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 |
|---|---|---|---|---|---|
Two-component signaling systems ( TCSs ) are major mechanisms by which bacteria adapt to environmental conditions . It follows then that TCSs would play important roles in the adaptation of pathogenic bacteria to host environments . However , no pathogen-associated TCS has been identified in uropathogenic Escherichia coli ( UPEC ) . Here , we identified a novel TCS , which we termed KguS/KguR ( KguS: α-ketoglutarate utilization sensor; KguR: α-ketoglutarate utilization regulator ) in UPEC CFT073 , a strain isolated from human pyelonephritis . kguS/kguR was strongly associated with UPEC but was found only rarely among other E . coli including commensal and intestinal pathogenic strains . An in vivo competition assay in a mouse UTI model showed that deletion of kguS/kguR in UPEC CFT073 resulted in a significant reduction in its colonization of the bladders and kidneys of mice , suggesting that KguS/KguR contributed to UPEC fitness in vivo . Comparative proteomics identified the target gene products of KguS/KguR , and sequence analysis showed that TCS KguS/KguR and its targeted-genes , c5032 to c5039 , are encoded on a genomic island , which is not present in intestinal pathogenic E . coli . Expression of the target genes was induced by α-ketoglutarate ( α-KG ) . These genes were further shown to be involved in utilization of α-KG as a sole carbon source under anaerobic conditions . KguS/KguR contributed to the regulation of the target genes with the direct regulation by KguR verified using an electrophoretic mobility shift assay . In addition , oxygen deficiency positively modulated expression of kguS/kguR and its target genes . Taken altogether , this study describes the first UPEC-associated TCS that functions in controlling the utilization of α-ketoglutarate in vivo thereby facilitating UPEC adaptation to life inside the urinary tract .
Urinary tract infection ( UTI ) is one of the most common bacterial infections in humans and is a significant clinical issue worldwide . Annually , UTIs are associated with 7 million office visits , 1 million emergency room visits , 100 , 000 hospitalizations , and $1 . 6 billion in healthcare costs [1] . In fact , ∼80–90% of community-acquired UTIs are caused by uropathogenic Escherichia coli ( UPEC ) [2] . Many virulence factors are required for UPEC to cause UTIs . Typically , UTIs begin with urethral contamination with UPEC from the bowel [3] . Successful attachment to the uroepithelium requires specific adhesins including P , type 1 and other fimbriae ( such as F1C , S , M , and Dr fimbriae ) [4] , [5] . UPEC may then ascend the urethra to enter the bladder and kidneys , where several highly regulated virulence factors , including fimbriae , secreted toxins ( hemolysin , Vat , Sat , and CNF ) and polysaccharide capsule , may contribute to colonization and pathogenesis [6] . It is likely that UPEC's ability to colonize the urinary tract and cause disease is affected by its adaptive responses to local environmental cues , including changes in nutrient availability . Urine is an important host environment that UPEC encounter after transition from the intestine to urinary tract . From the standpoint view of E . coli nutrition , urine is a mixture of amino acids and small peptides in low concentrations , with the notable exception of urea , which is abundant [7] . UPEC adapt to growth in urine by using peptides and amino acids as their primary carbon source for fitness [8] . About 85% of UPEC strains retain a complete dsdCXA locus [9] and these genes are responsible for the detoxification of D-serine allowing UPEC to use D-serine as the sole carbon and nitrogen source in urine [9] . Iron is another vital nutrient for UPEC but free iron concentrations in the mammalian host are extremely low . Consequently , UPEC have evolved multiple systems and strategies to obtain needed iron during infection including ent , iro , chu , sit , iutA and fyuA [10]–[13] . In addition , the urinary tract is considered to be a nucleotide-deficient environment . Study of pyrD and guaA deletion mutants showed that metabolism of nucleobases is required for UPEC colonization of the bladder [14] , [15] . Thus , UPEC metabolism appears to play a critical role in colonization and invasion of the host . Therefore , it is not surprising that acquisition of genomic islands encoding metabolic pathways is often essential for the colonization of host niches . Gene cluster ttrABC and ttrRS , responsible for utilizing tetrathionate as an electron acceptor in the intestine , are encoded on Salmonella enterica subsp . enterica serotype Typhimurium pathogenicity island II and contribute to detoxification of tetrathionate and confer a competitive advantage in growth [16]–[18] . Gene nixA within a transferable genomic island in Helicobacter pylori aids in transport of nickel and consequently results in enhanced activity of urease , a key factor for fitness and virulence [19]–[22] . Besides islands encoding iron uptake systems , no other metabolic island has yet been characterized in UPEC even though their study would help understand the pathobiology and facilitate the identification of potential targets for novel therapeutic approaches to prevent this prevalent pathogen . The pathways , conferring metabolic adaptation of bacterial pathogens to the host milieu , are usually controlled by both global and specific regulators . One mechanism used by most bacterial pathogens to sense and respond to nutrient availability is two-component signaling systems ( TCSs ) . TCSs , composed of a membrane-bound sensor histidine kinase ( HK ) and a cytoplasmic response regulator ( RR ) , have been implicated in regulating the response of bacteria to a wide array of signals and stimuli , including nutrients , quorum signals , antibiotics , etc . The recognition of physical or chemical signals by HK sensor domain typically triggers the modulation of its autophosphorylation activity . The phosphoryl group is then transferred to the RR , usually a DNA binding protein that functions by altering gene expression [23] , [24] . In E . coli K12 , more than 30 TCSs have been identified and characterized to some extent [25]–[27] . Of these , several were shown to be involved in UPEC pathogenesis . TCS BarA/UvrY controlled efficient switching between glycolytic and gluconeogenic carbon sources and contributed to UPEC virulence [28] . Deletion of QseC , the kinase of a well-known TCS QseC/QseB , dysregulated nucleotide , amino acid , and carbon metabolism and resulted in attenuation of UPEC virulence [29] , [30] . In addition , PhoQ/PhoP [31] and the AirS system [32] have been linked to UPEC pathogenesis . To date , all TCSs linked to UPEC pathogenesis have been common to E . coli and are not pathogen-associated . Given the distinct features of environments colonized by UPEC , we hypothesize that UPEC may possess special TCSs that sense urinary tract-specific metabolic signal ( s ) and adapt UPEC metabolism to available nutrients . Here , we describe the first TCS significantly associated with UPEC and its direct target genes encoded on a small genomic island . In addition , we demonstrate that the expression of these island genes is induced in response to α-KG , a nutrient found in high concentration in renal proximal tubule cells , and that these genes are involved in the utilization of α-KG under anaerobic conditions . Our results suggest that KguS/KguR contributes to UPEC fitness in vivo by facilitating the utilization of a host abundant metabolite α-KG . This study provides a new perspective for understanding UPEC pathogenesis in vivo .
TCSs enable bacteria to sense , respond , and adapt to a wide range of environments , stressors , and growth conditions [23] . Since UPEC encounter unique environmental conditions found within their host , as compared to commensal E . coli and diarrheagenic E . coli , we hypothesize that specific TCSs may exist and facilitate UPEC adaptation to certain host niches . Indeed , analysis of the UPEC CFT073 genome identified a putative TCS c5041/c5040 by the presence of specific domains in their respective predicted proteins ( Fig . 1 ) . Domains and three-dimensional structures of C5041 and C5040 were predicted by the threading method using the I-TASSER online server ( http://zhanglab . ccmb . med . umich . edu/I-TASSER/ ) and subsequently refined using four-body contact potentials . C5041 is predicted to be a typical sensor kinase harboring a dimerization and histidine phosphotransfer ( DHp ) domain and a histidine kinase-like ATPase domain ( Fig . 1A ) . The C-terminal 288 amino acids of C5041 contains five short signature segments , H- , N- , G1- , F- , and G2-boxes , which are conserved across most histidine kinases [33] , [34] . The histidine residue within the H-box , which is very likely phosphorylated during the signaling process , was found at position 394 . In the N-terminal section of C5041 , two transmembrane helices were identified at position 15–35 and at position 297–317 , respectively . Between the two transmemebrane helices , a periplasmic domain , which might serve as a signal sensor , was identified according to the distribution of charged residues ahead and behind the postulated transmembrane helices . The periplasmic domain of C5041 is 25 . 3% identical to that of DctB , a sensor kinase protein for dicarboxylates transport in Sinorhizobium meliloti ( of which crystal structure is available ) [35] , suggesting C5041 may sense analogs of succinate , which is sensed by DctB [36] . C5040 is 37% identical to the nitrogen regulatory protein C ( NtrC ) [37] , [38] at the amino acid sequence level . C5040 harbors the CheY-like receiver domain , phosphorylation site , AAA+ ATPase domain , and helix-turn-helix ( HTH ) domain ( Fig . 1B ) . The receiver domain ( position 13–123 ) catalyzes phosphoryl transfer from the HK to itself and regulates the output . The aspartate residue , which is predicted to be phosphorylated in the signal transduction cascade , was found at position 58 . In the C-terminal part of C5040 , a helix-turn-helix domain was identified , which is a DNA-binding domain that recognizes enhancer-like sequences . The CheY-like receiver domain , AAA+ ATPase domain , and helix-turn-helix ( HTH ) domains of C5040 are 32 . 5% , 51 . 7% , and 41 . 7% identical to that of NtrC , respectively . Such characteristics are consistent with our expectations for a TCS . Additionally , DNA sequence analyses were performed using the search engine at http://blast . ncbi . nlm . nih . gov/Blast . cgi and c5041 and c5040 did not show any homology to any characterized histidine kinase or response regulator genes , suggesting that C5041/C5040 is a novel TCS . Sequence analysis of all E . coli genomes available online showed that c5041/c5040 and its orthologs are absent from commensal and diarrheagenic E . coli , but present in UPEC strains . Also , genotyping of 317 E . coli clinical isolates including human diarrheagenic E . coli and UPEC ( Table S3 ) for c5041 and c5040 revealed their differential distribution . Both genes of this TCS were amplified in 139 of 200 UPEC ( 69 . 5% ) but were not detected in 12 enterotoxigenic E . coli , 66 enterohemorrhagic E . coli or 39 enteropathogenic E . coli isolates . Therefore , c5041/c5040 appears to be significantly associated with UPEC , raising the possibility that C5041/C5040 contributes to UPEC fitness or virulence in vivo . To study the role of C5041/C5040 in UPEC fitness in vivo , we compared the UPEC CFT073 wild-type strain with its Δc5041/c5040 double deletion mutant strain for the ability to colonize the mouse urinary tract at 48 hours post-inoculation using an in vivo competition assay . The double deletion mutation was constructed by replacement of c5041/c5040 with the cat cassette conferring chloramphenicol resistance . Though Δc5041/c5040 mutants grew as well as the wild type ( WT ) in LB , the mutant was significantly outcompeted by the WT , with a nearly 10-fold reduction in the median CFU/g from both bladders and kidneys at 48 hours post-inoculation ( Fig . 2A , P<0 . 05 for both organs ) . To verify that the impact on colonization is not due to a secondary mutation , in vivo complementation experiments were performed . A stable low-copy plasmid pGEN-MCS was used as it was shown to be well maintained in CFT073 up to 48 h even in the absence of antibiotic pressure [39] . Coding regions of c5041 and c5040 plus their predicted promoter region were cloned into pGEN-MCS , and the resultant construct was named pc5041/c5040 . As shown in Fig . 2B and C , the Δc5041/c5040 mutant containing the empty vector ( pGEN-MCS ) demonstrated an expected colonization defect in bladder and kidneys colonization as compared to the WT ( empty vector ) ( P<0 . 05 ) while mutants harboring the complementation plasmid ( pc5041/c5040 ) were able to colonize the bladder and kidneys at wild-type levels at 48 hours post-inoculation . These results clearly demonstrate that c5041/c5040 has a role in UPEC colonization and fitness in vivo . To determine how C5041/C5040 affects UPEC fitness in vivo , 2D-DIGE was used to identify differentially expressed gene products in the proteomes of the WT and single deletion mutants of c5041 and c5040 grown aerobically in human urine . Human urine was used as the growth medium to simulate the urinary tract environment [40]–[42] . c5041 and c5040 single mutants were constructed by replacement with Chlr gene , and the resultant mutants were named LMP100Chl ( CFT073 Δc5041::Chlr ) and LMP101Chl ( CFT073 Δc5040::Chlr ) . Using a cutoff of 1 . 5-fold change , seventeen proteins in the mutants were differentially expressed , as compared to the WT ( Fig . S1 ) . Of these , five were induced and five were repressed in Δc5041::Chlr mutant; while six were induced and five were repressed in the Δc5040::Chlr mutant . To determine the identities of the differentially expressed proteins , protein spots in the gel were excised and subjected to enzymatic digestion with trypsin followed by tandem mass spectroscopy . Functions of these proteins primarily fell into three categories: metabolism , cell envelope constituents , and translational machinery ( Table 1 ) . C5035 , which showed a 24 . 6-fold induction , is a putative 2-oxoglutarate dehydrogenase presumably involved in tricarboxylic acid ( TCA ) cycle . XylA , a xylose isomerase , was increased 8 . 7-fold . It catalyzes the interconversion between xylose and xylulose . SitA , induced 4-fold , is a putative periplasmic iron-binding protein . Several outer membrane proteins including OmpA , NmpC , OmpF and OmpX were uniformly induced in Δc5040::Chlr mutant . 50S ribosomal proteins , such as Rpll , RplQ , and RplF , were also differentially expressed in the mutants as compared to the WT . These results indicate that C5041/C5040 regulate multiple genes , suggesting that they may serve as a pleiotropic two-component system . The most differentially expressed protein identified by proteomic analysis was C5035 , which was induced nearly 25-fold in the Δc5041::Chlr mutant . qRT-PCR results confirmed that c5035 was up-regulated in Δc5041::Chlr mutant , but not in Δc5040::Chlr mutant , as compared to the WT when grown aerobically in human urine ( Fig . 3A ) . To rule out the possibility of polar effects , the Chlr gene was removed from both Δc5041::Chlr and Δc5040::Chlr mutants to generate mutants LMP100 ( CFT073 Δc5041 ) and LMP101 ( CFT073 Δc5040 ) , respectively . Interestingly , the expression of c5035 was detected at extremely low levels not only in the mutant strains Δc5041 and Δc5040 , but also in the WT . No significant differences in the c5035 expression were detected among the WT , Δc5041 and Δc5040 mutants when grown in human urine under aerobic conditions ( Fig . 3A ) . Note that in the Δc5041::Chlr mutant , c5041 was replaced by the Chlr gene , which transcribes in the same orientation as its downstream gene , c5040 . We further compared the expression levels of RR c5040 in the WT , Δc5041 , and Δc5041::Chlr mutants . The results showed that expression of c5040 was extremely low in the WT and Δc5041 mutant , but it was significantly upregulated in the Δc5041::Chlr mutant when grown in human urine under aerobic conditions ( Fig . 3B ) . Therefore , we suspect that the upregulation of c5035 in Δc5041::Chlr mutant was due to the constitutive expression of c5040 caused by the upstream Chlr gene . If so , overexpression of c5040 in the Δc5041 mutant should induce the expression of c5035 . Thus , c5040 was cloned into plasmid pMAL-MCS under the control of promoter Ptac and the resultant plasmid transformed into the Δc5041 mutant . Indeed , induction of c5040 significantly upregulated the expression of c5035 in the Δc5041 mutant ( Fig . 3C ) . We also tested such regulation under different culturing condition and the results revealed similar regulatory pattern . In summary , constitutive expression of c5040 induced the expression of c5035 regardless of growth conditions . Further analysis of genomic localization of c5035 in CFT073 revealed that genes c5032-5039 including c5035 , together with TCS genes c5041 and c5040 , are encoded by a genomic island inserted between E . coli K12 genes tyrB and aphA ( Fig . 4A ) . Not surprisingly , island genes c5032-5039 were strongly associated with UPEC , and about 91 . 4% of UPEC strains tested containing c5040/c5041 loci also possess c5032-5039 genes ( data not shown ) . Prediction by bioinformatics tools ( http://linux1 . softberry . com/berry . phtml and http://nostradamus . cs . rhul . ac . uk/~leo/sak_demo/ ) followed by reverse transcription PCR indicate that c5032-5037 forms one operon and c5038-5039 forms another ( Fig . S2 ) . We then examined if this genomic island contributed to UEPC fitness in vivo . An in vivo competition assay was used to compare the colonization levels of Δc5032-5039 mutants with that of wild-type strain . Δc5032-5039 mutant was constructed by replacement with the cat cassette . Fig . 4B showed that there is significant reduction in colonization levels of Δc5032-5039 mutants in bladder and kidneys at 48 hours post-inoculation , as compared to the WT ( P<0 . 05 ) . These results clearly demonstrated that island genes c5032-5039 are involved in UPEC colonization of murine urinary tract and it's tempting to presume that C5041/C5040 contributes to UPEC fitness through regulation of c5032-5039 genes ( The results that all other target genes in the genomic island are regulated by TCS C5041/C5040 were shown later in this paper ) . The proteins encoded by c5032 , c5034 , and c5035 are homologous to the E1 , E2 and E3 components of α-KG dehydrogenase in E . coli , with similarity of 64% , 69% , and 55% , respectively; whereas , the proteins encoded by c5036 and c5037 are 70% and 83% similar to the beta- and alpha-subunits of succinyl-CoA synthetase from E . coli . α-KG dehydrogenase converts α-KG to succinyl-CoA , which can be further transformed to succinate by succinyl-CoA synthetase [43] . c5038 is predicted to encode a putative dicarboxylate transporter with 13 transmembrane alpha-helices ( TMHMM program [44] ) , showing 49% similarity to citrate/succinate antiporter CitT , and c5039 encodes an enzyme belonging to malate/L-lactate dehydrogenase ( Ldh_2 ) family . To test the hypothesis that these enzymes and the metabolic transporter are involved in α-KG utilization , we compared the growth of the WT , the Δc5032-5037 mutant , the Δc5038-5039 mutant , and the Δc5032-5039 mutant in M9 minimal medium with α-KG as the sole carbon source . Under aerobic conditions , no significant growth differences were observed among these strains ( Fig . S3 ) . However , when tested under anaerobic conditions , deletion of c5032-5039 completely abolished the growth , and the Δc5032-5037 and Δc5038-5039 mutants also displayed statistically significant growth defects , as compared to the WT ( Fig . 5A ) . We further tested their growth in M9 medium containing glucose , glycerol , or four-carbon dicarboxylate compounds , including succinate , fumarate , and malate , as the sole carbon source . In these cases , no differences were found among the strains tested ( data not shown ) . These results indicated that genomic island genes c5032-5039 specifically contribute to α-KG utilization under anaerobic conditions but play no or a very limited role in α-KG utilization under aerobic conditions . The ability of α-KG to induce expression of c5032-5039 was also examined . Chromosomal c5032-lacZ and c5038-lacZ transcriptional reporter fusions in CFT073ΔlacZYA were constructed in order to study the expression of operon c5032-5037 and operon c5038-5039 , and β-galactosidase activity in M9 medium using α-KG ( 40 mM ) as the sole carbon source was also determined . Under anaerobic conditions , α-KG significantly induced the expression of c5032 and c5038 , as compared to the rich medium LB or M9 medium containing glycerol or glucose as the sole carbon source ( Fig . 5B ) ; while the analogs of α-KG such as 2-HO-glutarate , glutamate , glutarate , fumarate , or succinate could not . To verify that other target genes on the genomic island are also induced by α-KG , qRT-PCR was performed . RNA was isolated from bacteria grown in M9 medium ( glycerol and TMAO ) with or without inducer α-KG . As shown in Fig . S4 , all of the genes tested were greatly induced in the presence of α-KG . We then examined the effect of α-KG concentration on the stimulation of c5038 expression . Bacteria were grown in M9 medium plus glycerol and TMAO in the presence of various amounts of α-KG . We found that the induction of c5038 by the presence of α-KG was dose-dependent with concentrations as low as 200 µM able to induce their expression under anaerobic conditions ( Fig . 5C ) . These results indicate that island genes c5032-5039 are induced by α-KG and involved in anaerobic utilization of α-KG . To determine the regulatory role of C5040 in α-KG induction of target genes , chromosomal c5032-lacZ and c5038-lacZ transcriptional reporter fusions in the Δc5040 mutant were constructed . Strains were grown in M9 medium containing glycerol and TMAO in the presence or absence of inducer α-KG under anaerobic conditions . As shown in Fig . 6A , in wild-type genetic background , c5032 and c5038 expression was considerably induced in response to α-KG; in contrast , deletion of c5040 abolished the expression of c5032 and c5038 to the background level under anaerobic conditions . Re-introduction of a plasmid construct pc5041/c5040 carrying c5041/c5040 genes with their native promoter into c5040 mutants ( LMP206 and LMP209 ) markedly increased expression of c5032 and c5038; whereas , the empty plasmid vector did not affect the expression of c5032 and c5038 . In addition , the regulation of other target genes ( c5034 , c5035 , c5036 , c5037 , and c5039 ) in the genomic island in response to α-KG by C5040 was confirmed using qRT-PCR ( Fig . S4 ) . Together , these results indicate that α-KG induction of c5032-5039 is C5040-dependent ( Fig . 6A ) . To determine if C5040 directly regulates c5032 and c5038 expression , an electrophoretic mobility shift assay ( EMSA ) was performed . The promoter regions of c5032 and c5038 were predicted by BProm program ( http://linux1 . softberry . com ) [45] . The potential binding sites for c5032 and c5038 were identified and found to be highly similar , with TGTGTG-N13-CGCGCA for c5032 and TGTGCG-N8-CGCACA for c5038 . DNA fragments containing the potential binding sites were then PCR amplified for use as probes ( 305 nucleotides in size for c5032 , starting from −272 to +32 relative to translational start codon; 196 nucleotides in size for c5038 , starting from −163 to +32 relative to translational start codon ) . Fragments amplified from the coding region of c5036 were used as negative controls . Gene c5040 encoding the RR was cloned into the expression vector pMal-c2x and fused with MBP-His6 . MBP-C5040-His6 fusion protein retained its regulatory biological function since introduction of the plasmid construct expressing MBP-C5040-His6 fusion protein into Δc5040 mutant activated the expression of c5038; whereas , the empty plasmid did not affect expression of c5038 ( Fig . S5 ) . As shown in Fig . 6B , the purified MBP-C5040-His6 fusion protein was able to shift the promoter fragments of both c5032 and c5038 , but not the control fragment . Meanwhile , use of the purified MBP-His6 fusion protein without C5040 was not associated with a detectable protein-DNA complex ( Fig . 6B ) . These results demonstrate that C5040 directly binds to the promoters of the two operons . To determine if C5040 is involved in the utilization of α-KG as the sole carbon source under anaerobic conditions , the growth properties of the WT , Δc5040 mutants , and the complemented strains were compared . When grown in M9 medium with α-KG as the sole carbon source under anaerobic conditions , the Δc5040 mutant strains showed a growth defect as compared to the WT . The c5040-complemented strain ( Δc5040 carrying pc5041/c5040 ) significantly increased growth in this medium; while the empty plasmid vector ( pGEN-MCS ) did not improve growth of the mutant ( Δc5040 ) ( Fig . 6C ) . The regulatory role of HK C5041 was also determined . Chromosomal c5032-lacZ and c5038-lacZ transcriptional reporter fusions in the Δc5041 mutant ( LMP106 ) were obtained to create LMP205 and LMP208 , respectively . Strains were grown in M9 medium ( supplemented with glycerol and TMAO ) with or without inducer α-KG under anaerobic conditions . Deletion of c5041 abolished the expression of c5032 and c5038 to background levels in the presence of α-KG . Re-introduction of the plasmid pc5041 carrying the c5041 gene with its native promoter into LMP205 and LMP208 mutants significantly increased their expression ( P<0 . 05 ) , indicating that C5041 positively affects expression of the target genes ( Fig . 7A ) . This suggests that phosphorylation is very important for the regulatory function of RR C5040 . In addition , the expression of other genes ( c5034 , c5035 , c5036 , c5037 , and c5039 ) in the genomic island affected by C5041 was also confirmed using qRT-PCR ( Fig . S4 ) . The growth phenotypes of the c5041 mutant and its complementation strain in difference medium were then tested . Interestingly , the growth of the WT and its derivative mutants ( c5041 mutant and complementation strain ) in all media tested was statistically indistinguishable ( Fig . 7B ) . Island genes c5032-5039 contribute to anaerobic utilization of α-KG but play a limited role under aerobic conditions causing us to suspect that oxygen might modulate expression of these genes . Indeed , the c5032-5037 operon encoding a putative α-KG dehydrogenase and a putative succinyl-CoA synthetase was upregulated >5-fold in a reduced oxygen environment . Similarly , the expression of the c5038-5039 operon , encoding a putative transporter of α-KG and a protein with unknown function , was upregulated 2-fold under anaerobic conditions , as compared to aerobic conditions ( Fig . 8A ) . To determine if oxygen regulates c5032-5039 expression via C5041/C5040 but not through other independent pathways , the effects of c5041/c5040 double deletion on the expression of c5032-5039 were examined . Deletion of c5041/c5040 completely abolished expression of the two operons under both aerobic and anaerobic conditions , supporting the hypothesis that oxygen regulates these two operons via TCS C5041/C5040 ( Fig . 8A ) . In addition , we wished to determine if oxygen deficiency regulates expression of the TCS itself . Under anaerobic conditions , the expression of c5041 was about 3-fold greater than that observed under aerobic conditions ( Fig . 8B ) , suggesting that c5041 expression is significantly induced in response to anaerobiosis . Similarly , under anaerobic conditions , c5040 expression is dramatically increased as compared to that under aerobic conditions ( Fig . 8B ) . These results suggested that oxygen deficiency upregulated the expression of c5041 and c5040 and raised the possibility that oxygen modulates the expression of c5032-5037 and c5038-5039 operons via controlling c5041/c5040 expression . However , these results did not rule out the possibility that anaerobiosis stimulates the typical TCS phosphotransfer reactions between C5041 and C5040 , leading to increased DNA binding and up-regulation of target gene expression .
One of the greatest challenges confronted by all microorganisms is adapting to rapid changes of nutrient availability in different habitats . In the course of evolution , bacteria have developed several mechanisms to sense and utilize available nutrient sources associated with particular niches or to favor the most efficiently metabolizable nutrient sources when exposed to a range of choices . TCSs are major mechanisms enabling bacteria to couple environmental stimuli to adaptive responses [46] . The TCSs in E . coli K12 have been extensively studied [25]–[27] . Many TCSs that are common to both commensal and pathogenic E . coli , such as PhoQ/PhoP [31] , QseC/QseB [29] , [30] , BarA/UvrY [28] , and the AirS system [32] have been shown to contribute to virulence by mediating bacterial adaption to the host environment . However , no pathogen-associated TCS has yet been characterized in E . coli . Driven by the hypothesis that UPEC-associated TCSs exist to sense and respond to host environment signals distinct from those of intestinal E . coli , we identified a novel TCS C5041/C5040 significantly associated with UPEC strains , but not with EHEC , EPEC or ETEC strains . This TCS activated the expression of a genomic island involved in transport and metabolism of α-KG under anaerobic conditions . In view of these findings , C5041 was renamed KguS ( α-ketoglutarate utilization sensor ) and C5040 renamed KguR ( α-ketoglutarate utilization regulator ) . These results indicated that UPEC might actively import α-KG and that α-KG could be an important carbon source for UPEC in vivo . Consistent with this observation , deletion of the TCS or its target island genes resulted in a significant reduction in a UPEC's colonization of the murine urinary tract . To our knowledge , this is the first report of a pathogen-associated TCS in E . coli that contributes to UPEC pathogenesis . α-KG is an intermediate in the citric acid/tricarboxylic acid ( TCA ) cycle , which is a metabolic pivot for both catabolic and anabolic processes that supply key metabolic intermediates and energy in most eubacteria [47] . Many intermediates in the TCA cycle can be sensed by TCSs , which elicit an adaptive response in E . coli . Four-carbon ( C4 ) dicarboxylate-sensing DcuS/DcuR was shown to control fumarate transportation ( dcuB ) and respiration ( frdABCD ) under anaerobic conditions , and under aerobic conditions affect the utilization of most of C4-dicarboxylates like succinate , fumarate , malate , and aspartate [48] , [49] . In addition , tricarboxylate-sensing CitA/CitB was demonstrated to regulate anaerobic citrate fermentation genes citCDEFXGT encoding an active holo-citrate lyase catalyzing the breakdown of citrate to acetate and oxaloacetate and a citrate carrier [50]–[52] . This study has demonstrated that a novel TCS KguS/KguR responds to and only to α-KG , a five-carbon ( C5 ) dicarboxylate by activating metabolic enzymes and transporter involved in anaerobic metabolism of α-KG . This is the first-time that such a TCS has been identified responding to C5-dicarboxylate in E . coli . Identification of the TCS regulating utilization of C4 , C5-dicarboxylates and tricarboxylate suggested the tight control of the pathways involved in the metabolism of intermediates in E . coli TCA cycle . As found in this study , α-KG can induce expression of the TCS target genes c5032-5039 , while deletion of kguS or kguR abolished their expression in response to α-KG . By extrapolation from what we know of other TCSs and the results obtained in this study , it is very likely that α-KG serves as the signal molecule and HK KguS senses it via its periplasmic domain , then phosphorylates itself , and subsequently trans-phosphorylates KguR , which can then induce expression of the target genes . Very interestingly , overexpression of KguR can activate the expression of target genes independent of inducer α-KG . Similar observations have been reported for Salmonella where induction of target genes depends on the intracellular concentration of the response regulator PhoP and overexpression can substitute for phosphorylation and enhance the functionality of PhoP [53] . We further confirmed that oxygen tension modulated the expression of this pathway by regulating the expression of KguS/KguR itself . In E . coli , regulation by oxygen is usually mediated by the cytoplasmic regulator FNR and/or by TCS ArcB/A [54] . Currently , it is unknown if FNR and/or ArcB/A have any effect on the expression of KguS/KguR . Genome sequencing has shown that the TCA cycle is a readily modified pathway in bacteria with many strains harboring variant TCA cycles . The common feature of these variants is the absence or repression of the α-KG dehydrogenase with the oxidative branch terminating in the production of α-KG and the reducing branch in succinate [55] , [56] . Like any other E . coli , UPEC operate a complete TCA cycle aerobically , but the α-KG dehydrogenase shared by all E . coli is neither functional nor expressed under anaerobic conditions [47] . The KguS/KguR-controlled metabolic pathway in UPEC includes a putative α-KG dehydrogenase ( c5032-5035 ) and a putative succinyl-CoA synthetase ( c5036-5037 ) , which were induced and contributed to utilization of α-KG under anaerobic conditions . To our knowledge , this is the first variant TCA cycle described in E . coli that links the oxidative and reducing branches under anaerobic conditions . To maintain carbon flux under oxygen-limiting or anaerobic conditions , many microaerophilic and obligate anaerobic organisms that lack α-KG dehydrogenase possess similar bypass pathways to metabolically link oxidative and reducing branches . Helicobacter pylori preferably growing under microaerophilic conditions use α-KG ferredoxin oxidoreductase to connect oxidative and reducing half-cycles [57] . Likewise , in Mycobacterium an α-KG decarboxylase structurally related to the E1 component of 2-oxoglutarate dehydrogenase can produce succinic semialdehyde ( SSA ) , which is subsequently oxidized to succinate by succinic semialdehyde dehydrogenase ( SSADH ) [58] . This study and other groups' findings have highlighted the importance of the branched variants in adaptation to oxygen limiting or obligate anaerobic conditions and suggest that UPEC have different metabolic niches and needs compared to other E . coli . The acquisition of additional metabolic traits often increases adaptability and competitiveness of pathogens in a new niche [18] , [22] . This study found that TCS KguS/KguR and its controlled-anaerobic utilization pathway of α-KG were encoded by a genomic island in UPEC , which was absent from the genomes of commensal E . coli and intestinal pathogenic E . coli . These results were consistent with the previous report that c5032 to c5040 were among the 131 UPEC-specific genes identified using comparative genomic hybridization and in silico analysis of ten UPEC and four fecal/commensal isolates [59] , and also suggested that this genomic island may facilitate UPEC's adaptation to in vivo colonization . This was confirmed by the observation that deletion of kguS/kguR resulted in a significant reduction in UPEC colonization in the bladders and kidneys of mice . However , deletion of kguS/kguR or its target genomic island genes did not affect growth in human urine in vitro under both aerobic and anaerobic conditions ( data not shown ) . We reasoned that human urine might only be able to partially imitate the in vivo conditions of UPEC infection , whereas the contributions of KguS/KguR to UPEC pathogenesis may be restricted to some particular site of infection or pathogenic process , which cannot be represented by urine . Interestingly , Chlamydia trachomatis , an obligate intracellular bacterial pathogen of urogenital infections [60] , possesses an incomplete TCA cycle . Because of their small genome , chlamydiae lack many metabolic pathways retaining only the functions for performing key steps and interconversions of metabolites from the host cells . The fact that chlamydiae lack the first three enzymes of the TCA pathway and obtains α-KG from their host cell via a membrane transporter [56] , [61] suggests that utilization of intracellular α-KG is important for the survival of this urogenital intracellular pathogen . It could be inferred from Chlamydia and hypothesized that α-KG might be one of the preferred carbon and energy sources for UPEC during infection . Previous studies supporting this hypothesis showed that the oxygen tension ( PO2 ) in UPEC-infected tissue dropped to 0 mmHg within 3 . 5–4 h [62] , suggesting UPEC encountered anaerobic conditions during infection . It was also reported that the concentration of α-KG in human cells and blood is about 10 µM , a concentration that was able to induce the expression of KguS/KguR-controlled pathway in vitro under anaerobic conditions in this study . The KguS/KguR-controlled transporter ( C5038 ) differs from the constitutively expressed KgtP encoded by all E . coli [63] , [64] and was induced under anaerobic conditions . During infection , this transporter allows UPEC to import α-KG from the infected tissue into the cell , and α-KG dehydrogenase and succinyl-CoA synthetase allow UPEC to take advantage of α-KG under anaerobic conditions , thus contributing to UPEC's in vivo fitness . More interestingly , proximal tubule cells in the kidney are able to accumulate α-KG . Proximal tubules are the primary site where organic anions ( OA ) of physiological , pharmacological , and toxicological importance are cleared from the body . Their clearance involves uptake of OA via OA-dicarboxylate exchange driven by α-KG gradients . Thus , intracellular concentration of α-KG in the proximal tubule cells is ∼100–400 µM , which is 10 to 40-fold higher than that found in any other cells [65]–[67] . UPEC undergo an intracellular lifestyle where they form biofilm-like communities [68] , and UPEC are able to infect proximal tubule cells [62] , [69]–[71] . However , very little is known about the metabolism of UPEC within these host cells . It is very likely that KguS/KguR-controlled anaerobic utilization pathway of α-KG would increase UPEC fitnessduring infections of renal proximal tubule cells . Future studies in our laboratory should help clarify these questions . In summary , our findings suggested a model that describes a novel regulatory and metabolic pathway in UPEC . For optimal growth during infection , HK KguS may sense α-KG , an abundant metabolite in UPEC infection site-renal proximal tubules , and then phosphorylate the RR KguR , which finally activates the import and utilization of α-KG ( Fig . 9 ) . These findings provide compelling evidence that this first UPEC-associated TCS enables E . coli to sense infection niche-specific stimuli and adapt to local nutrient availability , thus increasing its in vivo fitness . Hopefully , this and future studies on this regulatory system and metabolism pathway could lead to a better understanding of correlations between bacterial metabolism and virulence , but also the molecular pathogenesis of UPEC .
7All animal procedures were conducted in accordance with NIH guidelines , the Animal Welfare Act and US federal law . The experimental protocol ( Protocol number 4-11-7111-Z ) for handling animals was approved by Institutional Animal Care and Use Committee at Iowa State University ( IACUC ) . All surgery was performed under isoflurane anesthesia , and all efforts were made to minimize suffering . The study using human urine was approved by Institutional Review Board ( IRB ID: 04-171 ) . Written informed consent was obtained from human participants and/or their legal guardians for urine collection . Strains and plasmids used in this study are listed in Table S1 . Aerobic growth was achieved by shaking in air at 180 rpm and anaerobic growth by incubating in a Bactron chamber ( Sheldon Manufacturing , Inc . , OR ) filled with gas mixture ( N2 , 90%; CO2 , 5%; H2 , 5% ) . For genetic manipulations , all E . coli strains were grown routinely in lysogenic broth ( LB ) medium . For growth and gene expression studies , bacteria were generally grown aerobically or anaerobically in M9 minimal salts with certain carbon sources indicated , supplemented with 2 mM MgSO4 , 0 . 1 mM CaCl2 , and 1 mg/ml vitamin B1 . When used , trimethylamine N oxide ( TMAO ) and di- or tri-carboxylates were present at 40 mM . Glucose ( 0 . 5% v/v ) or glycerol ( 0 . 25% v/v ) was added as energy substrates , as indicated . Fresh mid-stream human urine was collected from six male and female consenting donors , pooled , and filter sterilized . Selective antibiotics and IPTG were added when necessary at the following concentrations: ampicillin ( Amp ) , 100 µg ml−1; kanamycin ( Kan ) , 50 µg ml−1;chloramphenicol ( Chl ) , 25 µg ml−1; IPTG , 1 mM . Polymerase chain reaction ( PCR ) , DNA ligation , electroporation and DNA gel electrophoresis were performed according to Sambrook and Russel [72] unless otherwise indicated . All oligonucleotide primers were purchased from Integrated DNA Technologies ( Iowa ) and are listed in Table S2 . All restriction and DNA-modifying enzymes were purchased from New England Biolabs and used based on the suppliers' recommendations . Recombinant plasmids , PCR products , and restriction fragments were purified using QIAquick PCR purification kit or MinElute gel extraction kit ( Qiagen , CA ) as recommended by the supplier . DNA sequencing was performed at the DNA facility , Iowa State University . DNA and amino acid sequence analyses were performed using CloneManager software ( Scientific & Educational Software , NC ) and the search engine ( http://blast . ncbi . nlm . nih . gov/Blast . cgi . ) was used to identify conserved domain structures of the two-component system . Deletion mutants were constructed using the lambda red recombinase system described by Datsenko and Wanner [73] . Chromosomal transcriptional lacZ fusion was constructed by homologous recombination of the suicidal plasmid pVIK112 carrying a fragment of complete 5′-region , 3′- region , or internal fragment of the target gene [74] . Briefly , PCR fragments of target genes were cloned into pVIK112 using EcoRI and XbaI sites . The resultant pVIK112 derivatives were introduced into CFT073ΔlacZYA::Chlr by conjugation and the integration was allowed to occur . Conjugants were selected and confirmed by PCR . Chlr was removed by transforming pCP20 plasmid carrying flippase [75] . For complementation , the coding sequences of genes plus their putative promoter regions were amplified from the CFT073 genome and independently cloned into pGEN-MCS [39] using EcoRI and SalI restriction sites . To construct the plasmid overproducing MBP-KguR-His6 fusion protein , a 1 . 4-kb fragment containing the coding region of kguR was obtained by PCR from genomic DNA using MalE/c5040Histag-F and MalE/c5040Histag-R carrying codons for 6×His and subsequently cloned into pMAL-c2x ( New England Biolabs ) using BamHI and HindIII sites . The resultant plasmid contains MBP-KguR-His6 under the control of the Ptac promoter . The construction of control vector pMAL-MCS was achieved by replacing the MalE coding sequence with the MCS fragment from pEGFP plamid using NdeI and EcoRI sites . The nucleotide sequences publicly available of c5041/c5040 genes were aligned by using the ClustalW2 program . The primers were selected from a relatively conserved region and on the basis of G/C content , annealing temperature and size of the amplicon . Multiplex PCR was carried out according to Johnson et al . [76] . Overnight LB cultures of E . coli containing the gene of interest-lacZ fusions were washed with PBS once and then were diluted 1∶100 in LB or M9 medium with the carbon sources indicated and grown at 37°C to log phase . These cultures were diluted 1∶10 in Z buffer and assayed for β-galactosidase activity using ortho-Nitrophenyl-β-galactoside ( ONPG ) as a substrate as described previously [77] . To prepare bacterial protein samples , isolated colonies were used to inoculate LB and cultures grown overnight at 37°C , the culture was further diluted 1∶50 to fresh LB medium and re-incubated; after OD 600 reached 1 . 0 , 2 ml of culture ( approximately 109 CFU ) was pelleted and the supernatant removed . Bacterial pellets were re-suspended in 20 ml of human urine and grown aerobically for 4 hours at 37°C . Bacteria grown in human urine were then collected by centrifugation and washed 3 times with sterile PBS . The resultant bacterial pellets were snap-frozen at −80°C and submitted to Applied Biomics , Inc . in California for Fluorescence difference in gel electrophoresis ( 2D-DIGE ) [78] . RNA from E . coli CFT073 and its derivatives was stabilized by RNAprotect Bacterial Reagent ( QIAGEN ) and extracted using an RNeasy Mini Kit ( QIAGEN ) with a one-hour in-tube DNase digestion ( QIAGEN ) to remove possible DNA contamination according to the manufacturer's instructions . Three biological replicates of the each sample were prepared . The concentration of RNA was determined using a Spectrophotometer ( ND-1000 ) ( NanoDrop ) . For the co-transcription test , one microgram of total RNA was reverse transcribed in triplicate using random hexamers and ImProm-II reverse transcriptase ( Promega ) . Reactions without reverse transcriptase were used as a DNA contamination control . cDNA was then used as template for subsequent PCR reactions . For quantitative real-time RT-PCR , melting curve analyses were performed after each reaction to ensure amplification specificity . Differences ( n-fold ) in transcripts were calculated using the relative comparison method , and amplification efficacies of each primer set were verified as described by Schmittgen et al [79] . RNA levels were normalized using the housekeeping gene tus encoding DNA replication terminus site-binding protein as an endogenous control [80] . Quantitative real-time RT-PCR ( qRT-PCR ) was performed with a Bio-Rad iQ5 iCycler detection system using iScript one-step RT-PCR kit with SYBR Green ( Bio-Rad ) according to the manufacture's instruction [81] . To study the binding of KguR to the DNA probe , electrophoretic mobility shift assays ( EMSAs ) were performed using the commercialized EMSA kit ( Invitrogen , California ) [82] . MBP-KguR-His6 fusion protein was purified to homogeneity using Ni-NTA Spin Columns and dialyzed against binding buffer . DNA probes were PCR amplified using specific primers and gel purified using a QIAGEN MinElute gel extraction kit . EMSAs were performed by adding increasing amounts of purified MBP-KguR-His6 fusion protein ( 0 to 300 ng ) to the DNA probe ( 60 ng ) in binding buffer ( 10 mM Tris ( pH 7 . 5 ) , 1 mM EDTA , 1 mM dithiothreitol , 50 mM KCl , 50 mM MgCl2 , 1 µg ml−1 bovine serum albumin ( NEB ) ) for 30 min at room temperature . Reaction mixtures were then subjected to electrophoresis on a 6% polyacrylamide gel in 0 . 5×TBE buffer ( 44 . 5 mM Tris , 44 . 5 mM boric acid , 1 mM EDTA , pH 8 . 0 ) at 200 V for 45 min . The gel was stained in 0 . 5×TBE buffer containing 1×SYBR Gold nucleic acid staining solution for 30 min . Mouse infection studies were performed according to the methods of Johnson et al [83] . Female CBA/J mice ( six to ten weeks of age ) were anesthetized and inoculated via transurethral catheterization with a 20 µl ( 2×109 CFU ) challenge inocula per mouse . Overnight LB cultures for CFT073 and the mutant strain were pelleted and resuspended in sterile PBS , mixed in equal number and adjusted to make challenge inocula . To determine the initial CFU/mL , dilutions of each inoculum were plated onto LB plates with and without chloramphenicol . After 48 h , the mice were euthanized and the bladder and kidneys were aseptically removed , weighed , and homogenized in tubes containing PBS . Dilutions of the homogenized tissue were then plated onto duplicate LB plates with and without chloramphenicol or plates with different antibiotics to determine the bacterial concentration ( CFU/g ) of tissue . After overnight incubation , distinct colonies on plates were enumerated . The numbers of colonies on selective plates were subtracted from those on LB plates to obtain the number of wild-type bacteria . In the case of in vivo complementation studies , recovered bacteria were plated on LB plates with ampicillin and LB with ampicillin and chloramphenicol . Similarly , the numbers of colonies on ampicillin and chloramphenicol plates were subtracted from those on LB plates with ampicillin to obtain the number of wild-type bacteria . A group of 10 mice for each dual-strains challenge were used to determine alterations in fitness . The WT/Δc5041/c5041::Chlr competition assay was performed twice while other assays were performed once . For statistical analysis , a two-tailed Wilcoxon matched pairs test was used ( Prism software , CA ) and the threshold for statistical significance was a P value <0 . 05 . | Successful colonization requires bacterial pathogens to adapt their metabolism to the conditions encountered in particular infection sites . Two-component signaling systems ( TCSs ) enable bacterial pathogens to sense and respond to environmental cues , thus mediating their adaptation to environmental change . Though many TCSs that have been characterized in commensal E . coli have been associated with UPEC pathogenesis , no characterized TCS has been significantly associated with UPEC strains . Here , we characterized a UPEC-associated TCS that was localized to a genomic island . This novel TCS and its target genes were involved in anaerobic utilization of α-ketoglutarate , an abundant metabolite in UPEC infection site-renal proximal tubules , thus contributed to UPEC fitness in vivo . Our results also suggest that this TCS controls a variant tricarboxylic acid ( TCA ) cycle , the first described in E . coli , which links the oxidative and reducing branches under anaerobic conditions . Similar TCA branches have been identified in other bacterial pathogens that have adapted to oxygen-limited , obligate anaerobic conditions , and/or intracellular carbon sources . Therefore , this study provides new insight into the adaptation of bacterial pathogens to nutrient availability in vivo and makes possible the discovery of targets for antimicrobial treatment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"genetics",
"biology",
"microbiology"
] | 2013 | A Novel Two-Component Signaling System Facilitates Uropathogenic Escherichia coli's Ability to Exploit Abundant Host Metabolites |
Enterotoxigenic Escherichia coli ( ETEC ) is a major diarrheal pathogen in developing countries , where it accounts for millions of infections and hundreds of thousands of deaths annually . While vaccine development to prevent diarrheal illness due to ETEC is feasible , extensive effort is needed to identify conserved antigenic targets . Pathogenic Escherichia coli , including ETEC , use the autotransporter ( AT ) secretion mechanism to export virulence factors . AT proteins are comprised of a highly conserved carboxy terminal outer membrane beta barrel and a surface-exposed amino terminal passenger domain . Recent immunoproteomic studies suggesting that multiple autotransporter passenger domains are recognized during ETEC infection prompted the present studies . Available ETEC genomes were examined to identify AT coding sequences present in pathogenic isolates , but not in the commensal E . coli HS strain . Passenger domains of the corresponding autotransporters were cloned and expressed as recombinant antigens , and the immune response to these proteins was then examined using convalescent sera from patients and experimentally infected mice . Potential AT genes shared by ETEC strains , but absent in the E . coli commensal HS strain were identified . Recombinant passenger domains derived from autotransporters , including Ag43 and an AT designated pAT , were recognized by antibodies from mice following intestinal challenge with H10407 , and both Ag43 and pAT were identified on the surface of ETEC by flow cytometry . Likewise , convalescent sera from patients with ETEC diarrhea recognized Ag43 and pAT , suggesting that these proteins are expressed during both experimental and naturally occurring ETEC infections and that they are immunogenic . Vaccination of mice with recombinant passenger domains from either pAT or Ag43 afforded protection against intestinal colonization with ETEC . Passenger domains of conserved autotransporter proteins could contribute to protective immune responses that develop following infection with ETEC , and these antigens consequently represent potential targets to explore in vaccine development .
Enterotoxigenic Escherichia coli ( ETEC ) are a major cause of diarrheal illness in developing countries where these organisms cause hundreds of millions of infections and an estimated 300 , 000–500 , 000 deaths in young children each year [1] . ETEC are perennially by far the most common cause of traveler's diarrhea [2] . Disease caused by ETEC is highly endemic in regions plagued by inadequate sanitation and a lack of clean drinking water , and prevention of ETEC is a high priority [1] , [3] . ETEC are genetically heterogeneous pathogens that share the ability to colonize the small intestine where they deliver the cholera toxin-like heat-labile toxin ( LT ) and/or small peptide heat-stable ( ST ) toxins that ultimately result in diarrhea [4] . In the classic paradigm for ETEC pathogenesis , small intestinal colonization requires plasmid-encoded colonization factors ( CFs ) [4] . A variety of more than 25 antigenically distinct fimbrial , or fibrillar CFs have been described to date [5] , [6] . These antigens , along with LT , remain central to ETEC vaccine development [7] . However , CF antigens are not appreciably cross-protective , and many ETEC strains do not appear to produce CFs [8] , [9] . Moreover , LT alone ( or the homologous cholera toxin ) do not appear to afford complete sustained protection [10] , while ST , typically only 19 amino acids in its mature form , is not suitably immunogenic . These constraints , as well as a growing appreciation of the complexity of ETEC pathogenesis [4] , [11] , have prompted searches for additional surface-expressed antigens . Use of classical genetic approaches including TnphoA mutagenesis to find novel molecules exposed on the surface of ETEC , recently led to the identification of several putative virulence loci , including the etpBAC two-partner secretion locus [12] , responsible for secretion of the EtpA adhesin molecule [13] , and the autotransporter ( AT ) protein EatA [14] . EatA and other AT proteins contain three essential domains: an amino terminal signal peptide , the secreted “passenger” domain , and a third carboxy-terminal beta barrel domain inserted into the outer membrane [15] . The variable passenger portion of the protein may be cleaved by surface proteases and freely secreted as in the case of EatA , or remain attached to the transport domain . The surface expression of AT passenger domains proteins make them attractive targets for vaccine development , while only limited portions of the beta regions are predicted to be exposed [16] . While a broadly protective ETEC vaccine remains outstanding , one approach currently being explored is a protein subunit vaccine based on multiple ETEC antigens . Present acellular pertussis vaccines [17] , [18] , [19] , subunit formulations containing a two-partner secretion ( TPS ) exoprotein adhesin ( filamentous hemagglutinin , FHA ) [20] , [21] , the pertactin autotransporter [22] , [23] , and pertussis toxoid offer a potential strategy that might be adopted to guide ETEC vaccine development . Indeed , recent investigations of EtpA [12] , [13] , [24] , [25] , an ETEC TPS exoprotein adhesin , were prompted by its similarity to FHA . Recent immunoproteomic studies of ETEC H10407 independently identified EtpA as well as several AT proteins including EatA , TibA and antigen 43 suggesting that these proteins are expressed during both experimental infection in mice and in humans [26] . Interestingly , it appears clear that children repeatedly exposed to ETEC infections are ultimately protected against subsequent symptomatic infections [27] . However , the precise composition of the protective antigens remains uncertain [4] , [28] , [29] . The present studies were performed to examine the possible contribution of conserved , chromosomally-encoded AT proteins to protective ETEC immune responses , and to evaluate passenger domains of these ATs as possible candidates for ETEC vaccine development .
A complete list of strains and plasmids employed in these studies is included in table 1 . ETEC strains H10407 and E24377A were originally provided by Marcia Wolf and Stephen Savarino , respectively , from cGMP lots maintained at Walter Reed Army Institute of Research . A number of parallel bioinformatics approaches were used to identify candidate AT genes in recently sequenced ETEC strains . Strains B7A and E24377A were searched for highly conserved autotransporter domains using the Pfam database including the autotransporter beta domain ( http://pfam . sanger . ac . uk//family/PF03797 ) and the pertactin domain ( http://pfam . sanger . ac . uk//family/PF03212 ) . The resulting sequences containing these domains were used to identify additional autotransporters in the genome of ETEC strain H10407 available in un-annotated form via the Sanger Institute ( http://www . sanger . ac . uk/Projects/E_coli_H10407/ ) , which was facilitated by interrogating the available sequence using the National Microbial Pathogen Database Resource ( NMPDR ) [30] on the Rapid Annotation Subsystem Technology ( RAST ) server ( http://RAST . nmpdr . org/ ) [31] . SignalP ( http://www . cbs . dtu . dk/services/SignalP/ ) was used to identify potential signal peptide encoding regions of the predicted AT coding sequences . BLASTP ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) and Pfam domain searches of each putative AT peptide were used to define the conserved beta barrel transport domain ( TIGR01414 , and pfam03797 , in the NCBI Conserved Domain Database http://www . ncbi . nlm . nih . gov/Structure/cdd/cdd . shtml and Pfam databases , respectively ) . Peptide sequence alignments of AT proteins common to the three ETEC strains used ClustalW [32] and MUSCLE [33] alignment algorithms performed locally ( Mac OS 10 . 5 . 8 ) with an additional alignment plug-in ( v1 . 06 ) for CLC Main Workbench software ( v 5 . 5 ) . Regions corresponding to the majority of each passenger domain ( region between the end of the putative signal peptide encoding sequence and the beginning of the beta barrel domain ) were amplified using primers bearing attB flanking sequences by high fidelity PCR ( Platinum PCR SuperMix , Invitrogen ) . A complete list of primers used in these studies is included in table 2 . Resulting amplicons were agarose gel-purified ( Ultra Clean 15 , MOBIO ) . Amplicons containing attB flanking sequences were cloned by recombination with the lambda attP sites on the entry vector pDONR221 ( BP Clonase II , Invitrogen ) , and transformed into DH10B One Shot ccdB Survival T1 Chemically Competent E . coli ( Invitrogen ) . Colonies were selected and patched onto kanamycin and chloramphenicol plates . Plasmid DNA extracted from kanamycin-resistant , chloramphenicol-sensitive colonies was analyzed by restriction digest to confirm presence of the appropriate insert . pDONR221 entry clones were then recombined with pDEST17 ( LR Clonase II ) placing the passenger domains in-frame with an amino-terminal polyhistidine tag in the resulting expression plasmid . Cloning reactions were used to transform DH5α to ampicillin resistance . Plasmid DNA from ampicillin-resistant , chloramphenicol-sensitive colonies was analyzed first by restriction endonuclease digestion , then sequenced using T7 promoter primer 5′-TAATACGACTCACTATAGG-3′ to confirm that the 6His- and passenger encoding sequences were in-frame . Resulting plasmids were then used to transform E . coli expression strain BL21-A1 . To produce recombinant polyhistine-tagged passenger proteins cultures of BL21-A1 containing pDEST17-encoded 6His-passenger clones were grown overnight at 37°C in Luria broth containing ampicillin ( 100 µg/ml ) . After diluting 1∶100 in 100 ml of fresh media , cultures were grown at 37°C , 225 rpm for approximately 3 hours to an OD600 of approximately 0 . 6 , then induced by the addition of L-arabinose to a final concentration of 0 . 02% . After approximately 2 hours , cultures were centrifuged for 5 minutes ( 4°C , 10 , 000×g ) , and the resulting pellet saved and frozen at −80°C for subsequent processing . Pellets were thawed on ice , and resuspended in binding buffer containing 20 mM Tris , 8 M Urea , 500 mM NaCl , 5 mM imidazole , protease inhibitor cocktail ( 1× , Sigma P8465 ) and PMSF ( 1 mM ) , at pH 8 . 0 . After lysis on a rotator for 30 minutes , the suspension was centrifuged for 10 minutes at 10 , 000×g at 4° C . Recombinant polyhistidine-tagged autotransporter proteins ( rATp ) were then purified from clarified lysates using immobilized metal affinity chromatography ( IMAC ) in small-scale ( His SpinTrap columns ( GE Healthcare ) or in larger scale by using Ni-Sepharose columns ( HisTrap HP , GE Healthcare ) . Proteins were eluted over an imidazole gradient produced on a low-pressure chromatography system ( BioLogic LP , BioRad ) . Purity of recombinant proteins was assessed by SDS-PAGE . Where necessary , fractions containing the protein of interest were further purified by ion exchange ( HiTrap Q ) or by SDS-PAGE and subsequent electroelution ( Mini Whole Gel Eluter , BioRad ) . Polyclonal rabbit antisera were produced as previously described ( Rockland Immunochemicals , Gilbertsville , PA ) . Mouse polyclonal antibodies were obtained upon sacrifice of mice following completion of intranasal immunization with either adjuvant control or adjuvant and recombinant protein of interest as described below . Antibodies were affinity-purified as previously described [13] , [34] . Briefly , antibodies were absorbed onto immobilized antigen of interest on nitrocellulose , then eluted in 100 mM glycine , pH 2 . 5 followed by neutralization with 1 M Tris , pH 8 . 0 . To detect polyhistidine-tagged rATp proteins transferred onto nitrocellulose , blocking was performed for one hour in 5% skim milk in tris-buffered saline ( pH 7 . 2 ) containing 0 . 05% Tween-20 ( TBS-T ) , and subsequent immunoblotting used purified rabbit polyclonal anti-His serum ( 1∶500 ) and goat anti-rabbit immunoglobulin G ( Fc ) -Horseradish peroxidase ( HRP ) ( 1∶10 , 000 ) . All wash and incubation steps were performed in Tris-Buffered Saline ( pH 7 . 2 ) - 0 . 05% Tween 20 . Detection used luminol-based chemiluminescent substrate . Assessment of immune responses to individual rATp proteins was carried out in a similar fashion using pooled primary convalescent sera ( 1∶500 ) from mice following infection with ETEC strain H10407 . Pooled , pre-immune sera ( 1∶500 ) from the same mice was used as a primary antibody control . Sera obtained from ETEC-infected children ( ICDDR , B ) Dhaka , Bangladesh , and uninfected ( age-matched ) controls ( LeBonheur Children's Medical Center , Memphis , TN ) , were tested at a 1∶2048 dilution . Primary antibody binding was detected using an HRP-labeled secondary ( anti-mouse or anti-human ) antibody that detects IgA , IgM and IgG ( total Ab ) and developed using sensitive chemiluminescent substrate ( SuperSignal West Femto , Thermo Scientific ) . Kinetic ELISA assays were performed as previously described [24] . ( For an in-depth discussion comparing kinetic and standard end-point ELISA techniques , the reader is referred to an earlier review available at http://www . biotek . com/resources/docs/KineticAppNoteFinal . pdf ) [35] . ELISA wells were coated overnight at 4°C with ( 100 µl/well ) of individual rATp proteins diluted to a final concentration of 4 µg/ml in 0 . 1 M NaHCO3 buffer ( pH 8 . 6 ) . Plates were washed three times with TBS-T , blocked with TBS-T containing 1% BSA ( Blocker , Thermo Scientific ) for 1 hour at 37°C . After washing briefly with TBS-T , plates were incubated with dilutions of primary antibody in TBS-T/1% BSA for 1 hour at 37°C , and again washed as above . Next , plates were incubated ( 1 h , 37°C ) with goat anti-human or goat anti-mouse secondary antibody , which recognizes IgA , IgM , and IgG , at a final concentration of 1∶10000 in TBS-T-BSA . After washing , plates were developed with TMB ( 3 , 3′ , 5 , 5′-Tetramethylbenzidine ) peroxidase substrate ( Kirkgaard and Perry Laboratories ) . Kinetic absorbance measurements [36] were determined at 650 nm , acquired at 60 second intervals ( Molecular Devices Spectramax 340PC microplate reader ) . SoftMax Pro v5 . 0 . 1 was used for data recording and processing and the rate of substrate development was expressed as the Vmax in milli-units/min . Mice were vaccinated intranasally with candidate antigens as previously described [37] . Briefly , groups of 10 ICR [38] mice received either IVX908 ( Protollin® ) [39] ( 7 . 5 µg ) alone ( controls ) , or IVX908 ( 7 . 5 µg ) + rATp domains ( 20 µg ) on days 0 , 14 , 28 . Sera were collected on day 0 prior to immunization ( pre-immune ) and again 7–14 days after the final vaccination ( post-immune ) . Prior to bacterial challenge ( day 42 ) , 5–6 fresh fecal pellets were collected from each mouse and immediately suspended in 1 . 5 ml of extraction buffer containing Tris ( 10 mM ) , NaCl ( 100 mM ) , Tween-20 ( 0 . 05% ) , and sodium azide ( 5 mM ) pH 7 . 4 . Approximately 1 week after completion of the immunization schedule , mice were challenged with jf876 , a derivative of ETEC H10407 containing a KmR marker in the lacZ gene as previously described [37] , [40] . Two independent challenge studies were completed; each involved a total of 30 mice including 10 controls ( IVX908 only ) , and groups of 10 mice vaccinated with either of the two rATp proteins . Briefly , mice were treated with streptomycin ( 5 g/liter ) in their drinking water to eliminate colonization resistance from competing normal enteric flora . Food was then withheld 12 hours prior to ETEC challenge and sterile water was used to replace the streptomycin solution . All mice received cimetidine ( 50 mg/kg via intraperitoneal injection ) 2 hours prior to administration of bacteria . Mice were then challenged by gavage administration of bacterial strain jf876 with total dose between 104 and 105 cfu/mouse . Previous studies have demonstrated that at 24 hours after inoculation with ETEC in the murine model , the number of colonizing colony forming units in the small intestine parallel values obtained at later time points ( 72 hours ) [38] . Therefore , in these studies , approximately 24 hours after challenge mice were sacrificed , samples of ileum were harvested , then solubilized in saponin solution to release bacteria , and finally plated onto Luria agar plates containing kanamycin ( 25 µg/ml ) . Colonies were counted after incubation overnight at 37°C . Intestinal lysates with no bacteria recovered were assigned a value of 1 cfu ( the lower limit of detection ) as previously described [37] . Care was taken to minimize the use of animals in these experiments where possible . When necessary , appropriate steps were taken to ameliorate suffering of mice during vaccination and testing . The Institutional Animal Care and Use Committees of the University of Tennessee Health Sciences Center , and the VA Medical Center approved studies presented here . All procedures involving mice complied with Public Health Service guidelines , and The Guide for the Care and Use of Laboratory Animals . To evaluate surface expression of autotransporter passenger domains , suspensions of H10407 in phosphate buffered saline ( PBS , pH 7 . 2 ) were first fixed with 2% paraformaldehyde for 15 minutes . After washing twice with PBS , cells were blocked with 1% BSA in PBS for 30 minutes . The resulting cell suspension was incubated with either pre- or post-immune mouse sera ( diluted 1∶50 in blocking buffer ) for 1 hour at room temperature ( RT ) . After washing three times with PBS , cells were incubated with Alexa-Fluor ( 488 ) -labeled anti-mouse IgG [1∶250] for 1 hour at RT , washed three times , and resuspended in PBS . Analysis of cell-bound fluorescence by flow cytometry used a BD FACSCalibur 4-color , dual-laser flow cytometer equipped with a FACStation data management system .
In silico analysis of available ETEC genomes using searches for one or more of the known highly conserved AT domains identified multiple autotransporters in the genome of each ETEC strain . From these , we selected AT genes that were present in ETEC , but absent in the recently sequenced E . coli HS commensal isolate [41] , for additional study ( table 3 ) . All three genomes shared at least one copy of both antigen43 , and the pAT autotransporter genes ( figure 1 ) . Similar to other pathovars [42] , we identified two copies of the agn43 gene in both H10407 , and E24377A , but a single copy in B7A . The amino acid sequence of the passenger domains of these Ag43 molecules was highly conserved ( supporting figure S1a ) , while the pAT passenger domains exhibited approximately 70% identity and 80% similarity ( supporting figure S1b ) . Although genes encoding proteins similar to pAT were found in the Enterobacteriaceae by BLAST searches of the prototype pAT molecule from H10407 ( Uniprot designation E3PFJ1 ) in the UniprotKB database , and were widely distributed among various pathovars of E . coli as well as Salmonella , and Shigella species ( taxonomic distribution of these proteins is included in supporting figure S2 ) , only those in E . coli exhibited more than 80% identity . As demonstrated in table 4 , pAT homologues were identified in other diarrheagenic E . coli pathotypes including enterohemorrhagic strains ( EHEC ) , enteropathogenic isolates ( EPEC ) , and uropathogenic E . coli ( UPEC ) , and extraintestinal pathogenic E . coli ( ExPEC ) isolated from meningitis . However , we also identified potential pAT homologues in non-HS commensal isolates from humans [43] as well as animals [44] , suggesting that while pAT may not be essential for commensalism , it is not unique to pathogenic strains . Passenger domains of autotransporter proteins are exported to the bacterial surface where they may be cleaved and appear in the supernatant as in the case of EatA [14] , or remain largely associated with bacterial cell surface as has been described for antigen 43 [42] . Antibodies raised against the recombinant passenger domains of Ag43 , pAT , and EatA were tested in kinetic ELISA assays against the corresponding antigen as well as each of the other rATp domains . These studies revealed little or no cross-reactivity between antigens , suggesting that these proteins are immunologically distinct ( figure 2a ) . Flow cytometry detected significant amounts of passenger domains of Ag43 and pAT associated with the bacterial cell surface of ETEC H10407 , but not with the HS commensal strain ( figure 2b ) . Likewise , pAT ( figure 2c ) and Ag43 ( figure 2d ) were identified on the surface of both the E24377A , and B7A ETEC strains . Because passenger domains of autotransporter proteins are exported to the bacterial surface , they often elicit an immunologic response in the host during the course of infection [45] , [46] , [47] . Furthermore , recent immuno-proteomic studies have indicated that autotransporters are recognized during the course of experimental infection in mice [26] . Therefore , we examined the immune responses to conserved , chromosomally encoded autotransporter proteins identified in ETEC , using sera obtained from mice following experimental infection and from humans following natural ETEC infections . Immunoblotting studies of both of the H10407 autotransporters tested H10407_Ag43 , and H10407_pAT , demonstrated that the passenger domain of both of these antigens is recognized during the course of experimental intestinal infection in mice ( figure 3 a ) . This was also true of TibA and EatA passenger domains ( not shown ) as predicted by earlier proteomic studies , however we chose to focus here on the conserved , chromosomally encoded antigens , Ag43 and pAT . Subsequent studies demonstrated that both Ag43 and pAT are recognized during the course of human infections with ETEC ( figure 3 b , c ) . To explore the utility of ETEC autotransporter proteins as vaccine candidates , we examined both the immunogenicity and protective efficacy of individual passenger domains for Ag43 and pAT in a murine model of ETEC infection . For these studies , we used the mucosal adjuvant Protollin ( ivx908 ) , a mixture of Shigella flexneri 2a LPS , and meningococcal outer membrane proteins . When delivered intranasally , Protollin ( IVX908 ) elicits high levels of S . flexneri LPS-specific fecal IgA [48] , and when combined as an adjuvant with other proteins , it promotes similarly robust immune responses to target antigens [39] , [49] . Immunization with either of the passenger domains for antigen 43 or pAT resulted in significant increases in total serum and fecal antibody levels in immunized mice relative to adjuvant-only ( ivx908 ) controls ( figure 4a–c ) . Importantly , immunization with either rPATp or rAg43p resulted in significant increases in fecal IgA directed at the respective antigens ( figure 4d ) . Immunization with either the antigen43 or pAT recombinant passenger domains provided significant protection against subsequent colonization with ETEC relative to adjuvant-immunized controls ( figure 4e ) .
Despite the global importance of enterotoxigenic E . coli infections , there is presently no vaccine against these pathogens that would offer sustained , broad-based protection [50] . Vaccine development for ETEC is a challenge for a number of reasons . First , the inherent plasticity of E . coli genomes makes discovery of conserved , pathotype-specific antigens difficult [51] , [52] . In addition , much of the ETEC vaccine development effort to date has focused on the plasmid-encoded colonization factors ( CFs ) . Unfortunately , antigenic heterogeneity and lack of appreciable cross protection between CFs have been impediments to this approach . To date , over twenty-five different CFs have been identified in ETEC [53] , and many strains do not appear to make any of the known antigens [9] , [54] . Finally , carefully conducted epidemiologic studies of natural ETEC infections have suggested that LT and perhaps other as yet unidentified chromosomally-encoded antigens [29] , in addition to the plasmid-encoded CFs , could be involved in protective immune responses . Because we have recently demonstrated that several autotransporter proteins are recognized following experimental and natural ETEC infections [26] , we chose to investigate the possible contribution of conserved chromosomally-encoded autotransporter proteins . The studies here suggest that the passenger domains of these autotransporters are recognized during the course of both experimental infections in animals and naturally-occurring infection in humans , and they validate recent immunoproteomic data obtained with the prototype H10407 ETEC strain using sera from infected mice or human convalescent sera [26] . Two additional autotransporters have previously been examined in ETEC pathogenesis . These include the chromosomally-encoded TibA adhesin [55] , [56] protein and , EatA [14] , a plasmid-encoded member of the SPATE family ( serine protease autotransporters of the Enterobacteriaceae ) . Both proteins are recognized during the course of experimental and naturally occurring infections [26] . Interestingly , the EatA protein appears modulate adhesion and colonization by digesting another recently described virulence molecule , the EtpA two-partner secretion exoprotein , an adhesin [12] , [13] . In turn , this modulation of adherence appears to be required for optimal delivery of heat-labile toxin ( LT ) , a critical ETEC virulence molecule [57] . Although recent data suggest that both EtpA and EatA are reasonably conserved within the ETEC pathovar [52] , [58] , [59] , the inherent plasticity of E . coli genomes , and the relative paucity of pathovar-specific virulence genes [51] identified to date suggests that additional effort is warranted to explore the potential utility of other highly conserved surface structures as vaccine candidates . Although it is likely that autotransporters contribute the overall fitness of ETEC as a pathogen , neither of the proteins under study here has been shown to contribute to the pathogenesis of ETEC . Antigen 43 does however appear to be of importance in the pathogenesis of other E . coli pathovars , including uropathogenic E . coli ( UPEC ) . In the urinary tract , Ag43 is expressed in intracellular biofilm-like pods [60] , and particular variants appear to contribute to biofilm formation , and colonization of the urinary tract [61] . Similarly , immunoproteomic studies demonstrate that this antigen is also expressed [62] and recognized [63] during the course of E coli urinary tract infections in humans . Interestingly , in a study of an E . coli laboratory isolate , Ag43 contributed to biofilm formation in vitro , but did not appear to play a role in intestinal colonization in a murine model [64] . Nevertheless , some studies have suggested that specific Ag43 alleles segregate with diarrheagenic E . coli pathogens compared to other isolates from other pathovars [65] , and that in general Ag43 was more commonly found in pathogens than in commensal strains . Assessing the precise contribution of given antigens to the protective immune responses that develop following infection , or even following vaccination can be challenging [66] , [67] . While serologic responses to some CFs such as CFA/I have previously been correlated with a protective immune response to ETEC [28] , it is likely that protection seen following natural infections reflects a composite response to a number of antigens . Additional studies will be needed to determine the utility of these antigens as well as other autotransporters in ETEC vaccines . The surface expression of the autotransporter passenger domains , their immunogenicity , and preliminary data presented here support the concept that this class of molecules could serve as protective antigens . Although the inherent plasticity of E . coli genomes [68] in general poses an impediment to vaccine development for ETEC , important data emerging from the DNA sequencing of multiple ETEC genomes does suggest that these pathogens maintain a core subset of relatively pathovar-specific genes , such as the eatA autotransporter gene , that might serve as suitable targets [52] , [59] . The suggestion that relatively few genes separate the ETEC pathovar from commensal E . coli [58] is an important consideration in moving forward with putative ETEC vaccines . The data presented here suggest that other autotransporters not unique to the ETEC pathovar contribute to intestinal colonization , a critical step in ETEC pathogenesis as well as the host immune response to these important pathogens . Whether the more widely distributed ATs such as Ag43 and pAT are truly dispensable for the commensal E . coli similar to the HS prototype strain will be an important consideration in designing both subunit and live-attenuated vaccine strategies . | Diarrheal diseases are responsible for more than 1 . 5 million deaths annually in developing countries . Enterotoxigenic E . coli ( ETEC ) are among the most common bacterial causes of diarrhea , accounting for an estimated 300 , 000–500 , 000 deaths each year , mostly in young children . There unfortunately is not yet a vaccine that can offer sustained , broad-based protection against ETEC . While most vaccine development effort has focused on plasmid-encoded finger-like ETEC adhesin structures known as colonization factors , additional effort is needed to identify conserved target antigens . Epidemiologic studies suggest that immune responses to uncharacterized , chromosomally encoded antigens could contribute to protection resulting from repeated infections . Earlier studies of immune responses to ETEC infection had identified a class of surface-expressed molecules known as autotransporters ( AT ) . Therefore , available ETEC genome sequences were examined to identify conserved ETEC autotransporters not shared by the commensal E . coli HS strain , followed by studies of the immune response to these antigens , and tests of their utility as vaccine components . Two chromosomally encoded ATs , identified in ETEC , but not in HS , were found to be immunogenic and protective in an animal model , suggesting that conserved AT molecules contribute to protective immune responses that follow natural ETEC infection and offering new potential targets for vaccines . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"immunizations",
"applied",
"microbiology",
"medicine",
"infectious",
"diseases",
"emerging",
"infectious",
"diseases",
"immunity",
"global",
"health",
"immunology",
"biology",
"microbiology",
"immune",
"response"
] | 2011 | Directed Evaluation of Enterotoxigenic Escherichia coli Autotransporter Proteins as Putative Vaccine Candidates |
Elimination of the proliferating germline extends lifespan in C . elegans . This phenomenon provides a unique platform to understand how complex metazoans retain metabolic homeostasis when challenged with major physiological perturbations . Here , we demonstrate that two conserved transcription regulators essential for the longevity of germline-less adults , DAF-16/FOXO3A and TCER-1/TCERG1 , concurrently enhance the expression of multiple genes involved in lipid synthesis and breakdown , and that both gene classes promote longevity . Lipidomic analyses revealed that key lipogenic processes , including de novo fatty acid synthesis , triglyceride production , desaturation and elongation , are augmented upon germline removal . Our data suggest that lipid anabolic and catabolic pathways are coordinately augmented in response to germline loss , and this metabolic shift helps preserve lipid homeostasis . DAF-16 and TCER-1 also perform essential inhibitory functions in germline-ablated animals . TCER-1 inhibits the somatic gene-expression program that facilitates reproduction and represses anti-longevity genes , whereas DAF-16 impedes ribosome biogenesis . Additionally , we discovered that TCER-1 is critical for optimal fertility in normal adults , suggesting that the protein acts as a switch supporting reproductive fitness or longevity depending on the presence or absence of the germline . Collectively , our data offer insights into how organisms adapt to changes in reproductive status , by utilizing the activating and repressive functions of transcription factors and coordinating fat production and degradation .
Organisms are constantly adapting to fluctuations in their internal and external milieus by employing transcriptional , translational and endocrine mechanisms that sense varying stimuli and activate or repress different cellular processes in response . But , how metazoans maintain metabolic homeostasis when faced with multidimensional challenges such as increasing age , altered reproductive status or competing physiological demands is poorly understood . Failure to execute metabolic adaptability under such conditions has come to be seen as the underlying cause of a host of human pathologies including age-related diseases such as diabetes and metabolic syndrome [1 , 2] . The nematode Caenorhabditis elegans has proven to be a versatile platform for investigating aging and its modulation by various factors ranging from insulin IGF1 signaling ( IIS ) to dietary intake and reproduction . In C . elegans , germline loss increases lifespan and improves stress resistance [3 , 4] . Similar observations have been made in other animals , including flies [5] , mice [6] , rats [7] , grasshoppers [8] and salmon [9] , and in some human population studies as well [10] , suggesting that the reproductive control of aging may be conserved . But , this longevity is not just a byproduct of sterility because lifespan is extended only upon the removal of a specific population of totipotent germline-stem cells ( GSCs ) that are the precursors of the entire adult germline [5 , 11] . GSCs must produce signals that coordinate reproductive status with rate of aging; upon their removal , the animal not only copes with the challenge of fertility loss , it also successfully reestablishes metabolic homeostasis and converts the drawback into a favorable lifespan increment . Mechanisms underlying this remarkable adaptability are poorly understood . Transcriptional changes facilitated by modular gene-regulatory networks ( GRNs ) constitute a major mechanism by which animals respond to changing environmental conditions . In C . elegans , removal of GSCs in the gonad activates a network of conserved transcription factors in the intestine- a tissue that serves as the main adipose depot in worms and subsumes roles of the liver and pancreas . This network includes DAF-16/FOXO3A [12] , PHA-4/FOXA [13] , HLH-30/TFEB [14] , SKN-1/NRF2 [15 , 16] , HSF-1/HSF1 [17] and nuclear hormone receptors ( NHR ) , DAF-12/VDR [4] , NHR-80/HNF4 [18] and NHR-49/ PPARα [19] . Many of these proteins , including DAF-16 , are shared longevity determinants that alter lifespan in response to other interventions such as reduced IIS or dietary restriction ( DR ) [14 , 20–22] . Previously , we identified TCER-1 , the worm homolog of a human transcription elongation and splicing factor , TCERG1 [23] , and showed that it specifically increases lifespan following germline loss , likely by facilitating a distinct pattern of DAF-16-dependent gene expression [24] . While DAF-16 targets have been identified in multiple contexts [25] , little is known about downstream effectors of TCER-1 . Additionally , studies so far have largely focused on genes upregulated following germline depletion , although , it is conceivable that adaptation to GSC loss also necessitates the downregulation of some genes . DAF-16 , like most transcription factors , both activates and represses transcription , but its negative targets are not well characterized . Transcription-repression functions of TCERG1 have been documented [26] but genes repressed by TCER-1 are unknown . Thus , the elements of gene-expression networks orchestrated by DAF-16 and TCER-1 in GSC-less animals , the extent of their overlap with one another , and the molecular pathways governed by them are as yet unidentified . Germline removal is accompanied by the enhancement of cellular processes that govern macromolecular homeostasis including proteasomal activity and autophagic flux [13 , 14 , 16] . Several lines of evidence have also indicated a crucial role for lipid metabolism in this longevity paradigm [27–30] . The lifespan extension of GSC-less animals is dependent on multiple components of a steroid signaling pathway that culminates in the production of bile acid-like steroids called dafachronic acids ( DA ) that control the activity of the NHR , DAF-12 [31] . NHR-80 mediates the increased expression of fatty-acid desaturases [18] , and in a recent study , we demonstrated the importance of NHR-49/ PPARα in elevating fatty-acid β-oxidation and desaturation to promote longevity of germline-depleted worms [19] . Germline-less worms display elevated lipid levels [32] , but paradoxically , a lipase , LIPL-4 , has been found to be essential for their long lifespan [33] , and the molecular basis of their dramatic fat accrual remains unknown . LIPL-4 promotes autophagy in germline-less mutants [13] and lipid signals appear to mediate SKN-1 activation [15] , but overall , little is understood of the broader significance of lipid-metabolic pathways in the lifespan increment associated with germline depletion . In this study , we set out to explore the GRNs governed by DAF-16 and TCER-1 following germline loss by identifying genes whose expression is modulated by the two factors . RNA-Sequencing ( RNA-Seq ) -based transcriptome mapping revealed that these proteins shared a significant fraction of their downstream targets . In studying these targets , we discovered that DAF-16 and TCER-1 augmented expression of genes involved in both lipid anabolism and catabolism . These gene-expression changes are biochemically and functionally relevant because our lipidomic analyses demonstrated that lipogenic processes are enhanced upon germline loss . Previous reports have shown increased lipase activity in these animals , and we found both classes of genes are important for germline-less longevity . Our observations suggest that germline loss may trigger the simultaneous enhancement of antagonistic lipid-metabolic pathways and this may enable adaptation to loss of reproductive capacity . We also found that DAF-16 and TCER-1 performed important repressive functions . Upon germline depletion , TCER-1 blocked the expression of somatic genes that facilitate reproduction as well as anti-longevity genes . DAF-16 suppressed translation by inhibiting the expression of genes encoding ribosomal subunits . In addition , we discovered an unexpected requirement for tcer-1 in ensuring the reproductive fitness of normal , fertile adults . Overall , these experiments provide mechanistic insight into how an organism adapts to change in reproductive status and maintains metabolic equilibrium by modulating anabolic and catabolic processes .
The lifespan extension resulting from germline elimination in C . elegans is faithfully simulated by mutations that cause GSC loss and sterility . One such temperature-sensitive mutant , glp-1 ( e2141ts ) , has been used widely as a model for this longevity [11 , 24] . To map the transcriptomes governed by DAF-16 and TCER-1 in germline-ablated , long-lived animals , we performed RNA-Seq on three strains ( i ) glp-1 ( e2141ts ) ( ii ) daf-16 ( mu86 ) ;glp-1 ( e2141ts ) and ( iii ) tcer-1 ( tm1452 ) ;glp-1 ( e2141ts ) ( henceforth referred to as glp-1 , daf-16;glp-1 and tcer-1;glp-1 , respectively ) . These comparisons allowed us not only to document the transcriptional changes mediated by DAF-16 and TCER-1 upon GSC ablation , but also to assess the overlap between their targets . Since many molecular landmarks associated with germline-less longevity are observed once the animal becomes an adult [12 , 24] , and since DAF-16 and TCER-1 both act during adulthood to promote glp-1 longevity [11] , we isolated RNA from day 2 adults grown under identical conditions . Sequencing data was analyzed using the publicly available Galaxy pipeline running Tuxedo Suite tools [34] ( S1A Table; see Methods for details ) . We found that DAF-16 and TCER-1 mediated the transcriptional upregulation and downregulation of overlapping groups of target genes , henceforth designated UP and DOWN genes , respectively . The global comparison of glp-1 and daf-16;glp-1 identified 801 genes that were differentially expressed ( ‘DAF-16 Targets’ ) ( Fig 1A and 1B; S1B , S1C and S1F–S1H Table ) and comparison of glp-1 and tcer-1;glp-1 revealed 835 such genes ( ‘TCER-1 Targets’ ) ( Fig 1A and 1B; S1D–S1H Table ) . 256 genes were shared between these two sets , constituting a newly identified group of genes regulated by both TCER-1 and DAF-16 ( ‘Joint Targets’ ) ( Fig 1A; S1F–S1H Table ) . Of these ‘Joint Targets’ , 123 genes were upregulated in both mutants ( ‘Joint UP’ , S1F Table ) , 73 were downregulated ( ‘Joint DOWN’ , S1G Table ) and 60 showed opposite regulation ( S1H Table ) . The overlap between DAF-16 and TCER-1 targets was 5 . 2 times greater than that expected by random chance and comprised significant portions of the total transcriptomes dictated by each regulator ( DAF-16 , 31 . 9% and TCER-1 , 30 . 7%; p<2 . 041e-88 ) . These findings are consistent with our earlier data that suggested that DAF-16 and TCER-1 function in the same genetic pathway to mediate the reproductive control of aging [24] . There has been limited agreement between DAF-16 targets identified previously through other genomic approaches [25] . Through a meta-analysis of DAF-16-related genomic studies , Tepper et al . , have arrived at a ‘Consensus’ list of DAF-16 positive and negative targets ( corresponding to our UP and DOWN classes , respectively ) [35] . We compared the DAF-16 targets obtained from our RNA-Seq analysis with these lists and found 162/349 DAF-16-Specific UP genes ( 46 . 4% , p<9 . 4e-80 ) and 70/123 Joint UP genes ( 56 . 9% , p<1 . 46e-42 ) to be shared with the DAF-16 ‘Consensus’ positive targets ( S2 Table ) . Of the genes repressed by DAF-16 , 110/196 DAF-16-Specific DOWN genes ( 56 . 1%; p<1 . 9e-63 ) and 41/73 Joint DOWN genes ( 56 . 1% , p<2 . 1e-24 ) were represented in the DAF-16 ‘Consensus’ negative target list ( see S2 Table for details and additional comparisons ) . This degree of overlap is striking in the light of the dissonance observed between various DAF-16-target compilations [25] . In a recent analysis of individual DAF-16 isoforms , Chen et al . , identified 399 targets of daf-16a , the transcript with the strongest impact on germline-less longevity [36] . 58 of DAF-16-Specific and 62 of the Joint UP targets were shared with this list ( 15%; p< 1 . 797e-59 ) ( S2 Table ) . McCormick et al . , reported the identification of 230 ‘DAF-16-regulated’ genes in a glp-1 background [37] . 85 of these genes were represented among our DAF-16 targets ( 37%; p< 7 . 840e-55 ) . In contrast , only 10 were included in the TCER-1 list ( 4%; p = 0 . 2 ) ( S2 Table ) . Together , these comparisons reinforced our confidence in the newly-identified DAF-16 targets as well as the TCER-1 downstream genes . It is noteworthy that a majority of the DAF-16 targets we identified were novel , particularly in the glp-1 background , underscoring the sensitivity of our approach . Based on previous observations , we reasoned that the expression of UP genes is likely to be elevated upon germline removal . Indeed , when we compared the mRNA levels between wild type and glp-1 adults using quantitative PCRs ( Q-PCRs ) , 26/30 UP genes tested showed elevated expression in glp-1 mutants ( eight did not achieve statistical significance ) and 22 of these depended on completely or partially DAF-16 and/or TCER-1 for their upregulation ( Q-PCRs in this study ) . Following this substantiation of the RNA-Seq data , we asked if the increased expression of the UP genes was essential for glp-1 mutants’ longevity . Using feeding RNAi , the levels of DAF-16-Specific , TCER-1-Specific and Joint UP targets were reduced in glp-1 mutants . To obviate developmental phenotypes , RNAi was initiated on the first day of adulthood . Based on the fold change in expression , we tested the top 80 candidates and found that RNAi of 64 genes caused a statistically significant suppression of glp-1 mutants’ longevity and 54 of these clones suppressed lifespan in at least two independent trials ( Fig 1C and S3 Table ) . This group included genes encoding multiple transcription factors with roles in lifespan regulation ( eg . , nhr-49 , hlh-30 , skn-1 ) [14 , 15 , 19] , previously-identified DAF-16 targets ( eg . , mtl-1 ) [38] as well as novel genes that implicated new cellular processes in influencing lifespan ( see Discussion ) . Overall , these experiments demonstrated that our RNA-Seq analysis had identified DAF-16 and TCER-1 targets whose expression was elevated following germline removal and that were important for the consequent longevity . We performed DAVID analysis [39 , 40] on the RNA-Seq data ( S4 Table ) and used REVIGO [41] to visualize the enriched GO term classes and their semantic similarities in two-dimensional space ( S5 Table ) . DAF-16-Specific , TCER-1-Specific and Joint UP classes revealed a shared enrichment of fatty acid and lipid metabolism genes ( Fig 1B; S4A , S4B , S4E , S4F , S4I and S5A , S5C and S5E Tables ) . A detailed examination of these gene lists revealed that ( a ) specific lipid-metabolic pathways were enriched in this data set and ( b ) these pathways included both lipid anabolic {initiation of fatty-acid synthesis , desaturation and elongation , and conversion of diglycerides ( DAG ) to triglycerides ( TAG ) } and catabolic ( TAG hydrolysis and fatty-acid β-oxidation ) processes ( Fig 2 ) . For instance , fifteen of the thirty genes predicted to collectively mediate fatty-acid synthesis and TAG production were represented in our RNA-Seq dataset ( 50%; p< 8 . 892e-06 ) , fourteen of which were UP genes ( Fig 2A ) . Similarly , of the 98 genes together predicted to be involved in TAG hydrolysis and fatty-acid β-oxidation , 24 were identified by RNA-Seq ( 25% , p<1 . 8e-4 ) and 17 of these were UP genes ( Fig 2B ) . The over-representation of genes involved in fat buildup was particularly noteworthy and has not been reported before . It prompted us to investigate the role of DAF-16 and TCER-1 in modulating lipid-metabolic pathways . The first , and rate-limiting , step in the initiation of fatty-acid synthesis is mediated by the enzyme acetyl CoA carboxylase ( ACC; encoded in C . elegans by pod-2 ) ( Fig 2A ) [42] . POD-2 catalyzes the synthesis of malonyl CoA ( MCA ) , a critical intermediate that provides two-carbon units for the generation of the fatty acid palmitate ( C16:0 ) , the substrate for fatty-acid synthase ( FASN-1 , encoded by fasn-1 ) . MCA levels are determined by the opposing activities of POD-2 and malonyl CoA decarboxylase 1 ( MLCD-1; encoded by mlcd-1 ) [42] , an enzyme that converts it back to acetyl CoA [43 , 44] . Both pod-2 and mlcd-1 were identified as DAF-16-Specific UP targets in our RNA-Seq data ( S1B Table ) . Q-PCRs showed that the expression of both genes was enhanced in glp-1 mutants and this upregulation was completely ( mlcd-1 ) or partially ( pod-2 ) impaired in daf-16;glp-1 mutants ( Fig 3A and 3B ) . fasn-1 , although not picked up in the RNA-Seq dataset , showed a similar trend towards increased expression in glp-1 mutants ( S1A Fig ) . Based on these gene-expression changes we asked if germline ablation was accompanied by increased de novo fatty-acid synthesis . Using a 13C isotope fatty-acid labeling assay [45] , we first compared de novo fatty-acid synthesis and absorption between wild-type worms and long-lived glp-1 mutants . The eggs and embryos contained within the gonads of fertile adults confound lipid estimation hence this comparison could not be undertaken in day 2 adults; instead we used late L4/early Day 1 young adults before they became reproductively active . glp-1 mutants exhibited a significant increase in de novo fatty-acid synthesis as compared to their fertile counterparts in both the neutral and phospholipid fractions ( Fig 3C and 3D ) . To examine the contribution of DAF-16 and TCER-1 to this process , we compared glp-1 , daf-16;glp-1 and tcer-1;glp-1 day 2 adults in the same assay . DAF-16 loss significantly reduced de novo fatty-acid synthesis and tcer-1 did not have a major impact , in accordance with the RNA-Seq and Q-PCR data ( Fig 3E and 3F ) . Next we addressed the functional relevance of this metabolic shift by examining the effect of inactivating these genes on glp-1 mutants’ longevity . We introduced , fasn-1 ( fr8 ) , a partial loss-of-function allele of fasn-1 [46] , into the glp-1 background and found that it completely abrogated their longevity , whereas , fasn-1 single mutant did not live significantly shorter than wild-type worms ( Fig 3G and S6 Table ) . Adult-specific RNAi knockdown of pod-2 and mlcd-1 produced modest lifespan reductions ( three of five trials and two of four trials , respectively; Fig 3H and 3I and S7 Table ) . fasn-1 RNAi suppressed longevity consistently ( Fig 3H and S7 Table ) . Together , these experiments revealed that germline loss triggers increased de novo fatty acid synthesis mediated by DAF-16 , and implied that this enhancement is important , at least in part , for the subsequent lifespan extension . Fatty acids constitute the building blocks for the production of neutral lipids that are stored in the form of TAGs . The key step in this pathway is the conversion of DAGs into TAGs , catalyzed by the enzyme diacyl glycerol acyl transferase ( DGAT ) [47] . In C . elegans , six genes are predicted to encode DGATs [48] . Strikingly , four of these were identified in our RNA-Seq analysis as UP genes , being upregulated either by DAF-16 ( mboa-2 ) , TCER-1 ( dgat-2 and acs-22 ) or both ( Y53G8B . 2 ) ( Fig 2 and S1 Table ) . Q-PCRs confirmed the upregulation of all four genes in glp-1 mutants , and at least partially in a DAF-16- and/or TCER-1-dependent fashion ( Fig 4A; acs-22 expression was reduced in tcer-1;glp-1 mutants , but did not achieve statistical significance ) . The fifth gene , K07B1 . 4 , was identified in our previous study as a common target of DAF-16 and TCER-1 [24] ( the sixth , dgtr-1 , is predicted to be germline-restricted and thus would not have been detected in these experiments with germline-less mutants ) . Thus , the expression of a majority of the rate-limiting enzymes that catalyze the conversion of DAGs into TAGs was elevated upon germline loss by DAF-16 and/or TCER-1 . We next explored the biochemical ramifications of this increase in ‘dgat’ expression . First , fat content and distribution were examined using the lipid-labeling dye Oil Red O ( ORO ) [32] . In agreement with previous observations [32] , young glp-1 adults exhibited significantly higher levels of ORO staining than age-matched , wild type controls ( Fig 4B–4E and 4J ) . We found that this increase was significantly attenuated in daf-16;glp-1 mutants by day 2 of adulthood ( Fig 4F and 4G and 4J ) , unlike a recent study that examined day 1 adults and reported an insignificant effect of the daf-16 mutation on glp-1 ORO levels [15] . We also discovered that the absence of DAF-16 caused a progressive depletion of intestinal fat with age . By day 8 of adulthood , ORO staining was markedly reduced in daf-16;glp-1 mutants ( Fig 4J ) . In addition , daf-16;glp-1 adults exhibited ectopic fat deposition in other tissues such as body wall muscles ( arrowheads in 4F , G ) . Surprisingly , tcer-1;glp-1 mutants did not show any obvious changes in ORO staining or TAG levels ( Fig 4H–4J ) . To verify the ORO observations , we used gas chromatography/mass spectrometry ( GC/MS ) and found TAG levels to be noticeably lower in daf-16;glp-1 mutants , as compared to glp-1 ( Fig 4K ) . tcer-1;glp-1 mutants , however , showed no significant difference from glp-1 ( Fig 4K ) . The reason for this discrepancy is unclear , but it is possible that this may reflect the partial loss of function nature of the tcer-1 ( tm1452 ) allele , unlike daf-16 ( mu86 ) which is a null . To evaluate the functional significance of these gene-expression and biochemical changes , we asked if the ‘dgat’ genes were essential for glp-1 mutants longevity . RNAi inactivation of four of the ‘dgat’ genes that we tested caused modest reduction in glp-1 mutants’ longevity but showed variability between trials . For instance , dgat-2 and Y53G8B . 2 RNAi suppressed longevity significantly in three out of four trials , K07B1 . 4 in two of three trials and acs-22 in one of three trials ( Fig 5A and S7 Table ) . We also introduced loss-of-function mutations of dgat-2 and acs-22 [48] into the glp-1 background . Neither dgat-2;glp-1 nor acs-22;glp-1 lived shorter than glp-1 mutants alone when fed the normal E . coli OP50 diet ( Fig 5B and 5C and S6 Table ) although ORO staining was modestly reduced by both mutations ( Fig 5D ) . Surprisingly , when these mutants were fed the E . coli HT115 bacteria commonly used for RNAi experiments , lifespan extension was completely prevented ( Fig 5E ) but ORO levels remained the same ( Fig 5F ) . The reasons for this discrepancy are unclear . It can possibly be explained by genetic redundancy between the dgat-2 and acs-22 in influencing germline-less longevity on E . coli OP50 diet , but independent , essential functions on E . coli HT115 . Overall , these experiments demonstrated that TAG production is enhanced in germline-less worms , at least partly , through DAF-16 and TCER-1-mediated increase in expression of ‘dgat’ genes , and these genes contribute at least modestly to germline-less longevity . Previous studies [18] , including ours [19] , have shown that germline ablation is accompanied by increased expression of ‘fat’ genes encoding the fatty-acid desaturase enzymes that mediate conversion of saturated fatty acids ( SFAs ) to unsaturated fatty acids ( UFAs ) [49 , 50] . In accordance with these reports , multiple ‘fat’ genes were identified in our RNA-Seq dataset ( Fig 2 and S2A Fig ) . Our lipidomic analysis confirmed that glp-1 mutants manifest increased UFAs in both the neutral and phospholipid fractions , as compared to fertile worms and this increase is attenuated in daf-16;glp-1 mutants ( S2D–S2K Fig ) . Fatty acid desaturation is closely linked to elongation of the carbon chains , mediated by elongase enzymes encoded by the ‘elo’ genes . Four out of nine elo genes encoded in the worm genome were picked up as DAF-16 and/or TCER-1 targets ( Fig 2A and S2B and S2C Fig ) and the abundance of long-chain ( >C18 ) fatty acids was increased accordingly in glp-1 mutants ( S2D and S2E Fig ) underscoring the importance of DAF-16 and TCER-1 in the increased abundance of UFAs and long-chain fatty acids in long-lived germline-ablated animals . Altogether , our experiments collectively identified key genes required for initiation of de novo fat synthesis , fatty acid desaturation and elongation , and TAG production as being upregulated by DAF-16 and TCER-1 upon germline removal , and demonstrated the biochemical and functional relevance of these changes . They suggest that the enhancement of lipid anabolic processes is an important aspect of the response to germline removal and the ensuing lifespan enhancement . Lipid breakdown is initiated by the hydrolysis of TAGs by lipases to produce DAGs and free fatty acids ( FFAs ) . Six lipases were identified as UP targets in our RNA-Seq analysis . In Q-PCR assays , five of these exhibited five-to-eighty fold increase in expression in glp-1 mutants in a partially daf-16-dependent manner ( Fig 6A–6E ) . We were unable to confirm the RNA-Seq data for the sixth lipase , atgl-1/ATGL-1 ( S1B Fig ) . Each of the five lipases we identified was essential for glp-1 mutants’ longevity , including lipl-1 and lipl-2 that were predicted to be upregulated by DAF-16 but downregulated by TCER-1 ( Fig 2 and S1H Table ) . Q-PCRs showed that lipl-1 was in fact partially upregulated by both factors and only lipl-2 was repressed by TCER-1; however , RNAi inactivation of both genes caused similar suppression of lifespan . The identification of multiple lipases is in keeping with a previous report that found glp-1 mutants display increased lipase activity partially dependent on DAF-16 and the lipase , LIPL-4 [13] . Taken together these data suggest that lipolysis may be broadly enhanced upon germline loss . The FFAs released upon TAG hydrolysis are broken down further to produce acetyl Co A through fatty-acid β-oxidation ( Fig 2B ) . Eighteen peroxisomal- and mitochondrial-β-oxidation genes [51] were represented in our RNA-Seq dataset ( thirteen UP , three DOWN and two with opposite effects ) , in keeping with our recent demonstration that the expression of multiple mitochondrial β-oxidation genes is enhanced in glp-1 mutants through NHR-49 activity [19] . Q-PCRs confirmed that the upregulation of these genes in glp-1 mutants was partially suppressed by daf-16 and/or tcer-1 mutations as well ( Fig 6F ) . nhr-49 was included in the Joint UP class , as predicted by our previous study [19] . This identification of these lipases and β-oxidation genes , along with the evidence above that multiple lipogenic processes were augmented , suggest that germline loss may trigger a simultaneous increase in lipid production and breakdown , although this remains to be demonstrated directly ( see Discussion ) . The physiology of fertile , young adults is highly invested in macromolecular synthesis to support growth and reproduction . Loss of the germline induces a fundamental change in this metabolic state , and an inability to suppress the growth programs already in place can be detrimental to the animal . Thus , genes downregulated by DAF-16 and TCER-1 following germline ablation are likely to be as important as those that are upregulated . With this perspective , we examined the DOWN genes and found them to be highly enriched for molecular processes associated with active procreation such as protein translation and reproduction ( Fig 1B and S3A–S3C Fig ) . DAVID analysis of the DAF-16-Specific DOWN genes revealed protein synthesis as the predominant category ( Fig 1B ) . One of the main GO categories ( Enrichment Score 3 . 08 ) included eighteen genes encoding proteins involved in translation subunits {nine large and seven small ribosomal protein ( RP ) subunits} , a translation initiation factor , EIF-6 , and the small mitochondrial ribosomal protein subunit ( MRPS ) , MRPS-23 ( S3A Fig and S4D Table ) . In addition , expression of three tRNA synthetase genes was repressed by DAF-16 ( S4D Table ) . Since protein synthesis is a key requirement for a proliferating germline , it is plausible that germline removal triggers DAF-16-dependent repression of translation . DAVID analysis of TCER-1-Specific DOWN class showed that , within the group receiving the highest enrichment score ( 2 . 3 ) , 12/53 genes encoded proteins involved in splicing or RNA processing ( Table 1 and S4G and S4H Table ) . This is in accordance with the role of human TCERG1 in regulating elongation-associated splicing [52] . A detailed examination of the genes in this group also revealed that the knock down of 37/53 genes has been reported to result in reproductive phenotypes including gonadal defects , reduced brood size and sterility ( Table 1 ) . Reproduction was also one of the highly represented groups in the REVIGO analysis ( S5D Table ) . These observations implied that once the germline is lost , TCER-1 actively repressed the somatic program of reproduction . Thus , DAF-16 and TCER-1 may together facilitate the adaptation to germline removal by terminating the gene-expression programs that support reproductive physiology in the somatic cells of the animal . The genes most repressed by TCER-1 also included several factors whose reductions-of-function have been reported to increase the lifespan of normal adults ( highlighted in bold in Table 1 ) . Indeed , ‘Aging and Adult Lifespan’ was one of the main groups that emerged from the DAVID analysis of TCER-1 DOWN class ( Fig 1B ) . This led us to ask if TCER-1 inhibited the expression of anti-longevity genes . Since the DOWN genes were predicted to be already repressed in germline-less animals , testing their anti-longevity roles in glp-1 mutants was not feasible . Hence , we examined their function in wild-type animals . We used a fer-15;fem-1 temperature-sensitive mutant strain that is sterile when grown at the non-permissive temperature but exhibits normal lifespan and has been used extensively as a surrogate for wild-type worms in large-scale lifespan analyses [16 , 17] . We asked if inactivation of TCER-1 DOWN genes had a beneficial effect on the lifespan of fer-15;fem-1 adults . To circumvent developmental requirements , RNAi was initiated on day 1 of adulthood . Upon RNAi inactivation of nineteen of the most highly repressed TCER-1-Specific DOWN genes , we observed a statistically significant lifespan extension in thirteen cases , eight of which were reproduced in at least two trials ( Fig 1C and S8 Table ) . To substantiate the RNAi data , we also tested the lifespans of six strains carrying mutations in five of these genes . Two mutants , gst-24 and dopy-6 , showed consistent and statistically significant lifespan extensions compared to wild-type worms . Two others ( numr-1 and lys-4 ) showed significant but variable lifespan increments ( S9 Table; unlike the RNAi strategy , adult-specific gene knockdown was not possible with mutants and sickness resulting from developmental defects could not be avoided , as was the case with lin-17 ) . These data lead us to posit that TCER-1’s repressor functions following germline removal may include suppressing aspects of reproductive physiology , by inhibiting the expression of reproductive genes , as well as the expression of anti-longevity genes . The human TCERG1 gene is highly expressed in oocytes where its mRNA levels decline with age [53] . These reports , together with our observations that genes with reproductive roles are enriched in the TCER-1-Specific DOWN class , and that tcer-1 mutants produced fewer progeny than wild-type animals , prompted us to investigate tcer-1’s function in reproduction in normal , fertile worms . We employed a collection of assays to assess reproductive health , including quantifying the brood size ( total number of eggs laid ) , viability ( fraction of eggs that hatch successfully ) and reproductive span ( the duration of adulthood for which reproduction is maintained ) [54] . As worms age , they begin to lay unfertilized oocytes . We quantified oocyte production to obtain another measure of reproductive health , ‘oocyte ratio’ ( the ratio of total number of eggs laid to the number of oocytes laid ) ( see Methods for details ) . We found that at normal growth temperature ( 20°C ) , tcer-1 mutants laid ~65% fewer eggs than wild-type worms ( Fig 7A ) and the hatching rate of these eggs was ~40% less than that of wild type ( Fig 7B ) . The oocyte ratio was also significantly increased ( Fig 7C ) . The mutants started laying oocytes earlier than wild type and continued to do so late into middle age ( S4A Fig ) . In contrast , daf-16 mutants did not exhibit any of these phenotypes ( Fig 7A–7C and S4A Fig ) . To identify whether tcer-1 mutants’ fertility phenotypes were due to defects in sperm , oocytes ( or both ) , we repeated these reproductive health assays on reciprocal crosses: wild type hermaphrodites mated to tcer-1 mutant males ( to assess sperm health ) or tcer-1 mutant hermaphrodites mated to wild type males ( to assess oocytes ) . In both cases , we observed reduced brood size and decreased viability ( Fig 7D and 7E ) and oocyte abnormalities ( Fig 7F and S4B Fig ) indicating that TCER-1 is necessary in both gametes for fertility , although this does not obviate the role of somatic TCER-1 in fertility . The tcer-1 reproductive phenotypes suggested that there might be underlying defects in germ cell development or meiosis . We explored this by examining the germline for the number of DAPI-staining bodies at diakinesis , a read-out for meiotic crossover formation . In both wild type and tcer-1 mutant animals , six DAPI-staining bodies were noticeable in all the oocytes imaged indicating that crossovers were properly made between all the chromosomal homolog pairs . However , we noticed that ~30% of day 1 tcer-1 mutant adults were still undergoing spermatogenesis ( evidenced by presence of sperm in the gonadal arm rather than being restricted to the spermatheca ) , which is normally completed during the early L4 larval stage ( Fig 7G and 7H ) . This indicated a defect in the developmental switch from spermatogenesis to oogenesis . To determine whether the animals failed to switch completely or were delayed in switching , we analyzed older adults . By day two , oogenesis was noticeable in all the animals ( S4C Fig ) . We interpret this data to suggest that tcer-1 mutants exhibit germline heterochrony or a delay in switching from spermatogenesis to oogenesis in sync with somatic maturation . These defects were further aggravated when the worms were grown at higher temperatures . By the second generation at 25°C , tcer-1 mutants showed a precipitous decline in fecundity accompanied by dramatic changes in germline morphology; >70% had germline defects including germline heterochrony , small germlines with few-to-no germ cells as well as reduced sperm count , despite prolonged spermatogenesis ( S4C–S4E Fig ) . Together this spectrum of defects revealed that tcer-1 is required for normal germ-cell growth , differentiation and germline-soma synchrony , the latter being most sensitive to loss of tcer-1 function . Taken together , our results indicate the tcer-1 promotes reproductive health in normal , fertile adults , but upon germline loss , it switches roles to repress reproductive physiology and promote the expression of genes that ensure metabolic homeostasis and longevity ( Fig 7I–7K ) .
RNA-Seq analysis allowed us to describe a detailed picture of the GRNs governed by DAF-16 and TCER-1 following germline loss , including the first compilation of TCER-1 downstream genes . Our Q-PCR and lifespan assays indicate that many of the DAF-16 and/or TCER-1 UP targets are genes whose expression is , at least partially , increased following germline depletion , and which promote GSC-less longevity . A majority these genes’ inactivations caused moderate lifespan reduction ( ~10 to 20% ) . While it cannot be ruled out that these modest effects are due to incomplete knockdown caused by RNAi , it is also possible that some UP genes make incremental individual contributions , and the collective result is a substantial life lengthening . Interestingly , the UP group included several transcription factors , some reported to be essential for glp-1 mutants’ longevity ( NHR-49 , HLH-30 , DAF-12 , SKN-1 and MDT-15 ) [4 , 14–16 , 19 , 37] ( highlighted in S1 Table ) and new ones that implicate novel cellular processes in this longevity pathway . For instance , atfs-1 , a key mediator of mitochondrial unfolded protein response ( mitoUPR ) , was included in the TCER-1-Specific UP group and was essential for glp-1 longevity ( S3 Table ) . Induction of mitoUPR by knockdown of MRPSs increases lifespan in worms and mice [55] . Notably , we identified multiple MRPS-encoding genes as being repressed by DAF-16 and/or TCER-1 ( Table 1 and S3 Fig ) . These observations raise the enticing possibility that GSC loss may also activate mitoUPR . Previously we showed that DAF-16 and TCER-1 partly mediated nhr-49 upregulation in glp-1 mutants . Multiple β-oxidation genes that we had found to be upregulated by NHR-49 in glp-1 mutants [19] were represented in the RNA-Seq dataset , along with nhr-49 itself . However , we noticed that the daf-16 mutation had a modest effect on the mitochondrial β-oxidation genes as compared to nhr-49 . One possible interpretation of this data is that DAF-16 and TCER-1’s predominant role may be upregulating key transcription factors that in turn activate specific cellular processes . Accordingly , two HLH-30-regulated autophagy genes , bec-1 and atg-9 , were included in the UP class . However , the activities of many of these factors are controlled at post-translational steps such as nuclear localization ( SKN-1 , HLH-30 ) [14 , 15] and we did not observe a strong or consistent influence of DAF-16 or TCER-1 on their expression using Q-PCRs ( S5 Fig ) , so the significance of these factors being identified in the RNA-Seq dataset remains unclear . The data underscore the complexity of the GRNs activated by GSC removal and suggest new relationships between previously known genes important for glp-1 mutants’ longevity . We found that multiple lipogenic genes were upregulated by DAF-16 and TCER-1 in glp-1 mutants , and that at least three lipid anabolic processes- de novo fatty-acid synthesis , TAG production and fatty-acid desaturation and elongation- were all significantly elevated upon GSC removal . The identification of increased de novo fatty acid synthesis was particularly intriguing , because recent evidence suggests that a large fraction of the surplus fat in glp-1 mutants is likely to be a result of unchecked yolk production [15] . Yolk lipids are similar to bonafide storage forms of fat but they also have distinct biochemical and functional attributes [56] , and our experiments did not distinguish between them . However , the fact that ( a ) de novo fatty-acid synthesis is elevated upon germline loss , and ( b ) fasn-1;glp-1 exhibited dramatic lifespan suppression , suggest that lipogenic pathways perform crucial functions upon germline removal . While DGATs can help immure fats ( normally designated for oocytes in fertile animals ) into lipid droplets , what purpose is served by elevating fat production per se ? The answer to this question is unknown , but based on emerging evidences , from vertebrate literature and worm studies , we postulate that enhanced de novo fatty-acid synthesis facilitates the production of lipophilic signals and ligands for transcription factors that enable the adaptation to germline loss . Fatty acids have been known to serve as signaling molecules for long . But , recent reports have begun to emphasize the importance of the ‘source’ of lipid signals . For instance , mice incapable of synthesizing ‘new fat’ due to fatty-acid synthase ( FAS/FASN-1 ) deletion in the liver or hypothalamus cannot activate PPARα , with which NHR-49 shares functions [57 , 58] . Multiple lines of evidence indicate that inter-tissue communication in worms , including glp-1 mutants , involves activation of transcription factors by steroid ligands . For instance , DA activates DAF-12 [31] and UFAs activate SKN-1 in glp-1 mutants [15] . Oleoyl-ethanolamide ( OEA ) binds NHR-80 and is essential for expression of NHR-80 and NHR-49 targets in fertile adults [59] . So , it is plausible that DAF-16-mediated elevation of de novo lipid synthesis helps produce ligands for factors that help adapt to germline depletion . Further studies will be needed to test this hypothesis . We noticed that inactivation of individual lipogenic genes did not have major effects on steady state fat levels , even in cases where glp-1 mutants’ lifespan was substantially reduced . For instance , fasn-1 ( fr7 ) that abrogated glp-1 longevity completely , and pod-2 RNAi that significantly shortened lifespan , had considerably smaller effects on ORO levels in young glp-1 adults , and the depletion was not retained in older animals ( Fig 5 and S6 Fig ) . Both fasn-1 ( fr8 ) and fasn-1 RNAi had strong lifespan effects , but only the mutation caused significant reduction in ORO labeling ( Fig 5A and S7A–S7C Fig ) . One possible explanation for this perplexing observation is genetic redundancy , at least amongst the DGATs . Alternatively , these observations align with a growing body of research suggesting that steady-state fat levels and lifespan are not directly correlated [27 , 29 , 30] . Indeed , it is illustrated by the fact that germline-less animals and IIS mutants , both have higher fat but are healthier and longer-lived than their leaner , fertile counterparts [32 , 60] . Interventions that increase fly lifespan , such as reduced IIS and TOR inhibition , also elevate fat [61 , 62] . Similarly , ‘metabolically healthy obese’ individuals are noteworthy because they retain excessive weight without developing metabolic disorders [63] . Together with these observations , our data emphasize that the relationship between adiposity and lifespan is nuanced and multi-layered . Besides demonstrating that lipogenic processes are augmented upon germline loss , our study also identified multiple lipolytic genes , including putative lipases . It has been reported by others that lipase activity is increased in germline-less adults dependent upon DAF-16 , and our previous study suggested that β-oxidation is similarly augmented . Taken together , these observations imply that lipogenesis and lipolysis are concurrently amplified following germline depletion , although it remains to be demonstrated directly . Lipid turnover has been observed in other organisms under conditions of metabolic flux . Flies and mice undergoing dietary restriction ( DR ) display enhanced fatty acids turnover [64 , 65] . Pertinently , inhibition of acetyl CoA carboxylase ( dACC ) , the fly POD-2 homolog , abrogates DR-mediated longevity [65] . It is plausible that a similar large-scale turnover of cellular lipids following loss of the germline helps worms preserve metabolic homeostasis . Since our experiments did not gauge lipid turnover directly , further studies are needed to test this possibility . These processes may also be important for maintaining metabolic homeostasis in normal animals , since RNAi inactivation of some of the lipid genes shortened wild-type lifespans as well ( S10 Table ) . Importantly , our study has identified multiple lipid-metabolic genes and pathways involved in the adaptive response to germline loss , many of which are orthologous to human genes , including ones implicated in diseases ( S11 Table ) . Active protein synthesis is critical for successful reproduction—interventions that impair translation often result in sterility . Similarly , a host of somatic genes and proteins are essential for successful procreation . If reproduction is prevented , an inability to reduce protein synthesis or stop the somatic programs that support reproduction can cause metabolic disarray . Our data implicate DAF-16 and TCER-1 as the enforcers of these repressive functions in glp-1 mutants . The enrichment of ribosomal genes in the DAF-16-Specific DOWN class implies that the increased longevity of germline-less animals may be attributable , in part , to a global decrease in translation rates , a phenomenon associated with increased lifespan in many organisms , including worms [66 , 67] . Accordingly , reduced protein synthesis in dietary-restricted worms and , in a DAF-16-dependent manner , in daf-2 mutants has been observed [68] . Furthermore , the kinase TOR , that promotes ribosome biogenesis and translation , is down regulated in glp-1 mutants by DAF-16 [13] . It remains to be investigated if DAF-16 represses ribosomal gene expression directly , or through TOR inhibition . TCER-1 , on the other hand , represses the somatic gene-expression program of reproduction in glp-1 mutants , and in addition , appears to impede the expression of many anti-longevity genes . Expectedly , there was some overlap between the two classes . Of the 13 TCER-1-Specific Class B genes whose inhibition increased wild-type lifespan , two are associated with fertility defects ( hel-1 and pcf-11 ) . Similarly , four of the 37 genes with sterility defects have been reported to be long-lived ( Table 1 ) . Unlike DAF-16 , TCER-1 represses the expression of more genes than it upregulates ( 366 vs . 213 ) so its predominant transcriptional function may be that of a repressor , similar to TCERG1 [26] . The two factors also appear to exert opposite effects on a significant number of UP genes , including lipl-1 and lipl-2 . While RNA-Seq identified both genes as DAF-16-Specific UP and TCER-1-Specific DOWN targets , Q-PCRs showed that lipl-2 is repressed by TCER-1 and lipl-1 is upregulated . Also surprisingly , both genes promoted glp-1 mutants’ longevity . Presently , the reason for this paradox , and the relevant importance of lipl-1 vs . lipl-2 , is unclear . Further studies are needed to unravel the mechanism by which TCER-1 and DAF-16 mediate transcriptional activation vs . repression , and to understand their mutual interactions in influencing the expression of the ‘Opposite’ targets . tcer-1 mutants were impaired in all aspects of reproduction including the number of progeny produced and the fraction that developed successfully . They manifested germ-cell defects , the most striking of which was germline heterochrony , and deficits in sperm and oocytes , whereas , a previous study that assayed only sterility reported no major defects [69] . tcer-1 is expressed in both the soma and germline , and , since we did not perform tissue-specific knockdowns , the somatic contribution of TCER-1 to germline defects cannot be ruled out . Nonetheless , our experiments clearly demonstrate its necessity for optimal fertility . Thus , TCER-1 appears to perform two antithetical functions- in normal , fertile animals , it facilitates reproductive success , possibly by mediating germ-cell proliferation and ensuring gamete maturation in coordination with the somatic development . But , when the germline is eliminated , the somatic TCER-1 protein represses the expression of ( somatic ) genes that support reproduction . In addition , and together with DAF-16 , it triggers the increased expression of lipid catabolic and anabolic genes that likely allows the animal to retain metabolic equilibrium and leads to increased lifespan . This discovery of TCER-1’s molecular adaptability , in promoting or repressing fertility based on the physiological requirement of the animal , provides a unique mechanistic insight into the cross talk between procreation and length of life ( Fig 7I–7K ) . Notably , TCERG1 is expressed in mice and human oocytes [53 , 70] and its levels are reduced in oocytes of menopausal women [53] . It will be interesting to investigate if TCER-1’s role in balancing fertility and longevity is conserved across evolutionary time scales .
Animals were grown and maintained at 20°C using standard techniques [71] . The strains used in this study are N2 , AGP162 {acs-22 ( hj26 ) I outcrossed to Ghazi lab N2} , AGP163 {acs-22 ( hj26 ) I; glp-1 ( e2141ts ) III} , AGP164 {outcrossed dgat-2 ( hj44 ) V} , AGP165 {glp-1 ( e2141ts ) III; dgat-2 ( hj44 ) V} , AGP166 {outcrossed IG346 ( fasn-1 ( fr8 ) I; frIs7 ( nlp29p::GFP+col-12p::dsRed ) IV} , AGP167 {fasn-1 ( fr8 ) I; glp-1 ( e2141ts ) III} , CF512 {fer-15 ( b26 ) II; fem-1 ( hc17 ) IV} , CF1903 {glp-1 ( e2141 ) III} , CF2154 {tcer-1 ( tm1452 ) II; glp-1 ( e2141 ) III} , CF1880 {daf-16 ( mu86 ) I; glp-1 ( e2141 ) III} , CF2166 {tcer-1 ( tm1452 ) II} and CF1038 {daf-16 ( mu86 ) I} . Strains carrying mutations in TCER-1-Specific DOWN genes that were used in lifespan experiments are listed in S9 Table . RNA was isolated on the second day of adulthood , using the mirVana miRNA Isolation Kit ( Ambion , AM1561 ) , from approximately 5000 worms each of CF1903 , CF2154 and CF1880 ( two biological replicates ) and was prepared for sequencing using the Illumina TruSeq RNA Sample Preparation Kit as per the manufacturer’s instructions . The samples were then multiplexed prior to cluster formation and subjected to 50 base pair single-end sequencing on a Hiseq 2500 Illumina sequencer ( Tufts University Genomics Core ) . The data was analyzed using the bioinformatics tools available at the Galaxy project [34] . Using FASTQ Groomer , the raw sequencing reads were initially converted to the Sanger FASTQ format that is compatible with TopHat Junction Mapper ( 1 . 5 . 0 ) , the splice-junction mapping tool . Tophat ( single end mating , default parameters ) was then used to align the RNA-seq reads onto the C . elegans reference genome ( WS190/ce6 ) . The Sequence Alignment Map ( SAM ) format Tophat-output files were used as input for assembling the transcripts using Cufflinks and Cuffmerge ( 0 . 0 . 5 ) . Cufflinks was run using 300000 as the max intron length , without quartile normalization and correcting for bias . Cufflink normalized and quantified the data to produce Fragments Per Kilobase of exon per Million fragments mapped ( FPKM ) . To facilitate combinatorial pairwise sample comparison , RNA-seq data was pooled with the reference annotation file using the meta-assembler , Cuffmerge , to result in a singular merged Gene Transfer Format ( GTF ) file of all the transcripts . The Cuffdiff Differential Gene Expression tool ( version 0 . 0 . 5 ) was used to calculate differential gene expression using a false discovery rate of 0 . 05 , a minimum alignment count of 100 and with bias correction to obtain gene and transcript expression level data along with fold change ( in log2 scale ) and P values ( raw and corrected for multiple testing ) . In the final lists of differentially regulated genes , a few loci had multiple genes mapping to them that necessitated manual curation ( highlighted in blue in S1 Table ) . The Wormbase identifiers of the gene lists with the significantly and differentially expressed genes obtained from RNA-Seq analysis were uploaded on the publically available bioinformatic platform , Database for Annotation , Visualization and Integrated Discovery ( DAVID ) to identify enriched gene groups [39 , 40] . The Functional Annotation Chart tool was used to identify the most overrepresented Gene Ontology terms associated with a given gene list , and these categories were probed for enriched gene groups reported as Gene Ontology ( GO-BP ) groups . REVIGO analysis was performed on the DAVID GO-BP groups to obtain summarized GO-BP-IDs [41] . Worm RNA was isolated as described above , DNAse treated ( DNAse kit , Sigma #AMPD1 ) and reverse transcribed into cDNA ( Protoscript m-MuLV First Strand cDNA Synthesis kit , NEB #E6300S ) according to the manufacturer's instructions . Quantitative real-time PCRs were performed using an Applied Bio Systems 7300 Real Time PCR System employing Sybr Green chemistry ( SensiMix SYBR Hi-ROX kit , Bioline #QT-605 ) . Gene expression data were normalized to housekeeping gene rpl-32 ( and in many instance relative to pmp-3 and Y45F10D . 4 as well ) after confirming that rpl-32 was expressed at the same level in all the strains . All data reported here were obtained by combining results from at least three independent biological replicates , each comprising 2–4 technical repeats . All primer sequences are listed in S12 Table . All lifespan experiments were carried out at 20°C unless otherwise noted . CF512 , and temperature sensitive strains carrying the glp-1 mutation , were initially grown at 20°C for 2-4hrs , transferred to 25°C till day 1 of adulthood , then returned to 20°C for the remainder of life . For RNAi experiments , worms were grown on E . coli OP50-seeded plates till mid-to-late L4 stage and then transferred to plates seeded with E . coli HT115 bacteria carrying an empty control vector ( pAD12 ) or relevant RNAi clones . Survival curves were generated based on the Kaplan-Meier method using STATA 10 . 0 and 8 . 0 ( Stata Corporation ) , or the online tool , OASIS ( http://sbi . postech . ac . kr/oasis ) . Statistical significance was calculated using the non-parametric log-rank Mantel-Cox method . The students’ t-test was used to calculate significance for the Q-PCR data and the Mann-Whitney test was used to calculate statistical power for the reproductive health assays . The statistical significance of the overlap between two gene sets was calculated using the hypergeometric probability formula with normal approximation available as a program at nemates . org ( http://nemates . org/MA/progs/overlap_stats . html ) . Gravid adults were bleached to obtain approximately 15000 eggs per strain and transferred to NGM plates seeded with E . coli OP50 , incubated at 20°C for 2-4hrs and then transferred to 25°C . Animals were incubated at this temperature and collected either at the L4 stage ( for comparing N2 and glp-1 strains ) or kept at 25°C till day 1 , moved to 20°C thereafter and collected on day 2 of adulthood . For labeling , the collected animals were transferred for 6 hours to stable isotope plates , harvested , washed in M9 three times and frozen in a dry ice/ethanol bath and stored at -80°C until processed . Total lipids were extracted and purified as previously described [45] . Purified lipids were dried under nitrogen , re-suspended in methanol/2 . 5% H2SO4 and incubated for 1 h at 80°C to create fatty acid methyl esters ( FAMEs ) that were analyzed by gas chromatography/mass spectrometry ( GC/MS ) ( Agilent 5975GC , 6920MS ) . TAG levels were determined as the NL/PL ratio based on the total abundance of the NL and PL fatty acids and recovery of the following internal standards: tritridecanoin ( Nu-Chek Prep ) and 1 , 2-diundecanoyl-sn-glycero-3-phosphocholine ( Avanti Polar Lipids ) . Similar results were obtained upon estimating the ratio of TAGs to total TAGs + PLs . The total abundance of NL and PL fractions and analyses of these data are shown in S8 Fig . The relative abundance of fatty acids in each class was determined for all the major fatty acid species in the nematode as previously described [45] . To quantify TAG and Phospholipid ( PL ) yields , total PL and TAG abundance was corrected for losses using added standards and data presented as a TAG:PL ratio , determined by measuring the sum of all major fatty acids found in TAGs versus the sum of the major fatty acids found in PLs . de novo fatty acid synthesis was calculated through a series of described equations which allow for the quantification of the amount of each fatty acid species generated from synthesis , distinguished by an isotope pattern from a mixture of 12C and 13C [45] . Numbers reported here represent the amount of 13C-labeled fatty acids derived from synthesis when compared to the total amount of fatty acids newly incorporated into the animal ( Synthesized FA + Absorbed FA ) . ORO staining was done as described earlier [19 , 32] using 30–40 animals per strain for each of two or three trials . Animals were mounted and imaged with using a Leica M165FC microscope equipped with a Retiga 2000R camera ( Q Imaging ) . Images were captured with the QCapture Pro7 software ( Q Imaging ) and quantified using ImageJ software ( NIH ) . To compare lipid levels between multiple genetic contexts , ORO staining results were converted into the fraction F of the area of each worm that was labeled with ORO . Comparisons of F between different genetic contexts ( wildtype , RNAi , knockout ) were performed in the framework of generalized linear models , which accounts for 0 < F < 1 , summarizes over the two biological replicates , and provides tests akin to the ( paired ) t-test . Specifically , we used the glm function in the R language ( R Core Team 2015 ) with a quasibinomial family , thereby accounting for overdispersion ( high variance ) in the observed data . Hypotheses about differences in F were tested using the multcomp package [72] , and results are reported as univariate p values ( shown in the tables associated with S6 and S7 Figs ) . Reproductive health was assessed using previously described assays [54] as well as new measures ( oocyte ratio ) . All experiments were conducted at 20°C and when matricide/bagging occurred the animal was censored from the experiment on that day . The data presented is obtained from aggregation of three independent trials , in each of which at least 10–15 animals per strain were examined . Individual synchronized L4 hermaphrodites were moved to fresh plates on a daily basis till the end of the reproductive phase ( reproduction cessation for a minimum of 2 days ) , and number of eggs produced each day counted to calculate fecundity ( total number of eggs laid by a hermaphrodite during its reproductive phase ) . Each day , once the parent was moved to a fresh plate , the older plate with eggs was stored at 20°C overnight , and the number of hatched worms counted the following day to calculate brood size . The above two parameters were used to determine viability ( ratio of the total number of eggs laid by a hermaphrodite in its lifetime to total number of eggs that hatched ) . Similarly , the number of oocytes laid each day was counted to obtain oocyte number . These data were used to estimate the oocyte ratio ( ratio of total number of oocytes laid by an animal to the total number of viable eggs it produced ) . Oocyte production span is the distribution of the oocyte ratio on a daily basis for the length of time that an animal lays any brood ( eggs and oocytes combined ) . This parameter was used to assess premature oocyte production . N2 males were crossed with tcer-1 hermaphrodites ( oocytes lacking tcer-1 ) and tcer-1 males ( sperm lacking tcer-1 ) were crossed with N2 hermaphrodites . As controls , N2 males and hermaphrodites were crossed to each other . In each case , approximately 15 individual crosses were performed with a male: hermaphrodite ratio of 3:1 . Only plates with successful crosses ( ~50% male progeny ) were considered for subsequent analyses . Hermaphrodites and males were transferred to fresh plates every day , and the various measures of reproductive health described above were calculated from the plates on which eggs were laid . Strains were maintained at 20°C , or shifted to 25°C as L4 larvae and grown for one or two generations as required , prior to fixation . All animals were hand picked off plates , washed in M9 prior to fixation with Carnoy’s fixative ( 60% ethanol , 30% chloroform and 10% glacial acetic acid ) . Whole animals were then stained with DAPI for 15 minutes , washed with PBST ( 0 . 1% Triton ) and mounted in Prolong Gold with DAPI ( Life Technologies , Inc . ) prior to imaging on a Nikon A1 confocal with 0 . 2micron Z-section . Images were visualized using Velocity software ( PerkinElmer ) . | The balance between production and breakdown of fats is critical for health , especially during reproduction-related changes such as onset of puberty or menopause . However , little is known about how animals retain a balanced metabolism when undergoing major life events . Here , we have used a C . elegans mutant that successfully adapts to loss of reproductive cells to address this question . Our data suggest that the conserved proteins DAF-16/FOXO3A and TCER-1/TCERG1 mediate a coordinated increase in fat synthesis and degradation when the reproductive cells are lost . This coupling likely helps the animal to manage the lipids that would have been deposited in eggs as yolk , thus preventing metabolic disarray . These proteins also inhibit processes that would have normally supported reproduction . Together the activities of these transcription regulators allow the mutant to convert a debilitating loss of fertility into improved health and longevity . We also report that TCER-1 promotes reproductive health in normal adults , whereas when procreation is impeded , it switches roles to repress fertility and enhance lipid equilibrium . These observations offer insights into how complex organisms coordinate their metabolism to suit their reproductive needs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"rna",
"interference",
"immunology",
"germ",
"cells",
"age",
"groups",
"oocytes",
"adults",
"epigenetics",
"antibodies",
"immune",
"system",
"proteins",
"lipids",
"genetic",
"interference",
"proteins",
"... | 2016 | DAF-16 and TCER-1 Facilitate Adaptation to Germline Loss by Restoring Lipid Homeostasis and Repressing Reproductive Physiology in C. elegans |
Recent development of benzoxaborole-based chemistry gave rise to a collection of compounds with great potential in targeting diverse infectious diseases , including human African Trypanosomiasis ( HAT ) , a devastating neglected tropical disease . However , further medicinal development is largely restricted by a lack of insight into mechanism of action ( MoA ) in pathogenic kinetoplastids . We adopted a multidisciplinary approach , combining a high-throughput forward genetic screen with functional group focused chemical biological , structural biology and biochemical analyses , to tackle the complex MoAs of benzoxaboroles in Trypanosoma brucei . We describe an oxidative enzymatic pathway composed of host semicarbazide-sensitive amine oxidase and a trypanosomal aldehyde dehydrogenase TbALDH3 . Two sequential reactions through this pathway serve as the key underlying mechanism for activating a series of 4-aminomethylphenoxy-benzoxaboroles as potent trypanocides; the methylamine parental compounds as pro-drugs are transformed first into intermediate aldehyde metabolites , and further into the carboxylate metabolites as effective forms . Moreover , comparative biochemical and crystallographic analyses elucidated the catalytic specificity of TbALDH3 towards the benzaldehyde benzoxaborole metabolites as xenogeneic substrates . Overall , this work proposes a novel drug activation mechanism dependent on both host and parasite metabolism of primary amine containing molecules , which contributes a new perspective to our understanding of the benzoxaborole MoA , and could be further exploited to improve the therapeutic index of antimicrobial compounds .
Encouraged by significant advances in the disease control since the close of the 20th century , the World Health Organization ( WHO ) targeted HAT for eradication by 2020 [1] . A pressing obstacle to achieving this goal is the limited and obsolete range of treatments available that are compromised by deleterious side-effects , poor oral bioavailability and an alarming increase in drug resistance in the field [2–5] . Several new candidate drugs have advanced through the development pipeline , including acoziborole ( SCYX-7158/AN5568 ) [6] , a lead compound currently in a phase 2/3 clinical trial , and AN7973/SCYX-1608210 and AN7119/SCYX1330682 as back-ups [7 , 8] . These compounds represent a class of hemiboronic acids with distinctive chemical and pharmacological features [9 , 10] . The cyclic boronic ester in the molecules provides a good balance between the Lewis acidity essential for forming interactions with biochemical targets and the physicochemical properties important for good bioavailability . Molecular insights into the mechanism of action ( MoA ) in pathogenic model organisms has greatly contributed to development and strategies for assessing potential risk of resistance for this series of molecules . However , the benzoxaborole core structure is highly adaptive to substitution of function groups , which not only contributes to great chemical diversity but also gives rise to a broad MoA spectrum . The latter is manifested in a variety of targets and efficacy factors proposed from study of benzoxaboroles in various diseases and conditions , including proteases , phophodiesterases , kinases , anhydrolases , aminoacyl-tRNA synthetases , reductases and RNA splicing factors [11–20] . Furthermore , how the uptake and metabolism of these compounds occur in the context of infections remains as a significant gap in our understanding of benxoxaboroles . An earlier study examining the impact of resistance to acoziborole , together with identification of possible interacting proteins was inconclusive with regard to MoA [21] . Here , we adopted an approach combining forward genetics , biochemistry and structural biology and identified a metabolic pathway critical for achieving the trypanocidal activity of a series of 4-aminomethylphenoxy benzoxaboroles . The pathway involves two oxidation reactions occurring sequentially in the host and the parasite . This highlights the importance of metabolic interaction between host and pathogen [22–25] in considering novel MoAs , and contributes to our improved understanding of benzoxaborole MoA .
We screened a set of benzoxoaboroles with variable substituents against a genome-scale RNAi library in T . brucei [26 , 27] to uncover the genetic factors that sensitize trypanosomes to the compounds . These compounds included AN3054 and AN3057 that share a 4-aminomethylphenoxy substituent linked via the 6- or 5-position of the benzoxaborole core , and acoziborole that contains a 6-carboxyamide substituent ( Fig 1A ) . We identified a high confidence hit , Tb927 . 6 . 3050 , specifically with the 4-aminomethylphenoxy derivatives ( AN3054 and AN3057 ) , in addition to a cohort of candidates determining the sensitivity of the parasites towards a broad range of benzoxaboroles ( unpublished data ) . We then examined the impact of Tb927 . 6 . 3050 RNAi on the sensitivity of trypanosomes to individual compounds ( Fig 1B ) . Upon knockdown , the trypanosomes were significantly desensitized to either AN3054 or AN3057 as opposed to acoziborole . Moreover , there was a further divergence in the impact of the knockdown between AN3054 and AN3057 , with more profound impact on the latter . This is also consistent with the results from the initial genetics screens where the signal for Tb927 . 6 . 3050 was more pronounced in the screen with AN3057 than with AN3054 ( Fig 1A ) . Taken together , these data suggest a specific structure-activity relationship ( SAR ) between Tb927 . 6 . 3050 and 4-aminomethylphenoxy derivatives . To define this SAR , we first investigated the correlation between the potency of related but distinctive phenoxy compounds and Tb927 . 6 . 3050 RNAi . Included were three aminomethylphenoxy-substitution ( -CH2NH2 ) isomers , i . e . AN3057 , AN3054 and AN3056 , and three carboxyphenoxy-substitution ( -CO2H ) isomers , i . e . AN2861 , AN3330 and AN3410 , with each aminomethylphenoxy isomer paired with the corresponding carboxyphenoxy isomer as shown in Fig 2 . The potencies of all three aminomethylphenoxy isomers were significantly compromised by the knockdown , in contrast to the impact of the carboxyphenoxy equivalents . Thus , the methylamine moiety shared by the aminomethylphenoxy derivatives is the key link to the function of Tb927 . 6 . 3050 in defining the potencies of this series of benzoxaboroles in T . brucei . Noticeably , among three aminomethylphenoxy isomers , there was a further variation in the impact of the knockdown on drug potency; the highest impact was observed with AN3057 ( ΔEC50≈120 ) , followed by AN3056 ( ΔEC50≈7 ) , and the least with AN3054 ( ΔEC50≈4 ) . This phenomenon suggests a differential contribution by Tb927 . 6 . 3050 to the anti-trypanosomal activities of aminomethylphenoxy bezoxaboroles with further chemical diversity . Additionally , we found in comparing the aminomethylphenoxy-carboxylate pair compounds that the methylamine containing compounds are more potent than the carboxylate counterparts , with ~8 fold difference between AN3057 and AN2861 , ~3 fold between AN3054 and AN3330 , and ~15 fold between AN3056 and AN3410 . Next , we took a comparative genomics approach to elaborate the function of Tb927 . 6 . 3050 , which was previously uncharacterised [28] . We constructed a phylogenetic tree for all members of the ALDH superfamily currently known in humans and other opisthokonts ( S1 Fig ) , and used this as reference to identify and categorise potential orthologues in the Trypanosomatida ( S2 Fig ) . All candidates identified were categorized into five distinct clades based on the similarity of each towards the corresponding orthologue in the opisthokonts; a trypanosomatid-specific ALDH subfamily , represented by Tb927 . 6 . 4210 in T . brucei , emerged along with ALDH1/2 , ALDH3 , ALDH4 and ALDH5 subfamilies , indicating evolutionary functional divergence ( Fig 3A ) . Importantly , Tb927 . 6 . 3050 falls into the clade comprising members of the ALDH3 subfamily that have been linked specifically to the metabolism of fatty acid and aromatic aldehydes [29] , and thus is designated as TbALDH3 ( Uniprot Q583M9 ) . Greater functional insight was attained from analyzing the TbALDH3 structure by X-ray crystallography . The structure of HsALDH3A1 [30] ( PDB 1AD3 ) was used as the searching template to obtain the initial phases of the structure by molecular replacement . In the solved structure ( PDB 5MYP ) , TbALDH3 appears as a dimer that consists of two monomeric polypeptides ( S1 Table ) . As shown in Fig 3B , the monomer structure comprises 485 residues from A6 to K490 , and adopts a canonical ALDH fold that is composed of an N-terminal NAD-binding domain , a catalytic domain and an oligomerization domain embracing the second subunit of the dimer . Several distinct features are also apparent in the structure , including an extended N-terminus with 15 residues preceding helix α1 , a side-extrusion from the catalytic domain formed by helices α10 and α11 that are bridged by a flexible loop ( G297-Q303 ) , and an C-terminal extension adjacent to the dimerization domain ( helix α16 , Y477-L489 ) . Interestingly , this C-terminal structural element strikingly resembles the ‘gatekeeper helix’ that is unique to FALDHs and is involved in regulating enzyme activity and substrate specificity [31] . Previously it has been noted that differential localizations of family members can contribute to the functional divergence in ALDHs [32–37] . By immunofluorescence , we found that TbALDH3 is predominantly co-localized with glyceraldehyde phosphate dehydrogenase ( GAPDH ) . The latter is a distinctive marker for glycosomes , a specialised membrane-enclosed organelle in a few protozoan species including T . brucei , derived from peroxisomes and with essential metabolic functions ( Fig 3C ) . This result is also supported by a recent quantitative glycosomal proteomics study that identified TbALDH3 with high confidence [38] . Overall this evidence strongly suggests that TbALDH3 functions as a glycosomal FALDH . To understand the reaction catalyzed by TbALDH3 at molecular level , we obtained crystals of the aldehyde-NAD-TbALDH3 complex . A micro-reaction chamber ( S5 Fig ) was designed to serve as a physical barrier between the two reactions as well as to minimize the reverse reaction; meanwhile , a cysteine to serine mutation ( C259S ) was introduced in TbALDH3 to entrap the hyperactive aldehyde metabolites by abolishing the nucleophilic attack on the carbonyl carbon of the aldehyde . The resulting crystals are isomorphous to the apo-TbALDH3 crystals and the structure was solved at 2 . 5 Å ( PDB 5NNO ) . As illustrated in Fig 5A , the NAD cofactor is located in an extended pocket rendered by a Rossmann fold [41] . There are two hydrogen bonds formed between E365 and the hydroxyl groups of the ribose of the nicotinamide , establishing the key interactions between NAD with TbALDH3 in addition to an H2O-bridge that connects the carboxylate group of the nicotinamide to Y444 and T200 . Interestingly , similar topological arrangements have been proposed as part of the catalytic mechanism for ALDH . In this structure , the benzaldehyde benzoxaborole substrate exhibits distinctive electron density ( S6 Fig ) , being nestled into the substrate funnel and adjacent to the nicotinamide ( Fig 5B ) . Specifically , in the catalytic pocket , its aldehyde group directly contacts the catalytic residue ( C259S ) , and also connects to N128 , a residue whose counterparts in other ALDHs are involved in the catalytic activity by directing the carbonyl carbon towards the nucleophilic carbon on the nicotinamide [42 , 43] . Noticeably , the position of the substrate is relatively flexible as indicated by the B-factor ( S1 Table ) . There is an additional interaction between the substrate and TbALDH3 , which is mediated by the hydrogen bond between the hydroxyl group of the hemiboronic ring in the substrate and a side-chain conformer of R473 in TbALDH3 , proximal to the gatekeeper helix of the adjacent protomer ( Fig 5B ) . In contrast , the boron atom remains in trigonal planar configuration , therefore capable of forming stable adducts with the potential targets , which suggests that the aldehyde metabolites serve as the substrates of TbALDH3 rather than the enzyme inhibitors as described in other cases [11 , 13 , 20] . The enzymatic specificity of TbALDH3 towards the aldehyde metabolite was further determined by superimposing the structure onto two other ALDH3 subfamily member structures , HsALDH3A2 and RnALDH3A1 ( Fig 5C; see S7 Fig for the sequence alignment ) . The C-terminal ‘gatekeeper’ helix in HsALDH3A2 , absent in RnALDH3A1 , has been shown to be involved in defining its substrate specificity by partially occluding the substrate entrance from the ‘substrate funnel’ . Interestingly , a similar structure is also defined in TbALDH3 , although it is distant from the ‘substrate funnel’ , probably imposing less restriction on the substrates ( Fig 5B and 5C ) . This structural element in TbALDH3 appears to be held in position via the hydrogen bond between K481 and the backbone carbonyl of F474; highly conserved among human FALDHs ( S8 Fig ) , and serving as a mechanical hinge to support the activity of FALDHs . A point mutation in HsALDH3A2 ( K447E ) , equivalent to the K481E in TbALDH3 , leads to a deficient enzymatic activity and is genetically linked to Sjögren–Larsson Syndrome [44] . More importantly , in the superimposition ( Fig 5B ) , the benzaldehyde-benzoxaborole substrate can be accommodated by TbALDH3 and HsALDH3A2 , but encounters steric clashes with RnALDH3A1 , indicating that the unique topological arrangement in the substrate funnel of FALDHs such as HsALDH3A2 and TbALDH3 determines the substrate specificity besides the positioning of the C-terminal ‘gatekeeper’ helix . This also raises the possibility that the conversion of the aldehyde metabolites into carboxylic acids could also occur in the host , as the structural similarity suggests that human FALDHs , like HsALDH3A2 , are likely to possess a substrate specificity similar to TbALDH3 . The MAO-like enzymatic activity of the AO-TbALDH3 pathway remained undefined in vivo , considering the fact that neither genetic nor biochemical evidence supports the presence of such activity in Trypanosoma [45] . We postulated that the host provides this enzymatic activity , which is supported by documented AO activity in animal plasma [46–48] . Fetal bovine serum ( FBS ) is the only host-derived metabolically active component in our parasite culture system , and is the most likely source . To test this , we first analyzed the metabolism of AN3057 using an enzymatic cascade in vitro , composed of FBS and TbALDH3 , by HPLC-MS ( Fig 6A ) . The chemical transitions revealed resemble those through the MAOa-TbALDH3 pathway ( Fig 4 ) . No significant metabolic changes were detected with AN3057 in the absence of FBS , although the same loss of a hydroxyl group ( -OH ) occurred as observed previously with MAOa . Conversely , in the presence of both FBS and TbALDH3 , two acidic metabolites ( F-A1 , F-A2 ) were identified , with F-A1 sharing the same MS profile with AN2861 and F-A2 subjected to the loss of boron ( S9 Fig ) . Furthermore , when TbALDH3 was absent , AN3057 was metabolized via FBS into two metabolites ( F1 and F2 ) , both of which are suggested as similar benzaldehyde entities distinguished primarily by the boron in the structure ( S9 Fig ) . In summary , AN3057 is converted by FBS first into the aldehyde metabolites ( F1 , F2 ) , which are then further metabolized by TbALDH3 to the carboxylate metabolites ( F-A1 , F-A2 ) . The metabolic pathway revealed here very likely mirrors the host-pathogen encounter in the host vascular system . To further validate this point , we examined the relative potencies of the metabolites , generated by the FBS-TbALDH3 in vitro cascade reactions , upon silencing TbALDH3 by RNAi . As shown in Fig 6B , with both the methylamine parent and aldehyde metabolites , the potency was significantly compromised following the RNAi . In contrast , no significant change in potency was observed with the carboxylate metabolites . Therefore , an FBS-TbALDH3 pathway can serve as a metabolic route through the host and the parasite for activating the aminomethylphenoxy benzoxaboroles . Primary amine oxidases are generally classified into two major groups that are distinguished by the co-factor-dependent catalytic mechanism; one group is dependent on Flavin Adenine Dinucleotide ( FAD ) and represented by MAOa , while the second group , exemplified by Semicarbazide-Sensitive Amine Oxidase ( SSAO ) , on copper-quinone [49 , 50] . This biochemical feature differentiates the enzyme sensitivity to inhibitors that are specifically targeting the respective co-factors . Following this notion , we chose a fluorescence-reporter enzymatic assay to characterize the AO activity in FBS with a panel of specific AO inhibitors . As shown in Fig 6C , the enzymatic activity was exclusively sensitive to the SSAO inhibitors , particularly semicarbazide ( Semi ) and bromoethylamine ( Bromo ) , and remained unaffected by MAO inhibitors , thus confirming that SSAO provides the predominant AO activity in FBS . More importantly , when we introduced semicarbazide into the trypanosome cultures , the potency of AN3057 was significantly compromised in a dose-dependent manner ( Fig 6D ) , indicating the SSAO deamination activity derived from the host is critical for activating the aminomethylphenoxy benzoxaboroles as potent trypanocidals . Therefore , we identified the SSAO as the primary metabolic enzyme in the host intravascular system for metabolizing the chemical compounds containing primary amines via the oxidative deamination .
Benzoxaboroles have opened new pharmaceutical opportunities by combining unique chemical properties of the benzoxaborole core structure with a repertoire of structurally diverse substituents . This gives rise to a wide range of chemical entities of pharmaceutical interest that potentially act via diverse mechanisms in targeting various medical conditions . So far , the effort in understanding the MoA has been focused on identifying potential functional targets , with which the boron atom , as well as additional functional groups of the compounds , form stable interactions . For example , fungal or bacterial Leucyl-tRNA forms covalent bonds through the tRNA’s adenosine with the boron [16 , 20]; CPSF3 , an essential RNA metabolism factor in several protist parasites , is targeted primarily by the boron and the carboxylate group in AN3661 that form non-covalent bonds with the surrounding catalytic structures [11 , 13] . However , a systemic understanding of the reciprocal actions between benzoxaboroles and targeted pathogens is lacking in general , particularly regarding the mechanisms underlying drug uptake and metabolism . Here we exploited a genome-scale loss-of-function screen to unravel the complex interplay between structurally diverse benzoxaboroles and T . brucei as a pathogenic model . Part of the initial identifications suggested a specific SAR between the function conferred by an individual gene and aminomethylphenoxy derivatives , mediated by the methylamine moiety shared by the compounds . This gene encodes a protein , TbALDH3 that potentially functions as a trypanosomal FALDH based on our crystallographic and biochemical characterizations . More importantly , further evidence indicates that TbALDH3 contributes to an oxidative deamination enzymatic pathway that is initiated by a host AO activity and is required to achieve full potency of aminomethylphenoxy benzoxaboroles as trypanocides ( Fig 7 ) . Overall , this finding brings a drug metabolism perspective to our understanding of the MoA for benzoxaborole-based pharmaceutical entities , as well as highlights the possibility of similar metabolic pathways present in broader pathological contexts . Indeed , alternative routes could be constructed by divergent host AOs and pathogen-derived ALDHs , metabolizing structurally diverse compounds containing an amine moiety . Interestingly , a novel chemical route has been proposed for improving the potency of the antibiotics against Gram-negative pathogenic bacteria [51] . One of the key aspects was to introduce an amine group into the candidate molecules , resulting in effective drug accumulation in the pathogen , which implies more general applications for the drug metabolism mechanism uncovered here . We suggest that the potential exploitation of similar metabolic pathways on aminomethyl-containing drugs could be of significant value . We also demonstrated that the carboxylate metabolites of the compounds exhibit the primary trypanocidal activity , although the same chemical entities , if applied directly , are less potent than the methylamine isomeric compounds and corresponding aldehyde metabolites . This lower potency with the carboxylate-containing benzoxaboroles is likely due to a less efficient drug uptake , as these compounds are negatively charged at neutral pH and hence might exhibit lower membrane permeability than uncharged aldehyde metabolites derived from methylamine pro-drugs . Noticeably , carboxylate benzoxaborole derivatives remain active as potent anti-parasitic candidates against Plasmodium falciparum and Toxoplasma gondii [11 , 13] . Therefore , the drug metabolism mechanism revealed here can be exploited to improve the drug potency of these candidate compounds by replacing the carboxylate group with a methylamine . It would be also interesting to explore the uptake mechanism of these carboxylate benzoxaborole derivatives . Our work indicates that the second oxidative reaction catalyzed by TbALDH3 likely occurs in glycosomes , therefore raising two possibilities of subcellular targeting mechanisms for the carboxylate benzoxaboroles as effective drugs , i . e . being either glycosomal or in alternative subcellular compartments . Nevertheless , the carboxylate derivatives such as AN2861 exhibit potent trypanocidal activities , independent of SSAO-TbALDH3 metabolic conversion . These data together suggest that the carboxylate benzoxaboroles are capable of overcoming the restriction , imposed by the negative charge of the compounds , on membrane permeability at both cellular and subcellular levels . This notion is also supported by our additional observation where there was no difference in the metabolomic profiles of parasites treated with AN3057 or with acoziborole ( S3 Table ) , suggesting that similar ultimate impacts on the parasite targets are likely shared by structurally diverse benzoxaborole derivatives . In addition to the oxidative deamination by AO-TbALDH3 , we also observed an oxidative deboronation process , which has also been described elsewhere , especially in host plasma , for divergent classes of benzoxaboroles , where it accounts for a compromise in the efficacy of these compounds [52 , 53] . Our observation raises a possibility that the oxidative enzymes in plasma such as AOs could be responsible for this oxidative deboronation reaction , therefore providing a possible route for improving the efficacy of benzoxaboroles in general . It also suggests the parasite culture system as a credible model in understanding drug metabolism through host and intravascular parasites . Meanwhile , the aldehyde metabolites derived from the amine-containing pro-drugs are relatively stable , allowing us to analyze them by MS and also to extract them for probing the activity of TbALDH3 directly in the parasites . It actually reflects the general feature of the ketone and aldehyde metabolites in various biological and pathological contexts such as diabetes and alcoholism[54] . Although further work is required to define the physiological functions of TbALDH3 , it is reasonable to speculate that long chain fatty aldehydes are the most likely substrates for TbALDH3 , based on phylogenetic and structural analyses , as well as a glycosomal localization of the enzyme . However , little is known in general about the role of glycosomal lipid metabolism in the adaptation of trypanosomes to the host environment , nor the potential roles of ALDHs in fatty acid metabolism in trypanosomes . Interestingly , adipose tissue has been recently demonstrated as a host reservoir for T . brucei [55] , suggesting that further understanding of TbALDH3 and related FALDHs could provide unique insight into host-parasite interactions and compartmentalization of metabolic functions .
Genetic determinants of drug resistance were identified using a T . brucei RNAi library screen as previously described [26] . Briefly , the library was grown under RNAi-inducing conditions ( tetracycline , Tet at 1 mg . ml-1 ) 24 h prior to benzoxaborole drug-selection at 1 . 5~2xEC90 . Through the screening , cultures were maintained and supplemented with fresh drug before extracting DNA from the drug-resistant population once it was established . RNAi target fragments within the population were amplified by PCR using LIB2f and LIB2r primers and the collective products were then subjected to high-throughput sequencing at BGI ( The Beijing Genome Institute ) . The workflow for sequencing was as follows; the sequence libraries were constructed with the PCR products fragmented to ~300 bp and subjected to Illumina HiSeqTM4000 . Reads were mapped to the T . brucei 927 reference genome ( v6 , tritrypdb . org ) with Bowtie 2 [56] using the parameters: very-sensitive-local–phred33 . Alignment files were manipulated with SAMtools [57] and a custom script [26] , and data were further assessed using the Artemis genome browser [58] . EC50 was determined by exposing cells to the test compound at serial dilutions ( 2X or 4X as indicated in the corresponding data ) with the highest concentration at 10 μM unless specified otherwise . The assays were conducted in 96-well plate format with three replicates for each sample and a non-treatment control in parallel [59] . The parasites were seeded at 1X104 cells/ml and cultured for 72 h before adding resazurin . After 6h of incubation with resazurin , the fluorescence from resorufin was measured with a Gemini Fluorescent Plate reader ( Molecular Devices ) at an excitation wavelength of 530 nm , an emission wavelength of 585 nm and a filter cut-off of 570 nm . Each dataset shown is representative of three independent experiments ( n = 3 ) ; in the case of RNAi , two independent RNAi clonal cultures were analysed . For expressing recombinant proteins , the complete Tb427 . 06 . 3050 ( TbALDH3 ) coding sequence was PCR amplified from genomic DNA and cloned into a modified pET27b vector ( Novagen ) creating pET27bTbALDH3 . The plasmid produces TbALDH3 with a N-terminal hexa-histidine-tag followed by a tobacco etch virus ( TEV ) protease cleavage site . Catalytic cysteine to serine ( C259S ) mutation was generated by site directed mutagenesis using the Quikchange protocol ( Stratagene ) . For RNAi , in brief , a gene-specific target sequence was selected through RNAit [60] and introduced into 2T1 cells using the pRPaiSL vector [61] . Upon Tet-induction , dsRNAs were generated from the target sequence and the sequence-dependent RNAi was initiated in the transformed cells . For tagging the native locus , a sequence encoding a 6xMyc epitope was introduced into the native loci of Tb427 . 06 . 3050 using the pNATxTAG vector , resulting in the strain expressing TbALDH3 with a C-terminal epitope . The integrity of all constructs was verified by sequencing . See S4 Table for the details of constructs and the primers for cloning . All the transformed T . brucei bloodstream strains were established and maintained as previously described [61] . Sequences of putative aldehyde dehydrogenases were retrieved from a BLASTp [62] search using Homo sapiens ALDH proteins as query against the genomes of the selected Opisthokont and Trypanosomatida species ( BLOSUM62 , Existence: 11 Extension: 1 Conditional compositional score matrix adjustment ) . The subsequent hits were analysed and outliers excluded from the phylogenetic assembly based on protein alignment generated using EMB-EBI Clustal Omega . The curated list of proteins was again analysed with Clustal Omega and manually edited to remove the poorly aligned regions of the C and N terminus . Phylogenetic reconstructions were generated using PhyML [63] and MrBayes [64] using the default parameters with a 1000 bootstrapping and 800000 generations , respectively . Resulting trees were visualized using FigTree ( http://tree . bio . ed . ac . uk/software/figtree/ ) . At least two members of each clade in the individual Opisthokont tree were selected and aligned with a representative of the Leishmania sp . , the Leptomonas sp . , the African and the American Trypanosoma sp . proteins for each clade to reconstruct the phylogeny of aldehyde dehydrogenases . No ALDH6 , ALDH16 or ALDH18 orthologues were identified in Trypanosomatida and their branches have been omitted from the Trypanosomatida-Ophistokont tree for simplification; ALDH families 7 , 8 and 9 are shown in grey for the same reason . See S5 Table for the details of the species and the sequences with accession numbers . Recombinant TbALDH3 protein was obtained by cultivating freshly transformed E . coli BL21 ( DE3 ) Rosetta Gami 2 ( Novagen ) in a starter culture of Luria-Bertani ( LB ) medium in 20 mL , supplemented with 50 μg/mL of kanamycin and 15 μg/mL of chloramphenicol . 5 mL of the starter culture was then inoculated with the selective LB culture ( 4 L ) with 1 mM magnesium chloride and 0 . 5 mM calcium chloride , grown at 37°C in 5 L Erlenmeyer flasks . OD600 was measured until it reached 0 . 6–0 . 8 , at which point expression was induced with 1 mM IPTG ( isopropyl-β-D-thiogalactopyranoside ) and the culture incubated overnight at 25°C . The cells were harvested by centrifugation at 5000 g for 30 minutes at 4°C and were resuspended in buffer A ( 50 mM sodium phosphate pH 7 . 8 , 300 mM sodium chloride , 10% glycerol ) along with a protease inhibitor tablet ( Roche ) . The cell-buffer A was passed through a pressure cell homogeniser ( SPCH-10 , Stansted ) , supplemented with 25 mM imidazole and passed through a 0 . 45 μm filter . A 1 mL CV ( column volume ) His-Trap HP column ( GE Healthcare ) charged with Ni2+ was calibrated by 4 CV washes with buffer B ( 50 mM sodium phosphate pH 7 . 8 , 25 mM imidazole , 300 mM sodium chloride , 10% glycerol ) after which the sample was loaded and fractions collected over a linear increase in imidazole concentration ( 25–250 mM over 18 CV ) . Fractions containing the eluted recombinant TbALDH3 protein were combined and concentrated in a spin concentrator ( Millipore ) exchanging the buffer to buffer A and the sample incubated overnight with His-tagged TEV protease ( molar ratio of 1:20 TEV:recombinant protein ) . The sample was passed through a His-Trap HP column previously equilibrated with a buffer A , to separate cleaved enzyme from non-cleaved protein and the TEV protease itself . Fractions containing TbALDH3 were concentrated and chromatographed through a 24 mL CV Superdex75 GL ( HR 10/300 , GE Healthcare ) size exclusion chromatography column equilibrated with buffer A . The size exclusion chromatography columns had been calibrated with molecular mass standards ( thyroglobulin , 670 kDa; gamma-globulin , 158 kDa; serum albumin , 67 kDa; ovalbumin; 44 kDa , myoglobin , 17 kDa; vitamin B12 , 1 kDa ) . Fractions containing purified TbALDH3 were combined , concentrated and used in crystallisation trials . The yield of TbALDH3 was estimated on the basis of theoretical molar extinction coefficients at 280 nm of 42860 M-1 cm-1 , respectively , calculated using the software VectorNTI ( Invitrogen ) . TbALDH3 crystals , maximum dimension 0 . 5 mm , were obtained in five days by sitting-drop vapour diffusion at 20°C in 3 μL drops in the ratio of 1:1 protein solution ( 20 mg/mL TbALDH3 in 50 mM sodium phosphate pH7 . 8 , 300 mM sodium chloride , 10% glycerol ) to reservoir condition ( 100 mM Bis-Tris pH 5 . 5 , 0 . 2 M LiSO4 , 25% PEG3350 ) . For co-crystallisation with the aldehyde form of AN3057 the purified catalytically inactive TbALDH3C259S was concentrated to 27 mg/ml in buffer A supplemented with 2 mM NAD+ , 2 mM MgCl2 and dialysed ( 3500 MWCO ) for 14 h at RT against the same buffer containing 10 μM MAO-A ( Sigma Aldrich ) and 1 mM AN3057 using a homemade microdialysis device , then used in the same crystallization setup as before . TbALDH3 crystals , though of good appearance , displayed diffraction indicating the presence of multiple crystalline components . This problem was more dominant in the co-crystals and numerous crystals had to be screened to identify a crystal with twinning-properties down to acceptable levels . Crystals were harvested using a glycerol or LV-oil ( MiTeGen ) cryoprotectant , flash frozen in liquid nitrogen and characterised in-house with a Rigaku MicroMax 007HF generator equipped with VariMax VHF optic , and a Saturn944 HG+ CCD detector . The data were indexed and merged using XDS [65] and SCALA [66] , respectively . The crystals displayed the space group P1211 . The crystal structure of the wild-type apoprotein was solved to 1 . 95 Å using the MR protocol of Auto-rickshaw [67] using a 432 residues search model built using PHYRE2 [68] . Density modification was performed using PIRATE [69] , as implemented in the CCP4 suite of programs [70] , and the model was extended to 920 residues with ARP/wARP [71] . Refinement with Refmac5 [72] resulted in Rwork and Rfree of 0 . 19 and 0 . 24 , respectively . This model was inspected , along with electron density and difference density maps , adjusted and extended to 973 residues using COOT [73] . Translation/Libration/Screw ( TLS ) refinement [74] in Refmac5 with multiple rounds of electron and difference density map inspection , model manipulation and the inclusion of water molecules , dual conformers and glycerol completed the refinement . The apoprotein structure served as a search model to phase native data collected from the isomorphous co-crystals to 2 . 5 Å by molecular replacement with PHASER . Rounds of model adjustment using COOT , interspersed with rounds of Refmac5 intensity based twin-refinement calculations , the addition and refinement of AN3057 aldehyde , NAD , water molecules , and inclusion of multiple conformers completed the refinement . Geometrical restraints for the AN3057 aldehyde intermediate were generated using the Grade Web Server ( http://grade . globalphasing . org ) . MOLPROBITY [75] and COOT were used to monitor model geometry during TbALDH3 refinement . Figures were prepared using PyMOL ( Schrödinger LLC ) . The DALI server was used to search the PDB for structural homologues and structural superpositions were performed using DALILITE [76] . Multiple sequence alignments were calculated using CLUSTALW2 [77] and edited using ALINE [78] . Crystallographic statistics are presented in S1 Table . AN3057 , ( 4- ( 1-hydroxy-1 , 3-dihydrobenzo[c][1 , 2]oxaborol-5-yloxy ) phenyl ) methanaminium chloride , was synthesized as described [79] , to a 4 . 78 mmol solution of 1 . 2 g of 4- ( 1-hydroxy-1 , 3-dihydrobenzo[c][1 , 2]oxaborol-5-yloxy ) benzonitrile in EtOH ( 150 mL ) under N2 was added Pd/C ( 10 wt . % , 0 . 178 g ) . The reaction mixture was hydrogenated for 26 . 5 hours using a H2 balloon at room temperature with stirring . The mixture was filtered , rotary evaporated and purified by silica gel column eluted with MeOH containing 0 . 6% volume NH4OH ( 3 mL 28–30% NH4OH to 500 mL MeOH ) . The white solid obtained was dissolved in water ( 80 mL ) and 6N HCl ( 2 mL ) was added , filtered and the filtrate was lyophilized to give the desired salt ( 4- ( 1-hydroxy-1 , 3-dihydrobenzo[c][1 , 2]oxaborol-5-yloxy ) phenyl ) methanaminium chloride as white solid ( 0 . 93 g , 3 . 19 mmol , yield 66 . 7% ) . M . p . > 250°C . 1H-NMR ( DMSO-d6 , 300 MHz ) : δ 9 . 18 ( s , 1H ) , 8 . 43 ( br . s , 3H ) , 7 . 74 ( d , J = 8 . 1 Hz , 2H ) , 7 . 52 ( d , J = 8 . 7 Hz , 2H ) , 7 . 08 ( d , J = 8 . 7 Hz , 1H ) , 6 . 98–6 . 94 ( m , 2H ) , 4 . 91 ( s , 2H ) and 3 . 99 ( br . q , J = 4 . 8 Hz , 2H ) ppm . Purity ( HPLC ) : 94 . 9% at 254 nm . MS: m/z = 256 ( M+1 , ESI+ ) and m/z = 255 ( M- , ESI- ) . The method is summarized in S10 Fig . AN2861 was described by Yong-Kang et al . [80] , while AN3054 , AN3056 , AN3330 , AN3410 were described by Akama et al . [81] . The chemicals purchased from Sigma are as follows , Tetracycline hydrochloride ( T7660 ) , Clorgyline ( M3778 ) , Semicarbazide ( S2201 ) , Tetraphenylphosphonium chloride ( 218790 ) , Sodium diethyldithiocarbamate trihydrate ( D3506 ) , 2-Bromoethylamine hydrobromide ( B65705 ) , Mexiletine hydrochloride ( M2727 ) , Ammonium tetrathiomolybdate ( 323446 ) , Resazurin ( R7017 ) , anti-Myc ( commercial ) . Samples were analyzed on a Waters Xevo Q-TOF Mass Spectrometer coupled to a Waters Acquity UPLC system using a Waters BEH C18 column 50x2 . 1mm 1 . 6 μm . Analysis was done in either acidic ( A: H2O + 0 . 1% formic acid , B: ACN + 0 . 1% formic acid ) or basic eluents ( A: H2O + 0 . 1% NH3 solution , B: ACN + 0 . 1% NH3 solution ) . The same gradient was used with both types of eluent: 98% A for 0 . 5 min then linear gradient to 65% A ( 3 . 5 min ) then to 5% A in 1 min , gradient held for 1 min before a step change back to the to the starting conditions and an equilibration time of 1 min ( total run time 7 min ) . The mass spectrometer was operated in both positive ( Capillary Voltage 2 . 3 kV ) and negative mode ( Capillary Voltage 1 . 5 kV ) . Source temperature and desolvation gas temperatures were constant at 120°C and 500°C respectively . In MS operation , spectra were acquired every 0 . 2 s over a 50–1000 amu range . MAO-Glo assay ( Promega , V1401 ) was applied to determine MAO and SSAO activities respectively according to manufacturer’s instructions . Recombinant human MAO A ( Sigma , M7316 ) was included as the positive control , and the assay was conducted with following the optimized protocol [82] . TbALDH3 localization was determined using a native C-terminal 6xMyc epitope fusion . Samples were prepared as previously described . Antibodies were used at the following dilutions: rabbit anti-myc epitope IgG ( Santa Cruz Biotechnology Inc . ) at 1:500 , and mouse anti-TbGAPDH ( a kind gift from M . A . J . Ferguson , Dundee ) at 1:1000 . Secondary antibodies ( Life Technologies ) were Alexa Fluor 568 conjugated anti-rabbit IgG ( 1:1000 ) and Alexa Fluor 488 conjugated anti-mouse ( 1:1000 ) . Coverslips were mounted using Prolong Gold mounting medium supplemented with 4’ , 6-diamidino-2-phenylindole ( DAPI ) ( Life Technologies ) . The cells were examined on a Zeiss Axiovert 200M microscope and images captured with a AxioCam MRm camera . Digital Images were captured and processed using Zen Pro software ( Zeiss ) and Adobe Photoshop CS3 ( Adobe Systems Inc . ) . Metabolite extractions were performed as previously described [83] . Briefly , bloodstream form cells were treated with 5xEC50 of drug in HMI-9 for the time required for growth to be inhibited ( 6 hours for AN5568 , 8 hours for AN3057 ) . 1 x 108 cells were centrifuged and extracted for one hour , shaking in 200 μL UPLC grade chloroform:methanol:water ( 1:3:1 ) on ice . Samples were centrifuged and stored at -80°C before being run on a QExactive mass spectrometer ( Thermo ) after separation on a zic-pHILIC column ( Sequant ) according to previously published methods . Raw data were filtered and aligned and annotated using the Orbiwarp algorithm in PiMP ( http://polyomics . mvls . gla . ac . uk/ ) was used to match masses and retention times to authentic standards , to provide annotations and to perform statistical analyses . | Human African Trypanomiasis ( HAT ) is among a list of Neglected Tropical Diseases ( NTDs ) that impose devastating burdens on both public health and economy of some of the most unprivileged societies across the world . To secure the long-term global control of the disease , it is critical to understand the mechanisms underlying the interactions of drugs and drug candidates with the causative agents as well as resistance potentially arising from use of the compounds . We demonstrated here a metabolic enzymatic cascade dependent on a host-pathogen interaction that determines potency against T . brucei of a series of benzoxaborole compounds . More importantly , this pathway represents a metabolic interaction network between host and pathogen , illuminating an important perspective on understanding mechanism of action . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"chemical",
"compounds",
"enzymology",
"organic",
"compounds",
"parasitic",
"protozoans",
"trypanosoma",
"brucei",
"protozoans",
"metabolites",
"enzyme",
"metabolism",
"pharmacology",
"epigenetics",
"drug",
"... | 2018 | Host-parasite co-metabolic activation of antitrypanosomal aminomethyl-benzoxaboroles |
Cell migration in the absence of external cues is well described by a correlated random walk . Most single cells move by extending protrusions called pseudopodia . To deduce how cells walk , we have analyzed the formation of pseudopodia by Dictyostelium cells . We have observed that the formation of pseudopodia is highly ordered with two types of pseudopodia: First , de novo formation of pseudopodia at random positions on the cell body , and therefore in random directions . Second , pseudopod splitting near the tip of the current pseudopod in alternating right/left directions , leading to a persistent zig-zag trajectory . Here we analyzed the probability frequency distributions of the angles between pseudopodia and used this information to design a stochastic model for cell movement . Monte Carlo simulations show that the critical elements are the ratio of persistent splitting pseudopodia relative to random de novo pseudopodia , the Left/Right alternation , the angle between pseudopodia and the variance of this angle . Experiments confirm predictions of the model , showing reduced persistence in mutants that are defective in pseudopod splitting and in mutants with an irregular cell surface .
Eukaryotic cells move by extending pseudopodia , which are actin-filled protrusions of the cell surface [1] . Pseudopod formation by Dictyostelium cells , like many other moving cells , shows a typical pseudopod cycle: upon their initiation , pseudopodia grow at a constant rate during their first ∼15 s and then stop . The next pseudopod is typically formed a few seconds later , but sometimes commences while the present pseudopod is still growing , giving rise to a cell with two pseudopodia . The fate of the pseudopod after its initial growth phase determines its role in cell movement: the pseudopod is either retracted , or is maintained by flow of the cytoplasm into the pseudopod thereby moving the cell body . The frequency , position and directions of the maintained pseudopodia form the basis of cell movement , because they determine the speed and trajectory of the cell . An important aspect of cell motility is the ability of cells to respond to directional cues with oriented movement . Gradients of chemicals give rise to chemotaxis [2] . Other directional cues that can induce oriented movement are temperature gradients ( thermotaxis ) or electric fields ( electrotaxis ) [3] , [4] . These signals somehow modulate basal pseudopod extension such that , on average , cells move in the direction of the positional cues . In this respect , studies on cell movement are critical for understanding directional movement . Cells in the absence of external cues do not move in random directions but exhibit a so-called correlated random walk [5]–[9] . This tendency to move in the same direction is called persistence , and the duration of the correlation is the persistence time . Cells with strong persistence make fewer turns , move for prolonged periods of time in the same direction , and thereby effectively penetrate into the surrounding space . Other search strategies for efficient exploration are local diffusive search and Levi walks [8] , [10] . Can we understand the cell trajectory by analyzing how cells extend pseudopodia ? To obtain large data sets of extending pseudopodia we developed a computer algorithm that identifies the cell contour and its protrusions . The extending pseudopod is characterized by a vector that connects the x , y , t coordinates of the pseudopod at the beginning and end of the growth phase , respectively [11] . A picture of ordered cell movement has emerged from the analysis of ∼6000 pseudopodia that are extended by wild type and mutant cells in buffer [12] . Dictyostelium cells , as many other eukaryotic cells , may extend two types of pseudopodia: de novo at regions devoid of recent pseudopod activity , or by splitting of an existing pseudopod [12] , [13] . Pseudopod splitting occurs very frequently alternating to the right and left at a relatively small angle of ∼55 degrees . Therefore , pseudopod splitting may lead to a persistent zig-zag trajectory [14] . In contrast , de novo pseudopodia are extended in all directions and do not exhibit a right/left bias , suggesting that de novo pseudopodia induce a random turn of the cells . We observed strong persistence for cells that extend many splitting pseudopodia . Conversely , mutants that extend mostly de novo pseudopodia have very short persistence time and exhibit a near Brownian random walk [12] . In this report we investigated the theory of correlated random walks in the context of the observed ordered extension of pseudopodia . The aim is to define the descriptive persistence time or average turn angle with primary experimentally-derived pseudopod properties . First we obtained detailed quantitative data on the probability frequency distributions of the size and direction of pseudopod activity . We then formulated a model that consists of five components: pseudopod size , fraction of splitting pseudopodia , alternating right/left bias , angle between pseudopodia and variance of this angle due to irregularity of cell shape . We measured the parameter values of these components for several Dictyostelium mutants with defects in signaling pathways or cytoskeleton functions . Subsequently , we used these observed parameters in Monte Carlo simulations of the model and compared the predicted trajectories with the observed trajectories of the mutants . The results demonstrate two critical components in these correlated random walks: the ratio of pseudopod splitting relative to de novo pseudopodia , and the shape of the cell .
The strains used are wild type AX3 , pi3k-null strain GMP1 with a deletion of pi3k1 and pi3k2 genes [15] , pla2-null with a deletion of the plaA gene [16] , sgc/gca-null cells ( abbreviated as gc-null cells ) with a deletion of gca and sgc genes [17] , sgc/pla2-null cells with a deletion of sgc and pla2A genes [18] , and ddia2-null cells lacking the forH gene encoding the Dictyostelium homologue of formin [19] . Cells were grown in HG5 medium ( contains per liter: 14 . 3 g oxoid peptone , 7 . 15 g bacto yeast extract , 1 . 36 g Na2HPO4⋅12H2O , 0 . 49 g KH2PO4 , 10 . 0 g glucose ) , harvested in PB ( 10 mM KH2PO4/Na2HPO4 , pH 6 . 5 ) , and allowed to develop in 1 ml PB in a well of a 6-wells plate ( Nunc ) . Movies were recorded at a rate of 1 frame per second for at least 15 minutes with an inverted light microscope ( Olympus Type CK40 with 20× objective ) and images were captured with a JVC CCD camera . Cell trajectories were recorded as the movement of the centroid of the cell as described [20] . Images were analyzed with the fully automatic pseudopod-tracking algorithm Quimp3 , which is described in detail [11] . Briefly , the program uses an active contour analysis to represent the outline of the cell using ∼150 nodes [21] . Extending pseudopodia that satisfied the user-defined minimum number of adjacent convex nodes and the minimum area change were identified . The direction of each extending pseudopod was identified by the x , y and time coordinates of the central convex node of the convex area at the start and end of growth , respectively . The tangent to the surface at the node where the pseudopod started was calculated using the position of the adjacent nodes . The automated algorithm annotates each pseudopod as de novo versus splitting using the criterion that the convex area of the new pseudopod exhibits overlap with the convex area of the current pseudopod or is within a user-defined distance . The output files containing the x , y-coordinates of the start and end position of the pseudopod , the tangent of the surface and the annotation of the pseudopod were imported in Excel to calculate pseudopod size , interval , direction to gradient , direction to tangent , etc for de novo and splitting pseudopodia ( see Fig . 1 ) , as well as fraction s of pseudopod splitting and alternating Right/Left bias a ( RL +LR ) /total splitting; Table 1 ) . The cell shape parameter Ψ was determined as follows: Using the outline of the cell with ∼150 nodes , two ellipsoids were constructed , the largest ellipse inside the cell outline and the smallest ellipse outside the cell outline . Then an intermediate ellipse was constructed by interpolation of the inner and outer ellipse . This intermediate ellipse makes several intersections with the cell outline , thereby forming areas of the cell that are outside the intermediate ellipse ( with total surface area O ) , and areas of the intermediate ellipse that do not belong to the cell ( with total surface area I; see Fig . S4 ) . The intermediate ellipse was positioned in such a way that ( this also implies that the surface area of the cell ( T ) is identical to the surface area of the intermediate ellipse ) . The cell shape parameter is defined as ; it holds that . For a cell with a regular shape that approaches a smooth ellipsoid , the surface areas O and I are very small and Ψ approaches zero . In contrast , O and I are larger for a cell with a very irregular shape; the largest value observed among ∼600 cells was Ψ = 0 . 92 . With the exception of 5h starved cells , each database contains information from 200–300 pseudopodia , obtained from 6–10 cells , using two independent movies . For 5h starved cells , we collected a larger database containing 835 pseudopodia from 28 cells using 4 independent movies , and typical databases for each mutant . The data are presented as the means and standard deviation ( SD ) or standard error of the means ( SEM ) , where n represents the number of pseudopodia or number of cells analyzed , as indicated in Table 1 . The probability density functions of angles can not be analyzed as the common distribution on a line . Angular distributions belong to the family of circular distributions , which are constructed by wrapping the usual distribution on the real line around a circle . The data were analyzed with two circular distributions , the von Mises distribution ( vMD ) , which matches reasonably well with the wrapped normal distribution , and the wrapped Cauchy distribution ( WCD ) , which has fatter tails [22] . The vMD is given by ( 1 ) where I0 ( κ ) is the modified Bessel function of the first kind of order zero ( 2 ) The WCD is given by ( 3 ) Pseudopod extension is an ordered stochastic event [12] . The position of the tip of the formed pseudopodia depends on pseudopod size λp , splitting fraction s , Left/Right alternating ratio α , angle between split pseudopodia φ and variance of this angle σφ . A Monte Carlo simulation starts with a random angle α ( 1 ) of the first pseudopod . For the next and all subsequent pseudopodia the simulation uses four uniformly distributed random numbers Ri , n ( i = 1 , . . , 4 ) to calculate α ( n ) , the angle of the nth pseudopod: with the decision to split if R1 , n <s; with the decision for alternating splitting if R2 , n <a; for direction of split after de novo with decision right if R3 , n <0 . 5; and for the direction of the de novo pseudopod . These probabilities result in a projected angle of extension in degrees . Finally , the actual pseudopod direction is drawn from a wrapped von Mises distribution with this projected angle as mean and σφ2 as variance ( κ = 1/σφ2; variance converted to radians ) . The obtained α ( n ) and the pseudopod size λp are used to calculate the x , y coordinates of the tip of the pseudopod , followed by a next round of four random numbers to calculate α ( n+1 ) . In the simulations reported here we did not include stochastic variation in pseudopod length and pseudopod frequency , since we observed that they had only minor effects on the trajectories over several cell lengths . Please note that in the simulations the direction of the simulated de novo pseudopodia is random; consequently , a small fraction of de novo pseudopodia are in the same direction of the previous pseudopod , which would be recognized in experiments as splitting pseudopodia . Conversely , a small fraction of the simulated splitting pseudopodia have angles much larger than 55 degrees and would be recognized in experiments as de novo pseudopodia . From the geometry of the cell , we estimate that the number of simulated de novo in the current pseudopod and the number of splitting pseudopodia outside the current pseudopod are approximately the same , suggesting that the simulations represent the observed ratio of splitting and de novo pseudopodia .
The angles between pseudopodia were analyzed in detail and the results are presented in Fig . 2 . For splitting pseudopodia , the angle between the current and next pseudopod ( φ1 , 2 ) has a clear bimodal distribution ( Fig . 2A ) . A probability density function ( PDF ) of angles belongs to the family of circular or wrapped distributions . The data reported in this study were all fitted well by a von Mises distribution ( vMD ) , which is the circular analog of the normal distribution . The wrapped Cauchy distribution has fatter tails and provided a poorer fit of the data ( data not shown ) . The bimodal vMD presented in Fig . 2A is symmetric , yielding two means ( φ1 , 2 = +/−55 ) that have the same variance κ = 1/σφ2; σφ1 , 2 = 28 degrees ) . Figure 2B shows the PDF of the angle between the current and next-next pseudopod ( φ1 , 3 ) , which is best described by a single vMD with a mean of φ1 , 3 = 2 degrees and σφ1 , 3 = 42 degrees . Figure 2C reveals that there is no significant correlation between the magnitude of angles between first/second pseudopod and the magnitude of the angles between second/third pseudopod ( thus e . g . splitting at a larger angle is not followed by a split at a smaller angle ) . The extension of splitting pseudopodia is summarized in Fig . 2D , and is based on the previous observation that a pseudopod split to the right is frequently followed by a split to the left and visa versa [12] . Thus the next pseudopod is extended at an angle of ∼55 degrees to the right or left relative to the current pseudopod , and the next-next pseudopod is extended in roughly the same direction as the current pseudopod . The angle between a de novo pseudopod and the previous pseudopod shows a very broad distribution ( Fig . 2E ) . Nearly all angles between −180 and +180 are well represented with a somewhat lower abundance of angles around 0 degrees . This suggests that a de novo pseudopod can be extended in any direction , but with slightly lower probability of the direction of the current pseudopod . To investigate the consequence of the observed ordered extension of pseudopodia for cell movement on a coarse time scale for many pseudopodia we recorded the movement of Dictyostelium cells during 15 minutes; in this period about 30 pseudopodia are extended . Previously we have presented the cell trajectories for several strains and developmental stages [12] ( see also Fig . S1 ) . The mean square displacement as a function time , , exhibits a slow approach to a linear function ( Fig . 3A ) , which is typical for a transition of a correlated random walk at short times to a Brownian random walk after longer times [6] , [23] . Previously , the often used equation for a correlated random walk were fit to the data points to estimate persistence time and speed of the cells [12] . The aim of the present study is to analyze the mechanism of cell movement from the perspective of the extending pseudopodia , which have a specific length and direction . A correlated random walk in two dimensions can also be described with steps and turns [24] , [25] . With the replacement of the number of steps ( n ) in Eq . 7 in reference [25] for n = Ft we obtain ( 4 ) where λ is the step size in µm , F is the step frequency , and γ is the correlation factor of dispersion ( 0<γ<1 ) , defined as the arithmetic mean of the cosine of the turn angle θ between steps ( 5 ) With three variables ( F , λ , γ ) the estimates of the parameters become uncertain . Fortunately , the step size can be deduced accurately from experimental data . As will be shown below in Eq . 10 , the step size is given by λ = λpcos ( φ/2 ) , where measurements for λp and φ are presented in Table 1 . Using this value for λ , the dispersion data were fitted to obtain the observed correlation factor of dispersion ( γobs ) with the corresponding turn angle ( θ ) . In cells starved for 1 or 3 hours the correlation factor is only ∼0 . 5 with turn angle of ∼60 degrees . At 5 and 7 hours of starvation , cells move with much stronger persistence ( correlation factor of 0 . 74 and 0 . 81 and a small turn angle of 42 and 36 degrees ) . Deletion of PLA2 or guanylyl cyclases prevents this increase of correlation factor , persistence is very low and cells disperse poorly . How is pseudopod extension related to the observed correlation factor of dispersion γobs ? As previously stated ( see Fig . 2 ) , Dictyostelium cells may extend either de novo pseudopodia in nearly random directions , or splitting pseudopodia in a direction similar to the previous direction . Therefore , cells that extend exclusively de novo pseudopodia are expected to exhibit a random walk with γobs = 0 ( turn angle θ = 90 degrees ) , whereas cells extending exclusively splitting pseudopodia will exhibit strong persistence with large γ and small turn angle θ . As a consequence , γobs is expected to depend on the ratio s of splitting/de novo pseudopodia . Fig . 3B demonstrates that within experimental error this relationship is approximately linear; this holds true for the mutants as well as for wild type cells at different stages of development . The linear regression of all data yields γobs = 0 . 921s−0 . 044 . Thus , when all pseudopodia are de novo ( s = 0 ) the correlation factor is small ( γobs = −0 . 044 ) giving a turn angle = 93 degrees , close to the expected value of 90 degrees for random turns . In contrast , when all pseudopodia are the result of splitting ( s = 1 ) the correlation factor is large ( γobs = 0 . 88 ) yielding a small turn angle ( = 29 degrees ) . The implication of small turn angles for splitting pseudopodia will be discussed later . The alternating right/left extension of splitting pseudopodia can be used to simplify a description of the movement of Dictyostelium cells over longer distances . In this approach , the simplification may be valid for movement on a longer time scale only , as we study here , but may not be appropriate over shorter time scales of a few pseudopodia . Because pseudopodia are frequently extended alternating right/left , we consider movement by pairs of two pseudopodia . Figure 4 shows four possibilities of pairs of splitting pseudopodia , which are the RL , LR , RR and LL , each with corresponding probabilities and angles as indicated . In addition to these splitting pairs , three combinations with de novo pseudopodia are possible: split-de novo , de novo-split , and de novo-de novo . The correlation factor of dispersion yields for the seven pairs: ( 6 ) De novo pseudopodia are extended in a random direction , i . e . , and equal zero . The turn angles of the four splitting pairs are 0 , φ and 2φ , as indicated in Fig . 4A , and the variance is approximately 2σφ2 ( see Fig . 2B ) . Consequently Eq . 6 reduces to: ( 7 ) where denotes the expected value of the cosines of the angles on a circle with weights given by the vMD with mean φ and variance given by κ = 1/ ( 2σφ2 ) . Since all splitting pseudopodia show the same variance this can be further reduced to ( 8 ) In this equation is obtained by calculating the probabilities of all turn angles on a circle with the vMD using Eqs . 1 and 2 and then taking the weighted average of the cosines of these angles . Although this procedure is straightforward , Eq . 8 can be further simplified , because for σφ smaller than ∼50 degrees a good approximation is ( see Fig . S2 ) . Finally , on a longer time scale and averaged over many steps , the correlation factor of pairs is related to the correlation factor of its underlying two steps by . With these replacements we obtain the analytical expression for the correlation factor ( 9 ) Thus , the correlation factor γ is the product of three terms: the splitting ratio s , a noise term with the variance σφ , and a term with right/left bias a and angle φ . Finally , by considering movement in pairs of steps , Fig . 4 reveals that the step size of the displacement is given by ( 10 ) We used Monte Carlo simulations to investigate how λ and γ depend on the pseudopod parameters size λp , splitting fraction s , alternating ratio a , angle between split pseudopodia φ and variance of this angle σφ2 . These simulations are also useful to inspect whether step size λ and correlation factor γ are correctly described by Eqs . 8–10 . The direction in which a pseudopod is extended appears to be an ordered stochastic event [12] that depends on multiple decisions according to the following scheme: The next pseudopod is a splitting or de novo according to the ratio s . A splitting pseudopod is extended with angle φ to the right or left relative to the previous splitting according to the alternating right/left bias a . A de novo pseudopod is extended in a random direction . The splitting pseudopod that appears after a de novo pseudopod has an equal probability to be extended to the left or right . Finally , the direction of the emerging pseudopod has a variance σφ2 . The Monte Carlo simulation starts with a random angle α ( 1 ) of the first pseudopod and then uses the probabilities for splitting fraction s , alternating ratio a , angle between split pseudopodia φ and variance of this angle σφ2 to stochastically simulate the angle of the next pseudopod ( see methods and Table 1 for pseudopod parameters of wild type and mutant cells ) . The simulated trajectories are qualitatively similar to the experimentally observed trajectories ( see Fig . S1B ) : Fed wild type cells or mutants with abundant de novo pseudopodia make many turns and have small displacement , whilst the trajectories of starved wild type cells with abundant pseudopod splitting are more persistent with large displacements . To investigate how the correlation factor γ depends on pseudopod parameters , the displacement was calculated from 100 , 000 trajectories obtained by MC simulation using a unit pseudopod size and different values of s , a , φ and σφ . The obtained displacement was then fitted to Eq . 4 to obtain estimates for the step size λ and Monte Carlo correlation factor , γMC . The symbols in Fig . 5 show the results of the MC simulation , whereas the curves are the result of Eqs . 8–10 . We first investigated the angle φ between splitting pseudopodia and the alternating right/left bias a . When all splitting pseudopodia are alternating ( a = 1 ) , the cells make a nearly perfect zig-zag trajectory , and therefore the angle φ has very little effect on the persistence factor γ ( Fig . 5A ) . When splitting pseudopodia are extended in a random fashion to the right or left ( a = 0 . 5 ) , the persistence factor γMC decreases sharply as φ becomes larger than ∼30 degrees . At an intermediate right/left bias ( a = 0 . 75 ) the persistence factor γMC remains relatively high as long as the angle between pseudopodia is below 60 degrees . The results of the MC simulation appear to be described very well by the simplified model ( Eqs . 8–10 ) . Furthermore , at the observed angle of φ = 55 degrees and alternating factor of a = 0 . 77 , the deduced persistence factor γMC is 0 . 88 ( see asterisk in Fig . 5A ) . The fraction of splitting pseudopodia has a major impact on the persistence factor γ . In the MC simulations , the value of γ declines approximately linearly with the value of s ( Fig . 5B ) , as was also observed experimentally ( Fig . 3B ) , and obtained in Eqs . 8 and 9 . Finally , we investigated the contribution of the variance σφ2 of the splitting angle to the persistence factor γ . This reveals that the persistence decreases strongly with increasing variance ( Fig . 5C ) , with γMC following an approximately linear relationship with cos ( σφ ) . The MC simulations are well described by Eq . 8 , but deviate from Eq . 9 at σφ>30 degrees , as expected ( see Fig . S2 ) . We also used these Monte Carlo simulations to obtain an estimate of the step size λ . It appears that λ does not to depend on s and a , but depends on φ according to ( Inset Fig 5A ) , as was obtained in Eq . 10 . In summary , the obtained correlation factor from the MC simulation ( γMC ) are nearly identical to the correlation factor calculated with Eq . 9 ( γstep ) . This suggests that the movement of Dictyostelium cells is qualitatively and quantitatively described very well by the model of persistent steps and random turns with the observed pseudopod parameters λp , s , a , φ and σφ . How does the movement of pseudopodia relate to the movement of the centroid of the cell ? The data presented in Table 1 reveal that the observed correlation factor γobs of the centroid for different cell types correspond well with the deduced correlation factors of the pseudopods ( γMC and γstep ) , but is always larger by ∼15% ( Table 1 ) . Apparently , the observed turn angle of the cell's centroid is smaller than the turn angle of the extending pseudopod . Inspection of movies of 5h starved AX3 cells confirm this notion: the average angle between splitting pseudopodia is 55±28 degrees ( Fig . 2A ) , while the centroid moves during period at an angle of only 31±23 degrees ( mean and SD ) . Equation 9 reveals that the correlation factor γstep increases by 15% when φ = 55±28 degrees for the pseudopod is replaced by φ = 31±23 degrees for the cells centroid . Probably two phenomena are responsible for the difference between pseudopod and centroid: extension of multiple pseudopodia and geometry of cells . When cells extend multiple pseudopodia it is likely that at any given instant of time , the front of the cell moves with a fixed fraction of the vector sum of velocities possessed by the pseudopodia active at that instant in time . The temporal overlap of two pseudopodia was deduced from the measured probability distributions of pseudopod extensions ( Fig 2F in [26] ) , which reveal that ∼25% of the pseudopodia overlap with another pseudopod during on average ∼40% of their extension time . This suggest that the tip of the cell moves at an angle that is ∼6 degrees smaller than 55 degrees . Secondly , geometry predicts that the rear of the cell makes smaller changes of direction than the tip of the cell , comparable to the differences in curvature made by the front and rear wheels of a car . Figure S3 indicates that for a stereotypic pseudopod at 55 degrees the directional change of the centroid is ∼40 degrees ( see Fig . S3 ) . Together , multiple pseudopodia and cell geometry can explain observed difference between pseudopod and centroid changes of direction , leading to the small 15% difference between deduced pseudopod correlation factor ( γMC and γstep ) and observed centroid correlation factor ( γobs ) . The directional displacement is the displacement after n steps in the direction of the first step . An expression for the directional displacement is especially relevant when the organism is exposed to positional cues leading to a drift in one direction , such as during chemotaxis . The directional displacement of a cell after extending one pseudopod at an angle θ is , and for a population of cells . By Eqs . 3 and 10 , the displacement at the first step may be written as , and at the ith step , see Eq . 6 of reference [25] . The cumulative displacement after n steps is ( 11a ) which at is given by ( 11b ) In essence , this equation describes the displacement of a cell population in which all cells extend the first pseudopod in the same direction . Subsequent pseudopodia are extended with a bias , which reduces geometrically with each step; the correlation factor γ indicates how many pseudopodia have correlated direction and therefore how far the population will disperse in the direction of the first pseudopod . Figure 6 presents the directional displacement as observed experimentally in wild type cells . The displacement in the direction of the first pseudopod slowly decreases at each subsequent pseudopod , approaching random movement after ∼10 pseudopodia . On average a cell moves ∼15 µm in the direction of the first pseudopod , which is the equivalent of about 3 pseudopodia ( given a pseudopod size of ∼5 µm ) . This figure also presents the directional displacement as modeled by Eq . 11a with observed data for λp , φ and γ , which is in very close agreement with experimental data , again suggesting that the movement of a cell is satisfactory described by the model with five pseudopod parameters . The variation in pseudopod direction σφ2 plays an important role in Eqs . 8–11 describing cell dispersal . Previously [12] we have shown that the next pseudopod emerges at a specific distance d from the tip of the current pseudopod , and is then extended perpendicular to the cell surface ( i . e . perpendicular to the tangent to the surface curvature at the position where the pseudopod emerges ) . The pseudopod direction is expected to have high confidence for cells with a smooth ellipsoid shape , because the local bending is very predictable . However , this confidence is much smaller for cells with a very irregular shape . We investigated the role of cell shape using three experiments . First we demonstrate that the variance σφ2 indeed depends on the variance of the tangent and the normal to the tangent . Second , we show that wild type or mutant cells with irregular shape exhibit increased variance σφ2 . Finally we show that , due to the increased variance , the mutant exhibits poor dispersal . Quimp3 was used to construct the tangent to the surface curvature at the position where the pseudopod emerges . We first determined for wild-type cells the angle αt of this tangent relative to the previous pseudopod ( αt = 34 . 5±24 . 9 degrees ) , and the angle β of the new pseudopod relative to this tangent ( β = 89 . 1±13 . 3 degrees ) . As mentioned above , the observed angle of the new pseudopod relative to the previous pseudopod is φ = 55 . 2±27 . 8 degrees . We expect that the angle of the tangent relative to the previous pseudopod is independent from the angle of the pseudopod relative to the tangent; therefore we expect . Indeed , the observed standard deviation of 27 . 8 degrees is close to this expected value of 28 . 2 degrees . Importantly , the largest contribution to σφ2 is derived from the variance of the tangent σt2 , which is related to the local shape of the cell . In the collection of Dictyostelium mutants , we selected a strain with an irregular shape . Mutant ddia2-null with a deletion of the forH gene encoding the formin dDia2 has a star-like shape ( Fig . 7C ) . In this mutant , new pseudopodia are extended at about the same frequency and distance from the present pseudopodia as in wild type cells , pseudopodia also grow perpendicular to the surface , and are extended roughly in the same direction of φ = 55 degrees as wild type cells ( Fig . 7A ) . However pseudopodia exhibit much more variation in direction ( σφ = 47 degrees compared to σφ = 28 degrees for wild type cells ) . Finally , we determined a shape parameter Ψ that indicates how much the cell outline deviates from an ellipse ( see method section and Fig . S4 ) . Figure 7B reveals that cells with increased irregular shape , either being wild-type or mutant , exhibit strongly increased variance σφ2 . Importantly , the distance d and angle φ of the pseudopodia does not change with cell shape ( Fig . 7A ) . Using the observed values for s , a , φ , and σφ for ddia2-null cells we expect from Eq . 9 to obtain γstep = 0 . 43 , significantly lower compared to γstep = 0 . 69 . for wild type cells . Fig . 7D shows that the dispersion of ddia2-null cells is strongly reduced . The observed mean square dispersion was fitted to Eq . 4 yielding a correlation factor of γobs = 0 . 53 ( Table 1 ) , close to the value that was predicted from the extension of pseudopodia from an irregular surface . In summary , these and previous results [12] suggest that a splitting pseudopod is induced at some distance d from the tip of the current pseudopod , and then grows perpendicular to the surface . In a cell with a regular shape , the tangent and therefore pseudopod direction can be approximated using the distance d; alternating R/L extensions lead to a relative straight zig-zag trajectory , providing strong persistence of movement . In a cell with a very irregular shape , the local curvature of the membrane at distance d is unpredictable . Consequently , alternating R/L splitting occur with large variation of directions , leading to frequent turns and poor persistence .
The movement of many organisms in the absence of external cues is not purely random , but shows properties of a correlated random walk . The direction of future movement is correlated with the direction of prior movement . For organisms moving in two dimensions , such as most land-living organisms , this implies that movement to the right is balanced on a short term by movement to the left to assure a long-term persistence of the direction . In bipedal locomotion , the alternating steps with the left and right foot will yield a persistent trajectory . Amoeboid cells in the absence of external cues show ordered extension of pseudopodia: a new pseudopod emerges preferentially just after the previous pseudopod has stopped growth [12] . Importantly , the position at the cell surface where this new pseudopod emerges is highly biased . When the current pseudopod has been extended to the left ( relative to the previous pseudopod ) , the next pseudopod emerges preferentially nearby the tip at the right side of the current pseudopod . Since pseudopodia are extended perpendicular to the cell surface , this next pseudopod is extended at a small angle relative to the current pseudopod [12] . Therefore , this ( imperfect ) alternating right/left pseudopod splitting resembles bipedal locomotion . Cells may also extend a de novo pseudopod somewhere at the cell body , which is extended in a random direction . In starved Dictyostelium cells , the probability of extending a de novo pseudopod is ∼10-fold lower than of pseudopod splitting ( probability calculated per µm circumference of the cell [12] ) . The model for pseudopod-based cell dispersion depends on five parameters , the pseudopod size ( λ ) , the fraction of split pseudopodia ( s ) , the alternating left/right bias ( a ) , the angle between pseudopodia ( φ ) and the variance of this angle ( σφ2 ) . With these parameters the experimental data on mean square displacement and directional displacement are well-explained using Eqs . 9 and 11 , respectively . Pseudopodia are the fundamental instruments for amoeboid movement . The notion that the trajectories are described well by the five pseudopod parameters probably implies that we have identified the basic concept of the amoeboid correlated random walk: persistent alternating pseudopod splitting and formation of de novo pseudopodia in random directions . The cells may modify one or more of these five pseudopod parameters in order to modulate the trajectories ( see Table 1 ) . Nearly all mutants , as well as wild type cells at different stages of starvation and development , have approximately the same average pseudopod size λp . In addition , the alternating right/left bias ( a ) fluctuates between 0 . 67 and 0 . 82 , and the angle between splitting pseudopodia ( φ ) between 50 and 62 degrees . [pla2-null cells are the only exception [12]; emerging pseudopodia in pla2-null cells exhibit longer growth periods ( ∼27 s ) than wild type cells ( ∼13 s ) , and are thus longer] . This suggests that all strains use the same mechanism for pseudopod splitting . In contrast to these constant properties of split pseudopodia , the fraction of split pseudopodia ( s ) changes dramatically upon starvation , and appears to be regulated by cGMP and PLA2 signaling . Well-fed cells extend pseudopodia that are predominantly de novo in random directions , leading to a nearly Brownian random walk [20] . Upon starvation , the appearance of cGMP and PLA2 signaling enhances splitting and suppresses de novo pseudopod extensions , which leads to more persistent movement . The important role of the fraction of splitting pseudopodia for cell movement is also depicted by the linear dependence of the correlation factor γ on the fraction s of splitting pseudopodia ( Fig . 3B and Eq . 9 ) . The variance of the angle of pseudopod extension ( σφ2 ) plays an important role in movement . In wild type cells , as well as in many mutant strains , σφ is about 28 degrees . The primary source of the variance of pseudopod angles lies in the variation of cell shape , by which the normal to the cell surface at a specific position on this surface will have significant variation . Since the direction of pseudopodia is given by this normal , it is predicted that a cell with irregular shape should have more variation in pseudopod direction , and consequently shows poor dispersion . The experiments with mutant ddia2-null cells strongly support this interpretation . Wild-type cells have a relatively regular spherical shape by which two nearby pseudopodia are extended in nearly the same direction ( small σφ ) . In contrast , mutant ddia2-null cells have an irregular star-like shape; therefore , two nearby pseudopodia are often extended in very different directions ( large σφ ) . The variance σφ2 can be regarded as the noise of the system . It indicates how fast a cell that extends only alternating splitting pseudopodia ( a = 1 and s = 1 ) will lose correlation of directionality . With σφ = 28 degrees for wild type cells it follows from Eq . 10 that after ten pseudopodia the correlation of direction is still ∼0 . 5 . In contrast , for ddia2-null cells we obtained σφ = 46 . 5 degrees , which implies that already after four pseudopodia the correlation of direction has declined to ∼0 . 5 . Supported by Monte Carlo simulations using the parameters of the mutant , we conclude that poor dispersion of ddia2-null cells is due to the increased variance of pseudopod angles , which is caused by its irregular shape . The correlation factor γ is the product of three terms ( see Fig . 5 and Eq . 9 ) , namely: splitting fraction ( s ) , alternating pseudopod angles ( a and φ ) , and the SD of the pseudopod angle ( σφ ) . Strong persistence of cell movement is attained when all three terms are large and about equal in magnitude . Starved wild type cells follow this strategy: each term is ∼0 . 9 , resulting in the observed correlation factor of 0 . 74 . Mutants in which one of these terms is compromised , such as reduced splitting in sgc/pla2-null cells or enhanced noise of ddia2-null cells , have poor dispersion . In summary , the correlated random walk of amoeboid cells is well described by the balanced bipedal movement , mediated by the alternating right/left extension of splitting pseudopodia . Cells deviate from movement in a straight line due to noise and because cells occasionally hop or make random turns . The turns in particular are used by the cells to modulate the persistence time , thereby shifting between nearly Brownian motion during growth and strong persistent movement during starvation . | Even in the absence of external information , many organisms do not move in purely random directions . Usually , the current direction is correlated with the direction of prior movement . This persistent random walk is the typical way that simple cells or complex organisms move . Cells with poor persistence exhibit Brownian motion with little displacement . In contrast , cells with strong persistence explore much larger areas . We have explored the principle of the persistent random walk by analyzing how Dictyostelium cells extend protrusions called pseudopodia . These cells can extend a new pseudopod in a random direction . However , usually cells use the current pseudopod for alternating right/left splittings , by which they move in a persistent zig-zag trajectory . A stochastic model was designed for the persistent random walk , which is based on the observed angular frequencies of pseudopod extensions . Critical elements for persistent movement are the ratio of de novo and splitting pseudopodia , and , unexpectedly , the shape of the cell . A relatively round cell moves with much more persistence than a cell with an irregular shape . These predictions of the model were confirmed by experiments that record the movement of mutant cells that are specifically defective in pseudopod splitting or have a very irregular shape . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"cell",
"biology/cytoskeleton",
"biophysics/cell",
"signaling",
"and",
"trafficking",
"structures",
"cell",
"biology/cell",
"signaling"
] | 2010 | A Model for a Correlated Random Walk Based on the Ordered Extension of Pseudopodia |
Reinstatement of dynamic memories requires the replay of neural patterns that unfold over time in a similar manner as during perception . However , little is known about the mechanisms that guide such a temporally structured replay in humans , because previous studies used either unsuitable methods or paradigms to address this question . Here , we overcome these limitations by developing a new analysis method to detect the replay of temporal patterns in a paradigm that requires participants to mentally replay short sound or video clips . We show that memory reinstatement is accompanied by a decrease of low-frequency ( 8 Hz ) power , which carries a temporal phase signature of the replayed stimulus . These replay effects were evident in the visual as well as in the auditory domain and were localized to sensory-specific regions . These results suggest low-frequency phase to be a domain-general mechanism that orchestrates dynamic memory replay in humans .
Episodic memories are dynamic , multisensory events that are coded in our memory system . If you remember the last time you had dinner at your favorite restaurant , you will probably recall the person you were with , the music playing in the background , and the smell and taste of that delicious food . Whenever we re-experience episodic memories this way , the events unravel in front of our mind in a temporal order . Even subparts of these episodes , such as the movement of lips in a conversation or parts of the background melody , have an inherent temporal dynamic to them . Given this abundance of temporal structure in our memories , it is rather surprising how limited our understanding is as to how human brains orchestrate such dynamic memory replay . Here , we address this question for the first time , to our knowledge , and identify a neural mechanism that carries the temporal signature of individual dynamic episodic memories . By cuing dynamic memories of auditory and visual content , we were able to detect the presence of phase patterns in the electroencephalographic ( EEG ) signal that indicate the replay of individual auditory or visual stimuli in memory . Temporal signatures were carried by a frequency that was markedly similar in two sensory domains ( ~8 Hz ) , they appeared in sensory-specific regions , and they were related to decreases in power in the same frequency . Previous findings suggest that perception is not continuous but , instead , is rhythmically sampled in discrete snapshots guided by the phase of low alpha ( ~7–8 Hz ) [1–3] , which suggests a pivotal role of low alpha phase for providing a temporal structure during perception [4–8] . Accordingly , recent studies showed that low-frequency phase carries reliable information about stimulus content [9 , 10] . This key role of oscillatory phase during perception makes it a prime candidate to also organize the replay of neural representations in episodic memory , which is an untested prediction to date . A ubiquitous electrophysiological signature of successful memory processing is a pronounced power decrease in low frequencies , especially in alpha [11–15] . On a theoretical level , decreases in alpha power affect neural processing in two ways . First , they promote increased neural activity , as reflected by increased neural firing rates and increased blood-oxygen-level dependent ( BOLD ) signal [16–18] . Importantly , even when alpha power is decreased , its phase still rhythmically modulates firing rates [7] . Second , alpha power decreases reflect a relative de-correlation of neural activity and thereby index an increase in information coding capacity [8] . Accordingly , a mechanism by which decreases in alpha power allow for the temporal organization of information via phase has been proposed in perception [7]; however , whether memory replay is guided by a similar mechanism is an open question [8] . The reinstatement of neural patterns in memory can be detected with multivariate analysis methods such as representational similarity analysis ( RSA ) [19] . This approach has been successfully applied in functional magnetic resonance imaging ( fMRI ) [20 , 21] , EEG/MEG [9 , 22–26] and intracranial EEG ( iEEG ) [27 , 28] . However , even though some previous studies were able to decode information from oscillatory patterns , the mechanism by which oscillations carry mnemonic information remains completely unclear . This is because most prior studies either settle for classification of reactivated memories and thus do not aim for mechanistic explanations of memory replay or because they use static stimuli and analysis procedures . We overcome these central limitations by testing whether a temporal signature that is present while a video or a sound clip is perceived is actively reproduced by the brain during retrieval . By temporal signature we mean a sequence of electrophysiological activity that is specific to an individual stimulus . To this end , we test the mechanistic hypothesis that low-frequency power decreases are linked with the reinstatement of such stimulus-specific phase patterns . A paradigm was used in which memories of dynamic content are cued by a static word ( see Fig 1a–1d ) . In a visual and in an auditory condition , we asked subjects to watch ( or listen to ) 3-second-long video or sound clips and then to associate the respective stimulus with a word . Importantly , only four videos/sounds were repeatedly associated with different words . In the retrieval block , we then only presented the word cue ( or a distractor word ) under the instruction to vividly replay the associated video or sound . Note that there was no overlap in sensory input between the video/sound and the word , enabling us to investigate purely memory-driven reinstatement of temporal signatures . We hypothesize that we will find content-specific temporal signatures in those frequency bands that show pronounced power decreases during episodic memory retrieval [8] . Applying the logic of RSA to measures of phase-based similarity , we designed a new method that can detect content-specific signatures in neural time series in which the exact onset of replay is not known ( see Fig 2 ) , which is the case in our retrieval phase . To our knowledge , this is the first time that a method can use oscillatory phase patterns to decode content from activity that is not time-locked . We assess reinstatement in the auditory and in the visual modality in order to validate our novel , dynamic RSA method and to test for a domain-general memory replay mechanism . Whole-brain activity was measured via high-density EEG , and individual MRIs were collected to increase the fidelity of source localization .
Behavioral results are shown in Fig 1e . In the visual session , participants remembered on average 53 . 92% ( standard deviation [s . d . ] = 17 . 56% ) of the video clips with high confidence ( rating > 4 ) , and they further remembered 9 . 97% ( s . d . = 7 . 62% ) of the clips with low confidence . However , in order to increase the signal-to-noise ratio , hits with a low confidence rating were not included in further analysis . In the auditory session , 44 . 44% ( s . d . = 19 . 8% ) of the audio clips were remembered with high confidence , which was significantly less than in the visual condition ( t23 = -2 . 81 , p = 0 . 01 ) . An additional 9 . 06% ( s . d . = 6 . 9% ) of the audio clips were remembered with low confidence . In accordance , the number of misses tended to be lower in the visual session ( mean 25 . 66% , s . d . = 17 . 56% ) than in the auditory session ( 31 . 46% , s . d . = 19 . 15% , t23 = -1 . 91 , p = 0 . 07 ) . Another trend was observed toward a better identification of distractor words in the visual session ( t23 = 1 . 92 , p = 0 . 07 ) , in which 86 . 88% ( s . d . = 13 . 03% ) of the distractors were correctly rejected , while subjects correctly identified only 82 . 43% ( s . d . = 15 . 86% ) of the distractors as new words in the auditory session . Keeping with the slightly better performance for visual compared to auditory memories , the wrong video clip was less frequently selected in the visual ( 9 . 4% , s . d = 6 . 33% ) condition , compared to the wrong sound clip in the auditory session ( 14% , s . d . = 9 . 69% , t23 = -2 . 86 , p < 0 . 01 ) . In order to identify oscillatory correlates of memory reinstatement , trials in which subjects were presented with a memory cue and strongly reinstated the content ( i . e . , high confidence hits ) were contrasted with trials in which participants were presented with a distractor item and correctly indicated it as a new item ( i . e . , correct rejections ) . As expected , successful memory retrieval was associated with strong power decreases in the low frequencies ( <30 Hz ) ; power increases did not survive statistical testing , including the gamma frequency range ( up to 140 Hz ) . The clusters that survived multiple comparisons correction ( see Materials and Methods ) are shown in Fig 3 . Stronger power decreases for hits were obtained when compared to correct rejections in the visual ( Fig 3a , p < 0 . 001 ) and in the auditory condition ( Fig 3b; p < 0 . 001 ) . The same results emerged when a contrast was built between high confidence hits and trials in which subjects failed to remember the corresponding video or sound clip; that is , when they either failed to retrieve the correct associate or judged an old item as new ( see contrast of hits and misses , S1 Text , S1 Fig ) . This further emphasizes the link of power decreases to successful memory reinstatement . To identify the frequencies that showed the strongest decrease in oscillatory power , the power difference across all electrodes and time points in the retrieval interval was averaged and subjected to a t-test . Power decreases peaked at 8 Hz ( see Fig 3c ) when contrasting hits and correct rejections in the visual condition ( t23 = -5 . 2696 , p < 0 . 001 ) and in the auditory condition ( t23 = -3 . 86 , p < 0 . 001 ) . In the visual condition , these 8-Hz power decreases displayed a broad topography that showed a parietal maximum over the left hemisphere and frontal maxima over both hemispheres ( Fig 3d ) . In the auditory condition , power decreases at 8 Hz were equally broad . Maxima were located over left parietal and right frontal regions ( Fig 3e ) . In order to identify brain regions in which power decreases were maximal at 8 Hz , sources of the difference between hits and correct rejections were reconstructed for that frequency ( see Materials and Methods ) . Statistical testing was run unrestricted on the whole brain level . After multiple comparison correction ( see Materials and Methods ) , a cluster of significant differences emerged in the visual ( p < 0 . 001 ) and in the auditory ( p = 0 . 002 ) condition . Clusters of power decreases were broad and did not show statistical differences between the visual and the auditory condition . In the visual condition ( Fig 3f ) , the cluster of significant differences spanned parietal , temporal , and frontal regions of the left hemisphere and mid-frontal and parietal regions of the right hemisphere . In the auditory condition ( Fig 3g ) , power decreases spanned left parietal , temporal , mid-frontal , and right frontal regions . An important requirement for the detection of replay of temporal patterns during memory retrieval is that the stimulus content itself elicits a distinct time course of activity in the first place , i . e . , while being perceived during encoding . In order to test this prerequisite , a modified version of the pairwise phase consistency ( PPC ) [29] was contrasted between pairs of trials in which the same content was presented and pairs of trials that were of different content . This method assesses the degree of phase similarity that is specifically shared by trials that are instances of the same stimulus ( i . e . , content specificity of phase ) . Content specificity of phase was assessed for every frequency band between 1 and 40 Hz . The time window for statistical testing was chosen between 500 ms pre-stimulus and 3500 ms after stimulus onset , to account for the temporal smearing of the wavelet decomposition . Importantly , the combination of trials was carefully balanced to avoid any possible bias ( see Materials and Methods ) . After correction for multiple comparisons , significant differences were obtained in both conditions in the form of two broad clusters in the visual ( p < 0 . 001 , p = 0 . 003 , Fig 4a ) and one broad cluster in the auditory condition ( p < 0 . 001 , Fig 4b ) . Importantly , the clusters included 8 Hz , which showed the strongest memory effects during replay ( see above ) . We hypothesized that we would later find reappearing temporal patterns in the frequency band of 8 Hz during retrieval; furthermore , content specificity during encoding is a requirement for the detection of these patterns . For these reasons , temporospatial clusters in the data were now identified in which the 8 Hz time course was maximally content specific . Hence , the statistical analysis was now restricted to 8 Hz only . After multiple comparison correction , the cluster at encoding in which content could most reliably be differentiated ( i . e . , the cluster with the lowest p-value ) was selected for further analysis . In the visual condition , this cluster was identified between -152 ms and 564 ms ( p < 0 . 001 ) . Note that post-stimulus effects are temporally smeared into the pre-stimulus interval due to wavelet filtering . One further cluster was observed between 2 , 650 ms and 3 , 300 ms ( p = 0 . 016 ) . In the auditory condition , the most reliable cluster of content specificity was identified in a time window between 22 ms and 871 ms ( p = 0 . 002 ) . Two further clusters were observed ranging from 1 , 818 ms to 2 , 627 ms ( p = 0 . 003 ) and from 1 , 203 ms to 1 , 504 ms ( p = 0 . 047 ) . Therefore , in both domains , early and later time windows distinguished between different stimuli , reflecting the dynamic nature of the stimulus material . A 1 s time window was then defined around the center of the most content-specific cluster . In the visual condition , this center was located at 206 ms; thus , the window ranged from -294 ms to 706 ms ( Fig 4c , right ) . Differences in phase similarity between trials of same and different content showed a clear visual topography within this window , i . e . , the highest t-values were observed over posterior regions of the scalp ( Fig 4c , left ) . In the auditory condition , the 1 s window was centered at 446 ms ( Fig 4d , right ) , ranging from -54 ms to 946 ms . The topography of differences within that window showed a typical auditory distribution ( Fig 4d , left ) , i . e . , high t-values were observed at fronto-central electrodes [30] . Sources of the average difference in phase similarity between same and different content combinations were reconstructed for the 1 s windows to identify the origin of content specificity in that window . t-tests were corrected for multiple comparisons on the whole-brain level . In the visual condition , a cluster of significant difference ( p < 0 . 001 ) emerged in visual regions of the cortex , covering the occipital lobe as well as parts of the parietal lobe ( Fig 4e ) . The cluster exhibited a peak in the right middle occipital gyrus ( MNI: 10; -100; 10; BA: 18 ) . In the auditory condition , differences in similarity ( p = 0 . 004 ) were lateralized to the right hemisphere , which is in line with studies finding lateralization of musical processing to this hemisphere [31] . Differences covered temporal and frontal areas , including primary and secondary auditory processing regions ( Fig 4f ) . The auditory cluster peaked in right sub-lobar insula ( MNI: 40; -20; 0 ) . The crucial quest to identify a replay of temporal patterns from encoding during retrieval is challenged by the non-time-locked nature of retrieval . Indeed , replay of memory content during retrieval could happen at any point after presentation of the retrieval cue , with the exact onset varying from trial to trial . Moreover , we assumed that any temporal pattern from encoding could be replayed at any time during retrieval . We therefore developed a procedure that is not affected by these time shifts; specifically , we assessed the similarity between encoding and retrieval with a sliding window approach . To this end , phase similarity between combinations of encoding and retrieval time windows was computed using a variation of the single-trial phase locking value ( S-PLV ) [32 , 33] , namely the similarity of phase angle differences over time ( see Materials and Methods ) . This method is less susceptible to noise and allows for an estimation of similarity between two time windows in non-time-locked data . Again , phase-similarity of encoding-retrieval pairs that were of the same content ( e . g . , perceiving A , remembering A ) was contrasted with the similarity of pairs that were of different content ( e . g . , perceiving B , remembering D ) . The time window that contained the temporal pattern from encoding was selected based on the highest content specificity of phase during encoding ( see above ) . The width of the window amounted to 1 s ( 8 cycles ) around the center of the cluster that was located at 206 ms ( -294 to 706 ms , see Fig 4c , right ) in the visual condition and at 446 ms in the auditory condition ( -54 to 946 ms , see Fig 4d , right ) . Since activity at encoding , under the null-hypothesis , would be independent from activity at retrieval , the specific selection of a time window , based on results from encoding , can be used to increase the signal to noise ratio , without risking circular inference . Phase-similarity to this predefined encoding window was now assessed by sliding the window over the whole retrieval episode . To slide the window into retrieval , the pre-stimulus interval between -500ms and 0ms was used as padding . To slide it out at the end of the trial , the pre-stimulus interval between -1 , 000 ms and -500 ms was used as padding ( this was done because later time points were unsuitable for padding due to contamination with similar perception and responses ) . Note that the similarity at time point 0 is then assessed by comparing the encoding window to the retrieval window between -500 ms and 500 ms , and similarity at 4 s is assessed by comparing the encoding window to the concatenated retrieval window of 3 , 500 ms to 4 , 000 ms and -1 , 000 ms to -500 ms . The phase-similarity of the encoding window to episodes of replay of the same video/sound was now contrasted with the phase-similarity to episodes of replay where a different content was replayed from memory . The t-statistic was computed for every electrode on the averaged difference between same versus different combinations between 0 and 4 s . A cluster-based permutation test indicated replay of encoding phase patterns during retrieval for both the visual ( p = 0 . 002 ) and the auditory condition ( p = 0 . 01 ) . In the visual replay condition , the cluster of significant differences emerged over left parietal regions ( Fig 5a , right ) . In the auditory replay condition , a cluster of significant differences was observed over right posterior temporal areas ( Fig 5b , left ) . This signifies strong evidence for mnemonic replay of temporal patterns . To test for frequency specificity , the same analysis was performed for 5 Hz and 13 Hz , which are approximately in a golden ratio relationship to 8 Hz ( i . e . , maximally different in phase ) [34] . Two further control frequencies were tested that showed peaks in power decreases in at least one of the conditions , namely 4 Hz and 15 Hz . To this end , time windows from encoding were selected with the same criteria as for 8 Hz; electrodes for testing were again restricted to the electrodes in the significant cluster from encoding . Furthermore , the time window was likewise built from 8 cycles of the corresponding frequency . However , no effects were found in the visual or in the auditory condition for any of the control frequencies , suggesting that temporal reinstatement of phase patterns was specific to 8 Hz . The temporal profile of the replay effect was then inspected by averaging phase similarity across electrodes within the cluster of significant differences . A t-test was computed at every time point , applying a probability of error below 0 . 01 . For visual material , four episodes of replay could be identified ( Fig 5a , left ) , in which a one-sided test exceeded the critical threshold ( t23 = 2 . 5 ) . These episodes peaked at 203 ms ( t23 = 3 . 09 , p = 0 . 003 ) , 547 ms ( t23 = 2 . 51 , p = 0 . 01 ) , 828 ms ( t23 = 2 . 75 , p = 0 . 006 ) , and 1 , 844 ms ( t23 = 3 . 65 , p < 0 . 001 ) . For the auditory material , three episodes exceeded the critical t-value ( Fig 5b , right ) , peaking at 1 , 406 ms ( t23 = 2 . 64 , p = 0 . 007 ) , 3 , 125 ms ( t23 = 2 . 7 , p = 0 . 006 ) , and 4 , 016 ms ( t23 = 3 . 47 , p = 0 . 001 ) . To reveal whether the encoding-retrieval similarity effects were maximal in material-specific ( i . e . , visual/auditory ) brain regions , encoding-retrieval similarity was assessed on the source level . Statistical testing was run unrestricted on the whole brain level , and for each condition the maximal cluster ( i . e . , with the highest summed t-values ) was plotted . For the visual material , the strongest cluster of encoding-retrieval similarity showed a peak in the superior parietal lobule ( MNI: 20; -50; 60 , BA: 7 , see Fig 5c ) , overlapping with the similarity effects during encoding ( compare to Fig 4e ) and in line with studies finding parietal lobe contributions to episodic memory retrieval [35 , 36] . For the auditory material , similarity effects showed a peak in the right inferior temporal gyrus ( MNI: 50; -10; -44 , BA: 20 , see Fig 5d ) also overlapping with the similarity effects during encoding ( compare Fig 4f ) and in line with previously reported effects on memory for music [37] . Since power decreases at 8 Hz spanned multiple brain regions in the visual and in the auditory condition ( compare Fig 3f and 3g ) , content-specific decreases were still statistically unsubstantiated . In order to link phase-based similarity at 8 Hz with the power decreases during memory replay , we therefore compared the power difference at 8 Hz between hits and correct rejections ( see above ) within the regions of visual and auditory “replay . ” We computed a 2x2 ANOVA contrasting 8 Hz power decreases on the source level , with the factors region ( visual/auditory ) and condition ( visual/auditory ) . If power decreases are relevant for information coding , stronger power decreases should be observed in those sensory regions where replay occurred . This hypothesis was confirmed by a significant interaction ( F1 , 23 = 6 . 58 , p = 0 . 017 , see Fig 6 ) , showing that power decreases in the auditory region of interest were stronger during replay of auditory memories , whereas power decreases in the visual region of interest were stronger during visual memory replay . To obtain a further understanding of the temporal dynamics of memory reinstatement , a follow-up analysis within the electrode clusters of significant differences was run for all combinations of retrieval and encoding time windows , resulting in retrieval time–encoding time diagrams ( see Fig 5e and 5f ) . It should be acknowledged that further analyses on this cluster will be biased toward being optimal for the time window on which the electrodes were originally identified . Therefore , the results are likely to show more reinstatement of phase patterns from early encoding . Primarily , this analysis reveals which parts from the original sliding window ( centered on the most content-specific cluster from encoding ) maximally contributed to the effect when we tested for content specificity of reactivation ( e . g . , mostly activity from the beginning of the window ) . Moreover , on a descriptive level , this analysis gives an idea about which phase patterns from encoding , in addition to the early ones , were also reactivated during retrieval . It is worthwhile to keep two issues in mind when interpreting these plots . First , similarity between two windows will always express temporal smoothing on the diagonal . The diagonal width can be seen as an indicator of the length of the episode that was replayed , but it is also affected by the length of the sliding window ( i . e . , longer windows will induce more smearing along the diagonal ) . Second , the peak in these diagonals indicates which temporal pattern at encoding was actually replayed at which retrieval time point . When two time windows are aligned and they share a temporal pattern in their first quarter , this pattern would appear temporally delayed in a one-dimensional plot; however , in two dimensions , we can inspect the diagonal peak of similarity . In the visual condition , a permutation test revealed significant differences in five clusters . The peaks of the clusters suggested that early , around 141 ms , during retrieval , activity from 672 ms during encoding was reinstated ( p = 0 . 017 ) . At 266 ms of the retrieval interval , encoding patterns from around 359 ms reappeared ( p < 0 . 002 ) ; around 719 ms , phase patterns from 31 ms during encoding were reinstated ( p = 0 . 004 ) . Later , during retrieval at 1 , 859 ms , the phase patterns from 672 ms during encoding were detected ( p = 0 . 012 ) , and at 1859 ms , the activity from 1 , 172 ms during encoding showed a similarity effect ( p = 0 . 022 ) . In the auditory condition , only three clusters could be identified . Peaks within the clusters suggested that 1 , 203 ms after the onset of the retrieval cue , content from 15 ms during encoding was replayed ( p = 0 . 01 ) . Later , at 3 , 781 ms , activity from 78 ms at encoding reappeared ( p = 0 . 017 ) ; and finally , at 3 , 797 ms into the retrieval time , late encoding phase patterns from 1 , 765 ms could be detected ( p = 0 . 014 ) . Even though results are biased toward detecting replay from the early encoding window that served to identify the electrodes on which memory replay took place , this analysis could still give an idea of the temporal dynamics of reinstatement and show the potential of our method . Reactivation of visual patterns was observed very early , as was expected given recent evidence for early reactivation [23 , 38] , and notably earlier than reactivation of auditory patterns , which is in line with a worse memory performance of participants in the auditory condition of this study .
