memories from one sub-system to another . Our neuronal model provides a novel interpretation of replay activity . Indeed , we showed that in order to improve memory performance , synapses should be copied from one stage to another . The copying process occurs via the generation of neuronal activity , that reflects the structure of the recurrent synaptic connections to be copied . The synaptic structure , and hence the neuronal activity , is actually correlated with all past memories , although most strongly with recent ones . Therefore while this activity could be mistaken for passive replay of an individual memory , it actually provides a snapshot of all the information contained in the upstream memory stage . There is already experimental evidence that replay activity is not a mere passive replay [28] . Our interpretation also implies that the statistics of “replay” activity should change more quickly in fast learning stages like the medial temporal lobe , than in slow learning stages like pre-frontal cortex or some other areas of the cortex [18] . Our analysis also reveals a speed-accuracy trade off that is likely to be shared by a large class of neuronal models that implement memory transfer: the faster the memories are transferred ( i . e . when a large number of synapses are transferred per “replay” and hence a small number of repetitions is needed ) , the higher the error in the process of synaptic copying ( Fig . 6a ) . Accuracy is achieved only when the number of synapses transferred per “replay” is small and is sufficiently large . This consideration leads to a few requirements that seem to be met by biological systems . In particular , in order to have a large , it is important that the transfer phases are brief , if the animal is performing a task . This implies that the synaptic mechanisms for modifying the synapses in the downstream stages should operate on short timescales , as in the case of Spike Timing Dependent Plasticity ( STDP ) ( see e . g . [29] ) . Alternatively , the transfer can occur during prolonged intervals in which the memory system is off-line and does not receive new stimuli ( e . g . during sleep ) . Although we have focused on the transfer of memories in our model , the neuronal model can additionally be used to read out memories . Specifically , the neuronal response of any stage ( or several stages ) to a previously encoded pattern is larger than to a novel pattern . This is true as long as the SNR , as we have used it in this paper i . e . synaptic overlap , is sufficiently large . This difference in neuronal response can be used by a read-out circuit to distinguish between learned and novel patterns , see Text S1 for a detailed implementation . Our theory led to two important results which generate testable predictions . The results are: 1 ) the memory performance increases linearly with the number of memory stages , and 2 ) the memory trace should vary in a non-monotonic fashion in most of the memory stages . The first suggests that long-term memory systems are likely to be more structured than previously thought , although we cannot estimate here what the number of partitions should be , given the simplicity of the model . Some degree of partitioning has already been observed: for example graded retrograde amnesia extends over one or two years in humans with damage to area CA1 of the hippocampus , but can extend to over a decade if the entire hippocampus is damaged [30] . Systematic lesion studies in animals should reveal further partitioning in the hippocampal-cortical pathway for consolidation of episodic memories . A second prediction is related , since once the partitions have been identified , our work suggests that most stages should exhibit non-monotonic memory traces , although on different time-scales . In fact , a recent imaging study with humans revealed non-monotonic BOLD activation as a function of the age of memories that subjects recalled [31] . Furthermore the non-monotonicity was observed only in cortical areas and not in hippocampus . Here multi-unit electrophysiology in animals would be desirable to obtain high signal-to-noise ratios for measuring the memory traces . An analysis such as the one proposed by [32] , [33] , in which spiking activity in rats during sleep was correlated with waking activity , should enable us to estimate the strength of a memory trace . We expect that the memory trace is a non-monotonic function of time in most memory areas . The initial trace is usually small or zero , it then increases because of the information transferred from the upstream memory stages , and it finally decreases as a consequence of the acquisition of new memories . The timescales of the rising phase should reflect the dynamics of the upstream memory stages , whereas the decay is more related to the inherent dynamical properties of the memory stage under consideration . Therefore , the position of the peak of the memory trace and the timescale of the decay give important indications on the position of the neural circuit in the memory stream and on the distribution of parameters for the different memory stages . The statistics of neural activity during memory transfer ( replay activity ) should reflect the synaptic connections and in particular it should contain a superposition of a few memory traces in the fast systems , and an increasingly larger number of traces in the slower systems . The statistics of the correlations with different memories should change rapidly in the fast systems , and more slowly in the slow systems ( e . g . in the hippocampus the changes between two consecutive sleeping sessions should be larger than in cortical areas where longer-term memories are stored ) . To obtain experimental evidence for these two sets of predictions , it is important to record neural activity for prolonged times , in general long enough to cover all the timescales of the neural and synaptic processes that characterize a particular brain area . This is important both to determine the time development of the memory traces and to understand the details of the neural dynamics responsible for memory transfer . To estimate the SNR , one can analyze the recorded spike trains during rest and NREM sleep , when memory transfer is expected to occur . We believe that the strength of memory reactivation is related to our SNR . The analysis proposed in [32] , [33] should allow us to estimate the templates of memories that are reactivated during one particular epoch ( the templates are the eigenvectors of the covariance matrix that contains the correlations between the firing rates of different neurons ) . The time development of the memory trace can be then studied by projecting the activity of a different epoch on the eigenvectors . The projections are a measure of the memory reactivation strength and they should be approximately a nonlinear monotonic function of the memory signal . This analysis not only would determine whether the memory trace is a non-monotonic function of time but it would also allow us to estimate the parameters that characterize its shape in different brain areas . The memory model studied here is a simple abstraction of complex biological systems which illustrates important general principles . Among the numerous simplifications that we made , there are three that deserve additional discussion . The first one is about the representations of the random memories and the second one is about the synaptic dynamics . The first simplification is that we implicitly assumed that the memory representations are dense , as all synapses are potentially modified every time a new memory is stored . In the brain these representations are likely to be sparse , especially in the early stages of the memory transfer model , which probably correspond to areas in the medial temporal lobe . Sparseness is known to be important for increasing memory capacity [3] , [34] , [35] and one may legitimately wonder why we did not consider more realistic sparse representations . However , in our simplified model sparser random representations are equivalent to lower learning rates if the average number of potentiations and depressions are kept balanced . If is the average fraction of synapses that are modified in the first stage ( coding level ) , then all qs of the model should be scaled by the same factor . This does not change the scaling properties that we studied , except for a simple rescaling of times ( the x-axis of the plots should be transformed as ) and SNR ( SNR→SNR· ) . In conclusion , sparseness is certainly an important factor and we are sure that it plays a role in the memory consolidation processes of the biological brain . However here we focused on mechanisms that are independent from the coding level and hence we did not discuss in detail the effects of sparseness , which have been extensively studied elsewhere [3] , [34] , [35] . The second simplification that merits a further discussion is that the model synapses studied here have a single time-scale associated with each of them . Our model can be extended to include synaptic complexity as in [11] . In fact , allowing for multiple time-scales at the level of the single synapse should lessen the number of stages needed for a given level of performance . Specifically , time-scales spanning the several orders of magnitude needed for high SNR and long memory lifetimes can be achieved through a combination of consolidation processes both at the single synapse , and between spatially distinct brain areas . The homogeneous and heterogeneous synaptic models are comprised of stochastically updated binary synapses which evolve in discrete time . In the homogeneous case all synapses have the same learning rate , while in the latter case there are groups of synapses each . Each group has a learning rate . At each time step all synapses are subjected to a potentiation or depression with equal probability . The N-bit word of potentiations and depressions constitutes the memory to be encoded . The memory signal at time , is the correlation of the synaptic states with a particular N-bit memory , and we use superscript to denote evolution in discrete time . The signal-to-noise ratio ( SNR ) is approximately ( and is bounded below by ) the signal divided by , see Text S1 for more details . To compare with these Markov models one can derive a mean-field description which captures the memory signal averaged over many realizations of the stochastic dynamics . This is done by considering the probability that a given synapse is in a given state as a function of time . Specifically , the probability of a single synapse with learning rate to be in the potentiated state at time is justwhere and . In the case of the homogeneous synaptic model there are synapses with the same learning rate . The expected value of the signal averaged over realizations is thenand so the expected signal-to-noise ratio isWe can approximate the finite-time equation for with a continuous ordinary differential equation which , using the definition of SNR givesthe solution of which is . This equation is used to plot the curves in Fig . 1b . The heterogeneous case is analogous withwhere is the expected signal at time in stage . This equation is used to plot the solid curve in Fig . 1c . The SNR in the heterogeneous model can be increased by reading out only some of the groups at any one point in time , as opposed to all of them . This optimal readout is used to plot the dashed curves in the top panel of Fig . 3 . Once again we assume there are a total of synapses divided equally amongst stages . Synapses in stage have learning rate and hence the fastest learning rate is and slowest is . Synapses in stage 1 are updated every time step in an identical fashion to those in group 1 of the heterogeneous model above . Synapses in downstream stages however , update according to the state of counterpart synapses in the upstream stage . Specifically , if a synapse in stage is potentiated ( depressed ) at time , then synapse in stage potentiates ( depresses ) at time with probability . As before , the signal at time in stage is written . This fully defines the stochastic model . As before we can derive a mean-field description of the stochastic dynamics . In this case , the probability of a given synapse in stage 1 to be in a potentiated state at time is as in the simple models . The probability of a given synapse in stage begin in a potentiated state can be writtensee Text S1 for details . These equations reflect the fact that only synapses in stage 1 are updated due to the presentation of random , uncorrelated memories , while synapses in downstream stages are updated only due to the state of synapses in the preceding stage . The expected signal in stage is given by . The continuous time approximation to the mean-field dynamics is given by the set of equationswith initial conditions , for and we write for the expected signal . These equations are used to plot the curves in Fig . 2a and the solid curves in the top panel of Fig . 3 . For sufficiently large we can furthermore recast this system of ODEs as a PDEwhere the spatial variable . An asymptotic solution to this equation valid for , and taking now the SNR , is ( 1 ) see Text S1 for details . This equation is used to plot the pulse solution shown in Fig . 2b . An optimal SNR , in which only some of the stages are read out , can be calculated based on Eq . 1 and is ( 2 ) which is valid for intermediate times where the SNR is powerlaw in form . This equation is used to plot the curves in Fig . 3 bottom left . Using Eqs . 1 and 2 one can calculate the lifetime of memories as ( 3 ) if the SNR of the pulse is above one before reaching the last stage or ( 4 ) is the SNR drops below one already before reaching the last stage . Eqs . 3 and 4 are used to plot the solid curves in Fig . 3 bottom right . There are stages . Each stage is made up of all-to-all coupled McCulloch-Pitts neurons . Each one of the synapses ( no self-coupling ) can take on one of two non-zero values . Specifically , the synapse from neuron to neuron , where . Furthermore , there are one-to-one connections from a neuron in stage to a neuron in stage . The model operates in two distinct modes: Encoding and Transfer .	Memory is critical to virtually all aspects of behavior , which may explain why memory is such a complex phenomenon involving numerous interacting mechanisms that operate across multiple brain regions . Many of these mechanisms cooperate to transform initially fragile memories into more permanent ones ( memory consolidation ) . The process of memory consolidation starts at the level of individual synaptic connections , but it ultimately involves circuit reorganization in multiple brain regions . We show that there is a computational advantage in partitioning memory systems into subsystems that operate on different timescales . Individual subsystems cannot both store large amounts of information about new memories , and , at the same time , preserve older memories for long periods of time . Subsystems with highly plastic synapses ( fast subsystems ) are good at storing new memories but bad at retaining old ones , whereas subsystems with less plastic synapses ( slow subsystems ) can preserve old memories but cannot store detailed new memories . Here we propose a model of a multi-stage memory system that exhibits the good features of both its fast and its slow subsystems . Our model incorporates some of the important design principles of any memory system and allows us to interpret in a new way what we know about brain memory .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "neuroscience", "learning", "and", "memory", "biology", "computational", "neuroscience" ]	2013	Efficient Partitioning of Memory Systems and Its Importance for Memory Consolidation
Schistosomiasis and opisthorchiasis are of public health importance in Southeast Asia . Praziquantel ( PZQ ) is the drug of choice for morbidity control but few dose comparisons have been made . Ninety-three schoolchildren were enrolled in an area of Lao PDR where Schistosoma mekongi and Opisthorchis viverrini coexist for a PZQ dose-comparison trial . Prevalence and intensity of infections were determined by a rigorous diagnostic effort ( 3 stool specimens , each examined with triplicate Kato-Katz ) before and 28–30 days after treatment . Ninety children with full baseline data were randomized to receive PZQ: the 40 mg/kg standard single dose ( n = 45 ) or a 75 mg/kg total dose ( 50 mg/kg+25 mg/kg , 4 hours apart; n = 45 ) . Adverse events were assessed at 3 and 24 hours posttreatment . Baseline infection prevalence of S . mekongi and O . viverrini were 87 . 8% and 98 . 9% , respectively . S . mekongi cure rates were 75 . 0% ( 95% confidence interval ( CI ) : 56 . 6–88 . 5% ) and 80 . 8% ( 95% CI: 60 . 6–93 . 4% ) for 40 mg/kg and 75 mg/kg PZQ , respectively ( P = 0 . 60 ) . O . viverrini cure rates were significantly different at 71 . 4% ( 95% CI: 53 . 4–84 . 4% ) and 96 . 6% ( 95% CI: not defined ) , respectively ( P = 0 . 009 ) . Egg reduction rates ( ERRs ) against O . viverrini were very high for both doses ( >99% ) , but slightly lower for S . mekongi at 40 mg/kg ( 96 . 4% vs . 98 . 1% ) and not influenced by increasing diagnostic effort . O . viverrini cure rates would have been overestimated and no statistical difference between doses found if efficacy was based on a minimum sampling effort ( single Kato-Katz before and after treatment ) . Adverse events were common ( 96% ) , mainly mild with no significant differences between the two treatment groups . Cure rate from the 75 mg/kg PZQ dose was more efficacious than 40 mg/kg against O . viverrini but not against S . mekongi infections , while ERRs were similar for both doses . Controlled-Trials . com ISRCTN57714676 Schistosomiasis , food-borne trematodiasis , and soil-transmitted helminthiasis are neglected tropical diseases that are of considerable public health relevance in Southeast Asia [1] . In Lao People's Democratic Republic ( Lao PDR ) , approximately 80 , 000 individuals are at risk for schistosomiasis mekongi , 2 million individuals are at risk for food-borne trematodiasis ( particularly opisthorchiasis ) , and 1 million school-aged children are at risk for soil-transmitted helminthiasis [1] . Praziquantel ( PZQ ) is the current drug of choice in the treatment of schistosomiasis and most of the food-borne trematode infections [1] . Deworming programs against schistosomiasis aim at morbidity control [2] . The World Health Organization ( WHO ) recommends a standard single dose of oral PZQ between 40 and 60 mg/kg for both schistosomiasis and food-borne trematodiasis [1] , [2] . In Lao PDR , a single dose of 40 mg/kg PZQ is recommended for mass treatment of schistosomiasis and opisthorchiasis [3] . For individual treatment , the PZQ dose to treat Opisthorchis viverrini infection is a total dose of 75 mg/kg divided into three doses [4] . PZQ is known to be effective against all six Schistosoma species causing disease in humans . However there have been just two small published clinical trials on PZQ cure rates against Schistosoma mekongi [5] , [6] . Both were non-randomized studies involving individuals relocated to non-endemic areas and given 60 mg/kg PZQ divided into two or three doses . To our knowledge , a controlled trial to treat S . mekongi using 40 mg/kg , the recommended dose for mass treatment in Lao PDR , and any comparison between different PZQ doses for superiority has so far not been undertaken . Several clinical trials have assessed PZQ efficacy against O . viverrini at the following dosages: single dose of 25 , 40 , or 50 mg/kg , or repeated 25 mg/kg doses for a total dose of 50 , 75 , or 150 mg/kg [7]–[13] . However , none has been conducted in Lao PDR , which also has S . mekongi co-endemic areas , and 40 mg/kg has not been compared with 75 mg/kg . Diagnosis of schistosomiasis , opisthorchiasis , and other intestinal or hepatobiliar helminth infections in epidemiological studies is commonly based on the detection of parasite eggs in stool specimens under a microscope . The Kato-Katz technique [14] , [15] is the recommended field method [16] and permits estimation of infection intensity expressed in eggs per gram of feces ( EPG ) . It is a relatively simple and rapid diagnostic method , but unfortunately , a single Kato-Katz thick smear has low sensitivity , particularly for light infections , and hence repeated stool examinations are necessary to improve the sensitivity of this technique [17]–[20] . This is especially important after treatment to avoid overestimation of cure rates . The low sensitivity of a single Kato-Katz thick smear results from the small amount of stool examined ( usually 41 . 7 mg ) , variation in helminth egg excretion over time in the same individual , and from variation in egg density within a stool specimen depending on sampling location , as recognized for Schistosoma mansoni [19] , [21] , [22] . The relative contribution of day-to-day and intra-specimen variation in fecal egg counts has been investigated for S . mansoni [19] , [21] where examination of repeated stool specimens , rather than examination of multiple Kato-Katz thick smears derived from a single stool specimen , was shown to be more appropriate to improve the sensitivity of detecting an infection [19] , [22] . While it is documented for S . mansoni that diagnostic sensitivity depends on the sampling effort , other helminth species are less well investigated . Repeated or multiple stool specimen collection is difficult in practice , particularly in rural community field surveys [20] , due to logistical requirements and cost implications . The current study pursued two objectives . First , we assessed the efficacy of two oral PZQ regimens ( i . e . , 40 mg/kg single dose , and 75 mg/kg divided dose , given as 50 mg/kg then 25 mg/kg 4 hours apart ) against S . mekongi and O . viverrini infections . Second , we determined the effect of multiple stool sampling on the diagnostic accuracy of the Kato-Katz technique before and after treatment , and assessed its impact on drug efficacy evaluation , considering both cure and egg reduction rates . Ethical clearance was obtained from the National Ethics Committee , Ministry of Health ( MoH ) in Vientiane , Lao PDR ( reference no . 027/NECHR ) and by the Ethics Committee of Basel , Switzerland ( EKBB; reference no . 255/06 ) . The study protocol is registered with Current Controlled Trials on controlled-trials . com ( identifier ISRCTN57714676 ) . Written informed consent was obtained by the parents or guardians of all pupils before participation in the study . The children had the opportunity to withdraw from the study at any time . Both doses of PZQ ( i . e . , single 40 mg/kg dose or total of 75 mg/kg dose ) are accepted within Lao MoH published guidelines . The 40 mg/kg single dose is mainly used in mass drug administration programs , while 75 mg/kg ( divided into three dosages ) is used for the treatment of individuals . In our study the 75 mg/kg dose was divided into two doses ( 50 mg/kg plus 25 mg/kg given 4 hours apart ) to simplify the regimen for a school setting where classes ended by the early afternoon . At the end of the follow-up period , all children were treated against soil-transmitted helminth infections with a single oral dose of 400 mg albendazole [3] . The primary objective of this study was to compare the efficacy of two different dose regimens of oral PZQ in school-aged children from southern Lao PDR in a S . mekongi and O . viverrini co-endemic area . The two regimens compared were ( i ) 40 mg/kg single dose and ( ii ) 75 mg/kg divided dose , given as 50 mg/kg then 25 mg/kg 4 hours apart . The secondary objectives were to determine the effect of multiple stool sampling to assess cure and egg reduction rates and to estimate the increased diagnostic sensitivity by multiple Kato-Katz thick smears from a single stool specimen compared with additional stool specimens obtained over several days before and after treatment . S . mekongi and O . viverrini were the species of primary interest , but hookworm was also included for the baseline analyses . Finally , the prevalence of the other intestinal helminth infections among our cohort of schoolchildren was also assessed . The dose comparison study was a randomized trial with 1∶1 allocation . It was conducted in February and March 2007 in the primary and secondary schools on Don Long Island , Khong district , Champasack province , Lao PDR . The 308 children registered at the Don Long school were invited for the dose comparison trial . Most of the pupils ( 60% ) lived in one of the four villages of Don Long Island , whereas the remaining children traveled from four villages on surrounding islands . In-depth stool examination was limited to 93 children aged 10–15 years ( two classes ) . Based on the asymptotic normal method ( formula 7 ) of Sahai and Khurshid [23] , this sample size has a 70% power to demonstrate a superiority of 20% of the highest PZQ dosage ( type I error: alpha = 5%; 1-tailed test ) when considering a 20% dropout rate . Analyses of the present paper are restricted to this in-depth cohort . Acutely ill or febrile children were excluded from the study . Don Long is a rural island in the Mekong River with about 1 , 500 inhabitants who practice subsistence farming and fishing . Previous studies on this island found the area to be co-endemic for S . mekongi and O . viverrini infections [24] , [25] . Laboratory facilities were established in Khong district hospital in Muang Khong , a village on the east side of Don Khong , the main island of Khong district . The children were assigned into two treatment arms following a 1∶1 allocation regardless of the baseline examination . Randomization was generated using a random number table in blocks of 10 . Randomization and supervision of the trial were conducted by the study leaders ( LL , TKM ) . Based on the child's weight , the dose was rounded to the nearest 150 mg by splitting the 600 mg PZQ tablets ( Distocide® , Korea ) in quarters using a pill cutter . Doses were prepared in advance by team members not involved in administrating the intervention . Each preparation was double verified for name , dose , and recorded weight for each child . Twelve hours before treatment , all doses were prepared and sealed in opaque envelopes that were labeled with the dose number , study unique identification number , the child's name , and weight . After the dose envelopes were prepared , the randomization and allocation list was sealed in an opaque envelope . Box 1 contained the envelopes with the first ( and only ) dose for children allocated in the 40 mg/kg arm and the first dose for those assigned to the 75 mg/kg arm , organized by school class and name . Box 2 contained the prepared envelopes for the second dose ( 25 mg/kg ) only for those children allocated for the total dose of 75 mg/kg PZQ . The drugs were administered by one of two paired teams of health care workers . The team confirmed that the child matched the identification on the drug envelope and then directly observed treatment . The drug administering teams were not involved prior to or after the study and not in any outcome assessments . As the different regimen was apparent ( single vs . a divided dose 4 hours apart ) neither the two health care teams nor the children were masked during treatment administration . The Lao physicians who assessed the children for adverse events following treatment were unaware of the dose allocation and were not involved with administering the intervention ( KP , PAS ) . Laboratory technicians assessing infection status were blinded to the dose allocation . The purpose and procedures of the study were explained to the school director , teachers , and to the village chief , who all agreed to participate . The study was explained during class to the children and written informed consent was received from their parents or guardians . Clinical baseline measurements and baseline laboratory determination of infection status were performed prior to treatment for each participating child . Clinical measurements included a morbidity questionnaire and physical examination . For laboratory procedures , plastic bags with pre-labeled 30 ml plastic containers were distributed to the children at enrolment and pupils were asked to return the containers the following day with a thumb-sized portion of their morning stools . Containers were collected each morning at the school from 07:30 to 08:30 hours , recorded on a line listing , and children were given new empty plastic containers for the following day . This procedure was repeated until 3 morning stool specimens per child were received . Fresh stool specimens were transferred daily to the laboratory on Khong Island for examination . From each stool specimens , triplicate Kato-Katz thick smears using standard 41 . 7 mg templates were prepared on microscope slides in accordance with the kit instructions ( Vestergaard Frandsen; Lausanne , Switzerland ) . The slides were quantitatively examined under a microscope within 1 hour following slide preparation . The number of eggs of O . viverrini , S . mekongi , hookworm , Trichuris trichiura , Ascaris lumbricoides , Taenia spp . , Enterobius vermicularis , and other helminths were counted and recorded separately . For quality control , 10% of the slides were randomly selected and re-examined by a senior technician without prior knowledge of the results . When discrepancies were observed ( e . g . , egg counts differing by more than 10% ) , the technicians received closer supervision by a more experienced colleague . Since O . viverrini cannot be easily distinguished from minute intestinal flukes ( MIF ) microscopically by the Kato-Katz technique [26] , infections reported here as O . viverrini infections are assumed to include some MIF co-infections . Following baseline data collection , children were treated with 40 mg/kg or 75 mg/kg oral PZQ as described . Immediately following the dose , the children were given two soupspoons of sticky rice ( ∼40 g ) to increase PZQ bioavailability and minimize potential adverse events [27] . Adverse events spontaneously reported within 3 hours after administration of the first dose were recorded . Additionally , a solicited questionnaire on adverse events was administered 24 hours following PZQ administration and graded for severity . All clinical and laboratory assessments were repeated 28–30 days after PZQ administration . Data were entered in EpiData software version 3 . 1 ( EpiData Association; Odense , Denmark ) and double-checked against the original data sheets . Data analysis was performed using Intercooled STATA release 9 . 0 ( StataCorp; College Station , TX , USA ) . For each helminth species , an infection was defined as the presence of one or more eggs in at least one of the Kato-Katz thick smears examined . Cumulative prevalence of each helminth infection detected after examination of 9 Kato-Katz thick smears ( 3 stool specimens with triplicate Kato-Katz per specimen ) was calculated . Tests for significant associations with gender were analyzed by negative binomial regression . Intensity of infection ( expressed in EPG ) was calculated by multiplying the observed number of eggs by a factor of 24 . Geometric mean intensity of infection was calculated on EPG . Infections with O . viverrini were classified into three groups [28]: light ( 1–999 EPG ) , moderate ( 1 , 000–9 , 999 EPG ) , and heavy infections ( ≥10 , 000 EPG ) . S . mekongi infections were grouped into the following three categories [29]: light ( 1–99 EPG ) , moderate ( 100–399 EPG ) , and heavy infections ( ≥400 EPG ) . Negative binomial regression was applied to compare infection intensities of S . mekongi and O . viverrini at baseline among the two treatment groups . Cure rates of S . mekongi and O . viverrini were calculated as the proportion of children with no egg excretion after treatment among those with eggs in their stool at baseline . Children found egg-negative prior to treatment but egg-positive after treatment were considered to be false negative and counted as infected at baseline . These infections were assumed to have been missed at baseline because the 28–30 days follow-up would not have provided adequate time for re-infection and patency between the two surveys . Cure rates obtained with the two tested doses were compared with Fisher's exact test . Egg reduction rates were determined by comparing the geometric mean egg output before and 28–30 days after treatment among children infected at baseline ( 1 - geometric mean egg output posttreatment/geometric mean egg output at baseline , multiplied by 100 ) . The effect of multiple sampling on the sensitivity of the Kato-Katz technique to detect S . mekongi and O . viverrini infections was assessed before and after drug administration . Hookworm infections were also included at baseline . Prevalences with 95% confidence interval ( CI ) were calculated for each sampling effort , the minimum effort being defined as the first Kato-Katz thick smear derived from the first stool specimen . The sampling effort increased with additional Kato-Katz thick smear examinations from the same stool specimen and with additional stool specimens . The McNemar test was used to compare prevalences assessed by different sampling efforts . The maximum sampling effort , 9 Kato-Katz thick smears , was taken as the diagnostic ‘gold’ standard to assess the sensitivity of increasing sampling efforts . Adverse event frequencies depending on treatment doses were compared with the exact χ2 test . Additionally , infection intensities were expressed in EPG and for each child the arithmetic means were computed for each sampling effort . At the cohort level , geometric mean fecal egg counts were calculated for each sampling effort considering only the children with complete datasets at each time point separately . The analysis was restricted to the egg-positive children , based on the examination of 9 Kato-Katz thick smears ( maximum sampling effort ) . Table 1 summarizes baseline infection prevalences and intensities of all helminth species diagnosed in the present study before PZQ administration . Results pertained to those children who had complete data records ( 9 Kato-Katz thick smears ) prior to treatment ( n = 90 ) and before and after treatment combined ( n = 64 ) . S . mekongi , O . viverrini , and hookworm were the most common parasitic infections at baseline , with prevalences above 85% for each helminth species , as assessed with the maximum sampling effort . Other intestinal parasitic infections , in descending order of prevalence , were T . trichiura , A . lumbricoides , E . vermicularis , and Taenia spp . One infection with Hymenolepis diminuta was detected . Infection prevalences for any of the aforementioned helminths did not differ between boys and girls . Cure and egg reduction rates were compared between two cohorts ( Figures 1a and 1b ) . First , children who complied with the maximum diagnostic effort ( 9 Kato-Katz thick smears before and after treatment , n = 64 ) and , second , children with a minimum diagnostic effort ( 1 Kato-Katz thick smear at each time point , n = 85 ) . Results are summarized in Tables 2 and 3 . For both cohorts , there was no significant differences in the infection intensities of S . mekongi and O . viverrini at baseline between the two treatment groups ( all P>0 . 05 ) . S . mekongi cure rates among children who had provided three stool specimens at baseline and follow-up were 80 . 8% ( 21/26; 95% CI: 60 . 6–93 . 4% ) after 75 mg/kg PZQ and 75 . 0% ( 24/32; 95% CI: 56 . 6–88 . 5% ) after 40 mg/kg PZQ , which was not significantly different ( P = 0 . 754 ) . With the minimum diagnostic effort , observed cure rates were considerably higher , 94 . 7% ( 18/19; 95% CI: not defined ) and 85 . 7% ( 18/21; 95% CI: not defined ) , respectively . S . mekongi egg reduction rates in both cohorts were >93% . Slightly higher egg reduction rates were observed at the minimum sampling effort ( 97 . 9% and 99 . 6% in the 40 mg/kg and 75 mg/kg treatment group , respectively ) , compared to the highest sampling effort ( 96 . 4% and 98 . 1% , respectively ) . Based on the maximum sampling effort , O . viverrini cure rates were 96 . 6% ( 28/29; 95% CI: not defined ) after 75 mg/kg PZQ and 71 . 4% ( 25/35; 95% CI: 53 . 4–84 . 4% ) after 40 mg/kg PZQ , showing a statistically significant difference ( P = 0 . 009 ) . Considering the minimum diagnostic effort , observed cure rates were 100% ( 35/35; 95% CI: not defined ) and 94 . 3% ( 33/35; 95% CI: not defined ) , respectively , with no statistically significant difference ( P = 0 . 493 ) . Egg reduction rates , regardless of treatment group and diagnostic efforts , were above 99% . Solicited 24-hour adverse event profiles in the two treatment groups are summarized in Table 4 . Fourteen children were not available to be interviewed ( n = 6 , 40 mg/kg dose; n = 8 , 75 mg/kg dose ) , corresponding to 15 . 1% lost to follow-up , but no serious adverse events were reported by the community when we returned days 28–30 for post-treatment follow-up . Most children reported one or more adverse events ( 76/79 , 96% ) . More cases were reported for most types of adverse events in the 75 mg/kg treatment arm , but did not reach statistical significance in this small sample when comparing the total number of events or those graded as severe . There were a total of 7 cases recorded as hypotension ( below 100 mm Hg systolic blood pressure ) in the 75 mg/kg treatment group compared with a single case in the 40 mg/kg group , which was statistically higher ( P<0 . 02 ) but no case was graded severe ( e . g . , no syncope ) . Children with hypotension associated with dizziness and vomiting were given rest and monitored; all cases were self-limiting . No serious adverse events required hospitalization . Figure 2 shows the cumulative prevalence of infected children over repeated stool specimens according to the number of Kato-Katz thick smears examined per stool specimen for S . mekongi and O . viverrini infections both at baseline and at the 28–30 day posttreatment follow-up survey . Baseline results for hookworm infections were also recorded although not the primary outcome of the study ( nor were hypotheses made on the efficacy of PZQ against this helminth species ) . The sensitivity of three different sampling efforts ( considering the maximum diagnostic effort of 9 Kato-Katz thick smears as the diagnostic ‘gold’ standard ) is presented in Table 5 . Figure 3 illustrates the results of increased sampling effort on the geometric mean fecal egg counts before and after PZQ treatment of all children infected with S . mekongi and O . viverrini . This was also assessed for hookworm at baseline . At baseline , the mean fecal egg counts gradually increased with increasing sampling efforts . Thus , egg count estimates for S . mekongi , O . viverrini , and hookworm increased 4 , 2 . 4 and 1 . 7-fold , reaching values of 25 EPG , 342 EPG and 321 EPG , respectively , when assessed with the maximum sampling effort . S . mekongi and O . viverrini mean fecal egg count estimates were considered low-intensity infections . After PZQ treatment , the benefit of the maximum sampling effort for EPG was 9-fold and 8-fold for S . mekongi and O . viverrini , respectively . When comparing the pretreatment baseline with the 28–30 day posttreatment follow-up , the mean fecal egg count for O . viverrini sharply decreased from 342 to 9 . 1 EPG . The decrease was less marked for S . mekongi , from 25 to 8 EPG . A 3×1 sampling effort yielded substantially higher estimates than a 1×3 sampling effort for S . mekongi and hookworm egg counts . By contrast , the same efforts showed only a minimal increase for O . viverrini egg counts . PZQ is the drug of choice against most trematode infections , including schistosomiasis and opisthorchiasis . To our knowledge , PZQ dose comparison studies have not been described for S . mekongi . Dose comparison studies for O . viverrini have been conducted , but most studies relied on an insensitive diagnostic approach , i . e . , single stool specimen examination before and after drug administration . The accuracy of diagnosis , which is particularly important for estimating cure rates , can be improved by examining multiple Kato-Katz thick smears derived from a single or multiple stool specimens [30] . In this study , S . mekongi cure rate after administration of 75 mg/kg PZQ ( 80 . 8% ) was not significantly higher than the cure rate obtained after a single dose of 40 mg/kg ( 75 . 0% ) when assessed with the maximum sampling effort of 9 Kato-Katz thick smears . The cure rate from either regimen was largely overestimated if diagnosis was based on a single Kato-Katz thick smear . Studies based on fewer Kato-Katz thick smears are more likely to overestimate cure rate and be less diagnostically sensitive to detect any differences in dose comparisons . Two small studies carried out in the 1980s on S . mekongi infection reported high cure rates with 60 mg/kg PZQ ( 90 . 9% and 97 . 5% , respectively ) [5] , [6] when analyzing 2–3 stool specimens but using different stool diagnostic techniques ( Kato-Katz+modified Ritchie's and Stoll's , respectively ) . Similarly in a recent multi-country randomized trial comparing single 40 mg/kg and 60 mg/kg PZQ in children aged 10–19 years , with infections diagnosed by two stool specimens ( duplicate Kato-Katz thick smears per specimen ) , the 21-day posttreatment follow-up was reported as 92 . 8% with 60 mg/kg , which was not a significant improvement against S . mansoni , S . haematobium , or S . japonicum infections compared to the standard 40 mg/kg [31] . Consistent with results obtained from this recent trial , our study did not document a significantly improved cure rate ( days 28–30 posttreatment ) with an even higher total dose ( 75 mg/kg dose ) for S . mekongi , even with higher diagnostic sensitivity from greater stool sampling efforts . However our additional sampling effort did observe a cure rate for 40 mg/kg about 15% lower than rates reported in the multicenter trial . O . viverrini cure rate after administration of 75 mg/kg PZQ ( 96 . 6% ) was significantly higher than the cure rate obtained after a single dose of 40 mg/kg ( 71 . 4% ) when assessed with the maximum sampling effort . However , if the cure rate had been based on results of single Kato-Katz thick smear before and after drug administration , as often the case in community-based surveys , no significant difference would have been found . Cure rate was particularly overestimated when based on a single Kato-Katz thick smear in this study for a 40 mg/kg dose ( 94 . 3% ) , similar to high , and most likely overestimated cure rates ( 91–100% ) reported from previous studies using the same dosage and only a single stool examination [9] , [10] , [12] . Cure rates which were reported as 100% after administration of 75 mg/kg PZQ ( divided into three doses ) were also likely overestimated in previous studies [7] , [13] . Our study therefore provides supportive evidence that a 75 mg/kg total dose of PZQ is highly efficacious against O . viverrini and S . mekongi infections in school-aged children from Lao PDR . The total dose was divided into two doses instead of three and had a 24-hour profile of common adverse events similar to a single 40 mg/kg dose . Two doses , instead of three , are operationally and logistically more feasible , but clearly single-dose regimens are the preferred option for large-scale preventive chemotherapy programs . The small size of our study , however , limits detecting a difference in the nature or frequency of adverse events between the two regimens . The non-significant difference between the two doses to cure S . mekongi infections should be interpreted with caution . Again , this may result from the study's small sample size and it would therefore be valuable to investigate a larger sample . In addition , most of the children included in our study only had low intensity infections while cure rate achieved by PZQ has been shown to be influenced by the infection burden [32] . Some authors have argued that egg reduction rate is a more appropriate indicator than cure rate for drug efficacy evaluation [33] , [34] . We assessed both cure and egg reduction rates . Importantly , we found very high egg reduction rates ( >99% ) against O . viverrini for both treatment regimens regardless of the sampling effort . For S . mekongi , considering 9 Kato-Katz thick smears as the diagnostic ‘gold’ standard , a somewhat lower egg reduction rate was observed with a single 40 mg/kg dose of PZQ compared to the higher split dose ( 96 . 4% vs . 98 . 1% ) . At the lower sampling effort , higher egg reduction rates were observed ( 97 . 9% and 99 . 6% , respectively ) . These data suggest that the worm burden sharply declined from either dose regimen , which was found using either minimal or maximal diagnostic effort . This may be explained by the low posttreatment infection intensity of the non-cured children given either dose . The geometric mean egg counts in the two PZQ regimens were very similar . The public health goal of preventive chemotherapy is to reduce morbidity , which is indirectly assessed using egg reduction rates . Our results suggest that PZQ , given at a single oral dose of 40 mg/kg , is suitable to achieve this goal , particularly against O . viverrini . At baseline , the relative increase of sensitivity by multiple sampling was relatively low , especially for O . viverrini and hookworm infections . By contrast , multiple sampling was important after treatment , when infection prevalence and intensity were much lower . As a result , the sensitivity of the first Kato-Katz thick smear was much lower after treatment than at baseline , with a 4-fold lower and 3-fold lower sensitivity to detect O . viverrini and S . mekongi infections , respectively . A single Kato-Katz thick smear is known to have a low sensitivity for the diagnosis of O . viverrini , especially for low intensity infections [20] . For S . mekongi , the low sensitivity of a single Kato-Katz thick smear to detect this fluke observed in the present study agrees with previous findings obtained from investigations focusing on S . mansoni and S . japonicum [18] , [19] , [22] , [35] , [36] . Studies on the sensitivity of the Kato-Katz technique for diagnosis of S . mekongi are generally lacking . For O . viverrini and hookworm diagnosis , the sensitivity of a single Kato-Katz thick smear to detect infection at baseline was fairly high . For hookworm , this was in contrast to previous studies from Côte d'Ivoire [37] , [38] , Ethiopia [18] , and Tanzania [39] , where the sensitivity of a single Kato-Katz thick smear varied from 18% to 53% . However , after drug administration , when the overall O . viverrini infection intensity of our cohort of children became low ( <10 EPG ) , this study indicates the need for multiple Kato-Katz thick smear examinations , ideally performed on stool specimens collected over consecutive days for a more accurate estimation of the cure rate . Helminth eggs are non-randomly distributed within a stool specimen because the intestinal content is not uniformly mixed [40] and may affect the sensitivity of detecting an infection and fecal egg count estimates from a single Kato-Katz thick smear . Important day-to-day variation in egg output has been thoroughly documented for S . mansoni and S . japonicum [19] , [21] , [22] , [35] . By contrast , O . viverrini egg output was found to be relatively consistent over a period of several days in hospitalized patients [41] . Of note , Schistosoma egg shedding dynamics are additionally affected by retention of eggs in intestinal and liver tissues and the lower fecundity of female worms . We have compared the relative importance of intra-specimen and day-to-day variation of fecal egg counts before and after PZQ administration and determined its effect on evaluating anthelmintic drug efficacy . Previous research has shown that the examination of fewer specimens from different days proved to be superior than examining multiple Kato-Katz thick smears from a single stool specimen for more accurate estimates of the ‘true’ infection status for S . mansoni [19] , [22] . In the present study for S . mekongi and hookworm infections , examination of one Kato-Katz thick smear per stool specimen , with specimens collected over a 3-day period ( 3×1 sampling scheme ) , resulted in higher prevalence and mean infection intensity than three Kato-Katz thick smears taken from the first stool specimen ( 1×3 ) . For O . viverrini , however , the 3×1 and 1×3 sampling scheme revealed the same prevalence estimates . Since repeating the collection of a stool specimen over consecutive days is more costly , logistically more cumbersome , and negatively impacts on study compliance , examination of multiple Kato-Katz thick smears from a single stool specimen should be considered as a suitable approach for community surveys of helminth infections . Similar observations have been made before for the diagnosis of Clonorchis sinensis [42] . S . mekongi is known to be endemic in certain areas of the Mekong River basin [25] , [43]–[45] , while O . viverrini and hookworm species are widely distributed across Lao PDR [46]–[48] . Point prevalences as high as those observed in the present study for S . mekongi ( 87 . 8% ) , O . viverrini ( 98 . 9% ) , and hookworm ( 96 . 7% ) , based on a rigorous diagnostic effort , have rarely been described in the literature . Yet , our findings corroborate with a recent risk profiling study in more than 50 villages of Champasack province , where O . viverrini prevalences were above 80% in most villages , with particularly high prevalences observed in villages in close proximity to the Mekong River [24] . WHO surveyed selected villages on Khong Island ( an island also situated along the Mekong River , only 10 km from our study site ) prior to starting schistosomiasis control campaigns in the late 1980s , and found a similarly high S . mekongi prevalence ( 87 . 8% ) as reported here [49] . Studies carried out in rural provinces of southern Lao PDR ( Champasack and Saravane ) reported prevalences of O . viverrini and hookworm ranging from 18 . 8% to 70 . 8% and from 12 . 5% to 46 . 1% , respectively [46] , [47] , [50] , [51] . Infection prevalence is known to vary locally [46] , which may partially explain the difference between prior estimates and those found in this study . However , previous prevalence estimates were based on a single Kato-Katz thick smear , while 9 Kato-Katz thick smears were examined in the present study . O . viverrini infection prevalence probably includes MIF infections since co-infections are common , and polymerase chain reaction ( PCR ) techniques on stool specimens taken from the same study area in southern Lao PDR [52] have demonstrated that MIF eggs cannot easily be distinguished microscopically from O . viverrini by the Kato-Katz technique [26] . In conclusion , the present study found that the added benefit of multiple Kato-Katz thick smear examination and repeated stool sampling depends on the helminth species and baseline infection intensity . Thus , in the present setting in Lao PDR , where O . viverrini , S . mekongi , and hookworm are all highly endemic , estimating the baseline prevalence and intensity of infection for these species with a single Kato-Katz examination may be acceptable . By contrast , estimating the prevalence of infection after treatment by the Kato-Katz technique requires multiple thick smears , ideally taken from multiple stool specimens because the positive predictive value is lower ( both lower prevalence and lower geometric mean fecal egg count after treatment ) . A single Kato-Katz thick smear after treatment will considerably overestimate cure rate , but only minimally influences egg reduction rates . A rigorous diagnosis approach is necessary for estimating ‘true’ cure rates , as it has been previously demonstrated in studies on S . mansoni [30] , [53] . For anthelmintic drug evaluations with emphasis on egg reduction rates , a single Kato-Katz thick smear before and after treatment might suffice . In our view , multiple stool examination should nonetheless be considered in a subsample of the population surveyed in order to improve the monitoring of large-scale control programs , provide reasonable estimates on infection prevalence and intensity , and detect subtle changes in drug efficacies that might indicate the emergence of drug resistance development .	Parasitic worm infections are of public health importance in Southeast Asia . Particularly , the blood-dwelling Schistosoma mekongi worm , which is acquired by skin contact with the infectious cercariae in freshwater , can lead to liver enlargement . An infection with Opisthorchis viverrini is obtained by consumption of undercooked freshwater fish , and this infection increases the risk of developing cholangiocarcinoma . A single oral dose of 40 mg/kg praziquantel is recommended for mass treatment of schistosomiasis and opisthorchiasis , while at the individual level , a total dose of 75 mg/kg divided into three doses , is currently common practice to treat O . viverrini infection . Diagnosis is based on stool examination under a microscope for detection of worm eggs , but is limited by the low sensitivity of the widely used Kato-Katz technique . In this study , we showed that a 75 mg/kg total dose of praziquantel ( 50 mg/kg+25 mg/kg given 4 hours apart ) cleared significantly more O . viverrini infections than a single 40 mg/kg dose , but no difference was observed for S . mekongi . Solicited adverse event profiles were mainly mild and similar in both groups . Repeated stool examination before and after treatment was essential for an accurate assessment of drug efficacy in terms of cure rate , but showed no effect on assessing egg reduction rates .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "public", "health", "and", "epidemiology", "clinical", "research", "design", "infectious", "disease", "epidemiology", "parasitic", "diseases", "food-borne", "trematodes", "parasitology", "preventive", "medicine", "neglected", "tropical", "diseases", "infectious"...	2012	Efficacy of Praziquantel against Schistosoma mekongi and Opisthorchis viverrini: A Randomized, Single-Blinded Dose-Comparison Trial
The replication time of Saccharomyces cerevisiae telomeres responds to TG1–3 repeat length , with telomeres of normal length replicating late during S phase and short telomeres replicating early . Here we show that Tel1 kinase , which is recruited to short telomeres , specifies their early replication , because we find a tel1Δ mutant has short telomeres that nonetheless replicate late . Consistent with a role for Tel1 in driving early telomere replication , initiation at a replication origin close to an induced short telomere was reduced in tel1Δ cells , in an S phase blocked by hydroxyurea . The telomeric chromatin component Rif1 mediates late replication of normal telomeres and is a potential substrate of Tel1 phosphorylation , so we tested whether Tel1 directs early replication of short telomeres by inactivating Rif1 . A strain lacking both Rif1 and Tel1 behaves like a rif1Δ mutant by replicating its telomeres early , implying that Tel1 can counteract the delaying effect of Rif1 to control telomere replication time . Proteomic analyses reveals that in yku70Δ cells that have short telomeres , Rif1 is phosphorylated at Tel1 consensus sequences ( S/TQ sites ) , with phosphorylation of Serine-1308 being completely dependent on Tel1 . Replication timing analysis of a strain mutated at these phosphorylation sites , however , suggested that Tel1-mediated phosphorylation of Rif1 is not the sole mechanism of replication timing control at telomeres . Overall , our results reveal two new functions of Tel1 at shortened telomeres: phosphorylation of Rif1 , and specification of early replication by counteracting the Rif1-mediated delay in initiation at nearby replication origins . Chromosomal DNA replication occurs according to a regulated program , with some replication origins initiating early and others late in S phase [1] , [2] . S . cerevisiae telomeres provide a good model for understanding molecular controls over the temporal regulation of DNA replication . The replication time of S . cerevisiae telomeric regions is regulated by telomere length; chromosome regions close to telomeres with a normal length terminal TG1–3 tract generally replicate late , but those close to telomeres with a shortened TG1–3 tract replicate early [3] , [4] . This control is mediated through altered initiation time of replication origins . Telomeres may be replicated either by replication forks from an origin within the subtelomeric repeat sequences ( X or Y′ ARS elements ) , or by a fork arriving from a nearby telomere-proximal origin ( such as ARS522 , close to chromosome V-right; previously known as ARS501 ) [5]–[7] . Normal length telomeres can direct the late activation of such origins , while telomeric and telomere-proximal origins activate earlier if next to a shortened telomere—as demonstrated by experiments using recombination-based excision of TG1–3 repeats or the mutation yku70Δ that causes shortened telomeres [3] , [4] , [8] . Telomere repeat length can affect origins up to 40 kb from the chromosome end [4] . Earlier replication is proposed to favor telomerase recruitment and TG1–3 repeat lengthening [9]–[11] . However , how cells detect and respond to telomere length in order to control the replication time of telomeres remains unclear . The end-replication problem causes shortening of terminal TG1–3 tracts in successive cell cycles , and a network of controls detects critically short telomeres and ensures they are preferentially elongated by telomerase enzyme [12] . The mechanisms that detect TG1–3 tract length to control replication timing are likely to overlap with mechanisms that ensure preferential elongation of short telomeres . Indeed , the Rif1 protein is already implicated in both pathways . S . cerevisiae Rif1 binds to the TG1–3 repeat recognition factor Rap1 and with Rif2 regulates telomerase recruitment in response to telomere length [13] , [14] . Rif1 and Rif2 appear to ‘count’ the telomeric repeats and repress telomerase recruitment if the TG1–3 tract does not require extension . Cells lacking either Rif1 or Rif2 have abnormally long telomeres due to uncontrolled lengthening by telomerase [13] . The molecular mechanism by which Rif1 represses telomerase recruitment is still under investigation . Long and short telomeres bind similar amounts of Rif1 [15]; one proposal is that molecular modifications occurring selectively at short telomeres may relieve the repressive effect of Rif1 on telomerase recruitment [16] . For Rif2 , number of molecules may determine the repressive effect on telomerase , since more Rif2 molecules are present at long than short telomeres [17] . As well as acting in the pathway that recognizes short telomeres for lengthening , Rif1 is involved in controlling telomere replication time in response to length [4] . Specifically , in cells lacking Rif1 the link between telomere length and replication time is broken , since the telomeres of a rif1Δ mutant replicate early despite being abnormally long . Recently , Rif1 has been implicated as a regulator of replication timing more generally , having a repressive effect on genome-wide DNA replication mediated through recruitment of Protein Phosphatase 1 [18]–[23] . Tel1 , a PIK ( phosphatidylinositol 3-kinase ) -related checkpoint kinase , is involved in short telomere recognition . Tel1 binds to short telomeres and contributes to their preferential recruitment of telomerase [15] , [24]–[26] . Tel1 is recruited by interacting with the C-terminus of Xrs2 , a subunit of the MRX ( Mre11-Rad50-Xrs2 ) nuclease complex , which is also enriched at short telomeres . The kinase activity of Tel1 is required for its role in telomere maintenance [27] . Potential targets for Tel1 phosphorylation at telomeres include Xrs2 , Mre11 [28] and the telomeric single-stranded binding protein Cdc13 [29] , but it is unclear whether phosphorylation of these targets is important for telomere lengthening [30]; discussed in [16] . There is some overlap in function between Tel1 and Mec1 , the other yeast PIK checkpoint kinase , but while telomeres in tel1Δ mutant cells are extremely short , lack of Mec1 causes only a mild telomere length defect [31] , [32] . In general , Tel1 seems to play the primary role in regulating telomere function while Mec1 is the major checkpoint kinase . Since Tel1 is preferentially recruited to short telomeres , we investigated whether Tel1 is also involved in the pathway that detects short telomeres to specify early replication . We show here that Tel1 is required to drive the early replication of short telomeres , and that it acts upstream of Rif1 in the pathway that controls telomere replication timing . We tested whether Tel1 phosphorylates Rif1 , and identified two SQ ( i . e . Tel1 consensus ) sites that are preferentially phosphorylated in a short telomere mutant . Phosphorylation of one of the sites , Serine-1308 , is completely dependent on the presence of Tel1 . Mutation of these sites did not prevent the early replication of short telomeres , suggesting that Rif1 phosphorylation is not the sole mechanism through which Tel1 drives early replication . Our results are consistent with a model in which Tel1 that is recruited to short telomeres counteracts the repressive effect of Rif1 on replication initiation at nearby origins , to promote early origin activation and advance the replication time of short telomeres . To investigate the mechanism linking telomere length with replication timing , we examined the role of Tel1 , since this kinase is implicated in telomere length detection . tel1Δ cells have very short telomeres as shown in Fig . 1A , due to defective telomerase recruitment . If telomere replication time is still correctly coupled to TG1–3 tract length in this mutant , we would expect the short telomeres of a tel1Δ strain to replicate early—like the telomeres of a yku70Δ mutant , which replicate earlier than normal because they are short [4] . Replication time can be measured using the dense isotope transfer method , in which cells blocked in G1 phase with α-factor are transferred from isotopically dense to light medium . Upon release into S phase the transition of specific sequences from heavy∶heavy to heavy∶light DNA fractions on cesium gradient centrifugation is then monitored . Replication kinetics of particular sequences are plotted ( Fig . 1B ) , and replication time assigned as the time at which half the final level of replication has occurred . Since kinetics of α-factor release show some variability between experiments , the replication program can be usefully summarized using ‘replication indices’ ( Fig . 1C ) , with the various replication times normalized to early and late-replicating marker sequences ( ARS305 and Chr XIV-int respectively ) [33] . Replication times plotted relative to ARS305 are shown in Fig . S1 . In wild-type cells , the subtelomeric Y′ repeat sequences ( indicative of average telomeric replication ) replicate late in S phase ( Fig . 1B; top panel , solid line with filled circles , Fig . 1C&S1 , filled circle ) , 3 . 4 min later than the internal late replication origin ARS1412 [33] , [34] . In the yku70Δ mutant that has shortened TG1–3 repeat sequences , the Y′ sequences replicate much earlier , at a similar time to early origin ARS305 ( Fig . 1B , C&S1 ) [4] , [8] . Examining replication kinetics in a tel1Δ mutant strain revealed that Y′ sequences replicated late , close to their normal replication time ( Fig . 1B; third panel from top & Fig . 1C ) . Telomere-proximal sequences ( ARS522 and proARS1202 ) show a similar trend ( solid curve with filled diamonds and solid curve with filled triangles , respectively; Fig . 1B ) , so that overall the replication program of the tel1Δ mutant resembles that of wild-type cells ( Fig . 1C ) . Since the tel1Δ mutant has very short telomeres ( even shorter than those of yku70Δ; Fig . 1A ) this result suggests that in the absence of Tel1 kinase , the replication time of telomeric regions is uncoupled from telomere length . Telomeres of a yku70Δ mutant are short due to defects in telomere capping and extension . yku70Δ cells can detect telomere length status—since yku70Δ telomeres replicate early and Tel1 is correctly recruited to the short telomeres of a yku70Δ strain [35] . Early telomere replication in a yku70Δ mutant appears to result from telomere shortness , since restoring telomeres to wild-type length in a yku70Δ background leads to recovery of normal , late telomere replication [4] . The yku70Δ mutant therefore provides a convenient tool to investigate the controls linking telomere length to replication control . Note that the strength of the effect on telomere length of the yku70Δ mutation differs in A364a ( used for timing replication in this study ) and BY4741 yeast strain backgrounds ( Fig . S2 ) . Strain dependence of the effect of yku70Δ on telomere length was previously observed ( compare [36] , [37] with [38]–[40] ) . The reason for the strain dependence is not known , but the effects of the yku70Δ mutation on replication timing appear similar regardless of whether the effect on telomere length is weak or strong [8] , [41] , [42] . To understand whether Tel1 is required to transmit the signal for the early replication of short telomeres , we examined the replication program of a yku70Δ tel1Δ double mutant . This mutant has extremely short telomeres similar to a tel1Δ mutant ( Fig . 1A ) , but subtelomeric ( Y′ ) and telomere-proximal ( ARS522 and proARS1202 ) sequences replicated later than in yku70Δ single mutant , with replication timing similar to that in wild-type or tel1Δ cells ( Fig . 1B , C&S1 ) . While precise replication times and order show some variability between experiments [33] , repetition of these experiments confirmed the general trends ( Fig . S3 ) . Overall , these results suggest that in the absence of Tel1 , yku70Δ telomeres are no longer sensed as short and hence not replicated early , implying that Tel1 is involved in specifying early replication of short telomeres . We cannot exclude the possibility that effects shown above result from mutant phenotypes unrelated to telomere length . For example , the effect on replication timing of telomere uncapping in yku70Δ has not been tested . We therefore examined whether Tel1 promotes early telomere replication using an alternative mode of telomere shortening . We utilized a yeast strain in which a short telomere can be created by induction of HO endonuclease in cells blocked in G1 phase , as illustrated in Fig . 2A and similar to the construct previously described [43] . In this system , an HO cut site close to the left end of chromosome VII is flanked by short ( 80 bp ) and long ( 250 bp ) TG1–3 tracts on its centromere- and telomere-proximal sides respectively . Cutting with HO endonuclease in G1-blocked cells creates a single shortened telomere which , following release into S phase , stimulates earlier initiation at the neighboring , normally late-replicating origin ARS700 . 5 . ARS700 . 5 is located 18 kb from unmodified telomere VII-left and lies 5 . 3 kb from the HO cut site in this construct [Cooley & Bianchi , personal communication] . In a small-scale experiment we found that HO cutting levels exceeded 67% 5 . 5 hr after galactose addition , confirming that short telomere induction occurred in the majority of cells ( Fig . S4A ) . When S . cerevisiae cells attempt S phase in the presence of the replication inhibitor hydroxyurea ( HU ) , early origins are activated but late origin initiation is repressed by the Rad53-mediated S phase checkpoint [44]–[46] . Two-dimensional gel analysis of origin activation levels [47] after release into hydroxyurea therefore provides a proxy for differences in origin initiation time . The short telomere was induced by HO cutting and cells were then released into HU-containing medium ( during which cutting levels reached 90%; Fig . 2B ) . At the control early-initiating replication origin ARS305 , 2-dimensional gel analysis revealed strong bubble arcs in both TEL1 and tel1Δ strains ( Fig . 2C , upper panels ) . In contrast only low levels of replication intermediates were observed at the control late origin ARS1412 ( Fig . 2C , lower panels ) , due to checkpoint-mediated late origin repression . At ARS700 . 5 close to the induced short telomere , a strong bubble arc was observed in the TEL1 strain , consistent with stimulation of early ARS700 . 5 initiation as expected . Bubble arc intensity was however substantially reduced in the tel1Δ mutant ( Fig . 2C , middle panels ) , revealing that Tel1 is needed to drive early , checkpoint-resistant initiation of ARS700 . 5 following nearby short telomere induction . Quantitation of the bubble arc signal ( as shown in Fig . S4B&C ) revealed 4 . 8-fold-reduction in bubble arc intensity at ARS700 . 5 in the tel1Δ strain . In a construct with ARS700 . 5 placed proximal to a long telomere repeat a bubble arc was almost undetectable ( Fig . S5 ) , confirming that early activation of this origin depends on the nearby induced short telomere . Our 2-dimensional gel analysis therefore confirmed that after nearby short telomere induction , the absence of Tel1 changes the character of ARS700 . 5 from that of an early-initiating origin to that of a late replication origin . The results were therefore consistent with the replication timing analyses in Figs . 1 & S3 , showing that Tel1 is required to specify early replication of chromosomal regions in proximity to a short telomere . We also attempted to use isotope labeling-based replication timing analysis to examine ARS700 . 