In real life , most of our episodic memories are dynamic with an inherent temporal structure and are not bound to a single modality . We can re-experience information-rich memory traces with auditory and visual content and habitually reinstate these events with an abundance of subjective impressions in their correct temporal order . Although some of these temporal aspects of memory replay have been investigated in spatial navigation experiments in rodents [39–41] , the temporal properties of episodic memory replay in humans were largely ignored in previous research . Consequently , little is known about the neural mechanisms that orchestrate the replay of dynamic memories in humans . In the present study , we identified content-specific temporal signatures of individual memories in the visual and in the auditory domain . These signatures were specific to a carrier frequency of ~8 Hz and could be localized to modality-specific regions , i . e . , overlapping with those regions that carried the information of the stimuli during encoding . Strikingly , the 8 Hz frequency also showed the strongest power decrease during retrieval in both modalities . Likewise , the power decrease in 8 Hz during retrieval was modulated in a sensory-specific manner in those regions where memory replay took place , i . e . , stronger power decreases in the parietal ( visual ) region during replay of videos and vice versa for replay of sounds in the temporal ( auditory ) region ( see Fig 6 ) . These findings provide a link to other studies in which a similar interaction between alpha power decreases and oscillatory phase has been proposed to temporally structure perceptual contents [7] . In line with these findings , our results suggest that similar oscillatory mechanisms that guide perception also guide the “re-perception”—that is , memory replay—of these sensory events . In order to detect the reinstatement of temporal neural patterns that indicate such replay of individual memories , we developed a novel dynamic phase-based RSA method that is robust against variations in the onset of memory replay . This method can therefore be applied in conditions when the exact time point of the reinstatement of a neural pattern is unknown , like , for example , during offline replay in resting state or sleep . RSA has been previously used to track episodic memories in EEG/MEG [9 , 22–26] and iEEG [27 , 28]; however , some important differences to these studies have to be considered . First , we go beyond mere classification of memory content , since we use similarity measures to test a mechanistic hypothesis: that alpha power decreases are associated with the reinstatement of temporal patterns . Hence , we can test whether temporal patterns reappear during retrieval , and we can link this replay to a specific frequency band . Importantly , the detection of temporal patterns was only made possible with our dynamic RSA approach . Second , in our design , we carefully avoided any sensory overlap between encoding and retrieval . We were therefore able to investigate mechanisms of purely memory-driven reinstatement , as opposed to studies in which there was a high overlap in sensory stimulation between encoding and retrieval [25 , 28 , 42] . This aspect of our experimental design allows us to conclude that the brain actively reproduces a temporal pattern that is specific to a stimulus in order to re-experience this particular memory . An important open question concerns how the hippocampus is involved in the replay of temporal patterns in the cortex , as observed here . A critical involvement of the hippocampus , and the phase of theta oscillations therein , for memory replay is implicated by recent models and frameworks [8 , 43 , 44] . Future studies are required that record simultaneously from both the hippocampus and the neocortex to investigate how the reinstatement of the temporal phase patterns described here interacts with , or relies on , the hippocampus . Studying the temporal aspects of memory replay has proven to be difficult , because methods or stimulus material in previous studies did not allow for investigating this question . Overcoming these previous limitations , we identified a potential domain general mechanism that orchestrates the replay of dynamic auditory and visual memories in humans . Specifically , our findings suggest an intimate relationship between power decreases in an 8 Hz frequency and a content-specific temporal code , carried by its phase . These results corroborate recent theories linking power decreases with the coding of neural information [8 , 45 , 46] . Our findings open up new ways of investigating the temporal properties of memory replay in humans , which we only begin to understand .
Twenty-four healthy , right-handed subjects ( 18 female and 6 male ) volunteered to participate . Seven further participants were tested , or partly tested , but could not be analyzed due to poor memory performance ( n = 2 ) , misunderstanding of instructions ( n = 2 ) , and poor quality of EEG-recording and technical failure ( n = 3 ) . All participants had normal or corrected-to-normal vision . The average age of the sample was 23 . 38 ( s . d . = 3 . 08 ) years . Participants were native English speakers ( 20 ) , bilingual speakers ( 2 ) , or had lived for more than 8 y in the United Kingdom ( 2 ) . Participants provided informed consent and were given a financial compensation of £24 or course credit for participating in the study . The cues amounted to 360 words that were downloaded from the MRC Psycholinguistic Database [47] . Stimulus material consisted of four video clips and four sound clips in the visual and auditory session , respectively . All clips were 3 s long; videos showed colored neutral sceneries with an inherent temporal dynamic , and sounds were short musical samples , each played by a distinct instrument . In both sessions , a clip was associated with 30 different words . Sixty words were reserved for the distractor trials , and 12 additional words were used for instruction and practice of the task . For presentation , words were assigned to the clips or to distractors in a pseudorandom procedure , such that they were balanced for Kucera-Francis written frequency ( mean = 23 . 41 , s . d . = 11 . 21 ) , concreteness ( mean = 571 , s . d . = 36 ) , imageability ( mean = 563 . 7 s . d . = 43 . 86 ) , number of syllables ( mean = 1 . 55 , s . d . = 0 . 61 ) , and number of letters ( mean = 5 . 39 , s . d . = 1 . 24 ) . Furthermore , lists were balanced for word frequencies taken from SUBTLEXus [48] . Specifically , “Subtlwf” was employed ( mean = 20 . 67 , s . d . = 27 . 16 ) . The order of presentation was also randomized , assuring that neither the clips and their associates nor distractor words were presented more than three times in a row or in temporal clusters . The presentation of visual content was realized on a 15 . 6-in CRT monitor ( Taxan ergovision 735 TC0 99 ) at a distance of approximately 50 cm from the subject’s eyes . The monitor refreshed at a rate of 75 Hz . On a screen size of 1 , 280 x 1 , 024 pixels , the video clips appeared in the dimension of 360 pixels in width and 288 pixels in height . “Arial” was chosen as the general text font , but font size was larger during presentation of word-cues ( 48 ) than during instructions ( 26 ) . In order to reduce the contrast , white text ( rgb: 255 , 255 , 255 ) was presented against a grey background ( rgb: 128 , 128 , 128 ) . Auditory stimuli were presented using a speaker system ( SONY SRS-SP1000 ) . The two speakers were positioned at a distance of approximately 1 . 5 m in front of the subject , with 60 cm of distance between the speakers . Upon informed consent and after being set up with the EEG-system , participants were presented with the instructions on the screen . Half of the subjects started with the auditory session; the others were assigned to undertake the visual task first . Both sessions consisted of a learning block , a distractor block , and a test block . The sessions were identical in terms of instructions and timing and differed only in the stimulus material that was used . During instruction , the stimulus material was first presented for familiarization and then used in combination with the example words to practice the task . Instructions and practice rounds were completed in both sessions . As a way to enhance memory performance , participants were encouraged to use memory strategies . The suggestion was to imagine the word in a vivid interaction with the material content , yet the choice of strategy remained with the subject . In the learning block , 120 clip-word sequences were presented . Each sequence started with a fixation cross that was presented in the center of the screen for 1 s , then the video clip played for 3 s . In the auditory condition , the fixation cross stayed on the screen and the sound clip played for 3 s . Immediately after the clip , a word cue was presented in the center for 4 s , giving the subject time to learn the association . After that , an instruction asked the subject to subjectively rate on a six-point scale how easy the association between the clip and the word was . After a press on the space bar , this scale was shown . Equidistant categories were anchored with the labels “very easy” and “very hard;” those labels were displayed at both ends above the scale . Participants used six response buttons to rate the current association ( see Fig 1 ) . In the distractor block , subjects engaged in a short , unrelated working memory task; namely , they counted down in steps of 13 , beginning from 408 or 402 , respectively . After 1 min , the distractor task ended . Following a short self-paced break , subjects refreshed the instructions on the retrieval block . In this block , either a cue or a distractor was presented upon a button press on the space bar . Subjects were instructed to try to vividly replay the content of the corresponding video clip or sound -clip in their mind upon presentation of the cue . The word stayed on the screen for 4 s , giving the subject the opportunity to replay the memory . Finally , a fixation cross was presented for a varying time window between 250 and 750 ms , to account for movement and preparatory artifacts , before the response scale appeared on the screen . The response scale consisted of six options . Four small screenshots of the videos or four black-and-white pictures of the featured instruments were presented in equidistant small squares of 30 x 30 pixels . Additionally , the options “new” and “old” were displayed in the form of text at the most left and most right position of the scale ( see Fig 1c and 1d ) . Subjects could now indicate the target ( video/sound ) they just replayed by pressing the button corresponding to that clip . In addition , subjects could also indicate that the word was a distractor by pressing the button corresponding to the option “new , ” or they would simply indicate that they remembered the word , but could not remember the clip it was associated with . In this last scenario , subjects would press the button corresponding to “old . ” The positions of “old” and “new” at the end of the scale , as well as the permutation of the four target positions in the middle of the scale , were counterbalanced across participants . Finally , after making a decision , a further six-point rating scale was presented on which subjects could rate their confidence in their response . Again , a scale with equidistant categories was presented , ranging from “guess” to “very sure . ” An additional possibility was to press “F2” in case of an accidental wrong button press . In this case , the whole trial was discarded from analysis . Following the retrieval block , individual electrode positions were logged , allowing for a break of approximately 30 min before beginning the second session . In addition to the two experimental sessions , all participants came to a separate session to record anatomical MRI scans at the Birmingham University Imaging Centre ( https://www . buic . bham . ac . uk/ ) . This was later used to facilitate source localization ( see below ) . The recording of behavioral responses and the presentation of instructions and stimuli were realized using Psychophysics Toolbox Version 3 [49] with MATLAB 2014b ( MathWorks ) running under Windows 7 , 64 Bit version on a desktop computer . Response buttons were “s , d , f , j , k , l” on a standard “QWERTY” layout . Buttons were highlighted and corresponded spatially to the response options on the screen , so participants didn’t have to memorize the keys . To this end , the shape of corresponding fingers was also displayed under the scale . To proceed , participants used the space bar during the experiment . Physiological responses were measured with 128 sintered Ag/AgCl active electrodes , using a BioSemi Active-Two amplifier . The signal was recorded at a 1 , 024 Hz sampling rate on a second computer via ActiView recording software , provided by the manufacturer ( BioSemi , Amsterdam , the Netherlands ) . Anatomical data was acquired using magnetic resonance imaging ( MRI ) ( 3T Achieva scanner; Philips , Eindhoven , the Netherlands ) , and electrode positions were logged with a Polhemus FASTRAK device ( Colchester , Vermont , USA ) in combination with Brainstorm [50] implemented in MATLAB . The data was preprocessed using the Fieldtrip toolbox for EEG/MEG-analysis [51] . Data was cut into trial-segments from 2 s pre-stimulus to 4 . 5 s after stimulus onset ( i . e . , onset of the clip at encoding and onset of the word at retrieval ) . The linear trend was removed from each trial , and a baseline correction was applied based on the whole trial . Trials were then downsampled to 512 Hz , and a band-stop filter was applied at 48–52 , 58–62 , 98–102 , and 118–122 Hz to reduce line noise at 50 Hz and noise at 60 Hz; additionally , a low-pass filter at 140 Hz was applied . After visual inspection for coarse artifacts , an independent component analysis was computed . Eye-blink artifacts and eventual heartbeat/pulse artifacts were removed , bad channels were interpolated , and the data was referenced to average . Finally , the data was inspected visually and trials that still contained artifacts were removed manually . MRI scans of each participant were segmented into four layers ( brain , cerebrospinal fluid , skull , and scalp ) using SPM8 ( http://www . fil . ion . ucl . ac . uk/spm/ ) in combination with the Huang toolbox [52] . On this basis , a volume conduction model was created with the Fieldtrip “dipoli” method; individual electrode positions were aligned to the head model for every participant . For behavioral analysis , correct trials were defined as those of the retrieval phase in which the target was correctly identified and the confidence rating of the response was high ( 5 or 6 ) . Trials were defined as correct rejections if a distractor word was correctly identified as new; misses were defined as trials in which a cue word was incorrectly identified as a new word or the response “old” was given to indicate that the subject recognized the word , but could not remember the target video or sound it was associated with . Hits of low confidence were not considered in subsequent analyses . Furthermore , selections of the wrong clip as well as accidental presses of the wrong button and distractor trials that were not recognized as distractors were discarded from further analysis . Power at retrieval was determined by multiplying the Fourier-transformed data with a complex Morlet wavelet of six cycles . Raw power was defined as the squared amplitude of the complex Fourier spectrum and estimated for every fourth sampling point ( i . e . , sampling rate of 128 Hz ) . For each contrast ( i . e . , hits versus misses , or hits versus correct rejections ) , baseline normalization was performed separately . Therefore , a baseline was computed as the average power between -1 and 4 s of all trials within the contrast [15] . Every trial was then normalized by subtracting the baseline and subsequently dividing by the baseline ( activitytf−baselinef ) /baselinef , where t indexes time and f indexes frequency . The relative power was calculated for all frequencies between 1 and 40 Hz . For every frequency between 2 and 40 Hz , the stationarity of phase was defined within a sliding window of one cycle ( see S2 Fig , S2 Text , S3 Fig , S3 Text , S4 Fig ) . Phase was estimated by multiplying the Fourier-transformed data with a complex Morlet wavelet of six cycles . The complex signal was then divided by its amplitude to standardize its power to 1 . At every time point , the deviation from an even circular distribution within one cycle around this point was assessed , i . e . , the circular variance ( CV ) of phase over time was computed . CV was interpreted as a measure of signal stationarity , since a perfectly stationary signal has an even distribution over one cycle and the circular variance within the cycle is maximal ( i . e . , reaches 1; see Fig 7 ) . Phase stationarity was baseline corrected in the same way as oscillatory power . A baseline was computed as the average stationarity between -1 and 4 s of all trials within the contrast . Every trial was then normalized again by subtracting the baseline and then dividing by the baseline ( stationaritytf−baselinef ) /baselinef , where t indexes time and f indexes frequency . While participants learned the associations in the encoding block , they repeatedly saw ( heard ) the same dynamic stimulus . Content-specific properties could consequently be identified if they were shared by trials of the same content but not by trials of a different content . Hence , content-specific phase was assessed by contrasting the phase similarity between pairs of trials in which the same content was presented , with the phase similarity of an equal number of trial pairs that were of different content . To achieve this , trials were grouped and combined in a random but balanced way ( see below ) . For each pair of trials , the cosine of the absolute angular distance was then computed and finally averaged across all ( same or different ) combinations [29] . This resulted in an average similarity value at every time point , at every electrode and in every frequency of interest . This similarity was derived separately for the same pairs and for the different pairs and could consequently be subjected to statistical testing in order to define content specificity of phase . Importantly , the way of combining the trials can result in bias . For this reason , the trial combinations were randomly selected in a carefully balanced way ( Fig 8 ) . Firstly , the trials were grouped into four sets that were of the same content ( SE1-4 ) , e . g . , the same video . These sets were then recombined such that each set of content , say , A , could be paired with a unique set of mixed content ( say , B , C , and D ) that was equal in size , i . e . , a contrast-set ( CE1-4 ) . To make this possible , some trials were discarded from further analysis ( Fig 8a and 8b ) . In order to form pairs of same content , all possible N* ( N-1 ) /2 pairs within each of the four stimulus-sets ( SE1-4 ) were built . Then , to form pairs of different content , only N* ( N-1 ) /2 pairs between the stimulus-set ( SEI ) and its contrast-set ( CEI ) were built . Importantly , wherever the second trial in the pairs of same content appeared in several combinations , it was replaced by instances of the same exclusive trial from the contrast set , while building the combinations of different content ( Fig 8c ) . Participants not only saw ( heard ) the same dynamic stimulus several times in the encoding block; they also repeatedly recalled the same memory content . This made it possible to detect content-specific properties of memories if they were shared by trials in which the same content was learned and remembered ( e . g . , encoding A , remembering A ) but not by trials in which different content was learned and remembered ( e . g . , encoding B , remembering C ) . Content-specific phase was consequently assessed by contrasting the phase similarity between encoding-retrieval pairs of same content , with the phase similarity of encoding-retrieval pairs that were of different content . Again , trials were grouped and paired in a balanced randomization procedure to avoid potential bias . First , the trials at encoding were grouped into four sets that were of same content ( SE1-4 ) . Likewise , the trials at retrieval were grouped into four sets of same memory content ( SR1-4 ) . These sets at retrieval were then recombined , such that each set of content A could be assigned a unique set of mixed content ( B , C , and D ) that was equal in size , i . e . , a contrast set ( CR1-4 ) . To make this possible , some trials were discarded from further analysis ( Fig 9a and 9b ) . Encoding-retrieval pairs of same content were then formed by building all possible pairs of trials between each set of a content at encoding ( SEI ) and the corresponding set of this memory content at retrieval ( SRI ) . In order to build the pairs of different content , the very same set of trials from encoding ( SEI ) was combined with the corresponding contrast set ( CRI ) at retrieval . Finally , pairs containing the same word cue were ignored . This occurs when the encoding trial that was originally associated with a word cue was combined with the retrieval trial in which this cue was actually presented . Accordingly , in the combinations of different content , the pair between the discarded encoding trial and a random trial was ignored ( Fig 9c ) . Between the pairs of same combinations , a similarity measure of phase was then computed ( see below ) and contrasted with the similarity between the pairs of different content . In order to maximize the signal-to-noise ratio in further analysis , several restrictions were applied to define frequencies , time windows , and electrodes of interest . The tested frequency was 8 Hz , because both conditions expressed the strongest correlates of memory in this frequency band . Furthermore , the time window at encoding was restricted to a 1-second episode in which phase patterns were maximally different between the stimuli . The window was defined around the center of the cluster in which phase patterns were most reliably content-specific during encoding ( i . e . , the cluster with the lowest p-value ) . Centering the encoding window on the most content specific time course of activity should increase the sensitivity to detect differences from encoding at retrieval . Likewise , the electrodes for further analysis were restricted to the electrodes within that cluster ( 128/128 electrodes in the visual condition and 107/128 electrodes in the auditory condition ) . It needs to be emphasized that none of these restrictions leads to circular inference , because all of these prior restrictions are independent of the similarity between encoding and retrieval trials . Most importantly , phase similarity at encoding , under the null hypothesis , is completely orthogonal to any neural activity at retrieval . Phase similarity between two windows was then assessed with the Single-trial Phase Locking Value ( S-PLV ) [32 , 33] . This measure defines similarity between two windows ( x and y ) as the constancy of phase angle difference over time , where n denotes the width of the window and φ is the phase: SPLV = n-1∑t = 1nei ( φxt-φyt ) If the two signals are very similar over time , the phase angle differences will not vary much ( i . e . , have low circular variance ) . In this way , the similarity of two windows can be quantified as 1 minus the circular variance of phase differences over time . S-PLV has the advantage of increased robustness for noisy data at the expense of temporal resolution . For the purposes of assessing similarity between two oscillatory patterns , this measure is convenient because it affords a high degree of temporal invariance and results in a value between 0 and 1 when two oscillatory patterns are compared . Therefore , despite the oscillatory nature of temporal patterns in the EEG , this makes it possible to assess the average similarity across time , trials , and subjects . In their paper , the authors suggest computing the S-PLV over 6–10 cycles of a frequency for a good signal-to-noise ratio [32]; for our purposes , S-PLV was applied to a time window of 8 cycles , which resulted in a 1-second window for 8 Hz . Phase values were extracted by multiplying the Fourier-transformed data with a complex Morlet wavelet of six cycles . Phase values were then downsampled to 64 Hz . The similarity measure was computed for every pair of trials in the combinations of same content and in the combinations of different content . Importantly , a sliding window approach was used to account for the non-time-locked nature of the data ( memory reactivation could happen at any time during retrieval ) . For every combination of trials , this resulted in a single similarity value for every electrode and every time point at retrieval , i . e . , the similarity to the 1-second encoding window ( a similarity value at a single time point represented the similarity of the surrounding 1-second window at retrieval , to that window from encoding ) . Additionally , the retrieval window was truncated at 4 s in order to avoid potential confounds from post-stimulus images or responses; to assess similarity at 4 s , the time window was instead continued beginning from 1 s pre-stimulus ( i . e . , similarity at 4 s reflects the similarity between the encoding window and the concatenated window from 3 , 500 ms to 4 , 000 ms and -1 , 000 ms to -500 ms at retrieval ) . In order to test for content-specific phase patterns , the difference in similarity between same-content combinations and different-content combinations was averaged across the whole retrieval episode ( between 0 and 4 , 000 ms ) , which resulted in a single value for every electrode for the same content combinations and for the different content combinations . Those values were then statistically tested across subjects , controlling for multiple comparisons with the fieldtrip permutation procedure [53] . Additionally , two control frequencies were tested that were approximately in the golden mean ratio ( i . e . , maximally different in terms of phase ) to 8 Hz [34] , namely 5 and 13 Hz . Two further control frequencies were tested that showed the next strongest power decrease in one of the conditions , namely 4 and 15 Hz . Encoding time windows were defined accordingly for these frequencies as eight cycles around the center of the most reliable cluster during encoding . The electrode clusters of significant differences that resulted for 8 Hz were subjected to further analysis in order to explore the temporal dynamics of reinstatement . In a first step , a series of post-hoc t-tests was computed on the difference between same and different content combinations during every time point of retrieval . This resulted in a time series that is comparable to a cross-correlogram and can be interpreted as a time course of reinstatement ( see Fig 5a and 5b ) . In a further step , the sliding window analysis was repeated with different time windows from encoding , but keeping with the electrodes in the cluster of significant differences . Thereby , similarity between any two time points could be estimated with a temporal uncertainty of +/-500 ms . The outcome of this analysis was a matrix of similarity between every time point at encoding and every time point at retrieval on each of the electrodes in the cluster ( see Fig 5e and 5f ) . The difference between combinations of same and different content was then averaged across electrodes and tested over subjects . The resulting clusters reveal the temporal relationship between presentation at encoding and reinstatement during retrieval; however , it should be said that tests on this encoding-retrieval matrix are not independent from the original identification of the electrodes . To reconstruct the activity on the source level , a linearly constrained minimum variance ( lcmv ) beamforming approach was used , as it is implemented in the Fieldtrip toolbox ( http://www . fieldtriptoolbox . org/ ) . Individual electrode positions were used together with boundary element models that were constructed from individual MRI scans . With lcmv-beamforming , filters will be more accurate for the data that they were constructed on and will also be more accurate if constructed on a long time interval [54] . This trade-off was addressed by computing each filter around the preprocessed data that contributed to the effect being localized . Power differences were localized with a filter based on -500 ms to 4 , 500 ms at retrieval; for the phase similarity at encoding , the filters were estimated on the time window between -500 ms and 3 , 500 ms of the encoding trials . Phase similarity between encoding and retrieval was reconstructed with a filter based on -500 ms to 1 , 000 ms at encoding and -500 ms and 4 , 500 ms at retrieval . Activity on 2 , 020 virtual electrodes was thereby reconstructed and the analysis of the data was repeated in the same way on the virtual data .
Group statistical data and analysis scripts of this project are deposited in the Dryad repository: http://dx . doi:10 . 5061/dryad . ch110 [56] . | A remarkable ability of the human brain is that it can mentally replay past episodes . For instance , if one remembers the last movie one has seen , one can vividly evoke parts of this event in a temporally highly structured manner . This implicates a neural mechanism that temporally guides the brain through memory retrieval as a sensory trace unfolds over time . However , those mechanisms are weakly understood , because little is known about the temporal dynamics of memory replay . Using a new analysis method , we detect the replay of temporal patterns in a memory task that requires participants to mentally replay short sound or video clips . We show that the phase of a low-frequency oscillation carries a temporal signature that is unique to specific stimulus content ( either visual or auditory ) and signifies its replay in memory . Importantly , the replay of temporal signatures is localized in processing regions specific for each sensory modality and is related to decreases in low-frequency power in the same regions . Our study provides first insight that our brain codes information of dynamic stimuli in the phase of low-frequency neural time series and that these time series can be replayed purely from episodic memory . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Data",
"Availability"
] | [
"learning",
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"engineering",
"and",
"technology",
"signal",
"processing",
"brain",
"electrophysiology",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"learning",
"and",
"memory",
"magnetic",
"r... | 2016 | The Temporal Signature of Memories: Identification of a General Mechanism for Dynamic Memory Replay in Humans |
Trafficking of human papillomaviruses to the Golgi apparatus during virus entry requires retromer , an endosomal coat protein complex that mediates the vesicular transport of cellular transmembrane proteins from the endosome to the Golgi apparatus or the plasma membrane . Here we show that the HPV16 L2 minor capsid protein is a retromer cargo , even though L2 is not a transmembrane protein . We show that direct binding of retromer to a conserved sequence in the carboxy-terminus of L2 is required for exit of L2 from the early endosome and delivery to the trans-Golgi network during virus entry . This binding site is different from known retromer binding motifs and can be replaced by a sorting signal from a cellular retromer cargo . Thus , HPV16 is an unconventional particulate retromer cargo , and retromer binding initiates retrograde transport of viral components from the endosome to the trans-Golgi network during virus entry . We propose that the carboxy-terminal segment of L2 protein protrudes through the endosomal membrane and is accessed by retromer in the cytoplasm .
The human papillomaviruses are small non-enveloped viruses responsible for approximately 5% of human cancer deaths worldwide [1] . The cellular mechanisms involved in HPV infection are poorly understood , but they represent potential sites of anti-viral intervention . HPV consists of an ~8000 base-pair double-stranded DNA viral genome packaged in the viral capsid , which is composed of 360 molecules of the major capsid protein , L1 , and up to 72 molecules of the minor capsid protein , L2 , which is largely buried inside the L1 shell [2–4] . Initial binding of virus to cells is mediated by the interaction between the L1 capsid protein and heparan sulfate proteoglycans [5–8] . After cell binding , L1 and L2 undergo conformational changes , which allow cleavage of the amino-terminus of L2 at the cell surface by the protease furin [9–12] . HPV is then transferred to an as-yet-unidentified cell-surface receptor and internalized [13–17] . Disassembly of the capsid is initiated by acidification of the endosomal lumen by the vacuolar ATPase , and L1 , L2 , and viral DNA then traffic via retrograde pathways to the Golgi apparatus and endoplasmic reticulum [5 , 16–24] . Cell cycle progression and nuclear envelope breakdown appear required for HPV entry into the nucleus , where viral gene expression and DNA replication occur [25 , 26] . During virus trafficking , the L1 protein dissociates from the viral DNA , but the 473-amino acid L2 protein is required for efficient trafficking of the viral genome to the nucleus and remains associated with the genome during nuclear entry [20 , 27–34] . We performed a genome-wide siRNA screen to identify host cell genes required for HPV16 infection and discovered that infection of HeLa cervical cancer cells and immortalized cervical keratinocytes requires retromer , a cytoplasmic endosomal coat complex that mediates export of cellular transmembrane proteins from the endosome to the trans-Golgi network ( TGN ) or plasma membrane [22 , 35 , 36] . The cargo recognition core of the retromer consists of three subunits , Vps26 , Vps29 , and Vps35 , all of which are required for retromer-mediated endosomal sorting and for HPV infection [22] . In cells depleted of retromer , HPV components fail to arrive at the TGN [22] . In addition , retromer is present in a stable complex with viral capsid proteins in infected cells [22] . The mechanism by which retromer supports HPV trafficking is unknown . All known retromer cargos are cellular , integral membrane proteins [35 , 36] . Retromer recognizes sorting signals located in the cytoplasmic domain of these transmembrane protein cargos at the endosomal membrane to effect packaging of the cargo into budding vesicles or tubules that later fuse with target membranes to deliver the cargo to its destination . Unlike known retromer cargos , the non-enveloped HPV capsid is particulate and lacks transmembrane proteins . Furthermore , when the capsid is in the endosomal lumen during the early stages of infection , it is separated from retromer in the cytoplasm by the endosomal membrane . It is possible that retromer acts indirectly by mediating trafficking of a cellular transmembrane protein that is essential for some step in HPV entry . Alternatively , HPV might have developed a strategy to access retromer even when the capsid itself is in the lumen of the endosome . Finally , virus might exit from the endosome into the cytoplasm where it can be recognized by retromer , which later mediates its entry into the Golgi . The experiments reported here demonstrate that retromer directly binds to the minor capsid protein of incoming HPV and that this interaction mediates export of virus particles from the early endosome into the retrograde vesicular pathway during the early stages of intracellular trafficking of incoming HPV .
Because retromer is required for the delivery of HPV16 to the Golgi apparatus and initiates endosome-to-Golgi transport of various cellular proteins , we hypothesized that a viral protein might bind to retromer . Inspection of the amino acid sequence of the HPV16 L2 protein revealed that its carboxy-terminal segment contains two short sequences , FYL and YYML , that resemble known retromer binding motifs , e . g . , aromatic followed by any amino acid followed by leucine or methionine [ФXL/M] , or phenylalanine or tryptophan followed by leucine followed by a valine or methionine [Trp/Phe-Leu-Met/Val] [37 , 38] ( Fig . 1A ) . To determine if these sequences are important for HPV infection , we constructed alanine scanning mutations across this segment of L2 ( Fig . 1A ) . Because it is difficult to introduce mutations into authentic HPV , we used pseudoviruses ( PsVs ) comprised of L1 and L2 encapsidating a reporter plasmid , which display the entry properties of authentic virus [39 , 40] . PsV containing wild-type L1 , wild-type or mutant L2 with a C-terminal HA or FLAG tag , and a GFP or HcRed reporter plasmid were produced in 293TT cells [24 , 40] . The assembly of mutant capsids was confirmed by encapsidation of the reporter plasmid and electron microscopy , which revealed no morphologic differences from wild-type PsV ( S1A Fig ) . In addition , when normalized by encapsidated reporter virus plasmids , wild-type and mutant pseudovirus preparations displayed a similar level of purity and contained similar levels of L1 and L2 ( S1B and S1C Fig ) . HeLa cells were infected with wild-type and mutant PsV stocks containing the same number of encapsidated reporter plasmids , corresponding to a multiplicity of infection ( MOI ) of approximately 0 . 5 for wild-type , and successful infection was measured two days later by flow cytometry for GFP fluorescence . As shown in Fig . 1B , several mutants were competent to infect cells . Strikingly , however , the FYL/AAA mutant lacking one of the putative retromer binding sites showed a >80% reduction in infectivity , and the YYML/AAAA mutant lacking the other site showed an approximately 50% reduction . When the FYL and YYML mutations were combined to generate the double mutant ( HPV16 . L2DM ) , infectivity was essentially abolished . The double mutant showed a similar defect when the HA tag on L2 was replaced with a FLAG tag . The double mutant was also defective in HaCaT cells , a human skin keratinocyte cell line commonly used in HPV entry studies ( e . g . , [19 , 20 , 25] ) ( Fig . 1C ) , which also require retromer for efficient infection ( S2 Fig ) . The defect caused by mutations in the putative retromer binding sites in L2 suggests that retromer may interact with these sequences to promote HPV infection . To explore this possibility , we replaced FYL with the sequence tryptophan-leucine-methionine ( WLM ) ( Fig . 1A ) , a retromer sorting signal from the cytoplasmic tail of the cation-independent mannose-6-phosphate receptor ( CIMPR ) , a cellular cargo of retromer [37] . As shown in Fig . 1A and 1C , insertion of WLM into the double mutant to generate HPV16 . L2WLM/DM partially restored the ability of HPV16 . L2DM to infect HeLa and HaCaT cells , and a mutant containing WLM and the endogenous YYML sequence ( HPV16 . L2WLM ) infected cells as well as wild-type PsV . Taken together , these results demonstrate that the putative retromer binding motifs in the carboxy-terminus of L2 are required for efficient infection and suggest that they act by binding to retromer . Alignment of the carboxy-terminal segment of the L2 protein from multiple HPV types showed that the FYL putative retromer binding site or a closely related sequence is highly conserved in all genera of HPV ( Fig . 1D ) , but the YYML sequence is not . This correlates with the more dramatic defect caused by removal of the FYL sequence . To test directly if the sequence FYL can act as a retromer sorting signal , we used an antibody internalization assay to determine if FYL is able to replace the endogenous sorting signal in a cellular retromer cargo [41] . The extracellular and transmembrane domains of the cell-surface protein CD8 were fused to the 160-amino acid cytoplasmic tail of CIMPR , which contains an endocytosis motif and the WLM retromer sorting signal identified by Seaman and colleagues [37] ( Fig . 2A ) . This signal mediates retromer-dependent trafficking of CIMPR from the endosome to the Golgi apparatus . HeLa cells were transfected with a plasmid expressing CD8-CIMPR containing a wild-type or mutant WLM sequence , and after 24 hours , live non-permeabilized cells were incubated with a CD8 antibody for three hours at 37°C . During this incubation , the CD8-CIMPR fusion protein containing the endogenous WLM sorting signal was endocytosed and trafficked to the TGN , which was scored by co-localization with the Golgi marker GM130 . As previously reported [37] , substitution of the WLM sequence with AAA abolished Golgi trafficking , resulting in a punctate distribution of the fusion protein throughout the cytoplasm . Strikingly , replacement of WLM with FYL restored Golgi localization ( Fig . 2B ) . Furthermore , siRNA-mediated knockdown of the Vps35 retromer subunit ( S3 Fig ) eliminated Golgi localization of the WLM and FYL CD8-CIMPR fusion proteins in the antibody internalization assay and resulted in dispersed punctate distribution of the fusion proteins ( Fig . 2B ) . These results demonstrate that FYL can act as a retromer sorting signal in a standard trafficking assay . The experiments described above imply that retromer sorting signals in the L2 protein are involved in HPV trafficking during virus entry . To determine the role of retromer and the putative retromer sorting motifs in HPV infection , we conducted experiments in infected cells . To first rule out the possibility that the defect in infectivity caused by the L2 mutations was due to an inability of the mutant to enter cells or undergo disassembly , we infected HeLa cells with PsV containing L2 with a carboxy-terminal FLAG epitope tag ( designated HPV16 . L2F ) . The FLAG tag is constitutively exposed on the surface of capsids and does not inhibit infectivity nor affect entry requirements , as assessed by sensitivity to several genetic and pharmacologic inhibitors of entry [24] . We stained cells at early times after infection with the anti-FLAG antibody or with the 33L1–7 antibody , which recognizes an epitope on the L1 protein that is inaccessible in intact capsids and reacts with L1 only after the capsid has entered cells and disassembly has begun [20 , 29 , 42] . Neither antibody stained uninfected cells , but cells infected with PsV containing wild-type or double mutant L2 displayed similar punctate intracellular staining with both antibodies ( S5 Fig ) , demonstrating that the inability of the L2 double mutant to infect cells is not due to a defect in internalization or initiation of capsid disassembly . To determine if retromer knock-down impairs exit of HPV PsV from the endosome , we used the proximity ligation assay ( PLA ) , a specific immune-based detection system in which a fluorescent signal is generated only when two proteins of interest are nominally within 40nm [43 , 44] . PLA was used to determine if L2 is in close proximity to the early endosome marker EEA1 or the trans-Golgi marker TGN46 during entry . At eight and 16 hours post-infection with HPV16 . L2F , we incubated fixed and permeabilized cells with anti-FLAG and the antibody recognizing the cellular component and then processed the samples for PLA . As expected , PLA did not generate signals in uninfected cells . When HeLa or HaCaT cells were infected with wild-type HPV16 . L2F and stained for L2-FLAG in proximity to EEA1 , a PLA signal was detected at eight hours post-infection , and by 16 hours the PLA signal associated with this compartment was reduced ( Figs . 3 , 4 , and S6 ) . In contrast , there was little L2/TGN46 PLA signal in the Golgi at eight hours post-infection , but there was abundant signal by 16 hours . These results indicate that virus enters the early endosome by eight hours after infection , but by 16 hours it has transited through this compartment and arrived at the Golgi . HPV trafficking was strikingly different in retromer knock-down cells . In this experiment , HeLa and HaCaT cells were transfected with siRNA targeting the retromer subunit Vps29 , infected with HPV16 . L2F two days later , and subjected to PLA at various times post-infection . An L2/EEA1 PLA signal was observed in both control and retromer knock-down cells eight hours after infection , confirming that retromer is not required for virus endocytosis . However , at 16 hours post-infection , the L2/TGN46 PLA signal was essentially undetectable , whereas there was a striking increase in the L2/EEA1 PLA signal ( Figs . 3 , 4 , and S6 ) . This is in marked contrast to infected parental cells , which contained little endosomal L2 at 16 hours after infection . These experiments demonstrate that retromer knock-down in both HeLa and HaCaT cells causes the accumulation of L2 in the early endosome and prevents the arrival of L2 into the Golgi . If the non-infectious phenotype of HPV16 . L2DM is due to interference with retromer recognition , this mutant should fail to exit from the early endosome in cells with intact retromer function . To test this , we infected HeLa and HaCaT cells with FLAG-tagged HPV16 . L2DM and used PLA to assess localization of incoming virus . As shown in Figs . 3 , 4 , and S6 , an L2/EEA1 PLA signal was observed in cells infected with HPV16 . L2DM at eight hours after infection . At 16 hours post-infection , an L2/TGN46 PLA signal was not observed , while the L2/EEA1 PLA signal markedly increased . Thus , mutation of the putative retromer binding motifs in the carboxy-terminal segment of L2 is functionally equivalent to knocking down retromer function . Importantly , when WLM was inserted into HPV16 . L2DM at the original position of FYL ( to generate HPV16 . L2WLM/DM ) , a substantial restoration of the L2/TGN46 PLA signal was observed at 16 hours after infection in both cell types , together with a reduction in the EEA1 signal at this time , confirming that the WLM retromer motif restores exit of L2 from the early endosome and trafficking to the Golgi . Similar results were obtained with multiple independent stocks of HPV16 . L2DM and HPV16 . L2WLM/DM . These results strongly suggest that the trafficking defect displayed by the double L2 mutant is due to impaired retromer binding . Recognition of cargo is a major factor underlying retromer recruitment to membranes [45] . To test whether the trafficking defect displayed by the L2 double mutant correlated with impaired association between L2 and endogenous retromer in infected cells , we performed PLA with anti-FLAG and anti-Vps35 . HeLa cells were infected with FLAG-tagged HPV16 PsV containing wild-type L2 , L2DM , or L2WLM/DM , and PLA was performed at eight and 16 hours post-infection . As shown in Fig . 5 , a PLA signal was observed for wild-type L2 and retromer at eight hours . This signal decreased substantially by 16 hours , consistent with a transient association between L2 and retromer as HPV exits the early endosome . In contrast , at eight hours after infection with HPV16 . L2DM , the PLA signal for L2 and Vps35 was decreased by 70% compared to cells infected with the wild-type PsV . The signal persisted at this level at 16 hours , despite the markedly increased amount of the mutant L2 protein in the early endosome at this time . Replacing FYL in the double mutant with WLM restored transient association of the L2 protein with retromer . Taken together , these results indicate that mutations that impair the association of L2 with retromer in infected cells also impair L2 export from the early endosome and demonstrate that the WLM sequence restores association between L2 and retromer , as well as restoring endosome exit . To determine if the carboxy-terminal segment of L2 is sufficient to bind to retromer , we conducted in vitro pull-down experiments . First , we employed biotinylated peptides , one from the amino-terminal portion of the L2 protein , one from the middle of the L2 protein , and one from the carboxy-terminal portion , including the two putative retromer binding sites ( Fig . 6A ) . These peptides were incubated with detergent lysates of uninfected HeLa and HaCaT cells , and cellular proteins that bound to the peptides were collected on streptavidin beads , subjected to SDS-polyacrylamide gel electrophoresis and immunoblotted for retromer subunits . The carboxy-terminal peptide containing the retromer motifs precipitated endogenous Vps29 and Vps35 , whereas the other two peptides were devoid of retromer binding activity ( Figs . 6B and 6C , left panels , and S7 ) . We also conducted pull-down experiments with carboxy-terminal peptides containing mutations in the retromer binding motifs . As shown in Fig . 6B and 6C , mutations in either retromer motif eliminated retromer binding in HeLa and HaCaT cell lysates , as did mutation of both sites . In some experiments , slight binding to the YYML/AAAA mutant was observed ( S8 Fig ) , but binding to the FYL/AAA mutant was never detected , suggesting that the FYL mutation causes a more severe defect in retromer binding , consistent with the more dramatic defect in infection caused by the FYL mutation . Importantly , replacement of FYL with the WLM retromer motif restored a significant level of retromer binding in extracts of either cell type ( Fig . 6 ) . Taken together , these results show that the retromer sorting motifs in the C-terminus of L2 bind to endogenous retromer in vitro , and that mutations that inhibit infectivity and endosome exit interfere with retromer binding . To determine if retromer directly recognizes L2 , we tested whether the carboxy-terminus of L2 was able to bind to active human retromer assembled from individual Vps26 , Vps29 , and Vps35 subunits purified from E . coli and immobilized on glutathione resin . We previously showed that retromer assembled in this way bound to the cellular retromer cargo , DMT1-II [45] . A 24-amino acid wild-type or mutant segment of L2 containing the retromer binding sites was fused to poly-histidine-tagged maltose binding protein ( MBP ) , which was also expressed and purified from E . coli ( Fig . 7A ) . The L2 fusion protein was incubated with immobilized retromer , and the L2 fusion protein bound to retromer was eluted and detected following SDS-PAGE . As shown in Figs . 7B and 7C , retromer captured the L2 fusion protein containing the carboxy-terminal segment of the wild-type L2 protein , indicating that retromer and this segment of L2 interact directly . In contrast , the FYL and YYML alanine substitutions , alone or in combination , drastically decreased retromer binding . Thus , the carboxy-terminal segment of the L2 protein binds directly to retromer via sites required for exit from the early endosome .
Viruses utilize cellular machinery to reach the site of viral genome replication . Therefore , studies of virus entry not only reveal important features of the virus life cycle , but also elucidate the mechanisms cells use to ensure that cellular components are present in their proper intracellular locations . We previously identified retromer as a factor required for trafficking of HPV16 to the Golgi apparatus during infection [22] , but our published experiments did not determine if retromer plays a direct or indirect role in HPV infection . HPV is a non-enveloped virus that lacks transmembrane proteins and is present in the endosomal lumen early during entry . Thus retromer , which is present in the cytoplasm and transports transmembrane proteins , could act indirectly on a cellular cargo to support HPV entry , or it could recognize the HPV capsid in an unconventional manner . The experiments reported here reveal that the papillomavirus capsid is a new class of retromer cargo and that a direct interaction between retromer and the L2 minor capsid protein is required for L2 to exit the endosome and traffic to the Golgi . Because L2 is closely associated with the viral genome throughout the entry process and manipulations that interfere with L2 trafficking also inhibit infectivity , we conclude that the observed behavior of L2 reflects the behavior of the viral components required for infectivity . Several lines of evidence demonstrate that L2 is a retromer cargo . Retromer knock-down causes HPV L2 to accumulate in the early endosome in HeLa and HaCaT cells . Furthermore , the carboxy-terminal segment of the L2 protein contains short sequences that resemble known retromer binding motifs , and mutations in these motifs interfere with the ability of L2 to associate with retromer in infected cells and inhibit the export of PsV from the early endosome and its delivery to the Golgi apparatus in HeLa and HaCaT cells . Importantly , these defects are rescued by replacement of the major retromer binding site in L2 with a retromer sorting signal from a cellular protein . In addition , this L2 sequence can function in a standard retromer sorting assay . Finally , in vitro binding studies showed that these sites in the L2 protein bind directly to retromer . Taken together , these results demonstrate that a direct interaction between the carboxy terminus of the L2 protein and retromer is required for exit of L2 from the early endosome and subsequent entry into the Golgi during HPV16 entry . These findings strongly suggest that retromer sorts HPV into an endosome-derived transport vesicle that ferries HPV or a subviral structure containing L2 and viral DNA to the TGN . Elsewhere , we showed that inhibition of γ-secretase blocks HPV trafficking after export from the early endosome but before delivery to the Golgi and the ER [24] , indicating that γ-secretase is required in HPV entry after retromer action . The peptide and fusion protein pull-down experiments suggest that both the FYL and YYML motifs are required for efficient retromer binding in vitro , although YYML appears less important ( S8 Fig ) . However , the more dramatic infectivity defect caused by the FYL mutation and the lack of YYML conservation indicates that the FYL sequence is more important in cells and during natural infection . FYL or a closely related sequence is present in the same position in all sequenced HPV L2 proteins examined and in the great majority of animal papillomaviruses , implying that the ability of retromer to recognize the L2 protein and export it from the early endosome arose early during papillomavirus evolution and remains an important feature of the virus life cycle . Some positions flanking FYL are also highly conserved , but they are not required for HPV16 entry in the assays used here . We also note that FYL and YYML do not match the canonical CIMPR retromer binding motif , Trp/Phe-Leu-Met/Val . FYL does appear similar to the ФXL/M motif in the mammalian iron transporter , DMT1-II [38] . However , comparison of this sequence in numerous HPV types reveals important differences . Although the ФXL/M motif can accommodate tryptophan in the first position , in 45 HPV L2 proteins examined , ~70% contain phenylalanine at this position , ~30% contain tyrosine , and none contain tryptophan . At the second position , ~75% HPV contain tyrosine and only one contains leucine , while wild-type DMT1-II contains leucine . Thus , although the HPV L2 and DMT1-II motifs are related , the examination of numerous HPV types suggests that evolution has selected specific residues at these positions in the viral protein that differ from the sequence in DMT1-II . Similar non-canonical retromer binding motifs may exist in additional retromer cargos from cells or other viruses . The L2 protein may be analogous to capsid proteins of other non-enveloped viruses that undergo conformation changes that expose hydrophobic peptides that insert into cell membranes and disrupt them to allow capsid entry into the cytoplasm [46 , 47] . We propose that the carboxy-terminal segment of L2 causes a more subtle perturbation of membrane structure , allowing it to protrude through the endosomal membrane where it is recognized by retromer on the cytoplasmic face of the membrane . Cells may contain other similar , as-yet-unrecognized , non-conventional retromer cargos , and other viruses may use a similar entry mechanism . A peptide derived from the carboxy-terminal segment of L2 including YYML and an adjacent , conserved highly basic sequence exhibits membrane bilayer-disrupting activity in vitro and can mediate integration of a reporter protein into cell membranes [31] . This peptide disrupts membranes only at low pH [31] . Therefore , endosome acidification may trigger its penetration through the endosomal membrane during virus entry . Such behavior may explain at least in part the requirement for endosome acidification during HPV entry [17 , 18 , 23] . It is possible that the carboxy-terminal segment of L2 cooperates with another segment of the L2 protein exposed on the surface of the capsid to mediate membrane penetration and/or retromer recognition in intact cells . The L2 amino-terminus contains a transmembrane-like domain required for infectivity [34] . This L2 segment may stabilize the association of L2 with the endosomal membrane or retromer or facilitate the passage of the carboxy-terminal segment containing the retromer binding sites through the membrane . In addition , a central segment of L2 binds SNX17 , another cytoplasmic protein required for efficient infection [30] , implying that this L2 segment is also accessible to the cytoplasm in some situations . Interestingly , SNX17 is reported to cooperate with retromer in mediating recycling of the Notch ligand , Jag1a [48] . Finally , the basic sequence adjacent to YYML resembles motifs present in cell-penetrating peptides , which can carry protein cargo across membranes into cells [49] . Further analysis will determine if any of these sequences play a role in retromer action during HPV infection . In summary , these studies elucidate the role of retromer in HPV entry , demonstrating that it binds directly to the HPV capsid and mediates export of this unconventional cargo from the early endosome on its journey into the cell . We propose that retromer recognition in this system involves a novel mechanism whereby a segment of a minor capsid protein of a non-enveloped virus protrudes through a cellular membrane into the cytoplasm . Further analysis of the role of retromer in HPV16 infection will provide new insights into virus entry , retromer function , and intracellular trafficking .
HaCaT cells are a spontaneously immortalized line of human skin keratinocytes obtained from G . Paolo Dotto ( Massachusetts General Hospital ) [50] . 293TT cells were obtained from Dr . Christopher Buck ( NIH ) . HeLa-Sen2 cells ( designated here HeLa cells ) are a previously described cloned strain of HeLa cells that infects efficiently with SV40 and HPV16 pseudovirus [51] . HeLa-M cells , a strain of HeLa-S3 cells that transfects efficiently , were obtained from Walther Mothes ( Yale University ) . All cells were cultured in Dulbecco’s MEM ( DMEM ) with 10% fetal bovine serum ( FBS ) , 10mM L-glutamine , 10mM HEPES pH 7 . 2 and standard antibiotics ( Pen/Strep ) . HPV16-GFP pseudovirus containing an HA tag at the C-terminus of L2 ( designated here HPV16 . L2HA ) was generated by using a plasmid obtained from Patricia Day . We also used PsV designated HPV16 . L2F , which contains an L2 protein with a carboxy-terminal 3xFLAG-tag constitutively exposed on the surface of the capsid [24] . The HPV16 L2 C-terminal mutants were produced in either the HA- or the FLAG-tagged p16sheLL expression plasmid using the QuickChange site-directed mutagenesis system and the primers listed in S1 Table . The L1 and L2 genes in each mutant were sequenced in their entirety . Mutations in the retromer sorting motif in the CD8-CIMPR fusion protein were inserted into a plasmid expressing the wild-type fusion protein ( a gift from Matthew Seaman , Cambridge Institute for Medical Research ) . siRNA targeting Vps26 , Vps29 , and Vps35 and the control scrambled siRNA were purchased from Dharmacon ( Lafayette , CO ) . Sequences of all oligonucleotides used in this study are listed in S1 Table . pCINeo-GFP plasmid ( obtained from Christopher Buck ( NIH ) and pCAG-HcRed plasmid ( purchased from Addgene , Plasmid 11152 ) were used as reporter plasmids . Pseudoviruses were produced by co-transfecting 293TT cells with a p16sheLL plasmid expressing L1 and wild-type or mutant L2 and a reporter plasmid , and purified by density gradient centrifugation in OptiPrep ( Sigma-Aldrich , #D1556 ) as previously described [22 , 40] . Encapsidated GFP or far-red genomes were quantified by qPCR as described [52 , 53] . Briefly , 5 μl of each pseudovirus preparation was treated with 4 μl of RQ1 DNAase ( Promega , M6101 ) in 100 μl DNAase buffer ( 50mM Tris HCl pH 7 . 6 , 10mM MgCl2 ) for one hour at 37°C . The DNAase was inactivated by incubation at 75°C for 30 min , followed by the addition of 50 μg of proteinase K ( PK , Roche ) for one hour at 37°C , in PK buffer ( 10mM Tris HCl pH 8 . 0 , 10mM EDTA , 0 . 25% SDS ) [52] . DNA was isolated using a PCR purification kit , and the number of encapsidated genomes was determined by qPCR using primers for the GFP or far-red gene , using a 10-fold serial dilution of pCINeo-GFP or pCAG-HcRed plasmid ( 109 to 103 genomes/μl ) analyzed on the same plate as a standard . Encapsidated genomes for all of the PsV stocks used in any one experiment were quantified in parallel . To examine the purity and content of L1 and L2 in PsVs , Optiprep-purified PsV preparations containing FLAG-tagged L2 were denatured in SDS-Laemmli sample buffer ( 108 packaged reporter plasmid genomes/lane ) and electrophoresed on a SDS-10% polyacrylamide gel for 1 . 5h at 150V . Proteins were subjected to silver staining or transferred from the gel to PVDF membrane ( Millipore Immobilon , 0 . 2 μm , ISEQ15150 ) , which was probed with 0 . 5 μg/mL primary anti-L1 ( BD Pharmingen , 554171 ) or anti-FLAG ( Sigma , F3165 ) antibody . Following incubation with 1:10 , 000 dilution of horseradish peroxidase-coupled secondary antibodies , bands were visualized by luminescence ( SuperSignal West Pico , Thermo Scientific , 34080 ) . Freshly glow-discharged 200 mesh Formvar/carbon-coated copper grids ( Electron Microscopy Services , CG200-Cu ) were inverted on drops of gradient-purified PsV diluted 1:9 in phosphate-buffered saline ( PBS ) , and virus allowed to adsorb for five minutes . The grids were washed twice in deionized water and stained by two one-minute incubations on drops of Nano-W ( Nanoprobes , Nephank , NY ) before removing excess stain by gentle blotting with Whatman #1 filter paper . The grids were air-dried before viewing on a FEI Tecnai Biotwin transmission electron microscope at 80Kv . Images were taken using a Morada CCD camera and iTEM ( Olympus ) software , and ImageJ was used to provide sizing information based on a scale bar embedded in the images . 5×104 HeLa or HaCaT cells were plated in 12-well plates . Cells were infected with wild-type HPV16 PsV at MOI ~0 . 5 GFP-transducing units per cell ( i . e . , enough virus to result in GFP expression in one-half of the cells as assessed by flow cytometry on a BD Biosciences FACSCalibur flow cytometer 48 hours post-infection ) . The number of packaged wild-type reporter plasmids required to achieve this MOI in unmanipulated cells was quantified by qPCR , and an equivalent number of mutant genomes were used to infect cells . Depending on the experiment , 150–300 reporter plasmid genomes per cell resulted in an MOI of ~0 . 5 for wild-type pseudovirus in HeLa cells; approximately five-fold more virus was required to attain this MOI in HaCaT cells . In some experiments , cells were transfected with siRNA prior to infection . To confirm retromer knock-down , 5×105 cells were plated in a 6-well plate and reverse-transfected with 40nM siRNA targeting Vps26 , Vps29 , or Vps35 . Forty-eight hours later , cells were lysed in sample buffer , electrophoresed , and analyzed by immunoblotting for Vps35 . 3 to 5×104 HeLa cells were plated in eight-well chambered glass slides and infected the next day with wild-type PsV at MOI 20 or mutant PsV containing the same number of reporter plasmid genomes . ( The lower sensitivity of immunofluorescence or PLA [see below] compared to reporter gene expression necessitated an MOI of 20 or 50 to visualize virus components . ) Eight hours post-infection , the cells were fixed for 15 min at room temperature with 4% paraformaldehyde ( Electron Microscopy Sciences , #15710 ) , washed with PBS and then permeabilized with 0 . 5% Triton X-100 for 20 min at room temperature in PBS . The cells were blocked for one hour in 1% bovine serum albumin and 3% goat serum , and immunostained with 1:200 33L1–7 ( obtained from Martin Sapp ( LSU ) ) . After extensive washes , AlexaFluor 594-conjugated goat anti-mouse secondary antibodies were added at 1:300 dilution for 40 minutes at room temperature . Nuclei were stained with 1:100 dilution of 4′ , 6-diamidino-2-phenylindole ( DAPI ) , cells were washed extensively , and slides were mounted in Prolong gold anti-fade ( Molecular Probes ) . Images were recorded on a ZEISS Axiovert 200 inverted fluorescent microscope using appropriate filters processed with ImageJ . 3×104 HeLa-M cells in 8-chambered glass slides were transfected with 10nM Vps35 or RISC-free siRNA . Twenty-four hours later , the Trans-IT HeLaMONSTER reagent ( Mirus Bio ) was used to transfect cells with 1 μg of a plasmid expressing a CD8-CIMPR fusion protein containing WLM ( wild-type ) , AAA , or FYL . Twenty-four hours later , live non-permeabilized cells were incubated at 37°C with a 1:400 dilution of an antibody that recognizes the extracellular domain of CD8 ( Ancell , 153–020 ) . After three hours , the cells were fixed for 15 min at room temperature with 4% Formalde-Fresh , permeabilized with 0 . 5% Triton X-100 for 20 min , and then blocked with 5% donkey and goat serum for 30 minutes at room temperature . The cells were then stained with anti-GM130 ( Abcam , ab52649 [1:200] ) overnight at 4°C , washed five times with the blocking solution , and incubated with conjugated secondary antibody ( Life Technologies [1:500] ) for 30 minutes at room temperature . Cells were mounted with Duolink in situ Mounting Medium with DAPI , imaged on a ZEISS Axiovert 200 inverted fluorescent microscope and processed with Image J . 5×104 HeLa cells grown overnight on glass coverslips in 24-well plate were transfected with siRNA targeting retromer subunit Vps29 or control scrambled siRNA with Lipofectamine RNAi Max reagent ( Life Technologies , Carlsbad , CA ) 48 hours prior to infection . The cells were then infected with wild-type or mutant HPV16 . L2F at MOI of 50 ( according to genome normalization ) , fixed with 4% Formalde-Fresh at eight or sixteen hours post-infection , and permeabilized with 1% saponin for one hour at room temperature . The cells were incubated with anti-FLAG mouse antibody ( Sigma , F3165 [1:1000] ) to label L2 and an antibody recognizing EEA1 ( Cell Signaling , C45B10 [1:100] ) or TGN46 ( Abcam , ab50595 [1:200] ) . Alternatively , cells were incubated with anti-FLAG rabbit antibody ( Cell Signaling , 2368 [1:500] ) and anti-Vps35 antibody ( Abcam , 57632 [1:500] ) . PLA was performed with Duolink reagents from Olink Biosciences ( Uppsala , Sweden ) as described [22 , 54] . Briefly , samples were incubated with a pair of suitable PLA probes at 1:5 in a humidified chamber for one hour and processed for ligation for 30 min at 37°C . DNA was then amplified with fluorescent substrates for 100 min at 37°C . The nuclei were stained by incubation with 5μg/ml DAPI for 10 min and images were acquired as described above . Approximately 100 nuclei were imaged in each sample . The images were processed with ImageJ and quantitatively analyzed with BlobFinder software to measure total fluorescence intensity in each sample . The average fluorescence intensity per cell in each sample was normalized to the control sample as indicated in each experiment . All the experiments were done independently three times with similar results , and one representative experiment is shown . Peptides shown in Fig . 6 were purchased from NeoBioLab ( Cambridge , MA ) at >95% purity . L2-N was biotinylated at its C-terminus with the N-terminus unmodified , while all the other peptides were biotinylated at their N-terminus and amidated at their C-terminus . L2-N was dissolved in sterile deionized water containing 0 . 01% sodium azide , L2-M was initially solubilized in a small amount of DMSO ( ~ 70–80 μl ) and then dissolved in sterile deionized water with 0 . 01% sodium azide . L2-C was initially resuspended in 30% acetic acid and DMSO ( ~ 70 μl each ) , and then dissolved in sterile deionized water with 0 . 01% sodium azide . The FYL , YYML , DM and WLM peptides were initially solubilized in a small amount of 30% acetic acid ( ~80 μl ) , and then dissolved in sterile deionized water with 0 . 01% sodium azide . Peptide stocks ( 3 . 5–5 . 6 mg/ml ) were aliquoted and stored at-20°C . HeLa or HaCaT cells plated in six-well plates were lysed at ~80% confluency with 500 μl RIPA-MOPS buffer ( 20mM morpholinepropanesulfonic acid [pH 7 . 0] , 150mM NaCl , 1% Nonidet P-40 , 1mM EDTA , 1% deoxycholic acid , 0 . 1% sodium dodecyl sulfate [SDS] ) supplemented with protease inhibitors ( 1X HALT protease and phosphatase inhibitor cocktail [Thermo Scientific] ) or with 500 μl HEPES buffer ( 20mM Hepes pH8 , 50mM NaCl , 5mM MgCl2 , 1mM dithiothreitol ( DTT ) , and 1 . 0% triton X-100 ) containing 1:100 Halt TM Protease & Phosphatase Inhibitors ( Thermo Scientific , Prod # 78443 ) . The lysate was centrifuged at 14 , 000 rpm for 20 min , and the supernatant was incubated with 10 μg of a biotinylated peptide for two hours at 4°C . 40 μl of streptavidin agarose beads slurry ( Pierce , cat# 20349 ) was added , and the mixture was gently rocked for 45 min at 4°C . Beads were recovered by centrifugation and washed four times with RIPA-MOPS buffer supplemented with NaCl to a total of 0 . 4M or with HEPES buffer . Samples were analyzed by SDS-PAGE and immunoblotting with Vps35 ( Abcam , ab57632 ) or Vps29 ( Santa Cruz , sc-98611 ) antibody . Individual human Vps26 , Vps29 , and GST-tagged Vps35 subunits were expressed individually in E . coli , and the assembled trimeric retromer complex was immobilized on GSH resin via the GST-tag on Vps35 as described [45] . Maltose binding protein ( MBP ) -L2–6His fusion proteins containing a C-terminal segment from wild-type or mutant HPV16 L2 ( amino acids 434–457 ) were expressed in bacteria and purified using the AKTA-Prime plus FPLC system equipped with a His-trap column . The sequence appended to the C-terminus of MBP in the pMal-C2 expression vector ( New England Biolabs ) was GSASPQYTIIADAGDFYLHPSYYMLRKHHHHHHC ( L2 sequence underlined ) . Purified proteins were exchanged into 20mM Hepes pH 8 , 50mM NaCl and quantified by bicinchoninic acid assay . Five or 10 μM of each fusion protein was incubated with assembled retromer trimer immobilized on GSH resin for two hours at 4°C in 20mM HEPES pH 8 . 0 , 50mM NaCl , 5mM MgCl2 , 1mM DTT , and 0 . 1% Triton X-100 . Beads were washed twice in HEPES buffer , suspended in SDS loading buffer , boiled , and subjected to SDS-PAGE and anti-His immunoblotting . Bands corresponding to the MBP-L2-His constructs were quantified by Image Lab . | The human papillomaviruses are important carcinogens , but little is known about how these non-enveloped viruses traffic to the nucleus , the site of genome replication . We use imaging , biochemical , and genetic techniques to show that a multi-subunit intracellular trafficking machine known as retromer binds directly to the L2 minor capsid protein in the virus particle to initiate its transport from the endosome to other membrane-bound organelles farther inside the cell . Most notably , knock-down of retromer expression or mutation of newly identified retromer binding sites in L2 cause the accumulation of incoming HPV16 capsids in the endosome and prevent trafficking to the Golgi . These defects can be corrected by insertion of a retromer binding site from a cellular cargo . Because all previously known retromer cargoes are cellular transmembrane proteins , the virus represents a new class of retromer cargo . In addition to elucidating the mechanism of viral endosome escape , these results suggest that retromer may play a more versatile role in cell biology than previously recognized . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Direct Binding of Retromer to Human Papillomavirus Type 16 Minor Capsid Protein L2 Mediates Endosome Exit during Viral Infection |
Adult Clonorchis sinensis lives in the bile duct and causes endemic clonorchiasis in East Asian countries . Phosphagen kinases ( PK ) constitute a highly conserved family of enzymes , which play a role in ATP buffering in cells , and are potential targets for chemotherapeutic agents , since variants of PK are found only in invertebrate animals , including helminthic parasites . This work is conducted to characterize a PK from C . sinensis and to address further investigation for future drug development . A cDNA clone encoding a putative polypeptide of 717 amino acids was retrieved from a C . sinensis transcriptome . This polypeptide was homologous to taurocyamine kinase ( TK ) of the invertebrate animals and consisted of two contiguous domains . C . sinensis TK ( CsTK ) gene was reported and found consist of 13 exons intercalated with 12 introns . This suggested an evolutionary pathway originating from an arginine kinase gene group , and distinguished annelid TK from the general CK phylogenetic group . CsTK was found not to have a homologous counterpart in sequences analysis of its mammalian hosts from public databases . Individual domains of CsTK , as well as the whole two-domain enzyme , showed enzymatic activity and specificity toward taurocyamine substrate . Of the CsTK residues , R58 , I60 and Y84 of domain 1 , and H60 , I63 and Y87 of domain 2 were found to participate in binding taurocyamine . CsTK expression was distributed in locomotive and reproductive organs of adult C . sinensis . Developmentally , CsTK was stably expressed in both the adult and metacercariae stages . Recombinant CsTK protein was found to have low sensitivity and specificity toward C . sinensis and platyhelminth-infected human sera on ELISA . CsTK is a promising anti-C . sinensis drug target since the enzyme is found only in the C . sinensis and has a substrate specificity for taurocyamine , which is different from its mammalian counterpart , creatine .
Clonorchis sinensis is an important food-borne trematode parasite , which causes clonorchiasis in human and mammalian animals . This parasite is endemic in China , Korea , Taiwan , and northern Vietnam . Globally , 35 million people are estimated to be infected by C . sinensis , with approximately 15 million of these cases being in China [1] . The highest prevalence of C . sinensis human infection is reported in southern and northeastern parts of China , especially in Guangdong , Guangxi , and Heilongjiang provinces [2] . In humans , clonorchiasis can provoke severe pathologic changes in the hepatobiliary tract , and was recently recognized as belonging to the group of biological carcinogenic agents that cause cholangiocarcinoma [3] , [4] . The public health and economic impact of clonorchiasis is considerable and has inspired the development of vaccines and drugs in addition to other public health measures of control and eradication the parasite itself [5] . Phosphagen kinases ( PKs ) catalyze a transfer of high-energy gamma phosphoryl group from ATP to a guanidino group on an acceptor molecule ( phosphagen+MgADP+H+M guanidine acceptor+MgATP ) . The phosphorylated guanidines , called as phosphagens , serve an important molecule in cellular energy homeostasis . These phosphagens provide high-energy phosphates , which are accessible for ATP generation at substrate-level phosphorylation during periods of high metabolic demands [6] . Members of this enzyme family play a key role in the interconnection between energy production and utilization in animals . Among vertebrates , phosphocreatine is the sole phosphagen , and the corresponding kinasing enzyme is creatine kinase ( CK ) . In invertebrates , seven unique phosphagens and corresponding kinases were identified in addition to phosphocreatine [7] , [8] , [9]: glycocyamine kinase ( GK ) , taurocyamine kinase ( TK ) , lombricine kinase ( LK ) , opheline kinase ( OK ) , hypotaurocyamine kinase ( HTK ) , thalessemine kinase ( ThK ) , and arginine kinase ( AK ) . Several studies have also described the presence of phosphagen kinases from parasites [10] . Studies on arginine kinase from the protozoa Trypanosoma cruzi have identified AK as a potential target for novel drug development for Chagas' disease [11] , [12] . AK has also been presented in Toxocara canis [13] , and Ascaris suum [14] . Recently , two-domain TKs were reported from parasitic trematodes Paragonimus westermani [15] and Schistosoma japonicum [16] . Absence of these invertebrate PKs from the mammalian hosts , including human , imply that these kinases area possible target for new candidate drug against parasites and for development of new diagnostic reagent to detect infections . We found a cDNA clone encoding a polypeptide ( CsTK ) from the C . sinensis transcriptome database , which was homologous with the two-domain TKs as well as the PKs of other organisms . This study was performed to elucidate biomolecular functions of CsTK such as catalytic activity comparing with TKs and PKs , tissue localization , and developmental expression .
BALB/c mice ( female , 7-week-old ) and rabbits ( New Zealand White , male , 2 . 2–2 . 5 kg ) were handled in an accredited animal facility at Chung-Ang University ( Korea FDA; Unit Number 36 ) . Approval for animal experiments was obtained from the Institutional Animal Care and Use Committee at Chung-Ang University ( Permit Number: CAU-2011-0052 and CAU-2013-0005 ) . This study was carried out in strict accordance with the recommendations provided in the Guide for the Care and Use of Laboratory Animals by the US National Institutes of Health . The adult worms of C . sinensis were recovered from infected rabbits . Total RNA was isolated from adult worms by acid guanidinium thiocyanate-phenol-chloroform extraction method [17] . Messenger RNA ( mRNA ) was purified from total RNA using a poly ( A ) + isolation kit ( Nippon Gene , Tokyo , Japan ) . Single-stranded cDNA was synthesized with Ready-To-Go You-Prime First-Strand Beads ( Amersham Pharmacia Biotech , NJ , USA ) with a lock-docking oligo-dT primer [18] . Polymerase chain reaction ( PCR ) was carried out with reaction mixture containing cDNA , 10 pmol of each primer , 2 µl of 2 . 5 mM of dNTPs , 1 U of Ex Taq polymerase , 2 . 5 µl of 10× Ex Taq buffer ( TaKaRa , Tokyo , Japan ) . Thermal cycles were prepared as follows: initial denaturation at 94°C for 2 min , followed by 35 cycles of 94°C for 30 s , annealing at 50°C for 35 s , and extension at 72°C for 2 min , and a final extension at 72°C for 4 min . PCR was done in a thermal cycler , MyCycler ( BioRad , Foster , USA ) . The 3′-half of the cDNA was first amplified with lock-docking oligo ( dT ) primer and an “universal” redundant oligonucleotide primer 5′-GT ( ACGT ) TGG ( AG ) T ( ACGT ) AA ( TC ) GA ( AG ) GA ( AG ) GA ( TC ) CA-3′ , designed for amplification of PK [19] . PCR products were purified using GENE CLEAN Kit ( Funakoshi , Tokyo , Japan ) . Purified PCR product ( 400 bp ) was ligated into pGEMT-vector ( Promega , USA ) and transformed into Escherichia coli JM109 cells . Positive clones were obtained and plasmid DNA was extracted . Nucleotide sequences were determined with an ABI PRISM 3100-Avant DNA sequencer using a BigDye Terminators v3 . 1 Cycle Sequencing Kit ( Applied Bio-systems , Foster , CA , USA ) with two T-vector-specific primers , T7 and SP6 . The 5′-half of the cDNA was amplified as follows: A poly ( G ) + tail was added to the 5′-end of the C . sinensis cDNA pool with terminal deoxynucleotidy transferase ( Promega , Madison , WI , USA ) . 5′-half of cDNA of C . sinensis PK was amplified using oligo-dC primer ( 5′-GAA TTC18-3′ ) and PK-csR0 primer ( 5′-CCA AAT TAC TCG GGC AAC AA -3′ ) designed on the sequence of 3′-half . To confirm a tandem connection of CsTK D1 and D2 domains , central region of the cDNA was amplified using inner specific primers that were designed on the sequences of cDNA obtained by 5′-RACE and 3′-RACE PCR until full sequence of the cDNA was obtained [20] . With the PCR products , T-vector cloning and 3′-sequence determination of C . sinensis PK were performed as described above . PCRs were done in total volumes of 50 µl . The reaction mixture contained cDNA of D1D2 , 10 pmol of csPKXbaI forward primer ( 5′-TCT AGA ATG CAG GTC GAA CCA CTG AAA TC-3′ ) , 10 pmol of csPKPstI reverse primer ( 5′-CTG CAG CTA TGG CAA GGA TTT TTC AAT AGC -3′ ) , 1 U of KOD Plus DNA polymerase ( Toyobo Co . , Ltd . , Tokyo , Japan ) , 5 µl of 10× KOD Plus buffer , 5 µl of 2 mM KOD dNTPs , and 4 µl of 25 mM MgSO4 . The amplified products were purified using QIA quick PCR purification columns ( QIAGEN GmbH , Hilden , Germany ) . A-tail was added to 3′-end of the purified PCR fragments ( blunt-ended ) . A-tailing was done in a total volume of 30 µl containing purified KOD PCR product , 15 U of Gene Taq DNA polymerase ( Wako Nippon Gene ) , 3 µl of 10× Gene Taq buffer and 1 . 2 µl of 5 mM dNTP . This mixture was incubated at 70°C for 30 min . The resulting product was purified , subcloned into pGEMT-vector , and sequenced as described above . Coding region of C . sinensis PK cDNA of D1D2 was cloned into XbaI/PstI site of pMAL-c2 ( New England Biolabs , Ipswich , MA , USA ) . Maltose binding protein ( MBP ) -C . sinensis PK fusion protein was expressed in E . coli TB1 cells by induction with 1 mM IPTG at 25°C for 24 h . The cells were resuspended and sonicated in 5× TE buffer . Soluble recombinant fusion protein was purified by affinity chromatography using amylose resin ( New England Biolabs , Ipswich , MA , USA ) . Homogeneity of the purified recombinant enzyme was verified by SDS–PAGE and placed on ice until assayed for enzymatic activity within 12 h . MBP-tagged CsTK D1 and D2 expressed and purified as described above . Using Easy-DNA Kit ( Invitrogen , Carlsbad , USA ) , genomic DNA was isolated from a C . sinensis adult worm . PCR was performed with Ex Taq polymerase ( TAKARA ) and primers designed on the ORF . PCR conditions were as follows: initial denaturation at 94°C for 2 min , followed by 35 cycles of 94°C for 30 s , annealing at 50°C for 30 s and extension at 72°C for 3 min and a final extension at 72°C for 4 min . PCR product was purified and sequenced as described above . Using programs CLUSTAL W ( http://www . ddbj . nig . ac . jp ) and GENETYXMAX ( ver . 6 . 0 ) , multiple sequence alignment was performed . Phylogenetic analysis was done using the distance method in MEGA ( ver . 5 . 0 ) . For distance analyses , the Kimura 2-parameter model was used to construct the distance matrix , and the tree was inferred from this using the Neighbor-Joining ( NJ ) approach . Bootstrap re-sampling was performed to assess the degree of support for groupings on the tree . Accession numbers of other amino acid sequences used in the present study are shown in Table S1 . The following amino acid substitutions were amplified in the template of pMAL/C . sinensis TK wild type ( WT ) : R58A , I60A , Y84A and Y84R of TKD1; H61A , I63A , Y87A and Y87R of TKD2; R58A , I60A , Y84A and Y84R of TKD1D2 in D1 region; H61A , I63A , Y87A and Y87R of TKD1D2 in D2 region . Substitution was made using KOD+-DNA polymerase under the subsequent PCR conditions: initial denaturation at 94°C for 2 min , followed by 35 cycles of 94°C for 15 s , annealing at 60°C for 30 s and extension at 68°C for 9 min and a final extension at 72°C for 5 min . The primer sequences designed were as follows: CsPKMutD1R58Af: 5′-GCT TGC ATC CTT CCT CGC G-3′ , CsPKMutD1R58Ar: 5′- CGG ATT ACG AGC ATT GTG ACT GAC-3′; CsPKMutD1I60Af: 5′-GCT CTT CCT CGC GCT TGT GAT TTG-3′ , CsPKMutD1I60Ar: 5′-GCA CCG CGG ATT ACG AGC ATT GTG-3′; CsPKMutD1Y84Af: 5′-GCT CAT AAG GTG AAA GGA GAC-3′ , CsPKMutD1Y84RAr: 5′-GTC TAT AAT AAC GGC GTC AAA G-3′; CsPKMutD1Y84Rf: 5′-CGA CAT AAG GTG AAA GGA GAC-3′; CsPKMutD2H61Af: 5′-GCT TCA ATC TGT CCA CGG TAC TGG-3′ , CsPKMutD2H61Ar: 5′-TGG GTT GTA AGC ACC GTT ACG-3′; CsPKMutD2I63Af: 5′-GCT TGT CCA CGT ACT GGA GAA GC-3′ , CsPKMutD2I63Ar: 5′-TGA ATG TGG GTT GTA AGC ACC GT-3′; CsPKMutD2Y87Af: 5′-GCT CAT GGA GTG AGT GAC CCA GCT T-3′ , CsPKMutD2Y87RAr: 5′-GTC CAA AAT CAC TGC ATC CAG GTA GTC-3′; CsPKMutD2Y87Rf: 5′-CGA CAT GGA GTG AGT GAC CCA GCT T-3′ . PCR products were purified by QIAquick PCR purification column ( Qiagen GmbH , Hilden , Germany ) . After blunting and phosphorylation , the DNA was self-ligated . Expression and enzyme assay of the mutated proteins were performed as described above . Enzyme activity was measured by absorbance at a wavelength of 340 nm ( with MBP , UV/Visible Spectrophotometer 4300 Pro , Amersham , Biosciences ) with an NADH-linked assay at 25°C [21] , [22] . The reaction mixture ( total 1 . 0 ml ) contained 0 . 65 ml of 100 mM Tris–HCl ( pH 8 ) , 0 . 05 ml of 750 mM KCl , 0 . 05 ml of 250 mM Mg-acetate , 0 . 05 ml of 25 mM phosphoenolpyruvate prepared in 100 mM imidazole/HCl ( pH 7 ) , 0 . 05 ml of 5 mM NADH prepared in Tris–HCl ( pH 8 ) , 0 . 05 ml of pyruvate kinase/lactate dehydrogenase mixture prepared in 100 mM imidazole/HCl ( pH 7 ) , 0 . 05 ml of an appropriate concentration of ATP prepare in 100 mM imidazole/HCl ( pH 7 ) , and 0 . 05 ml of recombinant enzyme . The reaction was started by adding 0 . 05 ml of an appropriate concentration of guanidine substrate made up in 100 mM Tris–HCl ( pH 8 ) . Initial velocity values were obtained by varying the concentration of guanidine substrate under fixed concentrations of ATP . Protein concentration was estimated from an absorbance at 280 nm ( 0 . 77 AU at 280 nm in a 1 cm cuvette corresponds to 1 mg protein/ml ) . To measure mRNA transcript level in developmental stages of C . sinensis , quantitative real-time PCR ( qRT-PCR ) was performed using SYBR Green I dye with LightCycler Carousel-Based System ( Roche Applied Science , Indianapolis , IN , USA ) . cDNAs of C . sinensis adults and metacercariae were employed as templates of qRT-PCR . Four pairs of forward and reverse primers were designed on each 5′- and 3′-end of CsTK D1 and CsTK D2 using Oligo6 program ( Figure S1 Panel A ) . For qRT-PCR , forward primer 5′- TTT CCA CAA TGC CAA CAA GAC -3′ and reverse primer 5′- GCT TGA ATA CCC TGG ATG AGT -3′ on 3′-end of CsTK D2 were employed , producing a 442 bp amplicon ( Figure S1 , Panels B and C ) . qRT-PCR mix was prepared as of 1× SYBR green master mix , 1 µM gene-specific primers , and 80 ng total cDNAs . Reference genes employed were β-actin , phosphoglycerate kinase , and calcyphosine [23] . Thermal cycling of qRT-PCR started with pre-incubation at 95°C for 15 min , then continued 40 cycles of 95°C for 10 sec , 48°C for 10 sec , and 72°C for 30 sec . To verify a specific amplication of target mRNA , one melting cycle was run , 65°C for 1 min , and increase at 0 . 1°C/sec to 95°C to dissociate double-stranded amplicons . LightCycler software 4 . 05 ( Roche Applied Science , Penzberg , Germany ) was used to analyze melting curves and to calculate CT values . A ΔCT of target gene was calculated by subtracting an average CT of three reference genes from an average CT of the target gene . The ΔΔCT of target gene was analyzed using equation ΔΔCT = ( ΔCTtarget−average of ΔCTadult ) . The 2−ΔΔCT shows relative gene expression level [24] . To produce and purify recombinant CsTK D1 , a corresponding cDNA was subcloned into pET-23-Cs28GST vector . In this format , recombinant CsTK D1 was expressed as a fusion protein Cs28GST-tagged at its N-terminus . E . coli BL 21 ( DE3 ) pLysS was transformed with the expression construct and induced to produce the fusion protein using IPTG at 0 . 1 mM for 3 hrs . The fusion protein was absorbed to glutathione sepharose 4B column ( GE Healthcare , Uppsala , Sweden ) and washed with PBS . Recombinant CsTK D1 was cleaved off from Cs28GST tag on bead with 10 U/ml thrombin protease ( GE Healthcare , Buckinghamshire , UK ) overnight , and then eluted in PBS . Residual tagged protein was eluted in 5 mM reduced glutathione/PBS . Recombinant CsTK D1 was mixed with same amount of either complete or incomplete Freund adjuvant . BALB/c mice were injected peritoneally once with 40 µg antigen/200 µl of complete adjuvant mix and again with the same amount of incomplete adjuvant mix 2 weeks later . A final buster , 1 . 2 µl antigen in 30 µl PBS each mouse , was injected into a tail vein after 2 weeks . After 4–5 days , blood was taken and checked for antibody production toward recombinant CsTK D1 by western blotting . For soluble extract , Clonorchis sinensis adults were washed with PBS several times and homogenized in PBS/1% Triton X-100/proteases inhibitor ( 1× Complete Mini , EDTA-free , Roche Diagnostics ) . After keeping 4°C overnight , the homogenate was spun at 20 , 000×g for 60 min at 4°C and supernatant was saved as soluble extract or crude antigen of C . sinensis . The soluble extract , 20 µl , was deployed in 12% SDS-PAGE and transferred onto Hybond ECL membrane ( GE Healthcare , Uppsala , Sweden ) . The blotted membrane was incubated overnight in CsTK D1-immune mouse serum at 1∶5 , 000 dilution in skim milk/PBS , then in the secondary antibody , alkaline phosphatase-conjugated goat anti-mouse IgG at 1∶5 , 000 . Color was developed using BCIP/NBT substrate ( Sigma Co . , St . Louis , MO , USA ) and stopped in water . Adult C . sinensis flukes within a rabbit liver were fixed in 10% neutral formalin and processed for paraffin blocks . The sectioned flukes in paraffin ribbons were deparaffinized and rehydrated . The C . sinensis ribbons were incubated in CsTKD1-immune mouse sera at 1∶200 dilution for 30 min at room temperature . Then , the sections were incubated in peroxidase- and antimouse IgG antibody-conjugated dextran polymer ( to the dextran backbone , about 70 enzyme molecules and 10 primary antibodies were conjugated; Dako , Glostrup , Denmark ) for 30 min at room temperature . Color was developed in AEC+ substrate for 5 min . Normal mouse sera were used as negative control . Recombinant CsTK D1 protein , 1 µg/ml in carbonated buffer , was coated on 96-well plate at 4°C overnight . The wells were washed three times with PBS containing 0 . 1% Tween-20 ( PBS/T ) and incubated with human sera at 1∶100 dilution at 37°C for 1 hr . A secondary antibody , peroxidase -conjugated anti-human IgG ( MP Biomedicals , Santa Ana , CA , USA ) of 1∶4000 dilution was added to the wells and incubated at 37°C for 1 hr . Color was developed with a substrate , ophenylene diamine ( Sigma Co . , St . Louis , MO , USA ) , and optical density was measured at a wavelength of 490 nm . Human sera used were from 47 patients with clonorchiasis , 20 with opisthorchiasis viverinii , 14 with cysticercosis cellulosae , 14 with sparganosis erinacei and from 14 patients with paragonimiasis westermani . As a control group , serum samples from 28 parasite-free human subjects were employed . A cut-off line was set at an average+doubled standard deviation which was calculated with OD values of the control group .
During large-scale cDNA sequencing of an adult C . sinensis cDNA library , a 2 , 154 bp long cDNA was successfully amplified by PCR , which encoded for a polypeptide of 717 amino acids ( Fig . S2 ) and had 5′-UTR of 52 bp and 3′-UTR of 288 bp . Molecular mass of the polypeptide was estimated to be 80 , 274 Da with a pI of 7 . 88 , using ProtParam ( http://www . expasy . ch/tools/protparam . html ) . cDNA sequence analysis showed that this polypeptide consisted of two repetitive domains ( D1 and D2 ) homologous to known sequences of TKs ( Fig . S2 ) . D1 contained 360 amino acids with calculated molecular mass 40 , 573 Da and a pI of 8 . 09 , and D2 consisted of 357 amino acids with calculated mass 39 , 719 Da and a pI of 7 . 39 . This cDNA sequence was archived in GenBank under accession number JX435779 . cDNA sequence analysis using BLASTn revealed that C . sinensis TK D1 and D2 had 72 . 3% identity with P . westermani TK D1 and D2 . Further , the peptide sequence analysis , using BLASTp , showed that CsTK D1 and D2 were homologous with the respective domain in many TKs of different species ( Fig . S2 ) . C . sinensis TK polypeptide shared 79 . 2% sequence identity with P . westermani TK . CsTK D1 shared 77 . 7% identity with P . westermani TK D1 and 71 . 7% identity with S . mansoni TK D1 . Meanwhile , CsTK D2 shared 82 . 3% identity with P . westermani TK D2 and 67 . 9% identity with S . mansoni TK D2 ( Table 1 ) . With this sequence information , the polypeptide conceptually deduced from the C . sinensis cDNA clone was considered as a new member of PK of C . sinensis . Residues in GS region are highly conserved across animal PKs . I60 in GS region of C . sinensis TK was replaced by Val in all CKs . CsTK had not a mitochondrial targeting signal peptide in N-terminus ( Fig . S2 ) . A phylogenetic tree , constructed using NJ algorithm ( Fig . 1 ) , indicated that PKs can be grouped into two clusters . Trematode TKs including CsTK were grouped in Cluster 1 with mulluscan AK group and nematode , protozoan , and arthropod AK group . Cluster 2 was comprised of annelid PK group and CK group of CKs , GKs , LKs , and TKs from protozoan and various insect species . C . sinensis TK gene had 13 exons and 12 introns , encoding D1 and D2 domains . Introns of C . sinensis TK were located at amino acid positions 97 . 1 , 140 . 0 , 202 . 0 , 243 . 2 , 300 . 0 , 354 . 0 , 426 . 1 ( bridge intron ) in D1 , and at 97 . 1 . 1 , 167 . 1 , 243 . 2 , 300 . 0 , 366 . 1 in D2 ( Fig . 2 ) . The introns of C . sinensis TK began with GT and ended with AG ( GT–AG pattern ) , except for introns at position 202 . 0 ( GC–AG pattern ) in D1 . Size of the introns was variable from 115 bp to more than 4 , 000 bp ( Table S2 ) . Positions of C . sinensis TK introns were conserved between P . westermani TK and S . mansoni TK . Introns of C . sinensis TK 97 . 1 , 243 . 2 , and 300 . 0 shared equivalent positions with P . westermani TK and S . mansoni TK in D1 and D2 domains . Moreover , introns in C . sinensis TK shared additional intron positions with mollusk AKs at 97 . 1 , 243 . 2 , 300 . 0 , and 366 . 1 ( Fig . 2 ) . The presence of these conserved intron positions supported the phylogenetic relationship of platyhelminth PKs with molluscan AKs and provided evidence for a distinct lineage of taurocyamine kinase from trematodes . Recombinant whole and truncated variants of CsTK were successfully expressed as MBP-fusion proteins . Each set of recombinant fusion proteins was purified by affinity column chromatography to homogeneity , and appeared as a single band in SDS-PAGE CsTK D1 and D2 ( truncated domain+MBP ) about 80 kDa and whole CsTK 120 kDa ( Fig . 3 ) . Enzymatic activity of the recombinant CsTK was measured by NADH-linked assay for substrates taurocyamine , glycocyamine , creatine , D-arginine , and L-arginine . Whole and truncated D1 and D2 of CsTK showed enzyme activity 0 . 84–1 . 36 µmol/min·mg protein toward taurocyamine ( Table 2 ) . Kinetic parameter of whole CsTK KmTc 0 . 49 mM was higher than that of individual domain D1 or D2 , 0 . 35 and 0 . 48 mM each , indicating that whole CsTK had lower affinity for substrate taurocyamine . CsTK D1 had stronger affinity for ATP as its KmATP , 0 . 46 mM , was lower when compared to that of CsTK D2 , 0 . 75 mM , and of whole CsTK , 0 . 78 mM . Kcat value of CsTK D1 , 22 . 59 s−1 , was higher than that of D2 , 4 . 50 s−1 , and of whole CsTK , 17 . 78 s−1 . Similar results were also recorded for Vmax and kcat/KmTc , reflecting that CsTK D1 had more efficient catalytic activity than D2 domain or whole CsTK did . This enzymatic feature was different from that of P . westermani TK , of which whole P . westermani TK is catalytically more efficient than either of the truncated individual domains , D1 or D2 ( Table 3 ) . As appeared in the alignment of multiple PK polypeptide sequences ( Fig . S2 ) , guanidino specificity ( GS ) region of CsTK had 2–3 less amino acids than that of the other known AKs , which is common feature of trematode TKs . To characterize the substrate recognition system in GS region of C . sinensis TK , residue substitution was introduced in CsTK ( Fig . S2 ) . Parameters of affinity , activity , and catalytic efficiency are presented in Table 4 . For CsTK D1 , mutations in GS region decreased its affinity for taurocyamine as evidenced by the increase of KmTc values . Most significant decrease was observed in the Y84R mutant showing no enzymatic activity . Substitutions of equivalent residues in truncated D2 domain ( Y84A , Y84R ) , whole CsTK ( Y84A , Y84R in D1 and Y87A , Y87R in D2 ) resulted in the loss of detectable enzyme activity . Substitution of tyrosine in GS region might have affected stabilization of the closed structure of PK , suggesting that this amino acid plays an important role in taurocyamine binding . Another substitutions in D1 ( I58A , R60A ) , and whole CsTK ( I58A , R60A in D1 and H61A , I63A in D2 ) also decreased enzyme activity . However , mutation of equivalent position ( H61A and I63A mutants ) in truncated D2 increased catalytic efficiency , as evidenced by higher values of kcat , kcat/KmTc and Vmax . To compare relative gene expression level between developmental stages by using ΔΔCT equation , three reference genes were employed such as of β-actin , phosphoglycerate kinase , and calcyphosine . CsTK D1D2 mRNA level was 1 . 2-fold higher in the metacercariae than in the adults of C . sinensis ( Fig . 4 ) The recombinant Cs28GST-CsTK D1 fusion protein was produced as a major component and soluble form in the E . coli host . The recombinant CsTK D1 protein was cleaved off efficiently from the tag , Cs28GST , by thrombin treatment on bead . The cleaved CsTK D1 protein was eluted at high concentration and purity with a molecular mass of 42 kDa in SDS-PAGE gel ( Fig . 5 ) . This CsTK D1 was used for downstream experiments such as immune serum production and antigenicity tests on ELISA . Anti-CsTK D1 mouse immune sera reacted strongly to recombinant CsTKD1 protein . These mouse sera reacted to and detected native CsTKD1D2 protein from adult C . sinensis soluble extract . The native CsTKD1D2 protein was revealed as a distinctive major band of 80 kDa protein . Anti-CsTK D1 antibody was able to detect CsTK D1D2 as well as CsTK D1 , since CsTK D1 and CsTK D2 are fused in tandem ( Fig . 6 ) . Anti-CsTK D1 mouse immune sera were used for localization of CsTK D1D2 in the adult C . sinensis by immunohistochemical staining . The CsTKD1D2 was localized in tegument , oral and ventral suckers , testes , seminal vesicle and sperms , intra-uterine eggs and intestinal contents in adult C . sinensis . Dregs between C . sinensis were stained with strong positive color ( Fig . 7 ) . Acini of biliary epithelium had contents reacting positively to the immune sera . Recombinant CsTK D1 protein was evaluated for serodiagnostic antigen toward IgG antibody in C . sinensis-infected human sera by ELISA . A positive cut-off value , derived from the normal control sera , was set at A490 = 0 . 21 . The CsTK D1 protein gave 29 . 7% positive rate from sera of clonorchiasis patients . On the other hand , this positive rate was 65 . 0% from sera of opisthorchiasis patients , 14 . 3% of paragonimiasis patients , 0% from cysticercosis , 35 . 7% from sparganosis and 7 . 1% from the normal control sera ( Fig . 8 , Table S3 ) .
Taurocyamine kinase ( TK ) is a member of the phosphagen kinase family , which was first isolated from the body wall muscle of polychaete lugworm , Arenicola marina [25] . Two types of TK , cytoplasmic TK and mitochondrial ( Mi ) TK were previously purified from Arenicola brasiliensis [26] and from deep-sea Riftia pachyptila [27] . As such , TKs were believed to be restricted to certain marine annelids [26] . However , P . westermani and S . japonicum TK showed activity only towards taurocyamine [15] , [28] . Additionally , a two-domain phosphagen kinase with enzyme activity to taurocyamine was also reported from S . mansoni [29] . Thus , it was found that TK was not exclusive to marine annelids . In the present study , C . sinensis TK also showed exclusive activity towards taurocyamine . In this study , we identified a cDNA clone encoding a polypeptide of 717 amino acids . The deduced amino acid sequence showed a high similarity to TKs previously reported from P . westermani and other helminthic parasites . It should be noted that C . sinensis TK had higher sequence identity with molluscan AKs than with other annelid TKs . The neighbor-joining tree revealed two major groups: CK and AK isoenzyme groups . Both domains of C . sinensis TK fell in AK cluster , together with P . westermani TK , S . mansoni TK , molluscan AKs and sipunculid HTK . Recombinant C . sinensis TK showed high enzymatic activity toward the taurocyamine substrate . With these results , it was identified for the first time that the C . sinensis cDNA encoded a phophagen kinase . C . sinensis TK was considered to be a cytoplasmic TK , since it did not have the N-terminal signal peptide of 40 residues which were the probable mitochondrial targeting sequence present in characterized mitochondrial CKs [30] . Annelid PKs had considerable catalytic efficiencies for the guanidino substrates , glycocyamine and lombricine , in addition to its original target substrate , taurocyamine [26] , [27] . This was in contrast to C . sinensis TK , P . westermani TK , and S . mansoni TK , all of which showed exclusive enzyme activity for taurocyamine . Low degree of substrate specificity in annelid enzymes supported flexibility in substrate recognition , which might have been a driving force in the evolution of phaphagen kinases . Eisenia TK had obtained remarkable diversity supported by the mutation K95Y LK , dramatically changing guanidino substrate specificity from lombricine to taurocyamine [31] . Amino acid sequences of GS region of mitochondrial TK were quite different from cytoplasmic TK [32] , which reflects independent evolutionary processes . Sequential difference of these two enzymes did translate to differences in enzymatic activity and substrate specificity toward the guanidino substrates , taurocyamine , lombricine , glycocyamine , and arginine [26] , [27] . Number of amino acids in GS region of C . sinensis TK , P . westermani TK , and S . mansoni TK was smaller than those of annelid TKs ( Fig . S2 ) . Cytoplasmic TKs of A . brasiliensis and R . pachyptila were missing five residues [26] , [27] , and this might affect the differences in guanidino substrate specificity . The GS region was described as a possible candidate for the guanidine recognition site , and a number of amino acid deletion in this region correlated with the size of phosphagen substrates utilized [20] . Amino acid substitutitions in the GS region resulted in a significant decrease of enzyme activity for arginine [11] , [33] . However , functional properties and substrate binding mechanism of TK is not well defined yet . To characterize substrate recognition property of C . sinensis TK , amino acid substitutions were put in the GS region of truncated D1 and D2 domains , and in the D1 or D2 domain of whole enzyme . The residue 140 ( Fig . S2 ) was conserved across phosphagen kinases: even Tyr was replaced by Arg in CK , Ile in GK , His in TK , and Lys in LK [20] , [26] . This residue was not directly associated with substrate binding , as revealed by the CK and AK crystal structures . However , its position was close to the guanidine substrate-binding site [33] , [34] , [35] , and functionally , this residue determines guanidino substrate specificity [31] , [36] . The residue 140 was replaced by His and Lys in cytoplasmic and mitochondrial TKs , respectively in nature . The equivalent residues in other phosphagen kinases , which correspond with the residue 95 in Danio CK , had the roles of distinguishing guanidino substrates and organizing the hydrogen-bond network around this position , which offered an appropriate active center for high catalytic turnover . The mode of development of this network appeared to be unique in each phosphagen kinase , reflecting the evolution of each enzyme [27] . The equivalent residue was replaced by Tyr at position 84 and 87 in C . sinensis TK D1 and D2 domains , and similarly in TK of P . westermani and S . mansoni . A substitution of Y84R in C . sinensis TK D1 caused loss of affinity for taurocyamine . However , Y84R in D1 domain of whole CsTK still retained low enzyme activity , suggesting that two-domain structure of TK have synergistic role for enzymatic activity . Moreover , substitutions of Y84A in TK D1 , Y87R and Y87A in D2 , Y84A and Y84R in D1 of whole TK , Y87A and Y87R in D2 of whole TK decreased affinity for taurocyamine . It is suggested that Tyr84 in D1 domain was not a key residue for substrate recognition since replacement of this amino acid residue did not alter substrate specificity from taurocyamine to glycocyamine , but the residue is still important for taurocyamine binding . Amino acid residues in GS region were conserved among phosphagen kinase subgroups [37] . R58 and I60 were in GS region of C . sinensis TK , which were key residues for catalytic activity or substrate binding among other PKs . I60 of C . sinensis TK was replaced by Val in all CKs , which showed low enzymatic activity for glycocyamine . The fact that the equivalent amino position of Arenicola MiTK V71A mutant revealed strong activity for glycocyamine suggested that the Val71 in CK might minimize its kinase activity for this substrate [38] . Substitution of H61A or I63A in CsTK D2 enhanced the enzyme's affinity for taurocyamine by about 2-fold increase . These results suggested that substitution of these two residues in the GS region affected stability of the closed structure , and that these amino acids were important for taurocyamine binding . High catalytic efficiency and strong affinity of C . sinensis TK toward the substrate suggested that it had a significant role in the energy metabolism for the parasite organism . A previous study [39] reported that major AK and CK clusters could be categorized . Through phylogenetic analysis , a broad spectrum of animal PKs grouped into either an annelid-specific phosphagen kinase cluster ( lombricine kinase , glycocyamine kinase , and cytoplasmic and mitochondrial TKs ) or a sister-group of CKs from vertebrate and invertebrate animals . It appeared that the annelid-specific phosphagen kinases , including cytoplasmic and mitochondrial TKs , evolved from a CK-like ancestor ( s ) early in the divergence of the protostome metazoans . Furthermore , these results suggested that the cytoplasmic and mitochondrial isoforms of TK evolved independently [27] . It was proposed from tree topology and sequence identities that C . sinensis TK was in the AK subcluster with P . westermani TK , S . mansoni TK , molluscan AKs , and sipunculid HTK . Genomic organization of C . sinensis TK DNA , 13 exons interrupted by 12 introns , was remarkably conserved with those of P . westermani TK and S . mansoni TK . C . sinensis TK shared more intron positions with molluscan AKs than trematode TKs , and did not share any intron position with taurocyamine kinase from the annelid A . brasiliensis nor with other representative PKs belonging to the CK cluster . This result suggested that C . sinensis TK evolved from AK gene clad and supported the phylogeny or evolution of CsTK as shown in Fig . 1 , which was different from annelid TK evolved from CK group . Phylogenetic and gene structure analyses showed that trematode TKs had evolved from a different lineage of taurocyamine kinase . Close phylogenetic relationship had been reported between flatworms and mollusks as molecular data grouped them together in the Lopotrochozoa [40] . It was hypothesized that through horizontal gene transfer and exon shuffling trematodes acquired arginine kinase from gastropod intermediate hosts , which eventually became the class of annelid taurocyamine kinases . It was hypothesized arginine kinase of the protozoa trypanosoma was a product of horizontal gene transfer from arthropods [41] . It has also been proved that P . westermani TK and other trematode TKs represent a distinct lineage of TKs which evolved from a molluscan AK gene rather than from a CK gene from phylogenetic and gene structure analyses [28] . 13 extron/12 intron of C . sinensis TK had complex gene structures compared with other C . sinensis genes . Phospholipid hydroperoxide glutathione peroxidase ( PHGPx ) and myophilin-like protein of C . sinensis had three and five intron , respectively [42] , [43] . The results provided further insight into the evolution of taurocyamine kinase in the family of phosphagen kinases . In our experiments , CsTK was transcribed in the metacercaria and adults of C . sinensis . Quantitative realtime PCR analysis revealed that the transcription level of CsTK mRNA was lower in the adult stage than in the metacercaria . This was possibly due to that the protein played an important role in growth and development of the juvenile flukes . . Moreover , immunolocalization results showed that CsTK was distributed in the tegument , testes , ventral sucker , and intestinal contents in adult C . sinensis . The extensive distribution and developmental stage-independent expression may imply that CsTK is a multifunctional molecule in the developmental biology of C . sinensis , especially in organogenesis . The tegument is an interface between parasite and its host , which was a dynamic host-interactive layer which played an important role in modulation of the host response and parasite survival [44] , [45] . Moreover , tegument was one of the most active sites of energy metabolism involved with signal transduction , modulation , excretion , and osmoregulation . Previous studies have reported that proteins identified at the tegument of helminthes were suggested as important candidate antigens for drugs , immunological diagnosis , and vaccine [45] , [46] . In this study , anti-CsTK D1 mouse immune sera reacted strongly to recombinant CsTKD1 protein , and these mouse sera reacted to and detected native CsTKD1D2 protein from adult C . sinensis soluble extract . The CsTK in the seminal vesicle , intestine and uterus of C . sinensis is passed out along sperms , eggs and intestinal contents into bile . The bile containing CsTK is percolated in the biliary acni and stagnated as dregs between the flukes in the biliary track , then could be presented as immunogen to the host . Through this way , the CsTK could evoke the host immune system and had decent immunogenicity . Anti-C . sinensis antibody detection by enzyme-linked immunosorbent assay ( ELISA ) has been used for epidemiological surveys of clonorchiasis for convenience and celerity , but an ideal diagnostic and/or treatment-response assay using C . sinensis specific antigen or anti-body subtype could improve diagnostic sensitivity and specificity in the clinical setting . In the present study , the sensitivities of specific IgG detection and cross-reactions were measured . Recombinant CsTK D1 had low specificity and sensitivity toward sera of C . sinensis infected individuals . In conclusion , a novel gene coding of C . sinensis TK was identified from cDNA library for the first time . CsTK gene and gene products were characterized using phylogenetic analysis , gene structure , enzyme activity , mutation , immunogenicity , antigenicity , and immunolocalization . Our current study might enhance the deduction that TK could play an important role in the growth of C . sinensis organism and provides clues for a promising novel candidate drug target in the control of clonorchiasis . | The food-borne clonorchiasis imposes public health problems on inhabitants in endemic areas . Praziquantel has been employed as an efficacious anthelminthic in large-scale campaigns as well as for individual treatment of Clonorchis sinensis human infections . Although praziquantel continues to have good efficacy , new drug development for this parasite has been recognized as a crucial issue to be investigated intensively . Clonorchis sinensis adults generate energy through glycolysis , actively utilizing exogenous glucose , and produce a large amount of eggs each day . Taurocyamine kinase ( CsTK ) is distributed abundantly in the locomotive and reproductive organs , and is an important enzyme in energy generation and homeostasis in adult C . sinensis . Enzymes of the glycolytic pathway are also expressed abundantly in these organs and in tegument , implying these organs play central roles which are essential for survival and reproduction of C . sinensis . The TK enzymes , including CsTK , are found only among invertebrate organisms and have substrate specificity for taurocyamine , which are significantly different from phosphagen kinases of vertebrate animals . With these molecular biological , enzymatic , and evolutionary characteristics , we propose here that CsTK could be a target for development of chemotherapeutic agents against C . sinensis and be a biomolecular model for other human-infecting trematodes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Molecular Cloning and Characterization of Taurocyamine Kinase from Clonorchis sinensis: A Candidate Chemotherapeutic Target |
In this work we develop a microscopic physical model of early evolution where phenotype—organism life expectancy—is directly related to genotype—the stability of its proteins in their native conformations—which can be determined exactly in the model . Simulating the model on a computer , we consistently observe the “Big Bang” scenario whereby exponential population growth ensues as soon as favorable sequence–structure combinations ( precursors of stable proteins ) are discovered . Upon that , random diversity of the structural space abruptly collapses into a small set of preferred proteins . We observe that protein folds remain stable and abundant in the population at timescales much greater than mutation or organism lifetime , and the distribution of the lifetimes of dominant folds in a population approximately follows a power law . The separation of evolutionary timescales between discovery of new folds and generation of new sequences gives rise to emergence of protein families and superfamilies whose sizes are power-law distributed , closely matching the same distributions for real proteins . On the population level we observe emergence of species—subpopulations that carry similar genomes . Further , we present a simple theory that relates stability of evolving proteins to the sizes of emerging genomes . Together , these results provide a microscopic first-principles picture of how first-gene families developed in the course of early evolution .
Molecular biology has collected a wealth of quantitative data on protein sequences and structures revealing complex patterns of the protein universe , such as markedly uneven usage of protein folds and near–scale-free character of protein similarity networks [1−5] . On a much higher level of biological hierarchy , ecology , evolution theory , and population genetics established a framework for studying speciation , population dynamics , and other large-scale biological phenomena [6−8] . While it is widely accepted that gene families and the protein universe emerged during the course of molecular evolution through selection [9−11] , there is a substantial gap in our conceptual and mechanistic understanding of how molecular evolution occurred or what the determinants of selection are . Indeed , evolution , as we understand it , proceeds at the level of organisms and populations but not at the level of genomes . Evolutionary selection at the molecular level occurs due to a relation between genotype and phenotype , although a detailed understanding of this relation and its consequences for molecular evolution remains elusive . A number of phenomenological models ( e . g . , Eigen's quasispecies model ) were developed where fitness of an organism was related to the sequence of its genome [12–14] . A standard definition of fitness in phenomenological models is the growth rate of a population that is higher for the more fit species . However , the relationship between genotype and phenotype in quasispecies ( QS ) and similar population genetics ( PG ) models is purely phenomenological . For example , in single-fitness peak models , one specific genotype is postulated to be most fit , while deviations from it confer selective disadvantage . Despite providing several important insights , these types of approaches lack a fundamental microscopic connection between fitness and quantities of proteins that are both easily justifiable on biological grounds and measureable ( e . g . , structure/stability , function , or regulation ) . Therefore , such models cannot accurately describe molecular evolution of proteins . On the other hand , a number of models were proposed that focus on emergence and evolution of sequences of model proteins and RNA under direct pressure on their molecular properties such as stability [11 , 15−18] , folding kinetics [19 , 20] , and mutational robustness [21] . Schuster and Stadler [22] first studied the evolution of biological macromolecules , RNA , in the context of population dynamics . Later , in a series of papers , Taverna and Goldstein [10] used an Eigen model of reaction flow to grow populations of proteins modeled as 2-D 25 mers . These authors showed that when the requirement to exceed certain stability thresholds is imposed , the resulting distribution of structures in the evolved population appears highly skewed toward more designable structures [23] and more robust ( i . e . , less susceptible to mutations ) proteins [24] . One of the most surprising features of the protein universe is an uneven and broad distribution of proteins over folds , families , and superfamilies . While this fact had been noted by many researchers long ago [1 , 4 , 25 , 26] , the quantitative descriptions of these distributions began to emerge only recently . Huynen and van Nimwegen found that sizes of paralogous gene families follow a power-law distribution [3] . Gerstein and coworkers [5] observed a power-law distribution of frequencies of several other properties of gene families as defined in the Structural Classification of Proteins ( SCOP ) database [27] . Dokholyan et al . [2] studied a network of structural similarities between protein domains ( called the protein domain universe graph , or the PDUG ) and found that distribution of connectivities within the PDUG follows a power law ( within a limited range of connectivity variance ) , making it a finite size counterpart of a scale-free network . This is in striking variance with an expectation from random distribution of folds , which would result in an ( approximately ) Gaussian distribution of connectivities of the PDUG . The ubiquitous nature of power-law dependencies of many characteristics of gene families and of the protein universe may suggest their possible common origin from the fundamental evolutionary dynamics and/or physics of proteins . Huynen and van Nimwegen [3] , Gerstein and coworkers [5] , and Koonin and coworkers [9 , 28 , 29] proposed dynamical models ( the version proposed in [28] is called the birth , death , and innovation model [BDIM] ) on the basis of gene duplication as a main mechanism of creation of novel types . Such models , while providing power-law distribution of family sizes in some asymptotic cases , are sometimes based on assumptions that call into question their generality . In particular , as pointed out by Koonin and coworkers , for gene duplication dynamic models to provide nontrivial power-law distributions of paralogous family sizes , one has to assume that the probability of gene duplication per gene depends in a certain regular way on the size of an already existing gene family . Further , even under this assumption , the power-law distribution in the birth , death , and innovation ( BDIM ) model arises only asymptotically in a steady state of evolutionary dynamics [29] . In contrast , the duplication and divergence phenomenological model of Dokholyan et al . [2] did not use such dramatic assumptions . However , this model is limited to an explanation of the scale-free nature of the PDUG , and it does not provide any insight as to the nature of the power-law distribution of gene family sizes . In protein sequence space , a similar approach has been employed by Qian et al . [5] . However , models like the ones proposed in [2 , 3 , 5 , 28] and other works are purely phenomenological in nature , whereby proteins are presented as abstract nodes and where sequence–structure relationships are not considered . Here , we present a microscopic physics-based model of early biological evolution ( Figure 1 ) with a realistic generic population dynamics scenario where fitness ( i . e . , life expectancy ) of an organism is related to a simple necessary requirement of functionality of its proteins—their ability to be in native conformations . Since the latter can be estimated exactly in our model from sequences of evolving genomes , the proposed model provides a rigorous , microscopic connection between molecular evolution and population dynamics . We demonstrate that the model indeed bridges multiple evolutionary timescales , thus providing an insight into how selection of a best-fit phenotype results in molecular selection of proteins and formation of stable long-lasting protein folds and superfamilies . Furthermore , the coupling of molecular and organismal/populational scales results in the emergence of species—subpopulations of evolved organisms whose genomes are similar within their groups and dissimilar between groups . The resulting protein universe features power-law distribution for gene family and superfamily sizes closely matching real ones . The proposed model can be viewed as a first step toward a microscopic first-principles description of emergence and evolution of the protein universe .
Our evolution dynamics runs start from an initial population of 100 organisms , each having the same one primordial gene in their genomes . Initial gene sequence is random . Runs proceed according to evolutionary dynamics rules as described in the Methods section ( see also Figure 1 ) . The life expectancy of an organism is directly related to the stability of its proteins as explained in the Methods section; briefly , the death rate d is inversely related to protein stability Pnat , This equation expresses a postulate that all genes of early organisms were essential at the given time; no a priori assumptions about the number of these genes are made . We found that out of 50 simulation runs starting with different starting sequences , 27 runs successfully resulted in a steady exponential growth of the population , whereas in 23 runs the population has quickly gone extinct . A typical behavior of the population growth and protein structure dynamics in a successful evolution run is shown in Figure 2 . After a period of “hesitation” lasting for about 100 time steps , a steady exponential growth of the population sets in ( Figure 2B ) . In Figure 2C , we present the mean native state probability Pnat of all proteins present in the population at a given time . Owing to mutations and selection , Pnat steadily increases with time and dramatically exceeds the mean Pnat for random sequences , Pnatrand = 0 . 23 . In contrast to earlier models [10] , the selection pressure is applied to whole organisms rather than to individual protein molecules . The genotype–phenotype feedback , which we model by Equation 1 ( see also Methods ) , transfers the pressure from organisms to individual proteins to gene sequences . Figure 2B and 2C show that our selection mechanism works and results in the discovery of stable proteins due to evolutionary pressure . Using our model , we can follow each structure in the population . In Figure 2A , color hue encodes the number of genes in the population corresponding to each of the 103 , 346 lattice structures ( ordinate ) as a function of time ( abscissa ) . Structures marked in green are the most abundant in population at a given time , while black background corresponds to structures not found in any of the evolving organisms . The most important feature of this plot is the appearance of specific structures that correspond to highly abundant proteins comprising a significant fraction of the gene repertoire of the population . In what follows we will call them dominant protein structures ( DPSs ) . Such proteins visually appear as bright lines on Figure 2A . What is the genesis of DPSs , and how is their appearance related to population growth or decay ? To answer this question , let us track the development of the population of structures in time by comparing the structure repertoire , the population size , and Pnat plots . At t = 0 , the proteome consists of a single sequence–structure combination ( a single line on the structural repertoire plot ) , which corresponds to all individuals in the initial population having that structure in the genome . Over time , random mutations diverge sequences in each organism such that the dominance of a single structure is lost . This can be seen as a smeared line on the structural repertoire plot , as shown in Figure 2A , t < 100 . However , at a certain point , very favorable sequence–structure combinations are discovered . They represent DPSs whose incorporation into the genome leads to an abrupt increase of Pnat and explosive exponential growth of the population through increase in fitness . Shortly after the discovery of that DPS , the diversity of the structural space abruptly collapsed , as most of the organisms converge toward the newly discovered DPS . Such a dramatic event , discovery of a limited number of dominant proteins and ensuing exponential growth of the population , can be called the “Biological Big Bang , ” following a loose analogy with astrophysics . As seen on Figure 2A , the emerged dominant folds are very persistent in time . Nevertheless , fold discovery can occur at later stages of evolution . For example , in this particular simulation , at t ≈ 1 , 300 , new folds were discovered ( white arrow in Figure 2A ) , they become new DPSs , and the initial DPSs are completely replaced by the new folds by t ≈ 1 , 600 . This switchover , accompanied by a marked increase of Pnat , is a clear manifestation of punctuated discoveries of new folds , coupled with selection at the organismal level . Even though the number of organisms increases exponentially , the number of genes in each genome increases very slowly ( and stabilizes after the discovery of DPSs ) ( Figure S1 , red curve ) . Indeed , large genomes are not very advantageous in our model , as mutations occur in all of the genes , whereas the death rate is controlled by the gene with the lowest Pnat . Thus , it is only this gene that bears the brunt of selective pressure . Therefore , the rest of the genome accumulates mutations and is more prone to deleterious mutations . Unless all of the genes are very carefully selected ( or formation of pseudogenes is allowed ) , a larger number of genes means that there is a substantial probability that a point mutation will result in a sequence–structure combination with a very low Pnat , immediately killing the organism . The observed slow increase of the size of the genome reflects the subtle balance between the selection pressure and gene duplication and is analyzed in more detail below . Remarkably , the average number of genes in the surviving organisms decreases with increasing mutation rates ( Figure S7 ) . Indeed , if every protein is essential , then the probability of organism death due to a deleterious mutation is lower in organisms with shorter genomes . This result allows direct experimental verification and clearly sets the current model apart from the previous sequence evolution simulations [10] , which focused on the properties of individual proteins . Figure S2 shows the structural repertoire and population size of an unsuccessful simulation run , where the population quickly became extinct . This simulation did not result in a discovery of a stable fold , and the structural space was evenly filled until the extinction of the population . We found ( unpublished data ) that the choice of starting sequence does not have any significance in determining whether a particular simulation run will result in exponential growth or extinction . Furthermore , in the case of most unsuccessful evolution runs , the genome size rapidly increases with time ( Figure S1 , blue curve ) , decreasing the average evolutionary pressure per gene and making the discovery of DPSs less likely . Based on these observations , we conjecture that biological evolution , exponential population growth , and existence of stable genomes are possible only after the discovery of a narrow set of specific protein structures . To quantify the persistence of the DPS during evolution , we calculated the distribution of DPS lifetimes , defined as the timespan during which a structure comprises more than 20% of the genes present in the most populated structure , i . e . , the time between emergence of a DPS and its extinction in the population ( see Figure 3A ) . We consider only DPSs that already completed their “lifecycle , ” i . e . , the DPSs that emerged and went extinct over the time of an evolutionary simulation . It is clear from Figure 3B that the lifetime of a DPS is much greater than that of an organism or the average time between successive mutations . Moreover , the distribution of DPS lifetimes clearly follows a power-law–like distribution . The long nonexponential tail of the distribution demonstrates that some protein folds are extremely resistant to mutations and may persist over thousands of generations . Over such a long time , diverse protein ( super ) families are formed around the DPS folds . This is illustrated on Figure 4A , which shows the distribution of sizes of evolved families and superfamilies of proteins . To avoid confusion , we note that families and superfamilies here are defined not necessarily as sets of paralogous sequences but in the same way as they are defined in SCOP ( Structural Classification of Proteins ) [27]: protein families are defined as sets of all ( not necessarily belonging to the same organism ) homologous sequences that fold into a given domain structure , and superfamilies are defined as all monophyletic sets of sequences whose homology may not be detectable by sequence comparison methods , but which nevertheless fold into structurally similar domains . The statistics of protein families is dominated by orthologous genes in contrast to paralogous families studied in [5 , 9] . As shown in Figure 4A , both family and superfamily size distributions of evolved proteins follow almost perfect power laws with power-law exponent being greater for superfamilies ( −2 . 92 ) than that for families ( −1 . 77 ) . To compare this result with real proteins , we plotted the distribution of family and superfamily sizes of real proteins . As a measure of family sizes , here we estimated the number of homologous sequences that fold into a given domain ( see Methods ) , and as a proxy for superfamily size we estimated the number of functions performed by each domain . Clearly the distributions in Figure 4B follow power-law statistics , and as in the model , the exponent for the superfamily distribution ( −2 . 2 ) is greater than that for families ( −1 . 6 ) . Quantitatively , the slopes of the model and real distributions are similar . As mentioned above , the genomes of model organisms from exponentially growing populations are rather short ( about three genes ) in contrast to the extinct populations , where uncontrolled gene duplication is observed ( Figure S1 ) . To better understand this phenomenon , one should consider the distribution of protein stabilities Pnat before and after a round of mutations . Suppose each genome has N genes , and the fitness ( inversely related to the probability of death of an organism ) of the genome is defined by . At each time step , each gene in the genome has an equal probability of mutation . For simplicity , we assume that the distribution of stability Pnat of a lattice protein after a point mutation ( i ) does not depend on the stability before the mutation and ( ii ) is uniformly distributed between 0 and 1 . Such a crude approximation works surprisingly well for lattice proteins ( see Figure S3 ) and allows for an analytic calculation of the average genome fitness f′ after a round of point mutations ( see Methods ) . The average fitness after a point mutation depends on the number of genes N and original fitness f , and since larger genomes accumulate more mutations , they are more prone to a decrease in fitness after the mutation . In particular , if the fitness of an original genome was f , then after one round of point mutations the average new fitness f′ will be no less than f only if the genome is sufficiently short , namely if In other words , on average , organisms with more than N genes will decrease their fitness after a point mutation and will be eventually washed out from the population . Thus , Equation 2 establishes an upper boundary on the number of genes per organism at a given level of stability in the weakest link model of evolution with lattice proteins . In Figure 5 , we plotted the predicted boundary from Equation 2 and the results of 50 simulation runs , where we show the scatter between the average number of genes per organism N in a population and average stability Pnat of proteins in a population at every time step during each of the runs . As predicted by Equation 2 , only organisms with sufficiently short genomes survive at a given level of protein stability; the higher the stability , the lower the maximum possible number of genes per organism . Indeed , in a genome consisting of very stable proteins , most of the mutations are deleterious and confer a lethal phenotype in our evolutionary model . In this particular model , no more than three genes can be present in a genome at very high values of Pnat . It should be noted that this consideration applies only to the equilibrium size of the genome at large evolutionary times and does not describe the entire course of its evolution in time ( Figure S1 ) . A more realistic distribution of changes of Pnat upon mutation and a more detailed consideration of the effect of mutations on the fate of organisms result in realistic estimates of genome sizes for real organisms ( K . B . Zeldovich , P . Chen , and E . I . Shakhnovich , unpublished data ) . The exact nature of the model gives us direct access to the genomes of all evolved organisms , and an interesting question is whether all evolved genomes are similar ( monoclonal or single-species population ) , and , if not , can they be clustered into distinct clonal lines or species . It turns out that the number of DPSs in the evolved population is a very good indicator of species formation . In many cases , there is only one DPS in the evolved population . Then , the genomes of all organisms are similar , and the population is monoclonal . A more interesting case is presented in Figure 6A , where two different DPSs corresponding to structures 10 , 107 ( “A” ) and 15 , 550 ( “B” ) ( in our arbitrary numbering ) have evolved . Are these structures randomly distributed between organisms , or are there groups of organisms preferentially using structure A , but not B , and vice versa ? In the latter case , one could argue that two clonal lines or species have evolved , as each of the groups will have its own and distinct set of protein structures , and correspondingly , sequences . It turns out that in 1 , 536 organisms in the population , at least one gene encodes for the structure A , in 2 , 767 organisms , at least one gene encodes for B , but there are no organisms that include both A and B in their genomes . A total of 697 organisms have neither A nor B in their genomes . Therefore , organisms having the DPS of fold A in their genomes are very distinct from the organisms with the fold B . This difference is further illustrated in Figure 6B , where we plotted the histograms of pairwise Hamming distances between the sequences encoding for structures A and B . The black curve represents the distribution of all pairwise Hamming distances between the sequences encoding for structure A; the red curve corresponds to structure B . Both curves are shifted toward lower values of the Hamming distance , illustrating a certain degree of similarity of multiple sequences encoding for the same structure . However , the Hamming distance between the sequences encoding for A and sequences encoding for B ( green curve ) is much larger and is very close to that of purely random sequences . Therefore , in sequence space , we can identify two groups of sequences that are similar within each group and dissimilar across the groups . Thus , the genomes of our model organisms can be classified according to their membership in the two well-defined groups of sequences , which is our model analogue of genome-based taxonomy . It is interesting to note that since our model is purely divergent and lateral gene transfer is not allowed , the evolving lines ( or species ) of organisms remain isolated , each evolving around its own DPS . In the discussion of structures of evolved proteins , an important global characteristic of the set of evolved proteins is the PDUG [2] . In this graph , nonhomologous proteins are linked by an edge if their structural similarity score exceeds a certain threshold . It is known that in natural proteins [2] , the size of the largest cluster ( giant component ) of the PDUG abruptly shrinks at some value of the threshold , similar to the percolation transition . The degree distribution of the graph , ( i . e . , the probability p ( k ) that a protein has k structurally similar neighbors ) , is a power law at the transition point . The scale-free character of this graph is believed to be a consequence of divergent evolution [2 , 30 , 31] as suggested by simple phenomenological “duplication and divergence” models [2] . Therefore , it is important to test whether our model can reproduce the global features of the natural protein universe that are manifest in the unusual properties of the PDUG . Here we plot the PDUG of evolved proteins using Q score , the number of common contacts between a pair of proteins , as a structural similarity measure [31] . The degree distribution of the evolved PDUG at similarity threshold Q = 17 ( the midtransition in giant component of the evolved graph , see Figure S4 ) is shown in Figure 7 . The degree distribution plot clearly shows that the graph consists of two components , a scale-free–like component at lower k and a small but very highly connected component at high k . As a control , we computed p ( k ) for a divergent model without the genotype–phenotype feedback , with the fixed death rate of organisms equal to the death rate in the exponential growth regime of the evolution model . The degree distribution of the PDUG obtained in this control simulation where death rate is constant and independent of the stability of evolving proteins is shown in Figure S5 . The control graph is weakly connected , indicating randomness of the discovered structures . The degree distribution of the control graph is well approximated by a Gaussian distribution , in contrast to the one obtained from evolution simulation or the real PDUG [2] . Therefore , evolutionary selection has a profound effect on the global structure of the evolved protein universe . In the model , the structural similarity graph ( PDUG ) splits into a scale-free–like part and a highly connected part , corresponding to the DPS , populated by many dissimilar sequences . Our simulations predict that new folds emerge as offspring of DPSs , and in this picture the DPSs serve as prototypes of the first ancient folds . Following this logic , one should expect that ancient protein folds , being closer to prototypical DPSs , should be highly clustered and more connected than later diverged folds . To test this prediction , we analyzed the subgraph of the PDUG corresponding to the last universal common ancestor ( LUCA ) domains [32] . There are 915 LUCA domains . We compared the connectivity and clustering coefficient in the PDUG subgraph corresponding to LUCA domains with distributions for the same characteristics for 915 randomly selected domains as a control . The null hypothesis is that a random subset of protein domains has connectivity and clustering coefficients similar to that of the LUCA domains . In Figure S6 we present the histograms of mean connectivity k ( average degree of the node ) and clustering coefficient C found in 20 , 000 subsets of n = 915 randomly chosen protein domains ( out of a total of 3 , 300 distance matrix alignment [DALI] domains constituting the PDUG , see [2] ) . For random subsets of 915 domains from the PDUG , k = 2 . 91 and C = 0 . 197 , while the average values of the same parameters for the 915 LUCA protein domains are: k = 4 . 61 and C = 0 . 267 . The values of k and C for the LUCA domains are statistically much greater than corresponding values for the random subsets ( see Figure S6 ) , yielding extremely low p-values ( p < 10−10 ) . LUCA domains are connected and clustered just as a random subset of the PDUG ( assuming Gaussian distributions of mean connectivities and clustering coefficients for random subsets of the PDUG in Figure S6 ) . This proves that LUCA domains are statistically more connected and clustered than an equivalent set of random protein domains as predicted from our simulations . In Figure 8 , we summarize the divergent evolution scenario as observed in our model . Divergence and selection lead to the infrequent discovery of new protein folds ( dashed circles ) . Within these folds , mutations result in the formation of protein ( super ) families . The size of protein families steadily increases with time , so older families are generally larger . However , fold formation can occur at any time , branching off any family , so the newly formed families will be necessarily small . At the same time , the structures corresponding to superfamilies are all pairwise similar to each other , and for that reason they are highly clustered in the PDUG . Therefore , at any moment , the snapshot of the evolving protein universe will comprise tightly clustered families of all sizes . This picture of protein families that are “tightly knit” within each fold leads to a prediction of a peculiar property of the PDUG: that each node , ( i . e . , protein domain ) , with connectivity k , is primarily connected with nodes of similar connectivity , i . e . , members of its own fold family . To test this prediction , we follow the approach proposed by Maslov and Sneppen [33] . Connectivity correlation in the PDUG is defined as the probability P ( k1 , k2 ) , that two evolved proteins that have k1 and k2 structural neighbors are structurally similar to each other , i . e . , are themselves connected in the PDUG . To normalize P ( k1 , k2 ) , we created 1 , 000 realizations of the rewired graph where each node has exactly the same connectivity as in the original graph of evolved structures , but with randomly reshuffled links to other nodes . The rewired graphs allow us to calculate the average value Pr ( k1 , k2 ) and the standard deviation σr ( k1 , k2 ) of the probability that nodes with connectivities k1 and k2 are connected in a particular network . In Figure 9A , we present the Z score for connectivity correlations Z ( k1 , k2 ) = ( P ( k1 , k2 ) − Pr ( k1 , k2 ) ) /σr ( k1 , k2 ) for the natural PDUG . It follows from this plot that in PDUG , nodes of similar degree tend to be connected to each other: high values of Z ( k1 , k2 ) ( red ) are grouped along the diagonal k1 = k2 . While at low k , this property is simply a consequence of the transitivity of the measure of structural similarity ( if structure A is similar to B , and B is similar to C , then C must be similar to A ) , it is highly nontrivial to observe this property for highly connected nodes . The pattern of connectivity correlations where similarly connected nodes tend to be connected to each other is very different from the one found in protein–protein interaction , communication , and social networks , where low-connected nodes tend to be connected with highly connected hubs , but not to each other [33] . As seen from Figure 9B , our simple evolution model perfectly reproduces this unusual pattern of connectivity correlations . The reason for such unusual property of connectivity correlation is in the punctuated character of fold discovery and evolution both in the model and in real PDUG .
Here , we introduced a microscopic physics-based model of early biological evolution , which directly relates evolving protein sequences and structures to the life expectancy of the organism . We used a simple physical model of protein thermodynamics and a simple Malthusian model of the population dynamics . The main assumption of our minimalistic model is that the necessary condition of survival of a living organism is that its proteins adopt their native conformations . Therefore , the death rate of the organisms decreases when their proteins become more stable against thermal denaturation or unfolding . In other words , we assume that all genes of our model organism are essential . Biological function is not explicitly present in the model , but protein stability is the necessary condition for its evolution . Genes in our model have high mutation rates , conducive to rapid innovation . As such , our model can be directly applicable to ( and can be experimentally tested on ) the evolution of RNA viruses , which often encode for a handful of proteins , all of which are essential for the virus . The absence of an error correction mechanism results in very high mutation rates and heterogeneous , quasispecies-like populations of RNA viruses [34 , 35] similar to what is found in this model . Rapid evolution makes RNA viruses an ideal system for experimental studies along the lines of our model , where the simulation algorithm propagates model organisms almost like an infected host cell produces new viral particles . The low number of genes ( three to ten proteins per genome depending on simulations conditions , see Figure 5 ) observed in our model is in part related to the extremely high mutation rates , about six mutations per genome per replication . In modern life , such a high rate is observed only in populations of RNA viruses that lack the error correction mechanism . Remarkably , the genomes of RNA viruses are rather short and normally encode for fewer than ten proteins . More complex DNA-based viruses and all cellular organisms invariably possess much lower mutation rates due to error correction and , correspondingly , longer genomes [36] , in qualitative agreement with our model . Thus , our model suggests that protein stability requirements , together with mutation rates , play a crucial role in determining the size of the genomes of surviving organisms . There is a common belief that the experimentally observed moderate stability of natural proteins is a result of positive selection for function . However , no experimental proof for this conjecture is available . Rather , a circular argument that natural proteins are not extremely stable is offered to support this claim [37] . On the contrary , a recent study demonstrated that the higher the stability of a protein , the more likely that it confers selective advantage to the protein by making it more evolvable , by enhancing its ability to tolerate more mutations and as a result evolve a new function [38] . A more plausible explanation of moderate stability of natural proteins is that it is a direct result of a tradeoff between stability in the native conformation and entropy in sequence space , which opposes an evolutionary optimization beyond necessary levels [39] . We observe exactly this phenomenon in our model: while organisms with more stable proteins have selective advantage , the opposing factor—enormity of search in sequence/structure space—results in a compromise level of stability that corresponds to stable but not overstabilized proteins ( see Figure 2C ) . By not “overstabilized” we mean here that for the same structure , standard sequence design methods [40 , 41] can provide sequences with Pnat values that are much closer to 1 than observed in evolved model proteins ( unpublished data ) . Unlike in many previous attempts , our model explicitly describes the interplay of the evolution of individual genes and that of genomes ( organisms ) as a whole , since the death of an organism leads to a complete loss of its genome . The model gives important insights into the interplay between molecular evolution , protein fold evolution , and population dynamics . In combination with selection pressure , random diffusion in sequence and structure spaces eventually leads to the discovery of specific structures , DPSs , that are resistant to mutations and form very evolvable proteins . This , in turn , immediately leads to the “Big Bang” event whereby discovery of viable proteins is coupled to exponential population growth , as mutations are no longer a big threat to viability . The DPSs persist over many generations and may be infrequently replaced or augmented by other even more favorable structures , in a process similar to punctuated evolution . The remarkable separation of timescales between frequent mutations and rare DPS formation allows for the formation of the protein families and superfamilies . Our model suggests that the DPSs may be superseded by more advantageous folds during evolution . A similar domain-loss phenomenon has been discussed in [42] in the context of structure-based prokaryotic phylogenies . The model and simulations presented here provide a quantitative first-principles description of evolution of the universe of protein families . Despite the simplicity of the structural model of proteins and the phenotype–genotype relation invoked , it is able to quantitatively reproduce the power-law distributions that are observed in the natural protein universe . Earlier phenomenological models reproduced some aspects of power-law behavior , always at the expense of invoking dramatic assumptions about the dependence of the rates of gene duplication on the sizes of already existing gene families . Here no such assumptions are made , as the model is fully microscopic in nature . Furthermore , our simulations are capable of reproducing not only “power-law”–like behavior but also marked deviation from it . Indeed , as seen on Figure 4B , there is an inflection point in the distribution of family sizes ( blue curve ) where the apparent slope changes . A similar inflection point was found in a recent clustering analysis of more than 7 million global ocean sampling sequences [43] . Strikingly , the distribution of family sizes of evolved proteins ( Figure 4A ) features a noticeable inflection as well . It is not clear whether phenomenological duplication-growth models are capable of reproducing such fine details of the family size distribution . The most intriguing ( and relevant ) question is the origin of the universally observed power-law distributions in our model . Clearly an explanation proposed in many phenomenological models [3 , 28] is not applicable here because the rates of all processes , including gene duplication , are constant in the model and do not depend on sizes of already existing gene families . Therefore there are no ad hoc assumptions about the gene birth/death dynamics in the model that could result in power-law distributions . The only plausible reason may be that the underlying dynamics in sequence and structure spaces , coupled with selective pressure , is responsible for the emerging power-law distributions . Indeed , our key finding concerns dynamics of fold discovery and death; that the lifetimes of DPSs are power-law distributed ( Figure 3B ) . The size of a protein family ( and superfamily on longer timescales ) is proportional to DPS lifetime . Indeed , power-law exponents for family size distribution and DPS lifetimes are very similar . While these observations are suggestive , a more detailed future analysis of our model will make it possible to find a definite answer as to the origin of ubiquitous power-law distributions in sequence and fold statistics . Several earlier studies modeled the evolution of proteins by applying pressure directly on the proteins , assuming that the probability of replication of a protein in a population of proteins depends on its molecular properties such as stability [10 , 11 , 22 , 39] , folding kinetics [19 , 44] , or both [16] . In contrast , in the present model , biological ( or as will be argued below “physiological” ) constraints are applied to organisms as a whole , not to individual proteins . Evolutionary simulations and simple theory presented here highlight the importance of this distinction; the genome sizes are closely connected with maximum and average stability of evolved proteins . Therefore , biological pressure is “distributed” in the genome , and all genes act in concert in response to it . Furthermore , no DPSs were found in earlier simulations [10] , despite the fact that the overall population of evolved lattice proteins was somewhat skewed toward more designable structures . In contrast , our key finding is that evolution of population is strongly coupled with protein evolution , as population growth is contingent upon discovery of a very limited set of protein structures . The difference here may be due to the fact that simpler , 2-D lattice models were used in previous simulations , or due to the differences in how biological pressure is applied—on whole organisms here and on individual proteins in earlier works [10 , 45] . The presented model is markedly different from standard models of PG such as Fisher-Wright and QS [46 , 47] . PG and QS models are phenomenological descriptions of evolution , attributing certain values of fitness for the genomes with predetermined combinations of alleles . These models conveniently sidestep the important question of the molecular origins of the change of fitness upon recombination or mutation . The genotype–phenotype relationship in our model is not phenomenological but physiological: when a gene product loses stability ( and by implication functionality ) , the whole organism is likely to die . This assumption is justified by recent high-throughput experiments that use RNAi to determine the impact of gene knockout on phenotype [48 , 49] . As in an experiment where knockout of essential genes results in death of an organism , in our model the deterioration of stability of any gene of an organism confers the lethal phenotype . In the present implementation , our model assumes that all genes are essential . However , this assumption can be relaxed ( unpublished data ) , making it possible to study differentially the impact of biological constraints on the evolution of genes [50] . Another critical distinction between our approach and traditional phenomenological models is that in PG and QS approaches a single genotype is assumed to be advantageous [46 , 51] . While the outcome may be that genomes of the populations are peaked around the most fit one ( as in the standard QS model [47] ) or that a broader distribution among genotypes may emerge ( as in the “survival of the flattest” scenario [52] ) , it is always an implication of the key assumption that a certain genotype confers the highest fitness . In contrast , the present model makes no a priori assumptions about the fitness advantage of a certain genotype . Strikingly , sets of organisms distributed around dominant genomes and proteomes—species—emerge here as a result of evolution at longer evolutionary times . A key factor determining the emergence of species in this model is that productive evolution occurs only when the structural diversity of proteins collapses into a small set of DPSs . Our model of natural selection is minimalistic and is limited in its scope . It does not take into account such important biological processes as horizontal gene transfer , gene recombination , sexual reproduction , “death of a gene” ( via pseudogenisation ) , and Darwinian selection due to competition of populations for limited resources . Also , to make the minimum possible number of assumptions , the modern amino acid alphabet is used in the model , although it has been suggested that the amino acid alphabet itself had evolved over time [53–55] . However , we believe that our model is an important step toward the unification of microscopic physics-based models of protein structure and function and the macroscopic ( so far , phenomenological ) description of the evolutionary pressure . Its extensions are straightforward and may include a more explicit consideration of protein function , protein–protein interactions , and fitness function that rewards functional ( and therefore , structural ) innovations . Furthermore , since habitat temperature enters the model explicitly , it can be used to study thermal adaptation of organisms as well as adaptation to variable mutation rates . This work is in progress .
In our model , an organism is completely described by the set of its genes . The genetic code then defines amino acid sequences , and the exact nature of the lattice protein-folding model makes it possible to find the native structures of the encoded proteins . We assume that for an organism to function properly , it is imperative that its proteins spend a significant part of the time in their native conformations at a given environmental temperature . Let Pnat ( i ) be the thermodynamic probability that protein i is in its native conformation ( see protein model below ) . As the simplest approximation , we assume that the probability that an organism is alive is proportional to the lowest Pnat ( i ) across all of its proteins: i . e . , the longevity of an organism is determined by the least stable protein in the genome ( “weakest link” model ) . Our model of population and genome dynamics includes four elementary events: ( i ) random mutation of a nucleotide in a randomly selected gene with constant rate m per unit time per DNA length; mutations leading to the stop codon are rejected to ensure the constant length of protein sequences; ( ii ) duplication of a randomly selected gene within an organism's genome with constant rate u; ( iii ) birth of an organism via duplication of an already existing organism with constant rate b ( the genome is copied exactly ) ; and ( iv ) death of an organism with the rate d per unit time ( Figure 1 ) . For simplicity , we do not allow for the formation of pseudogenes or any other mechanism of removal of the genes from a genome; in every organism the number of genes increases ( or remains constant ) with time . However , the average number of genes per organism in the population can either decrease or increase due to enhanced survival of organisms with shorter ( longer ) genomes . Condition ( Equation 3 ) translates into the dependence of organism death rate d on the stability of its proteins: where d0 is the reference death rate . This relation gives rise to an effective selection pressure on proteins since organisms that have at least one unstable protein live shorter and thus produce less progeny . This simple , direct , and physically plausible relationship between the genotype ( thermodynamic properties of the proteins ) and the phenotype ( life expectancy ) is the key novel feature of our model . Another implication of this relationship is in the “collective punishment” effect that genes do not evolve independently; a very unfavorable mutation in a gene will likely lead to a quick death of an organism , so its complete genome will not be able to proliferate . Such cooperativity creates an important selection pressure toward mutation-resistant genes encoding stable and evolvable ( see below ) proteins . Interestingly , purely physical factors ensure that resistance to mutations , evolvability of a new function , and thermostability are well correlated [24 , 38] , so little or no tradeoff may be needed to satisfy both requirements . To ensure that a sufficient selection pressure is applied , we set d0 = b/ ( 1 − Pnat ( 0 ) ) , where Pnat ( 0 ) is the native state probability of a protein encoded by the primordial gene , which is the single gene in all organisms from which evolution runs start . Therefore , the Malthus parameter b − d of population growth is zero for neutral mutations ( not changing Pnat with respect to the primordial sequence ) , positive for favorable mutations that increase Pnat , and negative for deleterious mutations . In principle , the relationship between growth rate and protein stability can be experimentally verified by analyzing the growth rate of bacteria at elevated temperatures . While the exact biochemical mechanisms leading to slower replication and eventual death are complicated , they all originate in the loss of protein function or enzymatic activity due to thermal denaturation [56] . A sequence evolution model , also using the protein stability Pnat as fitness parameter has been recently proposed by Goldstein and coworkers [54] . In our model , each organism is represented by a list of its genes , 81 nucleotide sequences that are translated into amino acid sequences according to the genetic code . There can be up to 100 genes per organism; the gene duplication rate is chosen so that this limit is never reached in a simulation; typically , organisms have fewer than ten genes each at the end of a simulation . Initially , 100 organisms are seeded with one and the same primordial gene; Pnat ( 0 ) is the native state probability of the protein encoded by the primordial gene . At each time step of the evolution , each organism can undergo one of the five events: no event at all or the four events described in the main text ( duplication of an organism with probability b = 0 . 15 , death with rate d , gene duplication with probability u = 0 . 03 , and point mutation of a randomly chosen gene with probability m = 0 . 3 per gene ) . The organism death rate is calculated according to Equation 4 , , with d0 = b/ ( 1 − Pnat ( 0 ) ) . Every 25 time steps , an entire set of genes of all currently living organisms is recorded for analysis . The simulation stops after 3 , 000 time steps . Whenever the population size N exceeds 5 , 000 , we randomly remove N − 5 , 000 organisms to ensure constant population size , simulating a turbidostat; despite the artificially constrained population size , the growth regime remains exponential . To simulate the thermodynamic behavior of evolving proteins , we use the standard lattice model of proteins , which are compact 27-unit polymers on a 3 × 3 × 3 lattice [57] . The residues interact with each other via the Miyazawa-Jernigan pairwise contact potential [58] . It is possible to calculate the energy of a sequence in each of the 103 , 346 compact conformations allowed by the 3 × 3 × 3 lattice and the Boltzmann probability of being in the lowest energy native conformation , where E0 is the lowest energy among the 103 , 346 conformations , Ei are the energies of the sequence in the remaining 103 , 345 conformations , and T is the environmental temperature ( in the simulation , we assumed T = 0 . 5 in Miyazawa-Jernigan dimensionless energy units ) . Suppose each genome has N genes , and the fitness of the entire genome is then defined by . Based on the sequence design simulation , we find that it is a reasonably good approximation to assume that in our lattice model the distribution of stability Pnat of a lattice protein after a point mutation ( i ) does not depend on the stability before the mutation and ( ii ) is uniformly distributed between 0 and 1 . Performing a point mutation , we can either mutate the gene with the lowest fitness value , with probability ( case A ) , or select any one of the other more stable genes and mutate it with probability , case B . In case A , because the mutated gene was the original least-stable gene , there are two possible outcomes after the mutation: ( i ) if the new gene fitness value is less than f , then this new gene fitness value would be the new minimum among all gene fitnesses in the genome , therefore this new fitness will become the fitness of the new genome . This occurs with a probability f , and since the new fitness follows a uniform distribution in the region [0 , 1] , the expectation value in this case is f/2 . So this part's contribution to the expectation value of the new genome fitness is . ( ii ) If the new gene stability is greater than f , which happens with a probability of ( 1 − f ) , we can calculate the probability distribution for the new genome fitness being x is The significance of this equation is that when one gene has fitness f < x < 1 and is the new minimum , also under the condition that all fitnesses are within the region of [f , 1] , the probability for all the other ( N − 1 ) genes has to have fitness greater than x is . The multiplicity of this condition is . This is because we can pick any one of the N genes to be the new least-fit gene , and the fitness value is within the region [f , 1] with uniform probability distribution . The contribution for the new genome fitness in this situation is therefore the total product of the probability of this situation , the multiplicity MN , f , and the expectation value ∫f1 xp ( x ) dx , In case B , we also have two possible situations , situation B1 states when the mutated gene has a fitness less than f , and situation B2 states when the mutated gene has a fitness greater than f . In situation B1 , similar to the derivation in case A1 , the probability for the new stability to be smaller than f is f , and the expectation value of the new genome fitness in this case is f/2 . So we have . In situation B2 , if the stability of the mutated gene is greater than f , then the original gene with stability f would remain the least stable in the genome . Therefore , the genome fitness in this situation is still f , so the value of B2 reads , where is the probability to choose one of the ( N − 1 ) genes with stability greater than f , ( 1 − f ) is the probability to mutate this gene with fitness greater than f , and f is the expectation of the final fitness under this condition . Finally , summing up A1 , A2 , B1 , and B2 , we obtain the expectation value of the genome fitness after one point mutation: Now , if the average genome fitness after a single point mutation must be greater than the original fitness , the condition f′ − f > 0 must be satisfied . Solving this inequality , we find an upper limit on the number of genes in a genome ( Equation 2 ) . We took sequences of all structurally characterized domains from HSSP [59] . We used BLAST [60] with threshold 10−10 to identify all sequences with significant homology to each HSSP domain in a nonredundant sequence database NRDB90 [61] . We combined each set of sequences with homology into a single gene family . The number of nonredundant sequences matching the domain is the number considered in that family . We then used cross-indexing between NRDB90 [61] , Swiss-Prot [62] , and InterPro [63] to define the set of different functions each gene family performs . The number of different functions as defined by InterPro becomes the number of superfamilies folding into the same domain . In the model , the superfamily size is defined as the number of nonhomologous sequences with all mutual pairwise Hamming distance of 16 or more ( i . e . , 40% sequence identity or less ) having the same native conformation . The family size is defined as the number of all sequences folding into a given structure , without removing the homologous sequences . To construct the PDUG from the simulation data , we considered only the nonhomologous amino acid sequences . The selection is based on the Hamming distance between the sequences , which should exceed 18 ( i . e . , less than 33% sequence identity ) . To calculate the structure similarity in the PDUG , we used the Q score similarity measure . The Q score measure between the two structures i and j is the number of all pairs of monomers ( k , m ) that are in contact both in structure i and structure j . As there are always 28 contacts in compact 27 mers , Q score varies from 0 for completely dissimilar structures to 28 for two identical structures . The Q score is analogous to the distance matrix alignment ( DALI ) Z score , used as a structural similarity measure for real proteins . The simplest construction of the LUCA that still yields useful information is the delineation of the very old domains . Any domain shared by the three kingdoms of life can be placed in the LUCA [64] . If any such domain were not placed in the LUCA , multiple independent discovery ( or horizontal transfer ) events would be required to explain the occurrence of this domain in all kingdoms . The “extra” evolution involved in this case would result in a less parsimonious scenario . Inclusion of other domains is more probabilistic and depends on the exact form and method of parsimony construction used [64] . We thus define the structural content of the LUCA to be all domains that have homologs in at least one archaeal , at least one prokaryotic , and at least one eukaryotic species . This yields approximately one-third of the PDUG members . | Here , we address the question of how Darwinian evolution of organisms determines molecular evolution of their proteins and genomes . We developed a microscopic ab initio model of early biological evolution where the fitness ( essentially lifetime ) of an organism is explicitly related to the evolving sequences of its proteins . The main assumption of the model is that the death rate of an organism is determined by the stability of the least stable of their proteins . A lattice model is used to calculate stability of all proteins in a genome from their amino acid sequence . The simulation of the model starts from 100 identical organisms , each carrying the same random gene , and proceeds via random mutations , gene duplication , organism births via replication , and organism deaths . We find that exponential population growth is possible only after the discovery of a very small number of specific advantageous protein structures . The number of genes in the evolving organisms depends on the mutation rate , demonstrating the intricate relationship between the genome sizes and protein stability requirements . Further , the model explains the observed power-law distributions of protein family and superfamily sizes , as well as the scale-free character of protein structural similarity graphs . Together , these results and their analysis suggest a plausible comprehensive scenario of emergence of the protein universe in early biological evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"viruses",
"biophysics",
"archaea",
"computational",
"biology"
] | 2007 | A First-Principles Model of Early Evolution: Emergence of Gene Families, Species, and Preferred Protein Folds |
Comparative genomic analyses of primates offer considerable potential to define and understand the processes that mold , shape , and transform the human genome . However , primate taxonomy is both complex and controversial , with marginal unifying consensus of the evolutionary hierarchy of extant primate species . Here we provide new genomic sequence ( ∼8 Mb ) from 186 primates representing 61 ( ∼90% ) of the described genera , and we include outgroup species from Dermoptera , Scandentia , and Lagomorpha . The resultant phylogeny is exceptionally robust and illuminates events in primate evolution from ancient to recent , clarifying numerous taxonomic controversies and providing new data on human evolution . Ongoing speciation , reticulate evolution , ancient relic lineages , unequal rates of evolution , and disparate distributions of insertions/deletions among the reconstructed primate lineages are uncovered . Our resolution of the primate phylogeny provides an essential evolutionary framework with far-reaching applications including: human selection and adaptation , global emergence of zoonotic diseases , mammalian comparative genomics , primate taxonomy , and conservation of endangered species .
The human genome project has revolutionized such fields as genomics , proteomics and medicine . Markedly absent from these many advances however , is a formal evolutionary context to interpret these findings , as the phylogenetic hierarchy of primate species has only modest local ( family and genus level ) molecular resolution with little consensus on overall primate radiations . The exact number of primate genera is controversial and species counts range from 261–377 [1]-[3] . Although whole genome sequencing of 12 primate species are now completed , or nearly so , broader genome representation of man's closest relatives is necessary to interpret human evolution , adaptation and genome structure to assist in biomedical advances . Primate taxonomy has undergone considerable revision but current views [1]-[3] concur that 67–69 primate genera originated from a common ancestor during the Cretaceous/Paleocene boundary roughly 80–90 MYA . An Eocene expansion formed the major extant lineages of 1 ) Strepsirrhini , which is composed of Lorisiformes ( galagos , pottos , lorises ) , Chiromyiformes ( Malagasy aye-aye ) and Lemuriformes ( Malagasy lemurs ) ; 2 ) Tarsiiformes ( tarsiers ) and 3 ) Simiiformes composed of Platyrrhini ( New World monkeys ) and Catarrhini , which includes Cercopithecoidea ( Old World monkeys ) and Hominoidea ( human , great apes , gibbons ) ( see Figure 1 ) . Primate taxonomy , initially imputed from morphological , adaptive , bio-geographical , reproductive and behavioral traits , with inferences from the fossil record [1]-[3] is complex . Recent application of molecular genetic data to resolve primate systematics has been informative , but limited in scope and constrained to just specific subsets of taxa . Efforts to overcome this deficiency using a supermatrix approach [4] , [5] with published sequences culled from these prior studies are inherently flawed by a prohibitively large proportion of missing data for each taxon ( e . g . 59–85% see [5] ) . Here we employ large-scale sequencing and extensive taxon sampling to provide a highly resolved phylogeny that affirms , reforms and extends previous depictions of primate speciation . In turn , the clarity of the primate phylogeny forms a solid framework for a novel depiction of diverse patterns of genome evolution among primate lineages . Such insights are essential in ongoing and future comparative genomic investigation of adaptation and selection in humans and across primates .
The relative placement of suborder Strepsirrhini and infraorder Tarsiiformes at an early stage of primate evolution has been difficult to resolve [8]-[11] . Presently distributed in the islands of Borneo , Sumatra , Sulawesi and the Philippines , Tarsiiformes had a broad Holarctic distribution during the Eocene [10] . Phylogenetic placement of tarsiers has alternatively been defined as 1 ) sister taxa to Strepsirrhini to form Prosimii [2] , [8] , [12] , 2 ) allied with Simiiformes ( Anthropoidea ) to form Haplorrhini [1] , [13] , [14] and 3 ) a separate relict lineage with an independent origin [15] . Here we provide strong evidence that strepsirrhines split with suborder Haplorrhini approximately 87 MYA ( node 185 ) . The ancient lineage is monophyletic and defined by a long branch and eight shared insertions/deletions ( indels ) ( node 144 ) . Rooted by Lagomorpha , the phylogeny affirms Dermoptera as the closest mammalian order relative to Primates , followed by Scandentia [16] , [17] . A long continuous Tarsiiformes branch ( node 142 ) , marked by 25 synapomorphic indels , is consistent with a relict lineage of ancient origin . The sequence phylogeny unambiguously supports tarsiers as a sister lineage ( albeit distant ) to Simiiformes ( BS = 85 MP; 98 ML; 0 . 99 PP ) to form Haplorrhini ( node 143 ) . A few indels ( Table S5 ) define alternate evolutionary topologies , such as tarsiers aligned with Strepsirrhini ( 1 indel , ZFX ) or Scandentia ( 1 indel , DCTN2 ) , compared with those that support an ancestral grouping of Tarsiiformes +Strepsirrhini +Dermoptera +Scandentia ( 2 indels , PLCB4 , POLA1 ) . These incongruent alternatives suggest further investigation of more complex rare genomic changes as cladistic markers of ancient speciation is needed [17] , [18] . Aided by samples of rare taxa , the phylogeny expands upon recent findings [19]-[21] to better resolve long-standing questions on the evolution of Lorisiformes and the two endemic Madagascar infraorders of Chiromyiformes and Lemuriformes . Our data affirm the ancient split of Strepsirrhini , approximately 68 . 7 MYA ( node 144 ) , into the progenitors of Lemuriformes/Chiromyiformes ( origin 58 . 6 MYA , node 174 ) and Lorisiformes ( origin 40 . 3 MYA , node 184 ) . Lorisiformes evolution includes the radiation of Lorisidae ( pottos and lorises , 37 MYA , node 179 ) and Galagidae ( 19 . 9 MYA , node 183 ) species . Within Lorisidae , the four extant genera split into the African subfamily Perodicticinae ( Arctocebus , Perodictus ) and the Asian subfamily Lorisinae ( Nycticebus , Loris ) and are the most divergent within all of primates . For example , mean nucleotide divergence between Lorisidae species is 4–5 times that observed in family Hominidae ( Figure 3 ) and significantly ( p<0 . 05 ) exceed the average genetic divergence across all of Strepsirrhini ( nodes 176–178 , Table S7 , Figure 3 ) . Galagidae are found only in Africa and currently are divided into four genera . However , the Otolemur lineage ( node 180 ) is placed as part of a paraphyletic grouping ( node 182 ) along with two other extant Galago lineages ( nodes 181 , 183 ) , suggesting that further taxonomic investigation of Galago is warranted . Common ancestors of Chiromyiformes and Lemuriformes likely colonized the island of Madagascar prior to 58 . 6 MYA ( node 174 ) . Noted for extensive adaptive evolution , the relative hierarchical branching patterns of the four Lemuriformes families ( Indriidae , Lepilemuridae , Lemuridae , Cheirogaleidae ) recognized by taxonomists , has proven difficult to resolve conclusively . Inferences on species versus subspecies classification are controversial with as many as 97 Malagasy lemurs [22] under taxonomic review . Chiromyiformes diverged from a common ancestor with Lemuriformes shortly after colonisation of Madagascar [14] , [19] and today consists of a single relict genus Daubentonia defined by a long branch with high indel frequency ( N = 14 ) ( Figure 2 , Figure S1 , Table S7 ) . The evolution of the four Lemuriformes families began 38 . 6 MYA ( node 173 ) with the emergence of Lemuridae , followed by Indriidae and a monophyletic lineage that split 32 . 9 MYA ( node 152 ) to form the sister lineages of Lepilemuridae and Cheirogaleidae . This branching pattern among families agrees with earlier nuclear gene segment findings [20] that differ from studies using mtDNA sequence and Alu insertion variation which were unable to resolve these hierarchical associations [19] . Further , relatively weak nodal support here collapses Lemuriformes into an unresolved trichotomy of Lemuridae , Indriidae , and the Lepilemuridae + Cheirogaleidae lineage ( node 158 ) . Optimal resolution of this node is observed with exon sequences ( Figures S8 and S9 ) , indicating that intron sites may be saturated , while more conserved coding regions remain informative and reflect the ancient rapid radiation of Lemuriformes families . The phylogeny clarifies formerly unresolved questions concerning New World primate evolution including branching order among families , relative divergence of genera within families , and phylogenetic placement of Aotus , and provides genetic support for examples of adaptive evolution that led to nocturnalism , “phyletic dwarfism” and species diversification within the Amazonian rainforest . Here , Platyrrhini clearly diverged from a common ancestor with Catarrhini ( node 141 ) roughly 43 . 5 MYA during the Eocene . Although questions remain about the route and nature of primate colonization of the New World [23] , [24] and the impact of historic global climate change in neotropical regions [25] , [26] , the phylogeny unambiguously resolves the relative divergence pattern among families from a common ancestor 24 . 8 MYA ( node 78 ) . The common ancestor to Pitheciidae ( uakaris , titis and sakis ) originated 20 . 2 MYA ( node 140 ) and the majority of these species currently are distributed in the neotropical Amazonian basin extending from the Andean slopes to the Atlantic . Next to radiate are the Atelidae ( node 126 ) , with the most basal lineage leading to Alouatta ( howler monkeys ) , currently widely distributed from Mexico to northern Argentina , followed by the divergence of Ateles ( spider monkeys ) from South American lineage comprised of sister genera ( node 121 ) of Lagothrix ( woolly monkeys ) and Brachyteles ( muriquis ) . The Cebidae radiation initiated with the emergence of sister taxa Cebus ( Cebinae ) and Saimiri ( Saimirinae ) approximately 20 MYA ( node 113 ) , in agreement with other molecular studies [27]-[30] . Subsequently , during a relatively brief interval ( ∼700 , 000 years ) a lineage arose ( node 112 ) that split to form the Callitrichinae ( marmosets and tamarins ) and Aotus ( night monkeys ) . The Aotus lineage ( node 98 ) radiated with unusually high numbers of synapomorphic indels ( N = 15 ) , the most observed in Simiiformes ( Table 2 and Table 3 ) , to form a complex species group of controversial taxonomic designation as subfamily or family and uncertainty over its exact placement relative to other Cebidae lineages . Here , Aotus is the sister lineage to Callitrichinae ( marmosets , tamarins ) as originally hypothesized by Goodman ( 1998 ) [1] , [28] . Aotus species divide into sister lineages , with the “grey-necked” species ( A . trivirgatus + A . lemurinus griseimembra ) distributed north of the Amazon River , and “red-necked” species A . nancymaae , A . azarae species and associated subspecies located most to the south ( nodes 98 , 101 , 102 ) . The unusual depth of divergence ( i . e . sizeable nucleotide substitutions/site; high indel frequency ) may exemplify adaptive speciation as Aotus are the only nocturnal Simiiformes [31] , and thereby may have reduced competition with diurnal small-bodied platyrrhines inhabiting the same neotropical environments . Another case of adaptation termed “phyletic dwarfism , ” defined as a gradient in morphological size partially correlated with evolutionary time [32] , is supported in Cebidae . Aotus , Cebus and Saimiri species are larger than the more derived and smaller squirrel-sized Callitrichinae of Saguinus , Leontopithecus , Callimico , Mico , Cebuella and Callithrix . In Callitrichinae , Saguinus is the first to diverge with S . fuscicollis currently distributed south of the Amazon River . Subsequently , the genus diversified into northern ( S . bicolor , S . midas , S . martinsi , S . geoffroyi , S . oedipus ) and south Amazonian species ( S . imperator , S . mystax , S . labiatus ) ; a trend generally similar to findings based on mtDNA [33] and single nuclear genes [34] . The hierarchical branching order among the remaining Callitrichinae of Leontopithecus , Callimico , Callithrix and Mico mirrors decreasing body size and culminates with the smallest platyrrhine species , Cebuella pygmaea , as most derived . This phylogenetic depiction of Callitrichinae is concordant with several other morphological and reproductive traits [32] , [35] related to dwarfism and perhaps reflects adaptive evolution selected by fluctuating resource availability within the Amazon and Atlantic coast rainforests [36] . Cercopithecoidea ( family Cercopithecidae ) speciation patterns are confounded by symplesiomorphic traits in morphology , behavior and reproduction , and are further confused by hybridization between sympatric species , subspecies and populations ( summarized in [2] ) . Cercopithecidae includes two subfamilies , Colobinae and Cercopithecinae , which diverged 18 MYA ( node 62 ) , but classification schemes [2] are marked by inconsistencies between morphological [37] , [38] and genetic data , as well as differences among genetic data studies [4] , [27] , [39]-[44] . Colobinae radiation started approximately 12 MYA ( node 42 ) with species adapted to an arboreal , leaf-eating existence . Asian ( tribe Presbytini ) and African ( tribe Colobini ) genera are monophyletic ( nodes 53 and 61 , respectively ) , supporting earlier genetic findings [4] , [40] over morphology-based taxonomy [2] , [45] . Whilst African genera Piliocolobus and Colobus are commonly recognized , the taxonomic schemes for the critically endangered Asian langur and leaf monkeys , all sharing digestive adaptations for an arboreal folivorous diet , have ranged from a single genus Presbytis to three distinct genera ( Trachypithecus , Semnopithecus , Presbytis ) . Here , the Presbytis lineage , distinguished by 3 indel events ( node 56 ) , diverged first within Asian Colobinae , followed by the odd-nosed group ( Rhinopithecus , Nasalis , Pygathrix ) , Trachypithecus and Semnopithecus . As odd-nosed species are not exclusively arboreal and folivorous , the results indicate either 1 ) morphological convergence between Presbytis with Trachypithecus and Semnopithecus , 2 ) adaptation for an expanded diet in the odd-nosed group , or 3 ) that a folivorous diet is a symplesiomorphic trait within Asian colobines . Trachypithecus and Semnopithecus genera consist of closely related , often sympatric species ( node 51 ) , distributed in the Indian subcontinent and SE Asia , with inconsistent phylogenetic resolution among species [4] , [27] , [40] , [44] , [46] . Nonetheless , all genetic studies , including the present , place Trachypithecus vetulus ( monticola ) nested within the Semnopithecus clade ( node 50 ) , suggesting the need for taxonomic revision . Further , previously ambiguous associations between Trachypithecus and Semnopithecus ( nodes 43–51 ) are clarified . Inter-specific genetic differences are roughly half those observed among other colobine genera ( Figure 2 , Figure S1 , Table 3 , Table S9 ) and may indicate that recent speciation , taxonomic over-splitting , reticulate evolution , or a combination thereof , ( e . g . see [40] , [44] , [46] ) are common within the Asian Colobinae radiation . The remainder of Old World monkeys ( tribes Papionini and Cercopithecini ) [2] arose from a common ancestor approximately 11 . 5 MYA ( node 41 ) . Considerable interest in Cercopithecinae speciation is motivated not only by primate conservation , but increased biomedical surveillance for novel zoonotic agents and comparative research of host-pathogen adaptation relevant to the study of deadly human viral pandemics such as HIV/SIV . Cercopithecini ( guenons , patas monkey , talapoin , green monkeys ) include lineages rooted by divergent monotypic genera followed by more recent speciation , characterized by transition from an arboreal to a terrestrial lifestyle . Generally arboreal , Miopithecus and Allenopithecus are early offshoots with respect to the two Cercopithecini subclades formed approximately 7 MYA . The Cercopithecus lineage ( node 34 ) radiated after Miopithecus and retained an arboreal lifestyle . The second , rooted by Allenopithecus , forms a terrestrial clade of Erythrocebus patas and Chlorocebus species , with Cercopithecus l'hoesti separated the other Cercopithecus . This paraphyly , also reported in earlier genetic studies [39] , [47] , [48] and counter to initial morphological classifications [2] , suggests taxonomic revision of Cercopithecus . Further , resolution of Allenopithecus ( node 40 ) and Miopithecus ( node 35 ) speciation herein suggests a single evoluiontary transition from an arboreal to a terrestrial lifestyle in E . patas , C . l'hoesti , and Chlorocebus species . Papionini ( macaques , mandrills , drills , baboons , geladas , mangabeys ) is a taxonomically complex tribe [2] . One of the more familiar genera within Cercopithecoidea , Macaca ( macaques ) diverged 5 . 1 MYA and today is represented by an African lineage comprised of a single species M . sylvanus , and an Asian lineage consisting of well-defined species groups ( fascicularis , sinica , mulatta , nemestrina , Sulawesi ) inhabiting India and Asia , SE Asia and Sundaland [49] . With the exception of the fascicularis group , which is split in this study whereby M . arctoides [fascicularis] is more closely aligned with M . thibetana [sinica] rather than M . fascicularis as expected , our data otherwise strongly support these macaque species groups ( nodes 6 , 11 ) . Moreover , the phylogeny affirms Groves [2] proposal that Lophocebus and Theropithecus are distinct clades apart from Papio ( nodes 18 , 19 ) , although the average nucleotide divergence among these three genera are generally less than between other recognized Papionin genera ( Macaca , Mandrillus , Cercocebus ) ( Figure 2 , Figure S1 , Table 3 , Table S9 ) . Lastly , sequence divergence between tribes is unequal with Cercopithecini nearly twice that of Papionini ( mean branch length = 13 . 1 , 7 . 43 , respectively , p<0 . 005 ) and there are numerous instances of discordance between the present phylogeny with previous mtDNA studies [4] , [5] suggesting that continued resolution of Cercopithecinae speciation and of Papionini in particular , will likely include evidence of reticulate evolution represented by ongoing and historic episodes of hybridization ( e . g . see [39] , [48] ) . Once contentiously debated , the closest human relative of chimpanzee ( Pan ) within subfamily Homininae ( Gorilla , Pan , Homo ) is now generally undisputed . The branch forming the Homo and Pan lineage apart from Gorilla is relatively short ( node 73 , 27 steps MP , 0 indels ) compared with that of the Pan genus ( node 72 , 91 steps MP , 2 indels ) and suggests rapid speciation into the 3 genera occurred early in Homininae evolution . Based on 54 gene regions , Homo-Pan genetic distance range from 6 . 92 to 7 . 90×10−3 substitutions/site ( P . paniscus and P . troglodytes , respectively ) , which is less than previous estimates based on large scale sequencing of specific regions such as chromosome 7 [50] . The highly endangered orangutan forms the single genus Pongo in subfamily Ponginae ( nodes 75–76 ) , the sister lineage to Homininae . Currently restricted to the islands of Borneo and Sumatra , orangutans once inhabited all of Southeast Asia during the Pleistocene [51] . Differences in behavior , morphology , karyology , and genetic data between the two island populations [2] support the taxonomic designation as two separate species of Bornean ( P . pygmaeus ) and Sumatran orangutans ( P . abelii ) , and these designations are upheld by the data presented here . Hylobatidae ( siamang , gibbons , hoolock ) are noted for exceptional rates of chromosome re-arrangement [52] , [53] , 10–20 times faster than in most mammals [54] . Classification schemes of the 12 species range from two genera ( Hylobates and Symphalangus ) to four subgenera and/or genera ( Hylobates , Nomascus , Symphalangus , Hoolock ) , defined by unique numbers of chromosomes [54] , [55] . The eight species included in this study form three clades that coincide with genus designation ( absent is Hoolock; nodes 64–69 ) that diverged rapidly 8 . 9 MYA . Moreover , Nomascus species appear more recent than Symphalangus and Hylobates , with node divergence dates estimated at less than 1 MY ( Table 3 , Table S9 , Figure 2 ) . Thus , Hylobatidae exhibits episodes of rapid divergence perhaps related to excessive genome re-organization and warrants additional investigation . The clarity of the primate phylogeny here can be used to assess nucleotide divergence patterns , rates of substitution and accumulation of synapomorphic and autapomorphic indels . Genome divergence varies across primate lineages , with the least inter-specific differences observed in Cercopithecidae lineages and the most in Lorisidae , reflecting recent speciation in the former and the more ancient origins of the latter ( Figure 3 , Table 1 , Table 3 , Tables S7 and S9 ) . The global rate of nucleotide substitution across the entire primate phylogeny is 6 . 163×10−4 substitutions/ site/ MY , but exhibits significant heterogeneity across lineages ( Figure 3 ) and among branches ( Table 1 , Table 2 , Table 3; Tables S6 , S7 , S8 ) . For example , the “hominoid slow-down” hypothesized to have occurred in human evolution , is confounded by the reduced rates observed in all Catarrhini ( not just Homininae ) compared with Platyrrhini and Strepsirrhini ( Figure 3 , Table S10 ) . By contrast , the “phyletic dwarfism” of the Callitrichinae ( nodes 97 , 85 ) and the evolution of nocturnalism in Aotinae are correlated with increased rates along specific branches ( see nodes 99 , 100 ) rather than an being a function of an average rate among all branches within the lineage ( Figure 3 ) , suggesting that an adaptive “speed-up” occurred in the common ancestors of these extant species . The genome accumulates indels over evolutionary time , altering the degree of sequence homology between taxa . Further , large-scale genome sequence analysis demonstrate that indel formation is an indicator of genome plasticity , positively correlated with adjacent nucleotide substitution rate [56] , [57] , gene segmental duplication , chromosomal position , hybridization between species and speciation , and is enhanced by molecular mechanisms of recombination among repetitive elements [58]-[60] . Here , the distribution of indels is ubiquitous in both coding and noncoding segments ( Tables S4 , S5 , S6 ) , but is markedly disjunct among primate lineages ( Figure 3 ) . Excluding the infraorders Tarsiiformes ( 25 indels ) and Chiromyiformes ( 14 indels ) due to statistically inadequate sampling , the indel frequency per branch varies by a factor of 20 ( Table 1 , Table 2 , Table 3; Tables S7 , S8 , S9 ) with the greatest accumulation within Lorisidae ( particularly Arctocebus calabarensis ) and the least in Cercopithecidae ( Figure 3 ) . The major correlate of indel frequency is not substitution rate , but overall genome divergence represented by branch length ( R-square = 0 . 659 Lorisiformes; 0 . 610 Lemuriformes; 0 . 3286 Simiiformes; P<0 . 05 ) . The molecular genetic resolution of the primate phylogeny provides a robust comparative genomic resource to affirm , alter , and extend previous taxonomic inferences . Approximately half of the 261–377 species and 90% of the genera are included facilitating resolution of long-standing phylogenetic ambiguities . Early events within primate evolution are resolved such as: Dermoptera is the closest mammalian order to Primates; Tarsiiformes are sister taxa with Simiiformes to form Haplorrhini; Chiromyiformes ( Daubentoniidae ) and Lemuriformes are monophyletic indicating a common ancestral lineage colonized the island of Madagascar once; and the hierarchical divergence pattern among New World families Pitheciidae , Atelidae , and Cebidae is clarified . Additional insights are possible because the relative branching patterns among infraorders , parvorders , superfamilies , families , subfamilies , genera and species are resolved with high measures of support for all but three nodes . For example , Old World monkeys ( Cercopithecoidea ) display remarkably low levels of divergence , particularly within Papionini , consistent with reticulate evolution , recent speciation and possibly augmented by taxonomic over-splitting . By contrast , the Lorisidae are marked by extraordinary divergence relative to other primate lineages . In the New World , the phylogenetic placement of the unique , nocturnal Aotinae is unambiguously resolved , diverging rapidly after the sister lineage of Cebinae+Saimirinae and prior to the Callitrichinae within the family Cebidae . Further , the pattern of divergence of Callitrichinae is correlated with a gradation in species size , supporting “phyletic dwarfism” [32] , [35] . In the context of human evolution , the large amount of sequence available here for each well-recognized species in Hominidae provides a baseline estimate of average genetic divergence per taxonomic level in primates . However , deviations from these values observed across diverse lineages illustrate the remarkable biodiversity and species richness within the Primate order . One of the more intriguing unresolved questions is the origin of primates . Generally concordant , most molecular data suggest extant primates arose approximately 85 MYA from a common ancestor . However , the debate continues over the geographic locale most consistent with the existing fossil record [9] , [10] , [12] , [16] , [23] , [26] , [61]-[63] . A parsimonious interpretation of the present data would suggest an Asian origin as the ancient Asian Tarsiiformes and the strepsirrhine Lorisinae are most basal and the closest relatives of primates , Dermoptera and Scandentia , are also exclusive to Asia . Primate genomes harbor remarkable differences in patterns of speciation , genome diversity , rates of evolution and frequency of insertion/deletion events that are fascinating in their own right , but also provide needed insight into human evolution . Advances in human biomedicine including those focused on changes in genes triggered or disrupted in development , resistance/susceptibility to infectious disease , cancers , mechanisms of recombination and genome plasticity , cannot be adequately interpreted in the absence of a precise evolutionary context or hierarchy . Resolution of the primate species phylogeny here provides a validated framework essential in the development , interpretation and discovery of the genetic underpinnings of human adaptation and disease .
Primate DNA samples were obtained following the guidelines of Institutional Animal Care and Use Committee policies of respective research institutions ( see Table S1 ) . All tissue samples for the Laboratory of Genomic Diversity were collected in full compliance with specific Federal Fish and Wildlife permits from the Conservation of International Trade in Endangered Species of Wild flora and Fauna: Endangered and Threatened Species , Captive Bred issued to the National Cancer Institute ( NCI ) -National Institutes of Health ( NIH ) ( S . J . O . principal officer ) by the U . S . Fish and Wildlife Services of the Department of the Interior . Duke University Lemur samples ( J . E . H . ) were collected under research project BS-4-06-1 and Institutional Animal Care and Use Committee ( IACUC ) project A094-06-03 , and this paper is Duke Lemur Center publication #1192 . A complete list of individual and source DNA are presented in Table S1 . DNA was extracted from whole blood , buffy coat , hair or buccal swab samples using DNeasy Blood & Tissue Kit ( Qiagen ) following manufacture's protocol . DNA from different tissues ( muscle , kidney etc ) or cell culture pellets was extracted using standard phenol∶chloroform extraction methods . Proteinase K digestion in lysis buffer ( 100 mM NaCl , 10 mM Tris-HCl pH 8 . 0 , 25 mM EDTA pH 8 . 0 , 0 . 6% SDS , 100 µg/ml RNAse A ) at 56 °C for 3–12 hours rotating was followed by 30 minute phenol , phenol∶chloroform 70∶30 , and chloroform extractions using phase-lock gel tubes ( Eppendorf ) followed by ethanol precipitation and 70% ethanol wash . Dried DNA was reconstituted in TE pH 7 . 4 buffer and stored at 4 °C . DNA was quantified using Nanodrop ( Thermo Scientific ) and its quality was assessed using 0 . 7% agarose gel electrophoresis . DNA of limited quantity was used for whole-genome amplification using REPLI-g Midi Kit ( Qiagen ) . 50–100 ng of genomic DNA ( depending on its quality ) was used per one 50 µl reaction according to the manufacturer's protocol . A negative control ( no template ) was included in every WGA and was verified by downstream PCR and sequencing . Some strepsirrhine DNA was extracted and/or whole genome amplified as previously described [21] . A complete list of 54 primer sets used in this study is presented in Table S2 . This list includes primers from earlier studies [12] , [16] , [21] , [64]-[68] , as well as those designed specifically for this study using a unique bioinformatics approach ( Pontius , unpublished data ) . A panel of species representing the breadth of primate diversity was used in the testing and optimization of PCR primers and included: Gorilla gorilla , Pan paniscus , Nomascus leucogenys , Symphalanges syndactylus , Erythrocebus patas , Macaca fuscata , Macaca tonkeana , Chiropotes satanas , Saimiri boliviensis , Saimiri sciureus , Callithrix jacchus , Ateles fusciceps , Saguinus fuscicollis , Cheirogaleus medius , Lemur catta and Tupaia minor . All nuclear gene regions in all the samples were amplified with the following conditions . Either 30 ng of genomic DNA or 1 µl of WGA product was diluted 1∶10 with 0 . 1XTE per PCR reaction . DNA quantity was increased for poor quality DNA . Genomic and WGA DNA was aliquoted into plates , dried at room temperature and stored at 4 °C . Each 15 µl PCR reaction contained 2 mM MgCl2 , 250 µM of each dNTP , 150 µM of each forward and reverse primer , 0 . 8 units of AmpliTaq Gold polymerase ( ABI ) with 1X GeneAmp 10X PCR Gold Buffer . PCR was performed in PE ABI GeneAmp 9700 and Biometra T1 thermal cyclers . PCRs were carried out using a touchdown program with the following parameters: initial denaturation for 10 min at 95 °C; followed by 10 cycles of 95 °C for 15 s , 60–52 °C ( 2 cycles for each of the five down gradient annealing temperature steps: 60 °C , 58 °C , 56 °C , 54 °C and 52 °C ) for 30 s , and 72 °C for 1 min; and followed by 25 cycles of 95 °C for 15 s , 50 °C for 30 s , and 72 °C for 1 min; and a final extension at 72 °C for 30 min . PCR products were analyzed on 2% agarose gels . Only PCR products that produced single bands were sequenced . PCR products were purified using AMPure kit ( Agencourt ) or Mag-Bind EZ Pure ( OMEGA ) . PCR products were sequenced directly in two reactions with forward and reverse primers . The sequencing reactions were carried out with the BigDye Terminator v1 . 1 cycle sequencing kit ( Applied Biosystems , Inc . ) . For 10 µl sequencing reactions we used 0 . 25 µl of BigDye , 2 µl of 5X Sequencing buffer , 0 . 32 µM primer and 2 . 5 µl of PCR product ( we diluted PCR product if bands on the gel were too bright ) . Sequencing reactions were performed as following: 25 cycles of 96 °C for 10 s , 50 °C for 5 s , 60 °C for 4 minutes . Sequencing products were purified using paramagnetic sequencing clean-up CleanSEQ ( Agencourt ) or Mag-bind SE DTR ( OMEGA ) . PCR and sequencing cleanups were performed on Beckman Coulter Biomek FX laboratory automation workstation . The sequencing products were analyzed with an ABI PRISM 3730 XL 96-well capillary sequencer . Some of the prosimian PCR products and sequences were obtained following earlier published methods [21] . Consensus sequences for each individual were generated from sequences in forward and reverse directions using Sequencher 4 . 9 program ( Gene Codes Corporation ) . All sequences were deposited in GenBank under accession numbers presented in Table S11 . Multiple sequence files for each gene segment amplified were aligned by MAFFT version 6 [69] , [70] , imported into Se-Al ver 2 . 0a11 [71] and verified by eye . Regions of sequence ambiguity within the alignment were identified by GBLOCK version 0 . 91b [72] , and removed from subsequent phylogenetic analyses . A FilemakerPro database was created to manage all sequence records for each individual DNA specimen and the concatenated dataset was exported . The final , post-GBLOCK , edited , annotated PAUP* nexus alignment of the 54 concatenated genes used for this study is publically available at the following website: http://lgdfm3 . ncifcrf . gov/190Taxa_Rabbit_PAUP . zip The file is a compressed zip file that can be viewed in either a generic text editor , PAUP* , or alignment programs that read large nexus format files . Gene partitions were analyzed separately , as well as combined , for genome comparison and phylogenetic reconstruction . Six gene partitions were created , corresponding to X-chromosome , Y-chromosome , autosome , intron , exon and UTR segments . A separate phylogenetic analysis was conducted for each of the six data partitions to compare the concordance among tree topologies derived from each partition . It should be noted that the Y-chromosome tree is not directly comparable to the topologies of the other data partitions because the number of males ( N = 127 ) was a subset of the total ( N = 191 ) . In the concatenated data set of all 54 genes , females were coded as “missing” for the Y-chromosome gene sequence . Aligned multiple sequence files of either combined data or gene partitions were imported into ModelTest ver 3 . 7 [73] and the optimal model of nucleotide substitution was selected using the AIC criterion . Models are listed in Table S12 . Phylogenetic trees based on nucleotide data were obtained using a heuristic search with different optimality criteria of maximum likelihood ( ML ) and maximum parsimony ( MP ) as implemented in PAUP* ver 4 . 0a109 [74] for Macintosh ( X86 ) and additional runs of ML as implemented in GARLI ver 0 . 96 [75] . In PAUP* , conditions for the ML analysis included starting trees obtained by stepwise addition , and branch swapping using the tree-bisection-reconnection ( TBR ) algorithm . The MP analyses used step-wise addition of taxa , TBR branch swapping and excluded indels . Support for nodes within the phylogeny used bootstrap analysis with identical settings established for each method of phylogenetic reconstruction and values greater than 50% were retained . The number of bootstrap iterations consisted of 1000 for MP methods and 100 for ML . Detailed control files used for GARLI ML analyses are available from corresponding author . We estimated the phylogeny and divergence time splits simultaneously using a Bayesian approach as implemented in the program BEAST ver 1 . 5 . 3 [76] , [77] . Due to computational constraints , analyses were performed with 5 different sets of species: 1 ) genus-level data set including 61 Primate genera , two Dermoptera genera and one Scandentia genus rooted by Lagomorpha , 2 ) Catarrhini species with outgroups , 3 ) Platyrrhini species with outgroups , 4 ) Strepsirrhini species with outgroups and 5 ) genus-level analysis with a partitioned data set allowing for rate heterogeneity and different substitution models for autosome , X-chromosome , and Y-chromosome sequences . By using the uncorrelated lognormal relaxed-clock model , rates were allowed to vary among branches without the a priori assumption of autocorrelation between adjacent branches . This model allows sampling of the coefficient of variation of rates , which reflects the degree of departure from a global clock . Based on the results of ModelTest , we assumed a GTR+I+G model of DNA substitution with four rate categories . Uniform priors were employed for GTR substitution parameters ( 0 , 100 ) , gamma shape parameter ( 0 , 100 ) and proportion of invariant sites parameter ( 0 , 1 ) . The uncorrelated lognormal relaxed molecular clock model was used to estimate substitution rates for all nodes in the tree , with uniform priors on the mean ( 0 , 100 ) and standard deviation ( 0 , 10 ) of this clock model . We employed the Yule process of speciation as the tree prior and a Unweighted Pair Group Method with Arithmetic Mean ( UPGMA ) tree to construct a starting tree , with the ingroup assumed to be monophyletic with respect to the outgroup . To obtain the posterior distribution of the estimated divergence times , nine calibration points were applied as normal priors to constrain the age of the following nodes ( labeled A-H in Figure 1 of main text ) : A ) mean = 40 . 0 MYA , standard deviation ( stdev ) = 3 . 0 for time to most recent common ancestor ( TMRCA ) of galagids and lorisids [78] , B ) mean = 43 . 0 MYA , stdev = 4 . 5 for TMRCA of Simiiformes [79] , [80] , C ) mean = 29 . 0 MYA , stdev = 6 . 0 for TMRCA of Catarrhini [80] , D ) mean = 23 . 5 MYA , stdev = 3 . 0 for TMRCA of Platyrrhini [26] , [81] , E ) mean = 7 MYA , stdev = 1 . 0 for TMRCA of Papionini [82] , F ) mean = 4 . 0 MYA , stdev = 0 . 4 for TMRCA of Theropithecus clade [40] , [83] , G ) mean = 15 . 5 MYA , stdev = 2 . 5 for TMRCA of Hominidae [14] and H ) mean = 6 . 5 MYA , stdev = 0 . 8 for TMRCA of Homo-Pan [84] . A normal prior for the mean root height of 90 . 0 MYA with stdev = 6 . 0 was used based on molecular estimates of MRCA of all Primates [14] , [82] , [85] . The calibration points selected are based on fossil dates that have undergone extensive review in previous publications and are supported by a consensus of paleoanthropologists . Rather than re-iterate the considerable amount of information forming the basis for each calibration point , we list the respective citations with the most detailed overview and attendant references . Four to seven independent Markov chain Monte Carlo ( MCMC ) runs for each analysis were run for 20–100 million generations to ensure sampling of estimated sample size ( ESS ) values . The Auto Optimize Operators function was enabled to maximize efficiency of MCMC runs . Trees were saved every 1000 generations . Log files from each run were imported into Tracer ver 1 . 4 . 1 , and trees sampled from the first 1 million generations were discarded . Mixing of trees was assessed in Tracer by examination of ESS values . Analysis of these parameters in Tracer suggested that the number of MCMC steps was more than adequate , with ESS of all parameters often exceeding 200 , and Tracer plots showing strong equilibrium after discarding burn-in . Tree files from the individual runs were combined using LogCombiner ver 1 . 5 . 3 after removing 1000 trees from each sample . The maximum-clade-credibility tree topology and mean node heights were calculated from the posterior distribution of the trees . Final summary trees were produced in TreeAnnotator ver 1 . 5 . 3 and viewed in FigTree ver 1 . 3 . 1 . Heterogeneity in nucleotide substitution rates among primate taxa was assessed by a Bayesian approach , allowing for unequal rates of nucleotide substitution among lineages as implemented in BEAST . Rate estimates provided for each branch within the primate phylogeny were analyzed by ANOVA as implemented in SAS ( SAS Institute Inc . , SAS 9 . 1 . 3 ) . Significant differences among means used the Duncan multiple means test . Indels were assessed as possible indicators of genome plasticity among primate lineages . An a priori approach was developed that used the derived primate phylogenetic tree ( Figure 2 ) as a guide for identification of synapomorphic and autapomorphic indels . First , all indels were identified using FASTGAP on GBLOCKED alignments and verified by eye . Second , only indels that correctly conformed to the species associations of the primate phylogeny ( Figure 2 ) were used and identified as a subset of synapomorphic events ( Table 1 , Table 2 , Table 3; Tables S5 , S6 ) . Third , another subset of autapomorphic indels were identified and assessed as potential signatures of genome plasticity for a given species ( Tables S7 , S8 , S9 ) . Infrequently , some indels included in the analysis were positioned in regions that did not amplify across all species . In these cases , indels were identified as synapomorphic for a lineage providing ∼70% of the relevant species were successfully PCR amplified , and that species with missing sequence for the indel did not all occur on the same node within the lineage . The hypothesis that patterns of nucleotide substitution are influenced by indel frequency was tested by regression of ln-transformed branch length against ln-transformed indels per branch . Tests of the association between genome rates of evolution and indel frequency were conducted by regression of the rate of nucleotide substitution ( substitution/site/MY ) versus ln-transformed indel frequency per branch . Statistical software used was SAS ( SAS Institute Inc . , SAS 9 . 1 . 3 ) . | Advances in human biomedicine , including those focused on changes in genes triggered or disrupted in development , resistance/susceptibility to infectious disease , cancers , mechanisms of recombination , and genome plasticity , cannot be adequately interpreted in the absence of a precise evolutionary context or hierarchy . However , little is known about the genomes of other primate species , a situation exacerbated by a paucity of nuclear molecular sequence data necessary to resolve the complexities of primate divergence over time . We overcome this deficiency by sequencing 54 nuclear gene regions from DNA samples representing ∼90% of the diversity present in living primates . We conduct a phylogenetic analysis to determine the origin , evolution , patterns of speciation , and unique features in genome divergence among primate lineages . The resultant phylogenetic tree is remarkably robust and unambiguously resolves many long-standing issues in primate taxonomy . Our data provide a strong foundation for illuminating those genomic differences that are uniquely human and provide new insights on the breadth and richness of gene evolution across all primate lineages . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/genomics",
"evolutionary",
"biology/animal",
"genetics",
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"evolutionary... | 2011 | A Molecular Phylogeny of Living Primates |
Australia is the only high-income country in which endemic trachoma persists . In response , the Australian Government has recently invested heavily towards the nationwide control of the disease . A novel simulation model was developed to reflect the trachoma epidemic in Australian Aboriginal communities . The model , which incorporates demographic , migration , mixing , and biological heterogeneities , was used to evaluate recent intervention measures against counterfactual past scenarios , and also to assess the potential impact of a series of hypothesized future intervention measures relative to the current national strategy and intensity . The model simulations indicate that , under the current intervention strategy and intensity , the likelihood of controlling trachoma to less than 5% prevalence among 5–9 year-old children in hyperendemic communities by 2020 is 31% ( 19%–43% ) . By shifting intervention priorities such that large increases in the facial cleanliness of children are observed , this likelihood of controlling trachoma in hyperendemic communities is increased to 64% ( 53%–76% ) . The most effective intervention strategy incorporated large-scale antibiotic distribution programs whilst attaining ambitious yet feasible screening , treatment , facial cleanliness and housing construction targets . Accordingly , the estimated likelihood of controlling trachoma in these communities is increased to 86% ( 76%–95% ) . Maintaining the current intervention strategy and intensity is unlikely to be sufficient to control trachoma across Australia by 2020 . However , by shifting the intervention strategy and increasing intensity , the likelihood of controlling trachoma nationwide can be significantly increased .
Australia is the only high-income country in which trachoma , the worldwide leading cause of preventable blindness [1] , remains endemic [2] . In remote Aboriginal communities deemed to be at-risk of trachoma , an estimated 4% of adults suffer severely impaired vision or blindness [3] due to many years of repeated re-infection with the bacterium Chlamydia trachomatis—the infectious agent from which trachoma disease develops [4] . In 2009 , the Australian government pledged AUS$16 million over an initial four-year period towards the national goal of controlling trachoma by 2020 [3] . That is , to reduce the prevalence of trachomatous inflammation follicular ( TF ) to less than 5% amongst 5–9 year-old children within a community . This target closely aligns with the Global Elimination of Trachoma by 2020 ( GET 2020 ) initiative [5] developed by the World Health Organisation ( WHO ) . The Australian trachoma intervention effort combines annual surveillance activities with a Surgery , Antibiotics , Facial cleanliness and Environmental improvement ( SAFE ) control policy recommended by the WHO [3 , 6] . This four-component policy incorporates treatment for those with clinically detected disease and long-term solutions for reducing infection incidence and disease prevalence [7] . The WHO offers recommendations for the frequency and intensity of the screening and treatment programs integrated into the SAFE policy [8]; however , the Australian intervention effort involves a greater intensity of screening and treatment due to larger resource availability compared to other trachoma-endemic countries . Despite this , the prevalence of trachoma remains high in many Aboriginal communities [3] whilst several developing countries prepare to announce the national control or eradication of the disease [9] . In this paper , we assess the progress of recent trachoma intervention efforts in Australia and evaluate the possibility of achieving national control by 2020 . This is achieved by addressing the following three questions: ( i ) have past trachoma intervention efforts been effective in reducing infection incidence and disease prevalence ? ( ii ) what epidemiological impact can be expected if the current intervention strategy and intensity is maintained until 2020 ? ( iii ) how can a shift in strategy or increase in intensity improve this impact ? These questions are addressed through the development and analysis of a novel simulation model of trachoma transmission in remote Australia . Previous models of trachoma transmission have typically implemented population-based methods [10–14] . These traditional models can be useful for extracting general principles but often lead to an over-simplification of disease dynamics [15] . Recent studies have indicated that transmission between two individuals is influenced by factors such as age , with children younger than 10 years being the typical reservoir of infection [12 , 16] , and the presence of nasal or ocular discharge , i . e . a dirty face [17] . Demographic factors such as household overcrowding [10 , 11] and inter-community migration are also believed to contribute to trachoma persistence [18] . The temporary migration of individuals between communities is believed to be of particular importance in sustaining endemic trachoma in Australia [18] . Here , an individual-based simulation model is developed to incorporate these complexities . This is the most sophisticated trachoma transmission model to date , and the first model to specifically represent endemic trachoma in Australian Aboriginal communities . The parameters of the model are informed by the best available Australian and international data ( see S1 Table ) .
The model developed for this study simulates a population of Aboriginal persons within a remote Australian region . Each individual represented in the model is a member of an at-risk community encompassed by the region , and is also a resident of a household within a community ( Fig 1A ) . The temporary migration of individuals ( and potentially other members of their household ) is simulated based on rates of movement between communities [19] . The model is characterised by a five-state natural history structure ( Fig 1B ) . Upon infection with C . trachomatis an individual enters a short latent period where infection load is such that the newly infected individual is not yet infectious , whilst active disease ( trachomatous inflammation follicular/intense: TF/TI ) has yet to develop [20] . An immunopathological response then develops and the infected individual progresses to an infectious state where clinical disease appears [4] . Following the clearance of infection , the inflammatory disease state resolves slowly in the absence of reinfection . Individuals in this state are partially immune to re-infection , but if re-challenged will experience prolonged disease [21] . In the event that no re-infection occurs during this episode , the individual fully recovers to the susceptible state . The duration of each infection and disease state is dependent upon exposure to repeated episodes of infection . Exposure to reinfection is assumed to decrease with age[16] . The transmission of infection and the subsequent development of disease are stochastically determined at the individual-level [22] . That is , the probability of infection transmission between an infectious individual and a susceptible individual is calculated and a random number generated to determine whether transmission will occur at the relevant time point . The probability of transmission between two individuals is assumed to be influenced by the ‘clean face’ status of both the susceptible individual and the infectious individual , where facial cleanliness is assumed to reduce the probability of transmitting and contracting infection [17] . Age-stratified community-level facial cleanliness prevalence data has been recorded across remote Australia as a process of trachoma screening events since 2007 , and is directly entered into the model [23] . The probability of transmission is also affected by the infectiousness of the infected individual , assumed to be proportional to the bacterial load , which in-turn is assumed to be dependent upon the number of previous infections [12] . Two distinct settings for human interaction , and thus transmission potential , are considered in the model . The primary setting for transmission is the household , although transmission can also occur in the wider community . See S1 Text for further details and model equations . The model described was independently calibrated to empirical age-stratified community-level disease prevalence data from three Australian regions through first-order Monte Carlo filtering methods [24] . See S1 Table and S1 File for model parameters and further details of the model and the calibration process . The model source code is also available on-line [25] . Each modelled community was classified as hyperendemic ( ≥ 20% active trachoma disease prevalence in 5–9 year olds ) , mesoendemic ( ≥ 10% but < 20% ) or hypoendemic ( ≥ 5% but < 10% ) dependent on the mean community disease prevalence observed from 2007 to 2011 . The three modelled regions were selected to form a representative sample of trachoma-endemic remote Australia , with selection based on the endemicity of the communities within each region as well as the quantity and quality of surveillance data available . Throughout this paper , the simulated regions are de-identified and referred to as ‘predominantly hyperendemic’ , ‘predominantly mesoendemic’ and ‘predominantly hypoendemic’ based on the prevailing endemicity of the communities within the region . Communities with a consistent 5–9 year old disease prevalence of less than 5% are considered not-at-risk , with 5% also considered the threshold for control [8] . The timing and age-stratified intensity of 2007–2011 screening and treatment events were directly entered into the model according to programmatic monitoring data . The calibrated model was utilised to evaluate the impact of recent intervention efforts . This was achieved by simulating the model in the absence of past intervention efforts such as screening programs , treatment events , housing development initiatives and improvements in facial cleanliness prevalence . A direct comparison was made between the model calibrated to reflect observed conditions and the model output under the hypothetical scenario of no past intervention efforts . The model was then used to project the future impact of a series of potential intervention scenarios . A base-case future intervention scenario was compiled by extrapolating the trends from previously observed trachoma intervention events . The values obtained through this analysis are presented in S2 Table , whilst a description of the current National Guidelines for Trachoma Control in Australia are presented within S3 Table . This base case scenario was then analysed against a series of alternative intervention scenarios . The results obtained from a selection of these alternative scenarios , which are described in Table 1 , are presented in this paper . Each of the future intervention scenarios were simulated until 2020 and the community-level age-stratified prevalence of infection and disease were recorded in each modelled community . The likelihood of controlling trachoma was then calculated as the proportion of model simulations , for each community , in which the control criterion was satisfied by 2020 . To produce representative outputs which accounted for the stochasticity of the model , 1 , 000 simulations were produced using 1 , 000 distinct parameter sets . These parameter sets were sampled from the realistic range of plausible parameter estimates obtained through the model calibration process ( described in S1 File ) . These results were aggregated to form control likelihood estimates for communities of specific endemicity under a given intervention scenario . All numerical computation was performed using MATLAB [26] .
Model-based evaluations of the interventions implemented between 2007 and 2011 suggest that disease prevalence has generally been reduced through trachoma intervention efforts . However , the scale of impact of the past intervention measures was found to vary between regions . The greatest reductions were observed in the predominantly hyperendemic regions , where trachoma prevalence among 5–9 year old children was estimated to have been 23 . 5% ( mean from 1 , 000 simulations , with range 18 . 5%–30 . 7% ) in 2011 in the absence of interventions compared with 14 . 3% ( 10 . 5%–18 . 5% ) with interventions; in the predominantly mesoendemic region , trachoma prevalence was estimated to have reduced from 14 . 8% ( 10 . 3%–19 . 7% ) to 5 . 8% ( 3 . 2%–8 . 0% ) due to intervention efforts ( Fig 2 ) . However , the impact of intervention measures in the predominantly hypoendemic region is more modest: disease prevalence in 2011 was estimated to have reduced from 5 . 1% ( 2 . 2%–8 . 9% ) to 4 . 3% ( 2 . 3%–6 . 5% ) ( Fig 2 ) . This occurs despite a comparable , if not stronger , screening and treatment effort being observed in the hypoendemic region . Indeed , the mean 5–9 year old screening coverage from 2007 to 2011 in the predominantly mesoendemic and predominantly hypoendemic regions are 70 . 5% and 78 . 9% , respectively . The corresponding values for treatment coverage were 87 . 6% and 86 . 1% , respectively . A sensitivity analysis of model input parameters ( see S1 File ) suggests that this finding may be influenced by a higher baseline prevalence of child facial cleanliness in the predominantly hypoendemic region . Future projection simulations estimate the likelihood of achieving trachoma control in hypoendemic and mesoendemic communities by 2020 to be 85% ( 77%–89% ) and 70% ( 60%–79% ) , respectively , should current trachoma intervention efforts be maintained ( Fig 3 ) . However , the likelihood of satisfying the trachoma control criteria in hyperendemic communities under this scenario was calculated to be only 31% ( 19%–43% ) . The estimated likelihoods of controlling trachoma in hypoendemic communities by 2020 were found to be consistently high across each of the considered intervention scenarios . However , large differences were found in control likelihoods in the mesoendemic and , in particular , hyperendemic communities across the modelled future scenarios . This suggests that by optimising the intervention strategy and intensity , the likelihood of achieving trachoma control in highly endemic communities can be greatly increased . Increasing housing construction , and therefore easing the burden of household overcrowding , in addition to maintaining the current intervention strategy increased the estimated likelihood of achieving trachoma control in hyperendemic communities from 31% ( 19%–43% ) to 38% ( 30%–47% ) . Alternatively , assuming a 1 . 5-2-fold reduction in both infectiousness and susceptibility due to facial cleanliness[18] , achieving a consistently high facial cleanliness prevalence ( 90% ) amongst children was found to increase the likelihood of controlling trachoma from these worst-affected communities by 2020 to 64% ( 53%–76% ) . By additionally attaining consistently large screening and treatment coverages , this likelihood of control was estimated to be 75% ( 63%–85% ) . The epidemiological effect of this combination of interventions was found to be greater than the sum of the individual interventions , suggesting that a synergistic effect exists between screening , treatment and attaining high levels of facial cleanliness amongst children . The greatest likelihood for achieving control in the worst-affected communities occurred when these intervention intensities were further coupled with alterations in the treatment strategy; by introducing bi-annual mass drug administration ( MDA ) in hyperendemic communities , the likelihood of achieving trachoma control increased to 86% ( 76%–95% ) , with a corresponding control likelihood of 96% ( 92%–100% ) in mesoendemic communities . This most-effective future scenario was projected out until 2030 , and no rebounding of the epidemic was observed . The resources required to achieve control is an important consideration . Here , crude estimations of the resources required across the predominantly hyperendemic region were calculated for each scenario by the total number of people receiving treatment ( Fig 4 ) . An estimated total of 26 , 088 antibiotic doses are to be distributed between 2012 and 2020 under the current intervention strategy and intensity . This value compares with 12 , 855 treatments under the scenario in which ambitious yet feasible child facial cleanliness prevalence targets were also consistently satisfied . The significantly smaller number of treatments reflects the lower incidence rates attained when a substantially larger proportion of the young population had clean faces . Under the future scenario where ambitious yet feasible screening , treatment and facial cleanliness targets were consistently met , 15 , 312 antibiotic doses would be distributed . Despite this slight increase in antibiotic distribution compared with the previous scenario , the large increase in control likelihood that can be attained by implementing such a control policy and intensity , particularly in hyperendemic communities , makes a solid case for implementing such a control effort . A bi-annual MDA program would result in 25 , 989 antibiotic doses distributed . A large proportion of this total would be distributed within the first three years of implementing the strategy; however , the treatment effort required following this initial peak would decline over time to be less than that required under the current control strategy ( Fig 4 ) . The MDA strategy , with substantially greater control likelihood , emphasises the ‘hit hard , hit early’ principle for greatest effectiveness and cost-effectiveness .
Australia is the only high-income country to have endemic trachoma . Whilst being a signatory to the WHO’s GET 2020 initiative , the Australian government has responded to the health issue in recent years with large investment . However , as several developing countries with histories of trachoma prepare to announce the national control or eradication of the disease [9] , high prevalence levels of trachoma are still observed in remote Australian Aboriginal communities . Since 2006 , Australia has implemented national surveillance activities to collect age-segregated community-level data describing the timing , frequency and intensity of screening and treatment programs as well as disease prevalence , facial cleanliness prevalence , and more recently environmental conditions that may affect trachoma incidence and persistence [3 , 23 , 27] . Increases in community screening and treatment , along with recorded increases in facial cleanliness among children has correlated with declines in trachoma prevalence in Australia . Although our model estimates that current strategies and intensities of programs are unlikely to lead to national control , alternate scenarios appear to be feasible and effective means of achieving this goal . Our results suggest that to achieve control in the worst-affected communities , a more intense intervention effort may be required . For hypoendemic communities ( prevalence 5–10% ) , the model output indicates that continuing the current intervention strategy and intensity will likely be sufficient to control trachoma by 2020 . By assuming that facial hygiene programs can reduce transmission potential over 2-fold , our model estimates that a substantial increase in community facial cleanliness prevalence may be sufficient to control trachoma by 2020 in mesoendemic communities ( prevalence 10–20% ) . The model results indicate that annual screening events are appropriate in mesoendemic communities , whilst the long-term gain of implementing mass drug administration ( MDA ) as opposed to treating only the household contacts of index cases was found to have a negligible impact . The model predicts that the future intervention effort required to considerably raise the likelihood of achieving control in hyperendemic communities ( prevalence >20% ) by 2020 is more difficult . An increase in screening , treatment and facial cleanliness prevalence should be combined with an enhanced housing construction program . Continuous bi-annual MDA is also recommended for three consecutive years before resuming screening events to significantly raise the likelihood of controlling trachoma in these most endemic communities . However , it should also be noted that it may be possible for active disease amongst children to reach below 5% by 2020 even if all trachoma-specific interventions were discontinued immediately . Such decreases are not unusual in far less wealthy regions of the world where programmatic coverage has been poor . But without specific interventions in the past , the rates of trachoma have changed very little in these communities over many years . Thus , we believe concerted and targeted approaches—informed by this analysis—will increase the chance of trachoma control . Previous models of trachoma transmission have described the natural history of trachoma infection and disease using simple population-based systems of differential equations [10–14] . Although useful for extracting general principles , population-based models are limited by the level of complexity that they are able to incorporate . In the context of trachoma control in Australia , the range of complexities one can consider—such as an individual’s age , facial cleanliness status , usual residence and the environmental state of their home community—lends itself to a more flexible model with finer granularity . As such , this study is based on a detailed individual-based simulation model , informed by and calibrated to relatively large amounts of data for remote Australian Aboriginal communities experiencing trachoma . The model developed here attempts to accurately represent the natural history of trachoma infection and disease whilst also assimilating the demographic , cultural and biological factors that influence ocular C . trachomatis transmission . Whilst no model may ever be sophisticated enough to capture all of the heterogeneities involved in the transmission of an infectious disease , the usefulness of the output hinges on the optimality of the model’s balance between complexity and accuracy [22] . Extensive collaboration was sought to ensure that the model described in this paper achieved such a balance . However , it is imperative that the results presented must be consumed with perspective . Other limitations that should be addressed when assessing the validity of modelling results regard the empirical data that are used to inform the model parameters . For the purpose of this research , a large volume of nationally collated surveillance data was employed but these data are also potentially limited in completeness and representativeness . There exists a certain degree of uncertainty in the estimates of trachoma prevalence , particularly in the early years of data collection , as there was some degree of variation in screening coverage rates . However these coverages have progressively improved along with the accuracy of trachoma grading . Australian treatment coverage data can be difficult to interpret as the method of distribution has varied and has not always been clearly specified; these methods have included MDA , household contact based treatment and treating only affected children . Equally , the definition of the denominator and hence the coverage achieved have also varied somewhat . The model also assumes a steady-state equilibrium at baseline , which may be inaccurate as a result of previous trachoma treatment efforts . Although these are important limitations and have an impact on the precision of the forward estimates of effectiveness of the interventions , they do not influence the comparisons of the relative effectiveness of the different strategies . This is a strength of this research as it shows the greater effectiveness of a combined more intensive strategy compared with that currently employed . The world has a goal of eliminating blinding trachoma as a public health concern . The countries which have done so or are on the verge of announcing such success should be commended for their excellent public health efforts . However , current strategies may not be sufficient in other contexts and as we have demonstrated in this study , they may be insufficient in the trachoma-endemic country with the greatest amount of resources . Through detailed simulation modelling we have suggested some slight shifts in strategies and changes in intensities in the short-term which have the potential to yield substantial returns in the future in order to achieve this ultimate goal . | Australia is the only remaining high-income country reporting endemic levels of trachoma , with infections occurring predominantly within rural and remote Indigenous communities . Although the Australian government has recently invested large sums of money to combat the disease , it remains unclear whether the national goal of controlling trachoma by 2020 will be achieved . Here , we use a novel individual-based simulation model to estimate the impact of numerous potential future invention strategies and intensities . Our model is the most sophisticated trachoma transmission model to date , and the first to specifically represent trachoma in Australian Indigenous communities . Model projections suggest that although the current intervention strategy and intensity are unlikely to achieve the target of national control by 2020 , the likelihood of achieving this goal can be significantly increased by shifting the intervention strategy and increasing the intensity of key intervention components such as screening , treatment and facial cleanliness activities . Our findings that the most resource rich country with endemic trachoma may require a more intensive intervention effort to control the disease suggest that challenges may remain in the fight for the global control and eventual elimination and eradication of trachoma . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Control of Trachoma in Australia: A Model Based Evaluation of Current Interventions |
Cutaneous leishmaniasis ( CL ) is a skin disease caused by the protozoan parasite Leishmania . Few studies have assessed the influence of the sample collection site within the ulcer and the sampling method on the sensitivity of parasitological and molecular diagnostic techniques for CL . Sensitivity of the technique can be dependent upon the load and distribution of Leishmania amastigotes in the lesion . We applied a quantitative real-time PCR ( qPCR ) assay for Leishmania ( Viannia ) minicircle kinetoplast DNA ( kDNA ) detection and parasite load quantification in biopsy and scraping samples obtained from 3 sites within each ulcer ( border , base , and center ) as well as in cytology brush specimens taken from the ulcer base and center . A total of 248 lesion samples from 31 patients with laboratory confirmed CL of recent onset ( ≤3 months ) were evaluated . The kDNA-qPCR detected Leishmania DNA in 97 . 6% ( 242/248 ) of the examined samples . Median parasite loads were significantly higher in the ulcer base and center than in the border in biopsies ( P<0 . 0001 ) and scrapings ( P = 0 . 0002 ) . There was no significant difference in parasite load between the ulcer base and center ( P = 0 . 80 , 0 . 43 , and 0 . 07 for biopsy , scraping , and cytology brush specimens , respectively ) . The parasite load varied significantly by sampling method: in the ulcer base and center , the descending order for the parasite load levels in samples was: cytology brushes , scrapings , and biopsies ( P<0 . 0001 ) ; in the ulcer border , scrapings had higher parasite load than biopsies ( P<0 . 0001 ) . There was no difference in parasite load according to L . braziliensis and L . peruviana infections ( P = 0 . 4 ) . Our results suggest an uneven distribution of Leishmania amastigotes in acute CL ulcers , with higher parasite loads in the ulcer base and center , which has implications for bedside collection of diagnostic specimens . The use of scrapings and cytology brushes is recommended instead of the more invasive biopsy .
Cutaneous leishmaniasis ( CL ) is a parasitic disease of significant public health problem in at least 18 countries of Latin America; about 67 , 000 CL cases were reported to occur annually in the last decade [1] . The disease is caused by protozoan parasites of the subgenera Leishmania ( Viannia ) and L . ( Leishmania ) , with the former being responsible for most cases . The clinical phenotypes of CL are diverse and range from a single or few cutaneous ulcerative lesions at the site of infection that may heal spontaneously , diffuse and disseminated CL with multiple non-ulcerative lesions , to disfiguring mucocutaneous leishmaniasis that can be life-threatening [2 , 3] . The severity and outcome of the disease are dependent among others on the immune responses evoked by the host and the infecting Leishmania species [4 , 5] . Parasitological diagnosis of CL relies on either the microscopic demonstration of Leishmania amastigotes in cutaneous tissue or the isolation of parasites from lesions in culture [6–8] . While these techniques are highly specific for diagnosing CL , they are insufficiently sensitive [9] . Polymerase chain reaction ( PCR ) -based testing of skin lesion specimens has become an important tool to diagnose CL , because of its high sensitivity and specificity ( up to 100% ) [10–12] . Significant progress has been made towards evaluating molecular-based non-invasive methods for the diagnosis of CL that overcome the disadvantages of the traditional , invasive sampling methods such as punch biopsies , aspirates or skin slits/scrapings [13–16] . One such non-invasive method , cytology brush PCR , has shown potential for widespread use , both in the clinic [15] and field settings [17] . Several studies indicate that the sensitivity of diagnostic methods for CL can be dependent upon the number and dispersion of parasites in the lesion , the method used to sample ulcers , the stage ( chronicity ) of the lesion , and the technical skills of the personnel [6 , 9 , 10 , 11 , 18 , 19] . Conventionally , in accordance to guidelines established by the World Health Organization ( WHO ) [20] , tissue samples have been obtained from the lesion border , where parasite load and the density of inflammatory mononuclear cells harboring parasites are thought to be higher [21] . Evidences supporting that other sampling sites within lesions could result in comparable or even increased sensitivity of parasite detection by microscopy or PCR have been provided [22 , 23] . In a recent study using quantitative real-time PCR ( qPCR ) targeting Leishmania 18S rDNA , it has been reported that swab sampling over the ulcer allowed to recover a higher amount of parasite DNA as compared to aspirate samples taken from the lesion border [16] . Whether this indicates differences inherent to the sampling methods or truly reflects a higher parasite load in the ulcerated zone of the lesion [16] needs to be ascertained . The analysis of the load and distribution of Leishmania parasites within the skin lesions would be important not only for determining the best location within the ulcer to obtain samples for diagnostic purposes , but also for an eventual follow up of a patient’s response to treatment [24 , 25] . We herein applied a standardized qPCR assay targeting minicircle kinetoplast DNA ( kDNA ) [26] to detect and quantify Leishmania ( Viannia ) parasites in 3 sites within the CL ulcer ( raised border , base , and center ) . Paired lesion samples were collected by use of different sampling methods: a punch biopsy and a dermal scraping from each of the 3 lesion sites , and a cytology brush from each the base and center of the ulcer . The parasite load levels were compared according to the ulcer site , sampling method , and the infecting Leishmania species . We restricted this study to lesions originated from patients with acute CL ( ≤3 months ) , which characteristically have high parasite load in contrast to lesions from patients with chronic disease ( >6 months ) [10 , 26] . This fact enabled detection of Leishmania and quantifiable parasite load levels in most clinical specimens . Importantly , early diagnosis is considered a desirable control measure for CL . To our knowledge , this is the first report that quantitatively compares the parasite loads among different skin lesion sites and sampling methods by means of qPCR , thereby providing an insight into the likely distribution of Leishmania amastigotes in the ulcer for the L . ( Viannia ) species present in our sample set . The implications of our results on diagnosis of CL and the prognostic applicability are discussed , as well as how they may relate to the immunopathology of the disease .
This study was conducted according to the principles specified in the Declaration of Helsinki and under local ethical guidelines ( Universidad Peruana Cayetano Heredia Institutional Review Board ) . The study protocol , informed consent and sampling procedures were approved by the Institutional Review Boards of the Hospital Nacional Cayetano Heredia and Universidad Peruana Cayetano Heredia ( Lima , Peru ) for studies involving human subjects . Written informed consent was obtained from all participants prior to enrolment . The sample size was calculated using the G*Power 3 . 1 software ( release 3 . 1 . 9 . 2; available from: http://www . gpower . hhu . de/ ) [27] to assess the null hypothesis of no difference in parasite load levels between different sampling sites within the CL ulcer . Assuming a medium effect size of 0 . 5 , a significance level of 5% , and a power of 80% , 35 matched pairs of lesion samples were required to be examined ( two-sided , Wilcoxon signed-rank test for matched pairs ) . For significant results , the effect size was assumed to be ‘medium’ , which means an effect visible to the naked eye . Non-significant results were assumed to have a ‘small’ effect size . We managed to study 31 paired lesion samples from patients presenting with acute CL . Patients that attended the Leishmaniasis Clinic at the Instituto de Medicina Tropical Alexander von Humboldt , Hospital Nacional Cayetano Heredia , in Lima , Peru , between January and June 2013 for the examination of skin lesions were invited to participate in the study and evaluated for possible eligibility . Patients were considered for enrolment if they presented with ulcerative skin lesions of recent onset ( ≤3 months of evolution ) , with elevated and infiltrative borders and a lesion size over 1 cm in diameter; and were able to give written informed consent for the sampling procedures . We included adult patients with laboratory confirmed diagnosis of CL , as defined by a positive result on at least 1 of these 3 tests: direct microscopy on Giemsa-stained lesion smears [7] , lesion aspirate microculture [28] , and qualitative PCR targeting kDNA minicircles [29] on a biopsy specimen obtained from the ulcer border . This diagnostic PCR includes internal control primers for amplifying the human beta-globin gene as previously described [13] . The intradermal leishmanin skin test ( LST ) , used to assess exposure to Leishmania infection , was performed on CL patients before treatment , as described elsewhere [30 , 31] . We excluded patients allergic to local anesthetics , with clinical evidence of bacterial or fungal superinfection of the ulcer ( when possible ) , with any contraindication to skin biopsy and those undergoing active treatment for CL . In three cases with secondarily infected ulcers , patients were treated with a 5-day course of antibiotics before sample collection . In order to analyze the distribution and load of Leishmania amastigotes within the cutaneous lesion , samples were collected from 3 different sites , in the following order: the center of the ulcer , the base ( inner border ) of the ulcer , and the raised border of the ulcer ( Fig 1 ) ; using a randomly chosen coordinate defined as North , South , East or West , taking as reference the lateral and longitudinal axes of the human body . If the patient had more than one lesion , the most active and typical indurated ulcer was selected . Eight specimens were collected from a single lesion per patient: a punch biopsy and a dermal scraping from each of the 3 lesion sites , and a cytology brush from each the center and base of the ulcer . The order of sampling was: biopsy , scraping , and cytology brush . All samples were taken by the same physician in order to avoid inter-individual variation . Prior to sampling , lesions were cleansed with topical antiseptics , removed from any overlying scab or crust with saline solution and anesthetized with 1 cc of lidocaine 1% . Prior to DNA extraction , samples were centrifuged at 8000 g for 2 min and ethanol was discarded . Biopsied tissue was disaggregated with a sterile scalpel . Disaggregated tissue , lancets and cytology brushes were subjected to overnight lysis with Proteinase K and processed for DNA isolation using a column-based method ( High Pure PCR template preparation kit , Roche ) , according to the manufacturer’s instructions . The isolated DNA was then quantified by fluorometry using the Quant–iT Broad Range dsDNA Assay kit ( for biopsies ) and the Quant-iT High Sensitivity dsDNA Assay kit ( for scrapings and cytology brushes ) on the Qubit fluorometer ( Invitrogen ) . DNA samples were diluted to 5 ng/μL; those samples below this concentration were added directly into the PCR reaction . Parasites were typed using the heat-shock protein 70 gene ( hsp70 ) PCR-N variant followed by restriction fragment length polymorphism ( RFLP ) analysis using the restriction enzymes BsaJI and RsaI as in Montalvo et al . [33] . We applied a SYBR Green-based qPCR assay targeting kDNA minicircles to detect and quantify Leishmania ( Viannia ) parasites in clinical samples , as previously described [26] . Each kDNA-qPCR run included a standard curve of L . ( V . ) braziliensis ( MHOM/BR/75/M2904 ) DNA ranging from 5 × 104 to 5 × 10−3 parasite DNA equivalents/reaction ( run in duplicate ) ; a positive control with known amount of Leishmania parasites , which consisted of a mix of Leishmania DNA and human genomic DNA in order to mimic clinical specimens ( run in triplicate ) ; a negative control ( human genomic DNA from peripheral blood mononuclear cells of a healthy donor; run in triplicate ) ; and a blank ( no-template control; run in triplicate ) . The standard curves ( inter-assay reproducibility , n = 11 ) showed a mean square error ( MSE ) of ≤0 . 111 , correlation coefficient ( r2 ) of ≥0 . 998 and slopes of 3 . 28 ( mean ) ± 0 . 05 ( standard deviation ) , indicating a high amplification efficiency ( ≥1 . 99 ) ( 2 would indicate 100% PCR efficiency ) . The positive control showed a mean of 7 , 640 parasites and an inter-assay coefficient of variation of 7 . 8% ( n = 11 independent runs ) . All clinical samples were run in duplicate; if replicates differed by a standard deviation of >0 . 35 in Cq ( quantification cycle ) values ( >0 . 5 cycles ) , they were retested . A sample was quantified when it had a Cq value falling within the range of the standard curve . The highest dilution of template of the standard curve was defined as the lower limit of quantification ( LOQ ) . Samples with Leishmania DNA levels below the LOQ could be detected; they were considered positive ( qualitative detection ) only if their melting curves had the same profile as those of the standards included in the same experiment . The Leishmania parasite load was calculated as follows: [parasite DNA equivalents per reaction/amount of tissue DNA per reaction] × 103 , expressed as the number of Leishmania parasites per μg of tissue DNA . Frequencies and proportions were used to describe categorical variables while median and interquartile range or mean and standard deviation were used for numeric continuous variables . To assess whether the median parasite load in clinical specimens differed significantly according to the skin lesion site or the sampling method , analyses for paired samples using Friedman ( with Dunn’s post-hoc test ) and Wilcoxon signed rank tests were performed . The correlation degree between the parasite load measurements in scraping and cytology brush specimens with respect to those in biopsy specimens was calculated using the Spearman’s rank correlation test . The association between the Leishmania load and the parasite species was evaluated using the Mann-Whitney U test . Statistical analyses were performed under a 5% significance level , using the GraphPad Prism v5 . 02 software .
Demographic , epidemiological , and clinical characteristics of patients are summarized in S1 Table . Thirty-one patients with laboratory confirmed CL were enrolled: 29 ( 93 . 6% ) men and 2 ( 6 . 5% ) women , with median age of 34 years ( range 19–75 years ) and median disease duration of 2 months ( range 1–3 months ) . Median number of lesions was 1 ( range 1–10 ) , with 21 patients ( 67 . 7% ) presenting with single lesions and 9 patients ( 29% ) presenting with multiple lesions . Bacterial superinfection was present in only 3 ( 9 . 7% ) lesions . Twenty-eight patients ( 90 . 3% ) had a first episode of CL and only one patient ( 3 . 2% ) had a reinfection . Median duration of exposure in the risk area ( i . e . stay in area of endemicity ) was 3 months ( range 1 . 5 days–75 years ) . The kDNA-qPCR assay detected Leishmania DNA in 97 . 6% ( 242/248 ) of the examined lesion specimens . The overall qPCR positivity per lesion-analysis taking into account the 3 lesion sites and sampling methods ( 96 . 8%; 95% CI: 74 . 3–100 . 0% ) was higher than that of smear microscopy ( 80 . 6%; 95% CI: 62 . 5–92 . 6% ) , microculture ( 88 . 5%; 95% CI: 69 . 9–97 . 6% ) , and LST ( 72 . 4%; 95% CI: 52 . 8–87 . 3% ) , whereas it was comparable to that of the qualitative kDNA PCR ( 96 . 8%; 95% CI: 83 . 3–99 . 9% ) . The performance of the qPCR for Leishmania DNA detection ( no quantification at this stage ) was assessed in the 3 lesion sites and sampling methods . In the ulcer border , Leishmania DNA was detected by qPCR in 100% ( 31/31 ) of the scrapings and in 90% ( 28/31 ) of the biopsies . In the ulcer base , 100% ( 31/31 ) of the biopsy specimens , 97% ( 30/31 ) of the scraping specimens , and 97% ( 30/31 ) of the cytology brush specimens tested positive for Leishmania DNA . In the ulcer center , Leishmania DNA was detected in 100% ( 31/31 ) of the examined biopsies , in 100% ( 31/31 ) of the scrapings , and in 97% ( 30/31 ) of the cytology brushes . The kDNA-qPCR assay allowed the quantification of the parasite load in 238 out of 248 lesion specimens ( 96% ) . As for the 10 specimens that could not be quantified , 4 corresponded to dermal scrapings with detectable but not quantifiable parasite load , whereas 6 specimens among biopsies , scrapings and cytology brushes were qPCR negative . These 10 specimens corresponded to 3 patients . After exclusion of those 3 patients from the analysis , the quantified paired parasite load results ( 8 specimens per lesion ) corresponding to 28 patients were used for parasite load assessment across ulcer sites and sampling methods . The parasite load ( PL ) in skin lesion specimens varied from 2 . 53 × 101 to 5 . 72 × 106 parasites per μg of tissue DNA . The PL levels per skin lesion site and sampling method are given in Table 1 and depicted graphically in Fig 2 . Causative species was identified in 29 of 31 ( 93 . 5% ) patients having lesion specimens with sufficient amplifiable DNA: 20 patients were infected with L . ( V . ) braziliensis , 7 with L . ( V . ) peruviana , 1 with L . ( V . ) guyanensis , and 1 with L . ( V . ) lainsoni . There was no significant difference in PL according to the infecting species , taking into account in this analysis only the most well represented species ( i . e . L . ( V . ) braziliensis and L . ( V . ) peruviana ) ( P = 0 . 4 , Mann-Whitney U test ) ( Table 2 ) .
Here we assessed whether the Leishmania ( Viannia ) parasite load differs by sampling site within CL ulcers and sampling method by means of qPCR . We observed that a significantly lower amount of parasites was quantified from the ulcer border as compared to the ulcer base and center . The fact that this finding was similarly observed with lesion biopsy and scraping specimens points out its robustness . This finding called our attention because most of available studies in the literature on skin lesions caused by Neotropical Leishmania parasites are based on biopsies collected from the border of the ulcer , in accordance with WHO recommendations [20] , as this lesion site is regarded to likely concentrate a greater amount of parasites and viable infected mononuclear phagocytes compared to the necrotic center of the ulcer [21] . Nevertheless , consistent with our findings , in a recent study that evaluated the use of swab sampling over the ulcer coupled to qPCR for diagnosis of CL , Adams et al . [16] found indications of a greater quantity of parasite DNA in the ulcerated zone of the lesion ( as compared to that found in the ulcer border using aspirate samples ) . Furthermore , Ramírez et al . [23] reported a significant increase in the sensitivities of microscopy and conventional PCR of dermal scrapings when samples were collected from the central region of the bottom of the ulcer rather than from the margin of the lesion . This appeared to be related to a higher parasite load and easily detectable amastigotes in that area [23] . The differences in parasite load among sites within CL ulcers revealed herein may be related to the undergoing immunopathological process within the ulcer . Studies of lesion biopsies ( taken , where known , from the border of the lesion ) from patients infected with L . ( Viannia ) parasites have shown an inflammatory infiltrate in the dermis composed mainly of lymphocytes , macrophages , and plasma cells [34–38] . Leishmania amastigotes were seen within dermal macrophages located in the papillary dermis [35] and mid-dermis [37] . Notably , Gutierrez et al . [38] found a significant association between necrosis , relative abundance of tissue macrophages , and the presence of amastigotes in lesions of less than 6 months’ duration . The ulcerated zone of the lesion is mainly composed of dead cells , as shown by the presence of focal necrosis of the dermis as well as epidermal disruption [34–37] . In contrast , in the raised border adjacent to the ulceration , the epidermis exhibits hyperplasia and thickening [36 , 38] . Despite the fact that the quantified parasite loads varied widely among examined lesion samples , we found that there was no significant difference in parasite loads with respect to L . ( V . ) braziliensis and L . ( V . ) peruviana infections demonstrated in CL lesions , a finding consistent with a previous study [26] . As these 2 Leishmania species can lead to different clinical prognoses [39] , our data lends support to a lack of association between parasite load and the degree of pathology noted in studies on human [26] and experimental murine [40] CL . Thus , differences in pathogenicity should rely on other aspects of the host-parasite interaction . Remarkably , compared to biopsies , we found that a higher parasite load could be quantified from cytology brushes and dermal scrapings . There are two possible explanations for this finding . First , the difference in parasite loads quantified amongst sampling methods pointed out that the Leishmania amastigotes would not be homogeneously distributed along the skin compartments ( i . e . cross-sectional ) , with a higher abundance of parasites to be found in the upper layers of the dermis . This is corroborated by other studies that showed that the cell types in the CL lesion infiltrates were non-randomly distributed , with macrophages and parasites being most frequently found in the mid-dermis [37] as well as in the papillary dermis [35] . Second , there are differences inherent to the sampling instrument . Skin punch biopsy is the only method that allows to sample the full-thickness skin; therefore , the ratio of human host DNA to parasite DNA in this diagnostic specimen is several-fold higher compared to scrapings and cytology brushes . This can decrease the sensitivity of detection of the pathogen in clinical samples [19 , 41 , 42] . In contrast , scraping and cytology brush sampling is more superficial and allows to recover both cellular material and tissue fluid [15]; these features reduce the proportion of host cells resulting in improved parasite DNA representation relative to the human host DNA . Our finding that scrapings and cytology brushes outperform the invasive biopsy in terms of the parasite load quantified is particularly important when considering that invasive specimen collection is a traumatic procedure to the patient frequently associated with risks of bleeding and infection; it entails the risk of body fluid exposure to the healthcare worker via needle stick injury , and is difficult to perform in the pediatric population [13 , 15 , 43 , 44] . Furthermore , it is a complex medical procedure performed by trained medical personnel , normally a dermatologist , and is difficult to perform routinely in endemic settings . Cytology brushes offer the advantage over biopsies and scrapings of being non-invasive , easy to perform and well tolerated by the patient [15 , 44] . This makes them an attractive alternative not only for diagnosis of CL but also for monitoring patients’ response to treatment ( through assessment of parasite load kinetics ) . Such an applicability of cytology brush sampling coupled to kDNA-qPCR has been recently evaluated in a cohort of Peruvian patients with mucosal leishmaniasis [45] . From a practical point of view , our data herein also indicate that samples for routine laboratory diagnosis or an eventual post-treatment follow-up of CL patients can be easily and safely obtained from the ulcer base and center by use of less invasive means , thus obviating the need for any skin incision from the lesion border . The qPCR assay employed herein targets a multicopy conserved region of minicircle kDNA common to Leishmania ( Viannia ) species , present at about 10 , 000 copies per amastigote [46] , thereby allowing to quantify the number of parasites present in the ulcer with high sensitivity . The Leishmania kDNA levels detected and quantified in the lesion specimens are likely indicative of the presence of viable parasites , since nuclear and kinetoplast Leishmania DNA are rapidly degraded following amastigote death [47] . The variability of the number of kDNA minicircle targets [26 , 48] was not assessed in the lesion specimens examined here . Nonetheless , our data analysis took into account paired samples ( those taken from a same lesion of a patient ) , which allowed a more accurate estimation of parasite load levels in the ulcer across subjects . In this study , only patients with recent onset CL ( ≤3 months ) were enrolled . That early stage of CL is associated with a positive parasitological diagnosis and a high parasite load; conversely , chronic CL ( >6 months ) is characterized by a scarcity of parasites in lesions [10 , 23 , 26 , 38] . Thus , further studies covering both acute and chronic stages of CL caused by L . ( Viannia ) parasites are needed to confirm and expand our results . Nevertheless , this study is valuable as it is , to our knowledge , the first report that assesses quantitatively whether the Leishmania parasite load differs by both site of sample collection within the skin ulcer and sampling method by means of qPCR . Our results herein are applicable to ulcerative skin lesions , which represent the most frequent form of localized CL in Latin America . In areas where leishmaniasis is endemic , a smaller number of patients present with other types of cutaneous manifestations ( nodular , verrucous , plaques , and papular lesions ) , either as primarily presentation or in addition to the ulcerative lesion [13 , 49 , 50] . Future studies assessing the parasite load in these other types of lesions covering different stages of the disease will add to our understanding of a polymorphic skin disease as CL is . Our data reveal a picture of the CL ulcer being a complex place , where the process of survival of Leishmania amastigotes is occurring , with abundant amastigotes in a highly necrotic tissue . Future studies based on morphometric analysis of histopathological sections are needed to establish the in situ location and quantity of parasites in relation to cellular infiltrates in the ulcerated zone of the lesion , and during different stages of the disease . This may further our understanding of the dynamics of infection in human CL due to L . ( Viannia ) species . | Cutaneous leishmaniasis ( CL ) is a parasitic disease of the skin caused by obligate intra-macrophage protozoa of the genus Leishmania which usually presents as ulcerative lesions at the site of infection . Traditionally , histopathological and diagnostic studies on CL have employed samples collected from the border of the ulcer since this area is believed to contain the highest amount of parasites . However , no formal demonstration of the distribution of Leishmania parasites in the ulcer has been provided yet . Focusing on human skin lesions of recent onset ( ≤3 months ) caused by L . ( Viannia ) species , we estimated the parasite loads among different skin lesion sites by means of quantitative real-time PCR targeting the parasite kinetoplast DNA . Paired lesion samples collected by use of different sampling methods were analyzed . We found that the ulcerated zone of the lesion contained a higher parasite load than the ulcer border , and that scraping and cytology brush specimens presented higher parasite loads as compared to the more invasive biopsy . Our results have implications for bedside collection of diagnostic and post-therapeutic follow-up specimens from CL patients . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Quantification of Leishmania (Viannia) Kinetoplast DNA in Ulcers of Cutaneous Leishmaniasis Reveals Inter-site and Inter-sampling Variability in Parasite Load |
Forkhead box O ( FOXO ) transcription factors have a conserved function in regulating metazoan lifespan . A key function in this process involves the regulation of the cell cycle and stress responses including free radical scavenging . We employed yeast chronological and replicative lifespan assays , as well as oxidative stress assays , to explore the potential evolutionary conservation of function between the FOXOs and the yeast forkhead box transcription factors FKH1 and FKH2 . We report that the deletion of both FKH genes impedes normal lifespan and stress resistance , particularly in stationary phase cells , which are non-responsive to caloric restriction . Conversely , increased expression of the FKHs leads to extended lifespan and improved stress response . Here we show the Anaphase-Promoting Complex ( APC ) genetically interacts with the Fkh pathway , likely working in a linear pathway under normal conditions , as fkh1Δ fkh2Δ post-mitotic survival is epistatic to that observed in apc5CA mutants . However , under stress conditions , post-mitotic survival is dramatically impaired in apc5CA fkh1Δ fkh2Δ , while increased expression of either FKH rescues APC mutant growth defects . This study establishes the FKHs role as evolutionarily conserved regulators of lifespan in yeast and identifies the APC as a novel component of this mechanism under certain conditions , likely through combined regulation of stress response , genomic stability , and cell cycle regulation .
The evolutionarily conserved insulin-signaling pathway plays a critical role in multiple cellular processes [1]–[4] . Perhaps most important is the decisive role it plays in cellular , and organismal , survival . This pathway must be tightly regulated , as overactive insulin-signaling leads to increased survival of cells that would otherwise be shunted down the apoptotic pathway . This occurs by increased repression of stress response , pro-apoptotic , and DNA repair genes , thereby increasing the proliferative capacity , and oncogenic potential of these cells . Although the lifespan of damaged cells is increased under these conditions , this situation increases the probability that the organism will die prematurely due to disease onset . On the other hand , reduced insulin-signaling relieves repression of the stress response , cell cycle arrest and DNA repair pathways , increasing cell maintenance capacity and survival . It is now clear that there is a link between diabetes and cancer [5]–[7] , diseases associated with the insulin-signaling pathways , highlighting the importance of understanding the precise activity of this pathway . The insulin-signaling phosphorylation cascade activates AKT , which targets cellular factors that switch metabolism from catabolic to anabolic reactions , favoring growth and reproduction over maintenance and repair [8] . Major AKT targets include the forkhead box O family ( FOXO ) transcription factors ( reviewed in [9]–[14] ) . The FOXOs are believed to serve diverse rolls in longevity determination and tumor suppression in metazoans from nematodes to humans . The FOXOs integrate signals from energy , growth factor and stress signaling cascades to regulate cell differentiation , cell-cycle progression , apoptosis , autophagy , DNA-damage repair , and scavenging reactive oxygen species . FOXO proteins have been shown to interact with multiple cofactors that mediate their activity through posttranslational modifications . Phosphorylation , ubiquitination , methylation , and acetylation regulate transcription factor efficiency and nuclear shuttling ( reviewed in [9] , [10] ) . Specifically , nutrient ( insulin ) and growth factor signals lead to cytosolic sequestering and ubiquitination of the FOXOs , targeting them for degradation; conversely , internal reactive oxygen species ( ROS ) , DNA damage sensing and starvation signals can cause nuclear shuttling , and transcription factor activity . Thus , a dynamic and complex molecular network controls FOXO protein function , yet specific downstream targets remains speculative . Saccharomyces cerevisiae is often utilized to elucidate regulation of fundamental eukaryotic mechanisms . However , the individual deletion of any of the four Forkhead box orthologs does not affect lifespan [15] , suggesting a lack of functional conservation . However , two of the orthologs , FKH1 and FKH2 , show genetic redundancy , as deletion of both genes is necessary to alter growth , cell morphology and gene transcription phenotypes [16]–[20] . Further evolutionary conservation for FKH1 and FKH2 is suggested by their requirement for ROS-induced cell cycle arrest [18] , and cell cycle regulation through the regulation of both G1 and G2/M gene clusters [17] , hallmarks of the human FOXO genes . The Fkh1/2 regulated CLB2 gene cluster [17] encodes genes required for Anaphase-Promoting Complex ( APC ) activity ( APC1 , CDC5 , CLB2 , and CDC20 ) , as well as APC targets ( CLB2 , CDC5 , CDC20 and IQG1 ) [21]–[23] . The APC is a highly conserved multi-subunit ubiquitin-protein ligase ( E3 ) that promotes mitotic progression and G1 maintenance by targeting cell cycle regulators , such as the securin Pds1 and the cyclin Clb2 , for proteasome-dependent degradation [21] , [23] , [24] . The APC has been demonstrated to be critical for regulating genomic stability , and longevity in yeast and higher eukaryotic organisms [25]–[30] . In yeast , mutation to multiple APC subunits decreases replicative lifespan ( RLS; measures mitotic longevity ) and chronological lifespan ( CLS; measures post-mitotic survival ) , while over-expression of APC10 increases RLS [29] . In mice , mutations to the APC regulator BubR1 , a component of the spindle checkpoint , lead to premature aging defects [27] , [31] . Consistent with this , we and others have provided evidence that the yeast APC plays a role in stress response by possibly targeting proteins that block proper stress response for degradation [28] , [32] , [33] . Our data supports a model where FKH1 and FKH2 function is evolutionarily conserved with higher eukaryotic FOXO proteins with regards to lifespan and oxidative stress resistance . We show that the FKHs are required for increased stress resistance and survival in response to severe caloric restriction ( cultures maintained in water ) . Importantly , we identify the APC as a potential target of the FKHs under normal conditions , while functioning cooperatively under stress conditions .
FOXO transcription factors regulate processes involved in the homeostasis of metazoan cells and tissues with the net outcome being lifespan extension and tumor suppression , yet many of the downstream targets remain unknown [9]–[14] . The budding yeast Saccharomyces cerevisiae is a powerful tool used to elucidate genetic and molecular mechanisms mediating many cellular processes , but independent deletion of the four yeast forkhead box protein encoding genes ( FKH1 , FKH2 , HCM1 , FHL1 ) does not alter CLS [15] . However , Fkh1 and Fkh2 have been shown to be phenotypically redundant , as they are required for M/G1 progression and cell cycle arrest in response to hydrogen peroxide [18] . These characteristics lead us to examine the role of both Fkh1 and Fkh2 in the regulation of yeast lifespan . We investigated the FKHs using the RLS assay , a measure of the mitotic lifespan of individual cells , finding that deletion of either individual FKH gene had no effect on RLS , as reported for CLS [15] . Double mutant cells could not be assayed using RLS due to their flocculent phenotype ( data not shown ) . Therefore , we investigated the potential of the FKHs in regulating CLS , a measure of metabolic activity in post-mitotic stationary phase cells [34] . In cultures maintained in depleted complete media ( DM ) , we also observed that single deletion of the FKH genes did not impair CLS . Deletion of both FKH1 and FKH2 , on the other hand , reduced CLS ( Figure 1A ) , with mean ( 50% ) survival reached by day 8 . 5 for WT cultures , day 11 for fkh2Δ cultures , day 8 for fkh1Δ cultures , and only day 4 for fkh1Δ fkh2Δ cultures . Controversy exists as to whether higher eukaryotic FOXOs , downstream targets of nutrient/insulin signaling , are contributing factors to caloric restriction-induced lifespan extension [35]–[39] . To examine whether the yeast FKHs play a role in caloric restriction , we examined the CLS of the FKH mutants by maintaining the post-mitotic cultures in distilled H2O . Water is believed to act as a form of severe caloric restriction ( SCR ) that simulates the low-nutrient environment that yeast in the wild would most likely encounter [40] , [41] . Maintenance in H2O extended the mean survival of WT , fkh1Δ and fkh2Δ cultures to 19–21 days , while little change was observed in fkh1Δ fkh2Δ cultures with a mean survival of 5 days ( Figure 1B ) . The lack of response in fkh1Δ fkh2Δ cultures maintained in H2O suggests that Fkh1 and Fkh2 play a redundant role in SCR-induced lifespan extension . Although Fkh1 and Fkh2 have not previously been associated with longevity in yeast , they have been linked with stress response in mitotically active cells [18] , which is associated with an evolutionarily conserved role in long lifespan [42]–[44] . To address whether the Fkhs' role in longevity is a manifestation of their involvement in stress resistance in post-mitotic cells , we treated WT and fkh1Δ fkh2Δ day 5 stationary phase cells maintained in either H2O or DM with 100 mM hydrogen peroxide ( H2O2 ) for 1 hour ( Figure 1C ) . WT day 5 stationary phase cultures exhibited increased resistance to H2O2 when maintained in water compared to DM . However , this effect was nullified in fkh1Δ fkh2Δ cultures , indicating that the Fkh proteins are required for stress resistance during stationary phase . A plate assay confirmed that stationary phase cells exhibit increased stress response compared to mitotically active cells , and that deletion of FKH1 and FKH2 diminishes this effect ( Figure 1D ) . To further assess the role of the FKHs in normal and stressed post-mitotic lifespan , the endogenous genes encoding Fkh1 and Fkh2 were C-terminally TAP ( tandem affinity purification ) -tagged , with protein levels analyzed as cells aged in DM media and H2O ( Figure 2A ) . Fkh1-TAP was indeed expressed as cells aged during stationary phase in DM and H2O . Fkh2-TAP was also observed in aging stationary phase cells , but at much lower levels ( data not shown ) . Fkh1-TAP levels appear slightly lower in day 5 stationary phase cells compared to day 1 , with very little difference between H2O and DM . Nonetheless , Fkh1 and Fkh2 proteins are expressed in post-mitotic cells . Next , endogenous C-terminal GFP-tagged Fkh1 and Fkh2 were analyzed in aging cells to determine cellular localization . In day 5 stationary phase cultures maintained in DM , GFP fluorescence was observed to be nuclear in many cells ( Figure 2B ) . When Fkh1-GFP and Fkh2-GFP were monitored in progressively aging cells in DM and in H2O , we observed that a larger proportion of the Fkh protein remained nuclear in H2O compared to DM . An example is shown in Figure 2C , where day 13 stationary phase cells appear healthier , with a larger proportion of nuclear Fkh2-GFP when maintained in H2O ( Figure 2C ) . The percentage of cells harboring nuclear Fkh1-GFP or nuclear Fkh2-GPF was consistently observed to be greater when the cells were maintained in H2O compared to DM ( Figure 2D ) . This suggests the presence and nuclear localization of the Fkhs may be necessary for normal CLS and stress resistance , especially in an SCR environment . In higher eukaryotic systems increased expression of FOXO orthologues is associated with increased longevity and stress resistance [38] , [45] . Thus , we predicted that an increase in survival would be expected with the overexpression of FKH1 and/or FKH2 . Increased FKH expression was accomplished by integrating the GAL1/10 inducible promoter immediately upstream of the FKH1-TAP and FKH2-TAP start sites ( Figure 3A ) . Cells overexpressing both FKH1 and FKH2 were created by crossing the appropriate strains . Growth of these cells on 2% glucose was comparable to WT , but growth was diminished when FKH overexpressing ( OE ) cells were grown on 2% galactose-supplemented media ( Figure 3B ) . Fkh1-TAP and Fkh2-TAP were weakly expressed in 2% glucose , but massively expressed after 6 hrs in 2% galactose ( Figure 3C , 3D ) . Lower concentrations of galactose ( 0 . 05–0 . 1% ) did not influence vegetative growth ( data not shown ) , but did improve stress resistance and longevity ( Figure 4 ) . First , we measured the ability of 5 day stationary phase cultures maintained in DM to survive a 1 hour treatment of 100 mM H2O2 . For this experiment , once cells reached stationary phase the cells were split with one sample receiving a supplement of 0 . 05% galactose . After 5 days , samples were removed , treated with H2O2 , and then diluted onto YPD plates to determine colony forming units . Controls were cells that did not receive H2O2 . The FKH OE cultures exhibited increased survival in the absence of galactose , with improved resistance when supplemented with galactose ( Figure 4A ) . These observations are consistent with a role for the Fkh proteins in stress resistance during stationary phase . If the Fkh proteins do enhance stress resistance during stationary phase , then it is likely that increased FKH expression may prolong metabolic activity in these cells . CLS of WT and FKH OE cells was measured in DM in the presence and absence of 0 . 05% galactose ( Figure 4B ) . Cells were grown in 2% glucose to stationary phase , then split , with one sample receiving a supplement of 0 . 05% galactose . Unaltered WT cells were used as a control . We observed that the addition of 0 . 05% galactose increased the CLS mean lifespan ( 50% survival ) of the unaltered WT control from approximately 8 . 5 to 12 days . In the absence of galactose the OE strains exhibited mean lifespans of 9 . 5 to 12 days . However , in the presence of 0 . 05% galactose , the FKH1 OE strain experienced a 15 day mean lifespan , while FKH2 OE strains enjoyed mean lifespans of approximately 27 days . Yeast cell lifespan can be measured in post-mitotic stationary phase ( CLS ) , or in rapidly dividing mitotic cells ( RLS ) . Stress resistance plays a major role in determining both CLS and RLS , however not all genes that influence CLS also influence RLS . Sir2 in yeast is a good example of this [44] . To determine whether the Fkh proteins also influence RLS , we measured RLS in the cells employed in Figure 4B using 2% sucrose as a base carbon source in the presence and absence of 0 . 1% galactose ( Figure 4C ) . In the absence of galactose , RLS of all strains was relatively unchanged . Upon 0 . 1% galactose supplementation , FKH2 OE cells had a markedly longer RLS . The mean lifespan of FKH1 OE cells was also increased , but not to the same extent as the FKH2 OE cells . We also observed increased RLS in FKH2 OE , but not FKH1 OE cells using 0 . 05% galactose ( data not shown ) . The enhanced stress resistance and CLS observed in OE strains in the absence of galactose are not surprising as we previously documented the basal activity of the galactose promoter [33] . Our results are consistent with the Fkh proteins playing a role in responding to stress , which may indirectly lead to increased CLS and RLS . The effect appears to be greater during post-mitotic stationary phase cells , with Fkh2 perhaps playing a more pivotal role compared to Fkh1 . The advantage of using yeast for genetic studies is the relative ease of identifying interacting partners for proteins and genes of interest . Thus , we sought possible downstream Fkh targets that may be involved in stress response and longevity . One possible target of the Fkh transcription factors is the Anaphase-Promoting Complex ( APC ) . The APC is an evolutionarily conserved ubiquitin-protein ligase ( E3 ) that targets proteins that inhibit mitotic progression and exit , as well as G1 maintenance , for ubiquitin- and proteasome-dependent degradation [21] , [23] . We previously observed APC mutants to exhibit reduced CLS and RLS , while increased APC10 expression extended RLS [29] , [46] . Furthermore , APC mutants are sensitive to DNA damaging agents , and exhibit both chromatin assembly and histone modification defects [28] , [33] , [46]–[49] . Consistent with the APC's involvement in histone biogenesis and lifespan , we recently demonstrated that yeast cells harboring histone modification defects are subject to reduced RLS [50] . The APC appears to play an evolutionarily conserved role in lifespan , as mutations to mouse BubR1 , a component of the spindle checkpoint that inhibits APC function , resulted in inappropriate APC activity and premature aging phenotypes [27] . A possible link between the APC and the Fkhs was revealed by a previous microarray analysis of fkh1Δ fkh2Δ cells , where APC1 ( APC subunit ) , CDC5 ( APC activator/target ) , CLB2 ( APC activator/target ) , CDC20 ( APC activator/target ) , and IQG1 ( APC target ) , were identified as responsive to the Fkh transcription factors [17] , [21] , [22] . A subsequent analysis of CLB2 mRNA expression during the cell cycle ( CLB2 mRNA synthesis is cell cycle regulated ) showed that it was defective in fkh1Δ fkh2Δ mutants [16] . Clb2 , a mitotic cyclin that is an important activator of the APC , later becomes targeted by the APC for degradation to allow exit from mitosis [21] . To test our hypothesis that APC activity may be targeted and activated by the Fkh proteins , we created fkh1Δ fkh2Δ cells harboring a mutation in the APC subunit Apc5 to enable genetic analyses . APC5 encodes an essential APC subunit ( the apc5CA allele used in our studies contains a two basepair deletion at the 5′ end of the coding region , most likely creating an N-terminally truncated protein [28]; unpublished data ) . Cells with the apc5CA mutation grow slowly at temperatures above 36°C , which can be recovered or exacerbated by genetic alteration of negative or positive regulators , making this an excellent allele to identify APC5 interacting partners [29] , [33] , [47] , [49] , [51] . First , we examined the growth characteristics of apc5CA fkh1Δ fkh2Δ mutants . Deletion of both FKH genes severely impaired apc5CA growth at the restrictive temperature , but not at the permissive temperature ( Figures 5A ) . Deletion of both FKH genes also impaired the growth of apc10Δ cells and was lethal in the apc11-13 background ( unpublished data ) . This preliminary investigation provides evidence that the APC and the Fkhs may share a common function . If the Fkhs and the APC do share a common function , then increased expression of the FKHs may restore the apc5CA temperature sensitive ( ts ) growth phenotype . Thus , we expressed plasmid borne FKH1 and FKH2 under the control of the GAL1/10 promoter in WT and apc5CA cell . The cells were grown in 2% glucose supplemented media , then spot diluted onto plates containing either 2% galactose or 2% glucose ( Figure 5B ) . At 30°C , expression of the FKHs on glucose-supplemented media was slightly detrimental to WT cells , but beneficial to apc5CA cells . Galactose-driven genes do have basal activity in the presence of glucose [33] . On galactose plates , overexpression of either FKH was toxic , as observed above ( Figure 3B ) . At 37°C , expression of either FKH gene on glucose plates improved growth of both WT and apc5CA cells . We have also observed that increased FKH1 expression suppressed the ts defect in additional APC mutants , including apc10Δ , apc11-13 , cdc16-1 , cdc23-1 , apc5CA apc10Δ and apc5CA cdc26Δ cells ( data not shown ) . These observations suggest that under conditions of stress , whether temperature or impaired APC activity , moderately increased FKH expression improves the capacity of the cell to cope . Next , we asked whether deletion of both FKHs influenced apc5CA CLS . Cells expressing the different combinations of mutations were grown to stationary phase , then split , with one half resuspended in H2O , and the other half left in DM . Equal volumes were then plated on the days shown to generate survival curves ( Figure 5C , 5D ) . In DM , apc5CA cells rapidly lost viability compared to WT and fkh1Δ fkh2Δ cells ( Figure 5C ) . Interestingly , the triple mutant survived as long as fkh1Δ fkh2Δ cells . This suggests that deletion of both FKH1 and FKH2 is epistatic to the apc5CA allele under the conditions tested . In other words , under normal media conditions using DM , the Fkhs appear to act upstream of the APC . When the experiment was conducted by maintaining the cells in H2O for the duration of the experiment , a different survival profile was observed ( Figure 5D ) . WT ( 20 days vs 7 days ) and apc5CA ( 5 days vs 2 . 5 days ) cells both responded to H2O conditions by exhibiting a longer mean CLS . However , fkh1Δ fkh2Δ cells had the same CLS in H2O as in DM , which was similar to apc5CA cells in H2O , whereas the triple mutant had a greatly reduced CLS in H2O . As stated above , the failure of fkh1Δ fkh2Δ cells to survive longer under SCR conditions , such as H2O , suggests that the Fkhs are required for long life under SCR conditions . The similar H2O CLS observed in fkh1Δ fkh2Δ and apc5CA cells , and the dramatically reduced H2O CLS in the triple mutant indicates that the Fkhs and the APC may have redundant functions under stress conditions . This contrasts with the Fkh/APC epistatic interaction that appears to occur under DM conditions . This could reflect dual roles for the Fkh proteins; as cell cycle regulators under normal conditions , and as stress response proteins under stress conditions [17] , [18] , [52] , [53] . Lastly , these observations clearly identify another non-mitotic function for the APC . The APC has been shown to function in other non-mitotic activities , such as meiosis , quiescence , differentiation , metabolism , maintenance of post-mitotic neurons , and interestingly , memory in mice [30] , [54]–[58] . In addition to controlling CLS , the APC and the Fkhs are also involved in histone metabolism , but likely through very different mechanisms . The Fkhs are redundant activators of cell cycle dependent histone expression [17] . On the other hand , histones and histone modifications are reduced when genes encoding different APC subunits , such as APC5 , APC9 , APC10 , APC11 , CDC16 , CDC23 and CDC26 , are mutated [33] . The mechanism involved remains elusive , but it is likely post-transcriptional , as histone mRNAs are unaltered in APC mutants [33] . Our analysis of total histone levels in the different mutant combinations indicates that histone control is indeed through different redundant mechanisms , as histone levels are greatly reduced in the triple mutant compared to the single and double mutants ( Figure 5E ) . However , it remains possible that the Fkhs drive histone synthesis through direct transcriptional control and indirectly via the APC . As a direct assessment of whether the Fkhs control an APC function , we measured Clb2 stability in apc5CA and fkh1Δ fkh2Δ mutants . Clb2 is a B-type cyclin that is targeted by the APC for degradation in order to exit mitosis [21] . CLB2 transcripts are also controlled by the Fkhs [16] , [17] . WT , apc5CA and fkh1Δ fkh2Δ cells were grown to early log phase at 30°C , then arrested in G1 using α factor . The cells were determined to be arrested in G1 using microscopy ( data not shown ) and FACS analysis ( Figure 5F ) . The cells were then released into cycloheximide , with samples removed every 15 minutes for protein analysis using antibodies against endogenous Clb2 . In both WT and apc5CA cells Clb2 levels were decreased in G1 arrested cells when compared to asynchronous cells ( Figure 5F ) . In fkh1Δ fkh2Δ cells however , the degradation of Clb2 was reduced in G1 arrested cells . However , it should be noted that fkh1Δ fkh2Δ cells accumulated in G1 in asynchronously grown cells , as indicated by FACS , perhaps reflecting the inability to degrade mitotic cyclins . These observations are consistent with a model where the Fkh transcription factors act in a positive manner upstream of the APC , perhaps through transcriptional activation of APC subunits , activators and targets . Since the APC is associated with maintaining lifespan in multiple systems [27] , [29] , [30] , [31] , we tested whether the APC may also be involved in oxidative stress resistance . To test this hypothesis , we conducted CLS in the presence of oxidative stress in WT , apc5CA , fkh1Δ fkh2Δ and apc5CA fkh1Δ fkh2Δ cells . The cells were grown to stationary phase followed by the addition of 25 mM H2O2 ( Figure 6A ) , with cell counts determined every other day . The results show WT cells had a reduced lifespan in response to 25 mM H2O2 , while the mutants were further impaired . The lifespan of the triple mutant was dramatically reduced compared to apc5CA and fkh1Δ fkh2Δ cells in H2O2 , as it was in water ( Figure 5D ) . These observations provide additional evidence that the triple mutant is extremely sensitive to stress and likely perceives water as a severe stress , rather than a form of caloric restriction . To our knowledge , few mutations have been described that act in a negative manner to SCR . To investigate the roles of the APC and the Fkh proteins in stress resistance further , stationary phase cultures were treated with 100 mM H2O2 for 30 minutes at 30°C , and then plated to determine cell viability . While 53 . 2% of WT cells survived 100 mM H2O2 , only 21 . 3% of apc5CA cells and 31 . 3% of the fkh1Δ fkh2Δ survived the treatment ( Figure 6B ) . However , the apc5CA fkh1Δ fkh2Δ mutant was dramatically impaired , with only 3 . 5% surviving this treatment . Taken together , our data indicates that the APC and the Fkh proteins have overlapping functions in response to oxidative stress in post-mitotic cells , opposed to the epistatic interaction observed in unstressed cells ( Figure 5C , Figure 7 ) . Consistent with an APC involvement in post-mitotic CLS , the APC is also required for stress response in post-mitotic cells .
The initial focus of this work was to determine whether the conserved yeast Fkh proteins were involved in longevity , as shown with metazoan FOXOs . Our work clearly demonstrates a need for the Fkh proteins for extended CLS and RLS . However , our work also demonstrates that the Fkh proteins are necessary for extended lifespan in response to SCR . Recently , the Rim15 stress responsive transcription factor was identified as a major mediator of SCR lifespan extension [15] . Deletion of RIM15 blocked extended lifespan in ras2Δ , sch9Δ and tor1Δ strains , indicating that the phenomenon of SCR funnels through Rim15 . Interestingly , although deletion of RIM15 in the extremely long-lived ras2Δ sch9Δ mutant reduced lifespan under normal conditions , this strain could still respond to SCR , suggesting other factors can respond to SCR in the absence of Rim15 . Our data provides the possibility that Fkh1/Fkh2 may fulfill this role . Future work will require an analysis of strains lacking FKH1 , FKH2 and RIM15 . We tested the hypothesis that the Fkh proteins play a role in lifespan by contributing to APC activation . Our results support this hypothesis , as ( i ) low-level expression of FKH1 or FKH2 suppressed APC mutant growth phenotypes; ( ii ) deletion of both FKH1 and FKH2 exacerbated APC mutant histone and growth phenotypes in mitotically active cells; and ( iii ) deletion of FKH1 and FKH2 stabilized the APC substrate Clb2 . Furthermore , stress resistance and lifespan in the presence of stress were markedly worse in the triple mutant compared to the single and double mutants . These results demonstrate that the APC and the Fkhs function together in both mitotic and post-mitotic cells . While fkh1Δ fkh2Δ cells do not respond to SCR , apc5CA fkh1Δ fkh2Δ exhibit decreased CLS under these conditions . This suggests that H2O is perceived as much more than a low level stress in the triple mutant , which is consistent with the Hormesis hypothesis of aging , a theory that postulates low level stresses turn on stress defense mechanisms , leading to potenially longer life [60] . Nonetheless , these observations implicate the APC as a player in caloric restriction and stress response . We observed that APC and FKH mutants interacted differently depending on the growth conditions . Figure 7 presents a model describing how this may occur . Under non-stress conditions , such as growth on YPD or maintenance of stationary phase cells in the depleted media ( DM ) , there was little interaction ( growth on YPD at 30° ) or an epistatic interaction ( CLS in DM ) . The interaction could be interpreted to define a pathway where the Fkhs act upstream of the APC . It could be as simple as the Fkhs driving the transcription of the APC subunit Apc1 , and the APC activators Clb2 , Cdc5 and Cdc20 [16] . This interaction is likely more complicated than direct classical epistasis , perhaps involving stoichiometric alteration of APC activators , subunits and/or substrates . The deletion of the FKHs in APC mutant strains may bring about a homeostatic balance between APC activity and substrate levels . On the other hand , under stress conditions , such as high temperature or the addition of H2O2 , apc5CA fkh1Δ fkh2Δ cells exhibited a synergistic interaction . This type of interaction occurs when multiple proteins drive a similar activity in a redundant manner , thus requiring that more than one mutation must occur to expose a phenotype . Both the Fkhs and the APC are required for response to stress , but likely act in very different manners . The Fkhs drive expression of many stress response genes [16] , whereas the molecular mechanisms employed by the APC to combat stress may involve APCCdh1 , which has been found to regulate different stress responses through the degradation of substrates such as Clb2 and Hsl1 as a part of the APC's role in G1 maintenance [32] . An additional role the APC likely plays in stress revolves around histone metabolism . We have shown that APC mutants exhibit defects in histone maintenance , chromatin assembly and histone modifications [28] , [33] , [47]–[49] , [51] . It is well established that histone modifications , such as phosphorylation of H2A/H2AX , methylation of H3 Lys79 , and acetylation of H3 Lys56 and H4 Lys16 , are required for recruitment of DNA repair enzymes to sites of DNA damage [61]–[65] . Therefore , under stress conditions , the Fkhs and the APC likely play separate , yet overlapping roles in ensuring the cell responds to stressful damaging agents in a positive manner . Our studies provide a potential link between the Fkh transcription factors and the APC that ties the APC together with stress response . This provides insight into a molecular mechanism whereby the APC facilitates longevity . The evolutionarily conserved nature of our results , and the established role the APC plays in tumor development and progression , suggests that the APC may be a potential downstream target of the insulin signaling pathway . A recent report supports this scenario , as the insulin driven AKT1 kinase phosphorylates the APCCdh1 substrate Skp2 , an SCF component , which inhibits Skp2 degradation [66] . This enables SCF activity and promotes cell cycle progression . These studies offer the basis for further studies in understanding APC-dependent longevity . This study identifies the yeast forkhead box containing proteins Fkh1 and Fkh2 as regulators of lifespan , allowing for the characterization of upstream and downstream regulation by these factors . This may provide insight into how the highly related FOXO proteins in mammals regulate both lifespan extension and tumor suppression . Our data supports a model where FKH1 and FKH2 are functionally orthologous with the metazoan FOXOs , which opens the door to genetic manipulation in yeast for further exploration into the function and mechanisms controling these important metabolic and stress regulators .
The yeast strains used in this study are shown in Table 1 . The fkh1Δ::kanMX6 and fkh2Δ::kanMX6 strains , obtained from the ResGen library ( provided by W . Xiao , U . of Saskatchewan ) , were repeatedly backcrossed to our S288c background strains to generate apc5CA , apc10Δ and apc11-13 congenic partners . apc11-13 cells were a kind gift from T . Hunter ( Salk Institute ) . Cells harboring APC9 , APC10 and CDC26 deletions were acquired from W . Xiao and backcrossed repeatedly to our S288c background . cdc16-1 , cdc23-1 and the isogenic wild type were generously provided by D . Stuart ( U . of Alberta ) . The C-terminal FKH1-TAP and FKH2-TAP strains were generously provided by A . Ghavidel ( U . of Toronto ) . PCR based methods were used to TAP-tag FKH1 and FKH2 in various mutants . Cells expressing endogenously GFP-tagged FKH1 and FKH2 were obtained from Open BioSytems . The galactose inducible FKH1-HA and FKH2-HA encoding plasmids were obtained from the Research Genetics library of tandem affinity tagged plasmids purchased by W . Xiao ( U . of Saskatchewan ) . The GAL1/10 promoter was integrated upstream of the FKH1 and FKH2 genes by PCR-based homologous integration approach . Two sets of primers were designed for this approach . First , primers were designed to amplify the LEU2 gene ( plus 300 basepairs of the promoter ) flanked on the 5′ side by 60 nucleotides of sequence homologous to the immediate promoter regions of FKH1 or FKH2 , and the 3′ side by 60 nucleotides homologous to the GAL1/10 promoter . The second primer set was designed to amplify GAL1/10 promoter flanked on the 5′ side by 60 nucleotides homologous to the 3′ end of the LEU2 gene and on the 3′ end by 60 nucleotides homologous to the first 60 nucleotide of FKH1 or FKH2 . Primer sequences are available upon request . The two PCR fragments were transformed together into WT cells . Leu+ transformants were selected for further analysis . Media were prepared as previously described [28] , [33] . Segregation of the kanMX6 cassette was determined by patching spores onto 0 . 2 mg/ml geneticin-supplemented YPD plates . Mutants containing two or more kanMX6 marked alleles were generated by selecting tetrads in which G418 resistance segregated in a 2∶2 fashion . Escherichia coli strains JM109 and DH10B were used to propagate DNA plasmids . DNA manipulations such as DNA minipreps , and yeast and E . coli transformations were carried out according to standard protocols [67] . Spot dilution assays were conducted by pipeting 3 µl of cells from samples generated from a 10-fold dilution series onto the various media shown and grown at the temperatures indicated . The starting spot generally contained 5×104 cells . To assess resistance to oxidative stress , cultures were grown in 2% YPD at 30°C for 5 days and then diluted to an OD600 of 1 in depleted media ( DM ) . Each of these cultures was divided into two samples and 100 mM H2O2 ( EMD Chemicals ) was added to one sample . Both samples were then incubated at 30°C for 1 hour . Viability was determined by plating diluted cells onto 2% YPD and comparing the growth of the H2O2 treated culture to that of the non-treated control culture . Clb2 stability was determined by growing the indicated cells to early log phase at 30°C , then adding 2 µg/ml α factor , when using BAR1 strains , in media at pH 3 . 5 to arrest cells in G1 . After 1 . 5 hours , another 2 µg/ml was added , with continued incubation for another 1 hour . After this treatment , cells were arrested in G1 . Cell cycle arrest was confirmed by microscopic visualization of the cells and by flow cytometry . The α factor was then washed away and fresh media containing 10 µg/ml cycloheximide was added . The incubation was continued at 30°C , with samples removed every 15 minutes for Western analysis using antibodies against endogenous Clb2 and GAPDH . FACS was performed as described previously [28] . To characterize fkh1Δ fkh2Δ FACS profiles the following changes were made to overcome the persistent cell/cell contacts inherent to this mutant . 1 ml of culture ( OD600 0 . 4 ) was resuspended in 50 µl followed by the addition 10 µl of 12 . 5 U/µl lyticase . This solution was incubated for 15 minutes at room temperature , followed by the addition of 500 µl 50 mM Tris . The cells were then sonicated for 6 seconds at output 4 . The cells were centrifuged , resuspended in 1 ml 70% EtOH and processed as usual past this step . Chronological lifespan was performed as previously described [29] , [40] , [41] . Briefly , overnight CM ( 2% glucose ) cultures were diluted to OD600 0 . 5 in fresh CM with a flask to culture volume ratio of 5∶1 . The incubation continued ( 200 RPM ) at 30°C . Each day the same volume of culture was diluted and plated to evaluate colony forming units ( CFU ) as a measure of viability . When the CFU counts peaked , this was deemed stationary phase and denoted as Day 1 . Every two days CFU were determined and compared to Day 1 . For severe caloric restriction ( SCR ) experiments , once stationary phase was reached in CM ( Day1 ) , cultures were washed and resuspended in sterile distilled H2O , with washes of equal volume of water every two to four days to remove metabolites produced by the cells . Galactose and hydrogen peroxide were added to appropriate cultures upon reaching Day 1 to final concentrations of 0 . 05% and 25 mM respectively . For fluorescent localization , samples were obtained from CLS cultures , washed and mounted in PBS or Ultracruz mounting medium ( Santa Cruz Biotechnology sc-24941 ) and imaged using 100× oil immersion with an Olympus BX51 fluorescent microscope . Images were captured using an INFINITY 3-1UM camera , and analyzed with Infinity Analyse software version 5 . 0 . 3 ( Lumenera ) . Replicative lifespan experiments were performed as previously described [29] . The plates were stored at 4°C each night . The experiments were performed blind; the genotypes of the strains and conditions used were not revealed to the experimenter until the final mother stopped producing daughters . Yeast whole cell protein extraction was performed as previously described [28] . Samples were resolved on a 15% acrylamide SDS-PAGE gel , which was stained with Coomassie brilliant blue R-250 ( OmniPUR ) or transferred directly to nitrocellulose membrane ( PALL ) at 400 mAmps for 1 hour . Protein loading was analyzed using ImageJ 1 . 37v software ( NIH ) . Equalized samples were then transferred to nitrocellulose membrane . For Western analysis , the membranes were blocked in 5% non-fat milk ( Biorad ) and PBST overnight at 4°C . The membranes were then incubated with primary antibody in 5% non-fat milk and PBST for 1 . 5 hours at room temperature or overnight at 4°C . Rabbit polyclonal anti-H2B ( Abcam ) , polyclonal anti-H3 ( Abcam ) , and polyclonal anti-H4 ( Abcam ) were used at a 1∶4000 dilution . Rabbit polyclonal anti-Clb2 ( Santa Cruz; Y-180 ) was used at 1∶2000 . The TAP antibody ( Open Biosystems ) was used at a dilution of 1∶1000 . Mouse monoclonal anti-GAPDH ( Sigma ) was used at a dilution of 1∶20 , 000 . The membranes were then washed 3 times in PBST for 15 minutes , and incubated in 1∶10 , 000 dilution of secondary antibody in 5% non-fat milk and PBST for 1 hour at room temperature . After another 3 washes in PBST for 15 minutes , the membranes were processed with Enhanced Chemiluminescence reagent ( PerkinElmer ) and exposed to Kodak film . | Throughout human evolution , one question has remained constant: can we live forever ? We are continuously bombarded with products , diets , and exercise regimens that supposedly add years to our life . Is there an alterable program , whether genetic or environmental , that can be tweaked to increase longevity ? Medical advances have led to a dramatic increase in average lifespan over the last century . However , the maximum human lifespan has curiously remained constant . Recent research indicates that in many organisms a genetic program exists to control lifespan . The conservation of this genetic lifespan program extends into yeast where numerous longevity genes have been isolated and characterized . Interestingly , mutations that reduce genomic instability , glucose utilization , or oxidative damage extend lifespan in multiple organisms . Here we characterize one such set of genes , the FOXOs . In animals , these genes increase lifespan and suppress tumors , but have yet to be associated with longevity in yeast . By confirming that these genes play a similar role in yeast , we provide a tool to identify downstream factors triggered by the FOXOs , a feat which has not yet been accomplished in other systems . Considering the conservation of these factors , it is likely that our discoveries in yeast will be directly applicable to research into human cancer and aging . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biochemistry",
"model",
"organisms",
"molecular",
"cell",
"biology",
"global",
"health",
"genetics",
"biology",
"evolutionary",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | The Yeast Forkhead Transcription Factors Fkh1 and Fkh2 Regulate Lifespan and Stress Response Together with the Anaphase-Promoting Complex |
Many cell functions rely on the ability of microtubules to self-organize as complex networks . In plants , cortical microtubules are essential to determine cell shape as they guide the deposition of cellulose microfibrils , and thus control mechanical anisotropy of the cell wall . Here we analyze how , in turn , cell shape may influence microtubule behavior . Building upon previous models that confined microtubules to the cell surface , we introduce an agent model of microtubules enclosed in a three-dimensional volume . We show that the microtubule network has spontaneous aligned configurations that could explain many experimental observations without resorting to specific regulation . In particular , we find that the preferred cortical localization of microtubules emerges from directional persistence of the microtubules , and their interactions with each other and with the stiff wall . We also identify microtubule parameters that seem relatively insensitive to cell shape , such as length or number . In contrast , microtubule array anisotropy depends on local curvature of the cell surface and global orientation follows robustly the longest axis of the cell . Lastly , we find that geometric cues may be overcome , as the network is capable of reorienting toward weak external directional cues . Altogether our simulations show that the microtubule network is a good transducer of weak external polarity , while at the same time , easily reaching stable global configurations .
Despite their amazing diversity in shapes , biological organisms share some common structural components at the cellular level . Among those , one of the best conserved proteins across eukaryotes , tubulin , assembles into protofilaments , which in turn form 25 nm nanotubes known as microtubules , usually made of 13 protofilaments . The network of microtubules is highly labile and can reshape itself in a matter of minutes . In plants , microtubules form superstructures before ( the preprophase band ) , during ( the spindle ) and after ( the phragmoplast ) cell division . Plant microtubules also form dense and organized arrays at the periphery of the cell during interphase [1] and these arrays are known as cortical microtubules ( CMTs ) . The behavior of CMTs has been studied extensively at the molecular level [2] . One of their main functions is to guide the trajectory of the transmembrane cellulose synthase complex and thus to bias the orientation of cellulose microfibrils in the wall . This in turn impacts the mechanical anisotropy of the cell wall and controls growth direction [3–6] . This function explains why most mutants impaired in microtubule-associated proteins exhibit strong morphological defects [7] . Whereas this provides a clear picture on how microtubules impact cell shape , in turn how cell shape impacts microtubule behavior has been less explored . There is evidence that microtubule orientation depends on cell shape [8–10] , with microtubules being mostly transverse to the longest axis , but this might require specific regulation because microtubules orient along the longer axis of the confining domain in vitro [11] . There is also evidence that shape-derived mechanical stress can bias cellulose deposition , possibly through microtubule orientation towards the direction of maximal tension , both at the tissue and single cell scales [8 , 12–21] . The molecular mechanism behind remains largely unknown . Finally , cortical microtubules orientation may change in response to signals such as blue light or hormones , see for instance [17 , 22] and may be oriented by the hydrodynamic forces due to cytoplasmic streaming [23 , 24] . Here we use modeling to explore the relative contributions of cell geometry and external directional cues in the final microtubule organization . The molecular basis for microtubule dynamics is rather well established . Consistent with the absence of centrosome in land plants , microtubule nucleation is dispersed in plant cells , as it occurs at the cell cortex [25] , along existing microtubules during branching events [25–27] , and at the nuclear envelope [28] . As they grow , microtubules form stiff and polar structures . They can alternate growth , pause and shrinking at the so-called plus end [29] , whereas they mainly shrink or pause at the minus end [30] . The combination of an average shrinkage at the minus end and dynamic instability at the plus end leads to an overall displacement of the microtubule , also called hybrid treadmilling [30] , with a dominant contribution of short treadmilling microtubules in the final microtubule organization [31] . The growth of microtubules in persistent directions is the main cause for microtubule encounters . When one microtubule encounters another microtubule , different outcomes can be observed [9 , 32]: if the encounter angle is shallow , zippering can occur , i . e . the growing microtubule bends and continues its polymerization along the encountered microtubule , which leads to the creation of microtubule bundles; if the encounter angle is steep , crossover can occur , i . e . a microtubule polymerizing without deviating its trajectory and crossing over the encountered microtubule; or alternatively catastrophe is triggered , i . e . a rapid plus end shrinkage after contact with the encountered microtubule . Such selective pruning of microtubules may explain how microtubules can form parallel arrays from initially random orientations , and conversely change the net orientation of their arrays over time , through a phase of randomisation [33 , 34] . Selective pruning has indeed an essential ingredient of most models for microtubule dynamics [35–44] . The presence or absence of different microtubules associated proteins ( MAPs ) can modulate the stability of microtubules or their capacity to form bundles and to self-organize . For instance , the microtubule severing protein Katanin accounts for most of the pruning events at crossover sites [45] . The microtubule network is a typical example of a self-organizing system , where properties of individual elements and their interactions induce specific and sometimes counter-intuitive global properties . To predict how regulation at the level of each microtubule can give rise to specific global outcomes , one can resort to computational models . Modeling approaches have been developed , simplifying microtubule interactions by restricting them to the plasma membrane , i . e . a simpler 2D space [40 , 46] . In those agent-based models , several microtubule properties were coded and interactions between CMTs , based on these properties , were simulated . The outcome is an emergent network , whose characteristics can be analyzed . For instance , increased microtubule severing was predicted to generate a larger number of free microtubules , more amenable to bundle into aligned arrays [9 , 42] and this was observed in experiments [47] . So far , most of the microtubule models have been implemented in a 2D space or with microtubules confined to the surface of the cell . A major outcome of such models was to demonstrate that global orientations of the network can spontaneously emerge from the interactions between microtubules [48] . Many combinations of parameters and behaviors have been studied: instability at the plus end [35 , 37] , role of zippering [9 , 36 , 38 , 39] , nucleation modes [39 , 42] , and severing [44] . Beyond their differences , a global orientation emerges in most of these combinations suggesting that converging toward a global orientation is a robust feature of microtubule networks . Conversely , the diverse combinations of microtubule properties provide different scenarios for the fine-tuning of the network structure and stability of this emerging behavior . Some aspects of cell geometry were related to microtubule behavior in certain simulations . Simulations showed how different directional biases in nucleation can induce an ordering of the array toward directions that are correlated to cell geometry [35 , 43] . Further , branching nucleation rules can elicit handedness of the global direction of microtubule arrays , provided that the branching is biased toward one direction [41] . Other studies used a simulation space where borders , analog to cell edges , induce more or fewer catastrophe events or are more or less permissive toward microtubule growth [10 , 33 , 35 , 41] . Most studies concluded that a global orientation of microtubules can be correlated to cell face orientations . The contribution of the third dimension to microtubule behavior has started to be investigated . Computational models for animal systems have focused on 3D considerations but the nucleation hypotheses are too different from that in plants to be transposed directly [49 , 50] . Fully 3D models suited for plants are still lacking: almost all existing studies have confined microtubules to surfaces embedded in 3D [10 , 11 , 41 , 51] . In [35] , a 2D model was extended into a full 3D model but it did not include cell boundaries , which yielded microtubules distributed over the whole simulated domain , in contrast with the cortical localization of microtubules in planta . In this paper , we explore the influence of 3D cell shape on the basic properties of a dynamic microtubule network . We do not assume that microtubules are confined to the cell surface; rather , we simulate a closed volume where microtubules are more or less free to grow in all directions . Anchoring to the membrane is not imposed by the model and instead becomes a variable in the model . Using this framework , we investigate to what extent microtubule interaction with the membrane can influence microtubule dynamics . Our study also addresses the relative contributions of cell shape , microtubule interactions , and external directional cues in network organization .
Following previous studies , we modelled microtubules as a set of line segments that nucleate , grow , shrink , and interact with each other and with the cell surface represented as a triangular mesh ( see Methods for details ) . Nucleation of the minus end occurs randomly at the surface . Growth occurs from the plus end with a small directional noise that is related to the persistence length of microtubules ( Fig 1A ) . Shrinkage starts randomly at the nucleation site ( minus end ) and then continues at constant velocity . A microtubule that encounters a preexisting microtubule either changes direction to that of the preexisting microtubule if the encounter angle is shallow ( a process known as zippering , see Fig 1A ) , and otherwise starts shrinkage from the plus end ( “head-on” collision ) . We considered two types of interaction with the cell surface: strong anchoring , whereby microtubules remain on the surface , as in previous studies [10 , 11 , 41 , 51] , and weak anchoring , whereby microtubules are prevented from leaving the cell interior . More specifically , in the case of weak anchoring , the interaction between a growing microtubule and the nearby surface is similar to the interaction between two microtubules: the microtubule encountering the surface at a steep angle starts shrinking , and otherwise starts growing tangentially to the surface ( Fig 1A ) ; a microtubule may leave the surface because of the directional noise . We considered three base shapes: cube , “square” ( flattened cube ) , and “long” ( elongated small cube ) , of dimensions in the order of 10 μm , typical of plant cells . These shapes were smoothed so that the maximal curvature corresponded to a radius of either ∼1 μm ( “sharp” ) or ∼5 μm ( “smooth” ) , which corresponds to typical radii of curvature measured in root epidermis [10] , see Fig 1B . We also considered an ellipsoidal shape when investigating the effects of an external cue . Although these shapes are not fully realistic , they make it easier to disentangle the geometrical parameters influencing microtubule dynamics . A typical simulation with weak anchoring is given in S1 Video . Corresponding snapshots are shown in Fig 1C and 1D with various 3D and 2D views . A few first observations can be made: the microtubules tend to bundle; a well-defined local orientation appears; microtubules appear to be mostly close to the cortex . We first considered the effect of the anchoring of the microtubules to the membrane . There are proteins that have been shown to be associated with both microtubules and a plasma membrane component in plants [52] . For instance , CELLULOSE SYNTHASE INTERACTIVE PROTEIN 1 ( CSI1 ) interacts with both CMTs and the cellulose synthase ( CESA ) complex [6 , 53] and CSI1 was also proposed to stabilize the microtubule network [54] . Yet , the influence of CMT-CESA interactions on the microtubule network is still poorly understood . Thus , we took advantage of the 3D nature of our model to study the impact of the anchoring rule to the membrane on the global parameters of the microtubule network . We investigated the microtubule dynamics when the anchoring to the membrane is weak . As a reference case , we also considered strong anchoring , whereby microtubule are constrained to grow on the membrane , as if putative anchoring proteins were highly concentrated . As expected , strong interactions led to a surface-localized cortical zone with microtubules trajectories embedded in the plane parallel to the mesh ( Fig 2A ) . In the case of weak anchoring , microtubules grow in all directions , but occasionally , as they encounter the membrane , their direction may be transiently tangent to the membrane . The typical distance between microtubules and membrane ranges from 50 to 250 nm ( Fig 2A ) according to the shape . Even if weak anchoring allows microtubules to grow through all the cell volume , we find that such weak interaction with the membrane is enough to elicit the existence of cortical microtubules . Therefore , the three-dimensional nature of our model helps us demonstrate that strong anchoring is not required for the presence of large populations of cortical microtubules in plant cells: the directional persistence , together microtubule growth mode , can cause such sub-cellular localization . As the microtubules tend to stay at close to the membrane , they also bundle , with the proportion of tubulin in bundles varying from 30% to 75% ( Fig 2B ) . The ability for the microtubule network to generate a spontaneous bundled structure is consistent with previous models constraining microtubules to the surface . Strikingly , this effect is also present in the case of weak anchoring . Independently of the encounter rule , weak anchoring strength increases the total number and the size of microtubules ( by about 20% , in length and in number ) when compared to strong anchoring . Weak anchoring also yields less bundling ( reduction of about 20% ) . A likely explanation is that a weak anchoring to the membrane allows microtubules to “escape” inside the cell , thus diminishing the encounter probability . Consequently , microtubules weakly bound to the membrane have more space to grow and are less subject to shrinkage-induced collisions or bundling with other microtubules ( Fig 2B ) . Next , we used our model to determine the consequences of changing cell shape on global properties of the network . We simulated the network in three main sharp shapes represented on Fig 1A . The results from the simulations indicate that these shapes only have a small effect on the number of microtubules ( less than 10% difference ) , on the length of microtubules ( 5 to 15% difference ) and on the proportion of bundles ( less than 15% difference ) . Overall , elongated cells have more microtubules , more bundles , and longer microtubules , while cubes show the lower values ( Fig 3 ) . A likely explanation is that microtubules are more likely to follow the long axis in the long shape ( see below ) so that they are less affected by cell boundaries . We also analyzed the effect of cell shape on the microtubule array anisotropy , averaged over the cell ( see Methods ) ; this is quantified with an order parameter with values between 0 and 1 . As microtubule array anisotropy is skewed towards low values ( found inside the cell ) we used a non-parametric test based on ranks for statistical comparisons . Consistently with its effect on bundling ( see above ) , we found that anchoring strength slightly affects the anisotropy of the microtubule arrays . Weak anchoring decreases the anisotropy of the network by about 10% ( Fig 4C ) . The type of cell shape did not appear to influence the global anisotropy of the microtubule network ( Fig 4A ) . This is interesting as it suggests that different cell types with various shapes do not require differential regulation of the network in order to maintain the anisotropic properties of their CMT arrays . However , smooth and sharp shapes differ significantly in anisotropy of the microtubule network ( by 10% , p < 0 . 001 ) ; more curved cell edges in sharp shapes correspond to a lower anisotropy ( Fig 4B ) . Larger surface curvature induce more “heads-on” collisions when microtubules grow nearby , leading to a smaller microtubule density , as they become more distant from the surface ( Fig 2A ) ; this would lead to more spatial variations in the orientation of microtubules and hence lower anisotropy . In order to determine how the shape of the cell influences the global orientation of the network , we measured the distribution of orientations along two opposite faces of the cells . We chose two arbitrary faces for cube shape , the largest faces for square shape , and two of the largest faces for the long shape; accordingly , these two opposite faces are among the two largest faces in area for the shape of interest . An angle of 0° corresponds to the long axis in the case of the elongated cell . Angles of 90° or -90° correspond to directions perpendicular to that axis . First , we observed that in the case of square shaped faces , most of the microtubules align along the cell face diagonal , i . e . the longest path ( Fig 5 ) . This occurred whatever the anchoring strength and the shape of faces at the side . Second , we observed a strong correlation between the axis of the cell and microtubule orientation , and this correlation is the highest for the elongated cell: The microtubule distribution is always either maximal or minimal at an angle of 0 . This effect is higher in the case of strong anchoring , whereas in the case of weak anchoring , secondary peaks show that the diagonals are also overrepresented . These results indicate that the microtubule network is able to read the longest axis of the cell and orient toward that axis , by default . Interestingly , cortical microtubules become longitudinal in hypocotyl cells when growth stops [5 , 55 , 56] , suggesting that they may adopt their configuration by default in that situation . Next , we investigated the robustness of microtubule network . Microtubule arrays entirely reorganize during cell division [57] . Light and hormones can also completely reorient the microtubule network within minutes [33 , 34 , 58] , suggesting that the constraints on microtubules array must not be too strong to allow such rapid reorganization in vivo . We tested whether our model provides such adaptability , using the case of external , directional cues ( Fig 6 and S2 Video ) . We investigated the effect of an external cue , assuming that , as microtubules grow in the cell , their growth direction is biased toward the direction of cue , with a specific weight . We used an ellipsoidal cell with a circumferential cue , which is orthoradial to the long axis of the cell . Such a cue could , for instance , be related to cell polarity ( vector along the long axis ) or could be a proxy of mechanical tension [19] . In the absence of the cue , microtubules are on average parallel to the main axis of the cell . Strikingly , our simulations indicate that even a very low bias ( 0 . 1% ) could disrupt the main orientation of the network . When the weight reaches 1% or more , microtubules massively reorient toward the direction of the cue . Interestingly , the transition between the longitudinal and circumferential orientation occurs when the bias is comparable to the directional noise that is used to account for the persistence length of microtubules . This reinforces the idea that fluctuations of microtubules and the interactions between them lead to alignment with the long axis of the cell and that the small cue is sufficient to counteract these fluctuations and influence the orientation of microtubules . Accordingly , we found that anisotropy increases from no cue to a weight of about 1% and then seems to saturate ( S1 Fig ) . These results indicate that , despite an apparent robust organization , the microtubule network remains extremely sensitive to directional cues . As such , it is capable of reading slight directional cues and generating an ordered array in that direction . This ability to read directional cues is probably linked to the self-enhancement of microtubule orientation through their interactions: As more microtubules orient toward a direction , they prevent growth of microtubules in the perpendicular direction . Our simulations indicate that the network should exhibit two behaviors concerning its orientation . When no external directional cue is present , the network orients toward the main axis of the cell , and generates a polarity that is a direct reading of the global shape . When a directional cue is present , the network reorients so as to emphasize the direction of the cue .
The impact of external cues on the microtubule network has been extensively characterized in experiments . However , because real cells and tissues are never really devoid of external cues , the behavior of the microtubule network by default in a plant cell remains an open question . Our work provides some clues to address this question , by taking into account the 3D shape of the cell and by using a minimal set of parameters for microtubule behavior . For instance , the model with weak anchoring does not require a specific rule for the crossing of microtubules ( unlike when microtubules are confined to the cell surface ) , because microtubules naturally cross each other if they are farther than their diameter . We found that microtubule directional persistence largely determines the subcellular localization and orientation of the microtubule network in various cell shapes . We also identified parameters that seemed relatively insensitive to cell shape , such as microtubule length or number . Last , we found that microtubule dynamics yields at the same time stable orientation and high sensitivity to directional cues , even when such cues go against the default orientation . Altogether , this provides a conceptual framework to dissect the exact contribution of microtubule regulators to the microtubule network organization , in relation to 3D cell shape . Based on TEM images where microtubules are often seen very close to the plasma membrane , it is assumed that anchoring of microtubules to the plasma membrane is relatively strong . However it remains unknown how such anchoring would be mediated [52] . Many biochemical studies have been performed to extract proteins that would link the plasma membrane to cortical microtubules , and so far the only published candidate is a phospholipase [59] , for which no follow-up results have been obtained to the best of our knowledge . Other links have been put forward , such as CLASP [10] or CELLULOSE SYNTHASE INTERACTIVE PROTEIN 1 ( CSI1 ) [53 , 60] , but they might represent rather indirect regulators of the link between microtubules and plasma membrane . Does this mean that microtubule could be cortical without any anchoring module ? Our results suggest that in most of the measured cell shapes , microtubules do not need to be strongly connected to the membrane to remain cortical . This prediction implies that modulating anchoring would affect self-organizing properties tangentially to the cell surface , rather than modulating the density of microtubules inside the cell . This would allow molecular regulators to modify the microtubule organization without directly affecting the rate of cellulose deposition . In our simulations , we observed that anisotropy varied according to curvature of face-face contacts of the cell , which is an emerging property of the weak anchoring mode . Cells that have sharper edges exhibit decreased array anisotropy compared to smoother cells . This result is interesting , as , for instance , epidermal cells possess edges with different curvatures [10] . Similarly , in the L1 layer of the shoot apical meristem , the outermost wall has a stronger curvature than all the walls separating the cell from its neighbors . Pavement cells also display a broad range of curvature values . Based on our results , higher anisotropy would be produced in L1 cells without the need for specific regulation . Such results are difficult to account for in models with microtubules constrained to surfaces embedded in 3D , in which for instance additional hypotheses linking curvature and catastrophe rate are implemented [10] . Nevertheless , our results do not preclude microtubule associated proteins from additionally modulating microtubule dynamics according to curvature or to other membrane-localized cues [52]; we only account for the default behavior of microtubule networks according to surface curvature . Similarly , changes in the curvature of the epidermal wall can occur through changes in internal pressure , which in return influence the anisotropy of the network [61] . An increase in the microtubule organization could be a result of an increased pressure , that would increase the curvature of the epidermal cell . Further experimental work is required to investigate the role of curvature on microtubule behavior and its relation to mechanical stress in the epidermis . Our simulations show that the cell aspect ratio has an important impact on the global orientation of the network . The predicted default behavior of microtubules is their alignment parallel to the long axis of a cell , due to the directional persistence of microtubules . This is in agreement with previous models with microtubules confined to surfaces embedded in 3D [11 , 51] , where the default orientation is longitudinal for long cylinders . This default state was observed with microtubules polymerizing in vitro inside elongated 3D chambers [11] . In slowly growing cells of the hypocotyl , microtubules are oriented along the long axis of the cell , whereas microtubules are circumferential in rapidly elongating cells [55 , 56] . Our model suggests that directional cues are needed to avoid this default orientation is growing cells . At the boundary between the shoot apical meristem and the primordia , cell division leads to cell shapes that are elongated along the axis of the boundary; our model predicts that microtubules will be oriented along the same direction by default , amplifying their response to mechanical stress [19] . In this study , we show that the microtubule network is oriented by default along the longest axis of the cell . However , microtubules in plants often show supracellular orientation , independently of cell shape , a behavior that has been ascribed to tissue-level signals , notably mechanical stress [18–20] . Moreover , It has been demonstrated that inside the cell , microtubules orientation is coupled to polarity markers such as proteins from the PIN FORMED and RHO OF PLANTS families [62 , 63] . Simulations have assessed how localized membrane heterogeneity could result in a biased orientation of the microtubule network [10 , 41] . In this study , we show that a weak directional cue influencing microtubule growth rapidly modifies the orientation of the network towards the direction of the cue . This cue could be due to mechanical stress , hormone gradients [22] , or to cytoplasmic streaming [23 , 24] , for instance . As such , the network behaves as a sensor translating an external directional information into a structural polarity inside the cell . The coexistence of a default orientation and a strong ability to reorient could shed a new light on orientation changes in cells . Changes in microtubule orientation need not be always related to specific regulation but may also be related to the arrest of signals and the return to the default state . This could be occurring in the shift from transverse to longitudinal orientation in hypocotyls responding to light or hormones [33 , 34 , 58] . The shape of the cell has little influence on mean length , number of microtubules or bundles proportions . In addition , anisotropy of the network is not highly correlated to changes in global cell shape . This prediction of a robust network suggests that plant cells do not need specific regulations to compensate for their great variations in shapes . Accordingly , the microtubule network appears as a good polarity system , with a default orientation and a high sensitivity to directional cues . It was recently shown that , for global polarity to emerge in a tissue , an important requirement is the existence of internal cellular polarity [64] . In this work we show that the microtubule network is suited for such a requirement . Overall , microtubules and associated proteins form a complex self-organizing system that is difficult to comprehend without resorting to models . The results obtained here demonstrate that our three-dimensional model provides a framework to test hypotheses on the regulation of the microtubule cytoskeleton in plant cells . The model given here is only the beginning to a more complete analysis . We have not yet incorporated microtubule severing [44 , 58] , microtubule branching [26 , 27] , nucleation at the nuclear envelope [28] , or the possible effects of connections between cortical microtubules and cellulose fibrils outside of the cell as mediated by the cellulose synthase complex [65] . Severing in particular has been shown to be key to microtubule reorientation [44] following mechanical signals [66] . We also have not included limiting levels of tubulin [51] , which could affect overall microtubule number . Altogether , we expect our model to help progress in understanding how microtubule self-organization integrates directional cues with three-dimensional cell shape and how microtubule-associated proteins modulate this integration .
Microtubules were coded as 3D multi-segment vectors of constant length . A ring of tubulin of the length of a dimer is represented as a unit vector in the simulation . Microtubule growth in the model occurs by adding one vector element at the plus end of the microtubule , at the position of the end of the last vector . In plants , microtubules are considered to be mostly static , their growth and shrinkage are the result of treadmilling processes . To code for microtubule directional persistence ( which relates physically to bending stiffness ) , the direction at which a new vector is added to an existing microtubule changes by a small random amount . Microtubule shrinkage occurs at the minus end by removing the first vector from the list . The main model parameters are shown in Table 1 . Typically , a cell has a width of several micrometers , and we take the unit of length as 8nm , the height of a ring of tubulin . Considering a measured speed of growth at plus ends of 3-5 μm/min [67] , a simulation time step is approximately 0 . 1 to 0 . 2 s . A typical simulation of 10000 time steps thus represents 15 to 30 min of real time . The microtubules are nucleated on the cell surface [25–27] at a constant rate . The default value is np = 4 . 7 ⋅ 10−7 per time step per unit surface , corresponding to 1 to 2 nucleation events per time step , or 5 to 20 per second . Once nucleated , the microtubules do not immediately shrink . At each of the time steps that follow nucleation , a microtubule has a probability ns to begin shrinkage . Once a microtubule has started to shrink , one vector is removed from the minus end at each time step . The cell contour is described with a triangular mesh . Each vertex is endowed with the information of the vector normal to the surface , which is used during the simulation to calculate a local approximation of the tangential plane . It is possible to add other informations at the vertex level that can be read during the simulation and serve as extrinsic input . Inputs can be scalars or tensors . The distance from any point in space to the membrane is calculated using the nearest point on the surface . At this point , the membrane is approximated by the plane perpendicular to the normal of the mesh . The distance between a tubulin ring ( unit vector ) and the membrane is calculated as the shortest distance between the endpoint of the vector and this plane . A collection of standard cell shapes was generated for our simulations ( Figs 1 and 6 ) . The main shapes were constructed starting from square parallelepipeds of dimensions 8 . 8μm×8 . 8μm×8 . 8μm ( cube ) , 9μm×9μm×4 . 7μm ( “square” ) , and 4 . 8μm×4 . 8μm×15 . 6μm ( “long” ) , which are comparable to typical plant cell dimensions . Then the meshes were smoothed so that the minimal radius of curvature was 1 . 3μm and 4 . 7μm for sharp and smooth shapes , respectively , which roughly spans the range 0 . 5-5μm measured in plant roots [10] . The ellipsoid shape is an ellipsoid of revolution around the long axis; the short and long axis have dimensions of 10 . 3 μm and 16 . 8 μm , respectively . The simulation progress is made through discrete timesteps . At each timestep new vectors are added and vectors are removed from the simulation space according to the rules specified in the previous subsections ( Nucleation and minus end behavior; Plus end growth ) . Collision tests are performed so as to implement the different growth or shrinkage rules . In order to increase the simulation speed , space is divided into subelements and the vectors are identified according to the subelement to which they belong , which diminishes the number of particles involved in the collision test ( locality-sensitive hashing [73] ) . To visualize microtubule density and orientations an image is created using a matrix of resolution ( rx , ry , rz ) . rz is larger than rx = ry by typically a ratio of 10 to 1 , which mimics the anisotropic resolution of a confocal microscope . The simulation space is then screened . When a vector is located inside a cube of the matrix , the value of this cube is incremented by one , and the immediate neighbouring cubes are incremented by a lower number ( typically 0 . 3 ) . At the end of the process , a stack is formed where microtubules appear as blurred intensity signals . One can either visualize each sub-image from the stack by moving along the z axis , or create a projection that sums the matrix along the z axis . We used the standard nematic order parameter to quantify anisotropy . The space is subdivided into cubes of arbitrary size , typically segmenting the structure into circa 27 cubes ( S2 Fig ) . Segmenting the structure into 216 smaller pieces gives similar trends , with globally higher anisotropy values . All tubulin ring directions are extracted as a 3 × N matrix , D . We then compute the square symmetric ( 3 × 3 ) matrix M = DT ⋅ D/N , where T stands for the transpose . M is diagonalised , yielding three eigenvalues λi , i ∈ 1 , 2 , 3 . Local anisotropy , A , of the microtubule array in each cube of space is defined as A=32 ( λ1−λm ) 2+ ( λ2−λm ) 2+ ( λ3−λm ) 2λ12+λ22+λ32;λm=λ1+λ2+λ33 . ( 3 ) The value of A is such that 0 ≤ A ≤ 1 . Anisotropy value is then computed as the average of the local value A over the whole cell . The simulations were run 5 times for each parameter value or shape considered; to avoid artificial correlations between data , only the last snapshot was considered for further statistical analysis . As data for the three default values of np were statistically identical , they were pooled together for Figs 1–5 . The plots were produced with the boxplot function of R: the boxes extend between the first and third quartiles , the segments in the box indicate the medians , and the whiskers are representative of extreme values . | Plants exhibit an astonishing diversity in architecture and morphology . A key to such diversity is the ability of their cells to coordinate and grow to reach a broad spectrum of sizes and shapes . Cell growth in plants is guided by the microtubule cytoskeleton . Here , we seek to understand how microtubules self-organize close to the cell surface . We build upon previous two-dimensional models and we consider microtubules as lines growing in three dimensions , accounting for interactions between microtubules or between microtubules and the cell surface . We show that microtubule arrays are able to adapt to various cell shapes and to reorient in response to external signals . Altogether , our results help to understand how the microtubule cytoskeleton contributes to the diversity of plant shapes and to how these shapes adapt to environment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"microtubules",
"chemical",
"compounds",
"cellulose",
"microtubule",
"dynamics",
"cell",
"processes",
"condensed",
"matter",
"physics",
"anisotropy",
"organic",
"compounds",
"membrane",
"proteins",
"materials",
"science",
"microtubule",
"polymerization",
"tubulins",
"cellul... | 2018 | The self-organization of plant microtubules inside the cell volume yields their cortical localization, stable alignment, and sensitivity to external cues |
The morphogenesis of organs necessarily involves mechanical interactions and changes in mechanical properties of a tissue . A long standing question is how such changes are directed on a cellular scale while being coordinated at a tissular scale . Growing evidence suggests that mechanical cues are participating in the control of growth and morphogenesis during development . We introduce a mechanical model that represents the deposition of cellulose fibers in primary plant walls . In the model both the degree of material anisotropy and the anisotropy direction are regulated by stress anisotropy . We show that the finite element shell model and the simpler triangular biquadratic springs approach provide equally adequate descriptions of cell mechanics in tissue pressure simulations of the epidermis . In a growing organ , where circumferentially organized fibers act as a main controller of longitudinal growth , we show that the fiber direction can be correlated with both the maximal stress direction and the direction orthogonal to the maximal strain direction . However , when dynamic updates of the fiber direction are introduced , the mechanical stress provides a robust directional cue for the circumferential organization of the fibers , whereas the orthogonal to maximal strain model leads to an unstable situation where the fibers reorient longitudinally . Our investigation of the more complex shape and growth patterns in the shoot apical meristem where new organs are initiated shows that a stress based feedback on fiber directions is capable of reproducing the main features of in vivo cellulose fiber directions , deformations and material properties in different regions of the shoot . In particular , we show that this purely mechanical model can create radially distinct regions such that cells expand slowly and isotropically in the central zone while cells at the periphery expand more quickly and in the radial direction , which is a well established growth pattern in the meristem .
Mechanical forces are integral part of any living system and recent data is confirming their importance as signaling cues in animal and plant development [1]–[3] . This may be especially important for plants which have to sustain large loads while executing a developmental program that is optimal in their habitat [4] . Due to the lack of cell migration , plants must change the mechanical properties of their tissues on the cellular scale in order to facilitate directional growth of organs . The mechanical properties of plant tissue can be linked down to the properties of cell walls . The walls are composed of a network of cellulose microfibers interconnected by polysaccharides and xyloglucans [5]–[7] . They constitute the structurally strong element of plant tissue providing support against turgor pressure and internal tension . From a mechanical point of view , the walls can be considered to be thin visco-elastic elements . The epidermis of plant tissue is thought to play a special role in morphogenesis [8] , [9] . It is generally more mechanically stiff than internal tissues , which suggest a ‘tissue pressure’ model where tensional forces in the epidermis are generated by the pressure and growth of the internal cells [8] . Under the action of hormones or enzymes the epidermis can experience substantial changes in its mechanical properties [10]–[12] , which is determinant in the outgrowth of plant organs . The prevailing idea of how an isotropic tissue pressure generates anisotropic growth has to do with the anisotropy of plant material . The cellulose microfibers , which have been shown to have highly organized directional patterns in the epidermis [13] , [14] , restrict the elastic expansion of a tissue in the direction parallel to them . The organization of the wall fibers is regulated by cells via the deposition of cortical microtubules [15] . This fact has been exploited by experiments which often use microtubule direction as a proxy for fiber direction . While directional fibers can translate the isotropic forces into specific strain directions , additional mechanisms for long-term plastic anisotropic growth are also needed . The data suggests that such growth is the result of a molecular break and slip behavior with new material constantly being added to the walls [16] , [17] , where plastic growth is triggered by the stresses in the wall exceeding a yield threshold . When anisotropic material is generated by adding strong fibers , the picture becomes more complex , and the idea for how the growth proceeds is that weaker molecules connecting the fibers break and allow for extension in the direction perpendicular to the fibers [17] . While simple models of plant growth have been developed , a model for plant tissues that is compatible with the stress-based growth and anisotropic cell wall material has not been defined [18]–[20] . The composition of the plant cell wall is controlled by the genetic program of the cell which must allow for a large degree of adaptivity for the whole plant , the existence of specialized tissue types and the wealth of plant forms . However , as recent evidence [3] , [21] and previous ideas [22] suggest , it is likely that reciprocal signaling , linking mechanical states of the tissues and cell walls to biochemical processes takes place too , connecting growth rate and direction with mechanical properties of the plant tissue in a feedback loop . Molecular details of the mechanism of such two way relations between mechanics and cell functions are still elusive and require further investigation [7] . In particular , the organization of the cellulose fibers , which leads to a directional growth , may be determined by several cues . One suggestion is that the fibers align orthogonally to the maximal strain direction . This has been proposed for anisotropically growing tissues [23] , [24] . A more recent suggestion is that the fibers align in the maximal stress direction [25] , which is supported by the fiber patterns observed in the plant meristem [3] , [26] . For isotropic mechanical materials the two ideas would be easy to discriminate between , because then maximal principal strain and stress point in the same direction . However , for mechanically anisotropic plant walls , maximal strain and stress may very well be orthogonal and it may not be easy to discern between the two rules of fiber alignment . The situation is complicated further by the fact that a change in the fiber direction will lead to a change in stresses and strains resulting from the same external load . This complex feedback loop makes it difficult to predict a priori whether either stress or strain directions can act as stable inputs for shape generation , even if these directions are easily predicted given the material anisotropy . The intricate dynamics of fiber alignment resulting from such feedbacks has yet to be investigated in detail . Mechanical strains and stresses in tissues are not easy to measure , so there is a need for reliable mechanical models of biological materials that can quantitatively predict both magnitude and direction of the strains and stresses . Using such models , after prescribing the material properties and loading forces , one can accurately describe the mechanical response of the tissue and further test different scenarios for how the mechanical response is coupled to biochemical signals . There exists a large variety of finite element or particle based methods which can be applied to modeling mechanical responses of materials [27] . These methods , however , are usually quite computationally intensive and large scale models of biological cells are not always feasible within them . In addition these methods have not been designed or optimized to cope with the dynamic complexity of biological materials and the growth of tissues , which require rapid changes to the model's cellular topology and material composition . Given the geometry of a plant cell wall , where its thickness is often more than an order of magnitude smaller than its planar extension , finite element method ( FEM ) shell models provide an adequate description since they are specifically designed for thin curved surfaces and describe tensile and bending behavior ( Figure 1A ) [28] , [29] . More recently , Triangular Biquadratic Spring ( TRBS ) models have been developed to describe two-dimensional elastic elements [30] . TRBS has the benefit of simplicity: this class of model describes mechanical responses using just the resting and current lengths of the triangular edges ( {} , {} in Figure 1B ) . The TRBS implementation has been shown to accurately represent continuum properties of mechanics [30] . However , since in TRBS the bending energy is disregarded , it is not obvious that such models provide a good description of plant walls that typically consists of curved structures . In this paper we develop two implementations of a mechanical model for anisotropic plant wall material: a FEM shell implementation and a TRBS plate implementation . We compare the implementations both in in-plane loading simulations and in tissue pressure models of the plant epidermis , the latter leading to additional bending moments in shells ( Figure 1C–E ) . We analyze the relation between maximal ( first principal ) stress and strain directions under different loading forces . We use the method to analyze different proposed mechanisms of coupling between mechanical cues and alignment of material anisotropy of cells , based on perception of either maximal stress direction ( MSD ) or the direction orthogonal to maximal strain ( OsD ) . We apply the models to different geometries representing different tissues in plants in order to evaluate their potential for explaining cellulose fibers patterns and growth patterns observed in epidermal plant tissues ( Figure 1D–E ) .
One of our goals was to establish an efficient computational method and a sufficiently accurate material model that can be used to simulate the behavior of plant walls . In particular , we aimed to investigate whether a Triangular Biquadratic Spring method can provide a reliable description , given that it is a two-dimensional representation and that it does not explicitly include any bending resistance . To do this we developed a TRBS method and compared the results with a shell-based finite element method ( Methods and Text S1 ) . To describe the anisotropic wall material , we used a hyperelastic strain energy density formalism applicable to large strain deformations ( Text S1 ) . For the isotropic wall material we used a St . Venant-Kirchoff description [3] , [30] , and developed an anisotropic material model penalizing extension in a defined fiber direction ( Equations 1 , 3 and Text S1 ) . First , we tested a mechanically isotropic square patch of elements under different loading conditions and for different material properties ( Figure 1C ) . When we applied uniaxial tensions , the stress-strain relations completely agreed between the methods ( Figure 2A–B ) . Further , the two methods agreed for a wide range of Young moduli and Poisson ratios under isotropic loading forces with a difference of less than 0 . 1 percent between them ( Figure 2C , Figure S1A ) . Note that the principal stress value is a monotonically increasing function of not only the Young modulus but also of the Poisson ratio for this mechanical model ( Figure 2B ) . We extended the uniaxial tension tests into a large deformation regime to demonstrate the well known deficiency of the St . Venant-Kirchoff material model , where uniaxial loading forces can result in infinite stresses and zero volume at finite strains [31] . We found that this deficiency appears especially when the Poisson ratio is high ( Figure 2A–B ) . In simulations of plant tissues , we do not expect strains to exceed several percent , which corresponds to the typical values 5–10% encountered in experiments [32] , and as such the model provides an appropriate description of plant wall material . Next , we analyzed the response of the anisotropic material model for the square patch of elements under biaxial loading forces . Under isotropic loading forces an increased degree of material anisotropy led to an increased difference between the magnitude of principal stresses ( Figure S1B ) , and the maximal stress and strain directions were perpendicular to the fiber direction . Under anisotropic loading forces , the response depended on the angle between the maximal force direction and the direction of the axis of material anisotropy ( Figure 2D ) . When the material anisotropy direction coincided with the direction of the maximum loading force , the maximal principal stress value was lower than when those directions were perpendicular . This could have profound implications for plant wall mechanics . Since stresses trigger inelastic responses and breakage of brittle components of a material [33] , a plant cell's ability to control the amount of stress in the tissue by adjusting its anisotropy could be a way of directing growth given the stress magnitude's relation to the yield stress of the wall material [5] . To assess the importance of the lack of bending resistance in the TRBS method we compared principal stress pattern , principal strain value and deformation with the FEM shell method for a pressurized quadrilateral plate ( Figure 1C , 2E ) , different plant-like geometries ( Figure 1E , Figure S1C–D ) , and a saddle-like plate ( Figure S1E ) . The results showed good agreement between the two methods for the pressurized quadrilateral plate suggesting that the deformation in our tissue pressure model is dominated by tensile and not bending stress ( Figure 2E ) , although we found small quantitative differences . For example , the normalized distribution of equivalent von Mises strain for the TRBS method had a slightly higher average ( 0 . 052 vs . 0 . 049 ) ( Figure 2F ) probably owing to the lack of bending energy at the junctions . The agreement held for most geometries tested ( Figure S1C–D ) , with exceptions where compressive forces generated buckling ( Figure S1E ) –yet , even in such cases the qualitative pattern and distribution of stresses was in good agreement between both methods . The good agreement of the two methods indicates that tensile stresses dominate over bending moments , but also that although the TRBS approach does not explicitly account for bending energies at individual edges , the triangulated mesh structure may still incorporate a resistance towards bending via stretch and compression of the elements induced by bending . In conclusion , we have shown that TRBS and shell finite element methods strongly agree when applied to models of anisotropic wall material in two dimensions for a wide range of values of material anisotropy and applied forces . Although quantitative differences appear , the methods also show strong agreement in the case where two-dimensional structures are pressurized into three-dimensions and where bending forces are induced . We also found that for an anisotropic material under anisotropic loading forces a complex relation between the direction of maximal load and the directions of the maximal strains and stresses appear , indicating that plant cells can control these variables if they are able to control cellulose fiber directions . To analyze the relation between stress and strain directions under different loading forces and for different fiber directions we first investigated a situation where the direction of maximal applied force coincided with the fiber direction in a simple square . The maximal stress direction always followed the maximal loading force direction . Depending on the degree of anisotropy of the applied force and the material properties , the resulting maximal direction of strain could be either parallel or perpendicular to the maximal stress direction marking distinct regions in the ( force-anisotropy , material-anisotropy ) parameter space ( Figure 3A ) . As expected , for isotropic materials the maximal principal stress and strain directions both coincided with the maximal applied force direction . For anisotropic materials and anisotropic loads we obtained a region where maximal stress and strain directions can be perpendicular ( black region in Figure 3A ) . The extension of this region depended on the Poisson ratio of the material ( Figure S2A–C ) . Given a fixed material anisotropy ( dashed line in Figure 3A ) , isotropic loading leads to parallel directions of the maximal stress and strain . A higher directional force can be resisted by the stronger component of the material leading to a maximal strain direction perpendicular to the maximal force direction . However , when the forces are highly anisotropic then they overcome the resistance of the stronger component of the material and the maximal strain follows the direction of the applied force . This reveals that potential cellulose fiber orienting mechanisms based on the feedback from either stress or strain can behave differently from one another in some parts of a tissue while in other parts of the tissue they show the same behavior . We used the models to analyze the anisotropic growth of shapes resembling plant organs where the alignment of fibers in epidermal tissues is thought to guide growth . We simulated a cylindrically shaped tissue using the tissue pressure model and parameter values from experimental estimates [34]–[36] and recovered the expected stress about half in the longitudinal direction compared to the circumferential direction . We set the fiber direction to be circumferential to match observed microtubule directions in the epidermis of several plant tissues [3] , [37]–[39] . This led to a maximal direction of stress in a circumferential direction and of strain in a perpendicular , longitudinal direction ( Figure 3B–C ) . If we use elastic strain as a proxy for growth ( see Discussion ) , the result of this simulation corresponds to the idea that organ growth is perpendicular to the fiber direction , so extending the organ along its main axis . The experimentally observed circumferential direction of fibers seems to be explainable equally well by either the model where fibers orient perpendicularly to the direction of maximal strain ( OsD ) or by the model where fibers orient in the direction of maximal stress ( MSD ) ; both have been suggested as informative signals for fiber directions in plant tissues [3] , [23] , [24] , [40] . To analyze the different consequences of these different variables acting as signaling cues for the fiber directions , we introduced a dynamic description of the wall material properties , where the deposition of new cellulose fibers leads to changes in magnitude and direction of the mechanical anisotropy of plant walls ( Methods , Equations 8–10 ) . We assumed a constant addition of fibers to the walls with the anisotropy of the deposition guided by the anisotropy of the directional signal , i . e if the input signal is isotropic , the material will be isotropic , while an anisotropic input signal will result in an anisotropic material . Interestingly , the MSD and OsD hypotheses gave very different results , in spite of the fact that maximal strain and stress directions were perpendicular when fiber directions were fixed as descibed above . In the case of the stress based feedback , the fiber direction was identical to the fixed anisotropy direction case ( Figure 3E–F ) , whereas in the case of ( orthogonal ) strain based feedback the initial , circumferential fiber direction became unstable and subsequently reorganized into the longitudinal direction ( Figure 3H–K , Video S1 ) , in contrast to the circumferential orientation of microtubules observed in experimental data . A more detailed analysis of the influence of material and loading force anisotropy on the MSD and OsD material models showed that the former model results in regions of mutually parallel and orthogonal strain and stress ( Figure 3D ) . The extension of the region with perpendicular stress and strain directions was similar to the static anisotropy direction case ( Figure 3A , D ) , indicating that orthogonal directions of maximal stress and strain constitutes a robust stable situation for the MSD dynamical model ( Figure 3E , Figure S3 ) . In the OsD model , the region of orthogonality between stress and strain disappeared completely ( Figure 3G ) , indicating that this is an unstable situation for the OsD dynamical model ( Figure S3 ) . Independently of the anisotropy of the forces causing elastic deformation , the maximal stress and strain directions always became parallel ( Figure 3H–K ) . In conclusion , we have shown that in a situation where internal tissue is providing tension to the epidermis , an extension along the longitudinal axis of the organ can be explained by fibers resisting strain in the circumferential direction . This was clearly seen in a model where static fibers were laid out according to the experimentally observed pattern that results in a maximal strain that is orthogonal to the fiber direction . When the fiber directions were allowed to be reoriented by mechanical cues , more intricate dynamics was generated . A model where fibers aligned in the direction of maximal stress robustly preserved the circumferential directions of the fibers , as seen in experiments . On the contrary , a model where fibers aligned perpendicularly to the maximal strain direction led to the initial circumferential fiber pattern becoming unstable and reorienting into the longitudinal direction . To test the dynamic stress feedback fiber model on a template with varying curvature , we applied the tissue pressure model to a paraboloid template , as a proxy for a naked meristem , in which the curvature is isotropic at the apex and smoothly becomes anisotropic across the periphery ( Figure 1E ) . The dynamic changes of material properties in the cells resulted in a region of isotropic material at the apex and anisotropic material towards the periphery ( Figure 4C ) , corresponding to isotropic stresses at the apex and anisotropic stresses in the periphery ( Figure S5B ) . The dominant fiber direction oriented circumferentially around the central zone ( Figure 4A ) , as previously reported in experiments and models [3] , [40] . Remarkably , the switch from isotropic to anisotropic material ( and stresses ) in the radial direction was quite rapid , so creating a spontaneous zonation within the meristem purely from mechanical interactions . This corresponds to the very sharp transition between regions of parallel and perpendicular alignment of maximal stress and strain in the parameter space of material and loading force anisotropy ( Figure 4A , D , cf . circles in Figure 3D ) . Therefore , even though these parameters change smoothly in the radial direction of the meristem , the dynamic material model creates an abrupt transition between the regions . The extent of these regions depended on model parameters , but the switch-like behavior was a robust feature of the stress feedback model ( Figure S4 ) . The meristem has a central zone with slowly growing and dividing cells , and a peripheral zone where cells grow more quickly [41]–[43] . The cell expansion rates in the simulations also reflected the zonation ( Figure 4B ) . The model predicted a slow isotropic expansion rate in the central zone and a comparatively high radially oriented expansion rate in the periphery , correlating well with strain directions reported for meristems [42] . Next , we looked in more detail on the effects of the dynamic update of material anisotropy direction and intensity on a geometry where there is a primordium at the periphery of the meristem with a valley in between ( Figure 1E ) . Previously , we have shown that a tissue pressure model of the epidermis applied to a meristem shape leads to isotropic stress in the central zone while a valley in between the meristem and a primordium develops anisotropic stress . A simple spring model using a stress feedback generated fiber directions comparable to the measured microtubule directions in different areas of the meristem [3] . In the TRBS model , the stress feedback generated similar material fiber patterns ( Figure 4D , Video S2 ) , while the orthogonal strain feedback failed to generate these directions ( Figure S5E , Video S2 ) . In the valley between the meristem and the primordia , the stress feedback resulted in a fiber alignment along the valley and a high stress anisotropy ( Figure S5D ) : the model predicted parallel maximal stress and strain directions in the valley ( Figure 4D ) . While the alignment between maximal stress and strain directions in the central zone is a consequence of isotropic stresses in this region , alignment in the valley , in spite of anisotropic material , is caused by highly anisotropic stress ( Figure 3D , blue and red circles , respectively ) , resulting in a maximal strain direction along the valley . Maximal directions of stress and strain were perpendicular elsewhere ( Figure 4D , cf . Figure 3D , green and yellow circles ) , the same as in the periphery of the naked meristem simulation ( Figure 4A ) . In summary , we have shown that a stress feedback model is able to explain the patterns of microtubular organization seen in experiments . This feedback generates a relatively sharp zonation within the meristem , providing a purely mechanics-based explanation of strain magnitudes and directions inferred from experiments , where the central zone has a lower rate of isotropic expansion and the periphery a higher rate of radially directed expansion in spite of the circumferential stress direction . The model also predicts that highly anisotropic stresses generated in the boundary between the meristem and a primordium can lead to a maximal strain direction parallel to the maximal stress direction in this region . Next we analyzed how dynamic properties of wall material affect elastic deformations locally and at a tissue scale . When anisotropic forces are applied ( i . e . when curvature is higher in one direction in our tissue pressure models ) , the stress feedback model always aligns the fibers parallel to the maximal force so reducing the deformation in this direction and procuring locally a more isotropic deformation . Even in the case with strong anisotropy of the loading forces where maximal strain and stress are parallel ( Figure 3Bi , cf . boundary region between meristem and primordia ) the stress feedback model leads to a more isotropic strain field compared to an isotropic material of the same total elasticity . Since the strain feedback model aligns the fibers perpendicularly to the loading forces , this feedback tends to increase local strain anisotropies . To quantify these differences , we tested both material anisotropy feedback mechanisms within the TRBS model where geometries with different degrees of shape anisotropy were pressurized ( Figure 5 ) . When compared to isotropic material , the stress feedback model led to more isotropic strain , and the difference increased with the anisotropy of the geometry and hence with the loading force ( Figure 5C ) . In contrast , the orthogonal strain based feedback model led to increased strain anisotropy when compared to an isotropic material ( Figure 5C ) . This local difference had an impact on the resulting global deformation of the structure , where the stress feedback model promoted the maintenance of the geometrical anisotropy while the orthogonal strain feedback model decreased this anisotropy ( Figure 5A–B ) . Next we tested the different feedback models on our meristem-like template . The stress feedback model resulted in a more prominent anisotropic shape change at the meristem and primordium apices , promoting the upward movement of the shoot and a more directed shape change of the primordium ( Figure 5D ) . Also , the stress based feedback model resulted in a more pronounced valley between the meristem and the new organ ( Figure 5D , Video S2 ) . The changes of these features of the meristem have been seen experimentally when comparing wild-type plants and plants treated with oryzalin , a drug that depolymerizes microtubules and is assumed to lead to a more isotropic material [44] . In summary , a stress feedback to fiber directions enables plant walls to resist internal forces , which locally generates more isotropic elastic strains , and which at the same time counteracts to the tissue pressure forces acting towards isotropic curvature and hence a stress feedback maintains the shape of anisotropic structures .
The coordination of the changes in mechanical properties across a growing plant tissue is crucial for the creation of the complicated forms and shapes observed in plants [7] , [21] . In the shoot apical meristem a connection between the organization of the tissue's mechanical anisotropy and the perception of mechanical stress signals has been suggested [3] , [40] , while a competing idea that fibers organize perpendicular to the strain direction has emerged motivated by the correlation between growth and fiber directions in anisotropically growing organs [23] , [24] . In this study we analyzed two models in order to compare these mechanisms for how mechanical cues feed back to orient cellulose microfibers . In our simulations of the stress and strain patterns in the epidermis on a stem-like geometry , where fibers are fixed to be aligned circumferentially , maximal stress is circumferential and maximal strain is longitudinal , in accord with both experimentally motivated suggestions for organizing fiber directions . However , in models of dynamic orientation of cell wall mechanical anisotropy driven by stress or strain , we observed drastically different results for each of the two models . In the stress based feedback model the circumferential alignment of the fibers as well as the perpendicular orientation of maximal stress and strain directions can be robustly maintained ( Figure 3E ) . In contrast , the orthogonal to strain based feedback model results in a longitudinal alignment of the fibers and parallel , circumferential directions of maximal strain and stress , which contradicts the experimentally observed orientation of the cellulose fibers and microtubules ( Figure 3H ) . When simulating the more complex shapes appearing around the shoot apical meristem , the orthogonal strain feedback model again failed to explain the microtubule patterns seen in experiments . The stress feedback model translated the smooth increase of anisotropic curvature in the radial direction to a switching between different material properties in a central and a peripheral zone ( Figure 4 ) . It should be noted that even if there is an instant shift of strain and stress directions from mutually parallel to perpendicular , this does not represent a discontinuity in the model since the strain direction is degenerate ( isotropic ) when crossing these boundaries ( Figure S5A , C ) . A mechanical radial zonation has recently been suggested by experiments and models [32] , [45] , but in our model different properties of the material in different areas of the tissue are not dictated by an arbitrary specification of the separate regions . They are instead a natural consequence of the stress feedback model reacting to the differences in shape , curvature and stress response and stress anisotropy in different regions of the meristem . The alignment of maximal stress and strain directions in the central zone is a consequence of the material being isotropic in this region . The analogous alignment in the valley between the shoot and a primordium , which occurs in spite of the material's anisotropy , is caused by a highly anisotropic stress . The perpendicular maximal stress and strain directions in the periphery are a result of an anisotropic material and anisotropic forces , but where the forces are sufficiently opposed by the fibers to create a perpendicular strain direction . Such radial growth direction has been reported in experiments [42] . It is interesting to relate the spontaneously formed mechanical patterns to the known radial expression patterns in genes regulating differentiation [46] . It has been recently shown that the stem cell regulator WUSCHEL , expressed in the central regions of the shoot , moves between cells and directly activate the stem cell marker CLAVATA3 , expressed in the apical region of the meristem . WUSCHEL also represses genes that are important for differentiation . The combined gene regulatory network is sufficient to explain the radial expression zonation in the meristem [47] . How this molecular network interacts with mechanical properties is an interesting question for the future . While there might not be a direct interaction , both the mechanical and molecular models do depend on the shape of the meristem to generate a radial zonation and can hence affect each other's radial zonation via the geometry changes . Our simulations performed on templates resembling the shapes of the stem and meristem with outgrowing primordia show that a stress based feedback produces deformations which result in more elongated shapes of outgrowing organs while an orthogonal to strain feedback tends to round and level the protrusions of the surface ( Figure 5 , Video S2 ) . Interestingly , this is a consequence of the stress based feedback having more isotropic strain locally , compared to an isotropic material or a strain based feedback mechanism . We have compared the results of continuum mechanics simulations using a Triangular Biquadratic Spring method with more detailed simulations using a nonlinear shell Finite Element method . We found that the methods are in agreement for both stretching and bending pressurized tissue simulations that are used to represent epidermal plant tissue . This shows that TRBS , despite its simplified treatment of geometry and its lack of bending resistance , offers an adequate level of accuracy for the purpose of modeling plant tissue . Owing to its simplicity the TRBS method will prove useful for more complicated three dimensional models involving cell growth and proliferation and thus requiring changes in model topology . The assumption of modeling the internal cell layers as a simplified tissue pressure contribution can be analyzed further in future work , which can allow for a more complex interaction between internal layers and the epidermis in the meristem [9] , [11] . Our simulations suggest that this will improve the description mainly in situations with a negative curvature and compressive forces , e . g . in the boundary between the meristem and a primordium . Our simulations overestimate the strain rates in these regions and the absence of internal tissue can lead to buckling ( Figure S2E ) . For anisotropically shaped organs , this may also be important for the stress directions . For stem tissue it has been shown that the internal tissue exerts a longitudinal force on the epidermis with longitudinally oriented fibers [48] . The stress feedback model applied to our cylindrical template in the presence of large logitudinal forces will lead to fibers in the longitudinal direction ( Figure S6 , cf . red circle in Figure 3D ) , which is in accord with patterns seen in experiments [49] . However , the appearance of anisotropic longitudinal growth in internal tissues is still not understood in such a scenario [48] , and will probably require more data on fiber orientations in several cell layers [38] . Another challenge will be to integrate current models with long-term plastic growth of plant cell walls . Plastic growth is described as being triggered by wall stresses above a yield stress , which induces a break and slip behavior [50] , while we have compared elastic strain in the simulations with the plastic growth in experiments . While this might seem to be a contradiction , as we show that often the maximal stress and strain directions are perpendicular , it would be easy to remedy this difference . Either stresses in the isotropic matrix part of the wall could be used , which is the same as the strain , or the growth direction could follow the maximal stresses , but be oriented perpendicular to the fibers . Interestingly , our model predicts that a matrix stress idea and a stress perpendicular to fibers idea for growth can be discerned by a detailed measuring of growth directions in the boundary between the meristem and the new primordia , since there the fiber and strain directions are parallel . In any scenario , the maximal stress direction would not provide a good cue for plastic growth , since this would counteract the possibility to generate anisotropically shaped organs . There are no experimental molecular data on how stress or strain sensing mechanisms work , although several suggestions have been proposed for how it could be realized [7] , [51] . The recent data show that microtubule-severing protein katanin is required for the cell's response to mechanical signals in plants [26] , and several examples in animals show that proteins can act as mechanosensors , e . g . [52] . The development of detailed mechanical models will be integral for understanding morphogenesis in development . It will open up new venues of research for understanding whether mechanical cues are one of the main drivers of shape changes , and more importantly it will allow the development of integrated models where gene regulation and molecular signaling feed back to each other for describing the combined effects of differentiation and morphogenesis .
There exist many material models which parametrize elastic energy in terms of combination of deformation tensor invariants in different ways and describe behavior of different types of materials . In the simplest isotropic material case the TRBS uses a St . Venant-Kirchoff description , which is an extension of a linear material model . The strain energy density , , in this material model becomes ( 1 ) where and are Lame coefficients representing material elasticity and is the Green-Lagrange strain tensor . The advantages of this material model are the simple energy form and a clear interpretation of material properties . We assume plane stress condition where Lame constants can be expressed as ( 2 ) Here and are the Young modulus and Poisson ratio that represent elasticity and incompressibility of the material , respectively . In order to extend this material model for transversely isotropic materials we considered two sets of Lame constants , one for longitudinal and one for transverse to anisotropy direction [53] . To ensure that the energy expression is not over-penalized in the anisotropy direction we first equipartitioned the energy into three terms each corresponding to one of the principal directions ( longitudinal and two transverse directions with respect to fibers ) . Then we have penalized only the term corresponding to the anisotropy direction . A procedure which do not take into account equipartitioning of the energy overestimates the contribution of the anisotropic part [53] . The increased energy cost of deformation in direction of the fiber , , is then described by ( 3 ) where the anisotropic part contains invariants of a strain tensor constructed with a vector in the direction of the fibers . The and are the differences between longitudinal and transverse Lame coefficients which are in turn related to Young modulus in longitudinal and transverse directions ( , ) and Poisson ratio ( 4 ) where follows similar relation as in Equation 2 and is representing the elasticity of the material in the transverse directions . The total energy density , , including an isotropic term for the matrix and an anisotropic term for the fiber becomes ( 5 ) which can be used for calculating the stress tensor and forces applied on the nodes of the meshed structure ( Text S1 ) . The expression for St . Venant-Kirchoff energy ( Eq . 1 ) is based on the Green-Lagrange strain tensor , , which can be expressed in terms of a deformation tensor . The second Piola-Kirchhoff stress tensor , which is the energy conjugate of the Green-Lagrange strain tensor , yields the stress in the resting shape . For evaluating strain and stress in the deformed shape , which is the current configuration , we calculated Almansi strain and its energy conjugate , Cauchy or true stress tensors , respectively ( Text S1 ) . The stress in case of TRBS was calculated under the assumption of plane stress and in case of shell description we visualized the stress integrated over thickness in order to be comparable to the corresponding values for TRBS . All of these tensors are two dimensional for TRBS elements and three dimensional for shells . The relative stress ( strain ) anisotropy measure , , can be defined as ( 6 ) where and are first and second stress ( strain ) eigenvalues respectively . We consider only tensile stress ( strain ) to be relevant and compressive stress ( strain ) values were set to zero in the equations . In most simulations the magnitudes of stresses ( strains ) are of the same order of magnitude and such relative measure is appropriate . However , in the case of our meristem-like template with outgrowing primorium , stress ( strain ) magnitudes extend over a large range . In such a scenario , a relative value can overestimate an anisotropy measure in regions of low stresses ( strains ) , and it is more appropriate to include the stress ( strain ) magnitude itself in the measure . In this case we used ( 7 ) where is the largest stress ( strain ) value throughout the template ( excluding boundary effects ) . We have normalized the value of anisotropy measure since we use as an input to the material model an expression assuming the value of this parameter to be between 0 and 1 ( see next section ) . In most cases the anisotropy measures based on both definitions follow the same trend but in more complicated geometries , where strain and stress values are small , there can be significant differences between the two measures ( e . g . the primordial apical region in the meristem-like template ) . Since plant tissue is characterized by different and dynamically changing anisotropic material properties we have devised a model which allows for smooth temporal and spatial changes of anisotropy . The model assumes that stress anisotropy plays a role in defining the degree of material anisotropy while the average elastic strength of the material is maintained . We used a non-linear relation between stress and material anisotropy which saturates when stress anisotropy is maximal ( Figure 6 ) . The relations between longitudinal , , and transverse , , Young modulus and anisotropy measure , , can be written as ( 8 ) where and are model parameters and and are Young moduli of the isotropic matrix and anisotropic fiber part respectively . We implemented a delay in the update of longitudinal and transverse Young moduli ( fiber model ) as well as anisotropy direction of individual cells ( stress feedback ) to take into account different time scales of propagation of mechanical and biochemical interactions . Such approach also results in more stable simulations . The Euler steps for updating longitudinal and transverse Young modulus are ( 9 ) where determines the time delay , is the time step , is the current value and is the new values calculated from Equation 8 . Similarly the update for anisotropy direction is done based on ( 10 ) where is the current anisotropy direction vector and is the maximal stress direction vector , and again sets the time delay . The mechanical simulations have been performed with in house developed software optimized for simulations of cellular structures . Both methods used in our simulations ( TRBS and shells ) are based on the FEM approach , which relies on the division of the domain of interest into simpler geometrical elements ( meshing ) and looking for the solution of the continuous mechanics equations in the basis of the functions which are local to each element . In case of shell FEM simulations we have used quadrilateral shell elements within extensible director formulation [28] ( Figure 1A ) . The implementation of TRBS was based on the explicit procedure used previously in simulation of biological materials [30] ( Figure 1B ) . We triangulated the polygonal cells via adding a vertex at the centroid position . Since we used a single fiber direction in cells , we averaged stress or strain input from the individual triangles . In our simulations both explicit Newark and implicit solvers with Newton-Rapson iteration were used for the shell finite element implementation while explicit forth order and adaptive fifth order RungeKutta methods were used for TRBS . The material parameters used in the simulations of plant-like structures ( Figures 3 and 4 ) were matched to the experimental estimates from similar materials [34]–[36] . We have used Young modulus in range 40 –50 and 100 –120 for isotropic and anisotropic part of the material , respectively . Poisson ratio was set to 0 . 2 and turgor pressure 0 . 2 . We assumed a thickness of epidermal material of 1 and a cell size of order 10 to 20 . In the fiber model we have used and . For updating anisotropy directions and material properties using the equations 8 and 9 we used for anisotropy direction update and for material properties update . As long as small values for update rates were used the results were not sensitive to the exact value of those parameters . We have used smaller update rates for material properties update , assuming the change in material properties is a consequence of microtubular dynamics and should be delayed respect to the anisotropy direction update . These parameters resulted in the deformation of order 5% to 10% in agreement with experimentally reported estimates [32] . We have used fixed ( clamped ) boundary conditions for our simulations of pressurized templates , which means that there was no deformation on the open boundary edges of the simulated structures . Since such conditions are not exact for real plant organs and can affect the results of simulations close to the boundary we excluded those regions from the analysis . The effects of the boundary conditions can be seen in Video S2 . The simulation tools are in house implementations and the latest versions are publicly available via a subversion server upon request and the current versions are available as Supplemental Information . | Development and morphogenesis of tissues are dependent on a coordination between cell differentiation , proliferation and growth . Plants , which lack cell migration , control directional growth of tissues by adjusting cellulose fiber directions so forming the organ shapes . It has recently been shown that mechanical cues can guide these fibers . We developed detailed mechanical models to investigate how fiber directions may be responding to mechanical cues and what consequences this may have for positional and directional growth patterns . We show that a model in which fibers align to maximal stress directions spontaneously generates a radial zonation in the shoot , recapitulating the slowly growing center and more rapidly growing peripheral region previously observed in the meristem . These radial patterns emerging from mechanics are in striking correspondence to the expression patterns of the genes important for stem cell maintenance , which attain similar radial domains . We also show that the stress model can robustly define anisotropically growing organs , which emphasizes the potential importance of stress in generating correct organ shapes in plants . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Models"
] | [
"physics",
"tissue",
"mechanics",
"developmental",
"biology",
"growth",
"control",
"plant",
"growth",
"and",
"development",
"biophysics",
"simulations",
"biology",
"computational",
"biology",
"morphogenesis",
"biophysics",
"biomechanics",
"cell",
"mechanics"
] | 2014 | Stress and Strain Provide Positional and Directional Cues in Development |
Integrative approaches to studying the coupled dynamics of skeletal muscles with their loads while under neural control have focused largely on questions pertaining to the postural and dynamical stability of animals and humans . Prior studies have focused on how the central nervous system actively modulates muscle mechanical impedance to generate and stabilize motion and posture . However , the question of whether muscle impedance properties can be neurally modulated to create favorable mechanical energetics , particularly in the context of periodic tasks , remains open . Through muscle stiffness tuning , we hypothesize that a pair of antagonist muscles acting against a common load may produce significantly more power synergistically than individually when impedance matching conditions are met between muscle and load . Since neurally modulated muscle stiffness contributes to the coupled muscle-load stiffness , we further anticipate that power-optimal oscillation frequencies will occur at frequencies greater than the natural frequency of the load . These hypotheses were evaluated computationally by applying optimal control methods to a bilinear muscle model , and also evaluated through in vitro measurements on frog Plantaris longus muscles acting individually and in pairs upon a mass-spring-damper load . We find a 7-fold increase in mechanical power when antagonist muscles act synergistically compared to individually at a frequency higher than the load natural frequency . These observed behaviors are interpreted in the context of resonance tuning and the engineering notion of impedance matching . These findings suggest that the central nervous system can adopt strategies to harness inherent muscle impedance in relation to external loads to attain favorable mechanical energetics .
The capability of skeletal muscles to deliver mechanical power is key in determining the neuromechanical performance envelope of organisms . How fast and how far animals run , fly , swim , or jump is clearly limited by the mechanical power delivered by the muscle-tendon units to skeletal and environmental loads . Therefore , estimating the mechanical energetics of muscles ( henceforth simply called energetics ) has been of interest in diverse fields such as organismal biomechanics , biomimetic robotics and prosthetics [1]–[3] . Many factors influence the neuromechanical performance of organisms , including i ) the dynamics and mechanical properties of muscle actuators , ii ) skeletal mechanics , iii ) neural control and iv ) influence of loads external to the organism . Integrative approaches have been proposed to capture the interaction of all , or subsets of these factors . For example , the connection between muscle impedance ( particularly stiffness ) and neural control has been studied in depth with respect to postural and dynamic stability [4] , [5] , locomotory functions [6]–[9] , manipulation [10] , [11] , and other biomechanical tasks [12] . In this work , we adhere to the definition of muscle mechanical impedance as the “static and dynamic relation between muscle force and imposed stretch” [4] . Muscle impedance encompasses muscle stiffness , which is the static relation between muscle force stretch only . In the context of muscle energetics , most investigations focused on experimentally measuring the power output of individual muscles at a range of frequencies , phases and electrical stimulation parameters , and finding maximal power generating capability of muscles under prescribed motion trajectories . However , the role of muscle-load interaction on output energetics has not been formalized . The central premise of this work is that the mechanical energetics of a muscle-actuated system cannot be determined in a meaningful manner without considering the coupling of muscle properties , load dynamics and neural activation . By considering this coupling explicitly , we arrive at phenomena that cannot be captured using standard workloop testing methodologies , including the opportunity to harness muscle-load interaction in an energetically advantageous manner . Muscle energetics have been characterized under dynamic conditions , both in vitro [13] and in vivo [9] , [14] , [15] . In vitro measurements relied almost invariably on the workloop technique [16] . In this approach , isolated muscles are subjected to predetermined periodic length variations in time ( typically sinusoidal , but not always [17] ) by means of an external motion source . At a given phase of the imposed oscillation , an electrical stimulus is delivered synchronously , resulting in periodic muscle contractions . A plot of muscle contractile force versus displacement results in a cyclic workloop , with the integrated area within the loop being a measure of the net muscle work done . These and similar measurements have been reproduced in the muscle physiology literature for various muscle groups within various organisms [18]–[21] , and connections between the muscle function and its mechanical energetics have been made [22]–[24] . While such measurements provide useful energetic connections with muscle function , the experimental conditions do not capture representative in vivo conditions because motion profiles are imposed on single isolated muscles with no muscle-load interactions [25] , and without incorporating the effects of antagonist activity . In vivo measurements , on the other hand , capture all of the above effects in principle , but lack the experimental flexibility of varying load conditions in an unambiguous manner . Capturing the effect of muscle-load interaction on muscle energetics is critical . This interaction can be captured by considering the impedance of the muscles in relation to the impedance of the load . When a group of muscles acts on a common load , as exemplified by an antagonist pair acting on a common load , each muscle forms part of the load borne by the other muscles in its group . Because muscle impedance is activation dependent , neural control can be used to modulate the effective load observed by each muscle by modulating the impedance of the opposing muscles , thereby offering the opportunity to create favorable impedance conditions that maximize power transfer to the external environmental load . This is akin to the notion of impedance matching in engineering systems , where the driving source and the load are “matched” to provide optimal power transfer . In the context of neuromuscular control , impedance matching can enable groups of muscles to work synergistically to provide significantly higher energetics than the sum of individual muscles . Consequently , in this investigation we studied the influence of muscle-load interaction on muscle workloop energetics both computationally and experimentally . We set up a model problem consisting of a mass-spring-damper system actuated by either a single muscle ( Figure 1B ) , or a pair of symmetric , antagonist muscles ( Figure 1D ) . The input to the system ( either neural control or electrical stimulation ) can modulate the net force exerted by the two muscles as well as the net impedance . In the context of this problem , we investigated two hypothesis . Hypothesis 1 states that the power optimal oscillation frequency of a muscle actuated system is greater than the resonance frequency of the load . This is in direct contrast to an impedance-free actuator ( such as an ideal electric motor ) where the optimal oscillation frequency occurs exactly at the resonance of the load . Hypothesis 2 states that a pair of antagonist muscles can work together to produce more power synergistically than individually by margins that cannot be predicted without explicit incorporation of muscle impedance . We tested these hypotheses both computationally and experimentally . Our computational approach relied on optimal control solutions to the workloop maximization problem , which was based on a mathematical model of the problem . The experimental approach relied on in vitro measurements of workloop energetics of electrically-stimulated , frog muscle acting against emulated mass-spring-damper loads .
To investigate the role of muscle-load interaction and muscle impedance on output energetics , a mathematical model of the problem was developed . This model formed the basis for the ensuing optimization of workloop energetics . We modeled the case of Figure 1D . Note that the case of Figure 1B is a special case of the problem considered with the coefficients of the antagonist muscle set to zero . The key ingredient is a muscle model that captures activation and impedance characteristics of the muscle . The model of Equation ( 6 ) was treated as the basis for our analysis . Since our objective is to analyze optimal muscle workloop energetics , we maximize the average power transfer from the muscles to the load integrated over one periodic cycle . The instantaneous power delivered to the load is given by . The cyclic work done by the muscles on the load is the integral of the power over one complete cycle . Therefore the control inputs , , that characterize power-optimal oscillations are given by the solution of the following optimization problem: ( 7 ) where is defined in Equation ( 6 ) and is the control input vector . In this formulation , we assumed that the terminal time was given and defined by the objective task . Therefore , to optimize power at oscillations of frequency [Hz] , we set the solution time horizon [sec] . To derive necessary conditions for the optimal solution of Problem ( 7 ) , we applied the Pontryagin Minimum Principle [29] . We followed the following procedure: Details of this derivation , and the numerical methods employed therein are described as follows . The integrand of the Lagrangian cost function is given byWe augment the dynamical constraints to the cost function , and define the Hamiltonian scalar functionFrom the Pontryagin principle [29] , the evolution of the optimal co-state variables at the optimal solutions are governed by: The optimal control is given bywhere the last equality follows since is not a function of in this particular context . Substituting in Equation ( 6 ) , we getwhich implieswhere and are upper and lower bounds , respectively , on the control inputs . Depending on the signs of the switching functions and , the control assumes either the values or . This is a bang-bang control solution , and is an expected outcome in such power-optimal ( or maximum acceleration ) problems [30] . Mathematically , such solutions appear when the Hamiltonian is a linear function in the control , as is the case in this problem . In the absence of limits on the control , the optimization problem would be unbounded , implying that the muscles that can generate unbounded forces will add infinite power to the load . Therefore , for the optimization problem to be mathematically well-posed , upper and lower bounds on the control inputs and are necessary . In summary , the first order necessary conditions for power-optimal solutions are given by: ( 8 ) ( 9 ) ( 10 ) with cyclic boundary conditions: ( 11 ) ( 12 ) Equations ( 8 ) and ( 9 ) define a two-point boundary value problem ( 2-point BVP ) that is subject to the cyclic boundary conditions ( 11 ) and ( 12 ) and control constraints ( 10 ) . This 2-point BVP was solved to give the optimal state trajectory ( ) , the optimal control inputs , and the multipliers ( ) associated with the power optimal solution . Methods for solving this problem numerically are detailed in the supporting material Text S1 .
The optimal control problem ( Problem ( 7 ) ) was solved for various values of the time horizon that characterized the oscillation frequencies of interest . An example solution is shown in Figure 2 for an oscillation frequency ( 5 Hz ) that is greater than the load resonance frequency ( Hz ) . To investigate Hypothesis 1 computationally , successive optimizations similar to those of Figure 2 were conducted as the oscillation frequency was swept across the range of interest , and comparisons between optimal power generated by the bilinear muscle model and the optimal power generated by an impedance free actuator were drawn . As shown in Figure 3A , in the case of the system with = 2 Hz , the peak power was generated at = 2 . 4 Hz . In Figure 3B , in the case with = 4 Hz , the peak power was at = 4 . 8 Hz . This result is in direct contrast to the case when the load is driven by impedance-free actuators , where the optimal driving frequency is exactly equal to the resonance frequency of the load . The increase in optimal stimulation frequency is attributed to the contribution of active muscle stiffness to the net stiffness of the system ( shown in the stiffness sub-plots of Figure 2 ) , and thereby tuning the resonance of the combined muscle-load system . To investigate Hypothesis 2 computationally , we compared the power output of the optimal solutions of the single-muscle case against the optimal solutions of the case of a muscle pair in Figure 4 across the frequency range of interest . The computed power-optimal responses show that synergistic activation of antagonist muscles may produce more cyclic work than individual muscle activation by a factor of more than two ( Figure 4B ) . This is captured by the synergistic ratio , and is in direct contrast to constant impedance actuators where the ratio is exactly two . This model prediction implies that the energetics of individual muscles ( obtained by zero-admittance workloop tests ) cannot simply be summed to draw conclusions regarding the workloop energetics of the entire system . Figures 3C and 3D show the results of experimental workloops with single muscles acting on mass-spring-damper loads . To test Hypothesis 1 experimentally , that the peak normalized power output was indeed at , measurements were conducted on two load cases with different natural frequencies ( Hz and Hz ) . For both loads , we found that the normalized power measures and , with ( for all measurements ) . We attribute this increase in the optimal oscillation frequency over to the stiffness contribution of the muscles . This increase in optimal frequency over cannot be achieved via an impedance free force source , and can therefore be directly attributed to the increase in muscle stiffness due to the activation profile over the course of a full cycle . Figure 5 shows the power output measurements of a pair of antagonist muscles acting synergistically compared to their power output acting individually . When the oscillation frequency was set to 3 Hz , the value of the energetic ratio was not statistically different from . However , when the oscillation frequency was set to 4 Hz , we found to be . The ratio was significantly greater than 2 ( ) , showing that the energetics of the muscle pairs are greater than the sum of the energetics attained by individual activation . This is qualitatively compatible with the model predictions plotted in Figure 4 and is in support of Hypothesis 2 . This implies the possibility that energetic synergies may be achieved by a muscle-actuated system to enhance their energetic performance at particular frequency ranges . In the experimental measurements above , the absolute power value of the muscles , normalized by muscle mass , ranged between 17 [W/kg] and 81 [W/kg] at the optimal conditions .
One consequence of explicitly accounting for muscle-load interaction is the increase in the optimal stimulation frequency of the coupled system relative to the natural frequency of the uncoupled load . This is captured by Figure 3 where the maximal power was generated at a frequency higher than the uncoupled natural frequency of the load , which directly supports Hypothesis 1 . This is shown computationally ( Figure 3A & 3B ) where it is possible to scan the range of oscillation frequencies systematically to search for the frequency of peak power generations , and also experimentally ( Figures 3C & 3D ) where it is possible to do so only at select frequencies chosen to show the location of peak power . The increase in optimal power generation frequency is not an unexpected result since the stiffness contributions of the muscles should couple in with the overall frequency of the load . What this enables , however , is that resonance conditions can be tuned relative to the desired frequency of oscillation via an appropriate muscle activation pattern . Taken to the limit of zero load stiffness , we conjecture that this feature potentially enables creating resonance conditions out of non-resonant loads . The biomechanics of natural loads in many biological systems are non-resonating . Consider , as an example , the motion of a swimming fish . The external restoring force on a fish's body is negligible , therefore the sideways bending dynamics can be considered non-resonant . In the presence of muscle activation , however , significant activation modulate stiffness is added to the system , which can be tuned to the desired oscillatory frequency of the undulating motion . The importance of body bending stiffness in relation to the undulating frequency and speed of swimming fish has been reported in [32] , [33] . Another consequence of the coupling between muscle impedance and load dynamics pertains to energetic synergies that are observed in systems driven by multiple muscle systems . When multiple muscles act jointly on a common load , each muscle contributes to the effective load observed by the other muscles acting on that load . This contribution can be strongly modulated by the neural input to the muscles . Taking the simplest case of two antagonist muscles acting in parallel on a common load , Figure 5 shows that a pair of muscles can generate more power on a common load than the sum of them acting individually . The margins of collaboration were much higher than those theoretically predicted with impedance-free actuators . For a pair of identical impedance-free actuators , the ratio is exactly 2 at all frequencies of oscillation . When one impedance-free actuator is capable of producing more force than the other , the ratio ranges between 1 and 2 , but never exceeds 2 . The maximal value of 2 is achieved if the two muscles provide equal forces , and the minimal value of 1 is approached as the relative contributions of the two muscles vary widely . Ratios greater than 2 , as demonstrated in the 4 Hz oscillation case ( shown in Figure 5C ) , and as demonstrated in the maximal values of Figure 4B , are in direct support of Hypothesis 2 , and can only be achieved if additional muscle properties are introduced , such as activation dependent impedance . Our findings may be interpreted in the context of the engineering notion of impedance matching . In engineering systems , impedance matching plays an essential role when it is desired to maximize power transfer between two dynamical systems . When a power source is connected in series with a load ( in a Thevenin equivalent connection ) , maximal power transfer occurs when the internal impedance of the source is equal to the complex conjugate of the load impedance [34] . In a similar manner , neural activation of muscle modulates its stiffness to allow matching of muscle mechanical impedance to that of the load . When such a condition occurs , the power transfer is maximized . This implies that the mechanical work achieved by a single muscle is highly affected by the activation pattern of antagonist muscles , because such antagonist muscles form part of the load on the agonist muscle , and therefore the energetics of muscle-actuated systems must be considered holistically . The impedance of a linear mass-spring-damper load ( ) is the transfer function relating the velocity ( ) and force ( ) applied on the load , and can be expressed aswhere , , and are the mass , damper and spring coefficients of the load . Assuming that the source is primarily dominated by stiffness terms , as is the case of a bilinear muscle model , the impedance of the source ( ) is:Therefore , for this source impedance , which is purely reactive , we do not have the ability to arbitrarily change the phase . To maximize the power transfer from the source to the load , impedance matching conditions require that the reactive part of the source impedance is negative the reactive part of the load impedance [34] . ThereforeUnder such conditions , the total system natural frequency becomeswhich implies that the source stiffness is chosen so that the natural frequency of the system matches the desired oscillation frequency . Therefore , as the muscle pair modulates net stiffness , to a value that matches the desired load impedance , energetic advantages can be attained . Clearly there are limitations to the efficacy of impedance matching in helping maximize workloop energetics . For a pair of antagonist muscles to tune their stiffness to match the reactive impedance component of the load , certain amounts of co-contraction may be required . This was observed computationally with the time overlap of the control signals ( and ) . While co-contraction may attain the desired frequency tuning , it will decrease the peak-to-peak net forces produced by the muscle pair . Beyond a certain break-even point , the peak-to-peak forces will be greatly diminished to the point that impedance matching becomes non-optimal . Research in organismal motor control and biomechanics has reported extensively on the modulation of stiffness in limbs to enhance postural and dynamic stability . Our findings here provide further motivation to hypothesize that the central nervous system may utilize impedance matching as a means to enhance energetics against external loads . Prior studies support the notion that muscle stiffness is modulated to attain resonance tuning , though none have made an explicit energetic connection . Most of these investigations have focused on arm movements . In the context of rhythmic movements , perhaps the clearest evidence was provided in [35] , where forearm stiffness was found to increase quadratically with oscillation frequency , and that the stiffness was minimal at the resonance of the load . It was shown that by increasing the oscillation frequency above the load resonance , the arm stiffness increased in a manner that created resonance of the arm-load system . In other studies [36]–[38] , surface EMG measurements in horizontal arm reaching movements have shown that the overall co-contraction levels increase with increasing frequency of oscillation , and that co-activation increases with the square of frequency . Furthermore , in [39] , neuromuscular models of the forearm that predict qualitative resonance tuning behavior in rhythmic oscillations were proposed . These arguments have also been extended to the context of of non-rhythmic movements by comparing the average forearm stiffness during reaching tasks with the fundamental frequency content of these movements [40] . The degree to which impedance matching is utilized by organisms specifically for energetic purposes remains to be addressed in future studies . Using antagonist activation of variable impedance actuators can enable the central nervous systems to learn optimal impedances that , when coupled with external loads , can provide higher energetics . Viewed from this perspective , activation dependent muscle impedance may be regarded as a favorable biomechanical property . Furthermore , this postulates that the mechanical energetics of individual muscles cannot be directly summed to estimate the total energetics of a multiple-muscle system . | Movement in organisms is a result of the interplay between biomechanics , neural control , and the influence of external environmental loads . Understanding the interaction between these factors is important not only for scientific reasons but also for engineering robotic systems and prostheses that strive to match biological performance . Muscle mechanical impedance is key in defining the mechanical interaction between muscles and their loads . It is well known that neural activation modulates muscle impedance , particularly stiffness , and that such modulation can be used advantageously to stabilize the posture and motion in organisms . Here , we show computationally and experimentally that stiffness modulation can also be used to enhance the capability of muscle to generate mechanical power , which is key in determining how fast animals can run , fly , swim , or jump . When muscles are activated optimally in relation to their external loads , they can create resonance conditions at optimal frequencies that significantly enhance their mechanical energetics by up to 7-fold . These findings can be interpreted in the context of the engineering notions of impedance matching and resonance tuning , which are commonly used as guiding principles in the design of diverse power optimal systems , such as communication circuits and robotic systems . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"computer",
"science/systems",
"and",
"control",
"theory",
"computer",
"science/control",
"technology",
"biophysics/theory",
"and",
"simulation",
"computational",
"biology"
] | 2010 | Optimal Workloop Energetics of Muscle-Actuated Systems: An Impedance Matching View |
We present CATHEDRAL , an iterative protocol for determining the location of previously observed protein folds in novel multidomain protein structures . CATHEDRAL builds on the features of a fast secondary-structure–based method ( using graph theory ) to locate known folds within a multidomain context and a residue-based , double-dynamic programming algorithm , which is used to align members of the target fold groups against the query protein structure to identify the closest relative and assign domain boundaries . To increase the fidelity of the assignments , a support vector machine is used to provide an optimal scoring scheme . Once a domain is verified , it is excised , and the search protocol is repeated in an iterative fashion until all recognisable domains have been identified . We have performed an initial benchmark of CATHEDRAL against other publicly available structure comparison methods using a consensus dataset of domains derived from the CATH and SCOP domain classifications . CATHEDRAL shows superior performance in fold recognition and alignment accuracy when compared with many equivalent methods . If a novel multidomain structure contains a known fold , CATHEDRAL will locate it in 90% of cases , with <1% false positives . For nearly 80% of assigned domains in a manually validated test set , the boundaries were correctly delineated within a tolerance of ten residues . For the remaining cases , previously classified domains were very remotely related to the query chain so that embellishments to the core of the fold caused significant differences in domain sizes and manual refinement of the boundaries was necessary . To put this performance in context , a well-established sequence method based on hidden Markov models was only able to detect 65% of domains , with 33% of the subsequent boundaries assigned within ten residues . Since , on average , 50% of newly determined protein structures contain more than one domain unit , and typically 90% or more of these domains are already classified in CATH , CATHEDRAL will considerably facilitate the automation of protein structure classification .
Proteins comprise individual folding units known as domains , with a significant proportion containing two or more units ( multidomain structures ) [1] . Each domain adopts a specific fold , and it is estimated that there are up to several thousand such folds in nature [2–4] . As the domain is thought to be an important evolutionarily conserved unit , several structural classifications , such as SCOP [5] and CATH [6] , have sought to cluster them into fold groups and evolutionary families . Although a given pair of structures in these families can diverge below similarities of ≤30% in their sequence , these relatives still maintain a comparable topology or fold in the core of the structure [6 , 7] . More than 7 , 000 new proteins structures were deposited in the Protein Data Bank ( PDB; http://www . rcsb . org/pdb ) [8] in 2005 . Furthermore , analysis of version 2 . 6 ( April 2005 ) of the CATH database shows that nearly 50% of classified structures are multidomain . Although many are close in sequence to previously solved structures , the structural genomics initiatives have concentrated their resources on proteins with low sequence similarity to existing structures . As a consequence , they often require considerable manual analysis to be classified in the CATH domain database . That said , the vast majority of newly solved structures contain previously observed folds , although they are often quite remote homologues . In this situation , structural comparison algorithms can be essential to facilitate the automatic and semiautomatic classification of domains . The number of larger multidomain structures has been gradually increasing since the formation of the PDB , with improvements in techniques for structure determination . We can expect this trend to continue , as recent analyses of completed genomes have suggested that the proportion of multidomain structures in some organisms , particularly eukaryotes , may be as high as 80% [1] . Figure 1 shows that the majority of multidomain chains comprise two domains , although some structures have been solved with three , four , five , and six domains . A further complication is that approximately 20% of domains from multidomain proteins in the PDB are discontiguous [9] ( Figure 2 ) ; that is , the structure of the individual domains is formed from disconnected regions of the polypeptide chain . Both automated and manual approaches to domain boundary recognition have problems in assigning these domains . Various computational methods have been developed to automatically detect domain boundaries in multidomain structures ( see [10] ) through a posteriori knowledge of domain folding and interactions . Several approaches assume that a domain makes more internal contacts ( intradomain ) than external contacts ( contact with residues in the remainder of the structure ) . For example , the DOMAK algorithm of [11] derives a “split value” from the number of contacts measured when a protein is divided into two parts . Optimal values are obtained when the separate parts of the split structure are distinct domains . The protein domain parser ( PDP ) takes a similar approach and looks to divide multidomain structures so that the number of internal contacts in each putative domain increases . By contrast , the parser for protein unfolding units ( PUU ) algorithm by [12] uses a harmonic model to describe interdomain dynamics , and is used to define domain units in the FSSP database [13] . A further approach , used by the DETECTIVE algorithm [14] , attempts to determine the hydrophobic core at the heart of each domain unit . The original CATH classification protocol [9] used a consensus approach by combining the results from the three independent methods , described above ( PUU , DOMAK , and DETECTIVE ) . Although each method reports 70%–80% accuracy in benchmarking tests , our experience of updating the CATH database suggest that these methods frequently ( ∼80%–90% of the time ) produce results that are inconsistent with one another . As a consequence , manual validation becomes the only secure way to resolve these conflicting predictions . A recent analysis by Holland et al . [10] showed that all automatic methods run into difficulties when assigning boundaries for certain architectures that do not fit their chosen model of a domain unit—for example , an alpha horseshoe domain , which does not form a compact structure . The authors suggested improvements achieved by a heuristic method that accounts for exceptions to the theoretical domain model . An alternative approach would be to compare a given protein chain against a library of known domain folds . Although many of the algorithms described above effectively delineate domains for a large percentage of protein chains in the PDB ( even those which contain novel folds ) , they provide no indication as to how similar each predicted domain is to folds already classified within the CATH database . Therefore , it is still necessary to compare the excised domain against the CATH library to classify the fold . Since manual validation of domain boundaries and structure-based database scans are both slow , this has remained one of the major bottlenecks in the CATH classification process . As discussed above , there are a limited number of folds , and a novel multidomain structure could well comprise those that have already been classified . This concept of recurrence is not new , and has been successfully exploited by other structural classifications . For example , the DALI algorithm is used to detect recurrent folds for classification in the DALI Domain Database [13] , while the SCOP database uses manual inspection to locate known folds . Many methods exist to find recurring domains using pure sequence approaches ( e . g . , MKDOM [15] , SMART [16] , and PFam [17] ) . However , these are designed to identify individual protein families within gene sequences , rather than predict structural domains . Others , such as SnapDragon [18] and Rigden's covariance analysis [19] , attempt to infer domain boundaries through prior prediction of tertiary structure . Nagarajan and Yona [20] used a combination of PSI-BLAST multiple alignments , predicted structural features , and neural networks to identify the transition between domains in the sequence ( i . e . , the boundaries ) . The authors were able to correctly predict the domain architecture for 35% of multidomain proteins when compared with SCOP . Recent analyses of structures solved by the structural genomics initiatives—which are frequently targeted because they have no clear sequence similarity to existing structures and may adopt novel folds—show that approximately 90% are similar to those already observed in the PDB through sequence or structure comparison [21 , 22] . Therefore , exploiting the concept of domain recurrence to detect known folds in newly determined multidomain structures is a sensible strategy to classify the majority of new structures . Moreover , several fast and powerful algorithms for structure comparison now exist that could be used to perform this task . Some of these compare secondary structures between proteins [23–25] , while others operate at the residue level ( DALI [26] , SSAP [27] , COMPARER [28] , STRUCTAL [29] , and CE [30] ) . The performance of an automatic structural alignment method should be measured both on its ability to generate biologically meaningful alignments and its capacity to accurately detect fold similarities and homologous protein structures . As Kolodny et al . [29] highlight , not all structural comparison methods are as good at scoring their alignments as they are at producing them . A root-mean–squared deviation ( RMSD ) value , or any linear transformation of this , often remains dependent on the number of aligned residues . Some algorithms ( e . g . , CE [30] ) are optimised to find highly conserved regions between two protein structures with a low RMSD . This can be useful in detecting similarities within extremely diverse superfamilies and fold groups . However , this approach does not necessarily give a globally optimal alignment , and can assign high significance to matching small structural motifs that may not be in equivalent positions in the two structures being compared . Hence , for the purpose of domain boundary recognition , it is also vital to consider the number of aligned residues as a proportion of those residues in the larger of the two structures as well as the RMSD of superposed residues . This paper reports the development of the CATHEDRAL algorithm , a novel domain identifier that exploits the fold-recurrence philosophy . CATHEDRAL is an acronym for CATH's existing-domain recognition algorithm . It compares a novel multidomain protein structure against a library of previously classified folds in the CATH database [6] by modifying and combining features from two established structural similarity algorithms . A secondary-structure–matching algorithm , GT ( using graph theory ) [25] , which is very fast and reasonably accurate , is combined with a residue-based method that uses double-dynamic programming ( DDP ) [27] , and is therefore slower but very accurate . By combining these approaches , a 100-fold to 1 , 000-fold increase in speed is achieved , depending on the size of the query structure , at no cost to the quality of the domain alignments . This enables regular scans of newly determined protein structures and rapid classification of their constituent domains into the CATH database . To investigate the efficacy of CATHEDRAL in producing quality alignments , it has also been benchmarked against other publicly available structure comparison algorithms at the single-domain level . By aligning domains in a consensus SCOP/CATH dataset , CATHEDRAL was found to give comparable and , in many cases , superior performance for fold recognition . In addition , when assessing the fidelity of the structural alignments in comparison to hand-curated structural alignments with respect to BAliBASE [31] , it consistently performed better than other approaches by aligning more residues correctly .
The rationale behind CATHEDRAL was to use a fast secondary structure–based graph theory ( GT ) algorithm to discover putative fold matches for a given protein domain/chain structure , which could subsequently be more accurately aligned using a residue-based method exploiting DDP . To evaluate the performance of GT and DDP for fold recognition and accurate alignment , we first created a dataset of single CATH–SCOP domains and compared this approach with other publicly available methods before optimising the algorithm for discovering domain folds in multidomain chains . This benchmarking was also performed on a larger dataset of domains in the nonredundant CATH library ( version 2 . 6 ) and produced almost identical results . Once CATHEDRAL has identified putative domain matches for a query multidomain structure , all domain hits to the chain are ranked by the SVM score , and domain boundaries are assigned using the protocol described in Methods . CATHEDRAL was able to assign 90% of domains in the query dataset to the correct fold group , with 80% of these within ten residues of the actual boundary ( Figure 9 ) . Although our dataset only contained multidomain chains in which all component domains were represented in the CATH library , this is not always the case in classifying novel structures . Indeed , assigning erroneous folds to chains could adversely affect the quality of the domain boundaries . Figure 10 shows a plot of coverage according to the percentage of accurate boundaries ( i . e . , within 10 residues ) . It can be seen that once the SVM score cutoff is increased above 2 , the coverage drops dramatically . However , the accuracy of the domain boundaries does not increase significantly , suggesting that this is an appropriate threshold for CATHEDRAL . Figure 9 shows the coverage of all chains in the dataset with respect to the accuracy of their predicted domain boundaries . CATHEDRAL was developed as a method to be applied unilaterally to all protein chains to be classified into the CATH database . As it is not known a priori whether a given chain contains more than one domain , it is important that the algorithm does not split whole-chain domains unnecessarily . To analyse whether this would pose a problem , the iterative version of CATHEDRAL was also applied to the single-domain CATH–SCOP dataset . In less than 4% of cases , CATHEDRAL predicted that these structures contained more than one domain . The major speed increase in CATHEDRAL is due to the fact that GT preselects representatives for DDP to align to the query chain . By default , it takes all relatives ( nonredundant at 35% sequence identity level ) in each of the top ten top-scoring fold groups identified by GT , even if this results in thousands of comparisons , as occurs in large fold groups such as the Rossmann and TIM barrel folds . This can produce very long running times for some query chains . Nevertheless , it is important to find the closest structural relatives for each assignment to reduce the number of insertions and deletions and therefore increase the accuracy of the domain boundary . We explored whether only a limited number of relatives from each fold could be taken without compromising the fidelity of the domains boundaries . However , given that GT does not accurately discriminate between homologues and domains with the same fold , it was decided to take at least one relative from each superfamily in the target fold group and explore the effect of varying this number . CATHEDRAL was run as described above ( by targeting the top ten fold groups at each iteration ) , but the number of nonredundant representatives ( fr ) taken from each superfamily to be aligned by DDP was varied . Figure 11 shows a plot of the number of correctly assigned domain boundaries ( within ten residues of manually validated boundary ) at each of these levels . It appears that taking any more than seven representatives from each superfamily does not increase the number of good assignments , and hence appeared to be an appropriate level to set the fr parameter . Figure 12 shows the relationship between the accuracy of the domain boundary and the sequence identity between the assigned domain region and best structural match used to assign the boundary . When sequence identity exceeds 10% , there is an increase in the number of correct domain boundaries . It could be expected that the closer the relative from which the assignment is made , the greater chance of it being correct . However , it is encouraging to note that 60% of assignments with sequence identities between 5% and 10% show very little deviation from the manually verified boundaries . Structural embellishments of the core of a fold are responsible for the majority of examples where there is a disagreement between a manually assigned boundary and those predicted by CATHEDRAL . Figure 13 illustrates this problem by showing a domain assignment for a catalase HPII [32] ( PDB code 1iph ) domain , through similarity to its closest match in the CATH library [33] ( PDB code 1beb ) . The matched domain is much smaller than the query , and hence CATHEDRAL is only effective at aligning the core of the fold ( shown in red ) . A number of large insertions in the catalase domain cannot be assigned purely by structural comparison , and these sites are therefore not included within the domain , causing a substantial discrepancy from the correct boundary assignment . Recent analyses of CATH superfamilies has revealed that in 40% of well-populated superfamilies ( nine or more diverse relatives at <35% sequence identity ) , there is 2-fold or more variation in the sizes of the domains ( as measured by the numbers of secondary structures in the domain ) [7] . Therefore , in these superfamilies , it may be difficult to obtain accurate boundaries until a close structural relative is deposited in the PDB . To place the performance of CATHEDRAL in context , we compared its ability to assign domains boundaries with two other methods: hidden Markov models ( HMMs ) and domain predictions from structure ( PDP ) . Our dataset of protein chains was scanned against HMMs built from each structure in the CATH library using the HMMer suite of programs [34] . Domain boundaries were then assigned to the query chains in the same way as CATHEDRAL , but using the HMM E-value instead of the CATHEDRAL SVM score to rank hits . We found that the HMM method was only able to discover 65% of domain folds within the dataset chains . One of the main reasons for this low coverage was that 11% of the chains were not annotated with any domains using an E-value threshold of 0 . 001 . Of the domain boundaries assigned , only 33% were within ten residues , compared with 80% for CATHEDRAL . It is possible that the number of assigned domains could have been increased by using a less conservative E-value threshold . However , this is unlikely to improve the overall quality of the domain boundaries , given the low quality of those that were assigned by the HMM alignments . The domain recognition performance is on a par with the method of Nagarajan and Yona [20] , who predicted the correct domain architecture of 35% of a dataset of multidomain PDB chains . However , by incorporating structural information they were able to increase the percentage of boundaries within ten residues to 63% . CATHEDRAL finds domain boundaries for a query chain by using structural alignment to known folds in CATH . To compare our approach with other methods that do not exploit the concept of fold recurrence , but instead are based on ab initio analysis of structural properties such as residue contacts , we applied the PDP method to our multidomain chain dataset . PDP was able to predict correct domain boundaries ( within ten residues ) for 67% of the chains in the dataset . Although this is lower than CATHEDRAL , it is substantially higher than the 33% achieved by HMM methods . Furthermore , the performance of PDP is still impressive given the problem of distinguishing domain units in a chain based purely on structural properties such as internal contacts and hydrophobicity . More than 50 structural comparison algorithms have been published in the literature in the last 30 years , the vast majority of which are not in regular use by the bioinformatics or structural biology communities . Those which have gained popularity tend to have a Web-based interface for users to submit their own structures or structures from the PDB . CATHEDRAL has been implemented as a crucial part of the CATH classification protocol , and a new Webserver was created to provide users to investigate domain assignments and homologue recognition with their protein structure of choice ( http://cathwww . biochem . ucl . ac . uk/cgi-bin/cath/CathedralServer . pl ) .
We have developed a protocol for domain boundary assignment in multidomain proteins ( CATHEDRAL ) that exploits the recurrence of folds in different multidomain contexts . This was devised because a high proportion ( currently >90% [21] ) of domains in newly determined structures contain folds that have been previously classified in CATH . CATHEDRAL first scans a query structure against a library of folds from the CATH databases . The algorithm first uses GT to perform a secondary structure–based comparison and identify putative domain fold matches in the query structure . A representative sample of nonredundant superfamily relatives from the top ten folds are then recompared to try to obtain a better alignment and refine the domain boundaries . This latter step exploits a DDP algorithm that has been guided by information on equivalent secondary structures identified by the GT match . CATHEDRAL combines the power of two established structural comparison algorithms in order to develop a fast and accurate protocol for homologue recognition and domain assignment . CATHEDRAL misses ∼10% of the domains in the target dataset . Of these , ∼30% are too small ( fewer than three secondary structures ) and so are ignored by the CATHEDRAL protocol . Manual inspection revealed that a further ∼20% are distorted or irregular structures giving poorly defined graphs . The remaining ∼50% are missed because they do not pass the score similarity cutoff , as the relatives are too distant and related structural motifs in neighbouring fold groups are better matched . The CATH classification of protein folds gives a discrete description of fold space . However , there are difficulties in identifying distinct folds in some populated regions of fold space where the structural universe could be more reasonably represented as a continuum [6] . In many cases , as the size of the protein increases , the repertoire of folds appears to consist of extensions to existing motifs . It has been shown by Koppensteiner et al . [35] that it is possible to “walk” from one α/β sandwich fold to another , through the extension of α/β motifs . Furthermore , certain motifs , described as “attractors , ” occur as the core of a protein's structure more frequently than others [36] . Recent analyses of the overlaps between fold groups has shown that for some protein architectures ( αβ sandwiches and mainly β sandwiches ) , extensive overlaps between fold groups are observed due to large common structural motifs [37] . For nearly 80% of the test set , all domain boundaries within the multidomain were correctly assigned within ten residues . This is a considerable improvement over a previous consensus protocol ( DBS; [9] ) described above , for which on average only 10%–20% of domains could be identified as having reliable boundary assignments from agreement between three independent methods . Furthermore , as known folds are recognised by CATHEDRAL , individual domains can be simultaneously classified in the CATH database , without the need for further structure comparison as in previous classification protocols . The method is currently being extended to assign a confidence level or p-value to the boundary and fold assignments predicted by CATHEDRAL . Furthermore , at present , CATHEDRAL assigns domains to a query chain in an iterative fashion . It could be conceived that a better prediction of boundaries and fold assignments could be attained by considering a number of different classifications . The best of these could be identified as the prediction with the highest confidence value . Since CATH aims to maintain high quality domain boundary assignments [38] , results returned by the CATHEDRAL algorithm will continue to be manually assessed . However , the high accuracy of the approach will considerably facilitate this process . Since the proportion of domain folds classified within CATH is likely to continue to increase significantly in the next decade due to the progress of the structural genomics initiatives , the CATHEDRAL algorithm will considerably increase the speed of classification of new multidomain structures and their constituent folds within CATH .
CATHEDRAL and DDP ( a modified version of the SSAP algorithm [27] ) were benchmarked against other publicly available structural comparison methods , STRUCTAL [29] , DALI [26] , LSQMAN , and CE [30] ( see Text S1 for description of methods ) . An all-against-all structural comparison was performed on the 6 , 003 unique CATH domains ( <35% sequence identity to each other ) from 907 fold groups for each of the different structural comparison methods , giving more than 18 million individual comparisons . To minimise any bias in the CATH dataset , a dataset that was a subset of CATH version 2 . 6 . 0 and SCOP verson 1 . 65 was also constructed . Each of 6 , 003 CATH ( SRep ) domains was checked to see if it had an equivalent SCOP domain containing at least 80% of the same residues . All domains satisfying this criterion were mapped to their CATH and SCOP superfamilies . These superfamilies were then compared , and only those sharing 80% of the same members were identified . This restricted the CATH–SCOP dataset to 1 , 779 SReps encompassing 406 folds . There are several publicly available methods that have been endorsed by widespread community use and/or validation by comparative benchmarking against established methods . We selected a range of methods , many of which had been previously benchmarked by Kolodny et al . [29] for their performance in fold recognition and alignment accuracy . Protocol used to compare the performance of CATHEDRAL and full DDP in fold recognition and alignment accuracy with other established methods—fold recognition . Structure alignment methods were compared using ROC curves . These plot true positive rate ( sensitivity ) against the false positive rate ( 1 − specificity ) for different similarity scores returned by the individual methods . A binary classifier was defined by the CATH hierarchy whereby a positive match is one in which both domains share the same fold classification . The matches for each method were ordered by the structural similarity score of their alignment , and the number of true positives and false negatives were calculated at varying thresholds . Protocol used to compare the performance of CATHEDRAL and full DDP in fold recognition and alignment accuracy with other established methods—alignment accuracy . Kolodny and coworkers tested several measures for assessing the accuracy of structural alignments [29] . They identified redundant measures , and alignment accuracy was subsequently compared using the two geometric measures shown in Equations 1 and 2 below: SAS , and SiMin where Nmat represents the number of aligned residues and L1/L2 represents the length of the respective domains . The different measures attempt to balance the different properties that describe a “good” alignment , weighting the RMSD against the length of the alignment as a fraction of the length of the proteins aligned . As CATHEDRAL is exploiting fold recognition to obtain reliable domain boundary assignment , we developed a further measure that scores the global alignment accuracy . As opposed to SiMin , which gives a good score for a small fold appearing as a conserved motif within a much larger fold , SiMax ( Equation 3 below ) takes account of the proportion of residues aligned in the larger domain structure to determine whether a significant fold match has been achieved . All the measurements are in angstroms , and the percentage of alignments within a particular distance in angstroms were calculated for each measure ( SAS , SiMax , and SiMin ) . In addition to these geometric measures , alignment accuracy was also assessed by comparison against a set of manually curated alignments . BAliBASE is a database of manually refined multiple structure alignments specifically designed for the evaluation and comparison of multiple sequence alignment programs . The alignments in BAliBASE are selected from the FSSP [36] or HOMSTRAD [39] structural databases , or from manually constructed structural alignments taken from the literature . When sufficient structures are not available , additional sequences are included from the HSSP database [40] . The VAST Webserver [23] is used to confirm that the sequences in each alignment are structural neighbours and can be structurally superimposed . Functional sites are identified using the PDBsum database [41] , and the alignments are manually adjusted to ensure that conserved residues and secondary structure elements are correctly aligned . A total of 14 BAliBASE multiple alignments were selected , comprising 108 pairwise structural comparisons . All the alignments represented single-protein domain chains that shared less than 25% sequence identity , making alignment nontrivial . All three major protein classes were represented , and the quality of the alignments generated by the different structure comparison methods are measured by the score , fm , which quantifies the number of amino acids correctly aligned in the structural alignment divided by the total number of aligned residues in the BAliBASE alignment . CE was not included in this analysis , as it only identifies the largest continuous motif . CATHEDRAL was benchmarked to calculate its ability to delineate domains within multidomain proteins , as well as correctly recognising the fold of the constituent domains . A set of representatives from 1 , 071 multidomain S35 sequence families ( clustered by single linkage at 35% sequence identity ) was selected . From this set , those chains containing domains from folds with less than two S35 sequence families were removed . The remaining set contained 964 chains comprising 1 , 593 domains . These originated from 245 distinct fold groups and 462 superfamilies . To identify domain boundaries in a novel multidomain structure , CATHEDRAL scans the query structure against a library of folds classified in the CATH database ( see Text S1 for description of CATH hierarchy ) that are derived from contiguous domain representatives from each sequence family ( in which relatives have at least 35% sequence identity ) in version 2 . 6 of the CATH database . This comprised 4 , 707 domains , covering 907 folds . A secondary structure graph of each domain was generated as described in Harrison et al . [25] . Iterative protocol used by CATHEDRAL . CATHEDRAL uses an iterative protocol illustrated in Figure 14 . As described above , novel multidomain proteins are first scanned against a library of domain folds from CATH using the secondary structure GT algorithm . All folds containing hits in the top ten ranked fold hits are then selected for further analysis . To improve the alignment of the matched regions and thereby identify the closest structural neighbour , fr representatives from each superfamily in the selected folds are compared against the matched region using the DDP algorithm . As matches to small domains ( fewer than five secondary structures ) can produce insignificant E-values ( see [25] ) when compared to large chains , these were isolated from the original CATH library and scanned only after all large domains had been assigned by CATHEDRAL . A variety of different scoring schemes were assessed for their ability to recognise true matches , together with a combination of several measures using an SVM ( see below ) . If the score suggests that the match is valid , the region is accepted as a putative domain and the alignment used to indicate the residues that can be excluded from the multidomain structure ( and score matrix ) in future searches . A new graph is constructed from the remaining secondary structures , and the GT and subsequent DDP search is repeated to identify another putative domain . CATHEDRAL continues for up to ten iterations or until there are fewer than three secondary structures left to be assigned . Identification of corresponding secondary structures in the multidomain protein and a single domain structure using GT . GT was first introduced for protein structure comparison by Artymiuk and coworkers [42] . CATHEDRAL uses a new implementation of this approach [25] ( see Text S1 ) that includes further structural features ( e . g . , chirality ) to obtain a better resolution between related and unrelated folds . A robust statistical framework was also derived to calculate expectation values ( E-values ) that can be used to assess the significance of each comparison ( see [25] for a detailed description of the GT algorithm used in CATHEDRAL ) . Generating a residue alignment of the fold match using DDP . Once a putative domain within the multidomain structure has been matched to a fold in the CATH database , an accurate alignment between this domain and the target structure can be obtained using a residue-based method that exploits DDP . CATHEDRAL uses the global alignment version of the DDP algorithm , described in Taylor and Orengo [27] . This choice followed assessment of the performance obtained using the global and local alignment versions [43] . The global alignment version is better able to handle proteins with discontiguous domains , as the alignment produced by the local version in such cases was found to match only one of the fragments of the discontinuous domain . The break in the discontiguous domain appears to the alignment program as a large gap , and the local alignment score within the gapped region rapidly falls to zero , thus terminating the alignment incorrectly . Using secondary structure matches from the GT filter to guide residue alignment by DDP . The full DDP algorithm is computationally expensive because it makes an exhaustive search of all possible pathways through the residue and summary level matrices , although this search can be constrained by imposing a window on the score matrix [27] . Fortunately , it is not necessary to compare all the equivalent positions between two related proteins to obtain an accurate residue alignment . Therefore , the clique information identifying matching secondary structures can be used to exclude large regions of the score matrix by populating a binary matrix , which dictates which residues to compare . First , residues in equivalent secondary structures must pair with one another . As equivalent strands and helices can vary in length ( e . g . , a helix with 11 residues could be aligned to one with eight residues ) , it must be an all-versus-all pairing ( represented by a square of “1” values in the matrix ) . Similarly , residues on the end of aligned secondary structures could potentially be paired with residues in the loop regions , so the boundary is extended by 10 residues on either side . Second , although the alignment for residues outside the clique is unknown , it is possible to exclude certain pairings . The clique effectively orientates the alignment and dictates that if helix 1 in protein A is equivalent to helix 2 in protein B , it cannot simultaneously be equivalent to helix 3 in protein B . Moreover , it gives the overall direction of the alignment and allows the regions between the clique secondary structures to be linked together . Finally , the alignment of embellishments of the core clique secondary structures at the start and ends of the domains is unspecified . However , it is known that these cannot be aligned to any of the core residue pairs . Hence , the starts and ends of the domains are paired up for DDP to decide where the equivalences lie . As outlined in the DDP description in Text S1 , residue pairs possessing similar torsional angles and accessibility within these matching secondary structure blocks are then selected for comparison . Cliques indicate blocks within which residues in matching secondary structures should be aligned . Gaps between these blocks are also possible locations for the residue alignment algorithm to search . The rest of the score matrix can be ignored . This typically gives a significant reduction in the number of residue pairs that must be compared in the first pass of the DDP algorithm . As well as speeding up the alignment , it also reduces the amount of noise in the summary score matrix accumulated in the first pass , as fewer nonequivalent residue pairs are compared . Similarly , once a domain has been matched in the multidomain structure , the block associated with that domain need not be subsequently searched . These restrictions on the search space result in much faster comparisons without significantly affecting the ability to recognise equivalents . Adapting the CATHEDRAL protocol to favour global matches over local motif matches . The accuracy of the secondary structure–matching algorithm improves with clique size because for larger cliques there are more equivalent geometric relationships identified . This is because a clique that has N nodes contains N ( N − 1 ) / 2 edges . Matches identified using GT are therefore more secure when the clique is large , independent of the residue similarity score . Furthermore , because the scoring scheme for graph-matching breaks down for the very small folds ( fewer than three secondary structures [25] ) , to maintain the integrity of CATHEDRAL's predictions , these very small proteins are excluded by the algorithm . As CATHEDRAL iterates toward a solution , the CATH database is repeatedly scanned . However , some large folds contain structural motifs that match well to small folds . These motif matches sometimes rank higher in the match list because the geometry is very well conserved , and the selection of these matches over equivalent folds can therefore confuse the identification of domain boundaries . This effect can be avoided by attempting to match only large domains first; that is , two passes of CATHEDRAL are performed . The first pass only allows matches to folds in CATH that have graphs of five or more nodes . Once CATHEDRAL has reached its termination , it is applied again to the folds in CATH that have graphs with three or four nodes . This strategy results in the smallest folds only being compared against regions of the multidomain protein that are not part of a large fold , as well as typically increasing CATHEDRAL's speed by 50% or more since fewer searches are required . Hence , CATHEDRAL essentially assigns all large domains first before attempting to align smaller domains to any remaining unassigned regions . To aid the assignment of discontiguous domains , in the first iteration , the top hit is also required to be contiguous ( i . e . , the assigned region comprises one continuous sequence segment ) . Scoring the structural similarity of the domain region aligned by DDP . To assess whether a given structural hit represents a true fold match within the multidomain protein , several measures of similarity are calculated . The structural similarity score returned by the DDP algorithm is normalised to lie in the range of 0–100 ( with 100 for identical structures ) irrespective of the protein sizes [44] . This score is based on similarities in the vectors between Cβ atoms of equivalent residues in the aligned proteins and is normalised to take account of the size of the largest domain being compared . A rigid body superposition of the structures is also generated from the equivalent residues identified by the alignment . RMSD of the aligned Cα atoms is calculated , and a cutoff can be imposed on the local structural similarity ( see above ) to select only the most similar residue pairs when generating the superposition of the structures . A cutoff of 30 ( with 100 representing identical residues ) is used to ensure the most equivalent residues pairs are used to calculate the SAS . Using an SVM to validate structural matches . Determining domain boundaries in protein chains through iterative fold assignment presents several challenges . For example , there is the problem of mis-assigning folds that simply match a large structural motif that does not correspond to a significant “global” match to the domain region . Discontiguous domains can also present problems for structural alignment algorithms . Several similarity measures can be considered when gauging whether a match is valid . Manual experimentation can be used to explore and optimise the combination of these measures , or machine learning methodologies can be used . In CATHEDRAL , we exploited an SVM to perform the optimisation automatically and to determine when a significant domain structure match to a classified fold in CATH was occurring . In addition to the similarity measures provided by the GT and DDP algorithms , we also considered other features ( e . g . , the proportion of residues matched between the two structures , and similarity in domain sizes ) to help improve recognition of global similarity between domain structures . We used the SVMLight package [45] to combine these features using a linear kernel . To train the SVM , 5-fold cross-validation was used to assess the performance of the SVM models . That is , the dataset was split into five sets , and each one was successively used as the test set , while the model was trained on the other four sets . This reduces any potential bias caused by random fluctuations in the composition of the training and test sets . The error cost for positive examples was weighted according to their ratio to negative examples . Features included the raw score , E-value , and clique size ( number of matched secondary structures ) returned by the GT comparison . In addition , the raw score derived from the DDP algorithm together with the residue overlap ( percentage of residues in the CATH domain aligned against the putative domain region ) , CATH domain size , sequence identity , and SAS . To improve the ability of the classifier to avoid bias toward one feature , each was normalised between 0 and 1 . Identifying domain boundaries and handling discontinuous domains . The individual DDP similarity score of equivalent residue pairs , normalised to lie between 0–100 , indicates where residue similarity is good ( high ) , where it is poor ( low ) , and where it is nonexistent ( residue score is zero ) . Since only individual domains from CATH are scanned against the multidomain structure , the alignment can be used to find domain boundaries , because the residue pair score falls to zero at the boundary . When CATHEDRAL determines which fold to assign to a region of the protein chain , it is also making a judgment of where the domain boundaries lie . The fidelity of this latter process is arguably dependent on the structural similarity between the domain region in the chain and the domain it has matched in the library . Although domain boundaries can be assigned to the chain in the same step as taking the highest scoring hits to each region of the chain , the accuracy can be improved by modifying the boundaries once all assignments have been made . Subsequently , domain assignments that contain regions of the chain that overlap with one another are processed as a last step in the protocol . Conflicts are resolved by assuming that the highest-scoring domain is most likely to have the correct boundaries . The boundaries of the overlapping domain are cropped to exclude the shared region . Second , some chains may contain small regions at the start and end that are unassigned . This is often fewer than 20 residues and is unlikely to contain another domain , or comprise an additional segment of a discontiguous domain . In these instances , CATHEDRAL assigns the extra residues at the beginning and end of the chain to the first and last domains , respectively . Similarly , some chains contain small regions between assigned segments . In these cases , CATHEDRAL splits the unassigned residues equally between the two neighbouring segments . | Proteins comprise individual folding units known as domains , with a significant proportion containing two or more ( multidomain structures ) . Each domain is thought to represent a unit of evolution and adopts a specific fold . Detecting domains is often the first step in classifying proteins into evolutionary families for studying the relationship between sequence , structure , and function . Automatically identifying domains from structural data is problematic due to the fact that domains vary substantially in their compactness and geometric separation from one another in the whole protein . We present a novel method , CATHEDRAL , which iteratively identifies each domain by comparing a query structure against a library of manually verified domains in the CATH domain database through computational structure comparison . We find that CATHEDRAL is able to outperform the majority of popular structure comparison methods for finding structural relatives . Furthermore , it is able to accurately identify domain boundaries and outperform other methods of structure-based domain prediction for the majority of proteins . CATHEDRAL is available as a Webserver to provide domain annotations for the community and hence aid in structural and functional characterisation of newly solved protein structures . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"none",
"computational",
"biology"
] | 2007 | CATHEDRAL: A Fast and Effective Algorithm to Predict Folds and Domain Boundaries from Multidomain Protein Structures |
Focal degradation of extracellular matrix ( ECM ) is the first step in the invasion of cancer cells . MT1-MMP is a potent membrane proteinase employed by aggressive cancer cells . In our previous study , we reported that MT1-MMP was preferentially located at membrane protrusions called invadopodia , where MT1-MMP underwent quick turnover . Our computer simulation and experiments showed that this quick turnover was essential for the degradation of ECM at invadopodia ( Hoshino , D . , et al . , ( 2012 ) PLoS Comp . Biol . , 8: e1002479 ) . Here we report on characterization and analysis of the ECM-degrading activity of MT1-MMP , aiming at elucidating a possible reason for its repetitive insertion in the ECM degradation . First , in our computational model , we found a very narrow transient peak in the activity of MT1-MMP followed by steady state activity . This transient activity was due to the inhibition by TIMP-2 , and the steady state activity of MT1-MMP decreased dramatically at higher TIMP-2 concentrations . Second , we evaluated the role of the narrow transient activity in the ECM degradation . When the transient activity was forcibly suppressed in computer simulations , the ECM degradation was heavily suppressed , indicating the essential role of this transient peak in the ECM degradation . Third , we compared continuous and pulsatile turnover of MT1-MMP in the ECM degradation at invadopodia . The pulsatile insertion showed basically consistent results with the continuous insertion in the ECM degradation , and the ECM degrading efficacy depended heavily on the transient activity of MT1-MMP in both models . Unexpectedly , however , low-frequency/high-concentration insertion of MT1-MMP was more effective in ECM degradation than high-frequency/low-concentration pulsatile insertion even if the time-averaged amount of inserted MT1-MMP was the same . The present analysis and characterization of ECM degradation by MT1-MMP together with our previous report indicate a dynamic nature of MT1-MMP at invadopodia and the importance of its transient peak in the degradation of the ECM .
Metastasis is the major cause of death in cancer patients . If metastasis is blocked , 90% of patients will survive [1] . Thus hindering metastasis should be a main therapeutic target . The first step of invasion is the focal degradation of extracellular matrix ( ECM ) surrounding cancer cells , and MT1-MMP , a membrane metalloproteinase , plays a critical role in this process [2]–[4] . Invadopodia , which are tiny protrusions found on the surface of malignant cancer cells , are the machinery involved in the focal degradation of ECM , and this is where MT1-MMP is accumulating [5]–[8] , and causing degradation of ECM [5] . Thus , the activity of MT1-MMP at invadopodia is critical for ECM degradation and cancer cell invasion and metastasis . MT1-MMP is blocked by an endogenous soluble inhibitor TIMP-2 . On the other hand , TIMP-2 binds a proform of an endogenous soluble metalloproteinase MMP-2 ( proMMP-2 ) forming a ternary complex of MT1-MMP . TIMP-2 . proMMP-2 [9] . If additional MT1-MMP binds to the ternary complex through the hemopexin domain of MT1-MMP a quaternary complex is formed of MT1-MMP . MT1-MMP . TIMP-2 . proMMP-2; proMMP-2 in the complex is processed by the newly bound TIMP-2-free MT1-MMP , which results in the release of active MMP-2 [10] , [11] . Active MMP-2 degrades ECM together with MT1-MMP [12]–[16] . Thus , TIMP-2 plays dual roles as an MT1-MMP inhibitor and as an adaptor for MMP-2 activation , and thus acts as both inhibitor and enhancer of ECM degradation [9] , [17]–[20] . In spite of these possibilities , a high TIMP-2 concentration leads to almost complete inhibition of MT1-MMP . In order to overcome this inhibition , it is essential to recycle and replenish TIMP-free MT1-MMP to the membrane surface by vesicular trafficking whatever the route is . Thus the vesicular trafficking of MT1-MMP plays a more dynamic role in ECM degradation than simply transporting newly synthesized MT1-MMP to the membrane surface . In fact , pharmacological blockade of vesicle trafficking dramatically reduced the focal degradation of ECM at invadopodia [21]–[28] . However , the activation and deactivation mechanisms of MT1-MMP are not as simple as suggested in the previous report because various complexes can form from MT1-MMP , TIMP-2 and MMP-2 [23] . We suggested that these complexes play a role in the control of ECM degrading activity , and that the dynamic nature of these complexes is critical for the ECM degradation . However , the dynamics of the ECM-degrading activity of MT1-MMP still remains to be elucidated . Here we analyze the activity of MT1-MMP in our computational model , which also was used in the previous report , with the aim of finding a possible reason for the repetitive insertion required for the ECM degradation [23] . We have found a very narrow transient activity of MT1-MMP with a half-width of less than 5 sec followed by a steady state activity . The transient activity depends on the TIMP-2 concentration , and is prominent at higher TIMP-2 concentrations . Importantly , we have found that the transient activity is crucial for the degradation of ECM , because in the absence of the transient activity , ECM degradation by MT1-MMP was heavily reduced . Finally , we test the effect of pulsatile insertion of MT1-MMP , which is a simulated vesicular insertion of MT1-MMP [26] , [29] . The simulation shows a faster ECM degradation by insertion of a larger MT1-MMP content in a single vesicle , while the time-averaged amount of insertion is kept constant .
When head and neck squamous cell carcinoma SCC61 cells are cultured on glass plates coated with Dylight 633-labeled fibronectin over cross-linked gelatin , the loci of ECM degradation are visualized as dark spots ( left panel in Figure 1A ) . ECM degradation in our system depends on MT1-MMP and its vesicular trafficking to the surface of invadopodia , since knockdown of MT1-MMP or the inhibition of vesicular transport reduces the number of ECM-degradation spots ( right panel ) [23] . In the previous report , we have shown that the turnover of MT1-MMP is required for the effective ECM degradation [23] . However , it is not known why the turnover is required for the ECM degradation . Since MT1-MMP is inhibited by the endogenous inhibitor TIMP-2 , newly inserted MT1-MMP can readily be inhibited by TIMP-2 . Therefore , a repetitive insertion of MT1-MMP may be required to overcome this inhibitory effect of TIMP-2 and provide an effective ECM degradation . In order to investigate this possibility , we ran a series of computer simulations , and analyzed the dynamic activity of MT1-MMP , which can be too fast to observe experimentally . The computational model is shown schematically in Figure 1B [23] . All possible complexes and their state transitions are shown in Figure S1 , and an A-Cell model is shown in Figure S2 , where MT1-MMP , TIMP-2 , and MMP-2 are named M14 , T2 , and M2 , respectively . In this model , schemes of the ECM degradation are not involved . First we investigated the time course of the concentration of active complexes of MT1-MMP that play a role in ECM degradation . We defined M14a , which is the sum of the concentrations of M14 , M14 . M14 , M14 . M14 . T2 , and M14 . M14 . T2 . M2 , as a measure for the ECM degrading activity of MT1-MMP . The dimer M14 . M14 was assumed to possess twice the activity of that of other complexes . We assumed a surface expression of 100 nM MT1-MMP at t = 0 . As shown in Figure 1C , M14a decreases quickly followed by a steady state activity of MT1-MMP at a TIMP-2 concentration of 100 nM . The steady state level ( M14asteady ) continues for at least 50 , 000 sec ( Figure S3 ) . The level of M14asteady depends on the TIMP-2 concentration ( Figure 1C inset and Figure 1D ) . At a lower concentration of TIMP-2 , M14asteady gradually decreases as the TIMP-2 increases , which is followed by an abrupt decrease around 80 to 120 nM of TIMP-2 , and then by an additional gradual decrease after a further increase in TIMP-2 . ttransient , which is the half-width of M14a transient activity , has a peak value of 4 . 2 sec at TIMP-2 of 100 nM , and at lower and higher TIMP-2 concentrations ttransient decreases ( Figure S4 ) . These characteristics are almost identical in the absence of MMP-2 ( Figure S5 ) . Next we sought the reason for this sharp transient peak and analyzed the time courses for the concentration changes of active complexes ( Figure 2A ) . M14 quickly decreases within 10 sec after its surface expression , and M14 . M14 and M14 . M14 . T2 . M2 have much reduced delayed peaks . In contrast , inactive complexes begin to grow during or after the sharp transient peak of M14a ( Figure 2B ) . Among them M14 . T2 . M2 resembles the transient peak and it reaches a plateau , and then M14 . T2 . M2 . M14 . T2 . M2 , which is the highest order inactive complex in our model , grows monotonically and reaches a plateau . These computational analyses strongly suggest that the delayed inactivation of active MT1-MMP is the reason for the sharp transient peak of M14a . To confirm this , we ran simulations that deleted paths to and from inactive complexes and aimed at eliminating transient peaks and increasing the M14asteady level ( Figure 2C ) . If we delete M14 . T2 . M2 . M14 . T2 . M2 ( deletion 1 ) and M14 . T2 . M2 ( deletion 2 ) , the M14a transient peak is reduced and M14asteady is increased ( thick black line ) in comparison with the control ( thin black line ) , and if we delete M14 . T2 . M14 . T2 . M2 ( deletion 3 ) , M14 . T2 . M14 . T2 ( deletion4 ) , and M14 . T2 ( deletion 5 ) additionally , M14a transient and M14asteady are further decreased and increased , respectively ( thick red line ) . Even after the deletion of all inactive complexes , there still remains a transient in M14a ( thick red line ) . This is because M14a includes a half-inactive complex of M14 . M14 . T2 . The increase in M14asteady by the deletion of inactive complexes ( deletions 1 through 5 ) is clearly seen in Figure 2D , where the M14asteady is progressively increased by the increase in the number of deleted paths . These results clearly indicate that the M14a transient peak seen in our model is due to the delayed inactivation of active MT1-MMP complexes . In the previous report , we found that there were two pools X and D of MT1-MMP in the invadopodial membrane with turnover time constants of 259 sec and 26 . 0 sec , respectively [23] . MT1-MMP is docked and internalized to and from these sites , and the turnover of MT1-MMP at these sites was found to be crucial for the ECM degradation . Therefore , we next investigated whether the sharp transient peak of M14a is also observed in the presence of the turnover of MT1-MMP using the same model as that used in the previous report ( Figure S6 ) [23] . ECM degradation was not included in this simulation to compare the transient activity of M14a with those observed in the absence of the turnover of MT1-MMP in the previous section . Simulations show the existence of the sharp transient peak also in the presence of the MT1-MMP turnover ( inset of Figure 3A ) . However , the reduction in M14asteady by the increase in TIMP-2 concentrations is greatly reduced ( continuous red line in Figure 3A ) in comparison with the case in the absence of the turnover ( continuous black line ) . This suggests that M14asteady is modulated by the turnover of MT1-MMP . In fact , M14asteady depends on the turnover rate , as indicated by the broken red lines in Figure 3A , where M14asteady is decreased by the reduction of the turnover rate . It should be noted that the effect of the turnover rate is much more prominent in the region of TIMP-2 concentrations higher than 80 nM . At TIMP-2 concentrations lower than 80 nM , the difference in M14asteady in the presence or absence of turnover is only negligible . To clarify the mechanism responsible for the different behavior of the M14a transient peak in the presence or absence of MT1-MMP turnover , we next sought to deduce which active complexes are responsible for the increased M14asteady in the presence of the turnover . Although virtually no alteration is seen in M14 . M14 . T2 , there are increases in M14 , M14 . M14 , and M14 . M14 . T2 . M2 at steady state as the turnover rate increases ( Figure 3B ) . Among them , M14 and M14 . M14 showed the two largest increases by the increase in the turnover rate . Thus , the increase in monomeric and dimeric MT1-MMP is the reason for the higher steady state level of M14a at a higher turnover rate of MT1-MMP . Thus , the turnover of MT1-MMP abrogates the effective inhibition by TIMP-2 to some extent . While the ECM-degrading activity of MT1-MMP is high but transient ( ttransient<5 sec ) in the very beginning after the insertion of MT1-MMP , it is low but long-lasting in the following steady state . Which will contribute the most to ECM degradation , the sharp transient activity or the long-lasting steady state ? To explore this , we ran simulations of ECM degradation in the presence or absence of the sharp transient activity . For this purpose , we added a model for the ECM degradation to the one discussed in the previous section , which is the same as before ( Figure S14 in [23] ) . In this model , both MT1-MMP in pools X and D and active MMP-2 ( M2act ) degrade ECM . To realize simulations in which the sharp transient activity was eliminated , we modified the model ( See Materials and Methods ) . Briefly , the MT1-MMP that is inserted into the plasma membrane is apportioned to its complexes according to their relative fractions at steady state . We confirmed that there is no transient peak in M14a or in each of the complexes for pools X and D ( Figure S7 ) . The simulated ECM degradation proceeds much faster in the presence of the transient activity than in its absence at all TIMP-2 concentrations from 100 nM to 500 nM ( Figure 4A ) . Thus , the sharp transient activity of MT1-MMP is crucial for the efficient ECM degradation . This situation is clearly shown in Figure 4B , where the time to half-degradation of ECM τH is shown at TIMP-2 concentrations from 0 to 500 nM in the presence ( closed circles ) or absence ( open circles ) of the transient activity . At all TIMP-2 concentrations except at 0 nM , τH is larger in the absence of the transient activity than in its presence . Especially at TIMP-2 concentrations higher than 60 nM , τH in the absence of the transient activity shows a dramatic increase , and at TIMP-2 of 500 nM , the difference in τH is more than two-orders of magnitude . Thus , τH in the absence of the transient activity is increased as the TIMP-2 concentration is increased . In the model discussed above , the turnover of MT1-MMP is simulated as a continuous event . However , MT1-MMP is inserted by vesicular trafficking [5] , [21] , [23] , [26] , [29] . This indicates that the actual turnover of MT1-MMP at the invadopodial membrane will proceed in a pulsatile manner . Therefore , we constructed a model for the pulsatile turnover of MT1-MMP . Briefly , the pulsatile insertion intervals for pools X and D are set to be 25 . 9 and 2 . 6 sec , respectively , depending on the two time constants observed in our FRAP experiments [23] . These intervals were calculated by assuming that the concentrations of single insertions are 10% of the maximum available docking concentrations for pools X and D , which are 3 nM and 7 nM , respectively ( See Materials and Methods for detail ) . The time courses of M14a in the pulsatile insertion with regular intervals for TIMP-2 concentrations from 0 to 200 nM are shown in the top panel of Figure 5A . The model schemes are the same as those in Figure S6 except the insertion is pulsatile instead of continuous . The zigzag lines caused by the pulsatile insertion have long and short regular time intervals , which correspond to insertion intervals for pools X ( 25 . 9 sec ) and D ( 2 . 6 sec ) , respectively . The basic behavior of the M14a time course is the same as that of the continuous insertion model , that is , M14a decreases after the first insertion of MT1-MMP at t = 0 , and this decrease becomes larger as TIMP-2 increases . The decreased levels of M14a are identical to those of the continuous insertion model ( Figure S8 ) . It is not known whether the pulsatile MT1-MMP insertion occurs at regular or random intervals . Therefore , we ran simulations of random insertion intervals with their averages corresponding to those of regular insertion , which are 25 . 9 and 2 . 6 sec for pools X and D , respectively . The time courses of M14a in random insertion intervals ( bottom panel of Figure 5A ) differ largely from that of the regular insertion ( upper panel ) . The simulation results shown above suggest that the regular and random insertion intervals could result in different time courses for ECM degradation . To see this possibility , we ran simulations of ECM degradation with regular and random insertion intervals . Unexpectedly , however , the time course for regular and random insertion protocols did not differ greatly ( continuous and broken lines in Figure 5B for regular and random insertion intervals , respectively ) . These situations are clearly shown in Figure S9 , where histograms of τH are shown at TIMP-2 concentrations of 0 , 200 , and 400 nM with 120±9 . 66 , 173±16 . 2 , 277±32 . 9 sec . The SDs are around 10% of the average . The same situation is also true when amount instead of interval of MT1-MMP at single pulsatile insertion is varied randomly ( Figure S10 ) . The simulation results discussed above clearly show that the time course of ECM degradation does not change considerably despite the large variability in the time course of M14a . To explore the reason , we investigated the time course of the M14a . ECM complex ( = M14 . ECM+M14 . M14 . ECM+M14 . M14 . T2 . ECM+M14 . M14 . T2 . M2 . ECM ) , because the concentration of this complex determines the rate of ECM degradation . The random insertion protocol gives higher or lower concentrations of M14a . ECM ( black thin line in Figure 5C ) in comparison to the regular insertion protocols ( black thick line ) . If we take the integrated concentration of the complex , which is directly related to the amount of degraded ECM and hence the time course of ECM degradation , the difference in the amount of degraded ECM between regular and random interval insertion is reduced ( red lines in Figure 5C ) . Taken together , the difference in the time course of ECM degradation and hence τH is much reduced even if the insertion intervals or the amount at a single insertion of MT1-MMP are randomly changing . In the previous section we assumed the concentrations of single pulsatile insertions were 10% of the maximum available docking concentrations for pools X and D . However , the amount of MT1-MMP in a single vesicle is not known . Therefore , it is important to investigate the effect of different concentrations in single insertions . To this purpose , we set the time-averaged amount of the insertions unchanged but used different frequencies and amounts in a single insertion . The question is whether the two regimens shown in Figure 6A result in the same ECM degradation efficacy . To test this question , we ran simulations by changing the frequency and concentrations while keeping the average amount unchanged . In the control condition , the frequency/concentration for pools X and D are ( 1/25 . 9 /sec ) /3 nM and ( 1/2 . 6 /sec ) /7 nM , respectively . In the range from frequency/concentration of 100/0 . 01-folds of the control ( 1/1 ) to 0 . 2/5-folds , there is virtually no change in the τH ( Figure 6B ) . This situation is also shown in the time course of ECM degradation , where all 13 sets of frequency/concentration are shown , and the time courses for 9 sets from 100/0 . 01 to 0 . 2/5 are overlapping ( Figure 6B inset ) . When the concentration is larger than fivefold of the control , however , τH begins to decrease despite the average insertion amount is the same . Thus , the ECM-degrading efficacy is higher for lower frequency/higher concentration regimens . These results seem strange , since the average amount of insertion was not changed . To explore the reason , we analyzed the change in the initial rate of ECM degradation , which is the rate of the ECM degradation at t = 0 , by the change in frequency/concentration ( Figure 6C ) . The initial rate increases by the increase in the concentration as it was expected to do . Then , it would be expected that the initial higher rate of ECM degradation at the lower frequency/higher concentration regimen would be gradually reduced , and that the ECM concentration at the next insertion in this regimen would coincide with that of the higher frequency/lower concentration regimen ( blue and red lines in the inset of Figure 6C ) . To explore whether or not this is the case , we analyzed the time courses of ECM degradation at a very early stage ( Figure 6D ) . The expectation was almost met for the higher frequency/lower concentration regimen ( black and blue lines for frequency/concentration of 1/1 and 0 . 5/2 , respectively ) . If the frequency/concentration is changed to 0 . 1/10 ( red line ) , however , the ECM concentration at the second insertion time ( ta ) is significantly lower ( the difference is da ) . Thus , the expectation shown in the inset of Figure 6C is not met for the lower frequency/higher concentration regimen . If we look at Figure 6D carefully , the difference in the ECM concentration is larger at later times of insertion ( da<db at time tb ) , and the slope of the curve is larger during ta to tb than from 0 to ta . This indicates that the efficacy of the same concentration of inserted MT1-MMP is larger at later times of insertion . Since the ECM concentration at later times is decreased by the degradation , the relative value of inserted MT1-MMP/ECM and hence the efficacy to degrade ECM will be higher at later times . This will cause more efficient ECM degradation for lower frequency/higher concentration regimens , and result in a smaller τH . Thus , the equality shown in Figure 6A does not hold at later times , and the difference in the ECM degradation efficacy is significantly higher at lower frequency/higher concentration regimens , even if the average insertion amount is the same . All of the simulations discussed above were performed without any spatial distribution . Next we investigated the significance of the sharp transient activity both for continuous and pulsatile insertions in a spatially distributed model . In the pulsatile insertion , both the interval and the concentration of insertions were regular . The spatially distributed model of ECM degradation was the same as that discussed earlier [23] . Briefly , the extracellular space of 5×5×3 µm , which is assumed to be filled with ECM proteins , is divided into 51×51×1 compartments of identical size . At the center of one surface of the extracellular space , the focal degradation of ECM was defined ( red compartments in the inset of Figure 7A ) . Models for the activation of MT1-MMP , its turnover , the activation of proMMP2 , and ECM degradation by MT1-MMP were embedded to these compartments . ECM degradation by MMP-2 and deactivation of MMP-2 by TIMP-2 were embedded to all compartments . proMMP-2 , MMP-2 , TIMP-2 and TIMP-2 . MMP-2 complex are diffusing species . Black lines in Figure 7A show the time course of ECM degradation in the absence of the sharp transient activity . Virtually no ECM degradation is seen neither in continuous ( continuous line ) nor pulsatile insertions ( broken line ) at TIMP-2 of 200 nM . In the presence of the sharp transient activity ( red lines ) , however , ECM degradation proceeds both in continuous and pulsatile insertions with almost identical time courses . If we plot the spatio-temporal profile of the ECM degradation along a vertical white line crossing the invadopodial region shown in Figure 7B , virtually no ECM degradation is seen in the absence of sharp transient activity both in continuous and pulsatile insertions . In the presence of the sharp transient activity , however , ECM degradation proceeds significantly in both insertion regimens . The profiles of ECM degradation are almost identical in continuous and pulsatile insertions . These simulation results clearly indicate a crucial role for the transient peak in ECM degradation also in a spatially distributed environment .
The dynamics of the activity of MT1-MMP at the surface of invadopodia is currently difficult to measure experimentally . In this study we have analyzed the activity of MT1-MMP in terms of ECM degradation . First , our simulation suggests the existence of a sharp transient peak with a half-width of less than 5 sec followed by a steady state level of active MT1-MMP complexes . Second the steady state level of MT1-MMP complexes depends on the TIMP-2 concentration , being lower at higher TIMP-2 concentrations . Third , we found that this sharp transient peak is critical for the effective ECM degradation . If this sharp transient peak was eliminated computationally , the efficacy of ECM degradation was dramatically reduced . Fourth , in the pulsatile insertion of MT1-MMP , the effect of random insertion intervals or random insertion concentrations of MT1-MMP is small , resulting in a small variance in τH in comparison to the regular insertion regimen . Finally , the insertion of MT1-MMP at lower frequency/higher concentration resulted in a faster ECM degradation even if the time-averaged amount of insertion is the same . Thus we have shown a number of critical characteristics of MT1-MMP dynamics in the ECM degradation . These results are summarized in Figure 8 . There are two docking sites in the membrane of invadopodia into which MT1-MMPs are inserted . We call these sites pools D and X . The turnover rates of MT1-MMP in pools D and X are 26 . 0 sec and 259 sec , respectively [23] . MT1-MMP is transported to these pools by vesicular trafficking via various routes in the cytoplasm . MT1-MMP inserted into these pools is subjected to ECM degradation and also to inhibition by the endogenous inhibitor TIMP-2 . This inhibition is fast , and hence the half-life for newly inserted MT1-MMP to possess ECM-degrading activity is only about 4 sec depending on the TIMP-2 concentration ( Figures 1 , 2 and S4 ) . This short half-life is compensated by the fast turnover of MT1-MMP leading to the effective degradation of ECM at invadopodia . It should be noted that the model is not a simple enzymatic reaction system , in which one enzyme catalyzes a reaction of one substrate . MT1-MMP catalyzes both of the processing of proMMP-2 and at the same time the degradation of ECM . Thus the present model describes a one enzyme-two substrate system . This will result in the rivalry between two enzymatic reactions . In the previous report , we showed that reduction of the turnover rate of MT1-MMP is one possible therapeutic target [23] . This has been confirmed in the present study , since the steady-state concentration of M14a at higher TIMP-2 concentrations is dramatically decreased by a reduction of the turnover rate ( Figure 3A ) . In addition , the effectiveness of a combined treatment with an increased TIMP-2 concentration and elimination of the sharp transient peak is shown in Figure 4B . The difference in τH between the presence and absence of the sharp transient activity is much larger at higher TIMP-2 concentrations . Thus , at a TIMP-2 concentration of 500 nM , the difference in τH is more than two orders of magnitude . It is thought that the increased TIMP-2 concentration caused by a drug results in a reduced ECM degrading activity . However , it is not the case in the presence of the sharp transient activity as shown in Figure 4B , because the sharp transient activity as short as 4 sec is effective to degrade ECM . The prevention of this sharp transient activity will require a drug with much higher affinity and much faster binding kinetics to MT1-MMP than TIMP-2 . On the other hand , an abrupt increase in the inserted MT1-MMP concentration has only a small effect in ECM degradation . This is shown by histograms of τH for random insertion intervals and concentrations with relatively small variance ( Figures S9 and S10 ) . The vesicular content of MT1-MMP and its insertion intervals can be randomly fluctuated , and insertions at unexpectedly large or unexpectedly short intervals will then occur . From our simulation results , however , its effect is smaller than expected . In fact , an abrupt large amount of insertion has only a small effect on the time course of ECM degradation ( Figure S11 ) . In contrast , the concentration of MT1-MMP in progressively inserted vesicles has a significant effect on ECM degradation . Although the same ECM degradation efficacy was expected for different sets of insertion frequency/concentration with identical time-averaged amount of inserted MT1-MMP , our simulation results have shown that this is not the case , and instead ECM degradation is much efficient at higher concentration and lower frequency regimens ( Figure 6B ) . The time of half-degraded ECM τH was shorter with the higher vesicular content , because of the increased ratio of inserted MT1-MMP/residual ECM . This suggests the importance of reducing vesicular content of MT1-MMP in order to prevent accelerated ECM degradation . If we look at the spatio-temporal dynamics of free TIMP-2 concentration , we found its large spatial gradient at an initial TIMP-2 concentration of 20 nM with increasing spatial gradient with time ( Figure S12A ) . This indicates that the sharp transient peak is a phenomenon that is also observed in the limited accessibility of TIMP-2 at invadopodia . To further confirm this , we ran simulations with different diffusion coefficients from 2×10−11 to 2×10−19 m2/s and with an increased number of invadopodia ( Figures S12B and C ) . The sharp transient peak was still observed at conditions with different diffusion coefficients both in single- and five-compartment cases . It was suggested that the TIMP-2 concentration is lower at invadopodia around the cell center than at lamellipodia around the cell periphery , because of the limited accessibility of TIMP-2 at the cell center [30] . Our simulation results indicate that the sharp transient peak exists even at the situation of limited accessibility of TIMP-2 at the invadopodia . The surface MT1-MMP is subjected to the ectodomain shedding . Fragments of MT1-MMP diffuse freely and bound with TIMP-2 . Thus it acts as a scavenger of TIMP-2 and this reaction can be an additional mechanism limiting the availability of TIMP-2 at invadopodia . However , there is only a negligible effect both on the sharp transient peak and the ECM-degrading activity in our simulation ( Figure S13 ) . Thus in this report , we further characterized the model that we reported on earlier [23] , and found important characteristics in the activity of MT1-MMP in regard to ECM degradation that suggest potential therapeutic targets .
The SCC61 cell lines , which have been described previously [31] , were maintained in DMEM with 20% fetal bovine serum with 0 . 4 mg/ml hydrocortisone . The matrix degradation assay was performed as before [23] . Briefly , fibronectin ( BD Biosciences ) was labeled with Dylight 633 ( Fisher ) by dialysis in borate buffer [0 . 17 mol/L borate , 0 . 075 mol/L NaCl ( pH 9 . 3 ) ] , and the buffer was changed to PBS and dialyzed extensively for 3 to 4 days . A solution of 2 . 5% gelatin/2 . 5% sucrose in PBS was added to dishes to coat MatTek , followed by crosslinking with 0 . 5% glutaraldehyde in PBS . A solution of 50 µg/mL solution of fluorescence-labeled fibronectin was incubated with the crosslinked gelatin in MatTek dishes in the dark for 1 h . The dish was sterilized with 70% ethanol , washed with PBS , and equilibrated with medium [DMEM supplemented with 15% FBS and 5% Nu-Serum ( BD ) ] for 30 min before the addition of cells . A total of 7×104 cells were suspended in 2 mL of medium containing 100 µM EGF . The models were constructed using A-Cell [32] , [33] , and all models can be downloaded from http://www . ims . u-tokyo . ac . jp/mathcancer/A-Cell/index_e . html . The model for continuous insertion of MT1-MMP is the same as the model described earlier [23] . Briefly , the model has two independent pools of X and D that dock MT1-MMP , where docked MT1-MMP interacts with TIMP-2 and forms the ternary complex ( MT1-MMP-TIMP-2-MMP-2 ) . MT1-MMP is internalized and inserted . The model includes the activation of proMMP-2 by MT1-MMP , MMP-2 inactivation by TIMP-2 , and ECM degradation by MT1-MMP and MMP-2 at both pools . In the spatio-temporal simulations , biochemical reactions were embedded into the 3D shape . Details of the models and parameter values are shown in Figures S2 and S6 and Tables S1 and S2 . In the A-Cell model , MT1-MMP , TIMP-2 , and MMP-2 were designated as M14 , T2 , and M2 for simplicity . In our model of MT1-MMP dynamics , the ECM-degrading activity resembles a sharp transient activity followed by a steady state level ( Figure 1C ) . In order to elucidate the role of this sharp transient peak in ECM degradation , we computationally eliminate it by distributing amounts of newly inserted MT1-MMP among all complexes according to their fractions at the steady state level . This method for the elimination of the sharp transient peak was applied both for continuous and pulsatile insertion . Simulations for pulsatile insertion were run both for regular and random intervals . In the regular interval model , the insertion interval for pool X was so calculated such that the time-averaged amount of insertion was the same as that for the continuous insertion model using the following equation: ( 1 ) where PHx , kX and MF are , respectively , amount of MT1-MMP at a single insertion , the rate of insertion in the continuous insertion model , and the concentration of free docking sites for MT1-MMP on the invadopodial membrane , respectively . However , no insertion will occur if all docking sites are occupied by MT1-MMP and there is no available site . Therefore , Eq . 1 should be modified as follows: ( 2 ) where MF0 is the concentration of maximum docking sites . The same method was applied for pool D , and we got tint0 for pools X and D of 25 . 9 sec and 2 . 6 sec , respectively if we assume the concentrations of single insertions of 10% to the maximum available docking concentration for pools X and D , which are 3 nM and 7 nM . For the simulation of regular insertion , MT1-MMP was inserted at a constant interval of the calculated value above . For the simulation of random insertion , MT1-MMP was inserted at random intervals or random concentrations , whose time-average inserted amount of MT1-MMP was the same as in regular insertions . The simulation program was modified for the use of random insertion . A reaction-diffusion simulation program in C language was automatically generated by A-Cell from the constructed model , and compiled using Intel C++ Studio XE 2011 for Linux . The differential equations were numerically integrated by the fourth-order Runge-Kutta method . Simulations were run on a Linux-based system with Intel Xeon X5680 3 . 33 GHz . | Metastasis is the major cause of death in cancer patients . If metastasis is blocked , the survival rate will be greatly increased . Cancer cells are surrounded by ECM ( extracellular matrix ) , which prevents their free movement . MT1-MMP is a potent membrane proteinase that degrades ECM , which is the first step of cancer cell invasion . Thus , the control of MT1-MMP activity is a key to the prevention of metastasis . Here we found a sharp transient peak in the ECM-degrading activity of MT1-MMP by computer simulations , and computational elimination of this peak greatly prolonged the ECM degradation . MT1-MMP is transported intracellularly to the surface of the membrane of cancer cells , and its insertion into the membrane is thought to occur in a pulsatile manner . Therefore , we asked whether the ECM-degrading efficacy was the same in low-frequency/high-concentration and high-frequency/low-concentration insertions of MT1-MMP . Unexpectedly , the low-frequency/high-concentration regimen resulted in much faster ECM degradation even if the time-averaged amount of MT1-MMP insertion was the same . Thus , reduction of the sharp transient activity and vesicular content of MT1-MMP are important therapeutic targets . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"extracellular",
"matrix",
"biology",
"computational",
"biology",
"molecular",
"cell",
"biology"
] | 2013 | Critical Role of Transient Activity of MT1-MMP for ECM Degradation in Invadopodia |
The Y chromosome harbors nine multi-copy ampliconic gene families expressed exclusively in testis . The gene copies within each family are >99% identical to each other , which poses a major challenge in evaluating their copy number . Recent studies demonstrated high variation in Y ampliconic gene copy number among humans . However , how this variation affects expression levels in human testis remains understudied . Here we developed a novel computational tool Ampliconic Copy Number Estimator ( AmpliCoNE ) that utilizes read sequencing depth information to estimate Y ampliconic gene copy number per family . We applied this tool to whole-genome sequencing data of 149 men with matched testis expression data whose samples are part of the Genotype-Tissue Expression ( GTEx ) project . We found that the Y ampliconic gene families with low copy number in humans were deleted or pseudogenized in non-human great apes , suggesting relaxation of functional constraints . Among the Y ampliconic gene families , higher copy number leads to higher expression . Within the Y ampliconic gene families , copy number does not influence gene expression , rather a high tolerance for variation in gene expression was observed in testis of presumably healthy men . No differences in gene expression levels were found among major Y haplogroups . Age positively correlated with expression levels of the HSFY and PRY gene families in the African subhaplogroup E1b , but not in the European subhaplogroups R1b and I1 . We also found that expression of five out of six Y ampliconic gene families is coordinated with that of their non-Y ( i . e . X or autosomal ) homologs . Indeed , five ampliconic gene families had consistently lower expression levels when compared to their non-Y homologs suggesting dosage regulation , while the HSFY family had higher expression levels than its X homolog and thus lacked dosage regulation .
The human Y chromosome harbors 10 . 2 million bases ( Mb ) of ampliconic regions containing nine protein-coding multi-copy gene families [1] . These genes are important not only because of their association with male infertility [1 , 2] but also because they might hold the key to understanding the evolutionary forces that have shaped the Y chromosome . Ampliconic gene families show a high level of copy number variability [3–5] and , possibly , a similar variability in gene expression levels . Understanding the relationship between these two variabilities is an important step in the study of these genes . Yet , there has been no comprehensive investigation to-date that explores expression of these gene families and its connection to copy number at a large , population-level scale . Studying ampliconic gene families has been a considerable challenge because they exhibit a much higher intra-familial sequence similarity than other gene families . The majority ( eight out of nine ) of Y ampliconic gene families are located in palindromes—structures composed of highly similar inverted repeats ( arms ) around a relatively short unique sequence ( spacer ) . The arms within a palindrome are 99 . 9% identical to each other , which results in a high sequence identity among paralogous genes located on the arms [1] . The ninth family , TSPY , is present as an array of tandem repeats outside of palindromes [1] , however its genes still share sequence identity of >99% . It has been hypothesized that the Y chromosome has acquired its ampliconic structure as a way of facilitating gene conversion [6] , which can overcome the decay due to a lack of inter-chromosomal recombination [7 , 8] . Why these ampliconic gene families are preserved on the Y chromosome remains an open question . It has been suggested that this is due to sexual antagonism eventually leading to increased male reproductive fitness [6 , 7 , 9] . Sexual antagonism is expected to lead to the accumulation of genes and mutations benefiting males on the Y chromosome [10] . Consistent with the sexual antagonism hypothesis , all ampliconic genes on the Y are expressed exclusively or predominantly in testis . However , it is also possible that these genes have recently evolved under relaxed function constraints . The ability to analyze the expression levels of Y ampliconic genes at a large scale can help in exploring their potential functional constraints via comparing their testis expression level to that of their non-Y homologs ( when available ) . For instance , if a Y ampliconic gene family undergoes neo-functionalization , then its resulting expression level is expected to be independent of and potentially higher than that for its non-Y homologs ( which we assume retained the ancestral function ) . In support of some functional constraints is the observation that the loss or partial deletion of Y ampliconic gene copies is linked to infertility in humans . For example , TSPY copy number was linked to both infertility [11] and sperm count [11–13] . The long arm of the human Y chromosome includes three azoospermia factor regions ( AZFa , AZFb , and AZFc ) , which cover most of the ampliconic genes families and are active during different phases of spermatogenesis [14] . Complete or partial deletion of these regions is linked to azoospermia and arrest of spermatogenesis [2 , 12 , 14–16] . Presumably , copy number decrease linked with infertility is accompanied by a reduction in gene expression of the affected Y ampliconic gene families , however this is yet to be demonstrated . Recent studies indicated high variation in Y ampliconic gene copy number in healthy men [3–5] . Skov and colleagues [4] studied Y ampliconic gene copy number variation in 62 men of Danish descent and identified multiple copy number changes across all nine gene families among unrelated individuals , as well as copy number differences for the TSPY and VCY gene families between a father and a son . Ye and colleagues [3] assessed Y ampliconic gene copy number variation in 100 individuals from around the world . They observed that the size of gene family is correlated with its variation in copy number: larger families , such as TSPY and RBMY , have higher levels of variation , however the variation appears to be independent of the Y haplogroup . Two men rarely had the same Y ampliconic gene copy number profile and , when they did , this was likely a result of homoplasy . Lucotte and colleagues [5] used the data from the Simons Genome Diversity Project [17] and observed substantial variation in copy number in six out of nine human Y ampliconic gene families [5] . Teitz and colleagues [18] assessed copy number of full-length Y chromosome amplicons located in the AZFc region in men sequenced by the 1000 Genomes Project [18] . Their results suggest that selection has preserved the ancestral ampliconic gene copy number on the Y chromosome in diverse human lineages [18] . These multiple studies of copy number notwithstanding , there has been little investigation of gene expression of Y ampliconic genes . A recent study investigating the expression of Y ampliconic genes during male meiosis found that gene families with high variation in copy number also have high expression levels at different stages of sperm development [5] . Other than the results of this single study , there is a big gap in our understanding of variation in expression of Y ampliconic genes among humans , even though gene expression could be a better predictor of genes’ functions than copy number . Additionally , previous studies have reported that aging affects gene expression [19 , 20] . Even less is known about how variation in copy number of Y ampliconic genes affects their gene expression . Most parsimoniously , a gain of a complete gene copy should lead to an increase in gene expression levels , unless the extra copy obtains a new function through neo-functionalization , has decreased functional demands due to sub-functionalization or is lost due to pseudogenization . Indeed , this parsimonious hypothesis was supported by the data from the 1000 Genomes Project , where most genes overlapping multiallelic copy number variations ( CNVs ) display a positive correlation between copy number and gene expression [21] . However , studies across different model organisms have reported that differences in copy number result in increased , decreased or unchanged expression levels among individuals in a population [22] . This more complex relationship can be caused by several scenarios during duplication . For instance , a tandem duplication event may not include regulatory elements , may physically disrupt topologically associated domains ( TADs ) , which prevents the interaction of the gene with its enhancer in 3D space [23 , 24] , or may result in a new copy acting as a negative feedback loop to reduce transcription [22] . Moreover , a non-tandem duplication may occur to a site that is not transcriptionally active [22] . Which of these parsimonious or more complex scenarios occurs on the human Y chromosome ampliconic genes has not been explored . In this study , we explored the above questions by analyzing the largest data set available to-date consisting of expression data from testis , along with matched whole-genome sequencing data , from 170 men , as generated by the Genotype Tissue Expression ( GTEx ) consortium [25] . Simultaneously , we developed a novel computational tool AmpliCoNE to estimate the copy number of an ampliconic gene family from sequencing data . Such estimation is complicated by the presence of multiple highly-similar gene copies in the reference , which makes conventional tools inapplicable [26] . Custom strategies have been developed and shown to be effective [4 , 5 , 21 , 27–29] , but we did not identify any existing software that could be run directly on Y chromosome ampliconic gene families . Using AmpliCoNE , we explored whether variation in Y ampliconic gene expression levels could be explained by variation in gene copy number , Y haplogroup , and individual’s age . We correlated the estimated with AmpliCoNE copy numbers of Y ampliconic gene families to their expression levels in testis , and studied how this correlation is affected by Y haplogroups . Additionally , we investigated how testis-specific expression of Y ampliconic genes diverged from their non-Y homologs during evolution .