5 replication following short telomere induction , but inefficient and variable HO cutting after growth in the minimal medium required for this technique prevented satisfactory analysis of replication timing . Rif1 is implicated in the control of replication timing in response to telomere length , since in a rif1Δ mutant the link between telomere length and replication time is uncoupled . Specifically , in a rif1Δ mutant the TG1–3 tracts are over-extended ( Fig . 3A ) , but cells fail to detect the length of their telomeres and replicate them inappropriately early ( Fig . 3B , C & Fig . S6A ) [4] . Consistently , ARS700 . 5 initiates prior to the S phase checkpoint in a rif1Δ mutant with an induced short telomere ( Fig . S6B ) . Early replication of the long rif1Δ telomeres presents an interesting reversal of the effect in tel1Δ , where cells fail to detect the shortness of their telomeres and replicate them inappropriately late ( Fig . 1 ) . The opposite nature of these phenotypes implies that Tel1 and Rif1 have opposing actions in the pathway that controls telomere replication timing , with Rif1 enforcing the late replication of long or normal length telomeres , while Tel1 signals early replication of telomeres that are shortened . Loss of Rif1 impacts replication timing of many genomic regions [48] with subtelomeric regions most strongly affected [4] , probably because telomeres are the main genomic Rif1 binding locations [18] . To test the relationship of Tel1 and Rif1 in the telomere replication timing control , we examined a rif1Δ tel1Δ double mutant . Deleting RIF1 somewhat relieves the short telomere phenotype of tel1Δ ( Fig . 3A ) , presumably reflecting an effect of Rif1 on the backup mechanisms that recognize critically short telomeres in the absence of Tel1 [24] . We tested whether the rif1Δ tel1Δ strain replicates its telomeres early ( as in rif1Δ ) or late ( as in tel1Δ ) . We found that in rif1Δ tel1Δ cells , both Y′ and telomere-proximal sequences replicate very early , similar to their replication time in a rif1Δ single mutant ( Fig . 3B & C; replication times shown in Fig . S6 ) . The rif1Δ mutation is therefore epistatic to tel1Δ in control of telomere replication—consistent with the idea that Tel1 counteracts Rif1-mediated delay to telomere replication timing . Since Tel1 is actively recruited to shortened telomeres , we hypothesized that Tel1 may act to prevent or ‘switch off’ the delaying effect of Rif1 on nearby replication origins . The Rif1 protein sequence contains multiple S/TQ motifs , corresponding to the consensus sequence for Tel1-mediated phosphorylation [27] , [29] , so Rif1 is a potential target for Tel1 kinase activity . We therefore tested by mass spectrometry whether Tel1 phosphorylates Rif1 . Using Myc-tagged Rif1 that retains almost complete protein functionality ( as assayed by telomere length , Fig . 4A ) , we devised an immunoprecipitation procedure to pull down the majority of cellular Rif1 ( Fig . 4B ) . Initial high-resolution mass spectrometry identified multiple phosphorylated peptides in Rif1 from both YKU70 and yku70Δ strains , including two phosphorylation sites corresponding to Tel1 consensus sequences , one at Serine-1308 ( within the sequence…KVDSQDIQ… ) and the other at Serine-1351 ( …MNSSQQE… ) ( Fig . 4C ) . Rif1 S-1308 phosphorylation is not previously described; while S-1351 was identified as phosphorylated in response to DNA damage by MMS [49] . Identification of these phosphorylated SQ sites suggests that Rif1 may indeed be a target for Tel1 kinase , perhaps specifically at shortened telomeres which recruit Tel1 . We used the comparative proteomic method of SILAC to compare phosphorylation levels in wild-type cells with the short telomere yku70Δ mutant . Phosphorylation at both sites were increased in the yku70Δ mutant , by about 16-fold at S-1308 , and about 4-fold at S-1351 ( Fig . 4D–G; Dataset S1 ) . The corresponding unphosphorylated peptides were not increased in the yku70Δ strain ( Fig . S7; Dataset S1 ) . These results show that phosphorylation of these Rif1 SQ motifs is increased in the shortened telomere context of yku70Δ . To address whether Tel1 kinase mediates phosphorylation of Rif1 at S-1308 and S-1351 , we used a similar SILAC strategy to test whether the phosphorylation levels are decreased when Tel1 is unavailable . This experiment was carried out in the yku70Δ background where peptides containing phosphorylated S-1308 and S-1351 residues are reliably detected . Peptides from heavy-labeled yku70Δ tel1Δ cells were compared with those from light-labeled yku70Δ cells . The S-1308 phosphorylated peptide was abundant in the yku70Δ mutant , but was 10-fold reduced in the yku70Δ tel1Δ strain ( Fig . 5A; Dataset S2 ) . A longer peptide covering the same phosphorylated S-1308 residue was also greatly reduced in yku70Δ tel1Δ ( Fig . S8A; Dataset S2 ) , while its unphosphorylated equivalent showed no significant change ( Fig . S8C; Dataset S2 ) . In contrast , levels of the S-1351 phosphorylated peptide were largely unchanged in yku70Δ tel1Δ when compared to yku70Δ ( Fig . 5B ) . Based on this analysis , we propose that Tel1 directly phosphorylates S-1308 . However , we cannot exclude the possibility that Tel1 activates a different SQ-directed kinase that phosphorylates Rif1 S-1308 at short telomeres . If Tel1 contributes to phosphorylation at S-1351 , its role can be substituted by a different kinase ( probably Mec1 since [49] showed S-1351 phosphorylation requires one or other of Mec1 and Tel1 ) . Alternatively , Mec1 may be solely responsible for S-1351 phosphorylation . Fig . S9 provides a summary of the phosphorylation sites identified on Rif1 in these proteomic analyses . This study identified a cluster of phosphorylated DDK and CDK consensus sites close to the Rif1 N-terminus , which in a separate investigation were shown to regulate Protein Phosphatase 1 recruitment by Rif1 [21] . Our results suggest a model in which Tel1-mediated phosphorylation of Rif1 antagonizes the delaying effect of Rif1 on telomeric and telomere-proximal replication origins at short telomeres . We constructed a Rif1 mutant where the relevant serine residues are replaced by alanine , to test whether this non-phosphorylatable construct constitutively delays replication , preventing early replication of short telomeres . We replaced the serine or threonine residue with alanine at all seven of the potential Tel1 phosphorylation sites ( SQ and TQ motifs ) between 1308 and 1569 in the Rif1 amino acid sequence , to construct a rif1-7S→A allele . We mutated the entire cluster of S/TQ motifs since it could contain phosphorylation sites not detected proteomically and because preventing phosphorylation of one of these residues might re-direct kinase activity to a nearby consensus site . Telomere length was hardly affected by this rif1-7S→A allele , or by an phospho-mimetic equivalent rif1-7S→E glutamate substitution allele , in either YKU70 or yku70Δ backgrounds ( Fig . 6A ) —confirming that these substitutions do not ablate Rif1 protein function . We examined the replication program of the rif1-7S→A mutant in the short telomere ( yku70Δ ) background . We found that telomeres still replicate early ( Fig . 6B ) , with Y′ elements replicating at a similar time to the early marker sequence ARS305 ( Fig . 6C & S10 ) , equivalent to the yku70Δ mutant . The non-phosphorylatable RIF1 allele therefore does not prevent the early replication of short telomeres , implying that phosphorylation of the Rif1 S/TQ cluster is not essential for Tel1 to drive early replication of short telomeres . The rif1-7S→A mutation similarly caused minimal change to the replication timing program in a YKU70 background ( Fig . S11 ) . We also tested the replication program of the rif1-7S→E allele designed to mimic a phosphorylated form of Rif1 . In this rif1-7S→E mutant telomeres still replicate at approximately the same time as late origin ARS1412 ( Fig . S12 ) . The phenotypic analyses of the yku70Δ rif1-7S→A and rif1-7S→E mutant therefore suggest that Tel1-mediated phosphorylation of the Rif1 S/TQ cluster is not necessary or sufficient to drive early replication . They do not however exclude the possibility that Tel1-mediated Rif1 S/TQ cluster phosphorylation could contribute to early replication of short telomeres . Indeed a very slight advancement ( 3–4 min; Fig . S12 B&C ) in telomere replication time in the rif1-7S→E allele may be consistent with this idea . One possibility is that phosphorylation of Rif1 S-1308 , S-1351 and nearby S/TQ sites is integrated with other , redundant mechanisms to ensure that shortened telomeres replicate early . In investigating controls over telomere replication timing , we discovered that Tel1 specifies the early replication of short telomeres , as assessed either using a short telomere mutant ( yku70Δ; Fig . 1 ) or by analyzing origin activation close to an induced short telomere ( Fig . 2 ) . Rif1 specifies late replication of normal telomeres , and epistatic analysis indicated that Tel1 counteracts the delaying effect of Rif1 on telomere replication time . Phosphoproteomic analysis of endogenous S . cerevisiae Rif1 revealed at least two SQ motifs to be phosphorylated . Phosphorylation at these sites is increased in a short telomere mutant , with phosphorylation at Serine-1308 completely dependent on the presence of Tel1 . However , corresponding Rif1 alanine substitution mutants did not prevent early replication of telomeres in a yku70Δ background , indicating that phosphorylation of Rif1 by Tel1 at S-1308 , S-1351 , or nearby consensus sites within the Rif1 S/TQ cluster domain , cannot be the sole mechanism by which Tel1 drives early replication at short telomeres . While Rif1 phosphorylation could potentially contribute , Tel1 must mediate early replication of short telomeres through additional , possibly redundant , pathways . S . pombe , S . cerevisiae and human Rif1 proteins all negatively regulate DNA replication genome-wide [18]–[21] , and very recently it was shown that Rif1 recruits Protein Phosphatase 1 to control DNA replication [21]–[23] . The stimulatory effect of removing S . cerevisiae Rif1 on the overall replication program is reflected by a shortened S phase ( Fig . 3B & S6A: S phase duration is 21 . 5 min in wild-type but 15 . 3 min in rif1Δ ) . Within the generally shortened S phase of the rif1Δ mutant telomeres are more dramatically affected , with telomere-associated sequences shifting their replication time from the latter half to the early part of S phase ( Fig . 3C ) . Proximity of Rif1 binding sites has been suggested to determine the susceptibility of replication origin initiation to Rif1-mediated repression [18] , and the delaying effect of Rif1 on replication is probably focused at chromosome ends by the preferential association of Rif1 with telomeres , as illustrated in Fig . 7A , explaining why telomere regions show the largest shift in replication timing when Rif1 is removed ( Fig . 3C ) . It is possible that non-telomeric Rif1 also contributes to the late replication of telomere regions . Removing both Tel1 and Rif1 leads to a phenotype that is essentially equivalent to a rif1Δ single mutant—that is , in the absence of Rif1 , it becomes largely irrelevant for telomere replication timing whether Tel1 is present ( Fig . 3C ) . For this reason , our results support a model where Tel1 affects replication timing by counteracting the delaying action of Rif1 on telomere replication , as illustrated in Fig . 7A & B . If non-telomeric Rif1 contributes to late replication of subtelomeric regions , its effect is presumably also neutralized by Tel1 . We envisage two modes through which Tel1 could counteract the delaying effect of Rif1 on origin initiation . First , phosphorylation of Rif1 by Tel1 at SQ sites might ‘switch off’ the Rif1 repressive effect . We identified Rif1 as a target of Tel1 phosphorylation at shortened telomeres , but mutating the sites identified , along with neighboring potential phosphorylation sites , did not dramatically impact telomere replication timing . This observation argues that Rif1 phosphorylation cannot be solely responsible for Tel1-driven early telomere replication , while leaving open the possibility that Rif1 phosphorylation acts redundantly with other control mechanisms . It is possible that Rif1 contains additional functionally critical Tel1 phosphorylation sites not identified by our proteomic analysis . It is also conceivable that phosphorylation of Rif1 by Tel1 at non-consensus ( i . e . non-S/TQ ) sites might contribute to replication timing control . A previous study [28] showed that a Dun1 substrate lacking any SQ consensus was still phosphorylated by Tel1 kinase , and noted that ATM ( the mammalian homolog of Tel1 ) phosphorylates non-canonical sites in the tumor suppressor BRCA1 [50] . Intriguingly , in the yku70Δ mutant we observed a 2 to 4-fold increase in phosphorylation levels of five serine or threonine residues that are not followed by glutamine ( Rif1 S-1338 , S-1355 , S-1362 , T-1367 , and S-1694; Fig . S9 & Dataset S1 ) . Second , Tel1 could prevent the Rif1-mediated replication delay by phosphorylating a different telomeric protein . A number of telomeric proteins have been identified as likely or possible targets of Tel1 phosphorylation , including Cdc13 [29] , Xrs2 , and Mre11 [28] . While they cannot be formally excluded , none of these proteins is directly implicated in controlling replication origin activation . It seems more likely that Tel1 counteracts the Rif1-mediated delay by phosphorylating an unidentified component of the molecular pathway through which Rif1 restrains origin activation . Such a mechanism could act redundantly with Tel1-mediated Rif1 phosphorylation to neutralize the Rif1 replication-delaying signal . Tel1 appears to have multiple targets at telomeres [28] , [29] , which may act in concert to produce biological function , so that ablating any particular phosphorylation event has rather mild effects . A third possibility is that telomere replication timing control depends on multiple mechanisms some of which do not involve Rif1 , although the strong effect of Rif1 loss on telomere replication ( Fig . 3 ) does suggest it is the most central controller of telomere replication time . H2A-S129 phosphorylation depends on Tel1 in telomere-proximal regions [51] , and a non-phosphorylatable ( H2A-S129A ) allele caused a slight delay to telomere replication in a yku70Δ background ( unpublished observations ) ; however , H2A-S129 phosphorylation is not elevated at shortened telomeres [51] , inconsistent with H2A-S129 phosphorylation being a critical mediator of the early replication of short telomeres . Phosphorylation of Rif1 may contribute to other telomeric functions . One possibility is that Tel1-mediated Rif1 phosphorylation counteracts repression of telomerase recruitment , favoring TG1–3 tract extension . Telomere length is not greatly altered by the rif1-7S→A or rif1-7S→E mutants ( Fig . 6A ) —although very slight telomere lengthening in some rif1-7S→E isolates hints that Rif1 phosphorylation might contribute to telomerase recruitment . As with replication timing , Rif1 phosphorylation may be one of a series of redundant mechanisms through which Tel1 regulates telomerase recruitment—another potential pathway being phosphorylation of Cdc13 [16] . A further role for Rif1 phosphorylation might involve regulation of anti-checkpoint function at telomeric DNA damage sites [43] , [52] . To summarize , we have identified an important new function for Tel1—namely , driving the early replication of shortened telomeres . Our results suggest that Tel1 exerts this function by neutralizing the delaying effect of telomeric Rif1 on nearby replication origins . Tel1 also directs phosphorylation of Rif1 , which may contribute to replication timing control along with other mechanisms that impact on origin initiation time . Since Rif1 and Tel1 are conserved and play similar roles in replication timing control and coordination of DNA repair in higher eukaryotes as in yeast , our discoveries are likely to illuminate general functions of these proteins . Yeast strains are listed in Supplemental Table S1 . Gene knockouts and tagging used standard PCR-based insertion methods , confirmed by PCR analysis; see Supplemental material ( Text S1 ) for details of specific strain constructions . Primer sequences are available on request . The replication time of specific sequences was measured using the dense isotope transfer procedure [53] , [54] in cells released from α-factor at 30°C , probing for genomic EcoRI fragments as described previously [8] . Inducible HO cut strains were initially grown in YP medium containing 2% raffinose with 0 . 01% glucose ( to allow adaptation to raffinose ) , and then grown for 24–48 hours in 2% raffinose at 30°C before blocking with 200 nM α-factor . Then galactose was added to obtain a final concentration of 4% , to induce HO endonuclease . After 5 . 5–6 hr , the cells were then released by the addition of pronase with simultaneous addition of 200 mM hydroxyurea , and harvested 2 h later . DNA was prepared using the NIB-n-grab method [55] digested with HindIII followed by 2-dimensional gel electrophoresis under standard conditions [47] . HO cutting efficiency was confirmed by Southern blot analysis of XmnI-digested DNA . Immunoprecipitation of Rif1 was carried out as described [56] with modifications as described in Supplementary Material ( Text S1 ) . Protein concentrations were estimated using the RCDC kit ( Bio-rad ) . Immunoprecipitated proteins were eluted in 1× SDS sample buffer ( Invitrogen ) with 5% 2-Mercaptoethanol . Cellular equivalent protein samples were separated by SDS PAGE ( Novex 8–16% Tris-Glycine gels , Precast; Invitrogen ) and wet blotted using 1× Towbin buffer with 10% Methanol onto PVDF membrane ( Hybond-P , GE Healthcare ) . Rabbit anti-Myc ( ab9106 , Abcam ) was used to detect epitope-tagged RIF1 , with secondary antibody AP-conjugated anti-Rabbit IgG ( S3731 , Promega ) . Detection substrate was CDP-Star ( Perkin Elmer ) using Medical X-ray ( Fuji ) film . For quantification of amount of Rif1 protein , a similar gel was stained overnight using SyproRUBY total protein staining solution ( Bio-rad ) and quantified with a Fuji Phosphorimager ( FLA3000 ) at 473 nm with O580 filter and FujiFILM ImageGauge ( software V4 . 21 ) . SILAC samples were prepared based on the procedure described [57] . To compare yku70Δ with wild-type ( Fig . 4 ) , yeast strain AYS30 was labeled with heavy L-ARGININE:HCL ( U-13C6: U-15N4; CNLM-539-H; Cambridge Isotope Laboratory ) and L-LYSINE:2HCL ( U-13C6; U-15N2 , CNLM-291-H; Cambridge Isotope Laboratory ) [R10K8] and ASY25 was labeled with light alternatives [R0K0] for at least ten generations . To compare yku70Δ tel1Δ with yku70Δ ( Fig . 5 ) , ASY46 was labeled with heavy Lysine and Arginine [R10K8] and ASY30 was labeled with light alternatives [R0K0] for at least ten generations , and subjected to immunoprecipitation as described above . Immunoprecipitated Rif1 was quantified by SYPRORuby staining . Equal masses of Rif1 were then mixed and run on a Novex 8–16% Tris-Glycine gel , and the Rif1 band was excised for analysis by high-resolution mass spectrometry ( FingerPrints Proteomics , University of Dundee ) as described in Supplementary Information ( Text S1 ) . Genomic DNA was digested with XhoI , separated on a 1 . 5% agarose gel and transferred to neutral membrane ( MP Biomedicals ) by Southern blotting . Terminal restriction fragments were detected using a probe directed against the TG repeats .	The ends of chromosomes are protected by specialized structures called telomeres , which prevent their recognition as DNA breaks and enable recruitment of telomerase , the reverse transcriptase that maintains telomere length by replacing terminal TG-repeat sequences lost during successive rounds of DNA replication . Chromosomal DNA is replicated from initiation sites called origins , which are activated in a reproducible temporal sequence . Replication origins close to telomeres are subject to specialized temporal control that contributes to telomere stabilization: origins close to normal-length telomeres initiate replication late , while those close to shortened telomeres initiate early . Here we uncover the control mechanism that links telomere length with replication timing . Rif1 , one of the components of telomeric chromatin , directs late replication of normal telomeres by delaying the activation of nearby origins . Our experiments show that a kinase called Tel1 , which is recruited to shortened telomeres , neutralizes the origin-delaying activity of Rif1 . We also find that Tel1 phosphorylates Rif1 at short telomeres , although our investigation shows this phosphorylation is not the sole mechanism through which Tel1 prevents Rif1-mediated replication delay . Since correct telomere replication timing control is important for telomerase-mediated length maintenance , this discovery represents an important step towards understanding the molecular mechanisms that ensure proper long-term stabilization of chromosome ends , as well as the controls over the DNA replication temporal program .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "molecular", "cell", "biology", "cell", "biology", "biology", "and", "life", "sciences", "chromosome", "biology" ]	2014	At Short Telomeres Tel1 Directs Early Replication and Phosphorylates Rif1
Cockayne syndrome is an inherited premature aging disease associated with numerous developmental and neurological defects , and mutations in the gene encoding the CSB protein account for the majority of Cockayne syndrome cases . Accumulating evidence suggests that CSB functions in transcription regulation , in addition to its roles in DNA repair , and those defects in this transcriptional activity might contribute to the clinical features of Cockayne syndrome . Transcription profiling studies have so far uncovered CSB-dependent effects on gene expression; however , the direct targets of CSB's transcriptional activity remain largely unknown . In this paper , we report the first comprehensive analysis of CSB genomic occupancy during replicative cell growth . We found that CSB occupancy sites display a high correlation to regions with epigenetic features of promoters and enhancers . Furthermore , we found that CSB occupancy is enriched at sites containing the TPA-response element . Consistent with this binding site preference , we show that CSB and the transcription factor c-Jun can be found in the same protein-DNA complex , suggesting that c-Jun can target CSB to specific genomic regions . In support of this notion , we observed decreased CSB occupancy of TPA-response elements when c-Jun levels were diminished . By modulating CSB abundance , we found that CSB can influence the expression of nearby genes and impact nucleosome positioning in the vicinity of its binding site . These results indicate that CSB can be targeted to specific genomic loci by sequence-specific transcription factors to regulate transcription and local chromatin structure . Additionally , comparison of CSB occupancy sites with the MSigDB Pathways database suggests that CSB might function in peroxisome proliferation , EGF receptor transactivation , G protein signaling and NF-κB activation , shedding new light on the possible causes and mechanisms of Cockayne syndrome . Cockayne syndrome is a devastating inherited disease in which patients have features of premature aging , display increased sun sensitivity , and suffer from profound neurological and developmental defects [1] . Mutations in the gene encoding the CSB ( Cockayne syndrome complementation B ) protein are associated with the majority of Cockayne syndrome cases . CSB belongs to the SWI2/SNF2 ATP-dependent chromatin remodeling protein family [2] . ATP-dependent chromatin remodelers are conserved from yeast to human , and they are critical in regulating fundamental nuclear processes , such as transcription and DNA repair [3] , [4] . These proteins use ATP as energy to alter chromatin structure non-covalently , resulting in changes in nucleosome position , composition or conformation . By doing so , ATP-dependent chromatin remodelers can regulate the access of protein factors to DNA . Additionally , some ATP-dependent chromatin remodelers can assemble nucleosomes or create equally spaced nucleosomes to facilitate the formation of higher-order chromatin structure [5] . Most remodelers in isolation can alter chromatin structure in vitro . The additional proteins that form complexes with ATP-dependent chromatin remodelers are often involved in enhancing the specific activity of the remodeler or targeting the remodeling complex to specific genomic regions [6]–[8] . Additionally , some SWI2/SNF2 family members can alter contacts between DNA and non-histone proteins [9] . For example , the MOT1 remodeler can use the energy from ATP hydrolysis to dissociate TBP ( TATA box binding protein ) from DNA [9] . In vitro , the CSB remodeler appears to interact with DNA in a sequence-independent manner , and CSB displays both DNA and nucleosome stimulated ATP hydrolysis activity [4] , [10]–[12] . The biochemical activities that have been associated with CSB are quite diverse [2] . For example , CSB has been shown to alter DNA conformation and to actively wrap DNA [13] , [14] , both in an ATP-dependent manner . Additionally , CSB has been shown to use the energy from ATP hydrolysis to alter chromatin structure [8] , [13] . Most recently , the NAP1-like histone chaperone was shown to significantly enhance the chromatin remodeling activity of CSB and , together , these two proteins can centralize mononucleosomes , suggesting a potential function of CSB in nucleosome spacing [8] , [15] , [16] . Several ATP-independent activities have also been ascribed to CSB . Such activities include dissociating non-histone proteins from DNA , annealing single-stranded DNA , and modulating the activities of DNA repair proteins [17] . CSB is best known for its function in transcription-coupled DNA repair , a process that preferentially removes bulky DNA lesions that stall transcription , such as those created by UV irradiation [18] , [19] . CSB is one of the first proteins recruited to sites of lesion-stalled transcription , and the ability of CSB to hydrolyze ATP is essential for this association [20] . After its arrival to lesion-stalled transcription , CSB appears to have both chromatin remodeling-dependent and remodeling-independent functions . One function of CSB is to recruit the DNA repair machinery [20] , and this activity does not rely upon nucleosome repositioning by CSB [8] . The chromatin remodeling activity of CSB , on the other hand , is likely to be important for regulating protein DNA associations or nucleosome positioning necessary for the repair process or the resumption of transcription after repair . CSB is also involved in the repair of oxidative lesions in both nuclear and mitochondrial DNA [21]–[24] . In addition to its well-documented function in DNA repair , accumulating evidence indicates that CSB also participates in transcription regulation [25]–[28] . CSB has been found to be in complexes with both RNA polymerase II and RNA polymerase I . Interestingly , CSB does not appear to impact transcription in a general manner , as transcription-profiling experiments revealed that CSB has specific effects on gene expression . The results of that study has lead to the intriguing hypothesis that Cockayne syndrome might be , at least in part , a disease of transcription deregulation [28] , [29] . To understand better how CSB carries out its diverse functions and to gain potential insights into the underlying mechanism of Cockayne syndrome , we performed a genome-wide study of CSB occupancy to elucidate the mechanisms of CSB targeting and to understand the impacts of CSB occupancy on transcription regulation . To identify genomic regions that CSB occupies , we performed chromatin immunoprecipitation using a monoclonal antibody raised against CSB , followed by deep sequencing . CS1AN-sv cells that were reconstituted with CSB were used for these experiments , as CS1AN-sv cells do not express the alternatively spliced CSB-PiggyBac fusion protein ( Figure S1 ) [30]; this cell line is hemizygous for the CSB locus and the retained CSB allele has a premature stop codon at amino acid 337 [19] ( see Materials and Methods ) . Additionally , to understand the importance of CSB's nucleosome-remodeling activity in CSB function , we also included in these assays the remodeling-defective CSBΔN1 derivative ( Figure 1A ) [8] . The resulting sequencing reads were mapped to the human genome ( HG19 assembly ) using the Bowtie aligner [31] . Peaks were identified using HOMER ( Hypergeometric Optimization of Motif EnRichment ) with a default option ( FDR = 0 . 001 and Poisson p-value cut-off = 0 . 0001 ) on ChIPed samples against matching input samples . In total , we recovered 17 , 779 CSB peaks and 3 , 607 CSBΔN1 peaks ( Table 1 ) . We subsequently classified peaks unique to CSB or CSBΔN1 ( Figure 1B ) . To do this , we compared signal intensities for each CSB peak identified by HOMER to the CSBΔN1 signal intensity , regardless of whether the CSBΔN1 signal was identified as a peak by HOMER . Similarly , we compared signal intensities for each CSBΔN1 peak identified by HOMER to the CSB signal intensity , regardless of whether the CSB signal was identified as a peak by HOMER . Signal intensities were compared over a 200 bp region . If the difference between signal intensities was 4-fold or greater and the p-value for that difference was ≤0 . 0001 , the signal was classified as unique; the remaining signals were classified as common ( Tables S1 , S2 ) . Among them , we identified 6 , 398 peaks unique to CSB and 877 peaks unique to CSBΔN1 ( Table 1 ) . As shown in Figure 1B , ∼36% of the CSB occupancy sites were unique to CSB and ∼24% of the CSBΔN1 occupancy sites were unique to CSBΔN1 . We then classified the CSB and CSBΔN1 occupancy sites into seven functional categories ( see Materials and Methods ) , using the UCSC RefSeq gene annotations and the CEAS package [32] . The distribution of CSB peaks among those categories was largely consistent with a random distribution ( Figure S2 ) ; however , we observed modest enrichment of CSB occupancy at promoter regions ( 1 . 5% for CSB peaks vs . 1 . 1% for the genomic distribution , binomial test p-value of 7 . 3e-07 ) and modest depletion at 3′UTRs ( 1 . 1% for CSB peaks vs . 1 . 4% for the genomic distribution , binomial test p-value of 1 . 6e-05 ) . Interestingly , CSBΔN1 peaks displayed greater enrichment at promoter regions ( 3 . 1% for CSBΔN1 peaks vs . 1 . 1% for the genomic distribution , binomial test p-value of 6 . 8e-23 ) and 5′UTRs ( 1 . 1% for CSB peaks vs . 0 . 4% for the genomic distribution , binomial test p-value of 1 . 4e-09 ) ( Figure S2 ) . Given that CSBΔN1 is more significantly enriched at promoters and 5′ UTRs than CSB ( p-values of 5 . 2e-11 and 1 . 8e-7 , respectively ) ( Figure S2 ) , we calculated the number of CSB and CSBΔN1 peaks as a function of distance from transcription start sites ( TSS ) and plotted the results as a histogram using a bin size of 500 bp . As shown in Figure S3 , CSB displayed only a modest increase in occupancy around TSSs , while the remodeling-deficient CSBΔN1 derivative displayed greater occupancy around TSSs . These results are consistent with our functional genomic distribution analysis , which indicated the CSBΔN1 displayed a more significant enrichment at promoter regions ( Figure S2 ) . We next classified the genomic localization of the common and unique occupancy sites into the seven functional categories described above ( Figure S4 ) . We found that the common CSB and CSBΔN1 occupancy sites were over-represented at promoters and 5′UTRs as compared to the genomic distribution ( 2% vs 1 . 1% , p-value of 4 . 3e-20 and 0 . 7% vs 0 . 3% , p-value of 6 . 1e-8 , respectively ) . The occupancy sites unique to CSBΔN1 were slightly over-represented at promoter regions ( 3% vs 1 . 1% , p-value of 2e-06 ) , while the occupancy sites unique to CSB were similar to the genomic distribution . To gain further insight into the potential functions of CSB , we classified CSB and CSBΔN1 occupancy sites according to the 15 chromatin states defined by Ernst et al . ( 2011 ) , which are largely based on the presence or absence of specific histone marks ( Table 2 ) [33] , [34] . For this analysis , we used the classifications from normal lung fibroblasts , as CS1AN-sv cells are also a fibroblast cell line and , therefore , these two lines are most similar [34] . In agreement with the functional classification described above , CSB and CSBΔN1 displayed significant enrichment at transcribing promoters ( Table 2 ) : 2 . 7% of the CSB peaks ( p-value of 3 . 2e-58 ) and 7 . 3% of the CSBΔN1 peaks ( p-value of 4 . 1e-119 ) occupied active and weak promoters . Strikingly , sites of CSB and CSBΔN1 occupancy displayed a strong correlation with strong enhancers ( Table 2 ) . Moreover , while regions containing the H3K4me1 ( a mark often associated with enhancers ) represented only 4 . 4% of fibroblast chromatin , this histone mark was present at ∼19% of the CSB occupancy sites ( p-value of 7 . 9e-1141 ) and ∼26% of CSBΔN1 occupancy sites ( p-value of 1 . 7e-412 ) . Moreover , ∼29% of the top CSB peaks were associated with enhancer features ( p-value of 1 . 2e-486 ) . Taken together , these results are consistent with the notion that CSB is involved in the regulation of gene expression [25]–[28] . To gain insights into the molecular functions of genes that lie close to CSB occupancy sites , we searched for overlaps with the Molecular Signatures Pathways Database ( MSigDB ) using the Genomic Regions Enrichment of Annotations Tool ( GREAT ) . The top terms associated with occupancy common to CSB and CSBΔN1 involve the roles of epidermal growth factor receptor ( EGFR ) transactivation by G-protein coupled receptors ( GPCR ) , mechanism of gene regulation by peroxisome proliferators , G-alpha ( 12/13 ) signaling , and NFκB activation ( Figure 1C ) . To validate our ChIP-seq results , we selected seven regions to analyze by ChIP-qPCR ( Figure 2A ) . chr1-1 , chr2-2 , chr4-1 , and chr7-1 were four regions that were occupied by both CSB and CSBΔN1 . As negative control regions , we examined HES1 , chrX-1 , and chr17-1 . ChIP-qPCR confirmed that CSB is highly enriched at chr1-1 , chr2-2 , chr4-1 , and chr7-1 as compared to HES1 , chrX-1 , and chr17-1 as well as a “beads-only” control ( Figure 2B ) . Moreover , like CSB , CSBΔN1 was also enriched at the same four regions ( Figure 2B ) . Previously , we found that ATP hydrolysis by CSB is essential for the recruitment of CSB to UV-induced DNA lesions , a necessary and early step in the process of transcription-coupled DNA repair [35] . We , therefore , determined if the ATP hydrolysis activity of CSB is also important for its targeting to specific genomic regions in the absence of UV treatment . To accomplish this , we used a Cockayne syndrome-associated mutant CSB protein , CSBR670W , which contains a single amino acid substitution at position 670 [36] . This missense mutation , located within the ATPase domain of CSB , disrupts the ability of CSB to hydrolyze ATP [35] . As shown in Figure 2C , CSBR670W , like CSB , was targeted to chr1-1 , chr2-2 , chr4-1 , and chr7-1 , but not to the negative control regions ( HES1 , chrX-1 , and chr17-1 ) ( Figure 2C ) [35] . These results suggest that ATP hydrolysis by CSB is not critical for the recruitment of CSB to chromatin during replicative cell growth , in contrast to the recruitment of CSB to DNA lesion-stalled transcription upon UV irradiation . To better understand the mechanisms by which CSB is targeted to specific genomic regions during replicative cell growth , we used HOMER to determine if binding motifs for sequence-specific transcription factors were enriched at sites of CSB occupancy ( Figure 3A ) . This analysis revealed strong enrichment of the TPA ( 12-O-tetradecanoylphorbol-13-acetate ) -response element , TGASTCA ( where S denotes a G or C ) . This motif is most notable for binding AP-1 ( activator protein 1 ) transcription factors that consist of either Jun-Jun homodimers or Jun-Fos heterodimers ( Figure 3A ) [37] , [38] . We classified the CSB occupancy sites that contain TPA-response elements into the same seven functional categories and found that these sites are slightly over-represented at transcription termination sites ( TTSs ) ( 1 . 8% vs . 0 . 9% for a random genomic sequence , p-value 1 . 9e-05 ) . CSB occupancy sites containing TPA-response elements were also under-represented at 3′ UTRs ( 0 . 7% vs . 1 . 4% , p-value 9 . 3e-4 ) and exons ( 1 . 1% vs . 1 . 9% , p-value 6 . 1e-04 ) ( Figures S2 and S5 ) . Given that CSB does not bind to DNA in a sequence-specific manner [4] , [10] , [11] , but is enriched at regions containing the TPA-response element , this observation suggested that AP-1 transcription factors might target CSB to specific genomic loci . To test this hypothesis , we first determined if CSB could interact with c-Jun , the most potent transcriptional activator of the Jun protein family [37] , [38] . ChIP-western analysis was performed with cells that were pre-extracted to remove soluble CSB and c-Jun protein before formaldehyde cross-linking . As shown in Figure 3B , chromatin immunoprecipitation using an anti-CSB antibody revealed that c-Jun could be found together with either CSB or CSBΔN1 in the same protein-DNA complex . Co-immunoprecipitation analysis of unfixed cell lysates , in which the chromatin fraction had been removed by centrifugation , also revealed an association between soluble CSB and c-Jun , albeit weaker than that observed in the chromatin fraction shown by ChIP-western analysis ( compare Figure 3C to 3B ) . We next used shRNA-mediated RNA interference to directly examine the impact of c-Jun on the targeting of CSB to genomic regions containing a TPA-response element . As shown in Figure 4A , we were able to significantly reduce c-Jun protein levels through c-Jun shRNA expression . We used ChIP-qPCR to compare the recruitment of CSB in cells expressing c-Jun shRNA to cells expressing a control shRNA . As shown in Figure 4B , there was a dramatic decrease in CSB enrichment at regions containing the TGASTCA motif ( chr1-1 , chr2-2 , and chr4-1 ) or an AP-1-like motif TGAATCA ( chr7-1 ) in cells expressing c-Jun shRNA as compared to the control shRNA . On the other hand , there were no significant changes in the enrichment of CSB at the negative control regions ( HES1 , chrX-1 , and chr17-1 ) . To demonstrate that c-Jun does , indeed , occupy these four genomic regions ( chr1-1 , chr2-2 , chr4-1 , and chr7-1 ) , we performed anti-c-Jun ChIP followed by qPCR . As shown in Figure 4C , c-Jun was enriched at each of these loci . Taken together , these results indicate that CSB can be targeted to genomic sites containing a TPA-response element through an association with a c-Jun-containing AP-1 transcription factor . CSB-PGBD3 is a fusion protein that arises from alternative splicing between sequence encoding the N-terminal 465 amino acids of CSB with sequence encoding a piggybac transposase that lies within the fifth intron of the CSB gene [30] , [39] . The genomic sites of CSB-PGBD3 occupancy had been previously determined in UVSS1KO cells ( a CSB and CSB-PGBD3 null cell line ) that had been reconstituted with CSB-PGBD3 [39] . Given that both CSB and the CSB-PGBD3 fusion protein interact with c-Jun and are targeted to TPA-response elements , we determined the overlap in CSB and CSB-PGBD3 occupancy . To accomplish this , we used HOMER to analyze the published CSB-PGBD3 ChIP-seq data with the same parameters used to analyze the CSB ChIP-seq data , and we identified 1 , 590 CSB-PGBD3 ChIP-seq peaks . Of those peaks , 165 were common to CSB and CSB-PGBD3 ( Tables 3 and S3 ) . And , of the common peaks , 45% contained a TPA-response element . Gene ontology analysis using GREAT revealed the top three “Biological Processes” associated with common CSB and CSB-PGBD3 occupancy sites were tissue development , positive regulation of developmental processes and positive regulation of catecholamine secretion ( Table 4 ) . GREAT was also used to compare human genes enriched for CSB and CSB-PGBD3 to genes involved in mouse phenotypes . Of interest , the top three mouse phenotypes associated with common CSB and CSB-PGBD3 occupancy were decreased body weight , abnormal body weight and decreased bone marrow cell number ( Table 4 ) . Whether or not deregulation of the genes occupied by both CSB and CSB-PGBD3 contribute to the clinical features associated with Cockayne syndrome awaits further studies . We next determined the extent to which CSB occupancy can impact the expression of nearby genes . Using reverse transcription coupled with quantitative PCR ( RT-qPCR ) , we compared RNA expression levels of 10 genes , which displayed significant CSB occupancy , in CS1AN-sv cells expressing CSB to those in CS1AN-sv cells harboring an empty vector [8] . For these experiments , RNA expression was normalized to β-actin transcript levels . ZNFX-NC1 and MCPH1 are two examples of genes that were positively regulated by CSB ( Figure 5A ) . The ZNFX-NC1 gene expresses a noncoding RNA that is subsequently processed into three snoRNAs and is involved in cell proliferation and differentiation [40] , and our ChIP-seq results revealed that CSB binds to the first exon/intron junction of this gene . CSB is also associated with an intronic region of the microcephalin 1 ( MCPH1 ) gene , which encodes a DNA damage response protein that may be involved in neurogenesis and the regulation of cerebral cortex size [41] . RT-qPCR revealed that re-introducing CSB into CS1AN-sv cells increased the expression of ZNFX-NC1 and MCPH1 about two-fold relative to the vector-only control . Also shown in Figure 5A are examples of eight genes that are negatively regulated by CSB . ZNF507 and ZNF385B are two zinc finger proteins likely involved in transcription regulation [42] , [43] . MSANTD3 is a Myb/SANT-like DNA-binding domain containing protein and is associated with brain tumors [44] . MAD1L1 ( Mitotic Arrest Deficient-Like 1 ) is a component of the spindle assembly checkpoint [45] . SACM1L ( suppressor of actin mutations 1-like ) is a phosphoinositide phosphatase , which degrades phosphoinositides and plays a key role in signal transduction events [46] . PRMT5 is a protein arginine methyltransferase and plays a role in early development and pluripotency [47] . WDR74 ( WD Repeat Domain 74 ) likely plays an essential role in RNA transcription , stability , and/or processing [48] . DPP9 ( dipeptidyl peptidase 9 ) regulates signaling pathways that affect cell survival and proliferation [49] . Among these genes , CSB occupies the intronic regions of ZNF507 , MAD1L1 and ZNF385B , and the promoter regions of MSANTD3 , SACM1L , PRMT5 , WDR74 , and DPP9 . RT-qPCR analysis revealed that CSB decreased expression of these genes between 10% and 85% . Altogether , our results revealed that CSB can both positively and negatively regulate the expression of genes that lie adjacent to its occupancy sites . To identify , on a global level , potential direct transcriptional targets of CSB , we compared our ChIP-seq data to publicly available CSB microarray data [28] , [29] . CSB-dependent transcription profiling had been performed in both hTERT immortalized CS1AN fibroblasts ( as compared to SV40-transformed CS1AN cells used in our ChIP-seq study ) and SV40 transformed UVSS1KO fibroblasts . In the CS1AN/hTERT cells , Newman et al . ( 2006 ) identified 188 CSB up-regulated genes and 205 down-regulated genes . As shown in Table 5 and S4 , 37% of the up-regulated genes and 14% of the down-regulated genes were occupied by CSB; among these genes , 9% and 4% , respectively , were associated with TPA-response elements . In the UVSS1KO study , Bailey et al . ( 2012 ) identified 100 CSB up-regulated genes and 184 down-regulated genes [29] . As shown in Table 6 , 33% of the up-regulated genes and 34% of the down-regulated genes were occupied by CSB ( Tables 6 and S4 ) ; among them , 10% and 7% , respectively , were associated with TPA-response elements . Taken together , these analyses suggest that at least 25% of the CSB-mediated gene expression changes might directly result from CSB occupancy . The Bailey et al . ( 2012 ) study also examined the transcription profile of USS1KO cells coexpressing CSB and CSB-PGBD3 as well as USS1KO cells expressing CSB-PGBD3 alone . When we compared the CSB occupancy data with those transcription profiling data ( Tables 6 and S4 ) , we found that 22% of the genes up-regulated and 30% of genes down-regulated by co-expression of CSB and CSB-PDGB3 were associated with CSB occupancy , and 19% of the genes up-regulated by CSB-PGBD3 alone and 30% of the genes down-regulated by CSB-PGBD3 alone were associated with CSB occupancy . These observations are consistent with the hypothesis that CSB and CSB-PGBD3 may work together to regulate the transcription of certain genes [29] . Additionally , TPA-response elements were significantly enriched at CSB-occupied genes that were transcriptionally upregulated by both CSB and CSB-PGBD3 or by CSB-PGBD3 alone . To determine whether the remodeling activity of CSB is required for its function in transcription regulation , we used RT-qPCR to examine the effect of the remodeling defective CSBΔN1 protein on RNA expression [8] . For this analysis , we focused on genes that were occupied by both CSB and CSBΔN1 , as revealed by the ChIP-seq analysis . Western blot analysis using an antibody that recognizes the C-terminal region of CSB demonstrated that the levels of CSB or CSBΔN1 expression in the stable CS1AN-sv cell lines used for this analysis were similar ( Figure 5B ) . Additionally , immunofluorescence examination of those cell lines revealed that the number of cells expressing CSB or CSBΔN1 were similar ( >95% , data not shown ) . From this analysis , we found that expression of the ZNFX-NC1 gene , which was enhanced by CSB , was not enhanced by CSBΔN1 ( Figure 5A ) ; the presence of CSB resulted in a greater than two-fold increase in ZNFX-NC1 RNA levels , while cells expressing CSBΔN1 had ZNFX-NC1 RNA levels similar to that of CS1AN-sv cells harboring an empty vector . ChIP-qPCR confirmed that both CSB and CSBΔN1 were recruited to the ZNFX-NC1 locus ( Figure 5C ) . The other genes that we examined showed decreased RNA expression levels in response to CSB expression ( Figure 5A ) . Some of the transcript levels were also decreased by CSBΔN1 , but to different degrees ( Figure 5A ) . These results suggest that the chromatin remodeling activity of CSB is important for transcription regulation of some genes ( Figure 5A , groups ( 1 ) and ( 2 ) ) and dispensable for the regulation of others ( Figure 5A , group ( 3 ) ) . These results also reveal a chromatin remodeling-independent activity of CSB in transcription regulation . Of note , CSBΔN1 , unlike CSB , did not impact MCPH1 gene expression ( Figure 5A ) . Given that we did not see significant enrichment of CSBΔN1 at this locus ( data not shown ) , we cannot distinguish if the chromatin remodeling activity or another CSB-related function is important for up-regulating MCPH1 gene expression ( Figure 5A ) . To further demonstrate that the chromatin remodeling activity of CSB is important for regulating ZNFX-NC1 expression , we performed micrococcal nuclease ( MNase ) sensitivity assays . MNase is a nuclease that preferentially digests naked DNA and leaves nucleosomal DNA intact . ENCODE data obtained from the extensively studied K562 and GM12878 cell lines indicates that there are two MNase-resistant regions adjacent to the CSB occupancy site in the ZNFX-NC1 gene ( Figure 6A ) [33]: one MNase-resistant region is centered at position 47 , 895 , 320 on chromosome 20 and a less resistant region about 150 bp downstream ( Figure 6A ) . To determine the impact of CSB on local chromatin structure , we treated formaldehyde cross-linked cells with limiting amounts of MNase and isolated mononucleosomal DNA ( ∼150 bp ) ( Figure S6 ) . Quantitative PCR using primer sets that span the ∼670 bp region surrounding the CSB occupancy site were used to compare differences in MNase sensitivity among cells not expressing CSB ( CS1AN ) , expressing wild-type CSB , or expressing the remodeling-defective CSBΔN1 protein . All PCR reactions were normalized to naked genomic DNA . ChrX-1 and chr17-1 were used as two negative control regions , as they are not occupied by CSB . As shown in Figure 6A , cells expressing CSB or CSBΔN1 , as well as cells harboring an empty vector , had similar levels of amplicon enrichment from chr17-1 and chrX-1 , indicating similar MNase sensitivities . However , cells expressing CSB demonstrated a greater enrichment of amplicons 1 , 2 , and 3 than cells expressing CSBΔN1 or harboring an empty vector ( CS1AN ) . These results indicate that this region is more protected from MNase digestion in CSB expressing cells . On the other hand , enrichment of amplicons 4 and 5 was very similar between CSB and CS1AN , while CSBΔN1 displayed slightly less MNase resistance . Given that chromatin remodeling by CSB is required for transcription up-regulation of ZNFX-NC1 ( Figure 5A ) , our MNase-qPCR analyses suggest that CSB binds to the promoter of ZNFX-NC1 and enables the region that spans amplicons 1–3 to become more MNase-resistant through nucleosome repositioning or nucleosome assembly , which promotes transcription up-regulation of ZNFX-NC1 . We next examined the MNase sensitivity of a region of the PRMT5 promoter near a CSB occupancy site . Remodeling by CSB is partially required for suppression of PRMT5 expression ( Figure 5A ) . As depicted in Figure 6B , there are two MNase-resistant regions that lie on either site of the CSB occupancy site . The nucleosome to the right appears to be better positioned than the nucleosome to the left . As shown in Figure 6B , cells expressing either CSB or CSBΔN1 had similar sensitivity to MNase at the region covered by amplicons 1 and 2 , but both displayed greater resistance to MNase than CS1AN cells . The region covered by amplicons 3–5 appeared to be more sensitive to MNase , suggesting a relatively more open chromatin structure . Examining MNase sensitivity patterns at the regions covered by amplicons 6 and 7 , we found that cells expressing CSB showed greater MNase resistance to the region covered by amplicon 6 than cells not expressing CSB ( CS1AN ) . Interestingly , cells expressing CSBΔN1 demonstrated an intermediate enrichment of amplicon 6 ( less than cells expressing CSB but more than CS1AN cells ) . On the other hand , enrichment of amplicon 7 was very similar between cells expressing CSBΔN1 and CS1AN cells , while cells expressing CSB displayed less MNase resistance . Together , these observations suggest that CSB-expressing cells have a better-positioned nucleosome around amplicon 6 . Given that CSB suppressed PRMT5 more efficiently than CSBΔN1 ( Figure 5A ) , these results suggest that a better-positioned nucleosome at the region of the PRMT5 promoter covered by amplicon 6 may facilitate PRMT5 repression . We also examined MNase sensitivity near the site of CSB occupancy in an intron of the MAD1L1 gene , where several MNase-resistant regions were predicted based on ENCODE data [33] . RT-qPCR analysis revealed that the remodeling activity of CSB was partially required for suppression of MAD1L1 expression ( Figure 5A ) . As shown in Figure 6C , cells expressing CSB demonstrated a similar MNase resistance to cells expressing CSBΔN1 at the region covered by amplicons 2 , 3 , and 6 , but higher than CS1AN cells . Additionally , CSB-expressing cells displayed greater MNase sensitivity at the region covered by amplicons 4 and 5 than CSBΔN1 or CS1AN cells , which were similar to each other . These results suggest that increased nucleosome occupancy at the region of the MAD1L1 gene covered by amplicons 4–5 may facilitate MAD1L1 repression . Lastly , we examined MNase sensitivity at the promoter of the WDR74 gene , where remodeling by CSB was dispensable for CSB-dependent suppression of WDR74 expression . Cells expressing CSB or CSBΔN1 displayed very similar patterns of MNase sensitivity , agreeing with the results of our RT-qPCR analysis ( Figure 5A ) , which revealed that CSB and CSBΔN1 had a similar effect on WDR74 expression ( Figure S7 ) . Interestingly , the nucleosome structure at the promoter of the WDR74 gene appeared to be relatively more open , as the overall MNase-qPCR signals were lower than those obtained from the ZNFX-NC1 , PRMT5 and MAD1L1 genes ( Figure 6 ) . A more open nucleosome structure could account for , at least in part , our observation that suppression of WDR74 expression by CSB does not rely upon remodeling activity of CSB . Our results reveal that the mechanism that targets CSB to chromatin for transcription regulation is distinct from the mechanism that targets CSB during transcription-coupled DNA repair [35] . The association of CSB with specific genomic loci during replicative cell growth does not rely upon ATP hydrolysis by CSB ( Figure 2C ) ; however , ATP hydrolysis by CSB is essential for its targeting to sites of UV-induced DNA damage [35] . Therefore , stable chromatin association during transcription-coupled DNA repair appears to be an active process while stable chromatin association during transcription regulation appears to be a passive process . Current evidence suggests that during transcription-coupled DNA repair , ATP hydrolysis by CSB induces a conformational change that exposes a chromatin interaction surface , which is normally occluded by the N-terminal region of CSB [35] . Given that the association of CSB with chromatin in the absence of UV-induced DNA lesions does not rely upon ATP hydrolysis , this would suggest that the residues that mediate the c-Jun association are normally exposed . Recently , Gray et al . showed that the N-terminal region of CSB mediates the interaction between c-Jun and CSB-PGBD3 [39] . Based upon this observation and the knowledge that the CSB-PGBD3 fusion protein contains the N-terminal 465 amino acids of CSB , it is likely that the N-terminal 465 amino acids also mediates the association of full-length CSB with c-Jun . However , an interaction between full-length CSB and c-Jun was not detected in that study , although a robust interaction between CSB-PGBD3 and c-Jun was observed [39] . In agreement with that observation , we observed only modest association between endogenous CSB and c-Jun in chromatin-free cell lysates ( Figure 3C ) . However , a greater degree of association was observed when we specifically examined chromatin-bound proteins ( Figure 3B ) . These observations suggest that the CSB-c-Jun association may be preferentially established or stabilized on chromatin ( Figure 7 ) . CSB and CSB-PGBD3 both interact with c-Jun , and 45% of peaks common to CSB and CSB-PGBD3 contain a TPA-responsive element ( Table 3 ) [39] . These observations are consistent with results obtained from transcription profiling studies , in which it was observed that CSB and CSB-PGBD3 can co-regulate the expression of certain genes [29] . These results further suggest that AP-1 transcription factors might play a crucial role in modulating this co-regulation . Additionally , these results are also consistent with the notion that CSB regulates many genes independently of CSB-PGBD3 [29] . From comparisons between ChIP-seq and transcription profiling data , it can be seen that the majority of CSB occupancy sites are not associated with known CSB-responsive genes ( Tables 5–6 ) . This could arise from the different cell lines used in these studies and/or the different immortalization methods used to obtain these cell lines . It is also possible that some CSB-responsive genes were not covered by the microarrays used for the transcription profiling studies and , therefore , complementary approaches such as RNA-seq might offer additional insights , such as the influence of CSB on non-coding RNA expression . Furthermore , our analysis indicates a strong correlation between sites of CSB occupancy and chromatin regions that contain epigenetic signatures of enhancers ( Table 2 ) , and many of the CSB peaks ( 41% , p-value of 6 . 7e-1536 ) have DNase I hypersensitive sites lying within 100 bp , as judged by the digital DNase I hypersensitivity clusters in 125 cell lines [33] . These observations suggest that some of the intergenic CSB occupancy sites could function as enhancer elements that might lie at a great distance from their target genes . The composition of the dimeric AP-1 transcription factors that bind to TPA-response elements varies; for instance , there are three Jun and four Fos family members , and some of these members have variants that result from alternative splicing [37] , [38] . Future experiments will unveil the full extent to which different AP-1 complexes can target CSB and how this targeting might be modulated in different cellular contexts . AP-1 participates in a number of fundamental cellular processes , including cell proliferation and cell death . It is tempting to speculate that loss of CSB activity might , to some extent , compromise AP-1-mediated gene regulation , which in turn might contribute to the underlying mechanisms of Cockayne syndrome . Approximately 15% of the total CSB occupancy sites contain TPA-response elements . The mechanism that underlies the targeting of CSB to regions of the genome that do not contain TPA-response elements is not yet clear , but it is likely that CSB is delivered to or stabilized at these regions through associations with other DNA-binding proteins . A recent study examining the targeting of Isw2 in Saccharomyces cerevisiae revealed that this remodeler is primarily targeted to specific loci by sequence-specific transcription factors; however , more than half of the transcription factor-dependent occupancy sites did not contain a cognate binding motif [50] . Chromatin conformation capture suggested that DNA looping between regions that contain a transcription factor binding site with regions that do not is an integral component of the Iswi2 targeting mechanism [50] . Accordingly , DNA looping may also play an important role in the targeting of the CSB remodeler to sites that do not contain a TPA-response element . Interestingly , AP-1 , in conjunction with NFκB , was found to mediate DNA looping to regulate gene expression in macrophages [51] . Future studies examining CSB-containing protein complexes and higher-order chromatin structure will offer insights into other CSB targeting mechanisms . Our ChIP-seq data revealed that 36% of the CSB peaks were unique to CSB and 24% of the CSBΔN1 peaks were unique to CSBΔN1 ( Figure 1B and Table 1 ) . We do not yet know the reason underlying this difference . It is possible that some of the occupancy sites unique to CSBΔN1 might represent the initial sites of CSB binding to chromatin and that CSB would subsequently translocate away from these sites during chromatin remodeling , which in turn might contribute to some of the unique CSB peaks . Alternatively , but not mutually exclusive , some of the unique CSB peaks might represent targeting that relies upon functions related to the N1 region , which is deleted in the CSBΔN1 protein , such as mediating protein-protein interactions . Of interest , CSBΔN1 is more significantly enriched at promoters and 5′ UTR regions than CSB ( Figures S2 , S3 ) . By examining the effect of CSB on several genes that lie close to CSB occupancy sites , we provide evidence that CSB can directly influence local gene expression mediated by RNA polymerase II . CSB is a member of the SWI2/SNF2 ATP-dependent chromatin remodeling protein family and displays ATP-dependent chromatin remodeling activity in vitro; therefore , CSB likely repositions nucleosomes [8] , [13] and/or other protein factors [17] to regulate transcription . By mapping nucleosome positions at the ZNFX-NC1 , PRMT5 and MAD1L1 genes with MNase , we found that regions adjacent to the CSB occupancy sites are more resistant to MNase digestion in cells expressing CSB than in cells that do not express CSB or express the remodeling-defective CSBΔN1 protein . These results indicate that CSB alters chromatin structure in an ATP-dependent manner to regulate transcription ( Figures 6–7 ) . Furthermore , CSB appears to make the border of nucleosome-free regions more pronounced ( Figure 6A amplicon 2–3 and Figure 6B amplicon 6 ) , resembling the function of yeast Iswi2 in regulating the length of nucleosome-free regions to prevent cryptic transcription and regulate gene expression [52] . By creating a better-positioned nucleosome ( Figure 6C , amplicon 4–5 ) , CSB could also support either transcription activation or repression by preventing the binding of transcriptional repressors or activators , respectively , to the DNA occupied by the nucleosome [53] . Interestingly , in collaboration with NAP1-like histone chaperones , CSB has been shown to efficiently move histone octamers to the center of a DNA fragment in vitro [8] . It will be of great interest to investigate the function of NAP1-like chaperones in CSB-mediated transcription regulation . Additionally , the complete rescue of gene expression by the CSBΔN1 protein at certain genes ( e . g . WDR74 ) suggests additional CSB functions in transcription regulation; such functions could include protein recruitment through remodeling-independent mechanisms [12] , [54] or protein eviction [55] , which may not rely upon the N1 region . During replicative cell growth , approximately 10% of CSB associates with chromatin , and this likely represents the CSB population that participates in normal transcription regulation . However , in the presence of UV-induced DNA damage ( >25 J/m2 ) , approximately 90% of the CSB population can become stably associated with chromatin [35] . A fraction of these chromatin-associated CSB molecules would be stabilized at sites of DNA lesion-stalled transcription to participate in DNA repair . In addition , some of these CSB molecules would also be expected to localize to new transcriptional targets , as CSB has been implicated in UV-induced transcription regulation [17] . Additional ChIP-seq analysis of CSB in cells challenged with UV irradiation will reveal if the fraction of CSB that is used during normal transcription regulation is redistributed in response to UV irradiation , either for DNA repair or UV-induced transcription regulation . Taken together , the results of this study reveal that the CSB remodeler binds to specific regions of the genome to regulate chromatin structure and RNA polymerase II-mediated gene expression . These observations are consistent with the hypothesis that Cockayne syndrome might be , at least in part , a disease of transcription deregulation [28] , [29] , [56] . Moreover , the results of this study open up new avenues to explore the mechanisms that might contribute to the diverse features of Cockayne syndrome . CS1AN-sv cells were maintained in DMEM-F12 with 10% FBS . The CS1AN primary cells have mutations in both CSB alleles , but only one of these alleles was retained after SV40 immortalization [19]; the resulting CS1AN cell line is , therefore , hemizygous for CSB . The retained allele contains an A to T transversion at position 1088 , which introduces a premature stop codon at amino acid 337 . Accordingly , CSB-PGBD3 is predicted to be absent from the CS1AN-sv cell line , and our anti-CSB immunoprecipitation experiments agree with this prediction ( Figure S1 ) . Stable cell lines expressing CSB were generated by infecting CS1AN-sv cells with CSB-expressing lentivirus ( pLenti-PGK-Neo , Addgene ) [8] . Stable cell lines expressing CSB or harboring the empty vector were selected with 600 µg/ml G418 . CSBΔN1 was expressed from the pSVL vector [8] . CS1AN-sv cells stably expressing CSBΔN1 were generated by cotransfection with pLenti-PGK-neo . After selection with 600 µg/ml G418 , single colonies were cloned [8] . CSBR670W was expressed from MSCV-Puro . The stable cell line expressing CSBR670W was generated by transfection and selecting with 250 ng/ml puromycin [35] . Mission shRNA targeting c-Jun ( TRCN0000010366 , Sigma ) was used to decrease c-Jun protein levels . A non-targeting shRNA ( SHC002 , Sigma ) was used as a negative control . Virus was produced by cotransfecting a 10 cm plate of ∼90% confluent 293T cells with third generation lentivirus packaging plasmids ( pMGLg/pRRE , pRSV-REV , and pMD2 . G/VSV ) . A total of 20 µg of plasmid was transfected , with the individual plasmids at an equal molar ratio . The culture medium was changed 24 hours post-transfection , and virus-containing medium was collected 24 hours later . Medium from one plate of virus-producing cells was distributed to six 10 cm dishes of target cells: CS1AN-sv/CSB . The confluence of the target cells at the time of infection was approximately 20% . Infected cells were harvested 36–48 hours post-infection for RNA preparation and western blot analysis . Total RNA was prepared using TRIzol ( Invitrogen ) . AMV reverse transcriptase and random primers were used for first strand cDNA synthesis ( Roche ) . cDNA was analyzed by real-time PCR using a MyiQ thermal cycler and SYBR green ( BioRad ) . Expression was first normalized against β-actin and fold over vector-only control was then calculated using ΔΔCt method [57] . Primers used for RT-qPCR are listed in Table S6 . ChIP-qPCR assays were performed as previously described [12] . Primers used for ChIP-qPCR are listed in Table S5 . To increase ChIP efficiency we removed soluble CSB before cross-linking DNA and proteins [8] , [35] , [58] . Cells were collected in Buffer B ( 150 mM NaCl , 0 . 5 mM MgCl2 , 20 mM HEPES pH 7 . 8 , 10% Glycerol , 0 . 5% Triton X-100 ) and soluble CSB was separated from chromatin by centrifugation at 15 , 000 RPM for 5 min at 4°C . The resulting pellets were resuspended in Buffer B and fixed with 1% formaldehyde for 10 min at room temperature . Cross-linked cells were sonicated at 40% amplitude ( 30 sec on , 90 sec off , for 24 min total ) using the Branson 101-135-126 Sonifier . Chromatin IP ( ChIP ) was performed using a monoclonal anti-CSB antibody ( 1B1 ) that recognizes the N-terminal 507 amino acids of CSB [35] , [39] or an anti-c-Jun antibody ( Santa Cruz , sc-1694 ) . ChIP samples were reverse cross-linked in SDS sample buffer for subsequent western blot analyses [20] . Antibodies used for western blot analysis are rabbit anti-CSB antibodies ( kindly provided by Dr . Alan Weiner , U . Washington ) [35] , c-Jun ( Santa Cruz , sc-1694 ) and GAPDH ( Millipore , MAB374 ) . 10 ng of ChIPed DNA was used to prepare libraries for deep sequencing using the multiplexed ChIP-Seq sample preparation protocol described on the website of the Next-Generation Sequencing Core , Perelman School of Medicine , University of Pennsylvania ( http://ngsc . med . upenn . edu/ ) . The Next-Generation Sequencing Core at the University of Pennsylvania performed DNA sequencing using Illumina hiSeq2000 sequencers for single-end sequencing with a read length of 50 bps . The resulting sequencing reads were mapped to the human genome ( HG19 assembly ) using Bowtie version 0 . 12 . 7 . Peaks were identified using HOMER version 4 . 1 ( Hypergeometric Optimization of Motif EnRichment ) with a default option ( FDR = 0 . 001 and Poisson p-value cutoff = 0 . 0001 ) on ChIPed samples against matching input DNA samples . Raw and processed files ( GSE50171 ) have been deposited at the Gene Expression Omnibus ( GEO ) repository . ( http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? token=vfmfryagmygcony&acc=GSE50171 ) . To classify a peak as unique or common , we determined the intensities ( rpm ) of the CSB and CSBΔN1 signals within a 200 bp region around a peak center . If the difference between signal intensities was 4-fold or greater and the p-value for that difference was ≤0 . 0001 , the peak was classified as unique . The remaining signals were classified as common . CSB and CSBΔN1 peaks were classified as follows: ( 1 ) promoter ( from −1 kb to the transcription start site ) , ( 2 ) TTS ( from the transcription termination site to +1 kb ) , ( 3 ) 5′ UTR , ( 4 ) 3′ UTR , ( 5 ) exon , ( 6 ) intron , and ( 7 ) intergenic ( the remainder ) . The source of gene annotation was UCSC RefGene . The CEAS package [48] computes p-values using one-sided binomial test . To compute p-values for comparisons between CSB and CSBΔN1 , we considered a null model in which both classes of peaks form a single population . m and n are the total peak numbers of those classes , and m′ and n′ are the number of peaks in a specific annotation category , and a combined frequency f equal to ( m′+n′ ) / ( m+n ) . p-values for comparisons between CSB and CSBΔN1 are equal to the product of the two p-values from the one-sided binomial test for n , n′ and f , as well as for m , m′ and f . The Genomic Regions Enrichment of Annotations Tool ( GREAT , version 2 . 0 . 2 ) was used for pathway analysis of CSB occupancy sites , using the “MSigDB pathways' category” [59] . The assignment of peaks to genes was made using the following parameters: proximal 5 kb upstream , 1 kb downstream , plus distal up to 1000 kb . To compare the genomic occupancy of CSB to CSB-PGBD3 , we downloaded the ChIP-seq data for CSB-PGBD3 and the matching input from the GEO repository ( GSE37919 ) [39] . The CSB-PGBD3 peaks were called against input using HOMER , and we identified 1 , 590 peaks . Using binary peak calling ( +/−100 bp ) , we identified 165 peaks as common to CSB and CSB-PGBD3 , which represented 1% total CSB and 10% total CSB-PGBD3 peaks ( Tables 3 and S3 ) . The hypergeometric p-value was calculated for the AP-1 motif against total CSB as well as CSB-PGBD3 peaks ( Table 3 ) . The ontology of the nearby genes was obtained using GREAT ( Table 4 ) . CS1AN cells expressing CSB were lysed in Buffer B60 ( 20 mM HEPES ( pH 7 . 9 ) , 60 mM NaCl , 0 . 5 mM MgCl2 , 1 mM DTT , 0 . 5% triton X-100 and 20% glycerol ) with protease inhibitors ( 0 . 5 mM PMSF , 10 µM E64 and 3 µM pepstatin A ) and centrifuged at 15 , 000 rpm for 5 min at 4°C . The supernatant was used for co-immunoprecipitation assays . CSB-containing complexes were recovered using the monoclonal anti-CSB antibody 1B1 ( 1 hour at 4°C ) and protein G agarose beads ( 30 min at 4°C ) . The resulting immunocomplexes were washed four times in buffer B60 , and the immunoprecipitated proteins were eluted with Laemmli buffer . Beads-only control immunoprecipitations were conducted in parallel . Formaldehyde-fixed nuclei were isolated from each cell line as described previously [58] , [60] . Nuclei ( A260 = 500 ) were incubated with MNase ( final concentration 25 U/ml ) at 37°C for 10 min and reverse cross-linked at 65°C for 16 hours . After phenol-chloroform extraction , digested DNA was resolved on a 1 . 2% agarose gel in 1×TAE at 100 V for 3 . 5 hr . Mononucleosome-size DNA fragments were purified from gel slices and subjected to qPCR analysis ( Figure S6 ) . Primers used in MNase-qPCR assays are listed in Table S7 . CSB , top CSB ( rpm greater than 1 ) and CSBΔN1 peaks were classified using the epigenomic information derived from NHLF ( normal human lung fibroblast ) cells , via different chromatin features using chromHMM [34] , [61] . The region annotation used here is the result of unsupervised learning , finding an HMM with 15 states that minimizes the entropy of observed histone modifications , and afterwards interpreted using prior biological knowledge . Peaks were assigned to regions according to the location of the peak centers . p-values were calculated using a one-sided binomial test . To compare the CSB ChIP-seq data with the microarray data , we used the lists of up- and down-regulated genes generated by Newman et al . ( 2006 ) and Bailey et al . ( 2012 ) [28] , [29] . Newman et al . identified CSB-responsive genes in hTERT immortalized CS1AN cells . Bailey et al . ( 2012 ) identified CSB-responsive , PGBD3-responsive , and CSB+PGBD3-responsive genes in USS1KO cells . We determined the number of genes whose body or promoter regions overlapped with CSB peaks . p-values were calculated after randomly generating 17 , 779 peaks ( the same number of CSB ChIP-seq peaks ) . Hypergeometric p-values were calculated against the total number of CSB peaks for the c-Jun/Ap-1 motifs .	Cockayne syndrome is a devastating inherited disease , in which patients appear to age prematurely , have sun sensitivity and suffer from profound neurological and developmental defects . Mutations in the CSB gene account for the majority of Cockayne syndrome cases . CSB is an ATP-dependent chromatin remodeler , and these proteins can use energy from ATP-hydrolysis to alter contacts between DNA and histones of a nucleosome , the basic units of chromatin structure . CSB functions in DNA repair , but accumulating evidence reveals that CSB also functions in transcription regulation . Here , we determined the genomic localization of CSB to identify its gene targets and found that CSB occupancy displays high correlation to regions with epigenetic features of promoters and enhancers . Furthermore , CSB is enriched at genomic regions containing the binding site for the c-Jun transcription factor , and we found that these two proteins interact , uncovering a new targeting mechanism for CSB . We also demonstrate that CSB can influence gene expression in the vicinity of its binding sites and alter local chromatin structure . Together , this study supports the hypothesis that defects in the regulation of gene expression and chromatin structure by CSB might contribute to the diverse clinical features of Cockayne syndrome .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "biochemistry", "cell", "biology", "genetics", "biology", "and", "life", "sciences", "genomics", "molecular", "cell", "biology" ]	2014	The Sequence-Specific Transcription Factor c-Jun Targets Cockayne Syndrome Protein B to Regulate Transcription and Chromatin Structure