AmpliCoNE is composed of two programs . The first ( AmpliCoNE-build ) is executed only once to process the reference genome . It takes the location of all the gene copies in the reference genome , grouped by family , determines which positions in the genes are informative ( i . e . where read depth is an effective predictor of copy number ) and which positions in the reference can be used as a control ( where copy number variation is infrequent and the read depth has limited noise ) . The second step ( AmpliCoNE-count ) is then executed separately for every sample . It parses read alignments and measures the GC-corrected read depth at the informative positions . It then accumulates this information at a family-level and reports the copy number for each gene family , using the read depth at control positions as a baseline . We provide further details in the Methods . To evaluate AmpliCoNE’s accuracy , we ran it on simulated data and whole-genome short-read data from the Genome in a Bottle ( GIAB ) consortium [30] . Using the hg38 human genome reference , we simulated three datasets with varying copy numbers of RBMY , TSPY , and VCY gene families and kept the copy numbers for the remaining six gene families constant ( i . e . with the copy number found in the reference ) . AmpliCoNE estimated ampliconic copy numbers correctly 100% of the time in the simulated datasets ( S1 Table ) . We then compared gene family copy numbers between different GIAB experimental runs ( technical replicates ) for the same human sample ( S2 Table ) , as well as between a father and a son ( which can be treated as biological replicates because copy number differences between generations are expected to be rare [4] ) . AmpliCoNE consistently predicted copy numbers with a difference of less than 0 . 5 copies per family . We tested AmpliCoNE at different depths of coverage and showed that it can predict similar copy numbers ( estimates with difference of less than 0 . 5 ) even for datasets with the Y chromosome sequencing depth as low as 6x ( S3 Table ) . AmpliCoNE’s runtime is dependent on the number of reads it needs to process . For instance , it took AmpliCoNE 11 minutes to process the GTEx Y-chromosome-specific BAM file ( ~500 MB in size ) . To measure the concordance between AmpliCoNE’s copy number estimates and complementary non-sequencing assays , we used droplet digital PCR ( ddPCR ) . Both AmpliCoNE and ddPCR were applied to estimate Y ampliconic gene copy numbers for four males sequenced by the GIAB consortium ( Tables 1 and S4 ) [30] . The ddPCR estimates were identical to AmpliCoNE estimates for five out of nine gene families ( BPY2 , DAZ , HSFY , PRY , and XKRY ) in all four samples . The CDY and RBMY family copy numbers differed between the two methods in only one and two individuals , respectively . The VCY and TSPY family copy number estimates differed in three and four individuals , respectively . Compared with ddPCR , AmpliCoNE consistently underestimated the copy number for the VCY gene family . Previous studies have indicated presence of X-to-Y gene conversion between VCX and VCY [31 , 32] . We investigated this case in more detail and discovered that genes from the VCY family harbor only a very short ( 220-bp ) sequence distinguishing them from their VCX paralogs . This sequence has a low sequencing depth even after GC correction , which results in the underestimation of the VCY copy number by AmpliCoNE . In the case of TSPY , it is known to have many highly-similar pseudogene copies which may themselves vary in copy number , which can potentially confound both AmpliCoNE and ddPCR estimates . These caveats notwithstanding , AmpliCoNE’s biases in estimating copy numbers for TSPY and VCY are consistent across samples and thus should not affect our results in a systematic way . Using AmpliCoNE , we estimated copy numbers of Y chromosome ampliconic genes in 170 presumably healthy men whose genomes were sequenced in their entirety as part of the GTEx project [25] . These individuals ( S5 Table ) were selected because they had matched testis expression data . The individuals belonged to ten major haplogroups: B , E , G , I , J , L , O , Q , R , and T ( Table 2 ) . The majority of the samples in the dataset had European or African Y haplogroups , with a few Asian haplogroups present . We also used AmpliCoNE to estimate the copy numbers of X-degenerate genes , which are expected to be 1 in healthy samples . Three samples had copy number estimates close to zero for two or more ampliconic gene families , or had less than one copy for several X-degenerate genes , which could suggest an individual with a disease or could result from a technical artifact , and thus were removed from the downstream analysis . As a result , we retained 167 samples . Gene families with higher median copy number had higher variation when compared to gene families with lower median copy number ( R2 = 0 . 91; S1 Fig ) . RBMY and TSPY were the largest gene families and displayed the highest variation in copy number ( 5–14 and 20–64 copies for RBMY and TSPY , respectively ) . HSFY , PRY , VCY , and XKRY were the smallest gene families , which on average had two copies per individual , and displayed low variation in copy number . We observed a positive correlation in copy number among BPY2 , CDY , and DAZ gene families , which could be explained by their co-localization on palindrome P1; duplication or deletion involving P1 can affect the copy numbers of all three gene families ( Fig 1A ) . We expected to observe a higher probability for gene families with lower median copy number to be completely deleted due to random rearrangements . Therefore , we aimed to test whether the gene families with lower copy number in human had a higher chance of being deleted in non-human great ape species . It is known from previous studies that the VCY gene family is missing in bonobo , gorilla , and orangutan , whereas the HSFY , PRY , and XKRY families are missing in bonobo and chimpanzee [33] . Consistent with our hypothesis , the HSFY , PRY , VCY , and XKRY gene families had low copy numbers in humans ( S1 Fig; S6 Table ) . To explore the relationship between ampliconic gene copy number and their expression levels , we analyzed testis expression data from the same 167 humans whose Y ampliconic gene copy number was estimated with AmpliCoNE . After removing outliers ( see Materials and Methods ) , we retained 149 samples and obtained expression levels for each gene family—the sum of expression of all the gene copies within each family—in each of them . We found that , similar to our observation for copy numbers ( S1 Fig ) , families with higher gene expression levels had higher variation in gene expression ( R2 = 0 . 99; S2 Fig ) . The TSPY family had the highest gene expression level and the highest variation in expression across individuals , and XKRY—the lowest ( S6 Table; S2 Fig ) . The XKRY gene family could be considered to be not expressed ( as its expression levels are zero ) in 58 individuals or expressed at very low levels ( with DESeq2 normalized read count < 10 ) in the remaining 91 individuals . DAZ , HSFY , and RBMY gene families had similar median expression levels and variance among themselves ( S6 Table; S2 Fig ) . Within our dataset , we found two sets of ampliconic gene families whose expression levels were positively correlated with each other ( Fig 1B ) . The first set included BPY2 , CDY , HSFY , and PRY , and the second set—DAZ , TSPY , RBMY , and VCY ( Fig 1B ) . The expression of these sets of gene families could be co-regulated or might have cell-type specificity . When we investigated the relationship between expression levels and copy number among all 149 individuals across nine ampliconic gene families , we found that more copious gene families tended to have higher expression levels in comparison to the less copious gene families ( Fig 2 ) . Indeed , the expression levels were positively correlated with estimates of copy numbers ( Spearman's rank correlation rho = 0 . 43; P-value < 2 . 2x10-16 ) . The DAZ , HSFY , and VCY gene families appeared to be outliers in this analysis , as they had gene expression levels similar to the RBMY gene family even though their median copy number estimates were approximately half or less than half of RBMY gene family . The DAZ gene family had similar gene copy number yet higher expression levels when compared to the CDY gene family . The XKRY family consistently had very low expression levels , even though its median copy number per individual was two . Next , we tested whether copy number , as measured for each individual , is positively correlated with gene expression levels , again measured for each individual , within the same gene family . There was no significant correlation in any of the nine families studied ( all P-values were above the Bonferroni-corrected P-value cutoff of 0 . 05/9 = 0 . 006; S3 Fig; S7 Table ) . To control for genetic variation on the Y , we next compared copy number estimates to gene expression levels for individuals with the same Y subhaplogroup . We focused on the European R1b and I1a , and the African E1b subhaplogroups because they had more than 10 individuals in our dataset ( 77 , 15 and 22 , respectively; Table 2 ) . We still found no significant correlations between copy number and expression levels in any of the nine gene families for individuals from either of these three subhaplogroups ( all P-values were above the Bonferroni-corrected P-value cutoff of 0 . 05/9 = 0 . 006; S4–S6 Figs; S7 Table ) . We further asked whether the major Y haplogroup could at least in part explain the variation we observed in copy number and in gene expression levels of Y chromosome ampliconic genes . We focused our analysis on major haplogroups R ( European ) , I ( European ) , E ( African ) , and J ( Western Asian ) because they were represented by at least 10 samples in our dataset ( Table 2 ) . Using one-way ANOVA , we found that the copy numbers of BPY2 ( P = 2 . 34x10-3 ) , RBMY ( P = 2 . 97x10-8 ) , and TSPY ( P = 1 . 07x10-22 ) gene families had significant differences among the four major Y haplogroups analyzed ( Bonferroni-corrected P-value cutoff of 0 . 05/9 = 0 . 006; Table 3 ) . The remaining six gene families did not display significant differences among Y haplogroups ( Table 3 ) . When we compared the mean copy number differences between haplogroups in a pairwise fashion using a permutation test ( 1 million permutations; 9 gene families are tested for 6 cases—R vs E; R vs I; R vs J; I vs E , I vs J , E vs J—thus we performed 9 x 6 = 54 tests; Bonferroni-corrected P-value cutoff of 0 . 05/54 = 0 . 00093 ) , TSPY differed significantly in copy numbers ( Fig 3 ) between major European ( R and I ) vs . African ( E ) or vs . Western Asian ( J ) haplogroups ( P = 0 for R vs . E; P = 0 for I vs . E; P = 0 for R vs . J; P = 0 . 3x10-5 for I vs . J; S8 Table ) . RBMY copy numbers differed significantly between European ( R ) vs . African ( E ) or Western Asian ( J ) haplogroups ( P = 6 . 94x10-4 for R vs . E; P = 0 for R vs J; S8 Table ) . No significant differences between the two major European haplogroups ( R and I ) were observed ( S8 Table ) . In contrast , we found that gene expression levels of all nine Y ampliconic gene families were not significantly different among major Y haplogroups ( all P-values were above the P-value cutoff of 0 . 05/9≈0 . 006; one-way ANOVA; Table 3 ) . We observed a trend suggesting differences in expression values among haplogroups for the BPY2 and DAZ gene families , but these differences were small in scale . Nevertheless , out of the nine gene families , BPY2 ( P = 0 . 056 ) and DAZ ( P = 0 . 01 ) had low P-values for the ANOVA analysis ( Table 3 , Fig 3 ) and for the permutation test comparing mean expression levels between haplogroups ( P = 7 . 09x10-3 for E vs . R for BPY2; P = 1 . 36x10-2 for E vs . R for DAZ; P cutoff of 0 . 05/54 = 0 . 00093; S9 Table ) . When we compared the trend in copy number and gene expression differentiation among haplogroups , we observed that in the TSPY gene family both copy number and gene expression levels were lower for the European haplogroups ( I , R ) than for the African ( E ) or Western Asian ( J ) haplogroups ( Fig 3 ) . This trend was statistically significant for copy number , but not significant for gene expression . Analyzing a larger sample size might lead to finding this trend to be significant also for gene expression . To examine the potential role of aging in determining Y ampliconic gene expression , we compared the ages of individuals at the time of sample collection to the ampliconic gene expression levels and found no statistically significant relationship ( nine gene families were tested for correlation which results in Bonferroni correction P-value cutoff of 0 . 05/9 = 0 . 006; S7 Fig; S10 Table ) . Next , to perform a similar analysis for individuals with the same subhaplogroup , we limited our analysis to individuals with the European R1b and I1a , and African E1b subhaplogroups ( 77 , 15 , and 22 individuals , respectively ) . For the R1b and I1a subhaplogroups we found no significant relationship between age and expression levels for any of the nine Y ampliconic gene families studied ( S8 and S9 Figs; S10 Table ) . However , for the African E1b subhaplogroup , HSFY ( Spearman correlation = 0 . 57; P = 0 . 0061 ) and PRY ( Spearman correlation = 0 . 61; P = 0 . 0028 ) gene families had a positive correlation between expression levels and age , which was significant after Bonferroni correction ( S10 Fig; S10 Table ) . A larger dataset of African samples should be studied to validate this relationship . The presence of homologs outside of the Y for two groups of Y ampliconic gene families allows us to study evolution of their gene expression levels [34] . In particular , the CDY and DAZ genes were copied to the Y chromosome from autosomes [34]; the HSFY , RBMY , TSPY , VCY , and XKRY gene families have homologs on the X and were likely present on the ancestral autosomes giving rise to the two sex chromosomes [34] . In the analyses below , we assume that the testis-specific expression of Y ampliconic genes was acquired prior to their amplification on the Y [9] and that their autosomal or X-chromosomal homologs have maintained ancestral expression levels , i . e . they possess expression levels of Y ampliconic genes prior to their Y linkage [35] . The latter assumption is based on the overall slower rates of evolution of X-chromosomal and autosomal genes as compared to their Y-chromosomal homologs . We envision three possible scenarios for gene expression evolution of Y ampliconic gene families that have non-Y homologs ( Fig 4 ) . First , because of sexual antagonism , a gene on the Y could obtain beneficial mutations and diverge in function from its non-Y homolog to acquire new functions in testis ( i . e . neo-functionalization ) . The expression of such a gene family would be independent from , and potentially higher than that for , its non-Y homologs ( scenario A ) . Second , a gene family on the Y could retain function of the non-Y homolog , but acquire testis-specific expression ( i . e . sub-functionalization ) . In this case , either the non-Y copy represents the ancestral expression levels and the Y copies are expected to maintain low expression levels , or the sum of expression from the Y and non-Y copies is regulated to be at levels similar to those of the non-Y copy in the ancestor ( scenario B ) . In this case , the expression of both Y and non-Y homologs might be down-regulated . Third , genes on the Y might be under relaxed selective constraints and thus have low expression levels ( scenario C ) [36] . Below we test these three scenarios by comparing expression levels of both Y and non-Y ampliconic gene homologs in testis tissue . In addition to the analysis of such overall differences in the expression level ( Fig 4 ) , we can also examine the relationship between the Y ampliconic genes’ and their non-Y homologs’ gene expression across individuals , which should further assist in determining a particular evolutionary scenario ( S11 Fig ) . If the expression levels of Y ampliconic genes are higher than those of their non-Y homologs , and across individuals the expression levels of these two groups of genes are positively correlated , then this pattern is consistent with neo-functionalization of the Y ampliconic genes . This is because higher expression levels of ampliconic genes than those at the ancestral state suggest independent expression of Y ampliconic genes from their non-Y homologs , and a positive correlation between Y ampliconic genes and their non-Y homologs suggests co-regulation , e . g . they might share similar transcription factors [37] . A combination of these two patterns suggests an acquisition of a new function ( scenario A ) ( S11A Fig ) . If the expression levels of Y ampliconic genes are higher than those of their non-Y homologs , and across individuals the expression levels of these two groups of genes are negatively correlated , then the data are compatible with neo- or sub-functionalization ( scenario A or B ) . Indeed , the observed negative correlation could be explained by neo-functionalization , where ampliconic genes acquired a new function and inhibit the expression of the non-Y homologs . Alternatively , the negative correlation could be explained by sub-functionalization , where ampliconic genes acquired new transcription factors which limit their expression to a few cell types , and the negative correlation is due to the differences in the abundance of cell types in which ampliconic genes are expressed ( S11B Fig ) . If the expression levels of Y ampliconic genes are lower than those of their non-Y homologs , and across individuals the expression levels of these two groups of genes are positively correlated , then this pattern is consistent with any of the three scenarios A-C . This is because the lower expression levels of Y ampliconic genes could be due to down-regulation of gene expression by the Y chromosome to accommodate the multi-copy state of ampliconic genes [38] , evolution of which could still be compatible with any of the three scenarios A-C ( S11C Fig ) . If the expression levels of the Y ampliconic genes are lower than those of their non-Y homologs , and across individuals the expression levels of these two groups of genes are negatively correlated , then the data are compatible with scenario A or B . This is because negative correlation eliminates the scenario of relaxed selection , i . e . scenario C ( S11D Fig ) . Finally , if we observe no correlation in expression levels between Y ampliconic genes and their non-Y homologs , then we can conclude that their expression is independent from each other , which could be a result of neo-functionalization , sub-functionalization or random drift in expression levels under relaxed selection . To test these scenarios , we first compared testis expression levels between Y ampliconic gene families CDY and DAZ , which were copied to the Y from autosomes , and their autosomal homologs ( Fig 5 ) . The CDY autosomal homologs CDYL and CDYL2 are ubiquitously expressed; and the DAZ autosomal homolog DAZL has testis-specific expression [34 , 39–41] . The expression levels of CDY ( the sum of expression levels for the whole gene family ) were 89% lower than those for their autosomal homologs ( the sum of expression of CDYL and CDYL2 ) , and for DAZ they were 63% lower than those for their autosomal homolog DAZL ( Fig 5 ) . Next , we tested whether the expression levels for Y ampliconic genes and their autosomal homologs are regulated at the level of each individual . For each gene family , we examined a potential correlation in gene expression levels between the Y ampliconic genes and their non-Y homologs . We observed a significant negative correlation between CDY and CDYL+CDYL2 expression levels ( Spearman correlation = -0 . 31; P = 2x10-4 ) , which indicates that , across individuals , whenever the CDY expression levels increase , the CDYL+CDYL2 expression levels decrease ( Fig 6 ) . In case of DAZ , a positive correlation in expression levels ( Spearman correlation = 0 . 57; P = 0 ) was observed between DAZ and its autosomal homolog DAZL ( Fig 6 ) . Lower expression of CDY and DAZ than their non-Y homologs could be a result of down-regulation of gene expression by Y chromosome to maintain the multi-copy state , however the negative correlation in CDY vs . CDYL+CDYL2 expression levels indicates the presence of either neo- or sub-functionalization . DAZ could have undergone any of the three scenarios , which are difficult to differentiate based on the available data . We next examined how testis-specific gene expression of the HSFY , RBMY , TSPY , VCY , and XKRY gene families diverged from that of their X homologs . Most of the X homologs of ampliconic genes ( except for VCY and XKRY ) are expressed in multiple tissues along with testis . The XKRX gene , an X homolog of the XKRY gene family , is not expressed in testis and we omitted this gene family from our analysis ( S11 Table ) . Three Y gene families studied ( RBMY , TSPY , and VCY ) on average had lower expression levels in comparison to their X homologs ( 66% , 75% , and 71% lower , respectively; Fig 5 ) . HSFY was the only gene family that on average had higher expression in comparison to their homologs on the X ( 35% higher than X-homologs ) . This could imply that HSFY might have acquired a new function , which is selected for in testis ( scenario A ) . At the level of the studied individuals , all four studied gene families exhibited positive correlation in gene expression levels between their Y ampliconic and X homolog genes , suggesting a potential co-regulation ( Fig 6 ) . This correlation was particularly strong for the HSFY and VCY gene families ( Spearman correlation of 0 . 69 and 0 . 84 , respectively ) . The observed higher expression of HSFY than of its X homologs , as well as positive correlation in gene expression levels between these two groups of genes , is a strong indicator of neo-functionalization . In the case of RBMY , TSPY , and VCY , it is challenging to differentiate among the three scenarios we propose based on the available data .
Our results indicate that smaller Y ampliconic gene families maintain lower variation in copy number and , as the size of gene families increases , variation in copy number also increases , in agreement with previous studies [3–5] . The parsimonious explanation for this observation is that a greater number of gene copies leads to loss or gain of gene copies because of a higher probability of rearrangements via replication slippage and/or non-allelic homologous recombination ( NAHR ) [42–44] . On the human Y , the larger gene families are either spread across multiple palindromes ( e . g . , RBMY ) or are arranged as a tandem array ( TSPY ) , and such arrangements can result in multiple scenarios of NAHR , which will lead to gain or loss of gene copies . BPY2 has two functional copies on palindrome P1 and one copy outside of palindromes , and such an arrangement can also result in NAHR . We found that the large TSPY and RBMY gene families have not only a high level of variation in copy number , but also a significantly different number of gene copies among the major Y haplogroups analyzed . An earlier study also found significant differences in copy number for these two gene families among human Y haplogroups across the world and suggested that this observation cannot be explained by selection [3] . However , selection explanation might warrant a further investigation . Indeed , a recent molecular analysis of infertile men indicated a positive correlation between the number of RBMY copies and sperm count and motility [45] . Moreover , RBMY is a male-specific oncogene [46] . Therefore , it will be of interest to investigate whether variation in RBMY copy number across Y haplogroups influences these two disease-related phenotypes and might be subject to natural selection . Similarly , TSPY is a candidate proto-oncogene which can regulate its own expression via a positive feedback loop in gonadoblastoma and a variety of somatic cancers [47] . Thus , additional studies should be performed to test whether variation in TSPY copy number across haplogroups is associated with differential predisposition to gonadoblastoma . The smaller Y ampliconic gene families ( HSFY , PRY , VCY , and XKRY ) have lower variation in copy number compared with larger families . These gene families , for which the average family size is only two copies , are each present on an individual palindrome ( the two copies are present as inverted repeats on opposite palindrome arms ) . Recombination between inverted repeats is expected to result in an inversion keeping copy number constant [48] . In addition , the presence of only two copies increases the chances of a complete gene family elimination due to Muller’s ratchet or of rearrangements which involve the whole palindrome . Consistent with this prediction , we find these gene families to be deleted or pseudogenized in several great ape species [33] . Thus , the copy number of ampliconic genes is an important factor in determining the survival of a gene family on the Y chromosome . Too few copies can lead to a complete loss of a gene family ( see the preceding paragraph ) , whereas too many copies can lead to frequent NAHR which can rapidly increase or decrease copy number [49] . Consistent with this expectation , it was suggested that the human Y chromosome evolves under selection to maintain an optimal copy number for its amplicons in diverse human lineages [18] . Most likely both random genetic drift and natural selection contribute to determining the Y chromosome ampliconic gene copy number . Drift leads to smaller-scale changes in copy numbers , whereas selection might act at removing extreme copy numbers because too few copies might lead to infertility and too many copies might lead to genetic instability and thus both are selected against . Variation in Y ampliconic gene copy number in subfertile and infertile males should be investigated in future studies and should shed additional light on the balance between these two evolutionary processes . Note that in the present study we only examined complete gene copy gains or losses , but insertions and deletions inside a gene can also affect gene expression and functionality , and might be linked to infertility [50] . The effects of such smaller CNVs are more robustly evaluated from long-read data and we leave this exploration to future work . Here we studied the expression levels of the Y ampliconic gene families in testis tissue of presumably healthy individuals . The vast majority of cells in testis are germline cells in the seminiferous ducts , where spermatogenesis takes place . We primarily captured Y chromosome gene expression in spermatogonia prior to meiosis and throughout different spermatogenesis stages after meiosis [51 , 52]; this is because Y transcription is silenced at other stages of spermatogenesis due to meiotic sex chromosome inactivation [52 , 53] and postmeiotic sex chromosome repression [51 , 52] . As a tissue , testis is a mixture of germline cells at different stages of development , Sertoli cells , myoid peritubular cells , and interstitial Leydig cells . Thus , the expression values generated using testis tissue as a source represent cumulative gene expression of germline cells at different stages of spermatogenesis with a mixture of somatic cells . This potential limitation notwithstanding , our results indicate substantial variation in expression levels for Y ampliconic genes in testis among men and suggest that different levels of Y ampliconic genes’ expression are tolerated by presumably healthy individuals . When we compared copy number of ampliconic genes to their gene expression values , we found that across gene families the gene families with higher median copy number had higher expression levels . This is consistent with an observation made by Lucotte and colleagues [5] who reported on the expression of Y ampliconic genes at different stages of spermatogenesis with respect to variation in their copy number . Overall , the Y chromosome has higher copy number of genes for those gene families whose median expression levels are higher in testis , however it is important to note that this relationship might be different at individual cell types in testis and should be studied further . When we examined the relationship between copy number and expression within a gene family , our analysis revealed that expression of Y ampliconic gene families is independent of their copy number . Moreover , no significant differences in Y ampliconic gene expression levels were observed among Y haplogroups , even though we found significant differences among Y haplogroups in copy number for some gene families ( BPY2 , TSPY , and RBMY ) . This suggests that testis tissue might have evolved the ability to tolerate different Y ampliconic gene copy numbers , and also variable Y ampliconic gene expression levels . Approximately 77% of all protein-coding genes in the human genome are expressed in testis [54] , and some of these genes could regulate expression of the Y ampliconic genes . Understanding the 3D organization and chromatin structure on the Y is expected to aid in identifying the genomic regions and genes that ampliconic genes interact with and are regulated by in the genome . Future studies analyzing expression data at different stages of spermatogenesis in individuals with different Y ampliconic gene copy numbers will assist in deciphering the role of copy number variation in determining gene expression in more detail . Additionally , our findings should be confirmed by studies of gene expression at the protein level . A man’s advanced age has significant negative impact on reproduction [55] . Semen parameters such as daily sperm production , total sperm count , and sperm viability are negatively correlated with age [56] . However , within our dataset , we observed mixed results regarding age effects on Y ampliconic gene expression: age did not influence variation in gene expression of these genes in individuals with European Y subhaplogroups I1a and R1b , however HSFY and PRY expression had a positive correlation with age in individuals with an African subhaplogroup E1b . These findings should be validated with a larger data set to examine the role of Y ampliconic genes in changes in spermatogenesis with age . The Y chromosome degradation , which is common across eutherian mammals , has resulted in the loss of the majority of genes originally present on the proto-sex autosomal pair [57] . To balance the loss of genes on the Y in males , the mammalian X chromosome adapted its expression levels by inactivating one of its copies and increasing the expression of the other copy in females [57–59] . We wondered whether a similar process evolved at Y ampliconic genes that have non-Y homologs , namely whether the expression of Y ampliconic genes and their non-Y homologs has been co-regulated . Alternatively , Y ampliconic genes might have evolved new functions , and thus potentially high expression levels , independent of their non-Y homologs . Yet another alternative would be the overall low expression levels because of the relaxation of functional constraints on the Y ampliconic genes . The precise functions of Y ampliconic genes have been under-characterized ( S12 Table ) due to the repeated nature of the Y chromosome and scarcity of testable orthologs in model organisms . While Y ampliconic genes have testis-specific expression likely as a result of sexual antagonism , the majority of non-Y homologs of Y ampliconic genes have ubiquitous expression . Recently , a multi-step model for preservation of tandem duplicate genes was presented . According to this model , the expression of gene duplicates is down-regulated immediately after the duplication event , followed by dosage sharing which could lead to functional adaptations such as sub- or neo-functionalization [38] . Knowing that non-Y homologs of Y ampliconic genes are expressed in testis ( except for XKRX ) , we compared the expression levels of closely related homologs of ampliconic genes on both autosomes and X chromosome to the sum of expression levels for all the copies of a Y gene family . We demonstrated that , with the exception of the HSFY family , Y ampliconic gene families have consistently lower expression levels when compared to their non-Y homologs , thus not elevating the overall expression level of the family . We term this phenomenon dosage regulation of Y ampliconic genes . Lower expression of Y ampliconic gene families could be an adaptation of the Y to maintain the multi-copy state of ampliconic gene families . By lowering the expression of the whole gene family , the Y can buffer sudden loss or gain of gene copies . In addition to dosage regulation , the gene family should be expressed at optimal levels to maintain their functionality during spermatogenesis . Lower optimal expression of Y ampliconic gene families compared to their non-Y homologs could be a result of sub-functionalization ( e . g . , testis specificity in expression ) , which benefits germline cell development . Alternatively , such low expression could be a result of relaxed selection , and , in agreement with this possibility , Y ampliconic genes show a higher rate of nonsynonymous to synonymous substitution rates compared to single-copy X degenerate genes on the Y [7] . Alternatively , a gene family could be under positive selection or undergoing neo-functionalization even in their low-level expression state . The expression of ampliconic gene families is important for spermatogenesis because of an association between gene deletions and infertility , but relaxed selection can facilitate rapid differentiation of ampliconic gene function . We found that expression levels of the CDY ampliconic genes and those of their autosomal homologs are negatively correlated among individual men . This suggests that the CDY gene family might not be expressed at the same time during spermatogenesis as its autosomal homologs or that there is a coordinated down-regulation of CDY expression with a rise in CDYL and CDYL2 expression ( and vice versa ) . In humans , the CDYL and CDYL2 autosomal genes produce the ubiquitously expressed long transcripts , but lost the testis-specific short transcript which is now produced by CDY [40] . The combined tissue expression patterns of CDY , CDYL , and CDYL2 in human recapitulate the expression patterns of CDYL and CDYL2 in mouse or rabbit , which do not have CDY on their Y chromosome [40] . In contrast with CDY , we found that expression levels of DAZ , HSFY , and VCY gene families are strongly positively correlated with their non-Y homolog expression across individuals , which suggests a co-regulation in gene expression levels of these ampliconic gene families and their homologs ( the RBMY and TSPY families also show positive correlation , however it is not strong ) . When we examine the linear relationship between ampliconic gene families and their homologs among individual men , the Y ampliconic gene expression increases at a slower pace when compared to the expression of their non-Y homologs , except for HSFY where the expression increases at a similar rate for both Y and non-Y homologs ( Fig 6 ) . The VCY gene family is the most commonly lost gene family among great apes , however in our dataset the expression of this gene family is higher than for most other gene families on the Y and is higher than is predicted from its copy number ( Fig 2 ) . The homologs of VCY on the X chromosome ( VCX , VCX2 , VCX3A , and VCX3B ) are expressed in testis [60 , 61]—and we show that at higher levels than the VCY family itself . In addition , there is high sequence identity ( >95% ) between the VCX and VCY gene families , which could imply that both VCX and VCY could have been under selection to maintain function of the gene family , however , to balance the expression of the multi-copy VCX family , VCY might have lowered its expression . The role of both VCX and VCY in ribosome assembly in spermatogenesis has been suggested [62] . The loss of VCY in great ape species might have been compensated by functionally similar VCX family expression in testis . The expression levels of the VCX family across great apes must be studied to understand its role in the loss of VCY . A recent study found multiple distinct clusters of full-length Y ampliconic gene transcripts , likely originating from different copies of the same family [63] . Therefore , the presence of multiple full-length transcripts [63] and low expression levels for Y ampliconic gene families ( the present study ) suggest that individual gene copies within a family are down-regulated to accommodate the expression of the whole gene family on the Y chromosome and outside of it ( on autosomes and on the X ) . This hypothesis needs to be examined in future studies in which expression levels of individual gene copies will be evaluated with long-read sequencing technology . It will also be important to decipher the isoforms and their expression levels for Y ampliconic genes and their non-Y homologs to understand whether Y ampliconic genes and their homologs express the same isoforms , or whether Y ampliconic genes express their own , unique , testis-specific isoforms . It is essential to note that , in addition to evolution of expression levels of the whole gene family including its non-Y homologs , the Y ampliconic genes can diverge to acquire additional male-specific functionality because they are present on the Y , which is susceptible to accumulating genetic differences dictated by sexual antagonism . In other words , Y ampliconic genes could have gained secondary functions independent of their functions on the proto-sex chromosomes . This scenario might be exemplified by the case of the HSFY family , whose expression levels have increased in comparison to its X-chromosome homologs . This pattern suggests that this gene family underwent neo-functionalization . The exact function of HSFY is unknown , but its role in transcription regulation has been suggested because it harbors a DNA-binding domain [64] . In fact , it was shown that HSFY and HSFX share only this DNA-binding domain but not the rest of their sequences and thus indeed might have diverged in their functions [64] . Moreover , HSFY has stage-specific expression during spermatogenesis , suggesting that it acquired a function different than that of heat shock proteins it is homologous to [64] . The loss of HSFY was linked to infertility [64–66] . In another study , under-expression of HSFY was linked to arrest of maturation of nascent germ cells to motile sperm [67] . According to our study , the expression of HSFY gene family was positively correlated with age in the African E1b Y haplogroup , however such a relationship was not found in the R1b haplogroup . Further studies addressing transcription regulation by the HSFY family in individuals of varying age across different Y haplogroups are required to understand the HSFY functionality in more detail . We assume that non-Y homologs have retained the ancestral expression state because of the overall fast evolutionary rate on the Y chromosome [68] . However , the X chromosome and autosomes have also been evolving , albeit slower than the Y . Evolutionary changes acquired by non-Y homologs since they diverged from the Y homologs have not been addressed in this study due to the lack of ancestral expression data . To address this , future studies should identify species which have orthologs of human ampliconic genes in a single-copy state on their Y chromosome . In the case of CDY and DAZ gene families , future studies should identify species in which these genes’ orthologs are present in a single-copy state on their autosomes and absent from the sex chromosomes . Once such species are identified , their testis-specific expression data for these genes could be used as the ancestral expression state .
To estimate copy number in highly-similar multi-copy gene families , several strategies have been proposed . One can align each read to all possible locations in the reference genome [69] , identify sites in the reference genome that uniquely distinguish and tag paralogs of interest [18 , 21 , 27 , 29] , use simulated reads for mock genomes with human gene cDNAs at different gene copy counts to obtain a theoretical function of the coverage distribution with respect to copy number [28] , or customize the reference to keep a single copy of each gene family [4 , 5] . While these strategies were effective in their respective papers , we could not find software that could work on human Y ampliconic genes . We therefore combine the ideas from these strategies into AmpliCoNE , a tool for estimating copy number in highly-similar multi-copy gene families . The Results section contains an overview of AmpliCoNE , but we provide more details here . To evaluate the accuracy of AmpliCoNE , we ran simulations . There are nine TSPY genes ( six functional + three pseudogenes ) , six RBMY genes and two VCY genes in the hg38 reference . We added different copies of these three ampliconic gene families to the Y chromosome ( S1 Table ) to simulate reads . The total number of gene copies in the three custom references used to generate the simulated reads were 22/7/4 copies ( for TSPY/RBMY/VCY , respectively ) in set 1; 29/12/2 copies in set 2; 23/9/3 copies in set 3 . Using wgsim [0 . 3 . 2] [76] we simulated 666 million paired-end reads of length 101 bp and insert size of 260 bp ( the exact parameters were "-d 260 -N 666873346–1 101–2 101 -S 9 -e 0 -r 0 -R 0" ) . The reads from the three simulated datasets were aligned to the hg38 reference genome using BWA MEM[v0 . 7 . 15] [77] . The SAM files were sorted and PCR duplicates were removed using the PICARD toolkit [v1 . 128] [78] . Finally , samtools [v1 . 3 . 1] [79] were used to index the alignments . The sorted indexed BAM files were presented as input to AmpliCoNE-count . We used mRNA sequencing data for 170 testis samples with matched whole-genome sequencing ( WGS ) data from the GTEx project [25] . The GTEx RNA-seq libraries were generated with the Illumina TruSeq protocol and whole-genome sequencing was performed with paired-end reads ranging from 100 bp to 150 bp in length with target insert size of 350–370 bp [25] . As the validation dataset for AmpliCoNE , we used WGS data from four males ( depth of coverage ranging from ~45-50x in HG002 and HG003 , ~300x in HG005 and ~100x in HG006 ) sequenced by the GIAB consortium [30] . The Y-chromosome-specific alignments of the GTEx dataset were extracted from dbGAP using the SRA toolkit [80] . From the alignments , we extracted the reads and aligned them to the hg38 reference genome using bwa-mem [v0 . 7 . 15] [77] . The SAM files were sorted and PCR duplicates were removed using PICARD toolkit [v1 . 128] [78] . Finally , samtools [v1 . 3 . 1] [79] were used to index the read alignment files . The generated BAM files were presented as input to AmpliCoNE-count to estimate ampliconic copy number . AMpliCoNE-build requires the locations of all the gene copies , in the reference genome , for each ampliconic gene family . While the locations of functional copies are already annotated in hg38 , these do not include highly similar pseudogenized copies . These are necessary to include since they will affect the read mappings . For each family , we therefore took an arbitrary annotated copy of a gene , and used BLAT [81] to find all sites aligning with >99% identity ( S13 Table ) . These locations were given as input to AmpliCoNE-build . In order to validate the in silico ampliconic gene copy number count in four individuals sequenced by the GIAB consortium [30] , we acquired their DNA ( NA24385 , NA24149 , NA24631 , and NA24694 ) from Coriell and performed ddPCR for all nine Y ampliconic gene families . In order to infer the copy number of these gene families we used SRY , a single-copy gene on the Y chromosome , and RPP30 , a two-copy autosomal gene , as references . We ran ddPCR for each sample in triplicates using EvaGreen dsDNA dye ( Bio-Rad ) on the Biorad QX200 digital droplet platform with the protocol and primers from our previous publication [82] . The results were analyzed using QuantaSoft software . Subsequently , after removal of outliers , the concentration ( copies/uL ) of each ampliconic gene family of interest was divided by the concentration of the references , SRY and RPP30 ( S4 Table ) . Gene expression estimates were obtained using the kallisto-DESeq2 pipeline described below . The standard human ( hg38 ) RefSeq transcripts obtained from the UCSC Genome Browser [83] were used as reference . We generated an index for the reference using the kallisto [v0 . 43 . 0] index function with default parameters [84] . For each sample we obtained read counts per transcript using the kallisto quant ( —bias , —seed = 9 , —bootstrap-samples = 100 ) function . The hg38 refFlat file containing the transcript-to-gene mapping information was obtained from the UCSC Genome Browser [83] annotation database , which was used to convert the transcript-level read counts to gene-level expression levels using tximport package [v1 . 2 . 0] [85] . Since there were no replicates for the samples , we set the 170 sample ids as different conditions in the design , and the gene-level read counts for 170 RNA-seq samples were normalized using DESeq2 [v1 . 14 . 1] [86] . Additionally , read counts based on the vst ( Variance Stabilizing Transformation ) function in DESeq2 were used to check for outliers . To identify outliers in the dataset we performed Principal Component Analysis using the prcomp ( ) function on the vst-based normalized read counts . When we plotted the first and second principal components , we found 21 samples outside the main cluster of the remaining 149 samples ( S12 Fig ) . We followed steps described in DEseq2 vignettes and plotted the heatmap of sample-to-sample distance for the top 1 , 000 highly expressed genes to identify outliers visually and we found the same 21 samples as outliers . Thus , we filtered out these 21 samples and utilized the expression values for the nine ampliconic families in the remaining 149 samples in the downstream analysis . We summed the expression values for all the gene copies within a gene family to obtain family-level expression values . Yhaplo [v1 . 0 . 11] [87] was used to predict Y haplogroup of the samples . The version of Yhaplo[1 . 0 . 11] we used expects the SNP coordinates consistent with the hg19 [88] version of the human reference . The Y-chromosome-specific BAM files downloaded from dbGAP were aligned to the hg19 version of the human reference using BWA MEM . We directly converted the downloaded BAM files into pileup format using samtools mpileup function . A custom script was used to convert the pileup file into Yhaplo-compatible input format . We annotated the Y haplotype for all the samples in the dataset using Yhaplo default parameters . Code used in the manuscript is available at github link: https://github . com/makovalab-psu/GTEx_Testis_Analysis . Steps to install and use AmpliCoNE are available at github: https://github . com/makovalab-psu/AmpliCoNE-tool | The human genome harbors two sex chromosomes—X and Y . Among them , the Y chromosome is present only in males . Deletions of portions of this chromosome have been linked to male infertility , however exactly why the loss of these genes leads to this condition is not well understood . Here we study a group of Y chromosome genes called ampliconic genes , which are expressed in testis and are frequently deleted in males with infertility . These genes are organized in nine gene families , each of which harbors multiple copies of genes highly similar in sequence . In this study , we aimed to establish a baseline of their variation in copy number and in gene expression—one measure of genes’ functional output—by studying 149 healthy men . We found that testis tolerates a wide range of copy number and expression variation of Y ampliconic genes . Additionally , we demonstrated that gene expression within most Y ampliconic gene families depends on the expression levels of gene family members located outside of the Y chromosome , i . e . they undergo dosage regulation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"meiosis",
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"cell",
"cycle",
"and",
"cell",
"division",
"population",
"genetics",
"cell",
"processes",
"sociology",
"social",
"sciences",
"reproductive",
"physiology",
"human",
"families",
"population",
"bi... | 2019 | Dosage regulation, and variation in gene expression and copy number of human Y chromosome ampliconic genes |
The development of neutralizing anti-drug-antibodies to the Factor VIII protein-therapeutic is currently the most significant impediment to the effective management of hemophilia A . Common non-synonymous single nucleotide polymorphisms ( ns-SNPs ) in the F8 gene occur as six haplotypes in the human population ( denoted H1 to H6 ) of which H3 and H4 have been associated with an increased risk of developing anti-drug antibodies . There is evidence that CD4+ T-cell response is essential for the development of anti-drug antibodies and such a response requires the presentation of the peptides by the MHC-class-II ( MHC-II ) molecules of the patient . We measured the binding and half-life of peptide-MHC-II complexes using synthetic peptides from regions of the Factor VIII protein where ns-SNPs occur and showed that these wild type peptides form stable complexes with six common MHC-II alleles , representing 46 . 5% of the North American population . Next , we compared the affinities computed by NetMHCIIpan , a neural network-based algorithm for MHC-II peptide binding prediction , to the experimentally measured values and concluded that these are in good agreement ( area under the ROC-curve of 0 . 778 to 0 . 972 for the six MHC-II variants ) . Using a computational binding predictor , we were able to expand our analysis to ( a ) include all wild type peptides spanning each polymorphic position; and ( b ) consider more MHC-II variants , thus allowing for a better estimation of the risk for clinical manifestation of anti-drug antibodies in the entire population ( or a specific sub-population ) . Analysis of these computational data confirmed that peptides which have the wild type sequence at positions where the polymorphisms associated with haplotypes H3 , H4 and H5 occur bind MHC-II proteins significantly more than a negative control . Taken together , the experimental and computational results suggest that wild type peptides from polymorphic regions of FVIII constitute potential T-cell epitopes and thus could explain the increased incidence of anti-drug antibodies in hemophilia A patients with haplotypes H3 and H4 .
The bleeding disorder hemophilia-A ( HA ) is treated with infusions of plasma-derived- or recombinant ( r ) -Factor VIII ( FVIII ) [1] . The development of anti-drug antibodies that inhibit FVIII function , which occurs in approximately 20% of patients overall , is currently the most important impediment to the successful management of this chronic disease [2] . The development of inhibitory antibodies against FVIII is a complex process that involves both product- and patient-related factors [3] , [4] . Genetic data could help understand why some individuals develop inhibitory antibodies , while others do not , even when an essentially identical recombinant FVIII protein-drug is infused in all patients . Our recent work suggests that non-synonymous Single Nucleotide Polymorphisms ( ns-SNPs ) in the F8 gene of some HA patients could be a pharmacogenetic risk-factor for developing inhibitory anti-drug antibodies [5] . At least six variants of the F8 gene ( designated H1 through H6 ) are found in normal individuals who do not suffer from HA , but only two ( H1 and H2 ) match the recombinant FVIII products used clinically . While H1 and H2 were found in all racial groups studied , H3 , H4 and H5 have been found so far only in Americans of black African descent , and H6 was found in a few persons of Chinese descent [6] . In 78 black patients with HA , 24% had an H3 or H4 background haplotype and the prevalence of inhibitory anti-drug antibodies was higher among patients with the H3 or H4 haplotypes than among patients with haplotype H1 or H2 ( odds ratio , 3 . 6; 95% confidence interval , 1 . 1 to 12 . 3; P-value = 0 . 04 ) [5] . These results suggested that a mismatch in the sequences of the endogenous ( albeit non-functional ) FVIII synthesized by the patient and the infused FVIII protein drug could be the underlying cause for the high incidence of inhibitory anti-drug antibodies among black patients . This hypothesis has , however , not been experimentally verified and the underlying mechanism for this clinical observation is poorly understood . Early ( though indirect ) evidence that CD4+ T-cells play a central role in the development of inhibitory antibodies to FVIII came from the observation that inhibitory anti-drug antibodies spontaneously disappeared in conjunction with an HIV-associated decline in CD4+ counts [7] . Subsequently , it was demonstrated that it is possible to prevent anti-FVIII antibodies in hemophiliac mice by blocking co-stimulatory signals [8] indicating that CD4+ cells are critical . A CD4+ T-cell response to an exogenous protein requires that peptides derived from the therapeutic protein are presented by MHC-II alleles . Moreover , as individuals are tolerized to self proteins , it is necessary that these peptides be recognized as foreign . This is consistent with clinical studies which indicate that there is an association between the nature of the mutation in a patient's F8 gene and the frequency with which inhibitory anti-drug antibodies are developed . For example , while <5% of patients with missense mutations develop inhibitory anti-drug antibodies , >60% of patients with the deletion of one or more exons do so [9] . However , even HA patients with missense mutations can have a high risk of developing inhibitory anti-drug antibodies if the mutation occurs at certain locations . For example , inhibitory anti-drug antibodies are frequently associated with the mutation R2150H in HA patients [10] . Saint-Remy and coworkers characterized FVIII-specific CD4+ T-cell clones derived from such a patient [11] and demonstrated that all T-cell clones recognized synthetic peptides encompassing R2150 . Interestingly none of the T-cell clones recognized recombinant peptide with H2150 , demonstrating that T-cell response was directed only at the wild-type sequence . This study is consistent with other studies showing that the anti-FVIII antibodies recognize the wild type ( foreign ) but not the mutant ( self ) protein [12]–[15] . Similarly , a more recent detailed study carried out on two unrelated patients with the R593C mutation and the MHC-II , DRB1*11:01 genotype [16] showed that wild type peptides which contain the missense site , bound to the MHC-II protein with physiologically relevant affinities . Moreover , in this study the peptide-MHC-II binding algorithm ProPred [17] was used to identify binding peptides and synthetic peptides predicted to bind with the highest affinity were tested in a binding assay using recombinant MHC-II DRB1*11:01 . These reports lay the groundwork for the current study because although ns-SNPs alone do not cause HA , the amino-acid change due to an ns-SNP in the endogenous FVIII of patients , are analogous to missense mutations and can potentially result in T-cell epitopes . The current study is designed to test the hypothesis that some polymorphisms in the endogenous F8 gene of HA patients are risk-factors for the development of anti-drug antibodies to recombinant-FVIII used as a drug ( see above and [5] ) . A necessary , though insufficient , condition for development of such inhibitory anti-drug antibodies is that wild type peptides , i . e . those that have the sequence of the FVIII protein therapeutic at the locations where polymorphisms occur , would be T-cell epitopes . Identification of MHC-II epitopes is an important task in basic immunological research with considerable practical value . Peptide-MHC-II prediction algorithms , if accurate enough , can replace time consuming and expensive experimental approaches . Therefore , a large number of computational methods to predict MHC-II epitopes have been developed in recent years ( for review see [18] ) . In addition , so-called “meta” approaches have been devised , combining the results of individual methods to improve the predictive performance [19] , [20] . Due to the large number of overlapping peptides that could be tested as epitopes as well as the diversity of MHC-II variants in the population it is not possible to experimentally test all possible peptide-MHC-II combinations . On the other hand , the predictive performance of peptide-MHC-II binding algorithms is , often , not good enough to justify a purely computational approach [18] . We therefore selected a few wild type peptides from regions of common polymorphisms and experimentally determined their affinities to six common MHC-II variants . These results were transformed into binding promiscuity scores that suggested that at least some polymorphic regions of the FVIII protein could yield T-cell epitopes . Furthermore , we used the complete dataset of experimentally determined “positive” and “negative” binders to re-validate two leading peptide-MHC-II binding algorithms for each of the six MHC variants; using a computational method allowed us to include many more peptide-MHC-II binding estimations per each polymorphism , significantly increasing the power of the statistical analysis . Together , the experimental and computational results support the hypothesis that peptides with the sequence of the therapeutic FVIII protein at locations where polymorphisms occur , constitute potential T-cell epitopes and thus could potentially explain the higher prevalence of inhibitory anti-drug antibodies in African American HA patients which is observed in the clinic .
There are over 2500 reports describing 898 unique missense mutations in HA patients ( for comprehensive data base see http://hadb . org . uk/ ) , of which very few ( 4 . 9% ) developed inhibitory anti-drug antibodies . However , inhibitory anti-drug antibodies are detected in approximately 30–40% of patients with three specific missense mutations: Y205C , R2150H and W2229C . In addition , several wild type peptides from regions of the FVIII protein where these mutations occur have been extensively characterized and shown to be potent and promiscuous T-cell epitopes ( see introduction , Fig . 1A and [10] , [11] ) . In this study , we use the term “wild type peptide ( s ) ” to denote peptides which have the same sequence as the corresponding region on the H1 variant of the FVIII protein . We screened 25 wild type FVIII peptides that included T-cell epitopes from these three regions as well as peptides from regions of the protein where missense mutations are not associated with inhibitory anti-drug antibodies ( Fig . 1A ) . Thus , the former constituted a set of positive-control peptides and the latter negative-control peptides for evaluating peptide-MHC binding and stability as markers for the development of anti-drug antibodies . The binding and stability of each peptide was determined for six MHC-II proteins , viz . DRB1*03:01 , DRB1*04:01 , DRB1*07:01 , DRB1*11:01 , DRB1*15:01 and DRB1*15:03 in a REVEAL Assay . The binding of each peptide to an MHC-II variant was expressed as a “binding score” on a scale of 0–100; the score of each wild type FVIII peptide being relative to a unique positive control peptide , which is a known T-cell epitope for that MHC-II allele . The stability of each MHC-II-peptide complex was expressed as the normalized product of the half-life of the peptide-MHC-II complex and the binding score ( see Materials and methods ) . In addition , we computed , for each peptide , a “binding promiscuity score” defined as the fraction of MHC-II variants that bind ( binding score ≥15 ) or form a stable complex ( stability score >6 ) with that peptide . The median binding promiscuity scores for positive-control wild type peptides are significantly higher than those for negative-control wild type peptides when using either peptide-MHC affinity ( Fig . 1B ) or stability ( Fig . 1C ) as a marker ( one-sided Mann-Whitney-Wilcoxon ( MWW ) P-values = 0 . 018 and 0 . 003 , respectively ) . The endogenous FVIII sequences of patients with missense mutations and ns-SNPs differ from the sequence of an infused FVIII protein-therapeutic only at the locations where the mutations or ns-SNPs occur . As these patients are tolerized to the mutant or polymorphic form of the protein , the wild-type sequence of the infused FVIII might be considered “foreign” by their immune system . Using historical clinical and immunological data of FVIII missense mutations ( see http://hadb . org . uk/ , references therein and [9]–[12] , [15] ) , the analyses shown in Fig . 1 suggest that wild type peptide-MHC binding and stability are both markers for identifying those peptides that are foreign to the patient and could potentially be T-cell epitopes and thus trigger the development of anti-drug antibodies . Consequently , a similar analysis should be predictive of the potential of wild type peptides that correspond to the polymorphic regions of FVIII to elicit anti-drug antibodies . We screened the binding of 30 FVIII-specific wild type peptides to the 6 MHC-II alleles listed above . These 30 peptides ( Table 1 ) include at least 4 overlapping peptides from each location on the FVIII protein where common polymorphisms occur . Binding scores for these peptides ( rows ) to each of the six MHC-II proteins ( columns ) are shown as a heat map in Fig . 2A ( binding scores are also listed in Table S1 ) ; for comparison , the binding scores of positive-control and negative-control peptides characterized in Fig . 1 are shown as well . As ns-SNPs do not always occur individually but as haplotypes , in the heat map , sets of wild type peptides are grouped together as the haplotypes H2 to H6 described previously [5] , [6] . The heat map ( Fig . 2A ) shows that at least one peptide from each of the polymorphic regions of FVIII binds to several MHC-II variants with higher affinities than the negative-control wild type peptides . Moreover many of these peptides exhibit affinities comparable to or higher than the positive-control peptides . Positive binders identified using the same criterion used in Fig . 1B ( binding score ≥15% ) are depicted in a binary plot ( Fig . 2B ) . As expected , only a few binders were identified in the group of negative-control peptides while numerous binders were identified among the positive-control and test-peptides . Moreover , it is clear that even peptides that are potential T-cell epitopes do not bind to all MHC variants . Thus , to estimate the propensity of each peptide to initiate an immune response in a population we calculated the fraction of MHC-II proteins each peptide binds to ( with a binding score ≥15% ) . Fig . 2C shows that several positive-control peptides and wild type peptides from the H3 , H4 and H5 regions of FVIII bind to ≥50% ( and up to 83% ) of the MHC-II alleles used in this study . In contrast , the negative-control wild type peptides bind to ≤33% of MHC-II alleles . In addition to peptide-MHC-II binding per se , the stability of such complexes has been shown to be an important parameter for predicting the eventual immune response [21]–[25] . A heat map depicting the stability index for negative-control , positive-control- and test-peptides ( Fig . 3A ) shows that more stable peptide-MHC-II complexes are generated between MHC-II proteins and the positive-control and test-peptides than between MHC-II proteins and the negative-control peptides . The binary-plot ( Fig . 3B ) shows the peptides that bind with a stability score ≥6 . As expected , the stability score is more stringent than the binding score and thus several peptides form stable complexes with fewer MHC-II proteins ( compare Figs . 2C and 3C ) . The experimental approaches described in Figs . 1–3 utilized six MHC-II variants . Moreover , rather than synthesizing all overlapping 15 mer wild type peptides around the location of each ns-SNP we sampled only 4–5 of the 15 peptides that could be generated . Binding of a complete coverage of all overlapping wild type peptides ( Table 1 ) and a greater diversity of MHC-II variants can be evaluated using a computational approach . Publicly available prediction tools have been previously evaluated [18] , [19] and shown to perform well in some cases but poorly in others . We therefore evaluated the predictive performance of two computational tools , NetMHCIIpan-2 . 1 [26] and the Consensus method [19] using the experimentally determined affinities of 54 FVIII-wild type peptides for the six MHC-II alleles described above . Although this is a much smaller data-set than previous evaluations our purpose was to select a method that best fits our FVIII-based experimental data . Figs . 4 A , B show the prediction performance of the Consensus and NetMHCIIpan-2 . 1 methods in terms of area under ROC curve ( AUC ) . Both methods show good predictive performance ( with a slight advantage to the Consensus method with respect to the AUC ) : AUCs ranged from 0 . 778 to 0 . 972 and 0 . 898 to 0 . 949 for the NetMHCIIpan-2 . 1 and Consensus methods respectively . However , as NetMHCIIpan-2 . 1 predicts binding for a larger set of MHC-II variants , thus covering a larger portion of the population , we chose this method in our analyses . Using the computational tool NetMHCIIpan-2 . 1 we predicted the affinity of all overlapping wild type peptides from polymorphic regions of FVIII ( all peptides in Table 1 ) to 30 MHC-II alleles . Together , these MHC-II alleles occur in at least 90% of the North American , African or World populations . Similarly , the negative-control and positive-control peptides used in this computational analysis included all overlapping wild type peptides in the regions of FVIII depicted in Fig . 1 . Following previous studies , we considered peptides with predicted affinity ≤500 nM as binders [27]; as described above , binding promiscuity score was defined as the fraction of MHC-II variants each peptide binds to . A box and whisker plot ( Fig . 5 ) shows that the median binding promiscuity score for positive-control FVIII peptides is significantly greater than that for negative-control FVIII peptides ( one-sided MWW , P-value<10−6 ) . Interestingly , wild type peptides from the polymorphic regions of FVIII exhibit a range of median binding promiscuity scores . For example the median binding promiscuity score of all H2 and H6 wild type peptides is not significantly different from that of negative-control peptides while wild type peptides from locations where the H3 , H4 and H5 polymorphisms occur have significantly higher scores than negative-control wild type peptides ( one-sided MWW P-values of 0 . 008 , 0 . 004 and <10−4 respectively ) . In the data presented above , we computed an ( un-weighted ) binding promiscuity score to estimate the prevalence of an immunogenic response to a specific peptide within the population ( Figs . 1C & 5 ) . Next , we analyze a binding promiscuity score weighted for the frequency with which the different MHC-II alleles occur in the North American population; Fig . 6A compares the median of these weighted binding promiscuity scores for positive-control- , negative-control- , and test-peptides ( all of which carry the wild type sequence; Fig . S1B shows the corresponding scores per each ns-SNP position ) . Although the specific median binding promiscuity scores for the different groups of peptides vary , overall the pattern remains the same , i . e . wild type peptides from the H2 and H6 regions of FVIII do not have significantly higher median binding promiscuity scores than negative-control peptides while wild type peptides from the H3 , H4 and H5 locations have significantly higher binding promiscuity scores . It is also important to note that polymorphisms in the F8 gene exhibit racial differences . For example , studies have shown that the haplotypes H3 and H4 in FVIII occur very infrequently ( if at all ) in individuals of Caucasian descent while they occur in about a third of individuals of Black African descent . Moreover the distribution of MHC-II alleles is also different between different populations ( Fig . 7 ) . Thus , the distribution of MHC-II alleles in the African population may be more appropriate than the distribution in the North American population as a whole for studying the higher incidence of anti-drug antibodies to FVIII in African-American HA patients than in Caucasian patients . As shown in Fig . 6B , wild type peptides that occur at the location of the H3 , H4 and H5 variants have significantly higher weighted binding promiscuity scores than the negative control peptides . Breakdown for individual ns-SNP positions ( Fig . S1C ) suggests that the drivers for the significantly higher binding promiscuity scores of wild-type haplotypes H3 and H4 are peptides that incorporate M2238 and R484 respectively .
The prevalence of anti-drug antibodies in African American HA patients is twice that observed in white patients of Caucasian descent [5] . Recent clinical studies from one of our laboratories determined that the endogenous FVIII protein is much more polymorphic in black individuals than in white individuals [5] . Moreover , FVIII-protein-therapeutics in the market are more likely to be matched , vis-à-vis the sequence , to the endogenous FVIII of a Caucasian patient than an African American patient . Taking these observations into account we reported that polymorphisms in the endogenous F8 gene are a risk factor for developing inhibitory anti-drug antibodies ( odds ratio , 3 . 6; 95% confidence interval , 1 . 1 to 12 . 3; P-value = 0 . 04 ) . The clinical findings suggest that wild type peptide sequences that correspond to the regions of polymorphisms may constitute T-cell epitopes and trigger T-cell activation . In this study we use experimental and computational methods to evaluate these putative T-cell epitopes as biomarkers for the development of anti-drug antibodies . In the first , experimental part of the study , we measured peptide-MHC-II affinity and stability and use it as a surrogate marker for the development of anti-drug antibodies . To assess our method's capability to predict a clinical manifestation of anti-drug antibodies we used 16 positive-control peptides , i . e . well characterized T-cell epitopes from FVIII associated with the development of inhibitory anti-drug antibodies [12]–[15] and 9 FVIII-derived negative-control peptides . The former have significantly higher median binding promiscuity scores than the latter ( one-sided MWW P-values = 0 . 018 and 0 . 003 for binding promiscuity scores based on peptide-MHC-II affinity and stability , respectively ) ( Fig . 1B , C ) . Assessing the risk of developing anti-drug antibodies is a two-step process . The first step involves considering variables that are anticipated to affect the likelihood of developing anti-drug antibodies and the clinical consequences if such antibodies are generated . The second step is determining the risk associated with the development of anti-drug antibodies as a function of the severity of the clinical consequence and the probability of that occurrence in a patient population . We devised a binding promiscuity score based on these principles . We first determined whether or not a peptide was a “binder” for each MHC-II variant based on a predetermined cut-point ( Figs . 2B & 3B ) . This is a surrogate marker for developing anti-drug antibodies and extensive clinical experience with HA demonstrates that the consequences can be very severe [1]–[3] . Secondly we calculated the binding promiscuity score as the fraction of MHC-II molecules each peptide bound to based on the cut-point ( Figs . 2C & 3C ) . Assuming that the MHC-II variants used in the study are representative of those in the population , this score is a measure of the risk of developing anti-drug antibodies that is associated with that haplotype in the patient population . The caveat is that the predictive power of such a binding promiscuity score is limited by the coverage of the population with respect to the number of MHC-II proteins studied and their frequency of occurrence in the population . For example the 6 MHC-II proteins used in this study occur in 46 . 5% and 40 . 8% of the North American and African populations respectively . Thus , although our experimental studies suggest that several wild type peptides from the H3 , H4 and H5 regions of FVIII are plausible candidates for T-cell epitopes in a patient population , the data set is limited by both the number of foreign peptides and MHC-II variants tested . Cost constrains and the availability of recombinant MHC-II variants limit the generation of complete data sets ( i . e . , all possible overlapping wild type peptides and MHC-II protein combinations ) . It is for reasons such as these that considerable effort has been expended in the last decades to develop computational methods to predict MHC-peptide binding [18] , [19] . In general , computational tools to predict peptide-MHC binding perform better for class-I proteins than for class-II molecules . Recent years , however , have seen a significant improvement in the prediction accuracy of class-II tools [18] , [19] . In particular , methods that combine the results of several tools ( e . g . , the Consensus algorithm; [19] ) as well as pan-specific algorithms , which can predict binding for proteins with no experimental binding data ( e . g . , NetMHCIIpan; [26] ) have been devised . We evaluated the performance of the Consensus [19] and NetMHCIIpan-2 . 1 [26] prediction methods for the 54 FVIII wild type peptides used in this study . Both methods showed good performance for all six MHC-II alleles with AUC values in the range of 0 . 778 to 0 . 972 , which is consistent with previous reports [19] . As the NetMHCIIpan-2 . 1 method provides prediction for a larger set of MHC-II proteins , we used it to estimate the binding of all overlapping 15 mer wild type peptides in the polymorphic , as well as control , regions of FVIII to 30 MHC-II variants . The cumulative frequency of these MHC proteins is 97 . 7% and 90 . 5% in the North American and African populations , respectively . Compared to the negative-control peptides , the median binding promiscuity scores for the positive-control peptides and test-peptides from regions H3 , H4 and H5 are significantly higher ( Fig . 5 ) , consistent with the experimental conclusions that polymorphic regions of FVIII are likely to generate positive T-cell epitopes ( see above and Figs . 2 and 3 ) . Expanding the repertoire of MHC-II variants by using computational methods provides another important advantage for this study: We can better understand what the consequences of a particular polymorphism would be in the general population as well as in specific sub-populations . In our calculations , the binding promiscuity score for each peptide is the fraction of MHC-II alleles that the peptide binds to with a higher affinity than a predetermined cut-point . Thus , if peptides from one region of the FVIII protein have median binding promiscuity scores that are significantly higher than those of another region they are likely to be more immunogenic in the population represented by the MHC variants . Furthermore , as the peptide binding is MHC restricted , the added advantage of expanding the number of MHC-II variants studied is that , the better the coverage of the population vis-à-vis MHC variants the better the risk assessment . It is important to note that the distribution of MHC-II variants is not uniform in the North American population , the different variants occur at frequencies that range from <1% to >10% ( Fig . 7A ) . In addition there are considerable geographical differences in the distribution of the MHC-II variants ( Fig . 7 ) which are likely maintained , to some extent , in US sub-populations originating from those geographical regions . Thus the binding promiscuity scores weighted for the frequencies of the different MHC-II variants in the population ( s ) of interest are more likely to provide more accurate prediction of risk for that population ( Fig . 6 ) . Moreover , such an analysis can also provide some mechanistic insights . For example , in the analysis performed in this study the promiscuity scores weighted for distribution of MHC-II alleles in the North American and African populations are comparable . This observation could suggest the hypothesis that higher prevalence of inhibitory anti-drug antibodies in the African-American sub-population may be mediated through recognition of the MHC-II-peptide complex by the T-cell receptor and not restriction in MHC-II binding . A comprehensive experimental determination of all possible peptide-MHC-II binding affinities required to generate such a weighted binding promiscuity score would be prohibitive in cost and present a considerable technical challenge . Despite using a large repertoire of MHC-II variants , the analysis of peptide-MHC-II interactions in this study was limited to DR molecules . A previous study demonstrated that peptides that contained FVIII residues I2144-T2161 strongly stimulated T-cell clones . However this response was completely abrogated by a monoclonal antibody to MHC-II DR molecules but not by monoclonal antibodies to MHC class II DP and DQ molecules [11] . Nonetheless it has not been conclusively established that the MHC class II restriction is mediated by DR molecules . Thus future studies involving MHC-II DP and DQ molecules would be useful . In conclusion , this study provides proof-of-principle that predictive algorithms for peptide-MHC-II binding can be used in conjunction with limited experimental studies to obtain mechanistic insights that have potential clinical importance . Our analyses suggest that: ( i ) peptide-MHC-II affinity could be used as a predictive marker for clinical manifestations of anti-drug antibodies of FVIII molecules used as protein-therapeutics ( Figs . 1B , C , 5 and 6 ) ; ( ii ) Publicly available peptide-MHC-II prediction methods show good predictive performance with respect to our experimentally determined affinity ( Fig . 4 ) ; ( iii ) Both experimental and computational affinities of peptide-MHC-II binding suggest that wild type peptides from some polymorphic regions of the FVIII proteins can generate strong T-cell epitopes ( Figs . 2 , 3 , 5 and 6 ) ; and ( iv ) The HLA-restriction of these potential T-cell epitopes suggests extensive penetration in the North American population as well as among African Americans ( who constitute the more relevant population , as F8 polymorphisms occur predominantly in this group ) ( Fig . 6 ) .
We used a proprietary high throughput , REVEAL binding assay ( ProImmune , www . proimmune . com ) to estimate the binding of individual test peptides to the following MHC-II variants: DRB1*03:01 , DRB1*04:01 , DRB1*07:01 , DRB1*11:01 , DRB1*15:01 and DRB1*15:03 [28] . Detection of a peptide binding to each MHC-II protein is based on the presence or absence of the native conformation of the MHC-peptide complex . The peptide-MHC-II complex is detected by an increase in a fluorescence signal due to the binding of a specific monoclonal antibody . Each test-peptide is given a score relative to a positive control peptide , which is a known T-cell epitope for that MHC-II molecule . The score is reported quantitatively as a percentage of the signal generated by the test peptide compared with the positive control peptide . Assay performance is evaluated by including in each set of measurements an intermediate control peptide which is known to bind the MHC-II molecule but with a lower affinity than the positive control peptide . Note that the positive control peptide used to determine the binding score is different from the positive control peptides depicted in the figures , which are known T-cell epitopes identified in the FVIII molecule . We measured the half-life of each peptide-MHC-II complex using REVEAL peptide-MHC affinity assays developed by ProImmune Ltd . ( Oxford , UK ) [29] . On-rates of peptides were measured over a 96-hr period at 10°C . A unique positive control peptide was used for each MHC-II allele . Samples of assembling peptide-MHC complexes were taken at defined time points and snap-frozen in liquid nitrogen prior to analysis . Assembly of peptide-MHC complexes was detected using a conformational ELISA involving an anti-HLA antibody . On-rates were calculated by fitting data to a one-phase association curve: To determine the off-rates of the peptides , samples of dissociating complexes were taken at defined time points and frozen in liquid nitrogen prior to analysis . Percent denaturation of peptide-MHC complexes was also detected using a conformational ELISA . Off-rates were calculated by fitting data to a one-phase dissociation curve:On- and off-rate half-life values ( t1/2 ) were calculated from the rate constant ( k ) using the following equation: To obtain a normalized stability score , the half-life values were multiplied by the Binding Score ( see above ) and divided by 100 . Also because the assay cannot obtain accurate half-life values above 120 hours , all values above this time-point were taken to be 120 in the calculation . The primary source of the MHC-II allele frequencies used in determining the weighted binding promiscuity score is the Allele Frequency Net Database ( www . allelefrequencies . net ) . For a given geographical region all the available studies are taken and the specific HLA-allele frequencies of interest are averaged , weighting the analysis on the number of individuals in the study . Care is taken to exclude studies that are based solely on single families or anthropological studies that are based on groups accounting for small proportions of the population . Where possible , data was calculated from large-scale general population sampling studies such as donor banks , as these are assumed to be representative of the population as a whole . The binding promiscuity score of a peptide is computed using its binding score ( measured or predicted ) or the stability score , to a set of HLA-DRB1 alleles , and an associated cut-point . The ( un-weighted ) binding promiscuity score is defined as the fraction of HLA-DRB1 alleles with score greater than the predefined threshold . The weighted binding promiscuity score also considers the frequency of each HLA-DRB1 in the population of interest and is defined as the sum of the frequencies of binding HLA-DRB1s , normalized by the total cumulative frequency of the alleles in the set . In this study , the experimental binding promiscuity score is based on measurements for six HLA-DRB1s , representing 46 . 5% of the North American population , and binding cut-points of >15% for REVEAL binding score and >6 for REVEAL stability score . Predicted binding affinities were computed by NetMHCIIpan-2 . 1 [30] and are given in nM , where peptides with affinities ≤500 nM are considered binders . The findings and conclusions in this article have not been formally disseminated by the Food and Drug Administration and should not be construed to represent any Agency determination or policy . | The development of anti-drug antibodies to therapeutic proteins is a significant impediment to development and licensure of therapeutic proteins and limits their clinical utility . The development of such antibodies requires CD4+ T-cell activation , which is mediated by the recognition of epitopes presented by MHC class-II ( MHC-II ) molecules . Here , we use experimental measurements and computational predictions of peptide-MHC-II affinities to study the clinical observation that African-American hemophilia A patients have a higher incidence of anti-drug antibodies to Factor VIII than Caucasian patients . Specifically , we used the experimental data to select and validate a computational prediction method which , in turn , allowed us to expand our analysis to a larger repertoire of peptide-MHC-II complexes . We showed that wild type peptides spanning haplotype polymorphisms common in the African American population bind MHC-II proteins significantly more than a negative control , thus providing a mechanistic explanation of the phenomenon in this population . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"major",
"histocompatibility",
"complex",
"drug",
"licensing",
"and",
"regulation",
"drug",
"research",
"and",
"development",
"clinical",
"immunology",
"drugs",
"and",
"devices",
"pharmacogenetics",
"adverse",
"reactions",
"immunology",
"biology",
"computation... | 2013 | Polymorphisms in the F8 Gene and MHC-II Variants as Risk Factors for the Development of Inhibitory Anti-Factor VIII Antibodies during the Treatment of Hemophilia A: A Computational Assessment |
Covalently closed circular DNA ( cccDNA ) of hepadnaviruses exists as an episomal minichromosome in the nucleus of infected hepatocyte and serves as the transcriptional template for viral mRNA synthesis . Elimination of cccDNA is the prerequisite for either a therapeutic cure or immunological resolution of HBV infection . Although accumulating evidence suggests that inflammatory cytokines-mediated cure of virally infected hepatocytes does occur and plays an essential role in the resolution of an acute HBV infection , the molecular mechanism by which the cytokines eliminate cccDNA and/or suppress its transcription remains elusive . This is largely due to the lack of convenient cell culture systems supporting efficient HBV infection and cccDNA formation to allow detailed molecular analyses . In this study , we took the advantage of a chicken hepatoma cell line that supports tetracycline-inducible duck hepatitis B virus ( DHBV ) replication and established an experimental condition mimicking the virally infected hepatocytes in which DHBV pregenomic ( pg ) RNA transcription and DNA replication are solely dependent on cccDNA . This cell culture system allowed us to demonstrate that cccDNA transcription required histone deacetylase activity and IFN-α induced a profound and long-lasting suppression of cccDNA transcription , which required protein synthesis and was associated with the reduction of acetylated histone H3 lysine 9 ( H3K9 ) and 27 ( H3K27 ) in cccDNA minichromosomes . Moreover , IFN-α treatment also induced a delayed response that appeared to accelerate the decay of cccDNA . Our studies have thus shed light on the molecular mechanism by which IFN-α noncytolytically controls hepadnavirus infection .
Hepatitis B virus ( HBV ) is the prototype member of the Hepadnaviridae family and contains a relaxed circular ( rc ) partially double stranded DNA ( 3 . 2 kb in length ) genome [1]–[3] . Upon entry into a hepatocyte , the nucleocapsid delivers the genomic rcDNA into the nucleus , where the rcDNA is converted into a covalently closed circular ( ccc ) DNA . The cccDNA exists as an episomal minichromosome and serves as the template for the transcription of viral RNAs [4] , [5] . Hepadnaviruses replicate their genomes via a protein-primed reverse transcription of pre-genomic ( pg ) RNA in the cytoplasmic nucleocapsids , which are subsequently enveloped upon synthesis of rcDNA and secreted out of cells as virions [6] , [7] . During the early phase of infection , additional cccDNA are produced from newly synthesized cytoplasmic rcDNA through an intracellular amplification pathway [8] , [9] . These two pathways culminate in the formation of a regulated steady-state population of 5 to 50 cccDNA molecules per infected hepatocyte [4] , [10] , [11] . Persistent infection of hepadnaviruses relies on the stable maintenance and proper function of a cccDNA pool in the nucleus of an infected hepatocyte as the source of viral RNAs . Not surprisingly , the metabolism and transcriptional activity of cccDNA are subjected to regulation by host pathophysiological cues . For example , although the cccDNA is apparently stable in stationary hepatocytes [12] , the molecules can be non-cytolytically purged from infected hepatocytes during the resolution of an acute HBV infection in vivo , which is most likely due to the antiviral responses induced by gamma interferon ( IFN-γ ) and other inflammatory cytokines [13]–[15] . In support of this notion , alpha-interferon ( IFN-α ) and interleukin-6 have been shown to reduce the amounts of viral RNA transcribed from cccDNA in cultured hepatocytes or HBV-infected uPA-SCID mice in vivo [16] [17] , [18] . Moreover , clinical studies suggest that cccDNA transcription activity is approximately 10-fold lower in HBeAg-negative patients than that in HBeAg-positive patients [19] , [20] , suggesting that host immunopathological factors regulate the activity of cccDNA transcription in vivo . In addition , despite a profound reduction in its total amount , cccDNA becomes the dominant form of HBV DNA upon suppression of viral DNA synthesis by nucleoside analogue therapies [21]–[24] . Elimination of the residual cccDNA and/or silencing of its transcriptional activity are obligatory for a therapeutic cure of HBV infection , which has rarely achieved in patients receiving prolonged treatment of highly active viral DNA polymerase inhibitors [21] , [25]–[27] . On the contrary , pegylated IFN-α ( pegIFN-α ) is effective in achieving sustained virologic response , defined as HBeAg seroconversion and/or hepatitis B virus ( HBV ) DNA levels below 20 , 000 copies/mL at 6 months after completion of the therapy , in only 30% of hepatitis e antigen ( HBeAg ) -positive and 40% of HBeAg-negative cases [28]–[30] . However , the pegIFN-α therapy does promote HBsAg clearance or seroconversion in a small , but significant fraction of treated patients [28] . The antiviral mechanism of IFN-α and the reasons for the differential therapeutic responses among the treated patients remain to be elucidated . Hence , understanding the molecular mechanism of cccDNA metabolism and transcription regulation by inflammatory cytokines should advance our knowledge of viral pathogenesis as well as facilitate the development of antiviral therapeutics to cure chronic hepatitis B . However , due to the lack of convenient cell culture systems for efficient HBV infection [31] , [32] , the transcriptional regulation of hepadnaviruses was investigated primarily with hepatoma cells transfected with HBV reporter plasmids or transgenic mice harboring an integrated linear HBV genome in the host chromosomes [33]–[35] . It is most likely that certain features of cccDNA transcription regulation could not be recapitulated in these surrogate systems [36] , [37] . As an alternative approach , Levrero and colleagues studied HBV transcription in the linear full-length HBV DNA-transfected human hepatoma cells , in which cccDNA were ostensibly formed by circularization of input linear HBV DNA [38] , [39] . However , it is difficult to investigate cccDNA metabolism in such a transient transfection system . In this study , we took advantage of a chicken hepatoma cell line that supports tetracycline-inducible duck hepatitis B virus ( DHBV ) replication and established an experimental condition mimicking the virally infected hepatocytes in which DHBV pgRNA transcription and DNA replication are solely dependent on cccDNA . Unlike the transfected cells that cccDNA is derived from the input linear DNA , our assay allows to study the metabolism and transcription regulation of cccDNA synthesized from its authentic precursor , rcDNA in the cytoplasmic nucleocapsids . Obviously , due to the lack of the x protein in avihepadnaviruses , which is essential for the establishment of woodchuck hepatitis virus infection in vivo and plays an important role in regulation of HBV cccDNA transcription in virally infected human hepatocytes [37] , [39] , [40] , the cccDNA metabolism and transcription regulation of DHBV may differ in molecular details from that of mammalian hepadnaviruses . However , just as studying other replication steps of hepadnaviruses , such as nucleocapsid assembly , priming , reverse transcription and cccDNA formation [31] , [41]–[43] , the principles revealed by studying DHBV cccDNA metabolism and transcription regulation by inflammatory cytokines should provide valuable insight in HBV cccDNA biology and clues for the development of therapeutics to control chronic hepatitis B .
We previously established a chicken hepatoma-derived stable cell line harboring an integrated transgene for transcription of DHBV pgRNA in a tetracycline ( tet ) inducible manner , designated as dstet5 [44] . Upon removal of tet from culture medium , pgRNA was transcribed from the viral transgene integrated in the host cellular chromosome , which led to a sequential occurrence of viral protein translation , nucleocapsid assembly , DNA synthesis and cccDNA formation . To determine the transcriptional activity of the cccDNA , dstet5 cells were cultured in the absence of tet and presence of 2 mM foscarnet ( PFA ) , a reversible DHBV DNA polymerase inhibitor , to arrest viral DNA synthesis [45] . Three days later , tet ( 1 µg/ml ) was added back into culture medium to shut off pgRNA transcription from the transgene . Meanwhile , PFA was removed from the medium to resume the viral DNA synthesis in pgRNA-containing nucleocapsids and subsequent formation of cccDNA . As shown in Fig . 1 , upon removal of PFA and addition of tet , the steady-state levels of DHBV pgRNA transcribed from the viral transgene declined quickly and reached the lowest point on day 2 ( Fig . 1A ) . However , in parallel with the appearance and accumulation of core DNA since day 1 and cccDNA since day 3 ( Figs . 1B and C ) , pgRNA gradually increased and reached the level comparable with that before the addition of tet into culture medium on day 5 ( Fig . 1A ) . The results thus imply that the cccDNA synthesized in the chicken hepatoma cells are transcriptionally active . The observed high transcriptional activity of DHBV cccDNA in dstet5 cells encouraged us to investigate if and how inflammatory cytokines regulate cccDNA transcription and whether or not the cccDNA minichromosome transcription is distinctly regulated and could thus be therapeutically targeted for selective inhibition of the viral gene expression . To achieve these goals , we first sought an appropriate experimental condition to study cccDNA transcription without interference from the integrated transgene . In fact , the tet-off inducible expression system allows us to promptly shut off the transgene transcription by simply adding tet into culture medium . Accordingly , we designed an experimental procedure where dstet5 cells were initially cultured in tet-free media to allow pgRNA transcription , DNA synthesis and cccDNA accumulation . Three days later , tet was added into the cultures to shut off transgene transcription . Meanwhile , lamivudine , an irreversible DHBV DNA polymerase inhibitor , was also added to arrest viral DNA synthesis and continuous cccDNA formation ( Fig . 2 , Treatment Schedule B ) . Because the pre-existing transgene-derived pgRNA were degraded with a half life of approximately 3 h [44] and became undetectable by the Northern blot hybridization assay at 24 h post addition of tet ( Fig . 2A , Treatment Schedule A ) , the pgRNA detected after 24 h of tet addition ought to be primarily transcribed from cccDNA ( Fig . 2A , Schedule B ) . Interestingly , a RNA species running slightly ahead of pgRNA was accumulated in cells treated with lamivudine and was apparently less sensitive to IFN-α treatment than pgRNA ( Fig . 2A ) . Failure to remove the RNA species by micrococcal nuclease treatment of cell lysates suggests the RNA may be encapsidated in the nucleocapsids ( data not shown ) . Hence , a plausible hypothesis is that the synthesis of minus strand DNA is not arrested by lamivudine at the priming stage , but at a site near the 3′ DR1 of pgRNA after the viral DNA polymerase translocation and elongation of minus stranded DNA synthesis . The observed RNA species was most likely the 5′ fragment of a pgRNA cleaved by RNase H . Studies are currently under way to further investigate this hypothesis . Intriguingly , although lamivudine efficiently arrested DHBV DNA synthesis ( Fig . 2B , Schedule A ) , it did not immediately stop cccDNA formation . The amount of cccDNA increased continuously for another three days since the addition of lamivudine , suggesting that inhibition of viral DNA synthesis did not prevent cccDNA formation from the pre-formed rcDNA ( Fig . 2C , Schedule B ) . This is , in fact , consistent with a previous report that viral DNA polymerase activity was not required for cccDNA formation from rcDNA in virion particles [46] . Consequentially , although the transgene transcription was shut off and pre-existing pgRNA derived from the transgene were quickly degraded , the amount of DHBV pgRNA increased for at least three days after addition of tet , further suggesting that DHBV cccDNA molecules produced in the dstet5 cells were highly active in transcription . However , treatment of the cells with 100 U/ml chicken IFN-α , starting at 3 days after addition of tet and lamividine for 24 or 48 h , significantly reduced the amount of pgRNA , but not core DNA or cccDNA ( Fig . 2A–C ) . Additional results presented in Fig . S1 further demonstrated that IFN-α treatment dose-dependently reduced pgRNA and was effective at a dose as low as 1 U/ml . The observed reduction of viral RNA in IFN-α-treated dstet5 cells could be due to either inhibition of cccDNA transcription or accelerated post-transcriptional decay of viral RNA . Because the pgRNA derived from the transgene and cccDNA were identical in sequence , we thus determined whether IFN-α affected the stability of transgene-derived pgRNA . As demonstrated in Fig . 3A , IFN-α treatment of the dstet5 cells cultured in the absence of tet to allow pgRNA transcription from the transgene , but in the presence of 10 µM lamivudine to inhibit viral DNA synthesis and cccDNA formation for three days did not apparently alter the accumulation of pgRNA . In addition , when dstet5 cells were initially cultured in the absence of tet and presence of lamivudine for three days to allow the accumulation of pgRNA and followed by treatment with 100 U/ml IFN-α for 24 h , the cytokine also did not apparently alter the steady-state level of pgRNA . However , as expected , the levels of pgRNA were drastically reduced upon the addition of tet ( Fig . 3B ) . Moreover , IFN-α treatment did not alter the decay kinetics of pgRNA upon shut-off of transgene transcription in dstet5 cells ( Fig . S2 ) . The results thus strongly indicated that IFN-α affected neither the transgene transcription nor the stability of pgRNA , but efficiently inhibited cccDNA transcription . IFNs elicit an antiviral response by binding to their cognate receptors , which trigger a signaling cascade ( JAK-STAT signaling pathway ) leading to the expression of IFN-stimulated genes ( ISGs ) , whose products exhibit antiviral activities [47]–[49] . Interestingly , the HBV genome contains a typical IFN-stimulated response element ( ISRE ) , which was demonstrated recently to be essential for IFN-α inhibition of HBV cccDNA transcription in the unit-length linear HBV genome-transfected HepG2 cells , presumably through direct recruitment of STAT1 and STAT2 to the cccDNA minichromosomes [18] . However , a role of the putative ISRE in regulation of HBV transcription by IFN-α and interferon response factors 1 and 7 could not be revealed by another study [50] . Nevertheless , to investigate whether IFN-α inhibition of DHBV cccDNA transcription is mediated by the recruitment of pre-existing cellular proteins , such as STAT1/2 , to the cccDNA minichromosomes or requires the synthesis of new antiviral protein ( s ) , we tested the effect of cycloheximide ( CHX ) , a protein translation inhibitor , on IFN-induced antiviral response on DHBV cccDNA transcription . As shown in Figs . 4 and S3 , time course studies demonstrated that the reduction of pgRNA became evident as early as 9 h after the cytokine treatment . Interestingly , treatment of the cells with 10 µM of CHX completely abolished the inhibition of IFN-α on cccDNA transcription , but not the transcription of the two IFN-stimulated genes ( ISGs ) , myxovirus resistance 1 ( Mx1 ) and 2′-5′-oligoadenylate synthetase 1 ( OAS1 ) . The results thus imply that IFN-α inhibition of DHBV cccDNA transcription most likely requires induction of one or more cellular antiviral proteins , but the recruitment of STAT proteins and/or other pre-existing cellular proteins to the cccDNA minichromosomes might not be required or sufficient for the antiviral response . Introduction of pegIFN-α into the clinics dramatically improved the antiviral efficacy of IFN-α therapy against chronic hepatitis B , primarily due to the prolonged half life of pegIFN-α over the standard IFN-α [28] . This clinical observation seems to indicate that the IFN-α-induced antiviral response might be short-lived and continuing engagement of the cytokine with its receptor is required for a persistent inhibition of HBV replication . To determine the longevity of IFN-α induced suppression on cccDNA transcription , dstet5 cells , upon cccDNA accumulation and shut-off of transgene expression as described above , were treated with IFN-α for two days and observed for additional 7 days after the cessation of the cytokine treatment . Consistent with the results presented above , two days of IFN-α treatment significantly reduced the amount of pgRNA , but not cccDNA ( Figs . 5A and B ) . As expected , IFN-α induced the expression of Mx1 and OAS1 mRNA , which gradually decreased after the termination of the cytokine treatment ( Figs . 5C and D ) . However , to our surprise , the level of pgRNA in IFN-α-treated cells continued declining after the treatment for at least 7 days ( Fig . 5A ) . Furthermore , IFN-α appeared also to induce a delayed response that accelerated the decline of cccDNA , starting between 2 to 4 days after the treatment ( Fig . 5B ) . In order to further investigate the pleiotropic effects of IFN-α on the DHBV replication cycle , dstet5 cells were initially cultured in the absence of tet for five days to allow the pool of cccDNA to be established . The cells were then cultured in the presence of tet to shut off the transgene transcription and passaged upon confluence for at least three weeks . As in DHBV persistently infected hepatocytes , the viral DNA replication cycle in these cells was only supported by cccDNA . As shown in Figs . 6A to C , although DHBV replication activity gradually declined in the untreated control cells as the cultures approached over-confluence , eight days of IFN-α treatment induced 121 . 5- , 48 . 2- and 7 . 6-fold net reduction of pre-C mRNA , core DNA and cccDNA , compared with the untreated controls , respectively ( Figs . 6D–F ) . Because the CMV-tet promoter of DHBV transgene in the dstet5 cell line was designed to initiate transcription at the authentic initiation site of pgRNA , which precluded pre-C mRNA transcription [44] , we thus preferred the quantification of pre-C mRNA that could only be transcribed from cccDNA . Moreover , analyses of the decay kinetics of viral DNA replication intermediates revealed that a profound reduction of pgRNA or pre-C mRNA was observed within 24 h of IFN-α treatment ( Figs . 6A and D ) . Reduction of single-stranded core DNA and rcDNA also became evident at 24 and 48 h of the treatment , respectively ( Figs . 6B and E ) . In marked contrast , the amounts of cccDNA were slightly increased during the first two days of the treatment and gradually reduced thereafter ( Figs . 6C and F ) . Intriguingly , cccDNA transcription efficiency , expressed as the molar ratio of preC mRNA over cccDNA , remained relatively stable over the eight days in untreated cells , but decreased more than 10 folds in the cells treated with IFN-α ( Fig . 6G ) . The results thus imply that cccDNA transcription is a primary target of IFN-α antiviral response against DHBV in the chicken hepatoma cells . The reduction of core DNA is most likely the combined effect of reduced pgRNA and inhibition of pgRNA encapsidation or accelerated degradation of pgRNA-containing nucleocapsids [51] , [52] . However , the delayed reduction of cccDNA in the IFN-α-treated cells could be due to either the reduced amount of the rcDNA in the cytoplasmic nucleocapsids , the precursor of cccDNA formation [31] , [53] , or accelerated decay of cccDNA itself . To investigate these two possibilities , we compared the decay kinetics of DHBV DNA replication intermediates in cells treated with IFN-α and lamivudine , respectively . As shown in Fig . S4 , inhibition of DHBV replication by lamivudine resulted in the decrease of core DNA and cccDNA , starting on day 2 and day 4 of the treatment , respectively . However , the amount of pgRNA in the lamivudine-treated cells only slightly reduced in comparison with the untreated cells , which could be due to the slower decay of encapsidated pgRNA compared to that of free pgRNA [44] . However , a more profound reduction of both core DNA and cccDNA in IFN-α-treated cells was observed since day 4 of the treatment , suggesting that IFN-α-induced antiviral response did not simply inhibit the cccDNA transcription and DNA replication , but also actively purged nucleocapsids and/or cccDNA . To gain insight into the epigenetic regulation of cccDNA transcription and identify molecular probes helping dissect the mechanism by which IFN-α inhibits cccDNA transcription , we tested a panel of 47 small molecules that are the inhibitors or activators of DNA and histone modification enzymes . Among others , we identified histone deacetylase ( HDAC ) inhibitors that selectively inhibited DHBV cccDNA transcription . As illustrated in Figs . 7A to C , both Trichostatin A ( TSA ) , a broad-spectrum HDAC inhibitor , and apicidin , a class I HDAC-specific inhibitor [54] , dose-dependently reduced the amounts of pgRNA transcribed from cccDNA , but did not alter the amounts of cccDNA . The results thus suggest that DHBV cccDNA transcription may require class I HDAC activity . On the contrary , the two HDAC inhibitors enhanced the transcription of DHBV pgRNA from the viral transgene integrated in the host cellular chromosome , which is driven by a CMV-IE-tet promoter ( Fig . 7D ) . Considering that HDAC inhibitors were shown to activate transcription of retroviral episomal circular DNA and transcription of HBV cccDNA in the unit-length linear HBV genomic DNA transfecetd HepG2 cells [38] , [55] , [56] , our finding is rather surprising . However , the results are consistent with a previous report that n-Butyrate , an HDAC inhibitor , inhibited DHBV replication in primary duck hepatocyte cultures [57] . Although HDAC activity is commonly correlated with transcriptional repression , it was actually essential for the induction of many IFN-stimulated genes ( ISGs ) and establishment of an antiviral state [58] , [59] . In order to further characterize the effect of HDAC inhibitors on cccDNA transcription , we first tested if inhibition of DHBV cccDNA transcription by HDAC inhibitors required new protein synthesis . Consistent with the results presented in Fig . 4 , while the presence of CHX abolished the inhibitory effect of IFN-α on cccDNA transcription , CHX did not compromise the transcriptional suppression of DHBV cccDNA by an HDAC inhibitor , Vorinostat ( suberoylanilide hydroxamic acid , SAHA ) ( Fig . S5 ) . The results thus imply that unlike IFN-α that induces cellular antiviral proteins to suppress cccDNA transcription , HDAC inhibitors may directly target one or multiple pre-existing cellular proteins , such as HDACs , that are required for cccDNA transcription . We next investigated the interaction between IFN-α and HDAC inhibitors on regulation of DHBV cccDNA function . To this end , dstet5 cells were treated with IFN-α and TSA , alone or in combination for one day and observed for additional 4 days after the cessation of the treatment . As shown in Fig . 8A , treatment of the cells with IFN-α for one day significantly reduced the amount of pgRNA and the viral RNA remained at reduced levels for an additional four days after the treatment . As observed in the previous experiments ( Fig . 5B ) , the level of cccDNA did not change after one day of IFN-α treatment , but was reduced since the third day after the treatment ( Fig . 8B ) . On the contrary , although treatment of the cells with TSA for one day reduced the amount of pgRNA , the level of viral RNA rebounded after the cessation of treatment . Interestingly , although TSA treatment significantly delayed the induction of ISGs , such as Mx1 and OAS1 , by IFN-α ( Fig . S6 ) , it did not apparently affect the cytokine-induced reduction of viral RNA and cccDNA . To distinguish whether the prolonged reduction of pgRNA in the IFN-α-treated cells was due to the long-lasting suppression of cccDNA transcription or decrease in the amount of cccDNA , pre-C mRNA and cccDNA were quantified by real-time PCR assays and the transcription activity of cccDNA were determined as the molar ratio of the pre-C mRNA over cccDNA . As shown in Fig . 8F , while cccDNA transcription activity gradually recovered after the cessation of TSA treatment , IFN-α treatment for 24 h induced a prolonged suppression of cccDNA transcription , which was apparently unaffected by the TSA treatment . Taken together , the results presented herein clearly demonstrated that IFN-α and HDAC inhibitors suppressed DHBV cccDNA transcription with different characteristics and thus via distinct mechanisms . Specifically , while HADC inhibitors promptly inhibited cccDNA transcription by directly targeting cellular functions required for cccDNA transcription , IFN-α induced a long-lasting suppression of cccDNA transcription through induction of cellular antiviral protein synthesis . In addition , HDAC inhibitors did not apparently affect the stability of cccDNA , but IFN-α induced an accelerated decay of cccDNA . Epigenetic modification of DNA and histones has been demonstrated to play important roles in gene transcription regulation [60] , [61] . The prolonged suppression of cccDNA transcription by IFN-α evokes a hypothesis that the cytokine may induce certain epigenetic modifications of cccDNA minichromosomes to sustain the inhibition . To test this hypothesis , we first determined whether or not IFN-α treatment altered cccDNA methylation . As depicted in Fig . S7 , sequence analysis of DHBV genome ( GeneBank accession number K01834 . 1 ) identified three GC-rich islands spanning nucleotide 278 to 407 , 1038–1232 and 1559 to 1733 , respectively . Bisulfate sequence analysis of the three GC-rich islands and DHBV core promoter region ( nt 2172 to 2529 ) revealed that cccDNA were unmethylated in the untreated dstet5 cells and that IFN-α treatment for two days did not induce cccDNA methylation . The result thus suggests that DNA methylation does not play an essential role in IFN-α suppression of cccDNA transcription . Previous studies suggest that hyperacetylation of histone H3 is correlated with a high transcriptional activity of HBV cccDNA in the livers of HBV-infected patients [38] . In particular , epigenetic modifications of histone 3 lysine 9 ( H3K9 ) and 27 ( H3K27 ) have been shown to play important roles in regulating the expression of a variety of host and viral genes [62] , [63] . Accordingly , we tested the acetylation status of H3K9 and H3K27 in cccDNA minichromosomes , promoter regions of the DHBV transgene , OAS1 and β-actin genes in the dstet5 cells mock-treated or treated with IFN-α or apicidin . As shown in Fig . 9A , consistent with the high transcriptional activity of cccDNA , the cccDNA-associated H3K9 and H3K27 were hyper-acetylated in mock-treated cells . Interestingly , treatment of the cells with IFN-α significantly reduced the acetylation levels of the cccDNA-associated H3K9 and H3K27 . However , apicidin treatment only significantly reduced the acetylation levels of the cccDNA-associated H3K27 , but not H3K9 . In marked contrast , the basal levels of H3K9 and H3K27 acetylation were much lower in the promoter regions of the DHBV transgene , OAS1 and β-actin genes than that in the cccDNA minichrosmosomes . While apicidin treatment significantly increased the acetylation of H3K9 , but not H3K27 , in the promoter regions of all the three genes , IFN-α treatment only significantly reduced the H3K27 acetylation in the CMV-tet promoter region of DHBV transgene ( Figs . 9B to D ) . In addition , methylation of H3K9 and H3K27 has been demonstrated to be correlated with heterochromatin formation and polycomb repressive complex-mediated gene repression , respectively [64] , [65] . In order to determine if IFN-α induced gene silencing or heterochromatin formation of cccDNA , we tested the abundance of trimethylated H3K9 ( H3K9me3 ) and dimethylated H3K27 ( H3K27me2 ) -associated with cccDNA minichromosomes by ChIP analysis . Consistent with the actively transcriptional status of cccDNA and other three genes under the investigation , only very low levels ( less than 0 . 1% ) of H3K9me3 and H3K27me2 were detected in cccDNA minichromosomes and the promoter regions of HBV transgene , OAS1 and β-actin ( Fig . S8 ) . Interestingly , neither IFN-α nor apicidin treatment increased the abundance of the H3K9me3 and H3K27me2 in cccDNA and other three tested promoters . In summary , the results presented in this section demonstrated that IFN-α suppression of DHBV cccDNA transcription was associated with the reduced acetylation levels of H3K9 and H3K27 in cccDNA minichromosomes . However , the low levels of H3K9me3 and H3K27me2 in cccDNA minichromosomes strongly suggest that the suppression of DHBV cccDNA transcription by either IFN-α or apicidin is most likely not due to polycomb repressive complex-mediated gene repression or heterochromatin formation of cccDNA minichromosomes .
While the prompt inhibition of cccDNA transcription by IFN-α requires antiviral protein synthesis , the prolonged suppression of cccDNA transcription implies that the cytokine may induce a profound alteration of cccDNA minichromosomes , such as formation of a heterochromatin-like structure , to silence the transcription [63] , [68] , [69] . DNA methylation and histone modifications have been shown to play important roles in regulating the expression of a variety of viral genes and the establishment of latency [62] , [63] , [70] . In particular , previous studies suggested that increased CpG methylation of cccDNA was associated with low serum HBV DNA levels and hyperacetylation of histone H3 was correlated with a high transcription activity of HBV cccDNA in the livers of chronic hepatitis B patients [71] , [72] [38] . Our studies reported herein further supported this notion , by demonstrating that IFN-α treatment significantly reduced the acetylation levels of cccDNA-associated H3K9 and H3K27 ( Fig . 9A ) . However , the failure to induce H3K9me3 and H3K27me2 in DHBV cccDNA minichromosomes by IFN-α implies that IFN-α suppression of DHBV cccDNA transcription is most likely not due to polycomb repressive complex-mediated gene repression or heterochromatin formation of cccDNA minichromosomes . Hence , further understanding of the nature , spatial distribution and dynamics of cccDNA epigenetic modifications in response to IFN and other inflammatory cytokines and their relationship with transcription suppression and elimination of the episomes are essential to understand how the host immune system noncytolytically controls HBV infection . cccDNA can only be synthesized from relaxed circular ( rc ) or double-stranded linear ( dsl ) DNA in the incoming or newly formed nucleocapsids in the cytoplasm , but cannot replicate themselves through semiconservative replication in the nuclei [8] , [9] . In principle , elimination of the nuclear cccDNA could take place via either selective destruction of the minichromosomes by cellular nucleases or dilution and unequal partition into daughter cells during cell division [73] , [74] . It is conceivable that IFN and/or other cytokines may activate cellular response to epigenetically modify cccDNA minichromosomes , which marks the episomes for selective decay in stationary cells or prevents the minichromosomes to be re-enclosed into nuclei after mitosis and subsequently degraded by cytoplasmic nucleases . Alternatively , the epigenetic modification could also alter cccDNA partitioning into daughter cells . Considering the fact that the accelerated decline of cccDNA in IFN-α-treated cells could be observed when viral DNA replication was arrested by lamivudine treatment ( Fig . 5B ) and was significantly faster than that occurred in the cells treated with lamivudine ( Fig . S4C ) , we favor the hypothesis that IFN-induced antiviral response actively purges cccDNA . However , the nature of epigenetic modifications associated with the cccDNA decay and the role of cell division in the cytokine-induced cccDNA elimination remain to be determined . HDAC activity is commonly correlated with transcriptional repression and establishment of latent viral infection [54] , [75] . Accordingly , HDAC inhibitors have been employed to activate HIV transcription in the latently infected cells and eradicate latent HIV infection [76] . However , a recent microarray study revealed that treatment of cells with HDAC inhibitors affected the expression of a large fraction of cellular genes and the expression of approximately 50% of the affected genes were inhibited by the treatment [77] . In agreement with the previous reports [58] , [59] , IFN-αinduced expression of ISGs , such as of Mx1 and OAS1 , in the chicken hepatoma cells required HDAC activity ( Fig . S6 ) . In this study , we demonstrated for the first time that DHBV cccDNA transcription could be inhibited by multiple compounds that inhibit HDAC activity . Unlike IFN-α , inhibition of cccDNA transcription by HDAC inhibitors did not require new protein synthesis and cccDNA transcription gradually resumed upon removal of the inhibitors . The results thus suggested that HDAC inhibitors might directly target one or multiple pre-existing cellular proteins essential for cccDNA transcription . Moreover , the observed hyperacetylation of cccDNA associated histone H3 and requirement of HDAC activity for cccDNA transcription suggest that the dynamic acetylation and deacetylation of cccDNA associated histones might be essential for cccDNA transcription . Hence , identification of HDAC ( s ) and histone acetylases ( HATs ) involved in the dynamic processes not only should advance our understanding of cccDNA transcription regulation , but also provide potential therapeutic targets for selective inhibition of cccDNA transcription . Ironically , our observation that HDAC inhibitors inhibited DHBV cccDNA transcription is contradictory with a previous report showing that TSA increased the HBV cccDNA transcription in the unit-length HBV genomic DNA-transfected HepG2 cells [38] . To resolve this discrepancy , we took the advantage of a recent discovery that expression of human sodium taurocholate cotransporting polypeptide ( NTCP ) in HepG2 cells conferred susceptibility of HBV infection and investigated the effects of HDAC inhibitors on HBV gene expression in HBV infected HepG2 cells [78] . As shown in Fig . S9 , treatment of HBV infected cells at 24 h post infection with apicidin dose-dependently reduced the numbers of HepG2/NTCP cells that expressed HBcAg antigen and decreased the amounts of 3 . 5 kb HBV mRNA and secreted HBeAg . Similar results were obtained with TSA treatment ( data not shown ) . Although the mechanism remains to be further investigated , the results do suggest that a cellular function sensitive to the HDAC inhibitors is required for HBV RNA transcription and protein expression . Finally , the requirement of HDAC activity for DHBV cccDNA transcription seems to mechanistically conflict with the observed association of reduced H3K9 and H3K27 acetylation and IFN-α suppression of DHBV cccDNA transcription . Because HDAC inhibitors did not apparently affect the long-lasting suppression of IFN-α on cccDNA transcription and IFN-α-induced decay of DHBV cccDNA ( Fig . 8 ) , it is thus possible that the reduced acetylation of H3H9 and H3K27 in DHBV cccDNA minichromosomes by IFN-αis catalyzed by HDACs that are not sensitive to the conventional HDAC inhibitors , such as sirtuins . Alternatively , it is also possible that the reduced acetylation of H3K9 and H3K27 in DHBV cccDNA minichromosomes by IFN-α is due to the disruption of a dynamic acetylation and deacetylation of histone H3 through preventing the recruitment of histone acetyltransferases ( HATs ) into the minichromosomes . Hence , further investigations on the recruitment of individual HDACs , sirtuins and HATs to cccDNA minichromosomes and a comprehensive analysis of the nature , spatial distribution and dynamics of cccDNA-associated histone acetylation are essential to ultimately clarify the role of histone acetylation in cccDNA metabolism and transcription regulation . In conclusion , we have established a convenient cell culture system harboring abundant and actively transcribing DHBV cccDNA produced from its authentic precursor , the cytoplasmic nucleocapsid DNA . This assay system allows for the investigation of hepadnaviral cccDNA metabolism and transcription regulation without the interference of the large amount of transfected viral DNA . Although there are differences in many aspects of replication cycle between DHBV and HBV , it is conceivable that as investigation of viral DNA synthesis and other replication steps , the principles uncovered in the DHBV model system on cccDNA metabolism and regulation should provide insight on understanding HBV cccDNA biology and clues for the development of therapeutics to control chronic hepatitis B .
Dstet5 is a chicken hepatoma cell ( LMH ) -derived stable cell line supporting the replication of an envelope protein-deficient DHBV genome in a tetracycline dependent manner [44] . Dstet5 cells were maintained in DMEM/F12 medium ( Mediatech ) supplemented with 10% fetal bovine serum , 100 U/ml penicillin , 100 µg/ml streptomycin , 1 µg/ml tetracycline and 200 µg/ml G-418 . Chicken IFN-α was produced and titrated as described previously [44] . HepG2-derived cell line expressing Sodium taurocholate cotransporting polypeptide ( NTCP ) ( HepG2/NTCP ) was established and maintained as described previously [78] . Trichostatin A , apicidin and Vorinostat were purchased from ENZO Life Sciences . Antibodies against histone H3-pan ( Cat . No . 07-690 ) , Acetyl-histone H3 ( Lys27 ) ( Cat . No . 07-360 ) , Acetyl histone H3 ( Lys9 ) ( Cat . No . ABE18 ) , trimethyl-Histone H3 ( lys 9 ) ( Cat . No . 07-442 ) and dimethyl-Histone H3 ( lys 27 ) ( Cat . No . 07-421 ) were purchased from Millipore . Normal Rabbit IgG ( Cat . No . 2729 ) was purchased from Cell Signaling , Inc . Intracellular DHBV core DNA was extracted as described previously [44] , [53] . One half of the DNA sample from each well of 12-well plates was resolved by electrophoresis into a 1 . 5% agarose gel and transferred onto Hybond-XL membrane . Extraction of protein-free DHBV DNA was carried out by using a modified Hirt extraction procedure [79] , [80] . Briefly , cells from one 35 mm diameter dish were lysed in 3 ml of 10 mM Tris-HCl ( pH 7 . 5 ) , 10 mM EDTA and 0 . 7% SDS . After 5 minutes incubation at room temperature , the lysate was mixed with 1 ml of 2 . 5M KCl and incubated at room temperature for 30 min and followed by centrifugation at 10 , 000 g for 15 min at 4°C . The supernatants were extracted twice with phenol , and once with phenol∶chloroform . DNA was precipitated with two volumes of ethanol overnight at room temperature and dissolved in TE buffer ( 10 mM Tris-HCl , pH 8 . 0 , 1 mM EDTA ) . One half of the protein-free DNA sample from each well of 12-well plates was then resolved in a 1 . 5% agarose gel and transferred onto Hybond-XL membrane . For viral RNA analysis , total cellular RNA was extracted with TRIzol reagents ( Life Technologies ) . Five micrograms of total RNA was resolved in 1 . 5% agarose gel containing 2 . 2 M formadelhyde and transferred onto Hybond-XL membrane in 20× SSC buffer and blotted onto Hybond-XL membrane . For the detection of viral DNA and RNA , membranes were probed with either an α-32P-UTP ( 800 Ci/mmol , Perkin Elmer ) labeled minus or plus strand specific full-length DHBV riboprobe . Hybridization was carried out in 5 ml EKONO hybridization buffer ( G-Biosciences , St . Louis , MO ) with 1 hour pre-hybridization at 65°C and overnight hybridization at 65°C , followed by a 1 hour wash with 0 . 1X SSC and 0 . 1% SDS at 68°C . The membrane was exposed to a phosphoimager screen and hybridization signals were scanned and quantified with QuantityOne software ( Bio-Rad , Hercules , CA ) . Expression of chicken IFN-stimulated genes , Mx1 and OAS1 , and a house-keeping gene β-actin was quantified by semi-quantitative RT-PCR . Total cellular RNA was extracted with TRIzol and cDNA was synthesized with oligo- ( dT ) 12–18 primer and Superscript III DNA polymerase ( Invitrogen ) by following the manufacturer's direction . The PCRs were carried out in a 25-µl reaction mixture with the Advantage cDNA PCR kit ( Clontech ) . The PCR annealing temperatures selected varied depending on the primers selected for amplification . The prime sequences are listed in Table S1 . Mx1 , OAS1 , ACTB and DHBV preC mRNA were also quantitatively detected with SuperScript III Platinum One-Step qRT-PCR Kit ( Invitrogen ) . Forty nanograms of total cellular RNA was reverse transcribed and amplified in a 20-µl reaction mixture per reaction . One step qPCR was carried out as follows . cDNA was synthesized at 60°C for 10 minutes and denatured at 95°C for 2 minutes , followed immediately by 40 cycles of amplification: 95°C , 15 seconds; 59°C , 35 seconds . For preC mRNA measurement , because its amplicon is larger than 250 bp , the maximum the protocol recommended , its amplification condition was adjusted as the follows: 40 cycles of amplification: 95°C , 15 seconds; 60°C , 65 seconds . DHBV core DNA and cccDNA is measured by LightCycler 480 SYBR Green I Master PCR kit ( Invitrogen ) . The PCR was carried out as follows: denatured at 95°C for 5 minutes , followed by 40 cycles of amplification: 95°C , 15 seconds; 60°C , 30 seconds . The amounts of DHBV core DNA and cccDNA were calculated based on a standard curve with known amounts of DHBV DNA under the same amplification conditions . Dstet5 cells were cultured in the absence of tet for three days and in the presence of 1 µg/ml tet and 10 µM lamivudine for additional 4 days . The cells were then mock-treated or treated with 100 U/ml IFN-αfor 48 h . DHBV cccDNA was extracted with an alkali denaturing method described before [44] and modified using the EZ DNA Methylation-Gold Kit ( ZymoResearch , Orange , CA ) following the manufacturer's protocol . The sequence of each sample was determined using Chromas Lite 2 . 33 ( Technelysium Pty Ltd ) . ChIP assay was performed with an EpiTect Chip One-Day Kit ( Qiagen ) by following the procedures provided by the manufacturer with slight modifications . Briefly , Dstet5 cells were cultured in the absence of tet for three days and in the presence of 1 µg/ml tet and 10 µM lamivudine for additional 4 days . The cells were then mock-treated or treated with 100 U/ml IFN-αfor 48 h . Cells were harvested by trypsinization and pelleted by centrifugation at 800 g for 10 min . The cells were suspended in PBS at a density of 3×106 cells/ml and fixed in 1% formaldehyde at room temperature for 10 minutes . The fixed cells were pelleted at 800 g for 10 minutes at 4°C and resuspended by addition of immunoprecipitation lysis buffer supplemented with proteinase inhibitor cocktail at a density of 1×107 cells/ml . Five hundred microliters of the cell lysates were sonicated by cup horn ( Sonicator XL2020 , Misonix ) at a setting of 10W for 2 minutes on and 2 minutes off . The sonication was repeated once at the same setting . This sonication condition has been showed steadily breaking cellular DNA into 500–800 bp fragments . For pre-clear , immunoprecipitation and DNA extraction , we strictly followed the instruction provided in the EpiTect ChIP One-Day Kit ( Qiagen ) . The obtained DNA was subjected to quantitative analysis by real-time PCR with primers specified in the Table S1 . HepG2/NTCP cells in 48-well plate were incubated with HBV virions at a multiplicity of infection of 100 for 24 h . Subsequently , the cells were washed three times with culture medium and incubated with the indicated concentrations of apicidin for 8 h ( Figure S9 ) . The culture medium was then replaced with fresh medium and changed every other day . Cells were stained on day 7 post infection to detect HBV core antigen ( HBcAg ) by an indirect immunofluorescent assay [78] . Secreted HBeAg was measured by ELISA [78] . Intracellular HBV 3 . 5 kb RNA was quantified with a quantitative RT-PCR [78] . | Hepatitis B virus ( HBV ) infection affects approximately one-third of the world population and more than 350 million people are chronically infected by the virus , for which the currently available antiviral therapies fail to provide a cure . This is because the HBV DNA polymerase inhibitors have no direct effect on the nuclear form of HBV genome , the covalently closed circular ( ccc ) DNA . Elimination or transcriptional silencing of cccDNA is the prerequisite for either a therapeutic cure or immunological resolution of HBV infection . However , due to the lack of proper experimental systems , the molecular mechanism of cccDNA biosynthesis , maintenance and transcription regulation remains to be elucidated . We report herein the establishment of a cell-based assay where the replication of duck hepatitis B virus ( DHBV ) , a close relative of HBV , is supported by cccDNA . This experimental system not only allows us to demonstrate the unique property of alpha-interferon suppression of cccDNA transcription , but also shows for the first time that DHBV cccDNA transcription requires histone deacetylase activity . It is conceivable that the principles revealed by studying DHBV cccDNA metabolism and transcription regulation should provide valuable insight in HBV cccDNA biology and clues for the development of therapeutics to control chronic hepatitis B . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biology"
] | 2013 | Alpha-Interferon Suppresses Hepadnavirus Transcription by Altering Epigenetic Modification of cccDNA Minichromosomes |
Coevolution of viruses and their hosts represents a dynamic molecular battle between the immune system and viral factors that mediate immune evasion . After the abandonment of smallpox vaccination , cowpox virus infections are an emerging zoonotic health threat , especially for immunocompromised patients . Here we delineate the mechanistic basis of how cowpox viral CPXV012 interferes with MHC class I antigen processing . This type II membrane protein inhibits the coreTAP complex at the step after peptide binding and peptide-induced conformational change , in blocking ATP binding and hydrolysis . Distinct from other immune evasion mechanisms , TAP inhibition is mediated by a short ER-lumenal fragment of CPXV012 , which results from a frameshift in the cowpox virus genome . Tethered to the ER membrane , this fragment mimics a high ER-lumenal peptide concentration , thus provoking a trans-inhibition of antigen translocation as supply for MHC I loading . These findings illuminate the evolution of viral immune modulators and the basis of a fine-balanced regulation of antigen processing .
Coexistence of pathogens and their hosts represents a masterpiece of evolution , which relies on a fine-tuned balance between pathogen replication and clearance of pathogens by the host immune system [1] . To escape immune surveillance , viruses have developed sophisticated strategies [2] , [3] . For example , Herpes simplex viruses and varicella-zoster virus establish latency in trigeminal and dorsal root ganglia , which express only low levels of major histocompatibility complex class I ( MHC I ) molecules [2] , [4] , [5] . Exploitation refers e . g . to memory T-cells that circulate through the body and hence provide excellent vehicles for virus dissemination during primary simian varicella virus infection [6] . Sabotage is mediated e . g . by a C-type lectin-like gene product of cytomegalovirus that functions as a decoy ligand to subvert missing-self recognition by natural killer cells ( NK ) , thereby circumventing the elimination of the virus infected cell [5] , [7] . Recognizing virus-specific epitopes displayed on MHC I at the cell surface is the essential step in priming and execution of an adaptive immune response against infection . These antigenic peptide epitopes are derived from degradation of the cellular proteome , including virus or tumor associated gene products , via the ubiquitin-proteasomal pathway . The generated peptides are translocated into the ER lumen by the transporter associated with antigen processing ( TAP ) and subsequently loaded onto MHC I molecules [5] , [8] . This heterodimeric ATP-binding cassette ( ABC ) transport complex is composed of two transmembrane domains ( TMDs ) and two cytosolic nucleotide-binding domains ( NBDs ) , which couple the chemical energy of ATP binding and hydrolysis to the peptide translocation across the ER membrane [9] . TAP is the central component of the peptide-loading complex ( PLC ) , composed of TAP1/2 , tapasin ( Tsn ) , ERp57 , calreticulin , and MHC I . After guiding antigenic peptides to MHC I molecules [5] , [10] , peptide-loaded MHC I complexes dissociate from the PLC and traffic via the secretory pathway to the cell surface , where their antigenic cargo is inspected by cytotoxic T lymphocytes ( CTLs ) . TAP can be dissected into the coreTAP complex , which has been shown to be essential and sufficient for peptide translocation , and extra N-terminal transmembrane domains ( TMD0 ) , which bind Tsn and are essential for the assembly of the PLC [11]–[14] . CPXV012 encoded by cowpox viruses ( CPXV ) has been identified to inhibit antigen processing , extending the types of viruses beyond the Herpesviridae family that interfere with the delivery of antigenic peptides into the ER lumen [15] , [16] . CPXV012 encodes an ER-resident type II membrane protein [15] of 69 amino acids , harboring a signal anchor sequence and a short C-terminal region in the ER lumen . CPXV012 associates with the PLC and prevents peptide loading onto MHC I molecules by inhibiting peptide translocation into the ER lumen [15] , [16] . Notably , CPXV012 prevents CD8+ T-cell effector responses by inhibiting peptide translocation into the ER [15] , [16] . CPXV belongs to Orthopoxvirus ( OPV ) genus of the Poxviridae family that also includes clinically relevant pathogens such as the variola virus , which causes smallpox . In contrast to herpesviruses , poxviruses are acute viruses that use a “kiss-and-run” strategy of propagation among host cells . CPXV can infect and replicate in the cells of many different mammalian species , including humans . In Europe and parts of Asia , endemic CPXV is the most common cause of human OPV infections [17] . The increased number of CPXV infections indicates that the abandonment of the smallpox vaccination in 1977 may render the population more vulnerable to CPXV . This emerging zoonotic hazard for humans raises public health concerns . CPXV infections of healthy humans are generally self-limiting and cause only localized skin lesions [18] . However , severe CPXV infections with lethal outcome have been reported in immune compromised and eczematous patients , particularly children [19] , [20] . Given the medical relevance , it is of particular interest to understand the molecular basis of how CPXV mediates immune evasion . Here we identified the coreTAP complex as the direct target for CPXV012 , and revealed how antigen translocation is mechanistically blocked . We further provide an explanation on how this poxviral factor has evolved its evasion potential to escape surveillance by the adaptive immune system .
We first examined whether CPXV012 variants affect the antigen presentation monitored by MHC I surface expression . As shown by flow cytometry , both Flag-tagged flagCPXV012 and C8-tagged C8CPXV012 caused a down-regulation of MHC I surface expression in HeLa cells ( Fig . 1A and S1 Figure ) . The expression of other known immune evasins , such as ICP47 of herpes simplex virus type 1 ( HSV-1 ) that binds to the cytosolic face of the TAP complex and blocks peptide binding [21]–[23] , the human cytomegalovirus ( HCMV ) type I membrane glycoprotein US6 that inhibits ATP binding by TAP via its ER-lumenal domain [24] , [25] , and the tail-anchored Epstein-Barr virus ( EBV ) protein BNLF2a that prevents peptide and ATP binding to TAP [26]–[29] had an even stronger effect on MHC I processing ( Fig . 1B and S1 Figure ) . Other than herpesviridae that are known to express highly potent TAP inhibitors , CPXV might rely on a synergistic cooperation between CPXV012 and CPXV203 , which blocks MHC I trafficking [30] , to efficiently down-regulate MHC I expression at the cell surface . CPXV012 has been reported to associate with the murine PLC without disrupting its overall assembly [16]; however , the direct interaction partner within this macromolecular complex has not been identified so far . We therefore expressed flagCPXV012 or C8CPXV012 in insect cells , which provide the key advantage that any combination of PLC subunits can be expressed in a cellular background that lacks all components of the PLC [11] . As shown by coimmunoprecipitation using TAP1 , TAP2 , or flag specific antibodies , CPXV012 associated directly with the TAP complex ( Fig . 1C ) . Similar results were obtained for C8-tagged C8CPXV012 ( see below ) . The interaction with TAP is specific , because CPXV012 did not coprecipitate with TAPL ( Fig . 1D ) , a lysosomal peptide translocation complex , which shares approx . 40% amino acid sequence identity with TAP1 or TAP2 [31] . Recent studies proposed that CPXV012 interacts with Tsn , thereby impairing Tsn-mediated peptide binding to MHC I [15] , [16] . Notably , Tsn was coprecipitated with CPXV012 only in the presence of the TAP complex . These results suggest that CPXV012 does not directly interact with Tsn ( Fig . 1E ) , but is bound to the PLC via its direct interaction with TAP ( Fig . 1F ) . Moreover , the presence of CPXV012 does not block the TAP-Tsn interaction . CPXV012 inhibits the TAP-mediated peptide translocation into the ER lumen in the absence of other components of the PLC ( Fig . 2A ) . As reported previously , a more pronounced inhibitory effect was observed for the herpesviral inhibitors UL49 . 5 [16] and EBV-BNLF2a . In contrast to BNLF2a , which blocks both , peptide binding and translocation [27]–[29] , [32] , CPXV012 did not interfere with peptide binding to TAP ( Fig . 2B ) . Upon peptide binding , rearrangements within the TMDs of TAP induce a conformational change of the NBDs , which can be monitored by cross-linking using the amine-specific homobifunctional reagent EGS [24] , [33] . Hence , we examined the influence of CPXV012 on the peptide-induced conformational change by EGS cross-linking . In the presence of peptides , cross-linking of the TAP1/2 heterodimer was detected at ∼160 kDa ( Fig . 2C ) . In the presence of EBV-BNLF2a , the peptide-induced cross-linking of the TAP heterodimer is abolished [29] . In contrast , CPXV012 had no effect on the peptide-induced conformational change monitored by cross-linking of TAP1 and TAP2 . As the coreTAP complex is essential and sufficient for peptide binding and translocation [11] , we also examined whether CPXV012 can block this minimal functional unit of antigen translocation . Interestingly , CPXV012 inhibited peptide translocation via coreTAP in a similar way as the full TAP complex ( S2A Figure ) . Consistent with the data obtained for full-length TAP , CPXV012 did not interfere with peptide binding to coreTAP ( S2B Figure ) . As demonstrated by coimmunoprecipitation using TAP1 and TAP2 specific antibodies , CPXV012 binds specifically to coreTAP ( Fig . 2D ) . Taken together , our data demonstrate that CPXV012 interferes with antigen processing by blocking the coreTAP complex at a state after peptide binding and a peptide-induced conformational change . To shed further light on the inhibition mechanism of CPXV012 , we examined whether the TAP1/2•CPXV012 complex can bind ATP . TAP1/2 or TAP1/2•C8CPXV012 complexes were affinity purified and subsequently photo cross-linked with 8-azido-ATP[γ]biotin . Cross-linked proteins were analyzed by immunoblotting using anti-biotin extravidin-HRP . A single biotinylated band at 70-kDa was observed in the absence of C8CPXV012 , consistent with azido-ATP photo labeling of TAP1 or TAP2 ( Fig . 3A ) . The photo labeling of TAP was specific , because it was blocked by an access of unlabeled ATP . Strikingly , ATP cross-linking was inhibited at the TAP1/2•CPXV012 complex . These results demonstrate that CPXV012 inhibits ATP binding or the rate of nucleotide exchange by TAP . Notably , the TAP complex harbors a canonical ATP-binding site II and a non-canonical ATP-binding site I . However , the functional role of these non-equivalent sites is not completely understood [9] . Hence , we investigated whether CPXV012 affects ATP binding to both sites . To distinguish between TAP1 ( 71 kDa ) and TAP2 ( 72 kDa ) in SDS-PAGE , the N-terminus of TAP2 was fused to tapasin ( Tsn ) via a flexible linker of 34 amino acids ( TsnTAP2 ) . It is worth mentioning that the TAP1/TsnTAP2 transport complex is fully functional in peptide translocation and MHC I peptide loading [14] . Moreover , expression of flagCPXV012 inhibited the peptide transport of TsnTAP2/TAP1 ( S3 Figure ) . Subsequently , we examined the effect of CPXV012 on 8-azido-ATP photo cross-linking of the TAP1/TsnTAP2 complex . In the absence of CPXV012 , TAP1 and TAP2 were specifically photo-labeled by 8-azido-ATP ( Fig . 3B ) . In contrast , the TAP•flagCPXV012 complex showed a strongly reduced ATP photo labeling of TAP1 and TAP2 . Taken together , these results suggest that CPXV012 inhibits ATP binding or the rate of nucleotide exchange of TAP1 and TAP2 . We next investigated whether CPXV012 might recognize a similar TAP conformation as the immune evasin BNLF2a or US6 , which target specific , yet different conformations of the antigen translocation complex [29] . After coexpression of TAP1 , TAP2 , flagCPXV012 , and either US6myc ( Fig . 4A and B ) or BNLF2aC8 ( Fig . 4A and C ) , coimmunoprecipitations were performed using antibodies against flagCPXV012 ( anti-flag ) , US6myc ( anti-myc ) , or BNLF2aC8 ( anti-C8 ) , respectively . Strikingly , expression of either US6 or BNLF2a prohibits binding of CPXV012 to TAP , while TAP1/2 was affinity purified with either US6myc or BNLF2aC8 . Moreover , if flagCPXV012 was immunoprecipitated , almost no TAP was found to be associated with the cowpox viral factor , while TAP1/2•US6 or TAP1/2•BNLF2a complexes were detected . On the other hand , the formation of TAP1/2•CPXV012 complexes was not impaired in the absence of the viral TAP inhibitors ( S4 Figure ) . Hence , US6 and BNLF2a prevent the formation of TAP1/2•CPXV012 complexes . A phylogenetic comparison revealed a substantial genotypic and phenotypic diversity among the twelve CPXV strains sequenced so far [34] . CPXV012 used in this study is derived from the reference strain “Brighton Red” ( BR , Supplemental Table 1 ) . Notably , all CPXV012 orthologs share 45 N-terminal residues , which include the cytosolic region of residue 1–12 and the transmembrane helix ( aa 13–35 ) of the type II membrane protein ( Fig . 5A ) . In contrast , the C-terminal regions are divergent and the CPXX012 orthologs cluster in two main groups . The first group , represented by strain GRI90 or GER91 , contains proteins harboring an ER-lumenal domain of 50 to 100 residues , which encodes a C-type lectin-like domain ( CTLD ) . These proteins do not suppress MHC I antigen processing [15] . In contrast , the ER-lumenal domains of CPXV012 belonging to the “Brighton Red” subgroup are significantly shorter and lack similarity to any known domain . Why are the C termini of both subgroups distinct to each other ? Surprisingly , if analyzed on the DNA level , all CPXV012 ORFs turned out to be very similar ( S5 Figure ) . The ORFs of CPXV012 are localized in the terminal genomic region , where many non-essential genes are clustered that encode for proteins affecting virulence or interaction with the host immune system [35] , [36] . Unlike the central region of the Orthopoxvirus genome , which is conserved and encodes proteins for virus replication , the terminal genomic regions are diverse and are more likely prone to DNA rearrangements or mutations . Indeed , the “Brighton Red” CPXV012 ORF differs by a deletion of five A-T pairs , thus causing a frameshift . The alternative reading frame starts right after the TMD and ends with a stop codon after 25 residues . Since these last C-terminal residues are unique , we hypothesize that this region may be functionally important . Hence , we generated CPXV012 variants lacking five or six amino acids at their C or N terminus ( C8CPXV012-CΔ5 or C8NΔ6-CPXV012 , respectively ) . C8NΔ6-CPXV012 suppressed the MHC I surface expression of HeLa cells to the same extent as full-length C8CPXV012 , whereas C8CPXV012-CΔ5 was inactive ( Fig . 5B and C ) , even though almost equivalent amounts of each CPXV012 variant were expressed ( Fig . 5D ) . This lack of function was not due to an impaired interaction with TAP , because C8CPXV012-CΔ5 can still interact with heterodimeric TAP1/2 complexes ( Fig . 5E ) . Hence , the transmembrane region of CPXV012 is likely responsible for the interaction with TAP . Interestingly , the CPXV012 variant D10L , which contains a C-type lectin domain at its C-terminus and can therefore not inhibit TAP [15] , interacts with the transporter ( S6 Figure ) . This is due to the fact that both proteins share a high conserved N-terminal region including the transmembrane domain . In conclusion , the transmembrane domain mediates binding of the viral inhibitor to the TAP complex , whereas the last five C-terminal residues of CPXV012 are crucial for TAP inhibition . Because this active C-terminal domain of CPXV012 is located within the ER lumen , these results further imply that CPXV012 inhibits ATP binding by TAP via an allosteric crosstalk across the ER membrane . Interestingly , the C terminus of CPXV012 exhibits all hallmarks of a high-affinity TAP substrate [37]: Two positive residues ( Arg59 and Arg60 ) , which correspond to positions one and two of a 11mer substrate peptide that ends with the hydrophobic Ile69 ( S7 Figure ) . Strikingly , replacement of Arg59 and Arg60 by alanine ( C8CPXV012RR59AA ) diminished the ability of the viral inhibitor to suppress MHC I surface expression , because the peptide recognition motif of TAP within the C terminus of CPXV012 was abrogated ( S7 Figure ) . Since ATP hydrolysis of TAP is tightly coupled to substrate binding and translocation [38] , we assessed how the frameshifted C-terminal region of CPXV012 affects the peptide-stimulated ATPase activity of purified coreTAP complexes . Peptides representing the last 5 , 10 , 15 , 20 or 25 residues of CPXV012 were synthesized with an acetylated N terminus ( Fig . 6A ) . This modification is known to dramatically decrease the affinity for TAP avoiding that these CPXV012 fragments act as “classical” TAP substrates [37] , [39] . The 25 , 20 , 15 , and 10mer fragments significantly inhibited the peptide-stimulated ATP hydrolysis ( Fig . 6B ) . The IC50 value for TAP inhibition by the 10mer CPXV012 fragment was 72±20 µM ( Fig . 6C ) . This suggests that the C-terminal region of CPXV012 mimics a TAP substrate just before its release into the ER lumen , which induces the functional arrest of TAP by inhibiting ATP binding to the transport complex . The direct proof of a novel trans-inhibition mechanism of CPXV012 was obtained by functional reconstitution of purified components into liposomes . After reconstitution , approximately half of the TAP complex face their NBD to the outside [40] and hence were exclusively energized by external ATP , thereby assuring uni-directional translocation of fluorescently labeled peptides into the lumen of the proteoliposomes [41] . Peptide transport was inhibited by competition with excess of unlabeled peptides as expected , but not by a 100-fold excess of the C-terminal CPXV012 fragment ( Fig . 6D ) . This demonstrates that the active domain of CPXV012 is unable to block TAP from the cytosol . In contrast , encapsulation of the 10mer CPXV012 fragment into the proteoliposomes abolished peptide translocation by the TAP complex almost completely . Thus , the C-terminal fragment of CPXV012 blocks antigen translocation in trans . Since the tethered C-terminal peptide imitates a high-affinity peptide substrate , it is tempting to speculate that CPXV012 mechanistically exploits a novel trans-inhibition mechanism of the antigen processing machinery .
Virus spread is accomplished through subversion of antiviral host immunity by an arsenal of factors that target key molecular steps necessary to mount an immune response [42] . In particular , gene products encoded by Herpesviridae are known to attack multiple steps in MHC I antigen presentation , including the generation of peptides by the ubiquitin-proteasome pathway , translocation of peptides into the ER by TAP , and the mistargeting of MHC I [3] , [43] . The poxvirus gene product CPXV012 was classified as an inhibitor of antigen delivery to MHC I [15] , [16] . However , the direct target , the inhibition mechanism , and the evolutionary origin remained illusive . By expressing CPXV012 in mammalian and insect cells , we establish that CPXV012 directly inhibits the coreTAP complex . Notably , CPXV012 interferes with ATP binding to TAP1 and TAP2 and hence peptide induced ATP hydrolysis via an allosteric crosstalk across the ER membrane . This is mediated by the C-terminal region of CPXV012 located in the ER lumen , at the opposite membrane side of the NBDs ( Fig . 6E ) . Similar to US6 encoded by HCMV , CPXV012 does not influence peptide binding , but inhibits ATP binding to the TAP complex . However , besides using a related strategy , our results demonstrate that the inhibition mechanism of CPXV012 is unique and distinct from US6: US6 inhibits ATP binding to TAP1 [24] , whereas CPXV012 blocks ATP binding or nucleotide exchange of both TAP subunits . In addition , CPXV012 acts as a small peptide , whereas the active region of US6 consists of a folded ER-lumenal domain of 125 residues . In conclusion , CPXV012 and US6 prevent ATP binding to TAP via two distinct mechanisms , which presents a striking example of convergent evolution . Since BNLF2a and US6 , which target distinct conformation states of TAP [29] , prevent CPXV012 binding to TAP , it is tempting to speculate that CPXV012 transiently intercepts the translocation cycle at a state , which is not accessible if TAP is arrested by either US6 or BNLF2a . However , we cannot exclude entirely that binding of either US6 or BNLF2a to TAP mask directly the CPXV012 interaction site on TAP due to steric competition . In this study , we provide several lines of evidence that the active domain of the immune evasin is located at the C-terminal tip of CPXV012 , which is tethered to the ER membrane and inhibits the coreTAP complex . First , a C-terminally deletion CPXV012-CΔ5 is inactive in blocking MHC I antigen processing , whereas an N-terminal deletion is active to the same extent as full-length CPXV012 . Second , the ten C-terminal residues of CPXV012 are sufficient to inhibit the peptide-stimulated ATPase activity of coreTAP . Notably , inhibition by the 10mer fragment is compartment-specific , because it only arrests TAP from the lumen , not from the cytosol . This active C-terminal segment of CPXV012 , which emerged from a frameshift in the CPXV012 gene , exhibits all hallmarks of a high-affinity TAP substrate: positively charged or aromatic residues at position 1–3 and a hydrophobic residue at its C terminus [37] . Disturbing this consensus sequence by replacing the N-terminal positively charged residues ( Arg59 and Arg60 ) by an alanine ( C8CPXV012RR59AA ) rendered the viral inhibitor inactive . If combined , these results are in line with a recently discovered negative feedback mechanism of TAP ( trans-inhibition ) [41]: substrate peptides are transported against the gradient into the lumen of TAP-containing proteoliposomes until the transporter is inhibited in trans by accumulated lumenal peptides . In vivo , a significant proportion of TAP-translocated peptides does not associate with MHC I molecules , since they are more selective in peptide binding , if compared to TAP [44] . Accumulation of non-optimal MHC I substrates may compete with newly transported peptides for MHC I loading . Moreover , high concentrations of ER-lumenal peptides might even induce the unfolded protein response ( UPR ) and ER stress as a direct consequence of peptide accumulation , or indirectly as a result of competition between peptides and proteins for folding , assembly and retrograde translocation [44] . There is growing evidence that the UPR intersects with the functioning of cells of the immune system [45] . For instance , ER stress induces a decrease of MHC I surface expression [46] and a reduction in TAP1 protein levels , which is caused by the expression of microRNA-346 that directly targets human TAP1 mRNA [47] . In summary , it seems reasonable to assume that a trans-inhibition mechanism of TAP , as identified recently [41] , would prevent detrimental accumulation of peptides in the ER lumen and hence ER stress . In addition , trans-inhibition might synergistically cooperate with pathways that export TAP-imported peptides from the ER lumen [48] , [49] . Since the C terminus of CPXV012 resembles a high-affinity TAP peptide substrate that is sufficient to inhibit the transporter as soluble lumenal peptide , we propose that CPXV012 exploits this trans-inhibition mechanism of the TAP machinery . The active C-terminal fragment of CPXV012 is membrane-anchored and bound to TAP via its TMD , thereby mimicking a high ER lumenal peptide concentration . This high local concentration may trigger the ER-lumenal low-affinity [41] peptide binding site of TAP , causing the inhibition of the transport complex in trans ( Fig . 6E ) . CPXV is the suggested ancestor of other orthopoxviruses and contains the most complete set of immune modulators [35] . They target host pathways that regulate immediate immune responses , particularly the IFNs , the chemokines , the pro-inflammatory cytokines , the complement system , and mediators that orchestrate adaptive immunity [50] . With regard to this repertoire of mechanisms , two immune evasins of CPXV , namely CPXV012 and CPXV203 , are known to inhibit synergistically the MHC I antigen presentation pathway: As shown here , CPXV012 diminishes the peptide supply in the ER by exploiting a novel trans-inhibition mechanism , whereas CPXV203 interferes with trafficking of MHC I from the ER to the plasma membrane by reprogramming KDEL-receptor recycling [15] , [16] , [51]–[53] . Down-regulation of MHC I antigen presentation by CPXV012 and CPXV203 does not inhibit the priming of CPXV-specific CD8+ T lymphocytes ( CTLs ) during acute infection , but blocks these CTLs at the effector response level by preventing recognition of infected cells in vivo [54] . Since direct priming should be also inhibited by CPXV-induced MHC I down-regulation , these results supported a model in which cross-presentation is used for CD8+ T cell priming during CPXV infection [54] . On the other hand , MHC I down-regulation would also leave infected cells vulnerable to be killed by NK cells . Usually , mature NK cells are tolerated toward normal cells by integrating stimulatory and inhibitory signals transmitted via receptor-ligand interactions [55] . According to the “missing-self” hypothesis [56] , NK cells recognize and eliminate cells that fail to express MHC I molecules . Hence , it is not surprising that orthopoxviruses encode immune evasins that inhibit NK-cell activation . For example , CPXV expresses the orthopoxvirus MHC class I-like protein OMCP [57] , a high-affinity antagonist of the activating NK-receptor NKG2D [58] . An opposite strategy is the expression of decoy ligands for the NK cell inhibitory receptor NKR-P1B , which plays a critical role in the MHC I independent “missing self” recognition by detecting the C-type lectin-related ligand Clr-b [59] , [60] . Indeed , rat cytomegalovirus expresses a C-type lectin-like ( RCTL ) gene product that resembles Clr-b , which inhibits NK killing of infected cells via direct interaction with NKR-P1B [7] . Interestingly , in a subgroup of CPXV strains , CPXV012 codes for D10L that contains the C-type lectin domain at the C terminus ( Fig . 5A ) . D10L is highly homologous to Clr-b , but it is beyond the scope of this manuscript to determine whether D10L indeed functions as a decoy ligand for NKR-P1B . Nevertheless , D10L binds to the TAP complex , since the N-terminal cytosolic and transmembrane region of D10L and CPXV012 are highly conserved . The physiological relevance of this interaction remains to be determined . More importantly , D10L lacks the C-terminal alternative reading frame , which mimics a TAP substrate consensus sequence , and is therefore unable to inhibit MHC I surface presentation [15] . In combination with the fact that the transmembrane region of all CPXV012 variants mediate the interaction with TAP , their differences at the C terminus explains on a molecular level how a gene product adapts “incidentally” a new function , thereby crossing the frontier of sabotage and stealth . While this manuscript was undergoing final revision , a report by Luteijn et al . was published that also describes data relevant to the viral immune evasions strategy of CPXV012 [61] . Consistent with our findings , it was shown that CPXV012 inhibits ATP binding , but not peptide binding to TAP . Substitutions of long polyalanine stretches within the ER-luminal region indicate that residues 41–65 of the viral protein are involved in TAP inhibition [61] . However , these experiments could not distinguish whether the loss-of-function is either caused because these long alanine stretches replace functionally important amino acids within the CPXV012 functional domain , or induce secondary effects , which then render the protein inactive . Here , we developed an opposite experimental strategy based on maintaining CPXV012 function . We show that the isolated C-terminal 10mer CPXV012 fragment is sufficient to inhibit ATPase activity of purified TAP . Moreover , by reconstitution of the inhibition pathway of CPXV012 by purified components , we provide direct evidence that this fragment is only active if provided in trans , on the ER-lumenal side of TAP . Based on these results , we conclude that the stretch of ten residues at the C terminus of CPXV012 is the active domain of the viral protein , which is necessary and sufficient to inhibit TAP . Interestingly , the ER-lumenal domain of CPXV012 displays a strong affinity for phospholipids and might reside at the lipid-water interface of the ER membrane in close proximity to the TMDs of TAP [61] . Since it is currently completely unknown where and how peptides are released form TAP inside the ER lumen , it is plausible that the low-affinity ER-lumenal peptide sensor [41] , which initiates trans-inhibition , is located at the membrane-TMD interface of the translocation machinery . Several key findings are disclosed by the present study . First , we decipher the molecular basis of how CPXV012 inhibits antigen presentation via MHC I by direct inhibition of ATP binding to coreTAP . Second , we show that the inhibition mechanism is distinct of all other yet identified viral TAP inhibitors . Thus , CPXV012 is an excellent tool to study the structure and function of TAP as well as new antigen processing pathways . Third , we propose that CPXV012 blocks TAP by hijacking the newly identified negative feedback mechanism of the antigen translocation machinery . Therefore , our findings provide an important tool to uncover the yet uncharacterized negative feedback regulation of TAP . Finally , by elucidating the functional evolution of the viral immune evasive protein CPXV012 , this study provides the rare opportunity to decipher on a molecular level how nature plays hide and seek with a pathogen and its host .
CPXV012 ( Gene ID: 1485887 ) with an N-terminal His10-flag tag ( flagCPXV012 ) was codon optimized for expression in mammalian and insect cells and synthesized de novo ( GenScript , Piscataway , NJ , USA ) . flagCPXV012 was cloned into pFastBac1 ( Invitrogen , Carlsbad , CA , USA ) vectors for the baculovirus production or was used as a template for PCR amplification . PCR reactions were performed under standard conditions using Phusion DNA polymerase ( Finnzymes , Vantaa , Finland ) and synthetic oligonucleotide primers ( endonuclease cleavage sites are underlined ) . All constructs were verified by DNA sequencing . C8CPXV012 was generated using C8-CPVX012-BamHI-fo ( GCGGATCC ATGCCGCGCGGCCCGGATCGCCCGGAAfGGCATTGAAGAATTCATCATGAGAGAGTCTATC ) , and CPXV012-EcoRI-rev ( GCGAATTCTTAGATGATGCTATCCAGC ) primers and was subcloned into pFASTBac1 ( Invitrogen ) . Recombinant baculoviruses were generated using the Bac-to-Bac baculovirus expression system ( Invitrogen ) . For expression in human cells , PCR-generated products were cloned into pIRES2-EGFP ( Clontech Laboratories Inc . , Palo Alto , CA , USA ) via the respective restriction sites upstream of the internal ribosome entry site ( IRES ) and enhanced GFP . C8CPXV012 , flagCPXV012 ( also containing an N-terminal His10-tag ) , and C8CPXV012-CΔ5 were amplified with the following primers: C8-CPXV012-NheI-fo ( AATGCTAGCACCATGGCA CCGCGCGGCCCGGATCGCC ) , Flag-CPXV012-NheI-fo ( AATGCTAGCACCATGGCACACCAT CATCACCATCATCACCAC ) , CPXV012-full-length-BamHI-rev ( TTAGGATCCTCAGATGATG CTATCCAGCTTGTGGTAGTAC ) , and CPXV012-CΔ5-BamHI-rev ( TTAGGATCCTCACTTGTG GTAGTACCTGCGGAACAG ) , respectively . C8NΔ6-CPXV012 was generated by three sequential PCR reactions using the overlapping forward primers CPXV012-NΔ6-PCR1-fo ( CGGAAGGCATTGAAGAAAGCGGATCTATCTACCGTGTGATGATCG ) , CPXV012-PCR2-fo ( CCGCGCGGCCCGGATCGCCCGGAAGGCATTGAAGAAAGCGGA ) , and C8-CPXV012-NheI-fo , and the reverse primer CPXV012-full-length-BamHI-rev . Residues Arg59 and Arg60 of CPXV012 were simultaneously replaced by two alanines ( C8CPXV012RR59AA ) via site-directed mutagenesis of the template C8CPXV012-pIRES2-EGFP using the primers CPXV012-R59AR60A-fo ( GTCCCTGCTGTTCGCGGCATACTACCACAAGCTG ) and CPXV012-R59A/R60A-rev ( CAGCTTGTGGTAGTATGCCGCGAACAGCAGGGAC ) . D10L ( Gene ID: 90660243 ) with an N-terminal C8 tag ( C8D10L ) was synthesized de novo and cloned into pIRES2-EGFP using NheI and BamHI restriction sites . BNLF2aC8-NST-pIRES2-EGFP has been described previously [29] . Human coreTAP1mVenus-His10 ( TAP1 residues 227-808; Q03518 ) and coreTAP2mCerulean-StrepII ( TAP2 residues 125-704; Q59H06 ) have been described previously [62] . Spodoptera frugiperda ( Sf9 ) insect cells were grown in Sf900II medium ( Invitrogen ) following standard procedures . Sf9 cells were infected with recombinant baculovirus encoding for flagCPXV012 , C8CPXV012 , US6myc [29] , BNLF2aC8-NST [29] , codon-optimized human TAP1/2 [63] , and human wild-type TAP1His6/TAP2 [64] , coreTAP1/2 [11] respectively . 48 h after infection , cells were harvested , and membranes were prepared as described [64] . HeLa and HEK293T cells ( ATCC CRL-11268 ) were cultured in DMEM ( PAA Laboratories , Cölbe , Germany ) supplemented with 10% fetal calf serum ( FCS; Biochrom AG , Berlin , Germany ) at 37°C in a 5% CO2-humidified atmosphere . For flow cytometry , HeLa cells ( ATCC CCL-2 ) were seeded in 6-well plates with a density of 2×105 cells/well the day before transfection . Transient transfection was conducted according to the Magnetofection protocol ( Chemicell , Berlin , Germany ) using 2 µg of plasmid DNA and 2 µl of PolyMAG ( Chemicell ) per well . For the tandem-affinity purification of TAP complexes , HEK293T cells were seeded in 150 mm dishes with a density of 2×106 cells/dish . On the next day , cells were transfected using 90 µg of polyethyleneimine ( Sigma-Aldrich , Taufkirchen , Germany ) and 30 µg of plasmid DNA according to standard procedures . Monoclonal mouse anti-TAP1 ( mAb 148 . 3 ) and anti-TAP2 ( mAb 435 . 3 ) antibodies were used for immunoblotting and immunoprecipitation [64] . In certain experiments , polyclonal rabbit anti-TAP1 ( 1p2 ) and anti-TAP2 ( 2p4 ) antibodies were used for immunoprecipitation [65] . C8-tagged CPXV012 and BNLF2a were immunoprecipitated and detected by an anti-C8 antibody [66] . Flag-tagged CPXV012 was immunoprecipitated and detected by anti-flag antibodies purchased from Sigma-Aldrich ( F1804 ) . US6myc was immunoprecipitated and detected by a monoclonal anti-myc antibody ( 4A6 ) from Millipore ( Schwalbach , Germany ) . The HC10 antibody , which recognizes human MHC I heavy chain [67] , was kindly provided by Hidde L . Ploegh ( MIT , Cambridge , MA ) . Peptide-loaded MHC I molecules were detected by phycoerythrin ( PE ) coupled anti-human HLA-ABC ( W6/32 ) . Mouse IgG2a isotype control antibodies were purchased from BioLegend ( San Diego , CA , USA ) . Actin was detected by a mouse anti-actin antibody purchased from Sigma-Aldrich ( A2228 ) . Peptides were prepared by solid-phase synthesis using the Fmoc/tBu strategy . Peptides were purified by reverse-phase C18 HPLC , and their identity was confirmed by mass spectrometry . Peptides representing the last 25 , 20 , 15 , 10 , or 5 residues of CPXV012 were acetylated at their N termini . Transiently transfected HeLa cells were harvested 48 h after transfection and washed once with cold FACS buffer ( PBS/2% FCS ) . After blocking with 5% BSA in FACS buffer on ice , cells were washed twice and incubated with W6/32-PE or the PE-coupled isotype control antibody for 15 min on ice in the dark . Cells were again washed twice and FACS analysis was performed using an Attune Acoustic Focusing Cytometer ( Applied Biosystems , Life Technologies , Carlsbad , CA , USA ) . Data were analyzed with FlowJo software ( TreeStar , Ashland , OR , USA ) . For evaluation , a FSC/SSC gate was drawn around the main cell population and only GFP-positive cells were analyzed with respect to MHC I surface expression . To confirm protein expression , parallel samples of transfected HeLa cells were solubilized in RIPA Lysis and Extraction Buffer ( Thermo Fisher Scientific , Pierce , Rockford , IL , USA ) and analyzed by SDS-PAGE ( 12% ) and subsequent immunoblotting with anti-C8 antibodies . Membranes prepared from insect cells ( 1 mg of total protein ) were resuspended in 0 . 4 ml lysis buffer L1 ( 20 mM Tris/HCl , pH 7 . 5 , 150 mM NaCl , 5 mM MgCl2 , 2% ( w/v ) digitonin ( Carl Roth , Karlsruhe ) , 1 . 5 mM phenylmethylsulfonylfluoride , 3 mM benzamidine ) . After solubilization for 60 min on ice , non-solubilized proteins were removed by centrifugation at 100 , 000× g for 30 min at 4°C . The supernatant was incubated with M-280 sheep anti-mouse IgG Dynabeads ( Dynal Biotech , Hamburg ) , which had been pre-loaded with antibodies . Beads were washed three times with 1 ml of washing buffer ( 20 mM Tris/HCl , pH 7 . 5 , 150 mM NaCl , 2 mM EDTA , 0 . 2% ( w/v ) digitonin ) . Proteins were eluted in 1× SDS sample buffer ( 2% SDS , 50 mM Tris/HCl , pH 8 . 0 , 200 mM DTT , 10% glycerol , 0 . 05% bromophenol blue ) for 3 min at 65°C . Samples were denatured for 20 min at 65°C and separated by SDS-PAGE ( 12% or 6% ) . After electro transfer onto nitrocellulose membranes , proteins were detected with specific antibodies as indicated . Horseradish peroxidase ( HP ) conjugated secondary antibodies were detected with Lumi-Imager F1 ( Roche ) . One representative blot out of three independent experiments is shown . Membranes prepared from insect cells ( 0 . 1 mg of total protein ) were resuspended in 50 µl of AP buffer ( 5 mM MgCl2 in PBS , pH 7 . 4 ) in the presence of 3 mM ATP . The transport reaction was started by adding 1 µM of the peptide RRYQNSTC ( F ) L ( C ( F ) , fluorescein-labeled cysteine ) for 3 min at 32°C and terminated with 1 ml ice-cold stop buffer ( 10 mM EDTA in PBS , pH 7 . 0 ) . After centrifugation at 20 , 000× g for 8 min , the pellet was solubilized in 0 . 5 ml of lysis buffer L2 ( 50 mM Tris/HCl pH 7 . 5 , 150 mM NaCl , 5 mM KCl , 1 mM CaCl2 , 1 mM MnCl2 , 1% Nonidet P-40 ) for 30 min at 4°C . Non-solubilized proteins were removed by centrifugation and the supernatant was incubated with 60 µl ConA Sepharose ( 50% w/v , Sigma-Aldrich ) for 1 h at 4°C . After three washing steps with 0 . 5 ml lysis buffer each , ConA-bound peptides were specifically eluted with methyl-α-D-mannopyranoside ( 200 mM ) and quantified by a fluorescence plate reader ( λex/em 485/520 nm; Polarstar Galaxy , BMG Labtech , Offenburg , Germany ) . Background transport was determined in the presence of apyrase ( 1 U/sample ) . All measurements were performed in triplicate . Membranes prepared from insect cells ( 0 . 1 mg of total protein ) were incubated with 0 . 5 µM of the peptide RRYC ( F ) KSTEL ( C ( F ) , fluorescein-labeled cysteine ) in 50 µl of AP buffer for 15 min at 4°C . Free peptides were removed by washing the membranes twice with 1 ml of ice-cold AP buffer and subsequent centrifugation at 20 , 000× g for 8 min . Membranes were lysed with 0 . 3 ml AP buffer containing 1% SDS , and peptides were quantified by a fluorescence plate reader . All measurements were performed in triplicate . Background binding was determined in 100-fold excess of unlabeled high-affinity substrate peptide R9LQK ( RRYQKSTEL ) . Membranes prepared from insect cells ( 0 . 5 mg of total protein ) were resuspended in 100 µl of ice-cold PBS buffer and incubated with the homobifunctional cross-linker ethylene glycol bis ( succinimidyl succinate ) ( EGS , Thermo Scientific , Rockford , IL ) at a final concentration of 0 . 5 mM . After incubation for 30 min at 4°C , the reaction was stopped by adding Tris/HCl buffer , pH 7 . 5 ( 50 mM final concentration ) . Membranes were collected by centrifugation ( 20 , 000× g for 8 min at 4°C ) and analyzed by SDS-PAGE ( 6% ) and immunoblotting using TAP1 or TAP2 specific antibodies . As specifically indicated , membranes were pre-incubated with 10 µM peptide R9LQK at 4°C for 1 h prior to cross-linking . One representative blot out of three independent experiments is shown . TAP1/2 complexes were purified from Sf9 membranes ( 600 µg of total protein ) by immunoprecipitation using anti-TAP1 ( mAb 148 . 3 ) or anti-C8 antibodies as described above . Dynabead immobilized complexes were pre-incubated in ATP binding buffer ( 20 mM HEPES , pH 7 . 4 , 137 mM NaCl , 3 mM MgCl2 ) containing 15 µM of 8-azido-ATP[γ]biotin ( Biolog Life Science Institute , Bremen , Germany ) for 5 min on ice . Controls contained 15 µM 8-azido-ATP[γ]biotin and 5 mM unlabeled ATP . Photo cross-linking was initiated by UV irradiation ( 254 nm hand-held UV lamp ) for 5 min on ice . After three washing steps in ATP-binding buffer , samples were denatured for 20 min at 65°C and separated by SDS-PAGE ( 12% ) . After electro transfer onto nitrocellulose membranes , proteins were detected with specific antibodies as indicated . Biotinylated proteins were visualized using an extravidin-HRP conjugate . Alternatively , membranes prepared from insect cells ( 600 µg of total protein ) were resuspended in 150 µl ATP binding buffer and pre-incubated with 15 µM of 8-azido-ATP[γ]biotin in the presence and absence of an excess of ATP ( 5 mM ) for 5 min on ice . After subsequent photo cross-linking for 5 min on ice , membranes were collected by sedimentation ( 20 , 000× g for 8 min ) , washed three times in ATP binding buffer , and then solubilized in 300 µl lysis buffer L1 . Immunoprecipitations were performed using antibodies against either TAP1 ( mAb 148 . 3 ) or the C8 tag as described above . Samples were analyzed by SDS-PAGE ( 12% ) and subsequent immunoblotting with extravidin-HRP or the corresponding antibodies . One representative blot out of three independent experiments is shown . Since overexpressed TAP1 or TAP2 subunits might form homodimers , a orthogonal purification strategy was applied that ensures the isolation of heterodimeric TAP1/2 complexes [62] . TAP1/2-expressing HEK293T cells were resuspended in ice-cold buffer A ( 50 mM Tris/HCl , pH 8 . 0 , 250 mM NaCl , 10% glycerol , 0 . 05% ( w/v ) digitonin ) supplemented with 1% ( w/v ) digitonin and incubated for 1 h in an overhead shaker . All steps were carried out at 4°C . Non-solubilized proteins were removed by centrifugation at 150 , 000× g for 1 h at 4°C . Proteins were bound to 100 µl streptavidin agarose beads ( Pierce , Rockford , IL , USA ) for 1 h . The beads were washed two times with 1 ml buffer A . Proteins were eluted with 1 ml buffer A supplemented with 2 . 5 mM biotin . The streptavidin agarose eluate was incubated for 1 h with IgG Dynabeads , which had been pre-loaded with anti-TAP1 antibodies ( mAb 148 . 3 ) . After three washing steps , the affinity-purified protein was eluted with 2× SDS buffer for 30 min at 37°C and analyzed by SDS-PAGE and immunoblotting . To prepare coreTAP-containing insect ER microsomes , Sf9 cells were infected with recombinant baculovirus encoding for coreTAP1/2 [11] . 48 h after infection , cells were harvested , and membranes were prepared as described [68] . For the generation of mRNAs , C8CPXV012 or C8D10L was amplified directly from the corresponding pIRES2-EGFP plasmid using SP6-pIRES-fo ( GATTTAGGTGACACTATAGAATACCACCGTCTATA TAAGCAGAGCTGGTTTAGTGAACC ) and pIRES-GFP-rev ( GTTTACGTCGCCGTCCAGC ) primers . Purified PCR product was transcribed in vitro using SP6 RNA polymerase as before [29] . In vitro translation of purified mRNA ( typically 50 µl reactions , 30°C , 40 min ) was performed in the presence of rabbit reticulocyte lysate ( Promega , Madison , WI , USA ) , Sf9 microsomes , and [35S]Met ( 0 . 4 µCi/µl ) . After translation , samples were analyzed by immunoprecipitation and phosphoimaging . Peptide stimulated ATPase activity was determined by a colorimetric assay based on the complex formation of free inorganic phosphate and ammonium molybdate with malachite green [69] . CoreTAP1mVenus-His10 and coreTAP2mCerulean-StrepII were expressed in Pichia pastoris and solubilized in 2% ( w/v ) digitonin as described previously [40] , [62] . CoreTAP1/2 heterodimers were purified by an orthogonal strategy using metal-affinity chromatography and subsequent streptactin-affinity chromatography as described [62] . For ATP hydrolysis measurements , 0 . 2 µM of purified coreTAP1/2 was incubated with 1 mM MgATP and the high-affinity substrate peptide R9LQK ( 1 . 0 µM ) for 30 min at 37°C as described previously [63] . For reconstitution of purified coreTAP1/2 into liposomes , large unilamellar vesicles consisting of Escherichia coli polar lipid extract ( Avanti Polar Lipids Inc . , Alabaster , AL , USA ) and 1 , 2-dioleoyl-sn-glycero-3-phosphocholine ( Avanti Polar Lipids , Inc . ) with a molar ratio of 7∶3 were prepared as previously [63] . Detergent-destabilized vesicles and purified coreTAP1/2 were mixed with a lipid to protein ratio of 20∶1 ( w/w ) and incubated for 30 min at 4°C . Detergent was removed by incubation with polystyrene beads as described previously [40] . Proteoliposomes were pelleted for 30 min at 80 , 000 rpm at 4°C . Protein aggregates , empty vesicles and proteoliposomes were separated by centrifugation on a continuous Ficoll density gradient [40] . After washing , proteoliposomes were resuspended to a final concentration of 5 mg/ml in reaction buffer ( 20 mM HEPES , 200 mM NaCl , 50 mM KCl , 5% glycerol , pH 7 . 3 ) . For peptide transport , coreTAP1/2 ( 0 . 5 µg ) containing proteoliposomes were pre-incubated on ice in 45 µl reaction buffer containing 1 µM fluorescence labeled peptide RRYC ( F ) KSTEL . Transport was started by adding 3 mM MgATP and performed for 10 min at 37°C . The reaction was stopped as described previously [63] . For encapsulation of peptides , TAP proteoliposomes in reaction buffer containing the active CPXV012 fragment were snap frozen in liquid nitrogen and thawed on ice . After three freeze and thaw cycles , proteoliposomes were washed twice to remove the remaining , not encapsulated peptide . Peptide transport assay was performed as described above . | Virus-infected or malignant transformed cells are eliminated by cytotoxic T lymphocytes , which recognize antigenic peptide epitopes in complex with major histocompatibility complex class I ( MHC I ) molecules at the cell surface . The majority of such peptides are derived from proteasomal degradation in the cytosol and are then translocated into the ER lumen in an energy-consuming reaction via the transporter associated with antigen processing ( TAP ) , which delivers the peptides onto MHC I molecules as final acceptors . Viruses have evolved sophisticated strategies to escape this immune surveillance . Here we show that the cowpox viral protein CPXV012 inhibits the ER peptide translocation machinery by allosterically blocking ATP binding and hydrolysis by TAP . The short ER resident active domain of the viral protein evolved from a reading frame shift in the cowpox virus genome and exploits the ER-lumenal negative feedback peptide sensor of TAP . This CPXV012-induced conformational arrest of TAP is signaled by a unique communication across the ER membrane to the cytosolic motor domains of the peptide pump . Furthermore , this study provides the rare opportunity to decipher on a molecular level how nature plays hide and seek with a pathogen and its host . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"immunology",
"biophysics",
"molecular",
"biology",
"veterinary",
"science"
] | 2014 | A Negative Feedback Modulator of Antigen Processing Evolved from a Frameshift in the Cowpox Virus Genome |
Senataxin , mutated in the human genetic disorder ataxia with oculomotor apraxia type 2 ( AOA2 ) , plays an important role in maintaining genome integrity by coordination of transcription , DNA replication , and the DNA damage response . We demonstrate that senataxin is essential for spermatogenesis and that it functions at two stages in meiosis during crossing-over in homologous recombination and in meiotic sex chromosome inactivation ( MSCI ) . Disruption of the Setx gene caused persistence of DNA double-strand breaks , a defect in disassembly of Rad51 filaments , accumulation of DNA:RNA hybrids ( R-loops ) , and ultimately a failure of crossing-over . Senataxin localised to the XY body in a Brca1-dependent manner , and in its absence there was incomplete localisation of DNA damage response proteins to the XY chromosomes and ATR was retained on the axial elements of these chromosomes , failing to diffuse out into chromatin . Furthermore persistence of RNA polymerase II activity , altered ubH2A distribution , and abnormal XY-linked gene expression in Setx−/− revealed an essential role for senataxin in MSCI . These data support key roles for senataxin in coordinating meiotic crossing-over with transcription and in gene silencing to protect the integrity of the genome .
Ataxia oculomotor apraxia type 2 ( AOA2 ) , a severe form of autosomal recessive cerebellar ataxia ( ARCA ) is characterised by progressive cerebellar atrophy and peripheral neuropathy , oculomotor apraxia and elevated α-fetoprotein [1] , [2] . The gene defective in AOA2 , SETX , is also associated with amyotrophic lateral sclerosis 4 ( ALS4 ) , an autosomal dominant juvenile-onset form of ALS [3] . Senataxin shares extensive homology in its putative helicase domain with the yeast , Saccharomyces cerevisiae splicing endonuclease 1 protein ( Sen1p ) which possesses helicase activity and is involved in RNA processing , transcription and transcription-coupled DNA repair [4] . A role for senataxin in DNA repair is supported by the observation that AOA2 patient cells display sensitivity to DNA damaging agents such as H2O2 , camptothecin and mitomycin C and have elevated levels of oxidative DNA damage [5] . Senataxin also plays a role in transcription regulation by its ability to modulate RNA Polymerase II ( Pol II ) binding to chromatin and through its interaction with proteins involved in transcription [6] . mRNA splicing efficiency , splice site selection , and transcription termination were all defective in senataxin-deficient cells [6] . A recent study has described an additional role for senataxin in transcription elongation and termination [7] . Cells deficient in senataxin displayed an increase in RNA read-through and Pol II density downstream of the Poly ( A ) site and also exhibited increased levels of R-loops ( RNA:DNA hybrids that form over transcription pause sites ) formation [8] , [9] . The yeast ortholog of senataxin , Sen1p , has also been shown to protect its heavily transcribed genome from R-loop-mediated DNA damage [10] . More recently , a role for senataxin has been described at the interface between transcription and the DNA damage response [11] . They revealed that senataxin forms nuclear foci in S/G2 phase cells and these foci increased in response to DNA damage and impaired DNA replication . These foci disappeared upon resolution of R-loops or after inhibition of transcription . Evidence has also been provided for the association of the yeast ortholog of senataxin , Sen1 , with DNA replication forks across RNA polymerase II transcribed genes [12] . These data demonstrate a co-ordinating role for Sen1 between replication and transcription . We generated a Setx knockout mouse model to investigate further the role of senataxin . Our data revealed that this protein is essential for male meiosis , acting at the interface of transcription and meiotic recombination , and also in the process of meiotic sex chromosome inactivation ( MSCI ) .
Setx−/− mice were produced using a Cre-LoxP system to delete exon 4 as outlined in Figure 1A . Crosses between Setx heterozygotes produced all 3 genotypes ( Wild type , heterozygotes and homozygote knockouts ) as expected ( Figure 1B ) and a Mendelian inheritance pattern was observed ( Wild type 25%; heterozygote 54%; knockout 21%; n = 87 ) . Inactivation of the Setx gene was confirmed by RT-PCR and the absence of Setx mRNA in the knockout mouse as compared to the wild type ( Figure 1C ) . Immunoprecipitation ( IP ) with anti-senataxin antibodies from testes extracts confirmed the presence of the protein in Setx+/+ mice but senataxin was not immunoprecipitated from Setx−/− extracts ( Figure 1D ) . While progressive cerebellar degeneration is characteristic of senataxin-defective AOA2 patients [1] , [2] we failed to detect either structural alterations , general cerebellar degeneration , or specific loss of Purkinje cells in Setx−/− mice ( data not shown ) . Using a simple phenotypic scoring system that has been employed to evaluate mouse models of cerebellar ataxia [13] , we failed to reveal any significant neurological/behavioural difference and ataxia between Setx+/+ and Setx−/− animals ( Figure S1 ) . Multiple attempts to breed Setx−/− mice with each other or with wild type mice were unsuccessful . Male mutant mice had normal development of secondary sexual characteristics , and were capable of the mechanics of mating , but were infertile . Histological examination of Setx−/− female ovaries at various ages ( from 35 days to 8 months of age ) revealed no overt phenotypic difference from their wild type littermates , with normal structure and presence of follicles at all stages and an ability to ovulate ( Figure S2 ) . However , the yield of viable embryos at 0 . 5 dpc was very low suggesting that Setx−/− females are less fertile than their wild type littermates . The fertility of an individual female is a reflection of the number of eggs ovulated and their competence . To investigate the fertility of Setx−/− female mice , we carried out superovulation and time mating to harvest one-cell stage ( 0 . 5 dpc , fertilised egg ) embryos in order to compare their viability . A greater than 3 . 5-fold reduction in the yield of 0 . 5 dpc for Setx−/− was observed compared to wildtype animals ( 10–20 0 . 5 dpc embryos for Setx−/− compared to 50–70 for wild types ) . In addition , only 23% of viable embryos were obtained at 0 . 5 dpc for Setx−/− and most of the viable ones did not survive in culture , indicating that Setx−/− females have a reduced fertility . Since oligospermia and testicular abnormalities are a frequent finding of ARCA patients and the corresponding mouse models [14] , [15] , we compared the development of testes and seminiferous tubules from Setx−/− with those from wild type mice . Setx−/− testes were smaller in size ( 50–60% reduction in size ) than wild-type littermates ( Figure 2A ) and histological examination of testes from 35 day-old Setx−/− males revealed a severe disruption of the seminiferous tubules and the absence of germ cells compared to Setx+/+ males ( Figure 2B–2G ) . Morphologically , spermatocytes in Setx−/− mice appear to have halted development at pachytene stage of meiotic prophase ( Figure 2E ) , suggesting that meiotic arrest in Setx−/− mice occurs during prophase I . Overall the seminiferous epithelium from an 8-month Setx+/− mouse testis appears normal but there is evidence of some disruption in places , with few round or elongated spermatids and debris in the lumen ( Figure S3 ) . Histological examination of the epididymis from Setx−/− mice revealed the total absence of mature sperm ( Figure 2F–2G ) thus confirming the infertility of Setx−/− males mice . Elevated levels of apoptosis were detected in some tubules of Setx−/− mice following TUNEL staining ( Figure 2H–2I ) , suggesting that arrested cells are eliminated via this pathway . To monitor the development of spermatocytes we counted the number of spermatocytes in all stages of meiotic prophase I ( Figure 2J ) . Synaptogenesis appeared to be grossly normal in Setx−/− mice as determined by staining for synaptonemal complex protein 3 ( SCP3 ) [16] but while the earlier stages of meiosis were represented we failed to detect diplotene stage spermatocytes for Setx−/− , indicating a block at the pachytene-diplotene transition ( Figure 2J ) . Further analysis of the first meiotic division of prophase I revealed a significant reduction of pachytene spermatocytes from day 16 to day 22 in Setx−/− ( Figure 2K ) in line with the lack of diplotene spermatocytes in Setx−/− . Fragmentation of the synaptomemal complex ( SC ) at pachytene stage in Setx−/− was also observed . To investigate the cause of the meiotic defect in more detail , we also determined the expression of spermatogenesis stage-specific markers ( Figure S4A ) [17] , [18] . Expression levels of spermatogonial and early spermatocyte markers Dmc1 , Calmegin , and A-myb were similar in both Setx−/− and Setx+/+ . Pgk2 , a marker expressed from the beginning ( pre-leptotene ) and throughout meiosis ( leptotene , zygotene , pachytene , diplotene ) up to the round spermatid stage showed only a small reduction in expression in Setx−/− as compared to Setx+/+ . Markers for haploid mature germ cells Prm1 , Prm2 and Tnp1 showed markedly reduced expression in Setx−/− compared to Setx+/+ ( Figure S4B ) . Together , these data indicate that male germ cells proceed normally from spermatogonia up to the meiotic pachytene stage in Setx−/− but fail to enter into spermiogenesis and form mature spermatids . Thus , both gene expression of meiosis stage-specific markers and spermatocyte spread analysis confirmed the blockage of meiosis in Setx−/− male germ cells and indicate that senataxin plays an essential role in the development and maturation of germ cells . Meiotic recombination is initiated by the formation of DNA double strand breaks ( DSB ) catalysed by a type II topoisomerase-like protein Spo11 [19] . These breaks trigger phosphorylation of histone H2AX at ser139 ( γH2AX ) on large domains of chromatin in the vicinity of the break [20] . As meiosis proceeds to the pachytene stage , γH2AX disappears from synapsed chromosomes and is restricted to the largely unsynapsed sex chromosomes in the sex body [21] , [22] . Successful generation of DNA DSB and initiation of repair was observed in Setx−/− ( Figure 3A ) . At pachytene stage , DSBs disappeared from the autosomes in Setx+/+ mice and only the sex chromosomes stained positive for γH2AX as expected [23] , [24] . On the other hand , γH2AX foci remained on apparently synapsed autosomes at pachytene stage in Setx−/− , indicating the persistence of unrepaired DSBs . Both Setx−/− and Setx+/+ displayed γH2AX staining at the sex chromosomes at pachytene stage ( Figure 3A ) . Staining of Setx−/− testes sections for γH2AX confirmed the greater intensity of labelling ( Figure S5 ) . The repair of meiotic DNA DSB occurs via homologous recombination ( HR ) and involves the participation of various DNA repair factors including RPA , Dmc1 and Rad51 [16] . Both Rad51 and Dmc1 play key roles in the initial steps of HR by mediating strand invasion and homologous pairing . These proteins are normally observed as multiple foci decorating the chromosomes , first appearing at leptotene and sharply decreasing at pachytene [25] . This was the case for Rad51 in Setx+/+ with few foci labelling pachytene chromosomes ( Figure 3B ) . In contrast , multiple Rad51 foci persisted at pachytene stage in Setx−/− ( Figure 3B , Figure S6A ) , pointing to a defect in Rad51 filament disassembly as a consequence of unrepaired DNA DSB and likely to interfere with HR progression in Setx−/− . Indeed , quantitation of the number of Rad51 foci at pachytene stage revealed a 6-fold increase of these foci in Setx−/− compared to Setx+/+ ( Figure 3C ) . This was not due to an over expression of Rad51 since comparable mRNA levels are observed in both types of mice ( Figure S7A ) . In contrast , immunoblotting of testes protein extracts revealed reduced levels of Rad51 protein in Setx−/− testes as compared to Setx+/+ indicating that the absence of senataxin is affecting the translation or stability of Rad51 protein ( Figure S7B ) . A similar abnormal pattern of retention at pachytene stage was found for Dmc1 with a 10-fold increase in Setx−/− compared to Setx+/+ ( Figure S6B–S6C ) . Similar to Rad51 , comparable levels of Dmc1 mRNA levels were observed in both mice ( Figure S7C ) . However , we were not able to determine the levels of Dmc1 protein in testes . To assess whether meiotic recombination is completed in Setx−/− , we examined the distribution of the mismatch repair protein Mlh1 , which normally forms foci and marks the location of chiasmata [26] , [27] . We observed an average of 22 Mlh1 foci per pachytene-stage spermatocyte in Setx+/+ ( Figure 3D ) , where up to 78% of spermatocytes SC contain one Mlh1 focus , 19 . 2% contain 2 foci , 0 . 5% contain 3 foci and 2 . 5% had no foci at all , in agreement with previous report [28] . In contrast , no foci were observed in Setx−/− pachytene-stage spermatocytes ( Figure 3D ) , indicating the absence of crossovers . The lack of Mlh1 foci in Setx−/− spermatocytes was not due to a defective expression of Mlh1 gene , as similar levels of the Mlh1 mRNAs were detected in both Setx+/+ and Setx−/− testes ( Figure S7D ) . In contrast to Rad51 , similar levels of Mlh1 protein in both Setx+/+ and Setx−/− were shown by Mlh1 immunoblotting of testes protein extracts ( Figure S7E ) . These results confirmed an essential role for senataxin in meiosis . Sen1p , the yeast homolog of senataxin was recently found to restrict the occurrence of RNA:DNA hybrids , also known as R-loop structures , that form naturally during transcription , and can trigger genomic instability if left unresolved [10] . Furthermore the same group showed that senataxin resolves R-loop structures to facilitate transcriptional termination in mammalian cells [7] . We reasoned that the defective meiosis in Setx−/− testes and the consequent apoptosis at pachytene stage might be due to R-loop accumulation as a consequence of transcriptional abnormalities in the absence of senataxin . As shown in Figure 3E , a collection of pachytene-stage spermatocytes from Setx−/− mice showed a marked accumulation of R-loops compared to Setx+/+ . There was some variation in the R-loop-specific ( S9 . 6 ) antibody [29] , [30] staining intensity between individual pachytene-stage spermatocytes of Setx−/− , indicating heterogeneity in R-loop accumulation ( Figure 3E–3F ) . The fluorescence intensity of individual pachytene-stage spermatocytes detected by the R-loop-specific antibody was classified into three categories: faint-none , medium and strong as indicated in Figure 3E . No pachytene spermatocytes with strong R-loop staining intensity were observed in Setx+/+ ( Figure 3E ) . This was confirmed with Setx−/− testes sections which again showed very intense R-loop staining which was variable in different spermatocytes ( Figure 4A ) . Pre-treatment of testes sections with RNAse H prior to immunostaining reduced dramatically the staining intensity in Setx−/− confirming that these were indeed R-loops ( Figure 4B ) . Co-staining with TUNEL revealed that most cells with accumulated R-loops also undergo apoptosis ( Figure 4A , 4E ) . Although occasionally present in Setx+/+ , R-loop accumulation was dramatically increased in Setx−/− seminiferous tubules as shown by an increase in the number of R-loop positive cells per tubules ( Figure 4C–4D ) . Co-staining with TUNEL revealed that most cells undergoing apoptosis in Setx−/− had accumulation of R-loops ( Figure 4E ) . These data suggest that failure to resolve R-loops is responsible for the accumulation of DNA DSB and disruption of meiosis in Setx−/− . To investigate in more detail the role of senataxin in meiosis , we studied its localization by performing immunostaining on Setx+/+ spermatocyte spreads . As shown in Figure 5A–5B , senataxin localised mostly to the sex chromosomes at pachytene stage . Some background staining was also observed over the autosomes ( Figure 5A ) in line with its effect on meiotic recombination and R-loop resolution . As expected there was no senataxin labelling in Setx−/− spreads ( Figure 5A ) . Partial co-localisation between senataxin and Brca1 was observed albeit there was a more diffuse distribution of senataxin in the XY body ( Figure 5C ) . Brca1 labels the axis of unsynapsed sex chromosomes at pachytene stage to where it is recruited to initiate ( MSCI ) meiotic sex chromosome inactivation [22] . While Brca1 localised to the axis of the sex chromosomes in Setx−/− this was incomplete since it was excluded from part of the chromosome ( Figure 5D ) . It appears that this corresponds to the Y chromosome based on the structural morphology [31] . We also determined whether there was a dependence on Brca1 for localisation of senataxin to the sex chromosomes using a Brca1Δ11/Δ11 p53+/− mutant mouse . The results in Figure 5E show that while senataxin localises to XY chromosomes in wild-type mice it fails to do so in Brca1 mutant mice . The Brca1Δ11/Δ11 mutant protein still localises to the sex chromosome but is unable to recruit senataxin ( Figure 5F ) . We next determined whether senataxin and Brca1 interacted using Brca1 and senataxin co-immunoprecipitations from testes extracts . We failed to co-immunoprecipitate endogenous senataxin and Brca1 from mouse testes extracts ( data not shown ) . In addition , Proximity Ligation Assay ( PLA ) which allows for the in situ detection of endogenous protein-protein interactions failed to reveal a direct interaction between these two proteins ( Figure 5G ) . In contrast , we confirmed the previously reported endogenous Brca1 and ataxia-telangiectasia and Rad3 related ( ATR ) interaction [22] using PLA in situ over the XY body ( Figure 5H ) . A specific PLA signal for the Brca1/ATR interaction is observed on/around the axis of the unsynapsed XY chromosomes in Setx+/+ pachytene spermatocytes in agreement with the Brca1 and ATR distribution patterns over the sex chromosomes . At pachytene stage , ATR kinase , another marker of XY chromosomes , is recruited to the unsynapsed axis of the XY chromosomes through an interaction with Brca1 and then diffuses to XY chromatin where it phosphorylates serine 139 of histone H2AX to trigger chromosomal condensation and transcriptional silencing [22] , [32] . The results in Figure 6A show a diffuse staining pattern for ATR in Setx+/+ on the XY body . On the other hand , ATR decorates only part of the XY chromosome in Setx−/− and does not diffuse out into chromatin ( Figure 6A and Figure S8 ) . The mediator of DNA damage 1 ( MDC1 ) protein also plays a key role at this stage in MSCI [33] . Recognition of unsynapsed axis of the XY chromosomes by Brca1 , ATR and TopBP1 is independent of MDC1 but the chromosome wide spreading of these proteins is dependent on MDC1 . We observed that MDC1 labelled the X chromosome but as with Brca1 and ATR failed to decorate the complete XY chromosome ( Figure 6B ) . However , γH2AX labelling was localised to XY chromatin ( Figure 6C ) . As spermatocytes progress from early to mid pachytene the X chromosome appears elongated and sickle shaped prior to loop “curled bundle” formation in late pachytene [31] . These looped XY structures were observed in Setx+/+ but sickle shaped chromosomes appeared mostly in Setx−/− indicative of arrest in mid pachytene ( Figure 6D ) . Distinguishable “curled bundle” sex chromosomes in Setx−/− pachytene spermatocytes were seen only in half the percentage of wildtype ( Figure 6E ) . Thus , the absence of senataxin also affects XY body formation/structure and reveals a defect in the recognition and distribution of DNA damage response proteins on the sex chromosomes and thus results in MSCI failure . In mammalian spermatogenesis , the sex chromosomes are transcriptionally-silenced during the pachytene stage of meiotic prophase I , forming a condensed chromatin domain termed the sex body [31] , [34] . In the majority of Brca1 mutant pachytene cells sex bodies do not form and transcription is maintained , demonstrating a failure in MSCI [22] . To determine whether the absence of senataxin had a similar effect on MSCI , we analysed the expression of sex-linked genes in Setx−/− mice using RT-PCR as previously described [35] , [36] . As shown in Figure 7A , an increase in the expression of the X-linked Usp26 ( 2 . 44 fold ) , Fthl17 ( 1 . 4 fold ) , Tktl1 ( 1 . 36 fold ) and Ube1x ( 1 . 65 fold ) genes was observed for Setx−/− mice compared to Setx+/+ mice . This was also true for several Y-linked genes that include Ube1y ( 2 fold ) and Rbmy ( 2 . 14 fold ) indicating that MSCI is defective in Setx−/− . Normal expression for autosomal genes Actinb , Dazl , and Gapdh was also observed , confirming the specific nature of MSCI ( Figure 7A ) . In order to confirm that MSCI was induced , staining for the activated form ( Phospho-S2 ) of RNA polymerase II ( Pol II ) , which is engaged in transcriptional elongation , revealed a lack of staining at the XY body in Setx+/+ , confirming transcriptional silencing ( Figure 7B ) . In contrast , Pol II staining was visible over XY chromosomes in Setx−/− ( Figure 7B ) . Ubiquitination of histone H2A has been shown to be associated with transcriptional silencing of large unravelled chromatin regions of the XY chromosomes [37] . Because of the continued presence of RNA Pol II on the sex chromosomes in Setx−/− we predicted that ubiquitination of H2A would be defective in Setx−/− . The results in Figure 7C revealed marked localisation of ubi-H2A to the XY body in Setx+/+ spermatocytes . On the other hand the extent of ubi-H2A on the XY body of Setx−/− was much reduced but ubi-H2A was also distributed across the autosomes . These data suggest that senataxin plays a key role in the initial Brca1-dependent stage in MSCI .
This study provides compelling evidence for an essential role for senataxin in spermatogenesis . We showed that senataxin removes R-loops to maintain the integrity of the genome during meiotic recombination and it is also required for effective MSCI . In Setx−/− mutant mice , spermatogenesis was arrested in pachytene stage where R-loop accumulation in cells coincided with apoptosis , resulting in male infertility . Testicular atrophy , depletion of germ cells and sterility are common features of animal models with defects in meiotic proteins such as Spo11 [38] , strand exchange protein Dmc1 [39] , Brca1 [40] and mismatch repair proteins Msh4 , Msh5 , Mlh3 and Mlh1 [41]–[43] . The phenotype in Setx−/− male mice overlaps with but is distinct from that described for these mutant mice . Unlike that for Setx+/+ , where breaks were confined to the XY body in pachytene , breaks were still present in the autosomes as well as the XY body in Setx−/− mice indicating a defect in repair of DNA DSB and consequently a defect in meiotic recombination . This was confirmed by persistence of Rad51 and Dmc1 on autosomes and a failure to detect chiasmata at late meiotic nodules in Setx−/− pachytene cells . Failure to remove Rad51 , as seen in Setx−/− , prevents the completion of meiotic DSB repair . The meiotic phenotype of Setx−/− mice resembles that seen in Brca1Δ11/Δ11p53+/− mice [44] . In that model , chromosome synapsis occurred normally and cells progressed through to pachytene , however no chiasmata were observed [44] . Furthermore , DSBs were not repaired in the correct temporal framework , as demonstrated by persistent γH2AX foci . While the failure to complete meiosis due to persistence of unrepaired DSB and lack of cross-overs is common to the Setx−/− and Brca1 mutants , one obvious difference is diminished numbers of Rad51 foci and normal localisation of Dmc1 in the Brca1 mutant [44] This could be accounted for by the interaction of Rad51 with Brca1 which would be disrupted in the Brca1 mutant . Sen1 , the yeast homolog of senataxin , restricts co-transcriptionally formed R-loops which accumulate in sen1-1 mutant in a transcription-dependent manner [10] . Furthermore , Mischo et al [10] observed a genetic interaction between sen1 and various factors involved in HR such as rad50 , mre11 , sgs1 and rad52 and concluded that sen1 plays a pivotal role in preventing genomic instability by transcription-mediated recombination [10] . More recently , Skourti-Stathaki et al [7] provided evidence that senataxin , like sen1 , resolves R-loop structures formed at transcriptional pause sites to ensure effective transcription termination . In vivo accumulation of R-loops was evident in Setx−/− seminiferous tubules and in pachytene stage spermatocytes , supporting a role for senataxin in resolving such structures . Furthermore , partial co-localisation between R-loops and TUNEL staining in Setx−/− germ cells indicates that this accumulation may contribute to cell death ( Figure 4A–4E ) . Transcriptional R-loop formation in eukaryotes is highly correlated with DNA recombination and/or impairment of genome stability , indicating an inherent impact of R-looping on the integrity of the genome [9] , [45] . R-loop formation is capable of inducing hyper-recombination and/or hypermutation phenotypes in eukaryotes [8] , [4] . Recently , THO mutants from S . cerevisiae and C . elegans showed defective meiosis and an impairment of premeiotic replication as well as DNA-damage accumulation [46] . Gan et al . [47] have shown that R-loop formation impairs DNA replication which is responsible for the deleterious effects of those structures on genome stability . More recently , Alzu et al . [12] provided evidence that when transcription and replication collide Sen1 displaces R-loops to counter recombinogenic events . Thus , R-loop formation may be an intrinsic threat to genome integrity throughout evolution and species have evolved a variety of co-transcriptional processes to prevent the formation of these structures [46] . Senataxin represents a novel factor that minimizes the impact of R-loops that arise as part of normal transcription processes [7] , [11] and/or DNA-damage-induced transcription stalling [48] . In the case of Setx−/− spermatocytes , accumulation of R-loops occurs throughout leptotene and zygotene at a time when DNA DSB are being repaired by crossing over and other mechanisms . Consequently it is likely that R-loops collide with Holiday junctions and interfere with resolution of DNA DSB and thus meiotic recombination . Furthermore , the accumulation of R-loops throughout the chromatin would also affect the repair of DNA DSB that are repaired through non crossover mechanisms . Senataxin specifically localises to the XY body in pachytene stage partially co-localising with Brca1 , MDC1 and ATR suggesting that it might have a role in MSCI . While Brca1 lines the unsynapsed axes of the XY chromosomes , senataxin is associated with these chromosomes but also has a more diffused distribution on chromatin . Prior to MSCI initiation , Brca1 is targeted to the unsynapsed axial elements of the X and Y chromosomes where it remains [22] . It subsequently recruits ATR to the axial elements where it phosphorylates H2AX . In agreement with these findings we provided additional evidence for a direct endogenous interaction between Brca1 and ATR in situ over the XY body ( Figure 5H ) . It seems likely that recruitment of senataxin is Brca1-dependent since senataxin did not localize to the XY chromosome in Brca1Δ11/Δ11 p53+/− mutant mice even though the smaller protein mutant Brca1 ( lacking exon 11 ) lined the axes of these chromosomes . We did not detect a direct endogenous interaction between senataxin and Brca1 , suggesting that the Brca1-dependent localisation of senataxin to the XY chromosome may be indirect and mediated by other DNA damage response proteins involved in MSCI . On the other hand , Brca1 still localised to the sex chromosomes in Setx−/− mutant mice . However , this was incomplete since it was excluded from part of the XY structure in pachytene . This is consistent with a recent report that the X and Y chromosomes have different patterns of incorporation and release of recombination/repair and MSCI-related factors during different stages of meiosis [31] . In that study , they provided evidence that some MSCI steps are triggered much later on the Y chromosome than the X chromosome . Comparison with these results suggests that Brca1 has not localised to the Y chromosome in Setx−/− spermatocytes due to a block earlier in pachytene . Once Brca1 localises to the axial elements of the sex chromosomes it recruits ATR which phosphorylates H2AX and it subsequently diffuses out into XY chromatin to trigger MSCI [22] . In Brca1 mutant cells , these proteins do not localise to the surrounding chromatin [44] . Loss of senataxin did not change the overall distribution of Brca1 on the XY chromosomes but ATR is no longer diffusely distributed and is instead retained on the axial elements of the XY chromosomes , similar to Brca1 . The pattern of ATR staining in Setx−/− suggests that meiosis only proceeds from early to mid pachytene in these mice and that ATR re-localisation is dependent on senataxin ( Figure 7D ) . Recent data show that in the absence of MDC1 , the diffusion of ATR , γH2AX and TopBP1 into XY chromatin is defective [33] . In the absence of senataxin the failure of ATR to diffuse from the axial elements to XY chromatin might be explained by defective MDC1 function . Our observation that MDC1 fails to localize fully to the XY body is consistent with this . During leptotene and zygotene , the sex chromosomes are transcriptionally active [49] . However , at pachytene stage when meiotic synapsis is complete , the sex chromosomes are rapidly silenced and compartmentalized into a peripheral nuclear subdomain , the sex body [50] . MSCI then persists throughout the rest of pachytene and diplotene [49] . The second wave of phosphorylation only occurs on the chromatin of the sex chromosomes and is absolutely essential for MSCI [24] , [50] . This second wave of H2AX phosphorylation occurs in Setx−/− but breaks were still evident in the autosomes . This , together with failure to form chiasmata points to regions of asynapsis in Setx−/− autosomes . Extensive asynapsis has been shown to result in MSCI failure and pachytene stage IV apoptosis [51] , [52] . Expression analysis of X- and Y-linked genes revealed defective MSCI in Setx−/− . Furthermore , in contrast to that for Setx+/+ mice RNA Pol II staining was still visible on sex chromosomes in Setx−/− , consistent with continuing transcription . A reduction in ubi-H2A on the XY body of Setx−/− is also consistent with a failure of MSCI . While no ubi-H2A was observed associated with autosomes in Setx+/+ , in keeping with transcriptional reactivation at pachytene stage , significant staining is seen on Setx−/− autosomes pointing to widespread abnormalities in transcriptional activity in these cells . This is in agreement with the increased R-loop staining observed at pachytene stage in Setx−/− cells . Recent results show that Brca1 preferentially mono-ubiquitinates H2A at satellite DNA regions and Brca1 deficiency impairs the integrity of constitutive heterochromatin causing disruption of gene silencing very likely through loss of ubi-H2A [53] . The evidence presented here suggests that R-loops accumulate in Setx deficient , actively transcribing cells in the presence of unrepaired DNA DSB . This supports a role for senataxin in resolving R-loops ( Figure 7D ) . However , we previously showed that senataxin has a broader role in RNA processing since splicing efficiency , alternate splicing and transcription termination are abnormal in AOA2 cells [6] . It is unlikely that the extent of R-loops accumulation in actively transcribing/replicating spermatocytes will be duplicated in post-mitotic cells , such as Purkinje Cells . Neither DNA replication nor recombination is taking place in neuronal cells thus avoiding collisions with the transcriptional apparatus . Indeed , we were not able to detect R-loops in the cerebellum and brain of Setx−/− mice ( unpublished data ) . These data suggest that the major clinical neurodegenerative phenotype seen in AOA2 patients is more likely to be due to a more general defect in RNA processing leading to reduced transcription fidelity rather than a failure to resolve R-loops . Altogether , these findings reveal a complex and coordinated network between transcription , RNA processing , and DNA repair pathways ( Figure 7D ) , and support the emerging importance of RNA processing factors such as senataxin in the DNA damage response .
All animal work and experiments have been approved by The Queensland Institute of Medical Research Animal Ethics Committee To disrupt the Setx gene a highly-effective recombineering approach was employed [54] . Briefly , two cassettes , a loxP-F3-PGK-EM7-Neo-F3 ( Neo ) cassette was inserted into a BAC clone ( RP23-389D11 ) , Children's Hospital Oakland Research Institute corresponding to mouse chromosome 2 and covering the Setx genomic sequence . The Neo cassette which provides positive selection in ES cells was flanked by a 5′ homology arm of 6 . 8 kb and a 3′homology arm of 3 kb . ES cells were then transfected with the linearized targeting vector and selected with 150 µg/ml of G418 . Successful recombinant ES clones were determined by Southern blotting with a specific probe and PCR genotyping , and targeted cells ( +neo ) were subsequently micro-injected into C57BL6/129Sv mice blastocysts to generate chimeras . Excision of the Neo cassette was obtained by crossing the chimeras with a Cre deleter stain to generate Setx−/− mice containing only a LoxP site . The mice were weaned at 21 days post-partum and ear clipped for identification . Genotyping was carried out by PCR on genomic DNA isolated from tail tips . Tail tips were lysed in directPCR Lysis eagent ( Qiagen , USA ) as recommended by the manufacturer . The primers used were In3F: 5′-TTTAAGGAACAGTGCTGC-3′ , In3R: 5′-ATGAAGCAGGTAGGATT-3′ and LoxPR: 5′-CGAAGTTATATTAAGGGT-3′ . PCR Cycling conditions were as follows: 35 cycles , denaturation at 95°C for 30 sec , annealing at 49°C for 30 sec , extension at 72°C for 1 min , with a final cycle and extension of 7 min at 72°C . Two PCR products were generated , a wild-type PCR product of 600 bp , and the targeted PCR product of 339 bp . PCR products were electrophoresed at 100 V for 30 min on 2% TAE Agarose ( Boehringer Mannheim , Amresco , Lewes , UK ) stained with Ethidium bromide and visualised with UV transillumination using a GelDoc XR ( Biorad Laboratories Inc , UK ) . Testes from adult ( 35-day-old ) , 4 months , 8 months and 12 month-old mice were collected and fixed in PBS buffered 10% formalin , embedded in paraffin block and sectioned at 4 µm . Sections were stained with Hematoxylin and Eosin ( H&E ) and Toluidine blue . Slides were examined under light microscope and then scanned using Scanscope CS system ( Aperio Technologies , Vista , USA ) . Images corresponding to ×10 and ×20 magnification were captured and assembled into Adobe Photoshop 7 ( Adobe Systems Inc , USA ) . Total RNA was isolated from 35-day-old wild type and knockout mice testes using the RNeasy mini kit ( Qiagen , USA ) according to the manufacturer's protocol . RNA concentrations were determined by UV spectrophotometry using a Nanodrop ND-1000 ( Thermo scientific , USA ) . cDNA was made from 5 µg of purified RNA . Briefly , RNA was mixed with 1 µl of random hexamer primers ( Bio-Rad Laboratories Inc . USA ) , 1 µl of 10 mM dNTP mix and DEPC-treated water up to a 14 µl volume . The mixture was heat-denatured at 65°C for 5 min . 4 µl of First Strand buffer ( Invitrogen , USA ) , 1 µl of 1 mM DTT , 1 µl of RNAaseIN ( Promega , USA ) , and 1 µl of SuperScriptIII reverse transcriptase enzyme ( Invitrogen , USA ) was added to the mixture , and incubated for 10 min at 25°C , then 60 min at 50°C , 15 min at 70°C , and chilled on ice . 1 µl of RNAse H was subsequently added to each tube and incubated for 20 min at 37°C , followed by heat inactivation for 20 min at 65°C . The resulting cDNA were stored at −20°C prior to use . Gene expression analysis was performed by PCR in a 2720 Thermal Cycler ( Applied Biosystem , USA ) . Reactions ( 25 µl ) contained 14 . 5 µl of sterile water , 50 ng of cDNA template , 1× PCR Buffer II ( Roche , Switzerland ) , 2 . 5 mM MgCl2 ( Roche , Switzerland ) , 20 µM dNTPs , 1 µM of each primer , and 5 µl of AmpliTaq Gold DNA Polymerase ( Roche , Switzerland ) . The primer pairs used for gene expression analysis are described in Table S1 . Amplification was for 30 cycles and cycling conditions were as follows: denaturation for 5 min at 95°C for 30 sec , annealing at 55°C for 30 sec , elongation for 1 min at 72°C followed by a final extension step of 7 min at 72°C . PCR reactions were separated on 2% TAE agarose gels and visualised as above . Testes from 35 day-old mice were collected and ground with a pestle to disrupt their structure and lysed for 1 h at 4°C on a rotating wheel with lysis buffer ( 50 mM Tris-HCl pH 7 . 5 , 50 mM β-glycerophosphate , 150 mM NaCl , 10% glycerol , 1% Tween 20 , 1 mM PMSF , 5 mM DTT and 1× EDTA-free Complete Protease inhibitor ( Roche , Switzerland ) . Cellular debris were pelleted by centrifugation at 16 , 100×g at 4°C for 10 min , and protein concentration was determined using Lowry Assay ( Bio-Rad Laboratories , Inc , USA ) . 2 mg of total cell extract were pre-cleared with 50 µl of a mixture of 1∶30 protein G+A beads ( Millipore , Germany ) for 3 hours at 4°C on a rotating wheel . Extract were centrifuged for 5 min at 2000×g , beads were removed , and 20 µg of anti-human senataxin antibody ( Ab1/Ab-3 ) was added to the extract . Extracts and antibody were incubated overnight at 4°C on a rotating wheel to allow binding of the antibody to mouse senataxin . The next day , 50 µl of protein G+A beads were added to the extract and incubated for 1 h at 4°C on a rotating wheel . The immunoprecipitate was subsequently washed 3 times with lysis buffer and the beads were resuspended in gel loading buffer and separated on 5% SDS-PAGE at constant current ( 20 mA per gel ) for 1 . 5 h . Once separated , proteins were transferred onto a nitrocellulose membrane ( Hybond C , Amersham ) for 1 h at 4°C with constant voltage ( 100 Volts ) . Immunoblotting with anti-senataxin ( Ab-1 ) antibody was performed using standard procedure as previously described [5] . All spreads were made from testes collected from adult 35-day-old mice or at day 16 , 20 and 22 post partum . Briefly , testes were decapsulated , finely chopped and rinsed in GIBCO DME medium ( Invitrogen , USA ) . Large clumps were removed by centrifugation at 6780×g for 5 min at room temperature . The remaining supernatant was centrifuged to pellet the cell suspension and mixed with 0 . 1M sucrose and spread onto glass slides pre-wetted with 1% paraformaldehyde and 0 . 1% Triton X-100 in PBS . Cells were fixed on the glass slides for 2 h at room temperature . The slides were subsequently washed with PBS and air-dried in the presence of a wetting agent , 1∶250 Kodak Photo-Flo 200 ( Kodak professional , USA ) . Once dried , spreads were stored at −80°C . For immunostaining , slides , were rehydrated in dH20 , and blocked in blocking buffer ( 0 . 2% BSA , 0 . 2% gelatine in PBS ) for 30 min at room temperature . Spreads were incubated with primary antibodies overnight at 4°C in a humidified chamber . Primary antibodies used included anti-SCP3 ( 1∶100 , NB300-230 , Novus Biologicals ) , anti-SCP1-DyeLight conjugated ( 1∶100 , NB300-2201R , Novus Biologicals ) , anti-γH2AX ( 1∶100 , Y-P1016 , Millipore ) , anti-Rad51 ( 1∶100 , SC-33626 , Santa Cruz Biotechnology ) , anti-Dmc1 ( 1∶50 , 2H12/4 , Sapphire Bioscience ) , anti-Mlh1 ( 1∶10 , G168-15 , Sapphire Bioscience ) , anti-ATR ( 1∶100 , SC-1887 , Santa Cruz Biotechnology ) , anti-senataxin ( 1∶100 , Ab-1 , [5] ) , anti-R-loop ( 1∶100 , S9 . 6 ) , anti-RNA Pol II ( phospho S2 ) ( 1∶100 , H5 , ab24758 , Abcam ) , anti-mouse Brca1 ( 1∶300 , David Livingston ) , anti-ubi-H2A ( 1∶100 , Clone E6C5 , Millipore ) . Slides were subsequently washes 4 times for 3 min each in PBS on a rocker , and probed with the appropriate Alexa-Dye488 or Alexa-Dye594-conjugated secondary antibodies ( 1∶250 , Invitrogen , Molecular Probes , USA ) . Slides were washed again 4 times for 3 min each in PBS . DNA was stained with Hoechst 33342 ( 1∶10 , 000 ) for 10 min at room temperature , and slides were mounted in Celvol 603 medium . Images were captured at room temperature using a digital camera ( AxioCam Mrm , Carl Zeiss Microimaging Inc . , Germany ) attached to a fluorescent microscope ( Axioskop 2 mot plus , Carl Zeiss Microimaging Inc . , Germany ) and the AxioVision 4 . 8 software ( Carl Zeiss , Microimaging Inc . Germany ) . The objective employed was a 63× Zeiss Plan Apochromat 1 , 4 Oil DIC ( Carl Zeiss , Germany ) . Images were subsequently assembled in Adobe Photoshop 7 ( Adobe Systems Inc , USA ) , and contrast and brightness were adjusted on the whole image panel at the same time . Terminal deoxynucleotidyl transferase dUTP nick end labeling ( TUNEL ) is a method for detecting DNA fragmentation by labeling the terminal end of nucleic acids . TUNEL is a common method for detecting DNA fragmentation that results from apoptotic signaling cascades . The assay relies on the presence of nicks in the DNA which can be identified by terminal deoxynucleotidyl transferase ( TdT ) , an enzyme that will catalyze the addition of Fluorescein-labeled dUTP . Paraffin sections were dewaxed and rehydrated with Shandon Varistain Gemini ES ( Thermo Scientific , USA ) . TUNEL assay was performed using the Fluorescence in situ Cell Death Detection Kit ( Roche , Switzerland ) following the manufacturer's instructions . Slides were visualised under a fluorescent microscope and images were captured as previously described . The objective employed was a Zeiss Plan Neofluar ×10/0 . 30 ( ×10 magnification ) . For double staining , TUNEL was carried out first followed by immunostaining as described below . Slides with tissue sections were dewaxed and enzymatic antigen retrieval was performed by incubating the sections with 1∶10 Trypsin dilution in PBS for 20 min at 37°C . Slides were washed 3 times for 5 min with PBS at room temperature for 5 min each . Tissues sections were blocked in ( 20% FCS , 2% BSA , 0 . 2% Triton X-100 ) for 1 h at room temperature . Slides were incubated with anti-R-loop ( 1∶100 , S9 . 6 ) [29] or anti-γH2AX ( 1∶100 , Y-P1016 , Millipore ) antibody overnight at 4°C in a humidified chamber . Slides were washed 5 times with 1× PBS containing 0 . 5% Triton X-100 for 5 min each at room temperature . Alexa-Dye488 or Alexa-Dye594-conjugated secondary antibody was added for 1 h at 37°C in a humidified chamber . Subsequently , slides were washed 3 times as before and Hoechst 33342 was added for 10 min to staining nuclei . Slides were finally washed twice and glass coverslips were mounted for imaging . Imaging was performed as described above . Confirmation of R-loop specific staining was obtained by pre-treating Setx−/− testes sections with RNAse H ( New England Biolads , USA ) . To investigate a possible interaction between Brca1 , ATR and senataxin we employed in situ Proximity Ligation Assay ( PLA ) ( Duolink , Olink Bioscience , Uppsala , Sweden ) on wild type ( Setx+/+ ) spermatocytes spreads . PLA allows the monitoring of protein interactions and modifications with high specificity and sensitivity . Protein targets can be readily detected and localized with single molecule resolution and objectively quantified in unmodified cells and tissues . Utilizing only a few cells , sub-cellular events , such as transient or weak interactions are revealed in situ . Two primary antibodies raised in different species recognize the target antigens of interest . Species-specific secondary antibodies , called PLA probes , each with a unique short DNA strand attached to it , bind to the primary antibodies . When the PLA probes are in close proximity , the DNA strands can interact through a subsequent addition of two other circle-forming DNA oligonucleotides . After joining of the two added oligonucleotides by enzymatic ligation , they are amplified via rolling circle amplification using a polymerase . After the amplification reaction , several-hundredfold replication of the DNA circle has occurred , labeled complementary oligonucleotide probes highlight the product . The resulting high concentration of fluorescence in each single-molecule amplification product is visible as a distinct bright spot when viewed with a fluorescence microscope . The assay was performed according to the manufacturer's protocol using rabbit anti-mouse Brca1 ( 1∶200 , David Livingston ) , sheep anti-senataxin ( 1∶200 , Ab-1 ) and goat anti-ATR antibody ( 1∶100 , SC-1887 , Santa Cruz Biotechnology ) antibodies and the corresponding anti-goat PLA Probe MINUS and anti-rabbit PLA probe PLUS . Identification of pachytene stage spermatocytes was determined by counterstaining with SCP3 antibody . PLA was also carried out on Setx−/− spermatocytes spreads as a negative control . Slide mounting and imaging was performed as described above . | Ataxia with oculomotor apraxia type 2 ( AOA2 ) caused by a defect in the gene Setx ( coding for senataxin ) is part of a subgroup of autosomal recessive ataxias characterized by defects in genes responsible for the recognition and/or repair of damage in DNA . Cells from these patients are characterized by oxidative stress and are defective in RNA processing and termination of transcription . Recent data suggest that senataxin is involved in coordinating events between DNA replication forks and ongoing transcription . To further understand the role of senataxin , we disrupted the Setx gene in mice and demonstrated its essential role in spermatogenesis during meiotic recombination and in meiotic sex chromosome inactivation ( MSCI ) . In the absence of senataxin , DNA double-strand breaks persist , RNA:DNA hybrids ( R-loops ) accumulate , and homologous recombination is disrupted . Senataxin localised to the XY chromosomes during pachytene . This was dependent on Brca1 , which functions early in MSCI to recruit DNA damage response proteins to the XY body . In the absence of senataxin , there was incomplete accumulation of DNA damage response proteins on the XY chromosomes and no MDC1-dependent diffusion of ATR to the broader XY chromatin . The end result was a defect in MSCI , apoptosis , and a failure to complete meiosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"model",
"organisms",
"genetics",
"biology",
"molecular",
"cell",
"biology"
] | 2013 | Senataxin Plays an Essential Role with DNA Damage Response Proteins in Meiotic Recombination and Gene Silencing |
Plants under pathogen attack produce high levels of ethylene , which plays important roles in plant immunity . Previously , we reported the involvement of ACS2 and ACS6 , two Type I ACS isoforms , in Botrytis cinerea–induced ethylene biosynthesis and their regulation at the protein stability level by MPK3 and MPK6 , two Arabidopsis pathogen-responsive mitogen-activated protein kinases ( MAPKs ) . The residual ethylene induction in the acs2/acs6 double mutant suggests the involvement of additional ACS isoforms . It is also known that a subset of ACS genes , including ACS6 , is transcriptionally induced in plants under stress or pathogen attack . However , the importance of ACS gene activation and the regulatory mechanism ( s ) are not clear . In this report , we demonstrate using genetic analysis that ACS7 and ACS11 , two Type III ACS isoforms , and ACS8 , a Type II ACS isoform , also contribute to the B . cinerea–induced ethylene production . In addition to post-translational regulation , transcriptional activation of the ACS genes also plays a critical role in sustaining high levels of ethylene induction . Interestingly , MPK3 and MPK6 not only control the stability of ACS2 and ACS6 proteins via direct protein phosphorylation but also regulate the expression of ACS2 and ACS6 genes . WRKY33 , another MPK3/MPK6 substrate , is involved in the MPK3/MPK6-induced ACS2/ACS6 gene expression based on genetic analyses . Furthermore , chromatin-immunoprecipitation assay reveals the direct binding of WRKY33 to the W-boxes in the promoters of ACS2 and ACS6 genes in vivo , suggesting that WRKY33 is directly involved in the activation of ACS2 and ACS6 expression downstream of MPK3/MPK6 cascade in response to pathogen invasion . Regulation of ACS activity by MPK3/MPK6 at both transcriptional and protein stability levels plays a key role in determining the kinetics and magnitude of ethylene induction .
The gaseous phytohormone ethylene profoundly impacts plant growth , development , and response to environmental stimuli [1]–[7] . Studies from a number of labs have defined a signaling pathway—from ethylene receptors to downstream signaling components to transcription factors—that alters gene expression and leads to ethylene-induced phenotypes ( reviewed in [1] , [2] , [5] , [8] , [9] ) . Ethylene-regulated responses are suppressed in the absence of ethylene , and such suppression is released upon plant sensing of ethylene . As a result , all ethylene-regulated processes begin with the induction of ethylene biosynthesis [10] . Plants under stress , including wounding , flooding , drought , osmotic shock , ozone , and pathogen/insect invasion , produce elevated levels of ethylene [1] , [6] , [7] , [11] . For this reason , ethylene is also known as a plant stress hormone . The biosynthetic pathway of ethylene has been fully elucidated for over two decades . Two enzymatic steps are unique to ethylene biosynthesis: conversion of S-adenosyl-methionine ( SAM ) , a common metabolic precursor , to 1-amino-cyclopropane-1-carboxylic acid ( ACC ) by ACC synthase ( ACS ) and oxidative cleavage of ACC to form ethylene by ACC oxidase ( ACO ) [1] , [4] , [12] . ACS activity is very low in tissues that do not produce a large amount of ethylene and is enhanced under conditions that promote ethylene formation [1] , [4] , [12]–[14] . In contrast , ACO is constitutively present in most vegetative tissues . As a result , ACS is believed to be the committing and generally rate-limiting enzyme in ethylene biosynthesis . ACS is encoded by a small gene family in plants . In Arabidopsis , there are nine ACS members . Based on the presence/absence of phosphorylation sites in their C-termini , ACS isoforms are classified into three types [15] . Type I ACS isoforms , which include Arabidopsis ACS1 , ACS2 , and ACS6 , have phosphorylation sites by both mitogen-activated protein kinases ( MAPKs ) and calcium-dependent protein kinases ( CDPKs ) [16] , [17] . Type II ACS isoforms , which include Arabidopsis ACS4 , ACS5 , ACS8 , and ACS9 , only have putative CDPK phosphorylation sites . In contrast , Type III ACS isoforms have shorter C-terminal extension and lack both phosphorylation sites . ACS7 and ACS11 are the two Type III ACS isoforms in Arabidopsis . ACS1 has a short deletion with the highly conserved tripeptide Thr-Asn-Pro ( TNP ) missing . It is enzymatically inactive as a homodimer , but can form functional heterodimers with other Type I isoforms and may contribute to ethylene biosynthesis [18] , [19] . ACS isoforms show cell- and tissue-specific expression and are developmentally regulated . In addition , expression of some members is highly responsive to extracellular stimuli [1] , [20] . More recent studies have highlighted the importance of ACS protein stability regulation by protein phosphorylation and dephosphorylation . MAPK cascades are signaling modules downstream of sensors/receptors that transduce extracellular stimuli into intracellular responses in eukaryotes . A basic MAPK cascade is composed of three interconnected kinases . MAPKs function at the bottom of the three-kinase cascade and are activated by MAPK kinases ( MAPKKs ) through phosphorylation on the Thr and Tyr residues in their activation motif between the kinase subdomain VII and VIII . The activity of MAPKKs is , in turn , regulated by MAPKK kinases ( MAPKKKs ) via phosphorylation of two Ser/Thr residues in the activation loop of MAPKKs . MAPKKKs receive signals from upstream receptors/sensors , most of the time indirectly with additional components involved [21] , [22] . The outputs of a MAPK cascade are dependent on the substrates of the MAPK ( s ) in the cascade . A subset of MAPKs in plants , represented by tobacco SIPK/Ntf4/WIPK and Arabidopsis MPK3/MPK6 , is activated under various stress conditions that elevate ethylene production ( reviewed in [21] , [23]–[26] ) . A gain-of-function analysis in tobacco revealed that activation of SIPK/WIPK induces high levels of ethylene production [27] . More detailed analyses in Arabidopsis have demonstrated that ACS2 and ACS6 , two Type I ACS isoforms , are substrates of MPK3 and MPK6 [16] , [28] . Phosphorylation of ACS2/ACS6 by MPK3 and MPK6 stabilizes the ACS protein in vivo , resulting in increases in cellular ACS activity and in ethylene production . The degradation machinery targets the C-terminal , non-catalytic domain of ACS6 and possibly ACS2 because of their sequence similarity [29] . Phosphorylation of ACS6 introduces negative charges to its C-terminus , which reduces the turnover of ACS6 by the ubiquitin-proteasome degradation machinery . In addition to protein phosphorylation , protein dephosphorylation also plays critical role in ACS stability regulation . Recently , it was demonstrated that protein phosphatase 2A dephosphorylates ACS2/ACS6 and destabilizes them , a critical process that counteracts with MAPK phosphorylation [30] . Members of the Type II group , including ACS5 and ACS9 , are also regulated at protein stability levels , possibly by protein phosphorylation as well [15] , [31]–[33] . However , the kinase ( s ) involved remain unidentified . Because of the complex regulation of ACS protein/activity at multiple levels , many details about the up-regulation of ethylene biosynthesis remain unclear , including the specific ACS isoforms involved in the ethylene induction in response to a specific stimulus , the regulatory pathways that control the expression of ACS genes , and the components involved in the regulation of ACS protein stability . It has been known for decades that a subset of ACS genes , including Arabidopsis ACS6 , is transcriptionally activated in plants under stress or pathogen attack . However , the importance of this transcriptional activation and the underlying regulatory mechanism are not known . Furthermore , ethylene induction by different stimuli exhibits different kinetics and magnitude . The underlying molecular mechanism of such differential induction is also unclear . We are interested in the regulation of ethylene biosynthesis in plants infected by pathogens . ACS2 and ACS6 , two Type I ACS isoforms , are involved in Botrytis-induced ethylene production [28] . The residual levels of ethylene induction in the acs2/acs6 double mutant suggest involvement of additional ACS isoforms . In this study , we investigated ( 1 ) the potential involvement of all ACS isoforms in ethylene induction triggered by B . cinerea infection , ( 2 ) the importance of transcriptional activation of ACS gene expression , ( 3 ) the signaling pathways involved in the ACS gene activation , and ( 4 ) the molecular mechanism underlying the differential kinetics and magnitude of ethylene induction by different stimuli . We found that members in all three ACS groups are involved in pathogen-induced ethylene production , with ACS2 , ACS6 , and ACS7 contributing the most to B . cinerea-induced ethylene production . Based on analyses of an ACS6 knockdown mutant and of conditional gain-of-function ACS6 transgenic lines , we also can conclude that the transcriptional activation of the ACS6 gene plays a critical role in sustaining high levels of ethylene induction . Interestingly , MPK3 and MPK6 not only function in the phosphorylation-induced stabilization of ACS2/ACS6 proteins , but also signal the ACS2 and ACS6 gene activation after B . cinerea infection . WRKY33 , a MPK3/MPK6 substrate that regulates camalexin biosynthesis [34] , is also responsible for turning on ACS2/ACS6 expression downstream of MPK3/MPK6 cascade . WRKY33 binds to the W-boxes in the ACS2/ACS6 promoters in vivo and is directly involved in MPK3/MPK6-induced ACS2/ACS6 gene expression . The duration and magnitude of MPK3/MPK6 activation vary with different stimuli and correlate well with the duration and magnitude of ethylene induction . Regulation of ACS activity at multiple levels by the MPK3/MPK6 cascade is an important mechanism by which the levels/kinetics of ethylene production are regulated during plant stress/defense response .
Our previous research demonstrated involvement of ACS2 and ACS6 in ethylene induction in B . cinerea-infected Arabidopsis [28] . This research also implicated the involvement of additional ACS genes since there was still an approximately 25% residual level of ethylene induction in the acs2/acs6 double mutant . To identify the ACS isoforms involved , we profiled the expression of all nine ACS genes in Arabidopsis infected with B . cinerea . As shown in Figure 1A , transcripts of ACS2 , ACS6 , ACS7 , and ACS8 accumulated approximately 1600 , 200 , 50 , and 1200 fold , respectively , over their basal levels . ACS11 transcript also accumulated about 6 fold . ACS5 and ACS9 transcripts could be reliably detected , but no increases were observed . In contrast , ACS1 and ACS4 transcripts were not detectable . To better assess the potential contribution of each ACS gene to ethylene production , we also calculated their expression levels relative to that of EF1α ( Figure 1B ) . This calculation allowed us to compare the relative levels of expression between different ACS genes . From this dataset , we found that the expression levels of ACS2 , ACS6 , and ACS7 were among the highest . ACS8 and ACS11 had lower levels of expression after induction , while ACS5 and ACS9 expression remained very low . The expression of ACS8 increased more than 1200 fold relative to its basal level ( Figure 1A ) . However , because of its low basal level expression , the induced level of ACS8 transcript was still much lower than those of ACS2 , ACS6 , and ACS7 ( Figure 1B ) . If regulation at other levels is the same , ACS8 is likely a minor contributor despite the high-fold induction . In contrast , levels of ACS7 transcript were considerably elevated ( Figure 1B ) , despite a relatively low fold induction ( Figure 1A ) , a result of a relatively high basal level . Based on these results , we speculated that ACS7 might be a major contributor to ethylene induction after plant sensing of pathogen invasion besides ACS2 and ACS6 . To establish the involvement of ACS7 , we identified two null mutant alleles of ACS7 , acs7-1 ( FLAG_431D05 , in Ws-0 background ) and acs7-2 ( CSHL_ET5768 , in Ler-0 background ) . In both mutant alleles , B . cinerea-induced ethylene production is slightly reduced ( Figure S1 ) , similar to that in the acs2 or acs6 single mutant [28] . This result suggests that ACS7 also contributes to B . cinerea-induced ethylene production . In our previous publications [16] , [28] , we did not assign allele numbers to the acs2 and acs6 mutants . To be consistent with the nomenclature used in Dr . Theologis's lab [35] , the acs6 allele ( Salk_090423 ) was given an allele number of acs6-2 . This allele turned out to be a knockdown mutant ( more discussion later ) . In contrast , the acs6-1 mutant allele ( SALK_025672 ) in the study by Tsuchisaka et al . ( 2009 ) is a null mutant with a T-DNA insertion in the open reading frame ( ORF ) [35] . We failed to identify any plant with a T-DNA insertion when we initially ordered this line from the Arabidopsis Biological Resource Center ( ABRC ) in 2003 . The acs2 and acs7-1 mutant alleles we used [16] , [28] ( Figure S1 ) are the same as the acs2-1 and acs7-1 alleles , respectively , in Tsuchisaka et al . ( 2009 ) report [35] . We then crossed acs7-1 into the acs2-1/acs6-2 double mutant background and identified an acs2-1/acs6-2/acs7-1 triple mutant in the F3 generation . As shown in Figure 2 , only about 10% residual ethylene production was observed in the acs2-1/acs6-2/acs7-1 triple mutant , confirming the importance of ACS7 in B . cinerea-induced ethylene production . Residual ethylene induction in the acs2-1/acs6-2/acs7-1 triple mutant again points to involvement of additional ACS members . To identify them , we utilized the high-order acs mutants generated in Dr . Theologis' lab [35] . We found that acs2-1/acs6-1 seedlings produced a lower level of ethylene than our acs2-1/acs6-2 double mutant after challenged with B . cinerea ( Figure 3 and Figure S2 ) , which is consistent with the knockdown nature of our acs6-2 allele . Additional mutation of ACS4 , ACS5 , and ACS9 genes , either one at a time or all three at once , in the acs2-1/acs6-1 background did not further reduce the ethylene induction . This finding is consistent with our previous conclusion that ACS5 and ACS9 are not involved in the ethylene induction triggered by B . cinerea infection [28] . In contrast , mutation of the ACS7 gene in the acs2-1/acs6-1/acs4-1/acs5-2/acs9-1 background resulted in further reduction in the ethylene induction . In this sextuple mutant ( acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1 ) background , mutation of ACS11 , but not ACS1 , slightly reduced the ethylene production . The very low level of ethylene induction in the acs1-1/acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1/acs11-1 plants also implicates the involvement of ACS8 . In the absence of B . cinerea infection , seedlings of all genotypes produced less than 15 nL ethylene per gram of seedlings within 24 hours , a very low level in comparison to the B . cinerea-induced ethylene production ( Figure S3 ) . From this dataset , we can also conclude that ACS7 contributed the most to the basal level ethylene production . Seedlings without a functional ACS7 gene including the acs1-1/acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1/acs11-1 failed to produce a detectable level of ethylene . As a result , the low-level ethylene production in this octuple acs mutant in response to B . cincerea infection is indeed a contribution of ACS8 gene . In summary , we can conclude that ACS2 and ACS6 are major contributors of ethylene induction and that ACS7 , ACS8 , and ACS11 contribute less , with a total of ∼15% of the ethylene induction in B . cinerea-infected Arabidopsis . Among the minor contributors , the role of ACS7 and ACS8 in B . cinerea-induced ethylene production is clear . In contrast , the contribution of ACS11 is somewhat uncertain because of the conclusion of its involvement is based on a small quantitative difference . One potential mechanism underlying each ACS isoform's contribution to ethylene induction is through the up-regulation of their gene expression ( Figure 1 ) . There is a good correlation between the transcriptional activation of ACS gene expression ( Figure 1 ) and involvement in B . cinerea-induced ethylene production ( Figure 3 ) . Previously , we demonstrated the importance of phosphorylation regulation of ACS2 and ACS6 by MPK3 and MPK6 in Arabidopsis in response to pathogens/pathogen-associated molecular patterns ( PAMPs ) [16] , [28] , [29] . It is well known that the ACS6 gene is highly induced by stress , including wounding and pathogen infection [1] , [4] , [6] , [11] , [36] , [37] . However , the importance of ACS gene activation in pathogen-induced ethylene production remains unclear . After our discovery that phosphorylation of ACS2 and ACS6 proteins by MPK3/MPK6 is required for ACS2/ACS6 protein stabilization and accumulation , we started to explore the potential contribution of ACS gene activation . Theoretically , an increase in ACS transcript levels is likely to increase the rate of de novo ACS protein synthesis , which , in turn , will increase the net ACS protein/activity after MAPK phosphorylation and protein stabilization . To determine whether ethylene induction in the conditional gain-of-function GVG-NtMEK2DD ( DD , for short ) plants [16] , [28] is associated with ACS gene activation , we profiled the expression of ACS genes in DD plants after dexamethasone ( DEX ) treatment . As shown in Figure 4 , the expressions of ACS2 and ACS6 were highly induced . Different from B . cinerea-infected seedlings , no induction in ACS7 , ACS8 , and ACS11 expression levels were observed . In addition , we noticed the kinetics of ACS6 induction in the DD plants were different from that in B . cinerea-infected plants ( Figure 4 versus Figure 1 ) . This difference is likely a result of the synchronous response in DD plants . In contrast , the infection process of B . cinerea is progressive . As more cells sensed the growing hyphae , higher levels of ACS6 transcript were induced at later time points . To confirm that MPK3 and MPK6 are responsible for induction of ACS2 and ACS6 expression levels in the gain-of-function DD plants , we examined the expression of ACS2 and ACS6 in DD/mpk3 and DD/mpk6 plants . As shown in Figure 5 , induction of ACS2 and ACS6 was compromised in both the mpk3 and mpk6 single mutant backgrounds . This finding demonstrates that induction of ACS2 and ACS6 in DD seedlings after DEX treatment is indeed a result of MPK3/MPK6 activation . Based on the fact that MPK3 and MPK6 are highly activated after B . cinerea infection [28] , [38] , we speculate that the MPK3/MPK6 cascade is involved in regulating the B . cinerea-induced ACS2/ACS6 gene expression and that induction of ACS7 , ACS8 , and ACS11 expression is regulated by pathway ( s ) other than MPK3/MPK6 cascade . To provide loss-of-function evidence to support the role of the MPK3/MPK6 cascade in B . cinerea-induced ACS2/ACS6 gene activation , we compared the induction of ACS2 and ACS6 gene expressions in wild type , mpk3 single mutant , mpk6 single mutant , and rescued mpk3/mpk6 double mutant . The rescued mpk3/mpk6 double mutant was obtained by transforming a DEX-inducible promoter-driven MPK6 ( GVG-MPK6 ) into mpk3−/−/mpk6+/− plants . When the T3 mpk3−/−/mpk6+/−/GVG-MPK6+/+ plants began to flower , DEX was sprayed every other day to rescue the embryo lethality of the mpk3−/−/mpk6−/−/GVG-MPK6+/+ zygotes . Progenies with mpk3−/−/mpk6−/−/GVG-MPK6+/+ genotype were called rescued mpk3/mpk6 double mutants [39] , and were used for this experiment . As shown in Figure 6 , B . cinerea-induced ACS2 and ACS6 expressions were little affected in either the mpk3 or mpk6 single mutant . In the rescued mpk3/mpk6 double mutant , the induction of both genes was dramatically reduced , which supports the conclusion that MPK3 and MPK6 regulate expressions of ACS2 and ACS6 based on the gain-of-function analysis . Different from B . cinerea-induced ACS2/ACS6 gene activation , gain-of-function DD-induced ACS2/ACS6 gene activation was compromised in either mpk3 or mpk6 single mutant background ( Figure 5 and Figure 6 ) . There are several potential explanations for this seemingly contradictory observation . First of all , MAPK-phosphorylation regulation of ACS2/ACS6 gene activation will be affected by both the phosphorylation of the downstream transcription factor ( s ) such as WRKY33 ( more discussion below ) , and their dephosphorylation by the unidentified phosphatase ( s ) . It is possible that in the gain-of-function DD plants , both MPK3 and MPK6 are needed to overcome the action of the phosphatase ( s ) to maintain the phosphorylation of there transcription factor ( s ) and the subsequent up-regulation of ACS2/ACS6 expression . In the absence of either MAPK , the signaling strength is below the threshold to counteract the phosphatases and the activation of ACS2/ACS6 expression is severely compromised . It is possible that , in addition to the activation of MPK3/MPK6 cascade , pathogen infection may also inactivate the dephosphorylation process as a mechanism to promote higher levels of ethylene production . In this situation , the absence of only one MAPK may not be sufficient to block ACS2/ACS6 activation . Alternatively , it is possible that the activation of pathways other than MAPK cascade can compensate the weakened MAPK pathway in the single mpk3 or mpk6 mutant , making it necessary to mutate both MPK3 and MPK6 to see the loss-of-function phenotype in response to B . cinerea infection . Similar phenomenon was also observed in MPK3/MPK6-mediated camalexin induction in response to B . cinerea infection [38] . WRKY33 is a substrate of MPK3/MPK6 in regulating the pathogen-induced phytoalexin biosynthesis [34] . WRKY33 functions as a transcriptional activator downstream of MPK3 and MPK6 in promoting the expression of camalexin biosynthetic genes . To determine whether WRKY33 also is involved in activation of ACS2 and ACS6 genes downstream of the MPK3/MPK6 cascade , we quantified the expression of these two genes in DD and DD/wrky33 plants . As shown in Figure 7B , the induction of ACS2 and ACS6 mRNA by the gain-of-function DD transgene was compromised in wrky33 mutant background . Associated with this , the induction of ethylene biosynthesis was mostly inhibited ( Figure 7A ) . Previously , we showed that DD protein induction and MPK3/MPK6 activation in DD/wrky33 plants are indistinguishable from those in DD plants [34] , which strongly supports the conclusion that WRKY33 also functions downstream of MPK3/MPK6 in promoting the expression of ACS2 and ACS6 genes . The low residual levels of ACS2 and ACS6 gene activation is likely to be a result of other WRKY transcription factors that can partially substitute for the loss of WRKY33 . In contrast to the gain-of-function DD plants , mutation of the WRKY33 gene had a minor impact on ethylene induction triggered by B . cinerea infection . As shown in Figure 8A , we observed only about a 20% decrease in ethylene production in both alleles of the wrky33 mutant . A comparison of ACS2 and ACS6 gene expressions in the wild type and in the wrky33 mutant revealed that about one-third of the induction in ACS2/ACS6 expression remained in the wrky33 mutant ( Figure 8B ) . This suggests that ACS2 and ACS6 still could be partially activated in the absence of WRKY33 . Because of the low residual ACS2/ACS6 gene activation in the mpk3/mpk6 mutant ( Figure 6 ) , we speculate that the residual levels seen in the wrky33 mutant are MPK3/MPK6-dependent but WRKY33-independent , again pointing to additional transcription factors , possibly WRKY33 homologs that partially replace WRKY33 in its absence . No major reductions were observed in the induction of ACS7 , ACS8 , and ACS11 expression in wrky33 infected with B . cinerea ( Figure S4 ) , which is consistent with the conclusion that their activation is MPK3/MPK6 and WRKY33 independent . The normal activation of ACS7 , ACS8 , and ACS11 expressions , together with the residual level of ACS2/ACS6 gene activation and protein phosphorylation stabilization , may explain the observation that the induction of ethylene in the wrky33 mutant was reduced by only about 20% after B . cinerea infection ( Figure 8A ) . In contrast , the wrky33 mutation almost completely blocked induction of ethylene biosynthesis in the gain-of-function DD plants ( Figure 7A ) . B . cinerea can activate multiple signaling pathways in plants . It is possible that pathway ( s ) other than MPK3/MPK6 cascade are able to partially compensate the loss of WRKY33 . It is known that pathogen infection induces a large number of WRKY genes [40] , [41] , some of which might be able to partially compensate the loss of WRKY33 gene in activating the expression of ACS2/ACS6 . Genetic analysis revealed that WRKY33 is essential for gain-of-function MPK3/MPK6-induced ACS2/ACS6 gene expression ( Figure 7B ) . Examination of the ACS2 and ACS6 promoters revealed the presence of eight and seven W-boxes , respectively ( Figure 9A ) . To further substantiate the role of WRKY33 in the activation of ACS2 and ACS6 gene expression , we performed chromatin immunoprecipitation-quantitative PCR ( ChIP-qPCR ) analysis to determine whether ACS2 and ACS6 genes are direct targets of the WRKY33 transcription factor . For this experiment , we used DD/wrky33 mutant plants complemented with a 35S promoter-driven WRKY33 transgene , which contains a four-copy myc epitope tag at the N-terminus ( DD/wrky33/35S:4myc-WRKY33 ) [34] . The presence of the myc tag allowed us to immunoprecipitate the WRKY33-DNA complex by using a commercial anti-myc antibody . As shown in Figure 9B , immunoprecipitation with the anti-myc antibody greatly enriched ACS2 and ACS6 promoter regions containing the W-boxes . In contrast , the IgG control antibody failed to enrich either gene promoter . This result demonstrates that WRKY33 directly binds to the promoters of ACS2 and ACS6 in vivo , suggesting that WRKY33 is the transcription factor downstream of the MPK3/MPK6 cascade involved in the activation of ACS2 and ACS6 expression . To provide further direct evidence in support of the role of ACS6 gene activation in B . cinerea-induced ethylene production , we transformed a DEX-inducible promoter-driven ACS6 ( GVG-ACS6 ) construct into the acs2-1/acs6-2/acs7-1 mutant background and then compared the ethylene induction in acs2-1/acs6-2/acs7-1/GVG-ACS6 plants with and without DEX treatment . Two independent lines ( #5 and #12 ) with different levels of ACS6 transgene induction after DEX treatment were used for this experiment to establish a correlation between the levels of ACS6 gene expression and the levels of ethylene induction . As shown in Figure 10A , without DEX treatment , both GVG-ACS6 transgenic lines produced about the same levels of ethylene after B . cinerea treatment in comparison to the acs2-1/acs6-2/acs7-1 triple mutant . In the presence of DEX , which induced ACS6 expression ( Figure 10B ) , the ethylene production was greatly enhanced . The higher level of ACS6 induction in line #5 in the presence of DEX correlated with a higher level of ethylene induction than that in line #12 . Furthermore , ethylene induction in Line #5 was higher than that in the wild type , indicating that transgene induction after DEX treatment not only complements the loss of ACS6 , but also compensates the loss of ACS2 and ACS7 genes . In the absence of B . cinerea infection , DEX treatment only slightly elevated the ethylene production ( Figure S5 ) to a level similar to the basal level ethylene production of Col-0 ( ∼10 nL/g FW in 24 hrs ) . This low-level ethylene production is likely a result of high-level ACS6 gene induction after DEX treatment ( and associated higher-level of de novo ACS6 protein synthesis ) in combination with the basal level activity of MPK6 , which can phosphorylate and stabilize ACS6 protein . MPK6 has very low basal activity even in the absence of stress/pathogen infection [16] . This is consistent with our previous conclusion that the overexpression of ACS6 gene in the absence of MPK3/MPK6 activation is not sufficient to induce ethylene production due to the lack of phosphorylation stabilization [16] . As a result , we can conclude that the high level of ethylene production seen in acs2-1/acs6-2/acs7-1/GVG-ACS6 lines after DEX and B . cinerea treatment is a combination of high level of gene expression ( as a result of DEX treatment ) , and phosphorylation stabilization due to MPK3/MPK6 activation by B . cinerea infection . Our attempts to identify the T-DNA insertion line in the coding region of the ACS6 gene ( SALK_025672 , acs6-1 ) failed to reveal a true mutant plant from the seeds received . As a result , we have been using the SALK_090423 line ( acs6-2 ) , which has a T-DNA insertion 170 bp upstream of the ATG start codon ( Figure 9A ) [16] , [28] . In the past , we routinely used the double ΔCt method to quantify gene expression in real-time PCR analysis , which indicated the acs6-2 mutant allele as a knockout mutant ( Figure 11B , upper panel ) . However , a more careful analysis performed in this study revealed that it is actually a knockdown mutant with an elevated basal level expression . As shown in Figure 11B ( lower panel ) , acs6-2 seedlings showed a higher basal level of ACS6 expression , but no increase in its transcripts was detected after B . cinerea infection . In contrast , no transcript was detected in the acs6-1 mutant before and after B . cinerea infection . Side-by-side comparison demonstrated that ethylene production levels in acs6-2 and acs6-1 single mutants after B . cinerea inoculation were similar ( Figure 11A ) . In contrast , acs2-1/acs6-2 seedlings produce higher levels of ethylene than acs2-1/acs6-1 ( both have the same acs2 mutant allele ) ( Figure S2 ) . The observable difference between acs6-2 and acs6-1 alleles in the acs2-1 mutant background could be due to the reduction of total ethylene production in the absence of ACS2 gene , which makes it possible to observe a small difference . These results suggest that acs6-2 mutant allele is not a null mutant as acs6-1 allele , and that the high level induction of ACS6 is important to pathogen-induced ethylene production . Together with the gain-of-function evidence shown in Figure 10 , we can conclude that ACS6 gene activation plays an essential role in promoting ethylene production in plants challenged by pathogens .