Replication fork stalling and collapse is a major source of genome instability leading to neoplastic transformation or cell death . Such stressed replication forks can be conservatively repaired and restarted using homologous recombination ( HR ) or non-conservatively repaired using micro-homology mediated end joining ( MMEJ ) . HR repair of stressed forks is initiated by 5’ end resection near the fork junction , which permits 3’ single strand invasion of a homologous template for fork restart . This 5’ end resection also prevents classical non-homologous end-joining ( cNHEJ ) , a competing pathway for DNA double-strand break ( DSB ) repair . Unopposed NHEJ can cause genome instability during replication stress by abnormally fusing free double strand ends that occur as unstable replication fork repair intermediates . We show here that the previously uncharacterized Exonuclease/Endonuclease/Phosphatase Domain-1 ( EEPD1 ) protein is required for initiating repair and restart of stalled forks . EEPD1 is recruited to stalled forks , enhances 5’ DNA end resection , and promotes restart of stalled forks . Interestingly , EEPD1 directs DSB repair away from cNHEJ , and also away from MMEJ , which requires limited end resection for initiation . EEPD1 is also required for proper ATR and CHK1 phosphorylation , and formation of gamma-H2AX , RAD51 and phospho-RPA32 foci . Consistent with a direct role in stalled replication fork cleavage , EEPD1 is a 5’ overhang nuclease in an obligate complex with the end resection nuclease Exo1 and BLM . EEPD1 depletion causes nuclear and cytogenetic defects , which are made worse by replication stress . Depleting 53BP1 , which slows cNHEJ , fully rescues the nuclear and cytogenetic abnormalities seen with EEPD1 depletion . These data demonstrate that genome stability during replication stress is maintained by EEPD1 , which initiates HR and inhibits cNHEJ and MMEJ . Maintaining genome stability depends on faithful DNA replication [1–3] . Since DNA damage from endogenous and exogenous sources creates barriers for the replication fork , replication is not a smooth , continuous process , but rather one of intermittent stress , with stops and restarts [4–6] . Replication fork reactivation after stalling at DNA damage is best characterized in E . coli , where forks are restarted by recombination-dependent or -independent pathways requiring RuvABC or the PriA/C complexes , respectively [5–7] . Eukaryotic replication fork restart is more complex and less understood , with the canonical repair pathway mediated by RAD51-dependent homologous recombination ( HR ) [1–3 , 8] . HR is best characterized for the repair of DNA double-strand breaks ( DSBs ) . It is initiated by a litany of components mediating 5’ end resection to create 3’ single-stranded ( SS ) DNA , which then use BRCA2/RAD51 to create heteroduplexes with homologous sequences on sister chromatids [3 , 4 , 8–12] . After an invading strand re-initiates DNA synthesis , Holliday junctions may be resolved by either Gen1 or Mus81 , with Slx4 serving as a scaffold [11–15] . End resection directs DSB repair toward HR , preventing the competing DSB repair pathway , classical non-homologous end-joining ( cNHEJ ) from occurring [16–19] . Similar to DSB repair , repair of stressed replication forks also requires 5’ end resection to initiate HR , but how this is regulated in fork repair and restart is less well defined [1–3 , 16 , 17] . End resection at a replication fork requires a free DNA double strand ( DS ) end structure to initiate 5’ exonuclease activity . This DNA DS end can be created at stressed forks in at least two ways: the fork can reverse into a chicken foot structure with a single DS DNA end [2 , 3 , 20] , or a nuclease can cleave the fork , directly creating a free DS end [3 , 13 , 14 , 17] . If a stressed fork is not repaired in timely manner , it may convert into toxic structures that make fork restart difficult [1 , 13 , 14 , 19] , leading to cell death or genome instability and neoplastic transformation [1 , 4 , 6] . Repair pathway choice at stalled forks is important for genome stability , because unopposed cNHEJ , as seen in malignancies with inherited deficiencies in HR proteins BRCA1 or BRCA2 , results in fusion of these DNA DS ends at damaged replication forks [21–26] . These chromosomal fusions cause severe genome instability , resulting in catastrophic mitoses revealed as gross nuclear abnormalities including nuclear bridges and micronuclei [1 , 21 , 22 , 25 , 27] . The tumor suppressor p53-binding protein 1 ( 53BP1 ) promotes cNHEJ at least in part by preventing end-resection . Preventing cNHEJ by repressing 53BP1 rescues HR-deficient cells from these nuclear defects [21–23] There is accumulating evidence that DSB pathway choice between cNHEJ and HR is mediated by 5’ end resection [16–18] . End resection appears to be a two-step process , with CtIP and Mre11 nucleases responsible for short end resection , and Dna2 and Exo1 catalyzing longer resection for HR [16 , 17 , 19 , 28 , 29] . It is thought that short end resection may lead to MMEJ and long range end resection to HR [17 , 19 , 30 , 31] . Although it is clear that end resection is important for regulating pathway choice at DSBs , key questions remain on how end resection is initiated at stressed forks . In this study we identify a previously uncharacterized 5’ endonuclease , EEPD1 ( endonuclease/exonuclease/phosphatase family domain-containing 1 ) , by its up-regulation in embryonic stem cells after DNA damage . We found that EEPD1 initiates end resection , thereby enhancing HR at the expense of cNHEJ , and also of MMEJ . Consistent with an upstream role in end resection , EEPD1 depletion markedly reduces stress-induced ATR and Chk1 phosphorylation and the formation of RPA , gamma-H2Ax , and RAD51 foci , while NBS1 , 53BP1 , and BRCA1 foci are intact . Depletion of EEPD1 results in severe chromosomal abnormalities , made worse by replication stress . This places EEPD1 at the apex of pathway choice in repair of stressed replication forks , where it is required for maintenance of genome integrity . In a survey of proteins induced by the topoisomerase IIα poison VP-16 in embryonic stem cells , we found that expression of EEPD1 , an uncharacterized human protein ( Uniprot Q7L989 , AAH65518 . 1 ) , was markedly increased . EEPD1 is a 569 aa protein with two amino terminal helix-hairpin-helix ( HhH ) DNA binding domains related to RuvA , a carboxy terminal DNase I-like domain that places it in the exonuclease-endonuclease-phosphatase ( EEP ) family , and a conserved D-D-N/D/E nuclease active site that overlaps with the HhH domain and the DNase I-like domain ( S1 Fig ) [32] . It is located at 7p14 . 2 , but is not involved in any known neoplastic translocations ( Catalogue of Somatic Mutations in Cancer ) . EEPD1 is evolutionarily conserved from some insects to humans and expressed at variable levels in a wide variety of primary human tissues and human cell lines ( S1B Fig ) . It is more highly expressed in the testis , leukocytes , and brain , as are many other DNA DSB repair components [33 , 34] . EEPD1 depletion moderately altered cell cycle progression in asynchronous or synchronized cells ( S2 Fig ) , increasing the fraction of cells in S and G2 phases in both situations . EEPD1 alone is required for proper clonogenecity; plating efficiency is reduced by almost 50% from EEPD1 depletion alone ( Fig 1A ) . EEPD1 deficiency also significantly slows cell growth ( Fig 1B ) , and increases the fraction of cells expressing cyclin A , without an increase in the fraction of cells with phosphorylated histone H3 ( Fig 1C and 1D ) . This suggested a potential role in DNA replication . To investigate whether EEPD1 is important for survival after exposure to agents that stress replication forks , we tested whether EEPD1 regulates sensitivity to VP-16 , hydroxyurea ( HU ) , camptothecin ( CPT ) , UV light , cisplatin , and ionizing radiation ( IR ) ( Fig 1E ) . EEPD1 depletion resulted in 3 . 5-fold less clonogenic survival after 18 h exposure to 10 uM VP-16 , compared to controls . EEPD1 depletion also decreased survival to continuous 0 . 4 mM HU ( 12-fold ) , 18 h exposure to 10 uM CPT ( 6-fold ) , continuous 0 . 4 mM HU ( 10-fold ) , 18 h exposure to 10 uM CPT ( 6-fold ) , continuous 5 uM cisplatin ( 4-fold ) , 15 J/m2 UV ( 12-fold ) , and 4 Gy IR ( 4-fold ) . To investigate the mechanism by which EEPD1 promotes cell survival during replication stress we used two techniques to measure replication fork restart after stalling . First , BrdU incorporation into nascent DNA after release from HU replication stress was measured by immunofluorescence ( [35 , 36] . By 2 h after release from an 18 h HU exposure , when replication fork restart was maximal in control cells ( as indicated by the number of BrdU foci ) , EEPD1-depleted cells restarting forks were reduced by 5-fold ( Fig 2A ) . This is a specific EEPD1 effect , as the fork restart defect in EEPD1 depleted cells was rescued by expression of an siRNA-resistant version of EEPD1 ( Fig 2A ) . We next used DNA fiber analysis to measure replication fork restart after release from a 1 h HU treatment , as well as replication speed and replication fork symmetry [32 , 35] . We found that 20 min after HU release , EEPD1 depletion reduced replication fork restart by 2 . 3-fold ( Fig 2B and 2C ) . Interestingly , over-expressing EEPD1 increased fork restart; however , by 30 min nearly all forks restarted even in EEPD1-depleted cells . New fork initiation is rare under these conditions , and EEPD1 depletion had no significant effect on this endpoint ( Fig 2C ) . By measuring fiber lengths , we determined that EEPD1 depletion significantly reduces replication speed ( Fig 2D ) . Consistent with EEPD1 promoting fork restart , EEPD1 depletion significantly reduced the percentage of bidirectional forks , reflecting restart at both ends of a replicon ( Fig 2D ) . These results indicate that EEPD1 accelerates restart of stressed replication forks , and that it increases the speed of replication during recovery from stress , implying that EEPD1 also assists in normal fork progression . Based on the above observations , we investigated the role of EEPD1 in the major DNA DSB repair pathways by using two previously described assays . EEPD1 depletion increased cNHEJ by 2 . 3-fold in the EJ5 cell reporter system ( Fig 3A ) [37 , 38] , implying that EEPD1 inhibits cNHEJ . EEPD1 depletion reduced HR repair of I-SceI induced DSBs by 6 . 4-fold in the HT256 reporter system ( Fig 3B ) [39] . This reduction in HR raised the question of whether EEPD1 depletion increased gene conversion tract lengths . Cells with defects in HR components display longer gene conversion tracts among residual HR products [30 , 31 , 40–46] . Consistent with these prior studies , HR products from EEPD1-depleted cells had significantly longer conversion tracts compared to controls ( Fig 3C and 3D ) . The longer gene conversion tracts are thought to reflect unstable heteroduplexes [40] and/or defective resection preventing efficient 5’ end-capture by the invaded template [30 , 46] . In the case of EEPD1 depletion , we hypothesize that defective end resection ( shown below ) results in less efficient 5’ end-capture . If true , then this implies that less efficient SS end-capture reactively stimulates synthesis along the invaded template , an idea supported by several published reports [30 , 31 , 42] . Cells with HR defects , such as those with BRCA1 or BRCA2 mutations , are hypersensitive to PARP1 inhibitors , due to an increase in unrepaired DSBs arising during replication [47–49] . We therefore repressed EEPD1 in BRCA1/2 proficient cells and assessed the effect of the PARP1 inhibitor olaparib on cell survival . EEPD1 repression markedly increased the cytotoxicity of olaparib ( 19-fold , Fig 3E ) , in the absence other genotoxins , consistent with EEPD1 playing a significant role in HR repair . There are two DSB repair pathways that use 5’ end resection to initiate the repair cascade , microhomology-mediated end joining ( MMEJ ) , and HR . The frequency of utilization of these two pathways can be compared at a single induced DSB in the EGFP-based MMEJ/HR-Mlu1 reporter ( Fig 4A ) [19] . Upon DSB induction with I-SceI transduction , repair by either MMEJ or HR results in loss of the I-SceI site and generation of EGFP , allowing repaired cells to be sorted by flow cytometry ( Fig 4B ) . The repaired EGFP loci were PCR amplified , and analyzed for repair by HR versus MMEJ . Cells repaired by MMEJ have a 9 nt duplication containing a BssHII site , while cells repaired by HR have an MluI site ( Fig 4C ) . The fraction of BssHII cleaved products among the total PCR products represent the fraction repaired by MMEJ , while the fraction cleaved by MluI represents HR repair . Depletion of EEPD1 resulted in an average decrease of 2 . 5-fold in EGFP-positive cells in the MMEJ/HR-MluI reporter system . The EGFP locus was PCR amplified from EGFP-positive cells , and digested with BssHII or MluI . This revealed that EEPD1 depletion resulted in a 9-fold reduction in HR and a 50% increase in MMEJ ( Fig 4D ) . This implies that when MMEJ and HR are competing at a single DSB site , EEPD1 pushes that repair decision towards HR , and away from MMEJ . This may mean that EEPD1 is important for initiating long range end resection used in HR , consistent with its interaction with Exo1/BLM as noted below . Since the key step for determining DSB and replication fork repair pathway choice is 5’ end resection [16–18] , we therefore assessed the role of EEPD1 in 5’ end resection after DSB formation using two approaches . First , we measured the generation of SS DNA at IR-induced DSBs by immunostaining newly incorporated BrdU in non-denatured SS DNA [50 , 51] . We found that depletion of EEPD1 reduced the number of cells with SS BrdU after IR by 5-fold ( Fig 5A and 5B ) . In the second approach , we assessed resection around an induced I-SceI DSB [52 , 53] . Using this technique , we found that EEPD1 depletion reduced end resection by 3-fold , nearly the same extent as CtIP depletion ( Fig 5C ) . Thus , EEPD1 is important for end resection after both a transduced restriction enzyme ( I-SceI ) and exogenous IR . We next tested whether EEPD1 functions in the same end resection pathway as Exo1 or CtIP ( Fig 5C ) . We depleted EEPD1 with or without co-depletion of Exo1 or CtIP . There is no significant difference in end resection when EEPD1 and Exo1 are co-depleted compared to individual depletion , suggesting that these enzymes function in the same resection pathway . Co-depletion of EEPD1 and CtIP also yielded similar results as individual depletions . These results suggest that EEPD1 functions in the same resection pathway ( s ) as Exo1 and CtIP . When a replication fork collapses , SS DNA arises by end resection , or by uncoupling of the polymerase complex from the helicase [10 , 54 , 55] . Such ss DNA is coated by RPA , which recruits ATRIP , leading to ATR activation and phosphorylation of RPA , H2Ax , and Chk1 which mediate cell cycle arrest and replication fork repair [56–58] . To define the epistatic position of EEPD1 in HR , confocal immunofluorescence microscopic studies of fork repair components were performed in cells treated with HU for a prolonged period , which causes replication fork collapse . We found that EEPD1 depletion significantly decreased foci formation by RPA32 ( 2 . 7-fold ) , gamma-H2Ax ( 3 . 3-fold ) , and RAD51 ( 3 . 5-fold ) ( Fig 5D and 5E ) . The decrease in RPA32 foci was consistent with its decreased phosphorylation , detected by Western blotting ( Fig 5F and 5G ) . Consistent with the decreased formation of RPA32 , gamma-H2Ax , and RAD51 foci , and the requirement for SS DNA to trigger ATR and Chk1 signaling , EEPD1 repressed cells also showed decreased phosphorylation of ATR ( 3-fold ) and Chk1 ( 11-fold ) ( Fig 5F and 5G ) . EEPD1 did not co-localize with gamma-H2Ax foci after damage ( S3 Fig ) , not surprisingly , since EEPD1 appears to act upstream of gamma-H2Ax . The Mre11-Rad50-NBS1 ( MRN ) complex is a first responder to DSB damage [59–61] . NBS1 recruits BRCA1 to stalled replication forks in an alternative pathway to the canonical gamma-H2Ax/MDC1/RNF8/BRCA1 recruitment pathway [62 , 63] . Confocal immunofluorescence studies were performed to investigate whether EEPD1 depletion impairs these early regulators of DSB repair . While HU-induced RPA32 , gamma-H2Ax , and RAD51 foci were significantly decreased in EEPD1 depleted cells ( Fig 5D and 5E ) , BRCA1 , 53BP1 , and NBS1 foci were unaffected ( S3 Fig ) , indicating the initial 53BP1 recruitment step of cNHEJ was functional , and that NBS1 and BRCA1 have upstream roles in repair of stressed forks . These data imply that MRN/BRCA1 and EEPD1 act in distinct repair pathways . For 5’ end resection to take place at a stalled replication fork , there must be a free DNA DSB end [3 , 4 , 8–12] . This can occur via fork reversal to form a chicken foot structure , but the majority of stalled forks do not reverse [20] . For those stalled forks that do not reverse the fork must be nicked to create this required free DNA DSB end [3 , 13 , 14 , 17] . If this is true , then DNA nicking should be increased after HU nucleotide depletion to stall replication forks . We assessed the occurrence of DNA nicking using alkaline single cell electrophoresis assays with and without EEPD1 depletion ( Fig 5H ) . We found that after HU exposure , DNA nicking increases 3 . 5-fold . However , EEPD1-depletion completely abolishes this increase , implying that EEPD1 is directly or indirectly responsible for DNA nicking in response to HU-induced replication stress . EEPD1 is expressed primarily in the nucleus , consistent with it functioning as a nuclease ( S4 Fig ) . Since EEPD1 has homology to RuvA , which binds to heteroduplex chicken foot structures , and it has a nuclease domain , we examined whether EEPD1 has nucleolytic activity on chicken foot structures . We found that recombinant EEPD1 protein did not nick any of the four double-stranded regions of the regressed fork , but it does have specific 5’ overhang endonuclease activity ( S5 Fig ) , cleaving at a single site at the joint of the overhang . Chicken foot structures with 5’ overhangs are difficult for Exo1 to process [64]; EEPD1 could promote further 5’ end resection by presenting Exo1 with a more amenable structure . These data demonstrate that EEPD1 is a 5’ endonuclease , consistent with the marked reduction in HU-induced nicks in EEPD1 depleted cells . We next assessed the effect of EEPD1 depletion and co-depletion of three other resection components , CtIP , Exo1 , and Dna2 , on cell proliferation with or without HU-induced replication stress ( S6 Fig ) . EEPD1 depletion alone suppressed cell proliferation , as did depletion of CtIP and Dna2 . Co-depletion of EEPD1 with CtIP or Dna2 did not further affect proliferation , with or without replications stress . By contrast , Exo1 depletion only modestly suppressed proliferation , and only with replication stress . Again , there was no further effect on proliferation with co-depletion of EEPD1 and Exo1 than with EEPD1 depletion alone . We also compared the effects of EEPD1 , CtIP , Exo1 , and Dna2 depletion singly and in pairs on the formation and resolution of after HU-induced replication stress ( S7 Fig ) . As above ( Fig 5D and 5E ) , EEPD1 depletion strongly suppressed the formation of gamma-H2Ax foci 4 and 24 h after HU , with similar or greater effects than CtIP depletion . Co-depletion of EEPD1 and CtiP did not further suppress gamma-H2Ax foci 4 h after HU . Exo1 depletion had similar effects as CtIP depletion , with or without co-depletion of EEPD1 . Dna2 depletion enhanced gamma-H2Ax focus formation , even in the absence of replication stress , indicating that Dna2 plays a key role in preventing endogenous DNA damage . Co-depletion of Dna2 and EEPD1 did not significantly suppress gamma-H2Ax foci in untreated cells or after HU exposure . The enhanced gamma-H2Ax foci with co-depletion of EEPD1 and CtIP may also reflect enhanced or more persistent DNA damage caused by the genomic lesions , independently of induced replication stress . Western analysis revealed that after replication stress with HU , EEPD1 is enriched in the nuclear chromatin fraction ( Fig 6A and 6B ) , suggesting that EEPD1 is recruited to chromatin containing damaged replication forks . We next assessed EEPD1 recruitment to stalled replication forks using Isolation of Proteins on Nascent DNA ( iPOND ) [65 , 66] . iPOND showed that EEPD1 is recruited to replication forks within 30 min of HU treatment , coinciding with the appearance of gamma-H2Ax ( Fig 6C ) which marks DSBs at stalled/collapsed replication forks [65] . PCNA was absent from the HU-stalled forks , consistent with replisome unloading from collapsed fork Okazaki fragments [65] . A control iPOND assay using a thymidine chase confirmed the specificity of EEPD1 recruitment to stalled forks ( Fig 6D ) . By using chromatin immunoprecipition [52] we also demonstrated that EEPD1 is recruited to an I-SceI induced DSB ( Fig 6E ) . Interestingly , EEPD1 constitutively co-immunoprecipitates with Exo1 , RPA32 , and BLM in the presence of DNase , whether or not replication stress is present , indicating that these proteins reside in the same complex ( Fig 6F and 6G ) . However , EEPD1 does not co-immunoprecipitate with Dna2 , indicating that it is likely not in the RPA/Dna2/MRN end resection complex [28] . Significantly , depleting EEPD1 reduced Exo1 and BLM protein levels , suggesting that EEPD1 promotes stability of the complex in which EEPD1 , Exo1 , and BLM reside ( Figs 6H and S6B ) , indicating that EEPD1 resides in an obligate complex , perhaps to prevent aberrant nuclease activation and improper DNA cleavage . Proper replication stress responses are required to prevent gross chromosomal instability , which can be assessed by the formation of micronuclei and nuclear bridges that result from mis-segregation of fused chromosomes [27 , 67] . We found that EEPD1-depleted cells display severe nuclear anomalies , with 6- and 7-fold increases in nuclear bridges and micronuclei , respectively ( Fig 7A–7D ) . Chromosome fusion events occur when collapsed forks are aberrantly repaired , as in BRCA1-deficient cells with unopposed 53BP1 [22 , 23] . 53BP1 depletion alone did not alter nuclear anomalies , but 53BP1 depletion largely suppressed both bridges and micronuclei associated with EEPD1 depletion . Metaphase analysis further demonstrated EEPD1 repression causes genome instability , revealed as significant increases in chromatid breaks and radial chromosomes , both of which arise in S/G2 cells ( Fig 7E and 7F ) . Interestingly , IR induced these S/G2-associated events , and at lower frequencies , G1-associated chromosome breaks and double minutes . EEPD1 depletion alone did not increase G1-associated events , nor did it affect the frequency of IR-induced these events ( Fig 7F ) . Thus , EEPD1 specifically suppresses S/G2 events . Although more chromatid breaks were observed in EEPD1-depleted cells treated with HU than with HU alone , the difference was not significant ( P = 0 . 17 ) ; EEPD1 depletion did significantly increase IR-induced chromatid breaks compared to IR alone . Interestingly , 53BP1 repression fully suppressed spontaneous and HU-induced chromatid breaks seen in EEPD1 depleted cells as well as HU-induced chromatid breaks in cells with normal EEPD1 expression ( Fig 7F ) . These results indicate that EEPD1 plays a critical role in maintaining genome stability , under stressed and non-stressed conditions , and suggest that EEPD1 promotes genome stability by mediating accurate HR repair of stressed replication forks . Cancer cells experience continuous replication stress due to metabolic alterations and checkpoint defects that permit DNA replication despite significant DNA damage . To manage this stress , it is reasonable to suppose that EEPD1 would be up-regulated in cancers . We tested this by analyzing mRNA expression of EEPD1 in newly resected colorectal cancer versus adjacent normal tissue . In the present study we analyzed EEPD1 expression in 181 new colorectal cancers , and found that EEPD1 was expressed an average of 2 . 3-fold higher than adjacent normal tissue in 171 of 181 cases ( Fig 7G ) . This study demonstrates that the uncharacterized EEPD1 nuclease plays a key role in repairing stressed replication forks via HR . Interestingly , while EEPD1 confers resistance to replication stress , it only appears to accelerate fork restart by 10 min . Thus , nearly all forks still restart within 30 min of release from stress in EEPD1-depleted cells compared to 20 min in wild-type cells ( Fig 2B and 2C ) . These results imply that when fork repair is repressed , even a relatively brief delay in fork restart can be lethal , perhaps because toxic recombination intermediates form if stalled forks fail to restart in timely manner [12 , 32 , 35] . EEPD1 is also important under non-stress conditions , as EEPD1 depletion significantly slows cell growth rate . Thus , EEPD1 probably promotes restart of replication forks that encounter DNA lesions arising spontaneously during normal cellular metabolism . There are numerous reports demonstrating that most cells do not tolerate long delays in the restart of stalled replication forks , if repair is impaired . BRCA1 deletion results in cell death after even a brief period of replication stress [68] . There is evidence that forks blocked by interstrand crosslinks are restarted via lesion by-pass long before the lesion is repaired [69] . Thus , restarting replication forks appears to be a higher priority for the cell than repair , at least in some situations . When an ATR inhibitor is combined with a fork stalling agent , cells in S-phase lose all ability to recover within 45 min [70] . Removing the replication stalling agent and ATR inhibition after that point does not restore cell viability . Reintroduction of the DNA damage checkpoint in yeast mutants after a brief period of replication stalling does not rescue cell viability [71] . Our data here show that the largest effect of EEPD1 on stalled fork repair and restart is 10 min after HU release , yet EEPD1 is required for survival to many replication stress agents . Thus , when replication fork repair is impaired , even a brief period of fork stalling can be lethal . Interestingly , cells proliferate more slowly when EEPD1 is depleted . In addition , even in the absence of replication stress with HU , cells with depleted EEPD1 have increased gamma-H2Ax and decreased RPA foci . Thus , cells lacking EEPD1 appear to experience spontaneous replication stress . This may be due to EEPD1 having a role in normal replication fork progression , perhaps in nucleolytic processing of replication fork lagging strand intermediates [1–3 , 13 , 14 , 16 , 17] . A not mutually exclusive alternative is that EEPD1 promotes restart of forks stalled by spontaneous lesions , which may arise frequently in rapidly growing cultured cells . This would require constant repair of replication forks stalled by the continuously generated DNA lesions within their paths , which would be reliant on EEPD1 . EEPD1 represses cNHEJ and enhances HR rates significantly . This indicates that EEPD1 plays an important role in DNA DSB pathway choice . By promoting 5’ resection EEPD1 would direct repair away from cNHEJ and towards resection-dependent repair pathways , namely HR and MMEJ [3 , 8–10 , 12 , 16 , 17 , 19] . PARP1 competes with the Ku complex for free DNA DSB ends to promote MMEJ over cNHEJ [72] . PARP1 inhibition with olaparib was synthetic lethal with EEPD1 depletion . This implies that EEPD1 depletion is not epistatic with PARP1 in replication fork repair or in MMEJ repair [49 , 73] , given that there is additional cytotoxicity when both are repressed . Both HR and MMEJ require 5’ end resection to begin their repair cascades [17 , 19] . Both pathways can repair and restart replication forks after stalling [19] . Recent reports indicate that DNA polymerase ( pol ) theta may suppress HR and promote MMEJ repair of DNA breaks [74] . HR-deficient tumors rely on pol theta for DNA DSB repair [75] . Pol theta enhances MMEJ by tethering free DSB ends after short range end resection for the microhomology search [76] . EEPD1 would seem to have the opposite effect , promoting HR at the expense of MMEJ ( Fig 4 ) . This would be beneficial because unopposed pol theta−mediated MMEJ repair of replication forks would increase non-conservative repair , and more importantly , chromosomal fusions [74 , 75] . It is possible that EEPD1 promotes HR over MMEJ by enhancing long range end resection [17 , 19 , 30 , 31 , 76] , perhaps via its interaction with Exo1 . Exo1 seems to be important for long range end resection during HR [17 , 19 , 77] . End resection promoting HR at DSBs is thought to initiate when BRCA1/CtIP displaces Rif1/53BP1 at DSBs [23–26] . CtIP has an important non-nuclease role in initiating 5’ end resection at undamaged DSB ends . CtIP may also have nuclease activity important for resection of damaged DSB ends , but this is controversial [24 , 30 , 60 , 78] . Thus , in addition to its role in cleavage of stressed replication forks , EEPD1 may also be important for initiating end resection at undamaged DSB ends in HR , where CtIP may not have a role . While the competing activities of Rif1/53BP1 and BRCA1/CtIP determines cNHEJ vs HR repair pathway choice , less is known about the role of these components in the repair decisions at stalled replication forks . The data here implies that EEPD1 directs the cell away from cNHEJ towards HR ( Fig 3A and 3B ) , probably by enhancing end resection . EEPD1 functions in a distinct end resection pathway from BRCA1/CtIP , or downstream of BRCA1/CtIP , since HU-induced BRCA1 foci are intact in EEPD1-depleted cells ( S3 Fig ) . Replication forks require a free DNA end with which to initiate 5’ end resection for repair by either HR or MMEJ [8 , 10 , 13 , 16 , 20] . For approximately one quarter of replication forks stalled with HU , this occurs via fork reversal , with Dna2 then mediating 5’ end resection [20 , 79] . However , the majority of replication forks require a nick in one of the parent strands at the replication fork itself to create a free DNA end . HU-induced replication stress results in rapid DNA nicking , which is mediated by EEPD1 ( Fig 5H ) . Given that EEPD1 is rapidly recruited to stalled replication forks , promotes 5’ end resection , HR , and replication fork restart , it is possible that the lack of DNA nicking seen after EEPD1 depletion is due to the failure of replication fork cleavage . This failure to cleave the stressed fork may prevent 5’ end resection for HR-mediated fork repair and decrease fork restart . This implies that some stressed replication fork nicks , rather than contributing to cell death , may instead promote cell survival by accelerating fork restart , perhaps by preventing accumulation of toxic HR intermediates [4 , 6] . Since many stalled replication forks do not reverse to form a one sided DNA free end for end resection [20] , such cleavage is often necessary to initiate end resection and HR [8 , 10 , 13 , 16] . Placing the various end resection nucleases epistatically within the context of 5’ end resection is challenging [13 , 14 , 17] . From a biochemical standpoint , there appear to be two end resection complexes , BLM-DNA2-RPA-MRN and EXO1-BLM-RPA [28] , with EEPD1 as a component of the latter . However , there is functional overlap between these complexes , and both are likely essential for HR and cell survival in response to replication stress [16 , 18] . From the data presented here , there is little additional deficiency in end resection after replication stress when EEPD1 is doubly depleted with Exo1 or CtIP . Interestingly , Dna2 depletion increases gamma-H2Ax formation de novo and after replication stress , while Exo1 and EEPD1 depletion reduce gamma-H2Ax formation after replication stress . Thus , Dna2 appears to operate downstream of gamma-H2Ax signaling , while EEPD1/Exo1 are upstream , but perhaps both complexes are needed to repair distinct forms of stressed replication forks [20] . End resection creates SS DNA that can signal replication stress and cell cycle arrest . EEPD1 depletion abrogates gamma-H2Ax foci ( Fig 5D and 5E ) , indicating resection promoted by EEPD1 precedes phosphorylation of H2Ax during replication fork repair . Similarly , ATR is activated by RPA/ATRIP loading onto SS DNA , ultimately activating Chk1 . There are two possible explanations for the finding that EEPD1 is required for ATR/gamma-H2Ax/Chk1 phosphorylation after HU . First , the resection defect in EEPD1-depleted cells may account for a fraction of the ATR and Chk1 activation defects after replication stress . In this scenario the SS DNA created by end resection plays a key role in RPA/ATRIP activation of ATR . Second , the SS DNA that signals ATR activation may arise not from end resection but from the disassociation of the helicase from the polymerase complex [54] , and in this case EEPD1 might play a role in RPA/ATRIP signaling to ATR . In either case , at least for replication stress induced by HU , EEPD1 is an important factor in the activation of ATR/Chk1 . Interestingly , EEPD1 is in a constitutive and obligate complex with Exo1 and BLM ( Fig 6F and 6G ) . The long resection exonuclease Exo1 requires a free 5’ DNA end to initiate resection at a damaged fork . Since EEPD1 and Exo1/BLM constitutively co-immunoprecipitate , this implies that EEPD1 is a partner in the Exo1/BLM/RPA end resection complex [28] . It also implies that when EEPD1 is recruited to the damaged fork , it is accompanied by Exo1/BLM , which are needed for completion of end resection . The obligate nature of this complex is not surprising , since nuclease function must be exquisitely balanced to prevent wide-spread and unregulated genome incision . It is imperative the cell tightly control all nucleases to prevent inappropriate or untimely DNA cleavage to suppress translocations , and to maintain genome stability . Our results also indicate that EEPD1 helps maintain genome stability . As proposed for cells with defects in other HR proteins like BRCA1 and BRCA2 [21–23] , the genome instability seen in EEPD1 depleted cells is likely a direct consequence of unopposed cNHEJ causing aberrant ligation of DNA ends at distinct collapsed replication forks . This hypothesis is supported by the fact that genome instability observed in EEPD1 depleted cells is suppressed by depletion of the cNHEJ promoting factor 53BP1 ( Fig 7A–7F ) . There is another potential reason why EEPD1 may prevent chromosomal instability- The MMEJ pathway mediates chromosomal translocation events when replication stress overcomes the ability of the cell to repair such stress [74 , 75 , 77] EEPD1 promotes HR over MMEJ ( Fig 4 ) and this could suppress chromosomal instability during replication stress by directing fork repair toward HR which is less prone to chromosomal fusions . Thus , EEPD1 may prevent chromosomal translocations by promoting HR and suppressing MMEJ during repair of chromosomal DSBs . Our results indicate that EEPD1 is an important guardian of genome stability that functions by regulating replication fork repair pathway choice . Although the mechanism by which EEPD1 depletion sensitizes cells to replication stress is not well defined by the present study , many prior studies that demonstrate HR defects increase sensitivity to replication stress correlate increased sensitivity with increased gamma-H2Ax [reviewed in refs . 1 , 2 , 3 , 8] . While it is widely accepted that stressed replication forks can collapse into aberrant structures in cells lacking HR machinery to repair them [1 , 13 , 14 , 19] , whether these aberrant structures actually cause cell death is not known . We demonstrate here that EEPD1 depletion increases cell death in the face of replication stressors , and that it reduces HR , slows replication fork restart , reduces DNA nicking , and creates cytogenetic and nuclear abnormalities . The marked increase in micronuclei and mitotic bridges in EEPD1 depleted cells is exacerbated by replication stress , and this suggests the following mechanism for cell death from EEPD1 deficiency during replication stress: collapsed replication forks end-join aberrantly , creating chromosomal fusions that are manifest as mitotic bridges and micronuclei . These gross chromosome abnormalities would intuitively be difficult for cells to recover from , and therefore may serve as a better correlation between HR and cell death than gamma-H2Ax . One would predict that malignancies would require intact EEPD1 to proliferate , and that loss of function mutations would be rare in human cancers . This is indeed the case; there are only 31 coding changes in EEPD1 out of 8273 individual cancer genome sequences in the COSMIC database ( http://cancer . sanger . ac . uk/cosmic ) , and the vast majority of these are conservative , and are not predicted to alter function . This is not surprising , as a malignancy with EEPD1 functional loss would have difficulty proliferating given that tumors often survive despite significant replication stress caused by oncogene activation , hypoxia , and/or nutrient deprivation [6] . Thus , even though loss of EEPD1 results in genomic instability , EEPD1 should not be viewed as a tumor suppressor in the same sense as BRCA1 and BRCA2 , two HR components that show loss of function mutations in cancer . It is likely that loss of EEPD1 function would be too detrimental to replication fork restart and fork progression to be selected for during oncogenesis , because of the fundamental importance of DNA replication to malignant cells . On the other hand , given its role in replication fork rescue , EEPD1 could be an excellent target for treatment of human malignancies . EEPD1 is over-expressed in nearly all colorectal cancers [80] ( Fig 7G ) and large cell lymphomas [81] , cancers whose treatment is based on agents that create replication stress . Targeting EEPD1 could block proliferation of cancers that depend on EEPD1 , or sensitize tumors to chemotherapeutics that cause replication stress . Such agents are the foundation for treating both of these types of malignancies [81–83] . A549 , HEK-293 , HEK-293T , HT256 reporter cells , and the various U2OS reporter cells ( EJ5 , MMEJ/HR ) , were cultured in D-MEM supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin . HT256 cells were cultured in Alpha-MEM supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin . EEPD1 was depleted using two mechanisms , shRNA and siRNA , to control for variation in the method of mRNA destruction . EEPD1 was depleted either by 1 ) EEPD1 lentivirus shRNAs produced from 293T cells ( pLKO . 1 , Thermo Scientific , Pittsburgh , PA ) ; or 2 ) SMARTpool ON-TARGETplus EEPD1 siRNA from Dharmacon RNAi Technologies ( GGACUGACCUUCACCGCCA; CUGAGAAGCCCUCGAGUCA , GGAAGUUGACCUCGGGGUA; UGCGAGAGGUGGUGUGCAU ) ( Pittsburgh , PA ) . EEPD1 3’UTR On-Target plus siRNA also from Dharmacon ( GGAAGUUGACCUCGGGUA ) . 53BP1siRNA ( h ) is a pool of 3 different siRNA duplexes from Santa Cruz Biotechnology ( sc-37455 ) . All other siRNAs were from Dhamarcon SMARTpools . All nucleic sequences are listed 5’ to 3’ in this Supplement . Polyethylenimine ( PEI ) was used to perform plasmid transfections according to the manufacturer’s instructions ( Thermo Scientific ) . Briefly , PEI was incubated with plasmid DNA at 3:1 ratio in Opti-MEM at RT for 20 min before addition to cells . After 6 h incubation , cells were washed and placed in fresh media . RNAiMAX ( Invitrogen , Grand Island , NY ) was used to transfect siRNA pools . Briefly , RNAiMAX was incubated with 50 nM of siRNA in Opti-MEM at RT for 20 min before addition to cells . After 24 h cells were washed and placed in fresh media . EEPD1 repression was confirmed by western blotting for every experiment . At least two clones were used for each Lentiviral shRNA experiment , to control for clonal variation in repression . There was no difference in phenotypes obtained between the shRNA and the siRNA repression of EEPD1 . Experiments were repeated using both techniques for EEPD1 depletion to control for off-target effects of the mechanism of repression . All experiments were performed at least three times , in at least two cell lines , to control for experimental and cell line variation . A549 lung cancer cells were used for most replicative experiments since this cell line has high EEPD1 expression . All studies in A549 were also repeated at least once in HEK-293 cells to control for cell lineage variability . Clonal survival after treatment with DNA damaging agents was determined by seeding 2 , 000 cells per 10 cm dish in either control media or media with varying concentrations of genotoxic chemicals , or exposure to varying doses of IR or UV light . Cells were exposed to etoposide or olaparib for 18 h , then washed and incubated in fresh media for 12 days . Colonies were stained with 0 . 1% crystal violet in methanol and counted . A colony greater than 50 cells was counted as a surviving clone . For HU , cells were treated continuously for 12 days before colonies were stained and scored . Plating efficiency was calculated as the number of colonies divided by the number of cells plated without genotoxin treatment . In all of these assays , survival was normalized to untreated cells transfected with control or EEPD1 si or shRNAs . Survival fractions were calculated as the number of colonies formed after exposure to a given genotoxin divided by the number of cells plated , then multiplied by the plating efficiency . Unpaired Student t tests were used for all statistical analysis , unless otherwise indicated . Each experiment was performed 6–9 times in triplicate . EEPD1 expression was monitored by standard western blotting protocol [36] using a custom-produced rabbit polyclonal antibody to EEPD1 peptide ( CAEFYTEKDWSKKDAPRNHS , Lampire Biological Laboratories , Pipersville , PA ) . Phosphorylated RPA32 ( S4/S8 ) and total RPA32 antibodies were from Bethyl Laboratories ( Montgomery , TX ) . Phosphorylated ATR ( T1989 ) antibody was from Genetex ( Irvine , CA , cat . Gtx128145 ) . Total ATR , phosphorylated Chk1 ( S345 ) , and total Chk1 antibodies were from Cell Signaling Technology ( Danvers , MA ) . 53BP1 and BLM antibodies were from Abcam ( Cambridge , MA ) , Exo1 antibody was from Proteintech ( Chicago , IL ) , and beta-actin antibody was from Sigma-Aldrich ( St . Louis , MO ) . When protein levels were quantitated , each western analysis was performed at least 3 times , with densitometric measures of band intensities normalized to loading controls . Student t tests were used for statistical analysis of the protein intensity differences . Immunoprecipitation was performed with the Pierce Crosslink Magnetic IP/Co-IP kit according to manufacturer’s instructions ( Thermo Scientific Cat . 88805 ) as we described [36] . Briefly , HEK-293 cells overexpressing V5-tagged EEPD1 were treated , harvested and washed by PBS before lysis using IP lysis/wash buffer , then 5 ug of V5 mouse antibody ( Invitrogen ) were coupled to protein A/G magnetic beads and cross-linked with 20 uM disuccinimidyl suberate . The antibody cross-linked beads were incubated with cell lysate ( 0 . 8–1 . 2 mg ) in a 500 ul of diluted lysate solution for 1 h at RT on a rotator . Beads were collected , washed and incubated with 100 ul of elution buffer for 5 min at RT . Antigen recovery was achieved by collecting the supernatant on a magnetic stand . Protease and phosphotase inhibitors were present in all buffers . ChIP was performed in HT256 cells using the procedure and GAPDH primers as we described [52] . ChIP primers for neo in HT256 152 nt from the I-SceI DSB site: Neo671 Forward: GACGGGCGTTCCTTGCGCAGCTG; Neo830 Reverse: CCAGATCATCCTGATCGACAAGAC . Primers 650 nt distant from the I-SceI site: Neo1 Forward: AAGCTTCACGCTGCCGCAAGCAC; Neo152 Reverse: GAACCTACCTGCTTTCTCTTTGC . GAPDH Forward: TCGGTTCTTGCCTCTTGTC; GAPDH Reverse CTTCCATTCTGTCTTCCACTC . Each immunoprecipitation was performed at least 3 times . Real time PCR to quantify immunoprecipitated sequences was performed using the SYBR green reagent ( Applied Biosystems , Thermo Scientific ) with the ABI 7000 sequence detection system , normalized to GAPDH amplification . Two methods were used for measuring stalled replication fork restart . In the first method , replication fork restart after arrest was measured by immunofluorescent detection of BrdU foci after DNA denaturation ( BrdU in DS DNA ) , as we described previously [36] . Log phase A549 cells expressing normal or repressed levels of EEPD1 , with or without expression of siRNA-resistant FLAG-tagged EEPD1 were incubated with 10 mM HU for 18 h and then released into media with 10 uM BrdU for 30 min . After washing , cells were fixed at different time points . Replication recovery was shown as percentage of cells with ≥ 3 BrdU foci 2h after release from HU . Cells without HU treatment served as controls for background staining from normal cell proliferation , which was used as threshold for measurement . Values are averages ( ± SEM ) for 11–23 distinct determinations ( >100 cells scored per condition ) . The second method was DNA fiber analysis , as we previously described [32 , 35] . Both A549 and HEK-293 cells were tested to control for cell line differences . 600 , 000 cells were incubated overnight at 37°C in six-well plates . 20 mm IdU was added to growth medium and incubated for 20 min at 37°C . The IdU media was removed and cells washed in fresh medium , cells were treated with 5 mm HU for 60 min or mock-treated . The HU-containing medium was replaced with fresh medium containing 100 mm CldU . Cells were then incubated for varying times at 37°C . The CidU medium was removed , cells harvested , resuspended in PBS , and 1 , 000 cells were transferred to a positively charged microscope slide ( Superfrost/Plus , Daigger ) , and processed for DNA fiber analysis as we described previously [32] . Slides were mounted in PermaFluor aqueous , self-sealing mounting medium ( Thermo Scientific ) , and DNA fibers were visualized using a confocal microscope ( Olympus , FV1000D , 63× oil immersion objective ) . Images were analyzed using the Olympus Fluoview software . Confocal immunofluorescence foci assays were performed as we described [35] with minor modifications . In brief , cells were cultured on coverslips followed by siRNA transfection and HU treatment . Cells were pre-extracted with 0 . 5% Triton X-100 and fixed with 4% paraformaldehyde for 20 min . Coverslips were then blocked with 1% BSA for 1 h before incubating with primary antibodies overnight . After washing twice , coverslips were incubated with secondary antibodies conjugated with Alexa Fluor dye ( Invitrogen ) , mounted in anti-fade solution containing DAPI and stored at 4°C . All samples were analyzed within 24 h with a laser confocal scanning microscope ( TCS-SP5 , Leica Microsystems , Exton , PA ) . Cells with >5 foci were counted as positive . Photomicrographs of distinct cell populations were taken at equal magnifications and equal fluorescence intensities . For NBS1 and BRCA1 foci , the cells were fixed in 100% methanol and incubated with 1% BSA in 0 . 1% PBS-Tween for 1 h before incubating with primary antibodies overnight . RAD51 antibody was obtained from Santa Cruz Biotechnology ( Dallas , TX ) . BRCA1 and RPA32 antibodies were obtained from Bethyl Laboratories ( Montgomery , TX ) ; gamma-H2AX ( S139 ) antibody from Millipore ( Billerica , MA ) , phosphorylated NBS1 ( S343 ) and BrdU antibodies from Cell Signaling ( Danvers , MA ) , and 53BP1 antibody from Abcam ( Cambridge , MA ) . To assess nuclear structural abnormalities ( micronuclei and post-mitotic bridging ) , control or HU-treated cells , with or without EEPD1 depletion , were grown on coverslips and fixed as above , and stained with 300 nM DAPI ( Beckman ) in PBS for 5 min . After washing thrice with PBS , coverslips were mounted in anti-fade solution and analyzed within 24 h . Of note , EEPD1 was located in the nucleus , but did not form discrete foci before or after damage . Each immunofluorescence assay was performed at least 3 times in triplicate . iPOND was performed as described by Sirbu and colleagues[65 , 66] , with minor modifications to improve protein capture . In brief , HEK-293T cells over-expressing V5-tagged EEPD1 were seeded in three 150 mm plates/condition 24 h before the experiment . After 24 h incubation , 10 uM EdU ( Invitrogen ) was added to the medium for 10 min . EdU treatment was followed with 3 mM HU ( Sigma , St . Louis , MO ) at indicated times . The cells were then fixed with 1% formaldehyde ( Sigma ) for 10 min at RT , quenched by 0 . 125 mM Glycine ( Sigma ) , and collected by scraping . The cells were permeabilized with 0 . 25% Triton X-100 for 30 min , and then subjected to click–iT reaction using Biotin azide ( Invitrogen ) for 90 min at room temperature . Lysis conditions were modified to reduce background: lysis was performed in 0 . 25% SDS lysis buffer for 10 min at RT , followed by sonication at 4°C using Bioruptor ( Diagenode ) for 10 min with 30 s on/off cycles at high intensity . This treatment consistently yielded fragments between 80–100 bp . Finally , EdU-labeled DNA was pulled down by incubation with Streptavidin-agarose beads ( Millipore ) overnight at 4°C . The beads were washed once with lysis buffer , once with 1 M NaCl , and twice with lysis buffer . Bound proteins were eluted in 2 x NuPAGE LDS sample buffer ( Invitrogen ) containing 1 x sample reducing agent ( Invitrogen ) at 95°C for 35 min before loading for western analysis . Protease and phosphatase inhibitors ( Thermo Scientific ) were added to all buffers . Each iPOND assay was performed 3 times . End resection was analyzed using two methods . First , end resection following gamma-irradiation was assessed using a single strand BrdU assay as described [51] . To detect single strand DNA formation , A549 cells were transfected with control and EEPD1 siRNAs , and plated on coverslips at 24 h , then incubated with 30 μM BrdU for 42 h before treatment with 20 Gy IR . At various times after irradiation , cells with native ( non-denatured ) DNA were analyzed by immunofluorescent confocal microscopy to detect BrdU in SS DNA created by end resection . Second , end resection was also measured adjacent to a specific I-SceI-induced DSB by quantitative PCR ( qPCR ) [50 , 53] . Genomic DNA ( gDNA ) was extracted from HT1904 cells [52] harvested 4 h after infection with adenovirus vectors that express I-SceI ( Adv-I-SceI ) or GFP ( Adv-GFP ) as control . Half of the gDNA was digested with XmaI ( NEB ) , and the remaining half was mock-digested . PCR reactions included XmaI-digested or undigested gDNA as template , 0 . 5 uM of each primer , 0 . 2 uM TaqMan probe , and 1X TaqMan universal master mix ( ABI ) . qPCR was performed on a 7900HT Fast Real-Time PCR System ( ABI ) under standard thermal cycling conditions . Results were analyzed with SDS2 . 3 ( ABI ) and Graph Pad 6 . For each sample , a ΔCT was calculated by subtracting the CT value of the undigested sample from the CT value of the XmaI-digested sample . The percentage of SS DNA was calculated with the following equation: SS DNA% = 1/ ( 2^ ( ΔCt-1 ) +0 . 5 ) *100 [50] . Primers and probes were: forward ( CGACCTTCCATGACCGAGTACAA ) , reverse ( TCCGGGTCGACGGTGTG ) , and probe ( 6FAMACCGCGACGACGTCCCCCGGGCC-TAMRA ) . All Ct values were corrected for different DNA concentrations , as determined by qPCR of a ‘No Cut’ amplicon on chromosome 22 that lacks XmaI sites: forward ( ACATTGTCTCTGTCATGGGC ) , reverse ( TGTGTCAGGGATTTGCTCAC ) , and probe ( 6FAM AGAGCATGGGTGGATCCTGGATATTCA-TAMRA ) . DSB induction by Adenoviral-I-SceI was measured by qPCR and calculated as described [52] using a primer set that flanked the I-SceI site , and primers to the chromosome 22 ‘No Cut’ amplicon as a negative control . The ‘No Cut’ amplicon was used to normalize the amount of genomic DNA in the qPCR reaction , and the percentage of DSBs in Adv-GFP treated cells was set to zero . Both end resection assays were performed three times in triplicate . Pure recombinant human FLAG-tagged EEPD1 protein was generated in 293 cells and purified as we described [84] . Nuclease assays were performed as we described [32 , 84] . 3’ overhang reversed fork ( “chicken foot” ) structures were obtained by annealing SHL101 , SHL108 , SHL109 , and SHL110 , and then gel-purifying the annealed structure . 5’ overhang reversed fork structures were obtained by annealing SHL101 , SHL108 , SHL111 , and SHL112 and then gel-purifying the intact annealed structure [32 , 84]: SHL101 ( 60mer ) : 5’-CGATACTGAGCGTCACGGACTCTGCCTCAAGACGGTAGTCAACGTGTTACAGACTTGATG-3’ SHL108 ( 60mer ) : 5’-CTAGACTCGAGATGTCAAGCAGTCCTAACTTTGAGGCAGAGTCCGTGACGCTCAGTATCG-3’ SHL109 ( 60mer ) : 5’-CATCAAGTCTGTAACACGTTGACTACCGTCGATCCACTAG AGGTCTAAGCGACCTCATTC-3’ SHL110 ( 40mer ) : 5’-CTAGTGGATCAGTTAGGACTGCTTGACATCTCGAGTCTAG-3’ SHL111 ( 40mer ) : 5’-CATCAAGTCTGTAACACGTTGACTACCGTCGATCCACTAG-3’ SHL112 ( 60mer ) : 5’-AGGTCTAAGCGACCTCATTCCTAGTGGATCAGTTAGGACTGCTTGACATCTCGAGTCTAG The HT256 reporter system was used to determine I-SceI-induced HR efficiency and gene conversion tract spectra as we described [39 , 85] . The EJ5-GFP U2OS system was used to assess NHEJ [37 , 38] . Both of these reporter systems have single , integrated copies of reporters with I-SceI target sites cleaved upon transfection of an I-SceI expression vector . Cells were transfected with either control or EEPD1 siRNAs or shRNAs , and then transfected 24 or 48 h later with pCBA-SceI or empty vector , using PEI . After 48 h incubation , EJ5 cells were trypsinized and washed with PBS and GFP-positive cells reflecting NHEJ frequencies were measured by FACSort ( Becton-Dickinson , San Jose , CA ) and analyzed with CellQuest ( Becton-Dickinson ) software . Productive HR in HT256 cells reconstitutes a functional neomycin phosphotransferase gene , generating G418 resistant colonies . Two thousand cells were plated in three 10-cm dishes per each condition , in non-selective media , 24 h after I-SceI vector transfection to establish plating efficiency . To assess HR , 500 , 000 cells were plated in media with G418 ( 325 ug/ml , 100% active ) added 24 h after transfection . DSB-induced HR frequencies were calculated as the number of G418-resistant colonies per viable cell plated in G418 medium after 21 days , normalized for plating efficiency . HR assays were performed 15 times in triplicate and the NHEJ assays 12 times in triplicate . Gene conversion tracts were analyzed as we described [39 , 40 , 85] on the above HR repaired neo-positive colonies . HT256 G418 resistant colonies were stained and counted , or expanded under continuous G418 selection for gDNA isolation and molecular analysis . Genomic DNA was extracted using the DNeasy Tissue Kit ( Qiagen , Valencia , CA ) . Primers A ( CCTTCACTTTCCAGAGGGTC ) and B ( GCGAAGAACTCCAGCATGAG ) were used to amplify a 1 . 5 kb fragment comprising the recipient neo allele ( MMTVneo ) by using standard PCR conditions . The donor neo allele carries 12 silent single-base mutations at approximately 100 bp intervals that create restriction fragment length polymorphisms ( RFLPs ) . These RFLP markers allow high-resolution analysis of gene conversion tract length , directionality , and continuity . The 12 silent RFLP markers and the natural BanII site were mapped in PCR fragments amplified from HR products . The analysis of MMEJ versus HR competitive repair from a single DSB was performed in modified U2OS cells as we described [19] . To directly compare MMEJ with HR , an EGFP-based MMEJ and HR competition reporter system , termed EGFP-MMEJ/HR-MluI , was generated . This reporter had the EGFP ( R-EGFP ) cassette of EGFP-HR replaced with the EGFP-MMEJ cassette . A unique MluI site in the parent EGFP ( D-EGFP ) cassette was created via a silent mutation at the BssHII site . Upon I-SceI cleavage , restoration of a functional EGFP cassette results in loss of the I-SceI site after cells undergo repair by either MMEJ or HR . PCR analysis of the sorted green cells using primers specific for R-EGFP was performed . The primers were: EGFP MMEJ/HR Forward:5’-ACGGGGTCATTAGTTCATAGCCCA , EGFP MMEJ/HR Reverse: 5’-GGGATTTTGCCGATTTCGGCC . Repair of the I-SceI DSB by MMEJ would retain one copy of the 9-bp duplication with an intact BssHII site . The percentage of the BssHII-digestible bands within the total PCR amplified product reflects the MMEJ frequency . Repair of that I-SceI-induced DSB by HR transfers the MluI site from D-EGFP to R-EGFP , and thus the percentage of MluI-digestible bands of the total PCR product reflects the HR frequency . To assess nuclear structural abnormalities , micronuclei and post-mitotic bridging from aberrant chromosomal segregation , control or HU-treated cells , with or without EEPD1 depletion , were grown on coverslips and fixed as above , and stained with 300 nM DAPI ( Beckman ) in PBS for 5 min [27] . After washing with PBS , coverslips were mounted in anti-fade solution and analyzed within 24 h . Of note , EEPD1 was located in the nucleus , but did not form discrete foci with or without DNA damage . At least ten distinct determinations ( 142–190 nuclei per determination ) were performed for each treatment group . Structural aberrations in metaphase chromosomes were scored by Solid Giemsa staining as described [86 , 87] . EEPD1 and/or 53BP1 were repressed using siRNA in log phase A549 cells , with or without 18 h treatment with 10 mM HU . Cells were washed with PBS and fresh media with colcemid ( final concentration 0 . 25 ug/mL ) was added , and cells were incubated for 24 h before harvest . Chromosome preparations were made according to the standard air drying procedure as we described [87] . Cells were harvested , washed with pre-warmed PBS twice , hypotonically treated ( 0 . 56% KCl , 20 min at 37°C ) and subsequently fixed in freshly prepared acetic acid-methanol ( 1:3 ) . At least three changes of fixative were performed before the cell suspension was dropped on to a pre-cleaned chilled glass slide and dried at RT at least for 1 day before staining . Structural translocations such as dicentric and ring chromosomes , and Robertsonian translocations , were scored under 63x magnification [87] . Statistics were calculated using Fisher exact tests . Cytogenetic spreads were performed three distinct times with a total of 102–374 metaphase spreads were analyzed per condition . Alkaline single cell electrophoresis assays for SS nicking was performed as described [88] using the CometAssay kit ( Trevigen , Gaithersburg , MD ) . Briefly , A549 cells were transfected with siRNAs and treated with 10 mM HU for 1 h or mock treated . Cells were harvested , washed with cold PBS and mixed with molten 1:10 ( v/v ) LMAgarose and immediately spread over the sample area of comet slides . Cells were immobilized at 4°C in the dark for 30 min and immersed in lysis solution overnight . For the alkaline comet assay , slides were treated with alkaline unwinding solution for 1 h at 4°C in the dark before electrophoresis in alkaline electrophoresis buffer . Samples were rinsed with water and immersed in 70% ethanol before drying at 37°C for 15 min . SYBR Gold was used to stain dried agarose for 30 min at RT before rinsing and drying again . Slides were viewed with a Leica inverted epifluorescence microscope and analyzed by software Image J with OpenComet plugin [89] . Alkaline comet assays were performed five times in triplicate , counting >100 slides per experiment . Colorectal carcinoma biopsies were re-analyzed specifically for EEPD1 expression , compared to adjacent normal mucosa . Gene expression measurements were performed in 217 patients with colorectal carcinomas from pre-therapeutic biopsies as we described [80] . From 217 patients , tumor samples were extracted , and from 181 of these matched normal tissue ( mucosa ) samples were also obtained . Gene expression was measured on Agilent Human Microarrays . Microarray data was extracted as log2 intensities and quartile normalized . Gene expression of EEPD1 ( Agilent Probe: A_23_P333498 , Refseq: NM_030636 , Chr . Coord: chr7:36340858–36340917 , Probe: CAGCCTGTTCTTACTCCAGCTCAACCCATTGGGTGTTGGCTGTTTTTGGTTTTAGTTGTT ) was obtained . Significance was computed from matched tumor vs . mucosa samples using a paired Wilcoxon test .	The cell itself damages its own DNA throughout the cell cycle as a result of oxidative metabolism , and this damage creates barriers for replication fork progression . Thus , DNA replication is not a smooth and continuous process , but rather one of stalls and restarts . Therefore , proper replication fork restart is crucial to maintain the integrity of the cell’s genome , and preventing its own death or immortalization . To restart after stalling , the replication fork subverts a DNA repair pathway termed homologous recombination . Using any other pathway for fork repair will result in an unstable genome . How the homologous recombination repair pathway is initiated at the replication fork is not well defined . In this study we demonstrate the previously uncharacterized EEPD1 protein is a novel gatekeeper for the initiation of this fork repair pathway . EEPD1 promotes 5’ end resection , the initial step of homologous recombination , which also prevents alternative fork repair pathways that lead to unstable chromosomes . Thus , EEPD1 protects the integrity of the cell genome by promoting the safe homologous recombination fork repair pathway .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	EEPD1 Rescues Stressed Replication Forks and Maintains Genome Stability by Promoting End Resection and Homologous Recombination Repair