In addition to post-translational regulation , we found that transcriptional activation of the ACS genes is also critical to the high-level of ethylene induction , as depicted in our working model ( Figure 12 ) . Stress- and pathogen-activation of ACS genes , such as Arabidopsis ACS6 , is well established [1] , [36] , [42] . In this report , we delineated a signaling pathway involved in the transcriptional activation of ACS2/ACS6 in Arabidopsis after pathogen infection . It is interesting to find that MPK3 and MPK6 not only function in the phosphorylation-induced stabilization of ACS2/ACS6 proteins , but also regulate the expression of ACS2 and ACS6 genes . The MPK3/MPK6 cascade-induced ACS2/ACS6 gene activation is mediated by WRKY33 , another MPK3/MPK6 substrate [34] . WRKY33 binds to the W-boxes in the promoters of the ACS2 and ACS6 genes directly in vivo ( Figure 9 ) and is involved in the MPK3/MPK6-induced ACS2/ACS6 gene expression ( Figure 7 ) . Mutation of WRKY33 resulted in a smaller reduction ( ∼60% ) in ACS2/ACS6 gene activation in response to B . cinerea infection ( Figure 8 ) , possibly due to the presence of other WRKY ( s ) that can partially compensate the loss of WRKY33 . Conditional overexpression of ACS6 in the acs2-1/acs6-2/acs7-1 mutant background greatly enhances the ethylene induction ( Figure 10 ) . Furthermore , reduction in ethylene induction in the acs6-2 allele , a knockdown mutant ( Figure 11 ) , provides loss-of-function evidence that demonstrates the importance of ACS6 gene activation during pathogen invasion . Transcriptional activation of ACS2 likely has a similar role . Induction of ACS6 expression is associated with stress-induced ethylene production [1] , [20] , [36] , [42] . However , direct evidence supporting the role of ACS gene activation has been lacking . In Arabidopsis , overexpression of wild type ACS6 genes is not sufficient to elevate ethylene production because of the requirement of protein phosphorylation and stabilization [16] . In addition , overexpression of the ACS6 gene in the wild type background fails to enhance ethylene production upon B . cinerea inoculation ( Li , G . , Liu , Y . , and Zhang , S . , unpublished data ) , a result of the high-level gene activation of the endogenous ACS genes ( Figure 1 ) . In this study , we expressed the ACS6 gene in an acs2-1/acs6-2/acs7-1 mutant background . The use of a DEX-inducible promoter and two independent lines with different levels of ACS6 gene induction after DEX treatment allowed us to demonstrate the importance of ACS6 gene activation ( Figure 10 ) . Our acs2-1/acs6-2 double mutant produces about 25% of the wild type level of ethylene after B . cinerea infection ( Figure 2 ) [28] . In contrast , the acs2-1/acs6-1 line only produces ∼15% of the wild type ethylene ( Figure 3 ) . The difference between these two double mutants is likely a result of different acs6 mutant alleles since both have the same acs2-1 mutant allele . The difference between acs2-1/acs6-2 and acs2-1/acs6-1 also suggests that the acs6-2 allele is not a complete null mutant , which is supported by the presence of ACS6 transcript in acs6-2 allele ( Figure 11B ) . Since the T-DNA insertion in this allele is in front of the transcriptional starting site ( Figure 9A ) , functional transcript is likely to be produced in this mutant allele . Nonetheless , the reduction of ethylene induction in the acs6-2 single mutant or in the acs2-1/acs6-2 double mutant [28] ( Figure 2 ) demonstrates the importance of high-level induction of ACS6 expression in pathogen-induced ethylene production . Interestingly , the reduction in ethylene induction in the acs2/acs6 double mutant is always more than the sum of the reduction in each single mutant ( Figure 3 ) [28] . It is possible that the heterodimers of ACS2 and ACS6 are less active than each of homodimer . In the absence of one isoform , only the homodimer can be formed , which partially compensates for the loss of the other isoform . Although very inefficient energy-wise , regulation of the ACS protein at the protein stability level by phosphorylation/dephosphorylation allows rapid induction of ethylene biosynthesis , which can occur within minutes after plant sensing of external stimuli [43]–[45] . Such rapid response could be important to plant response to the stress/pathogen stimuli . However , even after phosphorylation stabilization , the ACS6 protein may not be very stable . The half-life of the phospho-mimicking ACS6DDD is only ∼3 hours [29] . In the meantime , protein phosphatase 2A will counteract with the MAPKs by dephosphorylating the phospho-ACS protein [30] . Under this circumstance , transcriptional activation can provide another mechanism to further enhance the ethylene induction in response to pathogen infection . It is well known that stress-induced ethylene production follows different kinetics depending on the stimuli . However , the molecular mechanism underlying this difference is unclear . A good correlation exists between the kinetics of MPK3/MPK6 activation and ethylene induction . For instance , wounding induces a transient ethylene production , which is associated with a transient activation of MAPKs [46] , [47] . In contrast , infection of plants by pathogens , especially necrotrophic fungal pathogens , triggers a long-lasting and high-level induction of ethylene biosynthesis , which correlates with a long-lasting and high-magnitude activation of MPK3/MPK6 [28] . As depicted in Figure 12 , MPK3 and MPK6 regulate ethylene induction via two different mechanisms: by direct phosphorylation and stabilization of ACS2 and ACS6 proteins [16] , [28] , [29] and by activation of ACS2 and ACS6 gene expression ( this study ) . Transient activation of MPK3/MPK6 by wounding is also associated with the activation of ACS2/ACS6 gene expression [36] . However , due to the transient nature of MAPK activation , which returns to basal level within ∼0 . 5 hr to 1 hr [47] , the de novo synthesized ACS protein may not have the chance to be phosphorylated and will be degraded quickly in the absence of MAPK phosphorylation . In B . cinerea-infected plants , the induction of ACS2 and ACS6 gene expression will result in high rates of de novo protein synthesis . On top of this , the high-level and long-lasting activation of MPK3/MPK6 [28] , [38] can maintain de novo synthesized ACS2 and ACS6 proteins in a phosphorylated state and thereby stabilize the protein against proteasome-mediated degradation [16] , [29] . This dual-level regulatory mechanism can maintain a greatly enhanced level of cellular ACS activity and ethylene production in pathogen-infected plants . Recently , it was shown that a PP2A protein phosphatase can counteract with MPK3/MPK6 by dephosphorylating ACS2/ACS6 and can destabilize the ACS protein [30] . In this situation , it is even more important to have the high-level , long-lasting activation of MPK3/MPK6 in order to maintain the ACS2/ACS6 protein in a phosphorylated state to ensure the high rate ethylene biosynthesis observed in plants challenged by pathogens . Activation of MPK3/MPK6 and their orthologs in other plant species induces the expression of large number of stress/defense related genes [48] , [49] , suggesting the involvement of downstream transcription factors . ERF104 is a substrate of MPK6 . Phosphorylation of ERF104 by MPK6 results in release of ERF104 from the complex , which allows ERF104 to activate the expression of genes further downstream [50] . Recently , we identified the transcription factor WRKY33 as the substrate of MPK3 and MPK6 [34] . WRKY33 is involved in the induction of camalexin biosynthesis by promoting the expression of camalexin biosynthetic genes [34] , [51] . In this report , we demonstrate that WRKY33 is involved also in activation of ACS2 and ACS6 gene expression and ethylene induction . In the wrky33 mutant background , gain-of-function DD-induced ACS2 and ACS6 gene activation is essentially abolished . Furthermore , a ChIP-qPCR analysis demonstrated that WRKY33 directly binds to the promoters of ACS2 and ACS6 genes ( Figure 9 ) . These results reveal that WRKY33 regulates gene expression in multiple stress/defense responses and may function as a master transcriptional regulator downstream of the MPK3/MPK6 cascade . The expression of both ACS2 and ACS6 genes are regulated by the MPK3/MPK6 cascade and its downstream WRKY33 . However , the induction kinetics of ACS2 and ACS6 genes are different in both gain-of-function DD transgenic plants ( Figure 4 ) and wild type plants after pathogen treatment ( Figure 1 ) . One possibility is that one or more transcription factors , other than WRKY33 , are involved . The differential involvement of these unknown transcription factors could result in the different kinetics observed in the induction of ACS2 and ACS6 genes . These transcription factors may or may not be regulated by the MPK3/MPK6 cascade . The activation of MPK3 and MPK6 proteins in the gain-of-function DD plants is sufficient to induce the expression of ACS2 and ACS6 genes to levels similar to those observed in B . cinerea-infected plants ( Figure 1 and Figure 4 ) , suggesting that the transcriptional machinery controlling expression of ACS2 and ACS6 genes is fully turned on in DD plants . On the other hand , mutation of WRKY33 essentially blocks DD-induced ACS2/ACS6 gene activation ( Figure 7B ) but only partially blocks the induction of ACS2 and ACS6 genes in B . cinerea-infected plants ( Figure 8B ) , suggesting the activation of additional components by B . cinerea that cannot be activated by MPK3/MPK6 cascade alone , possibly homologs of WRKY33 that can partially replace the function of WRKY33 in its absence . The aforementioned discussions are focused on the role of the MPK3/MPK6 cascade in regulating ACS2 and ACS6 , two major contributors of pathogen-induced ethylene production , as depicted in our working model ( Figure 12 ) . A similar level of reduction in ethylene induction in the mpk3/mpk6 and acs2/acs6 double mutants ( ∼85% ) is consistent with our conclusion that the MPK3/MPK6 signaling cascade only controls ACS2 and ACS6 . Genetic evidence also supports the involvement of three additional ACS isoforms , ACS7 , ACS8 , and ACS11 , in B . cinerea-induced ethylene production ( Figure 2 and Figure 3 ) . ACS7 and ACS11 are the two members in the Type III ACS group in Arabidopsis . ACS8 belongs to the Type II ACS group . Based on mutant analyses , these three ACS genes contribute about 15% of the total ethylene produced in B . cinerea-infected plants ( Figure 3 ) [28] . Transcriptional activation of ACS7 , ACS8 , and ACS11 is not regulated by the MPK3/MPK6 cascade ( Figure 4 ) ; the signaling pathway ( s ) involved in the activation of their expression is unknown . It is also unclear whether ACS7 , ACS8 , and ACS11 are regulated at the protein stability level . Since ACS7 and ACS11 do not have a typical putative phosphorylation site in their C-termini , they are likely to be regulated at the transcriptional level only . ACS8 , similar to ACS5 and ACS9 , has a putative CDPK phosphorylation site in its C-terminus . It is possible that phosphorylation by CDPK ( s ) is involved in its protein stability regulation , similar to ACS5 and ACS9 [32] , [33] . Ethylene plays an important role in plant disease resistance . Using a high-order acs mutant , Tsuchisaka et al . ( 2009 ) demonstrated that ethylene production is essential to plant resistance against B . cinerea . However , ethylene induction was not examined in the study . In this report , we demonstrate that ethylene induction in acs1-1/acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1/acs11-1 mutant plants after B . cinerea infection is only at ∼2% of that in the wild type ( Figure 3 ) . Plant sensing of abiotic stress stimuli or invading pathogens triggers a number of signaling events . Among them , the activation of MAPK cascades and calcium influx are two of the earliest [21] , [24] , [52] . Our research demonstrates the regulation of ACS2 and ACS6 by a specific MAPK cascade at both transcriptional and post-translational levels . This pathway contributes ∼85% of the total ethylene induction in plants challenged by pathogens . Regulation of the remaining ACS isoforms is unclear at present . Additional studies , including identification of the signaling pathway ( s ) involved in regulation of ACS7 , ACS8 , and ACS11 expressions , protein phosphorylation , and protein stability , are needed to further our understanding of the complex regulation of ethylene induction during the plant stress/defense response .
Soil-grown plants were maintained at 22°C in a growth chamber with a 14-hr light cycle ( 100 µE/m−2 sec−1 ) . For experiments , seeds were surface-sterilized . After imbibition at 4°C for 3–5 days , seeds were sown in petri dishes with liquid half-strength Murashige and Skoog ( MS ) medium and grown in a growth chamber at 22°C with continuous light ( 70 µE/m−2 sec−1 ) . Five-day-old seedlings were transferred to 20-ml GC vials with 6 ml of liquid half-strength MS medium ( 10 seedlings per vial ) and maintained under the same growth conditions . Twelve- to fourteen-day-old seedlings grown in GC vials were used for experiments . Procedures for Botrytis cinerea ( Strain: DSM 4709 ) maintenance and spore preparation were described previously [28] . Twelve-day-old seedlings grown in GC vials were inoculated with B . cinerea spores at a final concentration of 4 . 0×105 spores/vial . Induction of DD and ACS6 expressions in GVG-NtMEK2DD and GVG-ACS6 transgenic plants was performed by the addition of DEX stock solution ( 5 mM in ethanol ) to a final concentration of 1 µM . An equal volume of ethanol was used as a negative ( −DEX ) control . At least two independent repetitions were performed with similar results for experiments with multiple time points . For single time-point experiments , at least three independent repetitions were performed . Arabidopsis thaliana Columbia ( Col-0 ) ecotype was used as the wild-type control , unless stated otherwise . T-DNA insertion mutant alleles of MPK3 ( At3g45640 ) , MPK6 ( At2g43790 ) , ACS1 ( At3g61510 ) , ACS2 ( At1g01480 ) , and ACS6 ( At4g11280 ) were described previously [16] , [28] , [39] . The two ACS7 ( At4g26200 ) mutant alleles , acs7-1 ( FLAG_431D05 ) and acs7-2 ( CSHL_ET5768 ) , were obtained from INRA and Cold Spring Harbor Laboratory , respectively . High-order acs mutants generated in Dr . Athanasios Theologis' laboratory [35] were obtained from the Arabidopsis Biological Resource Center ( ABRC ) . The stock numbers are CS16564 ( acs2-1 ) , CS16569 ( acs6-1 ) , CS16581 ( acs2-1/acs6-1 ) , CS16603 ( acs2-1/acs6-1/acs4-1 ) , CS16607 ( acs2-1/acs6-1/acs5-2 ) , CS16609 ( acs2-1/acs6-1/acs9-1 ) , CS16644 ( acs2-1/acs6-1/acs4-1/acs5-2/acs9-1 ) , CS16649 ( acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1 ) , CS16650 ( acs1-1/acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1 ) , and CS16651 ( acs1-1/acs2-1/acs6-1/acs4-1/acs5-2/acs9-1/acs7-1/acs11-1 ) . Conditionally rescued mpk3/mpk6 double mutant was generated by transformation of DEX-inducible promoter driven MPK6 cDNA ( GVG-MPK6 ) into mpk3−/−/mpk6+/− plants [39] . Double mutant seedlings were recovered from seeds of mpk3−/−/mpk6+/−/GVG-MPK6 plants sprayed with 30 µM DEX during the flowering stage . GVG-NtMEK2DD ( abbreviated as DD ) , DD/mpk3 , and DD/mpk6 lines were previously described [28] , [38] . The DEX-inducible promoter driven ACS6 construct ( GVG-ACS6 ) was generated by cloning the ACS6 ORF with a 4xmyc tag [29] into the Xho I/Spe I sites of the pTA7002 vector [53] . The binary vector was transformed into Agrobacterium tumefaciens strain GV3101 . Arabidopsis transformation was performed by the floral dip procedure [54] , and transformants were identified by screening for hygromycin resistance . Independent lines with ACS6 transgene induction were identified based on real-time qPCR analysis . GC vials with Arabidopsis seedlings were flushed and capped immediately after treatment . At indicated times , ethylene levels in the headspace of the GC vials were measured by gas chromatography as previously described [16] . Seedlings were then collected , weighed , frozen in liquid nitrogen , and stored at −80°C for future analysis . Total RNA was extracted using the Trizol reagent ( Invitrogen ) . After DNase treatment , RNA ( 2 µg ) was used for reverse transcription . Real-time PCR analysis was performed using an Opticon™ 2 real-time PCR machine as described previously [38] . The transcript of the EF1α gene was used to normalize the samples . Relative gene expression was calculated using two different methods . The first method is the commonly used double ΔCt method , which gives fold of gene induction relative to basal level before treatment ( 0 hr time point ) . The second method expresses the transcript level relative to that of the EF1α gene in the same sample , which is a better method when comparison of expression levels of different genes is necessary . The primers used for real-time PCR were ACS1 ( At3g61510 , 5′-ACGCTTTTCTCGTCCCTACTC-3′ and 5′-GGCCTTAAGGTACGCTGATTC-3′ ) , ACS2 ( At1g01480 , 5′-GGATGGTTTAGGATTTGCTTTG-3′ and 5′-GCACTCTTGTTCTGGATTACCTG-3′ ) , ACS4 ( At2g22810 , 5′-AACAACCTTGTGCTCACTGCT-3′ and 5′-AGATCCCTATCAAACCCTGGA-3′ ) , ACS5 ( At5g65800 , 5′-GACTCTCATGTTTTGCCTTGC-3′ and 5′-TTGGAAGCCATTAGAGCTTGA-3′ ) , ACS6 ( At4g11280 , 5′-GTTCCAACCCCTTATTATCC-3′ and 5′-CCGTAATCTTGAACCCATTA-3′ ) , ACS7 ( At4g26200 , 5′-ACGGTACGATACCATTGTGGA-3′ and 5′-GCTCGCCGTCTTTAGTTTTCT-3′ ) , ACS8 ( At4g37770 , 5′-CCTTCCTTCCTTCAAGAATGC-3′ and 5′-GAGAGTCTCGTTAGCCGGAGT-3′ ) , ACS9 ( At3g49700 , 5′-CATACCTCGACGAAAACCAGA-3′ and 5′-TCATGTCAACCCAACAGAACA-3′ ) , ACS11 ( At4g08040 , 5′-CAAACGATGGAGGTTGCTATG-3′ and 5′-TTGGAGACCCATTTGTTGATAAG-3′ ) , and EF1α ( At5g60390 , 5′-TGAGCACGCTCTTCTTGCTTTCA-3′ and 5′-GGTGGTGGCATCCATCTTGTTACA-3′ ) . F1 plants generated from the cross of wrky33/4myc-WRKY33 and DD lines were used for the ChIP assay . Two-week-old seedlings treated with 1-µM DEX for 12 hr were processed as previously described [55] . Briefly , chromatin was isolated from 0 . 8 g of frozen tissue and sonicated with a Bioruptor sonicator ( 15 s on and 15 s off cycles , medium-energy settings ) for 6 min . Immunoprecipitation was performed by incubating chromatin with 2 µg of anti-myc antibody ( Millipore ) or mouse IgG ( negative control ) for 1 hr at 4°C . The protein-chromatin immunocomplex was captured using Protein G-Dynal magnetic beads ( Invitrogen ) . After Proteinase K digestion , the immunoprecipitated DNA was purified using a ChIP DNA Clean and Concentrator kit ( Zymo Research Corporation ) . Immunoprecipitated DNA and input DNA were analyzed by qPCR using primers specific for the promoter regions of PAD3 and WRKY33 . The primer pairs ( forward and backward ) used for ChIP-qPCR were ACS2 ( 5′-AGGCCATAAGCCCATTCAAA-3′ and 5′-GCCTACAGTGCACGACTTCA-3′ ) and ACS6 ( 5′-AAAGTCGTTGAGATTGTGTTGG-3′ and 5′-TGGCAGCCTTAAAGACCAGT-3′ ) , which are in proximity of the W-boxes in the promoters . ChIP results are presented as percentage of input DNA . Sequence data for this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: MPK3 ( At3g45640 ) , MPK6 ( At2g43790 ) , EF1α ( At5g60390 ) , ACS1 ( At3g61510 ) , ACS2 ( At1g01480 ) , ACS4 ( At2g22810 ) , ACS5 ( At5g65800 ) , ACS6 ( At4g11280 ) , ACS7 ( At4g26200 ) , ACS8 ( At4g37770 ) , ACS9 ( At3g49700 ) , ACS11 ( At4g08040 ) , and WRKY33 ( At2g38470 ) . | Plant immunity , similar to that in animals , also involves mitogen-activated protein kinase ( MAPK ) cascades . However , plants use unique MAPK substrates and secondary signaling molecules in the process . Among them , ethylene , a gaseous plant hormone , plays critical roles . Ethylene-regulated responses begin with the induction of ethylene biosynthesis . 1-amino-cyclopropane-1-carboxylic acid synthase ( ACS ) catalyzes the committing and rate-limiting step in ethylene biosynthetic pathway . The Arabidopsis genome encodes nine different ACS isoforms . Two of them , ACS2 and ACS6 , were previously shown to be phosphorylated and stabilized by MPK3 and MPK6 , two Arabidopsis pathogen-responsive MAPKs . Using a genetic approach , we identified additional ACS isoforms including ACS7 , ACS8 , and ACS11 that also contribute to pathogen-induced ethylene production . In addition to direct phosphorylation modification and stabilization of ACS2 and ACS6 proteins , MPK3 and MPK6 also regulate the expression of ACS2 and ACS6 genes through another MPK3/MPK6 substrate , WRKY33 , a member of the plant-specific WRKY transcription factor family . Regulations of ACS isoforms at both transcriptional and post-translational levels contribute to the high-level ethylene production in plants challenged by invading pathogens . These findings shed light on our understanding of the regulation of the kinetics and magnitude of ethylene induction under different stress conditions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2012 | Dual-Level Regulation of ACC Synthase Activity by MPK3/MPK6 Cascade and Its Downstream WRKY Transcription Factor during Ethylene Induction in Arabidopsis |
Scabies and head lice are ubiquitous ectoparasitic infestations that are common across the Pacific Islands . Ivermectin is an effective treatment for both conditions , although the doses used vary . At a community level , mass drug administration ( MDA ) with ivermectin is an effective strategy to decrease prevalence of scabies . To what extent MDA with ivermectin will also reduce prevalence of head lice is unknown . Head lice prevalence was assessed before and after MDA with oral ivermectin ( at a dose of 200 micrograms per kilogram of body weight ) administered on day 1 and day 8 . The primary outcome was the change in prevalence of head louse infestation at two weeks compared to baseline . Longer term efficacy was assessed three months after MDA . 118 participants were enrolled . Baseline prevalence of active head louse infestation was 25 . 4% ( 95% CI 18 . 4–34 . 0 ) . At two-week follow-up , prevalence was 2 . 5% ( 95% CI 0 . 9–7 . 2 ) , a relative reduction of 89 . 1% ( 95% CI 72 . 7–91 . 4% , p<0 . 001 ) . At three-month follow-up , prevalence was 7 . 5% ( 95% CI 2 . 7–12 . 3 ) , a relative reduction of 70 . 6% ( 95% CI 72 . 7%-91 . 4% , p <0 . 001 ) . Head louse infestation was associated with younger age ( age ≤10 years: prevalence 46 . 7%; adjusted odds ratio compared to adults of 7 . 2 , 95%CI 2 . 0–25 . 9 ) and with having at least one other member of the household with active head louse infestation ( adjusted odds ratio 4 . 3 , 95%CI 1 . 7–11 . 1 ) . Head louse infestation is common in the Solomon Islands . This proof of principle study shows that oral ivermectin at a dose of 200 micrograms per kilogram can reduce the burden of active head louse infestation , offering an additional collateral benefit of MDA with ivermectin for scabies control . ClinicalTrials . gov NCT03236168 .
The human head louse , Pediculus humanus capitis , is an ectoparasitic insect found on the scalp that is endemic world-wide [1] . Head louse infestation can cause intense scalp itching and scratching can lead to secondary bacterial infections [2] . In low and middle income countries , this parasitic skin disease has received little attention in research [3] . Specifically , in the Pacific region , there is minimal research quantifying its prevalence or potential treatment strategies [4] . Ivermectin is an oral , semi-synthetic derivative of the avermectin family of lactones that selectively binds glutamate-gated chloride channels found in invertebrate muscle and nerve cells thus disrupting neurotransmission in a wide range of human parasites [5] . Oral preparations of ivermectin can be used to treat head lice [2 , 6] as well as other blood feeding ectoparasitic diseases , including scabies [7 , 8] . It should be noted that for both , head lice and scabies , oral ivermectin treatment does not have ovicidal action . A large trial evaluating the use of ivermectin monotherapy , in difficult-to-treat head louse infestation , showed that ivermectin at a dose of 400 micrograms per kilogram ( kg ) on day 1 and day 8 , had superior efficacy compared to 0 . 5% malathion lotion [5] . Small studies have suggested variable efficacy when a dose of 200 micrograms per kg is used . Single administration resulted in variable reductions in head lice of 45% - 100% at one week follow up but repeat administration on day 8 yielded higher reductions ( 87 . 5% - 100% ) at two-week follow up [9–13] . These studies have all focused on treatment of individual patients and not on whole communities where head lice are common and where re-infestation from other community members might be anticipated to result in a lower efficacy than that seen in individual treatment trials . Scabies , another ectoparasitic disease , has already been shown to be extremely common in the Pacific [14–16] . Ivermectin , used as part of mass drug administration ( MDA ) , at a dose of 200 micrograms per kg given at a 7-day interval , has been shown to be an effective strategy to reduce community wide prevalence of scabies in the region [17–20] . These programmes might have ancillary benefits such as treatment of headlice in the communities . The study we conducted aimed to establish baseline prevalence of head louse infestation in a rural community in the Solomon Islands and to assess whether MDA using ivermectin , at the lower dose of 200 micrograms per kilogram ( the same regimen already used for scabies treatment ) would be an effective method to lower community prevalence of head lice .
The study was approved by the London School of Hygiene and Tropical Medicine ( LSHTM ) ethics committee and the National Health Research and Ethics Committee of the Solomon Islands Ministry of Health and Medical Services . Written consent was obtained from all participants by a staff member fluent in both local languages ( Pijin and East Kwaio ) . Parents or guardians provided written consent for participants aged below 18 years . Children were additionally asked to provide verbal assent for participation . The study was prospectively registered on clinicaltrials . gov ( NCT03236168 ) . The study was conducted in the campus area of Atoifi Adventist Hospital ( AAH ) in the East Kwaio region of Malaita Island , Solomon Islands . The hospital staff and their families live on campus and create a self-contained community of approximately 200 individuals . The campus has a school for children aged up to 16 years . Students who attend Atoifi School live on campus ( staff children ) or live in villages that surround the campus . All individuals living on AAH campus and all children and families of children attending Atoifi School were eligible for the study . At baseline participants’ demographic and household data were collected and all individuals underwent a standardised examination of their skin and scalp . Scalp examination consisted of direct visual inspection of sites of predilection for head lice: the back of the ears , temples and neck . This was followed by parting the hair into 4 sections ( first down sagittal plane and then ear to ear ) . Each section was visually inspected and examined using a metal nit comb with plastic base and 0 . 1mm teeth size by the brand InfectoPedicul manufactured by InfectoPharm . Eggs and lice were as far as possible not deliberately removed from the hair during the examination to avoid biasing outcome measure assessment by physical delousing . Scalp examinations were carried out by trained nursing staff at Atoifi Hospital led by a physician ( SC ) . Prior to study commencement all staff received a standardised half-day training workshop including practical sessions on clinical examination and use of nit combs . Ten percent of hair examinations were repeated by an independent examiner to ascertain accuracy of diagnosis . Adult and nymphal stages of head lice were classified as ‘active head louse infestation’ . If only head lice eggs were seen , this was recorded as ‘egg infestation’ . No attempts were made to differentiate between viable and non-viable eggs . The diagnosis of scabies was made clinically based on presence of pruritic inflammatory papules or nodules with a typical anatomical distribution using previously validated criteria [21] . The distribution and number of scabies lesions were recorded and used to classify severity of scabies based on classifications used in previous studies in the Pacific region [16 , 22] . Participants were not re-examined for scabies as part of this study . Ivermectin was administered as directly observed therapy at a dose of 200 micrograms per kg of body weight . The first administration of ivermectin was at time of baseline examination . A second administration of ivermectin was given 7 days later to all study participants ( day 8 ) . Participants with contraindications to ivermectin treatment ( weight below 12 . 5kg , pregnancy or breastfeeding ) were offered malathion 0 . 5% lotion for head lice and permethrin 5% cream for scabies . Participants diagnosed with any non-scabies skin condition were briefly counselled and , if necessary , advised to attend the medical clinic to obtain treatment as per standard local protocol . To assess immediate efficacy , a subgroup of participants was re-examined at 48 hours . School children were chosen both for convenience sampling and because they were predicted ( from risk factors identified in literature citation ) to have highest rates of head louse infestation . The primary outcome of the study was change in prevalence of head lice at 2 weeks compared to baseline . Longer-term efficacy was assessed by examination at 3 months . For analysis , age was classified into three categories: less than or equal to 10 years , 11–20 years and over or equal to 21 years of age . Area of residence was classified into two categories: household in AAH or household out of AAH campus . We calculated the prevalence of individuals with adult or nymphal stages of head lice on examination ( active head lice ) and the prevalence of head lice eggs on examination . Univariable logistic regression was used to identify risk factors for the presence of head lice at baseline . Variables associated in the univariable regression analysis were included in a multivariable model . We considered age and gender as forced confounders to be included in the multivariable model . Treatment efficacy was assessed by comparing the proportion of individuals with head lice , using a two-sample test of proportion at baseline compared to 48 hours , two weeks and three months respectively . We calculated the absolute and relative reduction in the prevalence of both active head louse infestation and egg infestation . Data was analysed with STATA software version 14 . 2 ( Stata Corporation , College Station , TX , USA ) .
The baseline prevalence of active head louse infestation was 25 . 4% ( 95% CI 18 . 4–34 . 0 ) . In all cases where active head louse infestation was identified , head lice eggs were also seen . The baseline prevalence of head lice eggs ( with or without active head louse infestation ) was 42 . 3% ( 95% CI 30 . 8–48 . 8 ) . Scabies prevalence was 10 . 2% ( 95% CI 5 . 9–16 . 9 ) ( Table 2 ) . No cases of crusted scabies were identified . Other skin conditions diagnosed included tinea corporis ( n = 3 ) and eczema ( n = 4 ) . Active head louse infestation was associated with younger age , with increased risk compared to adults for both 0–10 years ( AOR 7 . 2 , 95% CI 2 . 0–25 . 9 , p = 0 . 003 ) and 11–20 years ( AOR 9 . 1 , 95% CI 2 . 4–34 . 2 , p = 0 . 001 ) . Infestation was also associated with the presence of other members of the household with active lice infestation ( AOR 3 . 9 , 95% CI 1 . 1–14 . 3 , p = 0 . 040 ) . No association was seen with gender or location of the house on or off the AAH campus ( Table 3 ) . The presence of head lice eggs was associated with age , with highest risk in children aged under or equal to 10 years ( AOR 5 . 9 , 95% CI 2 . 0–17 . 1 ) . Head lice eggs were also associated with female gender ( AOR 3 . 1 , 95% CI 1 . 3–7 . 6 ) ( Table 3 ) . Twenty-four out of 28 school-children ( 85 . 7% ) were re-examined at 48 hours following ivermectin administration . The prevalence of active infestation decreased from 45 . 8% at baseline to 0% . At two weeks after MDA , the prevalence of active infestation had declined significantly ( 25 . 6% vs 2 . 5% , relative reduction ( RR ) 89 . 1% , p <0 . 001 ) ( Table 4 ) . At three months , the prevalence of active head louse infestation remained significantly lower than at baseline ( 25 . 6% vs 7 . 5% , RR 70 . 6% , p <0 . 001 ) ( Table 4 ) . The prevalence of head lice eggs was unchanged at two weeks ( 42 . 3% vs 41 . 8% ) but reduced significantly at three months ( 42 . 3% vs 19 . 6% , RR 53 . 7% , p < 0 . 001 ) ( Table 4 ) .
This is the first study to demonstrate the efficacy of community-based ivermectin administration for head lice treatment using ivermectin 200 micrograms per kg , administered twice at day 1 and day 8 . This regimen significantly reduced prevalence of active head louse infestation at both the two-week and three-month follow-up visit . At 48 hours , none of the subgroup screened had active head louse infestation . This suggests that the 200 micrograms per kg dose is effective at killing nymphal and adult head lice stages in this population and that a higher dose ( 400 microgram per kg ) is unlikely to offer any additional short-term head lice killing benefits . At two weeks , the prevalence of active head lice in the whole study sample had decreased significantly ( from 25 . 6% to 2 . 5% ) . There was no reduction in head lice eggs at two weeks , however , this was expected given that head lice eggs are cemented , via glue-like glandular discharge , to the hair shaft and remain attached even after hatching . Therefore , head lice eggs will still be found even after the active head louse infestation has been treated . At three months , the effect of the intervention was sustained . The prevalence of active infestation remained significantly lower than baseline . Furthermore , the reduction in active head louse infestation translated into a reduction in egg prevalence at three months . Our study also demonstrates that head lice are an extremely common ectoparasitic infection in the Solomon Islands with prevalence comparable or higher than those reported in other lower income settings such as Nepal , Brazil and Egypt [3] . This is a significant burden of disease and further qualitative research is important to elucidate the local attitudes and perception towards head louse infestation . The findings of this study highlight an opportunity to concomitantly target two common ectoparasitic infections at community level with a single intervention strategy . Showing evidence for additional benefits of MDA may increase engagement of local communities and support from local authorities towards community-wide programs and enables control of head louse infestation to be integrated with growing control programs for neglected tropical diseases such as scabies . Our study had some limitations . First , MDA coverage was 54% of the eligible population . We did not record the age of non-participants and , as the prevalence of headlice is higher in children , this may have biased our overall prevalence estimate . Despite this our intervention demonstrated efficacy of MDA . Second , prevalence of active head louse infestation did increase slightly at three months compared to two weeks . This may well reflect re-introduction of head lice from untreated members of the community or surrounding villages . Expanding the population that receives treatment may reduce the risk of re-introduction and thus improve long-term reduction in head lice prevalence at community level . Finally , we did not differentiate between viable eggs and non-viable eggs . Being able to reliably distinguish viability of eggs would have yielded a more accurate estimate of active head louse infestation in the community and could be considered in future studies . Our study demonstrated the effectiveness of a two-stage administration of ivermectin at a dose of 200microgram per kg for the treatment of head lice . A reduction in head lice prevalence is likely to be an ancillary benefit of the scale-up of scabies control programmes in the Pacific region and elsewhere . As with other ectoparasitic infestations , head lice are a common and under-recognised parasitic infestation in the Solomon Islands . Fully measuring the impact of MDA programs can maximise the potential benefits to the community and aid community engagement with the intervention . Further delineating the full range of benefits of ivermectin MDA should be a priority for scabies control programmes . | Head lice and scabies are both caused by ectoparasites and lead to itchy skin conditions that are associated with secondary bacterial infections and social stigma . Both are common in developing countries . In the Solomon Islands , mass treatment of communities using ivermectin ( at dose of 200 micrograms per kilogram ) has been shown to be an effective strategy in reducing scabies prevalence . This study shows that the same dose of ivermectin and the same regime of mass drug administration is also effective at reducing the burden of head louse infestation , offering an additional benefit to community wide treatment . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"children",
"medicine",
"and",
"health",
"sciences",
"solomon",
"islands",
"education",
"scalp",
"sociology",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"parasitic",
"diseases",
"animals",
"age",
"groups",
"ectoparasitic",
... | 2018 | Impact of ivermectin administered for scabies treatment on the prevalence of head lice in Atoifi, Solomon Islands |
Phosphorylation at specific residues can activate a protein , lead to its localization to particular compartments , be a trigger for protein degradation and fulfill many other biological functions . Protein phosphorylation is increasingly being studied at a large scale and in a quantitative manner that includes a temporal dimension . By contrast , structural properties of identified phosphorylation sites have so far been investigated in a static , non-quantitative way . Here we combine for the first time dynamic properties of the phosphoproteome with protein structural features . At six time points of the cell division cycle we investigate how the variation of the amount of phosphorylation correlates with the protein structure in the vicinity of the modified site . We find two distinct phosphorylation site groups: intrinsically disordered regions tend to contain sites with dynamically varying levels , whereas regions with predominantly regular secondary structures retain more constant phosphorylation levels . The two groups show preferences for different amino acids in their kinase recognition motifs - proline and other disorder-associated residues are enriched in the former group and charged residues in the latter . Furthermore , these preferences scale with the degree of disorderedness , from regular to irregular and to disordered structures . Our results suggest that the structural organization of the region in which a phosphorylation site resides may serve as an additional control mechanism . They also imply that phosphorylation sites are associated with different time scales that serve different functional needs .
Phosphorylation is a ubiquitous post-translational modification that is known to be important for the regulation of a myriad of cellular processes , among which are cell growth , apoptosis , differentiation , signal transduction and transport [1] . Rapidly evolving mass spectrometry ( MS ) -based technologies , innovative labeling techniques and advances in computational proteomics provide powerful means for overcoming the low abundance problem of this modification and are making it possible to obtain large-scale , high-resolution quantitative data . With these advances , not only can single protein phosphorylation experiments be done with high accuracy , but also whole-phosphoproteome studies are becoming increasingly feasible [2] , [3] . Given the availability of these data , much research has been devoted to analyzing and understanding the structural features of phospho-sites . This includes creation of online resources containing structural information [4] , combining data on linear motifs and structural properties [5] , and development of software tools that use three-dimensional data for the prediction of phosphorylation sites ( DISPHOS [6] , Phos3D [7] ) . Large-scale studies of the structural characteristics of phosphorylation sites have focused on solvent exposure , local and global structure , amino acid context of the spatial surrounding , and structural motifs [7]–[9] . The mechanism of modification suggests that serine , threonine and tyrosine residues should be located on the protein surface where they are accessible for the modifying kinase [7] . The main challenge in studying structural properties of phospho-sites from experimental data is their preference for unstructured regions [6] for which electron density is often missing in X-ray structures . Disorder is strongly associated with protein-protein interactions [10] . Modified residues found within disordered regions can act as on/off switches , either promoting or inhibiting an interaction . Due to the specific structural organization of some protein kinases , in which the catalytic loop resides within a small cleft between two lobes , flexible regions within the substrate's interaction surface are well suited for binding to the kinase . However , a recent systematic study suggested that kinase preference for disordered regions is only marginal [8] . Furthermore , a computational study of kinase specificity reported that approximately 60% of the sites modified by protein kinase A lie within α-helical regions [11] . These considerations raise an interesting question: can a distinction between kinases be made with respect to the level of structural organization of their substrates . Since phosphorylation events both depend on the structural environment and influence its properties , protein structure and phosphorylation should be considered interrelated and mutually dependent . On one hand , disorder facilitates the access of a kinase to the residue to be modified . On the other hand , the addition of a phosphate moiety may lead to structural changes . Both order-to-disorder and disorder-to-order transitions upon phosphorylation have been observed in nature or studied via molecular dynamics simulations [12]–[15] . The major driving forces of conformational changes observed upon phosphorylation are the electrostatic interactions between the negatively charged phosphate group and the surrounding charged residues . The functional roles of charged residues range from stabilization to correct substrate identification and facilitation of conformational changes . Although numerous previous studies have focused on structural properties of phosphorylation sites [6]–[8] no systematic analysis has been performed combining large-scale quantitative data with structural features . To bridge this gap we here build on data from a recent study by Olsen et al . , which elucidated phosphorylation site occupancy during mitosis [16] . Quantitative data were measured at six time points , corresponding to major phases of the cell division cycle . The additional temporal dimension of these data makes it possible to examine how various phosphorylation sites are dynamically regulated . Olsen et al . clustered sites according to their distinct phosphorylation patterns and similarities in regulation with the aim to infer each site's functional importance . Here , in contrast , we focus on structural properties of the phosphorylation sites and , for the first time , distinguish between two groups of sites with respect to the overall variation of phosphorylation over time . We find that sites that lie within regular secondary structures exhibit less variable phosphorylation fold changes during the cell cycle than sites that are found in disordered regions . Analysis of the amino acid composition of the flanking regions of these two groups of sites revealed enrichment of positively charged residues and depletion of disorder-related residues such as proline , serine and threonine in the former group .
Using the data from the Olsen et al . investigation [16] , we here computed the overall variation of the phosphorylation ratios during six time points of the cell cycle and investigated the differences between the sites with small variation as opposed to the sites with large variation . The original data set comprised 6 , 027 proteins with 20 , 443 unique phosphorylation sites . We retained only those sites that had quantitative information for all six time points available ( 1 , 059 proteins with 5 , 173 sites ) . The phospho-site variability is calculated as the standard deviation of the phosphorylation ratios over the six time points measured during the cell cycle . We sought to investigate a possible relation between the structural organization of the environment in which a modified residue is found and the experimentally measured changes in phosphorylation during the cell cycle . To do so , we compared the phosphorylation variation of two groups of sites . These two groups were composed of sites that reside in ordered regions and sites that lie within disordered regions as predicted with DISOPRED [17] . In agreement with previously observed tendency we found over 90% of the modified residues to lie within disordered regions ( 4 , 675 sites versus 498 sites ) . Figure 1 shows three examples from our large-scale dataset , illustrating a non-variable site on a regular secondary structure ( α-helix ) , a slightly variable site on a short loop and a variable site in a disordered region . Our results revealed notable differences in the distributions of phosphorylation variations of the two sets ( Kolmogorov-Smirnov test p-value 6 . 6E-13 ) . The sites associated with structurally characterized regions were found to exhibit smaller changes in phosphorylation during the cell cycle ( median 1 . 77 ) as compared to sites located in disordered regions ( median 2 . 22 , Figure 2A ) . Having investigated the difference between ordered and disordered regions on a global scale , we next predicted protein secondary structure in more detail using PsiPred [18] . First we classified sites into regular structures ( 92 in α-helices or 53 in β-sheets ) and sites with irregular structures ( 5 , 028 in loops , turns and coils ) . Phosphorylation in regular secondary structures showed smaller variation over the six time points of the cell cycle . This effect was small but statistically significant ( ANOVA p-value 1 . 8E-04 ) . Although there is a large intersection between ordered structures and regular secondary structures , and the terms are often used interchangeably , the two sets are not identical . We observed that a large number of regions predicted as ‘coil’ by PsiPred are predicted as ‘ordered’ by DISOPRED . This reflects a distinction between ordered and disordered coils . A major difference between these two groups of coils is the length distribution of their elements ( p-value 4 . 12E-114 ) : ‘ordered coils’ are much shorter on average as they mainly correspond to turns and short loops connecting regular secondary structures . By contrast , ‘disordered regions’ are longer and represent large protein regions lacking defined structure ( see Text S1 for details ) . In order to take this distinction into account , we redefined the structural environments into three categories: regular structures ( predicted as helix/sheet and ordered ) , irregular structures ( predicted as coil and ordered ) , and disordered regions ( predicted as coil and disordered ) ( Figure 2B ) . We found significant differences in the variation of the phosphorylation ratios between these distinct structural groups ( ANOVA p-value 3 . 02E-09 ) . Sites within ordered structural environments appeared to be subjected to the lowest level of regulation during the cell cycle ( median 1 . 65 ) . Interestingly , a distinction emerged between coils ( median 1 . 83 ) and disordered regions ( median 2 . 22 ) , signifying that the latter exhibited the largest variation in phosphorylation changes . We speculate that the increased variation of phosphorylation in longer , disordered coils correlates with their higher solvent exposure , which makes them more easily accessible for both kinases and phosphatases . Overall , our data shows that the phosphorylation variation of a site clearly scales with the level of order of its structural context ( i . e . the tendency of a site to be found within a regular , irregular or disordered region ) . We wanted to investigate if sites with distinct phosphorylation patterns over the cell cycle differ not only according to structural context , but also with respect to the amino acid content in their local sequence environment . A two-sample logo [19] was computed to contrast the two data sets , using the highly variable sites as a negative set ( Figure 3 ) . For each position and each possible amino acid , a two sample t-test was used to evaluate the null-hypothesis that the vectors of residues at a given position in both the positive and negative data sets ( i . e . low and high variation ) come from the same distribution . We found statistically significant enrichment of charged amino acids and depletion of proline , serine and threonine in the surrounding of sites with small phosphorylation variability ( p-value<0 . 05 ) . Additional comparisons of the amino acid distributions of the two sets against a background distribution accounting for structural differences using the composition profiling technique [20] revealed similar trends ( see Text S1 for the detailed analysis and results ) . The enrichment of serine and threonine residues in the vicinity of the detected phosphorylation sites could correlate with additional modification events . To check this hypothesis , we determined if multiple phosphorylation sites are found with higher preference in disordered regions . Phosphorylation sites that had at least one neighboring phosphorylation site in both ordered and disordered regions were compared . A ‘neighbor’ was defined as any phosphorylated residue that lies within +/−1 , 2 , 3 , 4 , or 5 residue-long flanking region of a given modification site . Regardless of which of these five cut-offs was chosen , multiple phosphorylation sites were always highly significantly enriched in disordered regions ( Table 1 ) . Next , we were interested in potential differences in evolutionary constraints on the phospho-sites in structured and disordered regions . When analyzing conservation it is important to take into account the different evolutionary rates of disordered and ordered regions . We therefore compared conservation scores between phosphorylated serines , threonines and tyrosines with ‘control’ serine , threonine and tyrosine residues with a similar structural background . We define the set of ‘control’ residues , as all potential phosphorylation sites that were not found to be phosphorylated in the study of Olsen et al . As expected , phospho-sites that were predicted to lie in regular regions appeared significantly more conserved than phospho-sites in disordered regions ( p-value 3 . 23E-120 ) , due to the more conserved structural background of the former ( Figure 4 ) . In agreement with a previous study [21] modified residues in regions that lack defined structure were more conserved than the control serine , threonine and tyrosine residues with the same surrounding environment ( Mann-Whitney Wilcoxon test p-value 3 . 4E-03 ) . The same holds true for phospho-serine , phospho-threonine and phospho-tyrosine in ordered regions as compared to their equivalent control sets ( p-value 2 . 24E-16 ) . Despite the small size of the effect ( groups' means −0 . 38 , −0 . 28 , 0 . 14 and 0 . 22 for pS/pT/pY ordered , S/T/Y ordered , pS/pT/pY disordered and S/T/Y disordered respectively ) the higher evolutionary pressure on phosphorylated residues suggests functional importance of these sites in a broad range of species . We next asked if different groups of kinases would exhibit preferences for less variable or highly variable phosphorylation sites . To identify kinase recognition motifs that show similar behavior with respect to two variables – protein disorder and phosphorylation variation , we used the recently described 2D Annotation Enrichment technique ( see Methods and [22] ) . It employs a two dimensional generalization of the nonparametric two-sample test to detect preferences of a certain group of elements for two numerical attributes simultaneously relative to all other elements . The motifs separation is plotted in Figure 5 ( the complete data are available in Table 2 ) . The general trend between disorder and phosphorylation variation is reflected in the plot as sites with more disordered background show also higher variability . For individual kinases a very clear separation reflecting their preference for specific amino acids in their consensus motifs becomes apparent . Overall , four classes can be distinguished: ( i ) tyrosine kinases ( black squares ) , ( ii ) proline-directed kinases ( red circles ) , ( iii ) non-proline directed kinases with charged residues in their substrate recognition motif ( green and blue triangles ) and ( iv ) proline-oriented kinases , which contain a proline residue in their motif ( red triangles and pentagons ) . The class of tyrosine kinases shows a strong preference for low phosphorylation variability and structured regions , whereas the other three classes favor more disordered regions , but span a wider range of phosphorylation variation values . Also among the latter three groups higher quantitative variability is clearly associated with higher disorder . As seen in Figure 5 , basophilic and acidophilic kinases occupy the regions on the graph corresponding to low phosphorylation variation , whereas proline-directed kinases are located on the right part of the graph , demonstrating their preference for more variable sites . The motif corresponding to the highest variability is the consensus motif for the proline-directed CDK5 kinase , which is in agreement with the important regulatory role of this enzyme during the cell cycle . The proline-related class represented by two acidophilic , one basophilic and one atypical protein kinases shows preferences for intermediate level of disorderedness and phosphorylation variation properties . The Casein kinase II is characterized by various substrate recognition motifs [23] , but the main differences are related to presence or absence of a proline residue preceding the phosphorylated site . These two motifs show distinct structural and variation preferences – the former type being more similar to proline-directed kinases and the latter – to non-proline ones . The occurrence of the G protein-coupled receptor kinase 1 ( GRK1 ) near the proline-directed kinases ( red triangle ) can be explained analogously by the presence of a proline residue in the consensus sequence for that kinase . Interestingly , the reported consensus motifs of the MAPKAPK2 kinase ( blue triangle ) do not contain proline residues , however , it is still grouped together with the more variable kinase motifs . After careful examination of the amino acid composition of substrates of the MAPKAPK2 kinase in our data set we found multiple examples that contained proline within +/−6 residue window around the phospho-site . Together with our structural analysis , this suggests that this residue may play an important role in the substrate recognition . Overall , the group of proline-oriented kinases has similar preferences for disorder and phosphorylation variability as the proline-directed group . This observation also extends to the functional relevance of the member kinases to the regulation of the cell cycle . For example , the DNA-dependent protein kinase ( DNA-PK ) is involved in stress response and DNA repair and is known to play a role in the progression of the cell cycle [24] . Furthermore , the MAPKAPK2 kinase is involved in DNA repair processes and thus can provide an alternative to checkpoints activation [25] . Proline-directed kinases such as PLK1 are known to actively regulate the progression of the cell division cycle , thus implying that disordered regions ( which are enriched for prolines ) are subjected to regulation and therefore to variable phosphorylation patterns . We therefore checked if the tendency of phosphorylation variability to scale with the level of disorder persists if we control for proline-directed kinases and excluded all sites modified by such from the data set . Indeed , the effect of lower phosphorylation variability being associated with ordered regions and higher - with disordered regions remained the same .
Previous studies had already found a preference of phosphorylation to occur in loops or disordered regions [6] , [7] . However , those studies generally did not have access to the dynamics of phosphorylation and they therefore based their analysis on the absence or presence of phosphorylation sites alone . Here , we instead made use of a large-scale quantitative phosphorylation data set to investigate a possible relation between the structural features of phosphorylation sites with their degree of regulation . This allowed us to contrast the behavior of less variable sites to those that were dynamically regulated . Our data clearly demonstrate that the propensity of phosphorylation sites to be regulated during the cell division cycle is related to the level of structural organization of the environment in which these sites reside . Furthermore , we discovered that this effect occurs in a graded manner: regions with regular structure are least likely to harbor regulated phosphorylation sites , followed by irregular regions ( short loops or random coils ) . Note that over 90% of the sites were found within disordered structures and their high phosphorylation variability relates them to regulated phosphorylation events . Interestingly , the sets of sites within ordered loops and disordered structures showed significant differences . It has been shown before that different flavors of disordered regions exist with regards to their lengths , amino acid composition , and the conformational transitions that they undergo upon binding [26] , [27] . Liu et al . defined regions with no regular secondary structures ( NORs ) as one specific category of disordered regions . They demonstrated that NORs differ significantly from regular structured loops and argued that these might have different functional implications , a hypothesis which finds support in our study . Functional analysis of the highly variable set of sites revealed enrichment of cell cycle-related , biosynthesis and cellular organization and localization processes ( Table S1 ) . Some examples are RNA , DNA and mRNA processing , localization and transport , regulation of gene expression and biosynthesis . Cell cycle-associated processes such as regulation of the different phases of the cycle , DNA replication and repair , telomere organization and maintenance and chromatin assembly were also strongly over-represented in the variable set of sites . Phosphorylation is an important mechanism for regulation of a myriad of intricate processes during cell division . A detailed study of the cell cycle regulation through phosphorylation focused on functional analysis of protein groups that are up or down regulated at specific time points [2] . These were the proteins that contained sites that reached phosphorylation peaks at S or M phases . As expected , proteins involved in mitotic and cell cycle processes were shown to be maximally phosphorylated at mitosis . Interestingly , Olsen et al . found proteins that regulate metabolic processes to be weakly phosphorylated during S phase and highly phosphorylated at mitosis . An explanation to this discovery is the possibly inhibitory character of phosphorylation on proteins that regulate metabolic processes , as protein synthesis and related functions tend to shut down during mitosis . Furthermore , DNA replication takes place during S phase , which rationalizes the up-regulation through phosphorylation of various proteins involved in DNA replication repair . High phosphorylation of cytokinesis-related proteins in S phase appears to play an important role in the control of the correct segregation of the two daughter cells . The tendency of modification sites in regular structures to be less variable may be facilitated by proximal charged residues acting as stabilizers of the phosphate group . Charged flanking regions offer a suitable environment for hosting a phosphate group and allow for favorable interactions that potentially result in phosphorylation acting on a longer time scale . For instance , these favorable interactions could reduce the efficiency of phosphatases in removing a phosphate group , thereby contributing to the tendency for smaller variation in the phosphorylation level that we observe in our data . In contrast , negatively charged residues could lead to repulsion-driven conformational changes and polarization of the entire protein surface by creating clusters of negatively charged residues . Several mechanisms that are known from literature furthermore contribute to the observed tendency for structural rather than regulatory phosphorylation sites to be present in ordered regions . Specific structural changes due to phosphorylation include stabilization of the N-termini of α-helices via favorable interactions of the added phosphate group with the helix backbone [28] . This is effected by the interaction of the phospho group with the helix dipole moment . Yet , the same modification introduced at the C-terminus would have the opposite effect [29] . The optimal stabilizing position for the phosphate group was estimated as −2 relative to the N-cap of a helix . Additionally , favorable electrostatic interactions between proximal positively charged residues ( e . g . at a helix cap ) and the phosphate group can enhance helix formation . The stabilizing effect of salt bridges formed between a phosphate group and a lysine side-chain has been recognized as one of the strongest possible α-helix inducers [30] . In contrast , the phosphate-guanidinium interaction leads to disruption of the local regular structure [31] . Phosphorylation has also been reported to cause conformational changes in β-sheets and disruption of β-hairpins . In those cases repulsive interactions with an aromatic tryptophan residue in the spatial vicinity of the phospho-site are observed [32] . A related question that arises from our investigation is to what extent the phosphorylation variability of a site is connected to a role in the overall structural re-arrangements of a protein . A phosphorylation event can alter the energy that is required for a conformational change [15] , and thus hinder or facilitate it . Further experiments including 3D structural information or computational models are needed to increase our understanding of the interplay between structure and phosphorylation . Multiple experimental studies show the regulatory role of modification sites that show variation in their phosphorylation patterns and lie within intrinsically disordered regions For example , the cyclin-dependent kinase inhibitor 1B ( p27 ) is an intrinsically unstructured protein , which is multiply phosphorylated and regulates the cell cycle by inhibition of cyclin-dependent kinases ( CDKs ) [33] . The disorderedness of p27 plays an important role in keeping the complex formed between CDK and p27 flexible . Due to this flexibility the segment , which blocks the ATP binding site becomes exposed . This allows a tyrosine residue to become accessible for phosphorylation , upon which the space previously occupied by the inhibitor becomes available for ATP binding . Then the partially reactivated CDK phosphorylates p27 at another residue , which leads to its degradation and allows CDK to regain full activity and guide the progression through the cell cycle [34] . In another example , multiple phosphorylation sites on the transcription regulator Retinoblastoma protein ( Rb ) influence its ability to interact with transcription factors and other regulatory proteins . A detailed structural study reports that the different phospho-sites found within disordered regions induce distinct conformational changes and also serve different functional roles [35] . For instance , one of the modified residues decreases the affinity of Rb for binding the transcription factor E2F by reordering the pocket domain . At the same time another modified site at a loop in the pocket domain induces complete blocking of E2F binding . We found that the set of sites with varying phosphorylation patterns was enriched in amino acids associated with disorder , specifically Pro , Gly and Ser . Interestingly the same sites were more likely to have additional modified residues in their vicinity . Phosphorylation of a protein often occurs at several distinct residues and it has been reported that modification sites tend to cluster and function in a cooperative manner [36] . Mathematical models suggest that this phenomenon leads to an increase in the sensitivity and robustness of the cellular response [37] and may promote a switch-like behavior [38] . In such a case , the exact position of a modification site in a cluster would not be a determining factor on its own , but would rather contribute to a cumulative effect . It would be worth studying how different levels of phosphorylation variability in regions with different structural organization may be implicated in the cellular regulation of the cell cycle . Multiple phosphorylation sites with highly dynamic phosphorylation patterns may be suitable for both rapid and robust response . In contrast , the robustness of the response of sites within regular regions might be achieved on a longer time scale and be related to longer lasting effects of phosphorylation . We showed that phosphorylated residues tend to be more conserved than their equivalent non-modified residues . Conservation of phosphorylated residues has been a broadly debated issue [21] , [39] , but the general consensus appears to be that the overall conservation of phospho-sites is low . Even though statistically it is significantly stronger than that of the equivalent non-modified sites , the effect size is relatively small . Possible explanations include ( i ) loss and gain of phospho-sites at different positions in disordered regions , likely due to clusters of sites acting as functional units regardless of the exact sequence position [40] and ( ii ) potential silent phosphorylation events [39] . The idea that it is the cluster of phosphorylation sites that plays a functional role is becoming increasingly accepted [36] , [37] . The functional roles of multiple phosphorylated residues span a wide range: ( i ) targeting for sub-cellular localization , ( ii ) targeting for degradation , ( iii ) control of protein-protein and protein-nucleic acid interactions ( often through electrostatic effects ) and ( iv ) enhancement of a robust and rapid response to a stimulus [1] . Furthermore , mechanisms of ‘priming’ phosphorylation are also well-known [41] . Here we showed that disordered regions harbor variable sites , which tend to be surrounded by additional phosphorylated sites . This raises the possibility that the variability of these sites is related to some of the above-described phenomena . It is known that disordered regions can facilitate a large number of interaction partners , and that multiple sites can control their association and dissociation . Given the wide range of functions of multiple phosphorylated sites in disordered regions , a larger variability in their phosphorylation patterns may provide an adequate functional mechanism to effect the desired regulation . In contrast , structural regularity imposes certain constrains on the less variable sites . The necessity of evolutionary conservation of the structure tends to prevent the accumulation of disorder-associated serine and threonine residues and a consequent change of their positions . Furthermore , the more rigid structure implies a more limited number of interactions partners . Therefore , we reason that the requirement for regulation for these sites in structured regions can be smaller . Our data allowed us to investigate the kinase preferences of phosphorylation sites with high vs . low levels of regulation . Tyrosine kinases and kinases that require charged residues in their substrate recognition motives clearly preferred sites with smaller phosphorylation variation , whereas proline-directed kinases were clearly associated with sites that were dynamically regulated . Proline is known to be a helix and sheet breaker , due to the planarity of its side-chain . Proline lacks an NH backbone donor to form a hydrogen bond and thus disrupts the formation of regular hydrogen bond patterns , which are the basis of regular structure formation . Due to its unique stereochemistry the proline residue can adopt two different conformational states – cis and trans – and a large number of folded proteins contain both states of the residue . The intrinsic conformational changes resulting from the proline isomerization play an important role in determining the function , ligand recognition and interactions of the protein [42] . For instance , certain kinases , such as MAPKs and CDK2 preferentially modify substrates with the trans isomer [43] . Proline isomerization in a S/TP motif , where S/T is phosphorylated , can also control the opposite step – dephosphorylation , as some phosphatases appear to be conformation-specific and prefer the trans state [44] . Therefore , the preference of proline-directed kinases for sites with higher variation illustrates a connection between dynamic regulation and disordered regions . Our results highlight the central role of proline , as a disorder-promoting residue that is also part of regulatory motifs [45] . The directing role of proline together with the multiple functions associated with disorder explain the more variable character of phosphorylation of sites with these properties that we observe in our study . Furthermore , our statistical analysis of the interplay between structure and phosphorylation variation in relation to specific kinase recognition motifs presents a new approach of describing and classifying protein kinases . We showed that the combination of both properties can be used to gain conceptual and specific insights into regulation . We were able to reproduce known relations and to identify new links between kinases , which may reflect functional dependencies emerging from common regulatory behavior and structural preferences . In conclusion , we have related the tendency of phosphorylation sites to be dynamically regulated throughout the cell cycle with the structural features of the sites . While we have found clear relations between phosphorylation dynamics and protein structure , we are only scratching the surface of what we believe could be an exciting new area at the interface of proteomics and structural biology .
In the data set underlying our analysis [16] , human HeLa S3 cells were labeled with SILAC [46] , [47] to produce three different isotopic forms of lysine and arginine ( light , medium and heavy ) . The light and heavy isotopes were synchronized in six different stages of the cell cycle , while the medium one was kept non-synchronized as a reference . Relative quantification of protein abundances ( protein ratios ) and/or phosphorylation ( phospho-peptide ratios ) were computed by taking the ratio between two cell states at each time point ( i . e . synchronized heavy-labeled cells in S phase and non-synchronized medium-labeled cells ) . In order to account for the possible influence of protein abundance , changes in the phosphorylation ratios between the reference and the stimulated cells were normalized by the protein change . We mainly focused on the phosphorylation ratios as they were available for a larger number of sites compared to the absolute occupancy values . The data set contained information about the UniProt id of the phosphorylated protein , sequence positions of phosphorylated residues , and quantitative measures of phosphorylation ( normalized phosphorylation ratio ) at 6 time points ( i . e . cell cycle phases: G1 , G1S , Early S , Late S , G2 and M ) . In total 1 , 059 proteins and 5 , 173 phosphorylation sites with measured phospho-ratios for each of the six time points of the cell cycle were used in the analysis . In order to assure that the observed phenomena are not due to the properties of the chosen subset of sites , we repeated the analysis of phosphorylation variation between different structural groups with data sets containing five ( 5 , 254 sites ) , four ( 8 , 537 sites ) , and three ( 8 , 731 sites ) time points only ( see Text S1 for details ) . Although slight fluctuations were observed , the main tendencies remained stable and the conclusions did not change . Therefore , no bias in the reduced data set ( i . e . the one containing information about all six time points ) was found . Phosphorylation variation value for each modified site was computed as the standard deviation of the phosphorylation ratios over the six time points . High variation corresponds to sites with temporal variation of phosphorylation ratios ( e . g . a peak is observed in S phase ) , while low variation describes those sites that retain constant or slightly variable phosphorylation fold change during the cell cycle . The secondary structure of each site was predicted with PsiPred [18] . Each site was assigned one of three possible states: ‘H’ for α-helix ( 92 sites ) ‘E’ for β-sheet ( 53 sites ) , and ‘C’ for random coil , turn or loop region ( 5 , 028 sites ) . An intrinsic disordered state was also predicted for each site using DISOPRED [17] with standard settings . We found 498 sites to be in the ‘order’ state while the remaining 4 , 675 were predicted to be in the “disorder” state . Based on a combination of secondary structure and disorder predictions , we defined three distinct structural categories for each phosphorylated site: ( i ) regular regions ( helices and sheets in ordered regions , 145 sites ) , ( ii ) irregular regions ( coils in ordered regions , 353 sites ) , and ( iii ) disordered regions ( coils in disordered regions , 4 , 675 sites ) . Statistical analyses were performed within the R environment [48] and using the in-house statistics work frame Perseus . The lattice package was utilized for comparing distributions of phosphorylation variation in different structural categories . Differences between distributions were assessed with the standard non-parametric Kolmogorov-Smirnov test . In the case of three structural categories , analysis of variance of the phosphorylation fold change was performed using the structural category as an independent variable . Data on phosphorylation site variation and structure predictions are available in the Supporting material ( Table S2 ) . Enrichment of functional Gene Ontology ( GO ) categories was performed with the GOrilla tool [49] . We performed conservation analysis on phosphorylated residues in ordered and disordered regions . The proteins from our data set were mapped to pre-computed EggNOG groups of orthologs [50] . We used the maximum likelihood-based rate4site algorithm to build phylogenetic trees from the EggNOG clusters and to compute residue-based evolutionary rates [51] . Lower evolutionary scores correspond to stronger conservation . The ‘control’ sets of sites were defined as all serine , threonine and tyrosine residues from the phospho-proteins that were not measured to be phosphorylated in our data set with equivalent structural background ( i . e . disordered and ordered as predicted by PsiPred [18] ) . We tested if disordered regions are enriched in multi-phosphorylation sites , as compared to ordered regions . We considered phosphorylation sites with at least one modified neighbor as multi phospho-sites . A neighbor residue is defined as a phosphorylated serine , threonine or tyrosine located within +/−1 , 2 , 3 , 4 or 5 residue-long flanking regions of a central phospho-site . For each cut-off length , we built a contingency table . Each contingency table contained the number of sites with and without neighboring phospho-sites for both ordered and disordered regions . The significance of the enrichment was estimated with the Fisher's Exact Test . The 2D Annotation Enrichment technique [22] enables analysis of the preference of a certain group of elements ( i . e . phosphorylation sites , characterized by the same consensus motif ) for two numerical attributes simultaneously relative to all other elements ( in our case all other phosphorylation sites ) . It employs a two dimensional generalization of the nonparametric two-sample test and uses the Benjamini-Hochberg method to correct for multiple hypotheses testing . We used the default settings to distinguish the statistically significant groups , corresponding to false discovery rate <0 . 01 . We used the Human Protein Reference Database motif definitions in this analysis [23] . The difference between the amino acid content of the flanking regions of the sites with low and the sites with high phosphorylation variation was computed , assessed and visualized with the help of the Two Sample Logo method [19] . The highly variable set was used as the negative set . Residues significantly enriched in a certain position are shown above the horizontal line in the logo . | Cells employ protein phosphorylation – the addition of a phosphate group to serine , threonine or tyrosine residues – as a key regulatory mechanism for modulating protein function . Proteomics technologies can now quantify thousands of phosphorylation sites to reveal the dynamics of phosphorylation at each site in response to a biological process . It is known that phosphorylation does not occur randomly with regard to a protein's structure , but so far the relationship between the dynamics of phosphorylation and these structural properties has not been investigated . Here we relate the relative levels of phosphorylation for more than 5 , 000 sites through the cell cycle to the predicted structural features of the vicinity of the sites . We find that dynamic phosphorylation tends to occur in disordered regions , whereas phosphorylation sites that did not vary as much over the cell cycle are often located in defined secondary structure elements . Kinases that prefer charged amino acids in their substrate motives are more often associated with unchanging sites whereas proline-directed protein kinases phosphorylate cell cycle regulated sites in disordered regions more frequently . The structural organization of the region in which a phosphorylation site resides may therefore serve as an additional control mechanism in kinase mediated regulation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"sequence",
"analysis",
"systems",
"biology",
"spectrometric",
"identification",
"of",
"proteins",
"structure",
"prediction",
"biochemistry",
"genomics",
"protein",
"structure",
"biology",
"computational",
"biology",
"proteomics",
"structural",
"genomics",
"macromolecular",
... | 2013 | Phosphorylation Variation during the Cell Cycle Scales with Structural Propensities of Proteins |
The recombination activity of Escherichia coli ( E . coli ) RecA protein reflects an evolutionary balance between the positive and potentially deleterious effects of recombination . We have perturbed that balance , generating RecA variants exhibiting improved recombination functionality via random mutagenesis followed by directed evolution for enhanced function in conjugation . A recA gene segment encoding a 59 residue segment of the protein ( Val79-Ala137 ) , encompassing an extensive subunit-subunit interface region , was subjected to degenerate oligonucleotide-mediated mutagenesis . An iterative selection process generated at least 18 recA gene variants capable of producing a higher yield of transconjugants . Three of the variant proteins , RecA I102L , RecA V79L and RecA E86G/C90G were characterized based on their prominence . Relative to wild type RecA , the selected RecA variants exhibited faster rates of ATP hydrolysis , more rapid displacement of SSB , decreased inhibition by the RecX regulator protein , and in general displayed a greater persistence on DNA . The enhancement in conjugational function comes at the price of a measurable RecA-mediated cellular growth deficiency . Persistent DNA binding represents a barrier to other processes of DNA metabolism in vivo . The growth deficiency is alleviated by expression of the functionally robust RecX protein from Neisseria gonorrhoeae . RecA filaments can be a barrier to processes like replication and transcription . RecA regulation by RecX protein is important in maintaining an optimal balance between recombination and other aspects of DNA metabolism .
A given segment of chromosomal DNA may be subjected to repair , transcription , replication , and recombination , some or all of these processes occurring within a single cell cycle . Each of these processes poses real or potential molecular problems for the others , and many sources of genome instability lie at the interfaces [1–4] . The interface between replication and transcription has been the subject of numerous studies [5–7] . The role of collisions between replication forks and transient template discontinuities created by DNA repair events in the creation of double strand breaks is now well appreciated , as is the importance of recombinational DNA repair of those breaks [8–16] . In contrast , the potentially negative effects of recombinational DNA repair on other aspects of DNA metabolism have not been systematically investigated . The study described here is based on the following premise: ( a ) recombination systems can have negative impacts on DNA metabolism; ( b ) for that reason , recombinases such as RecA have not evolved to promote their characteristic DNA pairing and strand exchange activities optimally , but instead reflect an evolutionary compromise between the positive and negative effects of recombination; ( c ) substantial increases in recombinase functionality should be possible; and ( d ) since they were not selected during evolution , increases in recombinase functionality may have deleterious effects on cellular DNA metabolism . The bacterial RecA recombinase plays a key role in recombinational DNA repair in E . coli [11 , 17–23] . Inactivation of recombination functions results not only in DNA repair defects , but also in more general genomic instability such as stalled or collapsed replication forks [12 , 14 , 16 , 24–29] . The major activity of RecA protein in homologous genetic recombination reaction is the promotion of DNA strand invasion and strand exchange [30–36] . RecA functions as a helical nucleoprotein filament , which assembles on and dissociates from DNA in several steps [37–41] , and displays a diversity of conformations and dynamics [42 , 43] . RecA has several additional cellular functions . Its filaments , when formed on DNA , act as a coprotease to promote the autocatalytic cleavage of the LexA repressor leading to induction of the bacterial SOS response [44–54] . RecA also activates the mutagenic DNA polymerase V , a function induced late in the SOS response [45 , 55–59] . In this final role , RecA again acts as a coprotease to promote the autocatalytic cleavage of the UmuD subunit [60–62] , and acts itself as an essential subunit of the final activated enzyme [63–67] . In vitro , RecA protein is a DNA-dependent ATPase and promotes DNA strand exchange reactions that mimic its presumed roles in vivo [18–20 , 36 , 68–70] . RecA filament formation on single-stranded DNA ( ssDNA ) begins with a slow nucleation step , followed by rapid 5′ to 3′ extension . RecA filament extension occurs predominantly at the 3′-proximal end . When ATP is hydrolyzed , RecA monomers disassemble primarily at the 5′-proximal end [71–75] . For the strand exchange reaction , RecA filaments form on ssDNA first , followed by pairing with a homologous double-stranded DNA ( dsDNA ) [30 , 35 , 76] . Exchange of strands ensues and can encompass thousands of DNA base pairs as ATP is hydrolyzed [34 , 36 , 70 , 77–79] . Whereas recombination is necessary for double strand break repair and can produce genetic advantages via conjugation , recombination can also lead to genomic damage; e . g . , by aberrant elimination of genomic segments due to recombination between repeated sequences . In principle , the RecA protein can also harm cells and contribute to genome instability in at least three other ways . First , RecA could inappropriately induce the SOS response , with its accompanying cell division regulation and mutagenesis , when it is not needed [80 , 81] . Second , if RecA filaments were not efficiently removed from the DNA when no longer needed , replication and/or transcription could be inhibited . Third , a replication fork or transcription bubble collision with a branched DNA segment undergoing recombinational DNA repair could have a myriad of deleterious consequences . A multi-level regulatory system thus constrains and directs productive RecA-mediated recombination processes in the cell [25] . Several regulatory proteins , RecFOR [72 , 82–87] , RecX [88–92] , DinI [91 , 93 , 94] , RdgC [95] , PsiB [96] , DinD [97] , RadA [80] , and UvrD [98–100] are known to be involved in the regulation of RecA activity through interactions with the RecA nucleoprotein filament . An understanding of this regulation network is one prerequisite to an optimal in vivo harnessing of the recombination capacity of RecA . A particular focus of the current study is the E . coli RecX protein , a conserved and well-characterized RecA regulator expressed from a gene located immediately downstream of the recA gene in E . coli and many other bacteria . The RecX protein is a negative regulator of RecA , required to overcome deleterious effects of overexpression of RecA protein [101–104] . Deletion of the recX gene does not cause a clear phenotype in E . coli [105] , but overexpression reduces induction of the SOS response [106] . In vitro , the E . coli RecX ( EcRecX ) protein inhibits RecA-mediated ATPase and strand exchange activities [106] . The EcRecX protein binds deep within the major helical groove [107] and blocks the extension of a RecA filament by capping its 3′-proximal end while allowing filament disassembly to proceed at the 5′-proximal end [88] . The RecX protein from the bacterium Neisseria gonorrhoeae ( NgRecX ) exhibits a substantially more robust inhibition of RecA protein [90] . Instead of simply capping the growing filament end , the NgRecX appears to create breaks in the filament and increase the number of disassembling ends [90] . In spite of the often modest phenotypes seen in E . coli strains lacking recX function , interaction with RecX protein may be one of the key mechanisms that regulate the stability and recombination function of RecA nucleoprotein filaments in most bacteria . The bacterial RecA protein was first identified from the analysis of mutagenized colonies of an F- culture that were unable to form recombinants after conjugation with an Hfr strain [21] . Conjugational recombination is thus a classic function of RecA that helps define its recombination potential . During bacterial conjugation , once the mating pairs are established , rolling circle replication initiates at the F- plasmid oriT site . Then a nascent single stranded Hfr DNA with a 5′ end enters the F- recipient where it provides a template for lagging strand synthesis [108] . Transfer of DNA ceases at random points and leaves a linear double-stranded Hfr DNA fragment with a leading end and a single stranded overhang of variable length at the distal 3′ end because of the failure to complete synthesis of the complementary strand [109] . In the recipient , genetic crossovers promoted by RecA protein and auxiliary proteins integrate the Hfr fragment into the host genome . Two or more recombination events may occur concurrently or divergently , and the size of the integrated Hfr DNA varies . The recA gene has the capacity to evolve to meet extraordinary cellular challenges such as radiation damage [110–117] . Specific amino acid changes at the subunit-subunit interface produce RecA variants that promote higher levels of conjugational recombination [118] . These results are the basis for the hypothesis that RecA has not evolved for optimal recombination function but instead for an optimal balance between the necessary and potentially damaging consequences of recombination within a particular environmental context . We therefore set out to explore the limits of RecA recombination function . Conjugational recombination has been employed as a selection for RecA variants with the potential to generate higher numbers of crossovers between unlinked genetic markers . Based on the demonstrated functional enhancement observed in some RecA variants with alterations at the subunit-subunit interface [118] , our first effort has focused on this region . In this study , we demonstrate a facile generation of RecA variants that enhance recombination function . The results begin to define some of the resulting biochemical changes that potentially contribute to the enhancement and highlight some of the constraints placed on RecA function in vivo . We also explore the sometimes deleterious cellular consequences of these functional enhancements .
Three questions are addressed below in three successive sections . ( 1 ) Can increases in RecA functionality be obtained ? This involves a directed evolution experiment focused on increasing conjugational recombination function . ( 2 ) What changes in RecA activity give rise to the functional enhancements ? A thorough in vitro characterization of several RecA variants is carried out to address this question . ( 3 ) What are the cellular consequences of a functionally enhanced RecA recombinase ? Cell growth deficiencies associated with RecA functional enhancements are documented and explained .
There are three main conclusions to this work . First , the wild type E . coli RecA protein has not evolved to optimize the genetic exchanges required for conjugational recombination . Substantial increases in recombinase function can be obtained . Second , the observed functional improvements in conjugational recombination may involve many , sometimes subtle changes in protein activity . In this study , not all the changes are subtle . The one feature found in common for the three RecA variants arising most prominently in this study is a more persistent binding to DNA . This is reflected in substantially more rapid displacement of SSB for nucleation onto ssDNA ( Fig 2 ) , and a greatly reduced sensitivity to the RecX inhibitor protein ( Figs 3 and 4 ) . Third , the improvements in conjugation function come only at the cost of a growth deficiency evident for all three RecA variants in competition experiments . For RecA V79L , that growth deficiency reflects the increased DNA binding persistence . Normal growth is restored by overexpression of the more robust RecX protein from Neisseria gonorrhoeae ( Fig 7 ) . The growth deficiencies displayed by cells expressing the other two RecA variants might be explained by a similar mechanism . The work reveals a critical evolutionary compromise between necessary DNA repair processes and potentially deleterious genomic effects . We previously noted the existence of RecA mutant proteins with enhanced recombination potential [118] . In the current study , we have used a selection to generate variants with this same capacity for greater recombination . The selection protocol is robust and reproducible . The improved function of these RecA variants may provide a more robust platform for the continued investigation of key recombinase activities . The increases in recombination documented in this study are reflected in many changes in RecA protein activities , but many of them are subtle and unlikely to account for the enhancement on their own . ATP hydrolytic rates are increased , but the coupling between ATP hydrolysis and DNA strand exchange appears to be reduced ( DNA strand exchange actually proceeds slower rather than faster ) . DNA pairing is improved for some of the RecA variants , but this is evident only at high pH . Our working hypothesis is that the enhanced conjugational recombination reflects an overall increase in RecA filament persistence on the DNA . This is seen in multiple assays in which the rates of filament nucleation on SSB-coated ssDNA are increased for the RecA variants , and the rates of RecA filament disassembly ( in the presence of RecX or RecA K72R ) are decreased . Since ATP hydrolysis rates increase , a reduction in filament disassembly must come about via a decreased coupling between ATP hydrolysis and RecA subunit dissociation at the 5'-proximal end . That persistence in binding is perhaps best encapsulated by the greater overall binding of the RecA protein variants to short oligonucleotide DNA substrates . To promote conjugational recombination , RecA protein must bind to the transferred single stranded DNA and carry out a complete genomic search for homology . In this context , more persistent binding by a RecA filament makes sense . Improvements in this parameter should increase the length of time available for a homology search and improve chances that a productive pairing will occur . In the context of recombinational DNA repair at a replication fork , persistent binding of a RecA filament to DNA is not necessary and probably detrimental . At a replication fork , the homologous DNAs to be paired are generally in close proximity; a widespread genomic search for homology is not needed . A RecA filament that overstays its welcome will simply be a barrier to productive replication restart . A substantial reduction in sensitivity to the inhibitory EcRecX protein , seen for all three characterized RecA variants , makes a significant contribution to the overall DNA binding persistence that would occur in the cell . Elimination of recX function does not have major phenotypic consequences in wild type E . coli cells [105] . RecX helps to maintain an optimal balance between active ( bound ) and disassembled RecA protein in the cell . A decline in sensitivity to EcRecX helps lead to growth deficiencies , and expression of a more robust version of RecX protein can restore balance . The evolutionary significance of RecX is thus rendered more apparent . In addition to RecX , the UvrD helicase has a major role in removing RecA filaments from the DNA to keep them from impeding other aspects of DNA metabolism [98–100 , 148 , 149] . Recent work has shown that UvrD is defective in displacing RecA variants with enhanced DNA binding properties such as RecA E38K [100] . This might also help explain the observed growth deficiencies in strains expressing our RecA variants . The current work begins to build a case that RecA filaments can represent substantial barriers to replication and possibly to transcription , and that those barriers have cellular consequences . An earlier and extreme example of RecA as a barrier came in the form of the RecA K250R mutant , which hydrolyzes ATP six times more slowly than the wild type protein [126] . This leads to a six-fold decrease in rates of filament dissociation from DNA , and an accompanying six-fold decline in cell growth rate [126] . Suppressors arise quickly in strains expressing RecA K250R , most of them inactivating the mutant recA gene [126] . Collisions between replication forks and bound recombinase filaments could have genome instability implications in all cells . The mutagenesis and selection method used here focused on one region of the protein representing about 17% of the amino acid residues in RecA . Within this region , we have queried every possible single base substitution with 90% confidence , and the library included some double , triple , and quadruple mutant proteins ( the library did not cover nearly all the possible combinations of multiple mutations ) . The region selected , between residues 79 and 137 , is not the only part of the protein with the potential to generate variants with increased recombination potential . It was selected due to the presence of changes in the region that were previously shown to produce the desired phenotype . A complete assessment of changes that could affect RecA function in this way will require screens focusing on other recA gene segments . RecA was originally discovered due to its effects on conjugational recombination [21] , and many early studies of recA were carried out in this context . The lack of optimization for conjugation during evolution , coupled to the growth deficiency that accompanies enhancement of this process , provides yet another argument that recombinases did not evolve to promote chromosomal genetic exchanges per se [11–14 , 16 , 126 , 150] . Instead , recombination evolved to repair double strand breaks [11–14 , 16 , 126 , 150] . Genetic exchanges during conjugation , and perhaps eukaryotic meiosis , reflect an evolutionary repurposing of pre-existing systems . The functional compromise between the positive and negative effects of recombinases and recombination seems likely to take different forms in different species . The residues affected by the more prominent mutations identified in our two separate selective screens are highlighted in Fig 8 ( orange/red ) . Key residues bracketing the ATPase active site at the subunit-subunit interface ( K72 on one side [151 , 152] and K248 and K250 on the other [126 , 136] ) are shown in blue . Some of the residues identified in this study are at the subunit-subunit interface ( D100 , D102 , E86 , C90 , A131 ) , but others are not ( V79 , I93 ) . We hypothesize that the variants in all of these residues may affect coupling of ATP hydrolytic events to conformational changes and/or general allosteric communication between subunits . This communication may in turn affect rates of filament disassembly . Continued work should elucidate subtle structure-function relationships that affect all aspects of the coupling of ATP hydrolysis to RecA function . The wild type RecA protein of Escherichia coli seems to have evolved to do its job quickly and get out of the way .
Supercoiled double-stranded DNA and circular single-stranded DNA from M13mp18 bacteriophage were prepared as described previously [153] . Linear double-stranded DNA for strand exchange reactions was generated by complete digestion of supercoiled DNA with PstI restriction endonucleases . For D-loop forming reaction assays , 8 units of T7 exonucleases per μg of DNA were used for an additional digestion of double-stranded DNA to form 150 nt long 3' overhang . The concentration of dsDNA and ssDNA substrates were determined using absorbance at 260 nm and the conversion factors 108 μM A260-1 and 151 μM A260-1 , respectively . DNA concentrations are expressed in terms of total nucleotides . Donor EAW175 and recipient EAW188 strains were constructed by P1 transductions from several strains . EAW175 was made by a consecutive P1 transduction of , first , the Δ ( metA ) ::kan llele from SS6311 into CAG5052 ( KL227 btuB3191::Tn10 metB1 relA1 89′→6′ ) to obtain an intermediate strain EAW173 , checked by Tetr and Kanr phenotypes , then followed by kan flipping out and , second , the ilvO::kan allele from SS4761 into EAW173 strain , checking for both Tetr and Kanr phenotypes . To make recipient strain , kan was flipped out first from SS338 ( Δ ( attB ) ::psulA-gfp Δ ( metE ) 100::kan ) strain and intermediate strain EAW174 was made by P1 transduction of the Δ ( aroB ) ::kan allele from SS2495 to SS3388 . The ΔrecA::kan allele from EAW20 was then transferred to EAW174 by P1 transduction to make recipient EAW188 . EAW188 was transformed with pT7POL26 . EAW334 = MG1655 with recA I102L on the chromosome in the recA locus EAW334 was constructed using a variation of the procedure of Datsenko and Wanner [154] . A plasmid with the MG1655 region from the 200 bases upstream of the recA gene to 210bp downstream of the stop of the recX gene was constructed . A cassette containing the KanR gene flanked by a mutant FRT and a wt FRT was added just downstream of the stop of the recX gene to use as a removable marker . This plasmid was designated pEAW675 . A plasmid containing the recA gene with an I102L mutation was digested with NcoI and EcoRI and the mutant DNA fragment was ligated into pEAW675 digested with the same enzymes . The plasmid , designated pEAW884 was directly sequenced to confirm the presence of recA I102L . pEAW884 was used as a template in a PCR with primers consisting of bases 200–180 before the start of recA , and 210–192 after the end of recX . The PCR product was electroporated into EAW20 , which is MG1655Δ recA , containing the plasmid pKD46 . A Kanamycin resistant colony was used as template in a PCR , and the product was sequenced to confirm the presence of recA I102L . The KanR cassette was popped out by transforming the strain with the FLP expression plasmid pLH29 , and incubating with IPTG . EAW394 = MG1655 with recA V79L , and EAW410 = MG1655 with recA E86G+C90G on the chromosome in the recA locus . EAW394 and 410 were constructed in a manner similar to EAW334 , with the plasmid containing recA V79L , or E86G/C90G digested with NcoI and EcoRI and ligated into pEAW675 digested with the same enzymes . pEAW947 = Ng recX in pBAD/Myc-HisA . Plasmid Ng recX ( Siefert Lab ) was used as the template in a PCR with a primer consisting of a BspHI site followed by the bases 5–32 of the Ng recX gene . The BspHI site contains bases 1–4 of the start of the Ng recX gene . A change was made for better codon use at Leu7 . The other primer consisted of a BamHI site followed by the last 24 bases of the Ng recX gene . The PCR product was digested with BspHI and BamHI and ligated to pBAD/Myc-HisA ( Invitrogen ) digested with NcoI and BglII , enzymes having compatible cohesive ends with BspHI and BamHI . The resulting plasmid , designated pEAW947 was directly sequenced to confirm the presence of Ng recX . Construction of ΔaraBAD strains EAW214 , 564 , 568 , 569 . EAW214 was constructed using a variation of the procedure of Datsenko and Wanner [154] . pEAW507 , a plasmid containing a mutant FRT-KanR- wt FRT cassette , was used as a template in a PCR . The primers consisted of the 51 bases before the start of the araBAD promoter +20 bases before the mutant FRT , and the 51bases after the stop of araD+21 bases after the other FRT . The PCR product was electroporated into MG1655/pKD46 , and a Kanamycin resistant colony was selected . DNA from this colony , designated EAW214 , was used as a template in a PCR to confirm the presence of the FRT-Kan R-FRT replacing the araBAD promoter and genes on the chromosome . P1 transduction was used to transfer the araBAD deletion into EAW394 , 334 , and 410 . The resulting strains were designated EAW564 , 568 , and 569 . DNA from these strains was used as templates in PCRs to confirm the presence of the FRT-Kan R-FRT replacing the araBAD promoter and genes on the chromosome . EAW575 , and 578 = Gc recX on the chromosome in the Ec recX locus of wt recA , and recA V79L . EAW575 , and 578 were constructed using a variation of the procedure of Datsenko and Wanner [154] . The mutant FRT-KanR- wt FRT cassette from pEAW507 was excised by EcoRI and SalI digestion , and inserted after the end of the Gc recX gene of plasmid pEAW947 , which was digested with the same enzymes . The resulting plasmid , designated pEAW1016 , was used as template in a PCR with primers consisting of the 51bp of the E . coli chromosome before the start of recX +the first 21 bp of the Gc recX gene , and the 51bp of the E . coli chromosome after the end of recX+21 bases after the wt FRT of pEAW1016 . The PCR product was electroporated into MG1655 , and a kanamycin sensitive version of EAW394 , both containing the plasmid pKD46 . DNA from these strains was used as templates in PCRs , and sequenced to confirm the presence of wt recA + Gc recX for EAW575 , and recA V79L+Gc recX for EAW578 . The E . coli RecX , Neisseria gonorrhoeae RecX [90] , SSB [90] and the wild type RecA protein [155 , 156] were purified as previously described . The RecA V79L , RecA I102L , RecA E86G/C90G mutant proteins were purified by the same means as the wild type RecA protein with the following modifications . The plasmids encoding the mutant recA genes were transformed into the ΔrecA and nuclease-deficient strain STL2669 . The Polyethylenimine pellet was washed with R Buffer ( 20 mM Tris-Cl buffer ( 80% cation , pH 7 . 5 ) , 0 . 1 mM EDTA , 10% ( w/v ) glycerol , 1 mM dithiothreitol ) and extracted twice with R Buffer plus 300 mM ammonium sulfate . After precipitation by ammonium sulfate to 50% saturation , the pellet was resuspended in R buffer plus 1 M ammonium sulfate . Proteins were purified using chromatography on some combination of Butyl-Sepharose , Ceramic HAP , Source 15S , Source 15Q , DEAE sepharose columns . Between columns , peak fractions were identified by SDS-PAGE and pooled together before dialyzing , if necessary . The concentrations of E . coli RecX , SSB and RecA proteins were determined from the absorbance at 280 nm using the native extinction coefficient 2 . 57 × 104 M−1 cm−1 [88] , 2 . 38 × 104 M−1 cm−1 [157] and 2 . 23 × 104 M−1 cm−1 [158] , respectively . The purified proteins were free of detectable nuclease activities on double stranded DNA and single stranded DNA . The oligonucleotide cassettes-directed method [159] from earlier study was modified to create randomized libraries for E . coli RecA protein . The cassette mutagenesis procedure involves the synthesis of a small , double-stranded DNA molecule that can be ligated into a larger vector fragment to reconstruct the gene of interest [160] . As an in vivo expression vector , the plasmid pACYC184 with T7 promoter and recA gene was digested with restriction enzymes SapI and PstI to generate a backbone fragment . A double-stranded DNA fragments corresponding to the region between sites SapI and PstI was made by annealing three separate oligonucleotides . Only two oligonucleotides were randomly mutated through incorporation of degenerate DNA sequence using different molar ratios of four nucleotides as mixtures during synthesis . The ratios were 99% to 0 . 33% and 98 . 5% to 0 . 5% , corresponding to wild-type base to each of the other bases . These two oligonucleotides were placed abreast and annealed to the other complementary oligonucleotide to make randomized double-stranded DNA molecule . The last oligonucleotide was synthesized without mutations to avoid too many mismatches , thus increase annealing efficiency between complementary strands . This small DNA inserts were then ligated to the backbone fragment to generate the mutagenized plasmid library , which was transformed into the DH5α cells by electroporation . More than 27 , 500 grown colony isolates were combined in a pool and the population of recA gene-bearing plasmid was purified . The recipient cells were transformed with the purified plasmid DNA pool for conjugational assay . In order to determine the number of clones necessary to achieve 90 and 95 percent confidence of the presence of all 531 clones , a Monte Carlo simulation was designed . A simple code was written , using Python ( http://www . python . org ) , to pick a number out of 531 at random and keep picking numbers until the entire set of numbers 1 to 531 was selected . The total number of selections needed to complete the set was recorded for each trial , with the trial ending when the entire set of numbers was selected . With each number from 1 to 531 representing a different possible mutation , the total number of clones needed to get all 531 mutations for each trial was represented by this total . One million trials were run using this code . A 90 percent confidence level that all 531 mutations are present in a group of clones means that more than 900 , 000 trials must have a number of clones less than the group in question . Similarly , to be 95 percent confident , 950 , 000 trials must have a number of clones less than the group in question . This was done in Excel by totaling up the histogram data from the simulation and finding the minimum number of clones needed to obtain the entire set of 531 mutations in at least 900 , 000 and 950 , 000 trials , respectively . See the supplementary materials for the source code for this simulation . Conjugation was carried out essentially as described before [161] with following exceptions . Donor strain was grown at 37°C in Luria-Bertani ( LB ) broth with Tetracycline until an optical density ( OD600 ) of 0 . 7 was reached . Recipient strain was grown with Chloramphenicol , Kanamycin and Streptomycin until an optical density ( OD600 ) of 0 . 4 was reached and then induced for 40 minutes with Isopropyl β-D-1-thiogalactopyranoside ( IPTG ) of final 40 μM . The concentration of IPTG was optimized for producing about 700–800 recombinant colonies per 1 , 000 , 000 donors at a cross requiring transfer of one marker ( two crossovers ) , using recipients expressing wild type RecA protein from the same expression used in library construction . Both strains were spun down and gently resuspended in the amount of initial volume of fresh LB broth to remove antibiotics . Mating was carried out by mixing 200 μl of donor cells with 1800 μl of recipient cells and incubating 100 min at 37°C . The 200 μl of the mating mixture was mixed with 3 ml of pre-warmed 0 . 7% Bacto agarose solution to prevent additional mating and immediately poured onto a selective media plate . The plate was sat for a few minutes at a room temperature and turned upside down and incubated for 40 hours at 37°C . During repeated rounds of conjugation , all of mating mixtures were poured onto selective media plates . After each round , the resulting recombinant colonies were combined in a pool and the population of recA gene-bearing plasmids was isolated and stored . The isolated plasmid pool was introduced into a new batch of recA- recipient cells for the next round of conjugational cross . The plasmid population pools isolated after 4th , 5th and 6th rounds of conjugational assay were selected and subjected to Illumina deep sequencing . Each pool of mutated recA gene was PCR-amplified with lower number of cycles and submitted to University of Wisconsin Biotechnology Center ( UWBC ) for amplicon library preparation and sequencing using the Illumina genome analyzer . Libraries were prepared for sequencing according to the manufacturer’s instructions with the following modifications . The initial input into each reaction was 100 ng of amplicon DNA , size selection procedure was omitted since library samples were single PCR products and PCR amplification was performed with 11 total cycles . Data analysis was performed at the UWBC Bioinformatics Resource Center as following . Paired-end HiSeq data was merged using FastqJoin ( http://code . google . com/p/ea-utils/wiki/FastqJoin ) . The un-joined reads were trimmed for low quality bases using the fastx toolkit ( http://hannonlab . cshl . edu/fastx_toolkit ) and joined by concatenating the reverse complement of reverse read to the end of forward read . The merged and joined sequences were then aligned to the recA sequence using the classic Smith—Waterman algorithm . The alignment adjusted for gaps and missing sequence data to produce a nucleotide counts by position summary . The unique reads were also counted and translated using the standard codon translation table . Sequences and the corresponding translations were evaluated for the variant and effects and then ranked according to the number of supporting combined reads . A modified Alcian method was used to visualize RecA filaments on cssDNA . Activated grids were prepared as described previously [133] . All reactions were prepared by pre-incubating 3 μM RecA and 5 μM M13mp18 cssDNA , 25 mM Tris-OAc ( 80% cation ) buffer , 5% ( w/v ) glycerol , 3 mM potassium glutamate , and 10 mM Mg ( OAc ) 2 . All reactions were carried out at 37°C . For an ATP regeneration system , 10 units/ml pyruvate kinase and 3 . 0 mM phosphoenolpyruvate were also added to pre-incubation mixture . After 10 min pre-incubation , 3 mM of ATP and 0 . 5 μM of SSB were added . After another 7 min , the RecX protein to a 100 nM or the equivalent volume of RecX storage buffer was added . ATPγS was then added to 3 mM , followed by 1 minute incubation . The reaction solution was then diluted to a final DNA concentration of 0 . 0004 μg/μl with 200 mm ammonium acetate , 10 mm HEPES ( pH 7 . 5 ) and 10% glycerol and adsorbed onto Alcian grids for 3 min . The grid was then touched to a drop of the above buffer , followed by floating on a drop of the same buffer for 1 min . The sample was then stained by touching to a drop of 5% uranyl acetate followed by floating on a fresh drop of 5% uranyl acetate for 30 seconds . Finally , the grid was washed by touching to a drop of double distilled water followed by immersion in two 10 ml beakers of double distilled water . After the sample was dried , it was rotary-shadowed with platinum . This protocol is designed for visualization of complete reaction mixtures , and no attempt was made to remove unreacted material . Although this approach should yield results that provide insight into reaction components , it does lead to samples with a high background of unreacted proteins . To determine the proportion of the molecules observed that were either fully or partially coated by RecA protein or bound only by the SSB protein , at least two separate regions of two to three independent experiments were counted at an identical magnification for each sample . "Full" filaments completely encompassed the circular DNA molecule or had small discontinuities in the regular striated pattern of the filament . A molecule was considered gapped if it had a detectable region of SSB-coated DNA of any size . Imaging and photography were carried out with a TECNAI G2 12 Twin Electron Microscope ( FEI Co . ) equipped with a GATAN 890 CCD camera . Digital images of the nucleoprotein filaments were taken at X 15 , 000 and X 26 , 000 magnification as is evident from the scale bar . The observed lengths of the RecA filaments and the length of SSB-coated DNA were used to assign counted molecules to five categories: full filaments , medium filaments , small filaments , very small filaments or SSB/DNA molecules . Linearized DNA molecules , likely originating from shearing force during pipetting , were also counted . A RecA filament was considered a full filament if it does not have a detectable region of SSB coated DNA or a region that appeared to reduce the filament length by less than 10% . Medium filaments were smaller in length than full filaments , but still had substantial regions of nucleoprotein filament . Small filaments were generally less than half the length of full filaments , and often had regions of obvious SSB binding . Very small filamented molecules are those with just detectable segments of RecA filamented regions , with the rest of the molecule coated with SSB . With the total number of molecules counted as 100% , the percentage of each type of nucleoprotein filament was calculated . At least four separate regions of the grids encompassing at least 500 DNA molecules for each time point were counted at the identical magnification for each sample . For each RecA variant , length measurements were carried out using Metamorph analysis software on 10 molecules selected at random from each of the five categories ( excepting linears ) that represented more than 10% of the total molecules in a given sample . In total , between 20 and 70 molecules from each of these five classes were measured , bound to the same ssDNA substrate . The complete set of measurements is provided in Table 2 . Each filament was measured three times , and the average length was calculated . The 500 μm scale bar was used as a standard to calculate the number of pixels per μm . Each nucleoprotein fragment length , originally measured by Metamorph in pixels , was thus converted to μm . A coupled spectrophotometric enzyme assay [162 , 163] was used to measure the DNA-dependent ATPase activities of the RecA protein . In this assay , the regeneration of ATP from ADP by pyruvate kinase and phosphoenolpyruvate was coupled to the oxidation of NADH by lactate dehydrogenase . The conversion of NADH to NAD+ was monitored as a decrease in absorbance at 380 nm rather than 340 nm , in order to remain in the linear range of the spectrophotometer for the duration of the experiment . The amount of ATP hydrolyzed over time was calculated using the NADH extinction coefficient at 380 nm of 1 . 21 mM-1cm-1 . The assays were carried out on either a Varian Cary 300 dual beam spectrophotometer equipped with a temperature controller and a 12-position cell changer or Perkin Elmer Lambda 650UV/Vis spectrometer with 9+9 cell changer . The cell path length was 1 . 0 cm and the band pass was 2 nm . All reaction samples contained 25 mM Tris-OAc ( 80% cation , pH 7 . 4 ) , 1 mM DTT , 3 mM potassium glutamate , 10 mM Mg ( OAc ) 2 , 5% ( w/v ) glycerol , an ATP regeneration system ( 10 units/ml pyruvate kinase and 3 . 0 mM phosphoenolpyruvate ) , 10 units/ml lactate dehydrogenase , 2 . 0 mM NADH , 5 M M13mp18 cssDNA or poly ( dT ) and 3 μM RecA proteins unless otherwise specified in the Fig legends . DNA three-strand exchange reactions were carried out at 37°C in 25 mM Tris-OAc ( 80% cation , pH 7 . 4 ) , 1 mM DTT , 3 mM potassium glutamate , 10 mM Mg ( OAc ) 2 , 5% ( w/v ) glycerol , an ATP regeneration system ( 10 units/ml pyruvate kinase and 2 . 0 mM phosphoenolpyruvate ) . The final pH after the addition of all reaction components was 7 . 4 . The wild-type RecA protein and RecA mutant proteins ( 3 . 5 μM ) were preincubated with 10 μM M13mp18 cssDNA for 10 min . The mixture of SSB protein ( 1 μM ) and ATP ( 3 mM ) was then added , followed by 10 min of incubation . DNA strand exchange reactions were initiated by the addition of M13mp18 lds ( 20 μM ) . Strand exchange reactions with EcRecX proteins were also carried out with the same concentration of DNA and proteins . For this reaction , RecX protein ( 0 . 1 μM ) was added and incubated for 10 min before the reactions were initiated by adding ldsDNA . A 15 μl reaction aliquots were mixed with 5 μl of a solution containing 3 μl of Ficoll ( 0 . 4% bromophenol Blue , 0 . 4% xylene cyanol , 25% Ficoll , 120 mM EDTA ) and 2 μl of 10% ( w/v ) SDS , and incubated for 40 min at 37°C to stop the reaction . Aliquots were loaded on a 0 . 8% agarose gel , and electrophoresed at 50 mA overnight at room temperature . The DNA was visualized by ethidium bromide staining and exposure to UV light . Gel images were captured with GE Typhoon FLA 9000 biomolecular imager and quantified using ImageQuant TL software from GE healthcare . D-loop forming reaction assays were carried out at 37°C in 25 mM Tris-OAc ( 80% cation , pH 7 . 4 ) , 1 mM DTT , 3 mM potassium glutamate , 10 mM Mg ( OAc ) 2 , 5% ( w/v ) glycerol , an ATP regeneration system ( 10 units/ml pyruvate kinase and 2 . 0 mM phosphoenolpyruvate ) . The final pH after the addition of all reaction components was 7 . 4 . The wild-type RecA protein and RecA mutant proteins ( 2 μM ) were preincubated with 10 μM 3' overhung M13mp18 ldsDNA for 10 min . The mixture of SSB protein ( 1 μM ) and ATP ( 3 mM ) was then added , and incubated for an additional 10 min . The reactions were started by adding 10 μM M13mp18 cds . A 15 μl reaction aliquots were mixed with 5 μl of a solution containing 3 μl of Ficoll ( 0 . 4% bromophenol Blue , 0 . 4% xylene cyanol , 25% Ficoll , 120 mM EDTA ) and 2 μl of 10% ( w/v ) SDS , and incubated for 40 min at 37°Cto stop the reaction . Aliquots were loaded on a 0 . 8% agarose gel , and electrophoresed at 50 mA overnight at room temperature . The DNA was visualized by ethidium bromide staining and exposure to UV light . Gel images were captured with GE Typhoon FLA 9000 biomolecular imager and quantified using ImageQuant TL software from GE healthcare . For UV irradiation sensitivity test , cells ( EAW 105 , 334 , 394 and 410 ) were grown , serially diluted , and 100 μl of appropriate dilutions were spread onto LB plates . Dilutions for samples/treatments were empirically determined . The plates were then exposed to UV in a calibrated Spectrolinker XL-1000 UV crosslinker ( Spectronics Corp ) to the dose indicated . After incubating at 37°C overnight , the colonies were counted and divided by the dilution factor to get cfu/ml . For percent survival , colony counts on the treated plates were divided by the counts on untreated plates . For ciprofloxacin experiments , plates were poured with LB agar containing the ciprofloxacin ( 0 . 01 μg/ml ) . Cells were grown , serially diluted , and spot plated ( 10 μl , 10−2 through 10−6 ) on the ciprofloxacin-containing plates . Pictures were taken after growing overnight at 37°C . Wild type cells , and in cells expressing any of several variant forms of RecA protein at the normal recA chromosomal locus , were modified to carry a neutral Ara– mutation ( which confers a red color on colonies when grown on tetrazolium arabinose ( TA ) indicator plates ) to permit color based scoring of mixed populations [147] . Cells from a fresh single colony of each strain were cultured in LB broth [161] at 37°C with aeration . After growth overnight , competition cultures were started by inoculating 3 ml fresh LB broth with 30 μl of competition Ara+ or Ara– strains and grown overnight at 37°C with shaking . Equal amounts of strains to be compared were mixed . A sample of the mixture was taken , diluted by a factor of 10−6 , and plated on tetrazolium arabinose indicator plates . Then , 3 ml fresh LB broth was inoculated with 30 μl of the mixture , and grown overnight . The plating , inoculation , and growth cycle was repeated two more times . For experiments using cells containing plasmid pEAW947 ( expressing NgRecX protein from the araBAD promoter ) , media was supplemented with 1% arabinose . White and red colonies were counted on plates containing 40–300 colonies , and the % of cells expressing mutant RecA proteins was determined . For counting colonies , plates with fewer than 20 colonies of either competitor were excluded to reduce the effect of outliers caused by low counts [164] . Overnight cultures were diluted 1:100 in fresh LB , and 200 μl was added to the wells of a black-walled , clear-bottom 96 well plate ( Corning ) . For each sample , three overnights were grown from separate colonies , and each overnight filled three wells in the plate ( three biological and three technical replicates , for nine total wells per sample ) . The plate was inserted into a Tecan infinite M1000 Pro plate reader . A program was used to incubate the plate at 37°C with orbital shaking . Every 10 min , the plate was briefly shaken linearly , and the OD600 and 509 nm emission ( with 474 nm excitation ) was read . SOS response was induced by adding ciprofloxacin ( 0 . 005 μg/ml ) 3 hours after inoculation . | The genetic recombination systems of bacteria have not evolved for optimal enzymatic function . As recombination and recombination systems can have deleterious effects , these systems have evolved sufficient function to repair a level of DNA double strand breaks typically encountered during replication and cell division . However , maintenance of genome stability requires a proper balance between all aspects of DNA metabolism . A substantial increase in recombinase function is possible , but it comes with a cellular cost . Here , we use a kind of directed evolution to generate variants of the Escherichia coli RecA protein with an enhanced capacity to promote conjugational recombination . The mutations all occur within a targeted 59 amino acid segment of the protein , encompassing a significant part of the subunit-subunit interface . The RecA variants exhibit a range of altered activities . In general , the mutations appear to increase RecA protein persistence as filaments formed on DNA creating barriers to DNA replication and/or transcription . The barriers can be eliminated via expression of more robust forms of a RecA regulator , the RecX protein . The results elucidate an evolutionary compromise between the beneficial and deleterious effects of recombination . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Directed Evolution of RecA Variants with Enhanced Capacity for Conjugational Recombination |
The TAR DNA-binding protein 43 ( TDP-43 ) has been identified as the major disease protein in amyotrophic lateral sclerosis ( ALS ) and frontotemporal lobar degeneration with ubiquitin inclusions ( FTLD-U ) , defining a novel class of neurodegenerative conditions: the TDP-43 proteinopathies . The first pathogenic mutations in the gene encoding TDP-43 ( TARDBP ) were recently reported in familial and sporadic ALS patients , supporting a direct role for TDP-43 in neurodegeneration . In this study , we report the identification and functional analyses of two novel and one known mutation in TARDBP that we identified as a result of extensive mutation analyses in a cohort of 296 patients with variable neurodegenerative diseases associated with TDP-43 histopathology . Three different heterozygous missense mutations in exon 6 of TARDBP ( p . M337V , p . N345K , and p . I383V ) were identified in the analysis of 92 familial ALS patients ( 3 . 3% ) , while no mutations were detected in 24 patients with sporadic ALS or 180 patients with other TDP-43–positive neurodegenerative diseases . The presence of p . M337V , p . N345K , and p . I383V was excluded in 825 controls and 652 additional sporadic ALS patients . All three mutations affect highly conserved amino acid residues in the C-terminal part of TDP-43 known to be involved in protein-protein interactions . Biochemical analysis of TDP-43 in ALS patient cell lines revealed a substantial increase in caspase cleaved fragments , including the ∼25 kDa fragment , compared to control cell lines . Our findings support TARDBP mutations as a cause of ALS . Based on the specific C-terminal location of the mutations and the accumulation of a smaller C-terminal fragment , we speculate that TARDBP mutations may cause a toxic gain of function through novel protein interactions or intracellular accumulation of TDP-43 fragments leading to apoptosis .
Transactive response DNA binding protein with a molecular weight of 43 kDa ( TDP-43 ) is a ubiquitously expressed nuclear protein encoded by the TARDBP gene , located on chromosome 1p36 . TDP-43 was identified as the major disease accumulated protein in ubiquitinated neuronal cytoplasmic ( NCI ) and neuronal intranuclear inclusions ( NII ) , that define a growing class of neurological diseases , collectively referred to as TDP-43 proteinopathies [1]–[5] . These diseases include amyotrophic lateral sclerosis ( ALS ) , frontotemporal lobar degeneration ( FTLD ) with ubiquitin immunoreactive , tau negative inclusions ( FTLD-U ) and FTLD with motor neuron disease ( FTLD-MND ) . In TDP-43 proteinopathies , TDP-43 is relocated from the nucleus to the cytoplasm and sequestered into inclusions that are mainly composed of hyperphosphorylated and C-terminally truncated TDP-43 fragments [4] , [6] , [7] . TDP-43 immunoreactive histopathology has also been reported in 20–30% of patients with Alzheimer's disease ( AD ) , 70% of patients with hippocampal sclerosis ( HpScl ) , 33% of patients with Pick's disease and in a subset of patients with Lewy-body related diseases [8]–[12] . TDP-43 is a highly conserved protein , containing 2 RNA recognition motifs and a C-terminal glycine-rich domain , known to promote protein-protein interactions [13] . TDP-43 can bind to the common microsatellite region ( GU/GT ) n in RNA and DNA , with proposed functions in transcriptional regulation [13] . Most recent research has focused on the role of TDP-43 in the regulation of exon 9 alternative splicing in the cystic fibrosis transmembrane conductance regulator gene , however , additional targets have been identified and others likely await identification [14] , [15] . TDP-43 has also been implicated in microRNA biogenesis [16] . ALS and FTLD-U are etiologically complex disorders with genetic as well as environmental factors contributing to the disease . A positive family history is reported in 5–10% of ALS patients and in up to 50% of FTLD-U patients , often with an autosomal dominant pattern of inheritance [17]–[19] . Mutations in the Cu/Zn superoxide dismutase gene ( SOD1 ) account for ∼10–20% of familial and 1–2% of apparent sporadic ALS patients [20] . However , TDP-43 inclusions were not present in SOD1 mutation carriers , suggesting a distinct disease mechanism in these patients [21] . The genetic basis of FTLD-U is just starting to be understood [19] . Loss-of-function mutations in the gene encoding the secreted growth factor progranulin ( PGRN ) are a major known cause of familial FTLD-U [22] , [23] , explaining up to 25% of patients worldwide [24] . Other rare genetic causes of familial FTLD-U include mutations in the valosin containing protein gene ( VCP ) and the gene encoding the charged multivesicular body protein 2B ( CHMP2B ) , while some families with a combination of FTLD and ALS show genetic linkage to a locus on chromosome 9p [25]–[29] . Since rare missense mutations and multiplications have been identified in genes encoding the major constituents of the pathological deposits in several neurodegenerative diseases , we hypothesized that mutations in TARDBP may contribute to the development of TDP-43 proteinopathies . In fact , the first missense mutations in TARDBP were recently discovered in 2 autosomal dominant ALS families and 2 sporadic ALS patients , supporting the central role for TDP-43 in disease pathogenesis [30] , [31] . A large population-based study further identified 8 different missense mutations in 3 familial and 6 sporadic ALS patients and showed accumulation of a detergent-insoluble TDP-43 protein product of ∼28 kDa [32] . Here , we report on the extensive mutation screening of TARDBP in a diverse cohort of clinical and pathological confirmed patients with neurodegenerative diseases characterized by TDP-43 pathology , which led to the identification of 3 additional ALS families with TARDBP mutations . We further show accumulation of proteolytic cleaved fragments with a molecular weight of approximately 35 and 25 kDa in lymphoblastoid cell lines derived from TARDBP mutation carriers .
We performed in silico analyses of the TARDBP gene structure by alignment of human spliced expressed sequence tags listed in the UCSC genome browser ( http://genome . ucsc . edu/ ) . This led to the identification of a novel 5′ non-coding exon ( exon 0 ) in addition to the known non-coding exon 1 and the 5 coding exons that are included in the TARDBP reference mRNA sequence ( NCBI accession number NM_007375 ) . Sequencing analyses of the 5 coding and 2 non-coding exons of TARDBP in our initial cohort of 176 clinical patients and 120 patients with pathologically confirmed TDP-43 pathology revealed 3 heterozygous missense mutations in 3 of the 116 analyzed ALS patients ( 2 . 6% ) , while no mutations were detected in 180 patients affected with FTLD-U , FTLD-MND , AD , HpScl and Lewy-body disease ( Table 1 , Figure 1 ) . Since all mutation carriers were index patients of autosomal dominant ALS families , the frequency of TARDBP mutations increased to 3 . 3% in the subpopulation of familial ALS patients ( 3/92 patients ) . One silent mutation ( p . Ala66 ) and 18 additional sequence variants in intronic and non-coding regions were further identified , none of which was predicted to affect the TDP-43 protein ( Table S1 ) . Genomic TARDBP copy-number analyses in 208 patients including all 116 ALS patients did not reveal deletions or multiplications . All TARDBP mutations identified in this study are located in exon 6 ( Figure 2 ) . In the index patient of family A ( ND10588 ) , we identified the known c . 1009 A>G mutation , predicted to substitute valine for methionine at codon 337 ( p . M337V ) , and previously reported to segregate with disease in a large British autosomal dominant ALS kindred . In the index patient of family B ( ND08308 ) , a novel mutation c . 1035 C>A was identified , predicted to change asparagine to a lysine at codon 345 ( p . N345K ) . Finally , in the index patient of family C ( ND08470 ) , a novel mutation c . 1147 A>G which predicts an isoleucine for a valine substitution at codon 383 ( p . I383V ) was identified . Sequence analysis of TARDBP exon 6 in 185 healthy control individuals did not identify these or other sequence variants . Using custom made TaqMan genotyping assays , the presence of p . M337V , p . N345K and p . I383V was further excluded in 640 US control individuals . Genotyping 652 sporadic ALS patients for these mutations did not identify additional mutation carriers . Since all 3 mutation carriers were obtained from the National Institute of Neurological Disorders and Stroke ( NINDS ) Human Genetics Resource Center DNA and Cell Line Repository ( Coriell ) , DNA samples of relatives were unavailable for genetic studies and segregation of the mutations with disease could therefore not be determined . All 3 TARDBP mutation carriers were identified in the clinical patient series and were diagnosed by El Escorial criteria with probable or probable-lab supported ALS . Electromyography ( EMG ) examination was performed in 2 patients ( ND10588 and ND08470 ) and was supportive of the diagnosis of ALS . A detailed overview of the distribution of upper and lower motor neuron signs in the TARDBP mutation carriers is included in Table S2 . Patients ND10588 and ND08308 showed early onset ages of 38 and 39 years , respectively , while patient ND08470 showed symptom onset at 59 years ( Figure 1 ) . The initial presenting symptom in patients ND10588 and ND08470 was upper-limb ALS , while ND08308 suffered from lower-limb onset ALS . No signs of dementia or other atypical features of ALS were reported in any of the mutation carriers or their affected relatives . No autopsy of TARDBP mutation carriers was available . To investigate whether our US p . M337V mutation carrier and the previously reported p . M337V family from the UK are descendants of a common founder , we did an allele sharing study with 12 short tandem repeat ( STR ) markers spanning a region of 6 . 7 Mb flanking TARDBP , including 5 markers within 1 . 0 Mb of TARDBP ( Table 2 ) . We determined the disease haplotype in the UK family and compared this to the genotypes observed in ND10588 to detect allele sharing . Shared alleles were observed for 6 out of 12 STR markers in the region , however , only one marker ( Chr1 ( AC ) _11 . 06 ) directly flanking TARDBP was shared and the 264 bp allele identified at this marker was common in the population ( 62 . 4% ) . In addition , potentially shared alleles at all other markers in the region were also common ( >28% ) . These results make it unlikely that p . M337V originated from a common founder . Kabashi and colleagues previously reported a substantial increase in a ∼28 kDa fragment in lymphoblastoid cells with TARDBP mutations in the presence of the proteasomal inhibitor , MG-132 , but not in lymphoblastoid cells derived from control individuals or ALS patients suggesting an increase aggregation property of these TDP-43 mutants [32] . Based on this result , we performed a similar study and analyzed the 3 patients with TARDBP mutations identified in our study , 2 sporadic ALS cases and 5 control individuals in the presence or absence of MG-132 . Consistent with the previous report , a marked increase in the accumulation of detergent insoluble TDP-43 protein fragments were observed in the lymphoblastoid cell lines treated with MG-132 derived from patients with TARDBP mutations but not those derived from control individuals . In our study , we sized the higher and lower TDP-43 C-terminal fragments at approximately 35 and 25 kDa respectively ( Figure 3 ) . A similar increase was also found in individuals with sporadic ALS ( Figure 3 ) . We previously demonstrated that the proteolytic cleavage of TDP-43 by caspases can generate insoluble C-terminal fragments ( 35 and 25 kDa ) similar to those found in diseased brains . Therefore , we investigated whether proteasome-induced toxicity was associated with proteolytic processing of endogenous TDP-43 in cell culture models . H4 neuroglioma cells were treated with either vehicle ( DMSO ) or proteasome inhibitor I ( PSI ) ( 10 µM ) for 24 hours . In the presence of PSI , TDP-43 was cleaved into ∼35 and ∼25 kDa fragments ( Figure 4 ) , similar to the 35 and 25 kDa fragments found in the lymphoblastoid cell lines derived from the TARDBP mutation carriers ( Figure 3 ) . Similar results were obtained using MG-132 ( data not shown ) . The inhibitory activity and toxicity of PSI also led to a marked increase in cleaved ( active ) capase-3 levels , which promotes apoptotic cell death and accumulates upon such inhibition . Furthermore , when we co-treated the cells with PSI and the caspase inhibitor , Z-VAD ( OMe ) -FMK , the generation of proteolytic TDP-43 fragments was inhibited ( Figure 4 ) . HSP70 immunoblot analysis was used to verify the inhibition of the proteasomal machinery . As expected , HSP70 levels were increased after PSI treatment and the levels persisted in the presence of caspase inhibitor Z-VAD ( OMe ) -FMK ( Figure 4 ) . Taken together , these data strongly suggest that proteasome inhibition is sufficient to promote proteolytic cleavage and accumulation of TDP-43 through a mechanism that implicates programmed cell death .
The identification of rare mutations in genes encoding the major protein component of the pathologic brain depositions observed in familial neurodegenerative diseases has played a critical role in our current understanding of the molecular pathways underlying AD ( APP ) , FTLD ( MAPT ) and Parkinson's disease ( SNCA ) [33] , [34] . In this study , we performed mutation analyses of TARDBP , encoding TDP-43 , in a large cohort of patients with neurodegenerative diseases characterized by TDP-43 pathology to determine if rare mutations or multiplications in TARDBP are involved in the genetic etiology of TDP-43 proteinopathies . Patients with a clinical diagnosis of ALS , FTLD or FTLD-ALS , and patients with pathologically confirmed TDP-43-proteinopathy were included in the analyses . In support of our hypothesis , 14 different pathogenic TARDBP missense mutations were reported by other researchers during the course of this study in familial and sporadic ALS patients [30]–[32] , [35] . We identified 2 novel TARDBP missense mutations ( p . N345K and p . I383V ) and the known p . M337V mutation in 3 out of 92 familial ALS patients ( 3 . 3% ) , while no mutations were identified in 24 sporadic ALS patients or 180 patients with other neurodegenerative diseases . p . M337V , p . N345K and p . I383V were excluded in 825 US control individuals and in 652 additional sporadic ALS patients . The TARDBP mutation frequency in our familial ALS cohort is comparable to the frequency reported by Kabashi and colleagues [32] ( 3/80 patients = 3 . 8% ) but considerably higher than the frequency reported by Sreedharan and colleagues ( 1/154 patients = 0 . 6% ) [31] . This may reflect the difference in study design , as a significant number of our patients were index patients of autosomal dominant ALS families , including all 3 patients carrying TARDBP mutations . Unfortunately , since all mutation carriers were index patients obtained from the NINDS Human Genetics Resource Center DNA and Cell Line Repository , DNA of affected relatives was not available to determine segregation of the mutations with disease . The absence of TARDBP mutations in patients with neurodegenerative diseases other than ALS in our study , confirms the lack of mutations and genetic association of TARDBP in FTLD populations [30] , [36]–[38] . However , without extensive TARDBP sequence analyses in additional cohorts of FTLD and AD patients , TARDBP mutations cannot be excluded as a rare cause of these disorders . All TARDBP mutation carriers identified in this study presented with probable ALS according to El Escorial criteria in the absence of atypical clinical signs , in agreement with the previous reports on TARDBP mutation carriers . The p . M337V mutation has previously been reported to segregate with disease in a British autosomal dominant ALS family [31] . We identified p . M337V in an index patient from a US family with a strong family history of ALS . Our mutation carrier showed upper limb-onset ALS at 38 years of age , 6 years younger than the earliest onset age reported in the British p . M337V family . Signs of dementia were not reported in any of the family members , consistent with the previous report . An allele sharing study using 12 STR markers flanking TARDBP did not support a common ancestor for the UK family and our US patient , although our set of analyzed markers would not have detected a very distant common ancestor [39] , [40] . In addition , we cannot exclude the rare possibility that marker Chr1_11 . 28 mutated in patient ND10588 or that the genomic position of this marker is incorrect , which would leave open the possibility of a shared region of maximum 1 . 3 Mb ( D1S1635-D1S434 ) . In anyway , the identification of p . M337V in two genealogically unrelated ALS families adds further strength to the pathogenicity of TARDBP mutations and justifies mutation screening for TARDBP in patients with familial ALS . Similar to 13 of the 14 previously reported TARDBP mutations , both novel missense mutations identified in this study were located in exon 6 encoding the highly conserved C-terminus of TDP-43 , known to be involved in protein-protein interactions ( Figure 2 ) . p . N345K was identified in a 43 year old male with a 4 year history of ALS and an autosomal dominant family history . The p . I383V mutation was also identified in a familial ALS patient; however the onset age was 59 years , 2 decades later than the other 2 mutations identified in this study . This may reflect the more conservative amino acid substitution ( Iso→Val ) or its more C-terminal location in the TDP-43 protein compared to the other mutations , which may induce a different disease mechanism . Alternatively , additional genetic and/or environmental factors may determine the disease expression of TARDBP mutations , as suggested by the wide onset age range ( 48–83 years ) observed in the recently published family with the p . A315T mutation in TARDBP [30] . Finally , although there is strong evidence supporting that p . N345K and p . I383V are pathogenic , there remains the possibility that these mutations in fact represent rare benign polymorphisms . Definitive confirmation of their pathogenic nature will depend on finding additional ALS patients carrying these mutations . To determine the pathological significance of TARDBP missense mutations on the post-translational processing of TDP-43 , we examined human lymphoblastoid cell lines derived from all 3 familial TARDBP mutation carriers identified in this study , 2 ALS patients without TARDBP mutations and 5 control individuals ( Figure 3 ) . Patient cell lines revealed a substantial increase in a proteolytic cleaved fragment with a molecular weight of approximately 35 and 25 kDa consistent with caspase cleavage [7] . These data suggest that TARDBP mutations may cause a toxic gain of function through novel protein interactions or intracellular accumulation , particularly of caspase fragments . Kabashi and colleagues previously reported a similar substantial increase in a fragment of approximately 28 kDa in lymphoblastoid cell lines of TARDBP mutation carriers . This fragment accumulated in the presence of a proteasome inhibitor ( MG-132 ) , which led the authors to speculate that this TDP-43 product is likely degraded by the ubiquitin-proteasome system ( UPS ) [32] . While we can't exclude the enhanced aggregation of their mutants in the presence of the inhibitor , our data suggests that proteasome-induced toxicity enhances proteolytic cleavage of TDP-43 into 35 and 25 kDa fragments , resulting in cleavage fragments similar to those observed in ALS patients ( Figure 4 ) . Although we can't exclude the possibility that these fragments may be degraded by the UPS , it is likely that the accumulation of these fragments is primarily mediated by caspase cleavage . In conclusion , our findings support that TARDBP mutations are a rare cause of ALS , but so far are not found in other neurodegenerative diseases . Since all reported TARDBP mutations cluster in exon 6 encoding a highly conserved region of the TDP-43 protein , selective mutation analyses of TARDBP exon 6 in familial and sporadic ALS may be warranted .
Our initial study population comprised a total of 296 patients with TDP-43 related neurodegenerative diseases , including 176 clinically diagnosed patients with ALS , FTLD and FTLD-ALS and 120 patients with pathologically confirmed TDP-43 proteinopathy . The average age at onset in the clinical cohort was 57 . 8±10 . 7 ( range 31–81 years ) and the average age at death in the pathological cohort was 74 . 8±13 . 8 ( range 38–100 years ) . Among patients with known ethnicities ( N = 214 ) , 95% were Caucasian ( N = 203 ) , 3% were Hispanic ( N = 7 ) and 2% were others ( African/American ( N = 2 ) , East-Indian ( N = 1 ) and Caribbean ( N = 1 ) ) . A summary of the primary diagnoses and family history of the patients is provided in Table 1 . The majority of the pathological confirmed patients ( N = 87 ) were derived from the Mayo Clinic Jacksonville Brain Bank and primarily ascertained through The State of Florida Alzheimer's Disease Initiative funded through the Department of Elder Affairs , The Einstein Aging Study , The Udall Center for Excellence in Parkinson's Disease Research , CurePSP/The Society for Progressive Supranuclear Palsy , the Mayo Alzheimer's Disease Patient Registry ( ADPR ) and the Florida Alzheimer's Disease Research Center ( ADRC ) . Additional clinical and pathological confirmed patients were ascertained through the Mayo Clinic Jacksonville and Rochester ADRC ( N = 60 ) , Mayo Clinic Scottsdale Alzheimer's Disease Center ( ADC ) ( N = 4 ) , the Neurological Institute of New York , Columbia University ( N = 2 ) , the University of California , Los Angeles ( UCLA ) ADC ( N = 23 ) , the University of British Columbia ( N = 58 ) , the Harvard Brain Bank ( N = 5 ) , the Sun Health Research Institute ( N = 4 ) , the Drexel University College of Medicine ( N = 1 ) , the Northwestern Feinberg School of Medicine ( N = 13 ) and the Coriell Institute for Medical Research ( N = 39 ) . A list of the specific samples from the Coriell Institute included in the TARDBP mutation screening is provided as Table S3 . To determine the frequency of the TARDBP mutations identified in our initial cohort , an additional cohort of 652 sporadic ALS patients was obtained from the University of British Columbia ( N = 140 ) , the Neurological Institute of New York , Columbia University ( N = 48 ) and the Coriell Institute for Medical Research ( N = 464 ) . All control individuals ( N = 825 ) included in the study were Caucasian and ascertained through the Mayo Clinics in Jacksonville , Florida and Scottsdale , Arizona . The 5 coding and 2 non-coding exons of TARDBP were amplified by polymerase chain reaction ( PCR ) in standard 25 µl reactions using Qiagen PCR products ( Table S4 ) . PCR products were purified using the Agencourt Ampure method and sequenced using Big dye terminator V . 3 . 1 products . Sequencing products were purified using the Agencourt CleanSEQ method and analyzed on an ABI 3730 DNA analyzer ( Applied Biosystems , Foster City , CA , USA ) . The presence of TARDBP mutations c . 1009A>G ( p . M337V ) , c . 1035 C>A ( p . N345K ) and c . 1147 A>G ( p . I383V ) in sporadic ALS patients and control individuals was determined with custom-designed TaqMan SNP genotyping assays ( Applied Biosystems ) ( Table S5 ) and analyzed on an ABI7900 genetic analyzer using SDS2 . 2 . 2 software . TaqMan gene expression assays to exons 2 , 4 and 6 of TARDBP and to exon 5 of PSEN2 ( for use as endogenous control ) were designed using File Builder 3 . 1 software ( Applied Biosystems ) ( Table S6 ) to test for the presence of genomic TARDBP copy-number mutations in 208 patients selected from our population . This approach was used to detect copy-number mutations affecting exons 2 , 4 or 6 , as well as complete TARDBP and large N- and C-terminal TARDBP deletions and multiplications . Real-time PCR with 25 ng genomic DNA as template was performed on an ABI7900 using the TaqMan method according to standard procedures . All samples were run in triplicate . The FAM-fluorescent signal was analyzed using SDS2 . 2 . 2 software , and genomic copy number determined by relative quantification ( ΔΔct method ) . To examine whether the US and UK families carrying the p . M337V mutation shared a common founder , we typed 12 STR markers spanning a region of 6 . 7 Mb flanking TARDBP in 3 patients and 8 unaffected relatives of the previously published UK family , in the US patient ND10588 and in 2 CEPH samples . STR markers were amplified with one fluorescently labeled primer and PCR fragments were analyzed on an automated ABI3100 DNA analyzer . Alleles were scored using the Genemapper software ( Applied Biosystems ) . CEPH allele frequencies were used to estimate the allele frequency of the shared alleles in control individuals ( CEPH genotype database; http://www . cephb . fr/cephdb/ ) . The 2 novel markers were PCR amplified using Chr1_11 . 06-F: FAM-CAGCATCATGTGGTTTGGCAGT , Chr1_11 . 06-R: CAGCTCGCAGGGAAGATGAAA , Chr1_11 . 28-F: FAM-TGGCCATCTTAACAGGAACAGC and Chr1_11 . 28-R:TTCAAGGGCTTTCGAGGTGAA and allele frequencies were estimated in a population of 93 unrelated US control individuals . H4 neuroglioma cells were grown in Opti-Mem plus 10% FBS and 1% pen-strep . Cells were plated in 6-well plates and at 90% confluency treated with 10 µM proteasome inhibitor I ( PSI ) ( EMD Chemicals , Inc . San Diego , CA ) or 100 µM pan-caspase inhibitor ( Z-VAD-FMK ) ( EMD Chemicals , Inc . San Diego , CA ) separately or in combination . Twenty-four hours after treatment , cells were harvested for subsequent Western blot analysis in the Co-IP buffer ( 50 mM Tris-HCl , pH 7 . 4 , 1 M NaCl , 1% Triton-X-100 , 5 mM EDTA ) plus 1% SDS , PMSF , protease and phosphatase inhibitors . A similar experiment was performed using 10 µM MG-132 ( Calbiochem , San Diego , CA ) instead of PSI . Lymphoblastoid cells from 5 healthy control individuals , 3 familial ALS patients with TARDBP mutations and 2 ALS patients without TARDBP mutations were grown in RPMI1640 plus 10% FBS and 1% pen-strep . Cells were plated in T25 flasks and treated the following day with MG-132 ( 20 µM , 6 hours ) . Cell pellets from each cell line were lysed with the 0 . 2% Triton X-100-PBS with PMSF , protease and phosphatase inhibitors on ice for 10 minutes . After sonication , samples were centrifuged at 10 , 000 g for 15 minutes at 4°C . The supernatant was saved as the soluble fraction and the pellet was resuspended , sonicated in 2% SDS-PBS-Urea and saved as the insoluble fraction . The soluble and insoluble fractions were subjected to Western blot analysis . Protein concentrations of cells lysates were measured by a standard BCA assay ( Pierce , Rockford , IL ) . Next , samples were heated in Laemmli's buffer and equal amounts of protein were loaded into 10-well 10% or 4–20% Tris-glycine gels ( Novex , San Diego , CA ) . After transfer , blots were blocked with 5% nonfat dry milk in TBST ( TPS plus 0 . 1% Triton X-100 ) for 1 hour , and then incubated with rabbit polyclonal TDP-43 antibody ( 1∶500; ProteinTech Group , Inc , Chicago , IL ) , rabbit polyclonal caspase-3 antibody ( 1∶1000; Cell Signaling , Beverly , MA ) , HSP70 ( 1∶2000; Stressgen , Ann Arbor , MI ) or mouse monoclonal β-actin antibody ( 1∶5000 , Sigma , Saint Louis , MS ) overnight at 4°C . Membranes were washed three times each for 10 minutes with TBST and then incubated with anti-mouse or anti-rabbit IgG conjugated to horseradish peroxidase ( 1∶2000; Jackson ImmunoResearch , West Grove , PA ) for 1 hour . Membranes were then washed three times each for 10 minutes , and protein expression was visualized by ECL treatment and exposure to film . | The abnormal accumulation of disease proteins in neuronal cells of the brain is a characteristic feature of many neurodegenerative diseases . Rare mutations in the genes that encode the accumulating proteins have been identified in these disorders and are crucial for the development of cell and animal models used to study neurodegeneration . Recently , the TAR DNA-binding protein 43 ( TDP-43 ) was identified as the disease accumulating protein in patients with frontotemporal lobar degeneration with ubiquitin inclusions ( FTLD-U ) and in amyotrophic lateral sclerosis ( ALS ) . TDP-43 was also found in the brains of 20–30% of patients with Alzheimer's disease ( AD ) . Here , we evaluated whether mutations in TDP-43 cause disease in a cohort of 296 patients presenting with FTLD , ALS or AD . We identified three missense mutations in three out of 92 familial ALS patients ( 3 . 3% ) , and no mutations in AD or FTLD patients . All the identified mutations clustered in exon 6 , which codes for a highly conserved region in the C-terminal part of the TDP-43 protein , which is known to be involved in the interaction of TDP-43 with other proteins . We conclude that mutations in TDP-43 are a rare cause of familial ALS , but so far are not found in other neurodegenerative diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neurological",
"disorders/movement",
"disorders",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"neurological",
"disorders/cognitive",
"neurology",
"and",
"dementia",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses"
] | 2008 | Novel Mutations in TARDBP (TDP-43) in Patients with Familial Amyotrophic Lateral Sclerosis |
In the current era of malaria eradication , reducing transmission is critical . Assessment of transmissibility requires tools that can accurately identify the various developmental stages of the malaria parasite , particularly those required for transmission ( sexual stages ) . Here , we present a method for estimating relative amounts of Plasmodium falciparum asexual and sexual stages from gene expression measurements . These are modeled using constrained linear regression to characterize stage-specific expression profiles within mixed-stage populations . The resulting profiles were analyzed functionally by gene set enrichment analysis ( GSEA ) , confirming differentially active pathways such as increased mitochondrial activity and lipid metabolism during sexual development . We validated model predictions both from microarrays and from quantitative RT-PCR ( qRT-PCR ) measurements , based on the expression of a small set of key transcriptional markers . This sufficient marker set was identified by backward selection from the whole genome as available from expression arrays , targeting one sentinel marker per stage . The model as learned can be applied to any new microarray or qRT-PCR transcriptional measurement . We illustrate its use in vitro in inferring changes in stage distribution following stress and drug treatment and in vivo in identifying immature and mature sexual stage carriers within patient cohorts . We believe this approach will be a valuable resource for staging lab and field samples alike and will have wide applicability in epidemiological studies of malaria transmission .
One of the tenets of the recently released Malaria Eradication Research Agenda ( malERA ) is the development of new diagnostics specifically addressing transmission reduction [1] . Individuals harboring the Plasmodium falciparum transmissible parasite stage , or gametocyte , are the primary reservoir for malaria transmission , and thus proper surveillance of gametocyte carriers is critical to transmission reduction . Surveillance is difficult , however , because gametocytes comprise only a small fraction of the total body parasite load during active infection and are only observed in the bloodstream in their mature form , while developing stages are sequestered in tissues [2] . For these reasons , quantifying gametocytes in mixed parasite populations has been an ongoing challenge ever since they were first identified more than a century ago . Gametocytes do execute substantially different transcriptional programs from asexual parasite stages , however , as has been well-studied in vitro [3] . Like the sequential dynamics of the asexual Plasmodium life cycle [4] , [5] , gametocytes develop in a staged progression from immature ( young and intermediate stages ) to mature transmission-competent cells in preparation for meiosis and further development in the mosquito vector . The switch between asexual replication and sexual development does not occur ubiquitously in vivo or in vitro , as even the most synchronized gametocyte induction protocols result in partially asynchronous and mixed gametocyte stages [3] , [6] . This problem is compounded in vivo , as blood sampled during infection is likely to contain both gametocyte and asexual parasite populations , leading to a highly convolved transcriptional mixture . In addition to the need to dissect these signatures for analysis of microarray data , it is also of interest to develop a field-friendly approach for detecting and quantifying both immature ( indication of conversion to sexual development ) and mature ( indication of infectiousness to mosquito vector ) gametocyte stages . Transcriptional approaches such as RT-PCR , QT-NASBA and RT-LAMP have been developed [7] , [8] , [9] using the established mature gametocyte marker Pfs25 and the putative immature gametocyte marker Pfs16 . While these approaches enable sensitive detection of these transcripts , it is unclear how the detection of these transcripts - particularly Pfs16 - relates to actual gametocyte carriage [8] . The development of a qRT-PCR-based assay has thus far been impeded primarily because this approach cannot distinguish transcript from genomic DNA when sequences are identical; the majority of P . falciparum genes lack introns and thus have identical sequences for both RNA and DNA . It is therefore worth identifying novel intron-containing markers for which exon-exon junction-spanning primers can be designed so that this approach can be used for in vivo gametocyte quantification . Our goal was thus to develop a new transcript-based gametocyte model that addressed these challenges . Using a deconvolution approach , we quantified the stage-specificity of Plasmodium transcripts genome-wide and subsequently identified intron-containing markers from across the full range of asexual and sexual development . In order to identify expression patterns specific to different gametocyte stages , particularly the immature stages , existing in vitro asexual and sexual developmental time course samples were re-analyzed to account for their mixed stage composition . We further developed a qRT-PCR assay based on these results and , applying the model in reverse , established an algorithm to estimate the amounts of immature and mature sexual and asexual stages in a patient sample based on the expression of a small set of stage-specific markers . This was inspired by related approaches that have been used successfully in dealing with mixed cancerous/non-cancerous tissue samples [10] , [11] and with mixed stages of budding yeast [12] . This framework is implemented for public use at http://huttenhower . sph . harvard . edu/malaria2013; as a transmission-focused tool , this system can be applied in epidemiological settings , and as such will ideally support efforts directed toward reducing malaria prevalence worldwide .
To gauge how this modeling process performed on patient microarray samples , we applied it to microarray data from two patient cohorts , i ) a previously published cohort of severe malaria patients from Blantyre , Malawi collected in 2009 [23] and ii ) a cohort of uncomplicated malaria patients from Thies , Senegal collected in 2008 . While no staging information was available for the Senegal patients , a subset of Malawi patients were previously identified as gametocyte-positive by thick smears . The model inferred that the majority of patients from both cohorts have a strong ring-dominated profile , with the next largest subset being late asexual stages ( trophozoites and schizonts ) ( Figure 4A/B ) . For the 10 Malawi samples in which gametocytes were observed by thick smear , our model correctly identifies 4 ( 40% ) as such , with 0 false positive developing or mature gametocytes predicted among the 48 thick smear-negative patients ( Figure 4A ) . Interestingly , two thick smear-negative patients are predicted to have young gametocytes , which are difficult to identify by thick smear microscopy due to their morphological similarities with asexual stages . A subset of the uncomplicated malaria patients from Senegal were also predicted to be gametocyte carriers ( 6 of mature , 1 of developing and 1 of young gametocytes ) ( Figure 4B ) . As microscopy-based information was unavailable for the Senegalese cohort , we assessed how our gametocyte inferences correlated with patient parameters . Of the 6 parameters we measured for this cohort , illness duration and hematocrit differed significantly between the group of patients inferred to be gametocyte carriers and those inferred to be gametocyte-negative . The former had a longer duration of illness ( 6 . 33 days±1 . 02 SEM ) than the latter ( 3 . 84 days±0 . 25 SEM , t-test p = 0 . 0014 ) as well as a lower hematocrit measured in percent cell volume ( 34 . 86%±2 . 17 SEM ) than the latter ( 40 . 41%±1 . 06 SEM , t-test p = 0 . 031 ) ( Supplementary Table S4 ) . This finding agrees with published data on clinical correlates of gametocyte carriage: long illness duration ( greater than 2 days ) and anemia ( hematocrit less than 30% ) were both independently found to be risk factors of gametocytemia in uncomplicated malaria [24] . We next sought to test a variation of the microarray-based model for application to transcriptional measurements obtained by PCR , which might eventually be more appropriate for a field assay . As no MG marker that achieved our filtering criteria ( see Figure 1C ) for qRT-PCR also matched both the high expression levels and stage-specificity of the existing Pfs25 marker for gametocyte detection , we assessed the utility of the YG and DG markers in combination for the prediction of immature and mature gametocyte quantities when applying the model to qRT-PCR data . Specifically , PF14_0748 and PF14_0367 were likely to represent immature ( IG ) and mature ( MG ) gametocytes in combination , as PF14_0367 had a βg , s parameter similar to that of Pfs25 in mature gametocyte stages . We therefore cross-validated this 5-marker PCR set ( Table 1 ) comparably to the 6-marker microarray set , using the in vitro microarray time courses as described above . The simplified model remained able to predict stage distribution accurately , with a root mean squared error comparable to that of the 6-marker model ( Supplementary Figure S1 ) . In order to create a qRT-PCR assay for our sentinel transcripts , we designed exon-exon junction spanning primers ( distinguishing transcripts from genomic DNA ) and sequence-specific probes ( distinguishing transcripts from non-specific background amplification ) . Following confirmation that our primer/probe sets selectively amplified cDNA and not genomic DNA or non-specific products , we validated the stage-specific expression using in vitro-derived asexual and sexual stage RNA ( Table 1 , and supplementary Figure S2A , and Table S5 for optimization and validation of qRT-PCR parameters ) . For these experiments , we used the gametocyte-producing reference line 3D7 and a gametocyte-deficient clone thereof ( termed F12 [29] ) to confirm the stage-specificity of each of our sentinel markers . Normalized expression data from time courses of 3D7 and F12 confirmed stage-specificity of our sentinel marker set ( Supplementary Figure S2B ) . The asexual markers alternate with respect to time points in which there were predominately rings or trophozoites and schizonts in the culture , with similar results for both the F12 and 3D7 lines . The sexual markers demonstrate stage-specificity within the 3D7 time course and no appreciable expression in the F12 line once normalized . Specifically , PF14_0748 expression is detected in the early and mid gametocyte time points , while PF14_0367 expression is detected in both mid and late time points .
Several highly sensitive single-marker molecular assays are currently used to detect Plasmodium gametocytes . None of these existing tools have been appropriate for detection and quantification of the relevant range of parasite stages present during infection , however , due primarily to the lack of a sufficiently broad panel of stage-specific markers . Further , since malaria parasite populations exist as mixtures of the different phases of the life cycle , assays combining multiple markers require customized computational analysis methods for dealing with this complexity . We combined the development of such a bioinformatic deconvolution approach with panels of stage-specific , intron-containing markers appropriate both for microarray analysis and a newly developed qRT-PCR assay . This multi-marker platform enabled us not only to detect gametocyte carriers but primarily to infer the relative amounts of sexual and asexual stages within a sample . We provide an implementation of this platform for further development and application , particularly for refinement in field settings . This process can also be adapted bioinformatically by the exclusion or inclusion of markers to answer specific questions , such as determination of parasite sex ratios that are known to influence mosquito infectiousness [31] . Our deconvolution model provided the opportunity to define stage-specific gene sets and to characterize the biology of these stages' expression programs using tools such as GSEA , even in the absence of transcriptional data from pure stage populations . For example , our GSEA analysis confirms earlier studies that suggested increased mitochondrial and lipid metabolism during gametocyte development [3] , [32] . Interestingly , the analysis also suggests significant enrichment of several markers related to endocytic trafficking in late gametocyte development but not in any other parasite stage . The biological significance of this observation remains to be determined . To put such findings into context and ultimately describe the gametocyte transcriptome at high resolution , a systematic transcriptional re-analysis of the entire P . falciparum gametocyte cycle using isolated and synchronous gametocyte stages will be required . Transcriptional approaches have significantly increased the sensitivity of gametocyte detection in field-compatible assays [8] , [9] , [33] , [34] , [35] . However , these have been limited to either ( i ) qualitative assessments of multiple gametocyte markers , i . e . RT-PCR of immature and mature gametocyte markers [35] , or ( ii ) quantitative assessments of mature gametocytes only , i . e . QT-NASBA of the gamete surface antigen Pfs25 [8] . In order to properly define the reservoir of parasite and gametocyte carriers in the field , it is imperative to determine both the absolute parasite burden and the stage composition of parasites in the blood circulation . Challenges have prevented the development of a diagnostic that can measure the latter , such as ( i ) the lack of transcriptional analysis methods to identify gametocytes with high specificity in a sample containing a mixture of stages , ( ii ) the lack of validated immature gametocyte markers , and ( iii ) the lack of known intron-containing qRT-PCR compatible markers for all stages . We tackled these challenges by developing a model specific to the quantification process and ensuring that it was compatible with both microarray and qRT-PCR measurements . This is distinct , of course , from models that would focus only on sensitivity and specificity of gametocyte detection from such data , which represent a potentially fruitful course of future computational investigation . Instead , by incorporating relative expression values of the markers , the model allowed us both to identify a subset of patients as gametocyte carriers and to additionally quantify sub-categories of immature and mature gametocyte fractions within the mixture of stages in the bloodstream . Following validation of our model on samples for which stage composition was known , we applied our model to two microarray data sets in which stage composition was unknown: ( i ) a cohort of uncomplicated malaria patients , and ( ii ) two in vitro growth experiments in the presence of drug . In the former , we found that both mean illness duration and hematocrit differed between inferred gametocyte carriers and non-carriers , in agreement with published data demonstrating that long illness duration and low hematocrit is linked to gametocyte carriage [24] . In the latter , we observed an increase in the fraction of mature gametocytes as well as unexplained transcriptional signature upon the addition of drug treatment to parasites . The enrichment of mature rather than young gametocytes in response to drug treatment suggests that the drug selectively kills asexual stages , leaving gametocytes unaffected rather than inducing the development of new young gametocytes . The increase in unexplained signatures likely indicates the transition to unhealthy , dying parasite fractions . These applications demonstrate the range of potential uses for this inference tool . As the exon-exon junction spanning primer/probe sets for 5 markers designed here represent the first attempt at a multi-marker gametocyte-staged qRT-PCR assay , further modeling of PCR-specific measurement error and careful standardization of experimental protocol for this difficult task will both improve field inferences . Like the microarray expression model , however , this model successfully recapitulated the transition from asexual to sexual development across multiple in vitro experiments even on first application . When used initially in vivo for blood samples from a cohort of children with severe malaria in Malawi , the system successfully identified a subset of patients as immature and/or mature gametocyte carriers . Because immature gametocytes in particular are present in the body several days before the more mature forms emerge , our approach for detecting them could be used in further investigations into factors that influence gametocyte conversion in vivo . The assay and algorithm framework presented here has potential for use in epidemiological studies such as those of asymptomatic carriers , who likely represent a major reservoir for malaria transmission . Multiple such studies are already ongoing and will yield additional samples to further optimize computational models of gametocyte differentiation . This is also true of data generated from other sensitive expression platforms such as glass-slide arrays or Nanostring . The inference process may thus have applications in better understanding the natural progression of malaria in the human host , by identifying gametocytes earlier in the course of infection and determining the impact of specific drug treatments on gametocyte development . By scaling to future population-level screens , the resulting information will help better characterize the epidemiology of gametocytemia based on malaria transmission intensity , geography , climate and season .
The institutional review boards of the Harvard School of Public Health , Brigham and Women's Hospital , the University of Malawi College of Medicine , and the Ministry of Health in Senegal approved all or parts of this study . Consent was obtained from the patient or a child's guardian . A transgenic line , 164/GFP , of a gametocyte-producing clone of the 3D7 strain of P . falciparum was used to produce the mixed stage samples for model training and validation . This transgenic line , which aided in the quantification of gametocyte stages , produces stage-specific GFP expression under the PF10_0164 gene promoter , as described previously [22] . A previously characterized non gametocyte-producing clone , F12 , of the 3D7 strain was used to confirm stage-specificity of gametocyte markers [29] . Culture conditions were as described previously [38] , maintaining the parasite line in O+ blood at 4% hematocrit in RPMI-1640 media supplemented with 10% human serum . Cultures were kept at 37°C in a chamber containing mixed gas ( 5% CO2 , 5% O2 , 90% N2 ) . Prior to induction , asexual parasite cultures were synchronized for two cycles with 5% D-sorbitol [39] , and subsequently induction of gametocytogenesis was performed according to the Fivelman protocol [28] . Briefly , asexual parasites were grown to a high parasitemia in the presence of partially spent ( “conditioned” ) medium , and then sub-cultured at the schizont stage into new dishes containing fresh media and erythrocytes . One of two methods was used to reduce the amount of asexual stages in the cultures: Treatment with D-sorbitol was applied on two days later to lyse asexual trophozoite/schizont stages and selectively enrich for unaffected early gametocytes , or N-Acetyl glucosamine was added to the medium one day later and every subsequent day to selectively kill asexual stages . A 3D7 line was used to study the effect of drug perturbations on parasite growth . Culture conditions were performed as described above . Asexual parasite cultures were synchronized for three cycles with 5% D-sorbitol , and expanded to a parasitemia of 5–6% . Hematocrit was increased from 3 to 6% at the late schizont stage using fresh blood . Upon reinvasion drugs were added to the culture at a concentration of 5×IC50 . Drug-treated and control parasites were harvested at 10 , 20 , 30 , and 40 hours post-invasion and RNA was extracted ( Qiagen ) . In order to accurately quantify the stage distribution of parasites in our in vitro samples , we used a combination of standard and fluorescence microscopy . Parasite stage distribution was monitored throughout the parasite synchronization and induction protocol using Wright's Giemsa stain applied to thin blood smears . Quantification of asexual rings and trophozoite stages , as well as developing and mature sexual stages was done directly by light microscopy . In order to quantify early stages of sexual development that are morphologically similar to asexual stages , we used a combination of live imaging and immunofluorescence microscopy . Live imaging was performed using the transgenic 164/GFP line . Parasites were analyzed using the FITC channel on an inverted epifluorescence microscope ( Zeiss ) and quantification was done of the proportion of GFP ( + ) parasites out of the total number of Hoechst ( + ) parasites . Immunofluorescence assays were performed with cell monolayers on glass slides , prepared as described previously [40] . For labeling with the constitutive gametocyte marker Pfs16 , slides were fixed in ice-cold methanol , blocked with 5% nonfat dry milk powder , incubated with polyclonal mouse antibody against Pfs16 ( 1∶2500 ) [41] , washed and incubated with a secondary antibody conjugated to Alexa 488 . Parasite nuclei were labeled with DAPI and quantification was done on the proportion of FITC ( + ) parasites out of the total number of DAPI ( + ) parasites . For time points in which we had data from both live and immunofluorescence experiments [41] , the quantification of early gametocytes from both methods was averaged to give the final amount . RNA ( from peripheral blood of Senegalese patients and cultured in vitro drug perturbations ) was assessed by Bioanalyzer ( Agilent ) , and high quality RNA samples were labeled and hybridized to an oligonucleotide array ( Affymetrix ) custom-designed for the P . falciparum 3D7 genome , PlasmoFB , as published previously [5] . The raw CEL files were condensed into GCT expression files using RMA and the default parameter settings in ExpressionFileCreator in GenePattern [43] . Given the learned model parameters βg , s from stage-labeled data with known xs , the model was inverted to infer the unknown stage distributions xs in new samples . A quadratic programming approach was used to solve the system of linear equations with the constraint that the proportions of all stages must sum to 1 and that each stage contributes a non-negative fraction of expression: We implemented this process in the R function quadprog and solved for the stage distributions using the sets of six ( for microarrays ) or five ( for PCR data ) markers ultimately selected as follows . | The human malaria parasite Plasmodium falciparum is transmitted through a mosquito vector and causes over half a million deaths per year . The microorganism cycles through asexual and sexual life cycle stages , and its successful transmission relies on cells in the sexual stage . These stages are , however , present only at low levels during infection; most infecting cells are asexually reproduced . It can be challenging to assign biomolecular activity to particular parasite life cycle stages from typical gene expression profiles , given the mixed stage composition of most samples . We developed a deconvolution model to identify components of Plasmodium transcriptional activity contributed by sexual and asexual life cycle stages , initially using samples of known composition . From these , we optimized a small set of stage-specific genes with highly informative expression patterns and trained an inference model to predict the stage composition of new samples . The model successfully inferred the parasite's transition from asexual to sexual development over time under laboratory conditions and identified a subset of patient samples harboring transmissible sexual stages . The system presented here can aid in epidemiological or laboratory perturbation in which stage composition is an important step in understanding and preventing malaria transmission . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Inferring Developmental Stage Composition from Gene Expression in Human Malaria |
Chikungunya virus ( CHIKV ) is a re-emerging mosquito-borne Alphavirus that causes a clinical disease involving fever , myalgia , nausea and rash . The distinguishing feature of CHIKV infection is the severe debilitating poly-arthralgia that may persist for several months after viral clearance . Since its re-emergence in 2004 , CHIKV has spread from the Indian Ocean region to new locations including metropolitan Europe , Japan , and even the United States . The risk of importing CHIKV to new areas of the world is increasing due to high levels of viremia in infected individuals as well as the recent adaptation of the virus to the mosquito species Aedes albopictus . CHIKV re-emergence is also associated with new clinical complications including severe morbidity and , for the first time , mortality . In this study , we characterized disease progression and host immune responses in adult and aged Rhesus macaques infected with either the recent CHIKV outbreak strain La Reunion ( LR ) or the West African strain 37997 . Our results indicate that following intravenous infection and regardless of the virus used , Rhesus macaques become viremic between days 1–5 post infection . While adult animals are able to control viral infection , aged animals show persistent virus in the spleen . Virus-specific T cell responses in the aged animals were reduced compared to adult animals and the B cell responses were also delayed and reduced in aged animals . Interestingly , regardless of age , T cell and antibody responses were more robust in animals infected with LR compared to 37997 CHIKV strain . Taken together these data suggest that the reduced immune responses in the aged animals promotes long-term virus persistence in CHIKV-LR infected Rhesus monkeys .
Chikungunya virus ( CHIKV ) is a re-emerging member of the Alphavirus genus within the Togaviridae family . CHIKV was first isolated in the 1950's from the serum of a febrile patient in Tanzania during a dengue fever-like outbreak [1] . The virus and the associated disease were named chikungunya , meaning “that which bends up” in the local language in reference to the debilitating poly-arthralgia that accompanies the infection [2] . Minor outbreaks of CHIKV continued to occur until 2004 when a large epidemic in Kenya marked the beginning of a 4-year period in which CHIKV was imported into several new countries [3] , [4] . Several outbreaks occurred in India , where over one million cases were reported [5] , [6] . In 2006 , over 200 , 000 cases were reported on the island of La Reunion [7] , [8] , the most significant aspect of that outbreak being that a single point mutation in the viral envelope glycoprotein allowed the virus to replicate to very high titers in both Aedes ( Ae . ) albopictus and Ae . aegypti mosquitos , which are widely distributed throughout the world [3] , [9] , [10] . Viremic travelers returning from India initiated a local outbreak in Italy through infection of Ae . albopictus mosquitoes [11] , [12] . CHIKV cases from travelers returning from endemic regions were also reported in France and the United States , demonstrating the potential of CHIKV spread to distant locales [13] , [14] . Clinical symptoms of CHIKV infection in humans include acute fever lasting up to two weeks and severe poly-arthralgia of the peripheral joints that can be very debilitating and last for several months . Additional symptoms include nausea , headache , rash and lymphadenopathy . Although CHIKV infection is usually a self-limiting disease , the outbreaks occurring after 2005 were explosive and exhibited complicated clinical-pathological manifestations including neurological involvement . Aged individuals and adults with underlying immunological conditions experienced severe morbidity and sometimes mortality [15] indicating that a functional immune system maybe important to control infection and promote recovery . Recent epidemics also showed the first CHIKV intra-partum maternal–fetal transmission with devastating outcomes [16] . In addition , age-matched studies during an outbreak in Singapore found that women are more susceptible to CHIKV-induced chronic arthralgia than men [17] , [18] . Treatment of CHIKV infection and disease is currently limited to supportive care . Therefore , the development of effective CHIKV vaccines and therapeutics are currently being intensely explored . The immune response to CHIKV infection has been relatively understudied . A limited number of studies characterized changes in plasma cytokine/chemokine levels following CHIKV infection in order to identify predictors of poor disease prognosis . These analyses revealed that the acute phase of infection ( 4 days post onset of illness ) is associated with the production of robust Type-1 interferon ( IFN ) [17] , [19] . The level of IFN directly correlated with viral load [20] . The chemokines IP-10 ( CXCL10 ) , MCP-1 ( CCL2 ) and RANTES as well as the cytokines IL-6 , IL-1β IL-1Rα and IL-12 are similarly expressed early following infection and their levels correlate with viral loads [21] . Similar findings have been observed in CHIKV-infected non-human primates [22] . High viral loads during the early viral convalescence phase have also been associated with increased IL-12 and IL-6 levels compared to patients with lower viral loads [17] . Interestingly , increased IL-6 and GM-CSF are observed in patients experiencing chronic joint pain ( 2–3 months post onset of disease ) whereas , increases in Eotaxin and hepatocyte growth factors are associated with a full recovery from CHIKV disease [17] . The functional relationship between expression of these two proteins and disease recovery warrants further investigation . The adaptive immune responses to CHIKV infection have not been as widely investigated . In one study , the acute phase of CHIKV infection in a human cohort from a 2007 CHIKV outbreak in Gabon was found to be associated with CD8+ T cell activation . However , the antiviral specificity of the CD8 T cell response and its contribution to disease resolution were not investigated [19] . CHIKV-associated antibody responses have been better characterized . CHIKV infection induces robust neutralizing antibody responses in both humans and animal models . In fact , neutralizing antibodies directed against E2 have been found to be protective and the early presence of IgG3 antibodies have been correlated with protection from persistent arthritis in patients from a 2008 CHIKV outbreak in Singapore [23] , [24] . Unfortunately , we have a limited knowledge as to the viral protein ( s ) are targeted or by what specific immune cell subsets , nor do we know the timing of these events . Here , we describe the infection of adult and aged Rhesus macaques with either the Senegal strain ( CHIKV-37997 ) or the reemergent strain ( CHIKV-LROpy-1 ) . We detail the innate and adaptive immune antiviral responses following infection . Rhesus macaques infected with CHIKV become viremic within 24 hours of inoculation and this was typically resolved by 5 days post-infection ( dpi ) . All animals generated robust T and B cell responses that peaked by 14 dpi . However , our study demonstrates that aged animals persistent viral RNA in the spleen . Interestingly , this inability to clear the virus in the aged animals correlated with reduced innate and adaptive immune responses compared to adult animals . Our findings suggest that defects in the aged immune system play a critical role promoting CHIKV persistence .
All Rhesus macaques were handled in accordance with good animal practice as defined by relevant national and/or local animal welfare bodies . The use of non-human primates was approved by the Oregon National Primate Research Center ( ONPRC ) Institutional Animal Care and Use Committee ( IACUC #0826 ) . The ONPRC is fully accredited by the Assessment and Accrediatation of Laboratory Animal Care-International . For blood collection monkeys were anesthetized with ketamine by intramuscular injection . Monkeys were humanely euthanized by the veterinary staff at ONPRC in accordance with endpoint policies . Euthanasia was conducted under anesthesia with ketamine followed by overdose with sodium pentobarbital . This method is consistent with the recommendation of the American Veterinary Medical Association . The full-length infectious clones for the West African ( Senegal ) strain CHIKV-39779 and the recent outbreak strain La Réunion ( LR2006 OPY1; CHIKV-LR ) were obtained from Dr . Stephen Higgs University of Texas Medical Branch-Galveston . The viruses produced from the infectious clones retain their original viral phenotypes of CHIKV in both cell culture and in mosquitoes . RNA was transcribed using the T7 mMessage mMachine kit ( Ambion ) . The transcribed RNA was transfected into BHK cells and the resultant virus was propagated in C6/36-insect cells to produce low passage , high titer CHIKV stocks by pelleting through a 20% sucrose cushion by ultracentrifugation ( 22 , 000 rpm , 82520×g for 1 . 5 hrs ) . Plaque titration assays on Vero cells using a carboxy-methylcellulose overlay were performed to determine titers ( plaque forming units ) of the stock virus . RNA transcribed from the infectious clones of CHIKV-LR and CHIKV-37997 was transfected into BHK cells and passaged one time in insect cells to generate the virus stock used to infect the animals . Adult male and female rhesus macaques ( ranging from 6–13 years old ) and aged female rhesus macaques ( >17 years old ) were utilized for infection studies as described in Table 1 . Animals were infected intravenously with 1×107 , 1×108 or 1×109 plaque forming units ( pfu ) CHIKV-LR or CHIKV-37997 strain in 1 ml of phosphate buffered saline ( PBS ) , and a sample of all viral inoculums was back-titered to confirm the delivery of correct dosage . Blood samples were obtained following ketamine sedation ( 10 mg/Kg ) at 4 and 2 weeks prior to infection , on the day of infection ( day 0 ) , 1 , 2 , 3 , 5 , 7 , 10 and14 days post infection and on a weekly basis thereafter . Complete blood cell counts were obtained at every time point to monitor changes in immune cell numbers/µl of blood . Peripheral blood mononuclear cells ( PBMC ) were isolated from whole blood by centrifugation over Histopaque gradient ( Histopaque , Sigma-Aldrich , St Louis , MO ) . Viral load in peripheral blood ( plasma and PBMC ) as well as tissues at necropsy of the rhesus macaques was quantified using real-time RT-PCR . Primers and probes include: CHIKV-LR fwd: GGAACGAGCAGCAACCTTTG , rev: ATGGTAAGAGTCTCAGACAGTTGCA , and probe: GGAATAAGGGCTTGT; CHIKV-37997 fwd: CGGAGAGCCGTAAGCTTCTTAA , rev: TTCACACGAAACCACTGTATCACA , and probe: CTTCAGTGTTCCATCTAAA . Total RNA was prepared from 100 µl of animal sera using the ZR Viral RNA kit ( Zymo Research , Irvine , CA ) or from 5×106 total PBMC using Trizol . RNA was prepared from tissue specimens by the Trizol method . The isolated RNA was quantified using a Nanodrop spectrophotometer . RNA was first treated with RNAse-free DNAse and then single stranded cDNA was generated using random hexamers and Superscript III RT ( Invitrogen , Carlsbad , CA ) . Gene amplicons served as quantification standards ( limit of detection is 10–100 copies ) . Quantiative RT-PCR results were analyzed using ABI StepOne Plus Real-Time PCR system ( Applied Biosystems , Foster City , CA ) . CHIKV peptides were designed to be 18 mers that overlapped by 12 amino acids covering the entire CHIKV proteome ( peptides were purchased from Thermo Fischer Scientific ) . Peptides were resuspended in DMSO at 1 mg/ml . Separate peptides pools were made for each of the nine viral protein products . PBMC were stained with surface antibodies against CD4 , CD8β , CD28 , and CD95 to delineate naïve , central memory ( CM ) , and effector memory ( EM ) T cell subsets . All antibodies with the exception of CD8β ( Beckman Coulter ) were purchased from Biolegend ( San Diego , CA ) . Cells were then fixed and the nuclear membrane permeabilized as per manufacturer's recommendation ( BD Pharmingen , San Diego , CA ) before staining with anti-Ki67 ( BD Pharmingen , San Diego , CA ) . Samples were acquired using the LSRII instrument ( BD bioscience , San Jose , CA ) and data analyzed by FlowJo ( TreeStar , Ashland , OR ) . Gamma interferon ( IFN-γ ) enzyme linked immunospot ( ELISPOT ) assays were performed using the ELISpotPlus for Monkey IFN-γ kit ( MabTech ) . Plates were washed one time with 200 µl/well of PBS and then blocked for 30 minutes in 200 µl/well of RPMI containing 10% fetal calf serum ( FCS ) and penicillin , streptomycin and glutamine ( RPMI-10 ) . Overlapping CHIKV peptide pools ( 1 µg/well ) for each of the nine CHIKV proteins were added to wells containing 2×105 rhesus monkey PBMC ( n = 3 ) . Positive controls ( 1 µl of each phorbol 12-myristate 13-acetate ( PMA; 200 mg/ml stock ) and Ionomycin ( 7 mM stock ) ) and negative controls ( DMSO; 1 µl ) were also performed for each monkey PBMC sample . Plates were incubated for 18 hrs at 37°C . After incubation the plates were washed thoroughly with PBS and then incubated with a biotin-conjugated anti-IFN-γ antibody ( 1 µg/ml in PBS plus 0 . 5% FCS ) for 2 hrs at room temperature . Plates were washed five times with PBS and then incubated with Streptavidin-ALP in PBS plus 0 . 5% FCS for 1 hour . Plates were washed five times and then developed with filtered BCIP/NBT substrate solution . Flushing the wells with water stopped color development . The plates were air-dried and the number of spot forming cells was determined by subtracting the mean number of spots counted in negative controls wells ( n = 3 ) from the spots counted in each well . The magnitude of the response was determined as the sum of the spots counted per 6×105 PBMC . PBMC were stained with surface antibodies directed against CD20 ( Beckman Coulter ) , IgD ( Southern Biotech ) and CD27 ( Biolegend , San Diego , CA ) to delineate naive , marginal zone-like and memory B cells . The cells were then fixed and the nuclear membrane permeabilized as per manufacturer's recommendation before addition of Ki67 ( BD Pharmingen , San Diego , CA ) specific antibodies . Samples were acquired using the LSRII instrument ( BD bioscience , San Jose , CA ) and data analyzed by FlowJo ( TreeStar , Ashland , OR ) . Antiviral IgG levels were measured in circulating plasma using a standard ELISA assay using plates coated with CHIKV lysate or CHIKV particle preparations , which reacts with CHIKV Abs . In these experiments , serial three-fold dilutions of plasma were incubated in triplicates CHIKV lysate or virus-coated ELISA plates for 1 hr prior to washing , staining with detection reagents ( HRP-anti-IgG ) and addition of chromogen substrate to allow for detection and quantitation of bound antibody molecules . Log-log transformation of the linear portion of the curve was then performed , 0 . 1 OD units was the cut-off point to calculate end point titers . Each plate contained a positive control sample to normalize ELISA titers between assays , and a negative control sample to ensure assay specificity . PBMC were stained with surface antibodies directed against CD3 , CD11c , CD14 , CD20 , CD123 and DR to delineate plasmacytoid dendritic cells , myeloid dendritic cells , DR+ lineage negative other DCs and monocyte/macrophages . Samples were acquired using the LSRII instrument ( BD Bioscience , San Jose , CA ) and data analyzed by FlowJo ( TreeStar , Ashland , OR ) . Blood plasma ( 100 µl ) obtained from adult and aged CHIKV infected Rhesus macaques ( 3 dpi ) was UV inactivated ( 3× for 30 seconds at 600 µJ ) . Plasma was added to rhesus macaques fibroblasts cultured in 24-well plates for 24 hr . Recombinant universal Type-1 IFN ( RND Systems; 0 . 1 , 1 , 10 units/ml ) treated cells were used as positive controls and untreated cells were used as a negative control . Total RNA was isolated from the cells using the Trizol method ( Life Technologies ) . Quantitiative RT-PCR was used to measure Interferon-stimulated gene ( ISG ) expression using primers and probes specific for Mx-1 and ISG-56 . cDNA was generated using Superscript III ( Life Technologies ) and analyzed on an ABI StepOne Real-Time PCR system and normalized to L32 . Gene amplicons served as quantification standards ( limit of detection ≤100 copies/gene ) . ISG expression levels for each sample were expressed as fold-change by comparison to untreated cells . Data were analyzed by Prism Graph Pad software . Data were analyzed using Prism Graph Pad software using Two-way ANOVA , no post-test corrections were carried out due to the small sample size .
Recent outbreak strains of Chikungunya virus contain an adaptive mutation in the E1-protein that facilitates transmission via Ae . albopictus ( Asian Tiger ) mosquitoes as well as increased virulence . We sought to compare CHIKV-LR virus disease progression and host immune responses infection with that of a West African reference strain ( 37997 ) to determine whether CHIKV-LR was also more pathogenic in non-human primates . We also explored the impact of aging on virus replication and immune response to both of these strains . We first performed a titration experiment to determine the effects of virus strain and inoculation dosage on in vivo CHIKV replication and pathogenesis . Two cohorts of adult rhesus macaques were infected intravenously with either CHIKV-LR or CHIKV-37997 . Each cohort consisted of 3 groups of 2 animals each that received 1×107 , 1×108 or 1×109 plaque forming units of either strain and animals were observed daily for signs of disease . Animals developed lymphadenopathy 7–14 dpi as well as fever between 3–7 dpi . A rash was observed on the chest of some of the animals at 7 dpi ( data not shown ) . We compared viral loads in peripheral blood mononuclear cells ( PBMC ) and plasma at 0 , 1 , 2 , 3 , 5 , and 10 dpi using qRT-PCR ( Figure 1 ) . In each cohort viral loads were detectable by 2 dpi and viremia was resolved by 5 dpi . Interestingly , the inoculation dose had no detectable impact on peak viremia or on the kinetics of viral replication ( Figure 1 ) . We therefore treated all animals infected with a given strain as one cohort for the remainder of our analysis . While we did not detect statistically significant differences in peak plasma virus levels between CHIKV-LR and CHIKV-37997 , average viral loads were nearly 10-fold higher in CHIKV-LR infected animals ( 5 . 7×106 vs . 5 . 9×105 RNA genomic copies per 10 µl of plasma , respectively; p = 0 . 15 ) . CHIKV RNA levels were very low in PBMC but as described for plasma were slightly higher in CHIKV-LR infected animals ( 4 , 650 vs . 878 RNA genomic copies per 0 . 1 µg of total RNA; respectively , p = 0 . 29 ( Figure 1B vs . 1D ) . To investigate whether advanced age is associated with more severe CHIK disease , we infected 4 aged animals ( >17 years of age ) with 1×109 pfu CHIKV-LR ( n = 2 ) or CHIKV-37997 ( n = 2 ) . We compared peak viral loads and the kinetics of viral replication between the aged and adult cohorts ( Figure 1 ) . The kinetics of viremia were comparable in both young and aged animals for both virus strains . However , while no virus was detected after 5 dpi in adult animals , one of the aged CHIKV-LR animals had detectable persistent viral RNA in plasma at 10 dpi ( Figure 1A; Animal #20841 ) . Levels of virus detected in the PBMC of the aged CHIKV-LR infected animals tended to be higher compared to adult animals or aged animals infected with CHIKV-37997 ( Figure 1B &D ) . All animals were euthanized at 35–42 dpi , at which point , blood , spleen , lymph nodes and joint tissues were harvested to determine viral loads . At the time of necropsy , we observed no gross joint inflammation in any of the animals . In addition , we detected no virus in tissues harvested from adult animals at 35 dpi . However , viral RNA was present in spleen of aged CHIKV-LR infected animals but not in aged animals infected with CHIKV-37997 ( Figure 2 ) . Interestingly , aged animal #20994 infected with CHIKV-LR had both the highest levels of PBMC associated virus and highest level of spleen-associated virus compared to all other infected animals . These findings suggest that CHIKV-LR can persist in aged animals and is potentially more virulent than 37997 . The higher PBMC viral loads in CHIKV-LR infected aged animals as well as the persistence of viral RNA in the spleen could be due to increased viral replication and/or an ineffective immune response . Next , we investigated the impact of age on the adaptive immune response to CHIKV . One of the hallmarks of the immune response against a viral infection is rapid and robust proliferation of antigen specific lymphocytes . We therefore compared the kinetics and magnitude of the T cell proliferative burst in adult and aged animals by determining the frequency of T cells expressing the nuclear protein Ki67 , which is induced during cell cycle [25] . Peripheral blood mononuclear cells isolated at different time points after infection were first stained with antibodies directed against CD4 , CD8 , CD28 and CD95 in order to delineate naïve , central ( CM ) and effector memory ( EM ) T cell populations as shown in Figure 3A and previously described [25] . The cells were permeabilized , stained with antibodies directed against Ki67 , and analyzed by flow cytometry ( FCM ) . Proliferation of CD4 CM cells was evident by 7 dpi in both CHIKV-LR and CHIKV-37997-infected adult animals ( Figure 3B & C ) . The frequency of Ki67+ CD4 CM T cells peaked at 10 dpi and remained elevated in CHIKV-LR infected animals until 14 dpi before returning to baseline at 28 dpi ( Figure 3B ) . On the other hand , after peaking at 10 dpi , the frequency of CD4 CM Ki67+ T cells began to decrease in CHIKV-37997 infected animals , returning to baseline at 21 dpi ( Figure 3C ) . However , the CD4 CM T cell proliferative burst was delayed in aged animals . Specifically , CD4 CM proliferation was not initiated until 10 dpi in aged CHIKV-LR infected animals and 14 dpi in aged CHIKV-37997 animals ( Figure 3B & C; p = 0 . 037 and p = 0 . 005 , respectively ) . Moreover , the CD4 CM proliferative burst was sustained longer in aged animals infected with the LR strain compared to 37997 . Overall , the frequency of proliferating CD4 CM T cells was lower in aged animals compared to adults , and this difference was most noticeable in CHIKV-37997 infected animals ( Figure 3B & C ) . CD4 EM T cell proliferation was detected 10 dpi and peaked 14 dpi in LR infected young and aged animals before returning to baseline at 21 dpi ( Figure 3D ) . In 37997 infected adult animals , CD4 EM T cell proliferation was detected and peaked10 dpi returning to baseline by 21 dpi ( Figure 3E ) . As described for CD4 CM , CD4 EM T cell proliferation was delayed in aged animals infected with CHIKV-37997 the peak occurring 14 dpi and returning to baseline levels 21 dpi ( Figure 3E; p = 0 . 055 ) . In summary , proliferation of the CD4 CM subset preceded that of the CD4 EM subset . In addition , CD4 T cell proliferative burst was greater in animals infected with the LR strain compared to 37997 and aged animals infected with 37997 generated a smaller CD4 T cell proliferative response than adult animals . Similar patterns of proliferation were observed in the CD8 T cell subsets . Proliferation within the CD8 CM subset was detected in all LR-infected animals at 7 dpi ( Figure 3F & G ) . Proliferation peaked at 10 dpi and was sustained in both adult and aged animals infected with the CHIKV-LR strain through 14 dpi , returning to baseline by 21 dpi ( Figure 3F ) . In contrast , CD8 CM proliferative burst was shorter in adult animals infected with 37997 peaking 10 dpi and declining 14 dpi ( Figure 3G ) . Aged animals infected with 37997 exhibited a delayed and blunted CD8 CM proliferative response compared to adult animals ( Figure 3G; p = 0 . 061 ) . Proliferation within the CD8 EM subset was detected 10 dpi , remained stable until 14 dpi returning to baseline 21 dpi in CHIKV LR-infected animals ( Figure 3H ) . Adult and aged CHIKV-LR infected animals generated comparable CD8 EM proliferative bursts . CD8 EM in 37997-infected adult animals peaked 10 dpi before returning to baseline 14 dpi . As described for all other subsets , CD8 EM proliferation was delayed in 37997-infected aged animals ( p = 0 . 126 ) . In summary , aged animals infected with CHIKV-LR showed a comparable CD8 T cell proliferative burst following viral infection compared to adult animals . In contrast , T cell proliferation in aged animals infected with CHIKV-37997 was delayed and blunted compared to that of adult animals . To further characterize the impact of age on T cell response to CHIKV infection in vivo , we used IFN-γ ELISPOT assays to determine the frequency of CHIKV-specific T cells in PBMC . A complete overlapping peptide library to CHIKV was subdivided into peptides for each of the 9 proteins ( NSP-1-4 , Core , E1-3 and 6k ) . PBMC collected at 35 dpi were incubated with peptides and the mean numbers of spot forming units [26] per 6×105 PBMC were determined for four adult and four aged animals . A representative example of the data is presented in Figure 4A and the data summary is provided in Figure 4B . T cell response to CHIKV antigens was quite broad targeting both structural and nonstructural proteins . The CHKV-specific T cell response in adult animals was higher compared to the aged animals ( 403±88 vs . 256±73 , respectively; n = 4 ) . Similarly , the breadth of the anti-CHIKV T cell response to any given viral protein differed slightly between adult and aged animals ( Figure 4C ) . Frequencies of T cells specific for NS1 dominate the T cell response in both adult and aged animals , whereas the response to core protein was the second largest in adult animals , NSP4 was the second largest target for aged T cells . These data indicate that advanced age results in reduced frequency and altered breadth of anti-CHIKV T cells response . We next assessed the B cell response by measuring the kinetics and magnitude of the B cell proliferative response as well as IgG production following infection . B cells can be subdivided into three subsets based on the expression of CD27 and IgD as shown in Figure 5A and previously described [25]: naïve ( Na , IgD+ , CD27− ) , marginal zone like ( MZ-like , IgD+ , CD27+ ) and memory B cells ( IgD− , CD27+ ) . MZ-like B cell proliferation in adult animals infected with CHIKV-LR was detected at 10 dpi , peaked around 14 dpi and returned to baseline by 21 dpi ( Figure 5B ) . MZ-like B cell proliferation in aged animals was not detected until 14 dpi ( Figure 5B ) . Proliferation in the memory B cell subset was detected at 14 dpi in adult and aged CHIKV-LR-infected animals ( Figure 5D ) . As described for T cell subsets , B cell proliferation was delayed and reduced in CHIKV-37997-infected aged animals compared to adult ( Figure 5C & E ) . This difference was most noticeable for MZ-like B cells ( p = 0 . 074 ) . Of note , memory B cell proliferation seemed biphasic with detection of a second peak in adult animals infected with both virus strains at 28 dpi that was absent in the aged animals . This second proliferative peak could be indicative of MZ-like B cells acquiring memory phenotype and continuing to proliferate in adult animals . The lack of this second proliferative burst in memory B cells in aged animals is likely another indication of the failures of an aged immune system . To further characterize B cell response , we measure IgG end point titers by ELISA using 96-well plates coated with either purified virus or cellular lysates produced from infected Vero cells . When using purified virions as the antigen source , anti-CHIKV IgGs were detected at 14 dpi and peaked near 21 dpi in LR-infected adult animals ( Figure 6A ) . The IgG response was significantly reduced in aged animals ( p<0 . 05 for days 21 , 28 and 35 ) . When using viral lysate as antigen , IgG titers were first detected 14 dpi , continued to increase until 35 dpi and were comparable between adult and aged animals ( Figure 6B ) . Interestingly , the peak IgG titers to viral lysate were 3-fold higher than those to whole virions ( end point titer of 14545 vs . 4881 , respectively ) . Similar kinetics were observed for the IgG response to CHIKV-37997 in adult animals ( Figure 6C & D ) . The IgG titers were also higher in response to lysate antigen compared to virions ( 4-fold ) ( Figure 6C & D ) . However , in CHIKV-37997-infected aged animals , peak IgG titers to virions or lysate were comparable between aged and adult animals ( Figure 6C & D ) . These data taken together indicate that IgG response against viral determinants is reduced in LR-infected aged animals , possibly contributing to viral persistence in these animals . Further experimentation will be required to determine whether there are differences in specific antibody binding epitopes or avidity of antibodies that bind CHIKV between the aged and adult RM [23] , [24] . Given the proposed role for monocyte/macrophages in CHIKV pathogenesis [27] and formation of CHIKV immune response , we characterized the effects of CHIKV infection on myeloid cell populations in peripheral blood . For this assay , total PBMC were stained with antibody panels allowing separation of myeloid cells into 4 subpopulations: plasmacytoid dendritic cells ( pDC: CD3− , CD11c− , CD14− , CD20− , CD123+ and DR+ ) ; myeloid dendritic cells ( mDC: CD3− , CD11c+ , CD14− , CD20− , CD123− and DR+ ) ; non-plasmacytoid/non-myeloid dendritic cells ( other DC: CD3− , CD11c− , CD14− , CD20− , CD123− and DR+ ) ; and monocyte/macrophages ( CD3− , CD11c− , CD14+ , CD20− , CD123− and DR+ ) as depicted in Figure 7A . The frequency of all DC subsets as well as monocyte/macrophages , increased in the peripheral blood of adult animals after CHIKV infection ( Figure 7B–E ) . Of note , frequency of pDCs increased ∼14 dpi only in adult animals ( Figure 7B; p = 0 . 017 and p = 0 . 164 for LR vs . 37997 at 14 dpi respectively ) . This overall increase appeared more dramatic following infection with 37997 compared to LR strain . Frequency of mDCs increased in LR infected adult and aged animals 14 dpi before returning to baseline ( Figure 7C ) . In 37997 infected adult animals , mDC frequencies increased 7 dpi and remained fairly stable whereas aged animals did not experience any changes in mDC numbers . Frequency of other DC increased following infection with CHIKV-LR strain reaching comparable peak levels 7 dpi in both adult and aged animals before returning slowly to baseline levels 42 dpi ( Figure 7D ) . In contrast , infection with 37997 induced an increase in the frequency of other DCs only in adult animals ( p = 0 . 029 ) . Lastly , frequencies of monocyte/macrophages increased only in adult animals infected with both CHIKV strains ( Figure 7E; LR p = 0 . 019 and 37997 p = 0 . 048 ) . As described for pDCs and other DCs , this increase was more dramatic following 37997 than LR infection . To determine whether the innate immune response was also affected by age following CHIKV infection , we measured plasma levels of bioactive IFN . For this experiment , rhesus macaque fibroblasts were treated with UV inactivated plasma from CHIKV infected aged or adult animals at 3 dpi . Interferon stimulated gene ( ISG ) expression was determined by quantitative RT-PCR using specific primers . We detected a trend towards increased ISG expression in fibroblasts treated with plasma from the adult animals indicating that the levels of bioactive IFN was higher in those samples but this difference did not reach statistical significance ( Figure 8 ) .
Chikungunya virus is a re-emerging Alphavirus that causes debilitating arthralgia . The 2006 outbreak on the island of La Reunion involved a newly emerged CHIKV strain . This strain exhibits both an increased vector range making more widespread transmission highly likely but also more severe virus-induced disease . Ultimately these phenomena demonstrate that CHIKV is an adaptable pathogen capable of rapidly expanding its ecological niche and potentially emerging globally . Moreover , this new strain was associated with increased mortality in the elderly and newborns . Unfortunately , the host response to CHIKV infection and associated determinants of disease remain poorly understood . Characterization of the immune response following CHIKV infection in humans is hampered by the occurrence of symptoms at least a week after infection as well as severity of disease that can differ dramatically between patients , which both influence the time of presentation in the clinic . Thus , novel models that recapitulate human CHIKV disease are needed to facilitate the identification of the immune correlates of protection against CHIKV infection , and for evaluation of therapeutics and vaccines . Recently studies in Cynomolgus monkeys have shown that infection of non-human primates results in similar pathophysiology as that described for humans with the involvement of several joints , but the host immune responses were not evaluated [22] . Therefore , we compared the immune response following infection with the recent outbreak strain CHIKV-LR to the West African strain 37997 in adult and aged rhesus macaques . This is the first report describing the impact of age on CHIKV disease and immune response in non-human primates . Our results show that although viremia is detected within 24 hpi in both adult and aged animals , LR infection resulted in slightly higher peak viremia . Interestingly , at necropsy ( 35 dpi ) we were able to detect persistent viral RNA in spleens harvested from aged but not adult animals infected with the LR strain suggesting that aged animals were unable to fully clear viral infection . One possible explanation for persistent viral RNA in aged animals is a reduced immune response to CHIKV . Therefore , we characterized the innate and adaptive immune responses following CHIKV infection . Following infection with CHIKV-LR , adult and aged rhesus macaques generated comparable T and B cell proliferative responses , similar IgG antibodies to viral lysate but lower IgG antibodies to whole virions . In contrast , aged animals infected with 37997 generated weaker T cell and B cell proliferative responses ( delayed and reduced ) compared to adult animals , but the IgG responses were comparable . We also measured frequency of CHIKV-specific T cells using an overlapping peptide library for the entire CHIKV proteome . Overall CHIKV-specific T cell responses were lower in aged animals and the immunogenicity profiles differed . Aged animals also failed to exhibit an increase in the frequency of myeloid innate immune cells . This failure could be mediated by a defect in innate immune cell mobilization or proliferation . Lastly , production of bioactive Type 1 IFN was lower in the plasma of the aged animals at 3 dpi . Taken together , our findings suggest that aged animals have impaired ability to mount an effective immune response against CHIKV infection . Aging results in a general decline in immune function , commonly known as immune senescence , which contributes to the increased susceptibility to pathogens and cancer . Defects associated with immune senescence have been shown in both innate and adaptive processes . In the current report we found a reduction in innate immune responses in the aged animals infected with CHIKV compared to adult animals . Importantly , these responses were observed in animals infected with either the LR or 37997 strains . Of note , aged animals do not exhibit an increase in the frequency of pDC . Given that pDCs play a critical role in the antiviral response through the production of Type 1 IFN [28] , the results presented in this manuscript suggest that pDCs and production of type 1 IFN may contribute to the successful control of CHIKV viremia . Previous studies have shown a decrease in circulating pDC numbers with increasing age [29] , [30] . The data presented here suggest additional defects in pDC mobilization . We observed a corresponding reduction in the production of plasma levels of bioactive Type 1 IFN in aged animals . Both clinical data and rodent studies using IFNαR-/- and STAT1-/- mice demonstrate a critical role for type I IFN in CHIKV clearance [20] , [31] . Others have demonstrated an age-related decrease in IFN response through STAT1 , IRF1 and IRF7 signaling following infection with West Nile Virus [32] , [33] . Other cell types of the innate immune response such as macrophages and myeloid DCs play an important role in antigen presentation and the initiation of the adaptive immune response . Age-related defects in macrophage and DC antigen presentation and migration reduce their ability to prime naïve T cell responses to novel pathogens [34] , [35] . Aged animals infected with CHIKV did not show an increase in monocyte/macrophage and mDCs following CHIKV infection . This defect could in part explain the reduced T cell proliferation observed in 37997 infected animals In line with decreased magnitude of the T cell proliferative burst , we found reduced frequency of CHIKV-specific T cells in the aged animals by IFN-γ ELISPOT . Interestingly , the immunodominance profile also differed between adult and aged animals . Similarly , we observed a decreased IgG antibody response to whole virions . However , IgG titers measured using plates coated with viral lysate were higher than those measured using purified virions suggesting that some of the antibody response is directed against viral antigens that are not included in the virion . Collectively , these findings indicate that the aged immune system is incapable of mounting a broad and effective adaptive immune response to CHIKV . In summary , we present data indicating that the recent outbreak strain CHIKV-LR replicated , on average , to higher levels compared to the CHIKV strain 37997 . While we did not observe CHIKV-associated joint inflammation or disease for either virus in these animals at 35 dpi , we did identify the spleen as a potential site of CHIKV persistence in aged animals . This effect was limited to CHIKV-LR and did not occur in monkeys infected with CHIKV-37997 , which is in line with the increased severity of CHIKV-LR disease . We have also identified a number of differences in the ability of adult vs . aged animals to respond to CHIKV infection , with a general decline in both the innate and adaptive immune responses that could explain the increased disease severity observed in older individuals . | Chikungunya virus ( CHIKV ) is a re-emerging Alphavirus that has caused recent massive outbreaks in the Indian Ocean region . In addition , outbreaks have been documented in Europe and elsewhere in the world , initiated by infected travelers returning to their homelands . The recent outbreak strains possess extended vector range and as such , raise the potential of CHIKV outbreaks in the Southeastern parts of the United States . In this study , we examined CHIKV immunity in adult and aged Rhesus macaques following infection with two different CHIKV strains ( recent outbreak strain CHIKV-LR and a West African Strain CHIKV-37997 ) . CHIKV-LR causes persistent infection in the aged animals and replicates , on average , to higher levels than CHIKV-37997 . Irrespective of the viral strain used , aged animals had delayed and/or reduced immunity compared to adult animals . Our data support the clinical findings of CHIKV susceptibility in vulnerable populations including the aged and provide mechanistic evidence that an effective immune response directed against the virus is required for preventing persistent CHIKV infection . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biology"
] | 2013 | Chikungunya Virus Infection Results in Higher and Persistent Viral Replication in Aged Rhesus Macaques Due to Defects in Anti-Viral Immunity |
The NAD+-dependent histone deacetylase Sir2 was originally identified in Saccharomyces cerevisiae as a silencing factor for HML and HMR , the heterochromatic cassettes utilized as donor templates during mating-type switching . MATa cells preferentially switch to MATα using HML as the donor , which is driven by an adjacent cis-acting element called the recombination enhancer ( RE ) . In this study we demonstrate that Sir2 and the condensin complex are recruited to the RE exclusively in MATa cells , specifically to the promoter of a small gene within the right half of the RE known as RDT1 . We also provide evidence that the RDT1 promoter functions as a locus control region ( LCR ) that regulates both transcription and long-range chromatin interactions . Sir2 represses RDT1 transcription until it is removed from the promoter in response to a dsDNA break at the MAT locus induced by HO endonuclease during mating-type switching . Condensin is also recruited to the RDT1 promoter and is displaced upon HO induction , but does not significantly repress RDT1 transcription . Instead condensin appears to promote mating-type donor preference by maintaining proper chromosome III architecture , which is defined by the interaction of HML with the right arm of chromosome III , including MATa and HMR . Remarkably , eliminating Sir2 and condensin recruitment to the RDT1 promoter disrupts this structure and reveals an aberrant interaction between MATa and HMR , consistent with the partially defective donor preference for this mutant . Global condensin subunit depletion also impairs mating-type switching efficiency and donor preference , suggesting that modulation of chromosome architecture plays a significant role in controlling mating-type switching , thus providing a novel model for dissecting condensin function in vivo .
Since the first descriptions of mating-type switching in budding yeast over 40 years ago , characterization of this process has led to numerous advances in understanding mechanisms of gene silencing ( heterochromatin ) , cell-fate determination ( mating-type ) , and homologous recombination ( reviewed in [1] ) . For example , the NAD+-dependent histone deacetylase , Sir2 , and other Silent Information Regulator ( SIR ) proteins , were genetically identified due to their roles in silencing the heterochromatic HML and HMR loci , which are maintained as silenced copies of the active MATα and MATa loci , respectively [2–4] . The SIR silencing complex ( Sir2-Sir3-Sir4 ) is recruited to cis-acting E and I silencer elements flanking HML and HMR through physical interactions with silencer binding factors Rap1 , ORC , and Abf1 , as well as histones H3 and H4 ( reviewed in [5] ) . HML and HMR play a critical role in mating-type switching . Haploid cells of the same mating-type cannot mate to form diploids , the preferred cell type in the wild . Therefore , in order to facilitate mating and diploid formation , haploid mother cells switch their mating-type by expressing HO endonuclease , which introduces a programmed DNA double-strand break ( DSB ) at the MAT locus [6] . The break is then repaired by homologous recombination using either HML or HMR as a donor template for gene conversion [6 , 7] . This change in mating-type enables immediate diploid formation between mother and daughter . HO is deleted from most standard lab strains in order to maintain them as haploids , so expression of HO from an inducible promoter such as PGAL1 is commonly used to switch mating-types during strain construction [8] . There is a “donor preference” directionality to mating-type switching such that ~90% of the time , the HO-induced DSB is repaired to the opposite mating-type [9] . For example , MATα cells preferentially switch to MATa using HMR as the donor . However , while both silent mating loci can be utilized as a donor template , usage of HML by MATa cells requires a 2 . 5 kb intergenic region located ~17 kb from HML called the recombination enhancer ( RE ) [10] . Donor preference activity within the RE has been further narrowed down to a 700 bp segment containing an Mcm1/α2 binding site ( DPS1 ) and multiple Fkh1 binding sites [10] . The RE is active in MATa cells , requiring Mcm1 and Fkh1 activity at their respective binding sites [10–12] . The RE is inactivated in MATα cells due to expression of transcription factor α2 from MATα [13] , which forms a repressive heterodimer with Mcm1 ( Mcm1/α2 ) to repress MATa-specific genes [1] . Current models for donor preference posit that Fkh1 at the RE helps position HML in close proximity with MATa by interacting with threonine-phosphorylated H2A ( γ-H2AX ) and Mph1 DNA helicase at the HO-induced DNA DSB [14 , 15] . Sir2-dependent silencing of HML and HMR has two known functions related to mating-type switching . First , HML and HMR must be silenced in haploids to prevent formation of the a1/α2 heterodimer , which would otherwise inactivate haploid-specific genes such as HO [16] . Second , heterochromatin structure at HML and HMR blocks cleavage by HO , thus restricting its activity to the fully accessible MAT locus [17 , 18] . Here we describe new roles for Sir2 and the condensin complex within the RE during mating-type switching . ChIP-seq analysis revealed strong overlapping binding sites for Sir2 and condensin at the promoter of a small gene within the RE known as RDT1 . Here , Sir2 was found to repress the MATa-specific transcription of RDT1 , which is also translated into a small 28 amino acid peptide . RDT1 expression is also dramatically upregulated during mating-type switching when Sir2 is lost from the RDT1 promoter and instead associates with the HO-induced DNA DSB at MATa . Furthermore , eliminating Sir2/condensin recruitment to the RDT1 promoter disrupts chromosome III architecture such that donor preference is partially impaired . The RDT1 promoter region therefore functions like a classic locus control region ( LCR ) in MATa yeast cells , regulating localized transcription as well as long-range chromosome interactions .
We previously characterized global sirtuin distribution using ChIP-Seq to identify novel loci regulated by Sir2 and its homologs [19] . Significant overlap was observed between binding sites for Sir2 , Hst1 , or Sum1 with previously described condensin binding sites [19 , 20] , suggesting a possible functional connection . ChIP-Seq was therefore performed on WT and sir2Δ strains in which the condensin subunit Smc4 was C-terminally tagged ( 13xMyc ) ( Fig 1A ) . To avoid “hyper-ChIPable” loci that can appear in yeast ChIP-seq experiments , we also ran nuclear localized GFP controls [21] . Genes closest to Sir2-dependent condensin peaks after subtraction of GFP are listed in S1 Table , and are distributed throughout the genome . One of the strongest peaks overlapped with a Sir2-myc binding site on chromosome III between KAR4 and SPB1 that was not enriched for GFP ( Fig 1A ) . The specificity of Sir2 enrichment at this peak , as opposed to the adjacent SPB1 gene , was independently confirmed by quantitative ChIP using an α-Sir2 antibody ( Fig 1B ) , with enrichment comparable to levels observed at the HML-I silencer ( Fig 1A and 1B ) . Sir2-dependent condensin binding was also confirmed for Myc-tagged Smc4 and Brn1 subunits ( Fig 1C ) . The ~2 . 5 kb intergenic region between KAR4 and SPB1 was previously defined as a cis-acting recombination enhancer ( RE ) that specifies donor preference of mating-type switching in MATa cells [10 , 13] . Quantitative ChIP assays revealed that Sir2 and Brn1-myc enrichment at the RE was also MATa-specific ( Fig 1D and 1E ) , which was notable because the ChIP-seq datasets in Fig 1A happened to be generated from MATa strains . We next considered whether the condensin binding defect in the MATa sir2Δ mutant was due to HMLALPHA2 expression caused by defective HML silencing . To test this idea , we re-examined Brn1-myc ChIP signal at the RE in strains lacking HML , and found that deleting SIR2 no longer affected condensin recruitment ( Fig 1F ) . Similarly , a MATa condensin mutant ( ycs4-1 ) known to have an HML silencing defect [22] reduced Sir2 recruitment to the RE , but had no effect when HML was also deleted ( Fig 1G ) . Sir2 and condensin are therefore independently recruited to the RE only in MATa cells . Donor preference activity ascribed to the RE was previously narrowed down to a KAR4 ( YCL055W ) -proximal 700 bp domain defined by an Mcm1/α2 binding site ( Fig 2A , DPS1 ) [10 , 11 , 13] . The Sir2 and condensin ChIP-seq peaks we identified were located outside this region , between a second Mcm1/α2 binding site ( DPS2 ) and a small gene of unknown function called RDT1 [23] ( Figs 1A and 2A ) . We noticed the location of RDT1 coincided with the smallest of several putative non-coding RNAs ( ncRNA ) previously reported as being transcribed from the RE , but not annotated in SGD ( Fig 2A , [13] ) . Quantitative RT-PCR and analysis of publicly available RNA-seq data from BY4741 ( MATa ) and BY4742 ( MATα ) revealed that RDT1 expression was indeed MATa specific ( Fig 2B and S1A Fig ) . We next asked whether Sir2 and/or condensin regulate histone acetylation and RDT1 expression when recruited to the RE . Sir2 normally represses transcription at HML , HMR , and telomeres as a catalytic subunit of the SIR complex where it preferentially deacetylates H4K16 ( reviewed in [5] ) . Accordingly , deleting SIR2 , SIR3 , or SIR4 from MATa cells increased H4K16 acetylation at the RDT1 promoter ( Fig 2C ) , consistent with the observed enrichment of Sir3-myc and Sir4-myc at this site ( S1B Fig ) . Furthermore , re-introducing active SIR2 into the sir2Δ mutant restored H4K16 to the hypoacetylated state , whereas catalytically inactive sir2-H364Y did not ( Fig 2D ) . While H4K16ac is a preferred Sir2 substrate for silencing at HML and HMR , the recruited SIR complex also maintains lysine deacetylation of the other N-terminal histone tails [24] . Since RDT1 transcription is MATa specific , and Mcm1/α2 represses MATa-specific genes by recruiting the Ssn6/Tup1 corepressor complex and Class I/II HDACs such as the H3/H2B-specific histone deacetylase HDA1 [25 , 26] , we also tested the effect of deleting SIR2 on H3K9/14 acetylation at the RDT1 promoter , predicting it may remain hypoacetylated due to the SIR complex being replaced by Ssn6/Tup1/HDA1 . Indeed , H3K9/14 acetylation was reduced in the sir2Δ mutant relative to WT , but was significantly elevated in the hmlΔ sir2Δ double mutant ( Fig 2E ) . RDT1 expression was similarly reduced in the sir2Δ mutant and strongly upregulated when HML and SIR2 were both deleted ( Fig 2F ) . An even stronger expression effect was observed in an hmlΔ sir2Δ hst1Δ triple mutant that eliminates any possibility of redundancy between the Sir2 and Hst1 paralogs ( S1C Fig ) . On the other hand , RDT1 was not upregulated in an hmlΔ ycs4-1 condensin mutant ( Fig 2G ) . Taken together , these results provide strong evidence that the SIR complex represses RDT1 in MATa cells by establishing a generally hypoacetylated chromatin environment at the promoter , while condensin has a functional role independent of transcriptional regulation . We next attempted to prevent Sir2 and condensin recruitment to the RDT1 promoter by precisely deleting a 100bp DNA sequence underlying the shared enrichment region ( coordinates 30701–30800 ) , while not disturbing the adjacent Mcm1/α2 site ( Fig 3A ) . Sir2 and Brn1-myc binding to the RE as measured by ChIP was greatly diminished in this mutant ( Fig 3B and 3C ) , despite unaltered Sir2 , Brn1-myc , or Smc4-myc expression levels ( S2A–S2C Fig ) . Furthermore , RDT1 RNA expression level was significantly increased by the 100bp deletion exclusively in MATa cells ( Fig 3D ) , consistent with the loss of Sir2-mediated repression . Because Sir2 and condensin were not present at the RDT1 promoter in MATα cells , we reasoned that their binding should require a MATa specific transcription factor . This made the 2nd Mcm1/α2 binding site ( DPS2 ) upstream of the Sir2/condensin ChIP-seq peaks an ideal candidate , because it had not previously been ascribed a function other than redundancy with DPS1 for donor preference [13] . Deleting MCM1 is lethal , so alternatively , we precisely deleted the 2nd Mcm1/α2 binding site ( ChrIII coordinates 30595 to 30626 , S3A Fig ) and then retested for Sir2 and Brn1-myc enrichment . As shown in S3B and S3C Fig , respectively , Sir2 and Brn1-myc enrichment at the Mcm1/α2 binding site ( DPS2 ) and the RDT1 promoter ( defined as the Sir2/condensin peaks ) was significantly reduced in the binding site mutant . These results suggest that Mcm1 nucleates a complex that recruits the SIR and condensin complexes to the RDT1 promoter in MATa cells , and also provides a possible mechanism of blocking the recruitment in MATα cells due to the interaction of Mcm1 with α2 . Ribosome Detected Transcript-1 ( RDT1 ) was originally annotated as a newly evolved gene whose transcript was associated with ribosomes and predicted to have a small open reading frame of 28 amino acids [23] . Our work suggested that RDT1 and the putative non-coding R2 transcript were the same ( Fig 2A ) . To determine if RDT1/R2 codes for a small protein , the ORF was C-terminally fused with a 13x-Myc epitope in MATa and MATα cells . As shown in Fig 3E , a fusion protein was exclusively detected in exponentially growing MATa WT cells and also highly expressed in the 100bpΔ strain , correlating with the increased RNA level observed for that mutant in Fig 3D . Since additional MATa-specific RNAs are derived from the minimal 700bp RE domain ( Fig 2A; R1L and R1S ) [13 , 27] , we next tested whether Sir2/condensin recruitment to the RDT1 promoter had any effect on expression of these upstream ncRNAs from a distance . Quantitative RT-PCR for the R1L/S transcripts indicated their expression level in the WT strain was comparable to RDT1 , and was also reduced in a sir2Δ mutant because of the pseudodiploid phenotype ( Fig 3F ) . However , while RDT1 was strongly upregulated by the 100bp deletion of the Sir2/condensin binding site , R1L/S expression was unaffected ( Fig 3F ) . As a control , we also measured expression of the SPB1 gene located immediately downstream of RDT1 ( see Fig 2A schematic ) . SPB1 expression is not mating-type specific , it encodes a rRNA methyltransferase required for maturation of the large 60S ribosomal subunit [28] , and interestingly , also functions in silencing at HML and HMR [29] . It was therefore intriguing that SBP1 expression was increased 2- to 3-fold in the sir2Δ and 100bpΔ mutants ( Fig 3G ) , suggesting that Sir2 at the RDT1 promoter has a modest downstream repressive effect on SPB1 , but not on the upstream R1L/S ncRNAs . Notably , steady state RDT1 and R1L/S RNA levels were relatively low compared to SPB1 and the ACT1 loading control , even in the 100bpΔ mutant ( Fig 3G ) . We next asked if Sir2 played any role in regulating RDT1 during mating-type switching . Sir2 was previously shown to associate with a HO-induced DSB at the MAT locus during mating-type switching , presumably to effect repair through histone deacetylation [30] . SIR complex association with DSBs occurs at the expense of telomere binding [31] , so we hypothesized that the HO-induced DSB at MAT could also trigger loss of Sir2 from the RDT1 promoter , thus facilitating increased RDT1 transcription . To test this idea , HO was induced at time 0 with galactose and then turned off 2 hours later by glucose addition to allow for repair/switching to occur ( Fig 4A and 4B ) . By the 3 hr time point ( 1 hr after glucose addition ) , ChIP analysis indicated Sir2 was maximally enriched at the MAT locus ( Fig 4C ) , corresponding to the time of peak mating-type switching ( [30] and Fig 4B ) . Interestingly , Sir2 was significantly depleted from the RDT1 promoter within 1 hr after HO induction , and by 3 hr there was actually stronger enrichment of Sir2 at MAT than RDT1 ( Fig 4C ) . Critically , this shift in Sir2 distribution correlated with maximal induction of RDT1 mRNA and the Myc-tagged Rdt1protein ( Fig 4D and 4E , 3 hr ) . Once switching was completed by 4 hr ( 2 hr after glucose addition ) , RDT1 transcription was permanently inactivated and Sir2 binding never returned because most cells were now MATα . The Myc-tagged Rdt1 protein , however , remained elevated for the rest of the time course ( Fig 4E ) , suggesting that it is relatively stable , at least when epitope tagged . A parallel ChIP time course experiment was performed with condensin ( Brn1-myc ) , indicating significant depletion from the RDT1 promoter within 1 hr ( Fig 4F ) , similar to the timing of Sir2 loss . Unlike Sir2 , Brn1-myc enrichment at the HO-cleaved MAT site did not increase , suggesting that condensin normally associates with MATa in non-switching cells and is then displaced in response to the HO-induced DSB , perhaps to facilitate structural reorganization associated with switching . Since RDT1 was highly expressed during mating-type switching , we next asked whether the small protein encoded by this gene had a direct function during the switching process when using the galactose-inducible system employed in this study . The 28 amino acid ORF was precisely deleted using the delitto perfetto method [32] , and the efficiency of switching from MATa to MATα was then tested across a time course by PCR ( S4A and S4B Fig ) . No significant differences were observed between the WT and rdt1Δ strains . Next , donor preference was tested using a strain previously developed by the Haber lab [15] , whereby HMRa was replaced with HMRα containing a unique BamHI site ( HMRα-B ) ( S4C Fig ) . Following completion of Gal-HO-induced switching from MATa to MATα , the proportion of donor utilization was determined by BamHI digest of a MATα PCR product . As shown in S4D and S4E Fig , deleting RDT1 also had no effect on donor preference , indicated by low utilization ( ~9% ) of HMRα-B . Therefore , although RDT1 gene expression strongly correlates with switching , a specific function for its gene product remains elusive . Therefore , we shifted our attention to a possible function for the RDT1 promoter . The coupling of Sir2 and condensin distribution with RDT1 transcriptional regulation during mating-type switching was reminiscent of classic locus control regions ( LCR ) that modulate long-range chromatin interactions [33 , 34] . We therefore hypothesized that the RDT1 promoter region functions as an LCR to modulate long-range chromatin interactions of chromosome III . To test this hypothesis , we performed Hi-C analysis with WT , sir2Δ and the 100bpΔ strains . Genomic contact differences between the mutants and WT were quantified using the HOMER Hi-C software suite [35] , and the frequency of statistically significant differences for each chromosome calculated ( Fig 5A ) . Chromosome III had the most significant differences in both mutants , so we focused on this chromosome and used HOMER to plot the observed/expected interaction frequency in 10kb bins for each strain as a heat map ( Fig 5B ) . In a WT strain ( ML1 ) there was strong interaction between the left and right ends of chromosome III , mostly centered around the HML ( bin 2 ) and HMR ( bin 29 ) loci . Interestingly , HML ( bin 2 ) also appeared to sample the entire right arm of chromosome III , with the interaction frequency increasing as a gradient from CEN3 to a maximal observed interaction at HMR , thus also encompassing the MATa locus at bin 20 . This distinct interaction pattern was completely disrupted in the sir2Δ mutant , whereas some telomere-subtelomere contacts were retained in the 100bpΔ mutant ( Fig 5B ) , suggesting there was still limited interaction between the left and right ends of the chromosome . We confirmed the changes in HML-HMR interaction for these strains using a quantitative 3C-PCR assay to rule out sequencing artifacts ( Fig 5C ) , and to confirm an earlier sir2Δ 3C result from the Dekker lab [36] . Importantly , despite the loss of HML-HMR interaction in the 100bpΔ mutant , heterochromatin at these domains was unaffected based on normal quantitative mating assays ( S5A Fig ) , and unaltered Sir2 association with HML ( S5B Fig ) . We next analyzed the Hi-C data using an iterative correction method that reduces background to reveal interacting loci that potentially drive the overall chromosomal architecture , rather than passenger locus effects [37] . HML ( bin 2 ) and HMR ( bin 29 ) again formed the dominant interaction pair off the diagonal in WT , which was lost in the sir2Δ or 100bpΔ mutants ( Fig 5D , red arrows ) . Importantly , a prominent new interaction between HMR ( bin 29 ) and MATa ( bin 20 ) appeared in both mutants ( Fig 5D , black arrows ) , as would be predicted if normal donor preference of MATa cells was altered . We conclude that the RDT1 promoter does function like an LCR in MATa yeast cells , regulating localized transcription and establishing a long-range chromatin interaction between HML and HMR that appears to prevent HMR from strongly associating with MATa ( Fig 5E ) . Sir2/condensin binding was observed in the right half of the RE ( Fig 1A ) , but this region was previously reported as dispensable for donor preference activity [10] . Considering that HMR was aberrantly associated with the MATa locus in sir2Δ and 100bpΔ mutants ( Fig 5B–5E ) , we proceeded to test whether these mutants had any alterations in donor preference . As was done for the rdt1Δ mutant in S4C and S4D Fig , the 100bpΔ mutation was introduced into a reporter strain with HMRα-B on the right arm of chromosome III ( Fig 6A; [15] ) . After inducing switching to MATα with galactose , the proportion of HMLα or HMRα-B used for the switching was determined by BamHI digestion of a MATα-specific PCR product ( Fig 6B; [15] ) . As expected for normal donor preference , HMRα-B on the right arm was only utilized ~9% of the time in the WT strain ( Fig 6C and 6D ) . Strikingly , donor preference was lost in the sir2Δ mutant , similar to a control strain with the RE deleted ( Fig 6C and 6D ) , and consistent with the clear interaction between HMR and MATa bins observed for the sir2Δ mutant in Fig 5D and 5E . This interaction was less prominent in the 100bpΔ mutant ( Fig 5D ) , and the corresponding donor preference defect was also less severe ( ~25% HMRα-B ) , though still significantly different from WT ( Fig 6C and 6D ) . Since the donor preference assay is an endpoint experiment , we next tested whether the sir2Δ or 100bpΔ mutations temporally impacted the efficiency of switching from MATa to MATα in the same ML1 background strains used for Hi-C analysis . As shown in Fig 6E and 6F , switching efficiency was dramatically impaired in the sir2Δ strain , but unaffected in the 100bpΔ mutant . These results suggest a model whereby condensin and Sir2 recruitment to the RDT1 promoter supports donor preference by organizing chromosome III into a structure that limits HMR association with the MATa locus , but is not required for the mechanics of mating-type switching . We suspect at least part of the strong sir2Δ effect on chromosome III organization and donor preference is caused by HMLALPHA2 derepression , which inactivates the RE due to formation of the Mcm1/α2 heterodimer [13] . We also considered a possibility that the sir2Δ heterochromatin defect at HML and HMR could make them highly accessible to HO cleavage [38] , which would prevent their usage as donor templates . As a measure of HO cleavage at MATa , HML , or HMR , we assayed for reduced PCR amplification across the recognition site following Gal-HO induction ( S6 Fig ) . While MATa was equally cut by HO in WT and sir2Δ strains ( S6A and S6B Fig ) , HML was only cut in the sir2Δ strain ( S6A and S6C Fig ) , consistent with the idea of reduced HML availability for switching . Unexpectedly , HMR remained available as a template in the sir2Δ strain ( S6A and S6D Fig ) . We confirmed the difference in HO accessibility between HML and HMR using real-time qPCR ( S6E and S6F Fig ) , which could contribute to the extreme sir2Δ donor preference defect ( Fig 6C and 6D ) . In the 100bpΔ mutant , which locally eliminates condensin recruitment at RDT1 , the continued maintenance of heterochromatin at HML/HMR and telomeres ( S5 Fig ) , as well as residual telomere clustering ( Fig 5D ) , may partially buffer the resulting donor preference defect by still limiting contact between the subtelomeric HMR locus and MATa . Condensin recruitment to the RDT1 promoter does not appear critical for the mechanics of mating-type switching , as indicated by normal timing of switching in the 100bpΔ mutant ( Fig 6E and 6F ) . However , condensin could still potentially impact the switching process independent of the RE . In order to test this hypothesis , we C-terminally tagged the Brn1 condensin subunit with an auxin-inducible degron ( AID ) fused with a V5 epitope , which allows for rapid depletion of tagged proteins upon the addition of auxin [39] . Indeed , Brn1-AID was effectively degraded within 15 min of adding auxin to cells , as measured by western blotting ( S7A Fig ) , or ChIP at the RDT1 promoter ( S7B Fig ) . Importantly , even after 1 hr of auxin treatment , there were no significant changes in RDT1 or HMLALPHA2 gene expression as measured by qRT-PCR ( S7C and S7D Fig ) , thus indicating that silencing of HML was unaffected , unlike the ycs4-1 condensin mutant used in Fig 1G [22] . The efficiency of ML1 switching from MATa to MATα was then tested across a time course with or without auxin treatment ( Fig 7A ) . As shown in Fig 7B and 7C , auxin significantly slowed the efficiency of switching to MATα , indicating that the Brn1 subunit of condensin is important for normal mating-type switching . Since the 100bpΔ mutant caused a modest donor preference defect that we partially attributed to a loss of condensin ( Fig 6C and 6D ) , it was important to also test for a donor preference defect when condensin was depleted . Indeed , Brn1-AID depletion produced a significant defect in donor preference using the HMRα-B reporter strain ( Fig 7D ) that was similar in magnitude to that observed for the 100bpΔ strain ( Fig 6D ) . Taken together , these results support a working model whereby condensin recruitment to the RDT1 promoter in MATa cells organizes chromosome III into a conformation that limits the association of HMR with MATa , thus partially contributing to donor preference regulation . We hypothesize that upon HO cleavage of MATa , the increased expression of RDT1 caused by loss of Sir2 , displaces condensin from the promoter , which then allows the left half of the RE to physically direct HML to MATa for use as a donor ( Fig 8 ) .