Kynurenines , the main products of tryptophan catabolism , possess both prooxidant and anioxidant effects . Having multiple neuroactive properties , kynurenines are implicated in the development of neurological and cognitive disorders , such as Alzheimer's , Parkinson's , and Huntington's diseases . Autoxidation of 3-hydroxykynurenine ( 3HOK ) and its derivatives , 3-hydroxyanthranilic acid ( 3HAA ) and xanthommatin ( XAN ) , leads to the hyperproduction of reactive oxygen species ( ROS ) which damage cell structures . At the same time , 3HOK and 3HAA have been shown to be powerful ROS scavengers . Their ability to quench free radicals is believed to result from the presence of the aromatic hydroxyl group which is able to easily abstract an electron and H-atom . In this study , the redox properties for kynurenines and several natural and synthetic antioxidants have been calculated at different levels of density functional theory in the gas phase and water solution . Hydroxyl bond dissociation enthalpy ( BDE ) and ionization potential ( IP ) for 3HOK and 3HAA appear to be lower than for xanthurenic acid ( XAA ) , several phenolic antioxidants , and ascorbic acid . BDE and IP for the compounds with aromatic hydroxyl group are lower than for their precursors without hydroxyl group . The reaction rate for H donation to O-atom of phenoxyl radical ( Ph-O ) and methyl peroxy radical ( Met-OO* ) decreases in the following rankings: 3HOK ~ 3HAA > XAAOXO > XAAENOL . The enthalpy absolute value for Met-OO* addition to the aromatic ring of the antioxidant radical increases in the following rankings: 3HAA* < 3HOK* < XAAOXO* < XAAENOL* . Thus , the high free radical scavenging activity of 3HAA and 3HOK can be explained by the easiness of H-atom abstraction and transfer to O-atom of the free radical , rather than by Met-OO* addition to the kynurenine radical . The kynurenine pathway ( KP ) , the primary route of tryptophan degradation in mammalian cells , includes kynurenine ( KYN ) , kynurenic acid ( KYNA ) , 3-hydroxykynurenine ( 3HOK ) , 3-hydroxyanthranilic acid ( 3HAA ) , quinolinic acid ( QUIN ) , and other metabolites collectively called kynurenines ( Fig 1 ) . There are multiple mechanisms of kynurenines' action on nervous system . QUIN and KYNA , the ligands of ionotropic glutamate receptors [5 , 6] , modulate neurodegenerative processes in the brain [7] . The autoxidation of 3HOK and 3HAA leads to the hyperproduction of reactive oxygen species ( ROS ) which damage cellular lipids , proteins , and DNA [8–10] . Kynurenine 3-monooxygenase ( KMO ) , an enzyme producing 3HOK from KYN , has been linked to the pathophysiology of HD by a mechanism involving ROS [11] . Accumulation of 3HOK in the central nervous system of Drosophila cardinal mutant leads to the progressive memory loss [12] . Since 3HOK is capable of auto-condensation , the eyes of this mutant , as well as the color of mammalian lens cataract [13] progressively get brown on ageing . The Drosophila eye color mutants are started to be envisioned as a therapeutic tools for HD [14] . At the same time , both 3HOK and 3HAA were shown to be powerful antioxidants scavenging peroxyl radicals [15 , 16] . Xanthurenic acid ( XAA ) , a product of KYNA hydroxylation , has similar antioxidative properties , but its rate of interaction with free radicals is slower [15] . Tryptophan and its catabolites without aromatic hydroxyl group , such as kynurenine ( KYN ) , KYNA , and anthranilic acid ( AA ) have no effect on peroxy-mediated oxidation . Thus , phenolic hydroxyl group is important for antioxidant activity of kynurenines . Antioxidants are supposed to beneficially interfere with diseases-related oxidative stress , however , the interplay of endogenous and exogenous antioxidants with the overall redox system is far from clear [17] . Phenolic compounds suppress lipid peroxidation due to their ability to react with free radicals at a faster rate than with the substrate [18 , 19] . There are two main pathways of phenolic antioxidants quenching free radicals: electron transfer and H-atom transfer . H-atom easily abstracted from the aromatic OH-group interacts with peroxyl radical ROO* produced during lipid peroxidation and breaks the chain reaction: Ar−OH+ROO→Ar−O+ROOH ( 1 ) There are two pathways of hydrogen transfer: hydrogen atom transfer ( HAT ) and proton-coupled electron transfer ( PCET ) [20] . HAT is preferable when electron density of singly occupied molecular orbital ( SOMO ) in the transition structure ( TS ) lies along the same line as the O…H…O bond and H is transferred between the oxygens as a whole particle . PCET is preferable when SOMO is orthogonal to O…H…O bond , as in phenoxyl-phenol complex , and proton is transferred between oxygen σ lone pairs forming hydrogen bonds with them , while the electron is transferred between oxygen π-orbitals . Also , phenolic antioxidant radicals are able to quench peroxyl radical via its addition to the aromatic ring at ortho- or para-position . In order to trap the radical and not to react with hydrocarbon R-H substrate , an antioxidant should have less value for the homolytic O-H bond dissociation enthalpy ( BDE ) than ROO-H and R-H . Moreover , antioxidant radical should be kinetically stable to prevent its reaction with substrate [21 , 19] . Thus , the antioxidant power is not an absolute property of Ar-OH , but depends on the substrate which should be protected . The toxicity of 3HOK depends mainly on the products of its oxidative dimerization , such as hydrogen peroxide , xanthommatin ( XAN ) , 4 , 6-dihydroxyquinolinequinonecarboxylic acid ( DHQCA ) , their active free radical forms , and o-aminoquinone [22] . Ommochromes XAN and dihydroxanthommatine ( DXAN ) , the brown eye pigments , easily transform into each other under physiological conditions [23 , 24] . DXAN is synthesized from 3HOK by phenoxazinone synthetase ( PHS ) –the process disturbed by the cardinal mutation [23] . PHS catalyzes two consecutive abstractions of H-atoms from the hydroxyl group of o-aminophenols , 3HOK or 3HAA , followed by their non-enzymatic condensation to phenoxazinone [25] . The formation of ommochromes can also result from non-enzymatic oxidation of 3HOK [26] . High concentration of 3HOK catabolite hydrogen peroxide induces apoptotic cell death in neuronal cell cultures [27] . 3HOK and 3HAA generate superoxide anion and hydrogen peroxide in the presence of copper–the process leading to the formation of a quinoneimine structure [28] . Both amino and hydroxyl aromatic groups are important for lowering 3HOK and 3HAA oxidation potential . Initially , they can be two-electron donors with antioxidant activity , but their quinoneimine products are highly reactive and damage cell structures . Pro- and antioxidant power of o-aminophenols depends on the whole activity of the redox systems in cell [29] . Other kynurenine metabolites also possess pro- and antioxidant activity [4] . In particular , KYNA is able to scavenge hydroxyl radicals , superoxide anion radicals , and peroxynitrite , decreasing lipid peroxidation and ROS formation [30] . QUIN affects the ROS level only together with iron ions; the pro- and antioxidant effects of QUIN are concentration-dependent [31] . Free radicals scavenging mechanisms shown for non-o-aminophenol kynurenines include electron transfer , metal ion chelation , destruction of carbon skeleton , and radical addition to the aromatic ring [4] . Whereas experimental data regarding chemical and physiological properties of kynurenines are abundant and diverse , there are few computational studies on kynurenines . Quantum chemical calculations could provide a better understanding of the mechanisms of kynurenines' antioxidant activity . In this study , the redox properties of kynurenines and several synthetic phenolic antioxidants were investigated computationally using density functional theory ( DFT ) approach . The validity of B3LYP methods to model phenolic antioxidants and free radical reactions has already been proved [18 , 32] . The methodology was similar to that of [18 , 33]: the energies of frontier highest occupied and lowest unoccupied molecular orbitals ( EHOMO , ELUMO ) , phenolic O-H bond dissociation enthalpy ( BDE ) , and ionization potential ( IP ) were calculated and compared for structures fully optimized in the gas phase . We also studied the influence of water solvation on the chemical properties of antioxidants . Finally , we modeled the kinetic behavior of hydroxykynurenines interaction with phenoxyl and peroxyl radicals . Optimal geometries for kynurenines and synthetic antioxidants with substituted phenolic groups were calculated at different levels ( Table 1 ) . Six compounds with experimentally known BDE values are used as standards for the estimation of the validity of computational methods . Despite high diversity of chemical structures , Pearson correlation coefficient R is high for level II ( 0 . 870 and 0 . 867 for BDE and BDECOR , respectively; p<0 . 05 ) and III ( 0 . 865 , 0 . 863; p<0 . 05 ) , being less for level I ( S1 Table ) ( 0 . 717 , 0 . 710; p > 0 . 1 ) . Total spin <S2> shows small spin contamination ranging from 0 . 75 to 0 . 80 for all free radicals ( II , III ) , being abnormally high for some radicals calculated at level I . Thus , ( I ) computational data were omitted from further analysis . BDE and BDECOR ( II , III ) for phenol are greatly higher than the experimental value and the value previously calculated at the same level of theory; the cause is explained in Methods section . With the exception of phenol , the correlation of BDE/BDECOR with experimental values is very strong ( R = 0 . 959 and 0 . 974 for level II and III , respectively; p < 0 . 05 , n = 5 ) . The goal of this study was not the precise calculations of energy values , but rather the comparison of such values for different antioxidants . Thus , DFT calculations at level II or III can be used to predict the relative antioxidant power of the studied compounds . The values of EHOMO , ELUMO , BDE , and BDECOR at level II are highly correlated with those calculated at level III ( R = 0 . 98–1 . 00 ) In general , level III gives a slightly higher BDE/BDECOR than level II . BDE/BDECOR calculated at B3LYP and HCTH/407 levels of DFT ( basis set II ) are strongly correlated ( R = 0 . 999 and 0 . 998 , respectively; p < 0 . 05 , n = 16 , without phenol ) . HCTH/407 gives slightly lower values of BDE than B3LYP ( S1 Table; ΔE = 2 . 252±0 . 751 kcal/mol ) . For phenol , BDEHCTH/407 is 84 . 702 kcal/mol , which is much closer to the experimental value . Thus , both functionals can be used to estimate BDEs for kynurenines and phenolic antioxidants . The rankings for O-H homolytical BDECOR are nearly the same at levels II and III . O-H bond is the strongest in water and the weakest in negatively charged 3HAACO2- . BDECOR values for 3HOK and 3HAA are close to that for 2-aminophenol , their structural precursor . 2-aminophenol is an antioxidant with a large decrease in the O-H BDE compared to phenol [38] . L-3HOK and D-3HOK optical isomers have almost equal BDE values . 3HOK and 3HAA are characterized by the decreased energies of H abstraction compared to phenol and its derivatives DIBP and DTBP , both native and modified by propenoic acid ( DIBA , DTBA ) . Total energy for XAA oxo form is lower by 7 . 4 kcal/mol than for enol form ( level III ) ; therefore , we used the oxo form in the majority of calculations . XAAOXO is close to phenolic antioxidants in its H donating properties . B3LYP , as well as most DFT methods , is known to give EHOMO and ELUMO in a very poor agreement with experiment , significantly underestimating H-L gap . Using of tuned range-separated hybrid functionals can solve this problem [39 , 40] . We have computed EHOMO , ELUMO , H-L gap , and IP for five compounds optimized at level III , B3LYP ( L-3HOK , 3-HAA , XAAOXO , 2-aminophenol , and DTBP ) using tuned LC-BLYP range-separated functional . Indeed , LC-BLYP gives significantly higher absolute values for H-L gap ( ΔELC-BLYP–B3LYP = -69 . 849±8 . 135 kcal/mol ) , and IP values are close to -EHOMO ( ΔE-HOMO–IP = -3 . 333 ±2 . 929 kcal/mol ) . The optimal range-separation parameter μ values are close for four aminophenols ( ~0 . 20 ) and differ from that for phenolic antioxidant DTBP . At the same time , EHOMO , ELUMO , and H-L gap values calculated by III ( LC-BLYP ) and III ( B3LYP ) are highly correlated ( R = 0 . 911 , 0 . 989 and 0 . 955 , respectively ) , thereby the rankings for electron donating power are virtually the same in both cases . XAAOXO has the highest values for IP and -EHOMO , whereas L-3HOK is the best electron donor among the uncharged hydroxykynurenines . There is a moderate negative correlation between EHOMO and BDE/BDECOR ( levels II , III ) ( S2 Table ) . The correlation between H-L gap and BDE/BDECOR is even stronger . Hence the ability of O-H bond homolytical dissociation tends to increase along with the lowering of EHOMO and H-L gap absolute values . The correlation between ELUMO and BDE/BDECOR is not significant at all levels . For the compounds with an ionized group , such as 3HAACO2- , XAACO2- , and L3HKNH3+ , EHOMO significantly differ from those of uncharged compounds . 3HAACO2- and XAAOXO/CO2- are more powerful H donors than the uncharged forms . On the contrary , protonation of NH2 group in L-3HOK phenolic ring significantly complicates O-H dissociation . This is in agreement with the fact that electron-donating groups reduce O-H BDE , thus enhancing antioxidant activity , whereas electron-withdrawing substitutions raise it [41 , 42 , 18] . DIBP and DTBP , the substances with skeletal isomerism , have similar BDEs . However , for each of them , BDE is closer to that of its propenoic derivative than to BDE of its isomer . Hence the side chain isomerism significantly affects the H- donating properties of phenolic group . DXAN has the least stable O-H bond among the uncharged compounds , making it a potent anioxidant with the high H donating ability . In order to check the possible effect of the basis set superposition error ( BSSE ) on BDE , BSSE correction was performed for phenol , DIBP , and DXAN with a small , intermediate , and large hydrocarbon moiety of a radical . The values of BSSE ( III ) are -0 . 722 , -0 . 832 and -1 . 45 kcal/mol , respectively , being small and similar in all compounds . The decrease of BDE for bulky antioxidants cannot be explained by the growth of BSSE . The geometry of frontier molecular orbitals and spin-orbits was calculated at level III for hydroxykynurenines and their precursors , as well as for their derivatives without an aromatic hydroxyl group . The highest occupied molecular orbital ( HOMO ) of phenolic antioxidants and kynurenines is localized mainly on the phenolic ring . HOMO is divided into two parts: the first part occupies phenolic OH group and three approximate C atoms , and the second part occupies the opposite two or three C atoms ( Fig 2 ) . HOMO also occupies unsaturated and polar groups attached to the phenolic ring , such as the aromatic NH2 group of L-3HOK and 3HAA , which HOMO's and spin-orbit's geometry is virtually the same as that for 2-aminophenol . Together with the aromatic rings , OH and NH2 groups form a π-conjugated systems known to decrease IP [43] . DXAN has the largest conjugated system allocated mainly to phenoxazinone structure , which possibly facilitates H-atom and electron abstraction . For the ionized compounds , HOMO is moved from the aromatic hydroxyl group to the charged group . BDE , IP and frontier orbital energies for compounds optimized at level III were calculated at level IV ( B3LYP/6-311+ ( O ) +G ( d ) ) in the gas phase and water solution ( Table 2 , Fig 3A and 3B ) . A moderate negative correlation between BDE and EHOMO/H-L gap values was observed , as well as for levels I-III , both in the gas phase and in water solution ( S2 Table , second part ) . A strong correlation between adaibatic IP and -EHOMO or the so called vertical IP can be seen . This is in accordance with Koopmans' theorem , applicable in high approximation for outer valence Kohn-Sham orbitals [44] . The difference between IP and -EHOMO is 36 . 7±3 . 3 kcal/mol in the gas phase , in agreement with the fact that B3LYP underestimates the absolute values for EHOMO [39] . In water solution , IP becomes slightly lower than the negative of EHOMO . BDE is positively correlated with IP; thereby the electron and H donating capacities of the studied compounds are interrelated . BDE of aromatic antioxidants is strongly correlated with standard deviation of Mulliken spin density ( δSD ) on radical ( RGAS = 0 . 879 , RWATER = 0 . 917; p < 0 . 05 , n = 14 ) and spin density ( SD ) on radical O* atom ( RGAS = 0 . 910 , RWATER = 0 . 951; p = 0 . 05 , n = 14 ) after H abstraction . BDEs for phenol and DTBP were not considered due to significant deviations from experimental values . Also , BDE is strongly correlated with SD on radical CPARA aromatic atom after H abstraction ( RGAS = 0 . 849 , RWATER = 0 . 876; p < 0 . 05 , n = 13 , without KYNAENOL which has N instead CPARA ) . There is no significant correlation between IP and δSD for kynurenine radicals after single electron abstraction . Thus , electron delocalization on kynurenines seems to be more important for H-atom donation activity than for the electron donation activity . OH group bonded to the aromatic ring significantly increases the ability of kynurenines to donate H-atom and electron . L-KYN C3-H BDE is much higher than L-3HOK O3-H BDE: the difference is 40 . 4 kcal/mol in the gas phase and 40 . 3 kcal/mol in water solution ( level V: B3LYP/6-311++ ( d , p ) ) . IP value is lower for XAAOXO , L-3HOK , and 3HAA than for KYNAOXO , L-KYN , and AA , respectively ( ΔIPGAS = 6 . 4±2 . 9 kcal/mol , ΔIPWATER = 5 . 5±0 . 6 kcal/mol ) . However , the relative IP rankings are the same for compounds with and without OH group: KYNA has higher IP compared to L-KYN and AA , as well as XAA compared to L-3HOK and 3HAA . This is in agreement with experimental data: the electrochemical potential Epa for kynurenines has been shown to decrease in the following rankings: KYNA > KYN > AA > 3HOK > 3HAA [29] . There is a strong positive correlation between experimental Epa and IP calculated for non-ionized compounds in water solution at level IV ( R = 0 . 924; p < 0 . 05 , n = 5 ) . For compounds with an ionized CO2 group , the correlation is not significant ( R = 0 . 804; p > 0 . 1 , n = 5 ) , probably , due to the small sampling . QUIN is the least powerful electron donor among the uncharged kynurenines . IPGAS for XAN is close to that for 3HAA , hence XAN easily abstracts electron , but not H-atom . This possibly makes it a prooxidant with toxic effects [22] . The electron-donating substituents are known to decrease IP and to increase the antioxidant activity [33] . In general , compounds with ionized CO2 group have lower IP and higher EHOMO values than their neutral forms ( ΔIPGAS = -88 . 5±11 . 5 kcal/mol , ΔIPWATER = -8 . 7±9 . 5 kcal/mol , without KYNAOXO; ΔEHOMO/GAS = 88 . 6±8 . 4 kcal/mol; ΔEHOMO/WATER = 7 . 4±1 . 6 kcal/mol ) . KYNAOXO in water solution has lower IP than KYNAOXO/CO2 . Probably , geometry optimization of charged compounds in the gas phase leads to some distortions in KYNAOXO/CO2 structure . In water solution , IP becomes lower for the majority of compounds and higher for the anionic forms . This seems to result from a high dielectric capacity which decreases electrostatic interactions , stabilizes anions , and diminishes electron attraction to cations and neutral molecules . The change of gas–water -EHOMO and -ELUMO is correlated with the change of IP . BDE for XAAOXO/CO2- becomes higher than that for 3HAA and only 2 kcal/mol less than that for XAAOXO . The same trend is observed for 3HAA/3HAACO2- and L-3HOK/L-3HOKNH3+ . Thus , water solution significantly diminishes the influence of charged groups on BDEs and IPs . There are different pathways for ROS inactivation by antioxidants [19] . Most likely , kynurenines quench radicals by donating aromatic hydroxyl H-atom to radical O-group . We have computationally studied the kinetics of this process for the complexes of four hydroxykynurenines , 3HAA , L-3HOK , XAAOXO , and XAAENOL , with phenoxyl radical ( Ph-O ) and methyl peroxy radical ( Met-OO* ) . BDE difference for Met-OO* and buthyl peroxy radical is less than 0 . 7 kcal/mol ( levels II , III ) , hence Met-OO* can be used instead of the radicals with long aliphatic chain to simplify calculations . Ph-O–DTBP and Ph-O–DTBA complexes have been also calculated , as well as Met-OO* complex with XAA in ionized form . TSs for reaction pathways were located at level II . Reagent and product complex structures are in good agreement with the results of IRC calculations ( RMSD = 0 . 026±0 . 015 Å for all complexes and 0 . 019±0 . 008 Å for kynurenines' complexes ) . The values for the reaction rate and height of activation barrier were calculated in the gas phase and water solution ( Table 3 ) . k ( T ) values are significantly higher than those experimentally shown for phenolic compounds with BDE values of 70–80 kcal/mol , which are about 104−107 M-1s-1 [19] . This fits the fact that B3LYP underestimates the reaction barrier heights , whereas functional XYG3 is almost as accurate , as the highly precise CCSD ( T ) method [45] . It is rather difficult to calculate the exact value of the reaction rate , as multiple factors should be considered , and appropriate DFT level should be used [46] . However , the location of TS point calculated for 3HAA–Met-OO* and XAAOXO−Met-OO* by B3LYP ( level II ) is similar to that calculated by XYG3 ( level V ) ( S4 Fig ) . Thus , even B3LYP with the relatively small basis set II correctly describes the geometry of TS structure . ΔETS-R ( XYG3 ) for XAAOXO complex is higher than that for 3HAA complex . For both Ph-O* and Met-OO* , k ( T ) increases in the following rankings: XAAENOL < XAAOXO <3HAA ~ L-3HOK ( Table 3 , Fig 3 ( C ) ) ; the same rankings applies to -ΔEP-R . The structures of radical complexes with 3HAA , L-3HOK , and XAAOXO are very similar ( Fig 4 , Table 4 ) . For kynurenines in complex with Ph-O* , aromatic rings of reagents and products form the plane angle of ~50–70° . The geometry of Ph-O–DTBP is very different: aromatic rings are nearly perpendicular in reagent and product complexes . Ph-O–DTBP and Ph-O–DTBA complexes are rather similar . For 3HAA , L-3HOK , and XAAOXO in complex with Met-OO , the radical rotates in space along with the attachment of H-atom , so O-O* and C-O bonds in Met-OOH become nearly perpendicular to those in reagent complexes . The direction of Met-OO* rotation is different in complexes with XAAENOL and XAAOXO/CO2 . In all cases , O…H…O bond significantly shortens upon the TS formation . The influence of solvent and partial charges' distribution on antioxidant activity depends on whether HAT or PCET is the dominant mechanism of H transfer . The increase in H-atom charge in the TS compared to the parent antioxidant is specific for PCET [20] . The interaction of phenolic antioxidants with tert-buthyl-peroxy radical is known to occur via PCET [47] . PCET-TS is stabilized by the enhanced spin density ( SD ) and electron density on radical O2 and O3 atoms . Thereby ( O3+O2-O1 ) negative charge and Δ ( O3+O2-O1 ) TS-R negative charge correlate with the reaction rate [47] . In our study , positive charge on H-atom increases in all TSs ( ΔQ ( H ) > 0 ) , and the negative charge on O atoms in the gas phase moves towards the free radical ( Δ ( dQ ) TS-R < 0 ) ( Table 5 ) . In the gas phase , there is a strong correlation between ΔETS-R and SD on O1 and CPARA atoms of antioxidants . Hence the high SD on these atoms decreases the reaction rate . For Ph-O* complexes in water solution , the decrease of the negative charge on radical O atoms ( Δ ( dQ ) TS-R ) correlates with the growth of ΔETS-R , as it is typical for PCET . There is a negative correlation between ΔETS-R and ETS-SOMO . Thus , ESOMO may serve to predict the reaction rate , as shown by Nikolic [47] . The geometry of TS SOMO and spin-orbit on O1 and O2 differs from both classical σ- and π-orbitals: p-orbitals on O atoms form a sharp angle projected to plane passing through H atom perpendicular to O…H…O bond ( Fig 5 ) . In Ph-O–kynurenines' complexes , p-orbitals are nearly parallel to this plane and perpendicular to O…H…O bond . In Ph-O–DTBP complex , O1 and O2 protrude parts of the electron clouds towards H , and in Met-OO–kynurenines' complexes , the angle between O…H…O and O1 p-orbital is close to 45° . Hence kynurenine's SOMO in Ph-O complexes is closer to π-orbital than in Met-OO* complexes . Partial charges Q and ΔQ on H are higher , and the negative charge displacement to O1 is lower for Ph-O* complexes than for Met-OO* complexes . Thus , PCET seems to be more preferable for kynurenines' reaction with Ph-O* than for their reaction with Met-OO* . HAT may also occur in both cases , however , SOMO geometry and charges distribution character indicate that it is not the chief mechanism of H transfer for the studied complexes . Another possible way of free radical quenching is its addition to the aromatic ring of the antioxidant radical . We have modeled the products of Met-OO* addition to the aromatic ring of phenoxyl and kynurenines radicals at para-position relative to O* atom ( Fig 6 , Table 6 ) . The orientation of side chain Met-OO group varies , being closer for the different forms of XAA than for the different antioxidants . In the gas phase , all reactions are thermodynamically favorable ( ΔEP-R/COR < 0 ) , in contrast to H abstraction . In water solution , radical addition to 3HAA and L-3HOK radicals is slightly unfavorable . The rankings of -ΔEP-R and -ΔEP-R/COR are the same at all levels: 3HAA* < L-3HOK* < XAAOXO/CO2-* < XAAOXO* < DTBP* < XAAENOL* < Ph-O* . It is reverse to the rankings of -ΔEP-R and k ( T ) for H-atom donation: the affinity to Met-OO* is minimal for 3HAA* and maximal for phenoxyl radical . XAAOXO/CO2-* is less active than XAAOXO* and more active than L-3HOK* , both in the gas phase and in water solution . Thus , high Met-OO* scavenging activity of L-3HOK and 3-HAA is unlikely to be explained by Met-OO* addition to the aromatic rings of kynurenine radicals . The antioxidant power of a substance depends not only on its chemical properties , but also on its ability to penetrate into the surroundings where it displays its antioxidant activity . To inhibit lipid peroxidation , a substance should have high lipophilicity . It can be measured as a logP value , where P is the octanol-water partition coefficient [48] . We used the Molinspiration method of logP calculation reported to be robust and precious . Among the antioxidants studied , substituted phenols , such as DTBP and DTBA , have maximal lipophilicity , whereas the kynurenines' ions have higher water solubility compared to ASC ( Table 7 ) . Lipophilicity decreases in the following rankings: AA > 3HAA> XAAOXO > QUIN > L-3HOK; the rankings are the same for kynurenines' carboxylic anions . Hence XAA should penetrate through lipid bilayer better than 3HOK and 3HAA . This fact does not fit with the lower rate of XAA reaction with peroxy radicals , which is rather explained by the higher rate of H donation . Topological polar surface area ( TPSA ) is a molecular descriptor numerically close to PSA which can be used for the prediction of passive transport through membranes in intestines and blood-brain barrier [49] . Drugs that penetrate the brain by passive absorption typically have PSA < 70 Å2 , while the most non-CNS active drugs have much larger PSA values up to 120 Å2 [50] . In our study , TPSA is less than 120 Å2 for all compounds except L-3HOK , so they are more or less capable of penetrating passively through the plasma membrane . TPSA is minimal for phenolic antioxidants , which should be easily absorbed in intestines and penetrate into the brain . It is significantly higher for 3HOK ( > 120 Å2 ) than for 3HAA ( < 90 Å2 ) , while XAA has the intermediate TPSA . Hence in the case of absence of specific carriers 3HAA should more actively penetrate through lipid bilayer than 3HOK . A great need in studies of biochemical properties of KP is promoted by the fact that this very pathway plays an overwhelming role in physiology and pathology . Dysregulation of this pathway , resulting in hyper- or hypofunction of active metabolites , is associated with neurodegeneration and other disorders , such as depression and schizophrenia [51] , diabetes mellitus [52 , 53] , attention-deficit hyperactivity disorder [54] , and cataract [13] . Some KP metabolites are neuroactive , while others are molecules with prooxidant and antioxidant properties [3] . Therefore , it is necessary to understand the molecular and biophysical mechanisms of kynurenines' activity to elaborate the strategy of disorders' prevention and therapy . In this study , we investigated the antioxidant activity of kynurenines , namely their ability to donate electron and H atom . The hydroxyl group BDE and adiabatic IP are the most important determinants for the radical scavenging activity of substituted phenols [55] . According to our data , the antioxidant properties of 3HOK and 3HAA are determined by their 2-aminophenolic moiety . For the uncharged hydroxykynurenines , BDE and IP are maximal for KYNAENOL and miminal for DXAN , both in the gas phase and in water solution . 3HOK and 3HAA have lower BDE and IP than XAA , ascorbic acid , and some phenolic antioxidants , such as DTBP and DTBA . Aromatic OH group diminishes IP values for 3HOK , 3HAA , and XAA relative to KYN , AA , and KYNA . Our results confirm the correlation between BDE and IP also shown by Borges [56] . Negatively charged carboxylic group significantly diminishes BDE and IP values , while positively charged amino group enhances them . This phenomenon can be explained by electron-donating and withdrawing effects of substituents [18] . The effects of charged groups are significantly more pronounced in the gas phase than in water solution . Basis set and the type of density functional had a little effect on the rankings for BDE values of kynurenines . Adiabatic IP strongly correlates with -EHOMO , confirming that Koopmans' theorem can be used to calculate IP at DFT level [44] . However , B3LYP significantly underestimates the absolute values of EHOMO . We have used the tuned LC-BLYP range-separated functional to compute EHOMO , ELUMO , H-L gap , and IP for several antioxidants , including three hydroxykynurenines . The tuned LC-BLYP gives significantly higher absolute values for EHOMO and H-L gap , also , there is a minor difference between -EHOMO and adiabatic IP . However , the rankings for kynurenines' EHOMO values are the same as for those calculated using B3LYP . Also , BDEs calculated with B3LYP and HCTH/407 are highly correlated . B3LYP is significantly faster than the high quality functionals partly based on perturbation theory , such as XYG3 . B3LYP contains less empirical parameters than HCTH/407 , thereby it seems to be more universal . B3LYP has been successfully used to model both thermodynamic and kinetic properties of free radicals [18 , 32 , 33] . Thus , we used B3LYP in the majority of our calculations . At the same time , using LC-BLYP and other range-separated functionals may be favorable to predict the exact values for frontier orbital energies . High radical stability and even spin distribution are among the factors predisposing low BDE and IP values [55] . Conjugated bonds system facilitates electron delocalization after HAT or single electron transfer ( SET ) . Standard deviation of SD in kynurenine radicals is correlated with BDE; however , there is no significant correlation between SD and IP . BDE for kynurenines is correlated with SD on O* and CPARA atoms . SD on antioxidant O* atom in TS complex with Met-OO* is also strongly correlated with the height of the activation barrier . The same is true for OPARA atom in phenolic antioxidants [47] . The rate of H-atom donation to phenoxyl and methyl peroxy radicals is correlated with BDE: 3HAA and 3HOK are more active radical scavengers than XAA . Likewise , for phenolic compounds donating H to hydroxyl radical , the rate constant is negatively correlated with O-H bond straight , IP , and SET enthalpy [57] . The rankings for free energies of radical addition to kynurenine radicals in para-position relative to OH group is reverse: 3HAA* radical has the lowest affinity to Met-OO* . Thus , high antioxidant activity of 3HAA and 3HOK relative to XAA [15] rather can be explained by their lower BDEs and higher rates of H-atom donation to peroxy radical . PCET seems to be the chief mechanism for H donation by kynurenines to phenoxyl radical and , probably , to Met-OO* radical . Studies of oxidations of O-H bond usually invoke stepwise oxidation , where the positive and negative charges are transferred separately . Here , there may be a complex dependence of BDE and k ( T ) on solvent , biochemical surroundings , and pH [58] . Low BDE and IP values are not sufficient for Ant-OH to be a powerful antioxidant without toxic side effects . Some of the necessary conditions include: 1 ) O2 should not abstract H from Ant-OH; 2 ) Ant-OH should react with ROO* much faster than ROO* with R-H; 3 ) Ant-O* should not abstract H from R-H at an appreciable rate; 4 ) Ant* should not react with O2 to produce AOO; 5 ) Ant-OH and its products should not be toxic [21] . Though we did not concentrate on the study of these properties , our calculations performed for 3HAA at level II have shown that: Hence O2 can abstract H from 3HAA , but the reaction is dramatically slower than H abstraction by Met-OO ( 2 . 2 x 1011 M-1s-1; Table 3 ) . The aliphatic ethane interacts with Met-OO* faster than with 3HAA , and 3HAA interacts with Met-OO* dramatically faster than with ethane . Only the reaction of 3HAA with Met-OO* is thermodynamically favorable , so O2 or free radicals must be in high concentration to hinder the protective action of 3HAA . Therefore , it seems to be a potent antioxidant , as well as 3HOK . At the same time , the ability of 3HAA and 3HOK to form dimers leads to the production of toxic free radicals which damage the cell [23 , 26 , 22] . 3HAA undergoes three successive one-electron oxidative reactions: 1 . conversion to semiquinoneimine ( hydroxyl H abstraction ) ; 2 . conversion to quinoneimine ( amine H abstraction ) ; 3 . two quinoneimine molecules condensation to cinnabarinic acid [59 , 60] . The rate of 3HAA oxidation increases exponentially with increasing pH [59] . This corresponds to our data that BDE decreases and k ( T ) for H-atom donation increases for compounds with ionized carboxylic group . 3HOK autoxidation is similar to that of 3HAA [29] . BDE for 3HOK N-H is significantly higher than for 3HOK O-H , being 99 . 5 and 77 . 2 kcal/mol , respectively ( level III ) . Both of them are smaller for semiquinoneimine , becoming 91 . 8 and 69 . 4 kcal/mol after the other H-atom abstraction . Hence the first stage of oxidation facilitates the second one . o-Quinoneimine which is synthesized at the second stage may be responsible for the prooxidant effects of 3HOK and 3HAA [29] . The enzymatic oxidation of o-aminophenols leads to the concomitant reduction of oxygen to water [25] . Non-enzymatic oxidation produces the toxic reactive forms of oxygen . Therefore , it may be therapeutically important to enhance the antioxidant power of hydroxykynurenines by inhibiting their non-enzymatic dimerization and/or stimulating the enzymatic dimerization . The inhibition of tryptophan 2 , 3-dioxygenase , the key enzyme of KP , is neuroprotective in Drosophila huntingtin ( htt ) mutant . Feeding flies by 3HOK alone , in the absence of mutant HTT , did not cause neurodegeneration [14] . Thus , the high level of 3HOK is toxic , yet , it may be not sufficient for neurodegeneration which also requires the additional factors , such as the lack of neuroprotectant KYNA . Both 3HOK and 3HAA inhibit the spontaneous lipid peroxidation in the brain [61] . The dual redox activity of 3HOK makes it prooxidant at low concentrations ( 5–20 μM ) and antioxidant at higher concentrations ( 100 μM ) in the rat striatum slices . 3HOK seems to be a redox modulatory molecule which stimulate the increase in glutathione reductase and glutathione S-transferase activities [62] . Interferon-γ induces TRP degradation along the KYN pathway in mononuclear blood cells and inhibits the oxidation of low density lipoprotein ( LDL ) . 3HAA inhibits LDL oxidation in submicromolar concentrations , probably being a catalyst for the other antioxidants [63] . It is a highly efficient coantioxidant for plasma lipid peroxidation which can be initiated by α-tocopherol radical ( α-TO* ) . 3HAA in low concentration ( 5μM ) inhibits α-TO* production and accumulation of lipid peroxides . 3HOK inhibitory efficacy is the same as for 3HAA , but AA lacking the phenolic group can not reduce α-TO* [16] . This fits with low BDE for kynurenines having phenolic group . Monocytes in human blood can release 3HAA in concentration up to 30 μM [63] . Thus , 3HOK and 3HAA display the antioxidant activity under physiological conditions , not only in the brain , but also in blood plasma regulating the process of atherogenesis . In contrast to 3HAA , the antioxidant properties of XAA are in relation to its ability to chelate the transition metals which induce LDL oxidation [64] . Thus , kynurenines' action on redox conditions and physiological processes depends on their level in organism . The lack of kynurenines in Drosophila mutant vermilion , as well as the excess of 3HOK in cardinal , leads to the progressive loss of 3 h memory performance under conditioned courtship suppression paradigm [65 , 12] . The lack of kynurenines redox activity might partially cause these effects . Not only BDE and IP define the antioxidant power of substances , but also their ability to pass through the biological barriers , mainly the lipid bilayers . The lipophilicity of kynurenines is low , compared to phenolic antioxidants , due to their polar and charged groups . Hence they should rather act in water environment than in membrane . Besides , their surfaces are quite large , that should hamper their penetration through intestine and blood-brain barriers . Indeed , 3HAA , KYNA , and QUIN poorly cross the blood-brain barrier by passive diffusion , but KYN and 3HOK are taken up into the brain by a large neutral amino acid carrier [66] . AA easily penetrates into the brain by passive diffusion that can be explained by its high logP and low TPSA values . Kynurenine pathway enzymes in the brain are preferentially localized in astrocytes and microglia; however , the cerebral pathway is driven mainly by blood-borne KYN [2] . Thus , 3HOK , KYN , and AA may play an important role in the brain both as prooxidants and antioxidants . In our study , we did not consider several important factors affecting the antioxidant power of kynurenines , such as: 1 ) thermodynamics and kinetics of 3HOK and 3HAA dimerization; 2 ) energy and rate of proton abstraction from antioxidant OH group; 3 ) interaction between solvent and OH group of kynurenines; 4 ) hydrogen bond formation between functional kynurenine groups; 5 ) steric effects of side-chain groups on free energy and rate of kynurenine interaction with radicals , etc . Other functional groups of kynurenines can also donate H-atom , such as the 3HOK aromatic NH2 group , which BDE was shown to be significantly higher than that for OH group . It would be interesting to evaluate the activity of kynurenines and phenolic antioxidants in their native surroundings , such as lipid bilayer , affecting the dielectric capacity and hydrophobic interactions . Consideration of these factors is a task for the future . The structures of 2-aminophenol , anthranilic acid ( AA ) , ascorbic acid ( ASC ) , kynurenic acid ( KYNA ) , L-kynurenine ( L-KYN ) , D-3-hydroxykynurenine ( D-3HOK ) , L-3-hydroxykynurenine ( L-3HOK ) , quinolinic acid ( QUIN ) , 2 , 6-di-tert-butylphenyl-4-hydroxymetylphenol , xanthommatin ( XAN ) , dihydroxanthommatin ( DXAN ) , and xanthurenic acid ( XAA ) were taken from PubChem Compound database [67] . The structures of phenol , 2 , 6-di-isobutylphenol ( DIBP ) , 2 , 6-di-tert-butylphenol ( DTBP ) , β- ( 4-hydroxy-3 , 5-di-isobutylphenyl ) propenoic acid ( DIBA ) , β- ( 4-hydroxy-3 , 5-di-tert-butylphenyl ) propenoic acid ( DTBA ) , and 3-hydroxyanthranilic acid ( 3HAA ) were constructed on the base of PubChem structures using Vega ZZ 3 . 0 . 3 [68] . The systematic conformational search of low-energy geometry for the constructed structures was performed using Avogadro [69] . The ionic forms for kynurenines with α-carboxylic group ( total charge -1 ) were modeled as well as the uncharged forms . The major forms for 3HAA and XAA at physiological pH ( 7 . 4 ) are the forms with the ionized carboxylic group , while KYN and 3HOK are mainly in zwitterionic form with ionized α-amino and α-carboxylic group [70] . All quantum chemical calculations were performed using Firefly 8 . 1 . 0 partially based on the GAMESS ( US ) [71] source code . Firefly 8 . 1 . 0 was kindly provided by Alex A . Granovsky [72] . The geometries of molecular structures with neutral total charges were fully optimized using density functional theory ( DFT ) at 6-31G ( d ) level ( I ) , B3LYP/6-31G ( d ) level ( II ) , and B3LYP/6-311G ( d , p ) level ( III ) [73–75] . B3LYP1 version of B3LYP was used . Highly parameterized functional HCTH/407 [76] was also used to calculate BDE values for compounds at level II . Closed shell configurations were calculated with restricted Hartree-Fock or DFT methods; open shell configurations were calculated with unrestricted Hartree-Fock or DFT methods . All closed shell molecules were calculated in a singlet state , whereas doublet state was used for free radicals . The symmetry point group was set as C1 for all compounds . Hessian matrix , vibrational frequencies , and thermal corrections to the enthalpy were calculated with the same methods . The enthalpies and free energies were obtained from the vibrational frequency calculations at 298 . 15 K , using unscaled frequencies . In order to calculate the adiabatic ionization potential ( IP ) , cation-radical forms of corresponding molecules were fully optimized at level III . BSSE correction [77] was performed for several compounds at level III . The nature of all stationary points was determined by evaluating the vibrational frequencies . Standard deviation of Mulliken spin density ( δSD ) was used as an estimator of electron delocalization on the radicals . The following energy parameters were estimated: where ECAT is the energy of cation radical after single electron abstraction ( or the neutral form for compounds with ionized carboxylic group ) , EW is the total energy of the whole molecule . BDEs for methane , water , phenol , 2-aminophenol , water-soluble antioxidant ASC ( uncharged form ) , and phenolic antioxidant DTBP were used as standards and reference points to estimate the relative activities of antioxidants . For symmetric phenol and DTBP radicals , there were significant deviations of BDE from the experimental values . To exclude the possible artefacts , BDE was also calculated for several structural analogues of DTBP–DTBA , DIBP , and DIBA , which are believed to have similar BDE values . Zwitterionic form is not stable in the gas phase , therefore , the optimization of KYN and 3HOK in the neutral form was performed . To check the influence of positively charged group on BDEs and IPs , calculations were performed for L-3HOK with aromatic NH3+ group ( total charge +1 ) . 3HAA and AA cations with ionized carboxylic group are not stable in the gas phase , therefore , their optimization was performed in water solution at level IV ( see below ) , without cavitation , dispersion and repulsion free energies . NWChem software [78] was used to calculate EHOMO , ELUMO , H-L gap , and IP with the help of the tuned range-separated hybrid functional LC-BLYP for five compounds optimized with B3LYP ( III ) . The tuning of optimal range-separation parameter μ was done as in [40]: the single point energies were calculated using basis set III for antioxidant's cation , anion , and neutral form for different values of μ ranging from 0 . 05 to 0 . 9 with increments of 0 . 05 , and then the optimal parameter was obtained by minimizing the following function: J2 ( μ ) =[EμHOMO ( N ) +IPμ ( N ) ]2+[EμHOMO ( N+1 ) +IPμ ( N+1 ) ]2 ( 6 ) where N is the number of electrons in antioxidant . For XAAOXO , J2 ( 0 . 1 ) and J2 ( 0 . 15 ) were obtained by spline interpolation due to the problem with DFT convergence . The curves for J2 ( μ ) are shown in S5 Fig; the minimum of each curve ( optimal μ , see Table 1 ) was obtained by spline interpolation . Firefly 8 . 1 . 0 and Gaussian 98 [79] give almost equal values of total energy for phenol and significantly different values for phenoxyl radical . Gaussian 98 uses Harris functional for the initial orbital guess by default instead of extended Huckel calculations used by Firefly . Harris functional is a nonself-consistent approximation to Kohn-Sham density functional theory [80] , hence electron correlations should be partially taken into account . Nevertheless , the differences between Gaussian 98 ( S1 Table ) and Firefly ( Table 1 ) BDE values are generally small ( 0 . 092±0 . 03 kcal/mol; p < 0 . 05 , n = 16 , without phenol ) . For antioxidants in complex with phenoxyl radical ( Ph-O* ) and methyl peroxy radical ( Met-OO* ) , transition structures ( TSs ) and corresponding local minima were optimized at level II . Intrinsic reaction coordinates ( IRC ) calculations [81] were performed for all TS species at the same level to confirm that anticipated reagent ( R ) and product ( P ) are connected to TS on potential energy surface . The products of the Met-OO* addition to antioxidant radical in para-position were optimized at level II . ΔECOR is corrected reaction activation energy: ΔETS−R/COR=ΔETS−R+ΔGTS−R ( 7 ) ΔETS−R=ETS–ER ( 8 ) ΔGTS−R=GTS–GR ( 9 ) ΔEP−R/COR=ΔEP−R+ΔGP−R ( 10 ) ΔEP−R=EP–ER ( 11 ) ΔGP−R=GP–GR ( 12 ) where ΔGTS-R and ΔGP-R signify thermal correction to free energy at 298 . 15 K; ETS , ER , and EP are total energies of TS , R , and P; GTS , GR , and GP are thermal contributions to free energies of TS , R , and P . The rate of reaction ( M-1s-1 ) between antioxidant and radical was calculated as in [32 , 46] using conventional TS theory: k ( T ) =Ix ( kBT/h ) x[exp ( −ΔETS−R/COR/RT ) ]x24 . 3xA ( T ) ( 13 ) where I is the reaction pathway degeneracy ( equal to 1 for the all compounds ) , kB is Boltzmann's constant , h is Planck's constant , 24 . 3 is a multiplier used to convert the units from 1 atmosphere standard state to 1 M standard state , and A ( T ) is a temperature-dependent factor which corresponds to quantum mechanical tunneling , approximated by the Wigner method [82]: A ( T ) =1+ ( 1/24 ) x ( 1 . 44νi/T ) 2 ( 14 ) where νi is the imaginary frequency ( cm-1 ) whose vibrational motion determines the direction of the reaction . Atom coordinates of the optimized structures are given in S1 Dataset . For the structures optimized at level II ( TSs ) and III ( all other structures ) , the single point energy calculations were performed at B3LYP/6-311+ ( O ) +G ( d ) level with diffuse sp functions added only to O atoms ( level IV ) or at B3LYP/6-311++G ( d , p ) level ( V ) , both in the gas phase and in water solution at 298 . 15 K using dielectric polarizable continuum model ( DPCM ) [83] . Due to the DFT convergence problem , the calculations were performed at level V only for five compounds ( L-KYN , L-3HOK , 3HAA , KYNA , and AA ) . Pearson correlation coefficient R for IPs calculated at levels IV and V is 0 . 999 . Hence the lack of H ( p ) polarization functions and C , N ( sp ) diffuse functions at level IV did not change the rankings of IP values for different kynurenines . Single point energy calculations were performed for 3HAA–Met-OO* and XAAOXO−Met-OO* IRC at level V in the gas phase using XYG3 functional [45] . The values of the total free energy in solvent were used to calculate ΔE for the compounds in water solution . Since the value of the thermal correction to BDE ( III ) was very similar for different compounds ( -6 . 645±0 . 260 kcal/mol , p < 0 . 05 , n = 16 ) , it was not considered . Also , the value for the thermal correction to antioxidant IP ( III ) was small ( -0 . 20±0 . 25 kcal/mol , p < 0 . 05 , n = 21 ) and was not considered . For the TSs , the values of ΔGTS-R and νi obtained at level II were used to calculate the values of ΔECOR and k ( T ) at level IV . Statistical analyses were performed using Social Science Statistics online resource [84] . Illustrations were prepared with the help of MaSK 1 . 3 . 0 . [85] and VMD [86] . The lipophilicity ( logP ) of compounds was calculated using the Molinspiration server [87] .	Kynurenines , the tryptophan metabolites with multiple biological activities , regulate the production of reactive oxygen species ( ROS ) during several neurodegenerative diseases . Many experiments show that kynurenines can be both prooxidants and antioxidants depending on their concentration , mode of action , and cell redox potential . However , there is lack of computational studies of kynurenines properties which could help us better understand the biophysical mechanism of their antioxidant activity . We performed the computations of kynurenines' hydrogen and electron donating power , both in the gas phase and in water solution . We found that aromatic hydroxyl group facilitates hydrogen and electron abstraction by kynurenines , in agreement with experimental data and computations earlier performed for phenolic antioxidants . We revealed the correlations of kynurenines' antioxidant power with their electronic structure , charge , and surroundings . We also found that 3-hydroxykynurenine and 3-hydroxyanthranilic acid can fastly quench free radicals by hydrogen atom donation . Hence both of them are potent antioxidants . The therapeutic strategy may be to inhibit their oxidative dimerization leading to ROS production .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "chemical", "compounds", "density", "functional", "theory", "solutions", "enthalpy", "materials", "science", "chemical", "radicals", "quantum", "mechanics", "thermodynamics", "materials", "by", "structure", "physical", "chemistry", "chemistry", "free", "radicals", "oxidat...	2016	Antioxidant Properties of Kynurenines: Density Functional Theory Calculations
Staphylococcus aureus requires branched-chain amino acids ( BCAAs; isoleucine , leucine , valine ) for protein synthesis , branched-chain fatty acid synthesis , and environmental adaptation by responding to their availability via the global transcriptional regulator CodY . The importance of BCAAs for S . aureus physiology necessitates that it either synthesize them or scavenge them from the environment . Indeed S . aureus uses specialized transporters to scavenge BCAAs , however , its ability to synthesize them has remained conflicted by reports that it is auxotrophic for leucine and valine despite carrying an intact BCAA biosynthetic operon . In revisiting these findings , we have observed that S . aureus can engage in leucine and valine synthesis , but the level of BCAA synthesis is dependent on the BCAA it is deprived of , leading us to hypothesize that each BCAA differentially regulates the biosynthetic operon . Here we show that two mechanisms of transcriptional repression regulate the level of endogenous BCAA biosynthesis in response to specific BCAA availability . We identify a trans-acting mechanism involving isoleucine-dependent repression by the global transcriptional regulator CodY and a cis-acting leucine-responsive attenuator , uncovering how S . aureus regulates endogenous biosynthesis in response to exogenous BCAA availability . Moreover , given that isoleucine can dominate CodY-dependent regulation of BCAA biosynthesis , and that CodY is a global regulator of metabolism and virulence in S . aureus , we extend the importance of isoleucine availability for CodY-dependent regulation of other metabolic and virulence genes . These data resolve the previous conflicting observations regarding BCAA biosynthesis , and reveal the environmental signals that not only induce BCAA biosynthesis , but that could also have broader consequences on S . aureus environmental adaptation and virulence via CodY . Staphylococcus aureus is a serious human pathogen capable of causing infections that range from mild skin and soft tissue infections , to severe infections of the bone , muscle , heart and lung [1–4] . To survive and thrive in such diverse host environments , S . aureus must maintain sufficient levels of metabolites and co-factors to support virulence determinant production and replication [5 , 6] . The branched-chain amino acids ( BCAAs; Ile , Leu , Val ) represent an important group of nutrients for S . aureus metabolism and virulence , as they are required for synthesis of proteins and membrane branched-chain fatty acids ( BCFAs ) , which are important for S . aureus membrane homeostasis and environmental adaptation . In addition to their nutritional importance , the BCAAs are key regulatory molecules in low GC-content Gram-positive bacteria , as they are activators of the global transcriptional regulator CodY . CodY coordinates expression of nutrient scavenging and synthesis systems , as well as virulence genes , upon depletion of both BCAAs and GTP [7–13] . The requirement of BCAAs for both S . aureus replication and niche adaptation necessitates that it either synthesize these nutrients or acquire them from the environment . Indeed , both BCAA biosynthesis [14–18] and transport [19–22] have been linked to promoting the virulence of other important pathogens in host environments . Bacteria acquire BCAAs via dedicated active transporters , including BrnQ ( Gram-negative and–positive bacteria ) , BcaP ( Gram-positive bacteria ) , and the high affinity ATP-Binding Cassette ( ABC ) transporter LIV-I ( Gram-negative bacteria ) [23–36] . S . aureus encodes three BrnQ homologs ( BrnQ1 , BrnQ2 , BrnQ3 ) , and BcaP . BrnQ1 and BcaP transport all three BCAAs , with BrnQ1 playing a predominant role , and BrnQ2 is an Ile-dedicated transporter [24 , 25] . No appreciable BCAA transport function is associated with BrnQ3 [24] . Despite encoding the BCAA biosynthetic operon , S . aureus relies on the acquisition of BCAAs , most importantly Leu and Val , for rapid growth in media with excess or limiting concentrations of BCAAs , indicating that BCAA biosynthesis is typically repressed [24 , 25] . Paradoxically , biosynthesis remains repressed even in the absence of an exogenous source of Leu or Val , with growth of S . aureus observed only after a prolonged period , likely explaining why previous studies have been misled to conclude that S . aureus is auxotrophic for Leu and Val [37 , 38] . The molecular explanation for this phenotype in S . aureus has remained elusive . Both Gram-positive and Gram-negative bacteria repress BCAA biosynthesis when intracellular levels are sufficient to support growth . In the Gram-negative bacteria Escherichia coli and Salmonella enterica sv . Typhimurium , this is regulated by transcriptional attenuation , which couples translation of a BCAA-rich peptide upstream of the biosynthetic genes with transcriptional termination , such that high levels of BCAAs prevent transcription of the biosynthetic genes [39–44] . In Gram-positive bacteria , including Bacillus subtilis , Listeria monocytogenes , and S . aureus , CodY represses transcription of the biosynthetic genes by binding to a CodY box and inhibiting binding of RNA polymerase [8 , 9 , 45–49] . Additional levels of regulation of the ilv-leu operon in the Gram-positive bacterium B . subtilis include activation by CcpA in response to glucose and repression by TnrA in response to nitrogen levels [50] . Additional fine-turning of the operon in this species is mediated by a Leu-responsive T-box riboswitch [50–53] , as well as mRNA processing [54] . The BCAA biosynthetic genes in S . aureus , encoded by the ilvDBNCleuABCDilvA operon ( ilv-leu ) , and ilvE are similarly repressed by CodY; this regulator binds to two regions upstream of ilvD proximal to the transcriptional start site and two regions within the operon , proximal to ilvC and leuC ( Fig 1 ) [7–9] . Repression is also mediated by the essential genes gcp and yeaZ through an unknown mechanism [55 , 56] . Given that CodY transcriptional repression should be alleviated in the absence of BCAAs , it is unclear why in the case of Leu and Val specifically , growth remains inhibited when either of these two amino acids is absent from the growth medium . We therefore investigated the mechanisms governing these phenotypes in S . aureus to resolve this paradox and to identify the signals required to induce synthesis . Here , we unravel the complex regulation of BCAA biosynthesis in S . aureus , by demonstrating that control is mediated by both trans and cis acting mechanisms of repression . We identify the metabolic cues regulating each mechanism , therefore revealing how S . aureus controls its preference for exogenous BCAAs , and the conditions under which endogenous synthesis is induced . In doing so , we uncover an unappreciated role for Ile in CodY-dependent regulation , demonstrating that it is this BCAA that plays a dominant role in controlling the expression of genes involved in BCAA synthesis and transport . Moreover , since we show that Ile can dominate CodY-dependent gene expression , we highlight an important role for Ile limitation in virulence gene expression , where the absence of this BCAA can induce expression of nuclease , a known CodY-dependent virulence factor . S . aureus has previously been reported as auxotrophic for Leu and Val [37 , 38] , despite possessing a complete BCAA biosynthetic operon . In contrast to these reports , we have observed that S . aureus is indeed able to grow in the absence of Leu and Val following an extended growth period [24] , which might in part explain the discrepancy in these observations . Curiously , when investigating the kinetics of S . aureus growth in response to deprivation of each individual BCAA , we found differing growth phenotypes , even though all enzymes required for the synthesis of each of the BCAAs are encoded from the same biosynthetic loci . For example , when grown in chemically-defined media ( CDM ) lacking Leu , S . aureus exhibited a growth lag of 6–8 h , and when grown in CDM lacking Val S . aureus exhibited a growth lag of ~ 20 h , relative to its growth in complete CDM ( Fig 2A ) . In contrast , growth of S . aureus in CDM lacking Ile was comparable to growth in complete CDM ( Fig 2A ) . These observations were particularly perplexing given the known mechanism , via CodY , regulating BCAA biosynthesis . CodY represses the BCAA biosynthetic operon , such that inactivation of codY results in growth of S . aureus in media lacking either Ile , Leu , or Val ( Fig 2B ) . Given that all three BCAAs have been reported to individually activate CodY DNA binding activity in vitro [13 , 57 , 58] , it was surprising to observe the differences in growth upon omission of the individual BCAAs from the growth medium , and how different it was from that of WT S . aureus and a codY mutant ( compare panels A and B in Fig 2 ) . We therefore hypothesized that the individual BCAAs differentially regulate CodY activity during growth , and that at least one additional mechanism regulates BCAA biosynthesis . To uncover the molecular mechanisms governing these phenotypes , we first questioned whether the absence of Leu or Val selects for mutations that enable growth in the absence of these BCAAs . To address this question , we recovered cells that had grown up following the growth lag in media lacking Leu ( CDM-Leu ) or Val ( CDM-Val ) , and then sub-cultured these isolates back into the same medium from which they were recovered . Cells recovered from CDM-Leu medium exhibited the same growth delay upon sub-culture into the same medium , indicating that this condition does not select for mutations . Conversely , cells recovered from CDM-Val medium grew readily when re-inoculated into CDM-Val , suggesting that they were synthesizing Val ( -Val suppressors , referred to as ValSup ) ( Fig 2C ) . These results suggest that growth in the absence of exogenous Leu requires a regulatory adaptation , whereas growth in the absence of Val selects for a heritable mutation . We hypothesized that identification of the genetic mutations permitting growth of S . aureus in the absence of exogenous Val would reveal important regulators of the BCAA biosynthetic operon and , in-turn , would help reveal the mechanisms behind the BCAA-specific growth phenotypes . Since a codY mutant synthesizes BCAAs and is , thus , capable of growth in the absence of BCAAs , we reasoned that the absence of exogenous Val may select for mutations in codY . We again grew cells in the absence of Val , isolated mutants from twelve independent cultures ( ValSup mutants ) and amplified the codY gene by PCR . Five out of the twelve mutants contained mutations in codY; one had a point mutation resulting in a premature stop codon , two had a 60-bp deletion , and two had independent point mutations resulting in nonsynonymous mutations ( Table 1 ) and ( Fig 3A ) . We then mapped the mutations to identify their position within the CodY protein structure ( PDB ID:5EY0 ) [59] . All mutations occurred in the linker region between the metabolite sensing domain and the DNA-binding domain ( Fig 3B ) . We used secreted protein profiles as a read-out of CodY function , since CodY represses many secreted proteins [60] and therefore the secreted protein profile of a codY mutant differs substantially from WT . The secreted protein profiles of the ValSup mutants with confirmed mutations in codY resembled the protein profile of the codY mutant , except for ValSup-10 mutant ( Fig 3C ) , indicating that all but one of our codY mutations result in an inactive CodY protein , at least insofar as its ability to repress synthesis of secreted proteins . The growth phenotype for each of the unique ValSup strains with confirmed mutations in codY could be reverted to WT-like growth in CDM-Val through complementation with an intact copy of the codY gene in trans ( Fig 3D ) . These results confirm that codY inhibits Val synthesis , and that all of the ValSup codY mutants , along with the codY insertion mutant ( codY::ϕNΣ ) , alleviate this inhibition . The remaining seven ValSup mutants that did not have mutations in the codY gene may have acquired other mutations that indirectly affect the ability of CodY to regulate the ilv-leu operon ( e . g . mutations in GTP synthesis ) . To address this possibility , we assessed the secreted protein profiles of these mutants and all seven were found to exhibit profiles comparable to the WT strain ( S1 Fig ) . These results suggest the mutations occurring within these ValSup strains affect a CodY-independent mechanism of BCAA synthesis regulation . We therefore performed whole genome sequencing to identify the nature of these mutations . This revealed that all seven of these ValSup strains had mutations in the 5’ untranslated region ( UTR ) upstream of ilvD , with a total of three unique point mutations and one 27-bp deletion ( Table 2 ) . The mutations did not overlap with the known promoter features upstream of the ilvD gene ( i . e . the CodY binding motifs ) ( Fig 4A ) . To confirm that mutations in the 5’UTR of ilvD result in an increase in expression of the ilv-leu operon , which would yield the phenotype of growth in CDM-Val without delay ( Fig 2 , panel C ) , we generated a luminescence reporter of the ilvD promoter by cloning the 5’UTR of ilvD into the pGY::lux vector ( Fig 4A ) . Within this reporter construct , we then mutated the 5’UTR to contain the three point mutations identified from genome sequencing of the mutants ( ValSup-1 , ValSup-7 , and ValSup-9 ) . Two of the mutant sequences ( ValSup-1 and ValSup-7 ) resulted in a statistically significant increase in ilvD promoter activity , and the third mutant sequence ( ValSup-9 ) resulted in a trend towards increased promoter activity , although not significant ( Fig 4B ) . Furthermore , using qPCR , we observed that levels of ilvD and ilvC transcripts were elevated when we examined two of the mutants compared to the WT strain ( Fig 4C and 4D ) . Together , these data suggest that the mutations in the 5’UTR of ilvD relieve repression of the ilv-leu operon . We next investigated whether the 5’UTR of ilvD contained a cis-regulatory element , initially considering a T-box riboswitch , since the ilv-leu operon in B . subtilis is regulated by a tRNALeu-responsive T-box riboswitch [51 , 53] . Predictive structure analysis and sequence comparison of the ilvD 5’UTR to known T-box riboswitch sequences revealed that although the S . aureus ilvD 5’UTR contains some features that loosely resemble T-box riboswitches ( S2 Fig ) , it lacks the conserved Stem 1 motifs and structures essential to tRNA anchoring and decoding [61] . We next considered translation-dependent transcriptional regulation ( i . e . attenuation ) , since BCAA-rich leader peptides have been found to regulate BCAA biosynthetic genes in E . coli [42] and S . typhimurium [43 , 44] , and are predicted to regulate BCAA synthesis in Lactococcus lactis sp . lactis [62] , Corynebacterium glutamicum [63 , 64] , and Streptococcus spp . [65] . A search for open reading frames ( ORFs ) in the ilvD leader sequence revealed a short coding region that would be predicted to encode a 26-aa peptide . The predicted peptide contains a string of three Ile codons followed by two Leu codons , and an additional three interspersed Leu codons ( Fig 5A ) . A putative ribosome binding site was also identified 9 nucleotides ( nts ) upstream of the start codon and a putative terminator hairpin structure is located 52 nts downstream from the peptide stop codon , consistent with transcription termination ( S3 Fig ) . We found that the Ile and Leu codons in the peptide were highly conserved across the staphylococci ( Fig 5B ) , as was the predicted terminator stem loop structure ( Fig 5C and S3 Fig ) , suggesting that these features are biologically relevant . Secondary structure predictions revealed an alternative mRNA structure could also form that sequesters the terminator poly-U within an antiterminator ( Fig 5C ) . In the antiterminator fold , the terminator stem-loop is intact , but two upstream stem-loops refold into a new , long stem-loop shifted further upstream . This rearrangement frees a 5’-GAAUGG-3’ motif to pair with 5’-UUGUUU-3’ in the terminator poly-U tail ( Fig 5C ) . Ribosome pausing within these regions could foreseeably disrupt folding to favor antiterminator formation and drive transcription of the ilv-leu operon . When we considered how the mutations that were selected for in media lacking Val might disrupt these features , we found that the 27-bp deletion in ValSup-5/6/11 deletes the predicted terminator stem-loop and the T to A mutation in ValSup-1 changes a Leu codon in the leader peptide to a stop codon ( Fig 5A ) . These mutations would therefore be predicted to relieve repression of transcription , supporting that these are biologically relevant features . The T to C mutation in ValSup-9/12 and the C to A mutation in ValSup-7 occur in predicted secondary structural elements that stabilize terminator formation , and could also relieve repression . We therefore hypothesize that expression of the BCAA biosynthesis operon is regulated by Leu-dependent attenuation in S . aureus . We predict that , in conditions of high Leu availability , translation of the attenuator peptide promotes formation of the terminator hairpin and subsequently transcriptional termination . In conditions of low Leu availability , the ribosome stalls during translation of the attenuator peptide , promoting formation of the antiterminator hairpin and leading to transcriptional read-through . This region , upstream of ilvD , is henceforth referred to as the attenuator sequence and not the 5’UTR . The in vitro selection experiment revealed two mechanisms involved in repression of the ilv-leu operon , and yet the growth data ( see Fig 2A ) suggest that the operon is fully repressed only under conditions of Val deprivation and not Ile or Leu deprivation . We therefore continued to investigate how these mechanisms respond to deprivation of the individual BCAAs to explain these unique growth phenotypes . To test our hypothesis that Leu availability regulates expression of the ilv-leu operon via attenuation , we used our luminescence reporter construct containing the attenuator sequence cloned into the pGY::lux vector to examine how it responds to Leu deprivation . We were also interested to investigate how BCAA availability regulates CodY regulation of the ilv-leu operon , since the growth phenotypes of S . aureus in the absence of Leu or Val suggests that depletion of these nutrients alone is not sufficient to relieve CodY-dependent repression ( compare Fig 2A and 2B ) . To study CodY-dependent promoter activity in isolation of attenuation , we generated a second reporter construct ( partial promoter; pGYilvDP::lux ) that lacked the attenuator sequence and contained only the CodY binding sequence and compared this to the original construct ( complete promoter; pGYilvDC::lux ) that contained both regulatory elements ( Fig 6A ) . We first confirmed that both constructs responded to CodY and , indeed , observed higher promoter activity in the codY mutant compared to the WT strain that peaked during mid-exponential growth ( Fig 6B and 6C ) . All endpoint pGY::lux experiments from this point on are therefore the luminescence normalized to the optical density of mid-exponential phase cells ( RLU/OD600 ) . We note that , generally , the partial promoter fusion has higher activity than the complete promoter fusion , likely due to omitting the attenuator sequence . We next assessed promoter activity in response to depletion of each BCAA . Since complete omission of Leu and Val from the growth medium significantly attenuates S . aureus growth , we instead limited their concentrations to 10% of that in complete CDM to minimize differences in growth . We first examined CodY-dependent promoter activity using the pGYilvDP::lux construct . Promoter activity increased to levels comparable to the codY mutant only upon Ile limitation , and limitation of Leu or Val in combination with Ile did not alter ilvD promoter activity any further ( Fig 7A ) , indicating a predominant role of Ile in regulating CodY activity on the ilvD promoter . We next examined the effect of BCAA limitation on attenuator-dependent regulation . ilvD promoter activity of the pGYilvDC::lux construct also increased upon Ile limitation , however , we also observed a further increase in promoter activity when Leu and Ile were limited simultaneously ( Fig 7B ) . These data suggest that the attenuator sequence , which is unique to the pGYilvDC::lux construct , responds to Leu availability . This is consistent with our hypothesis that the attenuator represses the BCAA operon in response to Leu . We previously identified mechanisms of BCAA transport in S . aureus , including BrnQ1 and BcaP , which transport Ile , Leu and Val , and BrnQ2 , a dedicated Ile transporter [24 , 25] . To determine the contribution of each of these transporters to either CodY-dependent or attenuator-dependent regulation of BCAA biosynthesis , we assessed ilvD promoter activity in various BCAA transporter mutants . ilvD promoter activity of the pGYilvDP::lux construct increased only in the brnQ2 mutant ( Fig 7C ) , whereas the pGYilvDC::lux increased in the brnQ1 and brnQ1bcaP mutants ( Fig 7D ) . These data indicate that BrnQ2-dependent Ile transport is linked to CodY activity and BrnQ1/BcaP-dependent Leu transport is linked to attenuation . Notably , using the complete ilvD promoter region , we did not observe a change in promoter activity in the brnQ2 mutant . We have previously shown that brnQ1 is upregulated in a brnQ2 mutant and consequently a brnQ2 mutant takes up more Leu and Val permitting enhanced growth in media limited for these BCAAs [24] . We thus postulate that in a brnQ2 mutant , the increased Leu uptake causes repression of ilvD promoter activity via the attenuator and overrides the CodY-dependent Ile response . We next revisited the DNA binding activity of CodY at the ilvD promoter in the presence of each BCAA to compare CodY activity in vitro vs during growth . To test whether Ile activates CodY to bind DNA more efficiently than Leu or Val , we analyzed the interaction of CodY with a fluorescently-labeled DNA fragment ( ilvD266p+ ) containing the annotated CodY regulatory region of ilvD [9] . Ile-activated CodY formed DNA:CodY complexes with as little as 6 . 5 nM CodY monomer , whereas Leu- and Val-activated CodY formed similar , multiple DNA:protein complexes as Ile-activated CodY , but required ~4-fold more CodY protein ( S4A–S4C Fig ) . However , band densitometry analysis and fitting the data to a Hill equation revealed that the apparent binding constant values were essentially identical for all ligands tested ( Fig 8 ) . We did not observe an additive effect of all three BCAAs on CodY binding activity ( S4D Fig ) . Thus , CodY binds all three amino acids in vitro . Thus far , our data provide insight into the molecular mechanisms governing the BCAA-specific growth phenotypes of S . aureus observed in panel A of Fig 2 . In complete CDM , the ilv-leu operon is repressed in an Ile-dependent manner via CodY and in a Leu-dependent manner via the attenuator peptide . Omission of Ile from the growth medium relieves CodY repression of the ilv-leu operon , resulting in Ile synthesis , which supports rapid growth in the absence of an exogenous Ile source . In media lacking Val , CodY remains active and the presence of Leu triggers transcriptional termination of the ilv-leu operon via the attenuator peptide; thus S . aureus is unable to synthesize Val and consequently unable to grow unless either of the aforementioned mechanisms is mutated . Omission of Leu from the growth medium relieves attenuator-dependent repression , however CodY remains active , and consequently , Leu synthesis is only partially relieved , resulting in a reduced growth rate . It therefore follows that simultaneous omission of Ile and Leu or Val should permit growth of S . aureus due to de-repression of CodY . Indeed , we found that the growth of S . aureus in CDM lacking Ile and Leu initiated more rapidly than growth in CDM lacking Leu alone , indicating that the reduced growth rate in CDM–Leu is due to Ile-dependent CodY repression ( Fig 9A ) . Unexpectedly , S . aureus grown in CDM lacking Ile and Val resembled growth of S . aureus in media lacking Val alone ( Fig 9B ) . Since the presence of Leu also contributes to repression of the operon , we further examined growth of S . aureus in CDM lacking all three BCAAs , however the growth of S . aureus remained attenuated , with no observable growth until a prolonged period of ~ 16 hr ( Fig 9C ) . These data are curious given that the immediate precursor of Val is also a precursor of Leu , and the aminotransferase ( IlvE ) that converts ketoisovalerate to Val also produces Ile and Leu ( Fig 1 ) . Since our promoter:reporter data demonstrate that the ilvD promoter is active in media limited for all three BCAAs ( Fig 7A and 7B ) and thus the operon is presumed derepressed , we postulate that the growth impairment in media lacking all three BCAAs is related to enzymatic activity of the biosynthetic enzymes , whereby either the aminotransferase exhibits substrate bias towards Ile or Leu synthesis , or there is negative or positive cross-regulation between the pathways . For example , the threonine deaminase ( IlvA ) required only for Ile synthesis ( Fig 1 ) is activated by Val in E . coli and B . subtilis [66 , 67] . Therefore , it is possible that in the absence of Val , Ile synthesis is reduced , contributing to the growth impairment in media lacking both Ile and Val . We were curious to investigate whether mutations in the promoter region of ilvD arise in the environment , reasoning that S . aureus might encounter Val-limited environments that impair growth and therefore select for mutations in the regulatory mechanisms involved in repression . We compared the nucleotide sequence of the ilvD promoter region from USA300 FPR3757 to all complete genome sequences of S . aureus . Overall , there was high sequence conservation , however , several variants were identified in the putative regulatory regions ( S5 Fig ) . Intriguingly , two variants occur in the putative ORF upstream of ilvD and both alter the number of Leu codons in the peptide ( located at +151 and +162 in S5 Fig ) . Several sequence variants also occur in the first CodY binding region . Ongoing studies will investigate the consequence of these mutations on the level ilv-leu expression and subsequent BCAA biosynthesis in these strains . Our data revealed an unexpected role for Ile , and not Leu or Val , in regulating CodY activity on the ilvD promoter . The predominant role for Ile could have important implications for S . aureus physiology and virulence given that CodY is considered a master regulator of metabolism and virulence gene expression in S . aureus [7–9 , 49 , 60] . It was therefore of high interest to us to investigate whether the predominant role of Ile in regulating CodY activity was unique to ilvD or if it extended to other CodY-regulated genes . We selected the CodY-regulated brnQ1 gene as a representative metabolic gene [7–9] , and we modified the luminescent reporter experiment slightly , such that instead of limiting BCAAs , which can alter growth , we added back excess BCAAs and examined whether the addition of excess BCAAs has repressive effects on CodY target gene expression . We first confirmed the effect of excess BCAAs on the ilvD promoter . Indeed , we observed that excess amounts of Ile in the growth medium repressed ilvD promoter activity ( Fig 10A ) , whereas excess Leu had no effect ( Fig 10B ) and , intriguingly , excess Val had the opposite effect of increasing promoter activity ( Fig 10C ) . We repeated this experiment with a lux promoter:reporter containing the brnQ1 promoter . Consistent with the ilvD promoter:reporter , we observed excess Ile , but not Leu or Val , to have a repressive effect on promoter activity ( Fig 10D–10F ) . We next investigated whether Ile limitation results in relief of CodY-mediated repression of virulence gene expression , specifically , the secreted factor nuclease [7 , 9 , 68] . To do this , we took advantage of a previously constructed nuc-gfp reporter [7] and measured fluorescence during mid-exponential phase . In agreement with past results , we measured relatively low nuc-gfp fluorescence when WT cells were cultured in complete CDM; the fusion was derepressed ~17-fold in codY null mutant cells in the same medium ( Fig 11 ) . When WT cells were grown in CDM lacking Ile , nuc-gfp fluorescence increased ~5-fold over that observed in WT cells grown in CDM with excess Ile . Compared with complete CDM , we measured essentially the same amount of nuc-gfp fusion fluorescence in codY null mutant cells when Ile was omitted from the medium . Thus , Ile limitation results in a partial derepression of nuc-gfp in a CodY-dependent manner . Together , these data suggest that the role of Ile in regulating CodY activity is not unique to the ilvD promoter . In this study , we sought to determine the mechanisms by which each BCAA regulates expression of the ilv-leu operon to explain the unique growth phenotypes of S . aureus upon depletion of each of Ile , Leu and Val . By selecting for genetic variants of S . aureus that grew rapidly in the absence of an exogenous source of Val , we characterized two classes of mutations that relieve repression of the ilv-leu operon; mutations in the transcriptional repressor CodY and mutations in the region upstream of ilvD , the first gene in the ILV biosynthetic operon . We demonstrate that CodY activity is predominantly regulated by Ile availability during growth , an unexpected finding given that all three BCAAs activate CodY:DNA binding in vitro ( S4 Fig ) . Bioinformatic analysis revealed that the region upstream of the ilvD coding sequence contains a highly-conserved attenuator peptide that is rich in Leu codons and , therefore , presumably controls transcriptional read-through in response to Leu availability . This is supported by experimental evidence demonstrating that ilvD promoter activity increases in response to i ) Leu depletion , and ii ) mutations in the attenuator peptide . Therefore , Ile and Leu each regulate expression of the ilv-leu operon through unique mechanisms ( summarized in Fig 12 ) . The primary reservoir of S . aureus is the anterior nares . That Ile was not detected in human nasal secretions [69] lends support to the idea that Ile deprivation is perceived by S . aureus in vivo and is a signal , via CodY , to upregulate ILV synthesis that would presumably aid in bacterial survival in at least this niche . A predominant role for Ile in regulating CodY activity during growth has also been observed in another S . aureus strain [8] , as well as other species , including B . subtilis [12] , L . lactis [57] , and L . monocytogenes [70] . Since CodY has been linked with virulence factor expression in S . aureus [7 , 8 , 71–75] , including nuc as demonstrated in this study , it will be important to determine whether Ile is the predominant BCAA to modulate CodY activity on additional target genes , including virulence genes . It is also noteworthy that we demonstrated an important link between the BrnQ2 transporter , but not BrnQ1 or BcaP , and Ile availability to CodY activity . BrnQ transporters exist in other organisms , yet none function , as BrnQ2 does , as a dedicated Ile-transporter [32 , 34 , 35] . This suggests that BrnQ2 could provide an advantage to the adaptation of S . aureus to Ile-depleted environments . The fact that mutations in CodY are selected for when S . aureus is grown in the absence of exogenous Val supports the notion that Val contributes minimally to regulating CodY activity , at least on the ilvD and brnQ1 promoters , during growth . In addition to trans regulation , via CodY , of ilv-leu operon expression in S . aureus , we identified a cis-dependent mode of regulation of operon expression , via an attenuator . Attenuation regulates BCAA biosynthesis in E . coli and S . enterica [42–44] , and predicted attenuators can be found in various Gram-negative and Gram-positive bacteria [62–64 , 76] . In support of attenuation regulation of ilv-leu in S . aureus , one of the mutations we identified in our screen occurs in the predicted leader peptide and changes a Leu codon to a stop codon , which we predict would reduce Leu-dependent repression . The leader peptide also contains three Ile , suggesting that the level of uncharged tRNAIle also regulates ilv-leu expression . In support of this , expression of the ilv-leu operon is increased upon exposure to mupirocin , an antibiotic that binds to isoleucyl-tRNA synthetase and blocks the charging of tRNAIle [77] . CodY appears to be the dominant mechanism of repression , since S . aureus exhibits a growth delay in media lacking Leu , but not Ile , suggesting that Leu deprivation alone is not sufficient to fully relieve repression . Two additional CodY binding regions have been identified in the ilv-leu operon [9] , and therefore transcription of downstream genes in the operon would occur in a codY mutant , bypassing transcriptional termination at the ilvD leader . Alternatively , CodY repression could block further transcription upon relief of attenuation , resulting in shorter transcripts . Notably , these conditions did not select for mutations in other previously described regulators of the ilv-leu operon , such as Gcp and YeaZ [55 , 56] . Since Gcp and YeaZ are essential genes in S . aureus , we did not expect to isolate mutations in these genes , however , we cannot rule out the possibility that the mutations in the attenuator region upstream of the ilvD coding region reduce binding of YeaZ [56] . Our model predicts that expression of the ilv-leu operon would be de-repressed upon Ile and Leu deprivation , yet S . aureus exhibits a significant growth lag in the absence of all three BCAAs . One possible explanation for this observation is potential allosteric regulation of the BCAA biosynthetic enzymes . The last gene in the ilv-leu operon , ilvA , encodes a threonine deaminase ( TD ) , which catalyzes the first step in Ile synthesis by converting threonine to α-ketobuytrate . The E . coli TD enzyme is inhibited by Ile and activated by Val [78–80 , 66] . The B . subtilis TD enzyme is inhibited by Ile and it is proposed that Val activates TD in the presence of Ile and inhibits TD at high concentrations [67] . If TD activity in S . aureus is most efficient in the presence of Val , it follows that Ile synthesis would be reduced in the absence of Val . This could explain the absence of growth in media lacking Ile and Val . It would be of interest to investigate whether TD in S . aureus is similarly subject to allosteric regulation . Given the multiple physiological roles of BCAAs , another possibility is that simultaneous removal of all three BCAAs imparts enhanced stress on S . aureus compared to when the amino acids are omitted alone . In support of this , we demonstrated that in the absence of Leu and Val acquisition , the S . aureus membrane lacks Leu- and Val-derived iso-fatty acids , but this loss is compensated for by higher incorporation of Ile-derived iso-fatty acids [25] . Perhaps in the absence of all three BCAAs , such compensatory mechanisms are not achievable . Whatever the mechanism , it is evident that environments where Val is limited or absent poses a challenge to S . aureus and suggests that Val transport is critical for its growth . Indeed , we have previously shown that BCAA transporters BrnQ1 and BcaP , the only S . aureus transporters for Val , are required for S . aureus growth in vivo [24 , 25] . Altogether , this study details the molecular mechanisms regulating BCAA biosynthesis in S . aureus and uncovers environments where S . aureus engages in BCAA biosynthesis . In doing so , we reveal a predominant role for Ile in regulating CodY activity on the ilvD and brnQ1 promoter . Given the role of CodY in additionally regulating virulence genes , our data support the hypothesis that environmental availability of Ile is an important regulatory cue for S . aureus adaptation to nutrient limitation and virulence gene expression . All strains and plasmids used in this study are described in Table 3 . Methicillin-resistant S . aureus ( MRSA ) pulsed-field gel electrophoresis type USA300 LAC that has been cured of pUSA03 , a plasmid conferring macrolide and lincosamide resistance , was used in all experiments as the wild-type ( WT ) strain . S . aureus strains were grown in either tryptic soy broth ( TSB ) ( EMD Millipore , Billerica , MA ) or in a chemically defined medium ( CDM ) , described previously [24] . Final concentrations of Ile , Leu and Val in complete CDM were 228 μM , 684 μM , and 684 μM , respectively . Final concentrations were adjusted to 10% of their concentration in complete CDM in some experiments , as indicated . For growth experiments in TSB , S . aureus strains were pre-cultured in TSB until mid-exponential phase was reached , and then sub-cultured into fresh TSB to a starting optical density ( OD600 ) of 0 . 01 . For growth experiments in CDM , S . aureus strains were pre-cultured in CDM until mid-exponential phase was reached , and then sub-cultured into fresh CDM to a starting OD600 of 0 . 05 in either complete CDM or CDM where BCAA concentrations were limited or omitted , as indicated . Growth curves were performed in 100-well plates containing 200 μL/well of media and were read using the Bioscreen C visible spectrophotometer ( Growth Curves USA; Piscataway , NJ ) . End point growth assays were performed in tubes with a 7:1 v/v tube:media ratio . All growth experiments were performed at 37°C with shaking . Growth media were supplemented with chloramphenicol ( 10 μg mL-1 ) , ampicillin ( 100 μg mL-1 ) , or erythromycin ( 3 μg mL-1 ) , where required . Deletion of bcaP was constructed using the pKOR1 plasmid as described previously [24] . Primer sequences were based on the published USA300 FPR3757 genome and are displayed in Table 4 . The bcaP deletion was introduced into the markerless brnQ1 deletion mutant , described previously [24] . The pGYlux vectors were constructed using primers described in Table 4 . The pGYlux plasmid is derived from a low copy plasmid ( 5 copies/cell ) [85] , and we estimated 20 copies/cell in our experiments . lux plasmids were further used as templates for site-directed mutagenesis , using primers described in Table 4 . Briefly , PCR reactions containing the Phusion High-Fidelity DNA Polymerase ( ThermoFisher , Waltham , MA ) were set up such that half of the reaction mixture contained the forward primer and the remaining half contained the reverse primer . These reactions proceeded for 3 cycles of 98°C for 10 s , 60°C for 30 s , and 72°C for 12 min . After 3 cycles , the forward and reverse primer reactions were mixed together and the reactions proceeded for an additional 17 cycles . Plasmids were treated with DpnI ( New England Biolabs , Ipswich MA ) for 1 hr at 37°C and were then transformed into E . coli DH5α . Mutations were confirmed by PCR . All plasmids were first constructed in E . coli DH5α and subsequently electroporated into the restriction-defective S . aureus strain , RN4220 , prior to electroporation into the desired strain . To select for genetic mutations that permit adaptation to growth in media lacking Val , twelve independent colonies of WT S . aureus were grown in complete CDM to mid-exponential phase and sub-cultured into CDM lacking Val . Recovered cells were harvested and plated onto TSB agar and grown overnight at 37°C . Isolated colonies were grown in complete CDM to mid-exponential phase and sub-cultured into CDM lacking Val to confirm the occurrence of a heritable mutation . Genomic DNA was isolated from all twelve mutants , referred to as ValSup-mutants , as well as from two biological replicates of our laboratory WT USA300 , using the Invitrogen PureLink Genomic DNA Preparation Kit ( ThermoFisher Scientific , Boston MA ) per the manufacturer’s instructions . Primers used for the targeting sequencing of the ilvD promoter and the codY gene are listed in Table 4 . Samples were sent to the London Regional Genomics Center for sequencing on the MiSeq platform . Libraries were prepared using the Nextera XT DNA Library Preparation kit ( Illumina , San Diego , CA ) . 150 bp reads were mapped to the USA300 FPR3757 ( NC_007793 . 1 ) genome using the BWA-MEM aligner [86] and variants were determined using SAMtools [87] . Strains were pre-grown in TSB to mid-exponential phase and then sub-cultured into TSB to a starting OD600 of 0 . 01 and grown overnight . The OD600 of stationary phase cultures were determined and a supernatant volume equivalent to 5 OD units was harvested and incubated with trichloroacetic acid ( TCA ) ( Sigma-Aldrich , St . Louis , MO ) at a final concentration of 20% overnight at 4°C . Precipitated protein samples were dissolved , run on a 12% acrylamide gel and stained with Coomassie-Blue . Kinetic lux reporter experiments were performed in flat , clear-bottom 96-well white plates ( Thermo Fisher Scientific ) and read using a BioTek Synergy H4 Hybrid Multi-Mode Microplate Reader ( BioTek Instruments Inc , Winooski , VT ) . Pre-cultures were inoculated into either complete or limited CDM to a starting OD600 of 0 . 01 in 200 μL/well . Luminescence and OD600 values were read at hourly intervals . For end-point lux reporter experiments , pre-cultures were sub-cultured into either complete or limited CDM to a starting OD600 of 0 . 05 in tubes with a 7:1 tube:media ratio . At hourly intervals , aliquots of 200 μL were transferred to flat , clear-bottom 96-well white plates ( Thermo Fisher Scientific ) and luminescence and OD600 values were read . Samples of strains containing the lux construct with the complete ilvD promoter were supplemented with 0 . 1% ( v/v ) decanal in 40% ethanol and luminescence was measured immediately . Data presented are the relative light unit ( RLU ) values normalized to the OD600 of the sample when the cultures reached mid-exponential phase ( OD600 0 . 6–0 . 8 ) . Data were analyzed by one-way ANOVA with Dunnet’s multiple comparison test relative to the control sample in GraphPad Prism Version 7 . 0b . RNA was isolated from cells grown to mid-exponential phase ( OD600 of 0 . 6–0 . 8 ) in complete CDM using the Aurum Total RNA Mini Kit ( Bio-Rad; Hercules , CA ) per the manufacturer’s instructions . RNA ( 500 ng ) was reverse transcribed using SuperScript II ( Invitrogen , Carlsbad , CA ) per the manufacturer’s instructions using 500 μg mL-1 of random hexamers . cDNA was PCR-amplified using SensiFAST SYBR No-ROX Kit ( Bioline , Taunton , MA ) . Data were normalized to expression of the rpoB reference gene , and analyzed by an unpaired student’s t-test in GraphPad Prism Version 7 . 0b . Primers used are listed in Table 4 . The codY ORF ( QV15_05910 ) was amplified from S . aureus strain UAMS-1 using oligonucleotides oKM1 and oSRB410 . The PCR fragment was purified and subjected to a second round of PCR using oKM1 and oSRB411 to append a Tobacco Etch Virus ( TEV ) protease cleavage sequence followed by six histidine ( CAT ) codons and a TAA stop codon . The 830-nt fragment was digested with SacI/SphI and ligated to the same sites of pBAD30 [88] . The resulting plasmid was introduced into E . coli DH5α . CodY-His6 was overproduced by growing the strain carrying the plasmid in LB at 37°C until mid-exponential phase ( OD600 ~0 . 3 ) . At this time , L- ( + ) -arabinose was added to a final concentration of 0 . 2% ( w/v ) . After four hours of induction at 37°C , cells were pelleted by centrifugation ( 8 , 610 x g at 4°C ) and frozen at -80°C . The cells were thawed , resuspended in Buffer A ( 20 mM Tris-Cl [pH 7 . 9] , 500 mM NaCl , 5 mM imidazole , 5% [v/v] glycerol ) supplemented with 0 . 1% ( v/v ) nonidet P-40 and 1 mM phenylmethylsulfonyl fluoride ( PMSF ) , and lysed by sonication . CodY-His6 protein was purified from clarified soluble extracts using a computer-controlled ÄKTAPrime plus FPLC system equipped with a His-Trap FF column ( GE Healthcare Life Sciences ) using a linear gradient elution with Buffer B ( 20 mM Tris-Cl [pH 7 . 9] , 500 mM NaCl , 685 mM imidazole , 5% [v/v] glycerol ) . Fractions containing CodY-His6 protein were pooled and supplemented with glycerol to 50% ( v/v ) and stored at -20°C . A 266-bp fragment ( ilvD266p+ ) spanning -131 to +134 relative to the annotated ilvD transcriptional start site in S . aureus UAMS-1 [9] was synthesized by PCR using primers oNW025 and oAK031 ( Table 4 ) , simultaneously incorporating a 6-carboxyfluorescein ( FAM ) -label . EMSAs were performed with purified CodY-His6 protein and FAM-labeled ilvD266p+ fragment in binding buffer ( 20 mM Tris-Cl [pH 8 . 0] , 50 mM KCl , 2 mM MgCl2 , 5% [v/v] glycerol , 0 . 05% [v/v] Nonidet P-40 , 1 mM dithiothritol [DTT] , 0 . 025 mg ml-1 salmon sperm DNA ) . Samples ( 20 μl ) containing various amounts of CodY-His6 , 200 fmol of 6-FAM-labeled DNA fragment , 2 mM GTP , and 10 mM of the indicated BCAA ( s ) were incubated for 20 min at 25°C in a thermomixer ( Eppendorf ) with moderate agitation ( 250 rpm ) . The samples were separated on 8% non-denaturing 35 mM HEPES ( pH 7 . 4 ) -43 mM imidazole-10 mM BCAA polyacrylamide gels for 40 minutes at 200 V . Fluorescent DNA fragments were detected and quantified using a computer-controlled ImageQuant LAS 4000 biomolecular imager ( GE Healthcare Life Sciences ) using a SYBR filter set . Quantitative analysis of CodY binding to ilvD266p+ was performed using ImageJ software [89] . Since the binding curves appeared to have a sigmoidal shape , the data from two independent experiments were fitted to the Hill equation Θ = Ch/ ( Ch + K0 . 5h ) using Prism ( ver . 7; GraphPad Software ) . In this equation , Θ is the fraction of bound DNA , C is the concentration of CodY , K0 . 5 is the binding constant , and h is the Hill coefficient . K0 . 5 and h shown are from fitted data where r2 > 0 . 96 . Strains were grown overnight in CDM complete , then sub-cultured the next morning in CDM complete to a starting OD600 of 0 . 05 in 125 ml DeLong shake flasks ( 5:1 flask:medium ratio ) . Incubation was performed in an Innovo orbital shaking water bath ( New Brunswick ) with vigorous agitation ( 280 rpm ) . At an OD600 of 0 . 8 , cells were pelleted and resuspended in either CDM complete or CDM lacking isoleucine to an initial OD600 of ~0 . 05 . When cells reached mid-exponential phase ( OD600 of 0 . 4–0 . 5 ) , a 1-mL sample was removed , washed once with phosphate buffered saline ( PBS ) , and resuspended in PBS to minimize background fluorescence from the medium . Fluorescence was measured using a computer-controlled Tecan Infinite F200 Pro plate reader equipped with 485 nm excitation and 535 nm emission filters . GFP signal acquisition parameters were kept constant throughout the experiment ( gain , 49%; flash number , 10; integration time , 40 μs; lag time , 0 μs; settle time , 0 ms ) . Data are presented as relative fluorescence units ( RFUs ) after subtracting the fluorescence from USA300 LAC ( lacking the GFP reporter plasmid ) and dividing by OD600 to correct for cell density . Putative terminator structures in the ilvD 5’UTR were identified using the predictive software RibEx [90] and RNAfold [91] . Mfold [92] was used to identify putative antiterminator sequences . RNAfold and Mfold were used to search for conserved T-box riboswitch features . T-box riboswitch multiple sequence alignments were generated with predicted T-box riboswitch sequences in S . aureus subsp . aureus N315 ( NC_002745 . 2 ) that were annotated in the Rfam database [93] and the ilvB T-box riboswitch from B . subtilis ( NC_000964 . 3 ) using MUSCLE [94] with default parameters . The alignments were manually adjusted in JalView [95] with insight gained from experimentally characterized S . aureus T-box leaders: glyS , ileS and metI . The putative S . aureus ilvD leader was then added to the finished alignment using MAFFT [96] . The peptide multiple sequence alignments were generated by extracting the top 15 BLAST results for ilvD leaders from different staphylococci , translating the ORFs , and aligning the peptide sequences using MUSCLE [94] . The ilvD 5’UTR from USA300 FPR3757 was aligned to 168 S . aureus complete genomes using BLAST .	To infect its human host , the bacterial pathogen Staphylococcus aureus must either take up nutrients from the surrounding environment or produce them itself . Previous studies have reported that S . aureus does not produce the amino acids leucine and valine , despite it possessing the genes to do so . In this study , we show that S . aureus does indeed produce leucine and valine , but only under certain nutritional conditions . We select for mutants of S . aureus able to grow in valine-depleted environments to uncover genetic variants that enable valine production . We discover genetic variants in a repressor protein and a region of non-coding DNA that both , when properly functioning , inhibit production of leucine and valine under nutrient-rich conditions . We further identify the nutritional conditions where the inhibition of leucine and valine production is relieved , thus revealing a previously overlooked role for another amino acid , isoleucine , in influencing nutrient metabolism . We show that isoleucine levels also influence expression of genes involved in the ability of S . aureus to cause disease . These findings help to reconcile conflicting reports regarding leucine and valine production in S . aureus and reveal nutritional cues that could influence its ability to cause infection .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "gene", "regulation", "pathogens", "split-decomposition", "method", "microbiology", "operons", "staphylococcus", "aureus", "multiple", "alignment", "calculation", "mutation", "sequence", ...	2018	Repression of branched-chain amino acid synthesis in Staphylococcus aureus is mediated by isoleucine via CodY, and by a leucine-rich attenuator peptide
We suggest for the first time that the action of multivalent cations on microtubule dynamics can result from facilitated diffusion of GTP-tubulin to the microtubule ends . Facilitated diffusion can promote microtubule assembly , because , upon encountering a growing nucleus or the microtubule wall , random GTP-tubulin sliding on their surfaces will increase the probability of association to the target sites ( nucleation sites or MT ends ) . This is an original explanation for understanding the apparent discrepancy between the high rate of microtubule elongation and the low rate of tubulin association at the microtubule ends in the viscous cytoplasm . The mechanism of facilitated diffusion requires an attraction force between two tubulins , which can result from the sharing of multivalent counterions . Natural polyamines ( putrescine , spermidine , and spermine ) are present in all living cells and are potent agents to trigger tubulin self-attraction . By using an analytical model , we analyze the implication of facilitated diffusion mediated by polyamines on nucleation and elongation of microtubules . In vitro experiments using pure tubulin indicate that the promotion of microtubule assembly by polyamines is typical of facilitated diffusion . The results presented here show that polyamines can be of particular importance for the regulation of the microtubule network in vivo and provide the basis for further investigations into the effects of facilitated diffusion on cytoskeleton dynamics . Microtubules ( MTs ) are αβ-tubulin hollow polymers that perform major organizational tasks in eukaryotic cells [1] . These structures , which contribute to the rigidity and structural integrity of the cells , are implicated in motor-driven intracellular traffic of organelles and vesicles , in the formation of the mitotic spindle and in cell migration and motility . At least two physical parameters counteract microtubule assembly in cells: ( i ) their highly negatively charged surface due to the negative charge of the αβ-tubulin heterodimer itself ( 20–30 e− ) [2]–[4] , which induces an electrostatic self-repulsion between tubulin molecules [5] , [6] , ( ii ) the viscosity of the cytoplasm which slows down tubulin diffusion ( about 5 . 10−12 m2/s as measured in sea urchin extracts [7] ) . Consequently , due to the requirement of a proper orientation of tubulin to associate to MT ends [8] , [9] , the flow of tubulin arriving directly to microtubule ends by 3D diffusion is critical to sustain the rapid elongation of microtubules observed in cells ( >10 µm/s [10] ) . We propose here that facilitated diffusion may significantly enhance the association of tubulin to growing nuclei or microtubules . The mechanism of facilitated diffusion was already introduced to understand how DNA-binding proteins can find their target sites by sliding and hopping along DNA molecules [11] , [12] . However facilitated diffusion can also favour tubulin association to nucleus and microtubule ends provided that an attraction exists between tubulins and microtubules or nucleus to enable sliding . This attraction force can be triggered by the presence of small multivalent cations that are shared between tubulin dimers . Among small cations , polyamines such as divalent putrescine , trivalent spermidine and tetravalent spermine are reasonable candidates due to their high concentrations in all cells [13] , [14] . Surprisingly , despite the potential influence of polyamines on tubulin dynamics in cells [15] , [16] , only a few studies have addressed the mechanisms by which they act on microtubule assembly . The enhancement of the polymerization rate observed in vitro was generally attributed to the neutralization of the C-terminal tails of tubulin [2] , [17] , [18] . In addition to this effect , facilitated diffusion can also explain how multivalent cations actually act on the different aspects of microtubule assembly , notably nucleation and elongation . In this paper we thus decipher , both theoretically and experimentally , if facilitated diffusion can interfere with microtubule dynamics . Taking advantage of the recent advances in the description of the electrostatic properties of tubulin [3] , [4] , [19] , we show that the attraction force between highly anionic tubulins , mandatory for facilitated diffusion , can indeed occur due to the presence of polyamines . We then develop an analytical model to describe the influence of facilitated diffusion on microtubule assembly . Considering an attraction between tubulin dimers , we propose that facilitated diffusion can act on MT dynamics via two different ways: ( i ) a longer residency time of tubulin onto a growing nucleus which favours association , ( ii ) a facilitated elongation thanks to sliding of free tubulin to the MT ends . A series of in vitro experiments is conducted to test the model through the analysis of light scattering curves , sedimentation assays and high resolution AFM imaging of MTs . Finally , to approach more physiological conditions , we show some preliminary results for MAPs-tubulin . Multivalent cations like polyamines can induce an electrostatic attraction between two highly negatively charged surfaces . The theoretical treatment of this force has been the subject of extensive studies [20]–[23] in particular to explain the phenomenon of DNA condensation by polyamines [24]–[26] . Concerning microtubules , their self-attraction can also be triggered by multivalent salts and can lead to the formation of bundles [27] . The energy benefit of association between two like-charged bodies is generally due to the correlations between the multivalent counterions condensed on their surfaces . Here we will show that the interaction between two tubulin heterodimers , which are highly negatively charged proteins with a net charge about 20–30 e− , at pH around 7 [19] , may also be influenced by this mechanism . Let us first estimate the attraction energy between two tubulin heterodimers in the presence of polyamines . The αβ-tubulin heterodimer is a nonspherical globule , having dimensions of 46×80×65 Å with two long C-terminal tails ( ∼35 Å ) [3] . Most of the tubulin charge , at least 40% , is concentrated in the C-terminal tails so that the charge distribution on its surface is non homogeneous . We then need to investigate the electrostatic properties of the C-terminal tails and of the remaining of the heterodimer molecule separately . As described in details in Text S1 , we obtained after theoretical developments the energy gain for these two interactions . The attraction energy for two heterodimers without the C-terminal tails is positive for putrescine and spermidine and negative only for spermine ( −0 . 26 KBT ) , which means a weak energy benefit . As this result was obtained by considering that there is only spermine on heterodimer surfaces , we might expect that the resulting attraction force is negligible in physiological conditions where spermine competes with other cations or proteins for tubulin neutralization . On the other hand , the attraction energy becomes significantly larger between two interacting tubulin C-terminal tails . We obtained in this case UC values of +4 . 7 , −2 . 6 , −5 and −6 . 2 KBT for cation valence Z = 1 , 2 , 3 and 4 respectively , showing the presence of an attraction energy when Z>1 and which increases with the valence of ions . The energy benefit per counterion in correlations remains lower than KBT for divalent putrescine ( see Text S1 ) so that it could hardly induce an attraction force between C-terminal tails due to thermal agitation . On the other hand , we expect that trivalent spermine and tetravalent spermidine can induce a significant attraction force between the C-terminal tails of tubulin dimers ( see Text S1 and Figure 1A ) . As the correlations between monovalent cations cannot contribute to the attraction force , the replacement of multivalent cations at high monovalent salt concentrations by monovalent ones lowers the attraction force . The fractional surface concentration of the multivalent cations , which is the ratio of the multivalent counterion surface density to the total surface density of the counterions , is very sensitive to the counterion valence [28] , [29] . Higher valence cations are generally better competitors for the surface neutralization than lower valence ones ( see Text S2 and [28] ) . In addition , the monovalent salt concentration above which multivalent cations are removed from the surface increases with the surface charge density [28] . For the highly negatively charged C-terminal tails , assuming σ∼1 e−/nm2 and using the model of Rouzina and Bloomfield [28] ( see Text S2 , supplementary file ) , about 50% of the spermidine counterions are removed with 300 µM spermidine and 150 mM KCl . As facilitated diffusion requires tubulin sliding on microtubules , we investigated if polyamines can mediate an attraction of tubulin dimers on microtubules . To measure this interaction , fluorescent Cy3-labeled tubulin ( 0 . 3 µM ) was allowed to interact with non fluorescent taxol-stabilized microtubules ( 4 µM polymerized tubulin ) . This step was conducted without GTP at 22°C and without taxol to avoid the formation of new microtubules with Cy3-tubulin dimers . We then detected the attraction of fluorescent tubulin through the co-precipitation with microtubules at 20 , 000×g for 10 min at 22°C . The analysis of the fluorescence intensities of both supernatants and pellets shows than Cy3-tubulin was indeed co-precipitated with MTs at spermidine concentrations higher than 10 µM ( Figure 6 ) . Even if the supernatant fluorescence intensity is lower at higher spermidine concentrations , we note that the concentration of free tubulin in the supernatant remains close to its maximum at low spermidine concentrations . Therefore the attraction of tubulin onto microtubule is relatively weak under these conditions , with both free and adsorbed Cy3-tubulin coexisting . We shall see in the “elongation” section that the strength of this interaction is an important parameter which modulates the efficiency of facilitated diffusion . As a control , figure S1 in supplementary files shows more directly the attraction of Cy3-tubulin on a MT pellet mediated by 200 µM spermidine . The probability that free GTP-tubulin molecules encounter long microtubule cylinders can be significantly larger than that of a direct encounter with MT ends . Therefore , after the encounters of tubulin with the MT wall , if the GTP-tubulin can proceed to the random scan of the MT via surface diffusion , a higher association rate of the free GTP-tubulin to the MT ends is expected ( Figure 1C ) . A parallel can be drawn here between DNA and MTs . Indeed it has been shown that facilitated diffusion of site-specific proteins via 3D diffusion and 1D diffusion on long DNA increases their association rate to target site up to two orders of magnitude [40] . In recent theoretical studies [40] , [41] , simple equations have been developed to describe qualitatively the effect of facilitated diffusion on DNA/protein association constant , in particular , the useful model developed by Hu et al . [42] which describes in a simple manner the influence of the attraction energy on facilitated diffusion . Via straightforward modification of this model , we extend its domain of application to free GTP-tubulin diffusion to the MT ends . As the persistence length of MTs is larger than 1 mm [43] , the MT is considered as a rigid cylinder . In addition , contrary to DNA , its larger radius implies that facilitated diffusion proceeds through 2D diffusion on the MT wall . The main assumption of this model is to consider that the flux of proteins delivered by 3D diffusion to the MT extremities at a distance lower than λ from one end is equal to the flux of protein delivered to the end via surface sliding at the equilibrium [42] . Considering λ as the sliding distance on the MT wall , if the separation distance from the nearest MT end and the new coming GTP-tubulin is shorter than λ , the GTP-tubulin can find its target site through 2D diffusion whereas for larger distances it will be released from the MT wall before reaching it . Then: ( 1 ) where D3 , D2 are the tubulin diffusion constants in the bulk solution and at the MT surface respectively , λ is the 2D diffusion length , a is the microtubule radius , Cfree is the free tubulin concentration , Cs is the surface density of tubulin adsorbed on the MT surface . The first term of this equation is the 3D diffusion near the MT extremities , and the second term represents the sliding . Three different regimes can then be derived ( see Text S3 ) : To describe microtubule assembly and dynamics in vitro , a simple model with two phases , nucleation and elongation is generally advanced . Nucleation proceeds via the assembly of several tubulin dimers into a nucleus of critical size [33] , [34] . The size of this nucleus varies with the ionic composition of the buffer and the presence of tubulin partners . Above this critical size , a short microtubule can be formed and then the elongation of the microtubule takes place via the addition of free GTP-tubulin to its extremities [33] , [34] , [36] , [37] , which leads to the formation of a long MTs . The elongation phase is nearly completed when the flow of new incoming GTP-tubulin is not sufficient to maintain the GTP cap at the MT ends [45] , [49] , [50] and thus to prevent MT depolymerisation [45] . The important point is that microtubule assembly requires the constant supply of GTP-tubulin dimers via 3D diffusion to nucleus and MT ends . The rate of encounters between tubulin dimers and a body of radius b0 ( nucleus or MT ends ) is given by the well known Smoluchowski equation: Js = 4πD3C0b0 , where D3 is the 3D diffusion constant of tubulin and C0 is the concentration of free tubulin . Using b0≈4–10 nm ( 4 nm is the value previously used to describe the tip size at the MT ends [51] ) , D3≈5 . 10−12 m2/s and C0 = 10 µM ( measured in sea urchin egg extracts [7] ) , we obtain about 1500–3800 collisions per second . Such a high collision rate via 3D diffusion is largely enough to sustain typical elongation rates ( about 14 µm/min [10] , [51] ) . Indeed , there are about 1640 subunits per µm of MT and , if we assume that both ends grow at the same rate , thus neglecting treadmilling , the association rate of free GTP-tubulin to one end needs to be only larger than 190 s−1 to maintain elongation [51] . It could then be advanced that MT dynamics is not diffusion limited . However , this conclusion is in apparent contradiction with several experimental evidences linking the GTP-tubulin supply to the MT ends and the mean rate of elongation . First , upon dilution , MTs collapse [52] . Second , near critical concentration , the GTP-tubulin supply becomes so critical that elongation only compensate shortening [45] . Third , the larger is the free tubulin concentration , the faster is the elongation [45] and the rate of tubulin nucleation also increases sharply with the free tubulin concentration [33] , [36] . One possible explanation for the discrepancy between the calculated high collision rate and the high sensitivity of both nucleation and MT elongation on free tubulin concentration is that the probability for a new coming tubulin to dock correctly on its target site ( lateral or longitudinal bounds ) is low [9] and thus requires several collision trials . To add credit to this idea , theoretical and experimental arguments have shown that an encounter can evolve into a stable complex only if the two interacting biomolecules are correctly oriented and positioned , with respect to theirs binding sites [8] . This geometrical constraint , which cannot be neglected in the case of tubulin and microtubule , leads to an estimated probability of complex formation per encounter 3 to 6 orders of magnitudes lower for typical proteins , depending on the number of binding sites and their dimensions [6] , [8] . Taking into account this geometrical factor , the net tubulin association rate to the nucleus or MT ends in the conditions described above would rather be between 0 . 0015–0 . 0038 and 1 . 5–3 . 8 s−1 , which is not sufficient to sustain MT assembly ( 190 s−1 ) . In this context , it is difficult to longer consider that MT assembly is not diffusion limited since many more encounters need to occur to allow microtubule elongation and the formation of critical nucleus . As an attraction between tubulin dimers can be mediated by multivalent counterions , we propose that polyamines could significantly increase the elongation and nucleation rates via facilitated diffusion mechanisms . Facilitated diffusion increases the flow of free tubulin arriving at the MT ends through tubulin sliding . As we can reasonably assume that a larger flow of tubulin favors elongation , a significant increase of the elongation rate is then expected under the best conditions for facilitated diffusion ( see Figure 9A ) . Such theoretical predictions are in close agreement with the experimental results presented in Figure 10 . A threefold increase of the pseudo-first order rate constant of elongation was measured experimentally with 100 µM spermidine . Such an increase could be critical in vivo when tubulin supply is scarce . We also observed that the pseudo-first order rate constant of elongation constantly increases with tubulin concentration in the presence of spermidine . It indicates that microtubule ends are not saturated with free tubulin and , therefore , elongation is still diffusion-limited in the presence of polyamines . The elongation rate could then be further increased up to GTP-tubulin saturation , which may arise at very high concentrations of tubulin . For the nucleation step , an attraction force mediated by polyamines induces a longer interaction lifetime between a new incoming GTP-tubulin and the forming nucleus . In agreement with this , experimental results clearly indicate that nucleation duration in the presence of polyamines is significantly decreased compared to control ( Figure 7 ) , whereas , the critical size of the nucleus may not be affected ( Figure 8 ) . Interestingly , the tenth time ( lag time of assembly ) no longer decreases for tubulin concentrations higher concentration than CL ( ∼20 µM , Figure 7B ) . This is an apparent discrepancy with classical theory and with typical in vitro experiments as a slight reduction of the pool of free tubulin increases the lag time by a dramatic factor ( Figure 7 and [33] , [34] , [36] ) . In contrast to elongation , the formation of a stable nucleus is the result of many relatively stable intermediate aggregates [34] . During this slow process , free tubulin dimers could accumulate on growing nucleus in the presence of polyamines . Consequently , the rate of nucleation no longer depends on GTP-tubulin concentration because of over-supply . This results is in agreement with in vivo reports from yeast showing that these cells do not have very sensitive mechanisms to regulate their tubulin level ( after a two-fold reduction ) , and that such regulation was not necessary for normal microtubule function [53] . The discrepancy between the necessity for a close regulation of free tubulin in vitro and the apparent independence of the nucleation rate on free tubulin level in vivo could thus be partly explained by polyamines beside other nucleation factors such as γ-tubulin [54] . The reciprocal consequence is that modulation of the polyamine level should influence microtubule nucleation . Another interesting point raised by the present study is that polyamines do not significantly stabilize microtubules ( Figure 11 ) . Indeed , polyamines participate to MT dynamics to both enhance nucleation and the formation of long microtubules but , in the same time , they do not prevent MT collapse which would have been toxic for living cells . The absence of a significant stabilizing effect of polyamines on microtubules was at first glance surprising; however it is most probably due to the fact that , in the MT wall , the distance between the C-terminal tails of adjacent tubulin heterodimers is too long to allow their significant overlapping . The model of facilitated diffusion provides basis to enlighten the interplay between polyamines and microtubule dynamics in living cells . The polyamine levels vary greatly during the cell cycle and in relation to the proliferation status of tissues [14] , [15] , [17] , [55] , which may significantly impact microtubule dynamics in living cells [15] , [16] , [56] . Indeed it has been shown that polyamine concentrations increase significantly during mitosis with a sequential patterns reflecting the fact that putrescine is a precursor of spermidine and spermidine in turn is a precursor of spermine [15] . In parallel to the increase of polyamines during mitosis , the rate of microtubule nucleation also increases sharply during mitosis [57] to reach at anaphase a sevenfold increase compared to that observed in the G2 phase . Our results suggest that the increase of the polyamine content could support the increase of MT nucleation and elongation . Against this proposal it could however be argued that MAPs , which are known to neutralize the negative charge of microtubules and to stabilize the microtubule wall , are such strong competitors for the microtubule neutralization that they might inhibit the overall effects of polyamines . However our results show that polyamines influence MT dynamics even in the presence of MAPs ( Figure 12 ) . Interestingly we also observed that the combination of taxol and spermidine cooperate to increase the microtubule mass , which indicates that polyamines and taxol could use different mechanisms ( facilitated diffusion and stabilization , respectively ) . In summary , our results lead to the following conclusions: Polyamines ( putrescine , spermidine , spermine ) were purchased from Sigma-Aldrich and used without further purification . Tubulin was purified from sheep brain crude extracts as described previously [58] . For long term storage , aliquots were stored at −80°C in 50 mM MES-KOH pH 6 . 8 , 0 . 5 mM dithiothreitol , 0 . 5 mM EGTA , 0 . 25 mM MgCl2 , 0 . 5 mM EDTA , 0 . 1 mM GTP , 30% glycerol ( v/v ) . Before use , an additional cycle of polymerization was performed in 50 mM MES-KOH pH 6 . 8 , 0 . 5 mM dithiothreitol , 0 . 5 mM EGTA , 6 mM MgCl2 , 0 . 5 mM EDTA , 0 . 6 mM GTP , 30% glycerol . Microtubules were sedimented by centrifugation ( 52 , 000×g , 30 min at 37°C ) at the end of which the microtubule pellet was resuspended in 25 mM MES-KOH pH 6 . 8 , 0 . 5 mM EGTA , 1 mM DTT and disassembled at 4°C for 20 minutes . Tubulin aggregates were finally eliminated by a further centrifugation at 4°C ( 52 , 000×g , 20 min ) . Tubulin concentration was determined by spectrophotometry using an extinction coefficient of 1 . 2 mg−1×cm2 at 278 nm [59] . Pure tubulin was pre-incubated on ice for 2 min in buffer M ( 25 mM MES-KOH pH 6 . 8 , 20% glycerol , 1 mM EGTA , 2 mM MgCl2 , 0 . 5 mM GTP ) in the presence or absence of various concentrations of polyamines . For Figures 4 and 5 , tubulin polymerization was initiated by placing the ice-cold cuvette ( 1 cm light path ) at 37°C in a PTI QuantaMaster 2000-4 thermostated spectrofluorimeter . The kinetics of microtubule assembly was then immediately monitored by 90° light scattering at 370 nm . For Figures 7 , 8 , and 10 , the tubulin samples were immediately placed in a pre-warmed cuvette ( time t = 0 ) to capture the kinetics of nucleation and elongation . This eliminates the time necessary for heating the cuvette in the spectrofluorometer . S-tubulin was prepared as previously described [60] . Briefly , subtilisin was added to tubulin with a 1∶200 subtilisin∶tubulin ( w/w ratio ) . The mixture was then incubated for 40 min at 25°C . In these conditions , about 95 % of tubulin was correctly cleaved by subtilisin as determined by gel electrophoresis ( percentage of cleavage ) and MALDI-TOF mass spectrometry ( mass of cleaved tubulin ) . Aliquots of S-tubulin were used fresh or kept frozen in 25 mM MES-KOH pH 6 . 8 , 1 mM EGTA , 10% glycerol . 12 µM of pure tubulin was pre-incubated on ice for 2 min in ( 25 mM MES-KOH pH 6 . 8 , 20% glycerol , 1 mM EGTA , 2 mM MgCl2 , 0 . 5 mM GTP ) with various concentrations of KCl and spermidine ( as indicated ) . Tubulin assembly was obtained after incubation at 37°C for 20 min . Microtubules were then pelleted at 25 , 000×g for 15 min at 37 °C and resuspended in 25 mM MES-KOH at 4°C in a volume equivalent to that of the supernatant . Equal volumes of supernatant and resuspended pellet were loaded on SDS-PAGE . To analyze the attraction of Cy3-tubulin on microtubules , Cy3-tubulin was prepared as described previously to obtain 0 . 5 Cy3 molecules per tubulin dimer [61] . Non fluorescent microtubules were preassembled in the presence taxol and then pelleted at low speed ( 20 , 000×g for 10 min ) . The pellet was washed several times to remove free taxol . After resuspension in buffer M without GTP , taxol-stabilized microtubules ( 4 µM of polymerized tubulin ) were then allowed to interact with 0 . 3 µM of Cy3-tubulin at various concentrations of spermidine . 100 µl of the reaction solutions were centrifuged at 20 , 000×g for 10 min to pellet MTs and MT-adsorbed Cy3-tubulin . Pellets were resuspended in the starting volume . Fluorescence intensities of pellets and supernatants were then measured using a PTI QuantaMaster 2000-4 thermostated spectrofluorimeter ( λexcitation = 550 nm , λemission = 569 nm ) . 1 . 4 mg/ml of twice cycled tubulin containing 15% of MAPs was incubated 5 min on ice in 25 mM MES-KOH pH 6 . 8 , 5 mM MgSO4 , 1 mM EGTA , 1 mM GTP , 0 . 7 mM ATP , 1 mM DTT with or without 200 µM spermidine or/and 5 µM taxol . The samples were then placed at 37°C for 15 min . Microtubules were then pelleted at 25 , 000×g for 15 min . Pellets and supernatants were then analyzed by SDS-PAGE . Microtubules samples from assembly reaction ( in buffer M ) were directly deposited on NiCl2 pretreated mica surfaces . Samples were deposited on the mica surface for a short time of about 30 s to prevent microtubules from collapsing on the surface as previously reported [62] . Finally , the surface was then dried with filter paper . AFM imaging was performed in intermittent mode . We used silicon cantilevers AC160TS ( Olympus ) with resonance frequencies of about 300 kHz . All images were collected at a scan frequency of 1 . 5 Hz and a resolution of 512×512 pixels . A first or second order polynomial function was used to remove the background slope .	Interactions between biomolecules ( DNA , proteins , sugar , etc . ) represent the link between genome information and function of living organisms . For effective competition between organisms and adaptation to environmental changes , these interactions have to take place at very high rates . As such interactions require successive associations and dissociations between biomolecules , most of the time could be spent in between interactions when biomolecules diffuse freely in the intracellular space until the next interaction occurs . To reduce this waste of time , cells have developed adaptive mechanisms . First , the concentration of biomolecules in the intracellular medium is very high , which increases their frequency of interactions . Second , the possibility for biomolecules to diffuse along linear polymers , a process named “facilitated diffusion , ” increases the probability for biomolecules to find their targets . This mechanism was first described to understand how DNA-binding proteins could find their specific targets among thousands of putative binding sites . We show here that facilitated diffusion could also play a significant role in promoting the assembly of cytoskeleton proteins that are involved in critical cell functions ( e . g . , division or neuron architecture ) . Alteration of this mechanism may be of particular interest for cancer and neuropathologies .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "biophysics/macromolecular", "assemblies", "and", "machines", "biochemistry/experimental", "biophysical", "methods", "biophysics/experimental", "biophysical", "methods", "biophysics/biomacromolecule-ligand", "interactions", "biochemistry/macromolecular", "assemblies", "and", "machines"...	2009	Polyamine Sharing between Tubulin Dimers Favours Microtubule Nucleation and Elongation via Facilitated Diffusion
Synapses are dynamic molecular assemblies whose sizes fluctuate significantly over time-scales of hours and days . In the current study , we examined the possibility that the spontaneous microscopic dynamics exhibited by synaptic molecules can explain the macroscopic size fluctuations of individual synapses and the statistical properties of synaptic populations . We present a mesoscopic model , which ties the two levels . Its basic premise is that synaptic size fluctuations reflect cooperative assimilation and removal of molecules at a patch of postsynaptic membrane . The introduction of cooperativity to both assimilation and removal in a stochastic biophysical model of these processes , gives rise to features qualitatively similar to those measured experimentally: nanoclusters of synaptic scaffolds , fluctuations in synaptic sizes , skewed , stable size distributions and their scaling in response to perturbations . Our model thus points to the potentially fundamental role of cooperativity in dictating synaptic remodeling dynamics and offers a conceptual understanding of these dynamics in terms of central microscopic features and processes . Chemical synapses are sites of cell–cell contact specialized for the rapid transmission of signals between neurons and their targets—muscles , glands or other neurons . The vast majority of synapses in mammals are found in the central nervous system ( CNS ) where they typically connect the axon of one neuron to the dendrite or soma of another neuron . Structurally , axonal presynaptic compartments are characterized by clusters of synaptic vesicles facing specialized regions of the presynaptic membrane , known as active zones ( AZs ) [1]; these , in turn , are juxtaposed against electron-dense thickenings of the postsynaptic membrane known as postsynaptic densities ( PSD; [2 , 3] ) . The molecular composition of AZs and PSDs is now known in great detail; furthermore , much is now known on the synaptic molecules themselves and on their interactions with other synaptic proteins and membranes . In parallel , recent experiments have provided information on the dynamics of synaptic molecules; these experiments have led to the realization that AZs and PSDs are not static structures but dynamic assemblies of molecules that move in , out and between synapses over time scales of seconds to many hours [4 , 5] . Such extensive spontaneous dynamics would seem to question the ability of individual AZs , PSDs and synapses in general to maintain their specific sizes ( e . g . areas of PSDs and AZs , volumes of spines and presynaptic boutons ) and functional properties over behaviorally relevant time scales . Indeed , live imaging studies consistently reveal that instantaneous molecular contents of individual synapses , and by extension , their functional properties , change continuously in manners that are only partially activity dependent ( e . g . [4–14 , 16–20 , 23–25 , 27] ) . The dynamics of synaptic molecules and synaptic properties motivated the formulation of abstract models aiming to describe the properties of individual synapses and synaptic populations [11 , 14 , 15 , 18 , 19 , 21–24 , 26 , 28 , 29] . Many of these models were based on low dimensional statistical processes in which each synapse was represented by a single probabilistic variable ( e . g . synaptic size; Fig 1A ) , with causal or deterministic relations emerging at the population level . Somewhat surprisingly , these descriptive models faithfully capture major aspects of synaptic features: the random-like changes in synaptic sizes along time; the stability and skewed shape of synaptic size distributions; the scaling of such distributions in response to changes in network activity as well as other perturbations , and , at the extremes , the dynamics of synapse formation and elimination [14 , 15 , 18 , 19 , 23 , 24] . While these descriptive models are quite successful in reproducing many of the aforementioned features , they provide little insight on the principles by which such macroscopic features might emerge from microscopic molecular dynamics within synapses . Conceivably , such insights might be obtained by constructing highly detailed dynamical models , which include all known synaptic molecules , their quantities , their interactions , and the kinetics of all such interactions ( Fig 1C ) . At present , however , in spite of extensive protein-protein interaction databases ( e . g . [30] ) , such models are still not feasible , mainly due to the paucity of data concerning binding affinities and kinetics . More importantly , however , even if such high dimensional , realistic models were feasible , it remains unclear if they are capable of providing intelligible insights on the principles by which microscopic molecular dynamics give rise to the macroscopic phenomena mentioned above [31] . In the current study , we describe a mesoscopic level exploration that aims to address the following question: Can the spontaneous dynamics exhibited by synaptic molecules give rise to the key features of individual synapses and synaptic populations described above ? If so , what are the essential aspects of these dynamics that are necessary for such features to emerge ? Mesoscopic level models can be valuable in this respect as they may reveal conceptually tractable principles which would be difficult to glean from highly detailed microscopic models or macroscopic descriptive models [31] . To construct a mesoscopic model , we distilled from the myriad features of synapses and synaptic molecules a small number of key attributes common to practically all synapses , namely a spatially localized patch of membrane , molecules that continuously bind and unbind to this patch , and the strong tendency of such molecules to interact among themselves ( Fig 1B ) . We then use these components to formulate several mesoscopic models of increasing complexity and test their ability to recapitulate major features of synaptic size dynamics , distributions and organization . We show that these macroscopic features emerge naturally from a simple biophysical process based on stochastic binding and unbinding of proteins to spatially confined patches of membrane and to each other , as long as both binding and unbinding have significant cooperative components . The emerging view of the synapse as a dynamic molecular assembly implies that at any given moment its size , composition , microscopic organization and ultimately its function , reflect the outcome of myriad processes in which synaptic molecules are assimilated or removed . This applies not only to relatively mobile constituents such as neurotransmitter receptors [5] and synaptic vesicles [7 , 25 , 32 , 33] , but also to synaptic building blocks known as scaffolding molecules . These molecules are generally believed to confer a degree of stability to the sizes and function of synaptic assemblies [4 , 34]; Moreover , pre- and postsynaptic scaffold molecule contents strongly correlate with functionally important measures of synaptic size , namely AZ and PSD area , respectively ( e . g . ; [6 , 16 , 35–38] ) . Both PSD molecule content ( e . g . [35] ) and PSD area ( e . g . [36 , 39] ) strongly correlate with dendritic spine volume , which has been repeatedly shown to correlate positively with synaptic strength ( e . g . [40–44] ) ; reviewed in [45 , 46] ) . Intriguingly , when scaffold molecule contents are followed at individual synapses over time ( hours , days ) they are found to change considerably ( e . g . [6–9 , 13 , 16–19 , 23 , 25 , 27] ) , reflecting in all likelihood , fluctuations in synaptic sizes ( e . g . PSD areas , spine volumes ) and strengths . Experimental studies ( e . g . [14 , 18 , 19 , 23] ) have given rise to several key observations regarding these fluctuations and their consequences . The first is the observation that whereas sizes of individual synapses fluctuate significantly in time ( Fig 2A and 2B ) , distributions of synaptic sizes in a network are very stable over time ( Fig 2D ) . The second is that these fluctuations are state dependent , namely they depend on the momentary size of the synapse: small synapses tend to grow larger and large synapses to become smaller ( Fig 2C ) , thus supporting the stable population distribution . The third is the observation that these stable size distributions are non-Gaussian and rightward skewed ( Fig 2D ) . Fourth , perturbations of network activity can modify synaptic size distributions while preserving distribution shapes , resulting in the collapse of the different distributions one onto each other when plotted in scaled units ( Fig 2E ) [18 , 23 , 47–50] . Finally , very recent studies based on super-resolution imaging techniques have revealed that PSDs and AZs consist of multiple nanodomains which seem to be dynamic and short lived [51–57] . Can this set of observations be effectively explained by a minimal biophysical model , which views the synapse as the net product of continuous , spontaneous assimilation and removal of their molecular constituents ? Below we test this possibility by examining and progressing through a set of biophysical models that differ in the modes of assimilation and removal and , in particular , in the level of cooperativity exhibited by these processes . The feasibility of each model is then assessed by testing the degree to which its outcomes reproduce the experimental observations described above . Our general framework employs a biophysical model in which synaptic size dynamics stem from binding and unbinding of scaffold proteins to a patch of postsynaptic membrane ( Fig 1B ) . The modeling platform is based on a representation of the postsynaptic membrane as a matrix composed of M potential binding sites for synaptic scaffold molecules . Synaptic size at any given time is then estimated as the number of occupied sites , that is , the number of scaffold molecules bound to the matrix . Scaffold molecule binding and unbinding are in principle stochastic events characterized by probabilities per unit time . Consequently , the number of molecules binding to the matrix per unit time depends on the binding probability and on the number of vacant sites . Similarly , the number of molecules dissociating per unit time depends on the unbinding probability per unit time and on the number of bound molecules ( = occupied sites ) . In this stochastic description , the binding and unbinding of proteins result in temporal fluctuations in synaptic size , i . e . in the number of molecules bound to the matrix . An ensemble of synapses is represented by multiple realizations of this stochastic process , and thus a distribution of synaptic sizes emerges across a population of synapses modeled in this fashion . In the continuum approximation , where fluctuations are neglected , binding and unbinding are described by rate equations . On average , the change in synaptic size is then the outcome of the net effect of these two processes , and average synaptic size S follows the continuous equation: dSdt=kon ( M−S ) −koffS ( 1 ) where kon and koff are the rate coefficients , the continuous analogs of the binding and unbinding probabilities per unit time , respectively . In general , the binding and unbinding rates are not necessarily constant , and may depend on interactions between the binding molecules; we consider below several cases of such interactions . In what follows , we use the continuous equation to represent the interactions in a compact way and to identify steady states at the level of synaptic populations . However , as we were primarily interested in studying fluctuations and their effects , the results below were mainly derived from discrete numerical simulations based on the stochastic analog of Eq ( 1 ) . In such simulations , matrices with dimensions of 50x50 were used , with these dimensions loosely derived from measurements of average PSD diameter ( 360-400nm ) and the granularity imposed by widths of canonical PSD molecules and glutamate receptors ( 5-20nm ) [3 , 58–60] . Monte Carlo simulations were used to determine the occurrence of binding and unbinding events , using a random number generator and the probability per unit time of the relevant event and the current state of binding sites in the matrix ( further details are given in the Methods section ) . Whenever possible , numerical simulation results were compared to continuum approximations , or to the master equation and its solutions ( for a review on the relation between continuum and stochastic descriptions , see [61] ) . In the simplest model of binding and unbinding , the rate coefficients kon and koff in Eq ( 1 ) are constant in space and time: kon = α , koff = β . In terms of individual molecular events , this implies that scaffold proteins bind to and unbind from the matrix independently from each other and that there are no interactions between molecules ( Fig 3A ) . This model is known in physical chemistry as the Langmuir adsorption model and is used to describe the kinetics of gas adsorption and desorption on a solid surface where interactions between molecules are negligible [62] . Stochastic simulations of the Langmuir model show that the distribution of synaptic sizes in a population is symmetric and well approximated by a normal ( Gaussian ) distribution ( Fig 3B ) . This is expected in light of the Central Limit Theorem , which states that the sum of a large number of independent , identically distributed random variables is approximately normally distributed , regardless of underlying variables . Since site occupancies are independent random variables , and since synaptic size is a sum of thousands of such variables ( 2 , 500 in the simulation ) , its distribution will be indistinguishable from a normal distribution . These simulation results are corroborated by direct solutions of the master equation in the Fokker-Planck approximation ( S1 Appendix , Section 2 . 