While we do not yet know the molecular function of RDT1 in mating-type regulation or other cellular processes , the promoter region of this gene clearly controls the structure of chromosome III . Three-dimensional chromatin structure has long been proposed to influence donor preference [40 , 41] . However , deleting the minimal 700bp ( left half ) of the RE alters donor preference without a large change in chromosome III conformation . Furthermore , deleting the right half of the RE , which includes RDT1 , changes chromosome III conformation without a dramatic change in donor preference [10 , 13 , 42] . Based on these findings it was proposed that the RE is a bipartite regulatory element [42] , with the left half primarily responsible for donor preference activity and the right half for chromosome III structure . Our results support this view and narrow down the structural regulatory domain of the RE to a small ( 100bp ) region of the RDT1 promoter bound by the SIR and condensin complexes . Importantly , deleting this small region not only altered chromosome III structure , but also had a significant effect on donor preference , though not as strong as the sir2Δ mutation . The coordination of RDT1 expression with loss of Sir2/condensin binding at its promoter during mating-type switching , together with the loss of HML-HMR interaction in the 100bpΔ mutant , makes this site intriguingly similar to classic locus control regions ( LCRs ) in metazoans , which are cis-acting domains that contain a mixture of enhancers , insulators , chromatin opening elements , and tissue-specificity elements [33] . The minimal RE was previously described as an LCR in the context of donor preference [10] , and transcription of the R1S/R1L long non-coding RNAs via activation by the 1st Mcm1/α2 binding site ( DPS1 ) appears to be important for this activity in MATa cells [27] . We find that Sir2 indirectly supports donor preference from the left half of the RE in MATa cells by silencing HMLALPHA2 expression , which prevents transcriptional repression by an Mcm1/α2 heterodimer , and by protecting the HML template from HO cleavage . Similarly , the loss of Sir2 also represses RDT1 expression and condensin recruitment in the right half of the RE due to HMLALPHA2 expression . It remains possible that SIR-dependent heterochromatin formation also directly contributes to the HML-HMR interaction through clustering . More clearly , however , Sir2 directly represses RDT1 through localized histone deacetylation . How the loss of RDT1 regulation and condensin recruitment changes chromosome III structure in the sir2Δ mutant remains unknown , but we propose that the HMR-MATa interaction is a default state , while the HML-HMR association has to be actively maintained by condensin and likely additional factors co-localized to this element , as well as SIR-dependent heterochromatin . Interestingly , there also appears to be a functional relationship between the RE and silencing at the HML locus , such that deleting the left half of the RE specifically stabilizes HML silencing in MATa cells [43] . The mechanism involved remains unknown , but we hypothesize that eliminating this part of the RE could potentially allow the SIR and condensin complexes bound at the RDT1 promoter to encroach and somehow enhance the heterochromatic structure at HML . Under this scenario , the left half of the RE could be insulating HML from the chromosomal organizing activity that occurs at the RDT1 promoter . The RDT1 promoter is a major condensin binding site identified by ChIP-seq ( Fig 1 ) , and given the strong Hi-C interaction between HML and HMR , we initially hypothesized that condensin at the RDT1 promoter would crosslink with another condensin site bound on the right arm of chromosome III . However , this turned out to be unlikely because ChIP-seq of Smc4-myc did not reveal any strong peaks near HMR . The S . cerevisiae condensin complex was recently shown to catalyze ATP-dependent unidirectional loop extrusion using an in vitro single molecule assay [44] . The mechanism involves direct binding of condensin to DNA , followed by one end of the bound DNA being pulled inward as an extruded loop . Applying this model to the strong binding site at the RDT1 promoter , this region could act as an anchor bound by condensin , with DNA to the right being rapidly extruded as a loop until pausing at CEN3 . Extrusion would then continue at a slower rate toward HMR , allowing HML time to sample the entire right arm of chromosome III , until clustering with HMR ( Fig 8 ) . HOMER analysis of the Hi-C data in Fig 5B provides evidence for such a model because there is an ascending gradient of HML interaction frequency with sequences extending from the centromere region ( bin 12 ) toward HMR , suggesting that HML “samples” the right arm of chromosome III . Once brought in contact , HML and HMR would then remain associated due to their heterochromatic states and shared retention at the nuclear envelope [45] . Formation of this loop appears to limit HMR association with MATa , but since the 100bpΔ mutation has no effect on the timing of switching , we do not believe that condensin at the RDT1 promoter functions directly in the homologous recombination process . Rather , general Brn1 subunit depletion could slow the switching process by affecting chromosome III flexibility or conformational dynamics . Condensin , and Sir2 each strongly associated with the RDT1 promoter exclusively in MATa cells , though it is not clear if they bind at the same time , or are differentially bound throughout the cell cycle . Since DPS2 was required for Sir2 and condensin recruitment , and derepression of HMLALPHA2 from HML also eliminated binding , we hypothesized and then demonstrated ( S3 Fig ) that Mcm1 was a key DNA binding factor involved . Mcm1 is a prototypical MADS box combinatorial transcription factor that derives its regulatory specificity through interactions with other factors , such as Ste12 in the case of MATa haploid-specific gene activation , or α2 when repressing the same target genes in MATα cells [46] . This raises the question of whether Mcm1 directly recruits the SIR and condensin complexes , or perhaps additional factors that work with Mcm1 are involved . The latter is likely true because condensin and Sir2 are not recruited to the leftmost Mcm1/α2 binding site in the left half of the RE , as indicated by the ChIP-seq data in Fig 1A . At the RDT1 promoter , specificity for Sir2/condensin recruitment could originate from sequences underlying the condensin/Sir2 peaks . There are no traditional silencer-like sequences for SIR recruitment within the deleted 100bp ( coordinates 30702 to 30801 ) , and yeast condensin does not appear to have a consensus DNA binding sequence [47] . Closer inspection of the RDT1 promoter indicates an A/T rich region with consensus binding sites for the transcription factors Fkh1/2 and Ash1 , each of which regulates mating-type switching [11 , 48 , 49] . Fkh1 and Fkh2 also physically associate with Sir2 at the mitotic cyclin CLB2 promoter during stress [50] . Ash1 is intriguing because it represses HO transcription in daughter cells [49 , 51] , raising the possibility of RDT1 repression in daughter cells . Mcm1 activity in MATa cells could also indirectly establish a chromatin environment that is competent for Sir2/condensin recruitment , rather than direct recruitment through protein-protein interactions . In MATa cells , Mcm1 appears to prevent the strong nucleosome positioning across the RE that occurs in MATα cells [27] , and indicative of an actively remodeled chromatin environment . Perhaps condensin is attracted to such regions , which is consistent with the association of condensin with promoters of active genes in mitotic cells , where enrichment was greatest at unwound regions of DNA [52] . Furthermore , nucleosome eviction by transcriptional coactivators was found to assist condensin loading in yeast [53] , though the mechanism of loading remains poorly understood . Recruitment of condensin to the RDT1 promoter LCR therefore provides an outstanding opportunity for dissecting mechanisms of condensin loading and function .
Yeast strains were grown at 30°C in YPD or synthetic complete ( SC ) medium where indicated . The SIR2 , or HST1 open reading frames ( ORFs ) were deleted with kanMX4 using one-step PCR-mediated gene replacement . HML was deleted and replaced with LEU2 . Precise deletions of the 100bp condensin/Sir2 binding site within the RDT1 promoter ( chrIII coordinates 30701–30800 ) , DPS2 ( chrIII coordinates 30557–30626 ) , or the RDT1 ORF ( chrIII coordinates 30910–30996 ) were generated using the delitto perfetto method [32] . Endogenous SIR2 , BRN1 , or SMC4 genes were C-terminally tagged with the 13xMyc epitope ( 13-EQKLISEEDL ) . Deletion and tagged gene combinations were generated through genetic crosses and tetrad dissection , including Brn1 tagged with a V5-AID tag ( template plasmids kindly provided by Vincent Guacci ) . All genetic manipulations were confirmed by PCR , and expression of tagged proteins confirmed by western blotting . The pGAL-HO-URA3 expression plasmid was kindly provided by Jessica Tyler [30] . Strain genotypes are provided in S2 Table and oligonucleotides listed in S3 Table . Sir2 ChIP-seq was previously described [19] . For other ChIP-seq datasets , log-phase YPD cultures were cross-linked with 1% formaldehyde for 20 min , pelleted , washed with Tris-buffered saline ( TBS ) , and then lysed in 600 μl FA140 lysis buffer with glass beads using a mini-beadbeater ( BioSpec Products ) . Lysates were removed from the beads and sonicated for 60 cycles ( 30s “on” and 30s “off” per cycle ) in a Diagenode Bioruptor . Sonicated lysates were pelleted for 5 min at 14000 rpm in a microcentrifuge and the entire supernatant was transferred to a new microfuge tube and incubated overnight at 4°C with 5 μg of anti-Myc antibody ( 9E10 ) and 20 μl of protein G magnetic beads ( Millipore ) . Following IP , the beads were washed once with FA140 buffer , twice with FA500 buffer , and twice with LiCl wash buffer . DNA was eluted from the beads in 1% SDS/TE buffer and cross-links were reversed overnight at 65°C . The chromatin was then purified using a Qiagen PCR purification kit . Libraries were constructed using the Illumina Trueseq ChIP Sample Prep kit and TrueSeq standard protocol with 10ng of initial ChIP or Input DNA . Libraries that passed QC on a Bioanalyzer High Sensitivity DNA Chip ( Agilent Technologies ) were sequenced on an Illumina Miseq ( UVA Genomic Analysis and Technology Core ) . Biological duplicate fastq files were concatenated together and reads mapped to the sacCer3 genome using Bowtie with the following options:—best , —stratum , —nomaqround , and—m10 [54] . The resulting bam files were then converted into bigwig files using BEDTools [55] . As part of the pipeline , chromosome names were changed from the sacCer3 NCBI values to values readable by genomics viewers e . g . "ref|NC_001133|" to "chrI" . The raw and processed datasets used in this study have been deposited in NCBI’s GEO and are accessible through the GEO series accession number GSE92717 . Downstream GO analysis was performed as follows . MACS2 was used to call peaks with the following options:—broad , —keep-dup , -tz 150 , and -m 3 , 1000 [56] . GFP peaks in the WT or sir2Δ backgrounds were subtracted from the WT SMC4-13xMyc and sir2Δ SMC4-13xMyc peaks , respectively , using BEDTools “intersect” with the–v option . The resulting normalized peaks were annotated using BEDTools “closest” with the -t all option specified , and in combination with a yeast gene list produced from USCS genome tables . The annotated peaks were then analyzed for GO terms using YeastMine ( yeastmine . yeastgenome . org ) . Log-phase cultures were cross-linked with 3% formaldehyde for 20 min and quenched with a 2x volume of 2 . 5M Glycine . Cell pellets were washed with dH2O and stored at -80°C . Thawed cells were resuspended in 5 ml of 1X NEB2 restriction enzyme buffer ( New England Biolabs ) and poured into a pre-chilled mortar containing liquid N2 . Nitrogen grinding was performed twice as previously described [57] , and the lysates were then diluted to an OD600 of 12 in 1x NEB2 buffer . 500 μl of cell lysate was used for each Hi-C library as follows . Lysates were solubilized by the addition of 50 μl 1% SDS and incubation at 65°C for 10 min . 55 μl of 10% TritonX-100 was added to quench the SDS , followed by 10 μl of 10X NEB2 buffer and 15 μl of HindIII ( New England Biolabs , 20 U/μl ) to digest at 37°C for 2 hr . An additional 10 μl of HindIII was added for digestion overnight . The remainder of the protocol was based on previously published work with minor exceptions [58] . In short , digested chromatin ends were filled-in with Klenow fragment ( New England Biolabs ) and biotinylated dCTP at 37°C for 1 hr , then heat inactivated at 70°C for 10 min . Ligation reactions with T4 DNA ligase were performed at 16°C for a minimum of 6 hr using the entire Hi-C sample diluted into a total volume of 4 ml . Proteinase K was added and cross-links were reversed overnight at 70°C . The ligated chromatin was phenol:chloroform extracted , ethanol precipitated , then resuspended in 500μl dH2O and treated with RNAse A for 45 min . Following treatment with T4 DNA polymerase to remove biotinylated DNA ends that were unligated , the samples were concentrated with a Clean and Concentrator spin column ( Zymogen , D4013 ) and sheared to approximately 300bp with a Diagenode Bioruptor . Biotinylated fragments were captured with 30 μl pre-washed Streptavidin Dynabeads ( Invitrogen ) , then used for library preparation . Hi-C sequencing libraries were prepared with reagents from an Illumina Nextera Mate Pair Kit ( FC-132-1001 ) using the standard Illumina protocol of End Repair , A-tailing , Adapter Ligation , and 12 cycles of PCR . PCR products were size selected and purified with AMPure XP beads before sequencing with an Illumina Miseq ( UVA Genomic Analysis and Technology Core ) or Hiseq ( HudsonAlpha Institute for Biotechnology , Birmingham , AL ) . Raw and mapped reads are deposited at GEO ( GSE92717 ) . Iteratively corrected heatmaps were produced using python scripts from the Mirny lab hiclib library , http://mirnylab . bitbucket . org/hiclib/index . html . Briefly , reads were mapped using the iterative mapping program , which uses Bowtie2 to map reads and iteratively trims unmapped reads to increase the total number of mapped reads . Mapped reads were then parsed into an hdf5 python data dictionary for storage and further analysis . Mapped reads of the same strains were concatenated using the hiclib library’s “Merge" function . Both individual and concatenated mapped reads have been deposited in GEO . Mapped reads were then run through the fragment filtering program using the default parameters as follows: filterRsiteStart ( offset = 5 ) , filterDuplicates , filterLarge , filterExtreme ( cutH = 0 . 005 , cutL = 0 ) . Raw heat maps were further filtered to remove diagonal reads and iteratively corrected using the 03 heat map processing program . Finally , the iteratively corrected heatmaps were normalized for read count differences by dividing the sum of each row by the sum of the max row for a given plot , which drives all values towards 1 to make individual heatmaps comparable . Observed/Expected heatmaps were created using HOMER Hi-C analysis software on the BAM file outputs from the iterativemapping program of the hiclib library python package [35] . Tag directories were created using all experimental replicates of a given biological sample and the tbp -1 and illuminaPE options . Homer was also used to score differential chromosome interactions between the WT and mutant Hi-C heatmaps . The resulting list of differential interactions was uploaded into R where the given p-value was adjusted to a qvalue with p . adjust . An FDR cutoff of 0 . 05 was used to create a histogram of significantly different chromosome interactions in the mutants compared to WT . The histogram was further normalized by dividing the total number of significant differential interactions for a chromosome by total number of interactions called in the WT sample for that chromosome to account for size differences in the chromosomes . Thus , frequency represents the number of interactions that changed out of all possible interactions that could have changed . RNA-Seq data was acquired from GEO accessions GSE73274 [59] and GSE58319 [60] for the BY4742 ( MATα ) and BY4741 ( MATa ) backgrounds , respectively . Reads were then mapped to the sacCer3 genome using Bowtie2 with no further processing of the resulting BAM files visualized in this paper . Chromosome Conformation Capture ( 3C ) was performed in a similar manner to Hi-C with a few exceptions due to assay-specific quantification via quantitative real-time PCR rather than sequencing . Most notably , digested DNA ends were not filled in with dCTP-biotin before ligation and an un-crosslinked control library was created for each 3C library . Furthermore , all PCR reactions were normalized for starting DNA concentration using a PDC1 intergenic region that is not recognized by HindIII , in addition to PCR of the un-crosslinked control for all tested looping interaction primer pairs . Total RNA ( 1 μg ) was used for cDNA synthesis with oligo ( dT ) and Superscript II reverse transcriptase as previously described [61] . Expression levels are indicated in figures relative to the level of ACT1 mRNA , with this ratio then normalized to 1 . 0 for a specific strain or condition indicated for each experiment . Proteins were blotted using standard TCA extraction followed by SDS-PAGE as previously described [19] . Myc-tagged proteins were incubated with an anti-Myc primary antibody 9E10 ( Millipore ) at a 1:2000 dilution while tubulin was incubated with anti-Tubulin antibody B-5-1-2 ( Sigma-Aldrich ) at a 1:1500 dilution . The V5-AID tagged Brn1 was detected with anti-V5 antibody ( Invitrogen , R96025 ) at a 1:4000 dilution . Primary antibodies were detected with an anti-mouse secondary antibody conjugated to HRP ( Promega ) at 1:5000 dilution in 5% fat-free milk . Bands were then visualized with HyGlo ( Denville Scientific ) capture on autoradiography film ( Denville Scientific ) . For tracking the efficiency of switching , strains were transformed with pGAL-HO-URA3 , pre-cultured in SC-ura + raffinose ( 2% ) medium overnight , re-inoculated into the same medium ( OD600 = 0 . 05 ) and then grown into log phase . Galactose ( 2% ) was added to induce HO expression for 45 min . Glucose ( 2% ) was then added and aliquots of the cultures were harvested at indicated time points . Genomic DNA was isolated and 10 ng used for PCR amplification . MATα was detected using primers JS301 and JS302 . The SCR1 gene on chromosome V was used as a loading control ( primers JS2665 and JS2666 ) . PCR products were separated on a 1% agarose gel stained with ethidium bromide and then quantified using ImageJ . Donor preference with strains containing HMRα-B was performed as previously described [15] . Briefly , MATa was amplified with primers Yalpha105F and MATdist-4R from genomic DNA 90 after switching was completed ( 90 min ) , and then digested with BamHI . Ethidium stained bands were quantified using ImageJ . For the conditional V5-AID degron strains , degradation of V5-AID-fused Brn1 protein was induced by addition of 0 . 5 mM indole-3-acetic acid ( Auxin , Sigma # 13750 ) . For assaying HO cleavage across MATa , HML , and HMR , the WT and sir2Δ strains containing the pGAL-HO-URA3 plasmid were induced with galactose for 0 to 2 hrs following growth in raffinose . Genomic DNA was then isolated and PCR across the HO-cleavage site performed with primers specific to each locus , and SCR1 used as a loading control . MATa was detected with JS301 and JS854 , HML with JS3101 and JS3103 , and HMR with JS3097 and JS3100 . PCR was performed in the linear range and bands on ethidium stained agarose gels quantified with ImageJ . Real-time qPCR of HML and HMR with the same genomic DNA was performed with primers JS3101-JS3103 , and JS3097-JS3098 , respectively . | Sir2 is a highly conserved NAD+-dependent protein deacetylase and defining member of the sirtuin protein family . It was identified 40 years ago in the budding yeast , Saccharomyces cerevisiae , as a factor that silences the cryptic mating-type loci , HML and HMR . These heterochromatic cassettes are utilized as templates for mating-type switching , whereby a programmed DNA double-strand break at the MATa or MATα locus is repaired by gene conversion to the opposite mating-type . This directional switching is called donor preference , which in MATa cells , is driven by a cis-acting DNA element called the recombination enhancer ( RE ) . It was believed the only role for Sir2 in mating-type switching was silencing HML and HMR . However , we now show that Sir2 also regulates expression of a small gene ( RDT1 ) in the RE that is activated during mating-type switching . The promoter of this gene is also bound by the condensin complex , and deleting this region of the RE drastically changes chromosome III structure and alters donor preference . The RE therefore appears to function as a complex locus control region ( LCR ) that links transcriptional control to chromatin architecture , thus providing a new model to investigate the underlying mechanistic principles of programmed chromosome architectural dynamics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"chromosome",
"structure",
"and",
"function",
"gene",
"regulation",
"hormones",
"organisms",
"fungi",
"plant",
"science",
"model",
"organisms",
"plant",
"hormones",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"epigenetics",
"chromatin",
... | 2019 | A Sir2-regulated locus control region in the recombination enhancer of Saccharomyces cerevisiae specifies chromosome III structure |
Leprosy has a global presence; more than 180 thousand new cases were registered in 2013 , 15% of which were found in the Americas . The elderly are a very susceptible demographic in terms of developing illnesses , mainly because of characteristics natural to the senescence of the human organism . This study’s goals were to analyze leprosy in an elderly population from a hyperendemic region of the Brazilian Amazon in a historical series from 2004 to 2013 and to determine the clinical and epidemiological profile of a series of leprosy cases of elderly people in the period spanning from 2009 to 2013 . To achieve these goals , an observational , longitudinal , retrospective and descriptive study was put together to analyze leprosy in elderly people from data acquired from the Notification Aggravations Information System . Furthermore , a profile of the disease from a retrospective cohort based on data collected from medical records was developed . The number of new cases and the leprosy detection rate decreased across the observed period but remained stable among the elderly . The trend for the next ten years indicates decreases in the number of cases and in the detection rate in the general population and an increase in only the elderly . The overall profile was characterized by a predominance of males ( 64 . 32% ) , the multibacillary clinical form ( 87 . 57% ) , Type 1 reaction episodes ( 37 . 50% ) and some physical incapacity at diagnosis ( 49 . 19% ) . The risk of reaction was greater in the first six months of multidrug therapy , and the positive result from the skin smear was associated with the greater chance of reactional condition development . The resulting data demonstrate that leprosy amongst the elderly deserves attention because of the increased susceptibility to disability in this age group , with their higher risk of reaction and their greater level of co-morbidity .
Poverty is closely related to the occurrence of Neglected Tropical Diseases ( NTDs ) , and leprosy has high endemicity mainly in Africa , the Americas , Southeast Asia and the Eastern Pacific , where social and economic inequalities reflect adverse conditions of life and health for the population [1] . Brazil is number two in the world for leprosy , registering more than 30 thousand new cases a year [2] . In North Brazil , the state of Pará , a region with a low Development Human Index ( DHI ) in the country , is considered hyperendemic with 3 , 917 notified cases in 2014 alone [3] . Leprosy can affect people of all ages , including the elderly . The growth of the elderly demographic is a remarkable reality . In 2000 , there were already more people aged 60 and above than children aged below 15 in the world , and , according to projections , in 2050 , these numbers will hold[4] . When diagnosed and treated late , leprosy leads to physical disability [5] that , when combined with the aging process and other comorbidities , can cause the loss of personal autonomy to the elderly person [6] . Additionally , leprosy reaction episodes , the phenomena potentially responsible for the functional loss of the peripheral nerves [7] , also results in disability that contributes to greater vulnerability and dependency in the elderly [8] . With aging , alterations in the peripheral nervous system occur , such as a reduction in fiber myelination , decreasing nervous system conduction speed [9] and compromising the pressure and tactile senses . These alterations make the elderly person more susceptible to skin lesions that may render the leprosy diagnosis more difficult and interfere with the evaluation of these patients [10 , 11] . It is necessary for health professionals , relatives and caretakers to be attentive to leprosy symptoms in elderly people , mainly in endemic areas , such that it is necessary to perform a detailed clinical investigation to make sure that diagnosis and treatment occur as early as possible . Therefore , the objective of this study was to analyze leprosy in the elderly population from the state of Pará in a historical series spanning from 2004 to 2013 and to determine the clinical and epidemiological profile of a series of leprosy cases among the elderly in a highly endemic area inside the Amazon region in the period from 2009 to 2013 .
Since this was a retrospective cohort study with elderly people , the Human Research Ethics Committee from the State University of Pará accepted to proceed to data compilation and analysis with no previous informed consent obtained from the participants ( CAAE protocol number 41597615 . 5 . 0000 . 5174 CEP-CCBS/UEPA ) . With the objective of assuring the confidentiality of the collected data from the medical records , the people responsible for the health units surveyed signed a Consent Term for Database Usage . All clinical and epidemiological data were anonymized . This study consisted of two parts , the first being an observational , longitudinal , retrospective study by means of the complete analysis of leprosy cases diagnosed in elderly people in the state of Pará , so as to observe the trend of the disease over the timeframe during which it was observed . The second part is a clinical , epidemiological approach evaluating a retrospective cohort of leprosy cases in the elderly , so as to complement the profile observed in the evaluation of the longitudinal study . The sample of the longitudinal study was composed of new leprosy cases in the state of Pará in the historical series spanning 2004 to 2013 noted in the Notification Aggravations Information System ( SINAN ) and available at the Computing Department of the Unified Heath System website . These data were used to analyze leprosy trends and determine the detection rate in new cases of the disease per 10000 inhabitants among the general and elderly populations . Calculations of the detection rate were based on the values of the total resident population in the state of Pará taken from annual estimations of the Brazilian population elaborated by the Brazilian Institute of Geography and Statistics ( IBGE ) . The epidemiological clinical characterization was completed with data on a retrospective cohort among elderly patients diagnosed with leprosy that started and concluded treatment at the Sanitary Dermatology Specialized Reference Unit ‘Dr . Marcello Candia’ , the Tropical Medicine Center at Federal University of Pará , the Basic Health Unit of Guamá and the Health School Center of Marco in the period from 2009 to 2013 . Although the sample does not represent all of Pará State , we assume its importance in light of the study locations’ status as recognized leprosy centers in Pará State in addition to receiving patients from various locations and being themselves located in endemic areas . Though small , the sample proves enlightening as it relies on a multi-professional team equipped to treat leprosy , with each patient evaluated efficiently and each case handled adaptively . For the longitudinal study the sample was of the new leprosy cases in the state of Pará noted in the SINAN in the historical series spanning 2004 to 2013 . The epidemiological clinical characterization was of the medical records in the period from 2009 to 2013 . All patients 60 years-old or older were considered elderly , in accordance with the National Geriatric Policy of the Brazilian Ministry of Health . The types of data collected included demographics ( age group and gender ) , clinical ( operational classification , clinical form , therapeutic scheme used , disability grade at the time of diagnosis , associated comorbidities and presence of leprosy reactions ) and laboratorial ( smear skin index at the time of diagnosis and serology results for ELISA anti-phenolic glycolipid-1—anti-PGL-1 at the time of diagnosis ) . The clinical forms obeyed the Madri Classification [12] and the higher disability grade at the time of diagnosis followed the Disability Grade Classification from the World Health Organization ( WHO ) . The descriptions of the medical records were used to consider the reactional episodes made by a health professional and referring to clinical signs and symptoms typical to leprosy reactions or just to the type of reaction and were classified as mixed reaction ( types 1 and 2 ) in all the patients who presented , simultaneously or not , with reaction episodes of types 1 or 2 [13] . In the absence of indication on the medical record about the type of reaction , the described reactions were classified according to the clinical criteria found in the Directives Project from the Hansenology Brazilian Society and the Dermatology Brazilian Society [14] . Patients with unclassified reactions were the ones whose medical records lacked any report of reactional signs and symptoms or classification regarding the type of reaction were registered only as “in reaction” or “in reactional state” in the medical records , such as those who presented with only clinical signs and symptoms followed by treatment without the possibility of classifying the situation because of the absence of specific characteristics of a certain type of reaction . Only reactions that occurred during the multidrug therapy and up until 24 months after medical discharge were considered , excluding reactions present at the time of diagnosis . For longitudinal study , a the sample was composed of 50 , 094 new leprosy cases in the state of Pará noted in the SINAN , corresponding to all notification in the historical series spanning 2004 to 2013 . The calculation methods used for the presented detection rates were as follows ? 1 ) Leprosy detection rate in the general population = Number of confirmed new leprosy cases in residents / Total resident population in the given period X 100 , 000; 2 ) Leprosy detection rate in children under 15 years of age = Number of confirmed new leprosy cases in residentes under 15 years of age / Total resident population in the given period X 100 , 000; 3 ) Leprosy detection rate in the elderly = Number of confirmed new leprosy cases in resident elderly people / Total resident population in the given period X 100 , 000 . The epidemiological clinical characterization was completed with to 185 eligible medical records in the period from 2009 to 2013 . This study analyzed the leprosy trends for the next 10 years with variables that included the detection coefficients per 10 thousand inhabitants and the number of new cases among the general population and new cases among the elderly population in a ten-year period . To obtain these values , polynomial regression models for temporal series were used with modeling of third order polynomial regression and curve adjustment models . To analyze the data from the retrospective cohort study , measurements of central tendency ( arithmetic median ) and variability ( standard deviation ) were calculated . To verify intergroup differences , Chi-square tests or G Tests were used . Survival curves were generated using the Kaplan-Meier test to evaluate the occurrence of the first reactional episode according to the administered therapeutic scheme . The Odds Ratio ( OR ) was calculated between the final disclosure ( leprosy reactions during and after the treatment ) and the laboratorial exam results , with consideration of the 95% confidence interval ( IC ) . The Spearman correlation non-parametric test was used to verify the degree of association between the smear skin index and the number of reactional episodes via the Pearson correlation coefficient ( r ) . The data were analyzed with BioEstat 5 . 3 software considering a significance level of 5% ( p-value ≤ 0 . 05 ) .
The number of new leprosy cases registered among the general population of the state of Pará in the historical series spanning the years from 2004 to 2013 was 50 , 094 , and 5 , 447 of those cases included elderly individuals . There was a reduction in the number of new cases in the elderly population from 2004 to 2009 and a peak in the year 2012 . There was an increasing trend between 2010 and 2012 . The detection rate in the general population of the state of Pará was highly variable , diminishing from 9 . 61 to 4 . 89 per 10 thousand inhabitants . Among the elderly , the detection rate dropped throughout the years , with decreasing variation , and had a high of 0 . 87 and a low of 0 . 64 per 10 thousand inhabitants ( Fig in S1 Fig ) . The analysis of survival after the occurrence of leprosy reactions starting from the beginning of multidrug therapy treatment according to the operational classification and the number of doses administered to the elder individuals diagnosed with leprosy showed that the first reactional episode occurred mainly in the first six months of treatment , demonstrating that the risk for reaction was higher in the initial months of treatment and decreased progressively with time ( Fig in S2 Fig ) . Initially , a survey was made of 256 medical records from elderly patients diagnosed with leprosy in the aforementioned period . However , 62 patients who experienced interruptions in treatment at the surveyed units because of death , abandonment or transfer were excluded , which included 9 patients who had incomplete or inadequate medical records . The 185 remaining cases were followed from the start of treatment up until at least 24 months after the medical discharge . From the 185 elders surveyed , 64 . 32% were male and 69 . 73% were in the 60- to 69-year-old age group ( 67 . 50 years , on average ) . The predominant operational classification was multibacillary , and 62 . 70% of the elderly presented with a dimorphic clinical form . Among the therapeutic schemes used , 69 . 73% of the elderly went through 12 doses of multibacillary multidrug therapy ( MDT/MB ) and only 9 . 73% went through 6 doses of paucibacillary multidrug therapy ( MDT/MB ) ( Table in S1 Table ) . The occurrence of leprosy reactions in the elderly was 64 . 86% ( Table in S2 Table ) . From those reactions , 37 . 50% presented with type 1 reactions , with a predominance of the Borderline and Lepromatous clinical forms , and 65 . 83% of the patients were treated with prednisone during the episodes . Among the elderly that developed reactions , only 115 presented with reactional signs and symptoms described in the medical records , and 34 . 78% manifested new erythematous plaques , infiltrated and/or edematous , and 21 . 74% manifested signs and symptoms of neuritis in type 1 reactions ( Table in S3 Table ) . Among the 147 elderly that went through the smear skin examination at the time of diagnosis , 65% of the patients who developed reactions presented with a positive skin examination index at the time of diagnosis . The odds ratio demonstrated that individuals with a positive skin examination index at the time of diagnosis had a 6 . 07 times higher chance of developing a reaction compared to individuals with negative results ( p < 0 . 0001 ) . The chances of developing a reaction was 2 . 93 times higher in those patients with a positive ELISA anti-PGL-1 serology result performed at the time of diagnosis ( p = 0 . 1605 ) ( Table in S4 Table ) . Among the elderly , 49 . 19% already presented with some physical incapacity at the time of diagnosis ( Table in S5 Table ) . The presence of disability grade 1 or 2 at the time of diagnosis was more prevalent in the Multibacillary , Borderline and Lepromatous clinical forms . Considering the comorbidities present , systemic arterial hypertension ( 28 . 65% ) and diabetes mellitus ( 13 . 51% ) were the most prevalent ( Table in S6 Table ) .
The leprosy detection rate per 10 thousand inhabitants in the general population of the state of Pará showed a decreasing trend throughout the years . The elderly population of the state also showed a reduction in the number of cases , but the detection rate suffered a less significant drop and an approximately constant conformation . Therefore , the present study demonstrated a trend for a decrease in new cases and in the detection rate among the general population for the next ten years , and , in contrast , a trend for an increase when only the elderly population was evaluated . Such a situation among the elderly population can be explained by the higher life expectancy achieved in later decades , which resulted in a higher number of new cases diagnosed in this group . Concerning the gender of the elders in the series of cases , there was a predominance of males that corroborates the data from Monteiro et al . [15] , in which 60 . 3% of patients were males older than 60 years , and from Nobre et al . [16] , in which there were 15 . 11% more males than females . However , considering that the population studied was composed of elders and that there was a feminization of the aging process because of the higher life expectancy of women [17] , a predominance of females would be expected in this study . There was not a predominance of females , probably due to the long incubation period of leprosy , which can last up to seven years [11] . Maybe the men were infected previously and manifested the signs and symptoms only at an old age . It is also possible that a late diagnosis occurred , because the time between the appearance of signs and symptoms and the diagnosis can vary from a month to seventeen years [18] . The predominance of the aforementioned 60- to 69-year-old age group can be explained by the fact that it is the major age group among the elderly in the state of Pará and in Brazil , and , because these younger elders typically have more social contact , they are more susceptible to contracting leprosy . According to data from the Brazilian Institute of Geography and Statistics [19] , in the last ten years , the frequency of leprosy in individuals older than 60 years old in this age group was bigger than on all the other age groups . There was a predominance of the multibacillary forms over the paucibacillary forms of the disease , with a preponderance of the Borderline clinical form , which was in agreement with the study from Vieira et al . [20] that found a higher prevalence of the Borderline form in the leprosy profile of all ages , summing to 42 . 21% of the cases found . According to Miranzi , Pereira and Nunes [21] , the occurrence of multibacillary cases has a directly proportional relation to increased age . This relation could be due to the long incubation period of the disease combined with late diagnosis . Concerning the reactional episodes , the type 1 reaction was more frequent in the evaluated patients , such as in the studies of Pinto et al . [22] and Chabra et al . [23] who studied individuals from different age groups . In this reactional type , the active participation of T lymphocytes occurs , with tissue production of Th1 cytokines ( IL-2 and IFN-α ) and pro-inflammatory cytokines such as the TNF-α [24] . In the elderly , in turn , there was an increase in the number of memory T lymphocytes in relation to the naive T lymphocytes due to chronic exposure to infectious agent antigens throughout life , implicating greater cytokine production and contributing to the pro-inflammatory state in the elderly [25 , 26] , which may explain the higher occurrence of type 1 reaction in this group of individuals . The prednisone was the most used medicine to treat leprosy reactions , especially type 1 reactions , as expected , given that corticosteroids are recognized as the drug of choice in this reaction for its suppressive effect on the inflammatory process , diminishing the INF-ɣ and TNF-α pro-inflammatory cytokines , and for their importance in the recovery of neural functions in the post-reactional period [27] . Regarding treatment of type 2 reactions , there was greater use of thalidomide associated with prednisone both in the isolated reactions and the mixed ones , a result similar to the ones found by Teixeira , Silveira and França [28] and Nazario et al . [29] in research with patients from various age groups . In Brazil , thalidomide is the drug of choice for treating type 2 reactions because of its immunosuppressive effect , allowing most patients to reach full resolution of the skin lesions within seven days [30 , 13] . However , moderated and aggravated type 2 reactional episodes can occur with peripheral neuritis such that the associated use of systemic corticosteroids may be necessary [31] . Effects associated with the use of corticoids in the elderly relate primarily to comorbidities that accompany aging , such as hypertension , muscular atrophy , and osteoporosis , for example . Effects related to the use of corticoids in the elderly may be diminished by the use of prescriptions only in severe episodes , such as leprous reactions , besides gradual reduction of the dosage [32 , 33] . The analysis of survival based on the occurrence of leprosy reactions in relation to time demonstrated that the first reactional episode occurred mainly in the first six months , both in paucibacillary and multibacillary individuals . In the initial months of multidrug therapy treatment , the risk of occurrence was higher and diminished progressively throughout the months . In general , the reactional episodes appeared in the first six months of multidrug therapy in virtue of the rapid destruction of the bacilli by the medicine , which increases the risk of reactions considerably [34 , 35] . This study presented evidence of the importance of smear skin index elevation as a risk factor for the development of the reaction from the observation of a positive association between the smear skin index at the time of diagnosis and the number of reactional episodes during and after multidrug therapy . Additionally , the present study verified the higher chance of developing reactions when compared to individuals with negative results on this same test . Such findings are in agreement with the results found in the studies of Antunes et al . [36] and Brito et al . [35] and support the causal association between the bacillary load and the development of reactional states in the scientific literature . Positive ELISA anti-PGL-1 serology at the time of diagnosis did not represent a predictive factor relevant to the occurrence of reactions during and after MDT treatment in the present research . There is a possibility , however , that the studied variables did not have a positive association due to the low number of patients who had the test , given that the serology needed to investigate leprosy is not obligatory for a diagnosis in Brazil . In relation to the physical capacity evaluation , 95 . 68% of the elders were evaluated at the time of diagnosis in a way that the health units surveyed followed the recommendation of the Health Ministry , which indicate that a physical disability evaluation must be performed in at least 90% of the leprosy patients at the time of diagnosis and at the time of discharge to be considered active and of good service quality [37] . The greater prevalence of grade 1 physical incapacity among the patients with the presence of a physical disability is a piece of data that must be considered with caution because the evaluation of sensibility in the elderly can be compromised by the neurological alterations resulting from the senescence process in such a way that the altered sensibility as an outcome of leprosy can also be associated with aging itself . There may be reduction in the sense acuity in elders due to morphological alterations , size , density and location of the nociceptors in a way that , as aging progresses , more distance or touch pressure is needed for touch be perceived [38] . The more frequent comorbidities in the present study were arterial hypertension and diabetes mellitus , corroborating the data from Perry [39] , whose research about the life quality of people with leprosy from all age groups found that diabetes mellitus and systemic arterial hypertension were the most frequent comorbidities among these patients and can , combined with leprosy , contribute to the installation and aggravation of physical disability and interfere in the social and economic lives of the patients . Because the elderly constitute a population with a tendency to have health problems , in such a way that it is estimated that 80% of them suffer from at least one chronic disease , this increase in the number of chronic diseases is directly related to the higher functional incapability [40] . It would be valid that the health services that care for elders with leprosy incorporated their routine evaluation scales of functional capacity in this population . In this way , it would be possible to more completely evaluate the impact of the disease on the quality of life of these individuals , facilitating the institution of early rehabilitation . The limitations of the present study are related to its design as it is about a retrospective cohort performed from the review of medical records , and the quality and veracity of the data entries made by the medical professionals are factors that interfere with the trustworthiness of the analyzed data . In the face of research scarcity about leprosy in the elderly , it is suggested that new prospective studies be made to contribute to greater knowledge about this theme and to the creation of strategies for early diagnosis and disability prevention , seeking to decrease costs in the health system , loss of family relationships and compromises to the autonomy of the elderly .
The temporal analysis of leprosy among the elderly in the state of Pará demonstrated increasing trends for new cases and for the detection rate in the general population and a trend for an elevation in these values in the elderly population for the next ten years . Regarding the epidemiological and clinical profile , it was verified that there was a predominance of males in the 60 to 69 year-old age group and a predominance of the multibacillary operational classification . Leprosy reactions were highly prevalent , and the first reactional episode occurred most frequently in the first six months of multidrug therapy . Patients with positive smear skin at the time of diagnosis presented higher chances of developing leprosy reactions . However , positive ELISA anti-PGL-1 serology in the diagnosis was not a predictive factor relevant to the occurrence of the reactions . Prednisone was the most used medicine in the treatment of the reactional episodes . A high proportion of the elders already presented with some physical incapacity at the time of diagnosis . Systemic arterial hypertension and diabetes mellitus were the predominant comorbidities . Therefore , the leprosy amongst the elderly deserves attention because of the increased susceptibility to disability in this age group , with their higher risk of reaction and their greater level of co-morbidity . | Leprosy , despite being an ancient disease , still represents a challenge to public health systems today . There are still just a few studies about it , particularly among the elderly . It is known that they constitute a very heterogeneous group in terms of immune response to infections , alterations to the peripheral nervous system and predisposition to situations of vulnerability and functional dependency . The Amazon region is a hyperendemic region for leprosy and has been trying to address , along with the rest of Brazil , a rapid increase in the population’s life expectancy . This article surveys medical records from elderly people diagnosed with leprosy in a five-year period at the metropolitan region of Belém , state of Pará ( Brazil ) , identifying a predominance of the multibacillary forms of the disease , a high prevalence of leprosy reactions mainly during treatment with multidrug therapy , and the presence of some physical incapacity in most of the people evaluated . It is expected that this study will contribute to knowledge about the clinical and epidemiological characteristics of leprosy among the elderly and stimulate the making of new studies on the theme . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"elderly",
"tropical",
"diseases",
"geographical",
"locations",
"bacterial",
"diseases",
"age",
"groups",
"developmental",
"biology",
"physiological",
"processes",
"organism",
"developm... | 2019 | Leprosy in elderly people and the profile of a retrospective cohort in an endemic region of the Brazilian Amazon |
Seed-setting rate is a critical determinant of grain yield in rice ( Oryza sativa L . ) . Rapid and healthy pollen tube growth in the style is required for high seed-setting rate . The molecular mechanisms governing this process remain largely unknown . In this study , we isolate a dominant low seed-setting rate rice mutant , sss1-D . Cellular examination results show that pollen tube growth is blocked in about half of the mutant styles . Molecular cloning and functional assays reveals that SSS1-D encodes OsCNGC13 , a member of the cyclic nucleotide-gated channel family . OsCNGC13 is preferentially expressed in the pistils and its expression is dramatically reduced in the heterozygous plant , suggesting a haploinsufficiency nature for the dominant mutant phenotype . We show that OsCNGC13 is permeable to Ca2+ . Consistent with this , accumulation of cytoplasmic calcium concentration ( [Ca2+]cyt ) is defective in the sss1-D mutant style after pollination . Further , the sss1-D mutant has altered extracellular matrix ( ECM ) components and delayed cell death in the style transmission tract ( STT ) . Based on these results , we propose that OsCNGC13 acts as a novel maternal sporophytic factor required for stylar [Ca2+]cyt accumulation , ECM components modification and STT cell death , thus facilitating the penetration of pollen tube in the style for successful double fertilization and seed-setting in rice .
Rice ( Oryza sativa L . ) is not only the staple food for more than half of the world’s population , but also a model species for plant developmental and genetic studies [1] . Panicle numbers , grain number per panicle , grain weight and seed-setting rate constitute the major determinants of grain yield in rice [2] . During the past few decades , great progress has been made in elucidating the molecular genetic control mechanisms of panicle numbers , grain number per panicle and grain weight in rice [2] . Recently , increasing effort has been made to elucidate the genetic control of seed-setting rate in rice as low seed-setting rate of Indica-Japonica hybrid has become a bottleneck limiting further improvement of hybrid grain yield [3] . Low seed-setting rate in rice could result from spikelet sterility due to abnormal floret structures , defective pollen grain or embryo sac development , impaired anther dehiscence , gametophytic incompatibility or inappropriate temperature at the reproductive stage . Several genes underlying seed-setting rate have been characterized . For examples , Pollen semi-sterility1 encodes a kinesin-1-like protein important for male meiosis , anther dehiscence , and fertility in rice [4] . WA352 is a mitochondrial gene that confers wild abortive cytoplasmic male sterility in rice . WA352 inhibits the known nuclear-encoded mitochondrial protein COX11 functioning in peroxide metabolism , thereby triggering premature tapetal programmed cell death ( PCD ) in the anther and consequent pollen abortion [5] . The S5 locus contains three tightly linked genes that regulate indica-japonica hybrid fertility [6] . A recent study also showed that defective pollen tube growth in pistil could also cause reduced panicle fertility and seed setting [7] . Despite the tremendous progress , however , our understanding of the genetic control of seed-setting rate in rice still remains very fragmented . A critical determining factor of seed-setting rate in flowering plants is the rapid and directional pollen tube growth in the style transmission tissue ( STT ) to deliver the male gametes to the ovary of the pistils for successful double fertilization , and this process involves intensive communication between the pollen tube and the surrounding maternal sporophytic tissues [8] . In particular , pharmacological and genetic studies have shown that Ca2+ , composition of the extracellular matrix ( ECM ) and PCD of STT cells are all required for healthy growth of pollen tube in the style [9–13] . However , the functional relationships among these factors in controlling pollen tube growth and reproduction are still poorly understood . The cyclic nucleotide-gated channel proteins ( CNGCs ) were nonspecific , Ca2+-permeable cation channels and were first identified in vertebrate photoreceptors [14] . There are 20 members of CNGCs in Arabidopsis that are differentially expressed in all tissues [15] . AtCNGC2/5/6/7/8/9/10/16/18 have been demonstrated to mediate Ca2+ currents [16–19] , while AtCNGC1 and AtCNGC2 have been demonstrated to facilitate K+ or other monovalent cation fluxes [20] . A wide range of biological functions has been reported for CNGC proteins . For example , it has been reported that AtCNGC2 , AtCNGC11/12 , AtCNGC11 and AtCNGC12 play a role in PCD [21–26] . AtCNGC16 was shown to be critical for stress tolerance in pollen reproductive development and AtCNGC14 was reported to regulate root gravitropism [27 , 28] . Notably , the pollen tube tip plasma membrane-located protein AtCNGC18 was recently shown to be the long-sought essential Ca2+ channel for mediating external Ca2+ influx and pollen tube tip growth in Arabidopsis [18 , 29] . In rice , 16 CNGC members were identified , and these genes are nominated according to their phylogenetic placement [30] . However , there is no reported functional studies on these rice CNGCs yet . In this study , we isolated a dominant low seed-setting rate mutant in rice ( sss1-D ) and demonstrated that a mutation in a gene encoding cyclic nucleotide-gated channel ( OsCNGC13 ) is responsible for the mutant phenotype . We found that OsCNGC13 is preferentially expressed in the pistils , and that the mutation causes cytoplasmic calcium concentration ( [Ca2+]cyt ) loss , altered ECM components , and delayed cell death in the style . As a result , pollen tube growth is blocked in about half of the styles . Our results suggest that OsCNGC13 defines a novel maternal sporophytic factor that links stylar [Ca2+]cyt , ECM components modification and style cell death for promoting pollen tube growth and seed-setting in rice .
In an effort to dissect the mechanism underling rice fertility , we isolated a dominant low seed-setting rate rice mutant named semi-seed-setting rate1-Dominant ( sss1-D ) , by screening a 60Co-irradiated 9311 M2 population . Under normal field conditions , the seed-setting rate of the wild type ( 9311 ) was ~92% , while the mutant showed a much lower seed-setting rate ( ~51% ) ( Fig 1A and 1B ) . The floret structure , pollen fertility and pollen germination in vitro of the mutant appeared normal , compared to the wild type plants ( Fig 1C–1G ) . To determine the underlying cause of the fertility defect , we conducted reciprocal crosses between sss1-D and its wild type 9311 . The results showed that when sss1-D was used as the pollen donor , the seed-setting rate was normal; however , when sss1-D was used as the pollen receiver , varying degrees of reduced seed-setting rates ( ranging from 30% to 42% ) were observed ( S1 Table ) . This finding suggests that the low seed-setting rate of sss1-D is due to a maternal defect . Further , we found that the seed-setting rates of 9311/sss1-D F1 and sss1-D/9311 F1 plants were ~54 . 8% and ~54 . 4% , respectively; and the plants of an F2 population ( Q1 , n = 160 ) derived from sss1-D and 9311 segregated plants of normal and low seed-setting rate in a 1:3 ratio ( x2 = 0 . 075 < x20 . 05 , 1 ) ( S1 Fig ) . These results indicated that the low seed-setting rate phenotype of sss1-D is caused by a single dominant mutation . To dissect the cellular defects for the low seed-setting rate phenotype of sss1-D , we compared the behavior of pollen tube growth after pollination . At 5 min after pollination ( MAP ) , the germination rate of pollen grains and the ability of pollen tube generation in sss1-D were comparable to those of wild type ( Fig 1H–1K and S2 Fig ) . However , reduced pollen tube growth was observed in sss1-D at 30 MAP and thereafter ( Fig 1L–1O ) . At 120 MAP , about 87% wild type pistils had pollen tube tips arriving at the basal ovules and the micropyles , but only about 51% of the sss1-D pistils had pollen tube tips reaching the micropyles and about 49% of the sss1-D pistils had pollen tubes arrested in various positions in the styles ( Fig 1P–1U ) . These results suggest that the low seed-setting rate of the mutant most likely results from blockage of pollen tube growth in the style . On the other hand , these results also suggest that once the growth barrier in the styles was overcome , the sss1-D pollen tubes could reach the micropyles . To confirm this , we analyzed a large number of embryo sacs at 24 h after pollination ( HAP ) , when the double fertilization is completed in rice . In wild type , about 98% of the embryo sacs were normally fertilized , each with a multi-celled globular embryo and a layer of free endosperm nuclei . On the contrary , only about 46% of the embryo sacs were fertilized in sss1-D ( S3 Fig ) . These observations are consistent with the reduced seed-setting rate of the sss1-D mutant ( Fig 1B ) . To further confirm this , we examined pollen tube growth in the pistils of hand-pollinated reciprocal crosses . When wild type was used as the pollen receiver , pollen tubes grew normally and ultimately reached the micropyles . However , when the sss1-D mutant stigmas were sprinkled with wild type or its own pollen grains , the same abnormalities were observed as those in the self-pollinated mutant pistils ( Fig 1V ) . Taken together , our results demonstrated that the blocking of pollen tube growth in the styles and subsequent failure in double fertilization might be the main cause of low seed-setting rate in the sss1-D mutant . To test the female transmission efficiency , we analyzed the genotype of each individual plant in another F2 population ( Q2 ) derived from sss1-D and 9311 ( n = 234 ) . The result showed a 1:2:1 segregation ratio ( x2 = 1 . 385 < x20 . 05 , 2 ) of homozygous wild type plants ( 59 ) : heterozygous plants ( 124 ) : homozygous sss1-D mutant plants ( 51 ) , suggesting that both the wild type and mutant female gametophytes could be efficiently transmitted . These observations together suggest that the reduced seed-setting rate of the mutant is due to a defect in the female sporophytic tissue . A map-based cloning strategy was used to isolate the target gene locus using an F2 mapping population derived from the cross between sss1-D and the indica cultivar N22 . The target gene locus was initially placed in an ~300-kb interval between the markers Q-17 and Q-7 on the short arm of rice chromosome 6 and was further restricted to a 52-kb genomic region flanked by the markers Q-20 and Y-80 ( Fig 2A ) . Three putative open reading frames ( ORF1-3 ) were predicted in this mapping region . ORF1 ( LOC_Os06g10580 ) is predicted to encode the rice Cyclic Nucleotide-Gated Channel 13 ( OsCNGC13 ) with a pore-forming region and CNBD domain at the C-terminal region [15 , 30] . ORF2 ( LOC_Os06g10590 ) is predicted to encode a putative uncharacterized expressed protein , and ORF3 ( LOC_Os06g10600 ) is predicted to encode a putative homeobox and START domain containing protein ( http://www . tiger . org ) . Sequence analysis revealed that an ~44-kb genomic DNA segment was inverted and two ORFs ( ORF1 and ORF3 ) were interrupted in sss1-D , resulting in disruption of ORF3 and a mutated ORF1 ( mORF1 ) . ORF2 remained unchanged ( Fig 2B and 2C ) . The inversion causes mORF1 become a chimeric gene that is prematurely terminated ( Fig 2D ) . Quantitative real-time reverse transcription ( qRT ) -PCR assay showed that the transcript levels of ORF1/mORF1 and ORF2 between wild type and sss1-D were comparable , while the expression of ORF3 was dramatically reduced in sss1-D ( Fig 2E ) . To test whether the alterations in ORF1 or ORF3 might be responsible for the mutant phenotype , we performed a genetic complementation assay . As the sss1-D mutant ( in an indica background ) was difficult for transformation , we individually transformed pORF1::ORF1 , 35S::ORF1-GFP , gORF3 , and 35S::ORF3-GFP into W109 , a rice line derived from the N22/sss1-D F3 population and is homozygous at the sss1-D locus in the N22 genetic background ( Fig 3A ) . The low seed-setting rate of W109 was largely rescued when ORF1 was expressed ( either driven by its endogenous promoter or by the CaMV 35S promoter ) ; while the expression of ORF3 ( driven by its endogenous promoter or the CaMV 35S promoter ) showed no effect ( Fig 3B–3F ) . In addition , knockout of ORF1 by CRISPR/Cas9 in the wild type Kitaake caused much reduced seed-setting rates ( ranging from 37 . 8 ± 10 . 9% to 51 . 9 ± 5 . 7% ) and much decreased percentages of the ovules with pollen tube ( ranging from 38 . 7 ± 5 . 1% to 50 . 0 ± 2 . 7% ) in all positive transgenic lines , while knockout of ORF3 showed no effect on the seed-setting rate ( Fig 3G and 3H ) . These results suggest that ORF1/OsCNGC13 represents the target gene . Since the low seed-setting rate phenotype of sss1-D mutant was dominant , we reasoned that it might be caused by a dominant-negative effect of the mORF1 protein . Compared with the normal OsCNGC13 , mORF1 is predicted to encode a product ( OsCNGC13-D ) with only 440 amino acids with five transmembrane domains at the N-terminus . The C-terminal pore-forming region and CNBD domain were lost due to premature termination ( S4A and S4B Fig ) . To test whether the mutated OsCNGC13-D protein may interfere with the normal function of OsCNGC13 , we first examined whether OsCNGC13 can form homodimer and whether OsCNGC13-D can interact with OsCNGC13 . Both yeast two-hybrid assay and bimolecular fluorescence complementation ( BiFC ) assay failed to detect homomeric or heteromeric interaction between OsCNGC13-D and OsCNGC13 ( S5 and S6 Figs ) , indicating that it is unlikely OsCNGC13-D negatively affects OsCNGC13 protein function directly . Next , we compared the OsCNGC13 transcript levels in the wild type 9311 , 9311/sss1-D F1 plants ( heterozygous plants ) and sss1-D homozygous mutant plants . The result showed that , compared with the wild type or the homozygous mutant , the transcript level of OsCNGC13 in the heterozygous plants was reduced ( Fig 4A ) . qRT-PCR analysis using allele-specific primer pair also showed that expression of the wild type OsCNGC13 allele , but not the mutant OsCNGC13-D allele , was dramatically reduced in the heterozygous plants ( Fig 4B and 4C ) . Moreover , we artificially expressed OsCNGC13-D in wild type Kitaake . Positive transgenic plants showed reduced seed-setting rates and decreased percentages of ovules with pollen tube ( Fig 4D–4F ) , which is similar to the mutant phenotypes of sss1-D . More importantly , we found that the expression of OsCNGC13 allele was significantly reduced in the transgenic lines ( Fig 4G ) . Furthermore , we observed a correlation between the expression levels of OsCNGC13 , percentages of ovules with pollen tube , and the reduced seed-setting rates in the OsCNGC13 RNAi lines ( Fig 4H and 4I and S2 Table ) . These observations together suggest that the dominant nature of sss1-D mutation might be caused by haploinsufficency resulting from the suppression of OsCNGC13 by the mutated OsCNGC13-D allele . Search of the microarray expression data from the RiceXpro database ( http://ricexpro . dna . affrc . go . jp/ ) revealed that OsCNGC13 is expressed in various tissues with peak expression in the pistils ( S7A Fig ) . To examine its expression in more detail , we generated pOsCNGC13-GUS reporter gene transgenic plants . Histochemical staining detected expression in the vascular tissues of the primary root , leaf , leaf sheath , spikelet hulls ( S7B–S7H Fig ) , the stigmas and styles of the mature pistils , and the pistils at 30 MAP ( Fig 5A and 5B ) . Further , in situ hybridization confirmed preferential expression of OsCNGC13 in the floral primodia , the stigmas and styles of mature pistils before and after pollination ( Fig 5C–5L ) . To determine the subcellular localization of the OsCNGC13 protein , we transiently expressed the OsCNGC13-GFP fusion protein in rice protoplasts . The result showed that the control GFP signal was dispersed in the cytosol , while the green fluorescence of OsCNGC13-GFP merged well with red signal of the plasma membrane marker , PIP2;1-mCherry [31] , indicating that OsCNGC13 was located to the plasma membrane ( Fig 5M–5P ) . Similar result was also found in the tobacco leaf epidermal cells and mesophyll protoplasts transiently expressing the OsCNGC13-GFP construct ( S8 Fig ) . Consistent with these observations , the OsCNGC13-GFP fusion protein was also detected in the cell perimeter in the 35S::OsCNGC13-GFP transgenic plants ( Fig 5Q–5T ) . Together , these results indicate that OsCNGC13 is a plasma membrane localized protein . Multiple Arabidopsis CNGCs have been demonstrated to mediate the Ca2+ currents virtually [18 , 28 , 32 , 33] . To determine whether OsCNGC13 had permeability to Ca2+ , we conducted patch-clamp whole-cell recording in HEK293 cells to measure the OsCNGC13-mediated Ca2+ currents . Remarkable whole-cell inward Ca2+ currents were detected when OsCNGC13 , but not the negative control GFP or the OsCNGC13-D protein , was transfected into HEK293 cells ( Fig 6A and 6B ) . On the other hand , OsCNGC13 had barely detectable permeability to K+ compared with the positive control OsAKT1 [34] ( S9 Fig ) . Previous studies reported that application of compatible pollens to the stigma papilla cells could trigger an increase in cytosolic free calcium [35 , 36] . Thus , we speculated whether a similar response might happen in the style when pollinated . Scanning electron microscopy-energy dispersive x-ray spectrometry ( SEM-EDX ) of the cross-sections of the styles revealed that sss1-D mutant had continuous low Ca2+ concentrations in the style , whereas the wild type showed a maximum net value of ~9 . 2% at 5 MAP at the bottom of the styles , and this high Ca2+ concentration continued till 30 MAP ( Fig 6C ) . Staining with Fluo-3 acetoxymethyl ( AM ) ester , a Ca2+-sensitive fluorescent dye [37] , showed that in the styles of mature pistils before flowering , the [Ca2+]cyt maintained at a relative low level in both the wild type and the sss1-D mutant ( Fig 6D–6G ) . After pollination , the fluorescence intensity in the wild type styles increased gradually , while in the mutant , only background signal could be detected and the fluorescence intensity remained at a low level ( Fig 6F and 6G ) . Similar results were observed when yellow cameleon 3 . 6 ( YC3 . 6 ) was used as a ratiometric fluorescent calcium indicator [38] and when potassium pyroantimonate was used to precipitate the intracellular calcium [39] ( Fig 6H and 6I and S10 Fig ) , respectively . These results together demonstrate that the OsCNGC13 is indispensable for triggering [Ca2+]cyt accumulation in the style after pollination . In angiosperms , specialized cell files in the style , which form the STT , define the paths of pollen tubes to approach the ovary and produce a complex mixture of polysaccharides , glycoproteins , and glycolipids that is known as the ECM . The ECM has been reported to provide guidance signals ( i . e . , Ca2+ flux capacitor ) as well as nutrients for pollen tube growth [40–42] . Staining with H33342 , a nucleic acid fluorescent dye [43] , showed no difference in the overall cellular structure of the style tissue between the wild type and the mutant ( S11A–S11D Fig ) . Alcian blue , a stain for acidic polyanions [11] , was used to indicate the acidic polysaccharides ( major components of the ECM ) . The result showed that after pollination , most tissues and cells ( except for the vascular tissues and epidermal cells ) in the sections of wild type could be stained deeply , while the staining was weaker in the corresponding regions of the mutant styles ( Fig 7A–7D ) . Moreover , weak staining was also observed in the mature pistils before flowering ( S11E–11L Fig ) . These results demonstrate that in sss1-D , the female reproductive tract is defective in ECM components . Previous studies had also shown that pollination triggered calcium signaling could lead to modification of ECM components and the death and degeneration of transmission tract cells , allowing the pollen tube to penetrate the style and reach the ovules for fertilization [10 , 11 , 44 , 45] . Both plastic section observation and the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling ( TUNEL ) assay showed no cell degradation in the styles of wild type or sss1-D before flowering ( S11M–11T Fig ) . Cytological examination revealed that at 30 MAP , the wild type style had intense staining and notable intercellular space , indicating that some style cells had broken down after pollination , while in the sss1-D mutant , all the cells remained intact ( Fig 7A–7H ) . To confirm this , we performed transmission electron microscopy ( TEM ) analysis of the transverse sections of the wild type and mutant styles . Some bottom style cells in the wild type pistil lacked contents inside , whereas all the cells had well defined organelles and abundant cytosolic contents in the mutant style ( Fig 7I and 7J ) . Consistent with the above results , the TUNEL assay showed that positive signals were detected in the wild type style , but not in the mutant style ( Fig 7K–7P ) . Further , qRT-PCR analysis revealed that the expression of several genes related to PCD [46–48] was elevated in the wild type pistils at 30 MAP , but not in the sss1-D mutant pistils ( Fig 7Q ) . All these results indicate that STT cell death is defective in the sss1-D mutant after pollination .
In this study , we showed that sss1-D represents a novel dominant low seed-setting rate mutant in rice . Our cytological studies suggest that the low seed-setting rate of sss1-D is caused by blockage of pollen tube growth in about half of the pistils ( Fig 1H–1V ) . Further , our reciprocal cross assays , together with transmission analysis , pollen fertility assay , and cytological observations , demonstrate that the blocking of pollen tube growth in sss1-D styles is caused by a maternal defect . Similar cases have been reported before . PTB1 , which encodes a RING type E3 ubiquitin ligase in rice , positively regulates panicle seed-setting rate and grain yield by promoting pollen tube growth in the STT [7] . Differently , ptb1 exhibits extreme sterility because all the pollen tubes are blocked in the STT . While the seed-setting rate of the sss1-D mutant is ~51% , and pollen tube is blocked in ~51% of the mutant pistils at 120 MAP . Our molecular cloning and functional complementation assay revealed that a genomic inversion in OsCNGC13 forming OsCNGC13-D is responsible for the mutant phenotype . We show that OsCNGC13 possesses all the classical domains of the CNGC family , and is localized to the plasma membrane , similar to AtCNGC2 , AtCNGC12 , AtCNGC14 , AtCNGC17 and AtCNGC18 [49] . However , it should be noted that some CNGCs have been localized to other subcellular localizations . For example , three Medicago truncatula CNGC proteins ( MtCNGC15a/b/c ) were found to be localized to the nuclear membrane [33] . Surprisingly , AtCNGC19 and AtCNGC20 , the closest homologs of OsCNGC13 in Arabidopsis [30] ( S12 Fig ) , were shown to be localized in vacuole membrane in one study [50] , while AtCNGC20 was localized to the plasma membrane in another study [51] . The distinct subcellular localizations of CNGC proteins might be consistent with their broad range of biological functions . In general , CNGC have been shown to form tetrameric channels in animals [52] . Similarly , AtCNGC11/12 , AtCNGC2 and AtCNGC4 have been shown likely to form homotetramer and heterotetramer as well [53 , 54] . Recently , AtCNGC17 was shown to form homodimer or oligomer and can interact with AHA1 , AHA2 , and BAK1 to regulate Arabidopsis growth , suggesting that AtCNGC17 functions in a protein complex ( es ) [55] . Somewhat unexpectedly , our Y2H ( S5 Fig ) and BiFC ( S6 Fig ) experiments failed to detect direct homomeric or heteromeric interaction between OsCNGC13-D and OsCNGC13 , suggesting that OsCNGC13-D is unlikely to affect the function of OsCNGC13 directly . However , whether OsCNGC13 can form heteromultimer with other members of the rice CNGC family in vivo remains to be investigated in future studies . Notably , we found that the mutated allele , OsCNGC13-D , causes a significant reduction in the expression level of the wild type OsCNGC13 allele in the heterozygous background and in wild type plants expressing OsCNGC13-D driven by its native promoter ( Fig 4A–4G ) . Therefore , we deduced that the dominant nature of sss1-D mutant is likely caused by haploinsufficiency [56 , 57] . The precise mechanism of this regulation is currently not clear . It is possible the OsCNGC13-D allele might become a paramutagenic allele to alter the expression of the other allele at the same genetic locus in a heritable manner [58] . Further studies are required to resolve this issue . We also show that OsCNGC13 can mediate inward Ca2+ current ( Fig 6A and 6B ) . Consistent with being a maternal sporophytic factor required for normal pollen tube elongation in the styles , we found that OsCNGC13 is preferentially expressed in the stigmas and styles ( Fig 5A–5L ) . The sss1-D mutant is defective in triggering [Ca2+]cyt accumulation in the style after pollination ( Fig 6C–6I and S10 Fig ) . These results suggest that OsCNGC13 is likely required for Ca2+ influx in STT cells . Interestingly , we also observed altered ECM components and delayed STT cell death in the sss1-D mutant pistils ( Fig 7A–7Q ) . These observations are consistent with the earlier findings that characteristic calcium signaling in the principal sporophytic cells is essential for successful double fertilization [59 , 60] and that pollination triggered calcium signaling is functionally linked with ECM production/modification and PCD of STT cells to facilitate pollen tube growth in the stylar tissues [40 , 41 , 61] . In support of this notion , several AtCNGCs have been shown to be involved in the process of PCD . For example , AtCNGC2 was shown to possess a Ca2+ influx channel activity that can mediate Ca2+ influx into leaf cells [19 , 22] . The leaves of null mutant of AtCNGC2 , defense , no death ( dnd1 ) , display delayed PCD when challenged with pathogens [22] . Null mutants of AtCNGC4 ( the closest paralog of AtCNGC2 ) exhibit remarkably similar phenotypes to dnd1 [62 , 63] , supporting a notion that AtCNGC2 and AtCNGC4 may form a heteromeric Ca2+ channel complex [54] . Similarly , several recent studies also documented evidence for a role of AtCNGC11 and AtCNGC12 in PCD and plant immunity [21 , 23 , 24 , 26] . Thus , we speculated that the delayed cell death in the style caused by the OsCNGC13 defect might ultimately explain the blockage of pollen tube growth in the sss1-D mutant . Similar observations have been made with the Arabidopsis ntt ( no transmitting tract ) mutant , which has reduced fertility due to defects in ECM production and formation of the ovary transmitting tract , and consequently blockage of pollen tube growth [11] . Although the mechanistic details need to be further elucidated , our results suggest that OsCNGC13 play an important role in linking [Ca2+]cyt accumulation , ECM components modification , and PCD in the style after pollination to allow proper pollen tube growth and successful double fertilization ( Fig 7R and 7S ) , which ultimately affects seed-setting rate in rice .
The sss1-D mutant was identified from a 60Co-irradiated M2 population of the indica rice ( Oryza sativa L . ) cultivar 9311 . The sss1-D mutant was crossed with 9311 to produce two F2 populations ( Q1 and Q2 ) in 2012 and 2013 , respectively , which were adopted for genetic and gametophyte transmission analyses . An F2 mapping population was generated from a cross between the sss1-D mutant and the indica cultivar N22 . All plants were grown in paddy fields during the normal growing seasons or in a greenhouse at the Chinese Academy of Agricultural Sciences , in Beijing . More than 30 pollinated pistils of the mutant and wild type were collected and fixed in FAA ( containing an 18:1:1 ( by vol . ) mixture of 70% ethanol , formalin and acetic acid ) for 24 h . The spikelets were then processed through an ethanol series ( 70 , 50 , and 30% ) and washed three times with distilled water . The spikelets were incubated in 10 mol L-1 sodium hydroxide for 8 min at 56°C and then washed with distilled water three times and stained in 0 . 1% aniline blue solution for 12 h . Finally , the samples were examined using a scanning confocal microscope ( ZEISS LSM 700 ) . Embryo sac development analysis was performed as described previously [43] . To map the target gene locus , 1610 mutant individuals were collected from the F2 population of sss1-D and N22 . The selected mutant plants and newly developed molecular markers were used for primary and fine mapping . Inversion in the mutant was confirmed using PCR with the primer pairs 1F/1R , 2F/2R , 3F/3R , and 4F/4R . All primer sequences used for the map-based cloning are listed in S3 Table . Total RNA was extracted from fresh samples using a plant RNA extraction kit ( Tiangen Co . , Beijing , China ) according to the manufacturer’s instructions . 1 μg of total RNA of each sample was reverse transcribed to cDNA using a reverse transcription kit ( SuperScript II; TaKaRa ) . qRT-PCR was performed using a SYBR Premix Ex TaqTM kit ( TaKaRa ) on an ABI prism 7500 Real-Time PCR System . The rice ubiquitin gene ( Os03g0234200 ) was used as a reference gene with the primer pair Ubq . Each RNA sample was extracted from a pool of tissues collected from at least three individual plants . The 3’ RACE was performed using the SMARTer RACE cDNA amplification kit ( Clontech Laboratories ) , according to the manufacturer’s instructions . The first-strand cDNA was synthesized from total RNA of 7-d-old sss1-D mutant seedlings . The sequence of the obtained PCR fragment was used as a template for direct sequencing . All primer sequences used here are listed in S3 Table . For genetic complementation , a 2 . 9-kb genomic fragment upstream of the ATG start codon of ORF1 was amplified by PCR using the wild type genomic DNA as the template , and then the full-length CDS of ORF1 was amplified from the wild type cDNA . Both fragments were then cloned into the pCAMBIA1305 vector to generate the pORF1::ORF1 construct . A 6 . 7-kb ORF3 genomic fragment spanning the entire coding region , 2000-bp upstream sequence , and 500-bp downstream sequence was amplified and recombined into the pCAMBIA1305 vector to generate the gORF3 construct . The full-length CDSs of ORF1 and ORF3 were cloned into the pCAMBIA1305-GFP vector ( generated by insertion of a SacI-SalI fragment containing the GFP expression cassette of pAN580 into the pCAMBIA1305 vector ) to generate the 35S::ORF1-GFP and 35S::ORF3-GFP constructs , respectively . Subsequently , the plasmids pORF1::ORF1 , gORF3 , 35S::ORF1-GFP , and 35S::ORF3-GFP were individually introduced into the Agrobacterium tumefaciens strain EHA105 and used to infect the calli of W109 . Transgenic roots used for subcellular localization were analyzed by scanning confocal microscope ( ZEISS LSM 700 ) . For knockout lines , one 20-bp gene-specific sequence for ORF1 and ORF3 was synthesized and cloned into the entry vector pOs-sgRNA , and then cloned into the gateway destination vector pOs-Cas9 , respectively . The resulting plasmids were individually introduced into Kitaake . Positive lines were confirmed by PCR followed by sequencing . For expression of mORF1 in wild type , a 2 . 9-kb promoter fragment of ORF1 was amplified from the wild type , and then 1 . 3-kb full-length mORF1 was amplified from the sss1-D cDNA . Both fragments were cloned into the pCAMBIA1305 vector to generate the pORF1::mORF1 construct , which was introduced into Kitaake by Agrobacterium-mediated transformation . The construct pCUbi1390-ΔFAD2 ( ubiquitin promoter and a FAD2 intron inserted into pCAMBIA1390 ) was used as the RNAi vector [64] . Both sense and anti-sense versions of a specific 500-bp fragment from the coding region of the ORF1 were amplified and successively inserted into pCUbi1390-ΔFAD2 , to form the RNAi construct pUbi-RNAiORF1 , which was introduced into Kitaake . For expression pattern analysis , the 2 . 9-kb OsCNGC13 promoter fragment was amplified by PCR and cloned into the binary vector pCAMBIA1305 to generate the pOsCNGC13-GUS construct . The construct was introduced into Kitaake . Samples of transgenic lines were observed with a stereo microscope ( Leica MZ16 ) , and photographed ( Leica DFC490 ) after GUS histochemical staining . The primers and sequences used for constructing these vectors are listed in S3 Table . Gene prediction was performed using the Rice Genome Automated Annotation System ( http://ricegaas . dna . affrc . go . jp/ ) . Sequence analysis of OsCNGC13 and OsCNGC13-D was performed using the SOSUI program ( http://bp . nuap . nagoyau . ac . jp/sosui/ ) . Amino acid sequences of OsCNGC13 and OsCNGC13-D were used for alignment using the ClustalX 2 . 01 program with default settings . The BioEdit software was also used to perform multiple sequence alignments to confirm the ClustalX data output . The DUALhunter system ( Dualsystems Biotech ) was used for yeast two-hybrid assay . The coding sequences of OsCNGC13 , OsCNGC13-D and OsCNGC12 were amplified and fused to the Cub-LexA-VP16 fragment in the pDHB1 and pXGY18 vector and the Nub fragment in the pPR3-N and pXGY17 vector , respectively . Test constructs were transformed into the yeast strain NMY51 according to the manufacturer’s instructions . The growth state of each transformant was examined on the QDO ( SD/-Trp/-Leu/-His/-Ade ) medium . For the BiFC assay , the coding sequences of OsCNGC13 , OsCNGC13-D and OsCNGC12 were cloned into the binary BiFC vectors pSPYNE173 and pSPYCE ( M ) , respectively . AtCNGC2 was co-expressed with AtCNGC4 as the positive control [54] . For transient expression , the Agrobacterium tumefaciens strain EHA105 carrying the BiFC constructs were co-infiltrated into N . benthamiana leaves with the p19 strain and the PM marker , PIP2;1-mCherry fusion protein [31] . Infiltrated leaves were observed 48–72 h after infiltration using a laser scanning confocal microscope ( ZEISS LSM 700 ) . The eYFP and mCherry fluorescent signals from the expressed fusion constructs were monitored sequentially . The excitation and detection wavelengths for eYFP and mCherry are 514 and 587 nm for excitation , and 527 and 610 nm for detection , respectively . All primer sequences used for plasmid construction are listed in S3 Table . Young inflorescences at 2 weeks before flowering and spikelets at various stages were fixed using RNase-free FAA for 12 h at 4°C , dehydrated through an ethanol series , and then embedded in paraffin ( Paraplast Plus , Sigma ) . A 252-bp OsCNGC13 region was amplified and subcloned into the pGEM–T Easy vector ( Promega ) , which was used as a template to generate both the antisense and sense RNA probes . DIG Northern starter kit ( Roche ) was used to prepare the digoxigenin-labeled RNA probes . Slides were observed under light microscopy ( Leica DM5000B ) , and photographed using a Micro Color charge-coupled device camera ( Apogee Instruments ) . All primer sequences used here are listed in S3 Table . The full-length CDS of ORF1 was amplified from the wild-type cDNA and cloned into the pA7-GFP vector to form translational fusion with the N-terminus of the green fluorescent protein ( GFP ) . As a positive marker , the cDNA of a previously characterized plasma membrane protein , PIP2;1 [31] , was fused to the mCherry gene to generate 35S::PIP2;1-mCherry . Both constructs were co-transformed into rice protoplasts and incubated in the dark at 28°C for 16 h before examination . Meanwhile , the binary vectors 35S::ORF1-GFP and pCAMBIA1305-GFP were introduced into N . benthamiana leaves , respectively . Confocal imaging analysis was performed using a laser scanning confocal microscope ( ZEISS LSM 700 ) . All primer sequences used here are listed in S3 Table . The patch-clamping recordings from HEK293 cells were performed as described previously [34] HEK293 cells ( ATCC ) were cultured in DMEM ( Dulbecco’s modified eagle medium ) with 4500 mg L21 glucose ( Gibco ) and 10% fetal calf serum ( Gibco ) for 24 h at 37°C , 5% CO2 . OsCNGC13 and OsCNGC13-D were cloned into the pCI-neo vector to generate the pCI-neo-OsCNGC13 and pCI-neo-OsCNGC13-D constructs , respectively ( see S3 Table for primers and cloning sites ) . Then the pCI-neo-OsCNGC13 and pCI-neo-OsCNGC13-D plasmid were individually co-transfected with pEGFP-N1 into HEK293 cells using the Lipofectamine 2000 Transfection Reagent ( Invitrogen ) . Then , the transfected cells were treated with Trypsin ( Gibco ) , centrifuged at 160g for 5 min , and kept on ice for patch-clamp recording . The cells with GFP fluorescence were selected for whole-cell recording . The method for K+ current patch-clamp whole-cell recording was described previously [34] . For Ca2+ current recordings , standard whole-cell recording techniques were applied [18] . The components of the standard Ca2+ pipette solution were 8 mM NaCl , 120 mM CsCl , 6 . 7 mM EGTA , 3 . 35 mM CaCl2 , 3 mM MgCl2 , 10 mM Hepes , 2 . 5 mM Mg-ATP added freshly and D-sorbito ( л = 350 mosmol kg-1 ) , pH 7 . 2 adjusted with NaOH . The standard bath solution for Ca2+ current recordings contains 120 mM NaCl , 10 mM CsCl , 2 mM MgCl2 , 10 mM CaCl2 , 10 mM Hepes and D-sorbito ( л = 350 mosmol kg-1 ) , pH 7 . 2 adjusted with NaOH . The patch-clamp recordings were conducted at about 20°C in dim light . Whole-cell currents were recorded using an Axopatch 200B amplifier ( Axon Instruments ) . Samples were collected and rapidly frozen in liquid nitrogen and vacuum freeze dried at -80°C for 7 d . Hand-cut sections of the freeze-dried samples were gold coated in a high-vacuum sputter coater and analyzed using a Hitachi S-3400N scanning electron microscope equipped with an energy dispersive x-ray spectrometer ( EX-250; Horiba ) that was interfaced with the IXRF system under the following conditions: accelerating voltage , 10 kV; takeoff angle , 35°C; collecting time of x-ray counts , 50 s; working distance between sample and detector , 15 mm . Probe measurements of pistils were made with a broad electron beam covering the whole cross-section . The total amounts of C , O , Na+ , Mg2+ and Ca2+ were measured , and the relative amounts of Ca2+ were expressed as the weight fraction ( Wt . % ) . For Fluo-3/AM staining , fresh pistils were excised with a razor and immediately placed in 10 mM MES buffer or 10 mM MES buffer containing 20 μM Fluo-3/AM ( Molecular Probe , USA ) . After 5 h incubation at 4°C in the dark , the samples were washed with fresh Fluo-3/AM-free MES buffer before microscopic examination . Fluorescence from the pistils loaded with Fluo-3/AM was detected at 488 nm ( excitation ) and 520 nm ( detection ) under a laser scanning confocal microscope ( ZEISS LSM 700 ) . For YC3 . 6 fluorescence observation , the UBQ10::YC3 . 6 transgenic plants [38] were crossed with the wild type 9311 and the sss1-D mutant , respectively . The pistils of the wild type plants and the homozygous mutants in the F2 populations were excised with a razor and observed directly at 458 nm ( excitation ) and 525 nm ( detection ) under a laser scanning confocal microscope ( ZEISS LSM 700 ) . ZEN microscope and imaging software were used to measure the mean fluorescence intensity . Samples were fixed in a fixative of 2 . 5% glutaraldehyde / 2% potassium pyroantimonate ( pH 7 . 6 ) buffered in 100 mmol/L phosphate buffer for 12 h at 4°C . Then the tissues were washed in 2% potassium pyroantimonate buffered in 100mmol/L phosphate buffer for 2 h , post fixed in 1% osmium tetraoxide / 2% potassium pyroantimonate ( pH 7 . 6 ) at 4°C overnight . The postfixed tissues were washed with distilled water for 3 times with 30 min each time , followed by dehydration in a graded ethanol series and embedded in acrylic resin ( London Resin Company ) . Ultrathin sections ( 70 nm ) were double stained with 2% ( w/v ) uranyl acetate and 2 . 6% ( w/v ) lead citrate aqueous solution and examined with a transmission electron microscopy ( HT7700; Hitachi ) . The pistils were collected and fixed in FAA for 24 h at 4°C , dehydrated through an ethanol series , and then embedded in paraffin . For Alcian blue staining , tissue cross-sections ( 7 μM thickness ) were cut and hydrated with an ethanol series and stained with Alcian blue and 1% Nuclear Fast Red [11] before microscopic examination ( Leica DM5000B ) and photographing . For the TUNEL assay , sections were cut and hydrated and treated with proteinase K in PBS . The TUNEL assay with a Dead End Fluorometric TUNEL Kit ( Promega ) was performed following the manufacturer’s instructions . The green fluorescence of fluorescein ( TUNEL signal ) and red fluorescence of propidium iodide were analyzed at 488 nm ( excitation ) and 520 nm ( detection ) , and 488 nm ( excitation ) and 610 nm ( detection ) , respectively , under a laser scanning confocal microscope ( ZEISS LSM 700 ) . For plastic sections , the samples were fixed in 2 . 5% glutaraldehyde in a phosphate buffer for 24 h at 4°C , and dehydrated through an ethanol series; the samples were embedded in Technovit 7100 resin ( Hereaus Kulzer ) and polymerized at 45°C . Transverse sections of 1 μm were cut and stained with 0 . 1% toluidine blue O ( Chroma Gesellshaft Shaud ) before observation ( Leica DM5000B ) and photographing ( Apogee Instruments ) . The locus names ( RAP database ) for OsCNGCs are: LOC_Os02g15580 , OsCNGC1; LOC_Os06g33570 , OsCNGC2; LOC_Os06g33610 , OsCNGC3; LOC_Os03g44440 , OsCNGC4; LOC_Os12g28260 , OsCNGC5; LOC_Os04g55080 , OsCNGC6; LOC_Os02g41710 , OsCNGC7; LOC_Os12g06570 , OsCNGC8; LOC_Os09g38580 , OsCNGC9; LOC_Os02g54760 , OsCNGC10; LOC_Os06g08850 , OsCNGC11; LOC_Os02g53340 , OsCNGC12; LOC_Os06g10580 , OsCNGC13; LOC_Os03g55100 , OsCNGC14; LOC_Os01g57370 , OsCNGC15; LOC_Os05g42250 , OsCNGC16 . The locus names ( TAIR database ) for AtCNGCs are: AT5G53130 , AtCNGC1; AT5G15410 , AtCNGC2; AT2G46430 , AtCNGC3; AT5G54250 , AtCNGC4; AT5G57940 , AtCNGC5; AT2G23980 , AtCNGC6; AT1G15990 , AtCNGC7; AT1G19780 , AtCNGC8; AT4G30560 , AtCNGC9; AT1G01340 , AtCNGC10; AT2G46440 , AtCNGC11; AT2G46450 , AtCNGC12; AT4G01010 , AtCNGC13; AT2G24610 , AtCNGC14; AT2G28260 , AtCNGC15; AT3G48010 , AtCNGC16; AT4G30360 , AtCNGC17; AT5G14870 , AtCNGC18; AT3G17690 , AtCNGC19; AT3G17700 , AtCNGC20 . The locus names ( RAP database ) for the other genes in this article are: LOC_Os01g47530 , OsMPK20-4; LOC_Os03g12500 , OsAOS2; LOC_Os02g43010 , OsVPE3 . | Rice is not only the staple food for more than half of the world’s population , but also a model species for plant developmental and genetic studies . After pollination , rice pollen grains adhere and hydrate at the surface of stigmatic papilla cells . Then , the germinated pollen tubes invade the stigma and navigate through the style transmission tract to reach the micropyle of the embryo sac for fertilization . During this long and arduous process , pollen tube requires abundant communication with the surrounding sporophytic maternal tissues . However , how the growth of pollen tube is regulated by maternal tissue remains largely elusive . This work identifies a typical cyclic nucleotide-gated channel protein in rice , OsCNGC13 , which can mediate Ca2+ inward current . Our results suggest that OsCNGC13 acts as a novel maternal sporophytic factor required for stylar [Ca2+]cyt accumulation , extracellular matrix components modification and style cell death , thus facilitating the penetration of pollen tube in the style for successful double fertilization and seed-setting in rice . These findings provide new insights into the molecular genetic control mechanisms of seed-setting rate/grain yield in rice and expand our knowledge on the cyclic nucleotide-gated channel proteins in plant sexual reproduction . | [
"Abstract",
"Introduction",
"Result",
"Discussion",
"Materials",
"and",
"methods"
] | [
"biotechnology",
"plant",
"anatomy",
"plant",
"physiology",
"pollen",
"membrane",
"proteins",
"plant",
"science",
"rice",
"genetically",
"modified",
"plants",
"experimental",
"organism",
"systems",
"plants",
"cellular",
"structures",
"and",
"organelles",
"flower",
"anat... | 2017 | OsCNGC13 promotes seed-setting rate by facilitating pollen tube growth in stylar tissues |
Integration of multimodal sensory information is fundamental to many aspects of human behavior , but the neural mechanisms underlying these processes remain mysterious . For example , during face-to-face communication , we know that the brain integrates dynamic auditory and visual inputs , but we do not yet understand where and how such integration mechanisms support speech comprehension . Here , we quantify representational interactions between dynamic audio and visual speech signals and show that different brain regions exhibit different types of representational interaction . With a novel information theoretic measure , we found that theta ( 3–7 Hz ) oscillations in the posterior superior temporal gyrus/sulcus ( pSTG/S ) represent auditory and visual inputs redundantly ( i . e . , represent common features of the two ) , whereas the same oscillations in left motor and inferior temporal cortex represent the inputs synergistically ( i . e . , the instantaneous relationship between audio and visual inputs is also represented ) . Importantly , redundant coding in the left pSTG/S and synergistic coding in the left motor cortex predict behavior—i . e . , speech comprehension performance . Our findings therefore demonstrate that processes classically described as integration can have different statistical properties and may reflect distinct mechanisms that occur in different brain regions to support audiovisual speech comprehension .
While engaged in a conversation , we effortlessly integrate auditory and visual speech information into a unified perception . Such integration of multisensory information is a key aspect of audiovisual speech processing that has been extensively studied [1–4] . Studies of multisensory integration have demonstrated that , in face-to-face conversation , especially in adverse conditions , observing lip movements of the speaker can improve speech comprehension [4–7] . In fact , some people’s ability to perform lip reading demonstrates that lip movements during speech contain considerable information to understand speech without corresponding auditory information [1 , 8] , even though auditory information is essential to understand speech accurately [9] . Turning to the brain , we know that specific regions are involved in audiovisual integration . Specifically , the superior temporal gyrus/sulcus ( STG/S ) responds to integration of auditory and visual stimuli , and its disruption leads to reduced McGurk fusion [10–14] . However , these classic studies present two shortcomings . First , their experimental designs typically contrasted two conditions: unisensory ( i . e . , audio or visual cues ) and multisensory ( congruent or incongruent audio and visual cues ) . However , such contrast does not dissociate effects of integration per se from those arising from differences in stimulation complexity ( i . e . , one or two sources ) that could modulate attention , cognitive load , and even arousal . A second shortcoming is that previous studies typically investigated ( changes of ) regional activation and not information integration between audiovisual stimuli and brain signals . Here , we address these two shortcomings and study the specific mechanisms of audiovisual integration from brain oscillations . We used a novel methodology ( speech-brain entrainment ) and novel information theoretic measures ( the partial information decomposition [PID] [15] ) to quantify the interactions between audiovisual stimuli and dynamic brain signals . Our methodology of speech-brain entrainment builds on recent studies suggesting that rhythmic components in brain activity that are temporally aligned to salient features in speech—most notably the syllable rate [5 , 6 , 16–18]—facilitate processing of both the auditory and visual speech inputs . The main advantage of speech-brain entrainment is that it replaces unspecific measures of activation with measures that directly quantify the coupling between the components of continuous speech ( e . g . , syllable rate ) and frequency-specific brain activity , thereby tapping more directly into the brain mechanisms of speech segmentation and coding [17] . In the present study , we used a recently developed information theoretic framework called PID ( see Fig 1A and Materials and methods for details ) [15 , 19 , 20] . We consider a three-variable system with a target variable M ( here magnetoencephalography [MEG] ) and two predictor variables A and V ( here audio and visual speech signals ) , with both A and V conveying information about the target M . Conceptually , the redundancy is related to whether the information conveyed by A and V is the same or different . If the variables are fully redundant , then this means either alone is enough to convey all the information about M ( i . e . , obtain an optimal prediction of M ) , and adding observation of the second modality has no benefit for predicting the MEG signal M . The concept of synergy is related to whether A and V convey more information when observed simultaneously , so the prediction of M is enhanced by simultaneous observation of the values of A and V [15] . This means M also represents the instantaneous relationship between A and V . For example , if M is given by the difference between A and V at each sample , then observing either A or V alone tells little about the value of M , but observing them together completely determines it . The PID provides a methodology to rigorously quantify both redundancy and synergy , as well as the unique information in each modality . Unique information is the prediction of the MEG that can be obtained from observing A alone but that is not redundantly available from observing V . The PID framework therefore addresses a perennial question in multisensory processing: the extent to which each sensory modality contributes uniquely to sensory representation in the brain versus how the representation of different modalities interact ( e . g . , audio and visual ) . The PID provides a principled approach to investigate different cross-modal representational interactions ( redundant and synergistic ) in the human brain during naturalistic audiovisual speech processing—that is , to understand how neural representations of dynamic auditory and visual speech signals interact in the brain to form a unified perception . Specifically , we recorded brain activity using MEG while participants attended to continuous audiovisual speech to entrain brain activity . We applied the PID to reveal where and how speech-entrained brain activity in different regions reflects different types of auditory and visual integration . In the first experimental condition , we used naturalistic audiovisual speech for which attention to visual speech was not critical ( “All congruent” condition ) . In the second condition , we added a second interfering auditory stimulus that was incongruent to the congruent audiovisual stimuli ( “AV congruent” condition ) , requiring attention to visual speech to suppress the competing additional incongruent auditory input . In the third condition , both auditory stimuli were not congruent to visual stimulus ( “All incongruent” ) . This allows us to see how the congruence of audiovisual stimuli changes integration . We contrasted measures of redundant and synergistic cross-modal interactions between the conditions to reveal differential effects of attention and congruence on multisensory integration mechanisms and behavioral performance .
Next , we investigated how multimodal representational interactions are modulated by attention and congruence in continuous audiovisual speech . Here , we focus on an “AV congruent” condition in which a congruent audiovisual stimulus pair is presented monaurally together with an interfering nonmatching auditory speech stimulus to the other ear ( Fig 2B ) . This condition is of particular interest because visual speech ( lip movement ) is used to disambiguate the two competing auditory speech signals . Furthermore , it is ideally suited for our analysis because we can directly contrast representational interactions quantified with the PID in matching and nonmatching audiovisual speech signals in the same data set ( see Fig 2B ) . Fig 3 shows corrected group statistics for the contrast of matching and nonmatching audiovisual speech in the “AV congruent” condition . Redundant information is significantly stronger in left auditory and superior and middle temporal cortices ( Fig 3A; Z-difference map at P < 0 . 005 ) for matching compared to nonmatching audiovisual speech . In contrast , significantly higher synergistic information for matching compared to nonmatching audiovisual speech is found in left motor and bilateral visual areas spreading along dorsal and ventral stream regions of speech processing [24] ( Fig 3B; Z-difference map at P < 0 . 005 ) . Next , we tested attention and congruence effects separately because the contrast of matching versus nonmatching audiovisual speech confounds both effects . First , the congruence effect ( “AV congruent” > “All incongruent” ) shows higher redundant information in left inferior frontal region ( BA 44/45 ) and posterior superior temporal gyrus and right posterior middle temporal cortex ( Fig 4A; Z-difference map at P < 0 . 005 ) and higher synergistic information in superior part of somatosensory and parietal cortices in left hemisphere ( Fig 4B; Z-difference map at P < 0 . 005 ) . The attention effect ( “AV congruent” > “All congruent” ) shows higher redundant information in left auditory and temporal ( superior , middle , and inferior temporal cortices and pSTG/S ) areas and right inferior frontal and superior temporal cortex ( Fig 5A; Z-difference map at P < 0 . 005 ) . Higher synergistic information was localized in left motor cortex , inferior temporal cortex , and parieto-occipital areas ( Fig 5B; Z-difference map at P < 0 . 005 ) . In summary , theta-band activity in left pSTG/S represents redundant information about audiovisual speech significantly more strongly in experimental conditions with higher attention and congruence . In contrast , synergistic information in the left motor cortex is more prominent in conditions requiring increased attention . Therefore , the increased relevance of visual speech in the “AV congruent” condition leads to increased redundancy in left pSTG/S and increased synergy in left motor cortex . This differential effect on representational interactions may reflect different integration mechanisms operating in the different areas . For detailed local maps of interaction between predictors ( auditory and visual speech signals ) and target ( MEG response ) , see S3 Fig . Next , we investigated if the differential pattern of redundancy and synergy is of behavioral relevance in our most important condition—"AV congruent"—in which visual speech is particularly informative . To this end , we extracted raw values of redundancy for the location showing strongest redundancy in the left pSTG/S in Fig 5A and synergy for the location showing strongest synergy in the left motor cortex in Fig 5B for “AV congruent” condition . After normalization with surrogate data ( see Materials and methods section ) , we computed correlation with performance measures ( comprehension accuracy ) across participants . Both redundancy in left pSTG/S ( R = 0 . 43 , P = 0 . 003; Fig 5C ) and synergy in left motor cortex ( R = 0 . 34 , P = 0 . 02; Fig 5D ) are significantly correlated with comprehension accuracy . These results suggest that the redundancy in left pSTG/S and synergy in left motor cortex under challenging conditions ( i . e . , in the presence of distracting speech ) are related to perceptual mechanisms underlying comprehension .
In fMRI studies , audiovisual speech integration has been studied using experimental conditions that manipulate the stimulus modalities presented ( e . g . , [13 , 25] ) . Changes in blood oxygen level–dependent ( BOLD ) responses elicited by congruent audiovisual stimuli ( AV ) have been compared to auditory-only ( AO ) , visual-only ( VO ) , their sum ( AO + VO ) , or their conjunction ( AO ∩ VO ) . Greater activation for the congruent audiovisual condition ( AV ) compared to others has been interpreted as a signature of audiovisual speech integration . Comparison to auditory-only ( AO ) activation or visual-only ( VO ) activation has been regarded as a less conservative criterion for integration , since even if auditory and visual stimuli caused independent BOLD activity that combined linearly , this contrast would reveal an effect . To address this , comparison to the summation of the unimodal activations ( AO + VO ) has been used to demonstrate supra-additive activation , which is more suggestive of a cross-modal integration process . Rather than overall activation while the stimulus is present , the information theoretic approach instead focuses on quantifying the degree to which the changing speech time course is encoded or represented in the neural signals . The MI calculated here is an effect size for the ongoing entrainment of the MEG time course by the time varying speech—i . e . , it quantifies the strength of the representation of dynamic audiovisual speech in the neural activity . While the basic expression on which our redundancy measure is based ( Materials and methods , Eq 1 ) looks similar to an activation contrast ( e . g . , sum versus conjunction ) , it is important to keep in mind that this is about the strength of the dynamic low-frequency entrainment in each modality , not simply overall activation contrasts between conditions as in the classic fMRI approach . The PID can quantify the representational interactions between multiple sensory signals and the associated brain response in a single experimental condition in which both sensory modalities are simultaneously present . In the PID framework , the unique contributions of a single ( e . g . , auditory ) sensory modality to brain activity are directly quantified when both are present , instead of relying on the statistical contrast between modalities presented independently . Furthermore , the PID method allows the quantification of both redundant and synergistic interactions . In the context of audiovisual integration , both types of interaction can be seen as integration effects . Redundant information refers to quantification of overlapping information content of the predictor variables ( auditory and visual speech signals ) , and synergistic information refers to additional information gained from simultaneous observation of two predictor variables compared to observation of one . Both of these types of interaction quantify multimodal stimulus representation that cannot be uniquely attributed to one of the two modalities . Redundant representation cannot be uniquely attributed , since that part of the brain response could be predicted from either of the stimulus modalities . Synergistic representation cannot be uniquely attributed , since that part of the brain response could only be predicted from simultaneous observation of both modalities and not from either one alone . Note that these statistical interactions are quite different from interaction terms in a linear regression analysis , which would indicate the ( linear ) functional relationship between one stimulus modality and the MEG response is modulated by the value of the other stimulus modality . MI is an effect size that can be interpreted , because of its symmetry , from both an encoding and decoding perspective . From an encoding perspective , MI is a measure of how much an observer’s predictive model for possible MEG activity values changes when a specific auditory speech value is observed 100 ms prior . It quantifies the improvement in predictive performance of such an observer when making an optimal guess based on the auditory speech signal they see , over the guess they would make based on overall MEG activity without observing a stimulus value . From this perspective , redundancy quantifies the overlapping or common predictions that would be made by two Bayesian optimal observers , one predicting based on the auditory signal and the other the visual . Synergy is an increase in predictive power when both signals are obtained simultaneously . That is , it is possible to obtain a better prediction of the MEG with simultaneous knowledge of the specific combination of A and V observed than it is from combining only the predictions of the previous two unimodal observers . From considering the local plots ( i . e . , the values that are summed to obtain the final expectation value ) in S3 Fig , we can see that a better prediction of the MEG in left motor cortex is made from the joint multimodal input in the case in which the MEG signal is high ( above median ) , and the auditory and visual signals are in opposite ranges ( e . g . , high/low or low/high ) . Existing techniques like representational similarity analysis ( RSA ) [26] and cross-decoding [27] can address the same conceptual problem as redundancy but from the angle of similarity of representations on average rather than specific overlapping Bayesian predictive information content within individual samples , which the information theoretic framework provides . Techniques exploiting decoding in different conditions can show the degree to which multimodal representations are similar to unimodal representations [28 , 29] and whether there is an improvement in performance when the representation is learned in the multimodal condition . However , PID is explicitly a trivariate analysis considering two explicit quantified stimulus features and the brain signal . The information theoretic definition of synergy means there is enhanced prediction of neural responses from simultaneous multimodal stimuli compared to independent predictions combined from each modality ( but still presented together ) . This differs from typical multimodal versus unimodal contrasts , even those involving decoding , because it explicitly considers the effect of continuous naturalistic variation in both stimulus modalities on the recorded signal . Posterior superior temporal region ( pSTG/S ) has been implicated in audiovisual speech integration area by functional [30–32] and anatomical [33] neuroimaging . A typical finding in fMRI studies is that pSTG/S shows stronger activation for audiovisual ( AV ) compared to auditory-only ( AO ) and/or visual-only ( VO ) conditions . This was confirmed by a combined fMRI-transcranial magnetic stimulation ( TMS ) study in which the likelihood of McGurk fusion was reduced when TMS was applied individually to fMRI-localized posterior superior temporal sulcus ( pSTS ) , suggesting a critical role of pSTS in auditory-visual integration [14] . The redundant information in the same left superior temporal region in this study matches this notion that this region processes shared information from both modalities . We found this region not only in the congruence effect ( “AV congruent” > “All incongruent”; Fig 4A ) but also in the attention effect ( “AV congruent” > “All congruent”; Fig 5A ) . We found the left motor cortex shows increased synergy for the matching versus nonmatching audio stimuli of “AV congruent” condition ( Fig 3B ) . However , further analysis optimized for effects of attention and congruence revealed slightly different areas—with the area that shows strongest synergy change with attention ( Fig 5B; BA6 ) located more lateral and anterior compared to the area identified in the congruence ( Fig 4B ) . Previous studies have demonstrated increased phase locking of left motor cortex activity to frequency-tagged stimuli during auditory spatial attention [34 , 35] . We extend these findings by demonstrating attention-mediated synergistic interactions of auditory and visual representations in left motor cortex . The motor region in the attention contrast is consistent with the area in our previous study that showed entrainment to lip movements during continuous speech that correlated with speech comprehension [5] . In another study , we identified this area as the source of top-down modulation of activity in the left auditory cortex [23] . The definition of synergistic information in our context refers to more information gained from the simultaneous observation of auditory and visual speech compared to the observation of each alone . When it comes to the attention effect ( “AV congruent” > “All congruent” ) , “AV congruent” condition requires paying more attention to auditory and visual speech than the “All congruent” condition does , even though the speech signals to be attended match the visual stimulus in both conditions . Thus , this synergy effect in the left motor cortex can be explained by a net attention effect at the same level of stimulus congruence . This effect is likely driven by stronger attention to visual speech , which is informative for the disambiguation of the two competing auditory speech streams [5] . This notion is plausible because it is supported by directional information analysis that shows that the left motor cortex better predicts upcoming visual speech in the “AV congruent” condition , in which attention to visual speech is crucial ( S2B and S2D Fig ) . However , a number of open questions in need of further investigation still remain . First , the auditory speech envelope and lip area information used in our analysis only capture part of the rich audiovisual information that is available to interlocutors in a real-life conversation . Other , currently unaccounted features might even be correlated across modalities ( e . g . , a different visual feature that is correlated with the auditory envelope ) . Since our analysis is restricted to these two features , it is possible that with a richer feature set for each modality , the unique information obtained from each would be reduced . In addition , the auditory speech signal is available at a much higher temporal resolution compared to the lip area signal , leading to a potential bias in the information content of both signals . Since the analysis of speech-brain coupling is a relatively new research field , we envisage methodological developments that will capture more aspects of the rich audiovisual signals . But in the context of our analysis that is focused on syllable components in speech , it seems reasonable to use these two signals that are known to contain clear representations of syllable-related frequencies [18 , 36] . Second , it should be noted that we computed PID measures on the speech signals and 100 ms shifted MEG signal as in previous analyses [5 , 18 , 23] to compensate for delays between stimulus presentation and main cortical responses . We have confirmed that this ( on average ) maximizes speech-brain coupling . However , different aspects of multisensory integration likely occur at different latencies , especially in higher-order brain areas . This highly interesting but complex question is beyond the scope of the present study but will hopefully be addressed within a similar framework in future studies . Third , while an unambiguous proof is missing , we believe that converging evidence suggests that participants attended visual speech more in “AV congruent” condition than in the other conditions . Indeed , it seems very unlikely that participants did not attend to visual speech after being explicitly instructed to attend ( especially because visual speech provided important task-relevant information in the presence of a distracting auditory input ) . The converging evidence is based on behavioral performance , eye tracking results , and previous studies . Previous research indicates that the availability of visual speech information improves speech intelligibility under difficult listening conditions [1 , 6 , 37] . The “AV congruent” condition was clearly more difficult compared to the “All congruent” condition because of the presence of an interfering auditory stimulus . One could argue that participants could accomplish the task by simply using auditory spatial attention . However , our behavioral data ( see Fig 1B in [5] ) argue against this interpretation . If participants had ignored the visual stimulus and only used auditory spatial attention , then we would expect to see the same behavioral performance between “AV congruent” and “All incongruent” conditions . In both cases , two different auditory stimuli were presented , and only relying on auditory information would lead to the same behavioral performance . Instead , we find a significant difference in behavioral performance between both conditions . The availability of the congruent visual stimulus ( in the “AV congruent” condition ) resulted in a significant increase of behavioral performance ( compared to “All incongruent” condition ) to the extent that it reached the performance for the “All congruent” condition ( no significant difference between “All congruent” and “AV congruent” conditions measured by comprehension accuracy; mean ± s . e . m; 85 . 0% ± 1 . 66% for “All congruent , ” 83 . 40% ± 1 . 73% for “AV congruent” condition ) . This is strong evidence that participants actually made use of the visual information . In addition , this is also supported by eye fixation on the speaker’s lip movement , as shown in S5 Fig . In summary , we demonstrate how information theoretic tools can provide a new perspective on audiovisual integration , by explicitly quantifying both redundant and synergistic cross-modal representational interactions . This reveals two distinct profiles of audiovisual integration that are supported by different brain areas ( left motor cortex and left pSTG/S ) and are differentially recruited under different listening conditions .
Data from 44 subjects were analyzed ( 26 females; age range: 18–30 y; mean age: 20 . 54 ± 2 . 58 y ) . Another analysis of these data was presented in a previous report [5] . All subjects were healthy , right-handed ( confirmed by Edinburgh Handedness Inventory [38] ) , and had normal or corrected-to-normal vision and normal hearing ( confirmed by 2 hearing tests using research applications on an iPad: uHear [Unitron Hearing Limited] and Hearing-Check [RNID] ) . None of the participants had a history of developmental , psychological , or neurological disorders . They all provided informed written consent before the experiment and received monetary compensation for their participation . The study was approved by the local ethics committee ( CSE01321; College of Science and Engineering , University of Glasgow ) and conducted in accordance with the ethical guidelines in the Declaration of Helsinki . We used audiovisual video clips of a professional male speaker talking continuously ( 7–9 min ) , which were used in our previous study [5] . Since in some conditions ( “AV congruent , ” “All incongruent” conditions ) the auditory speeches are delivered dichotically , to ensure that there are no differences other than talks themselves in those conditions , we made all the videos with the same male speaker . The talks were originally taken from TED talks ( www . ted . com/talks/ ) and edited to be appropriate to the stimuli we used ( e . g . , editing words referring to visual materials , the gender of the speaker , etc . ) . High-quality audiovisual video clips were filmed by a professional filming company , with sampling rate of 48 kHz for audio and 25 frames per second ( fps ) for video in 1 , 920 × 1 , 080 pixels . In order to validate stimuli , 11 videos were rated by 33 participants ( 19 females; aged 18–31 y; mean age: 22 . 27 ± 2 . 64 y ) in terms of arousal , familiarity , valence , complexity , significance ( informativeness ) , agreement ( persuasiveness ) , concreteness , self-relatedness , and level of understanding , using Likert scale [39] 1–5 ( for an example of concreteness , 1: very abstract , 2: abstract , 3: neither abstract nor concrete , 4: concrete , 5: very concrete ) . Eight talks were finally selected for the MEG experiment by excluding talks with mean scores of 1 and 5 . Questionnaires for each talk were validated in a separate behavioral study ( 16 subjects; 13 females; aged 18–23 y; mean age: 19 . 88 ± 1 . 71 y ) . These questionnaires are designed to assess the level of speech comprehension . Each questionnaire consists of 10 questions about a given talk to test general comprehension ( e . g . , “What is the speaker’s job ? ” ) and were validated in terms of accuracy ( the same level of difficulty ) , response time , and the length ( word count ) . Experimental conditions used in this study were “All congruent , ” “All incongruent , ” and “AV congruent . ” In each condition ( 7–9 min ) , 1 video recording was presented , and 2 ( matching or nonmatching ) auditory recordings were presented to the left and the right ear , respectively . Half of the 44 participants attended to speech in the left ear , and the other half attended to speech in the right ear . The “All congruent” condition is a natural audiovisual speech condition in which auditory stimuli to both ears and visual stimuli are congruent ( V1A1A1; the first A denotes talk presented to the left ear , and the second A denotes talk presented to the right ear; the number refers to the identity of the talks ) . The “All incongruent” condition has three different stimulus streams from three different videos , and participants are instructed to attend to auditory information presented to one ear ( V1A2A3 ) . The “AV congruent” condition consists of one auditory stimulus matching the visual information , and the speech presented to the other ear serves as a distractor . Participants attend to the talk that matches visual information ( V1A1A2 for left ear attention group , V1A2A1 for right ear attention group ) . Each condition represents one experimental block , and the order of conditions was counterbalanced across subjects . Participants were instructed to fixate on the speaker’s lip throughout the presentation in all experimental conditions , and we monitored the eye gaze using an eye tracker . Furthermore , we explained the importance of eye fixation on the speaker’s lip movement during the instruction session . They were also informed that for this reason , their eye movement and gaze behavior would be monitored using an eye tracker ( see eye tracker data analysis in S5 Fig ) . A fixation cross ( either yellow or blue color ) was overlaid on the speaker’s lip during the whole video for mainly two reasons: ( 1 ) to help maintain eye fixation on the speaker’s lip movement and ( 2 ) to indicate the auditory stimulus to pay attention to ( left or right ear; e . g . , “If the color of fixation cross is yellow , please attend to left ear speech” ) . The color was counterbalanced across subjects ( for half of participants , yellow indicates attention to the left ear speech; for another half , attention to the right ear speech ) . This configuration was kept the same for all experimental conditions to ensure the same video display other than the experimental manipulations we aimed at . However , in “All congruent” condition ( natural audiovisual speech ) , in which 1 auditory stream is presented diotically , attention cannot be directed to left or right ear , so participants were instructed to ignore the color of the fixation cross and just to attend the auditory stimuli naturally . In addition , to prevent stimulus-specific effects , we used 2 sets of stimuli consisting of different combinations of audiovisual talks . These 2 sets were randomized across participants ( set 1 for half of participants , set 2 for the other half ) . For example , talks for “All congruent” condition in set 1 were talks for “AV congruent” condition in set 2 . There was no significant difference in comprehension accuracy between left and right ear attention groups ( two-sample t test , df: 42 , P > 0 . 05 ) . In this study , we pooled across both groups for data analysis so that attentional effects for a particular side ( e . g . , left or right ) are expected to cancel out . For the recombination and editing of audiovisual talks , we used Final Cut Pro X ( Apple , Cupertino , CA ) . The stimuli were presented with Psychtoolbox [40] in MATLAB ( MathWorks , Natick , MA ) . Visual stimuli were delivered with a resolution of 1 , 280 × 720 pixels at 25 fps ( mp4 format ) . Auditory stimuli were delivered at a 48 kHz sampling rate via a sound pressure transducer through 2 five-meter-long plastic tubes terminating in plastic insert earpieces . A comprehension questionnaire was administered about the attended speech separately for each condition . Cortical neuromagnetic signals were recorded using a 248 magnetometers whole-head MEG system ( MAGNES 3600 WH , 4-D Neuroimaging ) in a magnetically shielded room . The MEG signals were sampled at 1 , 017 Hz and were denoised with information from the reference sensors using the denoise_pca function in FieldTrip toolbox [41] . Bad sensors were excluded by visual inspection , and electrooculographic ( EOG ) and electrocardiographic ( ECG ) artifacts were eliminated using independent component analysis ( ICA ) . An eye tracker ( EyeLink 1000 , SR Research ) was used to examine participants’ eye gaze and movements to ensure that they fixated on the speaker’s lip movements . Structural T1-weighted MRIs of each participant were acquired at 3 T Siemens Trio Tim scanner ( Siemens , Erlangen , Germany ) with the following parameters: 1 . 0 × 1 . 0 × 1 . 0 mm3 voxels; 192 sagittal slices; field of view ( FOV ) : 256 × 256 matrix . Information theoretic quantities were estimated with the Gaussian-Copula Mutual Information ( GCMI ) method [42] ( https://github . com/robince/gcmi ) . PID analysis was performed with the GCMI approach in combination with an open source PID implementation in MATLAB , which implements the PID [19 , 20] with a redundancy measure based on common change in local surprisal [15] ( https://github . com/robince/partial-info-decomp ) . For statistics and visualization , we used the FieldTrip Toolbox [41] and in-house MATLAB codes . We followed the suggested guidelines [43] for MEG studies . | Combining different sources of information is fundamental to many aspects of behavior , from our ability to pick up a ringing mobile phone to communicating with a friend in a busy environment . Here , we have studied the integration of auditory and visual speech information . Our work demonstrates that integration relies upon two different representational interactions . One system conveys redundant information by representing information that is common to both auditory and visual modalities . The other system , which is supported by a different brain area , represents synergistic information by conveying greater information than the linear summation of individual auditory and visual information . Further , we show that these mechanisms are related to behavioral performance . This novel insight opens new ways to enhance our understanding of the mechanisms underlying multi-modal information integration , a fundamental aspect of brain function . These fresh insights have been achieved by applying to brain imaging data a recently developed methodology called the partial information decomposition . This methodology also provides a novel and principled way to quantify the interactions between representations of multiple stimulus features in the brain . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"acoustics",
"medicine",
"and",
"health",
"sciences",
"linguistics",
"engineering",
"and",
"technology",
"audio",
"signal",
"processing",
"visual",
"signals",
"signal",
"processing",
"brain",
"social",
"sciences",
"neuroscience",
"brain",
"mapping",
"vision",
"neuroimag... | 2018 | Representational interactions during audiovisual speech entrainment: Redundancy in left posterior superior temporal gyrus and synergy in left motor cortex |
Infection by the simian malaria parasite , Plasmodium knowlesi , can lead to severe and fatal disease in humans , and is the most common cause of malaria in parts of Malaysia . Despite being a serious public health concern , the geographical distribution of P . knowlesi malaria risk is poorly understood because the parasite is often misidentified as one of the human malarias . Human cases have been confirmed in at least nine Southeast Asian countries , many of which are making progress towards eliminating the human malarias . Understanding the geographical distribution of P . knowlesi is important for identifying areas where malaria transmission will continue after the human malarias have been eliminated . A total of 439 records of P . knowlesi infections in humans , macaque reservoir and vector species were collated . To predict spatial variation in disease risk , a model was fitted using records from countries where the infection data coverage is high . Predictions were then made throughout Southeast Asia , including regions where infection data are sparse . The resulting map predicts areas of high risk for P . knowlesi infection in a number of countries that are forecast to be malaria-free by 2025 ( Malaysia , Cambodia , Thailand and Vietnam ) as well as countries projected to be eliminating malaria ( Myanmar , Laos , Indonesia and the Philippines ) . We have produced the first map of P . knowlesi malaria risk , at a fine-scale resolution , to identify priority areas for surveillance based on regions with sparse data and high estimated risk . Our map provides an initial evidence base to better understand the spatial distribution of this disease and its potential wider contribution to malaria incidence . Considering malaria elimination goals , areas for prioritised surveillance are identified .
Malaria cases caused by the simian parasite , Plasmodium knowlesi , have been identified in at least nine Southeast Asian countries . In many parts of Malaysia , this parasite is the most common cause of malaria [1] and can lead to severe and fatal disease [2–4] . Despite the potential severity of infection [1 , 4–8] , diagnostics that identify P . knowlesi are not routinely used . Unless blood samples are tested using expensive nested PCR-based diagnostics , cases of P . knowlesi are often misdiagnosed by microscopy as one of the human malarias , principally P . malariae or P . falciparum [9–11] . Plasmodium knowlesi infection is routinely considered as a potential causal pathogen of malaria cases in three countries: Malaysia , Brunei and Singapore ( hereafter a region referred to as MBS ) , the latter two having already eliminated the human malarias . Plasmodium knowlesi malaria cases have also been reported in Cambodia , Indonesia , Myanmar , the Philippines , Thailand and Vietnam [8 , 12 , 13] , but sampling has been limited and the full geographical extent of disease risk across most of the region , including within these countries , is unknown . Understanding the geographical distribution of P . knowlesi is important to identify areas where residual malaria transmission could remain once the human malarias , namely P . falciparum , P . vivax , P . malariae and P . ovale , have been eliminated [14] . Human malaria parasites are primarily transmitted between humans via mosquitoes and are not frequently transmitted from other animals to humans . Many countries in Southeast Asia , including Malaysia , the Philippines , Thailand and Vietnam , are currently in the process of eliminating the human malarias [15] . Current control measures that reduce these malarias include mass anti-malarial drug administration , the provision of insecticide-impregnated bed nets ( ITNs ) and indoor residual spraying of houses ( IRS ) . These control measures do not , however , target transmission of the parasite within populations of the reservoir host species so P . knowlesi populations will not be eliminated . Further , ITNs and IRS are unlikely to offer the same degree of personal protection to humans , or community protection through reductions in mosquito longevity , since the vectors for P . knowlesi bite and rest outdoors [16] . If the presence of P . knowlesi is not considered when elimination strategies are developed , the impact of elimination measures and reduction in overall malaria cases in these areas will not match projections . In this study , we produced the first map of the geographical distribution of P . knowlesi malaria , using a niche modelling approach previously applied to the mapping of other vector-borne and zoonotic diseases , including dengue [17] , the Leishmaniases [18] , Ebola virus disease [19] , Lassa fever [20] , Marburg virus disease [21] , Crimean-Congo hemorrhagic fever [22] , and Zika virus [23] . Niche models are able to combine information on locations where diseases have been recorded with geographic data on environmental and socioeconomic factors hypothesized to affect disease transmission [24] . Once the model has been fitted , the potential presence of the disease can be predicted in regions where it has yet to be reported . To identify regions at risk from a disease with reservoirs in multiple host species , transmitted by multiple vector species , this modelling approach needs to be further refined . The spatial distribution of such diseases is restricted to locations where all species required for transmission coincide , so it is important to consider the distributions of these species [25] . The work presented here builds on previous work that assessed the evidence for the limits of transmission [12] . Here we refine those spatial limits and investigate the variation in risk within them . We recently defined the fine-scale species distributions of the known and putative reservoirs and vectors of P . knowlesi [26] , including the main macaque species identified as natural hosts of P . knowlesi , Macaca fascicularis and M . nemestrina [27–31] , and several anopheline mosquito species , all from the Leucosphyrus Group , implicated as vectors of P . knowlesi [16 , 32–35] . Our maps of these species are useful for defining the limits of zoonotic transmission , but an index of disease risk cannot be extrapolated directly from reservoir/vector maps . While co-occurrence of reservoir and vector species involved in a zoonotic disease system is necessary for transmission , it is not always sufficient , as many other factors contribute . We have developed a model that incorporates the relationships between P . knowlesi infection and the distributions of the reservoir and vector species , along with a range of other potential risk factors , to produce fine-scale evidence-based predictions of relative zoonotic P . knowlesi transmission risk . The final output provides an initial map that aims to identify locations where disease surveillance and epidemiological investigations would be most informative to improve our understanding of disease risk .
A schematic of the process we followed is given in Fig 1 . We collated and geo-positioned records of P . knowlesi infections in humans , and reservoir and vector species , from a variety of sources . The study area was a rectangle encompassing the locations of confirmed or putative P . knowlesi infections plus a minimum buffer zone of 300 km , giving an area from northeast Bangladesh to southwest Papua New Guinea . A map predicting the human risk of P . knowlesi malaria at every square ( pixel ) in a 5 km × 5 km grid was generated using an ensemble of boosted regression tree ( BRT ) models to carry out a niche modelling analysis . The model used the database of geo-positioned occurrence points for P . knowlesi infections combined with 19 gridded datasets of environmental and socio-economic explanatory covariates as well as probabilistic species distributions for M . nemestrina , M . fascicularis and the Leucosphyrus Group . Datasets comprising ad-hoc reports of disease occurrence ( as opposed to data from planned region-wide surveys ) are subject to spatial bias in reporting rates , which if unaccounted for may result in elevated risk predictions in the areas most likely to report [36] . Reporting bias is likely to be more pronounced for P . knowlesi malaria since significant resources are required to accurately diagnose infection and P . knowlesi infection is not routinely considered a possible cause of malaria in the region , with the exception of MBS . A model was therefore fitted using data only from MBS ( the model training region ) , where we could account for reporting bias through our selection of background data . This model was then used to predict the human risk of P . knowlesi infection across Southeast Asia . To assess the model’s predictive capacity outside its training region , we tested its performance on a set of disease presence and absence records from locations outside MBS . We also generated a multivariate environmental similarity surface to identify regions where the model was required to extrapolate to environments not found within MBS , and therefore evaluate the appropriateness of inferring risk in those regions . Records of P . knowlesi presence or absence for Southeast Asia were obtained from reports in the published literature from 2004 to 2015 , which were validated through personal communications with the authors to confirm details . Each presence record contains the coordinates of a point location or an area greater than 25 km2 ( polygons ) where a human , macaque or mosquito infection was confirmed using specific diagnostics that are able to distinguish P . knowlesi from the other Plasmodium spp . Presence points were excluded if P . knowlesi presence in the surrounding area ( within 300 km radius ) was not verified by a second , independent laboratory . Each absence record contains the coordinates for a site where an appropriate diagnostic for P . knowlesi was used , but no infections were found in a sample size of at least 500 individuals , or where no malaria cases were reported across an administrative division in 2012 [37] . Further details regarding data assemblage are included in the Supporting Information along with the complete data ( S1 and S2 Files ) . Nineteen 5 km × 5 km gridded data surfaces of a range of environmental and socio-economic factors , along with predicted reservoir and vector species distributions , were used as explanatory covariates ( Table 1 ) . No prior assumptions were made about the nature of any relationships between these covariates and disease risk . Further details regarding the construction of each covariate data surface is provided in the Supporting Information along with plots of each surface ( S1 File , and S1 and S2 Figs ) . To carry out the niche mapping analysis , we fitted an ensemble of boosted regression tree ( BRT ) models using the gbm R package [47] . This BRT approach has the ability to fit complex nonlinear responses including high-dimensional interactions between explanatory variables [48] , has been shown to have high predictive accuracy [24] and has been previously applied to disease distribution mapping [17–23] . Boosted regression trees combine two algorithms: regression trees ( which repeatedly split the data into two groups using a randomly selected predictor variable for each split ) and boosting ( which additively fits trees to the data , gradually prioritizing poorly modelled data to produce a set of trees that maximally reduce the loss function ) , to examine and quantify the relationship between explanatory variables and the response data [48] . The core setup used has been described previously [17 , 19 , 48] . The changes made to the method for the work presented here addressed sampling bias in the infection reports within MBS , incorporated host and vector data , allowed temporal changes in land cover to be incorporated , and improved handling of polygon data . Rather than exclusively using synoptic ( averaged across time ) covariate values for each of the occurrence locations irrespective of the occurrence date , we incorporated annual data surfaces describing land cover and reservoir and vector distributions from 2001 to 2012 . This was necessary to account for the substantial changes in land cover that have occurred in the region over the study period due to deforestation [49] , which is hypothesized to have impacted the distribution of the reservoir and vector species of P . knowlesi [50] . Using the annual land cover data surfaces , disease occurrence data collected between 2001 and 2012 were matched with covariate values for the relevant year; most data points ( 76% ) fell within this time period . Covariate values for 2001 were used for occurrence data prior to this date and covariate values for 2012 were used for post-2012 data . Final predictions were made to the most contemporary covariate values available to represent the current distribution of disease risk . We used a binomial likelihood function for the BRT model in order to incorporate both presence and absence records . Whilst records of disease absence are highly informative , they are rare because they require significantly greater sampling effort to ensure their reliability compared to presence data [51] , especially for diseases like P . knowlesi malaria where the appropriate diagnostics are rarely used . We therefore supplemented the dataset with a large number of background records to represent areas where the disease has not been reported within MBS . Six thousand background points were generated in total [51] with the same proportion of human , macaque and mosquito background points as the presence dataset . It has been demonstrated that predictive accuracy of presence-background niche models can be improved by selecting background data with similar spatial bias to the occurrence records [36] . Human infection background points were generated by randomly sampling across MBS , biased towards human population density , since more populous areas have a greater probability of reporting at least one case . This method was also used to generate background points for the mosquito infection data since all studies that looked for P . knowlesi infections in vector species , selected study locations based on the presence of human P . knowlesi cases in the immediate vicinity . Background points for the macaque infection data were randomly sampled from a macaque occurrence and mammal survey dataset that reflected the bias in locations chosen for macaque studies [26] . Covariate values for the specific times and locations of the background data were then extracted . Prior to covariate extraction , human and vector background points were assigned a year randomly sampled from the temporal distribution of presence points for each species . Since the occurrence dataset included data from humans , macaques and mosquitoes , a joint model was fitted for human , macaque and mosquito hosts that enabled all available infection data to be leveraged , whilst not constraining the model to assume that the distribution of infection risk would be identical for all three host organisms . As BRTs can fit high-dimensional interactions , the joint model is able to fit different environmental responses for each host organism , or if there is no difference in the signal , to fit the same response for all of them . To increase the robustness of model predictions and quantify model uncertainty , we fitted an ensemble of 500 BRT models ( sub-models ) , each trained to a separate bootstrap dataset randomly sampled with replacement from the complete presence/background dataset . To incorporate uncertainty in the precise location of infection for polygon occurrence records , each bootstrap randomly selected a 5 km × 5 km pixel within each polygon . Further information on model fitting can be found in the Supporting Information ( S1 File ) and the R code used to carry out the analysis is freely available at ( https://github . com/fshearer/pk_parasite ) . To generate the final prediction map , a mean predicted value of suitability for infection was calculated across the 500 sub-models ( each fitted using occurrence and covariate data from within MBS ) for each 5 km × 5 km pixel within and outside MBS . The model’s predictive performance was evaluated using the area under the receiver operator curve ( AUC ) statistic , i . e . the area under a plot of the true positive rate versus false positive rate , reflecting the ability to discriminate between presence and background records [52] . The overall statistic was calculated as the mean of the AUCs for each of the 500 sub-models , calculated under 10-fold cross validation . While each sub-model in the BRT ensemble was fitted using occurrence and background data from MBS ( the model training region ) , the goal of our analysis was to predict to a much broader study area from northeast Bangladesh to southwest Papua New Guinea . To assess the model’s predictive performance outside its training region , a separate AUC value was calculated for each sub-model using a validation dataset made up of presence and absence records from locations outside MBS . This AUC was calculated for each sub-model and then averaged across all 500 sub-models in the ensemble . Further information regarding the calculation of the AUC statistics is provided in the Supporting Information ( S1 File ) . The geographic regions outside MBS encompasses environments beyond the ranges of covariate values sampled within the training dataset . Model predictions to such environments are inherently less reliable than interpolations made to areas with environments within the range of covariate values in the training dataset . Thus it is important to assess the environmental similarities and differences between model training and prediction regions [53] . To investigate whether predictions to new geographic regions required extrapolation to covariate values beyond the range of the model training data , we computed and plotted a multivariate environmental similarity surface ( MESS ) [54] using R packages “dismo” [55] and “raster” [56] . This surface represents the similarity of the environment at each location to the covariate values at the presence and background locations in MBS ( the reference data ) . The MESS calculation produces negative values for novel environments , locations where at least one covariate has a value that is outside the range of reference values ( hereafter extrapolation ) , and positive values for locations within this range ( hereafter interpolation ) . We converted the raw MESS output into a binary map indicating areas in which model predictions used interpolation versus extrapolation . The model output was restricted to areas within the range of species known and hypothesized to be required for zoonotic transmission ( i . e . the overlap in range maps of at least one reservoir and vector species ) , using previously reported species range maps [26] . A high resolution map for the zone of zoonotic transmission was also generated ( S3 Fig ) using existing species distribution maps and occurrence datasets [26] . Threshold environmental suitability values for each of the species distribution maps for M . fascicularis , M . nemestrina , M . leonina , and the Leucosphyrus Group were determined to incorporate 90% of presence points in each species’ respective occurrence database . We used these thresholds to classify each continuous species map as either present or absent for every 5 km × 5 km pixel in the study region . These maps were combined to produce a final binary output showing areas of spatial co-occurrence of all species required for zoonotic , vector-borne transmission to humans i . e . presence of at least one macaque species plus at least one member of the Leucosphyrus Group .
A total of 439 P . knowlesi occurrence records were identified , consisting of 301 presence and 138 absence records . The evaluation dataset ( records falling outside MBS ) totaled 131 records , comprising 29 point locations and 102 polygons ( Fig 2A ) . The occurrence dataset used for model fitting ( records falling within MBS ) totaled 198 records , corresponding to 62 unique point locations and 136 polygons ( Fig 2B ) . The model fitting dataset consisted of human ( 166 ) , monkey ( 23 ) and mosquito ( 9 ) occurrence records . The model predictions for the geographical variation in P . knowlesi infection risk in humans are displayed in Fig 3A . Overall , 10-fold cross validation statistics for the model ensemble ( calculated using model training data ) indicated high predictive performance with an AUC of 0 . 833 ( SE ± 0 . 002 ) . A map of model uncertainty is displayed in the Supporting Information ( S4 Fig ) . The model output was restricted to areas within the geographic range of the species required for zoonotic transmission . An unmasked version of the mean output , showing relative suitability for zoonotic P . knowlesi transmission , is provided in the Supporting Information ( S5 Fig ) . The predicted map is presented alongside a projection of malaria eliminating countries in the year 2025 ( Fig 3B ) and together the two maps show countries where P . knowlesi transmission may persist after the human malarias are eliminated . The elimination projections , generated by the University of California San Francisco Global Health Group’s Malaria Elimination Initiative , are based on current national and regional goals as well as recent epidemiological trends for the human malarias , principally P . falciparum and P . vivax [57] . Within MBS , our model predicted considerable spatial variation in risk of P . knowlesi infection , with areas ranging from relatively low risk to high risk predicted within Peninsular Malaysia , and both Sabah and Sarawak States of Malaysian Borneo ( Fig 3A ) . The model also predicts areas of high risk for P . knowlesi infection in a number of countries that are forecast to be malaria-free by 2025 ( Malaysia , Cambodia , Thailand and Vietnam ) as well as countries projected to be eliminating malaria ( Myanmar , Laos , Indonesia and the Philippines ) ( Fig 3B ) . Large areas of high risk were predicted in Myanmar , Laos , Cambodia and Indonesia , with smaller areas predicted in Vietnam and Thailand . Human cases of P . knowlesi infection have been reported across this broad area ( Fig 2A ) . Regions for which we have no field data include areas of high predicted-risk , for example , eastern and western parts of Indonesia , and far eastern parts of India , although the predictions for the latter depend on whether M . leonina , included in the range of zoonotic transmission , is indeed a reservoir species ( Fig 3A ) . Our predictions outside MBS are a result of both interpolation within the environmental space and extrapolation . The binary MESS map ( Fig 3C ) shows that model extrapolation to novel environments occurred in large regions in Cambodia , Vietnam , Thailand , Myanmar , India , and the Andaman and Nicobar Islands , indicating that predictions in these areas should be interpreted with caution . The model did , however , demonstrate high predictive performance at sites outside the model-training region , with an AUC statistic of 0 . 796 ( SE ± 0 . 003 ) calculated using 131 presence/absence locations from the evaluation dataset ( Fig 2A ) . The predicted values for the evaluation data are presented in the Supporting Information ( S6 Fig ) . The main predictors for P . knowlesi infection risk were urban accessibility , human population density , elevation , proportional cover of land with croplands and environmental suitability for M . nemestrina . Marginal effect plots for each of these covariates are displayed in the Supporting Information ( S7 Fig ) .
Using a niche model informed by a spatial database of P . knowlesi infections , a range of environmental and socio-economic data , and reservoir and vector species distributions , we have produced the first map of the predicted geographical distribution for P . knowlesi malaria . Empirical data on P . knowlesi presence or absence is lacking for most of Southeast Asia and this map provides an initial evidence base to prioritize areas for disease surveillance and future epidemiological investigations . The predictive performance of the model was high and it also had a high capacity to predict suitability for infection in regions outside MBS . The latter result should , however , be treated with caution as data for model evaluation was only available from a limited number of locations outside MBS , and the selection of locations for which P . knowlesi has been tested is likely to be biased . Another important caveat is the large area to which model predictions were made , relative to the model training region , since this required the model to extrapolate to some novel environments ( see Fig 3C ) . Extrapolated predictions are inherently less reliable than those made in areas of interpolation and include large parts of continental Asia . Sampling for P . knowlesi infections in areas of extrapolation is likely to have the biggest impact on improving the disease risk predictions . The final map therefore shows the risk of zoonotic P . knowlesi transmission from known reservoirs ( specifically M . nemestrina and M . fascicularis ) and vectors of the Anopheles leucosphyrus Group . If human-to-human transmission were occurring , this form of the disease is likely to have a different niche to the zoonotic disease , i . e . a different relationship with environmental , socioeconomic and biological factors . Thus our model is not appropriate to predict human-to-human transmission risk . It is also important to note that the limits of zoonotic transmission , within which we have predicted infection risk , were defined using the reservoir and vector ranges generated by our earlier work and these ranges reflect the fact that species distributions are not fixed . Specifically , these ranges included introduced populations of the two macaque species , for example , pet M . fascicularis and M . nemestrina macaques are commonly found on Sulawesi where the environment is predicted to be suitable for the establishment of feral populations [26] . The predictions for infection risk that we present here therefore include locations on this island . Human P . knowlesi infections have been identified beyond the ranges of both M . nemestrina and M . fascicularis . Macaca leonina , whose range extends farther north into Myanmar where these human cases were reported [58] , has thus been implicated as a putative host species . We allowed predictions within the range of M . leonina but since this species is not found in MBS , it was not used as an explanatory covariate for model fitting . This may have impacted model predictions in the most northern parts of our study area where the environmental suitability for M . nemestrina and M . fascicularis is low , but high for M . leonina . Again , sampling in these areas , particularly northern Myanmar , would improve the predictions . Furthermore , two distinct P . knowlesi parasite populations , linked to M . nemestrina and M . fascicularis respectively , have been identified in human patients from Malaysia [59] . It is reasonable to assume that only the strain associated with M . fascicularis is circulating and infecting humans in areas of continental Asia , where M . nemestrina is absent , and it may have a distinct relationship with environmental and socioeconomic variables compared to the mixture of parasite infections in patients from Malaysia . The presence of Leucosphyrus Complex vectors in Malaysia and Dirus Complex vectors in continental Asia [26] further adds to the possibility of different relationships between disease risk and the environment in these two regions . Comparing our predicted P . knowlesi risk map ( Fig 3A ) with the map of current sampling efforts ( Fig 2 ) , and the map of malaria eliminating countries ( Fig 3B ) , allows us to identify relative surveillance priorities for P . knowlesi . These include a number of regions in Thailand ( Phisanulok , Phetchuban , Chaiyaphum , Prachan Buri , and southern Nakhon Ratchasima ) and Vietnam ( Lam Dong , Phu Yen , Gia Lai , and Kon Tum ) . We also propose that further surveillance in previously sampled areas of Thailand , Vietnam and Cambodia is required to fully understand the distribution of P . knowlesi in countries close to eliminating the human malarias . Among the countries next expected to eliminate the human malarias , our results highlight a need for surveillance in un-sampled , high-risk areas in Myanmar , Laos , and Sumatra and Kalimantan in Indonesia . Initial studies have reported cases in Aceh on Sumatra [13] , and South and Central Kalimantan [60 , 61] but no published reports are available from the other parts of these regions . Further surveillance is also needed in previously sampled areas , including Palawan in the Philippines . Importantly , our map predicts the environmental suitability for infection , not the prevalence of infection or incidence of cases in these places . The higher numbers of reported cases in Malaysia is not proof that the disease risk is higher here because most locations outside Malaysia simply have not been surveyed and P . knowlesi could be misdiagnosed as one of the human malarias . Studies that have investigated numbers of cases or infections have sampled a wide array of communities including malaria patients [4] , patients diagnosed as P . malariae by microscopy [62] , and whole communities [63] , meaning the disease prevalence indicators generated are not directly comparable . Until more locations are surveyed using a consistent measurement ( ideally infections in a cross section of the community ) and diagnostics that distinguish all human malarias , we cannot draw any firm conclusions about relative disease prevalence [64] . Studies of other diseases have , however , found a relationship between the environmental suitability for infection and infection prevalence or case incidence [17 , 65] and this is a potential use of future iterations of this map . It will be important to update the predictions presented here when new data become available , and systems are available to generate updated predictive maps [66] . Importantly the map presented here provides key information about the locations where new surveys for P . knowlesi infections would be most informative . As the volume and quality of geographical data on P . knowlesi infections increases across Southeast Asia , these maps will iteratively improve . For now , the work presented here provides the best evidence-base currently available for prioritizing P . knowlesi surveillance to better understand its spatial distribution and its wider contribution to malaria cases . | Plasmodium knowlesi is a malaria parasite found in wild monkey populations and transmitted from this animal reservoir to humans via infected mosquitoes . It causes severe and fatal disease in humans , and is the most common cause of malaria in parts of Malaysia . The geographical distribution of this disease is largely unknown because it is often misdiagnosed as one of the human malarias . Human malaria parasites are primarily transmitted between humans via mosquitoes and are not frequently transmitted from other animals to humans . Many countries in Southeast Asia , where P . knowlesi infections have been reported , are making progress towards eliminating the human malarias . Understanding the geographical distribution of P . knowlesi is important for identifying areas where malaria transmission will continue after the human malarias have been eliminated . In locations that have high volumes of P . knowlesi infection data , we modelled patterns of variation in the data linked to environmental predictors , and used this to estimate P . knowlesi infection risk in locations where data is lacking . The resulting map represents an initial evidence-base for identifying areas of human disease risk that should be prioritized for surveillance , particularly in the context of malaria elimination in the region . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"plasmodium",
"population",
"dynamics",
"tropical",
"diseases",
"geographical",
"locations",
"vertebrates",
"parasitic",
"diseases",
"animals",
"mammals",
"parasitology",
"parasitic",
"protozoans",
"primates",
... | 2016 | Estimating Geographical Variation in the Risk of Zoonotic Plasmodium knowlesi Infection in Countries Eliminating Malaria |
Lipid rafts in eukaryotic cells are sphingolipid and cholesterol-rich , ordered membrane regions that have been postulated to play roles in many membrane functions , including infection . We previously demonstrated the existence of cholesterol-lipid-rich domains in membranes of the prokaryote , B . burgdorferi , the causative agent of Lyme disease [LaRocca et al . ( 2010 ) Cell Host & Microbe 8 , 331–342] . Here , we show that these prokaryote membrane domains have the hallmarks of eukaryotic lipid rafts , despite lacking sphingolipids . Substitution experiments replacing cholesterol lipids with a set of sterols , ranging from strongly raft-promoting to raft-inhibiting when mixed with eukaryotic sphingolipids , showed that sterols that can support ordered domain formation are both necessary and sufficient for formation of B . burgdorferi membrane domains that can be detected by transmission electron microscopy or in living organisms by Förster resonance energy transfer ( FRET ) . Raft-supporting sterols were also necessary and sufficient for formation of high amounts of detergent resistant membranes from B . burgdorferi . Furthermore , having saturated acyl chains was required for a biotinylated lipid to associate with the cholesterol-lipid-rich domains in B . burgdorferi , another characteristic identical to that of eukaryotic lipid rafts . Sterols supporting ordered domain formation were also necessary and sufficient to maintain B . burgdorferi membrane integrity , and thus critical to the life of the organism . These findings provide compelling evidence for the existence of lipid rafts and show that the same principles of lipid raft formation apply to prokaryotes and eukaryotes despite marked differences in their lipid compositions .
The spirochete Borrelia burgdorferi is the causative agent of Lyme disease [1] , [2] , a tick-borne illness that can have manifestations in the skin , heart , joints , and nervous system of mammals [3] . B . burgdorferi has outer and inner membranes , and the periplasmic space between these membranes contains flagellar bundles . The flagella contribute to B . burgdorferi morphology [4] , and are not exposed to the extracellular environment unless the outer membrane is damaged [3] , [5] . B . burgdorferi membranes contain phosphatidylcholine , phosphatidylglycerol and lipoproteins [6]–[8] . They also contain free cholesterol , two cholesterol glycolipids ( acylated cholesteryl galactoside ( ACGal ) and cholesteryl galactoside ( CGal ) ) , and the glycolipid monogalactosyl diacylglycerol ( MGalD ) [9]–[12] . Only a few other bacteria are known to incorporate cholesterol into their membranes [13]–[17] . In eukaryotic cells , sterols ( together with sphingolipids having saturated acyl chains ) are believed to participate in the formation of ordered membrane domains called rafts , which co-exist with disordered membrane domains , and which are thought to play an important role in many membrane functions [18]–[23] . In model membranes , ordered sterol-rich domains are readily detected [24] . However , it has been difficult to characterize rafts in eukaryotic cells due to their small size and dynamic properties , and their existence remains controversial . We previously presented evidence that lipid microdomains containing cholesterol glycolipids exist in B . burgdorferi membranes [25] . In this study we demonstrate that the formation of these domains have all the hallmarks of lipid rafts and that the domains are present in living B . burgdorferi .
The hypothesis that B . burgdorferi domains are lipid rafts predicts that their formation should require lipids having the ability to form tightly packed domains . In previous studies we demonstrated that different sterols have a structure-dependent range of abilities to support formation of ordered raft lipid domains in model membrane vesicles [26]–[29] . Therefore , sterol substitution experiments were carried out in B . burgdorferi using sterols ( Table S1 in Text S1 ) ranging from those that are strongly ordered domain forming to those that are ordered domain inhibiting [26]–[29] . Free cholesterol and cholesterol glycolipids from B . burgdorferi can be substantially removed from cells with methyl-β-cyclodextrin ( MβCD ) while phospholipids and MGalD are unaffected [25] . When depletion is followed by incubation of the spirochetes with a diverse set of sterols , thin layer chromatography ( TLC ) analysis of B . burgdorferi lipid extracts indicated that sterol substitution had taken place ( Fig . S1 in Text S1 ) . Sterol substitution was confirmed by a strong correlation between the ability of a sterol to support ordered domain formation in model membranes [26]–[29] and membrane order in B . burgdorferi membranes ( in intact cells ) , as measured by the anisotropy of trimethylaminodiphenylhexatriene ( TMADPH ) fluorescence , subsequent to sterol substitution ( Table S1 and Fig . S2 in Text S1 ) . After sterol substitution , B . burgdorferi were prepared for immunogold negative stain TEM analysis to determine the effect of substitution upon cholesterol glycolipid-containing membrane microdomain formation ( Fig . 1 , Fig . S3 in Text S1 ) . For this , TEM grids were probed with a rabbit antibody to asialo GM1 , as this antibody cross-reacts with the cholesterol glycolipids of B . burgdorferi [25] , [30] , [31] . The ultrastructural appearance of the membrane microdomains in spirochetes whose native cholesterol was substituted with sterols that strongly promote lipid raft formation ( ergosterol , cholesterol , dihydrocholesterol , and stigmasterol ) was indistinguishable from the appearance of those in normal , untreated B . burgdorferi ( Fig . 1A and Fig . S3 in Text S1 ) [25] . Distinct clusters were observed with a diameter of ∼50 nm , similar to those seen in B . burgdorferi prior to sterol substitution [25] . This is consistent with the raft-promoting behavior of these sterols in model membranes [26] , [27] . Substitution of native cholesterol with sterols with intermediate raft-forming ability ( desmosterol , lanosterol , zymosterol and cholesterol formate ) also resulted in visually apparent glycolipid microdomains ( Fig . 1A and Fig . S3 in Text S1 ) . However , in the case of lanosterol and zymosterol unclustered sterol glycolipid molecules could also be seen ( Fig . 1A ) . Decreased microdomain organization was observed in cholesterol formate substituted spirochetes ( Fig . S3 in Text S1 ) , and was lost when B . burgdorferi were substituted with coprostanol or androstenol ( Fig . 1A ) , sterols that inhibit raft formation . The clustering of gold particles in the micrographs was also analyzed using Ripley's K-function ( Fig . 1B and Fig . S4 in Text S1 ) [32] . The clustering parameter L ( r ) -r is a measure of the excess of particles , relative to that expected if the distribution is random within a region radius r around each particle . Values significantly higher than zero indicate clustering , while a value of zero indicates random particle distribution within the distance r around each particle . The results were in agreement with visual assessment of ultrastructure , with clustering ( L ( r ) -r ) values above the 95% confidence limit for all sterols with the exception of coprostanol and androstenol , which showed no evidence of clustering , and cholesterol formate , which showed very weak clustering values near the confidence index . The y-axis values peak at points corresponding to domain radii close to 20–40 nm . We conclude that a sterol having the ability to form ordered domains is both necessary and sufficient for that sterol to induce TEM-detected domain formation in B . burgdorferi . To rule out the possibility that membrane domains are induced by the fixation used to prepare specimens for immunogold TEM analysis , a fluorescence resonance energy transfer ( FRET ) method was developed to detect domains in living cells . In this method , weak FRET is observed when co-existing ordered and disordered lipid domains are present because the donor , TMADPH , partitions moderately into ordered domains and so partially segregates from the acceptor , octadecylrhodamine B ( ODRB ) which partitions preferentially into disordered lipid domains [33] , [34] . The FRET method was calibrated in model membranes ( Fig . 2A ) . At low temperatures , there was higher TMADPH fluorescence ( higher F/Fo = weaker FRET ) in vesicles having a composition ( dipalmitoylphosphatidylcholine ( DPPC ) /dioleoylphosphatidylcholine ( DOPC ) /cholesterol ) in which segregation into DPPC-rich ordered and DOPC-rich disordered domains occurs [35] than in vesicles ( DOPC/cholesterol or palmitoyloleoylphosphatidylcholine ( POPC ) /cholesterol ) forming a homogeneous bilayer . At higher temperatures , in DPPC/DOPC/cholesterol samples lipid segregation is lost due to melting of the ordered domains , and FRET levels come close to that in homogeneous vesicles , which exhibit temperature-independent FRET . When TMADPH and ODRB were added to living B . burgdorferi cells , weak FRET which increases at higher temperatures is observed ( Fig . 2B and Fig . S5 in Text S1 ) . This indicates that there is formation of co-existing ordered and disordered domains up to at least ∼35–40°C . The FRET experiments were repeated after sterol substitutions ( Fig . 2B ) . Weak FRET which increased at higher temperatures was observed with sterols that strongly or moderately support ordered domain formation . In contrast , strong , temperature-independent FRET was observed after substitution with androstenol and coprostanol , sterols that do not support ordered domain formation , and for which domains were not observed in TEM . Thus , FRET-detected domain formation in living cells has the same dependence upon sterol structure as does TEM-detected domain formation , confirming that the latter is not an artifact of fixation . Partial depletion of cholesterol lipids with MβCD without subsequent sterol substitution did not totally abolish domain formation at lower temperatures , but did decrease their thermal stability , as shown by the transition from weak to strong FRET occurring at a lower temperature ( Fig . 2B ) . This indicates that residual cholesterol lipids remaining after extraction are sufficient to form some ordered microdomains . The ability to form ordered domains was evaluated next by isolation of detergent resistant membranes ( DRM ) from sterol-substituted B . burgdorferi cells . We previously showed that DRM containing cholesterol glycolipids , and the lipid-anchored proteins outer surface protein A ( OspA ) and outer surface protein B ( OspB ) , can be isolated from B . burgdorferi [25] . Resistance to solubilization by detergent is a characteristic property of ordered state domains in membranes [36] , [37] . Disordered lipid domains dissolve in detergents such as TX-100 while cholesterol-rich ordered domains do not . Although it has been claimed that the detergent TX-100 can induce domain segregation under some conditions , we recently found that this does not occur , and instead the effect of detergent is to coalesce pre-existing ordered domains into larger domains [38] . Ergosterol and cholesterol gave the highest yield of cholesterol glycolipids in DRM while androstenol and coprostanol , which inhibit domain formation , gave the lowest yield of DRM ( Fig . 3A ) . Lanosterol and zymosterol , with an intermediate ability to support ordered domain formation gave intermediate DRM levels . The levels of cholesterol glycolipids in the TX-100-soluble fractions showed an inverse pattern ( Fig . 3B ) . An analogous pattern for the sterol structure-dependence of DRM association was found for OspA and OspB levels in DRM and soluble fractions ( Fig . S6 in Text S1 ) . It should be noted that although DRM are prepared at 4°C , as is conventional [36] , DRM containing cholesterol glycolipids , plus both OspA and Osp B , could also be isolated from B . burgdorferi at 33°C ( Fig . S7 in Text S1 ) . Because B . burgdorferi is exposed to a wide range of ambient temperatures in ticks and mammals , both 4°C and 33°C are physiologically relevant . Lipids having only saturated ( e . g . palmitoyl ) acyl chains tend to associate strongly with ordered lipid domains while lipids with acyl chains having cis double bonds ( e . g . oleoyl ) , which act as kinks , pack poorly into , and so do not associate well with , ordered domains [19] , [39] . To determine if B . burgdorferi membrane domains specifically associate with lipids bearing saturated acyl chains , their association with biotin-polyethyleneglycol-dipalmitoylphosphatidylethanolamine ( biotin-PEG-DPPE ) and biotin-polyethyleneglycol-dioleoylphosphatidylethanolamine ( biotin-PEG-DOPE ) [40] was compared ( Fig . 4 ) . Previous studies showed that biotin-PEG-DPPE has a significant affinity for ordered membrane domains while biotin-PEG-DOPE partitions more strongly into disordered domains [40] . The association of the biotin lipids to the cholesterol glycolipid domains was evaluated by negative-stain immunogold TEM using antibodies to biotin . Biotin-PEG-DPPE colocalized with or adjacent to cholesterol glycolipid domains ( Fig . 4A , left ) , while biotin-PEG-DOPE did not ( Fig . 4A , right ) . Analysis of the distances between biotinylated and cholesterol glycolipids confirmed visual results ( Fig . 4B ) . This behavior confirms that the cholesterol glycolipid domains seen by TEM contain lipids in an ordered state . The association of biotin-PEG lipids with B . burgdorferi DRM was also measured . DRM and soluble fractions were probed with anti-biotin antibodies . Biotin-PEG-DPPE was found to partition primarily into B . burgdorferi DRM fractions while most of the biotin-PEG-DOPE partitioned into the soluble fraction ( Fig . 4C ) . These results indicate that B . burgdorferi DRM also have properties of ordered lipid domains . To test whether sterol substitution can affect other membrane properties , sterol-substituted B . burgdorferi were analyzed 5 h after sterol substitution by negative-stain TEM to assess sterol-dependent changes in spirochete morphology ( Fig . 5 ) . When substituted with strongly raft-forming sterols , B . burgdorferi cells showed normal planar wave morphology with a smooth surface and no release of periplasmic flagella ( Fig . 5 ) . Spirochetes substituted with sterols having an intermediate ability to support raft formation showed modest loss of planar wave morphology and with some of these sterols ( lanosterol , zymosterol and cholesterol formate ) membrane vesicles were observed protruding from the cells ( Fig . 5 ) . Spirochetes substituted with sterols that inhibit ordered domain formation completely lost their planar wave shape , forming straight cells or occasionally ( not shown ) a tightly curled morphology . They also exhibited greater formation and release of membrane vesicles , and exposure/release of periplasmic flagella into the external environment , suggesting loss of outer membrane integrity ( Fig . 5 ) . Spirochetes treated with MβCD but without sterol substitution coiled and showed a lesser release of periplasmic flagella after 5 h ( Fig . 5 ) . These experiments show that sterols having raft-forming properties contribute to the characteristic morphology of B . burgdorferi . To evaluate membrane integrity in more detail , the effect of sterol substitutions on B . burgdorferi membrane permeability was evaluated using propidium iodide . When membrane permeability increases or membrane integrity is lost , propidium iodide binds to DNA and becomes fluorescent . Substitution with sterols that strongly promote raft formation , as well as desmosterol , did not produce an increase in incorporation of propidium iodide by B . burgdorferi cells over a 5 h time period ( Fig . 6A ) . Other than desmosterol , sterols with an intermediate or inhibitory effect on raft formation caused a nearly linear increase in propidium iodide staining between 1 h and 5 h , as did sterol-depletion ( Fig . 6A ) , indicating increased membrane permeability . As propidium iodide staining does not distinguish between an increase in membrane permeability and a total loss of membrane integrity ( which would lead to spirochete death ) , B . burgdorferi substituted with different sterols were evaluated for protein release 5 h after sterol substitution . Protein release ( including that of cytoplasmic chaperone DnaK and FlaB ( from periplasmic flagella ) was not detected after substitution with strongly or moderately raft-forming sterols while the converse was true for raft-inhibiting sterols ( Fig . 6B and C ) . Thus , sterols with a strong raft-forming ability are necessary to maintain fully normal membrane impermeability , while either a strong or moderate raft-forming ability are necessary and sufficient to prevent total loss of membrane integrity and spirochete death ( see Discussion ) . This loss of membrane integrity appears to involve osmotic lysis ( data not shown ) as protein release was prevented in the presence of an osmoprotectant dextran500 [41] .
Our prior study showed membrane domains existed in fixed B . burgdorferi cells , and that cholesterol lipids played a role in their formation , suggesting they would be analogous to the lipid rafts proposed to form in eukaryotic cells . However , we did not show that these domains shared the characteristic properties of eukaryotic lipid rafts , or that they formed in living B . burgdorferi . The studies in this report show that B . burgdorferi membrane domains are true lipid rafts . Having raft-forming properties ( based on prior studies with eukaryotic sphingolipids [26]–[29] ) was both necessary and sufficient for sterols to support the formation of B . burgdorferi membrane domains as judged by TEM , detergent-resistance , and FRET . The observation of ordered domain formation by FRET was especially important as it indicates that domains form in live B . burgdorferi . In addition , membrane domains selectively incorporated molecules membrane-anchored via saturated acyl chains , but not analogous molecules having unsaturated acyl chains , the other key characteristic property of eukaryotic ordered lipid raft domains . Given the controversy concerning lipid rafts , it is noteworthy that these approaches may be useful for demonstrating the formation of membrane domains in other organisms , including eukaryotic cells . The use of complementary domain detection techniques involving different principles greatly reduces the possibility that experimental artifacts can explain these results . The possibility of artifacts is also greatly reduced by the use of a large set of sterols to establish a very strong correlation between membrane domain formation and raft forming physical properties , because it tends to rule out alternate explanations based on other variables ( e . g . a specific chemical feature found on one particular sterol ) . It is important that membrane domain properties in sphingolipid and sterol containing model membranes and those proposed for eukaryotic cells appear to be very similar to those of B . burgdorferi domains . Yet B . burgdorferi has no sphingolipids . The cholesterol glycolipid ACGal , which has been identified as being enriched in DRM from both B . burgdorferi [25] and plant cells [42] , is likely to play a role that combines that of sphingolipids and sterols in eukaryotes . However , we cannot rule out some role for proteins or other lipids in B . burgdorferi . The FRET protocol devised to detect lipid domain segregation in B . burgdorferi may be particularly useful for raft detection in other systems . TMADPH and ODRB are charged hydrophobic probes that readily incorporate into membranes when added externally to cells . Domains larger than Ro for the TMADPH/ODRB pair ( ∼23 Å calculated as described previously , and assuming all ODRB is membrane bound [38] ) can be detected with these probes . It is unlikely that the fluorescent probes used altered domain formation significantly . In addition to the consistency of FRET results with those obtained with other methods , FRET in untreated cells after halving ODRB concentration showed a temperature dependence similar to that observed at higher ODRB concentrations , and other controls showed that the fluorescent probes did not alter domain formation detected by TEM , cell morphology , or cell growth ( not shown ) . The parallels between the effects of sterol structure on both cell morphology and membrane integrity were striking . Not only were sterols with the ability to form ordered membrane domains both necessary and sufficient to maintain normal morphology and membrane integrity , but also sterols with an intermediate ability to support ordered membrane domain formation tended to show an intermediate disruption of impermeability and normal morphology . This is consistent with the conclusion that B . burgdorferi morphology and membrane integrity require the same sterol properties that support ordered membrane domain formation . Thus , the most obvious interpretation is that the formation of ordered domains is somehow required for membrane integrity . Alternately , these changes could be due to an overall decrease in how tightly outer membrane lipids are packed after sterol substitution . At a minimum , the observation that diverse sterols have effects on membrane integrity that strongly correlate with their physical properties , but not with any specific chemical feature of sterols , tends to rule out alternate hypotheses based on specific chemical features of sterols , as noted above . The change in membrane morphology and integrity appears to be linked to a change in the normal spirochete flat wave morphology . This raises the question of how these changes are related . It is possible that the loss of the flat wave morphology , which is created by the force exerted by the flagella on the cell wall/inner membrane complex , is due , in turn , to the loss of an external force produced by the outer membrane with ordered domains on the flagella itself , thereby decreasing the flagellar force on the inner membranes . Alternatively , these data may suggest that there is a physical connection between the outer membranes and the flagella . In this interpretation , a putative linking protein between the structures or a cofactor needed for their interaction could be lost when lipid rafts are not formed in the outer membrane . The change in these interactions could compromise flagellar structure . Finally , there could be an interaction between the outer membrane and peptidoglycan or inner membrane that when altered affects flagella . It is even possible that the inner leaflet has rafts perturbed by sterol substitution . Physical manipulations of the specimens are an unlikely explanation of the altered morphology , as all the experiments with the different sterols were done in the same manner and some sterols altered morphology whereas others did not . Based on this , we explain the release of proteins associated with the change in membrane permeability as due to outer membrane disruptions ( resulting in detection of released flagellar subunits ) followed shortly by the death of the spirochete ( resulting in detection of released cytoplasmic DnaK ) . We think that the initial changes in the outer membrane eventually render the bacteria sensitive to a secondary osmotic lysis event . Although this behavior by itself is not informative about the mechanism of raft or sterol interactions with other parts of the bacteria , it does tell us that the changes associated with substitutions involving sterols that cannot properly support raft formation can have catastrophic consequences for the bacterium , and therefore are functionally important . The observation that altering lipid composition lead to deleterious changes in B . burgdorferi could have biomedical implications , given the large fraction of antibiotics that act by attacking bacterial membrane integrity . Development of drugs interfering with B . burgdorferi cholesterol glycolipid synthesis or sterol uptake may represent a novel therapeutic approach to combat B . burgdorferi infections and that of other pathogenic cholesterol-containing bacteria/microorganisms . In this regard , it is interesting that a similar dependence upon sterol physical properties has been observed for maintenance of membrane integrity after freezing and thawing in yeast cells [43] . Modestly raft-supporting zymosterol increased yeast cell membrane permeability while the strongly raft-forming ergosterol did not [43] . Additionally , model membranes depleted of sterols or containing lanosterol have been shown to have a greater permeability than those containing cholesterol [44] . That domain-forming sterols are necessary to maintain membrane integrity and viability of B . burgdorferi is not surprising from an evolutionary standpoint . The life cycle of Borrelia involves infection of a tick vector or mammalian host , as these organisms are never free-living [3] . The cells of mammals and many species of ticks contain cholesterol , so it is logical to assume that B . burgdorferi has evolved a preference for domain-forming sterols , like cholesterol , as these would be the predominant sterols available during the spirochete life cycle .
1 , 2- dipalmitoyl-sn-glycero-3-phosphatidylcholine ( DPPC ) , 1-palmitoyl-2-oleoyl-phosphatidylcholine ( POPC ) , 1 , 2- dioleoyl-sn-glycero-3-phosphatidylcholine ( DOPC ) , zymosterol , and cholesterol were purchased from Avanti Polar Lipids ( Alabaster , AL ) . Other sterols were purchased from Steraloids Inc . ( Newport , RI ) . Lipids were stored in chloroform or ethanol at −20°C . ( Ethanol was the solvent for sterols used for substitution in cells whereas chloroform was solvent for sterols used for the model membrane experiments . ) Lipid concentrations were determined by dry weight . The fluorescent probes 1- ( 4-trimethylammonium ) -6-phenyl-1 , 3 , 5-hexatriene p- toluenesulfonate ( TMADPH ) and octadecyl rhodamine B ( ODRB ) were purchased from Molecular Probes division of Invitrogen , stored in ethanol at −20°C and concentrations determined by absorbance using ε = 84 , 800 M−1 cm−1 at 353 nm and 125 , 000 M−1 cm−1 at at 555 nm respectively . 1 , 6-Diphenyl-1 , 3 , 5-hexatriene ( DPH ) was purchased from Sigma-Aldrich ( St Louis , MO ) and its concentration determined by absorbance using ε = 84 , 800 M−1 cm−1 at 353 nm in ethanol . High performance thin layer chromatography ( HP-TLC ) plates ( Silica Gel 60 ) were purchased from VWR International ( Batavia , IL ) . HP-TLC analysis of 10 µg samples of sterols chromatographed in 1∶1 ethyl acetate/hexane [28] showed at most minor impurities after long term storage except in the case of stigmasterol , and especially ergosterol . However , ergosterol did not show impurities prior to storage , and anisotropy measurements on sterol-substituted cells confirmed that ergosterol retained its ability to support a high degree of membrane order after addition to cells . Analysis of total Borrelia lipids by TLC was performed as described previously [25] . Borrelia burgdorferi strain B31 was used for all studies and grown in BSK-H medium ( Sigma ) under microaerophillic conditions at 33°C . To substitute different sterols for cholesterol in live B . burgdorferi for various studies the following procedure was used . Cultures of 50 ml with spirochetes grown to about 1×108 cells/ml were pelleted from the BSK-H culture medium by centrifugation at 5000× g for 10 min at room temperature . The pellet was resuspended in 50 mL Hank's Balanced Salt Solution ( HBSS ) and washed twice . After the last wash , the pellet was resuspended in 48 . 5 mL of HBSS to which 2 . 5 mL of 200 mM MβCD ( Invitrogen ) in PBS was added for a final concentration of 10 mM MβCD . After incubation for 30 min at 33°C , the spirochetes were centrifuged and resuspended in HBSS . For propidium iodide assays and spirochetes intended for growth-recovery assays the cell concentration was 1×108 spirochetes/mL . For TEM analysis the cell concentration was 5×106 spirochetes/mL . For SDS-PAGE and Western blot analysis of supernatants , FRET analysis of live B . burgdorferi , and isolation of DRM from whole B . burgdorferi the cell concentration was 4×108 spirochetes/mL . Once spirochetes were resuspended in HBSS , sterols stored in 100% ethanol ( at 2 mg/ml ) or an equivalent amount of ethanol were warmed to room temperature and added to the bacteria at a final sterol concentration of 10 µg/ml and incubated at 33°C , generally for 30 min before proceeding to the next step , unless otherwise noted . For studies investigating osmotic lysis of B . burgdorferi was investigated , spirochetes were treated in the same manner with the exception of the osmoprotectant , dextran T500 ( 6% w/v , Pharmacia ) , being present in the HBSS . B . burgdorferi were tested for differences in membrane permeability following sterol substitutions by propidium iodide staining . Sterol-substituted spirochetes at a concentration of 1×108 spirochetes/mL incubated at 33°C in HBSS were stained with propidium iodide for 15 minutes at 33°C using the LIVE/DEAD BacLight Bacterial Viability kit ( Invitrogen ) according to the manufacturer's directions , at 45 min , 2 h 15 min or 4 h 45 min after . following the initial sterol substitution step . Propidium iodide fluorescence was measured in a SpectraMax M2 plate reader ( Molecular Devices ) at an excitation of 535 nm and an emission of 617 nm . Sterol-substituted B . burgdorferi were analyzed for protein release , a marker of spirochete death , by SDS-PAGE and Western blot analysis of supernatants from sterol-substituted B . burgdorferi at a concentration of 4×108 spirochetes/mL incubated in HBSS at 33°C for 5 hours . Supernatants were run on 12 . 5% SDS-PAGE gels and either stained with Coomassie blue R-250 for total protein staining or transferred to nitrocellulose for Western blot analysis . Blots were probed with the monoclonal antibodies CB 312 ( anti-B . burgdorferi DnaK , mouse IgG ) or CB1 ( anti-B . burgdorferi FlaB , mouse IgG ) followed by secondary anti-mouse IgG infrared conjugate ( IR800 , Rockland Immunochemicals ) . Blots were read in an Odyssey Infrared Scanner ( LI-COR ) at a wavelength of 800 nm . Detergent-Resistant Membranes ( DRM ) were isolated from sterol-substituted B . burgdorferi at a concentration of 4×108 spirochetes/mL using the Caveola/Rafts Isolation Kit ( Sigma ) as previously described [25] . DRM and soluble fractions were analyzed by both slot blot ( Hoefer ) and ELISA . For slot blots , fractions were loaded onto nitrocellulose and probed with anti-asialo GM1 ( for sterol glycolipids , rabbit polyclonal IgG , AbCam ) or monoclonal antibodies CB2 ( anti-OspB , mouse IgG ) or CB10 ( anti-OspA , mouse IgG ) [45] , [46] . Anti-Rabbit IgG IR700 conjugate or anti-mouse IgG IR800 conjugate were used as secondary antibodies . Blots were read in an Odyssey Infrared Scanner ( LI-COR ) at wavelengths of 700 and 800 nm . DRM and soluble fractions were also analyzed by ELISA by combining fractions with coating buffer ( 50 mM Na2CO3 , 50 mM NaHCO3 , pH = 9 . 6 ) , to coat the wells of ELISA plates at 4°C overnight . After blocking with 2% BSA , wells were probed with the antibodies mentioned above , except that anti-rabbit IgG or anti-mouse IgG alkaline phosphatase conjugates ( Sigma ) were used , followed by development with p-nitrophenyl phosphate ( pNPP , Sigma ) measured at 405 nm in a SpectraMax M2 plate reader . The lipid probes biotin-PEG-DPPE and biotin-PEG-DOPE [40] were used to label whole B . burgdorferi by adding aliquots from 10 mM ethanol stocks solutions of the probes to live spirochetes in HBSS to a final concentration of 60 µM . After addition of the probes , spirochetes were incubated at 33°C for 1 h . Sterol-substituted or biotin lipid probe-labeled B . burgdorferi at a concentration of 5×106 spirochetes/mL were adhered to grids , fixed in 1% ( v/v ) glutaraldehyde , and washed as described previously [41] to prepare organisms for negative-stain TEM analysis , and probed with anti-asialo GM1 ( rabbit IgG ) plus in some cases anti-biotin ( mouse IgG ) ( AbCam ) for 1 hour . Anti-rabbit IgG and anti-mouse IgG colloidal gold conjugates ( 6 nm and 15 nm , respectively; Jackson Immunochemicals ) were used as secondary antibodies . Following labeling with colloidal gold conjugates , grids were stained with PTA and analyzed by TEM as previously described [41] . To quantify the clustering of the cholesterol glycolipids , we analyzed the TEM micrographs using a common linear transformation of Ripley's K-function [47]–[49]where N ( r ) is the number of particles within a radius r of a given particle and D is the average particle surface density . When the particles are randomly distributed , L ( r ) -r = 0±the confidence interval ( CI ) for all values of r . If the function is positive for any value of r , the particles are clustered on that length scale . The confidence index ( CI ) was estimated by simulating 100 random particle distributions ( on a total surface and with D comparable to those of the examined samples ) and calculating each time L ( r ) -r . The CI ( r ) was then determined as two times the standard deviation from the average simulated L ( r ) -r curve . To avoid edge effects [47] we analyzed the number of neighbors only for particles that were >80 nm away from the edge of the image of the bacterial cell . Each image measured was chosen at random and covers approximately 400 nm along the length of the spirochete . For each sterol , we analyzed three different micrographs , one from each of three different bacteria , each bacterium being from a different batch and separate sterol substitution experiments . To quantify the co-localization of cholesterol glycolipids antibodies ( linked to “small” gold particles ) and biotinylated lipids ( linked to “large” gold particles ) , a slightly different analysis was used . We defined a co-localization parameter C40 . We first counted the number of small gold particles ( n40 nm ) within a radius of 40 nm of each large particle . C40 was then calculated by normalizing n40 as a percentage of the total number of small particles ( N ) , i . e C40 = n40 nm/N×100% . All quantitative image analyses and simulations were performed using self-written algorithms ( Matlab , The MathWorks , Natick , MA ) Sterol-substituted or sterol-depleted B . burgdorferi at a concentration of 4×108 spirochetes/mL were analyzed for lipid composition by extracting lipids after pelleting , using chloroform-methanol ( 1∶2 v/v ) [50] . Lipid extracts were resolved on HPTLC silica plates ( EM separations ) chromatographed in chloroform-methanol ( 85∶15 ) and stained with iodine vapor . Lipid profiles were analyzed using known Rf values from identical solvent systems [9] , [10] , [25] . Small unilamellar vesicles ( SUV ) were prepared by ethanol dilution in a manner similar to that described previously [38] . Lipids stored in chloroform at −20°C were warmed to 23°C , pipeted into glass tubes , mixed and dried under N2 , dissolved in 20 µL ethanol and dispersed in 980 µL 10 mM Na2HPO4 , 1 mM KH2PO4 , 137 mM NaCl and 2 . 7 mM KCl , pH 7 . 4 at 70°C . SUV contained 100 µM lipid and were incubated at room temperature for 1 h . Fluorescence was measured on SPEX Fluorolog 3 spectroflorimeter ( Jobin-Yvon , Edison , NJ ) using quartz semimicro cuvettes ( excitation path length 10 mm and emission path length 4 mm ) . Slit-width band-widths were set to 4 nm ( 2 mm physical size ) for excitation and emission . TMADPH fluorescence was measured at an excitation λ of 358 nm and emission λ of 430 nm , and values corrected for background fluorescence . Absorbance was measured using a Beckman 640 spectrophotometer ( Beckman Instruments , Fullerton , CA ) using quartz cuvettes . DPH fluorescence anisotropy measurements were made using SPEX automated Glan-Thompson polarizer accessory with slit-width band-widths set to 4 . 2 nm ( excitation ) and 8 . 4 nm ( emission ) . A 1 . 25 µL aliquot of ethanolic DPH from a stock solution of 198 µM was added per mL of cells in HBSS ( at 107 cells/mL ) , and incubated for 10 min at 24°C and immediately after stirring samples anisotropy was measured . Spirochetes or sterol substituted spirochetes at a concentration of 4×108/mL in HBSS were used for the FRET measurements . An aliquot of 5 . 2 µL ( from a 73 µM ethanolic stock solution ) of FRET donor TMADPH was added to 4 mL of spirochetes , mixed well by stirring and incubated at room temperature for 10 min . The spriochetes were divided into four 900 µL aliquots , placed in quartz cuvettes and heated to 35°C . Donor fluorescence intensity of all four samples was measured before adding acceptor . Two samples out of the four were defined as F samples , and to them 5 . 2 µL ( out of a 322 µM ethanolic stock solution ) of the acceptor , octadecyl rhodamine B ( ODRB ) , was added , followed by incubation at 35°C for 15 min . FRET measurements were then initiated on the F samples and Fo samples ( the two samples containing donor but not acceptor ) . Fluorescence in background samples lacking donor and ones lacking donor but containing acceptor was also measured . The temperature of the samples was measured using a probe thermometer placed in a cuvette . ( Fisher brand traceable digital thermometer with a YSI microprobe , Fisher Scientific ) . The cuvette temperature was slowly decreased from 35°C to 15°C in steps of 5°C and the cells were incubated at each temperature for 7 min once the temperature stabilized . Donor fluorescence intensity of Fo and F samples was measured at each temperature and the ratio of TMADPH fluorescence intensity in the presence of acceptor to its absence ( F/Fo ) was calculated and plotted as a function of temperature . The average values shown come from bacteria from two independent sterol substitution experiments . Background fluorescence was measured at 35 and 15°C and averaged for the two temperatures because they were found to be independent of temperature . The backgrounds were ≤2% of the TMADPH fluorescence signal and were subtracted from the FRET sample values . FRET measurement in model membranes was carried out similarly to that in cells with the following changes . FRET donor , 0 . 1 µM TMADPH was added to SUV prepared by ethanol dilution containing 100 µM lipid and was allowed to incubate at 50°C for 10 min . FRET acceptor , 3 µM ODRB , was added to the F samples and further incubated for 10 min after which the cuvettes were cooled to 45°C and fluorescence measurements were initiated . Cuvettes were further cooled slowly to 15°C in steps of 5°C increments and the fluorescence measured once the temperature stabilized . | Specialized domains ( “lipid rafts” ) rich in specific membrane lipids ( sphingolipids and cholesterol ) have been proposed to form in the cell membranes of higher organisms , and to be of functional importance . We recently found that domains can be detected in the membranes of the bacterium that causes Lyme disease , Borrelia burgdorferi . In this report it is shown that , despite a lack of sphingolipids in B . burgdorferi , these domains have all the characteristic properties of lipid rafts , and can be detected in living B . burgdorferi . This shows that true lipid rafts can form in bacteria . In addition , it is shown that sterols having a structure that promotes lipid raft formation are necessary and sufficient for those sterols to maintain B . burgdorferi membrane integrity . This is suggestive of a role for membrane domains in B . burgdorferi membrane integrity . Therefore , interfering with lipid raft formation may have biomedical applications in combatting B . burgdorferi infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biophysics",
"biochemistry",
"biology",
"microbiology"
] | 2013 | Proving Lipid Rafts Exist: Membrane Domains in the Prokaryote Borrelia burgdorferi Have the Same Properties as Eukaryotic Lipid Rafts |
Genetic information should be accurately transmitted from cell to cell; conversely , the adaptation in evolution and disease is fueled by mutations . In the case of cancer development , multiple genetic changes happen in somatic diploid cells . Most classic studies of the molecular mechanisms of mutagenesis have been performed in haploids . We demonstrate that the parameters of the mutation process are different in diploid cell populations . The genomes of drug-resistant mutants induced in yeast diploids by base analog 6-hydroxylaminopurine ( HAP ) or AID/APOBEC cytosine deaminase PmCDA1 from lamprey carried a stunning load of thousands of unselected mutations . Haploid mutants contained almost an order of magnitude fewer mutations . To explain this , we propose that the distribution of induced mutation rates in the cell population is uneven . The mutants in diploids with coincidental mutations in the two copies of the reporter gene arise from a fraction of cells that are transiently hypersensitive to the mutagenic action of a given mutagen . The progeny of such cells were never recovered in haploids due to the lethality caused by the inactivation of single-copy essential genes in cells with too many induced mutations . In diploid cells , the progeny of hypersensitive cells survived , but their genomes were saturated by heterozygous mutations . The reason for the hypermutability of cells could be transient faults of the mutation prevention pathways , like sanitization of nucleotide pools for HAP or an elevated expression of the PmCDA1 gene or the temporary inability of the destruction of the deaminase . The hypothesis on spikes of mutability may explain the sudden acquisition of multiple mutational changes during evolution and carcinogenesis .
The precise balance between genome stability and mutagenesis is vital for the survival of a species [1] , [2] , [3] . It ensures the maintenance of the optimal combinations and frequencies of alleles with high fitness and , simultaneously , the introduction of new mutations that are the raw material for the natural selection that drives adaptation in a changing environment . A wealth of data indicate that this balance shifts toward higher mutation rates during sub-optimal conditions , and then returns to normal levels ( [2] , [4] , [5] and references therein ) . Similar mechanisms have been proposed to explain the evolution of tumors [6] , [7] . Sequencing of cancer genomes shows that tumor genomes are highly enriched with mutations [8] , [9] . The accumulated mutation load cannot be explained by normal mutation rates and requires highly mutable cells ( [10] , [11]; reviewed in [12] ) . A stable mutator phenotype would inexorably reduce tumor fitness due to the accumulation of mutations in regulatory and essential genes . In order to account for this discrepancy , it has been hypothesized that the mutator phenotype in cancer is transient [13] , [14] . Spikes of hypermutability can be caused by epigenetic changes and/or the defective regulation of DNA repair and replication [6] , abnormally high expression of DNA editing deaminases [15] , [16] and other processes . Another layer of complexity is added by the fact that the mechanisms of the appearance of mutants are different in haploid and diploid organisms . In haploid cells , a mutation-causing defect of the gene product is expressed immediately . In diploid cells , a wild-type allele will mask a recessive mutation , and only the effects of dominant mutations will be observed ( Fig . 1 ) . For recessive mutations , the mutant phenotype will only be expressed in diploid cells when the second allele is inactivated . This can occur in various ways . First , either gene conversion or recombination between the mutated allele and the centromere will lead to a reduction to homozygosity . Second , chromosome loss or deletion of the region encoding the wild-type allele will result in a reduction to hemizygosity . Third , the wild-type allele may acquire an independent , typically heteroallelic mutation . The classic example illustrating the importance of two-step mutagenesis is Knudson's theory of retinoblastoma development via the inactivation of both alleles of a tumor suppressor gene [17] , [18] . If measured by phenotypic change , mutation frequency should be much lower in diploids than haploids ( Fig . 1 ) ; however , in yeast , it is only several-fold less ( [19] and references therein ) . Most mutagens act in yeast by a two-step mechanism involving mutation and segregation , because they induce a high frequency of recombination events [20] , while replication infidelity caused by non-recombinogenic base analogs or proofreading exonuclease defects somehow induces a high level of independent mutations in both homologs [21] , [22] . Most of our knowledge of the mechanisms of mutagenesis comes from classical studies in haploid models , such as E . coli , haploid yeast strains , or Drosophila germ cells . The molecular mechanisms of mutagenesis in diploid cells have not been studied in-depth . In this work , we induced mutations in isogenic haploid and diploid yeast using one of two different types of mutagens that generate non-canonical bases in DNA: the base analog 6-hydroxylaminopurine ( HAP ) , and ectopically produced editing cytosine deaminase PmCDA1 from sea lamprey . We have chosen these mutagens and genetic backgrounds to avoid an induction of recombination by mutagens . Yeast is characterized by high recombination . Our conditions were well-suited for study of mutagenesis more closely resembling the processes in human cells , when recombination is rare . HAP and PmCDA1 enhance replication infidelity and create a mutator phenotype on demand . HAP is incorporated during the growth in a media with analog and rapidly wiped out from cells after transfer to the medium without it . It is known that nucleotide pools are constantly and rapidly renewed in yeast cells [23] . The expression of PmCDA1 in our system is under the control of a regulatable promoter and could be turned on and off . After mutagenic treatment we selected forward mutants resistant to antibiotic canavanine or toxic drug 5-fluoroorotic acid ( 5-FOA ) and resequenced their genomes . This allowed for the determination of accumulated DNA sequence changes specific for each mutagen . The numbers of induced base substitutions were more than an order of magnitude higher in diploid mutants than in haploid mutants . The genomes of diploid clones treated with either mutagen but not selected for resistance also contained significantly less mutations than the diploid mutant clones . This indicates the heterogeneity in mutability between different cells and proves that selected mutants came from a fraction of cells that experienced the most dramatic mutagenesis . We call such cells hypermutable . Diploid hypermutated cells survived , because most of the induced mutations were recessive and did not result in phenotypic changes when heterozygous . Haploids with similar levels of mutagenesis die due to inactivation of essential genes . For the first time , to our knowledge , this work suggests that cells have a wide range of mutability in a genetically homogenous population of eukaryotic cells exposed to a mutagen . This may explain the rapid appearance of mutations ( mutation avalanches and recently discovered kataegis ) in evolution and disease progression , especially in sporadic cancer .
HAP is an adenine base analog that has an ambiguous base-pairing capacity . In imine form it can pair with thymine , whereas in its rarer amine form it pairs with cytosine . HAP is a universal mutagen that is active in most organisms , from humans to bacteria and their phages [24] , [25] . The conversion of HAP in cells to the corresponding deoxyribonucleotide triphosphate ( dHAPTP ) , followed by its incorporation into DNA by replicative polymerases , results in A-T to G-C and G-C to A-T transition mutations ( see Fig . 2A , 2B ) [26] , [27] , [28] , [29] , [30] . PmCDA1 belongs to the AID/APOBEC superfamily of editing deaminases [31] , [32] . These enzymes are found in different vertebrate species and perform a variety of functions , including immunoglobulin gene diversification ( AID ) , RNA editing ( APOBEC1 ) , restriction of retroviruses ( APOBEC3s ) , and possibly active DNA demethylation [33] , [34] , [35] , [36] . PmCDA1 is involved in the diversification of genes encoding immunoglobulin analogs in sea lamprey and is closely related to other APOBEC enzymes [31] . AID/APOBECs fulfill their functions by catalyzing cytosine deamination , which results in the formation of uracil in the substrate DNA or RNA . Uracil can then be processed by the base-excision repair pathway protein uracil-DNA-glycosylase , followed by repair , which may result in mutations and recombination . If uracil escapes repair during the next round of replication , a C-G to T-A transition occurs ( Fig . 2C ) [33] , [34] . Both HAP treatment ( reviewed in [24] ) and the ectopic production of PmCDA1 [31] , [37] are not very toxic but strongly mutagenic in wild-type yeast as measured by different reporter systems detecting base-pair substitutions . In contrast to the other organisms , HAP does not induce recombination in Saccharomyces cerevisiae [24] , [38] , [39] , most likely because a key enzyme required to excise HAP-containing DNA is absent in yeast . In addition , mismatch repair – one of the key safeguards of genome stability [40] – does not seem to recognize HAP in the DNA [24] , in contrast to the other base analogs such as dP [41] . These unique properties provide the opportunity to detect a genuine signature of base analog-induced mutations . PmCDA1 was chosen as a prototype of the AID/APOBEC1 family because it has the highest mutagenic effect in the group when produced in yeast [37] . PmCDA1 is recombinogenic in wild-type S . cerevisiae , but inactivation of uracil-DNA-glycosylase ( ung1 ) completely blocks deaminase-induced recombination [31] . Thus , HAP and PmCDA1 are perfect tools for studying mutagenesis in diploid cells under conditions when induced recombination is suppressed , by mechanism of the induction of independent mutations ( Fig . 1 , right panel ) . We examined the effect of ploidy on HAP- and PmCDA1-induced mutagenesis . The median frequency of canavanine-resistant mutants ( CanR , mutation in the CAN1 locus ) in the HAP-treated haploid strain LAN201 ( see Table S1 for genotypes of strains ) is 2 . 51*10−5 ( Table 1 ) , a 23-fold increase over the background level . The frequency of HAP-induced CanR mutants in the isogenic diploid strain LAN211 is 5*10−7 , 833-fold higher than expected based on the mutation frequency in the haploid strain ( both copies of the CAN1 gene have to be inactivated in order to produce an antibiotic resistant phenotype , Fig . 1 ) ( Table 1 ) . We did not find CanR clones in diploids in the absence of mutagen . This was consistent with previous observations that the spontaneous frequency of mutants in wild-type diploids is extremely low [21] , [42] . Overall , the results are in full agreement with our earlier genetic data on the mutagenesis of diploids with HAP with a different reporter , the LYS2 gene ( i . e . they are reporter-independent ) [21] . The expression of PmCDA1 in the ung1 ( uracil-DNA-glycosylase-deficient ) haploid strain LAN200 leads to a 22-fold increase in CAN1 mutagenesis over the background frequency ( 1 . 6*10−4 vs . 7 . 2*10−6 ) . Similar to HAP , the frequency of PmCDA1-induced mutations in the diploid ung1 strain LAN210 is much higher than expected based on the observed haploid rate ( 2 . 3*10−6 vs . 2 . 5*10−8 , a 92-fold increase ) ( Table 1 ) . Similar results has been obtained with the URA3 reporter gene ( mutants resistant to the 5-FOA ) . Frequency of FOAR mutants in diploid strain was much higher than predicted based on the measured frequency in the haploid strain ( see Table 1 ) . The viability of haploid and diploid cells treated with deaminase was 65% and 90% , respectively . High-throughput “next-generation” DNA sequencing ( NGS ) has revolutionized biomedical research . In order to better understand the phenomenon of an unexpectedly high mutation rate in diploid strains , we used NGS to determine the genome-wide spectra of mutations induced by HAP and PmCDA1 in yeast . To make the analysis of mutant clones possible , we first determined the sequences of the genomes of our wild-type strains . DNA from LAN201 , LAN211 , LAN200 and LAN210 ( Table S1 ) was extracted , sequenced on an Illumina HiSeq 2000 instrument , and reference genome sequences were de novo assembled from the sequencing data ( see Materials and Methods for details of sequencing and genome assembly ) . Since LAN201 and LAN211 — as well as LAN200 and LAN210 — are isogenic to each other , the sequences of their genomes were identical , with the exception of the MAT locus . However , the related UNG1 and ung1 strains ( LAN201 and LAN211 vs . LAN200 and LAN210 ) differ by seven single-nucleotide variations ( SNVs ) , in addition to disruption of the UNG1 gene by a cassette conferring hygromycin resistance ( Table S2 ) . Overall , the sequence of our LAN-specific reference genome contains 12 , 077 , 153 bp and covers 92 . 74% of the S288C nuclear genome . Other genome parameters , such as the number of genes and the GC percent , are similar between the LAN and S288C reference genomes ( Table S3 ) . Next , we resequenced the genomes of canavanine-resistant clones induced by HAP in LAN201 and LAN211 strains . Four haploid and 10 diploid genomes were sequenced . We detected numerous mutations in all 14 genomes ( Table 2 ) . All mutations detected in haploid clones have SNV frequencies of 80–100% . This confirms that all cells in the sequenced colony were derived from one mutated progenitor cell . Rare cases where SNV in haploid clones have a frequency between 40 and 80% were assembly errors ( see Materials and Methods ) . In diploid clones , most of the mutations are true heterozygous ( i . e . , frequencies of SNVs between 40 and 80% ) . Rarely , two or more SNVs in the same gene are found . They could be clustered mutations in one copy of the gene or changes in both copies , i . e . heteroallelic . Such cases in our reporter gene lead to a detectable phenotype due to the inactivation of both copies of the reporter and , therefore , were true heteroalleles . We cannot predict from the sequencing data whether the mutation will be recessive or dominant . However , most of the heterozygous mutations are expected to be recessive , because gain of function is a rather specific event . In addition , not all SNVs lead to the phenotypic changes ( also see section “Prediction of effects of multiple SNVs on viability” below ) . Therefore , it is likely that the functions of the majority of the genes with SNVs are not disrupted in diploid mutants , even in the cases where multiple SNVs were present . In the case of the CAN1 reporter gene , no dominant mutations have ever been reported , to our knowledge . This is expected because the resistance phenotype is due to the loss of function of Can1p , a one-subunit arginine permease ( www . yeastgenome . org ) . Because of the selection for the loss of function of permease in diploids , two copies of the CAN1 gene should be damaged ( Fig . 1 ) . The predominant mechanism of the appearance of CanR mutants was independent mutations in the two homologs . This results in heteroallelic mutations where both alleles are non-functional as nine clones out of 10 possessed heteroallelic mutations in the CAN1 gene . One clone had one homozygous mutation and is discussed below . Some mutations in the genomes were found with a SNV frequency of more than 80% and were classified as homozygous . The majority of these rare homozygous mutations in diploid clones apparently result from spontaneous recombination events ( see Table 2 ) . This includes the homozygous mutation in the CAN1 gene of clone LAN211-4 , which belongs to the group of 4 homozygous mutations localized on the distal end of the left arm of chromosome V and , therefore , is a result of a mutation-segregation mechanism via recombination ( Table 2 , Fig . 1 ) . The mutational load is strikingly different in the haploid and diploid clones ( P <0 . 005 , see Materials and Methods ) . Four haploid clones contain 54 to 356 mutations , whereas diploids had from 1020 to 1747 SNVs per genome ( Table 2 and Fig . 2D ) . The average number of SNVs per 100 Kb is 1 . 3 for haploids and 6 . 05 for diploids ( Table 2 ) . All mutations are A-T to G-C and G-C to A-T transitions , in agreement with the mechanism of HAP action during replication ( Table 2 and Table S4 ) . In most sequenced genomes , mutations in the G-C pairs were more abundant than mutations in the A-T pairs ( see right column in Table S4 and Fig . 2D ) , which is consistent with earlier data with specific reporters [26] , [43] . The bias toward mutations in G-C pairs suggests that most of the effects of dHAPTP are attributable to its misincorporation opposite C in the first replication cycle ( Fig . 2 ) . However , the variability of the ratio of mutations in the G-C pair to mutations in A-T pairs in individual genomes was high , from 0 . 5 in LAN211-1 to 5 . 3 in LAN211-7 . In particular , we observed a strong bias toward mutations in A-T pairs in one diploid HAP-induced mutant clone ( LAN211-1 ) . The reason for these differences is unknown and may reflect cell-to-cell variability in HAP metabolism and/or DNA replication ( see Discussion ) . This highlights the value of whole-genome resequencing studies , which provide a snapshot of the mutagenic process in individual cells . Analysis of the sequence context of these mutations did not reveal any strong biases toward any particular sequence contexts for HAP-induced SNVs ( Fig . 3 ) . However , we observed a slight preference for A/T rich sequences in our genome-wide data for both G-C to A-T and A-T to G-C transitions . Mutational spectra obtained using reporter genes shows different results depending on the substitution type and reporter used ( Fig . 3 , first column of consensus sequences; see Discussion ) . We sequenced four haploid CanR , seven diploid CanR and two diploid FOAR ( mutations in the URA3 locus confer resistance to 5-FOA ) mutant clones induced by PmCDA1 . Similar to the results obtained with HAP , all mutations in haploid PmCDA1-induced mutant strains have SNV frequency >80% , whereas the majority of mutations in diploid clones are heterozygous ( Table 3 ) . It is important to note that the average number of mutations in haploids was very close to what was found in yeast ung1 haploids after induction of hyper-active AID deaminase or APOBEC3B [44] . PmCDA1 induces slightly more homozygous mutants in diploids than HAP . As opposed to the results with HAP-induced diploid mutants ( see Table 2 ) , homozygous SNVs in PmCDA1-induced diploid mutants are mostly scattered throughout the genome and , therefore , are not due to recombination events . Even if the homozygous mutations were found very close to each other ( such as in the most hypermutable region on chromosome X , see Table S6 and [45] for details ) , they were always accompanied by the heterozygous SNVs in close proximity , and sometimes the heterozygous SNVs were found in between the homozygous ones . These data indicate that homozygous mutations in genomes of PmCDA1-induced mutants in diploids are unlikely to be due to recombination or gene conversion . It is plausible that regions of the genome that are prone to PmCDA1-dependent deamination can accumulate multiple independent mutations , sometimes leading to the homozygous SNVs . In the CAN1 reporter , heteroallelic mutations are present in six diploid CANR mutant clones , while only one mutant clone is homozygous . Both FOAR diploid clones possess heteroallelic SNVs in the URA3 reporter gene . Diploids accumulate more PmCDA1-induced SNVs than haploids ( 4 . 38 vs . 0 . 74 SNVs/100 Kb; 5 . 9-fold increase; p = 0 . 005 ) ; however , the variation of the number of mutations in PmCDA1-induced diploids is higher than in HAP-induced diploid clones ( Table 3 and Fig . 2D , 2E ) . All SNVs are C-G to T-A transitions , as expected from cytosine deamination ( Table S5 , Table 3 , and Fig . 2E ) . Interestingly , a small fraction of mutations ( about 0 . 6% ) are tandem , i . e . two consecutive cytosines or guanines are mutated ( CC→TT and GG→AA tandem transitions , see Table S5 ) . We found one triplet CCC→TTT mutation in clone LAN210-FOA-L1 . In addition , there are strong regional hot-spots in the genome-wide distribution of PmCDA1-induced mutations that are not present in HAP-induced mutants [45] . The observed local regions which are saturated with mutations cannot be associated with the recombinational hotspots and long regions of ssDNA formed during resection [44] , [46] , [47] because PmCDA1 does not induce recombination in ung1 yeast [31] . The high number of hotspots per genome cannot be explained in our system by the spontaneous DSB in yeast cells as it was recently proposed ( also see Discussion ) [44] . The hotspots of deaminase-induced mutations are described in detail in our recent paper [45] and the underlying mechanisms are currently under investigation . The mutation rate can be affected by the replication timing [48] . The mutagenic mechanism of HAP ( Fig . 2 ) allows for the discrimination of errors on the lagging versus the leading strand during DNA replication [49] . Previous studies examining site-specific reversions reported a preference for HAP-induced errors on the leading strand , when site-specific reversions are studied [49] , [50] . Our genome-wide analyses permitted us to reinvestigate this phenomenon independent from the selection for specific mutations . These new , genome-wide analyses of locations of C to T versus G to A mutations found that their distribution is random on the leading or lagging strands . In order to detect potential bias close to the origins of replication , we analyzed mutations around each known origin of replication in the region +/−2000 nucleotides . We extracted all cases of neighboring mutations where two or more mutations are found in the vicinity of the same origin . We assumed that if there is a strand-specific asymmetry of mutations near the origins of replication , this should be reflected by the distribution of the types of neighboring mutations . The changes on the opposite side of the origin of replication should be complementary , because the leading and lagging strands are swapped . For example , if two mutations G-C→A-T are located to the right of an origin , they both should be of the same type , either G→A or C→T , while mutations to the left of this origin should be reciprocal ( i . e . , C→T and G→A ) . Analysis of 489 pairs of such neighbor mutations revealed a marginally significant deviation from a random expectation ½: 270 pairs of mutations are consistent with the model of strand-specific asymmetry of mutations , whereas 219 are inconsistent with this model ( P sign test = 0 . 024 ) . This result is in agreement with the model that most errors induced by HAP occur with equal probability on lagging or leading DNA strands , while in some regions/sites the bias could be substantial . Earlier work with site-specific reversions may have only described a minor and specific pathway of HAP mutagenesis at such specific sites [26] , [43] . We recently reached the same conclusion for HAP-induced forward mutations in the URA3 reporter gene [51] . Resequencing of genomes of haploid and diploid HAP- and PmCDA1-induced mutants indicate that there is significant variability in mutation levels in yeast cell populations . Since diploid mutant clones were selected for concomitant mutations of the two copies of the CAN1 gene , we then investigated the mutational load in cells treated with either mutagen but not selected for canavanine or 5-FOA resistance . We have picked up arbitrary diploid clones from the same YPDU plates that were used to treat strains with HAP before replica-plating to the canavanine-containing media , and from synthetic complete plates that were used to estimate the viability in the case of PmCDA1 ( see Materials and Methods for details ) . We sequenced the genomes of eight HAP-treated and four PmCDA1-treated non-mutants . Analysis of SNVs revealed that HAP ( Table 2 and Table S4 ) and PmCDA1 ( Table 3 and Table S5 ) induce the same types of mutations in non-mutant clones as in selected CanR and FOAR mutants , albeit at significantly lower frequencies ( Fig . 4A and Fig . 4B , respectively ) . Most of the mutations in non-mutant clones are heterozygous . Interestingly , HAP-treated non-mutant diploid clones accumulate more SNVs than HAP-induced CanR haploid mutants , whereas PmCDA1-treated non-mutant clones contain fewer SNVs than CanR PmCDA1-induced haploid mutants ( Fig . 4 ) . These results provide additional evidence that levels of HAP- and PmCDA1-induced mutagenesis vary widely , even in the absence of selection ( see Discussion ) . Recessive heterozygous mutations in diploid genomes have no effect on survival but can cause lethality in haploids . We performed tetrad analysis to estimate the viability of the haploid progeny of wild-type diploid strains , as well as progeny from HAP- and PmCDA1-treated mutant and unselected mutagenized clones . Most of the haploid spores obtained from wild-type strains ( LAN211 and LAN210 ) are viable ( Table 4 and Fig . 4C , top row ) . On the other hand , most of the spores from HAP-induced mutants are inviable ( see example in Fig . 4C , second row ) . A few viable spores were detected for only two mutants tested ( LAN211-5 and LAN211-6 , see Table 4 ) . Similarly , the majority of spores obtained from PmCDA1-induced mutants do not grow ( see e . g . Fig . 4C , second row; see also Table 4 ) . HAP-treated , non-mutant clones show variable viability . All LAN211-NM1 progeny are inviable , whereas viability is very high in LAN211-NM2 and LAN211-NM4 progeny . LAN211-NM3 progeny display an intermediate level of viability ( 44 . 4% ) and considerable heterogeneity among viable spores . Some of the spores were of normal size , while others were small ( Fig . 4C , bottom row; Table 4 ) . The viability of the haploid progeny of PmCDA1-treated non-mutant clones is similar to that of the wild-type strains . About 75% of HAP-induced mutations were found in open-reading frames ( ORFs ) of protein-coding genes ( Fig . 5A ) , as expected , given that ORFs encompass about 73% of our reference genomes . Among these mutations , two-thirds ( comprising about 50% of all SNVs ) are non-synonymous , whereas about one-third ( ∼25% of all SNVs ) are synonymous . SNVs resulting in protein truncations range from 2% to 3% in different genome types ( Fig . 5 ) . Interestingly , we found eight mutations predicted to result in the extension of an encoded protein sequence ( Table S6 ) . Unexpectedly , we found no difference in the distribution of the types of substitutions between all types of clones - haploid mutants , diploid mutants and diploid non-mutants ( Fig . 5 ) . The same analysis was performed for SNVs in PmCDA1-induced mutants ( Fig . 5B ) . Here , many more SNVs are present in regions outside of CDS , as compared to the HAP results . Sixty-five and 56 percent of SNVs were found in non-protein coding regions in haploid and diploid mutant clones , respectively . These values are much greater than expected given that non-protein-coding regions comprise only about 25% of the yeast genome . Also , the fraction of non-synonymous SNVs is much less for PmCDA1-induced clones compared to HAP-induced clones ( 21% and 26% vs . 45–48% ) . The number of synonymous SNVs for PmCDA1-induced clones ranges from 11% to 16% . The fractions of truncation mutations were similar for HAP and PmCDA1 ( 3% in PmCDA1 genomes and 2–3% for HAP ) . We estimated from our data that 0 . 3 to 1 . 4% of all HAP-induced base substitutions cause lethal mutations in haploid cells . Our logic is as follows . Considering that about 18% of yeast genes are essential [52] , [53] , and given that about half of the SNVs in HAP-treated genomes are either non-synonymous or lead to protein truncation ( Fig . 5 ) , we estimate that up to 9% of all SNVs can potentially be lethal in haploid progeny . This translates into 43 , 4 , 40 and 15 such potentially lethal SNVs in the genomes of LAN211-NM1 , LAN211-NM2 , LAN211-NM3 and LAN211-NM4 , respectively . To get an estimate of how many of these potentially lethal SNVs are actually lethal , we performed the following calculations . Roughly half ( 44 . 4% ) of the spores obtained from LAN211-NM3 are inviable , indicating the presence of a single latent lethal heterozygous mutation in this clone . That means that about 2 . 5% ( one mutation out of 40 potentially lethal SNVs ) of non-synonymous SNVs in ORFs of essential genes lead to lethality . Strain LAN211-NM1 has a similar number of SNVs but none of its spores are viable ( 28 tetrads with 112 spores analyzed , all spores inviable; see Table 4 ) . Therefore , spore viability in this strain is less than ∼1% ( 1/112 ) , which translates into at least six or seven latent lethal heterozygous SNVs in this clone , assuming that the mutations are not linked ( viability of spores = ( 1/2 ) n , where n = number of heterozygous mutations lethal in homozygous state; for 1% viability n≈6 . 5 ) . At least ∼15% ( 6 . 5/43 ) of the non-synonymous SNVs in essential genes in this clone are lethal . Taken together , our data show that three to 15% of non-synonymous SNVs ( or 0 . 3 to 1 . 4% of all SNVs ) in our HAP-induced mutant clones are lethal in the homozygous state .
Earlier studies using next-generation sequencing in yeast documented rare spontaneous mutations in yeast haploid and diploid strains [54] , [55] . Here , we extend these findings by comparing strains with different ploidy and by applying two different mutagens . We found the intrinsic differences in the ability of cells from the same population to mutate after treatment with two different mutagens . Mutants conferring resistance to canavanine in diploid yeast induced by two types of mutagens accumulate many more SNVs than haploid mutants ( Figs . 2D , 2E , 3A , 3B , Tables 2 and 3 ) . This is in agreement with the high mutation frequency observed in diploids ( Table 1 ) . The canavanine-resistance phenotype ( CanR ) in diploids is a result of two genetic events needed to inactivate both copies of the CAN1 gene in diploid strains ( Fig . 1 ) . Since both HAP and PmCDA1 ( in ung1 strains ) do not induce recombination in our system , both alleles of CAN1 are inactivated by independent mutations ( right branch on Fig . 1 ) , except for rare cases of spontaneous mitotic recombination . Thus , by selection for can1 mutants in diploid cells , we essentially select the progeny of cells which experienced high levels of mutagenesis . The effect of transient hypermutability is not specific for CanR selection . First , PmCDA1-induced FOAR diploid mutants possess the same high level of mutations as their CanR counterparts ( Fig . 2E , 3B , Table 3 ) . Second , transient hypermutability is observed with other reporters , e . g . using the LYS2 forward mutagenesis reporter gene [21] . We demonstrated previously that the selection for mutants in haploid strains with a duplication of the reporter gene results in a much smaller number of mutants compared to normal diploids ( Fig . 6A ) [21] and [56] . The levels of HAP mutagenesis are the same in triploid strains and in diploid strains with a duplicated reporter gene on one of the homologous chromosomes ( Fig . 6B ) . Thus , in these model systems , high levels of mutagenesis require that cells be diploid or have higher ploidy . Observed spikes of mutability in individual cells are also not specific to only one particular mutagen . Progeny of such cells was observed in the case of both HAP and PmCDA1 , underscoring the fact that different mutagens can induce hypermutagenesis . However , the types of mutations found were mutagen-specific , suggesting that the principal mechanism of mutations in the hypermutable fraction is the same in all other cells . The genome resequencing and genetic results show that the distribution of the mutation load is highly uneven in cell populations . Some cells accumulate dramatically more mutations than others . In other words , the mutation frequency , as virtually any other variable , follows a certain distribution ( Fig . 7 ) . Cells that survive very high levels of mutagenesis constitute a hypermutable fraction of a population and impact the overall estimated mutation rate . For example , 1% of cells with a mutation rate three orders of magnitude higher than that in regular cells will elevate the detected rate for a given cell culture by ten-fold . These cells survive in diploid clones and were found as the canavanine- , 5-FOA or aminoadipic acid-resistant mutants that we selected . Haploid cells cannot tolerate such a high level of mutagenesis due to the inactivation of housekeeping genes . The nature and shape of the mutability distribution requires additional investigation with hundreds of genomes from mutagenized but randomly sampled ( i . e . non-mutant ) clones sequenced . Since the majority of prior studies on the molecular mechanisms of mutagenesis have been performed in the haploid model systems , the hypermutable fraction of diploid cells described here has evaded detection in the earlier literature . To our knowledge , the only exception is the detection of transiently hypermutable populations of cells that arise during adaptive mutagenesis in bacteria [57] , [58] , [59] , [60] . The existence of these hypermutable bacterial cells is restricted to the specific conditions of nutrient starvation . Importantly , hypermutable cells have never been directly detected in the eukaryotic species , although genetic studies are consistent with their presence [21] , [61] . Hypermutable cells can be potentially responsible for the accumulation of multiple mutations during carcinogenesis and evolution . We further corroborated our model by analyzing the genomes of several non-mutant clones treated with the mutagen . These clones have much less SNVs than their CanR mutant diploid counterparts ( Fig . 4 ) . When PmCDA1 is used , the number of SNVs in non-mutant clones is very low ( 10 , 14 , 4 and 34 mutations in LAN210-NM1 -LAN210-NM4 , respectively ) , indicating that only a small fraction of cells producing PmCDA1 experience extremely high levels of induced mutagenesis . Therefore , the distribution of cells with different mutation rates is narrower compared to HAP ( compare Tables 2 and 3 and Fig . 4A and 3B ) . It appears that every mutagen causes a different distribution of levels of mutagenesis among cells . The shape of this distribution may be modified by the type of organism , environmental conditions and degree of variation of the mutagen processing physiology in the cells . As a result , the size and parameters of the fraction of hypermutable cells is different for different mutagens . The shape of the “default” distribution of levels of mutagenesis ( that is characteristic of a certain cell population not treated with any mutagen ) is modified by the application of the mutagen . Mutagens not only increase the integral mutability in the cell population , but they also change the overall shape of the distribution of mutation rates in individual cells as evidenced by the comparison of mutation loads in non-mutant clones treated with HAP ( intermediate mutation load ) and PmCDA1 ( very few mutations ) . Several mechanisms could contribute to the uneven mutability of cells in a population . In the case of cells not treated with the mutagens , it could be fluctuations in DNA mismatch repair efficiency in strains with defective DNA polymerase proofreading from cell to cell [22] . In the case of mutagenized cells , the effective intracellular concentration of a mutagen may differ between cells . HAP-induced mutation rates can be influenced by differences in HAP uptake and subsequent metabolism ( conversion to dHAPTP by salvage and de novo nucleotide synthesis pathways and hydrolysis of dHAPTP by the Ham1 protein [25] ) . It is known that the deletion of the HAM1 gene leads to the increase of yeast sensitivity to the mutagenic action of HAP , by almost two orders of magnitude [39] . In the case of PmCDA1 , its mutagenesis level could be modulated by differences in deaminase gene expression , protein degradation and aggregation , availability of substrate ssDNA , and fluctuations in levels of proteins that protect the genome from deamination ( such as RPA [62] ) or stimulate deamination ( for example , [63] ) . The transient hypermutable cells are likely to exist in any cell population . Accumulating evidence suggests that gene expression profiles vary from between cells of the same type in the same tissue ( see recent paper about immune cells [64] and references therein ) . Such single-cell differences may affect the response of the cells to a particular mutagen or induce the expression of mutator proteins , such as APOBEC [9] , [10] , [14] , [16] , [42] , [44] , [45] . However , the mechanisms underlying these effects are different for different organisms , cell types and the mutagen or mutator backgrounds used . The types of mutations found in the progeny of hypermutable cells and their distributions over the genome depends on the conditions , whether cells were mutagenized and , if so , what mutagen was used . Even when the same mutagen was applied , the level of mutagenesis and its specificity are both variable between different cells . The ratio of mutations in C-G pairs to mutations in A-T pairs varies widely between different HAP-treated clones ( Table S4 ) . Moreover , one of the sequenced HAP-induced diploid mutants ( LAN211-1 ) shows a non-typical bias toward A-T to G-C transitions , whereas in other sequenced clones and in published reports using reporter genes , G-C to A-T transitions are more frequent [26] , [43] . It is hard to explain this extremely interesting phenomenon of clone-to-clone variability . One possibility is that it could be due to cell-to-cell differences in DNA replication . Eukaryotes replicate DNA with the aid of different polymerases [65] . One can speculate that there is a difference between the main replicative DNA polymerases δ and ε in the rules of HAP incorporation and replication of HAP-containing DNA by these enzymes . In this scenario , if partition between pol δ and ε varies from cell to cell , then this could account for the deviation from the expected behavior during HAP-induced mutagenesis , where more G-C to A-T transitions are typically observed . The use of genome-wide sequencing enabled the detection of both transiently hypermutable diploid cells and cell-to-cell variability in the type of changes induced by the same mutagen in the same population of cells . Similar to new paradigms emerging from single-molecule techniques in biochemistry , our analysis revealed that cells undergoing mutagenesis are not identical and differ significantly from the averaged sample estimates . Heterozygous mutations in diploid mutants have no effect on fitness as long as they are recessive . To estimate the effects of these mutations on viability , we induced sporulation of diploid yeast clones and dissected the resulting tetrads of haploid spores . The severe decrease in the viability of spores from CanR mutants ( Fig . 4C and Table 4 ) indicates that these diploids possess multiple lethal mutations in the heterozygous state . As expected from their low mutational load , the viability of spores derived from non-mutant PmCDA1-treated diploids is similar to the wild-type level . HAP-treated non-mutant clones show very interesting results after meiosis and tetrad dissection . Although all spores from clone LAN211-NM1 ( 474 heterozygous SNVs ) are inviable , LAN211-NM2 ( 40 heterozygous SNVs ) and LAN211-NM4 ( 161 heterozygous SNVs ) display near-wild-type spore viability . Of the spores from LAN211-NM3 , 55 . 6% ( 449 heterozygous SNVs ) are inviable . Among the LAN211-NM3 clone's spores , 38 formed colonies of normal size and 47 formed very small , barely visible colonies , which were not able to grow any further after being transferred into YPDAU broth and , thus were classified as inviable . Most likely , the ability of haploid spores to grow reflects the segregation of several lethal and conditionally lethal mutations . The segregation pattern differed from one individual spore to another ( see Fig . 4C ) . These results indicate that the upper threshold for the number of heterozygous SNVs per parental diploid genome after mutagenesis that haploid meiotic progeny will tolerate is somewhere around 460 . Analyses of the predicted effects of SNVs on genes in different types of HAP-treated clones did not reveal any significant differences in the ratio of synonymous to non-synonymous SNVs and to mutations outside the CDS . PmCDA1-treated clones show similar results , though variability is higher . On the other hand , deaminase induced many more mutations in non-CDS regions than HAP . This result is unexpected because AID/APOBEC deaminases are known to act on ssDNA , especially during transcription [33] . It is possible that PmCDA1 deaminates genomic regions corresponding to 5′- and/or 3′-UTRs of the genes . Another possibility is that deaminases may have access to the ssDNA formed during both transcription and replication in yeast , which results in mutation in transcribed and non-transcribed regions . Further studies are required to clarify the observed effect of preferential enzymatic deamination of non-CDS regions in yeast . Tandem CC→TT and GG→AA mutations are present in all seven diploid ( with one clone possessing triplet CCC→TTT mutation ) and one of the haploid PmCDA1-induced mutants . These tandem substitutions are indeed due to enzymatic deamination and not due to the oxidative damage to the DNA , because CC→TT mutations have been found exclusively in the genomes of clones treated with PmCDA1 . In addition , dense localized clusters of mutations are present in several loci . These clusters of mutations are highly similar to the clusters recently discovered in yeast under chronic exposure to a mutagen and in human cancers [9] , [42] . It has been hypothesized that AID/APOBEC deaminases are involved in the formation of these clusters . The existence of tandem SNVs and mutation clusters induced by PmCDA1 is likely a result of the processive action of deaminase on certain regions of the yeast genome ( i . e . , where it binds to ssDNA and slides back and forth , catalyzing multiple deaminations ) [66] , [67] . The processive action of deaminase in the genome may also help to explain the higher numbers of mutations in non-protein-coding regions ( Fig . 5 ) . Clones with a high frequency of mutations in ORFs that result from processive deaminase activity are likely to be counter-selected due to the dominant nature of the resulting mutant alleles . Our data provide the direct link between AID/APOBECs and mutational thunderstorms ( kataegis ) ; we concentrate on analysis of these clustered mutations induced by deaminase in our parallel paper [45] . Two other groups have recently used a yeast system to study deaminase-induced genome-wide mutagenesis and have come to similar conclusions [44] , [68] . Taylor and colleagues [44] proposed ( and demonstrated using SceI-induced double-strand break ( DSB ) , see also [47] ) that resection of DSB induced by the repair of deaminated cytosine or by other means ( independent of deaminase ) leads to exposure of ssDNA , which is preferentially deaminated by APOBEC , causing clustered mutations . The Gordenin group proposed similar mechanism [42] and recently reported clustered APOBEC3G-induced mutations in the reporter localized in the overhang resulting from uncapped telomeres [68] . Recombination induced by deaminase is completely blocked by uracil-DNA-glycosylase disruption in our strain [31] , but we still observe the genome-wide multiple mutation clusters ( this work and [45] ) . We conclude that the high level of mutagenesis in diploids allowed for the detection of clustered mutations induced independently from recombination . Considering genome-wide distribution on mutations induced by PmCDA1 , possible sources of ssDNA for deaminase could be intermediates of replication and transcription . Whole-genome resequencing provides an unprecedented opportunity to analyze the genome-wide distribution of mutations and their sequence context . We compared the genome-wide mutational sequence context data that we obtained for HAP and PmCDA1 mutagenesis with prior results obtained using reporter genes . We found that HAP has a slight preference for A-T-rich sequences in the genome compared to results obtained using the URA3 gene as a reporter ( Fig . 3 , left column of consensus sequences ) ( data from [26] and this work ) . An even stronger bias is evident in the spectrum of HAP-induced mutations in the LYS2 gene , where a major hotspot at position 3165 in the LYS2 ORF severely affects the results of sequence context analysis ( Fig . 3 , bottom consensus in the left column ) [43] . PmCDA1 mutagenesis shows a strong preference for deamination of cytosines at ATC motifs in the yeast genome , which agrees with the results of the PmCDA1-induced mutational spectra obtained from sea lamprey lymphocyte receptor gene variable regions , further corroborating evidence that PmCDA1 is responsible for VLR diversification [31] . Our genome-wide mutation sequence context results are very similar to the spectra of PmCDA1-induced mutations in the yeast URA3 and CAN1 genes when they are used as reporter genes ( in this work and [31] ) . In contrast , the CTC motif ( mutated base underlined ) is favored by PmCDA1 when the E . coli rpoB gene is used as a reporter , which is primarily due to a strong hotspot at position 1592 in the rpoB ORF ( Fig . 3 , bottom consensus , right column ) [31] . Taken together , we conclude that analyses of the sequence context preferences of mutagens using reporter genes should be interpreted carefully , especially when the number of detectable positions in the reporter is limited and strong hotspots are found for the reporter/mutagen combination under study .