2 ) . As mentioned above , one hallmark of synaptic populations is the broad and highly skewed distribution of their sizes ( e . g . Fig 2D ) . This observation rules out a model of independent binding and unbinding of scaffold molecules at the synaptic site , not a surprising conclusion in light of the multitude of interactions among synaptic proteins . This being so , we next examined whether molecular interdependency might defeat the Central Limit Theorem and give rise to the skewed distributions observed experimentally . Cooperativity is a key organizing principle in biology that represents a fundamental mechanism for accomplishing molecular interdependencies [63–69] . We here define cooperativity as the dependence of the binding affinity of one molecule on the state of the matrix in terms of other molecules already bound to it . In its most simple form , cooperativity could stem from the fact that many ( if not all ) synaptic molecules have multiple interaction sites through which they are capable of interacting simultaneously with other synaptic molecules; consequently , the presence of other molecules already bound to the matrix may increase the probability of an unbound scaffold molecule to bind as well . Conversely , their presence would be expected to reduce the probability of dissociating from the matrix due to a multiplicity of interactions with neighboring molecules . To examine whether this or other forms of cooperativity can give rise to the experimentally observed synaptic size dynamics and statistical features , we introduce unidirectional and then bidirectional cooperativity into the model as described next . A well-studied model of cooperative binding , which appears to extend the Langmuir model only slightly , is the Contact Process [70 , 71] . In this model , binding proceeds with a probability that increases with the number of occupied neighboring sites such that kon=λonχ ( 2 ) Here λon is a constant that indicates cooperative binding strength ( maximal value of kon ) and 0<χ<1 is the fraction of occupied nearest neighbors ( with the definition of nearest neighbors depending on matrix geometry as illustrated in Fig 4A ) . Specifically this implies that the probability of a molecule to bind the matrix at a given site increases if that site has neighboring sites that are already occupied . In contrast to binding , unbinding in the Contact Process is not cooperative and occurs independently from occupancy , consequently , koff = β . The introduction of cooperative binding changes the behavior of the model dramatically , and results in a transition between two qualitatively different phases . In the supercritical phase , ( λon/β is larger than a threshold critical value ) , the system approaches a stationary state . The resulting distribution of synaptic sizes in this phase still converges to a normal distribution ( See S1 Appendix , S4 Fig ) . In the subcritical phase ( λon/β is below this threshold ) , all realizations fall into an absorbing state in which all sites are vacant . A similar model was previously used to address the question of how synapses persist even though their molecular constituents continuously enter and leave the synaptic assembly [22] . It was found that this model extends the expected lifespan of the synaptic assembly relative to the timescales of single molecule binding and unbinding; yet , as expected , assembly sizes ultimately collapse to zero . To drive the system away from this absorbing state , that is , to obtain steady states at which assemblies have non-zero sizes , an additive constant term α can be incorporated into the binding rate , such that kon=λonχ+α ( 3 ) Biophysically , α represents low affinity binding to unoccupied matrix sites ( Fig 4A ) , serving , at the extreme , to seed synapse formation on an empty matrix . The additive component α thwarts the collapse into the absorbing empty state , resulting in stable limiting distributions . Unfortunately , as the stochastic simulations show , when α is large and dominant , limiting distributions of synaptic sizes become approximately Gaussian; this might be expected , as a major contribution comes from independent binding . Conversely , if α is relatively small , slightly skewed distributions can be found , but only for small values of λon/β ( Fig 4B and 4C ) . However , in this case , the mean synaptic size is very close to zero ( Fig 4B ) . Although mean synaptic size in our simulations has no particular biological meaning , the very low occupancy fraction found at steady states obtained by adding α to the Contact Process , would seem to imply very sparse occupancy of the postsynaptic membrane . This does not agree with what is known from quantitative ultrastructural studies [58–60] reviewed in [3] as well as others [51–57] . Furthermore , distribution skewness in this case is essentially the result of proximity to the lower limit on synaptic size , S = 0 . Collectively these findings indicate that unidirectional cooperativity in the form of the Contact Process , even with the addition of spontaneous binding to the matrix ( α ) , fails to robustly account for the experimental observations described above , due to the inevitable tradeoff between distribution skewness and mean size . The inadequacy of unidirectional cooperative binding to capture the statistical features of synaptic populations led us to examine bidirectional cooperativity . In fact , although the Contact Process has been studied extensively , from a biophysical standpoint it is highly unlikely that binding is cooperative while unbinding is not . In the bidirectional cooperativity model , the number of occupied neighboring sites modulates both binding and unbinding probabilities ( Fig 5A ) . As the occupancy of neighboring sites increases , the probability of binding increases and the probability of unbinding decreases . Thus , the binding rate coefficient kon and the unbinding rate coefficient koff are expressed as follows: kon=λonχ+α and koff=λoff ( 1−χ ) ( 4 ) where χ is the fraction of occupied neighboring sites as before , and λoff is the constant for cooperative unbinding ( the maximal value of koff ) . Stochastic simulations of the bidirectional cooperativity model reveal that , in contrast to the models described above , skewed synaptic size distributions can be obtained which are remarkably similar to those observed in experimental measurements ( compare Fig 5B and 5C with Fig 2D ) . For a given parameter set , these model distributions are stable over time ( Fig 5B ) . Moreover , synaptic size distributions in the model exhibit scaling similar to that observed experimentally ( Fig 2E ) : Increasing the cooperative binding coefficient λon leads to a broadening of synaptic size distributions , which approximately collapse one on another after scaling ( Fig 5D and 5E ) . Similar scaling is also observed when λoff is modified . We sought to identify the parameter regime that gives rise to such stable and skewed distributions in our model . Analysis of the continuum approximation of the model suggests that a stable distribution will be reached under the condition α < λoff−λon ( see S1 Appendix , Section 1 ) . To identify conditions for the emergence of skewness , we considered a simplified version of the bidirectional cooperativity model in which cooperativity acts globally on the entire synapse ( see S1 Appendix , Section 2 . 1 ) . This model still retains the main ingredient of cooperativity in both binding and unbinding processes but allows for solving the master equation , an equation for the probability for obtaining a certain occupancy state over time . The solution highlights a parameter combination that crucially affects the skewness of the steady-state distribution ( see S1 Appendix , Sections 2 . 2 , 2 . 3 ) . This is the cooperativity ratio C which characterizes the strength of cooperative binding and unbinding relative to the strength of the non-cooperative processes . In the model discussed here it is C=λon+λoffα ( 5 ) ( In S1 Appendix , a more general version is considered of which this is a special case ) . Fig 6A shows the steady-state synaptic size distributions for the global cooperativity model over a wide range of this parameter . As intuitively expected , when C is small , binding and unbinding are dominated by non-cooperative processes and steady-state distributions are symmetric , narrow and Gaussian-like ( orange distributions ) . When C is large , binding and unbinding are dominated by cooperative processes , and broad and skewed distributions can arise as observed in experiments ( blue distributions ) . Fig 6B depicts the same distributions in scaled units , highlighting the skewness as a dimensionless shape characteristic . These results are congruent with simulations of the bidirectional cooperativity model where interactions between scaffold molecules are local ( S1 Fig ) . To summarize , skewed distributions are obtained as long as cooperative binding is dominant relative to non-cooperative processes . We examined whether the bidirectional cooperativity model also captures the dynamic properties of synaptic ensembles shown in Fig 2 . Stochastic simulations show that sizes of individual synapses exhibit fluctuations that qualitatively resemble those observed for real synapses ( Fig 7A; compare with Fig 2B ) . The scatter plot of changes in synaptic size as a function of their original size shows the same dependence observed experimentally ( Fig 7B; compare with Fig 2C ) . Moreover , when plotting synaptic sizes as a function of their original sizes at increasingly greater time intervals ( Fig 7C and 7D ) the slopes and offsets of linear regression lines in such plots gradually decrease and increase respectively ( Fig 7E ) , just as observed for excitatory [23] and inhibitory [19] synapses . Similarly , the coefficient of determination , or R2 , gradually decreases ( Fig 7F ) , suggesting a gradual “deterioration” of synaptic configurations as previously shown for excitatory and inhibitory [19] synapses . The findings described so far suggest that introducing bidirectional cooperativity allows the model to recapitulate the experimentally observed statistical properties of synaptic sizes in a population of synapses . In addition to these properties , recent experiments show that synapses , both inhibitory and excitatory , are not uniform structures but are organized as “nanoclusters” that change over time [51–57 , 72] . Does the same model recapitulate this dynamic internal organization of individual synapses ? To test this , we examined the spatial patterns of bound molecules in our simulations and the changes in these patterns over time . We found that bound molecules do organize into nanocluster-like patterns ( Fig 8A ) ; moreover , “time-lapse” sequences revealed that these patterns “morph” in manners reminiscent of dynamics displayed by nanoclusters in glutamatergic synapses [51–53] . In fact , we note that spontaneous binding to the matrix ( through the parameter α ) creates transient “seeds” which can potentially nucleate nanocluster formation through cooperative binding to such seeds . To further quantify this clustered organization , we used autocorrelation analyses ( as in [56] ) to characterize the degree of spatial heterogeneity in the simulated synaptic structures . The spatial autocorrelation function g ( r ) provides a measure of the molecular density at a distance r from a particular molecule relative to the average molecular density of the whole structure . Thus , when molecules are randomly distributed on the matrix , g ( r ) = 1 . The extent to which g ( r ) exceeds 1 is related to the degree of clustering . As shown in Fig 8B , synaptic structures exhibit a much higher heterogeneity compared to randomly scattered molecules , indicating a high degree of clustering , in common with experimental observations [55 , 56] . We also quantified the number of nanoclusters formed in these simulations using the algorithm employed in [56] for identifying nanoclusters of synaptic scaffold proteins . As shown in Fig 8C , the average number of nanoclusters per synapse calculated by this method was 3 . 4±1 . 5 which is comparable to the number of PSD-95 nanoclusters observed experimentally ( 1 . 86 ± 0 . 07 per PSD ) [56] . While the number of clusters in the model clearly depends on matrix size and model parameters , we note that these were initially chosen according to experimental observations as described above , giving roughly several hundred scaffold molecules per synapse . With the same parameters , the resulting numbers of clusters are also comparable with observed values . The bidirectional cooperative model described here captures many features of synaptic dynamics previously observed in real neurons . Nevertheless , this mesoscopic model is undoubtedly simplistic and based on premises whose biological correctness is not obvious . Furthermore , its sensitivity to implementation details is not obvious either . Hence , matters of appropriateness and robustness warrant some discussion . The first matter concerns the existence of a matrix as a binding substrate . At first sight , this would seem to be an entirely artefactual construct . The concept of a matrix , however , finds substantial justification when considering the fact that synapses form at contacts between pairs of elongated structures , that is axons and dendrites or dendritic spines; such contacts define and circumscribe regions within which axonal and dendritic molecules can interact across the synaptic cleft while simultaneously interacting with intracellular molecules such as scaffold molecules [73–75] . Thus , an axodendritic contact defines a specialized membrane patch that is effectively the equivalent of a matrix . The dimensions and geometry of such membrane patches undoubtedly vary , yet it is notable that our model produces broad and skewed distribution of synaptic sizes , even for uniform matrix sizes . Our results were not particularly sensitive to matrix size , as long as α was maintained at sufficiently low values such that non-cooperative binding did not become dominant , and only small numbers of “nanoclusters” were formed . For any given value of α , increased matrix size was associated with reduced skewness , which can be understood when considering that in these cases , synaptic size was the sum of sizes of many nanoclusters formed independently of each other , reducing the dominance of cooperativity and increasing the dominance of the independent binding . Interestingly , nanocluster numbers in real synapses tend to be very low [55 , 56] in agreement with this observation and its expected effects . As to other possible matrix geometries ( such as hexagonal matrixes ) , alternatives were not explored; we did find , however , that smaller numbers of neighbors did not qualitatively affect our results ( S2 Fig ) . A second matter concerns the model’s simplicity—a single molecule type and only two types of interactions ( Fig 5A ) . Real postsynaptic densities contain hundreds of different molecule types [76] which typically bind to multiple other molecules , creating a bewilderingly complex interaction network [2 , 30 , 76] . The dynamics arising from such rich and complex collections of interacting molecules remain unknown , yet we tentatively suggest that the principles we outline here may hold in general: the more molecules bound to the postsynaptic matrix , the higher the probability of recruiting additional molecules to the same matrix . Conversely , the greater the number of molecules a particular molecule is bound to , the lower its probability of dissociating from the matrix . Indeed , in-vivo measurements of PSD-95 molecular dynamics [9] suggest that large PSDs capture more free PSD-95 and retain it for longer durations as compared to small PSDs . We thus expect that this form of cooperativity ( sometimes referred to as avidity [64] ) will give rise to qualitatively similar dynamics and population properties . A third matter concerns the linear dependence of binding and unbinding rates on the number of bound neighbors . The exact description of binding kinetics and their relation to physical interactions is a highly nontrivial aspect of surface science , even for relatively simple physical interactions [77] . Energy considerations and detailed balance impose some constraints but do not define the kinetics uniquely . All the more in our model , which is highly abstract and provides no more than a simplified sketch of a synaptic molecule assembly . We used linear dependence since it provided a simple realization of the principle of cooperativity as described above . Moreover , it enabled us to analyze a corresponding global cooperativity model , and obtain solutions of its master equation ( S1 Appendix ) . Nevertheless , we cannot exclude the possibility that our findings may not apply universally to all possible cooperativity models . A final matter concerns the parameter regimes used here . We noted that this regime is constrained by several considerations . We found that it is important to keep non-cooperative binding rates much smaller than cooperative rates in order to obtain skewed distributions of synaptic sizes; this is in line with the large number of interaction partners most synaptic molecules have . Additionally , values of λon very close to those of λoff were required in order to obtain “reasonable” mean synaptic sizes ( in terms of matrix occupancy; S3 Fig ) . At first sight , this requirement would seem to question the model’s robustness . We note , however , that from a biophysical standpoint , λon encompasses not only particular binding kinetics but also the concentration of free molecules that can potentially bind to the matrix; put differently , the rates at which molecules bind to the matrix are also proportional to free molecule concentrations . In our treatment so far , this dependence was not made explicit , and free molecule concentration was encapsulated in λon . Separating λon into these two components , however , gives rise to an interesting observation ( S4A Fig ) : For a broad range of total molecule concentrations , λon settles on values that are very close to those of λoff . Consequently , even when total molecule concentrations are changed several fold , the condition λon≈λoff is maintained and distributions of synaptic sizes remain skewed and stable . These same changes , however , affect mean synaptic size dramatically , ( S4B Fig ) . In summary , changing total molecule concentrations ( readily realized by altering protein synthesis or degradation rates , for example ) changes mean synaptic sizes and drives synaptic size distribution scaling ( as previously suggested , e . g . [48 , 50] ) , yet only minimally affects λon , which remains very close to λoff . Consequently , the parametric regime λon≈λoff is very reasonable in the context of our model . Mesoscopic models in which the synapse is described as an assembly of dynamic molecules have been put forward in several prior studies . Thus , for example , Shouval [22] in an approach already mentioned , depicted the synapse as a matrix to which neurotransmitter receptors can be added or removed . In a second study [21] the synapse was modeled as a three layer system divided laterally into synaptic and extrasynaptic regions . It was shown that cooperative interactions between synaptic molecules could give rise to persistent postsynaptic sites , which transiently trap receptors as they diffuse laterally in the plasma membrane . In a third approach [11] , a model based on reaction-diffusion equations for scaffold proteins and receptors was shown to give rise to postsynaptic domains ( via a Turing mechanism ) , that coexist with rapid receptor diffusion in the cell membrane plane . All these studies were aimed at explaining the long-term persistence of synapses in face of continuous diffusion , exchange and turnover of their molecular constituents . Very recently , a mesoscopic biophysical model based on diffusion , aggregation and removal of receptors and scaffold proteins in the membrane was used to explain the statistics of PSD molecule clusters [72] . None of these models , however , examined how such molecular dynamics may give rise to spontaneous synaptic remodeling or population properties such as size distribution shapes or their scaling . Conversely , the mesoscopic model described here shows how these properties emerge naturally from simple well-known biological processes , namely cooperative binding and unbinding , and by doing so provides a conceptually tractable explanation of these phenomena . Clearly , as mentioned above , it is an enormously simplified description of the postsynaptic specialization . However , its main ingredients—a postsynaptic membrane , dynamic molecules that continuously bind and unbind , and a strong tendency of such molecules to interact with multiple other molecules—are now well established facts . We thus carefully suggest that the nanoscale organization of synaptic scaffolds , the spontaneous , size dependent fluctuations in synaptic sizes , the gradual erosion of synaptic configurations , the skewed distribution of synaptic sizes and their scaling in response to global changes in synaptic molecule concentrations , are all likely to be driven , at least in part , by spontaneously occurring cooperative assimilation and loss of synaptic molecules . Naturally , real synapses will have many additional means of control through which they might change specific binding and unbinding affinities , the repertoire and abundance of synaptic molecules and the supply of metabolic energy required to fuel some of these reactions . Nevertheless , we conjecture that these additional means are layered upon foundations consisting of principles exposed by our simplistic model . Cooperativity is a ubiquitous and crucial regulation mechanism in a large variety of processes , including molecular recognition , enzyme catalysis , membrane transport , protein folding , and self-assembly of supramolecular complexes [63–69] . In the context of synaptic biology , cooperativity plays key roles not only in the formation of multi-molecular scaffolds ( e . g . [78–83] ) but also in synaptic function , where it is mostly appreciated in relation to neurotransmitter release [84] . Along these lines it is intriguing to note the considerable functional variability in space and time exhibited by presynaptic boutons as well as the skewed shape of various presynaptic property distributions ( e . g . [85–89]; reviewed in [90] ) . Interestingly , skewed distributions [91 , 92] as well as nonstationary properties ( e . g . [93 , 94]; reviewed in [95] ) feature prominently in neuronal functional and structural features . As a final note we wish to remark that our model , in its most generic and abstract sense , concerns the dynamics and statistical outcomes of stochastic , cooperative construction and deconstruction processes; consequently , the study’s conclusions are not necessarily limited to synaptic , neuronal , or , for that matter , biological settings . In fact , it is reasonable to expect that when collections of multiple instantiations of cooperative constructive and deconstructive processes are examined , these might exhibit features similar to those described here , that is , state dependent fluctuations in the properties of individual instantiations , and , at the same time , skewed and stable distributions of the same properties in populations of such instantiations . All simulations were performed using scripts written in Matlab ( MathWorks , MA , USA ) . A number of simulations were also repeated using code written in C . Monte Carlo simulations were performed to assess the dynamics and statistics that result from each one of the three models and test their congruence with experimental measurements . Specifically , for each model , the trajectories of 3500 synapses were simulated over 1500 time steps . At each time step and for each site , the fraction of occupied nearest neighbors χ was calculated by counting the number of occupied nearest neighbors and dividing it by the total number of nearest neighbors . The binding and unbinding probabilities for vacant and occupied sites , respectively , were determined by the mode of interactions presented by each model . A site changed its binding state if the probability calculated for this site was larger than a random number sampled from a uniform distribution between 0 and 1 . Unless stated otherwise , we used the following parameter values: λoff = 0 . 5 t−1 , λon = 0 . 493 t−1 , α = 0 . 0007 t−1 ( t stands for time ) . The geometry of the postsynaptic density was chosen , for reasons of simplicity , to be a 50x50 square matrix , giving a total of M = 2 , 500 sites . In this geometry , the maximal number of nearest neighbors is 8 and the fraction of occupied nearest neighbors χ was calculated accordingly . Parameters were chosen to give roughly several hundred scaffold proteins per synapse as observed for glutamatergic synapses in the mammalian central nervous systems [3] . Simulations were performed using a time step of 1 ( arbitrary units ) . Results were not significantly altered when time steps were decreased by factors of 2 to 100 . Code used for all MATLAB simulations is provided as S1 Code . Spatial autocorrelation analysis was used to quantify the clustering of synaptic proteins . The autocorrelation function g ( r ) is a measure of bound protein density at a distance r away from a given bound protein relative to the density of the whole matrix . The density at a certain distance r was calculated by averaging the number of occupied sites at distance r from each occupied site and dividing it by the total number of sites at distance r . The autocorrelation function was then obtained by performing the same calculation for different values of r and normalizing it by the density of the whole matrix . The case r = 0 was not considered due to its trivial contribution . For a higher precision , this analysis was performed for 3500 synapses and the autocorrelation function was taken as their average . The number of nanoclusters in the bidirectional cooperativity model was calculated using agglomerative hierarchical clustering algorithm as employed in [56] for analyzing scaffold proteins nanoclusters . Occupied sites were partitioned into sub-clusters using MATLAB functions pdist ( ) , linkage ( ) and cluster ( ) . The node height cut-off of the dendrogram was determined by the mean of nearest neighbor distances between occupied sites + 2 standard deviations . This analysis was performed for each time point to measure the morphing of clusters in time .	Neurons communicate through specialized sites of cell–cell contact known as synapses . This vast set of connections is believed to be crucial for sensory processing , motor function , learning and memory . Experimental data from recent years suggest that synapses are not static structures , but rather dynamic assemblies of molecules that move in , out and between nearby synapses , with these dynamics driving changes in synaptic properties over time . Thus , in addition to changes directed by activity or other physiological signals , synapses also exhibit spontaneous changes that have particular dynamical and statistical signatures . Given the immense complexity of synapses at the molecular scale , how can one hope to understand the principles that govern these spontaneous changes and statistical signatures ? Here we offer a mesoscopic modelling approach—situated between detailed microscopic and abstract macroscopic approaches—to advance this understanding . Our model , based on simplified biophysical assumptions , shows that spontaneous cooperative binding and unbinding of proteins at synaptic sites can give rise to dynamic and statistical signatures similar to those measured in experiments . Importantly , we find cooperativity to be indispensable in this regard . Our model thus offers a conceptual understanding of synaptic dynamics and statistical features in terms of a fundamental biological principle , namely cooperativity .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "medicine", "and", "health", "sciences", "engineering", "and", "technology", "nervous", "system", "condensed", "matter", "physics", "electrophysiology", "neuroscience", "simulation", "and", "modeling", "probability", "distribution", "mathematics", "statistics", "(mathematic...	2017	Cooperative stochastic binding and unbinding explain synaptic size dynamics and statistics
Human filarial infection is characterized by downregulated parasite-antigen specific T cell responses but distinct differences exist between patients with longstanding infection ( endemics ) and those who acquired infection through temporary residency or visits to filarial-endemic regions ( expatriates ) . To characterize mechanisms underlying differences in T cells , analysis of global gene expression using human spotted microarrays was conducted on CD4+ and CD8+ T cells from microfilaremic Loa loa-infected endemic and expatriate patients . Assessment of unstimulated cells showed overexpression of genes linked to inflammation and caspase-associated cell death , particularly in endemics , and enrichment of the Th1/Th2 canonical pathway in endemic CD4+ cells . However , pathways within CD8+ unstimulated cells were most significantly enriched in both patient groups . Antigen ( Ag ) -driven gene expression was assessed to microfilarial Ag ( MfAg ) and to the nonparasite Ag streptolysin O ( SLO ) . For MfAg-driven cells , the number of genes differing significantly from unstimulated cells was greater in endemics compared to expatriates ( p<0 . 0001 ) . Functional analysis showed a differential increase in genes associated with NFkB ( both groups ) and caspase activation ( endemics ) . While the expatriate response to MfAg was primarily a CD4+ pro-inflammatory one , the endemic response included CD4+ and CD8+ cells and was linked to insulin signaling , histone complexes , and ubiquitination . Unlike the enrichment of canonical pathways in CD8+ unstimulated cells , both groups showed pathway enrichment in CD4+ cells to MfAg . Contrasting with the divergent responses to MfAg seen between endemics and expatriates , the CD4+ response to SLO was similar; however , CD8+ cells differed strongly in the nature and numbers ( 156 [endemics] vs 36 [expatriates] ) of genes with differential expression . These data suggest several important pathways are responsible for the different outcomes seen among filarial-infected patients with varying levels of chronicity and imply an important role for CD8+ cells in some of the global changes seen with lifelong exposure . Infection with the pathogenic filariae , Loa loa , Brugia malayi , Wuchereria bancrofti , and Onchocerca volvulus , causes an enormous disease burden throughout tropical and sub-tropical regions of the world . Interestingly , however , the clinical manifestations of infection are often markedly different in those with lifelong exposure ( i . e . those born in filarial-endemic regions ) and those that acquire infection later in life through travel to or temporary residence in a filarial-endemic area [1] , [2] . Indeed , filarial infections are less likely to be subclinical in expatriates [3] or transmigrants [4] compared to those with lifelong exposure . Expatriates with loiasis , for example , are more likely to have Calabar swellings and other “allergic” phenomena – such as marked peripheral blood eosinophilia , elevated IgE levels , and urticaria – than is seen in the more chronically infected patients born and raised in endemic areas [5] . With availability of more sensitive assays for the definitive diagnosis of filarial infections , it is now known that infection occurs at much earlier ages than once believed [6] , although relatively intense exposure to the vectors that transmit these infections is typically required for acquisition of infection . However , expatriates who acquire infection are not subject to many of the environmental and familial factors that affect those born in endemic regions , the most notable being the alteration of immune responses specific for filarial antigens that occurs early in life [7] , [8] , [9] , and that can persist long-term ( decades ) [10] as a consequence of in utero exposure to filarial antigens . Moreover , polyparasitism is much more frequent among patients from filarial-endemic regions than in expatriates . That individuals living in an endemic area are exposed continually to the parasite , irrespective of the infection status , is evidenced by the Ag-specific antibody responses seen among filarial-uninfected endemic individuals [11] , [12] . Indeed , both susceptibility to infection and the nature of the immune response has a significant genetic component in helminth- and filarial-endemic populations [13] , [14] , [15] , [16] . Several studies have also demonstrated differences in immune responses to filarial antigens among filarial-infected travelers ( expatriates ) and those from filarial-endemic regions [1] , [2] . Filarial-infected individuals from endemic countries , while having increased antifilarial IgG4 antibodies [17] , have more profoundly diminished parasite-specific T cell responses [12] , [18] than those seen in expatriates [1] . This parasite-specific hyporesponsiveness is reflected not only in diminished proliferative and cytokine responses [12] , [18] , [19] , but also in the increased expression of molecules ( e . g . CTLA-4 , PD-1 ) known to inhibit T cell responses [20] , [21] . In addition , filarial Ags and live filarial parasites have themselves been shown to induce proliferative defects [22] , apoptosis of T cells [23] , and impairment of antigen presenting cell number and function [24] , [25] , [26] , that cumulatively may alter T cell responses . A number of studies have directly examined specific ( or candidate ) pathways in the cells of filarial-infected [24] , [25] individuals . To examine more globally the differences in responsiveness to filarial infections between persons with relatively newly acquired infection and those with lifelong exposure and to evaluate more comprehensively the T cell responses ( both CD4+ and CD8+ ) seen in these two groups , we utilized spotted , human microarrays and RNA from either CD4+ or CD8+ T cells ( ex vivo ) and in response to filarial and nonfilarial antigens . Our findings demonstrate a striking difference in gene expression between endemic and expatriate patients with the same filarial infection and demonstrate that these differences manifest not only in T cells ex vivo but also in response to both parasite and nonparasite Ag . All patients were seen under a protocol ( NCT00001230 ) that was approved by the Institutional Review Board of the National Institute of Allergy and Infectious Diseases , National Institutes of Health ( NIH ) , and informed written consent was obtained from all subjects . Three Loa loa-infected patients who had lived most or all of their lives in a region endemic for loiasis and 3 expatriate L . loa-infected individuals were chosen for study ( Table 1 ) . All patients were examined at the NIH , and all had demonstrable microfilariae in their circulation at midday . None of the patients tested positive for HIV . One expatriate patient ( but no others ) had intestinal parasites as well . PBMCs from all patients were collected prior to any treatment , cryopreserved using standard techniques and stored in liquid nitrogen until used . For this study , cryopreserved PBMCs were thawed and then layered over Ficoll/diatrizoate ( MP Biomedical , LLC , Solon , OH ) to separate viable cells from dead cells . Cells at the interface , were collected , washed , and counted ( >98% viable by trypan blue exclusion ) . Cells were cultured in 12-well plates in RPMI-1640 ( Invitrogen , Carlsbad , CA ) supplemented with 10% fetal calf serum ( Gemini BioProducts , Woodland , CA ) at 10×106 cells/well in the absence ( media alone ) or presence of a PBS extract of microfilariae ( MfAg , 10 mg/ml ) or with a non-parasite control Ag Streptolysin O ( SLO , 1∶100 final concentration; Difco , Detroit , MI ) . Wells with media alone were run for each antigen and for each cell type ( i . e . a total of 4 wells with media alone were cultured for each patient ) . Following a 16 hr . incubation , PBMCs were harvested and washed in PBS/0 . 1% BSA/2 mM EDTA . Primary selection for CD3+ cells was accomplished using the Dynal negative T cell isolation kit II ( Invitrogen , Carlsbad , CA ) known to retain activated T cells . After negative selection , positive selection for either CD4+ or CD8+ T cells was accomplished using Dynal beads , giving a >99% pure population of each cell type as determined by flow cytometry . After selection , cells were immediately homogenized in 1 ml Trizol ( Invitrogen ) followed by phase separation with chloroform . Following the addition of 70% ethanol to the RNA-containing aqueous phase , RNA was further purified using the RNeasy Kit ( Qiagen , Valencia , CA ) following the manufacturer's instructions . RNA concentrations were analyzed on a NanoDrop spectrophotometer ( NanoDrop Technologies , Wilmington , DE ) and quality was assessed using the Agilent 2100 Bioanalyzer ( Agilent Technologies , Santa Clara , CA ) . From a starting template of 40 ng RNA , cDNA was synthesized by reverse transcription , then amplified and labeled using the Ovation Aminoallyl RNA Amplification and Labeling System ( NuGen Technologies , San Carlos , CA ) following the manufacturer's instructions . Technical replicates were done for each sample . The resulting aminoallyl labeled cDNA was purified using QIAquick columns ( Qiagen ) and cDNA concentrations were measured . For each patient sample ( CD4+ and CD8+ cells ) , replicate cDNA from Ag-driven and media samples were processed concurrently for microarray analysis; because each of the two Ags were run in duplicate , cDNA from all four media samples was used in the analyis of unstimulated cells . For hybridization , aminoallyl cDNA was first labeled with cyanine ( Cy ) dyes ( Amersham , Piscatawey , NJ ) using Cy3 dye for antigen-driven samples and Cy5 dye for samples from media alone . Following purification on QIAquick columns , corresponding Ag and media Cy3 and Cy5 labeled samples were combined , concentrated , and then hybridized to human spotted arrays ( NIAID - Hsbb; Platform GPL1054 ) printed by the NIAID/Microarray Research Facility . The probe set for the microarrays was based on 70 mer oligonucleotides from the Human Genome Oligo Set V2 . 0 ( Qiagen ) . Each array contained 21 , 531 oligonucleotides . Microarray chips were imaged with a GenePix 4000B fluorescent scanner ( Molecular Devices , Sunnydale , CA ) . Data has been deposited in the National Center for Biotechnical Information ( NCBI ) Gene Expression Omnibus at http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? token=vnitzmsyymoigly&acc=GSE31894 , GEO accession number ( GSE31894 ) . Processed data can be accessed by using the “Series Matrix File ( s ) ” link . To confirm the quality of the microarray findings , identical RNA was used for comparison to quantitative real-time reverse transcription ( RT- ) PCR ( Taqman™ ) on a representative selection of differentially expressed genes in unstimulated cells ( 10 for CD4+ cells , 7 for CD8+ cells ) . Pre-developed assay reagents , primers , and probes were obtained from Applied Biosystems ( Branchburg , New Jersey ) and run per the manufacturer's instructions using an ABI 7900 Real-Time PCR System . All samples were run in triplicate and normalized to their own 18S ribosomal RNA . Gene expression for each patient was calculated as the antilog of ( [1/ΔCT]×100 ) where ΔCT = cycle threshold ( CT ) of the test gene minus the CT of the 18S ribosomal gene . To establish if there was a correlation between the expression of a given gene using RT-PCR and the expression determined by microarray , the 3 patient values for each gene within each group were averaged for both techniques and then plotted against each other . Correlation values were assessed by the Spearman Rank Test using GraphPad Prism 5 . 0 . To determine differences between numbers of genes expressed between the two patient groups , a statistical test for one proportion was done using the Normal approximation . For each category , the number of genes assigned to that category in the two classes ( endemic and expatriate ) was compared to each other . Assuming a null proportion of 0 . 5 ( i . e . that there is no difference in the number of genes of that category for the two classes ) , p-values were calculated for deviation from 0 . 5 using a Normal approximation . Among the 21 , 531 oligonucleotides on the arrays , expression by CD4+ and/or CD8+ T cells could be detected in ∼23% ( n = 5044 ) . As can be seen in Figure 1 , those subjects with lifelong exposure ( hereafter referred to as endemics ) had slightly greater numbers of differentially-expressed genes than did the expatriates ( 109 vs 80 for CD4+ cells; 136 vs 118 for CD8+ cells; at p<0 . 01 , corrected for FDR ) . Seventeen of these differentially expressed genes were independently validated using RT-PCR , as shown in Figure 2 and Table S1 in which the expression values from the microarray data were correlated with those data derived from the RT-PCR ( p = 0 . 0002 for CD4+ cells [10 genes] , and p = 0 . 035 for CD8+ cells [7 genes]; Figure 2 ) . When the genes expressed differentially between patient groups were analyzed further , the differences between endemic and expatriate patients could be inferred through functional assessments of these gene sets ( Figure 3 ) . For CD4+ cells , endemic patients over-expressed a significantly greater number of genes related to inflammatory disease ( 26 vs 3 for expatriates; p<0 . 0001 ) . Cell death-associated genes were also over-represented in endemic patients ( 31 genes vs 15 in expatriates ) although the difference did not reach statistical significance when corrected for multiple comparisons ( p = 0 . 09 ) . Although also not statistically different , expatriates had an overrepresentation of genes associated with transcriptional activity , cell function/maintenance and cell growth/proliferation . Overall , the number of genes that were differentially expressed by either patient group was greater in the CD8+ than in the CD4+ T cells . Similar to the findings for CD4+ cells , the CD8+ cells in endemic patients showed an expansion in the number of genes involved in pathways associated with “inflammatory disease” ( 35 genes in endemics vs 7 in expatriates , p = 0 . 0003; Figure 3 ) and with cell death ( 43 vs 29; p>0 . 05 ) . In contrast , the number of genes implicated in cell signaling ( 12 vs 2 ) and molecular transport ( 16 vs 6 ) , though not significantly different , were overrepresented in the CD8+ cells of expatriates . Both groups had a large number of differentially expressed genes related to cell death ( Figure 4 ) . As can be seen in this representative network , the endemic patients upregulated many of the genes involved in caspase activation whereas the expatriates were more likely to have relatively upregulated activation of the NFkB complex , including MAPK8 , part of the p38 MAPK activation complex . An additional network associated with IFN-γ , IFN-α , and IL-2 was also seen for the CD4+ cells of expatriates ( data not shown ) . Furthermore , expatriates showed a relative upregulation of several pro-apoptotic genes including PDCD4 ( programmed cell death 4; CD4+ and CD8+ T cells ) , HTRA2 ( HtrA serine peptidase 2; CD8+ T cells ) , and STK4 ( serine/threonine kinase 4 [MST1]; CD4+ cells ) and some involved in anergy induction ( DGKA ( diacylglycerol kinase-alpha; CD8+ cells; data not shown ) . While pro-inflammatory and activation-associated networks were also seen in endemic patients , represented by genes that included KLRD1 , CCL4 , CXCR4 , JAK2 and HLA-DRA [Figure 4 , Figure S1 , and Figure S2] ) , there was , in addition , a differential increase in the expression of several immunoregulatory molecules including IRF2 ( interferon regulatory factor 2; CD4+ and CD8+ cells ) , TIMP2 ( TIMP metallopeptidase inhibitor 2; CD4+ cells ) , PTGDR ( prostaglandin D2 receptor; CD8+ cells ) , and MAF ( musculoaponeurotic fibrosarcoma oncogene homolog; CD8+ cells ) . More importantly , the endemic patients differed in the number of genes directly linked to the increase ( DIABLO , CASP4 , MS4A1 , GNLY ) or inhibition ( BIRC3 , DHCR24 ) of caspase-dependent ( Figure 4 ) or caspase-independent ( CD99; data not shown ) apoptosis , suggesting there was a fine balance between pro- and anti-apoptotic molecules in the baseline CD4+ and CD8+ cell response to filarial infection in vivo . In addition to the prominent cell death/inflammatory networks discussed above , CD8+ T cells of endemic patients also demonstrated a differential increase in molecules coding for proteins with cytotoxic effector function , among them the killer cell lectin-like receptor molecules KLRB1 and KLRD1 , granzyme A ( GZMA ) and granulysin ( GNLY ) . Hierarchical clustering and Gene Set Enrichment Analysis ( GSEA ) was utilized to determine whether patient groups could be distinguished by their canonical pathway profiles . Analysis of CD4+ cells ( Figure S1 ) , showed significant enrichment ( enrichment of 56%; corrected p<0 . 01 ) in only the Th1/Th2 pathway among the endemic patients compared to the expatriate subjects . In marked contrast , several pathways were significantly and differentially enriched in CD8+ T cells from both patient groups ( Figure S2 ) . Among the canonical pathways in which there was enrichment in endemic patients were the IL-5 ( 150%; p<0 . 1 ) and eosinophil ( 75%; p<0 . 1 ) pathways . Two pathways enriched by ∼150% with respect to expatriate patients were the Asbcell and BBcell pathways , both which are involved in T cell-B cell interactions . The most highly enriched canonical pathways in expatriates included the PAC1 receptor pathway ( >150%; p<0 . 01 ) and the WNT Ca2 cyclic GMP pathway ( 95%; p<0 . 01 ) , as well as the complement/coagulation and granulocyte cell survival pathways . Figure 5 shows the number of genes that were either upregulated or downregulated in response to filarial MfAg or to the nonparasite Ag SLO compared to unstimulated cells ( based on a paired p-value<0 . 001 ) . As can be seen , the response to either Ag ( in general ) was different between the two patient groups . Interestingly , in response to MfAg , only the chemokine ligand CCL3 and tubulin folding cofactor A ( TBCA ) in CD4+ cells and glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) and TBCA in CD8+ cells were upregulated in common between the patient groups . In addition , there were relatively few ( n = 21 ) genes that were regulated similarly to both antigens ( MfAg and SLO ) inclusive of both groups of patients and both cell types , suggesting that the nature of the antigen ( and presumably the method of sensitization ) plays a major role in shaping the antigen-specific T cell response . In general , the more chronically infected patients ( endemics ) had a greater number of antigen-induced differentially regulated genes than did the expatriates; however , the most striking difference was in the number of genes differentially expressed by CD8+ cells ( 107 vs 15 , p<0 . 0001; Fig . 5 ) . Genes that were significantly up- or down-regulated ( p<0 . 001 ) in response to MfAg compared to their expression in unstimulated cells were assessed through pathway analysis . For CD4+ T cells , these functions included cell death , cell assembly and organization , cell development , and cell function and maintenance in endemic patients , while in expatriates the most prominent functions were cell assembly and organization , cell function and maintenance , and cell movement ( data not shown ) . Similarly , the principal functions associated with CD8+ T cells included cell growth and proliferation , cell development , and cell death in endemic patients , and cell death , inflammatory disease , and gene expression in expatriates . To establish an overall profile of genes upregulated in loiasis patients in response to MfAg , composite networks were identified ( representative networks shown in Figure 6 ) . Figure 6A illustrates that the majority of genes with altered expression from that seen in unstimulated cells were those associated with the NFkB ( both patient groups ) and Caspase ( endemic group ) complexes . For the endemic patients , however , an increase in the expression of genes linked to insulin and insulin signaling in both CD4+ and CD8+ T cells ( Figure 6A and data not shown ) could be seen to be induced by MfAg . A second composite network ( Figure 6B ) consisted chiefly of those genes upregulated in the T cells from endemic patients . This analysis clearly identified several molecules involved in activation , cell trafficking , and Ag presentation , including CD69 in CD4+ and CD8+ cells , CXCR4 and CCL3 ( MIP1α ) in CD4+ cells , and CD247 ( TCRζ ) and CCL22 in CD8+ cells . Further analysis of genes upregulated in endemic patients showed molecules associated with histone complexes ( such as histone h3 ) in both CD4+ and CD8+ cells ( Figure 6B ) , with the protein kinases AKT , P38MAPK , JNK , and ERK1/2 in CD4+ cells ( Figure 6B and data not shown ) , and with ubiquitin related molecules in CD8+ cells ( Figure 6B ) . As indicated in Figure 5 , because of the differences in numbers of genes with increased expression to MfAg , the contribution by expatriates to the overall composite profile was far less than that by the endemic patients . Moreover , when networks were examined by cell type , it was apparent that the majority of the response to MfAg in expatriates was associated with the CD4+ T cells ( Figure 5 and data not shown ) , differing from endemic patients in which responses were seen in both CD4+ and CD8+ cells . The most clear-cut contribution by expatriates to the overall response ( Figure 6A ) was an upregulation of genes associated with the calcium-binding protein calmodulin . Further analysis of individual networks derived from the CD4+ cells of expatriates determined that their response to MfAg was largely a pro-inflammatory one , comprised of genes either directly or indirectly associated with cytokines ( IL-6 and IL-12 ) , chemokines ( CCL3 [MIP1α] ) , or with the high mobility group box 1 ( HMGB1 ) molecule , known to be involved in the inflammatory response to antigenic stimuli ( data not shown ) . To establish the profile of downregulated genes in response to MfAg in loiasis patients , composite networks were again created ( representative network shown in Figure 7 ) . In the CD4+ cells of endemic patients , insulin-like growth factor I receptor ( IGF1R ) and RARRES3 , a member of the suppressive retinoid family ( Figure 7 and data not shown ) were markedly downregulated in response to MfAg , while in CD8+ T cells , members of the GTPase RAS superfamily of GTP binding proteins ( RAC1 ) and the reticulon family ( RTN4 ) , as well as inhibitors of the NFkB ( NFKB1A ) and AKT ( PTEN ) protein kinase complexes were decreased ( Figure 7 and data not shown ) . Further analysis of CD8+ networks showed that molecules either directly or indirectly associated with TNF , IL1B , and TGFB1 were also downregulated in endemic patients relative to their expression in media ( data not shown ) . The few molecules downregulated in expatriate T cells included FURIN ( in CD4+ cells ) and HLA-E ( CD8+ cells; Figure 7 ) . Genes that were significantly up- or down-regulated to MfAg with respect to unstimulated cells were also examined by GSEA . Pathways associated with two chemokine receptors ( CCR5 and CXCR4 [40% and 35% enrichment respectively] and for the transcriptional corepressor PELP1 ( 60% ) were enriched ( and upregulated ) in MfAg-stimulated cells from endemic patients ( Figure S3 ) . MfAg-driven downregulated pathways associated with the pro-apoptotic molecule BAD ( 52% ) as well as ERK ( 25% ) were also enriched in endemic patients . Interestingly , there were several canonical pathways associated with the insulin growth factor 1 molecule that were decreased in response to MfAg , including the IGF1 , Longevity , and IGF1 MTOR pathways ( 52% , 45% , and 23% enrichment respectively ) . In contrast to what was seen in GSEA analysis for endemic patient cells , expatriate patients' cells in response to MfAg had enriched pathways associated with the innate and adaptive inflammatory response ( Figure S4 ) . These included the pathways for Notch signaling ( most highly enriched at >500% ) , erythrocyte differentiation ( 180% ) , NKT activation ( 70% ) , and Toll-Like Receptor Signaling ( 25% ) . The DNA fragment pathway ( >500% ) was the major pathway downregulated in expatriates suggesting antigen-induced anti-apoptotic mechanisms . In marked contrast to the observations of responses to MfAg were the findings to the nonparasite Ag SLO . Most noticeable was the relative commonality in the responses seen in CD4+ T cells from both the endemic and expatriate patients ( Figure 5 ) with both the numbers of up- and down-regulated genes and functional pathways associated with transcriptional activity ( cell death , cell growth and proliferation , and gene expression ) being similar . Of note , the response to SLO in both groups consisted of increased expression of several chemokine ligands ( including CCL20 , CCL4 and CCL8 ) , interferon regulatory factors , genes associated with inflammatory cytokines , and importantly , the signaling molecule STAT1 , necessary for the production of interferons ( data not shown ) . The response in CD8+ T cells to SLO , however , was dramatically different between the endemic patients and the expatriates with the endemic population having many more genes altered ( n = 156 vs 36 ) . Moreover , 77% ( 120/156 ) of these genes were downregulated compared to Ag-unstimulated cells ( Figure 5 ) with many being either indirectly or directly associated with NFkB ( data not shown ) . Of particular interest was the observation that only endemic CD8+ T cells failed to upregulate STAT1 in response to SLO ( data not shown ) . For some time now , it has been known that differences in clinical manifestations exist between filarial-infected patients with lifelong exposure and those with significantly less exposure ( travelers/temporary residents; [1] , [2] , [5] ) . It has been felt that these disparities may reflect differences in immunologic responsiveness to the parasite in these patients . To address specifically the concept that chronicity of infection results in differences in the nature of immune responses to that infection , we examined the global gene expression of both CD4+ and CD8+ T cells from individuals who differed primarily in the length of time infected with the blood-borne filarial parasite , Loa loa ( Table 1 ) . Our data suggest that , while the expression of the majority of genes ( >5 , 000 ) examined by microarray was similar between the two groups , there were significant differences in the T cell responses ex vivo as well as in response to parasite antigen and even to a bystander antigen . Previous work has demonstrated that cells of filarial-infected endemic patients have markedly diminished parasite-specific T cell responses when compared to filarial-infected expatriate patients and even to uninfected endemic individuals [1] , [12] . In a study of transmigrants to an O . volvulus-endemic area from a non-endemic region , it was shown that recent infection was associated with vigorous parasite-specific proliferative and cytokine responses that differed in comparison to the diminished responses seen in the chronically infected patients [30] . Similar findings have been seen in patients with acute or subacute schistosomiasis infection who had higher parasite-specific proliferative responses than did those with longstanding , chronic infection [31] . In addition , filarial-infected patients from endemic regions of the world cured of filarial infection following treatment continue to show diminished T cell responses to filarial antigens [32] while expatriate patients cured of infection ( and not re-exposed ) recover many of their Ag-specific T cell responses [33] . A major finding in this study was the importance of the inflammatory and cell death networks in the ex vivo ( unstimulated ) cells of filarial-infected patients in both patient groups though individual genes within these networks segregated by patient group . For example , within cell death networks the expatriates were more likely to express genes associated with activation induced cell death whereas the endemic patients expressed genes associated with apoptosis . That increased cellular activation , cell death , and inhibition of cell death is occurring at a steady-state , suggests that under conditions of long-term Ag stimulation , a balance between pro- and anti-apoptotic transcriptional events ( e . g . DIABLO and BIRC3; [34] , [35] ) is seen in those with longstanding infection . Indeed , the finding that chronic filarial infection is associated with increased numbers of memory but decreased numbers of effector T cells [36] may support the idea of activation and T cell survival being tightly regulated through pro- and anti-apoptotic mechanisms . Many of the genes differentially expressed by unstimulated T cells of endemic patients have known regulatory and/or inhibitory roles in immune and inflammatory responses . Such molecules included TIMP2 , a suppressor of endothelial cell proliferation ( a clinical hallmark of filarial infection ) , the transcription factor IRF2 , a competitive inhibitor of IRF1 mediated transcription of IFN-α and β and a factor in the upregulation of FasL [37] , and the transcriptional activator and repressor MAF . The increase in MAF , a molecule that plays a role in increased T cell apoptosis as well as in the production of IL-4 and IL-10 ( a prominent regulatory cytokine in filarial infection [19] ) , but inhibits production of IFN-γ and IL-12 [38] , supports previous findings of an increased production of IL-4 IL-13 and IL-10 in microfilaremic loiasis patients [39] . Moreover , the interference of IRF2 and MAF with the IFNs strongly suggests an anti-inflammatory role for these molecules that , at the very least , impairs Th1 differentiation in chronically infected patients . In addition to these regulatory molecules , the receptor for prostaglandin D2 ( PTGDR ) was upregulated in the unstimulated CD8+ cells of endemic patients in comparison to expatriates . This receptor-ligand interaction decreases the migration of Langerhans' cells in the skin [40] , the cytotoxicity of NK cells [41] , and the expression of both IFN-γ and IL-2 [42] , all of which further serves to downregulate both the innate and adaptive immune responses . Moreover , CD8+ cells also overexpressed the chemokine ligand CCL4 , the secretion of which mediates CD8+ T regulatory cells to suppress T cell responses [43] . Taken together , the increased expression of all of these molecules in endemic patients may reflect the lack of clinical symptoms [1] , [2] and parasite-specific in vitro T cell responses [1] , [12] frequently observed in chronically infected patients with filarial infections and suggest possible mechanisms for the regulation of inflammatory activity . To examine the larger relationships between genes that were differentially expressed by either patient group GSEA was used . Interestingly , those pathways associated with CD8+ T cells were the most significant in the unstimulated cells of both patient groups ( Figure S2 ) . In endemic patients the IL-5 and eosinophil pathways [44] were significantly enriched in CD8+ T cells as well as CD4+ T cells . Two highly enriched pathways in endemic patients were associated with T cell-B cell interactions , the ASBCell Pathway ( involved in Ag dependent B cell activation ) and the BBCell pathway ( involved in the induction of apoptosis in Fas-expressing inactive B cells ) , suggesting a possible role for T cells in the Ag-induced activation and cell death of B cells in filarial infection . The CD8+ T cell pathways identified in the expatriates were related most often to metabolic and cell maintenance functions and included the PAC1 receptor pathway , associated with the activation of adenylyl cyclase and phospholipase C , and the WNT Ca2 cyclic GMP pathway . However , expatriate T cells also demonstrated an enrichment of those pathways involved in killing ( granulocyte cell survival and complement/coagulation ) as well as those associated with allergic functions ( adrenergic pathway ) . The enhancement of these latter pathways , might serve to explain the augmented pathology associated with infection in expatriates with Loa infection [5] , [45] seen to a much lesser degree in those with long-term infection . For MfAg stimulated recall responses , one common finding between the patient groups was that , unlike the enrichment of CD8+ pathways seen in unstimulated cells , it was the pathways in CD4+ cells that were significantly enhanced in response to parasite Ag , not surprising given the HLA-Class 2 restriction of many of the T cell responses [46] . Indeed , although the number of altered genes in response to MfAg was far fewer in expatriates in comparison to endemic patients , those genes that were upregulated were closely tied to pathways of the innate and adaptive inflammatory response ( i . e . Notch , and TLR pathways; Figure S4 ) . This increase in inflammatory pathways was further supported by the upregulation of molecules associated with NFKB activation and calmodulin-mediated responses ( Figure 6 and data not shown ) as well as by a downregulation of the apoptotic DNA fragment pathway ( Figure S4 ) . In endemic patients with long-term infection , the response to MfAg was difficult to synthesize fully . Compared with expatriates , where the recall response to MfAg was clearly of an inflammatory nature , the T cell response to MfAg by endemic patients appeared to be balanced between activation and regulation of the immune response . This balance was perhaps most clearly seen in the transcriptional regulation of molecules associated with cell death and apoptosis , similar to the findings in unstimulated cells . The downregulation of the pro-apototic BAD pathway ( Figure S3 ) and the molecule RTN-4 , an inhibitor of the anti-apoptotic factors Bcl-2 and Bcl-XL [47] , in addition to the upregulation of DHCR24 ( an inhibitor of caspase-3; [48] ) would counter mechanisms designed to increase cell death , including the upregulation of CXCR4 ( Figure 6B and Figure S3 ) , a mediator of CD95-independent cell death in CD4+ cells [49] , and the downregulation of the receptor and pathways for IGF1 , a potent proliferative and anti-apoptotic signaling system ( Figures 7 and Figure S3 ) . Numerous other examples of the opposing nature of certain functions associated with differentially regulated genes in endemic patients were also in evidence . Although several regulatory molecules were upregulated in the unstimulated cells of endemic patients , other such molecules as RARRES3 , a retinoic acid family member that functions as a negative regulator of cell proliferation , as well as NFKB1A and PTEN , suppressors of the NFkB and AKT complexes respectively , were all downregulated in MfAg driven cells ( Figure 7 ) . Furthermore , several molecules associated with activation ( CD69 and and CD247; Figure 6B ) and chemokine functions ( CCL3 [MIP1γ , also found in expatriate T cells] , CCL22 [a chemotactic molecule for chronically activated but not resting T cells] ) , as well as the canonical pathway for CCR5 were upregulated . Of particular interest is the interaction of CCR5 with its ligand CCL5 which induces the release of histamine from basophils and activates eosinophils , two common features associated with filarial infection [50] . Several other multi-functional complexes upregulated in the parasite Ag-driven cells of endemic patients may offer clues to mechanisms underlying the depressed T cell responses typically seen in these patients . Ubiquitin and its associated molecules ( Figure 6B ) function in a wide variety of cellular processes including Ag presentation and apoptosis and , through a post-translational modification , mark proteins for degradation . Recently , GRAIL ( the E3-ubiquitin ligase gene related to anergy in lymphocytes ) was shown to be responsible for the Th2 hyporesponsiveness in a mouse model of chronic schistosomiasis [51] , a finding suggested by data from human filarial infections [12] , [20] , [52] . In addition , several genes linked to the histone family of molecules were also upregulated , including the gene for histone h3 . This molecule , through an ERK dependent mechanism , allows the binding of transcription factors to the IL-10 promoter and subsequent expression of the IL-10 gene , which , as mentioned previously , is a prominent regulatory molecule associated with chronic filarial infection [53] , [54] . The additional finding of increased expression of molecules associated with insulin has parallels in two recent studies in which: 1 ) activation of the IL-4/Stat6 pathway , important in immunity to helminths [55] , [56] , has been shown to increase insulin action during helminth infection [57] , and 2 ) there was a relationship between eosinophil production of IL-4 , alternatively activated macrophages in adipose tissue , and enhanced glucose tolerance in a mouse model of Nippostrongylus brasiliensis infection [58] . Finally , when the CD4+ T cell responses to the nonparasite Ag SLO were analyzed , an upregulation of several molecules associated with activation and inflammation ( chemokines , cytokines , others ) was seen in the T cells of both endemic and expatriate patients . Indeed , since many studies have demonstrated the similarity in cytokine and proliferative responses to nonparasite Ags even between filarial-infected and -uninfected individuals [59] , the overlap of gene expression between patient groups in the CD4+ T cell microarray data would be expected . These similarities were not seen in the SLO response in CD8+ T cells , responses that were extremely different between the two groups . Of particular note was the observation that Stat-1 , important for IFN signaling , was not upregulated in the CD8+ cells of endemic patients as it was in CD4+ cells as well as in both T cell types of expatriates . With the recent findings that helminth-infected individuals have altered responses to Mycobacterium tuberculosis [60] , malaria [54] , HIV [61] , and even to vaccines [62] , [63] , [64] , it may be that CD8+ T cells play a larger role in the global modulation of the immune system seen in patients with chronic helminth infection . The present study thus demonstrates that the clinical and immunological differences previously observed between endemic and expatriate patients can be demonstrated at the transcriptional level in unstimulated ( ex vivo ) cells , during early recall responses to parasite Ag , and even to nonparasite Ag . Indeed , the transcriptional differences between the two groups of filarial-infected patients reflect many of the differences that are seen between acute and chronic viral infection [65] . While the microarray findings in this study by no means constitute the final analysis of the differences between patients with long-standing infection and those with more recently acquired infection , they do suggest several mechanisms that warrant further investigation . It must be argued , therefore , that chronicity ( and possibly in utero or neonatal exposure to filarial antigens ) helps define the disparities between these two groups of patients . Further characterization of the differences seen between long-term and newly acquired infection could help to define the natural progression of filarial infection and the responses that underlie this progression .	Infection with the filarial parasite Loa loa causes a parasite-specific downregulation of T cell responses . However , differences exist ( clinical and immunologic ) between patients born and living in filarial endemic regions ( endemics ) and those who become infected during travel or short-term residency ( expatriates ) . T cell responses are more depressed in endemics while expatriates have more clinical “allergic-type” symptoms . In this study , we showed that these differences reflect transcriptional differences within the T cell compartment . Using microarrays , we examined global gene expression in both CD4+ and CD8+ T cells of microfilaremic endemic and expatriate patients and found differences not only ex vivo , but also to parasite and , for CD8+ cells , to nonparasite antigens . Functional analysis showed that endemic patients expressed genes linked to inflammatory disease and caspase associated cell death at homeostasis while expatriates tended to have a more activation-induced gene profile at homeostasis and a CD4+ inflammatory response to parasite antigen . Patient groups were similar in their CD4+ response to nonparasite antigen but strongly differed in their CD8+ responses , demonstrating the potential global ramifications of chronic , longstanding infection . Our study describes potential transcriptional mechanisms for the variability seen in patients with different levels of exposure to and chronicity of filarial infection .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "medicine", "infectious", "diseases", "biology", "molecular", "cell", "biology", "parasitic", "diseases" ]	2012	Regulation of Global Gene Expression in Human Loa loa Infection Is a Function of Chronicity
Wnt proteins regulate many developmental processes and are required for tissue homeostasis in adult animals . The cellular responses to Wnts are manifold and are determined by the respective Wnt ligand and its specific receptor complex in the plasma membrane . Wnt receptor complexes contain a member of the Frizzled family of serpentine receptors and a co-receptor , which commonly is a single-pass transmembrane protein . Vertebrate protein tyrosine kinase 7 ( PTK7 ) was identified as a Wnt co-receptor required for control of planar cell polarity ( PCP ) in frogs and mice . We found that flies homozygous for a complete knock-out of the Drosophila PTK7 homolog off track ( otk ) are viable and fertile and do not show PCP phenotypes . We discovered an otk paralog ( otk2 , CG8964 ) , which is co-expressed with otk throughout embryonic and larval development . Otk and Otk2 bind to each other and form complexes with Frizzled , Frizzled2 and Wnt2 , pointing to a function as Wnt co-receptors . Flies lacking both otk and otk2 are viable but male sterile due to defective morphogenesis of the ejaculatory duct . Overexpression of Otk causes female sterility due to malformation of the oviduct , indicating that Otk and Otk2 are specifically involved in the sexually dimorphic development of the genital tract . Wnt proteins bind at the cell surface to transmembrane receptors , which transduce the signal to downstream components of the various branches of Wnt signal transduction [1] . In addition to members of the Frizzled receptor family , which were the first Wnt receptors to be identified [2] , Wnts were also shown to bind to the transmembrane proteins low density lipoprotein receptor-related protein 5/6 ( LRP5/6 ) [3] , [4] , receptor tyrosine kinase-like orphan receptors 1/2 ( Ror1/2 ) [5] , [6] , related to receptor tyrosine kinase ( Ryk ) [7]–[9] , muscle specific kinase ( MuSK ) [10] , syndecan [11] and protein tyrosine kinase 7 ( PTK7 ) [12] , reviewed in [13] . Some of these Wnt co-receptors form a receptor complex together with a Frizzled protein , whereas others are capable of binding Wnts in the absence of Frizzleds . In general , it is thought that the presence of different classes of Wnt receptors and co-receptors on the cell surface increases the specificity of the interaction of Wnts with their target cells and also serves to channel the Wnt signal into either the canonical , β-catenin-dependent branch or one of the so-called non-canonical branches of Wnt signaling [14] , [15] . One Wnt co-receptor of particular interest is PTK7 . PTK7 is required for the control of planar cell polarity ( PCP ) in vertebrates . Mice mutant for PTK7 show an open neural tube , defects in convergent extension movements during gastrulation and polarity defects of inner ear hair cells , which are classical PCP phenotypes in vertebrates [16]–[21] . PCP phenotypes were also observed upon knock-down or mutation of PTK7 in Xenopus [18] and zebrafish [22] . PTK7 knock-down in Xenopus furthermore caused defects in migration of cranial neural crest cells [23] , very similar to animals in which the function of Dishevelled , an intracellular component of Wnt signaling , has been impaired [24] . Regarding the mechanism of how PTK7 controls PCP , it was clearly demonstrated in Xenopus that PTK7 interacts physically with several components of Wnt signal transduction . As expected for a putative Wnt co-receptor , PTK7 binds to Wnt proteins and forms a complex with Frizzled7 , LRP6 and Dishevelled [12] , [23] , [25] . Somewhat unexpectedly , although PTK7 is supposed to promote the non-canonical PCP branch of Wnt signaling , it was found to bind to the canonical ligands Wnt3A and Wnt8 but failed to bind the non-canonical ligands Wnt5A and Wnt11 [12] . Additional experiments showed that co-overexpression of PTK7 with either Wnt3A or with Wnt8 blocked the capability of these Wnts to induce a second body axis in Xenopus , consistent with a function for PTK7 in suppression of canonical Wnt signaling , which may be essential for activation of non-canonical signaling [12] . This interpretation was recently strengthened by similar findings in zebrafish [22] , but is in conflict with other studies coming to essentially the opposite conclusion , according to which PTK7 promotes canonical Wnt signaling [25] , [26] . In Drosophila , mutations in genes that were reported to interact physically and functionally with PTK7 in vertebrates , including frizzled ( fz ) [27] , [28] , dishevelled ( dsh ) [29] and Van Gogh/strabismus ( Vang ) [30] all show PCP phenotypes . Surprisingly , until recently this was not reported for the proposed PTK7 homolog in Drosophila , encoded by the gene off-track ( otk ) [31] , [32] . These studies reported that a loss-of-function mutation of otk is embryonic lethal and causes axon pathfinding defects of certain embryonic motor nerves [32] and targeting defects of outer photoreceptor axons to the lamina in the developing eye [31] . Only recently it was reported that otk mutant embryos show cuticular defects pointing to a function in the determination of segment polarity [12] . These authors furthermore showed that Otk can bind Wnt4 and that mutation of otk blocks the occurrence of cuticular patterning defects observed upon overexpression of Wnt4 [12] . Together , these data led to the conclusion that Otk is a receptor for Wnt4 required for its function in embryonic patterning . Due to these obvious inconsistencies in the published literature we have reinvestigated the function of PTK7/Otk in Drosophila . We found that there are in fact two homologs of PTK7 in the fly that were most likely the result of a gene duplication that occurred only in the genus Drosophila but not in other insect species . In addition to the previously described otk gene we identified otk2 to encode a second close homolog of PTK7 . otk2 is positioned directly adjacent to otk on the second chromosome and is expressed in a pattern identical to otk . We generated null mutations for otk , otk2 and a double mutation that deletes both genes together . In contrast to the published literature , we found that both single mutants were homozygous viable and fertile without showing any of the previously reported mutant phenotypes . The otk , otk2 double mutant was homozygous viable but male sterile due to defective morphogenesis of the ejaculatory duct . The male sterile phenotype caused us to investigate potential interactions of both Otk and Otk2 with Wnt2 , which is also required for male fertility [33] . We indeed show that embryonic expression of otk depends on Wnt2 and that both Otk and Otk2 coimmunoprecipitate with Wnt2 , Fz and Fz2 , indicating that Otk and Otk2 are co-receptors for Wnt2 . Our results provide important information for unraveling the system of Wnt ligands and their specific receptors involved in male fertility . Furthermore , they may have implications for studying male fertility in humans , since mutation of Wnt7a , the mouse homolog of Drosophila Wnt2 , also causes male and female sterility [34] . Previous analyses of otk , the proposed Drosophila homolog of PTK7 , did not reveal any function in the control of PCP or β-catenin-dependent Wnt signaling [31] , [32] . Given that mutation of PTK7 causes strong loss-of-function phenotypes in vertebrates [18]–[20] , [23] , we speculated that there may be a second gene in the fly genome that could function redundantly with otk . Indeed , when we performed a BLAST search with the protein sequence of Otk we found that the protein encoded by the gene CG8964 is closely related to Otk ( 53% identity over 427 amino acids , Figure S1 ) . CG8964 is located right next to otk ( see http://flybase . org/ ) on the second chromosome [2R: 7 , 910 , 651–7 , 912 , 775 ( - ) ] , suggesting that it is the result of a gene duplication . Therefore , we named the gene CG8964 off-track2 ( otk2 ) . Phylogenetic analysis confirmed that Drosophila otk and otk2 are indeed two paralogs of the single PTK7 gene in mouse and human ( Figure 1A ) . To test whether the gene duplication is specific for Drosophila or occurred also in other arthropods , the sequences from different arthropod species homologous to Drosophila Otk and Otk2 were analyzed . The resulting phylogenetic tree clearly shows that two Otk paralogs can only be found in Drosophila species , but not in other arthropod species ( Figure S2 ) . Hence , a lineage specific duplication has generated two PTK7 homologs in Drosophila . Otk is a single-pass transmembrane protein of 114 kD consisting of five extracellular immunoglobulin-like domains , a single transmembrane domain and a kinase homology domain ( Figure 1B ) . Its paralog Otk2 has a molecular weight of 48 kD and only comprises three immunoglobulin-like domains , a single transmembrane domain and a short cytoplasmic domain of 69 amino acids ( Figure 1B ) . To answer the question which parts of the much shorter Otk2 sequence correspond to which parts of the Otk sequence , a dot plot was performed , visualizing matching residues in both sequences . This analysis demonstrated that nearly the entire Otk2 sequence ( except for the signal peptide ) matches to a contiguous stretch within the Otk sequence ( Figure S1B ) , ranging from the third immunoglobulin-like domain to the beginning of the kinase homology domain . To test whether Otk and Otk2 have the ability to interact in a homo- or heterophilic manner , we performed co-immunoprecipitation experiments with epitope-tagged proteins transiently transfected into S2r+ cells . To test for homooligomerization , Myc- and GFP-tagged Otk constructs were co-overexpressed in S2r+ cells under control of an actin promoter . Cell lysates were immunoprecipitated using anti-Myc antibody . Immunoblotting with anti-GFP antibody demonstrated that Otk-GFP co-precipitated with Otk-Myc ( Figure 2A ) . Interestingly , both Otk-Myc and Otk-GFP consistently migrated as two bands differing by about 50 kD in size in SDS-PAGE ( Figure 2 ) , pointing to modification of the Otk protein by posttranslational processing , most likely proteolysis . Co-IP experiments were also performed with Myc- and GFP-tagged Otk2 constructs and showed that Otk2-GFP co-immunoprecipitated with Otk2-Myc as well ( Figure 2A ) . Cells transfected with GFP-tagged constructs together with an empty vector were used as negative controls ( Figure 2A ) . Hence , both Otk and Otk2 are able to form homooligomers . The existence of the Otk paralog Otk2 in Drosophila raised the question of whether the two proteins can also interact with each other . To test for heterooligomerization , Myc-tagged Otk2 and GFP-tagged Otk constructs were co-overexpressed in S2r+ cells . Cell lysates were subjected to anti-Myc IP . Immunoblotting with anti-GFP antibody demonstrated that Otk-GFP co-precipitated with Otk2-Myc ( Figure 2B ) . The reciprocal experiment was performed with Myc-tagged Otk and GFP-tagged Otk2 constructs and showed that Otk2-GFP co-immunoprecipitated with Otk-Myc as well ( Figure 2B ) . From these results we conclude that Otk and Otk2 form both homo- and heterooligomers . Our co-IP experiments do not allow us to state whether these oligomers are dimers or higher order multimers . To roughly map the protein domains involved in the oligomerization of Otk and Otk2 , we generated deletion constructs of Otk and tested them in the same co-immunoprecipitation approach as described above . Both a version of Otk lacking the extracellular domain and a C-terminally truncated Otk protein were able to co-immunoprecipitate full length Otk as well as Otk2 ( Figure 2C ) . This result indicates that the interaction between both Otk proteins may be mediated by their transmembrane domains . It is beyond the scope of this paper to proof this hypothesis , but dimerization via their transmembrane domains has recently been demonstrated for several receptor tyrosine kinases [35] . Xenopus PTK7 was demonstrated to interact with Fz7 in Dsh membrane recruitment [23] . To test whether Otk and Otk2 could function as co-receptors with Fz and Fz2 we performed co-immunoprecipitation assays . GFP-tagged Otk or Otk2 were co-overexpressed with Myc-tagged Fz1 or Fz2 in S2r+ cells . Cell lysates were subjected to anti-GFP IP . Immunoblotting with anti-Myc antibody demonstrated that both Otk-GFP and Otk2-GFP robustly co-precipitated with Fz1-Myc and Fz2-Myc ( Figure 2D ) . To determine the expression pattern and subcellular localization of Otk and Otk2 during development , polyclonal antibodies were generated against both proteins . The highly dynamic expression pattern of Otk during Drosophila embryogenesis has been described before [36] and we were able to confirm the published data with our new antibody against Otk . Interestingly , the expression pattern of Otk2 is essentially identical to that of Otk ( Figure 3 ) , indicating that the expression of both genes may be controlled by common regulatory elements . Otk and Otk2 are first detectable in segmentally repeated stripes in embryos at stage 9–10 ( Figure 3A , E ) . Both proteins are also expressed in the developing central nervous system , the visceral mesoderm , the gut and the Malpighian tubules throughout embryogenesis ( Figure 3B–D , F–H ) . No expression was observed in the epidermis and the salivary glands . The expression data obtained by antibody stainings were confirmed by analysis of reporter lines for both otk and otk2 ( Figure S3 ) and by fluorescent RNA in situ hybridization ( FISH; Figure S4 ) . Otk and Otk2 were expressed in the larval brain ( Figure S3H–J ) , in the leg imaginal discs ( Figure S3G ) , in male and female genital discs ( Figure S6 ) and in developing photoreceptor neurons in third instar eye imaginal discs ( Figure S7 ) , but failed to be expressed in the wing imaginal disc ( Figure S7 ) At the subcellular level , both Otk and Otk2 were localized at the plasma membrane in all tissues analyzed ( Figure 3I , J ) . In the gut and in the Malpighian tubules , both proteins were present on the basolateral plasma membrane domain and showed little co-localization with Bazooka ( Baz ) , which is localized at the zonula adherens ( ZA; Figure 3I ) . Compared to the epithelial cells of the gut , expression levels of Otk and Otk2 were much higher in the visceral mesoderm surrounding the gut ( Figure 3I ) . In the central nervous system , both Otk and Otk2 were present on neuronal processes and on the plasma membrane of the neuronal perikarya ( Figure 3J ) . FISH analyses using antisense probes against otk and otk2 showed that in embryos at stage 9 the mRNA of both genes was present in segmentally repeated stripes that were in register with the stripes of Wingless expression in the epidermis ( Figure S4M–R ) . Closer inspection of co-stainings using antibodies against Otk and Wg revealed that Otk is in fact expressed in cells located below the Wg expressing cells in the epidermis , mainly in neuroblasts and their progeny , and in the visceral mesoderm ( Figure S5 ) . In the gut , Otk and Otk2 are expressed in three domains that overlap with the expression of Wg in the proventriculus , the Malpighian tubules and the region that will form the second midgut constriction ( Figure S8 ) [37] , [38] . To determine whether Otk expression itself might be a target of Wnt signaling , we compared Otk expression in embryos homozygous for mutations in different Wnt genes ( Figure 4 ) . Interestingly , Otk does not appear to be regulated by Wg , Wnt4 or Wnt5 , since both the segmental localization in early embryos as well as the localization in the developing gut and nervous system were not affected in the respective mutants ( Figure 4A–D ) . By contrast , overall Otk expression levels were strongly reduced in embryos homozygous mutant for Wnt2 ( Figure 4E , F ) . The reduction of Otk protein levels in Wnt2 homozygous mutant embryos was also detectable by Western blot of embryonic lysates ( Figure 4G ) . To determine whether Wnt2 controls the transcription of otk , we performed semiquantitative RT-PCR analysis . The levels of otk transcripts were unaffected by mutation of Wnt2 ( Figure 4H ) , pointing to posttranscriptional regulation of Otk protein levels by Wnt2 . We also tested whether Wnt2 affects the stability of Otk2 in embryos , but levels of Otk2 were unaffected by loss of Wnt2 ( Figure S9 ) . To test whether the expression pattern of Wnt2 was compatible with its function in regulating embryonic levels of Otk , we analyzed Wnt2 expression by whole mount in situ hybridization . As described before [39] , Wnt2 was expressed in a dynamic , segmentally repeated pattern during embryogenesis ( Figure S10 ) . Taking into account that Wnt proteins can spread from their source by diffusion , this pattern of Wnt2 expression is consistent with its apparent function in regulating levels of Otk . The fact that Otk protein levels are strongly reduced in embryos homozygous mutant for Wnt2 raises the possibility that Otk and Otk2 may be transmembrane receptors for Wnt2 and that Otk protein stability is regulated by a positive feedback loop . To test this possibility , we performed co-IP experiments between both Otk-GFP and Otk2-GFP and Wnt2-Myc in S2r+ cells . Cells were co-transfected with either Otk-GFP and Wnt2-Myc or with Otk2-GFP and Wnt2-Myc . Cell lysates were precipitated with anti GFP antibody and immunoprecipitates were subjected to Western blot with anti Myc antibody . These experiments revealed that both Otk and Otk2 co-immunoprecipitated with Wnt2 ( Figure 5A ) , consistent with a function for both Otk and Otk2 as receptors for Wnt2 . To further corroborate this finding , we used a cell binding assay described before [2] . In brief , S2 cells lacking Fz and Fz2 expression were transfected with either Otk-GFP or Otk2-GFP and were subsequently incubated with conditioned medium from S2 cells transfected with Wnt2-Myc . Cells were washed and fixed without permeabilization and stained with anti Myc antibody . Only cells expressing Otk-GFP ( Figure 5B ) or Otk2-GFP ( Figure 5C ) showed Wnt2-Myc staining on their cell surface , whereas cells without GFP fluorescence failed to stain with the Myc antibody . Control cells transfected with DE-Cadherin-GFP did not show Wnt2-Myc staining on their surface ( Figure 5D ) , demonstrating the specificity of this assay . Together , our results show that both Otk proteins bind to Wnt2 . To further investigate the function of Otk and Otk2 , null alleles were generated for both genes as well as a double knock-out . For this purpose the full coding sequences of otk and of otk2 were removed via FLP/FRT-mediated excision [40] . The peculiar chromosomal localization of both genes in tandem offered an easy way to generate a double knock-out for both genes using the same technique as for the otk and otk2 single knock-out . This method utilizes the ability of FLP recombinase to induce recombination between two FRT sites positioned in trans on two homologous chromosomes . Three suitable transposon insertion lines containing FRT sites were available from the Harvard stock collection . The P ( XP ) d01360 element is located upstream of the 5′UTR of otk and the PBac ( RB ) e03992 element is inserted downstream of the 3′UTR of otk ( Figure 6A ) . The PBac ( PB ) c01790 element is located in the second exon of Mppe , a gene located upstream of otk2 and P ( XP ) d01360 is downstream of the 3′UTR of otk2 ( Figure 6A ) . FLP recombinase-induced deletion of the genomic region between the FRT sites in P ( XP ) d01360 and PBac ( RB ) e03992 was used to remove the coding region of otk . Likewise , the genomic region of otk2 was removed by recombination between the FRT sites located in P ( XP ) d01360 and PBac ( PB ) c01790 . Finally , excision of the genomic region between the FRT sites in PBac ( PB ) c01790 and PBac ( RB ) e03992 deleted the genomic region of both otk and otk2 ( Figure 6A ) . Parts of the Mppe gene upstream of otk2 were also removed during recombination using PBac ( PB ) c01790 . Mppe is not an essential gene and encodes a metallophosphoesterase that functions in Rhodopsin 1 deglycosylation [41] . The following alleles were recovered: otkA1 removes the genomic region of otk and otk2C26 the genomic region of otk2 , while otk , otk2D72 is the double mutant of both genes ( Figure 6A ) . As the three putative deletion lines were all homozygous viable , loss of the respective genes was tested by several methods . Each deletion was verified by PCR on adult genomic DNA ( Figure S11A–C ) and by Western Blots with protein extracts from homozygous mutant embryos ( Figure S11D ) . In addition , whole mount immunofluorescent stainings were performed on homozygous mutant embryos ( Figure 6B , C , E ) . These analyses clearly demonstrated that no Otk protein could be detected in otkA1 embryos , while Otk2 localization was normal ( Figure 6B ) . In agreement with this , no Otk2 protein could be detected in otk2C26 embryos , while Otk expression was not changed ( Figure 6C ) . Besides showing that protein-protein interactions between Otk and Otk2 are not required for mutual stabilization of both proteins , these results also demonstrate that both single gene deletions leave intact the genomic regions responsible for regulation of the second homolog , respectively . Finally , neither Otk nor Otk2 could be detected in otk , otk2D72 embryos ( Figure 6E ) . We conclude that the obtained alleles otkA1 , otk2C26 and otk , otk2D72 are indeed null alleles for the respective genes . Flies homozygous mutant for otkA1 or otkC26 were homozygous viable ( Figure S12 ) and did not show defects in PCP of wings ( Figure S13 ) and eyes ( Figure S14 ) . This strongly disagrees with published data on the previously generated otk3 allele , which was reported to be embryonic lethal [12] , [31] , [32] . We also did not detect any patterning defects of the embryonic cuticle in otkA1 , otkC26 or otk , otk2D72 homozygous mutant embryos , in contrast to an earlier report [12] . Interestingly , also flies homozygous mutant for otk , otk2D72 were viable and did not show any PCP phenotype ( Figures S12 , S13 , S14 ) . Interestingly , we did observe genetic interactions between the otk , otk2D72 double mutant and different alleles of fz . Triple homozygosity for mutations in otk , otk2 and fz caused lethality ( Table 1 ) . This finding is consistent with a functional interaction between the corresponding proteins , as we already demonstrated by co-IP experiments ( Figure 2D ) . otk , otk2D72 homozygous mutant males showed fully penetrant sterility . Apparently , both copies of both genes need to be removed to render the males sterile , since transheterozygous otk , otk2D72/otkA1 or otk , otk2D72/otk2C26 males were fertile ( Table 2 ) . The sterility of the otk , otk2D72 double mutant was rescued by introduction of a full-length UASp-Otk-GFP transgene expressed under control of the ubiquitous driver daughterless::Gal4 , demonstrating the specificity of the male sterile phenotype ( Table 3 ) . Deletion mutants of Otk lacking either the extracellular or the cytoplasmic domain did not rescue sterility ( Table 3 ) , indicating that both domains are required for the function of Otk . To analyze whether sterility is caused by defects in sperm development , testes from males heterozygous and homozygous mutant for otk , otk2D72 were dissected and stained with Vasa , a marker for germline stem cells [42] and Fasciclin III , which marks the hub [43] ( Figure S15 ) . The hub consists of non-dividing stromal cells constituting the stromal niche for the germline stem cells and cyst stem cells [44] . Both markers localize normally in testes homozygous mutant for otk , otk2D72 ( Figure S15B ) , indicating that sterility is not caused by any defects in stem cell regulation . Furthermore , co-staining with the DNA marker DAPI revealed that all stages of sperm development , which can be distinguished by their characteristic packaging of the DNA , are present in testes from males homozygous mutant for otk , otk2D72 ( Figure S15B″ , D′ ) . To further confirm this , testes from males expressing a protamine B-eGFP [45] transgene were analyzed . During Drosophila spermatogenesis , histones are replaced by protamines to achieve sufficient chromatin condensation [45] . Indeed , testes from males homozygous mutant for otk , otk2D72 contain all stages of development ( Figure S15D ) , corresponding to testes from heterozygous control males ( Figure S15C ) . Live observation revealed that mature sperm from homozygous mutant males is motile ( data not shown ) . We conclude that any defects in spermatogenesis or sperm motility are unlikely to account for the observed sterility of otk , otk2D72/otk , otk2D72 adult males . However , after crossing of homozygous mutant otk , otk2D72 males expressing protamine B-eGFP to white− females , no sperm could be detected in the female reproductive tract , in contrast to the control group with heterozygous mutant males ( data not shown ) . This finding strongly indicates that male sterility of otk , otk2D72/otk , otk2D72 animals is caused by a structural or mechanical defect of the male reproductive tract . Wnt2 was shown to be expressed in genital discs and to be involved in the attachment of the testes to the developing seminal vesicle as well as subsequent myoblast migration [33] . Loss of Wnt2 was reported to result in male sterility due to defects in male reproductive tract formation [33] . Otk as well as Otk2 are expressed in both female and male genital discs as determined by antibody stainings and reporter expression ( Figure S6 ) . Because the testes of Wnt2 mutant males were reported to show gaps in their muscle layer [33] , the muscle sheath of the male genital tract of otk , otk2D72 mutant males was analyzed . In contrast to Wnt2 mutant male flies , the entire genital tract of otk , otk2D72 mutant males was surrounded by a contiguous muscle sheath ( Figure S16 ) and the filament organization of the single organs did not differ from that of heterozygous mutant control flies ( Figure S16 ) . It was recently described that the seminal vesicle and the sperm pump contain multinucleated striated muscles , whereas the paragonia and ejaculatory duct are enclosed by mononucleated striated muscle fibers . In contrast , the testes are encircled by smooth muscle fibers [46] . All of these types of muscle fibers could be identified ( Figure S16 ) and no difference between otk , otk2D72 homo- and heterozygous mutant males was observed . However , our analyses revealed that the ejaculatory duct of homozygous mutant males was severely malformed ( Figure 7 ) . Compared to the heterozygous control ( Figure 7A , D , E ) , in which the posterior ejaculatory duct is a long thin tube , the ejaculatory duct of otk , otk2D72 homozygous mutant males was much shorter and the posterior ejaculatory duct was severely thickened ( Figure 7B , C ) . This phenotype was 100% penetrant ( n = 35 ) . The morphology of all the other organs of the reproductive tract was normal ( Figure 7B ) . Consistent with these observations , sperm , which is normally only stored in the seminal vesicle ( Figure 7A ) , accumulated in the ejaculatory duct of otk , otk2D72 homozygous mutant males ( Figure 7B ) , pointing either to an obstruction of the lumen of the duct or to a defect in the transport of sperm through the ejaculatory duct lumen . Closer inspection of the ejaculatory duct revealed that its muscle sheath was strongly disorganized in otk , otk2D72 homozygous mutant males ( Figure 7C ) . We inspected the reproductive tract of Wnt2L/Wnt2O transgeterozygous mutant males in order to check whether their ejaculatory duct was also malformed . This was not the case , whereas we frequently observed missing or incompletely developed testes in the Wnt2L/Wnt2O mutant males ( Figure S17 ) . The penetrance of this phenotype was highly allele-dependent and even in the strongest allelic combination , Wnt2L/Wnt2O , the phenotype was variable with about 10% of the transheterozygous males being fertile ( Table S1 ) . To test whether overexpression of Otk had any effect on development , we ubiquitously overexpressed Otk using the daughterless::GAL4 driver line . Low level overexpression of Otk had no effect on viability and fertility and the corresponding ovaries ( Figure 8B ) were indistinguishable from wild type ovaries ( Figure 8A ) . By contrast , flies strongly overexpressing Otk ( see Figure 8D for quantitation ) were viable , but female sterile . Inspection of the ovaries of these flies revealed that all stages of oogenesis were present ( Figure 8C ) . However , mature eggs were never deposited by the females due to malformation and obstruction of the oviduct ( Figure 8C ) . Together , our data reveal a specific function for Otk and Otk2 in the sexually dimorphic morphogenesis of the reproductive tract in both males and females . Vertebrate PTK7 is by all commonly accepted criteria a bona fide regulator of PCP that functions in the Wnt signaling pathway as a Wnt co-receptor for canonical Wnt3A and Wnt8 in Xenopus [12] . However , it has not been clearly established whether PTK7 promotes signaling by these canonical Wnts or rather functions as an inhibitor of canonical Wnt signaling , which may be essential for activation of the non-canonical PCP branch of Wnt signaling [12] , [22] , [25] , [26] . Here , we investigated whether PCP in Drosophila also requires a homolog of PTK7 and if so , whether it functions in a similar manner as vertebrate PTK7 . Our analysis of the function of PTK7 in Drosophila revealed several unexpected results . We showed that otk is not the single ortholog of PTK7 in the fly genome , as has been previously proposed [12] , [18] , [31] , [32] . In addition to otk we identified otk2 as a second paralog of PTK7 that is most likely the result of a tandem gene duplication that occurred in a common ancestor of all Drosophila species with known genome sequences , but not in other arthropods . We furthermore showed that animals homozygous for the null allele otkA1 are viable and fertile , in contrast to the otk3 allele , which was reported to be embryonic lethal [12] , [31] , [32] . We also did not observe any patterning defects in the embryonic cuticles of homozygous otkA1 null mutant embryos , in contrast to a recently published report using the otk3 allele [12] . None of the previous reports using the otk3 allele showed a rescue of the lethality by a wild type otk transgene , which makes it likely that both the lethality and the reported cuticular phenotypes of otk3 are due to a lethal second site mutation on the same chromosome . In a recent paper from the Tolwinski lab [12] it was proposed that Otk may function as a coreceptor for Wnt4 required for proper cuticular patterning . However , we showed that neither Otk nor Otk2 are expressed in the embryonic epidermis , which strongly argues against this hypothesis . We also did not observe any cuticular patterning defects upon ubiquitous overexpression of Otk , as claimed recently [12] . The cuticular patterning defects upon co-overexpression of Otk and Wnt4 shown in the same paper are in fact very similar to those reported for overexpression of Wnt4 alone [47] , calling into question any functional interaction between Wnt4 and Otk . We did not try to thoroughly re-examine the occurrence of axon guidance defects as reported for the otk3 allele [31] , [32] in the otkA1 null mutant , but until this has been done there remains the possibility that the reported axon guidance phenotypes are also due to the second site lethal mutation on the otk3 mutant chromosome . The existence of otk2 as a second gene closely related to PTK7 offered the possibility that otk2 rather than otk is the functional ortholog of PTK7 in the Drosophila genome or that otk and otk2 function redundantly . However , like for the otkA1 null mutation , flies homozygous for the otk2C26 null allele were viable and fertile . Even the double mutant otk , otk2D72 was homozygous viable but male sterile and did not display any PCP phenotype in wings , legs and eyes . Both otk2C26 and otk , otk2D72 also remove the Mppe locus , which encodes a metallophosphoesterase required for deglycosylation of rhodopsin [41] . However , we can exclude the possibility that Mppe is responsible for the male sterile phenotype , because we can rescue sterility of the otk , otk2D72 double mutant by expression of Otk-GFP , and because a null mutation of Mppe is homozygous viable and fertile [41] . Thus , our data show that otk and otk2 function redundantly and are required for male fertility , but not for PCP signaling in wings , eyes and the adult cuticle . How is a function for both otk/otk2 genes in male fertility compatible with the proposed function of their vertebrate homolog PTK7 in Wnt signaling ? It was shown that flies homozygous for a null mutation in Drosophila Wnt2 are viable but male sterile [33] . We found that Wnt2 can form a complex with Otk and Otk2 upon co-expression in S2 cells . Moreover , expression of Otk in embryos is strongly reduced in Wnt2 mutant animals , but appears to be independent of Wg , Wnt4 and Wnt5 . We showed that Wnt2 stabilizes Otk at the posttranscriptional level , but we currently can only speculate about the mechanism responsible for this effect . In analogy to LRP6 , another important Wnt co-receptor , Wnt2 binding to Otk may induce the phosphorylation of the Otk cytoplasmic tail by a cytoplasmic protein kinase , e . g . Src [16] , thus stabilizing Otk in the membrane [48] . Although the expression patterns of Otk and Otk2 are distinct from that of Wnt2 in the embryo ( see Figures 3 and S10 ) , Wnt2 may reach areas of Otk and Otk2 expression by diffusion . In the case of tracheal development in the embryo , where Wnt2 functions redundantly with Wg , it has been demonstrated that Wnt2 can diffuse and influence cells in which it is not expressed [49] . Interestingly , the function of Wnt2 in tracheal development depends on Fz and Fz2 . This is consistent with our hypothesis that Otk and Otk2 may function as co-receptors for Wnt2 , because we also showed that Otk and Otk2 can form protein complexes with Fz and Fz2 . Wnt2 mutant male flies show defects in the muscle sheath surrounding the testis and lack pigment cells associated with the muscle sheath , which is likely to render the male reproductive tract nonfunctional [33] . Males homozygous mutant for otk , otk2D72 do not show exactly the same phenotype but do show an irregular architecture of the muscle sheath surrounding the ejaculatory duct . We indeed found that defective morphogenesis and obstruction of the ejaculatory duct is responsible for the sterility of otk , otk2D72 males . How exactly otk together with otk2 affects the morphogenesis of the ejaculatory duct is unclear at present and very little is known about the development of this region of the genital tract in wild type flies . However , it appears possible that the failure of the posterior ejaculatory duct to elongate and reduce its diameter may be caused by defects in convergent extension movements of the ejaculatory duct cells in otk , otk2D72 males . If that were true , it would be the first example of a component of Wnt signaling regulating convergent extension movements in Drosophila . Earlier studies showed that the ejaculatory duct develops from the male genital disc [50] and that the muscle sheath surrounding the testis and the ejaculatory duct develops from adepithelial muscle precursor cells attached to the genital disc [33] , [46] . The primary defect presumably causing sterility in Wnt2 mutant males is the defective migration of muscle precursor cells from the genital disc to the testis and the failure to induce pigment cells covering the testis [33] . When we analyzed the morphology of the ejaculatory duct and its muscle sheath in Wnt2 mutant animals we did not detect any obvious abnormalities . Instead , we found that in the strongest allelic combination Wnt2O/Wnt2L the testes were frequently missing or severely reduced in size . However , this phenotype was quite variable and some Wnt2O/Wnt2L males were in fact fertile . Other combinations of Wnt2 alleles were fully fertile , demonstrating that Wnt2 is not essential for male fertility . Despite of these apparent phenotypic differences between the Wnt2 mutant and the otk , otk2D72 double mutant , Wnt2 may nonetheless be one of the ligands for the Otk receptors relevant for assuring male fertility . In mammals , several Wnts are involved in the development of the female reproductive tract , including Wnt4 , Wnt5a , Wnt7a and Wnt9b [51] . Intriguingly , male and female mice mutant for Wnt7a , the closest homolog of Drosophila Wnt2 in mammals , are sterile due to defective morphogenesis of the genital tract . In Wnt7a mutant males the Müllerian duct fails to regress and this leads to a block of sperm passage in the vas deferens , which cannot connect properly at its distal end [34] . In Wnt7a mutant females , morphogenesis of the Müllerian duct derivatives , the oviduct and the uterus , is impaired , which does not allow proper transport and implantation of the ovum [34] . Very interestingly , Wnt7a interacts genetically with Van Gogh-like 2 ( Vangl2 ) in female reproductive tract development [51] , [52] . Vangl2 in turn interacts genetically with PTK7 in vertebrate PCP [18] , making PTK7 an excellent candidate to test for its involvement in reproductive tract development in mammals . Together , our findings reveal a redundant function for the transmembrane receptors Otk and Otk2 in male fertility . Our data furthermore support the hypothesis that Otk and Otk2 are co-receptors for Wnt2 and form complexes with Fz and Fz2 . These findings raise the question whether mammalian PTK7 interacts with Wnt7a and is required for the function of Wnt7a in reproductive tract development . The following stocks were used in this study: P ( XP ) d01360 , PBac ( PB ) c01790 , PBac ( RB ) e03992 ( Exelixis collection , Harvard University , MA ) ; otkCPTI000252 , CG8964SH1639 ( Kyoto stock center ) ; Df ( 3R ) BSC39 ( #7145 ) , Df ( 3R ) BSC199 ( #9626 ) , wgCX4 ( #2980 ) , Wnt4EMSS23 ( #6650 ) , Wnt4C1 ( #6651 ) , Df ( 2L ) DE ( #6653 ) , Wnt2L ( #6909 ) , Wnt2O ( #6958 ) , Wnt2I ( #6960 ) , daughterless-Gal4 ( #5460 ) , y w hs Flp; Sco/CyO ( #1929 ) , PhiC31 86FB ( #23648 ) ( Bloomington Drosophila stock center , Bloomington , IN; stock numbers given in parentheses ) . fzJ22 , fzH51 , fzP21 ( gifts from Paul Adler ) ; fzR52 , fzR52Df ( 3L ) Dfz2 ( gifts from Ken Cadigan ) ; Dfz2C2 , Df ( 3L ) 469-2 ( gifts from G . Struhl ) ; Wnt5400 [53] and protamineB-eGFP [45] were sourced as noted in the references . Null alleles of otk and otk2 were generated by FLP/FRT mediated recombination in trans of the P-element insertions P ( XP ) d01360 , PBac ( PB ) c01790 and PBac ( RB ) e03992 [40] . Transgenic fly lines for the constructs UASp::Otk-GFP29 , UASp::OtkΔC-GFP14 and UASp::OtkΔEx-GFP20 were generated as described in [54] , [55] by injection into an attP landing site at 86FB . The coding regions of full-length or partially deleted versions of otk , otk2 , fz , fz2 , Wnt2 and DE-Cadherin were amplified and the PCR products cloned into pENTR vector using the pENTR Directional TOPO Cloning Kit ( Invitrogen , Carlsbad , CA ) . For expression in S2 and S2r+ cells and for generation of transgenic flies , constructs were recombined into different expression vectors ( pAWG , pAWM , pPWG-attB; Murphy lab , Carnegie Institution of Washington , Baltimore , MD ) using Gateway technology ( Invitrogen ) . For expression as a GST-fusion protein , the extracellular domain of Otk corresponding to aa 159–338 was cloned into pGEX-4T-1 ( GE Healthcare , Piscataway , NJ ) using BamHI and EcoRI restriction sites . To raise polyclonal antibodies against Otk , a GST fusion protein corresponding to aa 159–338 of the extracellular domain was purified and injected into guinea pigs ( Eurogentec , Seraing , Belgium ) . The final bleed of one guinea pig was used for all experiments described in this study . Antibodies against Otk2 were generated by immunizing two rabbits with the peptides VELGRMDSTTSEPQLE ( aa 93–98 , internal fragment ) and ESTILEQESQVADDIV ( aa 418–433 , at C-terminus ) . Final bleeds were pooled and affinity purified against the C-terminal peptide ( Eurogentec , Seraing , Belgium ) . For immunohistochemical stainings , the following primary antibodies were used: guinea pig anti Otk , 1∶1000 ( this study ) ; rabbit anti Otk2 , affinity-purified , 1∶100 ( this study ) ; mouse and rabbit anti GFP , 1∶1000 ( A11120 and A11121 , Invitrogen , Carlsbad , CA ) ; rabbit anti Vasa , 1∶2000 , ( gift from Ruth Lehmann ) ; mouse anti beta-Galactosidase ( JIE7 ) , 1∶20; mouse BP 102 , 1∶50; mouse anti c-myc ( 9E10 ) , 1∶20; mouse anti-Fasciclin III ( 7G10 ) , 1∶20; mouse anti Wg ( 4D4 ) , 1∶20 ( DSHB , University of Iowa , IA ) . Secondary antibodies conjugated to Cy2 , Cy3 ( Jackson ImmunoResearch Europe , Newmarket , UK ) and Alexa Fluor 647 ( Invitrogen , Carlsbad , CA ) were used at 1∶400 dilution . To visualize F-actin , genital tracts were incubated with 2 U rhodamine-conjugated phalloidin in PBS with 0 . 1% Tween . DNA was stained with 4′ , 6-Diamidino-2-Phenylindole ( DAPI; Invitrogen , Carlsbad , CA ) . Genital tracts were fixed with 4% formaldehyde in phosphate-buffered saline ( PBS; pH 7 . 4 ) . Samples were examined using 25×0 , 8 NA Zeiss Plan-Neofluar and 63×1 , 4 NA Zeiss Plan-Apochromat oil immersion objectives on a confocal laser-scanning microscope ( Carl Zeiss LSM 510 Meta ) . Brightfield images were acquired with an AxioImager Z1 upright microscope using 10×0 , 3 NA Zeiss Plan-Neofluar and 25×0 , 8 NA Zeiss Plan-Neofluar oil immersion objectives . S2 cells transfected with GFP tagged Otk , Otk2 or DE-cadherin grown on acid treated coverslips were washed twice in PBS and incubated with 10× concentrated conditioned medium from S2 cells expressing Wnt2-Myc at 4°C for 3 h . After three 10 min washes with cold PBS the cells were fixed in 2% paraformaldehyde for 15 min at room temperature . After three more washes in PBS , fixed cells were incubated with monoclonal mouse anti c-Myc antibody 9E10 at a dilution of 1∶20 overnight at 4°C . After repeated washes in PBS cells were incubated with Cy3-conjugated donkey anti mouse secondary antibody ( Jackson ImmunoResearch Europe , Newmarket , UK ) at 1∶400 , washed in PBS and mounted for microscopy . Protein extraction and Western blots were performed according to standard procedures [56] . For Co-IPs , transiently transfected S2r+ cells were harvested and lysed in 1 ml cold Co-IP lysis buffer ( 50 mM Tris-Cl pH 7 , 5 , 150 mM NaCl , 1% NP40 , 0 . 2% Na-Deoxycholate with protease inhibitors ) by homogenization using a 26 G insulin syringe . Subsequently , the cells were disrupted by sonication with alternating bursts for 10 min . The lysates were then centrifuged and the supernatant was transferred into fresh tubes and pre-cleared with Protein A/G Sepharose beads ( BioCat , Heidelberg ) for 1 h on a rotator at 4°C . After pre-clearing , 20 µl of each sample were kept as an input control . The antibody-antigen reaction took place overnight on a rotator at 4°C . The beads were subsequently washed and 30 µl of 2× SDS buffer [56] were added to each sample followed by incubation at 95°C for 5 min for protein denaturation . The samples were stored at −20°C or used directly for SDS-PAGE and Western blot . Antibodies used for Western Blot were guinea pig anti Otk , 1∶1000 ( this study ) ; rabbit anti Otk2 , affinity-purified , 1∶100 ( this study ) ; rabbit anti GFP , 1∶1000 ( A11121 , Invitrogen ) ; mouse anti c-myc ( 9E10 ) , 1∶200; mouse anti alpha-Tubulin ( 12G10 ) , 1∶500 ( DSHB ) . Band intensities of blots were quantified with Photoshop© ( Adobe , San Jose , CA ) . Total RNA of 50 embryos was extracted using phenol/chloroform ( PeqGold TriFast , Peqlab ) or the Qiagen RNeasy Mini kit . cDNA was prepared using random hexamer primers . Gene expression was analysed by semi-quantitative RT-PCR using the GoTaq kit ( Promega ) and the following primers: Otk ( for 5′-CACCCTAAGCTTTGCCAGC-3′ and rev 5′-CTACATGGTCGGGTAAAGTGG-3′ ) and RpL32 ( for 5′-AAGATGACCATCCGCCCAGC-3′ and rev 5′-GTCGATACCCTTGGGCTTGC-3′ ) . Band intensities were quantified with ImageJ . Phylogenetic analyses were conducted with Mega software version 5 [57] using ClustalW alignment and neighbor-joining method with a gap opening penalty of 60 . Dot Plot analysis was performed using DottupP with word size 6 ( http://mobyle . pasteur . fr/cgi-bin/portal . py#forms::dottup ) . FISH was performed using the Tyramide Signal Amplification ( TSA ) Kit ( Molecular Probes ) according to the protocol provided . For preparation of a Digoxigenin ( DIG ) -labelled RNA probe , 5 µg of Otk-pOT2 ( LP17455; DGRC , Bloomington , IN ) , Otk2-pFLC-I ( RE41180; DGRC ) or Wnt2-pFLC-I ( RE36604; DGRC ) were linearized and purified using the High Pure PCR Product Purification Kit ( Roche ) . In vitro transcription and incorporation of DIG was performed with the DIG RNA labeling kit ( SP6/T7 ) ( Roche ) and the labeled probe was purified using the RNeasy Plus Mini Kit ( Qiagen ) . Wings were removed from adult flies and dehydrated in 100% isopropanol for at least 5 min . Wings were placed on a glass slide and the isopropanol was allowed to evaporate . Wings were mounted with a small drop of Roti Histokitt ( Roth ) for microscopic analysis . Viability was determined by aligning about 100 embryos on apple juice agar plates . The embryos were allowed to develop at 25°C and hatching rates were recorded after at least 24 h . Survival curves were recorded as previously described [58] . Cohorts of about 100 males or females were separated two days after hatching and were transferred to fresh vials . Subsequently , flies were transferred to fresh vials every four days and the number of dead flies was recorded .	Wnts are secreted , growth factor-like proteins that are important for the development of many tissues and organs in animals . They are also required in adult animals and humans for controlling the balance between growth and differentiation . Wnts are bound at the cell surface by Wnt receptors , which are dimers composed of a Frizzled protein and a co-receptor . Here we have analyzed the Drosophila Wnt co-receptors Off-track ( Otk ) and Off-track 2 ( Otk2 ) , which are closely related to vertebrate Protein tyrosine kinase 7 ( PTK7 ) . We found that in contrast to PTK7 in mice and frogs , which controls planar cell polarity ( PCP ) , Otk and Otk2 together are needed in males for development of the ejaculatory duct , a tube-like organ that transports the mature sperm . Our data furthermore indicate that Otk and Otk2 are co-receptors for Wnt2 . The sterile phenotype of Wnt2 mutant males is not identical to that of otk , otk2 double mutants , so additional Wnts may be involved in this process . Interestingly , the function of Wnt2 in male fertility appears to be evolutionarily conserved , because male mice mutant for Wnt7A , the vertebrate homolog of Drosophila Wnt2 , are sterile due to abnormal development of the vas deferens , which corresponds to the fly ejaculatory duct .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "biochemistry", "histology", "developmental", "biology", "medicine", "and", "health", "sciences", "urology", "model", "organisms", "anatomy", "cell", "biology", "evolutionary", "biology", "genetics", "biology", "and", "life", "sciences", "genomics", "molecular", "cell",...	2014	The PTK7-Related Transmembrane Proteins Off-track and Off-track 2 Are Co-receptors for Drosophila Wnt2 Required for Male Fertility
Natively unstructured regions are a common feature of eukaryotic proteomes . Between 30% and 60% of proteins are predicted to contain long stretches of disordered residues , and not only have many of these regions been confirmed experimentally , but they have also been found to be essential for protein function . In this study , we directly address the potential contribution of protein disorder in predicting protein function using standard Gene Ontology ( GO ) categories . Initially we analyse the occurrence of protein disorder in the human proteome and report ontology categories that are enriched in disordered proteins . Pattern analysis of the distributions of disordered regions in human sequences demonstrated that the functions of intrinsically disordered proteins are both length- and position-dependent . These dependencies were then encoded in feature vectors to quantify the contribution of disorder in human protein function prediction using Support Vector Machine classifiers . The prediction accuracies of 26 GO categories relating to signalling and molecular recognition are improved using the disorder features . The most significant improvements were observed for kinase , phosphorylation , growth factor , and helicase categories . Furthermore , we provide predicted GO term assignments using these classifiers for a set of unannotated and orphan human proteins . In this study , the importance of capturing protein disorder information and its value in function prediction is demonstrated . The GO category classifiers generated can be used to provide more reliable predictions and further insights into the behaviour of orphan and unannotated proteins . One of the challenges of the post-genomic era is to predict the function of a protein given its amino acid sequence . Most automated function prediction methods rely upon identifying well-annotated sequence and structural homologues to transfer annotations to uncharacterised proteins ( see [1 , 2] for a comprehensive review ) . Sequence similarity–based methods are relatively successful at annotating homologous proteins; however , they are not applicable to annotating orphan proteins or proteins whose relatives are not themselves functionally annotated . Currently , around 35% of proteins cannot be accurately annotated by homology-based transfer methods [3] , highlighting the need for function prediction methods that are independent of sequence similarity . ProtFun [4 , 5] is an ab initio feature based protein function prediction method that addresses the annotation of orphan proteins and is applicable to any protein whose sequence is known . The method makes use of sequence-based feature descriptors encoded from localisation , secondary structure , and post-translational modification predictions . Function category predictions were made using individual ensembles of neural networks trained to recognise feature patterns associated with particular functions . Similar approaches have been reported using structural properties and sequence information for prediction of enzyme classes [6 , 7] . One advantage of this type of approach is that features that are important in recognition of different function classes can be easily identified and quantified . Over the past few years , there has been a growing awareness of the fundamental importance of disordered proteins in many biological functions and processes . Disordered regions of proteins can be predicted from amino acid sequence [8 , 9] , allowing for rapid surveying of the occurrence of disorder in entire proteomes . The prevalence of disordered proteins in higher eukaryotes is thought to reflect the complexity of signalling and regulatory process within these organisms [10–12] . Disordered regions in proteins are defined as those which lack a stable well-defined 3-D structure in their native states [13 , 14] . Intrinsically disordered proteins may be either entirely disordered or partially disordered , characterised by long stretches of contiguously disordered residues . The presence of protein disorder is thought to confer dynamic flexibility to proteins , allowing transitions between different structural states [15] . This increased flexibility is advantageous to proteins that recognise multiple target molecules with high specificity and low affinity [13 , 15] . The functions of numerous disordered proteins have been characterised experimentally and include DNA and protein recognition , transcription and translation regulation , and targeted protein degradation [16–18] . The disordered regions of these proteins have been shown to be essential for their function , forcing a re-examination of the classical sequence–structure–function paradigm central to the field of structural biology and at the core of most automated function prediction algorithms [19 , 20] . The Protein Trinity hypothesis [19] states that protein function can arise from any of three states: ordered , molten globule , random coil , or from transitions between any or all of these states . Wright [17] defined a continuum of protein structures ranging from an unstructured conformational ensemble to mostly structured proteins containing only locally disordered regions . The functions of disordered proteins along the continuum are influenced by the presence and type of the unstructured regions . For example , disordered stretches can be flexible linker regions that allow movement between domains or can be sites of molecular attachment that become ordered on binding and give rise to functional specificity . In other proteins , disordered regions are associated with sites of post-translational modification that regulate protein–target interactions . It is clear that protein disorder is an important determinant of some protein functions; however , the value of this information remains unquantified and unexploited in current protein function prediction methods . To investigate the correlation of disorder with function , we considered the human complement of disordered proteins as predicted by DISOPRED2 [9 , 21] . Based on pattern analysis between the distributions of protein disorder and different function annotations , an encoding scheme for representing the occurrence of disorder in proteins is proposed . We then assess the direct influence of protein disorder in function prediction using single class Support Vector Machines [22] ( SVMs ) to predict individual Gene Ontology [23] ( GO ) categories . In this analysis , a protein was considered disordered if it contained a contiguous stretch of predicted disordered residues of ≥30 amino acids . GO categories were identified that were over-represented with disordered proteins as a positive control set of categories likely to be associated with protein disorder features . 31 MF categories and 33 BP categories ( Figure 1 ) were significantly enriched in disordered proteins at corrected p-values of <0 . 001 . The cutoff was purposefully stringent to ensure virtually no false positive terms were selected . “Transcription factor” , “DNA and protein binding” , “kinase signaling” , and “phosphorylation” molecular function ( MF ) categories were amongst those enriched in disordered proteins indicated by the highest log ratios of observed/expected occurrence of disordered proteins ( Figure 1A ) . Transcription factor categories were most enriched in disordered proteins , followed by Ion channel and phosphorylation related functions . Metal-ion and nucleotide binding functions exhibited smaller yet significant enrichment in disordered proteins . “Transcription regulation” , “kinase signalling” , “RNA metabolism” , and “phosphorylation” featured in the BP categories ( Figure 1B ) that were enriched in disordered proteins . These categories were consistent with those functions reported both experimentally [24–26] and those reported in similar analyses of other organisms [10] . We examined the distributions of protein disorder within different GO categories to ensure that the disorder features we used captured the trends and patterns relevant for function prediction . We used location descriptors to encode the position of disordered regions in proteins and length-based descriptors to distinguish short from long contiguous stretches of disordered residues . Correlations between location descriptors and GO categories were demonstrated by calculating the average frequency of disordered residues within different location windows for protein sequences annotated by a GO term ( see Methods for more detail , and Figure 2 ) . These averaged values were converted to Z-scores individually for each location window . This procedure normalised for the fact that the false positive rates for prediction of disordered residues are higher at the N and C termini of proteins than in the interior regions [10] . The Z-scores emphasized trends and sampling bias of frequencies of disordered residues directly attributable to the annotation categories . Clustering of annotation categories was performed using Ward's hierarchical method [27] , which minimizes within-cluster variance measured by sums of squares error . The location descriptors showed several trends associated with GO categories . “Transcription regulator” , “DNA binding” , and “RNA pol II Transcription factor” functions were associated with disordered residues in the protein interior , rather than at N and C termini ( Figure 2A ) . “Transcription factor activator” , “Transcription factor repressor” , and “Transcription factor” categories showed significant associations with disordered residues toward the C terminus . Disordered residues were over-represented at the N terminus within the set of Ion Channel and more specifically potassium channel annotated proteins . A further weak association was observed between disorder at the C terminus and the ion channel categories . These observations can be confirmed by crystal structure information . For example , it has been reported that the majority of voltage-gated potassium channel proteins contain intrinsically disordered residues at their N and C terminus [28] . At the N terminus , the residues are responsible for channel inactivation [29] . The disordered residues at the C terminus are adjacent to a PDZ motif mediating binding to scaffold proteins that support the assembly of multiple ion channel subunits into a fully functioning complex [28] . Descriptors for the occurrence of different lengths of disordered regions were also constructed . The link between the length of disordered regions and sequence composition has already been described [30] . To investigate whether this observation also corresponded with functional influences , a similar clustering was performed using descriptors derived from the length distributions of disordered regions within each GO category . The region ranges were selected to reflect the shape of the entire distribution of disordered regions in the human proteome and to avoid sparse descriptors at the upper tail of the distribution ( see Figure S1 ) . Clustering the GO categories by the lengths of their disordered regions ( Figure 3 ) revealed a greater degree of function association ( more significant Z-scores associated with GO categories ) than for the location descriptors . Long regions of more than 500 contiguous disordered residues were over-represented in transcription-related function categories . Shorter regions ( 50 residues or less ) were over-represented in proteins performing metal ion binding , ion channel , and GTPase regulatory functions . Proteins annotated with serine/threonine kinase and phosphatase categories were also over-represented with contiguous stretches of 300–500 disordered residue regions . Again these findings can be supported by structural evidence . Short disordered regions at the mid- to N-terminal regions in small GTPase regulatory proteins mediate a switching mechanism , enabling the protein to interact with multiple binding partners [31 , 32] . We demonstrate that these correlations are not simply a function of correlations between protein length and GO categories by considering “Ion Channel” and “Transcription factor binding” categories ( Figure 3A ) . We observed a statistically significant association between shorter disordered regions and the Ion Channel GO category , yet the average length of protein within this annotation category is more than 900 amino acids . In contrast , for “Transcription factor binding” , the opposite trend is observed . The average protein length for this class is closer to 700 amino acids , and we have reported an association with long ( more than 500 residue ) stretches of disorder . The correlations between function category and disorder region length may be symptomatic of the nature of the structurally disordered region . Tompa [33] described a general set of six functional classes for Intrinsically Unstructured Proteins ( IUPs ) that reflect their capacity to fluctuate freely in conformational space or their ability to partner molecules either permanently or transiently . It may be that the correlations displayed here between disordered region length and GO class represent the degree of structural malleability required by the protein to perform its function . For example , longer disordered regions observed in transcription regulator categories ( Figure 3B ) predominantly act as assemblers that are entirely unstructured and require great flexibility to function . GO categories that contain proteins whose disordered regions are predominantly display sites , for example those that are phosphorylated or involved in ubiquitination ( Ubiquitin cycle in Figure 3B ) , require only shorter disordered regions conferring local flexibility within the protein . The cluster groupings ( Figures 2 and 3 ) were symptomatic of the relationships between annotation terms in the GO graph structure . Specific terms inherit annotations from general parent terms and thus share protein sequences in common . The fact that inherited terms occupied the same or similar clusters provided evidence for the robustness of the observed trends between different annotation categories . Our systematic analysis of disordered regions in the human proteome revealed significant associations between both lengths and locations of disordered regions within proteins and their different GO categories . Many of the observations can be verified by available experimental structure information , highlighting the potential value in using these attributes of disordered proteins as feature descriptors in a method to predict protein function . Including highly correlated features as inputs to machine learning algorithms often results in little increase in performance , and can sometimes result in decreased performance . To investigate relationships between the disorder features and other features to be used in function prediction , a large set of general feature descriptors was assembled ( see Table S1 ) . These were grouped into biological concepts: glycosylation or secondary structure , for example . Redundancy between feature pairs was evaluated using a feature distance matrix ( 1-Pearson correlation ) . To represent the important information in the matrix in fewer dimensions , classical Multi-Dimensional Scaling ( MDS ) was performed . Visualisation of the matrix using the first three dimensions as orthogonal axes ( Figure 4 ) showed three clearly defined groupings . Amino acid composition , phosphorylation , and glycosylation features formed the first group , followed by secondary structure and transmembrane features . Disorder descriptors form a third group less extended from the origin of the plot . The shorter disorder axis reflects the fact that disordered residues are not predicted for all proteins , and , therefore , the information content within these features is comparably less than for amino acid or secondary structure features , which are generic to all proteins . The feature relationships agreed with biological knowledge . For example , sequence features such as hydrophobicity and charge were related to the frequencies of particular amino acids within proteins . The correlations between predicted phosphorylation sites and frequency of Ser , Thr , and Tyr residues ( Pearson correlation ∼0 . 2 ) were due to the fact that high frequencies of phosphorylated residues can only be observed when the relevant amino acid types occurred with a high frequency in the protein . Similarly , the frequencies of predicted O and N glycosylation sites displayed correlations with the occurrence of Asn and Ser/Thr residues . The features most closely related to disorder were random coils , PEST , and low-complexity descriptors with correlation values of 0 . 472 , 0 . 211 , and 0 . 307 , respectively , at the residue frequency level . These correlations , although relatively weak , indicated that some of the information within the disorder features is also encoded by these related feature descriptors . Disordered regions in proteins frequently contain residues that are also recognised as low sequence complexity [34]; however , a region of low complexity does not always imply structural disorder . For example , fibrous proteins such as collagens and silks are rigidly structured in their native state yet contain repetitive regions of low complexity [16] . PEST motifs are degradation motifs present in proteins involved in protein phosphorylation , protein–protein interactions , and cell adhesion [35] . These motifs have been shown to be enriched in an experimentally characterised database of disordered proteins [36] , and the residues that characterise the motifs represent a subset of those amino acids known to be disorder-promoting [18 , 37] . However , the correlations observed here between predicted occurrences of these features were small . The general spatial isolation of disorder descriptors in feature space suggested that they contain unique biological information not represented by the other features previously used in function prediction . Feature importance estimates for all features were collated across all GO categories using a leave-one-out elimination strategy . The histogram columns ( Figure 5 ) represent the average percentage loss in classifier accuracy for all GO categories belonging to MF and BP ontologies , regardless of their individual category performance . Secondary structure features contributed the most to classifier performance for the majority of MF and BP categories . Disorder features were the second most important feature for BP category recognition . Amino acid composition and secondary structure contributions were higher on average for MF categories than for BPs . For all other features , the importance estimates were higher for BP categories . Our results suggest that disorder patterns are more indicative of the biological process than the molecular activity of the protein . This is striking considering that only one-third of the proteins in the human proteome are predicted to contain significant disordered regions and the information content of the disorder feature set is comparably lower than that for secondary structure or amino acid composition . One possible reason for this observed difference lies in the respective ontology definitions . BP categories describe modules of functions that make up parts of a multi-step process [23] , whereas MFs describe a protein's biochemical activity . For example , the receptor tyrosine kinase signalling BP category annotation describes the series of molecular signals generated as a consequence of a transmembrane receptor tyrosine kinase binding to its physiological ligand . Three example proteins annotated by this term are neurterin precursor a neurotrophic growth factor , Rap guanine nucleotide exchange factor , and erb-B2 receptor tyrosine-protein kinase . These proteins are all unrelated at the primary amino acid sequence and secondary structure level , yet each sequence is predicted to contain at least one 30–50 disordered residue stretch ( exemplified in Figure 3B ) . The role of disordered regions in molecular recognition and in hub proteins in protein–protein interaction networks is well-defined [38–40] . Biologically , it would make sense that proteins that are part of the same multi-step process are more likely to co-localise and possess a common interaction surface such as a disordered region without sharing any similar sequence composition or secondary structure . To evaluate the contribution of disorder features in classification accuracy for individual categories , the performance loss was measured when disorder features were removed from each classifier using the Matthews Correlation Coefficient ( MCC ) . This measure represents the additional value of disorder features in function prediction , accounting for both interaction and compensatory effects between features . Classifier performances were reported for 26 GO categories ( Table 1 ) whose sensitivity at a false positive rate of 10% exceeded 50% . The significance of the improvements in correlation coefficients for individual categories were evaluated using Fisher's Z test , which considers both the magnitude of the performance increase and the strength of the correlation . The improvements that were significant at the 5% level ( p < 0 . 05 ) were marked in bold ( Table 1 , column MCC+diso ) . Eleven BP categories and 12 MF categories that were identified as enriched in disordered proteins ( Figure 1 ) showed improvements resulting from the addition of disorder features . Several additional GO classes were identified during feature selection that required disorder features for optimal performance . Seven categories: “UDP-glycosyl transferase” , “hormone” , “growth factors” , “transferase” , “hydrolase” , and “carboxylic acid transporters” were added to the MF set of categories , and “G protein signaling” was added to the BP category set of classifiers . The most notable performance gains were observed for “protein tyrosine kinase signaling , ” “G protein signaling” , “ubiquitin specific protease” , “transcription” , “protein kinase” , and “helicase” categories . For some categories ( “cation-channel” , “ion channel” , “metal ion transport” , “purine-nucleotide binding” , “nucleotide binding” , and “DNA binding” ) , little or no performance increase resulted from the addition of disorder features . Particularly for Ion channel , Metal Ion transport , and Nucleotide binding categories , other features such as transmembrane regions or secondary structure better characterised the relationship between the primary amino acid sequence of the protein and its function . The MCC diso–only values ( Table 1 ) showed the correlation observed when classifiers were trained with only disorder features . Some of the BP categories relating to transcription and the Transcription factor MF category could be recognised with sensitivities of >50% at false positive rates of less than 10% , yielding Matthews correlations of ≥0 . 