All S . cerevisiae strains used in this study ( see Table S1 for genotypes ) are derived from 1B-D770 [69] . The mutant ura3–4 allele in this strain was reverted to wild type by transformation with wild type URA3 DNA obtained by PCR , yielding the LAN201 strain . LAN211 is an auto-diploid of LAN201 obtained by HO endonuclease expression followed by selection for diploids . The haploid ung1-deficient strain LAN200 was described previously [62] . Auto-diploidization of LAN200 resulted in the diploid ung1 strain LAN210 . Standard yeast media were used [70] . For selection of mutants we have used synthetic complete ( SC ) agar plates without arginine with 60 mg/L of L-canavanine or 0 . 1% of FOA . For induction of deaminase expression , minimal synthetic media with addition of 1% raffinose and 2% galactose was used . Mutation frequencies were determined by fluctuation analysis as described previously [69] . For the HAP experiment , independent LAN201 or LAN211 clones were grown in rich YPD media overnight . HAP was added to the media , where applicable , to a final concentration 50 µg/ml . After overnight incubation at 30°C , cultures were plated undiluted on synthetic complete media with canavanine ( SC+CAN ) to select for can1 mutants , and with dilution to complete ( SC ) plates to estimate viability . The CAN1 gene encodes arginine permease , which transports the toxic arginine analog canavanine into cells . Inactivation of CAN1 renders cells resistant to canavanine . For the PmCDA1 experiments , plasmid pESC-LEU-PmCDA1 was constructed as follows . Total RNA was extracted from the blood of sea lamprey ( Petromyzon marinus ) and reverse-transcribed with oligo ( dT ) . PmCDA1 was amplified with primers NotICDA1N-F ( 5′-TTTGCGGCCGCACCATGACCGACGCTGAGTAC , location 118–135 in GenBank accession EF094822 ) and SpeICDA1C-R ( 5′- TTTACTAGTGCAACAGCAGGACTCTTAGTG , location 724–742 in EF094822 ) and cloned into pESC-LEU vector ( Stratagene ) . For yeast experiments , LAN200 or LAN210 strains were transformed with the pESC-LEU-PmCDA1 expression plasmid or with the vector only [31] , [37] . Colony-purified transformants were inoculated in 5 ml of synthetic liquid media without leucine containing 1% raffinose . After overnight incubation , galactose was added to cultures at a final concentration of 2% . Galactose activates the GAL1-10 promoter in the pESC-LEU vector which induces the expression of deaminase . After one day of incubation , culture suspensions were plated undiluted on SC+CAN and with dilution on complete plates . LAN201 and LAN211 were streaked on YPDAU plates and grown overnight . The next day , they were replica-plated on fresh YPD plates , and a drop of HAP was added to sterile filter paper placed on the agar surface so that different yeast patches receive a similar HAP dose . The next day , streaks were replica-plated on SC+CAN plates to select for mutants . One CanR colony was picked from one streak , then colony-purified and frozen as a glycerol stock at −80°C . To obtain non-mutant HAP-treated clones , yeast from YPD plates with HAP ( the same plates used to obtain CanR clones ) were streaked on YPDAU plates without HAP and then colony-purified . All of the isolated HAP-treated , non-mutant clones were confirmed to be CanS . In the PmCDA1 experiments , LAN200 and LAN210 were transformed with pESC-LEU-PmCDA1 . Individual transformants were then inoculated in 5 ml of liquid synthetic media containing glucose and without leucine , followed by incubation for one day at 30°C with shaking . Cells were then pelleted , washed once with sterile water , then resuspended in 12 ml of synthetic media without leucine containing 2% galactose and 1% raffinose , followed by incubation for 3 days at 30°C with shaking . Aliquots of the resulting yeast suspensions were plated on synthetic complete media containing canavanine to select for can1 mutants . Aliquots of diluted cultures were plated on synthetic complete ( SC ) plates to estimate viability . Individual CANR colonies ( one per each independent culture ) were colony-purified and stored at −80°C . PmCDA1-treated non-mutant clones were arbitrarily picked up from SC plates . These clones were confirmed to be CanS . We used the method described in [71] with slight modifications . Cells were collected from 30 ml of saturated culture ( OD600≈10 ) grown in YPDAU medium , washed once with water , and resuspended in 3 ml of lysis buffer ( 0 . 1 M Tris-HCl pH 8 . 0 , 50 mM EDTA , 1% SDS ) . Then 150 µl of 5 M NaCl and ∼1 . 2 ml of glass beads were added to the suspension . Cells were disrupted by vortexing ( 2 cycles , 2 min each ) in a cold room and then the lysate was centrifuged ( 13 , 000 g , 10 min ) . DNA was purified from the supernatant using phenol-chloroform extraction followed by ethanol precipitation . The DNA pellet was dissolved in DNA-grade water and treated with RNAse A ( Qiagen , 10 µl of 10 mg/ml per sample , 1 h at 37°C ) . DNA was purified again by phenol-chloroform extraction followed by ethanol precipitation , and finally resuspended in DNA-grade water . The concentration and quality of DNA preparations were monitored by agarose gel electrophoresis and the use of a NanoDrop spectrophotometer ( Thermo Scientific ) and a Qubit fluorometer ( Invitrogen ) . Isolated yeast genomic DNA was used to construct fragment libraries using the recommended kits for sequencing on the UNMC NGS Core Laboratory's HiSeq 2000 instrument . We multiplexed individual yeast libraries , each derived from an individual clone , in a single lane of an Illumina flow cell . Each of the yeast genomes was sequenced at 100× to 300× coverage ( depending on the run ) by sequencing 101 bp from each end of the individual DNA fragments in the library ( 101 bp paired-end sequencing ) , according to Illumina's recommendations . During the instrument run and after sequencing of the yeast libraries was completed , a variety of quality assurance ( QA ) measurements were made to ensure the integrity of the DNA sequence data . The DNA “bar codes” used for multiplexing were first used to partition the reads into their respective sample-specific bins , and then the bar codes were stripped from the reads to yield sample-specific DNA sequences of interest . Base-calling error correction was performed on each sample-specific set of de-multiplexed raw reads using Quake [72] . Raw Illumina resequencing data for the LAN211 strain and for the various mutant and non-mutant clones were deposited in the NCBI Sequence Read Archive ( www . ncbi . nlm . nih . gov/sra , accession numbers SRA057025 and SRP014741 ) . About ten million pair-end reads generated by sequencing of the whole-genome library obtained from the LAN201 reference strain were used for de novo genome assembly using CLC Bio's Genomics Workbench software ( CLC Bio , Aarhus , Denmark ) . This resulted in 458 contigs of various lengths ( from 200 to 217 , 386 bp ) . These contigs were aligned to the genome of the standard yeast strain S288C using batch BLAST . After sorting of the contigs by chromosome , each set was scaffolded ( ordered and oriented ) against the corresponding chromosome using Geneious Pro software ( Biomatters Ltd , Auckland , New Zealand ) [73] . Consensus sequences were extracted from the scaffolds and used in the next step of reference genome assembly . Raw sequencing data ( the same 10 million reads that were used for the de novo assembly ) were then assembled to the extracted consensus sequences , and SNVs were detected using Geneious Pro . After manual identification of false positives and the correction of alignments , a new consensus was obtained . This “version 1” LAN201 draft genome assembly covers 92 . 74% of the standard S288C yeast genome downloaded from the Saccharomyces Genome Database ( SGD , www . yeastgenome . org ) in October 2011 . This draft assembly has a GC content of 38 . 24% ( compared with 38 . 31% for S288C; see Table S2 ) . LAN211 is identical to LAN201 except for the mating type locus . To obtain references of ung1-deficient strains ( LAN200 and LAN210 ) , sequencing reads corresponding to LAN210 were assembled on the LAN211 draft in Geneious Pro . SNVs were called , manually checked , and then a consensus sequence was deduced . This resulted in a LAN210 draft genome assembly . The re-assembly of the LAN200 reads on the LAN210 reference followed by SNV calling confirmed that the two strains are isogenic . These draft genome assemblies were used to analyze the genomes of mutant clones . UNG1 and ung1 reference genomes are compared in Table S2 . Ten to 20 million pair-end reads per clone were comparatively assembled on the LAN211 ( for HAP ) or LAN210 ( for PmCDA1 ) reference genomes using Geneious Pro . SNVs were called using homozygous ( SNV frequency ≥80% ) and heterozygous ( 40% ≤ SNV frequency ≤80% ) modes . The threshold SNV call frequencies ( 40% and 80% ) were selected based on pilot experiments designed to optimize the detection of true SNVs and reduce the number of false positives . Regions of high and low coverage ( more than two standard deviations from the mean ) were excluded from the analysis . At this point , a fraction of non-expected substitution types was observed in the genomes . These included non-G-C→A-T and non-A-T→G-C mutations in HAP-treated genomes and non-G-C to A-T mutations in PmCDA1-treated genomes . The majority of these SNVs were found in the regions where reads were clearly misaligned to the reference ( i . e . having low mapping quality ) . The rest of the non-expected SNVs were found in the otherwise good regions of assembly . We PCR-amplified some of the representative genomic regions where expected and non-expected SNV types were detected , and then sequenced these amplicons using the Sanger method . All SNVs of the expected types were indeed present in the genomes , whereas all non-expected SNVs were found to be assembly errors . For example , we observed frequent putative A→C “transversions” in ACC or ACCCC sequence motifs , but these were not confirmed by Sanger sequencing . Based on these results , we removed all non-standard SNVs from the data set . Finally , detected SNVs were extracted from the alignments for further analyses ( Table S6 ) . The predicted effects of SNVs on proteins were analyzed in Geneious Pro [73] . To extract the genomic sequence context of mutations , we used ad hoc program Rseq1 . Consensus sequences , also called sequence logos ( Fig . 3 ) were created using WebLogo 3 ( http://weblogo . threeplusone . com/ ) [74] with adjustment for the GC composition of the corresponding genomes and reporter genes . The Mann-Whitney test as used to compare differences in mutation loads in different types of mutant and non-mutant clones ( see Fig . 4 ) . This result suggests that the difference between haploid and diploid strains is significant , reflecting differences in their ability to tolerate the high frequency of induced mutations . | Evolution and carcinogenesis are driven by mutations . Cells maintain constant mutation rates and can afford only transient mutagenesis bursts for adaptation . The nature of the mutational avalanches is not very clear . We sequenced the whole genomes of mutants induced in haploid and diploid yeast by nucleobase analog HAP and by DNA editing cytosine deaminase . Mutants selected in diploids are saturated with passenger mutations . Far fewer mutations are found in haploid mutants . Treatment with a mutagen without selection results in intermediate mutagenesis . The observed transient hypermutability of diploids under mutagenic insult helps to explain the wellspring of mutations that arise during evolution and carcinogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Genome-Wide Mutation Avalanches Induced in Diploid Yeast Cells by a Base Analog or an APOBEC Deaminase |
Intracellular pathogenic bacteria evade the immune response by replicating within host cells . Legionella pneumophila , the causative agent of Legionnaires’ Disease , makes use of numerous effector proteins to construct a niche supportive of its replication within phagocytic cells . The L . pneumophila effector SidK was identified in a screen for proteins that reduce the activity of the proton pumping vacuolar-type ATPases ( V-ATPases ) when expressed in the yeast Saccharomyces cerevisae . SidK is secreted by L . pneumophila in the early stages of infection and by binding to and inhibiting the V-ATPase , SidK reduces phagosomal acidification and promotes survival of the bacterium inside macrophages . We determined crystal structures of the N-terminal region of SidK at 2 . 3 Å resolution and used single particle electron cryomicroscopy ( cryo-EM ) to determine structures of V-ATPase:SidK complexes at ~6 . 8 Å resolution . SidK is a flexible and elongated protein composed of an α-helical region that interacts with subunit A of the V-ATPase and a second region of unknown function that is flexibly-tethered to the first . SidK binds V-ATPase strongly by interacting via two α-helical bundles at its N terminus with subunit A . In vitro activity assays show that SidK does not inhibit the V-ATPase completely , but reduces its activity by ~40% , consistent with the partial V-ATPase deficiency phenotype its expression causes in yeast . The cryo-EM analysis shows that SidK reduces the flexibility of the A-subunit that is in the ‘open’ conformation . Fluorescence experiments indicate that SidK binding decreases the affinity of V-ATPase for a fluorescent analogue of ATP . Together , these results reveal the structural basis for the fine-tuning of V-ATPase activity by SidK .
Acidification of intracellular compartments by vacuolar-type ATPases ( V-ATPases ) is crucial for numerous biological processes [1 , 2] . These processes include glycosylation in the Golgi [3 , 4] , loading of neurotransmitters in secretory vesicles [5 , 6] , protein trafficking in endosomes [7–9] , and amino acid sensing in lysosomes [10 , 11] . V-ATPases pump protons across a phospholipid membrane using energy from the hydrolysis of adenosine triphosphate ( ATP ) to adenosine diphosphate ( ADP ) and inorganic phosphate [1 , 12] . In the yeast Saccharomyces cerevisiae the complex is composed of subunits A3B3CDE3FG3Hac8c′c″def [13] , where subunits denoted by upper case letters form the soluble catalytic V1 region while subunits denoted by lower case letters form the membrane-embedded VO region . ATP hydrolysis occurs in the V1 region , where three catalytic heterodimers of A- and B-subunits assemble into a pseudo-symmetric trimer of AB heterodimers [14–16] . Each AB heterodimer contains a catalytic nucleotide-binding site and each is found in a different conformation [17] termed ‘tight’ , ‘loose’ , and ‘open’ with bound ATP , bound ADP and phosphate , and no nucleotide expected to be bound , respectively . Conformational changes in the AB heterodimers are coupled to proton translocation across the VO region by a rotary catalytic mechanism [18 , 19] where the rotor subcomplex , consisting of subunits DFc8c′c″d , turns relative to the rest of the complex . Under certain conditions , the V1 region can dissociate from the VO region to inhibit ATP hydrolysis [20–23] and prevent proton translocation . Aside from dissociation of the complex , little is known about how the activity of V-ATPases is regulated . V-ATPase activity has a central role in the clearance of material phagocytosed by immune cells . Killing of pathogens by phagocytic white blood cells , such as macrophages , occurs in phagolysosomes [24] , which are acidified by V-ATPases [25] . This acidification leads to the activation of enzymes that help to destroy phagocytized material . Some intracellular bacteria subvert this process by secreting effectors that inhibit either the formation or acidification of phagolysosomes [25–27] . The protein SidK , secreted by Legionella pneumophila [28 , 29] , interacts with the V-ATPase to inhibit acidification of the phagolysosome in the early stages of infection [25] . However , the molecular basis of this interaction and inhibition remain unclear . In this study , we determined the structure of the N-terminal domain of SidK at 2 . 3 Å resolution by X-ray crystallography , revealing SidK to be a flexible and elongated protein . Electron cryomicroscopy ( cryo-EM ) has recently emerged as a powerful method to analyse the structure of the V-ATPase at subnanometer resolution [13 , 30 , 31] . Although SidK normally binds to the human V-ATPase , to date it has only been possible to perform sub-nanometer resolution cryo-EM with V-ATPase from S . cerevisiae [30] or the insect Manducca sexta [31] due to abundance of the enzyme . Therefore , we used the S . cerevisiae V-ATPase to determine the structure of a V-ATPase:SidK3 complex at 6 . 8 Å resolution by single particle cryo-EM , showing that SidK binds the N-terminal region of the V-ATPase A-subunit . SidK binding to the V-ATPase inhibits V-ATPase activity by ~40% . Consistent with this subtle fine-tuning of V-ATPase activity , the structures do not reveal significant conformation rearrangement on binding . Instead , SidK binding reduces A-subunit flexibility and decreases the affinity of V-ATPase for the fluorescent ATP analogue TNP-ATP . The full neutralization of the Legionella containing vacuole ( LCV ) in the early stages of infection [32] therefore likely involves other effectors and the extended C-terminal domain of SidK seen in our model suggests that there are additional roles for this part of the protein .
In order to gain insight into the role of SidK in the intracellular survival of L . pneumophila , intact SidK ( 575 residues ) and different truncations of the protein were expressed heterologously in Escherichia coli , purified , and crystallization trials were performed . One SidK construct , consisting of residues 16–278 ( SidK-N ) , yielded two crystal forms that diffracted X-rays to 2 . 3 to 2 . 4 Å resolution and the structure of SidK-N was determined from these crystals . The first crystal form contained one molecule of SidK ( 16–278 ) in the asymmetric unit and the final model included amino acids 16–274 . The protein has an elongated and slightly bent α-helical structure consisting of three α-helical bundles ( Fig 1A ) . The three α-helical bundles contain four , four , and three α-helices , respectively , and are connected by loops in the bundles . The first α-helical bundle ( α1–2 , α5–6 ) also has an extension nearly perpendicular to the bundle axis that consists of two additional short α-helices ( α3-α4 ) ( Fig 1A , green arrowheads ) . Bioinformatic analysis of the SidK structure showed that the arrangements of α-helices in the second and third α-helical bundles are similar to the arrangement of α-helices observed in many other proteins , including DOCK2 ( PDB ID 3A98 ) and endo-α-N-acetylgalactoseaminidase [33] . However , the N-terminal bundle with its extensions , which from previous analysis was proposed to be the region that interacts with the V-ATPase [25] , does not resemble any other known protein structures . The second crystal form of SidK ( 16–278 ) contained two molecules in the asymmetric unit related by a non-crystallographic two-fold symmetry ( Fig 1B ) . In this crystal form SidK-N exists as a dimer , with the two molecules swapping their second and third α-helical bundles ( Fig 1A and 1B , red arrowheads and oval ) . This domain swapping is accomplished by a rotation of ~180° relative to the first α-helical bundle ( Fig 1C ) that occurs through conformational changes within the short linker connecting the first and second α-helical bundles ( residues 117–125 ) . The domain swapping retains the interface between the first and second α-helical bundles seen in the first crystal form and the domain-swapped monomer is similar to the monomer in the first crystal form ( Fig 1A and 1B ) . The monomeric SidK predominates in solution , even at high concentrations , and therefore the elongated monomer structure is expected to be the functional state of the protein . Only the N-terminal region of SidK interacts with the V-ATPase and consequently a highly-elongated SidK structure was unexpected . However , the structure on its own did not provide clues into how SidK interacts with or inhibits the V-ATPase . In order to understand how SidK interacts with the V-ATPase , purified full length SidK ( residues 1 to 573 ) and detergent solubilized and purified intact V-ATPase from S . cerevisiae [23] were incubated together and the resulting complex analysed by cryo-EM ( S1 Fig ) . In the absence of ATP , V-ATPase adopts three distinct conformations that correspond to different rotational positions of its rotor subcomplex relative to the rest of the enzyme [30] . Particle images were subjected to 3-D classification in order to separate these different conformations of the V-ATPase , as well as different occupancies of SidK binding to the V-ATPase ( S2 Fig ) . The cryo-EM maps showed additional density corresponding to SidK attached to the A-subunits of the V-ATPase ( Fig 2A , green arrowheads ) . While V-ATPase particles decorated with SidK were found in all three rotational states of the enzyme , the highest-quality cryo-EM map was obtained for V-ATPase in rotational state 1 fully decorated with three copies of SidK ( Fig 2A and S2 Fig ) and reached a resolution of 6 . 8 Å ( S3 and S4 Figs ) . The C-terminal region of SidK had a local resolution worse than the rest of the complex ( S4A Fig , 'C-term' ) . The distribution of particles in the different rotational states differed slightly from the distribution observed in the absence of SidK ( S4B Fig ) : 49% , 23% , and 28% for rotational states 1 , 2 and 3 with SidK bound versus 47% , 36% , and 17% without SidK bound [30] . A pseudo atomic model of the V-ATPase:SidK3 assembly was constructed using the cryo-EM map to guide flexible fitting of the crystal structure of SidK determined here , an earlier model of V-ATPase subunits A3B3CDE3FG3 [30] , and a model of the VO region subunits ac8c′c″def [13] ( Fig 2A ) . Fitting of SidK into each of the three corresponding densities in the map required flexing the crystal structure only at the interface between α-helical bundles I and II . The flexibility between the first and second α-helical bundles is consistent with the flexibility in the linker between these bundles observed by X-ray crystallography . Bound to the V-ATPase , the C terminus of SidK extends toward the expected position of the membrane ( Fig 2A ) . The flexing of the SidK model required to fit it into the cryo-EM map introduced a slight bend in the linker region around residues Lys123-Ser124 . The fitting shows that SidK binds the V-ATPase A-subunit primarily via its first α-helical bundle ( Fig 2B , blue bracket ) and the three copies of SidK each interact mostly with the corresponding N-terminal region of one of the three A-subunits [25] . Mapping the contact surfaces in all three SidK:A-subunit pairs in the cryo-EM map to the SidK crystal structure , the main contact surface appears to involve residues Gly24 , Tyr28 , Phe62 , Ser85 , and Trp122 from the first α-helical bundle of SidK . Although the SidK crystal structure spans almost the entire region that binds the V-ATPase , it is missing the first 16 residues of the protein , which the cryo-EM map shows to form a hook-like feature that penetrates the non-catalytic interfaces between AB heterodimers ( Fig 2C , 'hook' ) . The C-terminal region of SidK is poorly resolved in the map and has lower density than SidK-N , suggesting flexible tethering between the N- and C-terminal regions ( Fig 2A and 2C , grey density ) . Surprisingly , comparison of the V-ATPase:SidK3 and V-ATPase [30] structures revealed no major conformational rearrangements in the V-ATPase upon SidK binding ( S4C–S4E Fig ) . Furthermore , purification and structural analysis of substoichiometric V-ATPase:SidK assemblies ( S2C Fig ) did not show major conformational differences in the V-ATPase with substoichiometric SidK , compared to V-ATPase alone and V-ATPase that is fully-decorated by SidK . The main interaction between SidK and V-ATPase involves the ‘non-homologous regions’ of the A-subunits ( Fig 2B ) . These regions are found in the catalytic A-subunits of V-ATPases , but not in the corresponding catalytic β-subunits of ATP synthases , suggesting that SidK inhibition is V-ATPase-specific . To test this hypothesis we performed in vitro ATPase inhibition assays with V-ATPase and the F-type ATP synthase . These assays showed that SidK inhibits V-ATPase by ~40% ( Fig 3A ) but does not detectably inhibit the F-type ATP synthase ( Fig 3B ) . Partial inhibition of the V-ATPase by SidK is not surprising , as complete inhibition of V-ATPase activity is well known to kill mammalian cells [34] , which would not be advantageous for L . pneumophila infection . A construct consisting of residues 10 to 414 could inhibit V-ATPase to the same extent as the full length SidK ( Fig 3A ) , showing that the N-terminal ‘hook’ feature seen in Fig 2C is not essential for inhibition by SidK . SidK was first identified as a V-ATPase binding protein because its expression in yeast reduced growth in liquid medium buffered to pH 7 . 0 [25] , a characteristic of V-ATPase deficiency . We found that yeast expressing SidK from a plasmid were still able to grow on solid rich medium supplemented with 4 mM ZnCl2 , indicating residual V-ATPase activity and consistent with the partial inhibition seen in in vitro assays ( Fig 3A ) . Residues from the first α-helical bundle of SidK that appear to interact with V-ATPase ( Fig 2B , Gly24 , Tyr28 , Phe62 , Ser85 , and Trp122 ) are highly conserved in SidK orthologs from other Legionella species , with Ser85 substituted to Ala in some distant Legionella species [35] . α-helical bundles II and III of SidK had almost no exposed conserved residues , except for Glu220 . In order to test the importance of SidK residues as well as the model for V-ATPase binding presented above , SidK point mutations G24E , Y28A , F62A , S85E , and W122A were generated in both yeast and bacterial expression vectors . Using the bacterial expression vectors , mutant SidK was purified and all of the constructs were found to have melting temperatures within 0 . 5°C of the 54 . 1°C Tm found for wild type SidK . Mutant SidK constructs in yeast expression vectors were used to transform yeast , and yeast growth was monitored at pH 5 . 5 and 7 . 0 , the latter requiring fully functional V-ATPase for optimal growth . This assay showed that the point mutations F62A and S85E are sufficient to prevent SidK from inhibiting V-ATPase and allow normal yeast growth ( Fig 3C and 3D ) . Mutants Y28A and W122A were indistinguishable from wild type SidK while G24E expressed at a lower level than the other SidK constructs and consequently its effect on V-ATPase activity could not be characterized ( Fig 3D ) . The SidK mutants F62A and S85E ( S5A Fig ) . However , unlike wild type SidK ( S1A Fig ) , could not be co-purified with V-ATPase , suggesting significantly decreased affinity for the V-ATPase . The F62A and S85E mutations also prevent SidK from inhibiting ATPase activity of the V-ATPase in vitro ( S5B Fig ) . For these ATPase assays , which required large quantities of purified V-ATPase and simultaneous availability of wild type SidK and mutant SidK at identical concentration , SidK samples were frozen before use . Possibly as a result of this freezing , wild type SidK only inhibited V-ATPase by ~30% , not the ~40% inhibition seen with freshly purified wild type SidK ( Fig 3A ) . The ability of the F62A and S85E mutations to prevent SidK from inhibiting V-ATPase supports the proposed model of SidK:V-ATPase complex formation . As described , the V-ATPase:SidK3 structure by itself did not offer an explanation for the ~40% inhibition of V-ATPase activity that occurs on SidK binding . However , comparison of a map of V-ATPase alone [30] with the V-ATPase:SidK3 structure showed a subtle difference . In the earlier cryo-EM maps of the V-ATPase alone , the C-terminal region of the A-subunit in the ‘open’ conformation had a lower density relative to A-subunits in the ‘tight’ or ‘loose’ conformations ( Fig 4A upper , yellow densities ) . The cryo-EM map of V-ATPase alone was determined with identical specimen preparation , imaging , and image processing conditions . The decreased density indicates mobility of this protein domain . In comparison , the C-terminal region of the ‘open’ A-subunit in the V-ATPase:SidK3 cryo-EM map is well defined ( Fig 4A lower , yellow densities , and 4B , surface versus mesh ) , indicating a more rigid structure when SidK is bound . This higher density for the ‘open’ A-subunit with SidK bound compared to the ‘open’ A-subunit without SidK is seen in all three rotational states of the enzyme ( Fig 4C ) . Furthermore , subtraction of the V-ATPase map from the V-ATPase:SidK3 map , although dominated by the presence of SidK , shows residual density in all three rotational states that is consistent with a more rigid A-subunit with SidK bound ( S6 Fig ) . Estimation of the local resolution of the density maps also shows higher relative resolution estimates for the ‘open’ A-subunit C-terminal region in V-ATPase:SidK3 than in V-ATPase alone ( S4A Fig ) , further suggesting reduced mobility upon SidK binding . The ‘open’ A-subunit is poised to bind nucleotide and flexibility is essential for rotary catalysis in the V-ATPase and other rotary ATPases [30 , 36] . The binding of SidK to the A-subunit and the effect of SidK on the flexibility of this subunit suggested that SidK may alter binding of nucleotide to the AB heterodimers . The effect of SidK on the affinity of V-ATPase for ATP was investigated using the soluble V1 subcomplex purified from S . cerevisiae . As expected , the purified V1 subcomplex had no detectable ATPase activity [22] . However , the V1 subcomplex was still able to bind nucleotide as well as SidK ( Fig 4D and S1A Fig ) , suggesting that the catalytic A3B3 hexamer of the V1 subcomplex remains in a conformation similar to the intact V-ATPase assembly and that the auto-inhibition of the V1 region does not interfere with accessibility of the active site . The binding affinity of nucleotide to the V1 subcomplex was determined using the fluorescent TNP-ATP . Although the affinities of a protein for ATP and TNP-ATP may be different [37] , this fluorescent analogue of ATP has been used successfully to compare changes in nucleotide binding in a variety of experiments [37–39] . TNP-ATP was titrated into a solution containing the purified V1 assembly and fluorescence was measured ( Fig 4D ) . The increase in fluorescence was modeled to calculate the equilibrium dissociation constant Kd ( Fig 4D ) according to Eqs 1 and 2 ( see Methods ) . The calculated Kd of V1:TNP-ATP ( 160 +/- 50 nM ) was approximately two-fold lower than the Kd of V1:SidK:TNP-ATP ( 280 +/- 60 nM ) , indicating that SidK decreases the affinity of the V-ATPase soluble catalytic region for TNP-ATP . This decrease in nucleotide affinity is consistent with the decreased flexibility of the A-subunit of V-ATPase upon SidK binding . Removal of the C-terminal residues 415–573 of SidK did not significantly affect V-ATPase inhibition by the SidK mutants ( Fig 3A ) , but truncations of the N-terminal region of the protein has been demonstrated to prevent V-ATPase inhibition [25] . These data show that binding of the N-terminal domain of SidK to the V-ATPase is responsible for its partial yet specific inhibition of V-ATPase activity . The role of the C-terminal domain of SidK in vivo remains unclear .
In the data presented here , we determined the structure of residues 16–278 of SidK by X-ray crystallography . We showed that the N-terminal α-helical bundle of the protein binds with sufficient affinity to the catalytic A-subunit of V-ATPase to determine a cryo-EM structure of the V-ATPase:SidK3 complex . The interaction between SidK and the ‘non-homologous’ region of the V-ATPase A-subunit explains the specificity of SidK to V-type rotary ATPases . However , we observed that binding of purified SidK to purified and detergent solubilized yeast V-ATPase in vitro leads to only a ~40% inhibition of V-ATPase activity . It is possible that SidK induces a larger inhibitory effect on the macrophage V-ATPase than on the yeast V-ATPase . However , sequence alignment ( S4F Fig ) shows that the A-subunit from both human and yeast V-ATPase are highly conserved in their binding site for SidK , suggesting that at least binding affinity is similar . The partial inhibition of V-ATPase is consistent with the observed effects of SidK binding on V-ATPase structure and TNP-ATP binding affinity . Complete inhibition of purified V-ATPase by bafilomycin shows that the V1 region ( where ATP hydrolysis occurs ) and VO region ( where bafilomycin binds ) are fully coupled in the preparation of the enzyme used in these studies . The 40% inhibition appears to contradict the previous observation that purified SidK is a potent inhibitor of ATPase activity in yeast membrane vesicles [25] . However , the current in vitro assays with purified components [40] are more precise than the assays used previously , and complete inhibition of V-ATPase by an intracellular pathogen would be unexpected as it would ultimately kill the host cell . The amount of SidK expressed and translocated into the host cell cytoplasm could affect how SidK influences the host cell but it seems unlikely that inhibition would ever exceed the 40% observed in the in vitro assay without compromising the utility of the infected cells for the pathogen . Early in infection , L . pneumophila maintains a neutral pH in the Legionella containing vacuole ( LCV ) [32] . This stage of infection corresponds to the period where SidK expression is high [25] . However , it was also found that L . pneumophila with SidK deleted is still infectious and did not exhibit a detectable growth defect in either mouse bone marrow-derived macrophages or Dictyostelium discoideum [25] . This observation supports the idea that , as with most Legionella effectors , there are multiple functionally-redundant effectors that serve to neutralize the pH of the LCV during the early stages of infection . This idea is also consistent with the amount of inhibition of V-ATPase caused by SidK , which likely functions along with other , currently unknown , factors to control the pH of the LCV . SidK is not expressed at late phase of L . pneumophila infection , which may allow V-ATPase to lower the pH in the LCV , a condition that appears to benefit intracellular bacterial replication [32] . Thus , the coordinated expression of multiple effector proteins that affect the pH of the LCV would provide L . pneumophila with a level of control over LCV pH that would not be possible with a single effector protein that is a potent V-ATPase inhibitor . A recent report showed that SidK binds the V-ATPase with a Kd of ~3 . 5 nM [41] . This high affinity suggests that even translocated at a low level in the host cell , SidK will bind to available V-ATPase complexes and associate with the LCV , which is known to possess V-ATPase at its membrane [42] . V-ATPase has been proposed to interact with numerous other proteins in cells , including ARNO and Arf6 [7] , actin [43] , aldolase [44] , and ragulator [11] . It is possible that SidK binding to V-ATPase alters one of these interactions leading to downstream consequences in addition to V-ATPase inhibition . Although SidK clearly causes partial inhibition of the V-ATPase upon binding , it does so without causing significant structural changes in its target . Numerous other proteins inhibit rotary ATPases , but their binding usually involves more dramatic alteration of rotary ATPase structure and function . Inhibitory Factor 1 ( IF1 ) prevents ATP hydrolysis in the mitochondria F-type ATP synthase [45] and the ε-subunit of the bacterial F-type ATP synthase inhibits ATP hydrolysis in that enzyme [46] . However , unlike IF1 and the ε-subunit , SidK binds to an interface on the A-subunit that is far from the catalytic site of the enzyme . The SidK binding site also differs from that of PA1b from Pisum sativum , which inhibits insect V-ATPases by binding to the c-subunits [47] . It is tempting to speculate that the C-terminal regions of SidK , which extend toward the membrane in the V-ATPase:SidK3 structure , may play additional roles in infection . This hypothesis is particularly appealing because the majority of the intact SidK protein is not necessary for interaction with the V-ATPase and is accessible to serve some other function . The inability of SidK to cause significant structural changes in V-ATPase is consistent with the need for V-ATPase activity as the infection proceeds . Two of five point mutations tested were sufficient to prevent inhibition of V-ATPase by SidK . Three of the five mutations engineered to test the proposed SidK:V-ATPase interface did not alter SidK’s affect on yeast growth . For the G24E mutation , this null result is explained by reduced expression of the construct . Mutations Y28A and W122A appear to be simply insufficient to abrogate inhibition by SidK . Given their important role in the defense against intracellular pathogens , V-ATPases are likely modulated by other intracellular pathogens during infection . The hybrid structural analysis procedure described here is uniquely capable of understanding the formation of these host-pathogen protein complexes .
Truncated constructs of SidK ( Q5ZWW6_LEGPH ) were designed and created taking into account the secondary structure and disorder predictions [48 , 49] . Two constructs of SidK were cloned for crystallization trials: the full-length protein ( residues 2–573 ) and a truncated N-terminal domain ( residues 16–278 ) . The gene Lpg0968 was PCR-amplified from the genomic DNA of L . pneumophila strain Philadelphia 1 using the following primers: forward ( 2–573 ) 5’- TACTTCCAATCCAATGCCTCTTTTATCAAGGTAGGTATAAAAATGGG -3’ , reverse ( 2–573 ) 5’- TTATCCACTTCCAATGTTA AAGGCTTAGGCTTTCTTCCTGTACTTT-3’; forward ( 16–278 ) 5’- TACTTCCAATCCAATGCCGAGCAATATCATAGTCAAGTAGTCGGT -3’ , reverse ( 16–278 ) 5’-TTATCCACTTCCAATGTTATTTGCTTAAAGCATTTAATTTTTCGTTTTC-3’ . The PCR products were cloned by Ligation Independent Cloning ( LIC ) into the LIC vector pMCSG7 [50] with the standard protocol [50] . The pMCSG7 vector encodes an N-terminal 6×histidine tag , separated from the target gene by a TEV protease cleavage site . The constructs produced , pCX0016 ( SidK2-573 ) and pCX0038 ( SidK16-278 ) , were verified by DNA sequencing . Competent BL21 ( DE3 ) pLysS cells ( EMD Millipore Corp . , Billerica , MA , USA ) were transformed with the expression vectors described above . Single colonies were used to inoculate LB media ( 50 mL , 100 μg/mL ampicillin ) and were grown overnight at 37°C . 1L TB media ( 100 μg/mL ampicillin ) were inoculated with 50 mL of the overnight cultures and grown at 37°C with shaking at 220 RPM . Isopropyl β-D-1-thiogalactopyranoside ( IPTG ) was added to 0 . 5 mM when the OD600 of the cultures reached 2 . 0 , and cells were grown for another 16 h at 18°C before being harvested by centrifugation for 20 min at 5 , 000 g . SidK ( 16–278 ) labeled with selenomethionine was expressed in methionine auxotroph B834 ( DE3 ) ( Novagen ) . For selenomethionine labeling cells were grown in 1 L M9 medium supplemented with methionine at 37°C . When the OD600 reached 1 . 0 cells were centrifuged and resuspended in 1 L of M9 medium and the culture was grown at 37°C for another 4 h to deplete remaining methionine . After depletion , 50 mg/L of selenomethionine was added to the cell culture 30 min prior to the addition of IPTG . Cell pellets were resuspended in buffer A ( 50 mM Tris pH 7 . 8 , 400 mM NaCl , 0 . 5 mM tris ( 2-carboxyethyl ) phosphine ( TCEP ) ) with 1 mM p-aminobenzamidine and lysed with a high-pressure cell disrupter TS series Benchtop ( Constant Systems Ltd , UK ) at 35 psi . After centrifugation for 30 min at 16 , 000 g , the supernatant was loaded onto 5 mL of TALON Metal Affinity Resin ( Clontech Laboratories , Inc . , USA ) , incubated for 2 h at 4°C , and washed with 20 resin volumes of buffer A . Protein was eluted with buffer A with 100 mM imidazole . Protein was then dialyzed against 2 L of buffer B ( 20 mM 4- ( 2-hydroxyethyl ) -1-piperazineethanesulfonic acid ( HEPES ) pH 8 . 0 , 150 mM NaCl , 0 . 5 mM TCEP ) , and cleaved with TEV protease ( 1:50 ( w:w ) ratio , 16 h at 4°C ) to remove the 6×histidine tag . Protein was purified further by gel filtration with an ENrich SEC 70 10 x 300 column ( Bio-Rad Laboratories Inc . , USA ) in buffer C ( 20 mM HEPES pH 8 . 0 , 150 mM NaCl , 0 . 5 mM TCEP ) . The protein eluted in two peaks: ~ 10% as an apparent dimer and ~ 90% as a monomer . The dimer peak fractions were pooled and concentrated to 18 mg/mL for the crystallization screening . After initial unsuccessful crystallization trials , the monomer peak fractions were subjected to reductive methylation using the Reductive Alkylation Kit ( Hampton Research , Aliso Viejo , CA , USA ) , purified again by gel filtration , and the peak fractions were concentrated to 30 mg/mL for crystallization screening . SidK mutations were made by site directed mutagenesis ( New England Biolabs ) . Protein melting temperatures were measured with the fluorescent dye Sypro Orange ( Molecular probes ) using an Applied Biosystems StepOnePlus Real Time PCR Instrument ( Life technologies ) according to the standard TmTool protocol developed by the manufacturer . SidK ( 16–278 ) dimer fractions produced well-diffracting crystals during crystallization screening with a Crystal Gryphon robot ( Art Robbins Instruments , USA ) using a Hampton Research Crystal Screen . The best crystals grew at 20°C with the well solution containing 0 . 1 M sodium cacodylate trihydrate pH 6 . 5 , 0 . 2 M magnesium acetate tetrahydrate , and 20% ( w/v ) polyethylene glycol 8 , 000 . The best crystals of Se-Met derivatized SidK ( 16–278 ) dimer were obtained at 15°C by microseeding with well solution containing 0 . 1 M 2- ( N-morpholino ) ethanesulfonic acid ( MES ) pH 6 . 5 , 0 . 2 M magnesium acetate tetrahydrate , 0 . 13 M succinic acid pH 7 . 0 ( Hampton Research ) and 18% ( w/v ) polyethylene glycol 8 , 000 . The initial crystallization conditions for the methylated monomer of SidK ( 16–278 ) were also identified using the Hampton Research Crystal Screen and vapor diffusion method . The best crystals were obtained at 20°C with wells containing 0 . 1 M Tris hydrochloride pH 8 . 5 , 0 . 2 M lithium sulfate monohydrate , and 30% ( w/v ) polyethylene glycol 4 , 000 . For data collection , crystals were soaked in a cryo-protectant ( reservoir solution supplemented with 15% glycerol ) and flash-cooled in liquid nitrogen . The X-ray diffraction data were collected at 100 K on the 08ID-CMCF ( SidK-N dimer ) and CMCF-BM 08B1-1 [SidK ( 16–278 ) methylated monomer] beamlines at the Canadian Light Source ( Saskatoon , SK , CA ) [51] . Diffraction data from the Se-Met crystals were collected at the Se absorption edge wavelength of 0 . 9788 Å . The native and Se-Met datasets were processed and scaled with XDS [52] . The positions of heavy atoms were found with SHELXD [53] and the initial model of the SidK ( 16–278 ) dimer was built with Phenix AutoBuild [54] . The model was then refined against a higher-resolution native dataset using the PHENIX software package [54] combined with manual rebuilding using Coot [55] . The structure of the SidK ( 16–278 ) monomer was solved using the molecular replacement program Phaser [56] . Asp125-Pro235 of SidK ( 16–278 ) , was used as the initial model . This portion of the molecule was fixed as a partial solution , and then the N-terminal fragment of SidK ( 16–124 ) was used as a search ensemble . The resulting model was refined with phenix . refine and manually rebuilt in Coot . The structures were validated with MolProbity [57] . The details of data collection and refinement statistics are given in Table 1 . Search for structural homologs of SidK ( 16–278 ) was performed with the DALI server [58] , PDBeFold [59] ( http://www . ebi . ac . uk/msd-srv/ssm ) , and deconSTRUCT [60] . V-ATPase and V1 were purified from S . cerevisiae as described previously [23] via a C-terminal 3×FLAG tag on the A-subunit . The SidK gene was cloned into a pET28 plasmid containing an N-terminal 6×histidine tag followed by a tobacco etch virus ( TEV ) cleavage site . BL21 Codon+ cells containing the SidK plasmid were grown at 37°C with shaking ( 225 RPM ) in 1–4 L of LB media supplemented with 0 . 4% ( w/v ) glucose and 50 mg/L kanamycin . At an OD600 of 0 . 7 , protein expression was induced with 1 mM IPTG and cells were grown overnight at 16°C . All subsequent steps were performed at 4°C . Cells were harvested by centrifugation at 5 , 000 g and lysed by sonication in TBS ( 50 mM Tris-HCl pH 7 . 4 , 0 . 3 M NaCl ) containing 0 . 001% ( w/v ) phenylmethanesulfonylfluoride ( PMSF ) . Cell lysate was centrifuged at 38 , 000 g and the supernatant was loaded onto a HisTrap Ni-NTA column ( GE Healthcare ) . The HisTrap column was washed in TBS containing 25 mM imidazole and SidK was eluted in TBS containing 0 . 3 M imidazole . SidK was mixed with TEV protease and dialyzed overnight in 2 L of TBS buffer containing 1 mM dithiothreitol ( DTT ) . Cleaved protein was dialyzed in 2×1 L of TBS and passed through a HisTrap column . The HisTrap column was washed with TBS containing 25 mM imidazole and the flow through and wash were collected . Fractions corresponding to SidK were pooled and exchanged into ion exchange buffer ( 50 mM Tris-HCl pH 7 . 4 , 1 mM ethylenediaminetetraacetic acid ( EDTA ) ) by concentration and dilution in a centrifuged concentrating device ( EMD Millipore ) . SidK was loaded onto a HiTrap Q anion exchange column ( GE Healthcare ) and eluted with a gradient of 0 to 200 mM NaCl . Fractions corresponding to SidK were pooled and exchanged into TBS containing 5 mM DTT by concentration and dilution in a centrifuged concentrating device ( EMD Millipore ) . To purify substoichiometric V-ATPase:SidK assemblies , SidK from after TEV-cleavage was added to detergent-solubilized yeast vacuolar membranes and the V-ATPase:SidK complex was purified as described previously [23] except with twice the amount of washing . To purify fully-bound V-ATPase:SidK3 assemblies , TEV-cleaved SidK purified from the second HisTrap column ( see protein purification above ) was added to detergent-solubilized yeast vacuolar membranes and the protein complex was purified as described above . ATPase activity assays were performed as described previously [23] in assay buffer ( 50 mM Tris-HCl pH 8 , 0 . 05% [w/v] n-dodecyl β-D-maltoside ( DDM ) , 3 mM MgCl2 , 1 mM DTT , 0 . 2 mM NADH , 10 U pyruvate kinase , 25 U L-lactic dehydrogenase , 1 mM phosphoenolpyruvate , 2 mM ATP ) . Proteins were incubated on ice until assaying . 160 μL reactions were performed at room temperature using a 96-well plate in a Spectramax M2 UV/visible light plate reader ( Molecular Devices ) . Four replicate experiments were conducted for enzyme with SidK or enzyme with buffer using the same stock of freshly purified V-ATPase or frozen ATP synthase . The final concentrations of enzyme and SidK were ~1 nM and ~30 nM , respectively . The concentration of bafilomycin used was 6 μM . In order to assay multiple constructs of SidK simultaneously for S5 Fig , samples had to be frozen and thawed to coordinate experiments , which decreases the ability of SidK to inhibit V-ATPase . Yeast strain BY4741 [61] was transformed by the standard lithium acetate method [62] with plasmid p425GPD [63] carrying wild-type SidK or SidK mutants . The resulting strains were grown to saturation overnight at 30°C in Leucine dropout minimal medium ( pH 5 . 5 ) and diluted at 1:40 into either minimal medium at pH 5 . 5 or minimal medium buffered with 50 mM MES and 50 mM MOP to pH 7 . 0 . The cultures were incubated with vigorous shaking for 24 hrs at 30°C and yeast growth was monitored spectrophotometrically at 600 nm . Yeast cell lysates for western blotting analysis were prepared as described previously [25] . Fluorescence emission spectra of 2' , 3'-O- ( 2 , 4 , 6-trinitrophenyl ) adenosine 5'-triphosphate ( TNP-ATP ) from 485 nm to 600 nm ( 10 nm slit width ) were recorded with a Quantamaster QM-80 spectrofluorometer ( Photon Technology International ) using an excitation wavelength of 465 nm ( 4 nm slit width ) . The temperature was maintained at 10 . 0°C with a Peltier unit . Samples ( 0 . 5 mL ) contained 0 . 25 μM V1 , with or without 2 . 5 μM SidK , and varying amounts of TNP-ATP in assay buffer ( 25 mM Tris-HCl , pH 7 . 9 , 0 . 15 M NaCl , 1 mM MgCl2 , and 0 . 5 mM DTT ) . For each titration reading , 20 μL of the reaction sample was removed and replaced with 10 μL of trinitrophenyl-ATP ( TNP-ATP ) at different concentrations and 10 μL of 0 . 5 μM V1 in 2× assay buffer with or without 5 . 0 μM SidK . Fluorescence curves were corrected by baseline subtraction using the curve for [TNP-ATP] = 0 . No significant fluorescence from the protein was observed . The increase in fluorescence due to addition of TNP-ATP was modelled in GNUplot ( www . gnuplot . info ) with the fit command using the equation [64 , 65] I=[ES][ET]Ib+m ( [S]−[ES] ) ( 1 ) Where I is the fluorescence intensity , Ib is the fluorescence intensity of total enzyme bound to substrate , [ET] is the total enzyme concentration , [ST] is the total substrate concentration , and m is the increase in fluorescence per unit increase in free substrate . [ES] is the concentration of enzyme bound to substrate given by: [ES]=0 . 5{[ET]+[ST]+Kd− ( [ET]+[ST]+Kd ) 2−4[ET][ST]} ( 2 ) where Kd is the equilibrium dissociation constant given by Kd = [E][S]/[ES] . The m parameter was estimated from fitting a straight line through the points corresponding to the four highest [TNP-ATP] values in each titration dataset . Each titration dataset was then modeled with fixed m to calculate Ib and Kd . The experimental setup described here is similar to the study by Kubala et al . [65] . However , the tight binding affinity between TNP-ATP and the V1 subcomplex allowed parameters to be estimated from the titration data instead of from separate experiments . This experimental approach is advantageous in situations where protein availability is limited , which was the case in this study . Five replicate experiments were performed for each condition ( with and without SidK ) using the same batch of freshly purified V1 subcomplex . 3 μL of ~10 mg/mL V-ATPase:SidK sample was applied to nanofabricated holey carbon grids [66] previously glow discharged in air for 2 mins . Excess sample was blotted away and the grid was plunge-frozen in a liquid propane-ethane mixture using a modified Vitrobot Mark III grid preparation robot ( FEI company ) . Vitrified samples were imaged with a FEI Tecnai TF20 electron microscope operating at 200 kV and 34 , 483× magnification , resulting in a pixel size of 1 . 45 Å/pixel . Images were collected on a Gatan K2 Summit direct detector device operating in counting mode . 15 s movies were recorded at 2 frames/s and 5 e-/pixel/s . Movie frames were aligned using alignframes_lmbfgs and averaged with shiftframes [67] . Contrast transfer function ( CTF ) parameters of the averaged images were measured with CTFFIND3 [68] and corrected for magnification anisotropy using star_fixmaganiso [69] . Candidate particle image coordinates were identified automatically in averaged images with TMaCS [70] using templates that were 2D projections of an existing map of the V-ATPase that had been low-pass filtered to 20 Å [23] . Candidate particle images were extracted from the raw movies and corrected for local drift using alignparts_lmbfgs [67] . The aligned and averaged particle images were corrected for magnification anisotropy with correctmaganisotropy_fspace_list [69] and processed by 2D and 3D classification and 3D refinement in Relion [71] . To minimize the possible influence of SidK on the classification algorithm , image classification was performed with masked maps to focus classification on the V1 region . Maps were visualized and segmented in UCSF Chimera [72] . Atomic models of the V-ATPase ( PDB 3J9T , 3J9U , 3J9V ) were docked rigidly into the density maps using UCSF Chimera and flexibly fit into the maps with Molecular Dynamics Flexible Fitting ( MDFF ) [73] . Flexible fitting in MDFF was performed with implicit solvent and using backbone atoms only while restraining secondary structure . The atomic model of SidK was docked into the V-ATPase:SidK3 map using UCSF Chimera by rigidly docking alpha-helical bundles I and II/III into the map as independent domains . The domains were joined in UCSF Chimera and the geometry of the short connecting loop between the domains was optimized using Modeller [74] . The secondary structure of SidK was predicted using the online server JPred [75] . Where appropriate , results were analyzed using an unpaired , two-tailed Student's t-test ( TTEST function in Microsoft Excel 2007 ) to calculate a p-value . Values less than 0 . 05 were deemed statistically significant . | V-ATPase-driven acidification of lysosomes in phagocytic cells activates enzymes important for killing of phagocytized pathogens . Successful pathogens can subvert host defenses by secreting effectors that target V-ATPases to inhibit lysosomal acidification or lysosomal fusion with other cell compartments . This study reveals the structure of the V-ATPase:SidK complex , an assembly formed from the interaction of host and pathogen proteins involved in the infection of phagocytic white blood cells by Legionella pneumophila . The structure and activity of the V-ATPase is altered upon SidK binding , providing insight into the infection strategy used by L . pneumophila and possibly other intravacuolar pathogens . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"dimers",
"(chemical",
"physics)",
"medicine",
"and",
"health",
"sciences",
"crystal",
"structure",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"pathogens",
"condensed",
"matter",
"physics",
"microbiology",
"enzymology",
"electron",
"cryo-microscopy",
"phosph... | 2017 | Molecular basis for the binding and modulation of V-ATPase by a bacterial effector protein |
Oviparous animals across many taxa have evolved diverse strategies that deter egg predation , providing valuable tests of how natural selection mitigates direct fitness loss . Communal egg laying in nonsocial species minimizes egg predation . However , in cannibalistic species , this very behavior facilitates egg predation by conspecifics ( cannibalism ) . Similarly , toxins and aposematic signaling that deter egg predators are often inefficient against resistant conspecifics . Egg cannibalism can be adaptive , wherein cannibals may benefit through reduced competition and added nutrition , but since it reduces Darwinian fitness , the evolution of anticannibalistic strategies is rife . However , such strategies are likely to be nontoxic because deploying toxins against related individuals would reduce inclusive fitness . Here , we report how D . melanogaster use specific hydrocarbons to chemically mask their eggs from cannibal larvae . Using an integrative approach combining behavioral , sensory , and mass spectrometry methods , we demonstrate that maternally provisioned pheromone 7 , 11-heptacosadiene ( 7 , 11-HD ) in the eggshell’s wax layer deters egg cannibalism . Furthermore , we show that 7 , 11-HD is nontoxic , can mask underlying substrates ( for example , yeast ) when coated upon them , and its detection requires pickpocket 23 ( ppk23 ) gene function . Finally , using light and electron microscopy , we demonstrate how maternal pheromones leak-proof the egg , consequently concealing it from conspecific larvae . Our data suggest that semiochemicals possibly subserve in deceptive functions across taxa , especially when predators rely on chemical cues to forage , and stimulate further research on deceptive strategies mediated through nonvisual sensory modules . This study thus highlights how integrative approaches can illuminate our understanding on the adaptive significance of deceptive defenses and the mechanisms through which they operate .
Across most animal taxa , eggs are highly vulnerable to predators because they are immobile , highly nutritious , and defenseless . However , since egg production is costly [1] , losing them to predation greatly reduces Darwinian fitness [2] . Animals have thus evolved several parent-modulated strategies: camouflage [3 , 4] , communal egg laying [5 , 6] , egg clustering [7] , parental care [2] , chemical defenses ( toxins ) [8 , 9] , and aposematic signaling [10] to mitigate this loss of fitness . On the other hand , eggs are not just vulnerable to interspecific predators but are equally at risk of predation from older conspecifics ( cannibals ) , including parents and siblings [11–13] . Egg cannibalism is commonly dismissed as an aberrant behavior , resulting from unnatural breeding conditions . However , mounting evidence has demonstrated its adaptive value in several species , wherein the cannibals increase their fitness through both reduced competition and the supplemented nutrition [11 , 14] . In support of this argument , egg cannibalism is common even among noncarnivorous species [15] and has also been shown to have important ecological consequences on population dynamics and stability [11 , 16] . Nevertheless , egg cannibalism reduces direct fitness to parents and can additionally reduce inclusive fitness if the eggs consumed are genetically related to the cannibals [11] . Interestingly , most of the aforementioned parent-modulated strategies evolved in response to interspecific egg predators are often ineffective against conspecifics: while cannibals are generally resistant to conspecific toxins and aposematic signals [17] , other strategies like producing surplus eggs and communal egg laying might even facilitate egg cannibalism [18 , 19] . Additionally , since deploying toxic defenses against conspecifics would further reduce inclusive fitness [20] , natural selection should favor the evolution of anticannibalistic strategies that are likely to be nontoxic . Anticannibalistic strategies that deter egg cannibalism have evolved independently in species across taxa , and convincingly , none of them seem to be toxic . These strategies include laying of nondeveloping eggs ( trophic eggs ) within clutches by mothers to reduce cannibalism among offspring [11 , 21] , laying of eggs with protective coatings around the egg’s shell [22] , laying eggs on specialized structures ( like stalks ) [23] , nest guarding [24] , and synchronized egg hatching [25] . Nevertheless , there are several other communally egg-laying species , including insects , that avoid cannibalizing eggs despite lacking the above strategies [15] . This prompted us to speculate about the existence of alternative strategies that could modulate this behavior in nature . The understanding of such strategies is crucial , especially to the fields of conservation , epidemiology , and pest management . We recently reported predatory cannibalism among D . melanogaster larvae , wherein younger larval instars pack-hunt and consume older conspecific larvae under laboratory conditions [26] . Surprisingly , despite their predaceous nature , we never observed larvae attacking conspecific eggs , even upon starvation . In nature , D . melanogaster oviposit communally at sites already occupied by conspecific and heterospecific larvae to facilitate social feeding among larvae [27 , 28] . Although these oviposition sites ( decaying fruits ) are nutritionally rich , they at times risk desiccation [27] and immense larval competition [29] , both of which could coerce larvae to seek other food sources such as older conspecifics [26] and cadavers [30] . Interestingly , despite availability of several conspecific eggs in their vicinity , larvae never cannibalize them , either for food or for other benefits like reduced competition [15] and reduced risk of predation by younger larvae [26] . This observation thus raises important questions on why and how egg cannibalism is averted in this system . Given that parental care in D . melanogaster is mostly limited to oviposition site selection and egg provisioning [31 , 32] , we hypothesized that parental provisioning in some form protects D . melanogaster eggs from conspecific larvae . Virgin D . melanogaster females can lay nonviable unfertilized eggs , equivalent to trophic eggs laid by other species . However , given that mated D . melanogaster seldom lay such unfertilized eggs , their production more likely represents a “risk–return strategy” ( i . e . , the cost of producing such unfertile eggs is less than what is risked by aborting them ) rather than a strategy that mitigates cannibalism . In several insect species , the eggshell , its pigmentation and patterning , and specific extrachorionic modifications upon it are all known to deter predation [8] . Drosophila eggs are enveloped within a maternally provisioned eggshell during oogenesis that is composed of three distinct layers ( Fig 1D ) : chorion , wax layer , and vitelline membrane , which serve to deter pathogens , prevent dehydration , and facilitate respiration , respectively [8] . However , the extent to which these eggshell layers play a role in deterring cannibals is not known . In this study , we investigate the mechanisms that could deter egg cannibalism in D . melanogaster . To do this , we examine the protective role of the eggshell and its constitutive layers using an integrative approach . We consequently reveal a novel anticannibalistic strategy that is mediated through chemical deception and involves semiochemicals [33] present in the wax layer of the egg shell , a so-far overlooked strategy that could potentially be widespread across taxa .
To understand the extent to which D . melanogaster females provision nutrients within their eggs , we assayed larval survival upon hatching in the absence of any food . The first-instar larvae survived for up to five days posthatching in the absence of food ( Fig 1A ) , showing that the nutrients provisioned within an egg are surplus for embryonic development , and they can hence support larval survival well beyond hatching . Feeding two injured conspecific eggs to just-hatched larvae increased their survival ( Fig 1B ) , confirming the previous finding that eggs are nutritious and are thus worth cannibalizing [30] . When egg consumption by second-instar larvae ( food deprived for 2 h ) was assayed , larvae surprisingly did not cannibalize intact viable eggs ( Fig 1F ) . However , if the eggs presented to the larvae were injured ( by pricking ) , the eggs were immediately consumed ( Fig 1F ) , supporting the results from a recent study that used nonviable eggs [30] . The same larval response was observed when the eggs provided were from an unrelated D . melanogaster strain ( panel D in S1 Fig ) . To rule out the possibility that the eggs are toxic to the larvae , we assayed larval development and survival on standard fly medium laced with crushed eggs . Compared to the control larvae raised on the standard diet , larvae supplemented with crushed eggs developed faster ( Fig 1C ) and survived equally well ( Fig 1D ) , confirming that the eggs are nontoxic and nutritious . Given that injured eggs were readily cannibalized ( Fig 1F ) , we next tested whether breached eggshells facilitate cannibalism . We sequentially removed the two outer eggshell layers , the chorion membrane and the wax layer ( Fig 1E ) , by treating the eggs with sodium hypochlorite and hexane , respectively [34 , 35] , and then presented these chemically treated eggs to food-deprived larvae . Indeed , removal of the thin wax layer ( 4–5 nm , hexane treated ) but not the thick chorion layer ( 840–1 , 250 nm , dechorinated ) [36] made 60% of eggs vulnerable to cannibals ( Fig 1F ) . This refutes a recent report [30] that D . melanogaster larvae cannibalize dechorinated eggs , which we believe is due to the experimental procedure . The dechorinated eggs they use were killed prior to their assays by dyeing with tartrazine NaCl and storing them at 4°C . These treatments could have unintentionally damaged the eggs’ wax layer . In contrast , the consecutive treatment of eggs with sodium hypochlorite and hexane in our experiments had little or no effect on egg hatchability ( panel A in S1 Fig ) , egg’s time to hatching ( panel B in S1 Fig ) , and egg-to-adult viability ( panel C in S1 Fig ) . The experiments above thus suggest that in addition to the primary role of the wax layer in preventing desiccation of the embryo [37] , it might also serve to protect eggs from cannibal larvae . Next , to understand the mechanism underlying the wax layer’s anticannibalistic function , we extracted this layer from fertilized eggs and analyzed its biochemical composition using gas chromatography hyphenated with mass spectrometry ( GC-MS ) and high-resolution atmospheric pressure photoionization Fourier transform ion cyclotron resonance mass spectrometry ( APPI FT-ICR MS ) [38] . We used these advanced mass analyzers with high accuracy ( sub-parts–million mass accuracy ) and resolving power , mainly to mine for low-molecular–weight compounds ( for example , toxins ) . However , in addition , we aimed to establish methods that could unambiguously assign elemental formulas to metabolites for better characterization , especially in life sciences ( see Materials and Methods ) . We identified 13 compounds ( linear alkenes , alkadienes , and sterols ) that were mostly known cuticular hydrocarbons ( pheromones ) of adult D . melanogaster [39] ( Figs 2A and S2 and S1 Table ) , which are mainly known to be synthesized by specialized epicuticular cells called the “oenocytes” ( oe ) [39] . Below , we focus on four pheromones ( 7 , 11-heptacosadiene [7 , 11-HD]; 7 , 11-nonacosadiene [7 , 11-ND]; 7-tricoscene [7-T]; and 11-cis-vaccenyl acetate [cVA] ) , given that they are sex-specific , have known functions , and are commercially synthesizable . The pheromone profile we detected in the wax layer was similar to the pheromones already known to be deposited by adult flies ( both sexes ) on egg-laying sites to facilitate aggregation [40 , 41] . Furthermore , these pheromones are also known to be present in the reproductive tract of mated females [42] . Thus , to exclude the possibility of potential cross-contamination of our samples by adult flies , we analyzed the hexane washes of the outer chorion layer and that of dechorinated eggs and compared their hydrocarbon profiles . Most of the pheromones detected on the outer chorion ( layer exposed to environment and female reproductive tract ) were also found after dechorination ( in the wax layer ) ( panel A in S3 Fig ) . The male pheromone cVA produced by the male’s ejaculatory bulb [43] was an exception; it was abundant only on the chorion but greatly reduced upon dechorination , suggesting that eggs acquire cVA postchoriogenesis either from the environment or from male ejaculate within the female reproductive tract [42] ( panel A in S3 Fig ) . To further ascertain that these pheromones are indeed present in the wax layer , we analyzed three successive hexane washes of the dechorinated eggs and detected these pheromones at progressively decreasing concentrations across the washes , possibly reflecting the compact nature of the wax layer [37] ( panel B in S3 Fig ) . Interestingly , the presence of such pheromones within the eggshell having other physiological functions has been previously reported in insects and nematodes [44 , 45] . For over three decades , the wax layer has been considered to be synthesized by the follicle cells that surround the oocytes , exclusively based on electron microscopy observation of lipid endosomes within the follicle cells during oogenesis and their eventual deposition onto the vitelline membrane of the egg [46 , 47] . Most of the hydrocarbons we detect in the wax layer are so far only known to be synthesized in the oes [39]; for example , the biosynthesis of dienes like 7 , 11-HD and 7 , 11-ND ( female-specific pheromones ) requires the enzymatic action of a specific desaturase desatF ( Fad2 ) in the oes [48] . However , transcriptional data ( microarray and RNA-sequencing [RNA-seq] data from FlyAtlas and FlyBase , respectively ) show that this gene is not expressed in the ovary . Thus , further empirical investigations are necessary to clarify the role of follicle cells in the synthesis of the wax layer . The presence of hydrocarbons specific to both sexes in the wax layer motivated us to track the parental origin of these pheromones using mutant flies with ablated oes ( oe− ) [39] . Males and females with or without ablated oes were crossed , and the wax-layer composition of their eggs was analyzed . The label-free , semiquantitatively established hydrocarbon profile of eggs appears to be directed by the cuticular hydrocarbon composition of the parental cross [39] they were laid by; eggs of oe− parents had fewer hydrocarbons than those of oe+ parents ( Fig 2B and S2 Table , and panel B in S4 Fig ) . The eggs parented by oe+ males and oe− females had reduced female-specific hydrocarbons—for example , 7 , 11-HD and 7 , 11-ND—compared to male-specific hydrocarbons—for example , 7-T . Its reciprocal cross had reduced male-specific hydrocarbon ( 7-T ) compared to the female-specific hydrocarbons . However , the reason as to why cVA , the male pheromone of non-oe origin , was also reduced in the eggs laid by this cross is unclear . The presence of hydrocarbons corresponding to oe− parents at low concentrations in our samples could possibly be due to residual hydrocarbons produced prior to or during oe ablation . Thus , it seems that both parents contribute towards provisioning the pheromonal content of the wax layer . This suggests that wax-layer synthesis is likely to involve transportation of maternal and paternal hydrocarbons from the oes and deposited seminal fluid , respectively , to the ovary during oogenesis . Nevertheless , the existence of such transport mechanisms involving lipophorin molecules has been speculated in D . melanogaster [31] and other insects [45] . The deterrent effect of hydrocarbons present on the egg surface towards cannibals has been previously speculated about in the coccinellid Adalia bipunctata [49] . However , our system allowed us to ascertain the deterrent role of sex-specific hydrocarbons in egg cannibalism . For this , we first assayed the vulnerability of eggs laid by the above four crosses to cannibal larvae . Conspicuously , larvae only cannibalized eggs with oe− motherhood ( Fig 2C ) . The eggs from the three oe− mutant crosses had similar egg-to-adult viability ( slightly less than eggs from the oe+ control ) , thus excluding nonviability of eggs with oe− motherhood ( panel E in S1 Fig ) . We next independently verified this deterrent role of hydrocarbons by assaying the vulnerability of hexane-washed eggs perfumed with four commercially synthesized hydrocarbons found most abundantly on the egg ( 7-T , cVA , 7 , 11-HD , and 7 , 11-ND; panel A in S4 Fig ) to cannibal larvae . Since the actual concentration of the pheromones in the wax layer could not be determined ( because of the differential solubility of the wax layer in hexane ) , the synthetic pheromones were applied and tested at several ( serially diluted ) concentrations . Interestingly , larvae only refrained from cannibalizing eggs that were perfumed with the female pheromone 7 , 11-HD ( Fig 2D ) , even when present at very low concentrations . However , the other female pheromone 7 , 11-ND we used , despite being structurally very similar to 7 , 11-HD ( with just two additional carbon atoms ) , failed to deter cannibalism ( Fig 2D ) . Given that a ) the eggs of male oe+ and female oe− with 7 , 11-HD levels slightly lower than oe+ controls become vulnerable to cannibalism and b ) eggs perfumed with various concentrations of 7 , 11-HD remain protected , these results strongly suggest that the deterrent function of 7 , 11-HD is dose independent . These findings suggest that 7 , 11-HD , rather than the overall wax layer , deters egg cannibalism; nevertheless , the synergistic effect of other hydrocarbons and chemicals we detect within the wax layer ( S2 Fig and S1 Table ) cannot be ruled out . Next , we attempted to identify the larval sensory receptors associated with this 7 , 11-HD–mediated deterrence effect . In adult D . melanogaster , 7 , 11-HD mediates mate recognition [39] , maintains the species barrier with D . simulans [50] , and is used to mark sites suitable for mating and oviposition [51] . It was recently reported [52] that in D . melanogaster , 7 , 11-HD detection is dependent on the joined action of the three receptor genes pickpocket 23 ( ppk23 ) , ppk25 , and ppk29 that are necessary for the function of gustatory neurons located in their forelegs [53 , 54] . However , whether and how larvae respond to 7 , 11-HD remains elusive [55] . Anticipating that the role of ppk23 in 7 , 11-HD detection may be conserved in larvae , we tested whether egg cannibalism is promoted when ppk23 is mutated . For this , egg cannibalism by ppk23 mutant larvae was assayed when eggs were either dechorinated or hexane washed . Indeed , ppk23 mutant larvae cannibalized eggs from both treatments ( Fig 3D ) . In a subsequent assay , ppk23 mutant larvae also cannibalized hexane-washed eggs that were perfumed with 7 , 11-HD ( panel A in S5 Fig ) , thus suggesting that the function of ppk23 required for hydrocarbon detection is conserved across the developmental stages and , interestingly , modulates different behaviors in larvae and adults . Given that larvae might concurrently sense the wax layer as an aversive cue through other sensory pathways ( olfactory or gustatory ) , we further assayed whether larvae that are mutant for gustatory receptor ( Gr33a ) , odorant coreceptor ( Orco ) , and ionotropic receptor ( Ir25a ) cannibalized dechorinated eggs . All tested mutants and their respective control ( wild-type ) larvae abstained from cannibalizing dechorinated eggs but cannibalized eggs treated with hexane ( panels B–D in S5 Fig ) . All in all , these results suggest that the deterrent effect of the wax layer was not mediated by the function of the classical set of chemosensory systems ( Gr , Or , and Ir ) but is limited to the ppk-dependent sensory system . However , since our understanding of how larvae detect their gustatory and pheromonal environment at present is rather limited [55 , 56] , further work is necessary to identify the ppk23-dependent neuronal circuits that respond to 7 , 11-HD to regulate egg cannibalism . Earlier studies have reported the water proofing nature of an egg’s wax layer [37] that can be compromised using organic solvents [57] to facilitate eggshell permeability ( especially during histological preparations ) . We therefore next morphologically examined the surface of wax-layer–deprived eggs to understand why they become vulnerable to cannibal larvae and found that they extrude egg contents through their vitelline membranes . Fine fluid droplets appeared on the surface of hexane-treated eggs after about 20–25 min ( Fig 3B ) . However , such droplets were not present on dechorinated or intact eggs ( Fig 3A–3C ) . Cryo-scanning electron microscopy of the egg surface confirmed the droplets to be egg contents permeating through the vitelline membrane ( Fig 3C; panels A–D in S7 Fig ) . Interestingly , in most eggs , these droplets stabilize ( i . e . , they do not completely drain the egg ) and persist as such for several hours . This suggests that eggs might have repair mechanisms or that the extruded material could be blocking the permeating sites . To further ascertain the extent to which maternal and paternal pheromones in the wax layer contribute towards leak-proofing an egg , we examined dechorinated eggs of the four oe mutant crosses ( described earlier ) for leakage under both light and electron microscopy . We found that only eggs with oe− mutant motherhood that had reduced female pheromones were leaky , suggesting the involvement of maternal pheromones present in the wax layer in preventing egg leakage ( panel A in S6 Fig; panels E–H in S7 Fig ) . However , when hexane-washed eggs perfumed with 7-T , cVA , 7 , 11-HD , and 7 , 11-ND were examined for leakage using a similar setup , we found that all eggs leaked irrespective of the added pheromone ( panel B in S6 Fig ) . We nevertheless observed some differences among the eggs perfumed by the four hydrocarbons in a ) the extent to which they leaked and b ) the persistence of the leak . However , we were unable to quantify these differences empirically , and thus , despite demonstrating a strong link between maternally provisioned hydrocarbons and egg leakage , this assay could not independently tag 7 , 11-HD to the leak-proofing mechanism . This could also be possibly attributed to the arbitrary concentration of 7 , 11-HD we use or because perfuming fails to completely remodel the wax-layer structure artificially in the absence of other synergetic pheromones . The above finding that altering the wax-layer composition through either chemical treatment or genetic manipulation leads to extrusion of egg contents through the vitelline membrane led us to test whether this leaking egg content makes eggs vulnerable to cannibals . We first assayed larval movement in the presence of intact and injured eggs in agar-lined Petri plates to quantify changes in foraging patterns . However , unlike cannibalistic aggregation occurring around injured larva that we have previously reported [26] , groups of injured eggs did not elicit larval aggregation , and the injured eggs were only cannibalized when they were accidentally encountered ( Fig 3E ) . However , 30% of the generally invulnerable dechorinated eggs ( Fig 1F ) succumbed to cannibals when smeared with egg content leaking from injured eggs ( Fig 3F ) . These results imply that larval recognition and cannibalism of conspecific eggs relies on specific gustatory cues emanating from leaking egg content . Nevertheless , this does not rule out the possibility that leaky or injured eggs might additionally release other stress/injury responsive signals that act as cues , facilitating larval detection . Since several experiments above suggested that the protective role of 7 , 11-HD is dose independent ( Figs 1F , 2C and 3F ) , we speculated that the wax layer in general and 7 , 11-HD in particular form a physical layer that masks or conceals the egg’s nutrient content . To validate our hypothesis , we tested whether drops of yeast that are attractive to larvae lose their ability to do so when coated externally with a layer of 7 , 11-HD . We thus assayed larval detection and feeding of such 7 , 11-HD–coated yeast droplets that were dyed to score larval feeding . Larval feeding on yeast coated with 7 , 11-HD was much lower than in the control treatment in which the yeast was coated with the solvent hexane ( Fig 3G ) . Interestingly , when the 7 , 11-HD layer on the yeast was damaged with a needle ( analogous to injuring an egg ) , larval feeding was no longer reduced despite the presence of 7 , 11-HD ( Fig 3G ) . ppk23 mutant larvae , in contrast , showed no reduction of larval feeding on yeast droplets coated with 7 , 11-HD ( Fig 3G ) . These results suggest that 7 , 11-HD forms a physical layer that prevents emanation of cues from substances ( yeast or egg ) it envelops . However , damage to this layer compromises this effect , even though the same quantity of 7 , 11-HD remains on the surface . Our study shows that under laboratory conditions , the wax layer in D . melanogaster eggshell a ) protects eggs from cannibal larvae , b ) prevents egg contents from leaking out through the vitelline membrane , and c ) contains pheromones provisioned by both parents . Together with the already-reported result that ovipositing females mark egg-laying sites with a bouquet of pheromones [51] , our finding suggests a nonvisual deception mechanism , whereby larvae misclassify conspecific eggs as an inedible object . This deception seems to be mediated by the hydrocarbon-laden wax layer that conceals the embryo and consequently silences the cues that reveal its identity . However , when damaged or leaky , changes in tactile and gustatory cues might facilitate larval detection of these eggs . We speculate below on several possible mechanisms through which such a nonvisual deceptive egg defense could possibly operate: first , through “chemical insignificance” [58] , whereby cannibals may not detect or recognize an egg owing to a lack of a detectable chemical profile; second , through “chemical masquerade” [59] whereby , despite detection , the cannibals might misclassify eggs as objects they normally ignore , possibly empty eggshells . Third , since D . melanogaster mark oviposition sites with a similar bouquet of hydrocarbons [51] , the eggs might blend with the chemical profile of the environment , additionally preventing detection through “background matching” [60] . However , since eggs remained protected on the pheromone-free agar surfaces we used during our assays , background matching may be less crucial . Nonetheless , as female aggregation at these oviposition sites increases over time , the cumulative on-site pheromone concentrations might increase to extents that could provide enhanced background matching for eggs that are laid later . Thus , communal egg-laying behavior in D . melanogaster may facilitate chemical camouflage through impregnation of maternal pheromones onto the egg , a trait that is likely to be favored by natural selection . Interestingly , since oviposition sites are generally shared among several sibling Drosophila species [27] , such chemical protection of eggs may have extended effects on intraspecific predation and calls for further investigation [49] . Another potential and inclusive possibility is that the wax layer merely serves as a protective layer to the egg with a primary role in preventing desiccation and , as a byproduct , acts as a preventive barrier against cannibals , pathogens , and toxin permeability . This view is also supported by the parsimonious nature of arthropod pheromones [33] , which allows us to safely speculate that the pheromones in the wax layer of fly eggs could subserve other functions in diverse contexts [41] . For example , the pheromones in the wax layer could serve as conspecific cues for adult aggregation , modulate the number of eggs laid at a given site , act as a dispersal cue , and possibly mediate kin recognition . More broadly , our results empirically demonstrate the multiple independent context-dependent functions of a specific pleiotropic pheromone within a single species . In conclusion , deception as an antipredatory strategy has been predominantly studied in the visual sense but is certainly widespread in other sensory modalities , too [61] . The few studies describing chemical deception show chemical matching but generally do not study the sensory or the mechanism underlying such deception ( for example , masquerade , background matching ) [62 , 63] . Our work here possibly demonstrates chemical deception being mediated by pheromones , which might additionally match an environment created by the ovipositing females . This also mechanistically differs from “chemical mimicry , ” for example , in which social parasites mimic host individuals [64] . While studies on animal deception have rarely used model systems , our study demonstrates that doing so opens the way for research on the sensory , neural , genetic , and mechanistic basis of deception at an unprecedented resolution .
All fly lines were reared on standard cornmeal/yeast medium ( 15 g agar , 30 g sucrose , 60 g glucose , 12 . 5 g dry yeast , 50 g cornmeal , 0 . 5 g MgSO4 , 0 . 5 g CaCl2 , 30 mL ethanol , 6 mL propionic acid , and 1 g nipagin per liter of water ) at 25°C and 70% relative humidity [26] . Canton S larvae and eggs were used as the wild-type line for most of the behavioral and biochemical assays ( Figs 1 , 2A , 2E , and Fig 3 ) unless explicitly mentioned otherwise . The eggs and larvae used for the assays were randomly allocated to different treatments . PromE ( 800 ) -Gal4 [4M] , Tub:Gal80ts flies , UAS-StingerII , UAS-Hid/CyO flies , and UAS-StingerII flies were used to generate crosses for oe ablation experiments ( Fig 2B and 2C; panel C in S1 and S4 Figs , panel A in S6 Fig , panel E–H in S7 Fig ) as described previously [39] . Briefly , apoptosis was induced in adult flies after eclosion , by overnight heat shock at 30°C for four consecutive days . However , little fluorescence was observed in some of these mutant flies even after the heat-shock procedure . Orco; IR25a; Gr33a-Gal4; UAS-hid , rpr; w1118; and BAC-rescue constructs for Orco and Ir25a were used to assay the influence of smell and taste on larval ability to cannibalize conspecific eggs [65 , 66] . ppk23 mutant larvae were used to assay their response towards 7 , 11-HD [54] . Most experiments were analyzed using ANOVA in JMP v . 10 . The proportional data from the following assays were arcsine-square–root transformed and analyzed appropriately using either an ANOVA or a Welch’s t test for unequal variance: egg vulnerability , egg toxicity , yeast perfuming , and larval attraction to injured eggs . The data from the egg perfuming assay were analyzed by logistic regression with overdispersion , wherein the four pheromones were included with interaction ( PROC GLIMMIX ) and pheromone concentration was log transformed and treated as a continuous variable . To simplify the analysis , the three replicates per pheromone concentration were pooled . Statistical analysis of data from mass spectroscopy is described in detail within the respective section below . All the high-resolution mass spectra were acquired in the mass range of 280–500 m/z with the AGC value of 5 × 105 . The mass spectra were externally calibrated with the calibration mixture containing caffeine , MRFA , and ultramark ( Buchs , Switzerland ) . Further , mass spectra were internally recalibrated using monoisotopic peaks of the three reference compounds ( cVA , 7 , 11-HD , and 7 , 11-ND; Cayman Chemical ) and a standard two-parametric calibration equation . First , the mass spectrum of hexane extract of wax layer of intact eggs of wild-type flies was analyzed ( Fig 2A ) . The full list of compounds identified in this mass spectrum is shown in the S1 Table . In general , several ion types of each identified compound were detected: M+˙ , [M+H]+ , [M+H-H2]+ , [M+H-2H2]+ , [M+H-H2+H2O]+ , and [M+H-2H2+H2O]+ . The same ion types and corresponding ionization mechanisms using an APPI source have been previously reported [69] . The mass spectra of some ions of interest are shown in S2–S4 Fig . Only the monoisotopic peak ( 12C ) of the main ion ( the highest abundance ) of an identified compound is indicated in S1 Table . Additionally , several target compounds were commercially synthesized ( Cayman Chemical ) , diluted in the hexane with the concentration of 0 . 5 mg/mL , and MS analyzed under identical experimental conditions ( AGC = 1 × 105 ) . The corresponding mass spectra , displaying four major ( the highest abundance ) ions of commercially synthesized cVA , 7-T , 7 , 11-HD , and 7 , 11-ND , demonstrate the same spectral composition ( ion types ) of the compounds as in the mass spectrum of hexane extract of the wax layer of intact eggs of wild-type flies ( S3 Fig ) . Further , hydrocarbon profile of the wax layer was investigated on eggs laid by four parental crosses generated from males and females , with ( oe+ ) or without ( oe− ) oes . The intensities of cVA , 7-T , 7 , 11-HD , and 7 , 11-ND were calculated from mass spectra of hexane extracts of wax layer of eggs laid by transgenic mutant fly crosses with/without ablated oes ( oe− ) : C1 , ♂oe+ × ♀oe+; C2 , ♂oe− × ♀oe−; C3 , ♂oe− × ♀oe+; and C4 , ♂oe+ × ♀oe− . Three mass spectra of each hexane extract ( C1–C4 ) were acquired in separate runs , 12 mass spectra in total . The intensity of each target compound is the summation of the SNRs of the monoisotopic peaks of the four most abundant ions , averaged through the three corresponding mass spectra ( Fig 2B and S2 Table ) . SNR was calculated as follows: SNR = Ipeak/ ( 5 ∙ δnoise ) , where Ipeak is the absolute spectral intensity of an analyte peak and δnoise is the SD of noise . S4 Fig demonstrates the expanded views of single mass spectra of hexane extracts C1–C4 plotted on the same figure . To correct drift in MS response ( number of charges’ variation between measurements ) , all the mass spectra were normalized by employing the total SNR . The total SNR of a mass spectrum was calculated as sum of intensities of all peaks higher than 5 ∙ δnoise . In general , the total SNR variation between measurements was less than ±10% . The surfaces of five dechorinated and five hexane-washed eggs of Canton S were examined by cryo-scanning electron microscopy ( S7A–S7D Fig ) . Similarly , five dechorinated eggs of each oe mutant cross were also examined by environmental scanning electron microscopy ( S7E–S7H Fig ) . For cryo-scanning electron microscopy , we used a Quorum system PP3010T attached to a Helios 650 ( FEI Company , Eindhoven , The Netherlands ) . The eggs were mounted on aluminum stubs using a mixture of Tissue-Tek ( Sakura Finetek Europe , Alphen aan den Rijn , The Netherlands ) and colloidal graphite ( Agar Scientific , Stansted , Essex , UK ) , frozen in nitrogen slush at −210°C and then transferred to the preparation chamber of the Quorum system . The sample was freeze-dried at −80°C for 10 min and then sputter coated with platinum at 10 mA for 25 s . After transfer on the cryostage at −140°C in the scanning electron microscope , imaging was performed at 5 keV using an Everhart-Thornley electron detector [70] . Some experiments were done under low-vacuum 300–400 Pa conditions . Fresh eggs were directly mounted on a scanning electron microscopy stub and imaged with the low-vacuum large field detector in the Quanta 250 . | Egg-laying species that lack parental care often protect their eggs from predators by laying them in communal groups or by fortifying them with toxins . However , these strategies may backfire when the predators are from the same species ( cannibals ) since a ) there are plenty of available eggs in these sites , b ) the cannibals may be resistant to the toxins , and c ) poisoning cannibals who may be related would reduce inclusive fitness . Under these circumstances , natural selection should favor anticannibalistic strategies that are likely to be nontoxic . Here , we investigate how fruit flies ( Drosophila melanogaster ) , which oviposit communally , protect their eggs from cannibalism by their own larvae . We show that maternal hydrocarbons incorporated into the egg’s wax layer to make them waterproof interestingly also serve as a mask that conceals their identity from cannibal larvae . In particular , we identify one female sex pheromone that deters cannibalism by forming a layer around the egg to conceal it . We further demonstrate that this pheromone is nontoxic and can mask underlying substrates such as yeast when used as a coating . While deceptive strategies ( such as camouflage ) deployed to avoid predation are extensively studied from a visual perspective , our findings suggest that deceptive strategies operating through other nonvisual sensory systems might be equally common across taxa . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"chemical",
"compounds",
"hexanes",
"animals",
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"waxes",
"materials",
"science",
"experimental",
"organism",
"systems",
"zoology",
"drosophila",
"research",
... | 2019 | Drosophila melanogaster cloak their eggs with pheromones, which prevents cannibalism |
Surgery , Antibiotics , Facial cleanliness and Environmental improvement ( SAFE ) are advocated by the World Health Organization ( WHO ) for trachoma control . However , few studies have evaluated the complete SAFE strategy , and of these , none have investigated the associations of Antibiotics , Facial cleanliness , and Environmental improvement ( A , F , E ) interventions and active trachoma . We aimed to investigate associations between active trachoma and A , F , E interventions in communities in Southern Sudan . Surveys were undertaken in four districts after 3 years of implementation of the SAFE strategy . Children aged 1–9 years were examined for trachoma and uptake of SAFE assessed through interviews and observations . Using ordinal logistic regression , associations between signs of active trachoma and A , F , E interventions were explored . Trachomatous inflammation-intense ( TI ) was considered more severe than trachomatous inflammation-follicular ( TF ) . A total of 1 , 712 children from 25 clusters ( villages ) were included in the analysis . Overall uptake of A , F , E interventions was: 53 . 0% of the eligible children had received at least one treatment with azithromycin; 62 . 4% children had a clean face on examination; 72 . 5% households reported washing faces of children two or more times a day; 73 . 1% households had received health education; 44 . 4% of households had water accessible within 30 minutes; and 6 . 3% households had pit latrines . Adjusting for age , sex , and district baseline prevalence of active trachoma , factors independently associated with reduced odds of a more severe active trachoma sign were: receiving three treatments with azithromycin ( odds ratio [OR] = 0 . 1; 95% confidence interval [CI] 0 . 0–0 . 4 ) ; clean face ( OR = 0 . 3; 95% CI 0 . 2–0 . 4 ) ; washing faces of children three or more times daily ( OR = 0 . 4; 95% CI 0 . 3–0 . 7 ) ; and presence and use of a pit latrine in the household ( OR = 0 . 4; 95% CI 0 . 2–0 . 9 ) . Analysis of associations between the A , F , E components of the SAFE strategy and active trachoma showed independent protective effects against active trachoma of mass systemic azithromycin treatment , facial cleanliness , face washing , and use of pit latrines in the household . This strongly argues for continued use of all the components of the SAFE strategy together .
The World Health Organization ( WHO ) promotes the SAFE strategy for trachoma control ( Box 1 ) which comprises: 1 ) Surgery , eyelid surgery to correct in-turned eyelashes that stops pain and minimizes risk of corneal damage [1]; 2 ) Antibiotics , treatment for active trachoma using single-dose oral azithromycin or tetracycline eye ointment [2]; 3 ) Facial cleanliness , clean faces especially in children through sustained behaviour change [3]; and 4 ) Environmental improvement , to increase access to water and sanitation [4] . Antibiotic therapy in individuals and facial cleanliness in children , combined with environmental improvement ( A , F , E components of SAFE ) , have been designed to treat ocular Chlamydia infection and reduce the risk of transmission of ocular Chlamydia . There is evidence from randomised controlled trials that the individual A , F , E components of the SAFE strategy have an effect on active trachoma when applied on their own: effect of antibiotics on active trachoma at three months [2]; effect of face-washing on trachomatous inflammation-intense ( TI ) [3]; and effect of fly-control on active trachoma at three and six months [5] , [6] . In these trials , the effects of these components have been tested individually to avoid the use of hybrid interventions that generate findings that are difficult to interpret . The challenge of disentangling the relative contribution to an effect of a combination of components has prevented trials of such hybrid trials being conducted . However , the reality of program implementation is that SAFE is a comprehensive and integrated strategy that has three control components ( the A , F , E ) that should be applied simultaneously . In the biological and epidemiological context , it is reasonable to expect there to be an additive effect of the A , F , E components of the SAFE strategy . Mass azithromycin administration , plus an increase in clean faces among children , plus improved access to water , plus reduced vector populations will likely be more effective than any one component alone . Despite the need for evaluation of the combined effects of A , F , E components of SAFE being highlighted previously [7] , no studies have explored the relationship between programme delivery of the combined A , F , E interventions and prevalence of active trachoma . Therefore we aimed to investigate the association between active trachoma and community intervention with A , F , E components of the SAFE strategy .
Surveys were conducted in four districts ( Kiech Kuon , Padak , Katigiri and Tali ) in Southern Sudan between April to June 2005 . The sample size estimation and sampling of the surveys has been described previously [8] . In brief , population based surveys were undertaken to estimate the prevalence of active trachoma signs before and following three years of implementation of the SAFE strategy . A two-stage cluster random sampling design was used to select the sample . In each district , 6 villages ( clusters ) were selected in stage one and 30 households selected in each cluster at the second stage . All eligible persons residing in the household were examined for trachoma and/ or interviewed . Only children aged 1–9 years , who had been examined for trachoma signs , were included in the sample ( Figure 1 ) . Eligible children were examined for trachoma signs by integrated eye care workers ( IECW ) using the WHO simplified grading scheme [9] . Trainee examiners had to achieve at least 80% interobserver agreement in identifying trachoma signs compared to the senior examiner to participate in the survey [10] . Clinical signs of inflammatory trachoma ( TF , and TI ) were graded for each eye separately . An ordinal severity score of active trachoma comprising three categories was then assigned to all eligible participants based on the worst affected eye: where ‘1’ was no TF , no TI; ‘2’ TF only; and ‘3’ TI with or without TF . We have previously justified the ordinal nature of inflammatory signs of trachoma [11]: 1 ) pathogenesis of trachoma is initially characterised by lymphoid follicles ( stage TF ) , whereas papillary hypertrophy ( stage TI ) is seen with advancing severity [12]; 2 ) participants seen in longitudinal studies with persistent TI are more likely to progress to scarring ( stage TS ) than those with only TF [13]; 3 ) subjects with TI are more likely to provide ocular swabs positive for Chlamydia trachomatis than those with TF [14]; and 4 ) patients with TI provide ocular swabs that have a greater quantifiable load of C . trachomatis than those with TF [15] , [16] . Prior to implementation of the trachoma control programme , baseline surveys were conducted to establish the need for interventions and to define programme targets [17] . The SAFE strategy was implemented in accordance with standards advocated by the WHO [18]–[21] and have been described previously [8] . Non-governmental organizations ( NGOs ) responsible for implementing healthcare , in the absence of Ministry of Health infrastructure , undertook implementation of the SAFE strategy . Surgery: Identification and screening of trachomatous trichiasis ( TT ) cases was conducted by community health workers trained in using the WHO simplified grading scheme [9] . TT surgeons were trained by an experienced ophthalmologist from the Christoffel Blindenmission ( CBM ) . Eyelid surgery was conducted using the bilamellar tarsal rotation procedure at primary health care centers ( PHCC ) and at annual “surgery camps” [18] . TT surgery logs were maintained at the PHCC . Trichiasis surgery is not considered further in this study of associations between A , F , E interventions and active trachoma . Antibiotics: Azithromycin ( Zithromax , donated by Pfizer ) was targeted for distribution to all villages in the four districts [19] . During the annual treatment round , teams of distributors including health care workers and trained village volunteer moved from village to village providing directly observed treatment to all eligible persons . Persons not eligible to receive Zithromax were treated with 1% tetracycline eye ointment . A tally of number of people treated with Zithromax was maintained and reported to the programme monthly . Three annual rounds were distributed in Kiech Kuon , Katigiri and Tali; however , the third round had not been distributed in Padak by the time of the evaluation survey . Facial cleanliness: Health education and hygiene promotion on facial cleanliness was conducted by trained hygiene promoters , women-peer educators and health workers A flip chart with key messages on facial hygiene and prevention of trachoma was used routinely to deliver health education at health facilities , schools , churches and during community outreach . Within villages , trained women peer-educators coordinated facial hygiene promotion activities . Heath education was delivered at least once a month to target groups or communities . Health education and hygiene promotion activities were monitored and reported monthly . Environmental change: Hygiene promoters and local leaders spearheaded community mobilization for construction of household pit latrines . The programme provided digging tools and technical advice on pit latrine construction . In two districts ( Tali and Katigiri ) fero-reinforced concrete slabs were provided free of charge to households for latrine construction . The number of latrines constructed was reported monthly . Advocacy for water provision to target populations by development NGOs involved in water activities was also conducted . Structured interviews with mothers of children as principal household respondents and observations were used to measure the uptake of the A , F , E interventions using the following definitions . Antibiotics: reports by care-givers of the number of times azithromycin treatment a child had received over three years . Facial cleanliness: a clean face was defined as the absence of nasal and/or ocular discharge on examination; frequency of face washing was reported by care-givers as the number of times faces of children were washed per day; and whether heads of households had received health education on trachoma at home or elsewhere . Environmental improvement: water access was report by people responsible for water collection as the time for a return journey to collect water; and pit latrine availability and use was ascertained by observation . Other explanatory variables included age , sex , cattle ownership ( reported and confirmed presence of cattle in the vicinity of the household ) , and baseline prevalence estimates of active trachoma for each study site . Statistical analysis was conducted using Stata 8 . 2 ( Stata Corporation , College Station , Texas ) . Descriptive statistics were used to examine the sample characteristics , prevalence of active trachoma signs , prevalence of A , F , E interventions , and other explanatory variables . Differences in proportions were compared by chi-square test . To investigate the association between severity of active trachoma signs and A , F , E interventions , ordinal logistic hierarchical regression models were developed using generalized linear models ( GLM ) [22] . The multilevel structure of GLM allowed for non-independence of the household and district variables , enabled clustering of individual observations within households , and districts , and allowed for variability at individual , household , village and district levels . We fitted an ordinal logistic regression model to study associations between severity of active trachoma signs and A , F , E interventions [23] . This model allowed for analysis of a polytomous ordinal response on a set of predictors and computed odds ratios ( OR ) of having a more severe active trachoma sign compared to a less severe sign . In this model TI was considered more severe than TF , which is in turn a more severe sign of trachoma than a normal conjunctiva . This method did not assume that TF causes TI or that the relationship between the three orders ( normal , TF only , and any TI ) is linear . Univariate analysis was conducted for each explanatory variable . Multivariable models were then developed by stepwise regression analysis for model selection . This involved starting with a null model then proceeding in a sequential fashion of adding/deleting explanatory variables if they satisfied the entry/removal criterion which was set at 5% significance level using a log-likelihood ratio test . Age and sex were retained in all multivariable models to control for any potential confounding effects . Potential effect modification was evaluated by including interaction terms in the models . Sensitivity analyses were conducted to investigate the effect of missing data by multiple imputations [24] . The Sudan Peoples Liberation Movement Secretariat of Health ( SPLM/Health ) approved the protocol and clearance to conduct the surveys was obtained from the local authorities . Verbal informed consent to participate was sought from the head of the household , household respondents and parents of children in accordance with the declaration of Helsinki . Consent for household interviews and trachoma examination was documented by interviewers and examiners on the data collection forms . Personal identifiers were removed from the data set before analyses were undertaken .
A total of 25 clusters ( villages ) and 743 households in four districts were surveyed ( in Tali district an additional cluster was selected due to insufficient number of households in one of the six primary clusters ) . Of the 1 , 867 children aged 1–9 years examined for trachoma , 1 , 712 ( 91 . 7% ) in 656 ( 88 . 3% of 743 surveyed ) households were included in the main analysis ( Figure 1 ) . Of the excluded households , 36 ( 4 . 8% ) did not have eligible children , 26 ( 3 . 5% ) had all eligible children ( 98 ) missing data on clean face , whereas 25 households ( with 57 children ) had missing data on health education and face washing frequency . The overall proportion of children aged 1–4 was 47 . 8% , the overall proportion of boys was 51 . 0% , and the mean age [standard deviation ( SD ) ] was 4 . 9 years ( SD = 2 . 5 ) . The overall prevalence of active trachoma severity scores was: “no TF , no TI” = 64 . 1% , range by study district ( 29 . 7–94 . 8 ) ; “TF only” = 23 . 1% , range ( 4 . 7–41 . 0 ) ; and “any TI” = 12 . 8% , range ( 0 . 5–35 . 9 ) ( Table 1 ) . Table 1 summarises the uptake of A , F , E interventions by study district . Overall , 53 . 0% of the eligible children had received at least one treatment with azithromycin , whereas 62 . 4% had a clean face on examination . Among the 706 households with eligible children: 72 . 5% reported washing faces of children two or more times a day; 73 . 1% had received health education; 44 . 4% had water accessible within 30 minutes; and only 6 . 3% had pit latrines . Univariate ordinal logistic regression analysis of the associations between severity of active trachoma and A , F , E interventions is shown on Table 2 . Factors associated with reduced odds of a more severe active trachoma sign compared with not having the risk factor were: older age ( 5–9 years compared to 1–4 years ) [odds ratio ( OR ) = 0 . 3; 95% confidence interval ( CI ) ( 0 . 5–0 . 9 ) ]; female sex ( OR = 0 . 7; 95% CI 0 . 5–0 . 9 ) ; azithromycin treatment ( one , two , or three treatments compared with no treatments ) ( p-trend<0 . 001 ) ; clean face ( OR = 0 . 2; 95% CI 0 . 1–0 . 2 ) ; increased frequency of face washing ( twice or more frequent daily face washing compared to once ) ( p-trend = 0 . 001 ) ; and pit latrine ( OR = 0 . 4; 95% CI 0 . 2–0 . 9 ) . There was no association between active trachoma severity and health education , and water access . Cattle ownership was associated with increased relative odds of a more severe active trachoma sign; however , this was not statistically significant: OR = 1 . 4; 95% CI ( 1 . 0–2 . 1 ) ; p = 0 . 066 . Table 3 shows the multivariable associations between active trachoma signs and A , F , E interventions adjusting for age , sex and baseline prevalence . The number of individual treatments with azithromycin was associated with reduced relative odds of having less severe active trachoma ( p-trend = 0 . 036 ) . Having received one or two treatments was associated with a 20% reduction in the relative odds of active trachoma; however , this was not statistically significant . Receiving three treatments was independently and strongly associated with reduction on the relative odds of active trachoma: OR = 0 . 1; 95% CI ( 0 . 0–0 . 7 ) . Clean face was also strongly associated with independent reduction in the relative odds of active trachoma: OR = 0 . 3; 95% CI ( 0 . 2–0 . 4 ) . Reports of washing faces of children two or more times was independently associated with reduced relative odds of active trachoma ( p-trend = 0 . 001 ) ; whereas households using pit latrines had 60% reduction in relative odds of active trachoma: OR = 0 . 4; 95% CI ( 0 . 2–0 . 9 ) . There was no evidence of interaction between the A , F , E interventions based on our effect modification models . Sensitivity analysis of the effect of missing data by multiple imputation of missing data and analysis of association between severity of active trachoma and A , F , E interventions revealed effect estimates similar to those in which missing data had been excluded ( data not shown ) .
This study of associations between the A , F , E components of SAFE and active trachoma showed independent protective effects against active trachoma of mass systemic azithromycin treatment , clean face on examination , reported face washing , and presence and use of pit latrines in the household . This study provides important evidence for continued advocacy for implementation of the full SAFE strategy for trachoma control . | Trachoma is an infectious disease that is cased by a bacterium , Chlamydia trachomatis , and is the leading cause of preventable blindness estimated to be responsible for 3 . 6% of blindness globally . The World Health Organization ( WHO ) recommends a strategy for trachoma control known as SAFE—surgery , antibiotics , facial cleanliness , and environmental improvement . Regular evaluations of trachoma control activities are advocated for by the WHO for decision making , programme planning , and the rational use of programme resources . We undertook a survey to evaluate the effectiveness of the SAFE strategy following three years of interventions in four districts in Southern Sudan . In this paper , we aimed to find out the relationship between the antibiotics , facial cleanliness , and environmental improvement ( A , F , E ) and active trachoma signs . Our study revealed that prevalence of active trachoma was less in children who had received treatment with azithromycin , had clean faces , had faces washed more frequently , and used latrines compared to children who had not received these interventions . The study findings are important since they make the case for implementing the A , F , E interventions together . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"ophthalmology/eye",
"infections"
] | 2008 | Associations between Active Trachoma and Community Intervention with Antibiotics, Facial Cleanliness, and Environmental Improvement (A,F,E) |
The Gram-negative bacterium Bordetella pertussis is the causative agent of whooping cough , a serious respiratory infection causing hundreds of thousands of deaths annually worldwide . There are effective vaccines , but their production requires growing large quantities of B . pertussis . Unfortunately , B . pertussis has relatively slow growth in culture , with low biomass yields and variable growth characteristics . B . pertussis also requires a relatively expensive growth medium . We present a new , curated flux balance analysis-based model of B . pertussis metabolism . We enhance the model with an experimentally-determined biomass objective function , and we perform extensive manual curation . We test the model’s predictions with a genome-wide screen for essential genes using a transposon-directed insertional sequencing ( TraDIS ) approach . We test its predictions of growth for different carbon sources in the medium . The model predicts essentiality with an accuracy of 83% and correctly predicts improvements in growth under increased glutamate:fumarate ratios . We provide the model in SBML format , along with gene essentiality predictions .
B . pertussis is a Gram-negative bacterium that causes whooping cough , a respiratory infection responsible for significant annual mortality worldwide [1 , 2] , especially among infants and young children . B . pertussis is described as a fastidious organism . It does not metabolise sugars as carbon source as it does not possess an intact glycolysis pathway [3] . Amino acids appear to be the primary carbon sources for growth . B . pertussis can grow using most of the amino acids as a carbon source , however alanine , proline and glutamate are utilized preferentially suggesting that amino acids that are degraded to α-ketoglutarate or pyruvate are oxidized rapidly . Several studies have demonstrated that glutamate is by far the most efficiently metabolized and is considered to be the main carbon source for growth of B . pertussis [3–5] , which can be grown in the lab using solely glutamate as a carbon source and cysteine as a source of sulphur ( along with salts and some vitamins ) . It was a long-held view that the TCA cycle was not completely functional in B . pertussis . This stemmed from the inability of B . pertussis to utilise citrate as a carbon source along with observations of the build up of poly-hydroxybutyrate and release of free fatty acids in batch cultures . However , the B . pertussis genome contains genes that appear to encode a complete pathway [6] . Recently , demonstration of citrate synthase , aconitase and isocitrate dehydrogenase activities in B . pertussis gave a clear indication that the TCA cycle is fully functional , although it remains unclear why citrate does not support B . pertussis growth [7] . Commonly used media for broth growth , such as Stainer-Scholte ( SS ) broth [8] , contain glutamate as the main carbon source . Modified SS broth contains casamino acids and heptakis , and growth is enhanced by these additions . Casamino acids probably increase the level of glutamate and enable utilization of other amino acids . Heptakis , a cyclodextrin , absorbs free fatty acids that are inhibitory towards B . pertussis growth [9] . However , culture of B . pertussis in SS broth leads to an imbalance in N:C ratios leading to the formation of ammonium which is inhibitory to growth , resulting in relatively low final cell densities . Several studies have investigated parameters affecting the growth rate of B . pertussis using either batch cultures or steady state cultures in bioreactors ( for example see [3 , 5 , 10 , 11] ) . These informative studies revealed much of what is known about B . pertussis growth parameters , identifying the importance of balancing N:C ratios , avoiding excessively high substrate concentrations and the effect of salt concentrations for attaining high biomass yields . The slow growth and limited yields of B . pertussis in culture are important limitations to the efficiency of B . pertussis vaccine production . In particular , at least five times more culture volume is required to generate one dose of an acellular pertussis vaccine compared to a whole cell one . Expansion of B . pertussis vaccination programmes using acellular vaccines , either into the developing world that for the most part use whole cell vaccines , or to increase the use of booster doses for adolescents/adults would place strain on global production of these vaccines . Increased efficiency of B . pertussis culture would help to alleviate these strains but this requires greater knowledge of the growth characteristics of B . pertussis . Flux balance analysis ( FBA ) is an established approach for modelling the metabolic networks of organisms at the genome scale , and is a framework for integrating other ‘omics data layers with metabolism [12–16] . Briefly , the network of metabolic reactions in an organism is represented by an m × n stoichiometric matrix , S . Each row of S represents a metabolite and each column gives the stoichiometry for a particular metabolic reaction . There are m metabolites and n reactions . The list of metabolites includes both so-called “internal metabolites” , which are not exchanged with the growth medium or environment , and “external metabolites” , which are . External metabolites include nutrients in the modelled growth medium , metabolites that diffuse in and out of the cell , and by-products of growth that leave the cell . FBA models make the approximation that the time scale of interest ( hours or longer ) is long enough that short-term transients in the kinetics of individual reactions ( which would usually dissipate in seconds or minutes ) will have largely passed , so that reactions are running at steady state: there is no net production or consumption of ( internal ) metabolites . Mathematically , each reaction is associated with a flux v; the steady-state approximation is the constraint Sv = 0 . The specific growth medium and uptake rates mean that there are constraints on how fast the influx of nutrients can be; mathematically , this means that there are constraints on some or all of the reaction fluxes . Finally , FBA models describe the growth capacity of an organism using an objective function c: how much of the given objective could the metabolic network possibly produce , at steady state , under the given constraints ? The objective is typically a biomass vector , c , describing the major components of the dry weight of the cells . FBA models then approximate the metabolic network’s capacity to produce this biomass under various conditions . FBA is performed by solving a linear programming problem: max c · v s . t . S v = 0 a r ≤ v r ≤ b r . ( 1 ) where S is the stoichiometic matrix , v is a vector of reaction fluxes , c is the objective function , and ar and br are vectors of length n describing lower and upper constraints on the reaction fluxes . A growth medium is defined by setting constraints so as not to allow uptake of nutrients that are not present in the medium . In principle , an FBA model can be constructed directly from an annotated genome; where a gene’s enzymatic function is known , the relevant reaction and metabolites can be added to the system and the stoichiometric matrix can be constructed so as to capture the ( usually conserved ) stoichiometries of the included reactions . In practice , genome annotation and functional prediction is imperfect , and FBA models require substantial curation [17] . This typically requires first constructing a draft FBA model based on the annotation in an automated way , then examining each reaction in S and determining whether it describes realistic biochemistry , as well as examining the gene ( s ) associated with it , their annotation in the organism and whether the gene-reaction relationship is appropriate . This process requires considerable knowledge of the organism’s biochemistry , and is labour intensive [17–20] . Previously , dynamic models of limited compartments of B . pertussis were developed and demonstrated the utility of this approach for interrogating specific facets of B . pertussis metabolism . Here , we present the first published genome-scale metabolic reconstruction for B . pertussis . It is suited for flux balance analysis , and models B . pertussis’ metabolic reactions accordingly . We refer to this reconstruction as “the model” or “metabolic model” throughout . To demonstrate the use of the model to interrogate B . pertussis growth , we used it to predict reactions that are essential for growth on laboratory medium . We tested these predictions by performing a genome-wide screen for essential genes using a Transposon-directed Insertional Sequencing ( TraDIS ) approach [21] and demonstrate a high degree of concordance between model predictions and experimental observations . We used the model to investigate the reduction of ammonia production that occurs during growth in standard medium , and tested the predictions arising . The development of a genome-scale model provides a valuable tool for investigating the growth of this bacterium .
Essential genes were identified using Transposon Directed Insertion-Site Sequencing ( TraDIS ) [21] . Saturated transposon libraries were constructed using the pBAM1 delivery vector [28] , modified with PmeI restriction sites for digestion of vector-derived amplicons prior to sequencing . The details of construction of the transposon library , sequencing of insertion sites and analysis of insertion site frequency followed the approaches described previously for TraDIS [29] . Three independent transposon libraries were made . Each were plated on charcoal agar ( Oxoid ) supplemented with 50μg/mL kanamycin and incubated at 37°C for 72 hours . Between 300 000 and 500 000 transposon mutants were harvested per library and processed for TraDIS . Insertion indexes were calculated for each gene and essentiality calculated using the cut off point described previously [21] .
Extensive curation of the preliminary model was performed . Key changes are discussed below . The initial ModelSEED model file and a detailed curation history file are included as supplemental files to allow specific aspects of our curation and the effects of alternative curations to be investigated . Reactions that were automatically gap-filled were analysed . Based on known behaviours of B . pertussis , gap-filled reactions were removed to create true gaps ( e . g . nicotinate , cysteine auxotrophy [30] ) , removed as the reactions do not occur in B . pertussis ( e . g . 11 reactions specific to synthesis of E . coli rather than B . pertussis LPS ) , or genes identified that encode the probably missing function . This process left only 10 gap-filled reactions for which no gene assignment exists . These reactions are listed in S2 Table . PLP is an essential cofactor . The SEED model included two gap-filled enzymes corresponding to the PdxT/PdxS catalysed generation of PLP from glyceraldehyde-3-phosphate and ribulose-5-phosphate , as characterised in B . subtilis . However , there are no homologs of pdxT or pdxS in B . pertussis . An alternative well characterised pathway for PLP synthesis can occur via the activities of PdxB , PdxA and PdxJ . Clear homologs of both pdxA and pdxJ are evident in B . pertussis . PdxB is 4-phosphoerythronate dehydrogenase , an oxido-reductase enzyme . These enzymes generally show low levels of sequence conservation between homologs . Using BlastP of the E . coli PdxB sequence against the B . pertussis genome identified 4 putative dehydrogenases with scores in the range of 3e-10 to 5e-15 . Thus , it was concluded that there are potential PdxB candidates in B . pertussis and as PLP synthesis is expected to be essential , gap-filling of the PdxB-catalysed reaction was more logical than that of the PdxT/PdxS reaction . The SEED model filled gaps in the reactions catalyzed by quinolinate synthase , encoded by nadA and L-aspartate oxidase , encoded by nadB . There are no clear homologs of nadA or nadB in B . pertussis and this bacterium is auxotrophic for nicotinate , which is a component of the B . pertussis growth media . Thus , it is expected that the nicotinate synthesis pathway is incomplete in B . pertussis . These reactions were changed to true gaps in the model . The reaction catalyzed by this enzyme is a critical step in the synthesis of the cofactor heme . In B . pertussis there are no identifiable homologs of genes encoding the HemG or HemY members of this family of enzymes , although the remainder of the pathway appears to be present . In some other bacteria missing HemG/Y an alternative gene , hemJ , encodes this activity . BP2372 was identified as a potential hemJ homologue and was not associated with any other reaction in the model . Thus , the model was curated to include BP2372 as performing this step . Thiamine phosphate is a crucial cofactor . The SEED model contained thiamine phosphate biosynthesis based on the pathways described in E . coli in which ThiH catalyses the production of 4-hydroxy-benzylalcohol from tyrosine . However , in the model , 4-hydroxy-benzylalcohol is a dead-end metabolite , as it is not used in any pathway and the model constrains all fluxes producing dead-end metabolites to zero . There is no obvious homologue of ThiH in B . pertussis . It was reasoned that the biosynthesis more closely resembles the pathway described in B . subtilis involving ThiS , ThiF and ThiG for which there are obvious homologs in B . pertussis ( encoded by BP3690 , BP0610 and BP3597 respectively ) along with thiazole tautomerase , TenI ( BP3809 ) and ThiE ( BP0316 ) . The model was curated to include this biosynthetic pathway . The SEED metabolic models include LPS biosynthesis based on the E . coli LPS structure . The structure of B . pertussis LPS is known , and the genetics of its biosynthesis is well-characterised [31–33] . Reactions for synthesis and assembly of the B . pertussis LPS molecule were substituted for the E . coli-based reactions , and the associated B . pertussis genes were assigned to these reactions . This involved modifying the reactants and products of two reactions , the addition of nine new reactions and removing thirteen of the E . coli LPS-specific reactions . LPS is most abundant molecule in the outer leaflet of the outer membrane of gram negative bacteria . Constructing an accurate B . pertussis LPS biomass component enhances the accuracy of the model . Several reactions involving electron transfer were set by ModelSEED to operate in the opposite direction to the thermodynamically feasible direction for electron transport , producing unfeasibly large fluxes at no energetic cost . The direction of these transfers was reversed , S3 Table . Previous studies have identified a number of carbon sources that either can or can not be metabolised by B . pertussis [3 , 4 , 8 , 25] . Exchange reactions were modified to include the uptake of the metabolisable carbon sources , along with ammonia that can be used as a source of nitrogen by B . pertussis: pyruvate , L-aspartate , L-arginine’ , L-alanine , L-glycine , L-histidine , 2-oxoglutarate , malate , L-lactate , ammonia . The requirement that all metabolites remain at a constant concentration is a central approximation in FBA , and this places a basic limit that all metabolites must appear at least twice in the model if they are to take an active part in any fluxes . As a direct consequence , any reaction that contains a singularly-appearing metabolite ( a dead-end metabolite ) has its flux constrained to zero , regardless of the state of the rest of the network . Removing these metabolites and reactions from the model entirely has no impact on the model’s results . Our curated B . pertussis model contains 301 singleton metabolites , which take part in a total of 199 reactions , consequently all blocked . Assuming the annotations and associated genes are correct , their presence points to further missing reactions , completing the pathways from which they come . Alternatively , these reactions are the remnants of pathways from which enzymes are missing due to the extensive gene loss that has been a feature of B . pertussis evolution [6] . This extensive gene loss may have produced an unusually high number of degraded pathways . In this scenario , the reactions may be occuring but be producing dead-end metabolites . Given this uncertainty , they have been left in the model , but indicated with the note annotation blocked:True . The biomass composition of B . pertussis was measured using triplicate cultures ( see Methods ) : as percentage of dry cell weight , 53 . 9 ( +/- 2 . 7 ) protein , 5 . 5 ( +/- 1 . 9 ) carbohydrate , 4 ( +/- 0 . 5 ) DNA , 3 . 5 ( +/- 0 . 5 ) RNA and 9 . 5 ( +/- 1 ) lipids . The BOF was tuned to incorporate these proportions of macromolecules . Gene essentiality was determined using the TraDIS approach . Three independent transposon libraries containing 300 000–500 000 colonies each were constructed . Insertion indices were calculated for each genes as described previously [21] ( see Methods ) . This identified 415 genes as essential for growth under these conditions . A further 26 genes were ambiguous in terms of their essentiality but were not classed as essential in these studies . However , only 11 of the ambiguous genes appear in the model ( S4 Table ) . One ( BP3151 ) is associated with a singleton metabolite and thus a blocked reaction , and six others are part of multigene complexes ( ribosomes , NADH dehydrogenase , DNA replication ) formed by other essential genes and thus are associated with essential pathways/reactions , resulting in just four reactions associated with ambiguously essential genes appearing in the model . Fig 1 shows ROC curves for FBA classification of gene essentiality , comparing model predictions of essentiality with experimentally defined essential genes . The AUC score demonstrates good classification . Fig 1 also shows as a red dot the selected threshold , chosen as the closest point to the perfect performance of ( 0 , 1 ) . In Table 2 we give the raw scores for the chosen threshold , divided into true and false positives and negatives . We present the results in a standard contingency table , identifying the types of errors made , as well as giving an overall accuracy score ( calculated as ( TP + TN ) / ( TP + FP + TN + FN ) ) . The reactions for each of these categories are listed in S5 Table . When applying the FBA knockout approach to our network of metabolic reactions and associated genes , we achieve an accuracy of 83% in predicting the experimental essentiality . This compares well with scores achieved by other published metabolic models , and a perfect score is not to be expected , due both to experimental and theoretical considerations . While TraDIS is a state of the art approach , we cannot expect perfect results from TraDIS due to limitations in detecting extremely slow growing ( but viable ) mutants , and while our metabolic model reflects the current state of knowledge for B . pertussis metabolism , there remain uncharacterised proteins that may impact the performance of the network . Even accounting for errors in both TraDIS and the model , furthermore , FBA is an approach focused solely on the metabolic capabilities of an organism . There are regulatory and kinetic considerations that are beyond the scope of the FBA approach , but will nonetheless play a key role in the viability of knockout mutants . These considerations are likely to make perfect prediction an infeasible goal . Information on essential genes was used to refine some gene assignments for reactions . A number of reactions predicted to be essential had more than one possible gene assigned to them where it was not clear which gene was the correct assignment . In cases where one of the genes was shown to be essential , gene assignments were amended to show only this gene , as genes assigned to essential reactions also should be essential ( S6 Table ) . A key use of metabolic models is to be able to make predictions of organism metabolism that can be investigated experimentally . To test our model , we sought to make predictions of changes to media formulations that decrease the production of growth inhibiting ammonia , without diminishing predicted growth rate . Ammonia production is thought to arise from an imbalanced N:C ratio when B . pertussis utilises glutamate as its sole carbon source [3] . To investigate this , we modelled the effect of shifting from growth on glutamate towards growth using glutamine ( Fig 2a ) . Glutamine contains two amino groups compared to the one of glutamate . The model predicts that growth rate is unaffected whereas production of ammonia increases as the metabolism of glutamine over glutamte increases . Next , we modelled the effect of metabolising different ratios of glutamate and fumarate ( Fig 2b ) . Fumarate is an alternative carbon source but does not contain nitrogen . B . pertussis requires a nitrogen source to grow . If the uptake of ammonia as a source of nitrogen is prohibited then there is no growth in the model . However , as an increasing amount of glutamate is metabolised , with the corresponding decrease in fumarate metabolism , growth rate increases up to a point and the production of ammonia increases once a threshold ratio of glutamate:fumarate metabolised is reached . If this analysis is repeated allowing free uptake of ammonia , then the growth rate is unaffected by the ratio of glutamate:fumarate but ammonia is consumed up to a point when the metabolism of glutamate provides sufficient nitrogen , and ammonia is produced when the ratio of glutamate:fumarate metabolised reaches the point of imbalance between N:C ( Fig 2c ) . This identified an approximate 1:2 ratio of glutamate to fumarate ( in terms of contribution of carbon atoms rather than molecular mass ) as an N:C balance at which ammonia production was minimised , but growth rate was unaffected , when the medium does not contain available ammonia . We tested this prediction experimentally by growing B . pertussis in different SS medium formulations in which carbon was provided by different ratios of glutamate:fumarate . The growth of B . pertussis was followed by measuring the absorbance of the culture ( Fig 3a ) and the concentration of ammonia was measured in cultures at the end point of growth ( Fig 3b ) . Growth in media using solely glutamate as a carbon source resulted in relatively poor biomass yield and a relatively slow growth rate compared to media containing fumarate as a replacement for at least some of the glutamate . A glutamate:fumarate ratio of 5:1 produced moderate improvements in both rate and yield . Ratios of 2:1 , 1:2 and 1:5 all gave dramatic improvements in rate and yield . The total amount of carbon in each medium was the same , suggesting that differences in biomass yields between cultures was most likely due to differing levels of inhibition of growth as opposed to nutrient limitation . Interestingly , replacement of some of the glutamate in the medium with fumarate resulted in a significant reduction in the level of ammonia produced by B . pertussis , on a ammonia per OD unit basis . A glutamate:fumarate ratio of 5:1 gave the greatest reduction while other ratios resulted in similar levels of ammonia . We suggest that the poor growth of the culture growing solely on glutamate was due to inhibition of growth by the resulting ammonia that was produced . The data demonstrate the model prediction to be largely correct in that balancing N:C ratios by the addition of fumarate reduced the production of ammonia , but that additional factors are evident as the growth of the cultures were clearly different from each other . This highlights the need for development of genome scale metabolic modeling to incorporate regulatory and non-metabolic constraints on growth .
We have developed and curated the first published genome-scale FBA model for B . pertussis , and have included an experimentally-determined biomass . The model predicts essential genes with 83% accuracy , compared with the state-of-the-art determination of essential genes with the TraDIS technique . The model and related computations are available in python in the pyabolism module . In contrast with our curated model , the automated SEED model based on the annotated B . pertussis genome cannot produce biomass on the standard growth medium for B . pertussis ( SS broth ) . Extensive curation is typically required for genome-scale metabolic models [17] , and in our case , this curation made fundamental differences to the model metabolism , enabling both growth on SS broth and accurate classification of essential genes . While FBA models have extensive potential for applications , there are several remaining challenges . In particular , while genome annotation and function prediction are improving , the presence of genes classed as ‘hypothetical protein’ or with unknown function , and the presence of mis-classified genes , means that even with curation the accuracy of reconstructed models can be limited . This is a particular challenge for less-studied organisms; FBA models perform extremely well for well-characterized organisms such as E . coli . [20] . Even if the stoichiometric matrix were able to perfectly capture the metabolic reactions in an organism , there are reaction kinetics , regulatory interactions , the dynamics of transcription and translation and other important processes that are not captured in constraint-based models . Despite these limitations , the number of interesting applications in diverse micro-organisms has grown tremendously in recent years [34–38] . For this field to yield the results that have been promised , it is essential that the community develop and curate FBA models for more organisms—as we have done here . B . pertussis presents some unique challenges and opportunities for constraint-based metabolic modeling . For example , B . pertussis evolved from its ancestor ( B . bronchiseptica , or a B . bronchiseptica-like relative ) by a process of genome reduction and rearrangement [6] . This has resulted in a large number of pseudogenes , which were not always recognised as being non-functional by the automated model construction . Also , gene loss has resulted in a number of incomplete , presumably remnant , metabolic pathways which automated gap filling attempts to ‘correct’ by adding missing functions , on the assumption that a pathway that was mostly present must be fully functional . The raw SEED model was unable to produce biomass when simulations were run using the components of the standard growth medium for B . pertussis , SS broth , as inputs . Thus , the production of a metabolic model that mimics the known characteristics of the organism required extensive and laborious manual curation . B . pertussis is considered a re-emerging pathogen , with pertussis disease resurgent in numerous countries [39] . This has been associated with a change from the use of first generation , whole cell to second-generation , acellular pertussis vaccines . This resurgence has generated renewed interest in understanding the physiology and infection biology of B . pertussis . Understanding the basic growth of the bacterium is key to this , and a genome scale metabolic model is a widely applicable tool towards this goal . In addition , millions of doses of pertussis vaccines are used globally each year . An increase in demand for these vaccines , through either replacement of whole cell with acellular vaccines in more parts of the world , or expanded use of booster vaccinations to combat resurgence , will generate considerable strain on the global vaccine supply . Enhancement of the vaccine production process through shorter production times and increased yields from production will be important to meeting any increased demand . Understanding , and the ability to manipulate , B . pertussis growth characteristics is important towards this aim . The genome-scale metabolic model described here provides a novel tool to investigate B . pertussis growth and physiology . In particular , it allows the effects of altered medium formulations or genetic manipulation of metabolism to be investigated in silico , enabling much more targeted experimental investigations than are currently possible . The alteration of B . pertussis growth by substituting fumarate for some of the glutamate in standard media demonstrate the validity of this approach . | Metabolic flux models have been used to understand how organisms adapt their metabolism under different growth conditions , and are finding increasing application in synthetic biology and biotechnology . One barrier to progress in this field is the construction and curation of metabolic flux models for new organisms . Here we present a curated genome-scale metabolic flux model for Bordetella pertussis , the causative agent of whooping cough . Producing vaccines against whooping cough requires growing B . pertussis in large volumes . However , its growth is relatively slow , final yields of biomass are relatively low and growth characteristics can be variable . Understanding B . pertussis metabolism has applications to improving vaccine production , as well as in understanding the basic biology of this organism . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"bordetella",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"pathogens",
"metabolic",
"networks",
"microbiology",
"neuroscience",
"vaccines",
"metabolites",
"network",
"analysis",
"neurotran... | 2017 | A curated genome-scale metabolic model of Bordetella pertussis metabolism |
Rift Valley fever virus is an arbovirus that affects both livestock and humans throughout Africa and in the Middle East . Despite its endemicity throughout Africa , it is a rare event to identify an infected individual during the acute phase of the disease and an even rarer event to collect serial blood samples from the affected patient . Severely affected patients can present with hemorrhagic manifestations of disease . In this study we identified three Ugandan men with RVFV disease that was accompanied by hemorrhagic manifestations . Serial blood samples from these men were analyzed for a series of biomarkers specific for various aspects of human pathophysiology including inflammation , endothelial function and coagulopathy . There were significant differences between biomarker levels in controls and cases both early during the illness and after clearance of viremia . Positive correlation of viral load with markers of inflammation ( IP-10 , CRP , Eotaxin , MCP-2 and Granzyme B ) , markers of fibrinolysis ( tPA and D-dimer ) , and markers of endothelial function ( sICAM-1 ) were all noted . However , and perhaps most interesting given the fact that these individuals exhibited hemorrhagic manifestations of disease , was the finding of a negative correlation between viral load and P-selectin , ADAMTS13 , and fibrinogen all of which are associated with coagulation pathways occurring on the endothelial surface .
Rift Valley Fever Virus ( RVFV ) was recently re-classified in the Phenuivirdae family [1] , genus Phlebovirus [2] . Rift Valley Fever ( RVF ) outbreaks in animals can result in a large number of livestock deaths , including spontaneous abortions in pregnant livestock [3] . Transmission to humans can occur through direct contact with an animal’s infected tissue or bodily fluid or fomites . RVFV can also be transmitted via mosquito bite to both animals and humans [3] . Humans are at particularly increased risk of acquiring RVFV infection while assisting with animal birth or during animal slaughter . For these reasons , populations at high risk of RVF include herdsmen , butchers and abattoir workers [4] . Outbreaks of RVF have occurred throughout Africa and the Arabian Peninsula . Recent outbreaks have occurred in South Africa , Kenya , Sudan , Saudi Arabia , Tanzania , and Yemen [3 , 5] . Humans infected with RVFV can present with varying degrees of symptomatology ranging from asymptomatic or a mild influenza-like illness to severe disease with hepatitis , retinitis or encephalitis [6] . Laboratory findings in patients with RVF can include leukopenia , thrombocytopenia , and elevated liver enzymes . About 1% of RVF cases may progress to hemorrhagic disease . The overall fatality rate is estimated to be 0 . 5–1% , however in the 2006–2007 outbreak in Kenya , the case fatality rate was reported to be as high as 29% [7] . It is not well understood why there is such variation in the clinical presentation of RVF . The interaction of viral proteins with the host immune response may play a role in pathogenicity . RVFV is an enveloped negative strand RNA virus consisting of three RNA segments varying in size: large ( L ) , medium ( M ) and small ( S ) . The S segment encodes the N and NSs proteins; the M segment encodes the Gn , Gc , and NSm proteins; whereas the L segment encodes the polymerase [2] . The NSs protein is a non-structural protein that interacts with host transcription factors and may down regulate the host innate immune response during infection [8 , 9] . The interaction between the immune system , coagulation , and the endothelium during hemorrhagic fever virus infection has been previously studied in human subjects with Lassa , Ebola , and dengue virus infections [8 , 10–17] . However , there are few studies of human immune responses during RVFV infection and none that examine the endothelium or coagulation pathway [8 , 18 , 19] . A study of South African patients infected during a 2010/2011 RVF outbreak found that fatal RVF cases had increased levels of the pro-inflammatory cytokines IL-8 , IP-10 , CXCL9 , and MCP-1 as well as increased levels of the anti-inflammatory cytokine IL-10 [18] . Genetic polymorphisms in innate immune pathways , including IL-6 , may also contribute to variations in severity of RVFV symptoms [19] . These data suggest that dysregulation of the immune response during RVFV infection may contribute to pathogenesis . In March 2016 , two human cases of RVF were identified in Uganda[20]; these were the first RVF cases in Uganda since 1968 [21] . A third RVF case was later identified in June 2016 . These three cases provided a unique opportunity to examine the physiologic consequences of RVFV infection in the human host using blood samples collected serially as part of clinical care . We analyzed the concentration of a panel of cytokines/chemokines , as well as biomarkers of endothelial function and coagulopathy to look for associations between these pathways , host disease , and viral replication over time .
This study used de-identified pre-existing specimens and clinical records obtained during an outbreak and was determined to be exempt by the Centers for Disease Control and Prevention Human Research Protection Office . CDC Protocol Number: 6920 . Blood samples obtained in EDTA tubes as part of routine clinical care for RVF patients were analyzed using commercial multiplex assays according to the manufacturer’s instructions . The samples were aliquoted , frozen , and gamma irradiated on dry ice with 5x106 rads . In preparation for use in the assays , samples were thawed and centrifuged to remove membrane particles . The freeze thaw process and use of whole blood in these assays were previously demonstrated using 8 human blood samples from healthy controls obtained in the US [22] . Eight healthy human samples were also used to define the normal range for each analyte in this study . These samples were obtained via a normal healthy phlebotomy program in Atlanta , GA . Sex , age and race/ethnicity of the donors were unknown . Fifty-nine analytes were assessed in ten commercially available assays from Invitrogen ( Carlsbad , CA , USA ) , Millipore ( Billerica , MD , USA ) , and Affymetrix ( Santa Clara , CA , USA ) . The largest of these was a 30-plex assay for granulocyte colony stimulating factor ( G-CSF ) , granulocyte macrophage colony stimulating factor ( GM-CSF ) , interferon ( IFN ) -α , IFN-γ , interleukin ( IL ) -1β , IL-1RA , IL-2 , IL-2R , IL-4 , IL-5 , IL-6 , IL-7 , IL-8 , IL-10 , IL-12 ( p40/p70 ) , IL-13 , IL-15 , IL-17 , tumor necrosis factor ( TNF ) -α , Eotaxin , interferon gamma induced protein 10 ( IP-10 ) , monocyte chemoattractant protein ( MCP ) -1 , monokine induced by gamma interferon ( MIG ) , macrophage inflammatory protein ( MIP ) -1α , MIP-1β , regulated on activation normal T-cell-expressed and secreted ( RANTES ) , endothelial growth factor ( EGF ) , fibroblast growth factor ( FGF ) -basic , hepatocyte growth factor ( HGF ) , and vascular endothelial growth factor ( VEGF ) that was performed using an overnight incubation ( Invitrogen ) . Five additional assays also incubated overnight included three singleplex assays for Ferritin , ADAMTS ( a disintegrin and metalloproteinase with thrombospondin motifs ) -13 , and complement factor H ( CFH ) , a two-plex assay for tissue factor and thrombomodulin , and a four-plex assay for von Willebrand’s factor ( vWF ) , c-reactive protein ( CRP ) , Fibrinogen , and platelet factor ( PF ) -4 ( Millipore ) . One single-plex assay for IFN-β , an eleven-plex assay for E-selectin , fractalkine , granzyme B , melanoma growth stimulation activity alpha ( GRO-α ) , IL-29 , L-selectin , MCP-2 , MCP-3 , sCD40L , TNF-R2 , tissue plasminogen activator ( tPA ) , and a six-plex assay for D-dimer , plasminogen activator inhibitor ( PAI-1 ) , platelet endothelial cell adhesion molecule ( PECAM ) , P-selectin , sFas-ligand , TNF receptor ( TNF-R1 ) were performed using a 2 hour incubation ( Affymetrix ) . A two-plex assay for intercellular adhesion molecule ( ICAM ) and vascular cell adhesion molecule ( VCAM ) was run with an overnight incubation ( Affymetrix ) . Data were collected on a Luminex 200 ( Austin , TX ) and all assay results were reported in pg/mL , ng/mL , or mg/dL . For 11 of the biomarkers evaluated ( IFN-β , G-CSF , IL-5 , IL-13 , IL-17 , GM-CSF , TNF-α , IL-2 , IL-7 , IL-4 , and tissue factor ) levels of the biomarker were near or below the limit of detection in all patient samples , so data are not presented graphically . RNA was isolated from gamma-irradiated whole blood using MagMax Total RNA Isolation Kit ( Life Technologies , Grand Island , NY ) . qRT-PCR of RNA was conducted with established primer and probe sets for the RVFV L segment [23] . A standard curve , expressed as tissue culture infective dose 50 ( TCID50 ) /mL , was generated from normal human whole blood spiked with a wild-type RVFV stock . The standard curve was used to convert raw Ct values to relative TCID50/mL . Maxisorp plates ( Nalgene-Nunc , Rochester , NY ) were coated with RVFV lysate prepared from infected Vero-E6 cells as previously described [24] , diluted 1:2000 in PBS , and allowed to adsorb overnight at 4°C . Plates were blocked in 5% milk in PBS with 0 . 1% Tween-20 ( PBST ) with 5% FBS for 1 hour at 37°C . Patient samples and controls were serially diluted in blocking buffer and incubated on plates for 2 hours at 37°C . Plates were washed three times in PBST then incubated for 1 hour at 37°C with anti-human IgG HRP or anti-human IgM HRP ( Jackson ImmunoResearch Inc , West Grove , PA ) diluted 1:5000 in blocking solution . Plates were washed three times in PBST before incubation in TMB substrate ( KPL , Milford , MA ) for 10 minutes . Reactions were stopped with TMB stop solution , and plates were read at 450nm . Data were analyzed using Excel ( Microsoft Corp , Redmond , WA ) and Prism ( GraphPad Software Inc , La Jolla , CA ) . Descriptive analyses were performed using STATA 13 . 0 ( StataCorp , College Station , TX ) and Prism . Patient means were compared to those of the controls . P-values for all statistically significant comparisons are included in the supplementary table ( S1 Table ) . Patient biomarkers were analyzed for correlation with RVFV viral load using Spearman’s rank correlation . Control biomarker values were compared to case values during two disease periods: days 7–19 and days 20–33 . These time frames were chosen because virus was isolated in all cases until day 20 . Wilcoxon rank sum was used to determine the significance of differences in control biomarker levels versus case biomarker levels . The False Discovery Rate technique was used to adjust for multiple comparisons with an overall alpha of 0 . 05 [25] .
Case 1 was a middle aged male butcher who presented to a local hospital in Kabale District with a chief complaint of epistaxis . He had been well until a week prior to admission when he developed a high-grade fever associated with headache , neck pain , joint pain , and worsening fatigue . His admission labs were notable for an elevated white blood cell count of 16 , 100/μl , a decreased hemoglobin of 7 . 9 g/dl , and decreased platelets of 29 , 000/μl . Liver function tests revealed an elevated total bilirubin of 16 . 8 mg/dl and elevated transaminases . Creatinine and blood urea nitrogen were increased at 12 . 5 mg/dl and 240 . 5 mg/dl , respectively . Testing for malaria was negative . He was transferred to another hospital for transfusion and further management and recovered . Case 2 is a teenage student who presented to a local hospital in Kabale with a five day history of fever , myalgia , joint pain , headache and diarrhea as well as three days of epistaxis . He had cared for goats and sheep at home . On admission he was noted to have epistaxis and scleral icterus . Laboratory evaluation revealed an elevated unconjugated hyperbilirubinemia of 9 . 72 mg/dl , elevated AST of 470 U/L , decreased total protein of 3 . 43 g/dl , elevated amylase of 386 U/L , decreased hemoglobin of 7 . 3 g/dl , and decreased platelets of 45 , 000/μL . Renal function was normal . He was treated empirically with broad spectrum antibiotics and required blood transfusions . His hospital course was complicated by seizures and altered mental status , but gradually improved . Case 3 is a young adult builder who presented to a local hospital with fever , vomiting , abdominal pain , arthralgias , myalgias , and photosensitivity . He also had a history of epistaxis and hematemesis . Laboratory data were not available for this case . The relative viral loads ( log TCID50/mL ) in each of the three cases decreased over time ( Fig 1 ) . All patients still had detectable viral RNA after disease resolution , which likely represents the sensitivity of the assay rather than persistence of infectious virus since viral isolation attempts before day 20 post fever onset were successful and isolation attempts at time points 20 and more days post fever onset were unsuccessful . RVFV IgM was detectable on first date of blood sampling for Cases 1 and 2 , at day 13 for Case 1 , day 5 for Case 2 , but not until the second blood sample for Case 3 , on day 11 ( Fig 2 ) . RVFV IgG was undetectable in all three cases prior to 10 days post fever onset but was detectable at day 17 for Case 1 , day 11 for Case 2 , and day 20 for Case 3 ( Fig 2 ) . Correlation of biomarkers with viral load was assessed . Only those that were significant after adjusting for multiple comparisons are presented in Table 1 . sICAM-1 , CRP , IP-10 , eotaxin , tPA , D-dimer , granzyme B , and MCP-2 were all significantly positively correlated with viral load . P-selectin , ADAMTS13 , CFH , and Fibrinogen were all significantly negatively correlated with viral load . Of note , three of these factors ( CRP , P-selectin and CFH ) did not vary significantly outside the range of normal . Hemostatic derangement has been observed in human RVF cases , and was noted in all three of these patients . Although some markers of coagulopathy have been studied in rhesus monkeys [26] , there is limited data regarding human markers of coagulopathy during RVFV infection . For this reason , case biomarkers of coagulopathy were compared to control levels during two disease periods: days 7–19 ( virus isolation positive ) and days 20–33 ( virus isolation negative ) . Only data in which patient values were significantly outside the normal range are shown in the main figures . Any data obtained but not included in the main figures are presented in the Supplementary material for completeness ( S1 Fig ) . Levels of ADAMTS13 , D-dimer , PF4 , and vWF were significantly elevated in cases compared to controls during both disease periods ( Fig 3 ) . Thrombomodulin was significantly lower in cases compared to controls in both disease periods , but this data appears to be skewed by Case 2 with much lower levels than the others . Fibrinogen levels in cases were significantly lower than control levels early in the disease course , however mean fibrinogen levels increased in cases and were no longer significantly different from control levels after day 20 . In contrast , tPA levels in cases were significantly elevated early during the course of illness and decreased such that mean case levels were within the normal range later in the disease course . Endothelial expression of cell surface molecules can modify vascular stability , coagulation , and leukocyte recruitment [27] . Upregulation of these molecules can be associated with infection and inflammation . Given that microscopic changes in the endothelium have been noted in monkeys infected with RVFV [28] , case endothelial marker levels were compared to control endothelial marker levels in the two disease periods . Only data in which patient values were significantly outside the normal range are shown in the main figures . Any data obtained but not included in the main figures is presented in Supplementary material for completeness . Variations in endothelial marker levels were seen during both phases of illness among cases ( Fig 4 ) . L-selectin and VEGF levels were significantly higher in cases than controls throughout the illness . However , PECAM-1 was only significantly elevated early in the disease in cases compared to controls . In contrast , case E-selectin levels were normal early in the illness course but increased over time and were significantly elevated in cases late in the course of illness . Several endothelial markers were decreased among cases compared to controls . Mean VCAM levels were significantly lower in cases compared to controls throughout the course of illness . Soluble ICAM-1 levels were within the normal range early in the course of illness but decreased significantly late in the course of illness among cases compared to controls . Chemokines and cytokines play a vital role in recruitment , activation and maturation of the immune system; for this reason , we compared case levels of a series of cytokines and chemokines to control levels . Only data in which patient values were significantly different from controls are shown in the main figures . Any data obtained but not included in the main figures are presented in the Supplementary material for completeness . Case levels of IL-8 , MCP-2 , MCP-3 , fractalkine , GRO-α , and IP-10 were significantly elevated when compared to control levels throughout the course of illness ( Fig 5 ) . IL-10 and MIP-1β levels in cases were significantly elevated compared to control levels early during the illness course . However MIP-1β levels were within the normal range and did not exhibit marked variation during the disease course . Certain chemokines and cytokines were significantly lower in cases compared to controls . Case IL-1β and eotaxin levels were lower throughout the disease course . IL-12 levels in cases were significantly lower early in the course of illness only . RVFV infection in vitro and in animal models has been associated with decreased expression of type 1 interferons , secondary to the interferon antagonism activity of the RVFV NSs protein [8] . Consistent with this , IFN-α was lower in cases compared to controls . IFN- γ and IL-29 , a type III interferon , were noted to be higher in cases compared to controls but did not fluctuate significantly during the clinical course . Our assays also examined several different types of growth factors . Mean HGF levels in cases were significantly elevated early during the course of illness only . Mean case levels of FGF-basic were significantly elevated compared to mean control levels throughout the illness course; however , they were still within the range of control values . Granzyme B , a protease found in cytotoxic lymphocytes and natural killer cells that mediates apoptosis , and three other mediators of apoptosis , sFAS-ligand , sTNF-RI and sTNF-RII were found to be significantly elevated in cases compared to controls ( Fig 6 ) . We also measured non-specific markers of inflammation including C reactive protein ( CRP ) and ferritin ( Supplementary material and Fig 6 ) . Although case CRP levels were within the normal range and were not significantly elevated compared to mean control CRP levels , they did tend to be elevated early in the disease and decreased over time . Ferritin levels were significantly elevated in cases throughout the illness . sCD40L , which is mostly derived from activated platelets , was elevated in all patients throughout the course of infection .
In this study , we evaluated a series of biomarkers in three patients to examine several aspects of human pathophysiology during primary RVFV infection . All three cases had severe disease with hemorrhagic manifestations , required hospitalization , and all recovered from their illness . Although our study only included three cases , our findings illuminate the interaction of RVFV with the immune system , endothelium , and coagulation pathways . The pro-inflammatory cytokines/chemokines IL-8 , IL-10 , IP-10 , MCP-2 , MCP-3 , fractalkine , and GRO-α were all significantly elevated in cases compared to controls . MCP-2 and IP-10 were elevated in all three cases and correlated with viral load , suggesting that as viremia resolved , the inflammatory response normalized . IL-8 , MCP-1 , MCP-2 , TNF-R1 , IP-10 , and sFAS-ligand have been previously studied in Ebola , dengue , and Lassa patients and have shown some correlation with disease severity [11 , 13 , 17 , 29] . A previous study found IL-8 , IL-10 , IP-10 , CXCL9 and MCP-1 to be increased significantly in fatal RVF cases [18] however other cytokines/chemokines have not been studied extensively in RVF . Although we were not able to correlate these biomarkers with disease severity given the small numbers of patients , we observed elevated levels of these same pro-inflammatory molecules during the acute illness and trends of these biomarkers show normalization of levels as patients recovered . IL-1β , IL-12 , and eotaxin , were lower in cases than in controls . Of these only eotaxin was correlated with viral load . Eotaxin is a pro-inflammatory cytokine that mediates chemotaxis of eosinophils , basophils , mast cells , and Th2 lymphocytes during infection and has been noted to be lower in some patients with dengue infection when compared to controls , particularly patients with significant vascular leakage [30] . Further research using larger numbers of patients should be conducted to understand if RVF disease severity or clinical manifestations correlate with expression of these cytokines . Data from the literature suggests that RVFV infection is associated with decreased expression of interferons , secondary to the activity of the RVFV NSs protein [8] . IFN-α and IFN-β are an important part of the innate immune response to viral infection . Our study found that IFN-α was low throughout the course of illness and IFN-β was undetectable; it is tempting to speculate that this might be secondary to viral suppression via the NSs protein . Interestingly , IL-29 , a type III interferon , was elevated in our cases . Type III interferons primarily target mucosal epithelial cells and the liver; polymorphisms in certain Type III interferons have been associated with differential HCV clearance [31] . Our cases demonstrated elevated IL-29 and IFN-γ throughout the course of illness and no significant changes in levels were observed over time , therefore elevated IL-29 and IFN-γ levels observed in our study may be due to persistent immune activation or could simply be due to underlying differences in the Ugandan versus North American populations , rather than related to disease course . We found that levels of endothelial markers including P-selectin and sICAM-1 correlated with viral load . sICAM-1 levels decreased through the clinical course suggesting that endothelial activation and damage improved as viral infection resolved . In contrast , P-selectin levels increased as viral load decreased . The increase could contribute to resolution of coagulopathy in these RVF survivors , since P-selectin is involved in anchoring vWF to the endothelial cell surface thereby aiding in cleavage of vWF , which plays a role in clot formation [32] . Endothelial markers have not been studied previously in RVFV infections , but have been studied in Ebola [11 , 13] and dengue [15 , 33 , 34] infections . Previous studies have demonstrated that elevated sICAM levels in dengue [15] and Ebola cases [11 , 13] were associated with more severe disease . We were unable to correlate levels of endothelial markers with clinical severity; however our findings suggest that endothelial dysfunction may play a role in RVF pathogenesis . Given the fact that all three patients had hemorrhagic symptoms , it is not surprising that cases had higher levels of biomarkers of coagulopathy than controls . D-dimer and tPA were all elevated in cases compared to controls on presentation and were significantly correlated with viral load . Our observation of elevated d-dimer in cases is similar to findings in Ebola patients and dengue patients[12 , 35] and is indicative of ongoing fibrinolysis . Elevated levels of tPA are also consistent with the increased fibrinolysis . Laboratory evidence of decreased clotting and increased fibrinolysis on presentation is consistent with the bleeding that was initially observed in all three patients . tPA and D-dimer levels decreased over time and correlated with viral load decline , suggesting that clinical improvement and viremia resolution may correlate with decreased fibrinolysis . Elevated levels of vWF are less informative since the assay does not distinguish between the pro-coagulant multimers of vWF and the cleaved forms . It is interesting that levels of ADAMTS13 , which is a protease that cleaves vWF , increased with disease resolution in all three patients and then declined to normal in two of the patients . This could be an indicator of correction of the coagulopathy that was noted during acute disease . In support of this hypothesis , in a study of pediatric patients with severe dengue , ADAMTS13 activity was noted to be lower than normal , presumably secondary to consumption [14] . Deficiencies in ADAMTS13 levels or function can result in clinical manifestations of bleeding paradoxically via formation of microthrombii . As mentioned previously , P-selectin also negatively correlated with viral load ( i . e . levels increased as disease resolved ) . P-selectin is produced by activated platelets and increases in this protein could , similar to increases in ADAMTS13 , indicate a resolution of the coagulopathy that was noted clinically in the patients . This analysis suggests that hemorrhagic symptoms in RVF are due to complex interactions between the virus and both coagulation and fibrinolytic pathways . Granzyme B , a marker of cytotoxic T cell function , was also elevated in all patients and correlated with viral load decline . This finding is interesting because in patients with Ebola virus disease , both disease severity and viremia were correlated with granzyme B levels [11] . This could simply be a reflection of higher viral loads providing more antigen stimulus to T cells and thus more cytotoxic T cell activity and may be an indirect marker of disease severity . However , there are no data on the role of T cells during human RVFV infection , indicating a gap in our knowledge of RVFV pathophysiology that clearly warrants additional study . There are several limitations to our study . We had limited samples from only three patients , and we did not have complete clinical and laboratory information from all of the patients . Additionally , the healthy controls were from the United States and may have different baseline laboratory values of the biomarkers tested due to genetic or environmental differences than a healthy Ugandan . Given the limited number of patients and the fact that all three were ill enough to require hospitalization , we were unable to draw conclusions between biomarkers and severity of disease . Although our study had few patients , we were able to demonstrate similar trends in these cases in order to gain insight into the pathogenesis of symptomatic RVFV infection . We found that RVFV infection appears to have a complex and at times contradictory impact on endothelial markers , immune response , and coagulation . Further studies should be done to better characterize these biological pathways during RVFV infection . These types of studies have promise in determining if these biomarkers are associated with clinical outcome in patients with RVFV infection and whether they may serve as prognostic indicators of disease or possibly as clues that could guide the use of host-directed therapeutics . | Rift Valley fever virus is an emerging virus of public health significance . In endemic areas up to 60% of the population are seropositive by adulthood . This implies that most of the individuals living in these endemic areas become infected with this virus at some point in their lifetime . This highly prevalent virus is poorly understood likely secondary to lack of diagnostic capacity in affected areas . In fact , the majority of cases are undocumented and spontaneously resolve . However , in some individuals the disease can be very severe and at times associated with hemorrhage or encephalitis . In this report we identify three men infected during the acute stage who also exhibited hemorrhagic manifestations . These men had elevated cytokine and chemokine levels consistent with immune activation . Our data show that biomarkers of fibrinolysis correlated positively with virus while biomarkers of coagulation correlated negatively with virus , suggesting that opposing hemostatic pathways are altered during disease , and that they correct with resolution of viremia . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"rift",
"valley",
"fever",
"virus",
"immune",
"physiology",
"cytokines",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"pathogens",
"immunology",
"micro... | 2018 | Rift valley fever viral load correlates with the human inflammatory response and coagulation pathway abnormalities in humans with hemorrhagic manifestations |
Mutations in the LIM-homeodomain transcription factor LMX1B cause nail-patella syndrome , an autosomal dominant pleiotrophic human disorder in which nail , patella and elbow dysplasia is associated with other skeletal abnormalities and variably nephropathy and glaucoma . It is thought to be a haploinsufficient disorder . Studies in the mouse have shown that during development Lmx1b controls limb dorsal-ventral patterning and is also required for kidney and eye development , midbrain-hindbrain boundary establishment and the specification of specific neuronal subtypes . Mice completely deficient for Lmx1b die at birth . In contrast to the situation in humans , heterozygous null mice do not have a mutant phenotype . Here we report a novel mouse mutant Icst , an N-ethyl-N-nitrosourea-induced missense substitution , V265D , in the homeodomain of LMX1B that abolishes DNA binding and thereby the ability to transactivate other genes . Although the homozygous phenotypic consequences of Icst and the null allele of Lmx1b are the same , heterozygous Icst elicits a phenotype whilst the null allele does not . Heterozygous Icst causes glaucomatous eye defects and is semi-lethal , probably due to kidney failure . We show that the null phenotype is rescued more effectively by an Lmx1b transgene than is Icst . Co-immunoprecipitation experiments show that both wild-type and Icst LMX1B are found in complexes with LIM domain binding protein 1 ( LDB1 ) , resulting in lower levels of functional LMX1B in Icst heterozygotes than null heterozygotes . We conclude that Icst is a dominant-negative allele of Lmx1b . These findings indicate a reassessment of whether nail-patella syndrome is always haploinsufficient . Furthermore , Icst is a rare example of a model of human glaucoma caused by mutation of the same gene in humans and mice .
Nail-patella syndrome ( NPS ) ( OMIM 161200 ) is an autosomal dominant human disease , cardinal features of which are nail dysplasia , absent or hypoplastic patellae and abnormal elbows along with iliac horns . In addition , about 30–40% of patients develop nephropathy , which can progress to renal disease [1] . Open angle glaucoma is another feature of the disease that occurs in about 30–40% of patients [2] . Mutations of the transcription factor LMX1B have been found to be the underlying cause of NPS [3]–[6] . LMX1B is a member of the LIM-homeodomain ( LIM-HD ) family of transcription factors . The protein has two N-terminal LIM domains that are involved in protein-protein interactions followed by a homeodomain that binds to target DNA binding sites . Disease-causing mutations range from complete gene deletion to various frameshift , nonsense , splice and missense mutations . The majority of missense mutations are found in the homeodomain and the LIM domains . There is great variation in the severity and range of phenotypes both within families that carry the same mutation and between families that carry different mutations in LMX1B . Several missense NPS homeodomain mutations tested in vitro have shown no dominant-negative effect on the transcriptional activity of wild-type protein [7] , [8] and consequently NPS is thought to be a haploinsufficient disorder . Nevertheless , in a comprehensive study of 106 NPS patients from 32 families , patients with mutations in the homeodomain had more severe proteinurea than those with mutations of the LIM domains , although other aspects of NPS showed no phenotype-genotype correlation [9] . Association of haplotype with severity of nail dysplasia has also been reported [10] . Knockout studies in mice have shown that Lmx1b is required during development for dorsal patterning of the limb , the establishment of the midbrain-hindbrain boundary , the development of the cerebellum , for kidney development and for the specification of certain neuronal subtypes ( reviewed in [11] ) . Mice that lack Lmx1b have ventralised limbs , kidney abnormalities , calvarial bone defects and an absent cerebellum [12]–[14] . There are also anterior segment eye defects [15] . Postnatal conditional deletion experiments have shown that Lmx1b is required for formation of the trabecular meshwork , the maintenance of corneal integrity and transparency and loss results in corneal neovascularisation [16] . Heterozygous knockout mice are apparently normal indicating that in the mouse haploinsufficiency for Lmx1b does not lead to mutant phenotypes [12] , [15] . However , heterozygous knockout mice recover less well from unilateral nephrectomy than wild-type mice , suggesting some role in adult kidney regeneration and maintenance [17] . Here we report the identification of an Lmx1b missense mutation , Icst , which has a dominant-negative mode of action . In contrast to heterozygous knockout mice , heterozygous Lmx1bIcst mice have buphthalmic ( enlarged or bulging ) eyes and develop a glaucoma phenotype . In addition there is some post-natal lethality associated with kidney defects . We demonstrate that the difference in phenotypic consequence of the null and Icst alleles is due to LMX1BIcst protein acting in a dominant-negative manner . These findings have implications for the interpretation of the mode of action of mutant LMX1B in NPS .
We identified the N-ethyl-N-nitrosourea-induced mouse mutation , iris-corneal strands ( Icst ) , in a screen for dominant eye mutations . Mice that carry Icst have buphthalmic eyes suggestive of high intra-ocular pressure [18] . We had previously mapped Icst to proximal Chr 2 between the markers D2Mit365 and D2Mit372 [18] . Within this interval we considered Lmx1b to be a strong candidate for harbouring the Icst mutation because glaucoma occurs in about 30–40% of NPS patients [2] . We amplified and sequenced the exons and flanking regions of Lmx1b from Icst mutant mice and found a single nucleotide change , a T to A transversion in exon 5 in the Icst allele ( position 2:33 , 566 , 910 in Ensembl Release 74 mouse assembly GRCm38 ( http://www . ensembl . org ) ) which was absent from the reference mouse sequence and an additional 17 mouse strains [19] ( Figure 1A ) . This mutation causes substitution of hydrophobic valine with hydrophilic aspartic acid in the homeodomain ( V265D ) . The affected valine is in the recognition helix , and is very highly conserved across species and paralogues . In the crystal structure of the related paired-type homeodomain , the equivalent valine directly contacts the DNA recognition sequence by making hydrophobic contacts with the second thymine in the TAAT core of the recognition sequence [20] . The nature of the mutation in Lmx1b , coupled with the complementation tests described below , indicate that Icst is the causative mutant allele; Lmx1bIcst . Examination of the sequence traces of RT-PCR products spanning exon 5 of Lmx1b from embryonic samples indicated that equal amounts of mutant and wild-type transcript are produced in heterozygotes ( data not shown ) . We asked whether the mutant LMX1BIcst protein was able to bind to its recognition sequence . We produced recombinant His-tagged full-length LMX1B protein ( S form , see below ) and the homeodomain alone , in both wild-type and Icst mutant versions , and used these in bandshift experiments to determine if they could bind to a known LMX1B recognition sequence from intron 1 of the Col4a4 gene [21] . Both wild-type full-length and homeodomain proteins bound but the LMX1BIcst proteins did not ( Figure 1B ) . We then investigated the ability of wild-type and mutant LMX1B protein to transactivate transcription in a reporter assay . We tested two isoforms of the mouse LMX1B protein , a 372 amino acid protein ( S ) and a longer form ( L ) that includes an additional 29 N-terminal amino acids initiating from an upstream AUG ( see Materials and Methods ) [22] . Much of this additional sequence is conserved between humans and mouse but also includes a direct 18 bp repeat encoding a further six amino acids . We introduced the Icst mutation into the two isoforms and tested the ability of the wild-type and mutant versions to transactivate a luciferase reporter , previously shown to be activated by LMX1B , which has six copies of the LMX1B recognition sequence upstream of a minimal promoter [21] . Both L and S wild-type proteins transactivated the luciferase reporter construct on co-transfection , but neither of the LMX1BIcst proteins showed any transactivation activity ( Figure 1C ) . These results showed that the Icst mutation disrupts the ability of the LMX1B protein to recognise and bind to its target sequences and activate transcription . Mice that are heterozygous for the Lmx1b knockout allele do not have an eye phenotype ( Figure 2A ) [15] . In contrast , we observed that the eyes of Lmx1bIcst/+ mice had mild corneal opacity at a young age and by six to eight weeks Lmx1bIcst/+ eyes are buphthalmic and display a variety of abnormalities , as previously described [18] . Examples are shown in Figure 2B and 2C . Strands of tissue extend between the iris and cornea in some mice ( Figure 2B ) and corneal neovascularisation is seen , often associated with corneal ulcers ( Figure 2C ) . Other abnormalities in the cornea included scarring and inflammation , with wrinkling of Descemet's membrane and flattening of the endothelial cells with migration onto the anterior iris surface ( Figure 2E ) . It is likely that endothelial abnormalities contributed to the development of corneal oedema ( Figure 2F ) and ocular surface compromise which in some cases led to ulceration . Many Lmx1bIcst/+ eyes develop severe corneal ulcers with age . The bulging and distended appearance of the eyes suggested that there might be dysfunction of the drainage system at the iridocorneal angle resulting in raised intraocular pressure . We examined the angle by histology and found abnormalities that varied between eyes . In some Lmx1bIcst/+ eyes there was an open angle ( Figure 2H ) but in other cases the angle was closed with the iris adhering to the cornea ( Figure 2I and 2J ) . Although the angle is still open in some eyes , it is often narrow and sometimes Schlemm's canal appears abnormally short ( Figure 2L ) . The trabecular meshwork , a structure dependent on Lmx1b for its development [16] , typically appears abnormal , often being compressed and hypomorphic ( Figure 2L ) . In line with these histological observations , we found that intraocular pressure ( IOP ) was elevated in Lmx1bIcst heterozygotes ( Figure 3 ) . Between the ages of two and six months IOP was elevated in Lmx1bIcst heterozygotes compared to wild-types ( Figure 3A ) . The mean IOP measurements in older mice at 6–12 months are not statistically different between the two groups , in part due to an increased incidence of low IOP values ( <10 mmHg ) in older Lmx1bIcst heterozygotes , which is caused by significant corneal damage ( ulceration , sometimes with inflammation and perforation ) . Overall , about 35% of Lmx1bIcst heterozygotes were found to have high IOP as compared to about 5% of wild-types ( Figure 3B ) . Higher IOP is most commonly associated with poorly understood abnormalities of the iridocorneal angle [23] and it is a common and important risk factor for developing glaucoma in humans and can lead to the neurodegenerative hallmarks of the disease which are retinal ganglion cell loss and optic disc cupping . We observed optic nerve cupping in some , but not all , Lmx1bIcst heterozygous mice at various ages ( Figure 4A ) . There was also a profound reduction in the number of retinal ganglion cells in Lmx1bIcst heterozygous mice that was not seen in Lmx1bKO heterozygous mice ( Figure 4B ) . In addition , in Lmx1bIcst heterozygous mice optic nerve damage occurs and the loss of axons appear to be progressive ( Figure 4C ) . Homozygous knockout mice that lack LMX1B die on the day of birth [12] . They have ventralised limbs , absent cerebellum and kidney and eye defects [12] , [14] , [15] . We intercrossed heterozygous Lmx1bIcst mice and collected offspring for genotyping at E17 . 5 and at weaning . At E17 . 5 the three expected genotypes were present at Mendelian ratios indicating that Lmx1bIcst does not cause lethality in utero ( Table 1 ) . However , no Lmx1bIcst homozygous mice were present at weaning and four pups that were dead at birth were found to be homozygotes ( Table 1 ) . We examined homozygous mutant embryos and observed a phenotype very similar to that described for the knockout . Lmx1bIcst homozygotes lack a cerebellum , have abnormal kidneys and ventralised limbs ( Figure S1 and Figure 5D ) . We expected this cross to produce twice as many Lmx1bIcst heterozygotes as wild-type . However , at weaning we found equal numbers of these genotypes , indicating a deficit of heterozygotes and suggesting that the Icst allele is semi-lethal ( Table 1 ) . We also crossed heterozygous Lmx1bIcst mice with mice heterozygous for the knockout allele Lmx1btm1Rjo ( hereafter , Lmx1bKO ) and genotyped their offspring at weaning ( Table 2 ) . No double mutant mice were seen among 104 offspring , indicating that the two alleles do not complement each other and are thus allelic . There was a deficiency of about a third of Lmx1bIcst heterozygotes compared to Lmx1bKO heterozygotes , but this fell short of statistical significance ( χ2 = 3 . 358 , d . f . = 1 , P = 0 . 067 ) . To investigate this apparent dominant lethality further , we backcrossed both alleles of Lmx1b to C57BL/6J ( Table 2 ) and genotyped the offspring . We found the expected 1∶1 ratio of heterozygotes to wild-type for the Lmx1bKO allele backcross whereas for the Lmx1bIcst allele a third fewer than expected heterozygotes were seen ( χ2 = 26 . 3 , d . f . = 1 , P<0 . 0001 ) . All of the mice carrying the Icst mutation had abnormal eyes showing that the mutant eye phenotype , as described in the previous section , is completely penetrant on the studied genetic backgrounds . The observed deficiency in the number of Lmx1bIcst heterozygotes indicates that the Lmx1bIcst allele is semi-lethal when heterozygous , in contrast the knockout allele which does not elicit a heterozygous phenotype . Therefore the mutant LMX1BIcst protein must exert a dominant effect during development and/or early postnatal life that always results in ocular abnormalities and can be lethal . When and why do 30% of Lmx1bIcst heterozygous mice die ? The deficiency of Lmx1bIcst/+ mice at weaning ( Table 2 ) was present at P1 ( χ2 = 3 . 922 , d . f . = 1 , P = 0 . 048 ) ( Table 3 ) suggesting the lethality is perinatal and any Lmx1bIcst/+ mice that reach P1 survive as well as their wild-type littermates . We considered a possible cause of death to be kidney failure because of the importance of LMX1B in podocyte and slit diaphragm development [21] , [24] , [25] . We examined by electron microscopy the glomerular morphology of wild-type , Lmx1bIcst heterozygous and homozygous mice and , for comparison , glomeruli from Lmx1bKO heterozygous and homozygous mice at E17 . 5 and E18 . 5 ( Figure 5 ) . We took care to examine only the most mature glomeruli so that our analysis would not be confounded by comparing immature and mature podocytes . Both wild-type and Lmx1bKO heterozygous morphology was normal and no abnormalities were found ( Figure 5A and 5B ) whereas in the homozygous glomeruli the podocytes were effaced on the glomerular basement membrane ( GBM ) as previously reported ( Figure 5C ) [24] , [25] . In Lmx1bIcst homozygotes , podocytes appeared undeveloped; the morphology resembling that seen in the knockout ( Figure 5D ) . We also found glomerular abnormalities in all Lmx1bIcst/+ embryos examined although the degree varied between individuals ( Figure 5 E–L ) . In glomeruli from two Lmx1bIcst/+ E17 . 5 embryos there was normal development in many areas with foot processes forming normally ( Figure 5E ) . However , we found various abnormalities in both , representative examples are shown in Figure 5 F–I . In Figure 5F and 5H the GBM is split . In Figure 5F the podocyte is positioned flush against the GBM and foot processes have failed to develop and the GBM itself is fragmented , suggesting it is not adhering properly . In some glomeruli podocytes were immature and cuboidal in shape ( Figure 5I ) . These two individuals might have been Icst heterozygotes which would have survived as the ultrastructural changes are not extensive . In another Lmx1bIcst heterozygote examined at E18 . 5 there was extremely abnormal morphology ( Figure 5 J–L ) . The GBM is split ( Figure 5J and 5K ) and only some rudimentary foot process formation was observed ( Figure 5J ) . Areas where podocytes were effaced onto a split GBM were also found ( Figure 5L ) . As we found no evidence of normal glomerular development in this individual it is likely to be an example of one of the Lmx1bIcst heterozygotes which would die . This variability in the extent of the mutant kidney phenotype found in Lmx1bIcst heterozygotes is consistent with the finding that only one third of the Lmx1bIcst heterozygotes fail to survive . We have shown above that Lmx1bIcst induces gain-of-function heterozygous phenotypes of the eye and the kidney that are not found in heterozygotes for the knockout allele . Both alleles are homozygous lethal at birth . To investigate if wild-type Lmx1b could rescue these phenotypes we made mice that were transgenic for a bacterial artificial chromosome , RP23-305A12 , which contains the Lmx1b gene centrally located in a 225 kb insert ( Tg ( RP23-305A12 ) , hereafter BAC ) . This transgene was introduced into both the Lmx1bIcst and Lmx1bKO lines and mice crossed to assess if it could rescue the mutant phenotypes . First we examined the effect of the transgene on the Icst heterozygous and homozygous phenotypes . When hemizygous , the transgene rescued the Lmx1bIcst heterozygous gross eye phenotype ( Figure S2 ) . However , the hemizygous transgene did not rescue the perinatal lethality affecting Lmx1bIcst homozygotes ( Table 4 ) . In contrast , the hemizygous transgene could rescue the perinatal lethality of the homozygous knockout allele , although the number of Lmx1bKO homozygous rescued pups was low and they were small in size with ventralised limbs ( see below ) ( Table 4 ) . Interestingly , variable rescue of the eye phenotype of transgenic Lmx1bKO homozygous mice was observed; although the majority of eyes were normal , some eyes were very abnormal , with damaged corneas and optic nerve heads ( Figure S3 ) . We next asked if increasing the dose of transgenic Lmx1b could better rescue the mutant phenotypes ( Table 5 ) . We found that two copies of the transgene could rescue the Lmx1bIcst heterozygous semi-lethality; transgenic Lmx1bIcst heterozygotes survived in the normal Mendelian ratio . Two BAC copies could also rescue the homozygous perinatal lethality , albeit inefficiently ( Table 5 ) . The Lmx1bIcst/Icst rescued mice were smaller than their littermates and had the Icst mutant eye phenotype ( Figure S4 ) . For the knockout allele , a greater number of homozygous Lmx1bKO mice survived when the transgene was homozygous than when it was hemizygous ( Table 5 ) and in most cases the eyes are grossly normal . Nevertheless , a few did have a mutant eye phenotype and in these cases expression of wild-type Lmx1b from the transgenic BAC was reduced ( Figure S4 ) . All the rescued mice , whether Lmx1bKO/KO or Lmx1bIcst/Icst , had skeletal and limb defects ( Figures 6 and 7 ) . Lmx1bIcst/Icst mice homozygous for the transgene had paws that were clearly ventralised; the dorsal surfaces were largely devoid of hair and had thickened skin pads superficially , much like the ventral surface , although the skin was pigmented ( Figure 6 ) . Lmx1bKO/KO mice rescued by a hemizygous BAC transgene had the same ventralisation phenotype which was not substantially improved when the transgene was homozygous ( Figure 6 ) . However , when we examined the skeletons of these homozygous transgenic rescue mice by X-ray computed microtomography ( µCT ) , we found that aspects of the skeletal phenotype had been rescued ( Figure 7 ) . The paw skeleton of Lmx1bIcst/Icst mice homozygous for the BAC transgene was ventralised , as was that of Lmx1bKO/KO mice hemizygous for the BAC transgene ( Figure 7 ) . When homozygous the BAC transgene largely rescued the paw skeleton to near wild-type in Lmx1bKO/KO mice , despite the ventralised surface ( Figure 7 ) . Other skeletal abnormalities were also differentially rescued . Amongst other skeletal defects in homozygous Lmx1bKO mice the patella is absent and the scapula is very small [12] . We find the same defects in homozygous Lmx1bIcst embryos ( Figure S1 and data not shown ) , and absence of patella is , of course , a cardinal feature of NPS in human patients [1] . Lmx1bIcst heterozygotes ( and Lmx1bKO heterozygotes ) do have patellae ( data not shown ) . When homozygous , the BAC transgene does not rescue the patellar and scapular defects in Lmx1bIcst/Icst mice ( Figure 8B and 8E ) but does in Lmx1bKO/KO mice ( Figure 8C and 8F ) , again demonstrating that mutant phenotypes can be rescued more readily from the null background than when LMX1BIcst protein is present . These BAC transgenic experiments demonstrate that a higher level of wild-type Lmx1b expression ( i . e . two BAC copies ) is required to elicit rescue of the Lmx1bIcst mutant phenotype than the Lmx1bKO mutant phenotype . This is consistent with an Lmx1bIcst pathogenic mechanism with the LMX1BIcst protein exerting a dominant-negative effect on co-expressed wild-type protein . How does the LMXIBIcst mutant protein exert this dominant-negative effect on the wild-type protein ? LMX1B does not homodimerise [26] , [27] . Along with other LIM-HD proteins , LMX1B binds to co-factors via the two LIM domains [28] . One such co-factor is LDB1 which can itself homodimerise thus enabling the formation of homomeric or heteromeric LIM-HD complexes [27] . This raises the possibility that complexes containing both wild-type and mutant LMX1B could be formed in Lmx1bIcst heterozygous mice thus decreasing the level of functional complexes below that found in Lmx1bKO heterozygous mice . To test if such complexes can be formed we transfected cells with Myc-tagged wild-type LMX1B and FLAG-tagged LMX1BIcst either alone or together and immunoprecipitated using anti-Myc antibody . As expected the Myc-tagged wild-type LMX1B was present in the immunoprecipitated fraction but the FLAG-tagged LMX1BIcst was not , confirming that LMX1B indeed does not homodimerise ( Figure 9A ) . When LDB1 was included in the transfections , LDB1 was co-immunoprecipitated with the wild-type protein showing that LMX1B binds to LDB1 as expected ( Figure 9B ) . When all three proteins were present , FLAG-tagged LMX1BIcst protein was co-immunoprecipitated with the wild-type LMX1B ( Figure 9B ) showing that complexes containing both wild-type and mutant LMX1B are formed where the interaction is mediated by LDB1 . Consistent with this , less LDB1 appears to be in the bound fraction when FLAG-tagged LMX1BIcst was included along with the Myc-tagged wild-type LMX1B , probably due to competition between the wild-type and mutant protein for binding to LDB1 ( Figure 9B ) .
The Icst mutation , V265D , in the homeodomain of LMX1B abolishes DNA binding and transcriptional transactivation ( Figure 1 ) . Whilst heterozygous null ( knockout allele ) Lmx1b mice are phenotypically normal [12] , [15] and present in Mendelian numbers ( Table 2 ) , a fraction of Lmx1bIcst heterozygous mice die with associated kidney GBM defects ( Table 2 and Figure 5 ) . No morphological abnormalities have been found in glomeruli of Lmx1bKO heterozygotes up to one year of age [24] . Those Lmx1bIcst heterozygotes that do survive have a highly penetrant eye phenotype which is not seen in Lmx1bKO heterozygotes ( Figures 2–4 ) . Depletion of Lmx1b expression in adult mice causes corneal opacity and neovascularisation indicating that a threshold level of Lmx1b expression in the cornea is necessary for the maintenance of corneal integrity [16] . These differences between the Icst and knockout heterozygous phenotypes indicate that the LMX1BIcst protein exerts a dominant-negative effect on the wild-type protein . Dominant-negative mutant activity is typically mediated via protein complexes , usually dimers or higher-order structures , in which participation of one non-functional subunit inactivates the complex [29]–[33] . As previously reported [26] , [27] and confirmed by our results ( Figure 9 ) LMX1B does not homodimerise . We have shown that both wild-type LMX1B and LMX1BIcst proteins are found in protein complexes mediated by LDB1 ( Figure 9 ) . Complexes containing both wild-type and Icst mutant protein are likely to be non-functional and in Icst heterozygotes the level of functional complexes containing only wild-type protein would be 25% , compared to 50% in the null heterozygotes . This provides an explanation for the difference in the heterozygous phenotype of the two alleles and leads to the prediction that missense mutation in the LIM domains abolishing protein-protein interactions would be equivalent to null alleles . LMX1B has been shown to interact with LDB1 by yeast two-hybrid experiments [34] and complexes containing both proteins have been detected in rat glomeruli protein lysates [35] . The two genes have overlapping expression patterns [35] , [36] . In support of the role of LDB1 in LMX1B function , in mice specific inactivation of Ldb1 in podocytes leads to gradual loss of foot processes and GBM defects are found which lead to renal failure [35] . However , other binding partners for LMX1B are known , for example TCF3 [37] , [38] raising the possibility that proteins other than LDB1 may be responsible for mediating the dominant-negative effect in some cell types . There is a broad spectrum of disease severity both within and between NPS families and no clear genotype–phenotype correlation between the nature of mutations and severity of disease although in all cases skeletal abnormalities are found [5] . It is widely believed to be a haploinsufficient disorder and indeed , patients with a complete deletion of LMX1B have been found [6] . Variability in disease manifestation in patients with the same mutation is often observed [39] indicating genetic background modification . Furthermore , mutation of LMX1B does not always result in NPS . Two groups have reported the detection of novel missense mutations affecting R246 in the homeodomain of LMX1B in patients with isolated renal disease . In one report a patient with nail-patella-like renal disease was found to have an R246Q mutation that has residual transcription activity [40] . In the other report patients with autosomal dominant focal and segmental glomerular sclerosis either with the same R246Q mutation or with a different mutation affecting the same amino acid , R246P , were described [41] . None of the patients had any of the other classic symptoms of NPS . The reason why in these patients malformations are confined to the kidney is not clear . It may be that the residual activity of the R246Q protein is sufficient for normal development outwith the kidney or that these mutations of R246 only compromise binding to a subset of target sequences . The lack of phenotype in Lmx1b null heterozygous mice [12] , [15] contrasts with heterozygous Lmx1bIcst mice which have a strongly penetrant eye phenotype and have glomerular abnormalities that resemble defects found in human NPS patients ( Figure 5 ) . They do not , however , have any skeletal abnormalities , which is the most prevalent aspect of NPS . Valine 265 in LMX1B , the equivalent residue to that mutated in the mouse Lmx1bIcst allele , has been found mutated in NPS patients to phenylalanine and to leucine [8] , [22] . V265L ( originally reported as V242L ) along with other patient missense mutations have been tested for dominant-negative effects on wild-type protein in in vitro transcription reporter assays but none have been found [7] , [8] . Likewise , we have been unable to find , in in vitro experiments , a dominant-negative effect of the LMX1BIcst protein on transcriptional reporter assays ( data not shown ) , although it is clearly demonstrated by the phenotype in mice . It is apparent that mutant proteins can show dominant-negative activity in mice that is not seen in vitro and it is possible that some of the human LMX1B mutations may be dominant-negative . Indeed , it has been reported that patients with homeodomain mutations exhibit more severe proteinuria , and hence kidney defects , than patients with mutations in the LIM domains suggesting dominant-negative activity [9] . Glaucoma is usually associated with high IOP caused by dysfunction of the ocular drainage structures in the iridocorneal angle of the eye [23] . LMX1B mutations are well established to cause open-angle glaucoma in NPS patients , but due to the influence of modifier genes may also cause glaucoma without NPS . Although confirmation is required , LMX1B haplotype has been suggested to influence open-angle glaucoma in the general population ( without other aspects of NPS ) [42] . Similar to some of the Lmx1bIcst heterozygous mice , a narrow but open-angle phenotype with high IOP is present in some individuals with an LMX1B mutation [4] . The aetiology of IOP elevation in glaucoma remains poorly understood and it is likely to be mechanistically heterogeneous at the molecular level . Various studies have reported open-angle glaucoma phenotypes in mice [43]–[47] . However , in most cases and due to undefined multifactorial influences , these phenotypes have typically been mild or have not yet proven reproducible between laboratories . Lmx1bIcst/+ mice have variably open or closed angles , and may be a model of anterior segment dysgenesis leading to high IOP , rather than primary open angle glaucoma . Nevertheless they should provide a valuable mouse model of glaucoma caused by dominant point mutation in a gene that also causes a form of human open-angle glaucoma . The phenotype is highly penetrant and reproducible in our colonies at different institutions . Thus , the Icst mutation provides a new tool for dissecting the molecular and pathologic features of IOP elevation and glaucoma , and for testing new therapies .
The animal studies described in this paper at the MRC Human Genetics Unit were carried out under the guidance issued by the Medical Research Council in Responsibility in the Use of Animals for Medical Research ( July 1993 ) and Home Office Project Licence nos . PPL60/3124 , PPL60/3785 and PPL60/4424 . All experiments conducted at The Jackson Laboratory were approved by the institutional Animal Care and Use Committee . All animals were treated in accordance with the protocols established by the Association for Research in Vision and Ophthalmology . The Jackson Laboratory's pathogen surveillance program regularly screened for pathogens . Mice were housed in a 14 hour light to 10 hour dark cycle . Mutant and littermate control mice were housed together to control for cage-dependent differences . The Lmx1bIcst strain has been submitted to the European Mouse Mutant Archive ( http://www . infrafrontier . eu ) strain number EM:00114 . Both the Lmx1bIcst and knockout lines were maintained on the C57BL/6J background . Clinical examinations were carried out as previously described [18] , [48] . The BAC transgenic strain was made as described [49] . The exons and the immediate flanking sequences of Lmx1b were amplified from Icst , BALB/c , C3H and C57BL/6J genomic DNA using intronic primers that were also used for subsequent sequence analysis . PCR products were purified using Millipore Multi-screen PCR 96-well filtration system on a Biomek 2000 robotic platform and sequenced directly using Big Dye terminator cycle sequencing . Sequences were analysed using the Sequencher program . To produce N-terminally histidine-tagged fusion proteins of full-length 372 amino acid LMX1B and the homeodomain alone , wild-type and Icst cDNAs were amplified by PCR introducing Pci I and Nco I sites at the 5′ of the full-length and homeodomain respectively and a Bam HI site at the 3′ of both and cloned into pGEM-T Easy ( Promega ) . After digestion with Bam HI and Pci I or Nco I , as appropriate , the cDNAs were cloned into the Nco I and Bam HI sites of pET6H [50] . Recombinant histidine-tagged proteins were expressed in the Escherichia coli strain BL21 ( DE3 ) pLysS essentially as described [51] . Proteins were analysed by sodium dodecyl sulphate–polyacrylamide gel electrophoresis to assess yield and purity and equal amounts of wild-type and mutant proteins were used in bandshift experiments . For expression in mammalian cells wild-type Lmx1b and Lmx1bIcst cDNAs were cloned into the expression vector pcDNA3 . 1 ( Invitrogen ) . The original reported size of LMX1B protein is 372 amino acids [12] . An upstream in-frame ATG in the human sequence that would encode an additional 23 amino acids has been reported [22] . This sequence is conserved between mouse and human and the mouse LMX1B protein sequence in the database has been revised to include these extra 23 amino acids ( entry O88609 in http://www . uniprot . org ) . In addition , there is a direct duplication of 18 bp encoding the first 6 amino acids of the 372 amino acid protein in the mouse genome sequence that is not conserved in human . This 18 bp sequence has been found to be absent from some Lmx1b cDNAs ( e . g . BC125469 ) but it is found in the EST database ( BY741174 . 1 ) . By RT-PCR we found it to be present in Lmx1b transcripts ( data not shown ) . We therefore made two versions of LMX1B , one 372 amino acids long ( -S ) and one including the additional N-terminal 29 amino acids ( -L ) . Myc and FLAG-tagged mammalian expression vectors were made using the Gateway cloning system ( Life Technologies ) . In brief , Lmx1bWT and Lmx1bIcst were amplified from the -L constructs described above and cloned into the donor vector pDONR™221 ( Life Technologies ) and then into the destination vectors pcDNA3 . 1Myc-HisDEST [52] and pDEST/C-SF-TAP [53] to give WT-Myc and Icst-FLAG respectively following the manufacturer's instructions . Full-length Ldb1 was amplified from mouse embryonic cDNA using primers that introduced a Hind III site at the 5′ end and a Bam HI site at the 3′ end and cloned using these sites into pcDNA3 . 1 ( + ) to give pcDNA-LDB1 . All plasmids were verified by sequencing . The FLAT probe from the Col4a4 gene intron 1 [21] was made by annealing the two oligonucleotides 5′-GGTTCATGAAAGTAATTATTTTCA-3′ and 5′-GGTTTGAAAATAATTACTTTCATG-3′ and end-labelled by filling in the four base 5′ single-stranded extensions with 32P dATP and 32P dCTP using Klenow polymerase . Bandshift analysis was carried out essentially as described [54] . 20 , 000 cpm FLAT probe ( ∼1×107 cpm/µg radiolabelled DNA probe ) was used in each bandshift reaction . Bacterial extract protein concentrations were ∼5 µg/µl and we used 1 , 2 or 3 µl per reaction . Transfections were carried out using a MicroPorator MP-100 following the manufacturer's protocol ( Microporator ) . To correct for transfection efficiency and viability , 2 . 5 ng of renilla reporter vector was also transfected . The day following transfection luciferase assays were carried out using the Dual-luciferase reporter assay system ( Promega E1910 ) and readings were normalised using the renilla reporter . Each experiment was carried out in triplicate . For anterior segment examination and photography , a Nikon FS-3V zoom slit-lamp biomicroscope was used with an attached Nikon D300S digital still camera and digital images were saved using Adobe Photoshop CS5 ( Adobe , Inc . ) . Mouse paws were photographed using an imaging system comprising a Nikon AZ100 macroscope ( Nikon UK Ltd , Kingston-on-Thames , UK ) and a Qimaging Micropublisher 5 cooled colour camera ( Qimaging , Burnaby , BC ) . Image capture was performed using in-house scripts written for IVision ( BioVision Technologies , Exton , PA ) . Eye histology was carried out as previously described [18] , [48] . After dissection embryos were photographed and fixed overnight in 4% paraformaldehyde in PBS at 4°C . A small part of the tail was used for genotyping . After washing in PBS they were dehydrated by immersion in a series of increasing concentrations of alcohol , embedded in paraffin wax , sectioned and stained with haematoxylin and eosin . Slides were viewed on a Leica MZFLIII fluorescence stereo microscope fitted with a Coolsnap colour camera ( Roper Scientific , Tucson , Arizona , USA ) . Image capture was controlled by in-house scripting of IPLab Spectrum ( Scanalytics , Fairfax , VA , USA ) . For plastic-based processing , enucleated eyes were fixed ( 0 . 8% paraformaldehyde and 1 . 2% glutaraldehyde in 0 . 08 M phosphate buffer ( pH 7 . 4 ) ) and processed for plastic sectioning as previously described [48] . Serial sagittal sections passing through the optic nerve were collected , stained with hematoxylin and eosin , and analysed for pathologic alterations . For all mice , IOP was measured as previously described in detail [55] , [56] . Briefly , mice were acclimatised to the procedure room and anesthetized via an intraperitoneal injection of a mixture of ketamine ( 99 mg/kg; Ketalar , Parke-Davis , Paramus , NJ ) and xylazine ( 9 mg/kg; Rompun , Phoenix Pharmaceutical , St . Joseph , MO ) prior to IOP assessment . Eyes were collected into 2×PBS and fixed in 2% paraformaldehyde in PBS for two minutes . After rinsing in 2×PBS for five minutes the retina was dissected and laid flat by making radial incisions and fixed in methanol at −20°C for one hour . The retinas were then fixed again in 4% paraformaldehyde in PBS for five minutes , rinsed in 2×PBS and blocked in wholemount buffer ( 2×PBS , 1% Cohn fraction BSA , 3% Triton X-100 ) for one hour and then incubated with anti-BRN3 antibody ( SC6026 , Santa Cruz ) overnight . After three ten minutes washes in wholemount buffer the retinas were incubated in Alexa Fluor 594 donkey anti-goat secondary antibody ( 1/500 ) ( Molecular Probes ) in wholemount buffer for four hours . After three washes in wholemount buffer , retinas were rinsed with 2×PBS and post-fixed in 4% paraformaldehyde in PBS for five minutes , mounted in Vectashield hard set ( Vector Labs ) . All washes and incubations were carried out at room temperature on an orbital shaker . Four images were taken at 90° angles to each other around the optic disc using the Nikon TiE-C1Si confocal microscope . The number of BRN3-positive cells in each of the areas was counted using the Velocity Image acquisition and analysis software ( PerkinElmer , Waltham , MA , USA ) . Mice from two litters were analysed . In the first litter ( age four months ) two Lmx1bicst/+ mice were used and one each of the other two genotypes . In the second litter ( age one month ) one mouse of each genotype was used . Optic nerves were processed and analysed as previously reported [48] , [57] . Briefly , nerves were stained with paraphenylene-diamine which differentially stains single damaged axons allowing sensitive detection of axon injury . Nerves were determined to have one of three damage levels that are readily distinguishable by axon counting . E17 . 5 embryonic kidneys were fixed overnight in 3% glutaraldehyde in cacodylate buffer at 4°C then post-fixed in 1% osmium tetroxide for two hours at 4°C . After dehydration through ascending grades of alcohol and propylene oxide they were impregnated with TAAB Embedding Resin ( medium grade premix ) and cured for 24 hours . Ultrathin sections were stained with uranyl acetate and lead citrate and viewed on a JEOL JEM 100CXII fitted with an AMT Digital Camera using the AMTv600 image capture software . Live animals and whole mount embryos and pups were scanned using a Skyscan 1076 in vivo µCT system ( Skyscan B . V . , Aartselaar , Belgium ) . Live animals were scanned under fluothane anaesthesia . Animals were scanned at an isotropic resolution of 18 . 6 µm . Scans were performed at 50 kV , 200 µA using a 0 . 5° rotation step and a 0 . 5 mm aluminium filter . Higher resolution scans of limbs were performed using a Skyscan 1172 system at a resolution of 8 . 8 µm ( 60 kV , 167 mA , 0 . 6° rotation step , 0 . 5 mm aluminium filter ) . Scans were reconstructed using Skyscan NRecon software and analysed using Skyscan CTAn software . Three dimensional models were visualised using Skyscan CTVol software . HEK 293T cells cultured in 100 mm dishes were transfected with WT-Myc ( 5 µg ) , ICST-FLAG ( 2 µg ) , and pcDNA-LDB1 ( 5 µg ) as indicated in a total of 12 µg DNA adjusted with pcDNA3 . 1 ( - ) as necessary using Lipofectamine LTX PLUS ( Invitrogen ) . After 48 hours the cells were lysed using Cell Lysis Buffer ( Cell Signaling Technology ) and Myc-tagged complexes immunoprecipitated using the Profound c-Myc Tag IP/Co-IP Kit ( Thermo Scientific ) according to the manufacturer's instructions . Samples were separated on 4–12% NUPAGE gels ( Life Technologies ) , transferred to Hybond-P ( GE Healthcare ) and probed with anti-Myc ( Cell Signaling Technology , #2276 ) , anti-FLAG ( Cell Signaling Technology , #2368 ) and anti-LDB1 ( kind gift of Sam Pfaff [58] ) antibodies using standard protocols and visualised with horseradish-peroxidase secondary antibody ( GE Healthcare ) and SuperSignal West Pico Chemiluminescent Substrate ( Thermo Scientific ) following the manufacturer's instructions . Values were expressed as mean+standard error and P<0 . 05 was considered significant . For the data shown in Figure 1 a two-tailed unpaired Student's t-test was used for statistical analysis . For the data shown in Figure 3 a two-tailed unequal variance t-test was performed using JMP ( http://www . jmp . com ) . For the ganglion cell counts data shown in Figure 4 we used log transformation of raw data to allow for a mean-variance relationship . We then tested for a mean difference between replicate mice within genotype using ANOVA . As replicate mice within genotype differed significantly , we compared variation in mean cell count between pairs of genotypes against variation between means of replicate mice within genotype using a two-tailed Student's t-test . For the data shown in Tables 1–5 chi square tests were performed using http://graphpad . com/quickcalcs . | Nail-patella syndrome is a human genetic disease caused by an inactivating mutation in one copy of a gene called LMX1B , with the amount of protein produced from the remaining copy of the gene not being enough for normal function . Patients with this disease have malformations of their nails , elbows and kneecaps . Some patients also develop kidney disease and glaucoma . LMX1B controls where and when other genes are expressed and it is important during development . Studies in mice have shown that complete absence of Lmx1b is lethal at birth . In contrast to humans , mice with only one copy of the gene are normal . Here we describe a new mutant mouse , Icst , which has a mutation in Lmx1b that abolishes the ability of the protein to bind near genes that it controls . Mice with one normal and one copy of Lmx1b with the Icst mutation have eye defects and some die shortly after birth probably due to kidney failure . Therefore having one functional and one mutant copy of Lmx1b is more detrimental than having a half dose of functional protein . The Icst mouse is a model of human glaucoma where mutation of the same gene causes glaucoma in humans and mice . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"gene",
"function",
"mutagenesis",
"developmental",
"biology",
"mutation",
"model",
"organisms",
"animal",
"genetics",
"organism",
"development",
"cell",
"biology",
"genetic",
"screens",
"gene",
"identification",
"and",
"analysis",
"genetics",
"morphogenesis",
"biology",
... | 2014 | A Dominant-Negative Mutation of Mouse Lmx1b Causes Glaucoma and Is Semi-lethal via LBD1-Mediated Dimerisation |
Specific interactions between host genotypes and pathogen genotypes ( G×G interactions ) are commonly observed in invertebrate systems . Such specificity challenges our current understanding of invertebrate defenses against pathogens because it contrasts the limited discriminatory power of known invertebrate immune responses . Lack of a mechanistic explanation , however , has questioned the nature of host factors underlying G×G interactions . In this study , we aimed to determine whether G×G interactions observed between dengue viruses and their Aedes aegypti vectors in nature can be mapped to discrete loci in the mosquito genome and to document their genetic architecture . We developed an innovative genetic mapping strategy to survey G×G interactions using outbred mosquito families that were experimentally exposed to genetically distinct isolates of two dengue virus serotypes derived from human patients . Genetic loci associated with vector competence indices were detected in multiple regions of the mosquito genome . Importantly , correlation between genotype and phenotype was virus isolate-specific at several of these loci , indicating G×G interactions . The relatively high percentage of phenotypic variation explained by the markers associated with G×G interactions ( ranging from 7 . 8% to 16 . 5% ) is consistent with large-effect host genetic factors . Our data demonstrate that G×G interactions between dengue viruses and mosquito vectors can be assigned to physical regions of the mosquito genome , some of which have a large effect on the phenotype . This finding establishes the existence of tangible host genetic factors underlying specific interactions between invertebrates and their pathogens in a natural system . Fine mapping of the uncovered genetic loci will elucidate the molecular mechanisms of mosquito-virus specificity .
Most organisms engage in ecological interactions with organisms of different species that have profound effects on their fitness . These interactions , which can be antagonistic ( e . g . , parasitism , competition ) or mutualistic ( e . g . , cooperation ) , are major drivers of adaptive evolution and diversification . Understanding the evolution of traits mediating ecological interactions can be complicated by their genetic specificity , whereby fitness of a genotype depends on the genotype of the interacting species [1] , [2] . Such genotype-by-genotype ( G×G ) interactions , sometimes referred to as intergenomic epistasis , occur in both antagonistic [3] and mutualistic [4] relationships . Importantly , G×G interactions imply that the genetic basis of interaction traits is a composite entity that involves distinct genomes . Therefore , dissecting the genetic architecture ( i . e . , the number , position , effect and interactions between genetic loci underlying the phenotype ) of these traits requires accounting jointly for genetic variation in different species [5] . Among the most intriguing examples of G×G interactions are those involved in invertebrate host susceptibility to pathogens [6] . Indeed , specific interactions between host and pathogen genotypes have been documented in a wide variety of invertebrate systems [7]–[12] . This observation challenges the long-held view that invertebrate defense against pathogens relies on broad-spectrum recognition and effector mechanisms [13] , [14] . Lack of a mechanistic explanation , however , has questioned the nature of host factors underlying the observed G×G interactions [15] . For instance , the effect of host genotype can be confounded with that of symbiotic microbiota [16] , raising the possibility that G×G interactions may be environmentally driven . A critical question is whether G×G interactions observed at the phenotypic level truly result from the effect of discrete genetic factors within host and pathogen genomes . More generally , understanding the ecological and evolutionary dynamics of host-pathogen interactions requires a detailed knowledge of their genetic architecture [17] . In this study , we addressed this question in a natural insect-virus association that is relevant for human health . Aedes aegypti mosquitoes are the main vectors of dengue viruses , which cause the most prevalent mosquito-borne viral disease of humans [18] . Successful virus transmission requires that following mosquito blood feeding on a viremic host , infection is initially established in the insect's midgut cells and then disseminates throughout the rest of the body . The mosquito becomes infectious when the virus reaches the salivary glands and is released into the saliva . Vector competence defines the intrinsic ability of a mosquito to become infected following ingestion of infectious blood and to subsequently transmit the virus [19] . It varies substantially between and within Ae . aegypti populations throughout their wide geographical range [20] , [21] . The existence of genetic factors underlying the observed variation in mosquito susceptibility to dengue was initially demonstrated by artificial selection of resistant and susceptible inbred lines of Ae . aegypti [22] . This finding confirmed that , as for many other host-pathogen systems [17] , Ae . aegypti susceptibility to dengue has a genetic basis . Subsequent studies based on laboratory crosses of resistant and susceptible mosquito lines mapped several quantitative trait loci ( QTL ) controlling Ae . aegypti susceptibility to dengue virus infection and dissemination [23]–[25] . These QTL mapping studies , however , ignored the influence of viral genetic factors by exposing mosquitoes to a single , reference virus strain . A meta-analysis on the genetic architecture of host susceptibility in plants and animals revealed that QTL are recovered in only 25% of the cases when the mapping involves a different pathogen strain [17] . Dengue viruses exist in nature as four antigenically distinct serotypes ( DENV-1 through DENV-4 ) , which , in turn , consist of considerable genetic diversity [26] . Recently , we reported that several indices of Ae . aegypti vector competence for dengue viruses are governed by G×G interactions [9] , [27] . Thus , the efficiency of dengue virus transmission by Ae . aegypti depends on the specific pairing of mosquito and virus genotypes . Here , we surveyed genetic factors within the Ae . aegypti genome that are associated with G×G interactions influencing vector competence for dengue viruses . We developed an innovative genetic mapping strategy ( Fig . 1 ) based on wild-type Ae . aegypti families that were experimentally exposed to four different dengue virus isolates ( two DENV-1 isolates , designated as DV1-26A and DV1-30A , and two DENV-3 isolates , designated as DV3-10A and DV3-14A ) . The use of outbred families for genetic mapping was inspired from a validated study design previously developed to investigate the genetic basis of natural mosquito resistance to human malaria parasites [28] , [29] . To simulate a natural situation , we used naturally circulating virus isolates contemporaneous with the mosquitoes that were obtained from the serum of human patients . Their complete genome sequence confirmed that they were genetically distinct ( Fig . S1 ) . Genetic mapping was based on a set of microsatellite markers distributed across the Ae . aegypti genome , which consists of three chromosomes ( Fig . S2 ) . With one marker every 9 . 0 centiMorgans ( cM ) on average , marker density was entirely adequate for chromosomes 1 and 3 . For chromosome 2 , however , the paucity of valid and/or informative microsatellites resulted in poor coverage ( 1 marker every 23 . 4 cM ) . Therefore , we focus here on chromosomes 1 and 3 and provide mapping results for chromosome 2 as supplementary data . Our genetic mapping strategy allowed us to detect genetic linkage ( i . e . , non-independence between marker allele segregation and phenotype ) at two different levels for each marker . The first level measured the dependence of the phenotype on the mosquito genotype regardless of the virus isolate ( i . e . , the main host genotype effect across virus serotypes and isolates ) . The second level measured the dependence of the phenotype on the genotype conditional on the virus isolate ( i . e . , the interaction between virus isolate and mosquito genotype , a measure of G×G interactions ) . The methodology of our genetic survey ( Fig . 1 ) differs significantly from conventional genetic mapping strategies because it does not rely on controlled crosses between inbred lines that represent extremes of a trait . Although conventional strategies maximize QTL detection power , they are not best suited to identify multi-allelic QTL naturally segregating within unmanipulated populations [30] , [31] . The large number of progeny produced by a single parental pair of mosquitoes can be used as outbred families that are suitable for QTL mapping [28] , [29] . Vector competence was scored 14 days after an infectious blood meal according to three distinct phenotypes: ( i ) the proportion of mosquitoes that developed a midgut infection , ( ii ) the proportion of infected mosquitoes in which infection disseminated from the midgut to head tissues , and ( iii ) the infectious viral titer in virus-infected head tissues . Midgut infection and viral dissemination are prerequisites for virus transmission by mosquito bite [32] . Infectious titer of disseminated virus is used as a proxy for transmission potential [33] . All phenotypes were based on detection of infectious virus by standard plaque assay .
A total of 2 , 084 Ae . aegypti females from nine independent isofemale families ( mean sample size per family: 232; range: 104–403 ) were individually phenotyped and genotyped ( Table S1 ) . Five of the families yielded at least one QTL statistically significant at the genome-wide level for the midgut infection phenotype ( Fig . 2 ) . Significant linkage at the genome-wide level was detected on chromosome 1 at marker 71CGT1 ( 29 . 7 cM ) in family C01 ( genome-wide p-value = 9 . 44×10−4 ) and family 5 ( p = 2 . 9×10−2 ) , at marker 335CGA1 ( 38 . 2 cM ) in family C01 ( p = 5 . 55×10−4 ) , and at marker 88CA1 ( 44 . 9 cM ) in family 7 ( p = 4 . 94×10−3 ) and family 54 ( p = 4 . 0×10−2 ) . Linkage was also detected on chromosome 3 at marker 301ACG1 ( 0 . 0 cM ) in family 51 ( p = 7 . 47×10−5 ) and at marker B19 ( 13 . 6 cM ) in the same family ( p = 5 . 22×10−3 ) . The proportion of phenotypic variation explained by each significant marker ranged from 3 . 5% to 12 . 0% . Importantly , we also detected significant virus isolate-specific linkage on chromosome 3 at marker 301CT1 ( 0 . 0 cM ) in family 5 ( p = 1 . 95×10−2 , Fig . 2D ) . In this family , the proportion of infected females varied significantly among 301CT1 genotypes , but the genotype-phenotype relationship differed between virus isolates ( Fig . S3 ) . This isolate-specific genotype-phenotype association is interpreted as a G×G interaction between the mosquito and the viral genomes . An underlying assumption is that the isolate effect is primarily driven by genetic differences among isolates . When the isolate was replaced by the corresponding blood meal titer in the analysis , the interaction effect was no longer statistically significant ( p = 0 . 083 ) , which ruled out that uncontrolled variation in infectious dose among virus isolates ( Table S2 ) might have confounded our interpretation of the isolate effect . Significant linkage at the genome-wide level was detected in two of the nine families for the viral dissemination phenotype ( Fig . 3 ) . Linkage was significant on chromosome 1 at marker 335CGA1 ( 38 . 2 cM ) in family J07 ( p = 3 . 08×10−2 ) and family 42 ( p = 3 . 1×10−2 ) and on chromosome 3 at marker 69TGA1 ( 32 . 1 cM ) in family J07 ( p = 4 . 4×10−2 ) . The proportion of phenotypic variation explained by each significant marker ranged from 16 . 5% to 22 . 6% . Marker 335CGA1 on chromosome 1 was in linkage with the dissemination phenotype in two different families . In family J07 the marker effect was general across virus serotypes and isolates ( Fig . 3A ) , whereas in family 42 it was isolate-specific ( Fig . 3B ) . To verify that the isolate effect was not confounded with an effect of the infectious dose , we confirmed that the isolate by genotype interaction in family 42 was no longer statistically significant when the isolate was substituted by the blood meal titer ( p = 0 . 287 ) . For illustration , Fig . 4 shows the genotype-phenotype correlation for each virus isolate at marker 335CGA1 ( the allele segregation pattern at this marker is shown in Fig . S4 ) . Although marker genotype 439/439 confers protection against viral dissemination of isolates DV3-10A and DV3-14A , it does not have a detectable effect against isolates DV1-26A and DV1-30A . It is worth noting that because isolates DV3-10A and DV3-14A belong to DENV-3 whereas isolates DV1-26A and DV1-30A belong to DENV-1 , in this particular case the effect could be serotype-specific rather than isolate-specific . Significant linkage at the genome-wide level was detected in three of the nine families for the head titer phenotype ( Fig . 5 ) . Linkage was significant on chromosome 1 at marker 88CA1 ( 44 . 9 cM ) in family 51 ( p = 3 . 24×10−3 ) . Linkage was also detected on chromosome 3 at marker 17ATA1 ( 22 . 4 cM ) in family J07 ( p = 1 . 70×10−5 ) , at marker 69TGA1 ( 32 . 1 cM ) in family J07 ( p = 4 . 16×10−3 ) , at marker 201AAT1 ( 57 . 1 cM ) in family J06 ( p = 5 . 18×10−4 ) , and at marker 470CT2 ( 64 . 2 cM ) in family J07 ( p = 1 . 35×10−2 ) . The proportion of phenotypic variation explained by each significant marker ranged from 8 . 9% to 75 . 6% . The genotype-phenotype association was isolate-specific at marker 201AAT1 in family J06 and at marker 470CT2 in family J07 . Again , substituting the isolate by the corresponding blood meal titer ruled out a confounding effect of the infectious dose because the interaction was no longer statistically significant at marker 201AAT1 ( p = 0 . 434 ) or at marker 470CT2 ( p = 0 . 130 ) . For illustration , Fig . 6 shows the genotype-phenotype correlation for each virus isolate at marker 201AAT1 . Although marker genotype 338/338 confers protection against viral dissemination of isolates DV3-14A and DV1-26A , it results in increased head titer of isolate DV1-30A and no detectable effect against isolate DV3-10A . In this case the effect is truly isolate-specific ( not serotype-specific ) because isolates DV3-14A and DV1-26A ( DENV-3 and DENV-1 , respectively ) share the same pattern whereas isolates DV1-26A and DV1-30A ( both DENV-1 ) display opposite patterns . The isolate-specific genotype-phenotype correlation at marker 470CT2 is shown in Fig . S5 . Supporting information includes genetic mapping results for chromosome 2 ( Fig . S6 , S7 , S8 ) and for families that did not produce any significant linkage ( Fig . S9 , S10 , S11 ) .
Our genetic survey demonstrates that G×G interactions between dengue viruses and mosquito vectors can be assigned to physical regions of the mosquito chromosomes . To the best of our knowledge , this is the first study to successfully locate G×G interactions in an invertebrate genome by marker-based genetic mapping . In agreement with the conclusions of a previous meta-analysis [17] , we provide empirical evidence that the genetic architecture of host resistance depends on the pathogen strain . We establish the existence of tangible host genetic factors underlying G×G interactions in a natural invertebrate host-pathogen system . This is a critical first step towards their identification and characterization . This study also provides important new information on the biology of dengue virus transmission in a natural situation . Phenotypic variation in the ability of field Ae . aegypti populations to serve as vectors of dengue viruses was previously observed [20] , [21] . Genetic selection experiments [22] followed by QTL mapping studies using inbred selected lines [23]–[25] demonstrated a genetic basis for Ae . aegypti susceptibility to dengue virus infection and dissemination . Here , we provide direct evidence that a significant portion of natural phenotypic variation is genetically determined . We identify multiple genetic factors that control dengue susceptibility in a natural Ae . aegypti population , but show that the effect of these factors also depends on the virus genome . Irrespective of G×G interactions , the relatively large proportion of phenotypic variation explained by the individual mosquito markers ( up to 75 . 6% ) reveals the existence of QTL with major effects . Interestingly , QTL underlying the midgut infection phenotype explained a smaller proportion of the phenotypic variation than QTL underlying the viral dissemination and dissemination titer phenotypes , suggesting a different genetic architecture . This hypothesis is supported by a similar observation in an earlier QTL mapping study [23]–[25] . Alternatively , this could be due to differences in marker informativeness or because exclusion of uninfected mosquitoes ( on average , 57 . 5% of mosquitoes were uninfected in each family ) for analysis of dissemination reduces the contribution of other QTL to overall phenotypic variation . Genetic linkage observed in different mosquito families could result from distinct loci or different allelic variants of the same locus . Based on the present data , we show that midgut infection by dengue viruses is controlled by at least two QTL in this wild Ae . aegypti population . In infected mosquitoes , viral dissemination from the midgut to secondary tissues is also controlled by two or more QTL . Infectious titer of disseminated virus , a proxy for transmission potential [33] , is governed by three or more QTL . Our mapping strategy relies on marker-by-marker tests and does not generate a confidence interval of the QTL location on the chromosomes . In other words , conventional techniques of interval mapping cannot be applied . Therefore , we cannot ascertain at this stage whether QTL identified on chromosomes 1 and 3 match those previously mapped for a DENV-2 strain in laboratory systems . On chromosome 1 , a midgut infection QTL was previously identified at 19 cM [25] and a dissemination QTL at 31 cM [23] . On chromosome 3 , a dissemination QTL was previously identified between 44 and 52 cM [23] , [24] . No QTL was reported at the extremities of chromosome 3 in earlier studies . In the present study , significant linkage detected in the vicinity of the sex-determining locus ( 38 . 0 cM on chromosome 1 ) in four different families for the infection phenotype ( Fig . 2A , 2C , 2D , 2E ) , in two families for the dissemination phenotype ( Fig . 3A , 3B ) , and in one family for the head titer phenotype ( Fig . 5C ) , could point to a major gene , or cluster of genes , controlling mosquito-virus interactions . Another important limitation of our marker-by-marker mapping strategy is that epistatic interactions between mosquito loci could not be measured . Intragenomic epistasis is a major component of the genetic architecture of quantitative traits [34] , including host susceptibility to pathogens [17] . It is recognized as an essential determinant of the structure and evolution of complex genetic systems [35] . The main innovation of our study design was to explicitly account for viral genetic diversity in the genetic mapping of mosquito susceptibility loci . This allowed detection of both generalist and isolate-specific susceptibility loci . Several of the significant markers were in linkage with the phenotype independently of the virus isolate . Thus , the genetic basis of Ae . aegypti susceptibility to dengue viruses comprises a generalist component that is effective against diverse isolates , including isolates belonging to different serotypes . This result was previously unknown and gives hope to identify antiviral genes that confer a generalist protection against a diverse array of viruses . On the other hand , our genetic survey detected an isolate-specific component of the mosquito genetic basis for dengue susceptibility , which we interpret as G×G interactions between the vector and the virus . Markers associated with G×G interactions explained a significant proportion of phenotypic variation ( from 7 . 8% to 16 . 5% ) . Identification of QTL associated with G×G interactions rules out the possibility that genetic specificity in this system is solely driven by environmentally inherited symbiotic microbiota that could have been confounded with the host genotype [16] . Note that this does exclude an indirect role of microbiota because the type of microbiota itself might be controlled by the host genotype . It will be interesting to carry out fine-scale mapping experiments to identify the causal polymorphisms and their allelic profiles in the genomic regions where significant markers were found . An extension of the same protocol could be used to generate outbred isofemale lines beyond the F2/F3 generations to increase mapping resolution and locate candidate genes . Although several resistance mechanisms have been characterized in laboratory systems , mosquito genes underlying phenotypic variation in susceptibility to dengue viruses in nature have remained elusive . Leading candidates are genes known to be functionally involved in Ae . aegypti antiviral defense , including genes of the RNA interference ( RNAi ) , JAK-STAT and Toll pathways [36]–[38] . A key gene of the RNAi pathway was recently associated with G×G interactions in this system [39] . The extremely low frequency ( ∼0 . 1% ) of dengue virus infected Ae . aegypti in nature [40] and the relatively modest fitness cost of infection [41] make it unlikely that occasional challenge by dengue viruses is a strong enough selective pressure to drive the evolution of these genes . Rather , we speculate that their evolutionary dynamics are shaped by their concomitant role in the response to more prevalent pathogens in wild mosquito populations [42] . Conversely , natural selection of viruses that are able to evade or suppress resistance mechanisms is more likely to occur . Selection for enhanced transmission by mosquitoes has been proposed as a possible mechanism of adaptive evolution in dengue viruses [33] . Our results have at least two practical implications for the current development of novel strategies to interrupt virus transmission by genetically engineering resistant mosquitoes [43] , [44] . First , the observation that Ae . aegypti vector competence for dengue viruses is controlled by multiple segregating QTL in a natural population suggests that such strategies may need to knock-down a larger number of genes than previously thought to confer complete resistance . Second , our discovery that the effect of several QTL is dengue virus serotype- and/or isolate-specific highlights the requirement for engineered resistance to be effective across all possible virus serotypes and strains encountered in nature . In conclusion , our findings reinforce the idea that contributions from different genomes to the genetic architecture of ecological interactions cannot be fully disentangled because they depend on one another . By analogy with epistasis within the genome of a single organism , whereby the effect of a particular genotype on the phenotype depends on the genetic background , the direction and/or magnitude of the effect of host genes may depend on the pathogen genetic make-up . Like epistasis [45] , [46] , such G×G interactions between the genomes of two ( or more ) interacting organisms may constitute a significant component of the genetic architecture of complex traits resulting from ecological interactions . This may be true not only for antagonistically interacting organisms such as hosts and pathogens , but also for mutualistic interactions between , for example , animals and their gut microbiota or plants and their root microbiota [47] , [48] . Accounting for the contribution of such genetic interactions between genomes will advance our understanding of the full genetic architecture of complex interaction traits in nature .
Wild mosquito eggs were collected using ovitraps in several households in the Nhong Pling , Kon Tee , Mae Na Ree , Nhong Ping Kai , and Thep Na Korn subdistricts , Muang district , Kamphaeng Phet Province , Thailand , during May 2010 and February 2011 . Kamphaeng Phet Province is an agrarian , sparsely populated area located approximately 350 km northwest of Bangkok where dengue is endemic and the four dengue virus serotypes co-circulate [49] . All collections were made in rural villages located within a localized area of less than 850 km2 . F0 eggs were brought back to the AFRIMS laboratory in Bangkok and allowed to hatch in filtered tap water . F0 pupae were separated and allowed to emerge in individual vials . Aedes aegypti adults were identified by visual inspection . Single F0 pairs consisting of one virgin male and one virgin female were allowed to mate for 2–3 days following emergence . To avoid that F0 parents were siblings from the same wild mother , the male and the female of each pair were chosen from different collection sites . Inseminated females were offered daily blood meals and allowed to lay eggs . Egg batches from a single female were merged to obtain a pool of F1 eggs . F0 males and females were saved for later DNA extraction and typing . F2 and F3 families were produced by mass sib-mating and collective oviposition from the F1 offspring . Although the mass-mating step reduces statistical power to detect genetic linkage because parental genetic information is partially lost , it is traded for a considerable increase in sample size [28] . A single Ae . aegypti pair can produce several thousands progeny per generation after as few as 2–3 generations in the laboratory . F1 adults were allowed to emerge in the laboratory , mate randomly , and feed on defibrinated sheep blood ( National Laboratory Animal Center , Mahidol University , Bangkok , Thailand ) through a membrane feeding system . The F2 and F3 eggs were collected and stored on dry pieces of paper towel and maintained under high humidity no longer than 6 months . Although most Ae . aegypti females are inseminated by a single male in nature [50] , using single pairs of newly emerged mosquitoes instead of naturally inseminated females allowed us to genotype both F0 parents prior to phenotyping . Families are not equal in the information they bring to QTL detection . Only families with the highest proportion of polymorphic markers were retained for genetic mapping . The aim of choosing families was to maximize the number of informative ( i . e . , segregating ) meiosis at both marker and susceptibility loci . Out of a total of 184 initial mating pairs , nine families were selected that had >3 , 000 F2/F3 eggs and >80% polymorphic markers . Four low-passage dengue virus isolates ( two DENV-1 and two DENV-3 ) were used to orally challenge mosquitoes in vector competence assays ( Table S2 ) . They derived from serum samples collected between March and July 2010 during routine surveillance for diagnostic public health testing at AFRIMS from clinically ill dengue patients attending Kamphaeng Phet Provincial Hospital . Phylogenetic analysis assigned the viruses to known lineages of DENV-1 and DENV-3 that were circulating in Southeast Asia in the previous years ( Fig . S1 ) . Each isolate was amplified twice in Aedes albopictus cells ( C6/36 , ATCC CRL-1660 ) , which is the minimum required to obtain a viral titer sufficiently high to infect mosquitoes orally using an artificial blood meal . To prepare virus stock , 0 . 2 ml of human serum was inoculated onto 2-day-old confluent C6/36 cells in a 25-cm2 flask and incubated for 7 days at 28°C . The virus-infected cell culture supernatant was harvested and inoculated into a fresh flask of 2-day-old C6/36 cells for the second passage , of which supernatant was aliquoted and stored at −70°C . Viral genomic RNA was extracted from viral stock with the QIAamp viral RNA kit ( Qiagen , Valencia , CA , USA ) . RT-PCR was performed using the SuperScript One-Step RT-PCR kit with platinum Taq polymerase ( Invitrogen Life Technologies , Carlsbad , CA , USA ) according to the manufacturer's recommendations , with a set of primers covering the entire genome ( Table S3 ) . RT-PCR products were purified by ultrafiltration . Sequencing reactions were performed using the Big Dye Terminator v1 . 1 cycle sequencing kit ( Applied Biosystems , Foster City , CA , USA ) . Sequence chromatograms from both strands were obtained on an automated sequence analyzer ABI3730XL ( Applied Biosystems ) . For sequence analysis , contig assembly and sequence alignments were performed using BioNumerics v6 . 5 ( Applied-Maths , Sint-Martens-Latem , Belgium; www . applied-maths . com ) . Phylogenetic relationships were inferred using the maximum-likelihood method with the Tamura-Nei model implemented in MEGA v5 [51] . Reliability of nodes was assessed by bootstrap resampling with 1 , 000 replicates . The complete viral genome sequences were deposited to the GenBank database ( accession numbers HG316481–HG316484 ) . Ae . aegypti females of the F2 or F3 generation were used in vector competence assays to score their relative susceptibility to the four low-passage dengue virus isolates . Experimental infections were run in three large experiments that involved different triplets of mosquito families ( Table S1 ) . F2/F3 eggs were hatched synchronously by placing them under low pressure for 30 min . Larvae were reared in 24×34×9 cm plastic trays filled with 2 . 0 liters of filtered tap water at a density of approximately 200 first instars per tray and fed a standard diet of approximately 1 . 0 g of fish food pellets ( C . P . Hi Pro; Perfect Companion Group Co . Ltd . , Bangkok , Thailand ) per tray . Pupae were transferred to plastic screened 30×30×30 cm cages ( Megaview Science Education Service Co . Ltd . , Taichung , Taiwan ) and adults were maintained on a diet of 10% sucrose . They were kept in an insectary at 28±1°C , under a relative humidity of 70–80% and a 12∶12 h light-dark cycle . The day before the oral challenge , females were transferred from the rearing cage to 1-pint feeding cups of ∼100 females . Prior to experimental infections , 25-cm2 flasks of 2-day-old C6/36 cells were inoculated with a 1-ml aliquot from the viral stock and incubated for 45 min to 1 hour . At the end of the adsorption , 4 . 0 ml of maintenance medium were added and the cells were incubated at 35±1°C under 5% CO2 for 5 days . At day 5 , 1 . 0 ml of heat-inactivated fetal bovine serum containing 15% of sodium bicarbonate 7 . 5% solution ( HIFBS-NaHCO3 ) was added to the virus-infected cell culture supernatant , which was then harvested to prepare the infectious blood meal . The virus suspension was diluted 1∶3 or 1∶2 with RPMI 1640 medium containing 5% HIFBS and then mixed 1∶1 with defibrinated sheep blood ( National Laboratory Animal Center ) . The infectious blood meal was placed in water-jacketed glass feeders maintained at a constant temperature of 37°C and covered with a piece of desalted porcine intestine . Four- to 7-day-old Ae . aegypti females deprived of sucrose and water for 24 h prior to blood feeding were offered an infectious blood meal for 30 min . Samples of the blood meals were saved for subsequent titration . Blood meal titers ranged from 2 . 0×104 to 1 . 5×106 plaque-forming units per ml ( PFU/ml ) ; the majority ( 83 . 3% ) ranged between 1 . 0×105 and 1 . 0×106 PFU/ml ( Table S2 ) . Small differences in blood meal titers contribute to the isolate effect in the analysis , but we verified that it did not confound our interpretation ( see below ) . After blood feeding , mosquitoes were briefly sedated with CO2 from dry ice , and fully engorged females were transferred to clean 1-pint paper cups . Unfed or partially fed females were discarded . Engorged females were maintained for 14 days at 28±1°C , under 70–80% relative humidity and a 12∶12 h light-dark cycle and provided cotton soaked with 10% sucrose ad libitum . Vector competence was scored in the F2/F3 families at 14 days after the infectious blood meal according to three phenotypes: ( i ) midgut infection , ( ii ) viral dissemination from the midgut , and ( iii ) infectious titer in head tissues . Viral infection of midgut epithelial cells and subsequent dissemination to secondary tissues are two essential steps of dengue virus propagation in Ae . aegypti . Both events are prerequisites for virus transmission by mosquito bite and have been used to define a ‘midgut infection barrier’ and a ‘midgut escape barrier’ underlying Ae . aegypti susceptibility to dengue viruses [32] . These two vector competence indices were determined qualitatively ( i . e . , presence or absence of virus in mosquito bodies and heads , respectively ) . Although both phenotypes are binary traits ( all-or-nothing ) , they are assumed to be consistent with a multifactorial basis and to result from continuous variation on an underlying ( unobserved ) scale . Infectious titer of virus disseminated to head tissues is strongly correlated with the probability to detect virus in saliva samples collected in vitro [33] , and is therefore used as a proxy for transmission potential . Head titers were determined quantitatively by end-point titration . Upon harvest , the head of each female was cut off on a chill table and placed individually in 500 µl of mosquito diluent ( MD; RPMI 1640 medium with 10% HIFBS , 100 units/ml penicillin , 100 µg/ml streptomycin and 100 units/ml L-Glutamine ) . The remainder of the body ( thorax and abdomen ) was stored separately in 900 µl of MD with one 4 . 5 mm stainless steel bead in a 2-ml safe-lock tube . Samples were stored at −70°C until testing by plaque assay . They were quickly thawed in a water bath at 35±2°C and homogenized in a mixer mill ( Qiagen ) at 24 cycles/sec for 2 min . Four hundreds µl of each body homogenate were transferred into a new 1 . 5 ml safe-lock tube containing 400 µl of lysis buffer BQ1 ( Macherey-Nagel , Düren , Germany ) and stored at −20°C for DNA genotyping . Infectious virus was detected and quantified by plaque assay performed in rhesus monkey kidney epithelial cells ( LLC-MK2 , ATCC CCL-7 ) as previously described [52] . Briefly , the homogenized body and head samples were filtered individually through a sterile , syringe-mounted 0 . 22-µm membrane . The samples were placed in an ice bath , 100 µl/well were inoculated onto a monolayer of 3-day-old LLC-MK2 cells in 24-well plates . The virus was adsorbed at room temperature ( 20–28°C ) on a rocker platform for 90 min . The inoculum was removed and 0 . 5 ml/well of a first overlay of medium was added . The cells were incubated for 5 days at 35±1°C under 5±0 . 5% CO2 . The cells were stained with a second overlay of medium containing 4% neutral red ( Sigma Chemical Co . , Perth , WA , USA ) . Mosquito infection and dissemination status was determined based on the presence of plaques in their body and head homogenates , respectively . Mosquito whose bodies were negative by plaque assay were considered uninfected , and their heads were not processed further . Head titer of infected bodies was determined by plaque assay of 1∶10 and 1∶100 dilutions of head homogenates . QTL detection was performed in the outbred mosquito families using a set of 25 microsatellite markers broadly distributed across the genome ( Fig . S2 ) . Genetic position and PCR primers sequences for these markers were readily available from published literature [53] , [54] with the exception of markers 210TTC1 and 14ATT1 that we developed ( see below ) in an attempt to increase chromosome 2 coverage . In our Ae . aegypti population , few existing chromosome 2 markers were valid and/or informative , and despite our efforts to find additional markers , coverage remained too low to provide a sufficient mapping density of markers . The paucity of unique sequences among supercontigs mapped to chromosome 2 made it extremely difficult to design primer pairs resulting in unique PCR products . Efforts are currently being made to develop alternative markers based on single nucleotide polymorphisms ( SNPs ) . For each marker in the final map ( Fig . S2 ) , we verified that the pair of primers matched a unique supercontig of the unassembled Ae . aegypti genome [55] , which in turn was anchored to the reference genetic map [56] by the co-presence of another marker with known genetic position that uniquely matched the same supercontig . The only exception is marker B19 that falls in an unmapped supercontig but was independently assigned to chromosome 3 by linkage analysis [53] . The 25 microsatellites represent 18 distinct genetic positions along the Ae . aegypti genome . Twenty-two of these microsatellites ( 15 genetic positions ) are located on chromosomes 1 or 3 . Based on an estimated genome size of 1 , 376 Mbp and a genetic size of 205 centiMorgans ( cM ) , the relationship between physical and recombination distance is 6 . 71 Mbp/cM [55] , [56] . Estimated genetic sizes of chromosomes 1 and 3 are 70 . 6 and 64 . 2 cM , respectively [56] . For these two chromosomes , adjacent markers in our genetic survey were separated by an average distance of 9 . 0 cM ( 60 . 3 Mbp ) . Thus , an unknown QTL was on average less than 4 . 5% recombination away from a marker . The genetic survey was based on the analysis of outbred mosquito families at the F2 or F3 generation . Each mosquito family descended from a single pair of F0 parents collected in the field , providing an independent sample of up to four different alleles per locus from the original natural mosquito population . Based on the number of alleles present at the F0 generation , we verified at each marker that the correct number of genotypes was observed in the progeny . Three , six and ten different genotypes are expected in the progeny when F0 parents harbor two , three and four different alleles , respectively . The originality of the strategy is to use families with incomplete pedigree information due to the mass-mating step [28] . Mosquitoes are classified according to their genotype so that identity by state ( IBS ) is used as a surrogate for identity by descent ( IBD ) . Genetic linkage is not inferred from allele sharing proportions but from genotype-phenotype associations . Therefore , allele segregation in Mendelian proportions is not required by the study design . During mass mating and collective oviposition allele frequencies may be distorted because of random genetic drift or natural selection . Genetic drift is particularly likely to occur at the F1 generation because the number of reproducing adults is relatively small . Some genotypes could also be selected because they have a fitness advantage over other genotypes in insectary conditions . Departure from a neutral reproductive model may reduce the statistical power to detect marker-trait associations , but not the statistical significance of results . The same is true for null alleles or genotyping errors that would confound the observed genotypes . Our genetic model does not specify allelic codominance or recessivity . It simply compares genotypes ( or groups of genotypes if a null allele segregates ) regardless of their frequency . Statistical power is also limited by the extent of heterozygosity in the family . There is no guarantee that every F0 parent is heterozygous both at a QTL and at a linked segregating marker , which is a prerequisite to generate a marker-trait association in the progeny . We maximized statistical power by genotyping F0 parents and choosing the most informative families ( i . e . , with >80% of markers being polymorphic ) for phenotyping . In addition , the linkage phase between the marker and the QTL can vary in the progeny . This can reduce QTL detection power , if for example the same marker allele is associated with different QTL alleles in the F0 parents . Again , this would increase the probability to declare significant evidence against marker-trait association ( i . e . , in support of the null hypothesis ) but not the statistical significance of results . Microsatellite markers 210TTC1 and 14ATT1 on chromosome 2 were developed as previously described [54] . Briefly , supercontig sequences containing genetic markers mapped to chromosome 2 were retrieved from VectorBase ( http://aaegypti . vectorbase . org/ ) and submitted to the Tandem Repeats Finder program [57] using default parameters with the exception of a maximum period size of 3 . For tandem repeats with a consistent motif and a repeat copy number <30 , a ∼500 bp sequence encompassing the microsatellite was subjected to BLASTn analysis against the Ae . aegypti genome in VectorBase to verify their occurrence in single copy . PCR primers were designed in flanking sequences of selected microsatellites using Primer3 v0 . 4 . 0 [58] , with an amplicon size target of 100–500 bp in length . The primer sequences were 5′-TCATTCCCAGTACCACACAAACG-3′ ( forward ) and 5′-ACTCGTTACTGGATGTGCTATCCC-3′ ( reverse ) for marker 14ATT1 and 5′-GAACGCGCTCGTAAGCGAGA-3′ ( forward ) and 5′-CACTGTGCGTTGGTTTCGGCT-3′ ( reverse ) for marker 210TTC1 . Individual primer pairs were further subjected to BLASTn analysis to verify that they were predicted to amplify single copy sequences in the Ae . aegypti genome . PCR products were run by electrophoresis on 2% agarose gel to confirm that amplicons were unique . Genomic DNA was extracted from mosquito homogenates using the NucleoSpin 96 Tissue Core Kit ( Macherey-Nagel ) and stored at −20°C until use . Genotyping of microsatellite repeats was performed by PCR amplification using fluorochrome-labeled forward primers ( 5′-FAM , 5′-HEX or 5′-ATTO550 ) ( Eurofins MWG Operon , Ebersberg , Germany ) to generate fluorescent PCR products . Primer pairs producing different amplicon sizes were assembled into multiplex groups of 4–6 markers . Amplification was performed in 25 µl volumes in Thermo-Fast 96-wells PCR plates ( ABgene , Epsom , Surrey , UK ) in a Veriti thermal cycler ( Applied Biosystems ) . Each reaction contained 1× Taq buffer ( 50 mM KCl , 20 mM Tris pH 8 . 4 ) ( Invitrogen Life Technologies ) , 1 . 5 mM MgCl2 , 200 µM dNTPs ( Invitrogen Life Technologies ) , 0 . 2 µM of each primer , 1 unit of Taq DNA polymerase ( Invitrogen Life Technologies ) , and 2 µl of genomic DNA purified as described above . Thermocycling conditions were 5 min at 94°C , followed by 35 cycles of a 30-sec denaturation at 94°C , a 30-sec annealing at 50°C , and a 1-min extension at 72°C , followed by a 7-min final extension at 70°C . Multiplexed PCR products were examined by electrophoresis on 1% agarose gel and diluted 1∶10 in sterile water . Two µl of this dilution was added to 10 µl of Hi-Di Formamide ( Applied Biosystems ) containing 7 . 5% of 6-carboxy-X-rhodamine ( ROX ) -labeled Geneflo 625 size standards ( EurX , Gdansk , Poland ) . Capillary electrophoresis of multiplexed PCR products was performed on a 3730xl DNA Analyser ( Applied Biosystems ) . Sizes of microsatellite alleles were called and manually checked using the GeneMapper v4 . 0 software package ( Applied Biosystems ) . Our approach is a combination of linkage and association analyses . Linkage analysis generally uses pedigrees to infer the location of a susceptibility locus based on coinheritance of the disease phenotype with genetic markers whose chromosomal location is known . Association analysis does not rely on pedigree structure but assumes that strong associations between marker alleles and disease phenotype in a population will be due to linkage , rather than chance . In association studies , IBD due to coancestry is inferred from IBS in the form of observed allelic associations . In the present study , linkage was inferred from IBS as in association studies . Tests of genotype-phenotype associations , however , were performed in sibships ( single-generation families ) at the at the F2 or F3 generation . In contrast with association studies performed at the population level , high linkage disequilibrium in the families strongly reduces the marker density required for the genetic mapping . Genetic linkage was inferred from the significance of the genotype effect in a generalized linear model of the phenotype that included the factors mosquito genotype , virus isolate and their interaction as explanatory variables . Response variables were the three vector competence indices that we measured: ( i ) midgut infection status , ( ii ) viral dissemination status of midgut-infected mosquitoes , and ( iii ) head titer in mosquitoes with a disseminated infection . For binary phenotypes ( infection and dissemination ) , the model was fitted with a binomial error structure and a logit link function ( i . e . , a logistic regression ) . For the continuous phenotype ( head titer ) , the variable was log-transformed and the model was fitted with a normal error distribution and an identity link function ( i . e . , a linear regression ) . The model was fitted separately for each informative microsatellite marker in each mosquito family . Depending on the number of alleles of the marker , the factor genotype had from three to ten different categories , whereas the factor isolate always had four categories ( i . e . , the four isolates used in the study ) . If , due to random sampling effects in the progeny , one category of the genotype was not encountered in one or more categories of the isolate , this genotype category was excluded from the analysis so that the genotype by isolate interaction could be tested in the model . Depending on the marker , this could result in a different number of mosquitoes included in the analysis for the same family . Statistical significance of the genotype effect or the genotype by isolate interaction effect in the above model was determined differently for binary ( infection and dissemination ) and continuous ( head titer ) variables . For binary phenotypes , statistical significance was tested with an analysis of deviance [59] . The deviance measures the unexplained variation of the data for a given model . The difference between the deviances of two models measures whether the two models fit the data differently . We first tested whether a model with the factors isolate and genotype fitted the data significantly better than a model with only the isolate ( i . e . , testing whether the genotype is a significant predictor of the phenotype ) . Then we tested whether a model with isolate , genotype and genotype by isolate interaction fitted the data better than the model with only the main effects of isolate and genotype ( i . e . , testing whether the interaction is a significant predictor of the phenotype ) . To estimate the proportion of variation explained by a significant factor we compared the mean deviance ( deviance divided by the number of degrees of freedom ) of the model including the factor and the mean deviance of the model without the factor . For the continuous phenotype , statistical significance was tested with an analysis of variance . To estimate the proportion of variation explained by a significant factor we followed the approach described above for the binary phenotypes . We compared the residual variance ( sum of squares divided by the number of degrees of freedom ) of the model including the factor and the residual variance of the model without the factor . Because we performed multiple tests for each mosquito family , we used a Bonferroni correction of the p-values to ensure a genome-wide type I error of at most α = 0 . 05 ( i . e . , no more than 5% false positives overall ) . The genome-wide significance level of the test at each marker was α/N , where N is the number of informative markers tested in each family . A genotype-phenotype association was declared significant at the genome-wide level if the nominal p-value was smaller than α/N . When a significant genotype by isolate interaction was found , we verified that uncontrolled differences in the infectious titer of the artificial blood meal ( Table S2 ) did not confound our interpretation of the factor isolate as an approximation of viral genetic identity . We performed an analysis based on the same model as previously but replacing the isolate by the corresponding blood meal titer ( log-transformed ) . If the isolate effect were only due to differences in blood meal titer , we expect that the effect would remain statistically significant . Conversely , if the effect became insignificant , it would mean that the isolate effect resulted primarily from an effect of the viral genetic polymorphism rather than a simple effect of the infectious dose . All statistical analyses were performed in the statistical environment R [60] . | The outcome of invertebrate host-pathogen interactions often depends on the specific pairing of host and pathogen genotypes . This genetic specificity challenges our current understanding of invertebrate resistance to pathogens because it contrasts the limited discriminatory power of known invertebrate defense mechanisms . However , genetic factors underlying this observed specificity have remained elusive , questioning their very existence . In this study , we developed an innovative strategy to localize factors in the genome of the mosquito Aedes aegypti that govern specific interactions with dengue viruses . We used large mosquito families derived from a natural population in Thailand that we experimentally challenged with virus isolates obtained from human patients living in the same area . We identified several regions of the mosquito genome that control specific interactions with dengue viruses and contribute significantly to the observed variation in vector competence . Our study establishes the existence of tangible host genetic factors underlying specific interactions between invertebrates and their pathogens in a natural system that is relevant to human health . This represents a critical step towards identification of mechanisms underlying the genetic specificity of insect-virus interactions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"heredity",
"vector",
"biology",
"virology",
"genetics",
"host-pathogen",
"interaction",
"quantitative",
"traits",
"biology",
"evolutionary",
"biology",
"microbiology",
"genetics",
"of",
"disease"
] | 2013 | Genetic Mapping of Specific Interactions between Aedes aegypti Mosquitoes and Dengue Viruses |
Burkholderia pseudomallei is a soil saprophytic bacterium that causes melioidosis . The infection occurs through cutaneous inoculation , inhalation or ingestion . Bacteriophages ( phages ) in the same ecosystem may significantly impact the biology of this bacterium in the environment , and in their culturability in the laboratory . The soil samples were analysed for the presence of bacteria using culture methods , and for phages using plaque assays on B . pseudomallei strain 1106a lawns . Of the 86 soil samples collected from northeastern Thailand , B . pseudomallei was cultured from 23 ( 26 . 7% ) samples; no phage capable of infecting B . pseudomallei was detected in these samples . In contrast , phages capable of infecting B . pseudomallei , but no bacteria , were present in 10 ( 11 . 6% ) samples . B . pseudomallei and their phages were co-isolated from only 3 ( 3 . 5% ) of soil samples . Since phage capable of infecting B . pseudomallei could not have appeared in the samples without the prior presence of bacteria , or exposure to bacteria nearby , our data suggest that all phage-positive/bacteria-negative samples have had B . pseudomallei in or in a close proximity to them . Taken together , these findings indicate that the presence of phages may influence the success of B . pseudomallei isolation . Transmission electron microscopy revealed that the isolated phages are podoviruses . The temperate phages residing in soil-isolated strains of B . pseudomallei that were resistant to the dominant soil borne phages could be induced by mitomycin C . These induced-temperate phages were closely related , but not identical , to the more dominant soil-isolated phage type . The presence of podoviruses capable of infecting B . pseudomallei may affect the success of the pathogen isolation from the soil . The currently used culture-based methods of B . pseudomallei isolation appear to under-estimate the bacterial abundance . The detection of phage capable of infecting B . pseudomallei from environmental samples could be a useful preliminary test to indicate the likely presence of B . pseudomallei in environmental samples .