3 . For these categories , the increased performance resulting from the addition of disorder features ( difference between MCC+diso and MCC–diso columns in Table 1 ) was much lower than the correlation obtained from disorder features alone . This result can be explained by the representation of mutual information between random coil , low complexity , or PEST features reducing the magnitude of the effect of the disorder features . Conversely , for “G protein signaling” and “Receptor tyrosine kinase” BP categories and “Growth factor” , “Helicase” , “Hydrolase” , and “Ubiquitin specific protease” MF categories , the improvement resulting from the addition of disorder features was greater than the correlation obtained using disorder features alone . This finding indicates that disorder features interacted cooperatively with other features in the dataset to achieve a greater performance increase . Throughout this study , classification performance for GO categories has been reported using the MCC . This measure accounts for the imbalanced class frequencies encountered in the GO term classifiers . For completeness , the classification sensitivities obtained at 10% , 5% , and 1% false positives were reported ( Table S2 and Figure 6 ) . The number of positive class labels is also included to stress that different error rates are required for comparable performance between these classifiers . This fact is exemplified by the Receiver Operating Characteristic ( ROC ) curves ( Figure 6 and Table S3 ) which vary according to class size . The curves have been zoomed in to show the sensitivities at false positive rates of below 50% . The majority of reported classifiers were capable of achieving more than 50% sensitivity at false positive rates of less than 10% . Some categories were not recognised as enriched in disordered proteins using statistical tests due to small class frequencies and low occurrences of proteins containing disordered residues . This finding highlights the advantage in using a machine learning–based approach to assess patterns of disordered features over a simple statistical approach using frequency of occurrence in recognising GO categories for which disorder is an important determinant . In contrast to the finding that disorder features contributed more to BP category recognition , the improvements for MF and BP categories in Table 1 were slightly greater for MF than BP categories . However , these data reflect a subset of the categories for which we were able to produce accurate classifiers . This result highlights the fact that overall more BP categories utilised information from disorder features for classification than MF categories , resulting in a higher feature importance estimate overall . However , for most of these categories , we were not able to produce sufficiently accurate classifiers to be of practical use . Our method differed from the original ProtFun method [4] in several important ways . Firstly , our predictions for structure , disorder , and transmembrane regions utilised PSI-BLAST profiles rather than single sequence predictions as feature inputs . Encoding information from sequence profiles in this manner increased the accuracy of feature predictions for those proteins that belonged to unannotated families . Second , additional secondary structure features were encoded that recorded the frequencies of helices and sheets of particular length ranges within each protein . Despite these differences , we felt it was important to provide a benchmark comparison between our method and an independent method that did not utilise disorder information . To assess the performance of the ProtFun method , the ProtFun server GO category assignments used the 14 , 055 annotated proteins used in this study . Classifier accuracy was reported for eight common categories ( Figure 7 and Table S4 ) . The results indicated that our method outperformed the ProtFun server for all tested categories assessed using the MCC . All of these improvements were significant at the 95% level using Fisher's Z test for significance of correlation difference , except for the ion channel category . The performance of our method without disorder features ( Table 1 ) was also reported so that the improvements in accuracy could be attributed to the use of disorder features or to the use of different training datasets and machine learning algorithms . Four of the compared function categories; “Ion Channel” , “Voltage gated ion channel” , “Cation channel” , and “Metal ion transport” did not utilise information from disorder features; therefore , improvements resulted from other methodological differences . For the remaining categories , “transcription” , “regulation of transcription” , “hormone” , and “growth factor” , the source of performance improvements were a mixture of these effects and the addition of disorder features . The greatest accuracy increase resulting directly from the addition of disorder features was observed for the “growth factor” category . For the “hormone” category , the increased accuracy resulted equally from the addition of disorder features and the algorithm and encoding differences . “Transcription” and “Regulation of transcription” accuracies were improved more by the feature encoding and more recent training datasets used than the addition of disorder features . This result was not surprising considering that the ProtFun features included low complexity , PEST regions , and random coils that overlap considerably with disorder features within these categories . In this benchmark study , it was difficult to provide an unbiased performance measure that was comparable between the two methods . For ProtFun we were restricted to using the server output alone rather than individual neural network output scores , and any testing dataset was likely to have been used at least partially in the training of this method . However , these results indicate that our method represents a significant improvement in predicting protein function from sequence . The molecular recognition process and function classifiers reported have been used to classify a dataset of unannotated and orphan IPI proteins . A majority rule approach was applied to the annotations such that three of the five classifiers for each GO term must report a positive term assignment . At a confidence cutoff of 0 . 6 ( see Figure S2 for confidence distributions ) , we were able to assign putative functions to 317 proteins . The majority of high confidence predictions ( >0 . 9 ) were made by “transcription” and “DNA binding” MF classifiers ( Table 2 ) . Additionally , the hierarchical nature of the relationships between the GO classes can be exploited to distinguish more confident predictions . For example , many of the proteins predicted to be “regulators of transcription” also receive independent positive assignments from parent terms “transcription” and “regulation of cellular process” . The annotations have been made publicly available at http://bioinf . cs . ucl . ac . uk/anno/IPI . html . The aim of this study was to investigate the contribution of protein disorder features in protein function prediction . This work extended numerous survey studies that report the occurrence of protein disorder within entire proteomes by identifying relevant trends and patterns of disordered regions that can be used to predict the function of proteins . Additionally , we have extended and enhanced the repertoire of GO categories that can be recognised in prediction methods by incorporating disorder features . Disorder features contributed greater overall improvements in recognition of BP categories than MF categories . In fact , the disorder features were the second most informative feature set in BP category recognition whilst amino acid composition features were the least informative . The differences in feature importance were attributed to the differences in the descriptive nature of the two Ontologies . The anticorrelation observed between the importance of disorder features and amino acid composition for BP categories suggested that associations between disordered region length and location and BP category were not a function of similar amino acid compositions of proteins within BP categories . This finding is particularly relevant for methods that attempt to predict function or possibly protein interactions from amino acid sequence without the use of homologous sequence relationships . The performance of 26 GO category classifiers could be improved using disorder features . Using the disorder features alone , sensitivities above 50% at false positive rates of less than 10% were obtained for some transcription-related BP categories . The results for all other categories were significantly better than random using disorder features as the sole input . These findings were impressive considering that in this study disordered residues were predicted rather than experimentally confirmed . Consequently , the estimates of feature importance were conservative and restricted by the accuracy of the disorder prediction algorithm . DISOPRED2 currently predicts 57% of residues correctly at a false positive rate of 5% . Additionally , whilst structural and compositional subtypes of disordered region have been suggested in the literature [33 , 41] , such classifications have not yet been exploited in a method that predicts disorder from sequence . The potential value of encoding subtypes of disordered region in our function prediction method is indicated by the fact that in most cases the mutual information contained within PEST and low-complexity features was important for recognition of many of our reported GO categories . Finally , we have demonstrated the practical application of our classifiers in predicting function for orphan and unannotated human proteins . The classifiers are applicable to any protein sequence and are well-suited to predicting putative molecular recognition functions that can then be assayed in vivo for activity , or for the purpose of target prioritisation . For the better performing classifiers , such as DNA binding and transcription related categories , identification of function from sequence can be performed . Overall , our findings reflect the importance of capturing protein disorder information and demonstrate the value of disorder features in human protein function prediction . We used the International Protein Index ( IPI ) [42] as a comprehensive human protein dataset and the Gene Ontology Annotation ( GOA ) [43] for human . 28 , 057 proteins were annotated with one or more GO categories . The Cd-Hit [44 , 45] algorithm was used with a threshold of 60% identity to reduce overall sequence redundancy . The remaining 14 , 055 sequences were partitioned into five equally sized groups for cross-validation and testing . For rigorous cross-validation , the partitioning algorithm ensured that those sequences with significant homology relationships , defined as having a BLAST E-value ≤ 1e-6 , were allocated to either the same training set or the same test dataset but never both . This resulted in five equally sized training and testing sets for each GO term where the maximum sequence identity between pairs of training and testing proteins did not exceed 40% sequence identity or a BLAST E-value of 10−6 . Positive and negative training sets for each GO term with at least 50 representative proteins were generated . Positive training examples included those proteins annotated with a particular GO term or any of its child terms in the GO hierarchy . Negative training examples included those proteins not annotated with the particular GO term or any of its children . To avoid potential class labelling errors , proteins annotated with any of the parent or less specific terms in the GO hierarchy were subsequently removed from the negative training sets . These proteins represent incomplete annotations with respect to the GO category under consideration and may belong to either the positive or negative training set for the given term . Fisher's exact test was performed under the null hypothesis that the occurrence of the GO term annotation and presence of disorder in a protein were independent . The hypothesis was rejected at p-values of <0 . 001 after applying Bonferroni multiple testing correction . The calculations were performed using the R package for statistical computing [46] . The degree of over-representation for each GO category was compared using the log odds ratio of observed over expected numbers of disordered proteins . The expected number of disordered proteins represents the background frequency , or occurrence , of disordered proteins by random chance within a sample size equivalent to the size of the GO category . This calculation yields a scale whereby values of zero indicate equality between observed and expected numbers of disordered proteins and higher values indicate a larger difference between the observed and expected values . For a particular GO term , the set of proteins annotated by the term or any of its child terms was considered . For location-based measures , each protein was split into ten segments; N terminus , equally proportioned segments 1 through 8 , and C terminus . The frequency of disordered residues within each segment of each protein was calculated . Disordered residues were defined as those residues predicted to be disordered by DISOPRED2 at a threshold of 5% per residue false discovery rate . The set of frequencies of disordered residues within each location window for proteins annotated by each GO term was then averaged . This resulted in a set of ten average frequencies , one for each location region within each GO category . The average frequencies were Z-score normalised independently within each location window to account for the fact that the false positive rate for prediction of disordered residues is greater at the N and C termini than in the protein interior [10] . A similar approach was adopted to assess correlations between disordered region length in proteins and GO term annotations . Disordered regions in proteins were defined as contiguous stretches of ≥30 residues predicted to be disordered . The average frequency of regions that corresponded to each length range across all proteins annotated by the GO term was then calculated and converted to an independent Z-score for each length range . The Support Vector Machine [47] ( SVM ) is an efficient classification algorithm suitable for solving binary classification problems in high-dimensional spaces . The algorithm separates positive from negative class data by positioning a linear hyper-plane though the class examples . Often , the input data is not linearly separable , and a kernel function is required to map the data into a higher dimensional space to find the optimal separating hyperplane . The SVM was chosen over other machine learning methods of choice due to its capacity and ability to control error without causing overfitting to the data . The SVMlight [48] SVM package was used to train binary classifiers for individual BP and MF GO terms using the radial basis function kernel . Kernel parameters C and γ were selected by exhaustive grid searches performed on a 272 processor Linux cluster that maximised the MCC for each classifier . MCC was chosen as a more informative measure of classifier performance than percent accuracy or error as it avoids bias resulting from unbalanced class frequencies . For example , each of the five testing sets for “GO:0045449 regulation of transcription” comprised 356 positive class examples and 1 , 726 negative class examples . A classification accuracy of more than 82% can be obtained by setting all predicted outcomes to be negative , whereas the MCC balances and controls for the bias in class frequencies . The MCC is similar to the Pearson correlation coefficient where 0 represents random classification and 1 implies perfect classification . Feature selection was carried out using a recursive elimination strategy . Initially each classifier was trained and tested using all feature inputs . Optimisation of C and γ kernel parameters was performed at this stage . A single feature set was iteratively removed from the input data and the performance measured in terms of MCC . Feature attributes that did not contribute to classification performance or indeed caused improvements to performance when removed were permanently eliminated from the input data . When no further improvements were observed , a second round of parameter optimisation was performed on the final feature sets to produce final classification performance statistics . The results from feature elimination can be found in Table S2 . The features were divided into global ( single values per protein ) and spatial ( multiple descriptors describing feature location within the protein ) . Global features comprised amino acid composition , sequence features , signal peptides ( SignalP 3 . 0 [49] ) , and localisation information ( psortII [50] ) . The sequence features described general protein characteristics calculated directly from the protein sequence such as molecular weight , average hydrophobicity , iso-electric point , charge , and atom counts . Local features Disorder , PEST [51] ( motifs rich in proline , glutamate , serine , and threonine ) , coiled coils , and low-complexity residues were predicted using DISOPRED2 [52] , epestfind , coils [53] , and pfilt [54] algorithms with default parameter settings . Transmembrane and secondary structure content was predicted using Memsat3 [55] and PSI-Pred [56] algorithms . Post-translational modification features phosphorylation and glycosylation were predicted by NetPhos3 . 0 , Net-N-Glyc , and Net-O-Glyc software [57] . A detailed list of descriptors for these features can be found in Table S1 . All feature descriptors were scaled to between 0 and 1 before use in classification . Frequency-based descriptors such as the number of transmembrane regions were log-transformed prior to scaling . DISOPRED2 was used to predict disordered residues for the representative protein sequence set using three iterations of PSI-BLAST [58] against the UNIPROT database release 6 . 0 . Residues were predicted as being disordered at a false positive rate of 5% . Residue predictions were post-filtered for the presence of transmembrane regions predicted using MEMSAT 3 . 0 [55] set to default parameters . Predicted disordered regions were further filtered for stretches of at least 30 contiguous residues . A dataset comprising 2 , 157 orphan and unannotated IPI human proteins was compiled . These proteins contained one or more predicted disordered regions and represent a mixture of proteins that are either members of unannotated families or have no detectable sequence homologues by BLAST similarity searches . To calibrate comparable prediction accuracies between classifiers , the SVM outputs ( distances from the separating hyperplane ) were converted to posterior probabilities [59] . The probabilities were estimated from the testing datasets so that they reflect the performance of the classifiers on unannotated proteins . The predictions for the unnanotated disordered proteins have been made publicly available at http://bioinf . cs . ucl . ac . uk/anno/IPI . html .	As a result of high throughput sequencing technologies , there is a growing need to provide fast and accurate computational tools to predict the function of proteins from amino acid sequence . Most methods that attempt to do this rely on transferring function annotations between closely related proteins; however , a large proportion of unannotated proteins are orphans and do not share sufficient similarity to other proteins to be annotated in this way . Methods that target the annotation of these difficult proteins are feature-based methods and utilise relationships between the physical characteristics of proteins and function to make predictions . One important characteristic of proteins that remains unexploited in these feature-based methods is native structural disorder . Disordered regions of proteins are thought to adopt little or no regular structure and have been experimentally linked with the correct functioning of many proteins . Additionally , disordered regions of proteins can be successfully predicted from amino acid sequence . To address the requirement for protein function prediction methods that target the annotation of orphan proteins and explore the use of information describing protein disorder , a machine learning method for predicting protein function from sequence has been implemented . The inclusion of disorder features significantly improves prediction accuracies for many function categories relating to molecular recognition . The practical utility of the method is also demonstrated by providing annotations for a set of orphan and unannotated human proteins .	[ "Abstract", "Introduction", "Results/Discussion", "Materials", "and", "Methods" ]	[ "biochemistry", "none", "molecular", "biology", "computational", "biology", "homo", "(human)" ]	2007	Inferring Function Using Patterns of Native Disorder in Proteins
Forkhead box p3 ( Foxp3 ) -expressing regulatory T cells are key mediators of peripheral tolerance suppressing undesirable immune responses . Ectopic expression of Foxp3 confers regulatory T cell phenotype to conventional T cells , lending itself to therapeutic use in the prevention of autoimmunity and transplant rejection . Here , we show that adoptive transfer of polyclonal , wild-type T cells transduced with an inducible form of Foxp3 ( iFoxp3 ) can be used to suppress immune responses on demand . In contrast to Foxp3-transduced cells , iFoxp3-transduced cells home “correctly” into secondary lymphoid organs , where they expand and participate in immune responses . Upon induction of iFoxp3 , the cells assume regulatory T cell phenotype and start to suppress the response they initially partook in without causing systemic immunosuppression . We used this approach to suppress collagen-induced arthritis , in which conventional Foxp3-transduced cells failed to show any effect . This provides us with a generally applicable strategy to specifically halt immune responses on demand without prior knowledge of the antigens involved . Transplant rejection and autoimmune diseases ranging from rheumatoid arthritis , type I diabetes , and multiple sclerosis to inflammatory bowel disease—as diverse as they might appear—all have the same underlying problem: the launch of an undesirable immune response [1] . Equally similar are the current approaches to treat these conditions , which are generally based on drugs that lead to systemic immunosuppression [2] . Thus , the induction of specific tolerance is seen as the “Holy Grail” of therapeutic approaches [3] . The discovery that the immune system evolved regulatory T ( TR ) cells to stop undesirable immune responses , such as autoimmunity [4] and the rejection of the fetus [5–7] , is of obvious therapeutic promise [8] . Indeed , TR cells have already been shown to be capable of fulfilling such functions [9] . However , the translation of experimental findings into actual therapeutic approaches is hampered by a variety of problems . Under experimental conditions , antigen-specific tolerance can be achieved by using TR cells from TCR-transgenic animals or by ex vivo expansion of antigen-specific TR cells [9–11] . However , it is difficult to imagine how a TCR transgenic approach can be translated into a generally applicable therapy . The antigen-specific ex vivo expansion of TR cells [9–11] , or in vivo conversion of helper T cells ( TH ) into TR cells [12] , is more feasible , albeit still problematic . They not only rely on the knowledge of , or at least access to the antigens involved in the pathological immune response , but are also time consuming and complicated when applied in a therapeutic context [8 , 13] . There are also conceptual problems . The lack or malfunction of TR cells is suspected to be at the root of many autoimmune diseases [14 , 15] . In these cases , it might be impossible to obtain and expand functional , antigen-specific TR cells , as they may not exist in the host in the first place . In principle , this problem can be circumvented by the conversion of conventional T cells into TR cells , either by TGF-ß–mediated induction [16–18] or ectopic expression of the lineage factor Forkhead box p3 ( Foxp3 ) ( NP_473380 ) [19–21] . However , without enriching antigen-specific “induced TR cells , ” this is likely to be of limited benefit and may lead to systemic immune-suppression [11 , 22–24] . A further problem with TGF-ß–induced TR cells is that their phenotype seems to be unstable [25 , 26] , although the presence of retinoic acid appears to stabilize the conversion [27 , 28] . Here , we present a strategy to suppress undesirable immune responses in an antigen-specific fashion without prior knowledge of the antigens involved . We accomplish this by adoptive transfer of a small number of polyclonal TH cells transduced with a genetically engineered , inducible form of Foxp3 ( iFoxp3 ) . CD4+CD25− cells transduced with iFoxp3 ( TH::iFoxp3 ) initially retain their “proinflammatory” phenotype . They home “correctly” into the secondary lymphoid organs and partake in immune responses . Once the TH::iFoxp3 cells have expanded in an antigen-specific fashion , they can be converted to TR cell phenotype on demand by inducing iFoxp3 , thereby stopping the immune response they partook in . Encouraged by the initial finding that polyclonal CD4+CD25− T cells transduced with Foxp3 ( TH::Foxp3 ) can prevent and treat colitis in lymphopenic animals [19 , 29] , we , like others [23 , 30 , 31] , set out to test whether this can be used as a general strategy to prevent and treat autoimmune diseases . To test this hypothesis , we used collagen-induced arthritis ( CIA ) , which is a well-established murine model of human rheumatoid arthritis [32] . To obtain TH::Foxp3 cells , we transduced CD4+CD25− T cells with a murine leukemia virus ( MLV ) -based retroviral vector carrying a Foxp3-IRES-GFP cassette ( m6pg[Foxp3] ) ( Figure S1 ) . We immunized male DBA/1 mice with chicken collagen type II ( cII ) in complete Freund's adjuvant ( CFA ) . In this model , we observe the first clinical symptoms of arthritis on day 19 after immunization , with the average clinical score reaching a plateau around day 35 . Injection of 1 × 106 TH::Foxp3 cells 1 d prior to immunization did not have any significant impact on the outcome of the arthritis . It neither delayed the time of disease onset , nor did it alter disease progression ( Figure 1A ) . The failure of polyclonal TH::Foxp3 cells to show any beneficial effect on the outcome of CIA under these experimental conditions is in agreement with the findings of others [31] and led us to reassess the approach per se . Therefore , we decided to examine the homing , expansion , and participation of TH::Foxp3 cells in immune responses . The decision whether to launch or suppress an immune response is made within the secondary lymphoid organs [33] . This makes “correct” homing of the adoptively transferred cells an essential requirement for cytotherapy , as otherwise their participation in immune responses might be severely limited . We therefore compared the homing of TH::Foxp3 cells to that of m6pg[control]-transduced CD4+CD25− T ( TH::control ) cells ( Figure S1 ) and freshly isolated CFSE-labeled CD4+CD25− ( TH ) cells or CD4+CD25+ ( TR ) cells . A total of 1 × 106 cells were injected into wild-type Balb/c mice . After 48 h , we isolated the lymphocytes from the various tissues and analyzed them by flow cytometry . The transferred cells were identified based on either their green fluorescent protein ( GFP ) coexpression or CFSE label . TH::control cells , like TR and TH cells , could be detected at comparable frequencies in blood , and inguinal and iliac lymph nodes , as well as the spleen ( Figure 1B and 1C ) . In contrast , the homing of TH::Foxp3 cells into the lymph nodes appeared to be defective and their homing into the spleen slightly impaired . Instead , a large number of these cells could be found in the liver ( Figure 1C ) . The data suggest that ectopic expression of Foxp3 substantially altered the homing behavior of the transduced cells . The absence of T cells from the peripheral lymph nodes is one of the key features of CD62L-deficient ( sell−/− ) mice [34] . CD62L ( l-selectin ) plays a key role in the homing of lymphocytes into these tissues by allowing their attachment to high endothelial venules [35] . Activation of T cells leads to endoproteolytic shedding of CD62L from the surface of the cells , involving the matrix-metalloprotease Adam17 [36] . Therefore , we investigated whether the altered homing behavior of TH::Foxp3 cells is due to Foxp3-mediated effects on the surface expression of CD62L . We found that the majority of freshly isolated TH and TR cells are CD62Lhi ( Figure 2A and 2B ) . Activation of the cells for 72 h with anti-CD3/anti-CD28/IL-2 led to a down-regulation of CD62L surface expression , which was more marked in TR than TH cells ( Figures 2C and S2A ) . To assess whether this is due to an increase in Adam17 activity in TR cells , we activated freshly isolated splenocytes with PMA and compared the surface expression of CD62L on Foxp3+ ( TR ) and Foxp3− ( TH ) CD4+ T cells . The rate of CD62L shedding appeared to be very similar for both cell types and could be completely blocked by the Adam17 inhibitor TAPI-2 ( Figure 2D ) . This suggests that an additional Adam17-independent mechanism in TR cells is responsible for the difference in CD62L surface expression observed upon activation of TR and TH cells . To further investigate this , we examined CD62L expression in TH::Foxp3 cells . We transduced CD4+CD25− cells with either m6p8[Foxp3] or m6p8[control] . The cells carrying the vector were identified based on their coexpression of ratCD8α ( Figure S1 ) . Whereas TH::control cells exhibited some down-regulation of surface CD62L upon activation with anti-CD3/IL-2 , this was substantially more marked in TH::Foxp3 cells ( Figure 2E and 2F ) . For the first 24 h , TAPI-2 appeared to partially inhibit the loss of surface CD62L on TH::Foxp3 cells , but it did not halt the steady decrease in surface CD62L over an extended period of time ( Figure 2G ) . The CD62L down-regulation in TH::control cells was accompanied by an accumulation of soluble CD62L in the culture supernatant . This was not the case for TH::Foxp3 cells ( Figure 2H ) , suggesting that in these cells , CD62L surface expression is regulated by a mechanism other than shedding . As Foxp3 is known to be a transcriptional regulator [37–40] , we investigated whether it affects CD62L transcription . The CD62L mRNA expression level was reduced in both TH::Foxp3 and TH::control cells compared to freshly isolated TH and TR cells ( Figure 2I ) . However , the level of CD62L transcript was 7 . 2-fold lower in TH::Foxp3 cells than in TH::control cells . The data suggest that upon activation of the cells , CD62L is further down-regulated on a transcriptional level by Foxp3 . It is noteworthy that retroviral transduction requires at least some degree of activation of the cell to allow for transgene integration . In this context , the expression of Foxp3 led to a very marked and sustained down-regulation of surface CD62L expression . This is likely to be a major contributor to the altered homing behavior of TH::Foxp3 cells . Although the down-regulation of CD62L upon activation is similarly more evident in thymically derived TR cells than TH cells ( Figure S2A and S2B ) , albeit less marked than in TH::Foxp3 ( Figure 2I ) , it does not appear to interfere with the cells ability to home into peripheral lymph nodes ( Figure S2C ) . The “incorrect” homing of polyclonal TH::Foxp3 cells might well contribute to their lack of showing any beneficial effect in CIA [31] ( Figure 1A ) and other animal models of autoimmune disease [11] . However , one might question whether our initial approach had any merit in the first place , since the transfer of polyclonal TH::Foxp3 cells will only marginally increase the number of suppressive cells that recognize a particular antigen . Indeed , treatment with polyclonal TH::Foxp3 cells more or less mimics polyclonal TR cell therapy , which in contrast to approaches using antigen-specific TR cells , appears to be of limited benefit [22–24 , 41] . We decided to develop an alternative strategy , allowing us to convert the lineage commitment of conventional TH cells to that of TR cells after their antigen-specific expansion in vivo . To achieve this , we created an inducible Foxp3 ( iFoxp3 ) that is constitutively expressed , but only becomes functionally active upon induction . Polyclonal , primary TH cells transduced with iFoxp3 ( TH::iFoxp3 cells ) should act like conventional T cells , retain their homing behavior , participate in immune responses , and expand in an antigen-specific fashion . This antigen-specific in vivo expansion of TH::iFoxp3 cells should allow us to specifically switch off immune responses on demand by inducing iFoxp3 . We fused a modified estrogen receptor ( ERT2 ) to the C-terminal end of Foxp3 and cloned it into the m6p vector ( Figure 3A and 3B ) . ERT2 only responds to tamoxifen and its metabolites such as 4-hydroxytamoxifen ( 4-OHT ) , but not estrogen [42] . In the absence of induction , iFoxp3 is retained in the cytoplasm and kept inactive by heat shock proteins binding to the ERT2 part of the fusion protein [43] . To confirm the inducible nature of iFoxp3 , we transduced CD4+CD25− cells with m6p carrying a GFP-tagged iFoxp3 ( m6p8[GFP-iFoxp3] ) . This allowed us to assess the induction of iFoxp3 based on the translocation of the fusion protein from the cytoplasm into the nucleus . We induced iFoxp3 in vitro by exposure to 4-OHT for 48 h ( Figure 3C ) or in vivo after adoptive transfer of the transduced cells into wild-type Balb/c mice by intraperitoneal ( i . p . ) injections of tamoxifen ( Figure 3D ) . In either case , iFoxp3 translocated into the nucleus in about 60%–70% of the transduced cells at the time of microscopic analysis , confirming its inducible nature . A key requirement for our strategy is that iFoxp3 can be used to induce TR cell phenotype on demand . We therefore tested TH::iFoxp3 cells for hallmark features of TR cells such as sustained up-regulation of CD25 , in vitro anergy to anti-CD3-stimulation , and suppression of target cells [4] before and after induction of iFoxp3 . Whereas TH::Foxp3 cells were anergic ( Figure 3E ) , suppressed the proliferation of cocultured CD4+CD25− cells ( Figure 3F ) , and exhibited up-regulation of CD25 ( Figure 3G ) , TH::iFoxp3 cells did so only after induction of iFoxp3 with 4-OHT . This demonstrates that , at least in vitro , TH::iFoxp3 cells appear to behave like conventional TH cells and only assume the phenotype of TR cells upon the induction of iFoxp3 . From our observations with TH::control cells , we already knew that transduction per se did not appear to alter the homing behavior of the cells ( Figure 1B and 1C ) . Nevertheless , we wanted to verify that noninduced iFoxp3 neither changes the expression of CD62L nor significantly alters the homing behavior of the TH::iFoxp3 cells . We found that in the absence of iFoxp3 induction , CD62L expression remained unchanged in TH::iFoxp3 compared to TH::control cells ( Figure 3H and 3I ) . This is in stark contrast to our observations made for TH::Foxp3 cells ( Figure 2E to 2I ) . To assess the homing behavior of the cells , we used the same approach as described above . We found that the homing behavior of TH::iFoxp3 cells was comparable to that of TH::control cells ( Figure 3J ) and thus very similar to that of naive TH and TR cells ( Figure 1B ) . To assess whether TH::Foxp3 and TH::iFoxp3 cells expand upon antigenic challenge in vivo , we transferred transduced cells prepared from DO11 . 10xSCID/Balb/c mice that expressed an ovalbumin ( ova ) -specific TCR , into wild-type Balb/c mice . We transferred 5 × 104 cells containing a mixture of 2 × 104 TH::iFoxp3 cells and 3 × 104 nontransduced cells ( transduction efficiency of 40% ) with the transduced population being clearly identifiable based on the coexpression of GFP . TH::iFoxp3 cells expanded upon immunization with ova in CFA by a factor of 12 in the draining lymph nodes and by a factor of 37 . 5 in the spleen ( Figure 4A ) . In contrast , TH::Foxp3 cells only exhibited a very modest expansion , by a factor of 3 . 6 in the lymph nodes and 4 . 4 in the spleen . This could have been due to the TH::Foxp3 cells suppressing the ova-specific immune response and thereby impeding their own expansion . However , the levels of ova-specific antibodies in the serum were the same , independent of whether the mice had received TH::Foxp3 or TH::iFoxp3 cells , suggesting this was not the case ( Figure 4B ) . Our data demonstrates a clear expansion of TH::iFoxp3 cells , which is consistent with their participation in the immune response against ova . This in vivo expansion upon antigen exposure is considerably less marked in TH::Foxp3 cells . Next , we investigated whether the in vivo expanded ova-specific TH::iFoxp3 cells can be induced to suppress the very same immune response they partook in . We isolated splenocytes from these mice and exposed them to ova ex vivo . Although , in the absence of induction of iFoxp3 , we observed the expected antigen-induced recall proliferation , we could not detect any proliferation above background in the presence of 4-OHT ( Figure 4C ) . This suggests that upon iFoxp3 induction , the expanded TH::iFoxp3 cells became anergic and suppressed the proliferation of the cotransferred , nontransduced DO11 . 10 T cells as well as any endogenous ova-specific T cells . To assess to what degree polyclonal TH::iFoxp3 cells participate in an immune response , we transferred 1 × 106 wild-type TH::iFoxp3 cells into wild-type Balb/c mice . A week after immunization with ova , we analyzed the lymphocytes from various tissues by flow cytometry . Whereas the number of TH::iFoxp3 cells recovered from the blood , iliac lymph nodes , liver , and spleen did not appear to change upon antigenic challenge , we observed a marked increase in the inguinal lymph nodes of the immunized mice ( Figure 4D ) . This indicates that some of the TH::iFoxp3 cells expanded in the draining lymph nodes ( subcutaneous [s . c . ] immunization into the flanks ) . However , the number of endogenous cells in the inguinal lymph nodes increased equally ( Figure 4E ) , suggesting that both populations expand to a similar degree with their ratio remaining constant . To test the potential of TH::iFoxp3 cells in suppressing autoimmune responses , we turned to the CIA model in which TH::Foxp3 cells had failed to show an effect ( Figure 1A ) . We transferred 1–2 × 106 polyclonal TH::iFoxp3 cells into wild-type DBA/1 mice 1 d prior to immunization with cII in CFA . We induced iFoxp3 on day 15 after immunization , which lies between the peak of the T cell response to collagen around day 10 [44 , 45] and the onset of clinical symptoms around day 21 [46] . Mice that had received TH::iFoxp3 cells but did not receive tamoxifen injections to induce iFoxp3 showed the first signs of arthritis on day 19 , similar to the mice that received no transfer of cells ( Figure 5A ) . This effect was specific to the antigenic challenge ( cII in CFA ) inducing the autoimmune response , as mice receiving these cells without immunization did not exhibit any overt signs of developing autoimmune disease ( Figure S3 ) . Remarkably , 23 out of 25 of the mice that had received TH::iFoxp3 cells and tamoxifen injections to induce iFoxp3 did not show any clinical signs of arthritis ( scores <3; Figure 5B ) . This is in stark contrast to the other groups , in which the majority of animals developed arthritis ( scores ≥3; Figure 5B ) . Whereas tamoxifen has been reported to have anti-inflammatory properties [47] , we found that by itself , it had only a minor effect on the development of CIA ( Figure 5A ) and no effect on the activity of TH::control cells in vivo ( Figure S4 ) . Despite the clear suppression of the clinical signs of CIA , we could detect collagen-specific antibodies in the serum of the animals at day 52 , irrespective of the treatment they had received ( Figure S5 ) . Next , we investigated whether TH::iFoxp3 cells are capable of stopping already established CIA . To this end , we waited until the mice had reached a clinical score of 3 before inducing iFoxp3 . The induction appeared to completely halt , if not reverse , CIA , leading to a decline in the average severity score ( Figure 5C ) . None of the mice showed a further increase of symptoms after induction of iFoxp3 ( Figure 5D ) . To assess whether the conversion of TH::iFoxp3 cells to TR cell phenotype causes systemic immunosuppression , we compared “ex vivo recall reactions” to the antigen used prior to the induction of iFoxp3 ( cII ) to that of an unrelated antigen ( ova ) injected after induction . The collagen-specific T cell proliferation measured for mice in which iFoxp3 had been induced was significantly lower than that of mice that had received no transfer of cells , albeit still higher than that of naive mice ( Figure 6A ) . As we did not add tamoxifen to the ex vivo culture , this most likely reflects a lower number of cII-specific proinflammatory T cells in the animals that had received TH::iFoxp3 cells and tamoxifen induction , rather than a mere ex vivo suppressive effect of TH::iFoxp3 cells . Remarkably , we could not detect any difference in the T cell proliferation upon exposure to ova irrespective of whether the mice had received treatment or not ( Figure 6B ) . This suggests that the suppression only affects immune responses in which the TH::iFoxp3 cells have had the opportunity to participate prior to induction of iFoxp3 . Indeed , we were able to detect TH::iFoxp3 cells in the inflamed paw of cII-immunized mice , suggesting that in the absence of induction these cells can contribute to the inflammation ( Figure S6 ) However , once converted , the TH::iFoxp3 cells , despite still being present ( Figure S7A and S7B ) , seem to have lost the capacity to suppress further unrelated immunological challenges ( Figure 6B ) . This suggests that the conversion of TH::iFoxp3 cells by induction of iFoxp3 does not lead to a systemic immunosuppression . Having shown that induced TH::iFoxp3 cells do not suppress further unrelated immune responses postinduction , we wanted to investigate the suppressive activity of TH::iFoxp3 cells in a context in which both cII and ova are present prior to induction . We transferred 1 × 106 polyclonal TH::iFoxp3 cells into wild-type DBA/1 mice 1 d before immunization with a 1:1 mixture of ova and cII in CFA . We induced iFoxp3 on day 15 after immunization and assessed the antigen-induced proliferation of splenocytes prepared from these mice on day 28 . The recall proliferation against ova and cII were comparable . Equally similar was the reduction in proliferation in the cases in which iFoxp3 was induced ( Figure 6C ) . In combination , these results suggest that this approach enables selective suppression without affecting further unrelated immune responses after induction of iFoxp3 . It is noteworthy that we were able to detect TH::iFoxp3 cells 52 d after their transfer , independent of the level of arthritis and whether the mice had received tamoxifen treatment or not ( Figure 7A and 7B ) . An analysis of various tissues revealed that TH::iFoxp3 cells in blood were only marginally reduced between day 17 and day 52 ( Figure 7C and 7D ) , and could readily be detected in the auxiliary lymph nodes and spleen . Although this is likely to be of advantage with regard to actively suppressing immune responses , it poses the question whether continuous tamoxifen presence is required . Due to the long half-life of tamoxifen [48] , a direct assessment of this in vivo is not feasible . However , in in vitro suppression assays , TH::iFoxp3 cells had completely lost their suppressive activity 72 h after withdrawal of 4-OHT ( Figure 7E ) . To perform these experiments , we had to compensate for a marked reduction in the number of viable TH::iFoxp3 cells that could be recovered under these conditions . To formally address the effect of the withdrawal of 4-OHT on TH::iFoxp3 cell viability , we exposed the cells to 4-OHT for 48 h from the point of transduction and then cultured them for a further 48 h in the absence of 4-OHT . The number of viable cells was assessed by flow cytometry . Withdrawal of 4-OHT had no effect on TH::control cells , but led to a marked decrease in the number of TH::iFoxp3 cells ( Figure 7F to 7H ) . This suggests , that once induced , TH::iFoxp3 cells die upon tamoxifen withdrawal , but it remains unclear how this translates into an in vivo context . Indeed , it might be desirable to incorporate a suicide gene [49] into the retroviral vector as this allows the removal of the transduced cells if desired ( Figure S8 ) . Here , we have presented an approach that allows us to stop undesirable immune responses without prior knowledge of the antigens involved . TH::iFoxp3 cells participate in immune responses as conventional TH cells until iFoxp3 is induced . At this point , they change their phenotype from that of proinflammatory T cells to that of regulatory T cells and suppress the response they partook in . Ectopic expression of Foxp3 in conventional T cells leads to their conversion into cells with TR-like phenotype [19–21] . It was demonstrated early on that these TH::Foxp3 cells , like TR cells , could suppress the development of colitis in lymphopenic hosts [19 , 29] . However , it was noted that in this context , the effectiveness of both polyclonal TH::Foxp3 cells and TR cells [29 , 50 , 51] might be due to the regulation of homeostatic expansion of the cotransferred , proinflammatory cells , rather than to a true antigen-specific suppression [9 , 11 , 52] . Furthermore , adoptive transfer of polyclonal TR cells will only marginally increase the number of suppressive cells that recognize a particular antigen . Indeed , the use of polyclonal TR cell [22] or TH::Foxp3 populations [11 , 23] ( Figure 1A ) have been of limited efficacy , unless the immune pathology was caused by an absence of functional TR cells [20 , 53] or the experiments were performed in lymphopenic animals [11] . The restrictions imposed by the low frequency of antigen-specific TR or TH::Foxp3 cells in polyclonal populations can be circumvented by ex vivo expansion of antigen-specific TR cells and TCR transgenic TH::Foxp3 cells [9–11 , 41] . Both approaches have been successfully exploited in mouse models to treat diabetes [23 , 24 , 54 , 55] , arthritis [31] , and experimental autoimmune encephalomyelitis ( EAE ) [56] , as well as being used for the induction of transplantation tolerance [57 , 58] . Whereas TCR transgenic T cells are an invaluable research tool to improve our understanding of the regulation of immune responses [59 , 60] , it is unclear to what degree they can be used in a therapeutic context . Ex vivo expansion of antigen-specific TR cells [9 , 11] , or in vivo conversion of TH into TR cells [12] , promises to be more applicable . However , these approaches are technically challenging , time consuming , and most importantly , require knowledge of or access to the antigens involved in the immune response to be suppressed [8 , 13] . Our study of TH::Foxp3 cells revealed a further problem . Whereas TH::Foxp3 cells appear to adopt the characteristics of TR cells in vitro , we found their homing to be altered from that of endogenous TR and TH cells . This hinders the TH::Foxp3 cells from mimicking the homing behavior of endogenous TR cells , which has been shown to be important for their suppressive function in vivo [61–63] . Those TH::Foxp3 cells that fail to home to the secondary lymphoid organs might not receive the required antigen priming [63] and thus fail to expand like endogenous TR cells [64] . This might explain the difference in the efficacy of approaches that use polyclonal Foxp3+ cells and those that use antigen-selected or TCR transgenic Foxp3+ cells . The latter might circumvent the need for an antigen-specific expansion in vivo by ensuring that there are sufficient numbers of antigen-specific cells from the onset . The activation-induced , Foxp3-mediated down-regulation of CD62L might well be a key factor in the exclusion of TH::Foxp3 cells from the peripheral lymph nodes since T cells from CD62L-deficient mice exhibit a similar phenotype [34 , 35] . Further , it has been shown that CD62Lhi polyclonal TR cells have a more potent protective effect in vivo [65] . However , we cannot exclude that ectopic expression of Foxp3 also alters the expression of other homing receptors . Indeed , we found that the activation-induced down-regulation of CD62L in thymically derived TR and TH cells was not sufficient to exclude them from the peripheral lymph nodes . Here , we present an approach that addresses these problems by transducing polyclonal , conventional T cells with a retroviral vector encoding a genetically engineered inducible form of Foxp3 . TH::iFoxp3 cells retain their proinflammatory character and the ability to home to the lymph nodes . Those TH::iFoxp3 cells that recognize an antigen appear to participate in the immune response and expand . This in vivo expansion of antigen-specific TH::iFoxp3 cells circumvents the need for an ex vivo expansion and does not rely on any knowledge of the antigens involved . Upon induction of iFoxp3 , the in vivo expanded , antigen-specific TH::iFoxp3 cells assume a TR cell-like phenotype and suppress the undesirable response they initially partook in . We were able to demonstrate the efficacy of our approach by specifically halting CIA in a mouse model . Importantly , TH::iFoxp3 cell-mediated suppression appears to be restricted to the specific response , which is ongoing at the time of induction of iFoxp3 . Those TH::iFoxp3 cells that do not already participate in an immune response at the time of induction lose the capacity to suppress further unrelated immune responses despite still being present . Although we cannot exclude that other factors play a role , it appears that the antigen-specific expansion of the TH::iFoxp3 cells prior to induction is an integral part of the observed nonsystemic suppression . In a therapeutic context , it might be desirable to limit the exposure to tamoxifen to minimize possible side effects . Although it appears that most TH::iFoxp3 cells die upon withdrawal of tamoxifen , those that do survive lose their suppressive activity . To avoid possible deleterious effects , these “revertant” cells can be removed based on the incorporation of a suicide gene into the retroviral vector used for the delivery of iFoxp3 . We believe that this strategy of induced conversion of TH cells into cells with a TR cell-like phenotype using iFoxp3 is generally applicable and will allow us to stop a variety of undesirable immune responses . Balb/c and DBA/1 mice ( 8–12 wk old ) were purchased from Charles River and Harlan . DO11 . 10xSCID mice on the Balb/c background were kindly provided by Caetano Reis e Sousa , CRUK . Animals were maintained under specific pathogen-free conditions . Expert animal technicians provided animal care in compliance with the relevant laws and institutional guidelines . Cells used for in vivo and ex vivo experiments were purified ( >90% purity ) using an AutoMACS ( Miltenyi Biotec ) as previously described [66] . Flow cytometric analysis and proliferation assays were performed as described previously [66] using the following antibodies: ratCD8α ( BD Bioscience ) , CD62L ( BD Bioscience ) , CD4 ( BD Bioscience ) , CD25 ( BD Bioscience ) , and Foxp3 ( eBioscience ) . Foxp3 was amplified from total spleen cDNA and iFoxp3 was constructed by a C-terminal fusion of ERT2 in place of the stop codon . Both were cloned into m6p retroviral vectors coexpressing either GFP or a GPI-linked rat CD8α marker . For the measurement of in vivo translocation of iFoxp3 , GFP was cloned in-frame with Foxp3 after the first five codons in the 5′-end [67] in order to produce GFP-iFoxp3 . For the production of retroviral supernatant , 293eT cells were cotransfected with an equal amount of pCl-Eco packaging plasmid and the respective m6p retroviral construct . Supernatant was harvested at 36 h and 48 h after transfection , filtered , and then used immediately . For retroviral transduction , the freshly purified CD4+CD25− T cells were activated in the presence of plate-bound anti-CD3ε ( 0 . 6 μg/ml ) ( BD Bioscience ) and 10 U/ml of recombinant mIL-2 ( PeproTech ) . Cells were transduced at 24 h and 36 h after activation by resuspension in a 1:2 mixture of supernatant and complete medium ( RPMI/10% FCS/10 μM β-mercaptoethanol/50 μg/ml gentamicin ) supplemented with 10 U mIL-2 and 6 μg/ml protamine sulphate ( Sigma ) and 10 U/ml mIL2 , followed by centrifugation at 600×g for 2 h at 32 °C . Six hours after transduction , cells were resuspended in complete medium containing 10 U mIL-2 . A fixed ratio of transduced ( 50%–60% in all cases ) and nontransduced cells was adoptively transferred into mice 72 h after the last transduction . Male DBA/1 mice received 1–2 × 106 transduced cells intravenously ( i . v . ) ( day −1 ) and were immunized intradermally ( i . d . ) with 100 μl of cII ( Sigma , ) dissolved in 10 mM acetic acid and emulsified ( 1 μg/μl ) in CFA ( DIFCO ) the following day ( day 0 ) [46] . The mice were assessed ( blinded ) on a daily basis , and inflammation of the paws was scored as follows: grade 0—no swelling; grade 1—swelling in an individual joint; grade 2—swelling in more than one joint or mild inflammation of the paw; and grade 3—severe swelling of the entire paw and/or ankylosis . Each paw was graded , and all scores where totaled for a maximum score of 12 per mouse . Mice reaching a score of 8 or more were euthanized in accordance with restrictions imposed by UK legislation . For iFoxp3 induction , the mice were injected i . p . with 100 μl of tamoxifen ( in a 10:1 mixture of sunflower oil:ethanol [10 μg/μl tamoxifen] ) on days 15 and 16 , and ( in a 10:1 mixture of sunflower oil:ethanol [1 μg/μl tamoxifen] ) on days 23 , 29 , 30 , 36 , and 43 . Alternatively , iFoxp3 was induced once the mice had reached a score of 3 ( day 0 ) by i . p . injections with 100 μl of tamoxifen ( in a 10:1 mixture of sunflower oil:ethanol [10 μg/μl tamoxifen] ) on days 1 , 2 , 9 , and 16 . CD4+CD25− T cells were purified from 6–12-wk-old female SCIDxDO11 . 10 mice and transduced with Foxp3 or iFoxp3 as described above . Balb/c females received i . v . 5 × 104 of a 2:3 ratio of transduced and nontransduced cells . Three days later , each mouse was immunized s . c . with either ova ( Sigma ) in CFA ( 50 μg/mouse ) or just with CFA . The mice were sacrificed and analyzed 8 d after immunization . For ova-specific suppression assays , total splenocytes were prepared as described [66] , resuspended in complete medium , and then plated into round-bottom 96-well plates ( density of 2 × 105 cells/well ) . iFoxp3 was induced by adding 50 nM 4-OHT ( Sigma ) . Ova was added to the cells 16 h after induction . After 60 h , the cells were pulsed with 1 μCi of 3H-thymidine ( Amersham ) , collected at 72 h with a Filtermate Harvester ( Packard ) , and analyzed with a TopCount scintillation counter ( Packard ) according to the manufacturer's instructions . CIA and iFoxp3 induction was performed as described above . On day 28 , some of the mice received ova in CFA s . c . into both flanks ( 100 μg/mouse ) . Total splenocytes were prepared on day 35 and plated into round-bottom 96-well plates at a density of 5 × 105 cells/well . Proliferation of the cells was measured 72 h after addition of either ova ( 100 μg/ml ) or cII ( 100 μg/ml ) as described above . Alternatively , mice were immunized simultaneously with ova and cII on day 0 by i . d . injection of a mixture of 100 μg of ova and 100 μg of cII in CFA . Recall reactions were performed on day 28 as described above at a density of 2 × 105 cells/well . Ninety-six–well flat-bottom plates ( Nunc ) were coated with either ova ( 50 μg/ml ) or cII ( 2 μg/ml ) at 4 °C for 16 h and blocked with 1% BSA in PBS for 1 h . A total of 50 μl of serial dilutions ( starting at 1:50 for ova and 1:10 , 000 for cII ) of mouse sera in PBS were incubated for 2 h . Biotin-conjugated IgG1 , IgG2a , IgG2b , and IgG3 ( BD Bioscience ) were then applied for 2 h . For ova detection , IgM ( BD Bioscience ) was also included . The development of cII and ova-specific immunoglobulins was then measured using a DuoSet kit ( R&D Systems ) according to the manufacturer's instructions . Total RNA was extracted using an RNeasy kit ( Qiagen ) including DNaseI treatment ( Invitrogen ) . cDNA was synthesized with Superscript II ( Invitrogen ) with random hexamer primers ( Amersham ) following the manufacturers instructions . Real-time PCR was performed using Taqman SYBR green PCR master mix ( Applied Biosystems ) with primers specific for Sell ( CD62L ) and Hprt . The sequences used were: Sell primers: 5′-ATG CAG TCC ATG GTA CCC AAC TCA-3′ and 5′-CTG CAG AAA CAC AGT GTG GAG CAT-3′; Hprt primers: 5′-TTA AGC AGT ACA GCC CCA AAA TG-3′ and 5′-CAA ACT TGT CTG GAA TTT CAA ATC C-3′ . An ABI Prism 7900 sequence detection system ( Applied Biosystems ) was used for 45 cycles of PCR according to the manufacturer's instructions .	Autoimmune diseases come in many diverse forms—such as rheumatoid arthritis , type I diabetes , multiple sclerosis , and inflammatory bowel disease—yet all share the same underlying cause , the launch of a detrimental immune response . In healthy individuals , a specialized immune cell type called regulatory T cells prevents these undesirable immune responses . Here , we present a strategy to suppress undesirable immune responses using genetically modified proinflammatory T cells that participate in these inappropriate immune responses until they are activated with a drug . At this point , the genetic modification causes them to change their behavior to that of regulatory T cells . Using a mouse model , we demonstrate that this approach can be used to stop undesirable immune responses on demand with minimal intervention .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "immunology", "rheumatology" ]	2008	Specific Immunosuppression with Inducible Foxp3-Transduced Polyclonal T cells
It was shown recently using experimental data that it is possible under certain conditions to determine whether a person with known genotypes at a number of markers was part of a sample from which only allele frequencies are known . Using population genetic and statistical theory , we show that the power of such identification is , approximately , proportional to the number of independent SNPs divided by the size of the sample from which the allele frequencies are available . We quantify the limits of identification and propose likelihood and regression analysis methods for the analysis of data . We show that these methods have similar statistical properties and have more desirable properties , in terms of type-I error rate and statistical power , than test statistics suggested in the literature . Homer et al . [1] showed that it was possible in some circumstances to identify whether a person with observed genotypes at multiple loci was part of a sample from which only estimated allele frequencies were known . Such identification would be particularly useful in forensic science if the presence or absence of a person's DNA in a mixture of DNA could be established . The authors also discussed the relevance of their findings when summary statistics such as allele frequencies were available in the public domain as part of genotype-phenotype studies , because it possibly could be established that individuals , or their close relatives , were part of a particular study . As a result of the publication of Homer et al . , NIH and the Wellcome Trust added more restrictions to the access of such data to avoid potential identifiability ( http://grants . nih . gov/grants/gwas/data_sharing_policy_modifications_20080828 . pdf ) . The approach taken by Homer et al . was to have two samples with estimated allele frequencies , here called the “test” and “reference” sample , and to ask whether an individual was ‘close to’ either of these samples , using a statistic that measured a distance to the sample . The properties of the test statistic were not investigated theoretically ( although simulation studies were performed ) , and the difference between “sample” and “population” was not always clear . In this note we take a best-case idealised setting in which there is a single population from which there is a test sample with allele frequencies at a number of loci and from which there is a single individual , called the proband , with full genotypes . The question is whether the person was part of this test sample from which allele frequencies are available . We use both likelihood and linear regression theory , which illustrate different approaches to the problem , to draw inference about the hypothesis that a proband was part of the test sample . We show that the power of identification of a proband as part of a test sample is , approximately , proportional to the number of independent SNPs divided by the size of the sample from which the allele frequencies are available . The power is reduced by a predictable magnitude if the frequencies in the population are themselves estimated imprecisely . Properties of likelihood-ratios and regression test statistics and a comparison with the statistic used by Homer et al . were verified by simulation . There are m independent SNP markers with a population frequency of pi for allele B at the ith SNP . We assume Hardy-Weinberg equilibrium in the population , so that the genotype proportions for the ith SNP are ( 1−pi ) 2 , 2pi ( 1−pi ) and pi2 for genotypes AA , AB and BB , respectively . We have estimated allele frequencies based upon a test sample of N unrelated individuals . In the test sample of 2N alleles , ni is the number of B alleles at locus i . In this study we assume that N is known and individuals are equally represented in computing . Note that these conditions are unlikely to be fully met in forensic applications when the test sample may be a DNA pool and we consider the implications later . The genotype for proband X at the ith SNP is gi , which can take values of 0 , 1 and 2 for genotypes AA , AB and BB , and the expectation of yi = ½gi is the population frequency pi , i . e . E[½gi] = pi . To simplify derivations , we shall first assume the population frequencies pi , are known . More generally , we assume we have prior unbiased estimates of the allele frequencies from the same population from a different finite sample ( the “reference sample” ) of size N* , in which there are ni B alleles at locus i . As both the test and reference samples are drawn independently from the population , the best estimate of the frequency in the population is given by the pooled value , It is explained subsequently why this estimate , rather than say ni/2N* , the estimate of the allele frequency from the reference sample , is used in the statistical analysis . We show that the main results for the regression approach are based upon the expectation that the regression of the proband frequency , yi = ½gi , on , each expressed as deviations from population frequencies , is distributed about unity for all loci if the proband was part of the test sample , and about zero otherwise . Population allele frequencies on m markers were drawn from a uniform distribution with lower bound 0 . 05 and upper bound 0 . 95 ( i . e . , minor allele frequency ( MAF ) >0 . 05 ) . For the ith SNP , a genotype score ( yi ) of a proband was simulated from a binomial distribution with probability pi and sample size 2 . Allele frequencies in the reference and test samples were simulated from a binomial distribution with probability pi and sample size 2N* and 2N , respectively . If the proband was part of the test sample then the test sample was simulated on N−1 individuals and the allele count from the proband was added to that from this sample to create a sample from N individuals . Linear regression was performed as described previously , for a type-I error rate of 0 . 05 , and the Homer et al . [1] test statistic ( see Text S2 ) was also implemented . 1000 simulations were performed for combinations of N = 100 , 1000 , 10000 , N* = 100 , 1000 , 10000 and ∞ and m = 50 , 000 , when the proband was either part or not part of the test sample . The results are shown in Table 1 . The regression type-I error rates are well controlled when the hypotheses tested are true . As predicted ( Text S2 ) , the type-I error rates for the Homer et al . test statistic are not well controlled . In many cases the probability of rejecting the null hypothesis when it is true is close to zero . Power to determine whether the proband is part of the test sample is good for test samples of 1000 if the reference sample size is large . Inference from the regression and likelihood-ratio approach is similar , as expected ( Table S1 ) . In our derivations we have assumed that all samples ( proband , reference and test ) are from the same population and that within the population there is random mating . What if these assumptions are violated ? If all samples are from the same population but there is deviation from HWE then the tests are somewhat biased because HWE is assumed in computing the likelihood and the variance of sample allele frequencies . Population differences are more serious and can lead to the wrong inference . There are a large number of possibilities because , in principle , the proband , reference and test samples can all come from different populations . However , population differences between the reference and test sample can be tested explicitly using standard tests for differences in gene frequency . There seems little point in testing whether a proband was part of a specific test sample when there is no reference sample from the same population . Nevertheless , what can we predict if the reference population is not actually from the same population , but is used as if it is ? Then both the likelihood statistics for the hypothesis ‘in’ and ‘out’ are inflated , by essentially the same amount , so the problem is not the divergence between the two populations , but bias in the test statistic . If population frequencies are inappropriately or approximately estimated , the sample is more likely to be assigned as ‘in’ when it should not be . The reference sample is of little value if the divergence between the populations , expressed as Wright's FST , approaches 1/ ( 2N ) . Can we quantify the limits of identification in practical situations ? This is hard , because there are ( at least ) three difficulties in addition to the theoretical sample m/N criterion: For these reasons we cannot set a simple limit to identification without reference to other parameters ( or speculation ) . In the analysis we have not considered the possibility that the proband is not in the test sample , but is related to one or more persons who is . For example if a relative with relationship R ( e . g . R = ½ for full sibs ) is in the test sample , then the expectation of the regression coefficient is E ( b ) = R rather than 0 or 1 . Similar calculations can be done if , for example , there are several relatives in the test or reference samples . If many markers are used , a value of b of approximately one-half would raise suspicions that in fact a full sib , parent or child is in the test sample . Lower , but non-zero values could be consequences of sampling or relationship . The simulation results in Table 1 illustrate how sensitive the methods can be , and hence there seems a real possibility of identifying not just the proband but also his/her relatives . A problem frequently met in forensic applications is whether a particular individual's DNA appears in a mixture obtained at a crime scene , for example . In this case , it is usually unknown how many individuals' DNA is present in the sample ( i . e . , N is unknown ) , equal representation cannot be assumed , and there may be allelic drop out in the sample , although Homer et al . [1] showed empirically that probands could be detected even if their contribution to the DNA pool was small . We do not therefore consider the present results to be relevant for probabilistic inference in a forensic setting . However , exclusion of a proband from a pooled DNA sample is possible if many markers are used , the actual N is small and frequencies of alleles from the pool are estimated accurately . The likelihood framework is sensitive to genotyping errors in that false exclusions could occur , but the analysis could be adapted to model genotype counts with specified probability of errors or by assuming replacement sampling in computing P ( in ) . The linear regression approach is likely to be robust to genotyping error . In contrast to forensic applications , in the situation considered by Homer et al . in which the test sample is a database constructed using a specified number of individuals each with individual genotypes , and with the gene frequencies estimated as their average , our results support their conclusions . Probands that were part of a test sample could be identified even for samples sizes of 1000 . If , for example , there are both diseased case and healthy control samples in the association test , each assumed to be sampled from the same population , then it is possible to test whether an individual is present in either the case or control group using the analysis we have described , but using each sample in turn as the test sample . Current genome-wide association studies ( and meta-analyses based upon multiple studies ) are conducted on large samples , often of the order of 10 , 000 or so , and in this case our results show that the power to identify a proband who was part of such a large sample when the reference sample is of similar size is only about one-half ( Table 1 ) assuming 50 , 000 independent loci , even under the ideal circumstances considered in this study .	It was shown recently by Homer and colleagues that it may be possible to determine whether a person with known genotypes at a number of markers was part of a pool of DNA from which only frequencies of alleles at the markers are known . In this study , we quantify how well such identification can work in practice . The larger the size of the sample from which the allele frequencies are available , the more independent genetic markers are required to allow individual identification .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "genetics", "and", "genomics/population", "genetics" ]	2009	The Limits of Individual Identification from Sample Allele Frequencies: Theory and Statistical Analysis

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.