Burkholderia pseudomallei is a motile , Gram-negative , non-spore-forming bacterium that causes melioidosis [1] . The disease is endemic in Southeast Asia and Northern Australia . The clinical manifestations of melioidosis range from localized infection to sepsis and death , with pneumonia being the most common presentation [2] . Treatment with ineffective antimicrobials may result in case fatality rate exceeding 70% [3] . Currently , there is no licensed vaccine available . The bacteria are intrinsically resistant to many antibiotics . In a typical clinical case , parenteral treatment with ceftazidime is given for at least 10 days , followed by oral treatment with a four-drug combination ( chloramphenicol , doxycycline , trimetoprim-sulfamethoxazole ) for 20 weeks . Due to its aerosol infectivity , the high mortality rate and the absence of effective human vaccine available for the treatment of melioidosis [1 , 4] , B . pseudomallei has been recognised as a potential bio-threat agent , and it has been listed as category B disease/agents by the U . S . Centers for Disease Control and Prevention [1 , 5] . In the zones of endemicity , B . pseudomallei are commonly found in clay soils at a depth of about 25–45 cm , and the bacteria can move to the surface during the rainy season [6] . In Thailand , rice farmers rarely wear protective footwear , and thus they are exposed to a risk of infection with B . pseudomallei by cutaneous inoculation . Moreover , heavy rain may create aerosols contaminated with the bacteria , and this can result in the inhalation of the organism [7] . The existence of these risk factors is correlated with a high incidence of melioidosis in the rainy seasons . There are reports that temperature , pH , and water content in the soil might affect B . pseudomallei survival [8 , 9] . In addition to soil physicochemical properties , biological factors such as bacteriophages ( phages ) present in the same ecosystem as B . pseudomallei may affect the density of this bacterium in the environment . Our previous work has revealed that B . pseudomallei phages can be readily isolated from the soil environment [10 , 11] . Phages are the most abundant life form on Earth , and they are frequently found to co-exist with their bacterial hosts with approximately 10 phage particles for every bacterial cell . They can be found in most environments , such as sewage , soil , and water [12 , 13] . In general phages either follow a lytic or a lysogenic life cycle so they can shape microbial communities by lysing their hosts or alternatively by lysogenizing them where they may provide phenotypic advantages to recipient bacteria [14 , 15] . Phages are thus considered to have major role with respect to bacterial abundance , population structure and diversity in a variety of environments including soil . There is only limited information available about the role of temperate phages in the virulence of B . pseudomallei . The B . pseudomallei K96243 genome is organized into two circular chromosomes [16] . The genome is highly mobile with sixteen genomic islands constituting over 10% of the total genome . Eight of these islands have some prophage like characteristics , and one genomic island was shown to encode phiK96243 phage [16 , 17] . More recently , Gatedee et al [10] reported the isolation of ΦBp-AMP1 and other B . pseudomallei phages from soil samples . Almost all of the isolated B . pseudomallei phages were podoviruses , suggesting that such phages are dominant in the environment . Interestingly , there are no ΦBp-AMP-1-like prophages detected in the sequenced B . pseudomallei strains . This lack of such prophages may due to the commonly used culture temperature of 40°C for B . pseudomallei isolation , which results in the phage induction , and thus remaining bacteria are phage-free . Further characterization by our group [11] showed that a class of temperate podoviruses that target B . pseudomallei go through a lytic cycle at 37°C , whereas at 25°C they infect their host bacteria but remain temperate . These lysogens are relatively stable at lower temperature but when the bacteria are incubated at 37°C they are induced with a high frequency . Similarly , when lysogens are passaged through mice , in most cases the prophages are induced and thus the bacteria are killed . However , the small numbers of bacteria that are recovered are phage-free , suggesting that on entry to the mouse these bacteria have lost their phages without being killed . In order to assess whether B . pseudomallei could be co-isolated with their dominant phages from the environment , multiple soil samples were collected from northeastern Thailand where melioidosis is endemic . The presence of B . pseudomallei in the samples was assessed by culture . Phages were detected using spot assays and their presence confirmed by plaque assays on B . pseudomallei strain 1106a lawns . In the majority of samples that contained B . pseudomallei or phages , only one of the partners was present; bacteria and phages were co-isolated only in a small number of samples . The B . pseudomallei strains that were co-isolated with bacteria were found to be relatively stable lysogens and were resistant to infection from the dominant soil phage type . Furthermore their own resident temperate phages could be induced from the bacteria using mitomycin C ( MMC ) . The induced phages were compared to the soil-isolated phages by assessing their morphology using Transmission Electron Microscopy ( TEM ) and restriction analysis of the phage DNA .
Burkholderia pseudomallei strain 1106a was chosen as a propagation strain for phage isolation and purification because its genome lacks prophage islands [18] , and because it has a multilocus sequence type ( ST ) 70 , which is the most abundant genotype in northeastern Thailand [19] . All B . pseudomallei strains were cultured on Luria-Bertani ( LB ) agar ( Hardy Diagnostics , USA ) and incubated at 37°C for 18–24 hours . To obtain mid-log phase cells , 10 μl of overnight-cultured B . pseudomallei were sub-cultured into 3 ml of LB broth ( Thermo Fisher Scientific , USA ) and incubated at 37°C for approximately 4–6 hours . One hundred and one soil samples were collected from a backyard and a rice paddy field of a melioidosis patient in the rainy season from Roi-Et province , northeast Thailand , which is an endemic area of melioidosis disease . The soil in this area is sandy loam soil . Sharp-ended polyvinyl chloride tubes were used to collect soil from a depth of 30 cm and 2 . 5x2 . 5 meters apart according to Wuthiekanun et al [20] . For B . pseudomallei isolation , 10 g of soil samples were weighed and added directly into 10 ml of threonine-basal salt solution ( TBSS ) plus Colistin 50 μg/ml ( TBSS-C50 ) , mixed , and then incubated at 40°C for 48 hours . Ten microliters of the broth culture was subcultured onto Ashdown agar plates and incubated at 40°C in air for 4 days as described previously [21] . Putative B . pseudomallei showing typical colonies on Ashdown’s agar were confirmed by the latex agglutination test [22] and arabinose assimilation test [23] . B . pseudomallei were characterized as arabinose non-assimilators and latex agglutination test positive . Isolation of phages from soil was performed according to previously described [10] . Essentially , 2 g of soil were transferred into 10 ml LB broth plus 5 mM CaCl2 before mixing thoroughly by inversion and incubated at least 1 hour at room temperature . Then , the mixture was centrifuged at 4000xg for 20 minutes and supernatant was collected for filtration through a 0 . 22-μm filter membrane . This phage filtrate preparation was used for phages isolation by spot assay [10] and confirmed for the presence of phages by double agar plaque assay [24] using B . pseudomallei strain 1106a as bacterial host . Temperate phages were induced from B . pseudomallei according to Wright et al [25] . In brief , freshly prepared MMC ( Sigma-Aldrich , USA ) was added to 10 ml of mid-log phase B . pseudomallei culture to a final concentration of 250 ng/ml . The optical density ( OD600nm ) was measured every 2–4 hours over 24 hours . After 24 hours of incubation , the sample was centrifuged at 4000×g for 10 minutes and the supernatant containing phages was passed through a 0 . 45-μm filter membrane ( Whatman , UK ) . The presence of phages was confirmed using the double agar plaque assay [24] . Negative staining was performed for phage morphology examination by TEM ( JEOL 1230 , Japan ) . Phages ( approximately 106−108 particles/ml ) were fixed with 2 . 5% glutaraldehyde in 0 . 1 M phosphate-buffered saline ( PBS; pH 7 . 3 ) for 3 hours , and a drop of this solution was dropped onto a carbon-coated copper grid ( 200 mesh ) . The excess liquid was removed by filter paper and the grid was dried . A drop of 2% phosphotungstate was added for 1 minute , and the excess liquid was removed with a piece of filter paper before further drying . Digital images were captured using a GATAN Orius 1k camera with associated analysis software . Phages were concentrated using polyethylene glycol ( PEG; Sigma-Aldrich , USA ) precipitation as previously described [26] . After mixing with PEG8000 , the phage suspension was centrifuged at 11 , 000×g for 25 minutes at 4°C and the pellet was re-suspended in SM buffer ( Storage Media; 100 mM NaCl , 10 mM MgSO4 . 7H2O , 50 mM Tris-HCl , pH7 . 5 ) . DNaseI ( 14 mg/ml ) and RNaseA ( 30 mg/ml ) ( Thermo Fisher Scientific , USA ) were then added , and the mixture was incubated at 37°C for 1 hour to eliminate bacterial nucleic acids . Following phenol/chloroform extraction , the aqueous phase was precipitated using absolute ethanol . The precipitated phage DNA was washed , dissolved in sterile deionized water , and stored at -20°C until use . For restriction enzyme analysis , the isolated phage DNA was digested with BstBI or MluI restriction endonucleases using the manufacturer’s recommended conditions . Digested DNA was separated by agarose gel electrophoresis , and images were captured using GeneSnap acquisition software ( Syngene , UK ) . PCR was performed using primers specific for the phage tail tubular protein B gene . The forward ( 5′-TAAGGTAACAGGCAGCTACG-3′ ) and reverse ( 5′-ATTGAGCACGAAGCA GAACG-3′ ) primers were designed from B . pseudomallei ΦBp-AMP1 [10] . The reaction mixture contained 10 ng of phage DNA , 0 . 2 units of Taq DNA polymerase ( Bioline , London , UK ) , 0 . 4 μM of each primer , 250 μM of each deoxynucleotidetriphosphate ( dNTP ) , 1×PCR buffer , and 2 . 0 mM MgSO4 . Thermal cycling was performed in a TProfessional thermocycler machine ( Biometra , Germany ) with the following parameters: 95°C for 2 minutes , followed by 35 cycles of 94°C for 45 seconds , 51°C for 45 seconds , and 72°C for 1 minute , and a final extension at 72°C for 10 minutes . The amplified products were analyzed by agarose gel electrophoresis .
Northeastern Thailand is the endemic area of melioidosis where humans can be infected by having contact with B . pseudomallei residing in soil or water . Bacterial density in the environment is a crucial risk factor for infection . Despite a relatively high abundance of B . pseudomallei phages , little is known about their distribution and their impact on the B . pseudomallei population in natural habitats . To initiate research in this area , we collected 86 soil samples from a backyard of a melioidosis patient in the Roi-Et province , and processed the samples for the isolation of B . pseudomallei and their phages . Neither B . pseudomallei , nor phage capable of infecting B . pseudomallei could be detected in 50 samples . This could be due to the absence of bacteria and phages in these samples , or that their presence was at levels below the detection limit for the methods used in this study . Twenty-three ( 26 . 7% ) soil samples were positive for B . pseudomallei but not phages , as assessed by the growth of bacterial colonies on Ashdown agar plates ( Fig 1A ) . The isolation of B . pseudomallei from soil was carried out in accordance with recommendations provided by an authoritative paper in the field [27] . The authors of this paper clearly state that the incubation temperature of 40°C was suggested based on evidence that it allows good growth of B . pseudomallei but inhibits many other soil bacteria . In our project , we have also tried to incubate soil samples in liquid broth at 25°C ( instead of 40°C ) before plating at 25°C and found that many other bacteria over growth on the plate , even though there were antibiotics in the culture medium . B . pseudomallei colonies growing at 25°C on agar plate are hard to find due to the very small ( pin-point ) colony size and overgrowth of other bacterial species . In addition , it is hard to identify the suspected B . pseudomallei as colonies are not suitable for further confirmation by latex agglutination test due to the size and many contaminating colonies . To isolate phages , spot tests were used to test extracts from these soil samples , which were spotted on the B . pseudomallei 1106 a lawns . Ten ( 11 . 6% ) of the soil samples were positive for only the phages , as assessed by phage plaques formation on the lawns of B . pseudomallei strain 1106a ( Fig 1A ) . Note that we did not enrich for the presence of phages and thus would not have been able to observe phage presence if they were present at less than 5 plaque forming units ( PFU ) /ml of suspension which corresponds to 25 PFU/g of soil ( This is based on the detection method where 2 g of soil is suspended in 10 ml of LB broth , and then 200 μl aliquot is taken to quantify plaques by the double agar plaque assay ) . No B . pseudomallei colonies were grown on the Ashdown agar plates from these phage positive soil samples , when they were processed and plated . B . pseudomallei and their phages were only detected in 3 ( 3 . 5% ) of the samples analysed . Since phages capable of infecting B . pseudomallei could not have appeared in the samples without the prior presence of bacteria , or exposure to bacteria nearby , our data suggest that all phage-positive/bacteria-negative samples have had B . pseudomallei ( or possibly B . thailandensis ) in or in a close proximity to them , but these bacteria were killed by the phages either before or during the sampling itself or processing the samples . Thus , it could be concluded that currently adopted culture-based methods of B . pseudomallei detection in the environmental samples appear to under-estimate the bacterial abundance , at least in our sampling area . This under-estimation is likely to be at least due to the action of the phages e . g . the commonly used culture temperature of 40°C for B . pseudomallei isolation may lead to prophage induction and the bacteria are killed [11] . Since these phages appear to be abundant in the environment alongside with the bacteria , we propose that phage plaque assay on an indicator B . pseudomallei host could be a useful preliminary test to indicate the likely presence of B . pseudomallei in environmental samples . Although the results of such a test should be treated with some caution because many phages that infect B . pseudomallei can also infect the closely related B . thailandensis , they can provide a simple and rapid assessment of the likely presence of B . pseudomallei in the environment . To assess if the pattern of differential B . pseudomallei-phage detection in the soil samples was common , we repeated the experiment but sampled a smaller number of samples ( n = 15 ) in a rice field three kilometers from the first sampling site . A similar pattern of bacteria and phage distribution was observed: one sample was positive for only B . pseudomallei , four samples positive for only the phage , and three samples positive for both the bacteria and the phage ( Fig 1B ) . All three of the freshly soil-isolated B . pseudomallei ( RE1 , RE2 and RE3 ) which appeared to coexist with phages in the soil were chosen for phage induction by MMC treatment . The bacteria were subcultured at least 3–4 times to avoid phage contamination before culturing in LB broth with or without MMC treatment . Growth curves of MMC-treated B . pseudomallei isolates RE1-3 indicated the induction of temperate phages following treatment ( Fig 2A ) . The optical density ( OD600nm ) of MMC-treated B . pseudomallei strains was significantly reduced compared to the OD values of untreated bacteria . To confirm successful prophage induction , the supernatant of the MMC-treated B . pseudomallei cultures were subjected to a plaque assay . The results showed that the induced phages could infect a test strain of B . pseudomallei 1106a and yielded clear plaques approximately 3–5 mm in diameter; no plaque formation was detected when using B . pseudomallei RE1 , RE2 or RE3 as the bacterial host . The likely explanation for the apparent resistance of these strains is the presence of prophage within these host strains . We observed no difference in plaque morphology on the B . pseudomallei 1106a lawns for the MMC-induced phages and no difference in plaque morphology between the MMC-induced phages and our previously isolated free phages ΦBp-AMP1 [10] ( S1 Fig ) . The temperate phages derived from B . pseudomallei strains RE1-3 were subsequently designated ΦBp-RE1 , ΦBp-RE2 , and ΦBp-RE3 , respectively . No phage plaques were observed when samples from the control cultures incubated without MMC were spotted on the B . pseudomallei 1106a lawns , which indicates that there is no spontaneous release of the phages from the cultures of RE1-3 strains ( Fig 2A ) . However , it is possible that the induced phage ( if any ) is not able to infect B . pseudomallei 1106a indicator bacteria or the number of induced phages is lower the detection limit ( 5 PFU/ml ) . Together , these data suggest a successful temperate phage induction from all three soil-phage-resistant B . pseudomallei and that the induced temperate phages could infect B . pseudomallei strain 1106a . The temperate phages ΦBp-RE1 , ΦBp-RE2 , and ΦBp-RE3 were plaque-purified at least five times , and the phage morphology was assessed by TEM . All of the phages had icosahedral head ( diameter ~ 58 nm ) with a short and non-contractile tail , which is characteristic of phages in the Podoviridae family ( Fig 2B and 2C ) . The genome sizes of these phages , as assessed by pulsed-field gel electrophoresis , were estimated at approximately 45 Kb ( S2 Fig ) . This is similar with that for the ΦBp-AMP1 phages characterized in our previous work [10] . In addition , these data provide a likely explanation to why these three isolates were originally found co-isolated with the free soil phages: it is likely that they are lysogenized with the phages identical or at least highly similar to the dominant podovirus type abundantly found in soil , and this lysogeny provides a protection from the attack by the same or a related phage . To further compare the genetic features between MMC-induced B . pseudomallei phages ( ΦBp-RE1-3 ) and two newly isolated free phages ( ΦBp-RE4-5 ) , BstBI and MluI restriction enzyme digestion analyses of the phage genomic DNA were carried out . No significant difference between the BstBI restriction enzyme digestion patterns among the three MMC-induced phages and free phages were observed and these phages have the same calculated genome size of approximately 45 Kb ( Fig 3A ) . However , different DNA patterns were observed when the MluI restriction digestion profiles were analyzed ( Fig 3B ) . It appears that additional MluI recognition sites are present in the genomes of the induced phages , resulting in the alterations of the MluI digestion profile of the induced phages compared to that of the dominant soil-isolated ones . In addition to the restriction enzyme digestion analyses , PCR analysis with primers specific to a gene encoding the tail tubular protein B of ΦBp-AMP1 [10] were carried out on the MMC-induced temperate and the free phages . An approximately 325-bp-long DNA fragment was amplified when each of the phage samples was used as a template ( Fig 4A ) . To explore whether they have the same nucleotide sequences , DNA sequencing of the amplified DNA fragment was carried out . The amplified DNA fragment from MMC-induced temperate phages ( ΦBp-RE1-3 ) and free phages ( ΦBp-RE4-5 ) showed identical nucleotide sequences and had 100% nucleotide sequences identity to ΦBp-AMP1 ( Fig 4B ) , further suggesting the close similarity between these phages . The B . pseudomallei temperate podoviruses induced in this study differ from those observed in a previous study [26] , in which MMC treatment of a clinical isolated B . pseudomallei induced a temperate virus ( ΦP27 ) that belonged to the Siphoviridae family . In addition to MMC , Ronning et al [18] reported the induction of prophages from B . pseudomallei strains Pasteur 52237 , E12 , 644 by UV light , and found that the released viruses were from the Siphoviridae and Myoviridae families . Taken together , these data indicate the abundance of temperate phages in the environmental B . pseudomallei strains . The impact of such phages on B . pseudomallei biology is largely unknown . However , we have previously suggested that a group of temperature-dependent podoviruses are likely to significantly influence many aspects of B . pseudomallei existence in the environment . They are also likely to impact our ability to detect , and correctly enumerate bacteria from environmental samples [11] . The research presented here is consistent with our previous data and provides further evidence for the importance of this phage group . This is the first study that has assessed the presence of both B . pseudomallei and their phages in the same soil samples collected from the endemic area of melioidosis . The podovirus capable of infecting B . pseudomallei appear to be abundantly present in the soil . Currently adopted culture-based methods used for the detection of B . pseudomallei in the environmental samples appear to under-estimate the bacterial abundance due to the action of the phages . Some environmental isolates of B . pseudomallei appear to be relatively stably lysogenized by the podoviruses . Temperate phages could be induced from such B . pseudomallei strains isolated from soil by MMC . Restriction enzyme analyses indicated that MMC-induced phages are closely related to common free soil-isolated phages [10 , 11] . Further investigation of the phage-host infection networks and dynamics of complex phage-host communities will help us to reveal the role of phages player in shaping the density in the soil of the life-threatening bacterial pathogen B . pseudomallei . | Burkholderia pseudomallei is a motile , Gram-negative bacterium that causes melioidosis . The disease is endemic in Southeast Asia and Northern Australia and can be fatal . In the zones of endemicity , B . pseudomallei are commonly found in soils , and the bacteria can move to the surface during the rainy season . In Thailand , rice farmers rarely wear protective footwear , and thus they are exposed to a risk of infection with B . pseudomallei by cutaneous inoculation . Biological factors such as bacteriophages ( phages ) , viruses of bacteria , present in the same ecosystem as B . pseudomallei may affect the population dynamics of this bacterium in the environment . In order to study this , we have investigated the distribution patterns of B . pseudomallei and associated phages in nature and the impact of these phages on the bacterial culturability . This is the first study demonstrating that the presence of phages capable of infecting B . pseudomallei may affect the success of the pathogen isolation from the soil . Currently culture-based methods of B . pseudomallei appear to under-estimate the bacterial abundance . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] | [
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"bacteriophages",
"melioidosis",
"microbiology",
"viruses",
"bacterial",
"diseases",
"microscopy",
"molecular",
"biology",
"techniques",
"mammalian",
"genomics",
"microbial",
"genomics",
"research",
"and",... | 2016 | Analyses of the Distribution Patterns of Burkholderia pseudomallei and Associated Phages in Soil Samples in Thailand Suggest That Phage Presence Reduces the Frequency of Bacterial Isolation |
Bleeding tendency , coagulopathy and platelet disorders are recurrent manifestations in snakebites occurring worldwide . We reasoned that by damaging tissues and/or activating cells at the site of the bite and systemically , snake venom toxins might release or decrypt tissue factor ( TF ) , resulting in activation of blood coagulation and aggravation of the bleeding tendency . Thus , we addressed ( a ) whether TF and protein disulfide isomerase ( PDI ) , an oxireductase involved in TF encryption/decryption , were altered in experimental snake envenomation; ( b ) the involvement and significance of snake venom metalloproteinases ( SVMP ) and serine proteinases ( SVSP ) to hemostatic disturbances . Crude Bothrops jararaca venom ( BjV ) was preincubated with Na2-EDTA or AEBSF , which are inhibitors of SVMP and SVSP , respectively , and injected subcutaneously or intravenously into rats to analyze the contribution of local lesion to the development of hemostatic disturbances . Samples of blood , lung and skin were collected and analyzed at 3 and 6 h . Platelet counts were markedly diminished in rats , and neither Na2-EDTA nor AEBSF could effectively abrogate this fall . However , Na2-EDTA markedly reduced plasma fibrinogen consumption and hemorrhage at the site of BjV inoculation . Na2-EDTA also abolished the marked elevation in TF levels in plasma at 3 and 6 h , by both administration routes . Moreover , increased TF activity was also noticed in lung and skin tissue samples at 6 h . However , factor VII levels did not decrease over time . PDI expression in skin was normal at 3 h , and downregulated at 6 h in all groups treated with BjV . SVMP induce coagulopathy , hemorrhage and increased TF levels in plasma , but neither SVMP nor SVSP are directly involved in thrombocytopenia . High levels of TF in plasma and TF decryption occur during snake envenomation , like true disseminated intravascular coagulation syndrome , and might be implicated in engendering bleeding manifestations in severely-envenomed patients .
Snakebites , which have been considered a neglected tropical disease by the World Health Organizaton since 2009 , frequently evoke hemostatic disturbances . In Brazil , Bothrops snakes account for approximately 20000 snakebites annually [1] . Patients usually develop local inflammatory reactions at the site of the bite , e . g . , edema , local pain , ecchymosis , petechiae and necrosis , and also systemic bleeding manifestations , including gingival bleeding , hematuria , purpura , epistaxis , hemoptysis , among others . Thrombocytopenia , platelet dysfunction , and coagulation disorders are major laboratory findings observed in victims of Bothrops jararaca bites [2]–[4] . Eagle in 1937 [5] was the first researcher to notice that B . jararaca venom ( BjV ) contained at least two different principles that promoted the direct conversion of fibrinogen into fibrin , as well as the activation of prothrombin into thrombin , without the need of calcium or platelets . Snake venom metalloproteinases ( SVMP ) and serine proteinases ( SVSP ) , the two main protein families found in BjV with anti-hemostatic activity [6] , have been implicated in the hemostatic disorders associated with envenomation [7] . SVMP present in Bothrops venoms belong to a zinc-dependent enzyme family , which contributes to the inflammatory , proteolytic , hemorrhagic and procoagulant ( prothrombin and factor X activators ) activities in snake venoms [8]–[10] . Na2-EDTA completely inactivates the enzymatic activity of SVMP by chelation of divalent cations . The second most abundant enzyme class in BjV is SVSP [6] , which have a highly reactive serine residue . SVSP have been reported to affect platelet aggregation , blood coagulation and fibrinolysis , and several SVSP purified from BjV show anti-hemostatic activities [11] . Serine-modifying reagents , such as 4- ( 2-aminoethyl ) benzenesulfonyl fluoride hydrochloride ( AEBSF ) , are irreversible serine proteinase inhibitors [12] . The current model that explains how coagulant snake venoms promote consumptive coagulopathy was published more than one hundred years ago [13] . After the initial report by Felice Fontana in 1781 [14] that venom injection into animals caused paradoxical effects – i . e . , an initial phase of intravascular coagulation followed by a phase of blood incoagulability – , the mechanisms whereby this phenomenon occurred were not explained until the publication of the original observations of Mellanby in 1909 [13] . He noticed that rapid injection of small quantities of Notechis scutatus or Echis carinata ( sic ) venom caused massive intravascular coagulation , in virtue of the rapid production of thrombin and enormous production of fibrin; on the other hand , slow injection generated low quantities of thrombin that in turn produced gradual fibrinogen consumption . The complete consumption of plasma fibrinogen caused incoagulability , which was restored as fibrinogen levels augmented . Although various proteins isolated from BjV have been reported to cause systemic and local manifestation when injected into animals , no genuine approach has been used to evaluate the repercussion of local mediators generated or released at the site of venom inoculation on the induction of hemostatic disorders observed in vivo . Tissue factor ( TF ) , a 47-kDa transmembrane protein , is the cellular receptor for plasma factor VII/VIIa , and thus is an essential component for initiating blood coagulation in vivo . In steady state conditions , TF is usually excluded from the vascular compartment , and constitutive TF expression occurs particularly in vascular smooth muscle cells , adventitial fibroblasts and pericytes . In endothelial cells , monocytes and platelets , i . e . , cells in continuous contact with the bloodstream , TF is minimally expressed or is in an encrypted form; however , stimulation of these cells by various inflammatory mediators induces TF protein expression and activity in vitro . Although controversial , TF decryption has been attributed to the oxidoreductase protein disulfide isomerase ( PDI ) [15] , [16] . Interestingly , PDI has also been reported to be present in snake venom glands [17] , [18] , and may therefore be present in snake venoms [19] . Thus , snake venoms , by damaging tissues locally or systemically , and by promoting the activation of circulating platelets , endothelial cells and monocytes , might induce the expression and release of TF in bloodstream , resulting in the activation of blood coagulation . However , such a mechanism of coagulation activation has never been addressed in snake envenomation . Since fibrinogen consumption , thrombocytopenia , and secondary fibrinolysis are major hemostatic disturbances frequently observed in snakebite victims , the main objective of this study was to investigate the mechanisms that lead to the genesis of these laboratory signs in bites inflicted by B . jararaca . Furthermore , we evaluated whether TF levels were augmented in plasma and tissue samples obtained from animals during envenomation . We demonstrate that SVMP play a pivotal role in venom-induced coagulopathy and that the importance of TF release in plasma has been hitherto underestimated .
Lyophilized venom from adult specimens of B . jararaca snakes was obtained from the Laboratory of Herpetology , Butantan Institute . BjV was dissolved in sterile saline immediately before use . AEBSF , 1 , 10-phenanthroline ( o-phe ) , bovine serum albumin ( BSA ) , N-benzoyl-D , L-arginine-p-nitroanilide hydrochloride ( BAPNA ) , and bovine thrombin were purchased from Sigma ( USA ) , and sodium ethylenediamine tetraacetic acid ( Na2-EDTA ) from Bio-Rad ( USA ) . Aprotinin ( Trasylol ) was obtained from Bayer ( Brazil ) . To obtain rabbit anti-rat fibrinogen IgG , one rabbit was immunized i . m . with 500 µL of rat fibrinogen ( 4 . 32 mg/mL , [20] ) emulsified in 500 µL of Marcol-Montanide adjuvant; at fortnight intervals , the rabbit received four additional boosters in the same adjuvant . Anti-rat fibrinogen IgG was purified and biotinylated as previously described [21] . Rabbit anti-BjV serum was obtained as described elsewhere [22] . Rat thromboplastin was prepared as described elsewhere [23]; briefly , dried thromboplastin was diluted in saline ( 40 mg/mL ) , maintained at 50°C for 20 min , and then refrigerated at 4°C overnight; the supernatant was used in clotting assays . All other reagents were of analytical grade or better . Male Wistar rats , weighing 220–250 g , and two male 3 . 0-kg New Zealand rabbits , were obtained from the Animal House of Butantan Institute; they were supplied with free access to food and water . All procedures involving the use of animals were approved by the Animal Ethical Committee of Institute Butantan ( protocols 142/03 and 685/09 ) and were in accordance with the Guide for Care and Use of Laboratory Animals ( 2011 ) and the International Guiding Principles for Biomedical Research Involving Animals ( 2012 ) . Rats were anesthetized by intraperitoneal administration of xylazine ( 10 mg/kg b . w . ) /ketamine chlorohydrate ( 100 mg/kg b . w . ) . Prior to exsanguination , immunized rabbits were anesthetized with sodium thiopental ( 50 mg/kg , i . v . ) , and blood was collected through puncture of the carotid artery . SVMP and SVSP were inhibited by incubation with 13 mM Na2-EDTA and 4 mM AEBSF , respectively . In brief , 269 mM Na2-EDTA ( 52 µL ) or 200 mM AEBSF ( 19 µL ) was added to BjV solution ( 1 mL , 1 mg/mL ) , and incubated for 1 h at 37°C . As a control of inhibition , aliquots of saline were incubated with BjV , under the same conditions . In preliminary experiments , the effectiveness of o-phe in inhibiting SVMP was also tested . An aliquot ( 26 µL ) of 500 mM o-phe in ethanol was incubated with BjV solution , identically as described previously , and for control experiments , the same volume of vehicle was used . The effectiveness of Na2-EDTA in blocking the catalytic activity of SVMP was checked by assaying the minimum coagulant dose ( MCD ) [20] on citrated rabbit plasma , which is completely dependent on the coagulant activity of BjV metalloproteinases [24] . Clotting times were measured on a Start4 coagulometer ( Stago , France ) . Estimates and the associated uncertainty , expressed as a 95% confidence interval [25] , of MCD were calculated by linear regression analysis in Stata ( version 8 . 0 , USA ) , using logarithmic transformation of data . To test whether AEBSF , Na2-EDTA or o-phe blocked the catalytic activity of SVSP in BjV , the chromogenic substrate BAPNA was used [20] . Since BAPNA is hydrolyzed by snake venom serine proteases after the arginyl residue , we used it to detect the residual activity of SVSP in BjV . Animals were injected with aliquots of freshly-treated BjV , as described above , at the doses of 1 . 6 mg/kg b . w ( s . c . ) or 100 µg/animal ( i . v . ) . BjV doses were selected based on previous tests , and they reproduced the acute hemostatic disturbances characteristic of B . jararaca envenomation . Rats injected with saline-treated BjV or saline alone ( vehicle ) were used as positive or negative controls , respectively . To study acute hemostatic disturbances evoked by BjV , rats were anesthetized after 3 and 6 h , and blood was collected by puncture of the abdominal aorta and dispensed in plastic bottles containing anticoagulants . For complete blood counts , blood ( 500 µL ) was collected into plastic bottles containing 5 µL of 269 mM Na2-EDTA and 5 µL of Bothrops antivenin ( Institute Butantan , lot 1005107/C ) . Blood counts were determined in an automated cell counter BC-2800 Vet ( Mindray , China ) . To obtain plasma samples , blood ( 4 . 3 mL ) was collected into plastic bottles containing 700 µL of CTAD anticoagulant ( 75 mM trisodium citrate , 42 mM citric acid , 139 mM dextrose , 15 mM theophylline , 3 . 7 mM adenosine , 0 . 2 mM dipyridamole , and 2 µM imipramine ) [7] and 50 µL of Bothrops antivenin , and centrifuged at 2500 g for 15 min at 4°C . Serum samples were obtained by maintaining blood ( 500 µL ) without anticoagulant or antivenin at 37°C for 2 h , followed by centrifugation as mentioned above . One circular 4-cm diameter skin fragment , whose center was the point of BjV inoculation ( s . c . route ) , and one lung fragment ( s . c . and i . v . routes ) were also removed from each animal . Skin samples were sliced and used to determine BjV-induced hemorrhage , TF activity and protein expression , and PDI protein expression . TF activity and protein expression were also evaluated in lung samples . Skin ( 6 . 3 cm2 ) and lung samples ( 100 mg ) were immediately immersed in RIPA buffer ( 50 mM Tris-HCl , 150 mM NaCl , 1% Triton X-100 , 1% sodium deoxycholate , 0 . 1% SDS , pH 7 . 5 , containing 2 mM Na2-EDTA , 2 mM AEBSF , 2 µM aprotinin , 130 µM bestatin hydrochloride , 28 µM E-64 and 22 µM leupeptin ) , and frozen at −80°C . Tissues were macerated in an IKA T10 disperser ( Staufen , Germany ) , and frozen in dry ice and thawed in a water bath at 37°C for three times . The resultant emulsion was centrifuged at 13000 g for 10 min , and supernatants were frozen at -80°C until use . Plasma fibrinogen [26] , and hemorrhage in skin samples [27] were assayed as described elsewhere . TF activity in plasma , lung and skin samples was evaluated with Actichrome TF kit ( American Diagnostica , USA ) , according to manufacturer's instructions . In the case of tissue samples , the protein content was standardized to 1 . 95 mg/mL by the bicinchoninic acid protein method [28] prior to assays . Fibrin ( ogen ) degradation product ( FDP/fdp ) levels in plasma were evaluated by a home-made double-antibody sandwich ELISA assay , based on a previous protocol [21] , using rabbit anti-rat fibrinogen IgG for coating , rat fibrinogen ( 3 . 9-1000 ng/mL ) as standard , and biotinylated rabbit anti-rat fibrinogen IgG . To remove residual fibrinogen from plasma , samples ( 200 µL ) were initially incubated with aprotinin ( 10000 U/mL , 10 µL ) and thrombin ( 30 U/mL , 200 µL ) for 15 min at 37°C , and centrifuged at 10000 g for 15 min . Prothrombin time was assayed by incubating plasma samples ( 80 µL ) with rat thromboplastin ( 40 µL ) for 1 min at 37°C , and then 50 mM CaCl2 ( 40 µL ) was added and clotting times were measured . Factor VII ( FVII ) coagulant activity was determined using FVII-deficient plasma ( HemosIL , USA , 40 µL ) incubated with plasma samples ( 40 µL , 1/50 dilution ) and rat thromboplastin ( 40 µL ) at 37°C for 1 min; clotting time was measured after the addition of 50 mM CaCl2 ( 40 µL ) . A standard curve was constructed by using a pool of normal rat plasma diluted from 1/10 to 1/800 , and considering the 1/50 dilution as 100% of FVII . All clotting times were measured on a Start4 coagulometer . Circulating BjV levels in serum was assayed by a modification of a procedure previously described [29] . Briefly , Nunc 96-well microplates were coated with commercial Bothrops antivenin ( 100 µg/mL , Institute Butantan , lot 1001103/D ) , and blocked with 3% BSA in carbonate buffer , pH 9 . 6 . Then , 100 µL of diluted serum samples ( 1/10 in incubation buffer [PBS containing 1% BSA and 0 . 05% Tween 20] ) or venom standards ( 1 . 95-500 ng/mL BjV diluted in incubation buffer containing 10% of a pool of normal rat serum ) were added to wells . Subsequently , rabbit anti-BjV serum ( 1/1000 ) , and goat anti-rabbit IgG-peroxidase antibody were used , and reaction was developed using o-phenylenediamine . TF and PDI protein expression in skin samples ( 50 µg protein/lane ) was evaluated by western blotting . Briefly , proteins were electrophoresed under reducing conditions in 12% SDS-PAGE gels [30] and transferred onto 0 . 2-µm nitrocellulose membranes . Subsequently , membranes were blocked , incubated at room temperature for 2 h with either a 1∶1000 mouse monoclonal anti-TF antibody ( TF9-10H10 , Calbiochem , USA ) or 1/10000 rabbit polyclonal anti-PDI antibody ( Sigma P7372 ) in blocking solution , washed , and subsequently incubated with 1/10000 peroxidase-conjugated anti-mouse IgG ( Sigma A4416 ) or anti-rabbit IgG ( Sigma A0545 ) . Expression of glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) , used as a loading control , was evaluated using a peroxidase-conjugated anti-GAPDH antibody ( Sigma G9295 ) . Membranes were developed as reported elsewhere [20] , scanned with resolution of 300 dpi , and densitometric analyses were done with TotalLab TL100 software ( USA ) . For relative quantification [31] , optical densities of bands ( volumes ) were divided respectively by the total optical density of lanes in membranes stained with Ponceau S . One sample from a saline-injected animal was used as an internal control throughout experiments , and it was considered as 1 for determining relative expression of protein bands . The efficiency of preincubation of BjV with inhibitors , routes of BjV administration , and time periods were compared using ANOVA , followed by the Tukey test . TF activity in lung and skin samples was compared by Student's t test . Whenever necessary , data transformation was undertaken to obtain homocedasticity and normal distribution . Statistical analyses were performed using the softwares SigmaStat ( version 3 . 5 , USA ) and Stata ( version 8 . 0 , USA ) . Differences with p<0 . 05 were considered statistically significant . Data were expressed as mean ± standard error of mean ( s . e . m . ) . Accession numbers for proteins studied herein , according to UniProtKB/Swiss-Prot database , are: tissue factor ( P42533 ) ; protein disulfide isomerase ( P04785 ) , fibrinogen ( P06399 , P14480 , P02680 ) , hemoglobin ( P01946 , P02091 ) , and factor VII ( Q8K3U6 ) .
In this study , we used the specific inhibitor AEBSF to inhibit serine proteinases . In order to block the enzymatic activity of SVMP in the venom , we initially compared two non-specific inhibitors largely used in toxinology research , Na2-EDTA and 1 , 10-phenanthroline ( o-phe ) . Incubation of BjV with AEBSF inhibited the amidolytic activity of SVSP by 93% , whereas neither Na2-EDTA nor o-phe importantly blocked SVSP ( Table 1 ) . We also examined the efficiency of inhibitors in blocking the coagulant activity of BjV in rabbit plasma , which is almost exclusively dependent on procoagulant activators of BjV [24] . In rabbit plasma , the MCD of BjV incubated with saline was 3 . 6±1 . 4 µg/mL , and Na2-EDTA reduced this activity by ca . 70-fold ( MCD = 262 . 4±1 . 8 µg/mL ) . Incubation with o-phe also diminished the clotting activity of BjV by approximately 40-fold ( MCD = 336 . 7±2 . 0 µg/mL vs . MCD = 7 . 9±1 . 3 µg/mL for BjV incubated with ethanol , the vehicle for o-phe ) . In contrast , AEBSF did not inhibit the coagulant activity of BjV ( MCD = 4 . 0±1 . 4 µg/mL ) . Altogether , these results demonstrated that preincubation of BjV with either Na2-EDTA or o-phe inhibited SVMP , but not SVSP activity , whereas AEBSF markedly inhibited only SVSP activity . In preliminary experiments , we also evaluated whether o-phe and Na2-EDTA produced similar in vivo results , in order to choose one SVMP inhibitor for subsequent experiments . The results obtained for platelet count and fibrinogen assay at 3 h showed that both Na2-EDTA and o-phe provided similar results and experimental profiles ( Figure S1 ) . Based on these results , Na2-EDTA was preferred since it dissolved in aqueous solution , and no additional group was required for vehicle controls . In snakebites , venom is usually injected into victims via s . c . or i . m . routes . In order to assess whether local hemorrhage and an inflammatory reaction could modify the systemic hemostatic manifestations evoked by BjV , the i . v . and s . c . routes were used to compare hemostatic parameters in the acute phase of envenomation ( 3 and 6 h ) . Intravenous injection of 100 µg/animal defibrinogenated 100% of rats , and caused no clinical manifestations . After previous experiments , we elected the s . c . dose of 1 . 6 mg/kg b . w . , since hemostatic disturbances at 3 and 6 h were similar to those observed in patients on admission to hospital , and animals behaved normally . Using this dose , fibrinogen levels and platelet counts were progressively restored after 8 h , so that at 24 h they were hemostatically recovered ( mean platelet counts are higher than 600×109/L and mean fibrinogen levels are higher than 100 mg/dL , data not shown ) . Initially , circulating BjV levels were measured to test whether preincubation of BjV with Na2-EDTA or AEBSF modified the absorption of BjV from tissues into bloodstream , and could thereby interfere with subsequent analyses . Rats injected with BjV , regardless of the treatment or route used , exhibited statistically significant increases in venom levels compared with the saline group at 3 and 6 h ( Fig . 1a ) . Although some fluctuation was noticed in venom levels , no statistically significant difference was observed between the results of the Na2-EDTA- or AEBSF-treated groups compared with saline-treated BjV group ( p = 0 . 897 ) , both for 3 and 6 h . These results confirmed that neither Na2-EDTA nor AEBSF prevented BjV from entering the bloodstream , nor altered the levels of circulating BjV . When animals received BjV i . v . , circulating levels cleared more rapidly at 6 h ( p = 0 . 017 ) . As expected , subcutaneous injection of BjV into rats resulted in local hemorrhage ( Fig . 1b ) . Preincubation of BjV with AEBSF decreased the extent of local hemorrhage by approximately 30% . Na2-EDTA markedly reduced ( around 85% ) local hemorrhage at 3 and 6 h compared with saline-treated BjV ( p<0 . 001 ) . Furthermore , BjV induced a remarkable drop in plasma fibrinogen levels and a simultaneous sudden increase in FDP/fdp levels at 3 and 6 h ( p<0 . 001 ) , independently of the route of administration ( Fig . 1c , d ) . Preincubation with Na2-EDTA inhibited fibrinogen consumption at 3 h ( p<0 . 001 for s . c , and p<0 . 004 for i . v . route ) , and more markedly at 6 h ( p<0 . 001 for both routes ) . In addition , Na2-EDTA-treated BjV was less effective in reducing fibrinogen consumption at 3 h when given i . v . compared to s . c . However , AEBSF-treated BjV evoked the same extent of fibrinogen consumption as that of saline-treated BjV both at 3 and 6 h ( Fig . 1c ) . Likewise , the rise in FDP/fdp levels was promptly reversed by preincubation of BjV with Na2-EDTA , but not by AEBSF ( Fig . 1d ) . Together , these results show that the preincubation of BjV with Na2-EDTA drastically inhibited local hemorrhage , fibrinogen consumption and fibrinolysis activation , suggesting that SVMP have an essential role in inducing coagulopathy during envenomation . Independently of the route of venom administration , platelet counts decreased markedly ( around 80–90% ) in rats receiving BjV at 3 and 6 h ( p<0 . 001 ) ( Fig . 2 ) . In addition , regardless of the preincubation used , platelet counts at 6 h were lower when BjV was given s . c . than when given i . v . , perhaps because of the lower dose of venom injected and therefore a faster clearance of circulating venom . Neither AEBSF nor Na2-EDTA conspicuously attenuated the drop in platelet count , although platelet counts tended to be somewhat higher with Na2-EDTA treatment for the i . v . group at 3 h . These data demonstrate that neither SVMP nor SVSP were involved in venom-induced thrombocytopenia . Initially , PT was used to check activation of the extrinsic pathway of coagulation . As expected , PT was markedly prolonged in rats injected with saline-treated BjV at 3 h , and this increase in PT tended to subside at 6 h , for both routes of venom administration . Na2-EDTA abolished this increase ( Fig . 3a ) , except at 6 h for the i . v . route , because PT had already returned to basal levels in all groups; in contrast , AEBSF had no important effect on PT . FVII levels ( Fig . 3b ) were not reduced during envenomation , except at 6 h in rats treated with Na2-EDTA or AEBSF by the i . v . route; minor increases were noticed at 6 h in rats injected s . c . with BjV , regardless of the preincubation , and there was a major increase at 3 h in those receiving Na2-EDTA-treated BjV i . v . These data show that FVII consumption is apparently not an important event during envenomation , and that hypofibrinogenemia may be the primary cause of prolongation of PT . However , since FVII plasma levels are elevated during stressful conditions [32] , consumption might be masked by a simultaneous increase in the synthesis of this factor . Thus , to investigate whether the inoculation of BjV into rats induced a rise in plasma TF levels , a TF activity assay was employed . As shown in Figure 3c , rats injected with BjV showed a marked increase in plasma TF levels , in comparison with saline-injected animals ( p = 0 . 01 ) . Interestingly , Na2-EDTA , but not AEBSF , mitigated this increase ( p<0 . 05 ) . The i . v . and s . c . administration of BjV showed increased levels of TF activity in plasma , demonstrating that the local reaction , induced by s . c . injection , did not completely account for the rise in TF activity in plasma . To investigate where TF was being expressed , we analyzed samples from skin and lungs . After s . c . injection , high levels of TF activity were noticed in skin ( p = 0 . 048 ) and lung ( p = 0 . 015 ) at 6 h ( Fig . 4a ) . On the other hand , i . v . showed a trend for high levels in lung , but no statistically significant difference was observed . Thus , to understand whether elevated TF activity resulted from an increased protein expression in lung and skin tissues , semiquantitative western blotting was used to evaluate TF and PDI protein expression . Protein bands of 47 and 57 kDa , corresponding to TF and PDI , respectively , were observed in skin samples . In lung tissue , no bands were noticed , probably because the amount of PDI and TF proteins in the samples was below the detection limit of western blotting ( data not shown ) , and therefore only skin samples were analyzed further . Figure 4b , d depicts a statistically significant fall in TF protein expression in skin at 3 h , independently of the treatment used for BjV , in comparison with normal tissue ( p<0 . 001 ) . Inversely , TF expression was augmented ( p = 0 . 014 ) at 6 h in animals that received BjV in comparison with those that received saline; however , no statistically significant difference was noticed in TF expression among groups that received BjV pretreatments . On the other hand , PDI expression ( Fig . 4c , d ) was constant at 3 h , but a remarkable drop was noticed in all groups that received BjV at 6 h ( p<0 . 001 ) . GAPDH protein expression ( Fig . 4d ) was also used as a control , and no statistically significant difference was noted among groups at 3 or 6 h ( data not shown ) .
BjV is a rich source of proteins and enzymes that destabilize hemostasis . We reasoned that exposure , expression and/or release of TF induced by BjV might occur at the site of venom inoculation or systemically , in virtue of distant tissue damage evoked by venom toxins . To date , augmented TF expression has never been demonstrated in snakebites , although it has been claimed , without scientific evidence , that it does not occur [33] . We verified for the first time that a marked increase in TF levels occurs in plasma and tissues of envenomed rats . This evidence suggests that coagulopathy is not only due to the direct activity of snake venom toxins on coagulation factors , as demonstrated elsewhere [13] , but also by augmented TF expression and release . Irrespective of the location where TF is released , our results demonstrate that TF expression/activity is increased during envenomation , and may trigger the blood coagulation cascade following snakebite . However , the intensity and relevance of this finding to the overall picture of hemostatic dysfunction requires further investigation . In lieu of employing purified proteins [34] , we elected to use BjV , which contains a wide variety of toxins that act interactively , to evaluate whether SVMP and SVSP were important to induce hemostatic disturbances . Surprisingly , SVSP had no major role in B . jararaca-induced hemostatic disturbances . Only preincubation of venom with Na2-EDTA was able to substantially inhibit fibrinogen consumption , PT prolongation , FDP/fdp generation , local hemorrhage , and the increase in TF levels . Results obtained for the incubation of BjV simultaneously with Na2-EDTA and AEBSF were not different from those reported for Na2-EDTA alone ( data not shown ) . Local hemorrhage , which is usually attributed to the activity of SVMP [35] , is frequently observed in envenomed patients , and , as expected , was abrogated by treatment of BjV with Na2-EDTA , as reported previously in mice [36] . On the other hand , Na2-EDTA minimally blocked BjV-induced thrombocytopenia . Using batismastat , clodronate , and doxycycline to inhibit SVMP from Bothrops asper venom , similar conclusions were reached about the pivotal role of SVMP in venom-induced coagulopathy and hemorrhage , and their lack of involvement in thrombocytopenia in mice [34] , [37] , [38] . Fibrinogen consumption has been hypothesized to be the consequence of the direct defibrinogenating activity of thrombin-like enzymes , and/or of generation of intravascular thrombin promoted by prothrombin and factor X activators found in BjV [24] . Our findings show that SVMP play an essential role in inducing fibrinogen consumption in rats , and that the direct action of thrombin-like enzymes ( SVSP ) on fibrinogen contributes minimally to defibrinogenation after s . c . injection . In Brazil , most physicians and toxinologists immediately associate blood incoagulability observed in either humans or animals bitten by Bothrops snakes with the action of thrombin-like enzymes . This assumption is based on early reports , particularly from those that described the coagulant activity of Bothrops venoms [39]–[42] or that isolated thrombin-like enzymes [43]–[47] . Among the broad variety of coagulant enzymes found in snake venoms , various investigations have focused their research on isolating and characterizing thrombin-like enzymes , given the simplicity of isolating fibrinogen , the most abundant coagulation factor found in plasma . However , Eagle [5] had already observed that BjV , not only contained enzymes that clotted fibrinogen , whose action was similar to thrombin , but also prothrombin-activating enzymes . On the other hand , Gastão Rosenfeld , an eminent physician who worked at Hospital Vital Brazil , in Institute Butantan , characterized the coagulant and hemolytic activity of animal venoms in Brazil [41] . His teachings and paradigms still reverberate in Brazil , and physicians and scientists continue to believe that thrombin-like enzymes are the main enzymes involved in defibrinogenation in bites inflicted by Bothrops snakes . According to Rosenfeld et al . [48 , page 244] , “Among bothropic venom , B . jararaca and B . atrox were more extensively studied than others . No disagreement exits with respect to their thrombinlike activity ( Janszky , 1950 , 1956; G . Rosenfeld et al . , 1959; Nahas et al . , 1964 ) , which is responsible for the defibrination syndrome in the clinic . As a consequence of fibrinogen depletion , blood remains incoagulable” . This conclusion has prevailed henceforth , and no investigation has rigorously examined it . However , other signs apparentely strengthened this paradigm . For example , the cause of blood incoagulability in patients bitten by South American rattlesnakes ( Crotalus durissus spp ) , whose venom contains exclusively thrombin-like enzymes [49] , resembled the defibrinogenation evoked by Bothrops snakes . As a result , thrombin-like enzymes were erroneously considered as the main toxins that elicit fibrinogen consumption in the latter envenomation . Consolidation of this paradigm has been reinforced by the use of thrombin-like enzymes for the treatment of thromboembolic diseases [50] , [51] , which promote safe defibrinogenation , similar to that observed in most snakebites . However , our results do not agree with the traditional assumption that thrombin-like enzymes are the main toxins accounting for defibrinogenation , and indicate that SVMP are crucial enzymes promoting fibrinogen consumption , at least in rabbits [7] , [24] , mice [38] , and rats . However , which class of toxins accounts for most fibrinogen consumption in humans remains to be investigated , but we demonstrate herein that procoagulant enzymes and high TF plasma levels may exert a relevant role in the hemostatic disorder evoked by Bothrops bites , since intravascular thrombin generation , evidenced by raised plasma levels of TAT complex [52] , [53] , are noticed in patients . In Bothrops sp . venoms , few isolated SVMP and SVSP have been reported to interfere with platelet function and/or cause thrombocytopenia [34] , [54]–[60] . SVMP inhibition did not protect rats from the drop in platelet count observed during B . jararaca envenomation . Likewise , treatment of B . asper or Bothrops caribbaeus venom with SVMP inhibitors could not block thrombocytopenia [34] , [61] . Thus , this evidence suggests that SVMP have a minor role in directly inducing thrombocytopenia in rats , and that other pathophysiological mechanisms are involved in thrombocytopenia in B . jararaca envenomation . Interestingly , our findings indicate that SVSP reported to activate platelets ex vivo [57] did not seem to be important in vivo . The C-type lectin aspercetin , similar to botrocetin found in B . jararaca venom [62] , is apparently crucial to induce thrombocytopenia in mice injected i . v . with B . asper venom [34] . Whether botrocetin or other C-type lectins found in BjV account for thrombocytopenia in vivo is a matter for future investigation . Laboratory data from victims of B . jararaca snakebites show no correlation between fibrinogen consumption and thrombocytopenia [52] . Our findings corroborate such clinical observations , and do indicate that diverse and complex pathophysiological mechanisms are involved in this process . In line with this assumption , variation in the composition of toxins during the ontogenetic development of B . jararaca snakes [20] , [63] may explain why patients bitten by young snakes have a higher incidence of blood incoagulability and a trend to higher platelet counts on admission in hospital , compared to those bitten by adult snakes , who have a more accentuated fall in platelet counts and a lower frequency of blood incoagulability [2] , [64] . These findings demonstrate that there is no single mechanism or main toxin that may explain all events occurring in envenomation by B . jararaca . Local injury was reported to play a prominent role in sequestering platelets after s . c . or i . m . venom inoculation [65] . Two routes were used here to inoculate BjV , so that we could study the contribution of the local lesion to systemic hemostatic disturbances . BjV induced intense proteolytic activity and inflammatory reaction at the site of venom inoculation , demonstrated by the presence of intense local hemorrhage . However , animals injected with Na2-EDTA-treated BjV showed minimal local injury , but still demonstrated high platelet consumption , indicating that the local lesion minimally contributes to the sequestration of platelets from the circulation . Interestingly , neither Na2-EDTA nor AEBSF interfered with the kinetics or levels of BjV in circulation , although we have evaluated venenemia in only two time intervals . Anai et al . [66] reported that preincubation of B . jararaca venom with polyclonal antibodies anti-jararafibrase I ( i . e . , jararhagin [67] ) neutralized the hemorrhagic activity of crude B . jararaca venom , and prevented the development of hemostatic disturbances in vivo , demonstrating that hemorrhagic SVMP facilitate diffusion and absorption of coagulant toxins into the circulation . However , the mechanisms of inhibition of SVMP by antibodies and Na2-EDTA are different , since the latter directly inhibits the catalytic activity of SVMP , whereas the former bind to diverse epitopes in SVMP in order to inhibit their biological activity . Thus , our results suggest that other toxins . such as hyaluronidases [68] , may facilitate venom diffusion/absorption . TF has been reported to be involved in inflammation and thrombosis [69] , and several mediators , including proinflammatory cytokines and thrombin , induce TF expression [15] . In view of the intense inflammatory activity evoked by SVMP at the site of venom inoculation , tissue injury , cell necrosis/apoptosis , and/or the release of proinflammatory cytokines [70] may have accounted for the raised expression of TF in skin and lungs . Interestingly , berythractivase , a SVMP isolated from Bothrops erythromelas venom , but not jararhagin , from BjV , has been demonstrated to render endothelial cells highly thrombogenic in vitro , due to upregulation of TF activity and expression [71] . TF levels were raised in plasma as soon as 3 h after BjV injection , but statistically significant increases in skin and lung TF expression and activity were noticed exclusively at 6 h . These results suggest that activation of circulating cells , such as monocytes and platelets ( two cell types known to express TF when activated [15] ) by BjV might also be involved in the increase in plasma TF levels . Sustained raised levels of TF in plasma at 6 h were only noticed when BjV was administered s . c . , suggesting that this route provides additional stimulus for TF production/decryption . It is difficult to explain the significant decrease in TF protein expression in skin at 3 h , but one plausible explanation could be that BjV hydrolyses TF , although this possibility was not tested here . In skin , TF protein expression was markedly elevated at 6 h , and PDI was simultaneously diminished; however , the mechanism responsible for these findings remains to be demonstrated . Although the role of PDI in TF decryption is questionable , in models of thrombosis PDI accumulates at the site of vascular injury [16] , [72] . On the other hand , inhibition of PDI at the endothelial cell surface enhances TF pro-coagulating activity by affecting phosphatidylserine exposure [73] . Thus , the decrease in protein expression in PDI at 6 h may intensify the hemostatic disturbances during envenomation . Given the observed increase in TF protein expression in skin , lung and plasma , a drop in plasma FVII was expected . However , normal or increased FVII levels were observed here . In fact , FVII has the shortest mean-life ( 4–7 h ) of blood coagulation factors in the circulation , and the lack of diminished FVII levels may be ascribed to an increased hepatic synthesis that may occur in stressful situations [32] . A steady decrease in FVII levels has been reported for one patient bitten by Bothrops neuwiedi [74] , although most snakebites in humans do not induce a marked FVII consumption [75]–[78] . As shown elsewhere [42] , the coagulant activity of BjV does not depend on FVII in vitro . Disseminated intravascular coagulation ( DIC ) is characterized by a TF-mediated coagulation activation induced by cytokines , depletion of natural anticoagulants and PAI-1-mediated fibrinolysis inhibition [79] . In virtue of elevated FDP/fdp levels , thrombocytopenia , prolonged prothrombin time and fibrinogen consumption , which were also observed herein , snake envenomation has been associated with or may evolve to DIC [80] . Since high plasma TF levels have been described in patients with DIC [81] , our results suggest that raised TF levels may have an important role in activating the blood coagulation cascade , especially in more severely envenomed patients , in which the local lesion is more extensive [2] . On the other hand , since hemostatic disturbances are rapidly recovered after antivenom therapy in mildly or moderately envenomed patients [2] , the involvement of plasma TF in bleeding manifestations observed therein awaits clinical evaluation . In conclusion , we show that SVMP are crucially involved in the coagulopathy evoked by BjV in rats . Since BjV-induced thrombocytopenia was not mitigated by any inhibitor , our findings demonstrate that SVMP and SVSP are not directly associated with this phenomenon , and that other mechanism ( s ) or BjV toxins are involved . Our findings also indicate that TF is an additional component that should be considered when discussing snake venom-induced hemostatic disturbances . Moreover , the evidence of increased TF levels in plasma and skin is extremely important in dealing with hemostatic disturbances in severely envenomed patients bitten by B . jararaca snakes , since it approximates the so-called DIC-like syndrome , which occurs in snake envenomation , to the true DIC syndrome , initiated by increased TF expression . Our results suggest that plasma TF levels are likely to be elevated in patients bitten by poisonous snakes , and that their levels may be correlated with the frequency and intensity of hemostatic disturbances . Therapeutic interventions in this pathway should be tested as an ancillary treatment to antivenom therapy for allowing a prompt interruption to the development of true DIC syndrome . | Although the abundance of reports about hemostatic disturbances in snakebites , few studies have addressed how crude snake venoms evoke blood coagulation disturbances in vivo . Snake venoms contain several components that disturb hemostasis , and the prevailing model claims that coagulation disturbances observed in patients are triggered directly by those toxins . However , taking into account the physiological mechanisms that activate the coagulation cascade , tissue factor might also be generated and decrypted during snake envenomation . We investigated herein if tissue factor and protein disulfide isomerase , an enzyme that controls the encryption/decryption of tissue factor , were altered during experimental envenomation in rats . We observed increased activity/expression of tissue factor at the site of venom injection , as well as in lungs , and decreased expression of protein disulfide isomerase at the site of venom injection . Moreover , tissue factor levels were raised in plasma , demonstrating thereby that this via may be crucial to activate blood coagulation in patients , especially in those more severely envenomed . We also noticed that snake venom metalloproteinases accounted for most fibrinogen consumption . Our results clarify the mechanisms that activate blood coagulation during envenomation , evidencing that true intravascular coagulation syndrome , due to increased tissue factor expression , might occur during snake envenomation in human beings . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biochemistry",
"coagulation",
"disorders",
"medicine",
"and",
"health",
"sciences",
"blood",
"coagulation",
"body",
"fluids",
"blood",
"anatomy",
"coagulation",
"factors",
"proteins",
"plasma",
"proteins",
"thrombocytopenia",
"platelets",
"biology",
"and",
"life",
"sci... | 2014 | Bothrops jararaca Venom Metalloproteinases Are Essential for Coagulopathy and Increase Plasma Tissue Factor Levels during Envenomation |
Metabolomics studies use quantitative analyses of metabolites from body fluids or tissues in order to investigate a sequence of cellular processes and biological systems in response to genetic and environmental influences . This promises an immense potential for a better understanding of the pathogenesis of complex diseases . Most conventional metabolomics analysis methods exam one metabolite at a time and may overlook the synergistic effect of combining multiple metabolites . In this article , we proposed a new bioinformatics framework that infers the non-linear synergy among multiple metabolites using a symbolic model and subsequently , identify key metabolites using network analysis . Such a symbolic model is able to represent a complex non-linear relationship among a set of metabolites associated with osteoarthritis ( OA ) and is automatically learned using an evolutionary algorithm . Applied to the Newfoundland Osteoarthritis Study ( NFOAS ) dataset , our methodology was able to identify nine key metabolites including some known osteoarthritis-associated metabolites and some novel metabolic markers that have never been reported before . The results demonstrate the effectiveness of our methodology and more importantly , with further investigations , propose new hypotheses that can help better understand the OA disease .
Systems biology is an emerging research direction that takes a holistic approach to modeling complex biological systems [1–3] . It requires multidisciplinary efforts from research fields including biomedicine , statistics , and computer science . Systems biology approaches embrace the complexity of biological systems and focus on modeling the interactions among multiple components in biological systems including genome , transcriptome , proteome , and metabolome [4–8] . By integrating a variety of omics data , systems biology for human disease studies aims at better understanding the etiology of common diseases , discovering biomarkers that can help predict early disease onset , progression and severity , and identifying new drug targets [9 , 10] . Integrative data analysis and mining for systems biology often include hundreds to thousands of variables such as genes , proteins , and metabolites [11] . Most conventional tools adopt a univariate analysis strategy and may overlook the intertwined relationships among multiple variables . However , the high dimensionality has imposed a computational challenge for multivariate analyses since searching combinations of variables becomes prohibitive as the search space grows exponentially with the number of variables . It has brought about the realization that the development and application of powerful informatics and data mining methodologies for systems biology are critical and hold great potentials for the next generation of biomedical research [12 , 13] . Machine learning and heuristic search algorithms , including principal component analysis [14] , artificial neural networks [15] , and random forest [16] , have seen increasing and successful applications in omics data mining for biomarker discovery . However , such interdisciplinary research direction is still in a preliminary stage , and more learning and modeling algorithms are yet to be explored and developed in future investigations . In this article , we developed a new bioinformatics framework for high-dimensionality omics data analysis where we used an evolutionary learning algorithm to discover key metabolites and their combinations for an osteoarthritis ( OA ) metabolomics study . The non-linear synergistic effects of combining multiple metabolites were inferred using symbolic models that were trained through improving classification accuracies to predict the disease status . The key individual and combinations of synergistic metabolites are further visualized and analyzed using networks . Evolutionary algorithms define a collection of meta-heuristic optimization and modeling algorithms inspired by natural evolution [17–20] . An evolutionary algorithm maintains a population of diverse candidate solutions , which are compared with the desired outcome . Then , through multiple generations of variation , selection , and reproduction , the population adapts to the selection criterion ( the relative distance from the desired outcome ) , and produces fitter solutions . Evolutionary algorithms are highly robust and powerful in tackling imprecise and incomplete problems , thanks to their automated search mechanisms . They are also extremely parallelizable , due to their distinguishing feature of population-based search , which also allows them to scale to solve large and complex problems . Evolutionary algorithms have been successfully applied to modeling problems , where they can automatically derive a symbolic model of an aggregation of interrelated attributes through an evolutionary training process . OA is the most common form of arthritis . It causes substantial morbidity and disability in the elderly populations and imposes a great economic burden on our society [21 , 22] . Despite a high prevalence and societal impact , there is no medication that can cure it , or reverse or halt the disease progression , partly because its pathogenesis is still unclear and there is no reliable method that can be used for early OA diagnosis . Recent developments in the field of metabolomics provide an array of new tools for the study of OA . Metabolites are intermediate and end products of various cellular processes and their levels of concentration serve as a good indicator of a sequence of biological systems in response to genetic and environmental influences . A large number of small-molecule metabolites from body fluids or tissues can be quantitatively detected simultaneously , which promises an immense potential for early diagnosis , therapy monitoring and understanding the pathogenesis of complex diseases [23–25] . Our evolutionary algorithm and network analysis were able to identify nine key metabolites that appear most frequently in the best evolved models for predicting the disease outcome , four of which also serve as hubs and bottlenecks in the metabolite synergy network . Some of the nine metabolites were previously found highly associated with OA , and the rest are novel findings that could be very useful proposing new hypothesis to better understand the disease .
The study protocol was approved by the Health Research Ethics Authority ( HREA ) of the province of Newfoundland and Labrador , Canada , with reference number 11 . 311 and a written consent was obtained from all the participants . In the OA dataset used for the current study , knee OA patients were selected from the Newfoundland Osteoarthritis Study ( NFOAS ) initiated in 2011 [26] . The NFOAS aimed at identifying novel genetic , epigenetic , and biochemical markers for OA . The NFOAS recruited OA patients who underwent a total knee replacement surgery due to primary OA between November 2011 and December 2013 at the St . Clare’s Mercy Hospital and Health Science Centre General Hospital in St . John’s , the capital city of Newfoundland and Labrador ( NL ) , Canada . Healthy controls were selected from the CODING study ( The Complex Diseases in the Newfoundland population: Environment and Genetics ) , where participants were adult volunteers [27] . Both cases and controls were from the same source population of Newfoundland and Labrador . Knee OA diagnosis was made based on the American College of Rheumatology clinical criteria for the classification of idiopathic OA of the knee [28] and the judgment of the attending orthopedic surgeons . Controls were individuals without self-reported family doctor diagnosed knee OA based on their medical information collected by a self-administered questionnaire . We collected 153 OA cases and 236 healthy controls . Blood samples were collected after at least 8 hours of fasting and plasma was separated from blood using the standard protocol . Metabolic profiling was performed on plasma using the Waters XEVO TQ MS system ( Waters Limited , Mississauga , Ontario , Canada ) coupled with Biocrates AbsoluteIDQ p180 kit , which measures 186 metabolites including 90 glycerophospholipids , 40 acylcarnitines ( 1 free carnitine ) , 21 amino acids , 19 biogenic amines , 15 sphingolipids and 1 hexose ( above 90 percent is glucose ) . The details of the 186 metabolites and the metabolic profiling method were described in our previous publication [29] . Over 90% of the metabolites ( 167/186 ) were successfully determined in each sample . Prior to performing the informatics analyses , several steps of preprocessing were applied to the dataset . Batch correction was performed by multiplying each metabolite concentration value by the ratio of the overall mean and the batch mean for that metabolite . Then , covariate adjustment was performed to remove the variation due to individual’s age , gender , and body mass index ( BMI ) . The samples were randomly assigned to either a discovery or replication dataset , such that cases and controls were divided evenly between the two datasets . Finally , each metabolite concentration value was normalized to zero mean and unit variance across the population . In this study , the algorithm used to model the non-linear synergy among multiple metabolites associated with OA is a branch of evolutionary computation , termed genetic programming [30] . A population of diverse candidate prediction models is generated randomly in the step of initialization and will evolve to improve prediction accuracy gradually through a number of generations . After evolution halts , the best model of the population in the final generation will be the output . Each candidate prediction model takes the form of a symbolic computer program comprised of a set of sequential instructions . An instruction can be an assignment statement or a conditional statement . The conditional if instructions affect the program flow such that the instruction immediately following the if instruction is not executed if the condition is false . In the case of nested if instructions , each of the successive conditions needs to be true in order for the instruction following the chain of if instructions to be executed . A register r stores the value of a feature , a calculation variable , or a constant . A feature can be a predictor or an attribute used to make a prediction of the outcome . In the context of the current study , features are concentration levels of metabolites in the samples . A calculation variable serves as a temporary buffer that enhances the computation capacity . In an assignment instruction , only registers storing calculation variables can serve as the return on the left side of the assignment symbol “=” , but any register can serve as an operand on the right-hand side . This is to prevent overwriting the feature values . When a prediction model is evaluated on a given sample , feature registers take all the values of the sample , and the set of instructions are executed sequentially . The sigmoid transformation of the final value stored in the designated calculation register r[0] is used to predict the outcome of the sample , i . e . , if S ( r[0] ) is greater than or equal to 0 . 5 , the sample is predicted as diseased ( class one ) , otherwise the sample is predicted as healthy ( class zero ) . An example of classification model with eight instructions is given below . Here , the output register r[0] and calculation registers r[4] and r[5] are all initialized with ones . Feature registers r[1-3] take input values from three metabolite concentration levels m[1-3] respectively . For instance , when a sample with m[1-3] values as {0 . 2 , 0 . 01 , 0 . 085} is input to this classification model , the conditional statement r[1]>r[3] in instruction I1 becomes true , so in instruction I2 , r[0] changes its value to 0 . 51 . The rest of the instructions are executed sequentially , and the final value of r[0] is set to 1 . 0039 . Its sigmoid transformation S ( 1 . 0039 ) is greater than 0 . 5 , so this sample will be classified by this model as class one , i . e . , diseased . I1: if r[1]> r[3] I2: then r[0] = r[2] + 0 . 5 I3: r[4] = r[2] / r[0] I4: if r[0] > 4 I5: then if r[3] < 10 I6: then r[5] = r[3]—r[4] I7: r[4] = r[4] * r[1] I8: r[0] = r[5] + r[4] At the initial generation , a population of diverse classification models was generated randomly . The fitness of each model was evaluated using mean classification error ( MCE ) , computed as the average number of incorrectly classified training samples . A set of models were chosen as parents based on their fitness , and variation operators , including mutation and recombination , were applied to them . A mutation alters an element of a randomly picked instruction , i . e . , replacing a return or an operand register by a randomly generated one or replacing the operator . Recombination swaps segments of instructions of two parent models . Survival selection picks fitter models to form the population for the next generation . Such an evolution process iterates for a certain number of generations , and the model with the lowest MCE at the end is output as the final best model of a run . This evolutionary modeling algorithm was implemented using the Julia programming language [31] . The main parameters used in the implementation are shown in Table 1 . A five-fold cross-validation was used to prevent overfitting so that each run of the algorithm produced five best classification models as its output . For the first round of analysis , the evolutionary learning algorithm was run on the discovery dataset using 200 distinct seed values for the random number generator . As a result of the cross validation , our implementation gave five different best classification models for each seed value , resulting in a total of 1000 best classification models . We investigated the resulting classification models by calculating various statistics of the fitness ( MCE ) values , sensitivity , specificity and area under the curve ( AUC ) as computed on the testing fold for each run . In addition , we inspected the models by counting how often each of the 167 metabolites appeared as predictive variables in the set of 1000 best models . In addition to looking at the individual occurrence of single metabolites in the best classification models , the co-occurrence of metabolites in the models was studied by counting the number of times each metabolite pair appeared together in the same model . The top 1% of the resulting metabolite pairs , ranked by decreasing frequency , were used to construct a metabolite synergy network . Network science has seen increasing applications in biomedical research [32–34] , where biological entities are represented as vertices and their relationships can be modeled using edges linking pairs of vertices . Network modeling is a powerful tool to study interconnections among a large number of biological entities . In this study , vertices represent metabolites and an edge links two metabolites if they have a co-occurrence frequency in the set of 1000 best prediction models greater than the given cutoff . The network was rendered and analyzed using the Cytoscape software [35] . For the second round of a more focused analysis , only the subset of metabolites appearing in the metabolite synergy network was used as a restricted feature set in a repeated model learning implementation , allowing the evolutionary algorithm to only use these more important metabolites to construct the classification models . The analysis was performed on both the discovery and replication datasets , each resulting in another set of 1000 best classification models . The intersection of the top 20 most common metabolites from the discovery and replication runs was reported , and such metabolites are regarded interacting metabolites with high potential associations to the disease of OA . To evaluate the classification power of the best models found using our evolutionary algorithm , we trained logistic regression on both the discovery and replication datasets using the reduced feature set and compared its classification performance with our evolutionary algorithm .
In the first round of analysis on the discovery dataset , the full feature set of 167 metabolites was used for the evolutionary algorithm to search for the best classification models . The classification performance of those best models was then evaluated using testing samples . Statistics for the results are shown in Table 2 . It can be seen that the best results , among them the lowest MCE and the highest AUC value , suggest that some of the classification models found with the evolutionary algorithm during this first full feature scenario already achieve a reasonably high prediction accuracy . Although a feature register can be any of the full set of 167 metabolites in the NFOAS dataset , the final best models usually only contain a subset of those features , given the nature of the evolutionary algorithm . Moreover , recall that since the final value stored in the designated output register is used to compute the classification outcome , some instructions can be redundant and not have any effect on the output value . Those input features that appear in effective instructions , that is , instructions which do contribute to the outcome , are considered effective features . We looked into the number of effective features in the 1000 best models . A visualization of the distributions for fitness ( MCE ) and the number of effective features can be seen in Fig 1 . Most best models have their MCE values between 0 . 3 and 0 . 5 , while some runs can yield models with a classification error less than 0 . 1 . Most best models include around 20 to 40 effective features , about 10% to 20% of the total feature set . In addition , there turns out to be no correlation between the fitness and number of effective features of the classification models , with Pearson’s correlation coefficient being 0 . 044 and the associated p-value 0 . 16 ( S1 Fig ) . The top 20 metabolites and metabolite pairs most commonly contained in the 1000 best models are shown in Fig 2 . The top two individual metabolites taurine and arginine appear in about 30% of the models , with threonine and ornithine also being found in over 25% of the models . In addition to the highest individual appearance , taurine pairs up with threonine in about 10% of the best models , and with ornithine or arginine in about 9% of the best models . The top 1% metabolite pairs out of all ( 1672 ) possible combinations were used to construct the network of Fig 3 . Here each vertex is a metabolite and an edge links two metabolites if they have a high pairwise occurrence ( top 1% ) in the best models . The vertex size denotes the frequency of the individual metabolite’s occurrence in the best models , while the edge width shows the frequency of the occurrence of the corresponding metabolite pair . The network has 70 metabolites and 156 edges . There is one connected component and each vertex has an average of 4 . 5 connected neighbors . The vertex degree of the network follows a heavy-tail distribution ( Fig 3 inset ) . The metabolites that are individually most common in the best models also make up the vertices with the highest degree in the network , due to the methodology being used . Taurine has the highest degree of 44 , followed by arginine with a degree of 43 . Threonine and ornithine have degrees of 27 and 19 respectively . These four metabolites are also connected to each other in the network , forming a dense core of the network . In contrast , most peripheral vertices have degrees less than four . Those four vertices not only have the highest degrees but also the highest closeness and betweenness centralities , i . e . , they serve as essential hubs and bottlenecks of the network . The 70 metabolites appearing in the network of Fig 3 , which make up 1% of the most common metabolite pairs in the models for the first round of analysis , were used as the feature set for repeated evolutionary algorithm runs on both of the discovery and replication datasets . The statistics for the discovery dataset run are shown in Table 3 , and for the replication dataset in Table 4 . Comparing these statistics to those for the full feature set runs in Table 2 , it can be seen that the AUC value for the reduced feature set is higher on both the discovery and replication datasets . Thus the corresponding models achieve better performance in predicting the presence of OA based on the metabolite data . The 20 most common metabolites and metabolite pairs for the discovery and replication datasets , using the reduced feature set , are shown in Figs 4 and 5 respectively . The overlap between the 20 most common metabolites from the analysis using the full feature set is notable , with 14/20 and 10/20 metabolites being the same when comparing the full feature set runs to the reduced feature set runs on the discovery and replication datasets respectively . There are 9 metabolites that appear within the 20 most common ones on all three rounds of analyses: arginine , C16 , C18:1 , isoleucine , nitrotyrosine , ornithine , taurine , threonine , and tyrosine . This set also includes the four top hub and bottleneck metabolites in the previous network ( Fig 3 ) . The best model based on the AUC value was selected from the models found on each of the discovery and replication datasets . The ROC curves , as computed on the testing fold , for two sample best models are shown in Figs 6 and 7 . Both of these models achieved a perfect AUC value of 1 . Pseudocode representations of these two best models are shown in Listings 1 and 2 . Here each line contains one instruction , and the instructions are executed one after another like in any imperative language . The r[N] notation denotes calculation register at index N , and r[0] is treated as the output register . Calculation registers do not take input from feature values of training or testing data samples , but serve as buffers in the program to enhance its computational capacity . All calculation registers and the output register are initialized with ones at the start of the program’s implementation . After the program has been run on a data sample , the value contained in the register r[0] is converted to either zero or one , representing the prediction of healthy and diseased individuals respectively , by using the Sigmoid function and rounding to the nearest integer . This value is the classification prediction of the program . Listing 1 Pseudocode representation for the best model found on the reduced feature set ( discovery ) . if r[106] > Orn if PC aa C24:0 < r[65] r[68] = PC ae C40:5 + PC ae C44:5 r[108] = r[130] / Asn if PC ae C44:5 > SM ( OH ) C24:1 r[108] = lysoPC a C24:0 - r[18] if r[108] < r[133] r[98] = r[87] - Arg r[51] = PC ae C44:3 + Arg if Leu > Kynurenine r[68] = C5:1 ^ Taurine r[131] = Nitro-Tyr + r[6] r[125] = r[68] + r[51] if C5-DC ( C6-OH ) > r[33] r[98] = r[130] + C4 if PC ae C38:0 > r[131] r[98] = PC aa C32:1 * PC ae C40:1 if PC aa C34:3 < 4 . 0 r[0] = r[98] - r[125] Listing 2 Pseudocode representation for the best model found on the reduced feature set ( replication ) . r[15] = SM C16:1 ^ r[19] if PC aa C40:4 < r[15] if r[36] > C4 r[96] = r[125] * SM ( OH ) C24:1 r[59] = r[80] + Orn if r[61] < C4 r[59] = r[72] / r[127] if C5-DC ( C6-OH ) > Orn r[31] = r[59] - r[117] r[0] = r[31] - Arg r[59] = 10 . 687 / Nitro-Tyr if Ac-Orn > r[96] r[0] = r[59] - 10 . 595 To compare the classification power of the best models found using our evolutionary algorithm with a more widely used method , logistic regression was trained on the data . We used the logistic regression implementation from the scikit-learn Python library [36] . The same five-fold cross validation scheme and partitioning of data into the folds were used as with the evolutionary algorithm . ROC curves for the two best classification models found with logistic regression , one for each of the discovery and replication datasets , are shown in Figs 8 and 9 . As is apparent from the curves , the AUC values for these models are lower than of those evolved using the evolutionary algorithm .
The fast developing biomedical and computing technologies have brought research to a new era where multi-omics data are produced and mined in order to search for biomarkers of common human diseases . These omics data usually include hundreds to thousands of attributes and such high dimensionality has imposed a great computational challenge for bioinformatics studies . Most existing analyses look at one attribute at a time since the exponential increase of the possible combinations of attributes renders the exhaustive search prohibitive . Such a strategy , however , may overlook important interactions among multiple attributes with limited individual marginal effects . In this study , we developed an informatics framework that uses an evolutionary algorithm and network analysis to identify both the individual and combinations of metabolites associated with the risk of osteoarthritis ( OA ) . The evolutionary algorithm automatically searches for the models that can best predict the clinical outcome of OA using a feature set of metabolite concentration levels in healthy and diseased samples . Such an automatic learning algorithm performs feature selection systematically through the search for the best classification models and requires minimal prior assumptions on the models . The stochastic and population-based nature of the evolutionary algorithm produces a set of best models . Based on those best models , we constructed a metabolite network where vertices are high association metabolites and vertex sizes reflect their frequencies of occurrences in the best classification models . Edges link pairs of metabolites and their widths represent the co-occurrence frequencies of metabolite pairs . Such a network captures both the most important individual and combinations of metabolites associated with the disease . Moreover , it depicts the interconnected structure and patterns of multiple metabolites , and helps identify metabolic functions that may play a key role in explaining the OA disease . In the first round of analysis , the entire set of 167 metabolites in the metabolomics study was used for the evolutionary algorithm to search for the models that can best predict the disease outcome . The algorithm was run 200 times and a five-fold cross validation was used to avoid overfitting . Each run produced five best classification models on the five testing datasets , and thus a total number of 1000 best models were generated . The most frequent individual and pairs of metabolites that appear in those best models were reported ( Fig 2 ) . In the constructed metabolite network ( Fig 3 ) , four key metabolites taurine , arginine , threonine , and ornithine were identified as hubs and bottlenecks of the network , which indicates their important role in explaining the disease of OA . In the second round of analysis , we performed a more focused model search by using only the 70 metabolites included in the network as the reduced feature set . The evolutionary algorithm was executed again on the reduced feature set on both of the discovery and replication datasets . This round of analysis yielded improved classification models with higher prediction accuracies ( Tables 3 and 4 ) . Nine metabolites , arginine , C16 , C18:1 , isoleucine , nitrotyrosine , ornithine , taurine , threonine and tyrosine , were found most frequently appearing in the best models in the discovery dataset as the result of both rounds of analyses and were successfully replicated using the replication dataset , including the previous four key metabolites identified in the network . The results are interesting as arginine and its pathway related metabolites , such as ornithine , have been identified as being associated with OA in our previous analysis using traditional methods including pairwise comparison and regression technique [37] . Similarly , isoleucine was also previously identified as OA-associated metabolite [38] . The current analyses applied a novel analytic method , the evolutionary algorithm , which confirmed our previous findings and also identified additional novel metabolic markers for OA . These included four amino acids and two acylcarnitines , which could have potential utility in the clinical management of OA . For example , taurine is the most abundant free amino acid in humans , and may play an important role in inflammation associated with oxidative stress [39] . It has been reported to be associated with rheumatoid arthritis [40] . Nitrotyrosine is also associated with oxidative damage and has been found associated with aging and the development of OA in cartilage samples from both monkeys and humans [41] . The findings in the current study certainly warrant further investigation of the role of those novel metabolic markers in OA . Our bioinformatics framework , which uses an evolutionary algorithm and network analysis to identify key biomarkers for metabolomics studies , demonstrates the great potential of applying advanced computational techniques to biomedical data mining and model searching problems . Comparing to logistic regression , one of the most commonly used algorithms for such problems , our method was shown to be able to achieve better classification accuracies . Apart from most existing algorithms where metabolites are evaluated individually , our algorithm is able to examine which combinations of multiple metabolites can best predict the disease outcome . Some of the nine key metabolites reported in this study have very limited individual marginal effects , and could be overlooked using the traditional univariate analyses ( S2–S10 Figs ) . In addition , feature selection is embedded in our algorithm , so the search for the most relevant metabolites is systematically performed while evolving the best classification models . Classification models are represented as symbolic relationships between the metabolite concentration levels and the prediction outcome . Such a representation requires minimal prior assumptions on the models and can describe highly complex non-linear relationships . This can be even more important when multiple types of omics data are used in integrated analyses . Our methodology can be a very useful addition to the toolkit for bioinformatics research , and we expect to extend the applications of our methodology to a large range of data and problems in systems biology research . | Biomedical research has entered a new era where a large number of molecules and different components in biological systems can be quantitatively examined to investigate the causes of common human diseases . However , given the complexity of biological systems , those causes may not contribute to diseases individually but through interactions . The identification of those interactions , or the synergy of multiple factors , is a very challenging task due to the computational limitation , as well as the lack of effective methodologies for investigating multiple factors simultaneously . In this study , we proposed to model such an interaction effect through a self-learning algorithm using mechanisms inspired by natural evolution . Moreover , by constructing a synergy network using those evolved models , we were able to identify a set of interacting factors associated with a particular disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"rheumatology",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"applied",
"mathematics",
"metabolic",
"networks",
"organic",
"compounds",
"simulation",
"and",
"modeling",
"algorithms",
"evolutionary",
"computation",
"metabolomics",
"mathematics",
"metabol... | 2018 | An evolutionary learning and network approach to identifying key metabolites for osteoarthritis |
Rift Valley fever ( RVF ) is an emerging , vector-borne viral zoonosis that has significantly impacted public health , livestock health and production , and food security over the last three decades across large regions of the African continent and the Arabian Peninsula . The potential for expansion of RVF outbreaks within and beyond the range of previous occurrence is unknown . Despite many large national and international epidemics , the landscape epidemiology of RVF remains obscure , particularly with respect to the ecological roles of wildlife reservoirs and surface water features . The current investigation modeled RVF risk throughout Africa and the Arabian Peninsula as a function of a suite of biotic and abiotic landscape features using machine learning methods . Intermittent wetland , wild Bovidae species richness and sheep density were associated with increased landscape suitability to RVF outbreaks . These results suggest the role of wildlife hosts and distinct hydrogeographic landscapes in RVF virus circulation and subsequent outbreaks may be underestimated . These results await validation by studies employing a deeper , field-based interrogation of potential wildlife hosts within high risk taxa .
Rift Valley fever ( RVF ) is an emerging , vector-borne viral zoonosis that causes significant morbidity in humans and their livestock . The etiologic agent , Rift Valley fever virus ( RVFV ) , is a Phlebovirus in the Bunyaviridae family and is transmitted by several mosquito species that facilitate viral maintenance ( Aedes spp . ) or amplification ( Culex spp . ) [1 , 2] . Human infections are invariably asymptomatic or mild in early stages , however , severe cases can manifest as hemorrhagic fever or encephalitis [3 , 4] . Sheep , goats , and cattle experience fetal abortions as a result of RVFV infection , and the disease contributes to substantive economic losses to pastoralist communities during outbreaks [5 , 6] . Historically , most outbreaks in humans and domestic animals have occurred in the African continent , and in eastern Africa these typically follow periods of excessive rain in poorly draining arid or semi-arid landscapes [7] . The resultant flooding is conducive to the breeding and hatching of infected mosquitoes which transmit the virus to ruminant hosts followed by eventual secondary transmission to other ruminants and humans [5] . In more recent years , RVF has progressively expanded east into the Arabian Peninsula , with outbreaks in Saudi Arabia and Yemen [8] . The epidemiology and infection ecology of RVFV is complex and our knowledge of these incomplete . Several species in two distinct mosquito genera transmit infection . As primary vectors , Aedes mosquitoes maintain RVFV transovarially during dry periods; during these times there are little to no reported human or livestock infections . Aedes mosquito population explosion following wet periods leads to localized transmission to mammalian hosts [9 , 10] . Following this , Culex mosquitoes can expand ( amplify ) transmission to more dispersed livestock and human populations distant from the areas of local Aedes transmission [2 , 11 , 12] . Once RVFV becomes amplified in livestock , ongoing human infection occurs primarily through zoonotic transmission as a result of direct or indirect contact with animal tissues and body fluids , such as occurs during slaughtering or through performing obstetrical procedures on infected animals . Transmission from mosquitoes that feed on infected animals is also a viable though less important source of human infection [1 , 13] . While the role of vectors in RVFV infection ecology is well-established , the extent to which wildlife contributes to transmission as possible maintenance or amplification hosts is not well understood . Field investigations suggest that wild ruminants and rodents are the most likely RVFV reservoirs [14] . Nevertheless , data from these field surveys are limited , so definitive mammalian natural reservoirs for RVFV are not described [11 , 14] . The landscape epidemiology of RVFV is also incomplete with respect to abiotic systems of influence . For example , periods of excessive rain are strongly associated with RVF outbreaks in East Africa [7 , 15–18] , however very little is known regarding the interaction between climate and terrestrial or hydrogeographic profiles in mediating RVF outbreaks [19] . In addition , there has been a lack of attention to land cover characteristics , which have the potential to influence mosquito habitat , sylvan reservoir habitat , and the movement of domestic livestock through the landscape . Finally , anthropogenic influence , such as human migration , may introduce novel , or increase existing , exposures among pastoralist and/or other rural and peri-urban communities [20] . The study sought to expand our current understanding of RVF epidemiology and infection ecology by investigating the role of diverse hydrogeographic features and wild Bovidae and Muridae species richness in delineating the landscape suitability of future outbreaks across the African continent and Arabian Peninsula .
Occurrence data for RVF outbreaks in humans and livestock animals were obtained from the ProMED-mail electronic surveillance system . This surveillance system is maintained by the International Society of Infectious Diseases and provides near real time and archival documentation of formal and informal reports of infectious diseases [21] . The database was searched using the keywords “rift valley fever” , “rift valley fever virus” , “rvf” , and “rvfv” . Only those reports documenting RVF outbreaks in humans or livestock in unique locations were included ( i . e . duplicate outbreaks were not included ) . One hundred and three reports of laboratory confirmed , geolocated outbreaks of RVF in humans and livestock were documented by the ProMED system between January 1 , 1998 and August 31 , 2016 . Google Maps was used to capture the geographic coordinates for each outbreak and cross-checked against Open Street Map . Centroids of the reported outbreak locations were recorded to a spatial resolution of 4km2 . To test our landscape suitability model ( see Statistical Analysis section ) , a second source of RVF outbreak data were obtained from the World Organization for Animal Health ( OIE ) . OIE maintains an official biosurveillance mechanism for RVF in livestock . These data have been archived since 2004 and can be accessed via the World Animal Health Information System ( WAHIS ) web portal [22] . Reports included the location of each event by place name , the date , type of livestock affected , and the number of infected animals identified . Between January , 2005 and August , 2016 a total of 50 "immediate notification" and subsequent “follow-up” RVFV outbreak reports were submitted to OIE . The geographic coordinates for these events were obtained with Google Maps as above . Outbreaks from ten of these reports could not be located within this coordinate reference system . This left a total 40 OIE reports with 102 unique outbreak occurrences . Twenty-three of the OIE documented outbreaks were also recorded in the ProMED surveillance and therefore were not included in this testing dataset to prevent inflation of model performance . Thus , the final OIE sample of 79 was used for model testing . Altitude and four Bioclim climate rasters were obtained from the WorldClim Global Climate database and used as climate indicators for this investigation [23] . Aggregate spatio-temporal weather station data between 1950 and 2000 were used to calculate the mean temperature during the hottest and coldest quarters , and the mean precipitation during the wettest and driest quarters , and extracted as 30 arc second ( approximately 1 km2 ) resolution rasters [24] . Vegetation cover was assessed using the MODIS-based Maximum Green Vegetation Fraction ( MGVF ) , which is a data product from the United States Geologic Survey's Land Cover Institute [25] . The MGVF records the percentage of green vegetation cover per pixel as a function of the normalized difference vegetation index at a resolution of 1 km2[26] . Rasters were obtained at two time points , years 2001 and 2010 , and the difference between them calculated to determine vegetation loss over this 10 year period . Change in MGVF over this time period was considered a more robust representation of vegetation cover than mean MGVF , and therefore more appropriate in assessing its influence on RVF landscape suitability . The Global Lakes and Wetlands Database [27] was used to define surface water . This raster was derived from three discrete components . The first two comprised vector data of polygons . Component 1 represented lakes with area ≥ 50 km2 and controlled water reservoirs with volume ≥ 0 . 5 km3 , while component 2 represented all surface water with area ≥ 0 . 1 km2 . The third component combined and rasterized the polygon data from the first two components , while supplementing the wetland data . The final 1 km2 raster based on component three was used here . The surface water categories were: lake , controlled water reservoir , river , freshwater marsh , swamp , coastal wetland , brackish , bog , or intermittent wetland [28] . The surface water types were extracted and new distance rasters created . Distance was calculated in the QGIS geographic information system using the proximity function to produce separate 1 km2 resolution rasters for each water category[29] . Pixel values in these rasters convey the distance in kilometers between a given pixel and the nearest pixel occupied by each unique category of surface water . In this way the models can incorporate a spectrum of proximity to diverse hydrogeography across the metacontinent ( see Statistical Analysis section ) . Net human population migration was obtained as a 30 arc-second raster from the Socioeconomic Data and Applications Center ( SEDAC ) , which is part of the National Aeronautics and Space Agency's Earth Observing System Data and Information System [30] . This raster describes the net change ( increase vs . decrease ) in persons per km2 from the period 1990 to 2000 [31] . The global densities of cattle , sheep , and goats were represented as 1 km2 resolution rasters from the Gridded Livestock of the World ( GLW ) [32] . The GLW also classified ruminant livestock production systems by system ( livestock-only , mixed rain fed , and mixed irrigated ) and climate regime ( Hyper-arid , Arid , Humid , and Temperate/Tropical Highlands ) comprising 12 production system categories plus one additional category classified as Urban [33] . Rasters of Bovidae and Muridae species richness at 1 km2 resolution were acquired from the International Union for Conservation of Nature ( IUCN ) and Center for International Earth Science Information Network ( CIESIN ) [34] . Finally , all species of Aedes mosquitoes observed across the geographic range of RVF outbreaks were extracted from the Global Biodiversity Information Facility ( GBIF ) [35] . There were 215 field observations of Aedes mosquitoes geolocated within the African continent and Arabian Peninsula . However , of these 215 mosquito observations , 151 observations recorded the genus only without species designation , while 57 were Ae . africanus , and 7 were Ae . albopictus . As such , there was not sufficient species representation in the GBIF to produce valid models of the ecological niche of Aedes vectors . One generic ecological niche model of Aedes mosquitoes was included in an exploratory analysis , but this contributed very little to the loss function when modeling RVF landscape suitability ( see Statistical Analysis section ) , further suggesting that the mosquito data were insufficient to include in the current investigation . Similarly , this analysis did not include potential Culex amplification vectors as there were too few GBIF specimens across the region and those that were present were of too diverse an ecological and behavioral spectrum to be pooled for analysis . This study used maximum entropy ( Maxent ) machine learning to model the landscape suitability of RVF outbreaks in human and livestock hosts across Africa and the Arabian Peninsula at a resolution of 4 km2 . In the current study , risk is defined explicitly as the probability of landscape suitability to RVF outbreaks . Machine learning in general , and Maxent in particular , is analytically appealing because a specific model form is not assumed . Instead algorithms create rule-based data partitions that optimize homogeneity between predictors and outcomes [36] . Further , the Maxent machine learning algorithm does not require the locations of RVF outbreak absences which are effectively unknowable [37 , 38] . The full Maxent model ( based on ProMED data ) comprised the following landscape features: mean dry quarter precipitation; mean warm and cold quarter temperature; change in vegetation cover; proximity to the surface water features; wild Bovidae and Muridae species richness; cattle , sheep , and goat densities; and net human migration between 1990 and 2000 . Correlation between most of the landscape factors acquired for this study was low . However , there were a few exceptions ( wet quarter precipitation , ruminant production systems , swamp , and altitude ) , all of which were correlated with several other landscape factors and provided generally redundant information . Therefore , these factors were dropped from the original 22 predictors acquired from our data sources described above . Ten thousand background points were sampled , weighted according to human population density to adjust for any potential sampling bias in RVF occurrences derived from ProMED . A value of 1 . 0 was selected for the regularization parameter , to correct for overfitting of the model predictions . The Maxent models were trained using five-fold cross-validation . This approach divides the training set into k = 5 subsets , iteratively fits the model to 4-subset combinations , and then tests against the 5th . Each of the five subsets included approximately 20 RVF outbreaks selected randomly from the total number of available observations in the training dataset ( ProMed; n = 103 outbreaks ) . Landscape features used in the full Maxent model were ranked according to their permutation importance , which randomly permutes the values of the landscape factors between background and presence points in the training dataset . This is preferred over the direct percent contribution to the loss function because it is non-heuristic and more robust to any residual correlation in assessing the influence of individual features on RVF landscape suitability [37 , 39] . Finally , as a robust evaluation of prediction error , the trained models were tested against the data obtained from OIE . The difference in model predictions based on the training and testing data was used to assess the model prediction error , which was reported as the area under the curve ( AUC ) . The models were fit using the maxent function ( dismo package; v . 0 . 9–3 ) setting the distribution to Bernoulli [38 , 40 , 41] . All analyses were performed using R statistical software version 3 . 1 . 3 [42] .
The distributions of RVF outbreaks captured by ProMED and OIE are presented in Fig 1 . The clustering of these outbreaks in the Sahel and in eastern and southern African is demonstrated , as is the more recent emergence of RVFV in the Arabian Peninsula . All landscape features used in the ecological niche modeling are presented separately for the abiotic ( climate , vegetation change , and surface water ) and biotic ( livestock densities , Bovidae and Muridae species richness , and human population migration ) features in Figs 2 , 3 and 4 . The predicted landscape suitability of the African continent and Arabian Peninsula to RVF outbreaks is presented in Fig 5 . High risk landscapes were identified in Mauritania extending eastward into the Sahel , as well as in large portions of Sudan , Kenya , Tanzania , South Africa , Madagascar , and a corridor adjacent to the Red Sea in Saudi Arabia and Yemen . Moderate landscape suitability was predicted for northern parts of the Maghreb , the Horn of Africa and the broader Arabian Peninsula . The Maxent model identified proximity to intermittent wetlands ( permutation importance = 18% ) , Bovidae species richness ( 11 . 7% ) , sheep density ( 11% ) , dry quarter precipitation ( 10 . 2% ) , and proximity to freshwater marsh ( 9 . 1% ) as the most influential features to RVF landscape suitability ( Fig 6 ) . Muridae species richness was not as influential to suitability as Bovidae richness , but was impactful in the model with 8 . 6% permutation importance . Response curves for these features and RVF outbreak risk are presented in S1 Fig . Increasing wild Bovidae species richness was associated with an increase in landscape suitability , as was sheep density up to an approximate average of 100 animals per km2 , after which it decreased and remained constant at 250 animals per km2 . Muridae richness demonstrated a V-shaped relationship with high landscape suitability in areas of low and high species richness . Close proximity to intermittent wetlands and freshwater marsh was associated with greater suitability to RVF outbreaks , as was low precipitation during the driest quarter . The model performed well when tested against the OIE data with the AUC equal to 83% .
This study found that proximity to wetlands in landscapes that are rich in wild Bovidae species and high in sheep density , delineate the most suitable landscapes for RVF outbreaks . Moreover , this study found that the RVF risk surface may extend to regions beyond the historical range of past zoonotic experience should the virus be introduced to these regions via livestock transport or local invasion by infected mosquitoes . | Rift Valley fever ( RVF ) is a vector-borne zoonotic disease that imparts a substantial burden to the economy and public health of pastoralist communities across the African continent and Arabian Peninsula . Furthermore , RVF is also an emerging pathogen of growing global concern . Knowledge of the epidemiological and ecological factors that influence the geographic distribution of RVF outbreaks and determine risk for humans and animals is incomplete . The current study examined the distribution of RVF outbreaks from 1998 to 2016 and modeled their occurrence as a function of climate , surface water , land cover , livestock density , wild mammalian species richness , and human migration . The results indicate that wetlands , Bovidae species richness , and sheep density were associated with increased risk of RVF outbreaks . Our findings contribute to improved understanding of the spatial and ecological dynamics of RVF risk with a particular emphasis on the distribution of wetlands and potential wildlife reservoirs in designing RVF surveillance programs . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"geomorphology",
"invertebrates",
"livestock",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"rift",
"valley",
"fever",
"virus",
"surface",
"water",
"pathology",
"and",
"laboratory",
"medicine",
"landforms",
"ruminants",
"pathog... | 2017 | Wetlands, wild Bovidae species richness and sheep density delineate risk of Rift Valley fever outbreaks in the African continent and Arabian Peninsula |
Mutualisms between species play an important role in ecosystem function and stability . However , in some environments , the competitive aspects of an interaction may dominate the mutualistic aspects . Although these transitions could have far-reaching implications , it has been difficult to study the causes and consequences of this mutualistic–competitive transition in experimentally tractable systems . Here , we study a microbial cross-feeding mutualism in which each yeast strain supplies an essential amino acid for its partner strain . We find that , depending upon the amount of freely available amino acid in the environment , this pair of strains can exhibit an obligatory mutualism , facultative mutualism , competition , parasitism , competitive exclusion , or failed mutualism leading to extinction of the population . A simple model capturing the essential features of this interaction explains how resource availability modulates the interaction and predicts that changes in the dynamics of the mutualism in deteriorating environments can provide advance warning that collapse of the mutualism is imminent . We confirm this prediction experimentally by showing that , in the high nutrient competitive regime , the strains rapidly reach a common carrying capacity before slowly reaching the equilibrium ratio between the strains . However , in the low nutrient regime , before collapse of the obligate mutualism , we find that the ratio rapidly reaches its equilibrium and it is the total abundance that is slow to reach equilibrium . Our results provide a general framework for how mutualisms may transition between qualitatively different regimes of interaction in response to changes in nutrient availability in the environment .
Species in a community interact in a bewildering variety of ways , from parasitic to competitive to mutualistic . Mutualisms , in which two species engage in reciprocal cooperative behavior that benefits both partners , are thought to be particularly important for the stability of ecosystems [1 , 2] , although recent work questioned this role of cooperation in ecosystem stability [3] . Mutualisms in nature are common and diverse , including the pollination of crops and other plants by bees [4] , the cross-protection between clown-fish and anemone [5] , and the symbiosis between tubeworms and bacteria [6] . In the case of the tubeworm , the interaction is completely obligatory because it has no digestive system and acquisition of energy depends completely on bacterial symbionts . The mutualism between most plants and their pollinators , however , is typically facultative , as most plants have multiple pollinators and most pollinators feed from multiple plant species . Within the microbial realm , mutualisms can be due to cross-protection [7] or due to cross-feeding , in which each species supplies their partner with nutrients . Cross-feeding interactions can be present within a species [8] , between pairs of species [9–11] , or could represent a complicated network of dependencies [12] and possibly play a major role in driving the diversity of microbial communities in environments such as the soil [13 , 14] . In addition , cross-feeding could play an important role in determining the species composition and community-level functioning within the human gut microbiome [15] . Laboratory experiments are ideal for studying cross-feeding mutualisms , as they enable fine-grained control of microbial populations and the resources available in the environment . This provides the potential to integrate experiments and models in ways not possible in the field . For example , laboratory experiments have been used to show that cross-feeding can have a stabilizing effect on the relative abundance of two microbial species [9] , which can protect against invasion by cheater strains [16] . Although species in a mutualism generally benefit from interacting with each other , these benefits might decrease in different environments . A major focus of recent research on mutualisms has attempted to elucidate the conditions in which a mutualism can break down or switch to parasitism [17 , 18] . For example , the cross-protection mutualism between ants and the plants that house them can break down when grazing pressure on the plant is reduced [19] , and mycorrhizal mutualisms can become parasitic in the absence of abiotic stresses [20] . Theoretical work predicts that certain mutualisms can become competitive in high nutrient conditions [21] . Moreover , a global analysis of plant interactions concluded that interactions were often facilitative in the challenging environments present at high elevation , whereas the interactions became increasingly competitive in the more benign environments at low altitudes [22] . More generally , the mutualism–parasitism continuum hypothesis posits that a number of environments may cause a mutualism to degrade into a parasitic interaction [23] . Conversely , exposure to certain challenging environments that favor cooperation can stimulate establishment of novel mutualistic interactions [24 , 25] , and theoretical work predicted that almost any pair of species in a microbial ecosystem could establish cooperative interactions when grown in the right nutrient conditions [26] . Resource availability can also alter features other than the growth rate of cooperative strains . For example , resource availability can affect the spatial structure of cooperative species in a biofilm [27 , 28] , as well as the degree of intermixing of cooperative strains during a range expansion [10 , 29] . Although multiple studies have observed a shift in interaction because of varying environmental conditions , a detailed understanding of these changes is missing . It is currently unknown what the possible interaction shifts are and how the population dynamics of a mutualism are affected by these shifts . In our work , we use a synthetic cross-feeding yeast system in which we can modulate the relative strength of the mutualistic and competitive aspects of the interaction by supplementing the media with the amino acids that the strains cross-feed . By changing these two nutrient concentrations , we are able to switch between a surprisingly large number of different interaction types , including obligatory and facultative mutualism , competition , parasitism , competitive exclusion , and extinction of the population . Each of these regimes shows qualitatively different dynamics , which we can understand using a simple model . Our experiments shed light on the important question of how resource availability can modulate the types of interaction between species in a mutualism .
As a model system for mutualistic interactions , we used two non-mating Saccharomyces cerevisiae budding yeast strains that have been engineered to be deficient in the biosynthesis of an essential amino acid and also overproduce the amino acid required by its partner ( Fig 1A ) [10] . The red fluorescent protein ( RFP ) -tagged leucine auxotrophic strain ( Leu- ) overproduces tryptophan , whereas the yellow fluorescent protein ( YFP ) -tagged tryptophan auxotroph strain ( Trp- ) overproduces leucine . These strains have previously been demonstrated to form a cross-feeding mutualism when grown on solid agar , with each strain leaking out the amino acid needed by its partner [10] . To determine if we could establish a stable mutualism between these strains in well-mixed liquid batch culture , we inoculated monocultures and co-cultures at a range of leucine and tryptophan concentrations ( Fig 1B and 1C ) . Co-cultures were started with equal amounts of each strain at the same total density as monocultures . Each day we diluted by a factor of ten into fresh media containing the same defined concentrations of leucine and tryptophan ( Fig 1B ) . For a culture to survive , the growth of a population during the day should be at least as large as the decrease caused by dilution , and a population thus needs to divide at least log2 ( 10 ) = 3 . 3 times each day . In monoculture , Trp- cells required at least 2 μM tryptophan to avoid going extinct due to dilution , whereas Leu- cells required a minimum of 32 μM leucine . In contrast , co-cultures could survive on concentrations of leucine and tryptophan where the monocultures would each go extinct . Co-cultures survived eight of these growth-dilution cycles , indicating a stable mutualism . Even in concentrations where monocultures survived , we found that co-culture density was often much higher than the sum of monoculture densities ( Fig 1C ) , suggesting that in this regime the strains were interacting in a facultative mutualism . Understanding the relative benefits that each partner in the mutualism does or does not receive requires that we also determine the population abundance of each strain at different amino acid concentrations . We therefore co-cultured the strains and measured the population composition by flow cytometry at the end of each day . We tried to make both strains receive equal benefits from the amino acids being supplemented by adding leucine and tryptophan in a ratio of 8 to 1 , which is approximately the intracellular ratio of these amino acids [30] . We found that at low amino acid concentrations ( 1 μM tryptophan , 8 μM leucine; 1 and 8 μM ) , the strains indeed form an obligate mutualism with an apparently stable coexistence , because relative abundance changes little over time ( Fig 1D ) . At medium amino acid concentrations ( 8 and 64 μM ) , the strains form a facultative mutualism , with both strains benefiting from the presence of the other strain , yet also surviving when grown in monoculture . At high amino acid concentrations ( 32 and 256 μM ) , we observed coexistence of the two strains , but with the Trp- strain at an equilibrium abundance below what it would have reached in a monoculture . At this high amino acid concentration , we therefore found that the strains are forming an amensalism , in which the Leu- strain is relatively unaffected by the interaction but the Trp- strain performs worse in co-culture than in monoculture . This demonstrates that a simple microbial cross-feeding mutualism can transition into a qualitatively different interaction by a simple change in environmental conditions . Throughout our study , we compare the final population size of each strain in monoculture and co-culture to assess whether each strain is benefitted , harmed , or unaffected by the presence of its partner in each environmental condition . Once populations have reached an equilibrium size , all populations have the same mean growth rate over the course of the day , because reaching the same population size after a cycle of dilution and growth requires that each cell type undergo log2 ( 10 ) = 3 . 3 divisions over the course of the day . The division rate of a population is therefore not an appropriate measure of fitness or benefit/harm from a partner , as the division rate at equilibrium is always the same given the constant dilution rate present within the experiment . We also note that throughout each daily cycle of growth , the strains alter their habitat by consuming and producing amino acids . Therefore , the label for the different environments ( e . g . , 2 μM tryptophan and 16 μM leucine ) corresponds to the amino acid concentration of the media that we use to initialize growth at the beginning of each day . To gain insight into the transition between the different regimes of interaction in our cross-feeding strains , we implemented a simple phenomenological model designed to capture the essential elements of the interactions between the strains . We assumed that the two strains Trp- ( X ) and Leu- ( Y ) have a per capita growth rate that is modulated by the mutualistic partner as well as the supplemented amino acids: dXdt=rxX ( Y+aY+a+κ ) ( 1−X−Y ) −δX ( 1 ) dYdt=ryY ( βX+aβX+a+κ ) ( 1−X−Y ) −δY ( 2 ) Here rx and ry are the growth rates , a is the amount of supplemented amino acids , δ is the death rate imposed by dilution , κ is an effective Monod constant , and β quantifies the asymmetry of benefit that each strain receives from its partner . The growth rate of each strain increases with the abundance of the mutualist partner and the needed amino acid , but this benefit saturates via a Michaelis-Menten/Monod form as a function of both the concentration of the partner and the supplemented amino acid . This particular form for the interaction arises from a resource-explicit model in which the amino acid dilution/degradation is larger than consumption , but the qualitative predictions of the model are robust to this assumption ( S1 Information ) . We assume that the supplemented amino acids are always added at a fixed ratio , so we use a single variable “a” to capture the amount of supplemented amino acids ( despite the fact that the two strains are actually consuming different amino acids ) . Because the 1-to-8 ratio of tryptophan to leucine should give about equal “relative” amounts of amino acids , we used the same scaling constant ( κ = 0 . 12 ) for both equations . The two strains are also assumed to use other resources in the environment and hence saturate at a total population size , which is normalized to 1 . Additionally , we recapitulated our daily dilutions by introducing a fixed death rate , δ = 0 . 5 ( although our experiments are done in batch culture , for simplicity we model our mutualism in continuous culture ) . We incorporated only two aspects of the asymmetry between our two strains . First , based on competition experiments in saturating amino acid concentrations ( 200 and 1 , 600 μM ) , we calculated that Leu- has a fitness disadvantage of ~7 . 5% in optimal conditions ( S1 Fig ) , so we set the normalized growth rates to be rx = 1 and ry = 0 . 925 . Second , we assume that the Trp- strain contributes more nutrients to the mutualism than the Leu- strain ( β = 2 ) because the Leu- strain dominated co-cultures at non-saturating amino acid concentrations ( Fig 1D , also see below ) . This simple phenomenological model was able to explain the qualitative regimes of interactions that we observed previously ( Fig 1D ) and suggested that simply by varying the amino acid concentrations we may be able to observe an even larger number of qualitative outcomes between our two strains ( Figs 2 and S7 ) . Increasing amino acid concentrations from the region of obligatory mutualism ( Fig 2 , blue ) , the model predicts that the interaction should become a facultative mutualism ( green ) followed by a parasitism ( yellow ) , with the Leu- benefiting from the interaction and the Trp- being harmed . The model then predicts that the amensalism previously observed in Fig 1D corresponds to the boundary of the parasitism region and a competition region ( orange ) , in which the strains coexist but at an equilibrium density below what they would reach in monoculture . This outcome is achieved despite the fact that the force leading to coexistence of the strains is still the sharing of amino acids . Since these strains have complete niche overlap , coexistence is not possible without a stabilizing influence , which is provided by amino acid transfer [31] . At even higher amino acid concentrations the model predicts that the strain with a higher maximal growth rate ( Trp- ) should outcompete the slower dividing strain , because in this regime , amino acids are no longer limiting ( Competitive Exclusion , red ) . The model also predicts that due to the asymmetry in the strains , there will be a small region where the interaction is a facultative mutualism for one strain yet an obligatory mutualism for the other strain ( cyan ) . Finally , the model predicts that in the absence of supplemented amino acids , the mutualism will fail and both strains will go extinct ( dark blue ) . These results are not the result of a particular parameter setting , as the model predicts a shift through the same qualitative regimes over a large range of values for the death rate δ ( S8 Fig ) . This model , although exceedingly simple , therefore predicts the existence of a surprisingly wide range of different qualitative outcomes within a mutualist pair . To test these model predictions of many different interaction regimes , we experimentally measured the equilibrium abundances at a wide range of amino acid concentrations ( Figs 3 and S2 ) . As predicted by the model , we found that varying the amino acid concentration caused the mutualist pair to switch between seven different qualitative regimes , with the ordering of these regimes as predicted by the model . From low to high amino acid concentrations , we observed collapse of the mutualism , obligatory mutualism , obligatory/facultative mutualism ( different for the two strains ) , facultative mutualism , parasitism , competition , and competitive exclusion . Note that there are slight differences between the model and experiment in the behavior of the monocultures , as the Leu- strain is more abundant than the Trp- strain at high amino acid concentrations in our experiment . Nevertheless , it is remarkable that such a simple model provides such effective guidance in the outcomes that we observe in our experimental microbial cross-feeding system . In both the model ( Fig 2 ) and in the experimental system ( Fig 3 ) , the two strains coexist for intermediate values of supplemented amino acids , but one or both strains go extinct if the amount of supplemented amino acids is either too small or too large . This means that if the environment were to deteriorate ( for example , by decreasing nutrient availability ) , the system would go through a series of changes in the type of interaction ( e . g . , parasitism , facultative mutualism ) before becoming an obligatory mutualism and finally going extinct due to the environmental deterioration . Similarly , a rich environment would render the mutualism ineffective , so that the strain with lower fitness would eventually be outcompeted by the other . In principle , knowing the interaction type would indicate whether the system is approaching extinction , although this information requires knowledge of the equilibrium densities for both monocultures and co-cultures , which may not be easily available for many natural systems . An alternative way to detect an imminent population collapse consists of looking at early-warning signals , which are characteristic features exhibited by biological populations prior to an abrupt change of state [32] . To this end , we have analyzed the model behavior near the two onsets of extinction , namely in the obligatory mutualism and competition regimes . The equilibrium densities of the two strains in co-culture are given by the single non-zero equilibrium point of Eqs 1 and 2 ( Fig 2 ) . This equilibrium is stable , meaning that the system recovers from small perturbations in the way described by its eigenvalues and eigenvectors ( Fig 4 ) . The two eigenvalues , both negative , indicate how rapidly the equilibrium point is approached by the population trajectories along the directions given by the corresponding eigenvectors . A large negative eigenvalue indicates a rapid convergence ( i . e . , solid black line ) , whereas a small negative value indicates a slow convergence ( i . e . , solid magenta line ) . At nutrient concentrations near the onset of extinction , both in the obligatory mutualism or competition regimes , there is a separation of time scales: the slow eigenvalue goes to zero , indicating that the system takes a long time to reach the equilibrium point ( blue dot in insets ) along the slow eigenvector ( magenta arrow in insets ) . Simulations of the model confirm that near the onsets of extinction , the trajectories align parallel to the slow eigenvector before reaching the equilibrium point ( insets I , IV , and V ) —a phenomenon that does not occur when the eigenvalues assume similar values ( insets II and III ) . Finally , the orientation of the slow eigenvector indicates which quantity is slowly relaxing: close to collapse of the mutualism ( inset I ) , the ratio of the densities of each strain within the population ( i . e . , f = X/Y ) relaxes faster than the total population size ( i . e . , n = X+Y ) ; in contrast , before competitive exclusion occurs ( inset V ) , the population quickly converges to a fixed n , while slowly equilibrating f to the amount determined by the equilibrium point . In summary , our model predicts that the approach to equilibrium is very different when the cross-feeding strains interact in an obligatory mutualism as compared to when they interact competitively ( Fig 4 , see Materials and Methods section ) . Competitively interacting strains rapidly reach carrying capacity , and only later does the ratio of the strains reach equilibrium ( Fig 4 , inset V ) . In contrast , in the obligatory mutualism regime close to collapse , it is the ratio that first reaches equilibrium , and the total population size is the variable that is slow to reach equilibrium ( Fig 4 , inset I ) . In between these two interaction regimes there is no separation of timescales , and the approach to equilibrium is predicted to be approximately uniform from all directions ( Fig 4 , insets II and III ) . These changes in dynamics are expected very generally due to critical slowing down , in which the slow relaxation mode is associated with the direction of the eigenvector as the eigenvalue goes to zero ( Fig 4 ) . The model therefore predicts that simply measuring the dynamics of the partner strains allows for an estimate of the kind of interaction and , hence , how close the population is to collapse . In order to test these model predictions , we measured the dynamics of co-cultures initialized at a wide range of population sizes n and starting ratios f , spanning four and eight orders of magnitude , respectively ( Fig 5 ) . In accordance with the predictions of the model , in high amino acid concentrations ( 32 μM tryptophan and 256 μM leucine ) , we observed rapid convergence of n , whereas f did not equilibrate even after five days ( Fig 5C ) . In contrast , in low amino acid conditions ( 1 μM tryptophan and 8 μM leucine , Fig 5A ) , the interaction is an obligatory mutualism and the cross-feeding interaction resulted in a strong stabilizing effect on the relative abundances [9] , with the populations rapidly reaching a 1-to-1 ratio ( i . e . , f = 1 ) . As f equilibrated , the fate of the populations depended on the population size n: those that started at sufficiently high abundance slowly increased their total population size to the equilibrium point value , whereas populations that started too small or imbalanced were fated to extinction ( n = 0 ) . We were therefore able to experimentally observe the two different separations of timescale predicted by the model in the two different extreme regimes of interaction . Finally , we found that at intermediate amino acid concentrations ( 8 μM tryptophan and 64 μM leucine ) , there was a balance between the two relaxation timescales , thus causing the trajectories to converge to equilibrium from all directions ( Fig 5B ) as predicted by the model ( Fig 4 insets II and III ) . Therefore , the relaxation dynamics of the cross-feeding partners provide an early-warning indicator of population collapse .
We have established an experimental system that captures a multitude of interactions by simply varying the amount of nutrients freely available to two partners in a cross-feeding mutualism . Although it is tempting to conclude that this cross-feeding interaction should be an obligatory mutualism , we demonstrate experimentally that the interaction varies greatly with the environment . Depending upon the environment , we found that our cross-feeding strains could interact as an obligatory or facultative mutualism , parasitism , amensalism , or competition . A simple phenomenological model explained this range of outcomes , which we view as a significant success given that many models of mutualisms have difficulty shifting between such qualitatively different outcomes; indeed , the Lotka–Volterra model of interspecies interactions fails to even describe an obligatory mutualism without leading to ever-expanding populations [33] . Moreover , the model predicts different relaxation time scales on the brink of collapse that have been confirmed in our experimental system . Our experiments and modeling suggest that the interaction becomes increasingly cooperative as the environmental quality deteriorates via decreasing nutrient availability . This observation is consistent with work done on a range of other mutualisms and interspecies interactions [16 , 19 , 20 , 22 , 26] . However , our results show a much greater range of possible interactions than demonstrated previously and strengthen the idea of interactions between species being contingent upon the environmental conditions rather than being fixed . We also found that the population dynamics change drastically with changing nutrient availability . Low nutrient concentrations have a strong stabilizing effect on relative abundance , whereas high nutrient concentrations stabilize total population size . These dynamics provide a possible way to estimate the interaction and stability of a potential mutualism without having data regarding the viability of each species on its own . The strong stabilizing effect on either total population size or relative abundance suggests that variation will be predominantly on the variable that is not strongly stabilized . In particular , at low nutrient availability , fluctuations may lie primarily along the total population size , whereas in high nutrient availability , the fluctuations in relative abundance may be larger . These differences could provide a more accessible way of studying the stability of species with positive interactions , as it requires only studying the fluctuations of the populations around their equilibrium . Moreover , an experimentally tractable cross-feeding system such as ours could be used to explore counterintuitive effects predicted to occur as a result of noise , such as enhanced sensitivity to environmental fluctuations [34] and noise-induced oscillations [35] . In our study , we focused on the ecological dynamics of mutualisms ( changes in the number of individuals in a population ) rather than evolutionary dynamics ( changes in genetic structure ) . Rather than asking questions about how two strains would evolve cross-feeding , we simply assumed a priori that such an interaction had arisen evolutionarily . Given such an interaction as a starting point , we sought to understand the environmental conditions under which the mutualism would transition into competition . It would be fascinating to explore the evolutionary stability of the cross-feeding studied here , particularly because the evolutionary stability may depend strongly upon the environmental context [36] . In this paper , we have focused on the interactions between two auxotrophic strains , each of which produces the amino acid needed by its partner . However , in principle , this cross-feeding mutualism can be invaded by other strains , the most relevant of which would be the double-producer ( producing both tryptophan and leucine ) and the non-producer ( auxotroph for leucine and tryptophan ) . At least within the realm of our model , we predict that at intermediate amino acid concentrations the mutualism is non-invadable by either of these alternative strains ( S5 Fig , S1 Information ) . However , at higher amino acid concentrations the non-producer is predicted to invade and coexist with the single producers ( and , similarly , at lower amino acid concentrations the double producer is predicted to invade ) . It would be interesting to explore further the degree to which cross-feeding can stabilize the coexistence of multiple strains , particularly given the wide range of nutrients that can be shared in a microbial community . It is also worth noting that the two strains used in our study were able to form an effective cross-feeding mutualism without ever having previously grown together , i . e . , in the absence of coevolution . There is still considerable debate regarding whether mutualisms in natural microbial communities arise primarily from this sort of ecological fitting or via coevolution [19 , 25] . Laboratory experiments have demonstrated the stabilizing effects of coevolution on mutualism dynamics [11] . Regardless , we note that our mutualism dynamics are quite stable even in the absence of a period of coevolution . One important feature of our mutualism is that the two strains are almost genetically identical . This means they have near-perfect niche overlap , which results in very strong competition between the two strains when amino acid concentrations are high . In many other mutualisms , the partners will have less niche overlap and will therefore experience less competition . Incorporating this in our model predicts that the degree of niche overlap will have a strong influence on the outcome of the interaction and the degree to which different environmental conditions will switch the nature of the interaction ( S6 Fig ) . As perhaps expected , less niche overlap results in a larger range of parameters in which the species are mutualistic . Future studies in the field and in the laboratory will be needed to elucidate whether the wide range of interactions observed here is relevant for other mutualisms .
Both S . cerevisiae strains are from a W303 background and are genetically modified to cross-feed as described in [10] . The strains were adapted to growing with low amino acid supplementation through seven cycles of daily dilution ( 10X ) and growth in 2 μM tryptophan and 32 μM leucine . In these cycles , populations consisting of ~100 , 000 to 500 , 000 cells underwent bottlenecks in which as few as 10 , 000 cells survived . Monoclonal lines from adapted strain were derived through plating on 1 . 5% agarose plates and were used for all experiments except for comparison with unadapted strains . Strains were grown in batch culture in synthetic medium consisting of Yeast Nitrogen Base ( YNB , Sunrise Sciences ) , Complete Supplement Mixture lacking leucine and tryptophan ( CSM-leu-trp , Sunrise Sciences ) , and 2% glucose . Synthetic medium was supplemented with varying amounts of amino acids as indicated in experiments . All daily dilution experiments were performed in BD Falcon 96-well flat bottom plates . Cells were grown in 200 μl batch culture at 30°C and mixed by a shaker rotating at 900 rpm . Plates were sealed with Bemis Laboratory Parafilm to prevent evaporation . At the start of each co-culture experiment , single colonies were grown for 24 h until saturation in 3 ml synthetic medium containing 100 μM tryptophan and 1 , 000 μM Leucine . They were then diluted by a factor of ten and grown for 4 h to prevent cells from being in stationary phase at the start of the experiment . Cells were spun down and washed three times to remove any excess amino acids . Leu- and Trp- cells were then mixed in appropriate ratios and seeded in BD Falcon 96-well flat bottom plates in 200 μl medium . A daily dilution cycle consisted of 23 . 5 h of growth , after which density was measured by spectrophotometry ( Thermo Scientific VarioSkan Flash Multimode Reader ) and relative abundance was measured by flow cytometry ( Miltenyi MACSQuant VYB , minimum of 10 , 000 cells analyzed ) . Cultures were then diluted by a factor of ten into new 96-wells plates containing fresh medium Figs 2 and 4 have been obtained by computing analytical formulae for the equilibrium point , eigenvalues , and eigenvectors of Eqs 1 and 2 . Bifurcation analysis of the model is shown in S7 Fig . The analytical treatment has been carried out using a computer algebra system and can be found in the supplementary files ( S2 Information ) . Simulated trajectories in the insets in Fig 4 have been obtained by Gillespie simulations [37] of the corresponding stochastic model of Eqs 1 and 2 . The C code used for simulations is attached as supplementary material ( S2 Information ) . | Species often engage in mutualistic interactions that are beneficial for both partners . However , there is also a cost associated with cooperation , for example , in the form of energy required to make nutrients for a partner . When environments change , the costs and benefits of cooperating can change as well , and this can cause the mutualistic interaction to break down into other interaction types , such as parasitism . In this study , we varied nutrient availability to examine how changing environments can affect the interaction between two cross-feeding yeast strains . Lower nutrient concentrations made each strain more dependent on the nutrients provided by its partner strain and thus favored cooperation . Using both experiments and mathematic models , we found that in different environments , these yeast strains can interact in at least seven different qualitatively different ways , including obligate mutualism , facultative mutualism , parasitism , and competition . We also found that the dynamics of how the two strains influence each other change drastically in different nutrient concentrations . Examining the population dynamics could therefore potentially be used to predict the stability or collapse of a community . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"population",
"dynamics",
"organic",
"compounds",
"leucine",
"mathematics",
"chemical",
"equilibrium",
"algebra",
"amino",
"acids",
"population",
"biology",
"physical",
"chemistry",
"aromatic",
"amino",
"acids",
"p... | 2016 | Resource Availability Modulates the Cooperative and Competitive Nature of a Microbial Cross-Feeding Mutualism |
Zika virus ( ZIKV ) and chikungunya virus ( CHIKV ) are highly pathogenic arthropod-borne viruses that are currently a serious health burden in the Americas , and elsewhere in the world . ZIKV and CHIKV co-circulate in the same geographical regions and are mainly transmitted by Aedes aegypti mosquitoes . There is a growing number of case reports of ZIKV and CHIKV co-infections in humans , but it is uncertain whether co-infection occurs via single or multiple mosquito bites . Here we investigate the potential of Ae . aegypti mosquitoes to transmit both ZIKV and CHIKV in one bite , and we assess the consequences of co-infection on vector competence . First , growth curves indicated that co-infection with CHIKV negatively affects ZIKV production in mammalian , but not in mosquito cells . Next , Ae . aegypti mosquitoes were infected with ZIKV , CHIKV , or co-infected via an infectious blood meal or intrathoracic injections . Infection and transmission rates , as well as viral titers of positive mosquitoes , were determined at 14 days after blood meal or 7 days after injection . Saliva and bodies of ( co- ) infected mosquitoes were scored concurrently for the presence of ZIKV and/or CHIKV using a dual-colour immunofluorescence assay . The results show that orally exposed Ae . aegypti mosquitoes are highly competent , with transmission rates of up to 73% for ZIKV , 21% for CHIKV , and 12% of mosquitoes transmitting both viruses in one bite . However , simultaneous oral exposure to both viruses did not change infection and transmission rates compared to exposure to a single virus . Intrathoracic injections indicate that the selected strain of Ae . aegypti has a strong salivary gland barrier for CHIKV , but a less profound barrier for ZIKV . This study shows that Ae . aegypti can transmit both ZIKV and CHIKV via a single bite . Furthermore , co-infection of ZIKV and CHIKV does not influence the vector competence of Ae . aegypti .
Zika virus ( ZIKV; family Flaviviridae , genus Flavivirus ) is a pathogenic arthropod-borne ( arbo ) virus that causes neurological disease in humans and congenital syndrome in newborns and infants [1] . In the 60 years after its discovery in 1947 , sporadic ZIKV infections were reported in African countries and in parts of Asia [2] . The first larger ZIKV virus outbreak was reported in 2007 on the Yap Islands of Micronesia after which the virus quickly spread to other countries in south-east Asia , such as French Polynesia in 2013 , and Cook Islands and Easter Island in 2014 [3] . In 2015 , there was a dramatic increase of reported ZIKV cases in South America , especially Brazil where over 200 , 000 cases of infection , six deaths and over 2 , 200 incidents of ZIKV associated congenital syndrome were reported [4] . Prior to the ZIKV outbreak in the Americas , flavivirus infections linked to congenital disease were rarely reported . However , a causal relationship between ZIKV infection in pregnant women and subsequent birth malformations , such as microcephaly , has now been confirmed [1 , 4 , 5] . The main vector for ZIKV transmission is the Aedes aegypti mosquito [6–9] , while Ae . albopictus [9–11] , Ae . vittatus [12] Ae . luteocephalus [12] , and Ae . hensilli [13] can transmit ZIKV in laboratory studies . Other mosquito-borne viruses that circulate concurrently with ZIKV in South America include chikungunya virus ( CHIKV; family Togaviridae , genus Alphavirus ) , dengue virus ( DENV; family Flaviviridae , genus Flavivirus ) and yellow fever virus ( family Flaviviridae , genus Flavivirus ) . In 2013 , CHIKV was introduced into South America via the Caribbean . Since then over 319 , 000 cases of infection and 135 deaths have been reported in South America [14] . CHIKV strains that circulate in the Americas are predominantly transmitted by Ae . aegypti mosquitoes [15] . Since CHIKV and ZIKV co-circulate in the same geographical regions , individuals can become co-infected with both viruses [16 , 17] . Co-infections of patients with ZIKV and CHIKV already occurred in South America [18–20] , some even reporting triple infection with ZIKV , CHIKV , and DENV [21–23] . Whether a single bite of Ae . aegypti can transmit both ZIKV and CHIKV simultaneously , or whether sequential bites of two infected mosquitoes are required for such co-infections in humans , remains unclear . Here we designed a dual-colour immunofluorescence assay that can concurrently detect ZIKV and CHIKV infection in mammalian and mosquito cells . We analysed the effect of co-infections on virus growth kinetics in mammalian and mosquito cell lines . Furthermore , we studied the effect of co-infection with both ZIKV and CHIKV on the infection and transmission rates of both viruses in Ae . aegypti mosquitoes . Finally , mosquito transmission rates after an infectious bloodmeal and intrathoracic injections were compared to study the effects of the midgut and salivary gland barriers on co- and single-infections .
African green monkey kidney Vero E6 ( ATCC CRL-1586 ) cells were cultured in Dulbecco’s modified Eagle medium ( DMEM; Gibco , Carlsbad , CA , United States ) containing 10% fetal bovine serum ( FBS; Gibco ) , penicillin ( 100 U/ml; Sigma-Aldrich , Saint Louis , MO , United States ) , and streptomycin ( 100 μg/ml; Sigma-Aldrich ) ( P/S ) . Vero cells were cultured as monolayers in T25 cell culture flasks ( Greiner Bio-One , Kremsmünster , Austria ) at 37°C with 5% CO2 , and split every 3–4 days . Prior to infections , Vero cells were seeded in DMEM containing 4- ( 2-hydroxyethyl ) -1-piperazineethanesulfonic acid ( DMEM-HEPES; Gibco ) supplemented with 10% FBS , penicillin ( 100 U/ml ) , and streptomycin ( 100 μg/ml ) , hereafter named DMEM-supplemented . Aedes albopictus C6/36 cells ( ATCC CRL-1660 ) were cultured in Leibovitz L-15 medium ( Gibco ) supplemented with 10% FBS , 2% tryptose phosphate broth ( Gibco ) , and 1% nonessential amino acids ( Gibco ) , hereafter named Leibovitz-complete . Aedes aegypti Aag2 cells were cultured in Schneider’s Drosophila medium ( Lonza , Basel , Switzerland ) supplemented with 10% FBS , hereafter named Schneider’s-complete . Both C6/36 and Aag2 cells were cultured as monolayers in T25 flasks at 28°C , and split every 3–4 days . All proceedings involving infectious virus were executed in the biosafety level 3 laboratory at Wageningen University & Research . An infectious clone derived chikungunya virus 37997 strain ( CHIKV37997 ) was used in all studies . To prepare the chikungunya virus 37997 infectious clone ( pCHIKIC-37997 ) , the 37997 structural cassette was produced synthetically with AscI/EcoRI overhangs ( Baseclear , Leiden , The Netherlands ) and cloned into the previously described CHIKV 37997 replicon CHIKrep-FlucEGFP to replace the Fluc-EGFP fusion gene [24] . CHIKV 37997 RNA was in vitro transcribed from 5 μg PacI ( New England Biolabs ( NEB ) , Ipswich , MA , United States ) linearized pCHIKIC-37997 using SP6 RNA polymerase ( NEB ) following the manufacturer’s protocol . Vero cells were seeded one day prior to infection in 6 well cell culture plates ( Greiner Bio-One ) until a confluency of ~80% was reached . The culture medium was replaced for Opti-Mem ( Gibco ) and 3 μl of in vitro transcribed RNA was transfected into Vero cells using 2 . 5 μl Lipofectamine 2000 ( Invitrogen , Carlsbad , United States ) . Four days post transfection the cell culture medium was harvested , centrifuged and stored at -80°C until further use ( P0 ) . In total , 500 μl P0 was used to inoculate a T75 flask ( Greiner Bio-One ) of C6/36 cells . Four days post infection ( dpi ) the cell culture medium was harvested ( P1 ) , centrifuged and the supernatant was stored in aliquots at -80°C . Virus titers were determined by end point dilution assay ( EPDA ) on Vero cells . Zika virus Suriname strain 011V-01621 ( ZIKVSUR GenBank accession number , KU937936 ) [5] , was obtained through the European Virus Archive Goes Global catalogue ( www . european-virus-archive . com/virus/zika-virus-strain-suriname-2016 ) as a P3 stock grown on Vero cells . ZIKV P4 was generated by inoculating a pre-seeded T75 flask of Vero cells with 250 μl ZIKVSUR P3 . The supernatant was harvested ( P4 ) at 2 dpi , centrifuged to remove cell debris , and the supernatant was stored in aliquots at -80°C . Virus titers were determined by EPDA on Vero cells . Vero cell suspensions were retrieved by detaching Vero cells from a T25 flask with 1 ml of Trypsin-EDTA ( Gibco ) , after which 4 ml of DMEM-supplemented was added . Virus stocks were thawed , vortexed and serial dilutions were made in DMEM-supplemented . Vero cell suspensions were diluted 1:4 with DMEM-supplemented and added to the virus dilutions in a 1:1 ratio . 10 μl of the inoculated dilutions was plated in 6-fold in micro-titer plates ( Nunc , Roskilde , Denmark ) . EPDAs of samples infected with one virus were scored at 3 dpi based on virus induced cytopathic effect ( CPE ) . EPDAs of co-infected samples were fixed with 4% paraformaldehyde and scored by immunofluorescence assay ( IFA ) at 3 dpi . Cell monolayers were seeded in 6-well plates and infected on the same day for C6/36 and Aag2 cells , or the next day for Vero cells . The cell culture fluid was removed and infections were performed at an MOI of 0 . 1 ( 5 . 7 × 104–2 . 2 × 105 TCID50 ) in standard culture media in a total volume of 1 ml . After 1 h the inoculum was removed and the monolayers were washed twice with 1 ml of Phosphate Buffered Saline ( PBS ) , before addition of 2 ml fresh culture medium . C6/36 and Aag2 cells were maintained at 28°C and Vero cells were maintained at 37°C and 5% CO2 . Samples of 100 μl were taken at 0 , 24 , 48 , 72 and 96 hours post infection ( hpi ) and stored at -80°C until titration by EPDA on Vero cells . Cells were fixed with 4% paraformaldehyde/PBS for 1–3 h . Monolayers were washed 3x with PBS , permeabilized by 10 min incubation in 0 . 1% SDS in PBS , and washed 3x with PBS . Monolayers were stained with α-CHIKV-E2 ( Rabbit Polyclonal; 1:5000; [25] ) and pan-Flavivirus α-E ( 4G2; Mouse monoclonal; 1:50 [26] ) in a 5% FBS solution dissolved in PBS for 1 h at room temperature ( RT ) . Cells were washed 3x with PBS and stained with goat-α-mouse-Alexa Fluor 568 ( 1:2000; Invitrogen ) and goat-α-rabbit-Alexa Fluor 488 ( 1:2000; Invitrogen ) for 1 h at 37°C . Monolayers were washed 3X with PBS and visualized using an Axio Observer Z1m inverted microscope ( Zeiss , Jena , Germany ) in combination with an X-Cite 120 series lamp . Vero cell monolayers were seeded in 96-wells plates one day prior to infection and infected at an MOI of 0 . 1 . At the indicated time-point , the medium was removed and replaced with 100 μl of passive lysis buffer ( Promega , Madison , Wisconsin , USA ) . Cells were lysed by 10 min incubation at RT and lysates were stored at -20°C until further use . Twenty-five μl of reconstituted CellTiter-Glo Reagent ( Promega ) was added to 25 μl cell lysate and incubated at RT in the dark for 10 min before measuring the luminescence using a FLUOstar OPTIMA microplate reader ( BMG Labtech , Ortenberg , Germany ) . Cell viability was calculated by normalizing the average luminescence of the sample to the averaged luminescence of the mock . In all experiments female Aedes aegypti mosquitoes ( Rockefeller strain , obtained from Bayer AG , Monheim , Germany ) were used . Larvae and adults were maintained at 27±1°C with 12:12 light:dark cycle and 70% relative humidity . Adult mosquitoes were provided with 6% ad libitum glucose solution . Human blood ( Sanquin Blood Supply Foundation , Nijmegen , The Netherlands ) was provided through Parafilm using the Hemotek PS5 feeder ( Discovery Workshops , Lancashire , United Kingdom ) . Female mosquitoes were kept together with males for 3 to 6 days in Bugdorm-1 insect rearing cages ( 30 x 30 x 30 cm , Bugdorm , Taiwan , China ) , before females were transferred to buckets ( diameter: 12 . 2 cm , height: 12 . 2 cm; Jokey , Wipperfürth , Germany ) and transported to the Biological Safety Level 3 facility for virus infection assays . One day before blood feeding , the glucose solution was replaced by water in order to stimulate blood feeding of Ae . aegypti females . Virus solutions were made by diluting the virus to the indicated titer in DMEM-supplemented for ZIKV , and Leibovitz-complete for CHIKV . Since CHIKV was grown on C6/36 cells and ZIKV on Vero cells , we compensated for differences in cell culture media by mixing 250 μl of virus solution with 250 μl conditioned media from cultured C6/36 cells for ZIKV and 250 μl conditioned media from cultured Vero cells for CHIKV , after which 500 μl human blood was added . The infectious blood meal was offered through Parafilm using the Hemotek PS5 feeder . Mosquitoes were allowed to feed for 1 h ad libitum in light conditions , at 24°C and 70% relative humidity ( RH ) . Mosquitoes were anesthetized with 100% CO2 , placed on a CO2 pad and fully engorged females were selected . Immediately after selection , a selection of mosquitoes was frozen at -80°C to determine the amount of virus ingested by the mosquitoes . Exposed mosquitoes were maintained at 28°C . The glucose solution was refreshed every 2–3 days until 14 dpi . Virus dilutions of 4 × 107 TCID50/ml were prepared by diluting the ZIKV and CHIKV virus stocks 1:1 with conditioned media taken from cultured C6/36 and Vero cells , respectively , to compensate for differences in growth media . Ae . aegypti mosquitoes were anesthetized with 100% CO2 and placed on a CO2 pad . Female mosquitoes were selected and injected with 69 nl of the prepared virus stock using a Drummond Nanoject II Auto-Nanoliter Injector ( Drummond Scientific , Broomall , PA , United States ) . Infected mosquitoes were maintained at 28°C . The glucose solution was refreshed every 2–3 days until 7 dpi . Fourteen days post blood meal or 7 days post injection mosquitoes were anesthetized with 100% CO2 , and placed on a CO2 pad . Mosquitoes that died within the 7 or 14 days incubation period were discarded . Mosquitoes were immobilized by removing their legs and wings with forceps . The proboscis of each mosquito was inserted into a 200 μl yellow pipet tip ( Greiner Bio-One ) containing 5 μl of a 1:1 solution of 50% glucose solution and FBS , for a minimum of 45 min . After salivation , the mosquito bodies were added to 1 . 5 ml Safe-Seal micro tubes ( Sarstedt , Nümbrecht , Germany ) containing 0 . 5 mm zirconium beads ( Next Advance , Averill Park , NY , United States ) . Each saliva sample was added to a 1 . 5 ml micro tube ( Sarstedt ) containing 55 μl DMEM-supplemented with additional Fungizone ( 50 μg/ml; Invitrogen ) , and Gentamycin ( 50 μg/ml; Life technologies ) , hereafter named DMEM-complete . Mosquito bodies and saliva samples were stored at -80°C until further processing . Mosquito bodies were taken from the -80°C freezer and immediately homogenized for 2 min at max speed in a Bullet Blender Storm ( Next Advance ) . Homogenized bodies were centrifuged briefly and resuspended in 100 μl DMEM-complete medium . The homogenate was blended again for 2 min at max speed using the Bullet Blender and centrifuged for 1 min at 14 . 500 rpm . Mosquito saliva samples were thawed at RT . Of each body homogenate or saliva sample , 30 μl was used to inoculate a Vero cell monolayer in a 96 wells plate . After 2–3 h the inoculum was removed and replaced with 100 μl DMEM-complete . For mosquitoes infected with a single virus , the wells were scored for virus induced CPE at 3 dpi . This method was validated by comparing results based on CPE with IFA for the first replicate of mosquitoes , which gave identical results . For mosquitoes that were infected with both viruses , the supernatant was removed and monolayers were fixed with 4% paraformaldehyde in PBS at 3 dpi after which the wells were scored by dual-colour IFA . Bodies and saliva samples of a selection of mosquitoes with a fully disseminated infection of ZIKV , CHIKV , or both , were titrated by EPDA . Kruskal-Wallis tests were used to test for differences between engorged viral titers , and final titers of mosquito bodies and saliva samples . If the outcome of a Kruskal-Wallis test was significant , differences among groups were further tested with Dunn’s tests and corrected with the Bonferroni correction for multiple comparisons . Infection and transmission rates were calculated , respectively , by dividing the number of female mosquitoes with infected bodies or with infected saliva by the total number of female mosquitoes in the respective treatment . Mosquitoes with infectious saliva , but uninfected body ( <1% ) , were excluded from the analysis . Differences in infection and transmission rates were tested with Chi-squared tests . Multiple comparisons were corrected with the Bonferroni correction . All statistical analyses were done with the statistical software package R [27] . Power analysis was performed to confirm adequate sample sizes for the vector competence studies using G*Power software ( Düsseldorf , Germany ) .
A dual-colour immunofluorescence detection assay was developed to investigate whether ZIKV and CHIKV can infect and replicate in the same mammalian or mosquito cell . Vero ( green monkey kidney , mammalian ) cells and C6/36 ( Ae . albopictus , mosquito ) cells were seeded as monolayers and inoculated with ZIKV , CHIKV , or co-inoculated and stained for viral antigens at 48 hpi ( Fig 1 ) . Distinctions between ZIKV and CHIKV infected cells were clear after immunostaining in both Vero and C6/36 cells , indicating that the dual-colour immunofluorescence assay can be used to score co-infected samples for the presence of ZIKV , CHIKV , or both ( Fig 1A & 1B ) . ZIKV infected Vero cells displayed localization of the envelope ( E ( pan-Flavivirus α-E ( 4G2 ) ) ) protein predominantly near the perinuclear regions / endoplasmic reticulum , corresponding to the replication and assembly sites of flaviviruses [28] ( Fig 1A ) . Most Vero cells in the CHIKV infected sample already lysed due to the strong CPE of CHIKV infection , as indicated by the presence of the E2 envelope protein on the remaining cell projections . Viable CHIKV infected cells showed localization of E2 near the cell boundaries , related to the assembly sites of CHIKV [29] ( Fig 1A ) . The localization of E and E2 was similar in C6/36 cells , with ZIKV-E mostly present near the endoplasmic reticulum surrounding the nucleus , and CHIKV-E2 near the cell membrane ( Fig 1B ) . Co-infection did not alter the localization of ZIKV-E nor CHIKV-E2 , indicating that these viruses can co-infect the same cell without obvious interference . These results show that ZIKV and CHIKV are intrinsically capable to co-infect and co-replicate in cells of their mammalian and insect host . Viral co-infections can influence the replication rate in mosquito cell lines and may affect transmission in vivo by the mosquito vector [30] . In order to assess whether ZIKV and CHIKV interfere with each other’s replication , the growth kinetics of ZIKV and CHIKV were determined during co- and single-infections in mammalian Vero , Ae . albopictus C6/36 , and Ae . aegypti Aag2 cells ( Fig 2 ) . Cells were infected at an MOI of 0 . 1 , washed , and culture fluid samples were collected at 0 , 24 , 48 , 72 , and 96 hpi , and titrated by EPDA . In Vero cells , ZIKV reached a peak titer of 8 . 0 × 107 TCID50/ml within 48 hpi ( Fig 2A ) , whereas CHIKV only reached a titer of 8 . 7 × 105 TCID50/ml at 24 hpi ( Fig 2B ) . During co-infection in Vero cells , the titer of ZIKV was approximately 3 logs lower at 48 hpi and 72 hpi as compared to single-infection , whereas the titer of CHIKV was not affected by co-infection with ZIKV . In C6/36 cells , ZIKV reached a relatively low peak titer of 9 . 6 × 105 TCID50/ml at 96 hpi ( Fig 2C ) , whereas CHIKV reached peak titers of 7 . 7 × 108 TCID50/ml at 48 hpi ( Fig 2D ) . Co-infection resulted in approximately 1 log lower titer of ZIKV at 72 and 96 hpi as compared to single-infection , whereas CHIKV replication was not seemingly affected by co-infection . In Aag2 cells , ZIKV reached a peak titer of 6 . 7 × 107 TCID50/ml at 96 hpi ( Fig 2E ) , indicating that ZIKV replicates better in Ae . aegypti as compared to Ae . albopictus cells ( compare Fig 2C with 2E ) . In contrast , CHIKV reached peak titers of 1 . 2 × 106 TCID50/ml at 48 hpi in Aag2 cells , and CHIKV titers rapidly decreased at later time points ( Fig 2F ) . This suggests that CHIKV replicates better in Ae . albopictus than Ae . aegypti cells ( compare Fig 2D with 2F ) . Importantly , co-infections in Aag2 cells did not significantly affect the replication of either ZIKV or CHIKV ( Fig 2E & 2F ) . To investigate whether the observed difference in growth kinetics of ZIKV during co- and single-infections in Vero cells was due to altered cell viability , a cell viability assay was performed ( Fig 3 ) . Signs of virus induced cytopathic effect were readily observed by bright field microscopy at 24 hpi in the CHIKV and co-infected cells , whereas ZIKV induced cytopathic effects were only observed after 48 hours ( Fig 3A ) . Additionally , the cell viability of CHIKV infected and co-infected Vero cells was decreased dramatically to 20% at 48 hpi , whereas ZIKV maintained high cell viability up to 48 hpi ( Fig 3B ) . These results suggest that the reduction of ZIKV titers during co-infection with CHIKV is due to the rapid and extensive CPE resulting from CHIKV-induced host-shut-off [31] , which interferes with ZIKV virion production . Moreover , co-infections did not affect the cell viability in both C6/36 and Aag2 cells until 96 hpi ( Cell viability: 80–100% ) . The viral infectious dose in the blood meal is known to have a strong effect on the mosquito infection rates of mosquito-borne arboviruses [32 , 33] . Therefore , we determined the dose-dependent infection and transmission rates of ZIKV and CHIKV in Ae . aegypti . Female Ae . aegypti mosquitoes were offered an infectious blood meal containing 2 . 0 × 105 , 2 . 0 × 106 or 2 . 0 × 107 TCID50/ml of ZIKV or CHIKV . When comparing single- with co-infections it is important that the infectious blood meals contain equal virus titers and that the mosquitoes engorge similar numbers of infectious particles . To ensure that the infectious blood meals were completely homogenized and to validate our viral dilution series we froze a selection of engorged mosquito bodies directly after the blood meal and determined the virus titer by EPDA . Indeed , mosquitoes infected with increasing doses of infectious virus in the bloodmeal had increasing titers in their bodies . The ingested virus titers of mosquitoes that were infected with the lowest dose were significantly lower than those infected with the two highest doses ( ZIKV: P < 0 . 01 , CHIKV: P < 0 . 01; Fig 4A ) , although mosquitoes infected with the two highest doses were not significantly different amongst each other ( P > 0 . 05 ) . The infection rates were determined at 14 dpi by infectivity assay of mosquito bodies and transmission rates by infectivity assay of saliva samples . Inoculation with 2 . 0 × 105 TCID50/ml in the blood meal resulted in an infection rate of 65 . 3% for ZIKV ( Fig 4B and Table 1 ) . Increasing the infectious dose to 2 . 0 × 106 or 2 . 0 × 107 TCID50/ml significantly increased the ZIKV infection rate to 92 . 2% and 100% ( P < 0 . 01 ) . For CHIKV , inoculation with 2 . 0 × 105 , 2 . 0 × 106 or 2 . 0 × 107 TCID50/ml resulted in infection rates of 47 . 9% , 66 . 7% , or 81 . 2% , respectively . Inoculation with the highest CHIKV dose resulted in significantly higher infection rates compared to the lowest dose ( P < 0 . 001 ) . These results indicate that the mosquito infectious dose for Ae . aegypti is higher for CHIKV than ZIKV . In the same set of experiments , the mosquito saliva was collected by forced salivation assay and scored for the presence of virus to calculate the transmission rates . Transmission rates of 34 . 7% for ZIKV and 10 . 4% for CHIKV were reached with an infectious dose of 2 . 0 × 105 TCID50/ml ( Fig 4C and Table 1 ) . With a viral titer in the blood meal of 2 . 0 × 106 or 2 . 0 × 107 TCID50/ml the transmission rates of ZIKV increased significantly to 68 . 6% and 68 . 3% ( P < 0 . 01 ) , whereas transmission rates for CHIKV of 5 . 9% and 21 . 2% were not significantly different as compared to the lowest dose ( P > 0 . 05 ) . To observe whether increasing infectious doses in the blood meal lead to higher viral titers in the mosquito bodies and saliva we titrated both the mosquito bodies and saliva samples , of mosquitoes with a fully disseminated ZIKV or CHIKV infection ( positive body and saliva ) at 14 dpi . Median ZIKV titers in mosquito bodies reached 2 . 0 × 107 , 2 . 0 × 107 , and 7 . 1 × 106 TCID50/ml for the respective inoculation doses of 2 . 0 × 105 , 2 . 0 × 106 , or 2 . 0 × 107 TCID50/ml ( Fig 4D ) . Median titers of mosquito bodies inoculated with 2 . 0 × 107 of ZIKV were significantly lower compared to median titers of mosquito bodies inoculated with the lower doses ( P < 0 . 05 ) , indicating that a higher infectious dose in the blood meal does not necessary lead to a higher viral load in the mosquito . Median CHIKV titers were approximately 1–3 logs lower than ZIKV titers and reached values of 6 . 0 × 105 , 5 . 0 × 104 , and 5 . 7 × 105 TCID50/ml for the respective inoculation doses of 2 . 0 × 105 , 2 . 0 × 106 , or 2 . 0 × 107 ( Fig 4D ) . No significant differences were found between the median titers of mosquito bodies inoculated with different doses of CHIKV ( P > 0 . 05 ) . In addition , median viral titers in virus-positive mosquito saliva samples were determined . All median viral titers of saliva samples were below the TCID50 detection limit , and no significant differences between saliva samples could be observed ( P > 0 . 05; Fig 4E ) . These results show that Ae . aegypti is a competent vector for both ZIKV and CHIKV . Moreover , the relatively high mosquito infectious dose for CHIKV indicates that a higher viral dose in the blood meal should be used to study the effects of ZIKV and CHIKV co-infections . Co-infections of arboviruses can affect their transmission potential by mosquito vectors and even exclude transmission of one virus [34 , 35] . To investigate the effect of co-infection on the infection and transmission of ZIKV and CHIKV , female Ae . aegypti mosquitoes were offered an infectious blood meal containing a dose of 2 . 0 × 107 TCID50/ml of ZIKV , CHIKV , or both . Titrations of engorged mosquito bodies that were immediately frozen after the infectious blood meal , showed that the mosquitoes ingested equal amounts of CHIKV and ZIKV in the single- and co-infections ( P = 0 . 24; Fig 5A ) . At 14 dpi , saliva was collected from the mosquitoes and the infection and transmission rates were determined by infectivity assay . ZIKV infection rates were 100% for the single-infection and 97 . 9% for the co-infection , which was not significantly different ( P = 1 . 00; Fig 5B & Table 2 ) . Similarly , no significant difference was found between infection rates of orally exposed mosquitoes to CHIKV in single- ( 81 . 2% ) and co-infection with ZIKV ( 85 . 4%; P = 1 . 00 ) . In both single- and co-infection , infection rates of mosquitoes orally exposed to CHIKV were significantly lower than those of mosquitoes exposed to ZIKV ( P < 0 . 05 ) . In total , 84 . 4% of mosquitoes that were simultaneously exposed to both viruses , were infected with both ZIKV and CHIKV . These results indicate that co-infection of ZIKV and CHIKV does not affect the infection rates in Ae . aegypti . Transmission rates of mosquitoes orally exposed to ZIKV were 68 . 3% for the single-infection and 72 . 9% for the co-infection , which was not significantly different ( P = 1 . 00; Fig 5C & Table 2 ) . For CHIKV , transmission rates were 21 . 2% for mosquitoes with a single-infection and 14 . 6% for mosquitoes with a co-infection , which was again not significantly different ( P = 1 . 00 ) . However , transmission rates of mosquitoes orally exposed to CHIKV were significantly lower than ZIKV exposed mosquitoes ( P < 0 . 001 ) . Importantly , 11 . 5% of mosquitoes that were simultaneously exposed to both viruses , had both ZIKV and CHIKV in their saliva , showing that Ae . aegypti can transmit both ZIKV and CHIKV via a single bite . In summary , these results show that simultaneous exposure can lead to concurrent transmission of both viruses without affecting the infection or transmission rates of ZIKV or CHIKV in Ae . aegypti . Although no effect of co-infection on the infection and transmission rates was observed , there might be an effect on the viral titers in either the mosquito body or saliva that could have an effect on virus transmission . Therefore , viral titers were determined at 14 dpi for both mosquito bodies and saliva samples , of mosquitoes with fully disseminated infections of ZIKV , CHIKV , or both . ZIKV median titers in mosquito bodies reached 7 . 1 × 106 after single- and 2 . 0 × 107 TCID50/ml after co-infection , whereas CHIKV reached titers of 5 . 7 × 105 after single- and 6 . 3 × 105 TCID50/ml after co-infection ( Fig 5D & Table 2 ) . Median titers of both ZIKV and CHIKV in saliva samples reached 1 . 0–4 . 6 × 103 TCID50/ml ( Fig 5E & Table 2 ) . Compared to single-infection , co-infection did not influence the titers of ZIKV or CHIKV in mosquito bodies or saliva ( P >0 . 05 ) . ZIKV titers were significantly higher than CHIKV titers in both mosquito bodies ( P < 0 . 01 ) and saliva ( P < 0 . 05; Fig 5D & 5E ) . These results show that co-infection with ZIKV and CHIKV does not affect the transmission potential of Ae . aegypti for either virus . Importantly , these experiments demonstrate for the first time that Ae . aegypti is intrinsically capable of transmitting ZIKV and CHIKV via a single bite . We observed high infection rates for ZIKV and CHIKV , but the transmission rates for CHIKV were notably lower as compared to ZIKV . This substantial difference in transmissibility of ZIKV and CHIKV by Ae . aegypti could be due to the presence of a midgut escape barrier , a salivary gland barrier or both . To discriminate between these two possibilities , female Ae . aegypti mosquitoes were intrathoracically injected ( to by-pass the midgut barriers ) with 2 . 8 × 103 TCID50 units of ZIKV , CHIKV or both viruses . After 7 days , mosquito saliva was collected and bodies and saliva were tested for presence of virus by infectivity assay . Injection with ZIKV , CHIKV , and both viruses resulted in all cases in 100% infection rates ( Fig 6A & Table 3 ) . Transmission rates of mosquitoes injected with ZIKV were similar for the single-infection ( 77 . 6% ) and after co-infection ( 68 . 8%; P = 1 . 00; Fig 6B & Table 3 ) . For CHIKV the transmission rates were also similar for the single-infection ( 22 . 9% ) and the co-infection ( 27 . 1%; P = 1 . 00 ) , again suggesting that ZIKV and CHIKV do not interfere . Transmission rates of mosquitoes intrathoracically injected with CHIKV were significantly lower compared to ZIKV ( P < 0 . 001 ) . In total , 20 . 8% of the mosquitoes that were simultaneously exposed to both viruses had both ZIKV and CHIKV in their saliva . Viral titers were again determined for mosquito bodies and saliva samples of mosquitoes with a fully disseminated infection , which were injected with ZIKV , CHIKV , or both viruses simultaneously . ZIKV reached mosquito body titers of 2 . 0 × 107 TCID50/ml after single- and 5 . 4 × 107 TCID50/ml after co-infections , whereas CHIKV reached titers of 2 . 0 × 106 after single- and 6 . 3 × 106 TCID50/ml after co-infection ( Fig 6C & Table 3 ) . Median titers of both ZIKV and CHIKV in saliva samples reached 1 . 0–4 . 2 × 103 TCID50/ml ( Fig 6D & Table 3 ) . Compared to single-infection , co-infection did not influence the titers of ZIKV or CHIKV in mosquito bodies and saliva ( P > 0 . 05; Fig 6C & 6D ) . The low transmission rates of CHIKV as compared to the high infection rates after both blood meal infection and intrathoracic injection ( Figs 5C & 6B ) , indicate the presence of a salivary gland barrier that prevents the virus from dissemination into the saliva . For ZIKV , the transmission rates after intrathoracic injections and blood meal infections are only slightly lower than the infection rates . This suggests that for ZIKV the salivary glands form a minor barrier for accumulation of infectious virus in the saliva .
Since the start of the global spread of ZIKV , this virus co-circulates with CHIKV in many parts of the world . There is an increase in the number of reports describing co-infections of ZIKV and CHIKV ( and also DENV ) in human patients , but the extent of co-infection in field-collected mosquitoes is not clear . The aim of this study was to assess whether the predominant vector of ZIKV and CHIKV in the Americas , Ae . aegypti , is able to transmit both viruses simultaneously , and whether co-infection may change the vector competence for either virus . Here we show that Ae . aegypti mosquitoes can indeed simultaneously transmit ZIKV and CHIKV via a single bite . Infection with both ZIKV and CHIKV did not result in lowered infection or transmission rates for either virus , although Ae . aegypti was shown to be a more efficient vector for ZIKV as compared to CHIKV . Finally , we show that Ae . aegypti mosquitoes have a salivary gland barrier for both CHIKV and ZIKV . Triple co-infections with ZIKV , CHIKV and DENV have already been reported in patients from Colombia and Nicaragua [21–23] and several cases of ZIKV and CHIKV co-infections in patients have been reported [18–20] . These observations indicate that a single mosquito could take a blood meal containing multiple arboviruses , potentially resulting in the transmission of different viruses simultaneously . Simultaneous transmission of alphaviruses and flaviviruses by Ae . albopictus has been reported for DENV and CHIKV [36] , and for DENV and Sindbis virus [34] . In another study , CHIKV and DENV co-transmission by either Ae . aegypti or Ae . albopictus only occurred after sequential blood meals , but not after simultaneous infection in a single blood meal [37] . Furthermore , co-infection of Sindbis virus and DENV greatly decreases both the infection and transmission rates of both viruses [34] . In contrast to these studies , our results clearly show that ZIKV and CHIKV do not interfere with each other in either their infection or transmission by Ae . aegypti . Our results show that ZIKV and CHIKV can replicate simultaneously in a single cell of both the mammalian host and mosquito vector . Furthermore , we show that ZIKV and CHIKV can simultaneously disseminate to the saliva of Ae . aegypti mosquitoes , indicating that co-infections do not strongly interfere with virus replication . The effect of co-infection of arboviruses on virus replication is still poorly understood . One explanation for interference between different viruses is superinfection exclusion , where infection of a primary virus excludes secondary infection with the same or a different virus [30] . The primary virus infection could induce the host-immune response , claim important cellular factors required for viral replication , or produce defective interfering particles , that suppress replication of a secondary viral infection . Alternatively , primary infection may lead to suppression of the mosquito’s antiviral responses , leading to enhanced infection of a secondary virus ( reviewed in [30] ) . Potentially , asynchronous co-infections of CHIKV and ZIKV in Ae . aegypti mosquitoes could result in detectable interference and may affect the infection and transmission rates . However , viral interference was observed in growth curves in Vero cells with co-infection resulting in decreased ZIKV titers . The interference in Vero cells is likely due to a decrease in cell viability as a result of CHIKV infection , leading to high cell death and , thus , less ZIKV production . In contrast to what we observed in mammalian cells , virus replication and cell viability were not affected when C6/36 or Aag2 cells were infected or co-infected by both viruses . The absence of detectable interference between the two viruses on virus replication in mosquito cell lines supports our findings on vector competence of Ae . aegypti mosquitoes infected with both viruses . CHIKV replicates in spherules at the plasma membrane [38] whereas flaviviruses replicate in perinuclear regions of the endoplasmic reticulum [39] , which may explain the lack of interference between both viruses . Furthermore , tampering of virus replication by defective interfering particles , which often contributes to viral interference , is less problematic with alpha- and flavivirus co-infections [40] . Potentially , co-infections of multiple flaviviruses ( e . g . DENV and ZIKV ) may have a stronger effect on vector competence . Our results support the role of Ae . aegypti as a competent vector for ZIKV and CHIKV . Previous vector competence studies reported infection rates between 70–100% in Ae . aegypti for ZIKV [6 , 9 , 41] . Infection rates reported here for the ZIKVSUR strain are in line with these findings with 100% infection after administering a high dose of 2 . 0 × 107 blood meal , and 65–90% with a ten- to hundred-fold lower dose of ZIKV in the blood meal . The transmission rates of 35–70% for the ZIKVSUR strain are higher than some previously reported transmission rates , which ranged between 10–30% for the American strains of ZIKV in Brazilian and Mexican Ae . aegypti mosquitoes [9 , 41] . However , studies with Australian and Poza Rica strain Ae . aegypti reported transmission rates between 70–85% [7 , 41] , which is more in the range of our findings . For the CHIKV37997 strain , the infection rates reported here ( 50–80% ) are higher than some previous reports on CHIKV vector competence , which range between 10–30% [42 , 43] . However , another comprehensive study that investigated the vector competence of ten different Ae . aegypti populations for CHIKV reported high infection and dissemination rates between 90–100% [15] . Transmission rates of CHIKV by Ae . aegypti mostly range between 40–60% with exceptions to some virus-vector strain combinations that report low transmission rates [8 , 15 , 43] . These findings confirm that vector competence is highly variable and dependent on the specific combination of viral strain and mosquito population . In order for an arbovirus to accumulate in the saliva it has to pass the midgut infection- and escape barriers and the salivary gland infection- and escape barriers [44] . A midgut barrier for CHIKV has previously been reported in Ae . aegypti [45] and Ae . albopictus [46] . Here , we report infection rates of up to 80% for CHIKV , suggesting that the midgut does not form a strong barrier against CHIKV infection in our Ae . aegypti colony . However , CHIKV transmission rates of maximum 20% after an infectious blood meal or intrathoracic injections indicate the presence of a mosquito salivary gland barrier . The presence of a salivary gland barrier is indicated by a low percentage of mosquitoes with a salivary gland infection when a larger percentage of mosquitoes reaches a disseminated infection . We determined transmission rates of CHIKV of maximum 21% after oral exposure while an 81% infection rate was observed , suggesting that the mosquito colony used has a strong salivary gland barrier to the CHIKV37997 strain . Additionally virus-positive saliva samples had a low viral titer for both ZIKV and CHIKV . Although we did not quantify the amount of saliva that the mosquitoes excreted , this suggests that there is indeed a salivary gland barrier for both viruses that prevents the accumulation of high viral titers in the saliva . We determined that the transmission rates after intrathoracic injections were similar to the transmission rates after an infectious blood meal at 7 dpi . Potentially CHIKV and ZIKV may require longer incubation periods between 7 and 14 days to successfully infect the salivary glands and reach the saliva . However , a salivary gland escape barrier has earlier been described for the CHIKV37997 strain , where only 60% of the infected mosquitoes had virus-positive saliva at 7 dpi [47] . It is surprising that ZIKV spread so rapidly from its original , natural range to territories in the Pacific ( Micronesia , Eastern Island , French Polynesia ) , and subsequently to South- and North-America between 2007 and 2015 [2] . Several hypotheses have been proposed for the rapid dissemination of ZIKV throughout the Americas . First , genetic changes of ZIKV strains could result in adaptations that make the strain that circulates in the Americas more virulent . Evidence for this comes from a comparative genomic study that indicated 15 amino acid substitutions in epidemic strains compared to pre-epidemic strains [48] . However , an African ZIKV isolate was shown to outcompete the American strain in vitro and in vivo suggesting that transmission of the epidemic strains is not enhanced [49] . Secondly , the presence of a large naïve and susceptible human population in combination with high densities of anthropophilic mosquitoes might have accelerated the spread of ZIKV . And thirdly , co-infection of mosquitoes with ZIKV and other arboviruses such as CHIKV and DENV or both may have a positive effect on the vector competence of mosquitoes , resulting in increased transmission rates and faster spread of the viruses . We now show for the first time that mosquito co-infections of ZIKV and CHIKV can indeed occur , without altering the vector competence of Ae . aegypti for either virus . Importantly , our results suggest that patients reported with ZIKV and CHIKV co-infections could have been infected with both viruses via the bite of a single Ae . aegypti mosquito . However , the proportion of co-infected mosquitoes in a population of Ae . aegypti mosquitoes is expected to be extremely small . We therefore consider it unlikely that co-infections of multiple arboviruses contribute to the rapid dissemination of ZIKV across the Americas . In summary , this study shows that Ae . aegypti can transmit ZIKV and CHIKV simultaneously by a single mosquito bite . Additionally , we show that Ae . aegypti has a higher vector competence for ZIKV than for CHIKV and that co-infections do not affect the vector competence . By comparison of infections via the blood meal with intrathoracic injections we show that Ae . aegypti has a strong salivary gland barrier for CHIKV and a minor salivary gland barrier for ZIKV . Finally , studies with cell lines show that ZIKV virus production is decreased in mammalian , but not mosquito cells , due to induction of cytopathicity by CHIKV . The outcomes of this research provide novel insights into the effects of co-infections on the transmission of arboviruses by mosquitoes . | Zika virus ( ZIKV ) and chikungunya virus ( CHIKV ) are highly pathogenic arthropod-borne viruses that present a serious health threat to humans . Since 2015 , both viruses circulate in the same geographical regions of the Americas and are predominantly transmitted by the Yellow Fever mosquito Ae . aegypti . There is a growing number of case reports of ZIKV and CHIKV co-infections in humans , but it is uncertain whether co-infection occurred via single or multiple mosquito bites . Therefore , we infected Ae . aegypti mosquitoes via an infectious blood meal with ZIKV , CHIKV , or both and scored the saliva of ( co- ) infected mosquitoes 14 days post infection for the presence of ZIKV , CHIKV or both . Ae . aegypti was competent for both viruses with transmission rates up to 73% ( ZIKV ) and 21% ( CHIKV ) . A substantial proportion of mosquitoes became saliva-positive for both viruses ( 12% ) , suggesting that Ae . aegypti can transmit both CHIKV and ZIKV via a single bite . Additionally , co-infections did not influence the infection or transmission rates of either ZIKV or CHIKV . Our results indicate that co-infection of CHIKV and ZIKV can lead to simultaneous transmission by the same mosquito in the field . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"invertebrates",
"vero",
"cells",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"chikungunya",
"infection",
"pathogens",
"biological",
"cultures",
"tropical",
"diseases",
"microbiology",
"saliva",
... | 2017 | Mosquito co-infection with Zika and chikungunya virus allows simultaneous transmission without affecting vector competence of Aedes aegypti |
Peroxisomes are membrane-bound organelles within eukaryotic cells that post-translationally import folded proteins into their matrix . Matrix protein import requires a shuttle receptor protein , usually PEX5 , that cycles through docking with the peroxisomal membrane , ubiquitination , and export back into the cytosol followed by deubiquitination . Matrix proteins associate with PEX5 in the cytosol and are translocated into the peroxisome lumen during the PEX5 cycle . This cargo translocation step is not well understood , and its energetics remain controversial . We use stochastic computational models to explore different ways the AAA ATPase driven removal of PEX5 may couple with cargo translocation in peroxisomal importers of mammalian cells . The first model considered is uncoupled , in which translocation is spontaneous , and does not immediately depend on PEX5 removal . The second is directly coupled , in which cargo translocation only occurs when its PEX5 is removed from the peroxisomal membrane . The third , novel , model is cooperatively coupled and requires two PEX5 on a given importomer for cargo translocation — one PEX5 with associated cargo and one with ubiquitin . We measure both the PEX5 and the ubiquitin levels on the peroxisomes as we vary the matrix protein cargo addition rate into the cytosol . We find that both uncoupled and directly coupled translocation behave identically with respect to PEX5 and ubiquitin , and the peroxisomal ubiquitin signal increases as the matrix protein traffic increases . In contrast , cooperatively coupled translocation behaves dramatically differently , with a ubiquitin signal that decreases with increasing matrix protein traffic . Recent work has shown that ubiquitin on mammalian peroxisome membranes can lead to selective degradation by autophagy , or ‘pexophagy . ’ Therefore , the high ubiquitin level for low matrix cargo traffic with cooperatively coupled protein translocation could be used as a disuse signal to mediate pexophagy . This mechanism may be one way that cells could regulate peroxisome numbers .
Peroxisomes are single membrane organelles found in most eukaryotic cells [1] . They are involved in various anabolic and catabolic reactions including fatty acid oxidation , cholesterol biosynthesis , hydrogen peroxide metabolism , bile acid and plasmalogen synthesis [2] . Peroxisomal defects have been associated with serious genetic disorders such as Zellweger syndrome and neonatal adrenoleukodystrophy [3] . Peroxisomes are highly dynamic organelles , changing their numbers based on the specific metabolic needs of different tissues and cell types [4] . For example , in rodent livers , peroxisome numbers can rapidly increase two- to ten-fold in a matter of days by the activation of the receptor Peroxisome Proliferator-Activated Receptor-alpha ( PPAR ) [5] . In yeast , changing the carbon source to oleic acid from glucose induces the rapid proliferation of peroxisomes [4] . Conversely , removal of peroxisome proliferators results in degradation of peroxisomes in mammalian cells with peroxisome numbers returning to basal levels within a week [6] , [7] . Similarly , changing the carbon source from oleic acid back to glucose results in the decrease of peroxisome numbers in yeast within several hours [4] , [8] . Peroxisomal degradation in mammals is mostly mediated by selective autophagy , the process of targeting cytosolic components to lysosomes for degradation ( reviewed in [9] , [10] ) — called ‘pexophagy’ for peroxisomes . In pexophagy , superfluous or damaged peroxisomes are recognized by autophagic receptors that target peroxisomes either to autophagosomes or to lysosomes [11] . How peroxisomes are designated for degradation is not well understood . In mammalian peroxisomes , it has been hypothesized that sufficient ubiquitination of peroxisomal membrane proteins induces pexophagy by recruiting sufficient autophagy receptors such as NBR1 to peroxisomes [12] , [13] . There are indications that any ubiquitinated membrane protein can recruit NBR1 [13] , however the specific peroxisomal membrane protein ( s ) ubiquitinated to induce peroxisome degradation are not known . One candidate is the matrix shuttle protein PEX5 , as preventing its recruitment to peroxisomes prevents NBR1 mediated pexophagy [12] . PEX5 is a cytosolic receptor that binds newly translated peroxisomal matrix proteins ( cargo ) through their peroxisome targeting sequence 1 ( PTS1 ) [14] . PEX5 , with cargo , is imported onto the peroxisomal membrane via its interaction with two peroxisomal membrane proteins PEX14 and PEX13 [15]–[17] . On the membrane PEX5 is thought to form a transient pore via an interaction with PEX14 to facilitate subsequent cargo translocation [18] . On the membrane , PEX5 is ubiquitinated by the RING complex , which is comprised of the peroxisomal ubiquitin ligases PEX2 , PEX10 , and PEX12 . We call the RING complex , together with PEX13 and PEX14 , an ‘importomer’ . PEX5 can be polyubiquitinated , labelling it for degradation by the proteasome as part of a quality control system [19]–[21] , or monoubiquitinated , labelling it for removal from the peroxisome membrane and subsequent recycling [22] , [23] . Ubiquitinated PEX5 is removed from the membrane by the peroxisomal AAA ATPase complex ( comprised of PEX1 , PEX6 and PEX26 ) [24] . In mammals , monoubiquitinated PEX5 is deubiquitinated in the cytosol [25] , completing the cycle and leaving PEX5 free to associate with more cargo . The temporal coordination of cargo translocation , with respect to PEX5 ubiquitination by the RING complex and PEX5 removal by AAA , is not yet clear . This raises the basic question of how energy is provided to move cargo into the peroxisome . It has been suggested that there is no direct energy coupling , since it has been reported that cargo translocation happens before ubiquitination [26] . In this case , translocation of cargo would occur upon binding of PEX5 to the importomer . Subsequent removal of PEX5 would simply allow more PEX5-cargo to bind to the importomer , and the AAA ATPase is not necessarily involved in the energetics of cargo translocation . Conversely , an immediate or direct coupling of cargo import with PEX5 removal has been proposed in which energy for translocation would be provided by the AAA ATPase complex as it removes PEX5 from the membrane [27]–[29] . Using stochastic computational simulations , we have explored the implications of several models of how the PEX5 cycle couples cargo translocation with PEX5 removal by the AAA complex ( see Figs . 1 and 2 ) . The first , ‘uncoupled’ , model corresponds to no direct or immediate coupling [26] . The second , ‘directly coupled’ model translocates PEX5 cargo as the same PEX5 is removed from the membrane by the AAA complex [27]–[29] . Our third , ‘cooperatively coupled’ model translocates PEX5 cargo when a different PEX5 is removed from the peroxisomal membrane . While this can be seen as a qualitative variation of directly coupled import , we show that this novel model behaves significantly differently than both uncoupled and directly coupled models of PEX5 cargo translocation . We focus our modelling on accumulation of PEX5 and of ubiquitin on the peroxisomal membrane , as the traffic of PEX5 cargo in the cell is varied . This allows us to connect our models , of how PEX5 cargo translocation is coupled with PEX5 removal , with possible ubiquitin-regulated control of peroxisome numbers through pexophagy . Since both PEX5 levels and peroxisomal ubiquitination levels are accessible experimentally , this suggests an alternative approach to resolving how cargo translocation couples with PEX5 removal . Our modelling also shows that , regardless of what mechanism couples cargo translocation with PEX5 export , translocation coupling may have significant effects on ubiquitin levels of peroxisomes and so on regulation of pexophagy in mammalian cells . For example , both the uncoupled and directly coupled models lead to more ubiquitination with more cargo traffic . In contrast , the cooperatively coupled model leads to less ubiquitination with more cargo traffic . For cooperative coupling , this suggests a mechanism where lack of cargo results in the accumulation of ubiquitinated PEX5 on the peroxisomal membrane , thus leading to the degradation of underused peroxisomes . Our figures are organized as follows . In the Methods section , Figs . 1 and 2 illustrate the three translocation coupling models . In the Results/Discussion section , Figs . 3 and 4 compares the behavior of these models . We then focus on cooperative coupling . We explore the fluctuations around possible ubiquitin thresholds for pexophagy with Fig . 5 , and examine the role of numbers of peroxisomes with Fig . 6 . Finally we investigate the effects of PEX5 export complexes with Fig . 7 .
We model four processes in the PEX5 cycle , each with an associated rate: the addition of peroxisomal matrix proteins , or cargo , to the cytosol ( ) , binding of PEX5-cargo to an empty site of an importomer ( ) , ubiquitination of a PEX5 at an importomer ( ) , and export of ubiquitinated PEX5 from the importomer ( ) . Binding of PEX5-cargo is illustrated in Fig . 1 ( A ) , association of PEX5 with the RING complex in Fig . 1 ( B ) , and ubiquitination of bound Pex5 in Fig . 1 ( C ) . RING association is assumed to be immediate relative to other modelled processes , and so has no associated rate . Fig . 2 illustrates the three distinct models of cargo protein translocation that we consider , discussed immediately below: uncoupled ( Fig . 2 ( A ) and ( B ) ) , directly coupled ( Fig . 2 ( C ) ) , and cooperatively coupled ( Fig . 2 ( D ) ) . These cargo translocation models differ in the details of how cargo translocation coordinates with AAA ATPase activity . We implement the models of the PEX5 cycle computationally using the Gillespie algorithm [36] , for peroxisomes each of which has importomers , each with independent binding sites for PEX5-cargo , and all of which share a cytoplasmic pool of PEX5-cargo with concentration . We track the number of bound PEX5 for every importomer , together with ubiquitination status of every bound PEX5 . Association rates have not been determined experimentally , so we assume diffusion-limited association rates ( see next subsection ) . This allows us to explicitly avoid fine-tuning of parameters . Parameter definitions and values for the quantitative model are summarized in Table 1 . In the model the total number of cellular PEX5 ( ) is held fixed , as is the cytoplasmic volume ( ) , but the number of cytoplasmic PEX5 will vary as they cycle between the cytosol and the peroxisomes . We stochastically add cargo to the cytosol at fixed rate . We assume the association rate is fast , and so we immediately bind cargo to any cytoplasmic PEX5 without cargo . Cargo accumulates in the cytosol if free PEX5 is not available . PEX5-cargo is removed from the cytosol when it binds to a peroxisome importomer [37] with a diffusion-limited rate that depends on the number of importomers with available binding sites . We generally assume that for each importomer there can be at most one ubiquitinated PEX5 by not allowing the RING complex to associate with more than one PEX5 . We do not explicitly model RING complex motion or PEX5 motion within a given importomer , but once a ubiquitinated PEX5 has been removed from the peroxisome we allow ubiquitination of another PEX5 at a rate . We have checked that our results are qualitatively unchanged , though with slightly higher ubiquitin levels , if we instead allow the RING complex to ubiquitinate all of the PEX5 associated with an importomer ( see Fig . S1 ) . The AAA complex can remove ubiquitinated PEX5 from the peroxisomal membrane while the complex is transiently associated with the importomer [38] . This export occurs with a diffusion-limited rate that depends on the number of export complexes , together with the number of importomers with ubiquitinated PEX5 . Every importomer is initially primed with a single PEX5 that is not ubiquitinated , since we do not have peroxisome or importomer biogenesis processes in our model . For most of our results , the system is run for ten simulated minutes , but data is not taken until after the first 10 simulated seconds; the simulation has reached steady state after this time and is run longer for improved statistics . The peroxisomal PEX5 fraction and ubiquitin per peroxisome are recorded every simulated 0 . 1s . Average times above and below thresholds in Figs . 5 ( B ) and ( C ) were measured differently , as described below . Vertical bars indicate standard deviations . Statistical error bars are much smaller than the standard deviations , and are much smaller than the size of data points . To approximate the diffusivity of PEX5 in the cytosol we note that the diffusion constant of EYFP in the cytosol has been measured at for NLFK cells and in HeLa cells [41] . We assume globular shape , and scale the diffusivity with the inverse radius , and the radius with the cube-root of the molecular mass . The molecular mass of PEX5 is [42] with an additional for cargo [43] giving . Using with mass , this gives . Monoubiquitination of PEX5 in mammals is associated with the cytosolic UbcH5 family of proteins [39] , which have a molecular mass of [44] , [45] . Adding ubiquitin ( 8 kDa ) we have , which scaled from YFP gives a diffusivity . HeLa cell extracts have a UbcH5 concentration of [46] , assuming most of the E2 is activated with ubiquitin . Diffusion in membranes of rat basophil leukemia ( RBL ) cells has a measured diffusion constant of [47] . It has also been measured to be for mammals and in yeast [48] . Most recently membrane diffusivity has been measured in yeast as [49] . We use this most recent value , , for the diffusivity of the export complex within the peroxisomal membrane . The radius of a globular protein or protein complex can be approximated by for R in nm and M in Daltons [50] . We estimate the size of an importomer complex by including both the docking machinery involving PEX14 and the RING complex , which have masses of 800 kDa and 500 kDa respectively [33] . For a total mass of 1300 kDa we obtain a radius of . Since very little is known about the population structure of peroxisomes , we use a fixed peroxisomal radius in the middle of the range of reported peroxisomal sizes ( 0 . 1–0 . 8 in diameter [51] ) . We use peroxisomes , unless otherwise stated , which for purposes of computational efficiency is slightly smaller than the average number of reported for mammalian cells [52] . For a spherical cell of radius 10 , with 44 . 4 cytosol [43] , then . This is used to obtain concentrations of PEX5-cargo . A measured cytoplasmic concentration of PEX5 , [43] , corresponds to approximately PEX5 . We take a comparable but smaller number , corresponding to the slightly smaller number of peroxisomes in our system . We set the number of importomers per peroxisome . With , this works out to PEX5 per importomer when . This is much more than the number of possible PEX5 binding sites per importomer that we explore , which reflects the small proportion of PEX5 typically reported on peroxisomes [53] .
We first examined uncoupled and directly coupled models of protein translocation coupling , shown schematically in Figs . 2 ( A ) – ( B ) and ( C ) , respectively . As mentioned above , the dynamics of PEX5 and ubiquitin are indistinguishable for these two models . We consider different number of sites on each importomer for PEX5 binding in Fig . 3 , guided by studies showing distinct [18] , [54] , [55] PEX5∶PEX14 stoichiometries on the peroxisomal surface —— as well as explicit suggestions of multiple PEX5 sites at the importomer [30] . For each , we vary the cargo addition rate and consider both PEX5 populations and ubiquitination levels . As shown in Fig . 3 ( A ) , the cytosolic PEX5-cargo concentration increases approximately linearly for small then sharply increases before reaching a constant plateau at larger . The linear regime arises from a dynamic balance between cytosolic concentration and concentration-dependent binding to peroxisomes through . The plateau arises from saturation of the PEX5 cycling rates , together with complete binding of cytoplasmic PEX5 with cargo . The steep rise before the plateau occurs when the PEX5 cycling becomes rate limited by PEX5 removal through , and coincides with sharply increased peroxisomal PEX5 fraction ( see below ) — essentially more and more importomers are fully occupied by PEX5 and so cannot contribute to PEX5-cargo binding ( see Fig . 4 ( A ) inset ) . Increasing the number of binding sites per importomer , , decreases the cytosolic fraction of PEX5-cargo . The experimentally measured value of ( [43] ) is consistent with all , and roughly corresponds to where the PEX5-cargo concentration sharply increases due to saturation of importomer binding sites ( around ) . Mirroring cytosolic PEX5-cargo concentrations , Fig . 3 ( B ) shows that the peroxisomal PEX5 fraction also increases with . The mutual increase is possible with a fixed number of PEX5 ( ) at the expense of the reservoir of cytosolic PEX5 that is not associated with cargo . PEX5 accumulates on the peroxisome because of the increasing binding rate due to increasing cytosolic PEX5-cargo concentrations . Increasing the number of binding sites per importomer increases the peroxisomal fraction of PEX5 . Fig . 3 ( C ) shows us that we have a lower fraction of ubiquitinated PEX5 as the cargo addition rate increases . This reflects the higher peroxisomal PEX5 fraction , in combination with our restriction that at most one PEX5 can be ubiquitinated on each importomer . Since the peroxisomal fraction increases with the number of binding sites , while the restriction remains unchanged , the ubiquitinated fraction decreases with increasing . The number of ubiquitinated PEX5 per peroxisome is shown in Fig . 3 ( D ) . The number of ubiquitin increases roughly linearly with until it reaches a plateau slightly above 20 ubiquitin per peroxisome . The plateau value corresponds to the balance between ubiquitination ( ) and export ( ) . With the uncoupled and directly coupled models of translocation , neither of these processes depend on the number of PEX5 bound to an importomer — so the plateau is independent of . An exception is when , since the importomer is empty after every PEX5 export and this slightly decreases the ubiquitination rate . In comparison with the peroxisomal fraction of ubiquitinated PEX5 ( Fig . 3 ( B ) ) , there is a significantly larger standard deviation for the ubiquitin per peroxisome . The difference arises since each cellular fraction is averaged over peroxisomes while ubiquitin per peroxisome is not . We have measured the same quantities for the cooperatively coupled model as for the uncoupled and directly coupled models . The cooperatively coupled results for cytosolic PEX5-cargo concentration , shown in Fig . 4 ( A ) , are very similar to those for uncoupled and directly coupled , shown in Fig . 3 ( A ) . Results with only one binding site per importomer ( ) are not shown , as at least two PEX5 are needed for translocation and export with cooperative coupling . Peroxisomal PEX5 accumulation with cooperative coupling ( Fig . 4 ( B ) ) is also similar to uncoupled and directly coupled ( Fig . 3 ( B ) ) . One important difference is that at low cargo addition rates the peroxisomal PEX5 fraction vanishes for uncoupled and directly coupled but approaches a finite value ( approximately 5 ) with cooperatively coupled translocation . We see from Fig . 4 ( B ) that cooperative coupling implies a finite ratio between the peroxisomal fraction at high and low , and that this ratio is controlled by the number of binding sites per importomer . A 1∶5 ratio of PEX5∶PEX14 has been reported in normal conditions [54] , and a 1∶1 ratio when PEX5 export is blocked [18] . Assuming PEX14 levels do not change with cargo traffic , these observations imply a 1∶5 ratio of PEX5 in low∶high conditions , or for cooperatively coupled translocation . With this choice of , we also recover an absolute change of peroxisomal PEX5 between 5 in wild-type cells to 25 in those lacking a RING complex [53] , [55] . The 1∶5 ratio is also possible with uncoupled and directly coupled models , but requires fine-tuning of . The cooperatively coupled results for the fraction of peroxisomal PEX5 that is ubiquitinated , shown in Fig . 4 ( C ) , are also similar to those for uncoupled and directly coupled , shown in Fig . 3 ( C ) . One important difference is that the ubiquitinated peroxisomal fraction approaches 100 for small with cooperative coupling . Each importomer has at least one bound PEX5 , and small allows the bound PEX5 to be ubiquitinated long before a second PEX5 binds and allows cooperative translocation to occur . The number of ubiquitin per peroxisome vs . the cargo addition rate , shown in Fig . 4 ( D ) for cooperative coupling , shows strikingly different behavior from uncoupled and directly coupled translocation models . We see that the number of ubiquitin per peroxisome decreases with increasing . The amount of ubiquitinated PEX5 is high for low cargo addition rates because ubiquitinated PEX5 must wait for another PEX5 to arrive before it can be exported . Ubiquitinated PEX5 decreases as the cargo addition rate increases since PEX5-cargo arrives at the peroxisome more rapidly , allowing ubiquitinated PEX5 to be exported . At large , the asymptotic number of ubiquitinated PEX5 is approximately the same between the uncoupled and directly coupled , and cooperatively coupled translocation models . A slightly higher level is seen for cooperatively coupled translocation with , since after translocation the remaining PEX5 must wait for both ubiquitination and another PEX5 binding in the cooperative model . Similar results have also been obtained for the five-site cooperatively coupled model without the restriction of only a single ubiquitinated PEX5 on each importomer . Fig . S1 shows that the single ubiquitin restriction does not qualitatively change the PEX5 or ubiquitin behaviours . The cooperatively coupled model leads to high ubiquitin levels when there is little cargo addition . Since ubiquitinated peroxisomes will be degraded in mammals [13] , [56] through NBR1 signalling of autophagy [12] , high ubiquitin levels could be used as a degradation signal for peroxisomal disuse . We explore how a threshold level of ubiquitination could function as a trigger for specific peroxisomal autophagy ( pexophagy ) in greater detail below . We restrict ourselves to a five-site ( ) cooperatively coupled model of cargo translocation , since this recovers reported PEX5∶PEX14 stoichiometries [18] , [54] and a fivefold change in peroxisomal PEX5 when RING activity is absent [55] . A simple threshold model of pexophagy would trigger peroxisomal degradation when the number of ubiquitin on a peroxisome exceeds a certain threshold . While this appears straightforward in light of the average ubiquitin levels of Fig . 4 ( D ) , the substantial fluctuations around these averages must be considered . To illustrate the challenge , in Fig . 5 ( A ) we show a time-trace of the number of ubiquitin for a single peroxisome when and with cooperatively coupled translocation . This value of is chosen to lead to a relatively low level of ubiquitination ( see Fig . 4 ( D ) ) . Also shown with dashed lines are two example thresholds , at and at ubiquitin , which are below and above the rounded average of 58 ubiquitin . Stochastic fluctuations in the ubiquitination level lead to crossing of both thresholds . To investigate stochastic threshold crossing more systematically , we show in Figs . 5 ( B ) and ( C ) the average interval of time spent above and below various thresholds , respectively . We consider four thresholds , chosen between the minimum and maximum ubiquitin levels from Fig . 4 ( D ) , as indicated in the legend . For a given threshold , we only present data from a relatively narrow range of cargo addition rates . Beyond this range the threshold is only very rarely crossed , and any such crossings are very brief . This is true whether we are considering a threshold above or below the mean ubiquitin level . The ubiquitin level is able to fluctuate over a given threshold number only for a limited range of PEX5 cargo addition rates . Within this range , the amount of time spent on either side of the threshold changes by more than three orders of magnitude . Since the range is limited , if the system is outside of the range then a simple threshold model could give a clear signal for pexophagy . Even within the range , a simple threshold model may be sufficient because the time spent on either side of the threshold changes very rapidly with changing cargo addition rate . If the pexophagy response is sufficiently slow , rapid excursions across the threshold might be ignored . It would be interesting to study how NBR1 accumulation [12] might refine this scenario . In mammals , the proliferation of peroxisomes can be stimulated by treatment with peroxisome proliferators [57] . After treatment with the proliferators is stopped the expression of peroxisomal matrix proteins ( cargo ) and peroxisome biogenesis factors decrease [58] , [59] and the number of peroxisomes rapidly returns to normal levels [6] , [7] . In mammals , 70–80 of peroxisome degradation in these circumstances is performed by autophagy [10] . Because the degradation of ubiquitinated peroxisomes is by autophagy [12] , [13] , [56] , it is then plausible that the ubiquitin disuse signal we have proposed to signal degradation is involved in returning the peroxisome population to normal levels . To investigate whether the ubiquitin disuse signal could be involved in returning cells to normal peroxisome levels , we have held the number of total PEX5 in our system constant and varied the number of peroxisomes , considering both a halving and doubling of the number . The peroxisomal PEX5 fraction for 50 , 100 , and 200 peroxisomes is shown in Fig . 6 ( A ) and it behaves as expected: the increase from low PEX5 to high PEX5 is preserved , with the 50 peroxisome system halving and the 200 peroxisome system doubling the peroxisomal PEX5 fraction relative to the 100 peroxisome system . As seen in Fig . 6 ( B ) , the peroxisomal ubiquitin accumulation curve is a similar shape for all three , but with systematically lower ubiquitin accumulation for fewer peroxisomes at a given . This reflects the role of PEX5-cargo traffic in clearing ubiquitin from importomers , within the cooperative coupling model of translocation . This could then provide the cell with a straightforward feedback mechanism to adjust the number of peroxisomes to match the rate of matrix protein expression . At a given and a given ubiquitin threshold , between approximately and in this instance , an excess of peroxisomes would lead peroxisomes to be above the threshold and subsequently degraded . As they are degraded the ubiquitin level would decrease , until a stable number of peroxisomes was reached with ubiquitin levels below the threshold . Given that ubiquitin signals degradation through autophagy [12] , [13] , [56] , this mechanism is consistent with observations that autophagy is responsible for the degradation of excess peroxisomes in mammals [7] . Peroxisome proliferators increase the expression of PEX5 cargo proteins , and removing proliferators results in a decrease of cargo proteins [58] , [59] . We have shown that this decrease in cargo would increase the level of ubiquitinated PEX5 on peroxisomes , and could then induce peroxisome degradation through this simple threshold model . Once decreased peroxisomal numbers reduced ubiquitin numbers below the threshold , background levels of peroxisomal biogenesis would stabilize peroxisomal numbers . Decrease of peroxisomal numbers above the threshold would occur rapidly , while increase below the threshold would be slow in the absence of a proliferation signal . We have been unable to determine the number of AAA export complexes on each peroxisome from the literature . Since PEX1 and PEX6 only transiently associate with peroxisomes [60] we may not have , as we assume , . For example , the reduction in PEX26 expression during the removal of peroxisome proliferating signal [61] would result in the decrease of PEX1 and PEX6 on peroxisomes . Peroxisomal damage may also change the stoichiometry of . Fig . 7 ( A ) shows the peroxisomal PEX5 fraction vs for the different indicated by the legend . The peroxisomal PEX5 fraction is independent of larger ratios , indicating that our results will not be very sensitive to our choice of . Nevertheless , at smaller ratios the peroxisomal PEX5 fraction increases as export becomes impaired . This happens first at larger , as expected . Corresponding to PEX5 changes , the peroxisomal ubiquitin is shown in Fig . 7 ( B ) . Again , at larger ratios the ubiquitin levels are unchanged . However , as the ratios get smaller the ubiquitin per peroxisome increases — and this happens first at higher . This means that if the AAA complex numbers of a particular peroxisome are significantly decreased , the ubiquitination levels of that peroxisome will increase . Nevertheless , for smaller the ubiquitin levels do not change until the number of AAA complexes is below 5 of the number of importomers . This suggests that peroxisomes may be resilient to losses of export complexes , except at high . We have modelled PEX5 cycling through the peroxisomal importomer , and measured the temporal dynamics of both PEX5 and ubiquitinated PEX5 associated with peroxisomes , as the matrix cargo traffic is varied via . PEX5 cycling takes matrix proteins from the cytosol to the peroxisome , where they translocate into the peroxisomal matrix . However , the energetics of cargo translocation have remained unclear . We have implemented three models of cargo translocation , illustrated in Figs . 1 and 2 . The first is uncoupled cargo translocation , where the translocation of cargo happens spontaneously on PEX5-cargo association with a peroxisomal importomer . The second is directly coupled translocation , where cargo translocation happens at the same time as export of the ubiquitinated PEX5 to which the cargo is attached . The third is cooperatively coupled translocation , where cargo translocation happens at the same time as export of a different ubiquitinated PEX5 from the PEX5 to which the cargo is attached . Both directly coupled and cooperatively coupled models have cargo translocation driven by the AAA-dependent export of PEX5 from the peroxisomal membrane [28] , [29] . All three translocation models have peroxisomal ubiquitin numbers that strongly depend on matrix cargo protein traffic . Both uncoupled and directly coupled translocation models have indistinguishable PEX5 and ubiquitin dynamics in which peroxisomal ubiquitinated PEX5 increases as cargo traffic increases . In contrast , cooperatively coupled translocation has decreasing levels of peroxisomal ubiquitinated PEX5 as cargo traffic increases . Ubiquitin on the surface of peroxisomes leads to the recruitment of NBR1 , which recruits the autophagic machinery [12] and leads to peroxisome degradation [12] , [13] . For cooperatively coupled translocation , ubiquitin buildup at low cargo traffic could be used as a disuse signal to initiate autophagic peroxisome degradation . This feedback mechanism could be used to rapidly return peroxisome numbers to normal after induced peroxisome proliferation [7] , [10] , [57] . For uncoupled and directly coupled translocation models , the increase of ubiquitin levels at high cargo traffic levels means that to avoid unwanted pexophagy at high cargo traffic the autophagic response to ubiquitin must be insensitive to the maximal levels of PEX5-ubiquitin expected . This then provides a challenge to identify ubiquitinated peroxisomal membrane proteins other than PEX5 that could control pexophagy . If we assume that peroxisomal damage has a range of severity , with lightly damaged peroxisomes avoiding pexophagy , this also implies that additional pexophagy of lightly damaged peroxisomes would be quickly triggered by increases in matrix cargo traffic — as the PEX5-ubiquitin levels tipped the balance of these peroxisomes towards pexophagy . This work investigates only the cycling and mono-ubiquitination of PEX5 . We do not model the ubiquitination of other proteins or polyubiquitination of PEX5 . How might these effect pexophagy signalling and/or PEX5 cycling ? Polyubiquitinated PEX5 can be removed from the peroxisome membrane by the AAA complex [62] , and polyubiquitinated PEX5 is targeted for degradation [19]–[21] . We assume that this background process does not significantly change PEX5 levels as cargo traffic is changed . While the ubiquitination of other peroxisomal proteins , including the polyubiquitination of PEX5 , can contribute to the induction of autophagy [13] , [56] , we assume that these ubiquitination levels do not change significantly as cargo traffic is varied . If so , then they will simply bias or offset the PEX5 mono-ubiquitination signal and any threshold could be appropriately shifted as well . Here , we have focused on PEX5 and its accumulation on the peroxisomal membrane during changes in the import of matrix cargo . If ubiquitination of proteins other than PEX5 , or polyubiquitination of PEX5 , do change significantly as cargo traffic is varied , then they will need to be considered in conjunction with the PEX5 cycling of our model . A 1∶5 ratio of PEX5∶PEX14 is observed with normal conditions [54] , and a 1∶1 ratio in systems with no PEX5 export [18] . This fivefold change is also observed when peroxisomal PEX5 goes from 5 in wild-type to 25 in cells without a functional RING complex [53] , [55] , implying no ubiquitination and so no export . It is possible to recover this fivefold change with uncoupled and directly coupled translocation , but only by tuning parameters – and only for specific values . These ratios are more naturally recovered for a five-site importomer with cooperatively coupled translocation because with cooperative coupling the importomer cannot remove all PEX5 . The 1∶5 ratio would then correspond to low cargo traffic , and the 1∶1 ratio to high cargo traffic or no export . Miyata et al [63] were able to measure peroxisome associated PEX5 and ubiquitinated-PEX5 . Our modelling indicates that PEX5 cycling responds in just a few seconds to changes in matrix cargo traffic . This response is much faster than timescales to change other protein expression or peroxisome numbers , so we expect that changes in peroxisomal ubiquitin with traffic could directly distinguish between the contrasting predictions of uncoupled or directly coupled translocation models and cooperatively coupled translocation models . From Fig . 3 ( D ) and Fig . 4 ( D ) , we see that in the linear regime a doubling of matrix cargo traffic leads to a doubling of peroxisomal PEX5-ubiquitin for uncoupled or directly coupled models , and a halving of peroxisomal PEX5-ubiquitin for the cooperatively coupled model . Complicating this is that we might expect to be close to the end of the linear regime ( i . e . ) in normal conditions , so that the linear response would be seen only for a marked decrease of matrix cargo traffic . Nevertheless , we might expect to be in the linear regime after induced peroxisomal proliferation and before pexophagy has reduced the number of peroxisomes significantly . Our model is tuned for mammalian peroxisomes , since the E2 enzyme for monoubiquitination of PEX5 is cytosolic and is embodied in our model via a 3d diffusion-limited rate from Eqn . 1 . In yeast , the E2 for monoubiquitination of Pex5 is Pex4 , which is attached to the peroxisome membrane by Pex22 so that should be determined by a 2d diffusion-limited rate from Eqn . 2 . We do not expect any qualitative changes to the Pex5 cycling because of this , and cooperatively coupled translocation should lead to an increase of ubiquitinated Pex5 in yeast when matrix cargo traffic is reduced . This could be used to probe the translocation mechanism of peroxisomal matrix proteins in yeast . Nevertheless , the role of peroxisomal ubiquitin in pexophagy appears to be , at best , indirect in yeast [10] , [64]–[66] so that our discussion of ubiquitin thresholds and pexophagy is restricted to mammalian systems . | Peroxisomes are small organelles that must continually import matrix proteins to contribute to cholesterol and bile acid synthesis , among other important functions . Cargo matrix proteins are shuttled to the peroxisomal membrane , but the only source of energy that has been identified to translocate the cargo into the peroxisome is consumed during the removal of the shuttle protein . Ubiquitin is used to recycle peroxisomal shuttle proteins , but is more generally used in cells to signal degradation of damaged or unneeded cellular components . How shuttle removal and cargo translocation are coupled energetically has been difficult to determine directly , so we investigate how different models of coupling would affect the measurable levels of ubiquitin on mammalian peroxisomes . We find that for the simplest models of coupling , ubiquitin levels decrease as cargo levels decrease . Conversely , for a novel cooperative model of coupling we find that ubiquitin levels increase as cargo levels decrease . This effect could allow the cell to degrade peroxisomes when they are not used , or to avoid degrading peroxisomes as cargo levels increase . Regardless of which model is found to be right , we have shown that ubiquitination levels of peroxisomes should respond to the changing traffic of matrix proteins into peroxisomes . | [
"Abstract",
"Introduction",
"Methods",
"Results/Discussion"
] | [
"eukaryotic",
"cells",
"cellular",
"structures",
"subcellular",
"organelles",
"biophysic",
"al",
"simulations",
"cellular",
"types",
"biology",
"computational",
"biology",
"molecular",
"cell",
"biology",
"biophysics",
"simulations",
"biophysics"
] | 2014 | PEX5 and Ubiquitin Dynamics on Mammalian Peroxisome Membranes |
Meiosis is the cell division that halves the genetic component of diploid cells to form gametes or spores . To achieve this , meiotic cells undergo a radical spatial reorganisation of chromosomes . This reorganisation is a prerequisite for the pairing of parental homologous chromosomes and the reductional division , which halves the number of chromosomes in daughter cells . Of particular note is the change from a centromere clustered layout ( Rabl configuration ) to a telomere clustered conformation ( bouquet stage ) . The contribution of the bouquet structure to homologous chromosome pairing is uncertain . We have developed a new in silico model to represent the chromosomes of Saccharomyces cerevisiae in space , based on a worm-like chain model constrained by attachment to the nuclear envelope and clustering forces . We have asked how these constraints could influence chromosome layout , with particular regard to the juxtaposition of homologous chromosomes and potential nonallelic , ectopic , interactions . The data support the view that the bouquet may be sufficient to bring short chromosomes together , but the contribution to long chromosomes is less . We also find that persistence length is critical to how much influence the bouquet structure could have , both on pairing of homologues and avoiding contacts with heterologues . This work represents an important development in computer modeling of chromosomes , and suggests new explanations for why elucidating the functional significance of the bouquet by genetics has been so difficult .
There are two major forms of cell division among eukaryotes . Mitosis is used for cell duplication and meiosis is used to produce gametes and spores ( Fig . 1 ) . Meiosis is preceded , like mitosis , with a round of DNA synthesis that replicates all chromosomes . A major difference between these methods of cell division resides in the number of nuclear ( and chromosomal ) divisions . In mitosis there is a single nuclear division , restoring the normal chromosome complement in two daughter cells . In meiosis there are two rounds of nuclear division creating four daughter cells , with half the chromosome complement ( Fig . 1 D to 1 G; review [1] ) . The sexual life cycle is completed when two of these haploid gametes , or spores , fuse to rebuild a diploid cell . Each of the four gametes/spores produced during meiosis are genetically unique . During cell division allele combinations are assorted through two mechanisms . The first mechanism involves recombination between parental copies of each chromosome , brought about by a process called crossing-over ( Fig . 1 C; arrows ) . Recombination occurs before the first division , so after division chromosomes become a patchwork mixture of paternal and maternal DNA ( Fig . 1 D ) . The second mechanism comes from two rounds of independent assortment of chromosomes . During the first meiotic division one copy of each chromosome segregates away from its homologous partner , and there is no link between unrelated chromosome pairs ( Fig . 1 D ) . During the second meiotic division sister chromatids segregate in opposite directions , and again there is no link in the direction of segregation for unrelated chromosomes ( Fig . 1 F ) . Prior to meiosis the parental pair of homologous chromosomes are relatively dispersed throughout the nucleus ( Fig . 1 A ) . In order that they become close enough to form crossovers , and then segregate from each other , homologues line up in pairs ( reviews [1] , [2] ) . Reorganising premeiotic chromosomes into ordered pairs involves chromosome condensation and movement that ultimately leads to their synapsis ( Fig . 1 B ) . Studies mainly from budding yeast , Saccharomyces cerevisiae , show that precursor molecular events to crossing-over are concomitant with and probably part of the pairing process [1] . Once chromosomes are brought close enough they synapse , and subsequently mature crossover products can be detected in molecular assays [1] . Many genes are involved in causing and regulating the chromosome movements and crossing-over , and chromosome architecture itself plays a role in this physical process [3] . It has been known for over a century that chromosomes can adopt a highly polar organisation in which centromeres are clustered with chromosome arms occupying different latitudes according to arm length ( Rabl organisation [4]; Fig . 2B ) . The Rabl configuration is regularly seen in interphase cells of plants and other species , but it is not universal and is absent from most mammalian interphase cells ( e . g . see [5] , [6] , [7] , [8] , [9] , [10] , [11] , [12] ) . The Rabl organisation is a well established feature in S . cerevisiae , demonstrated by cytological , genetic and physical techniques [13] , [14] , [15] , [16] , [17] . In S . cerevisiae the Rabl configuration chromosome ends ( telomeres ) are dispersed but those from chromosomes of a similar size are more likely to be close to each other [17] . During the leptotene/zygotene transition of meiosis telomeres attach to the nuclear envelope ( Fig . 2 C ) and move to a telomere-clustered bouquet formation ( Fig . 2 D; reviewed in [2] , [13] , [18] , [19] , [20] , [21] ) . Most organisms studied have a demonstrable bouquet stage , but the degree of polarisation varies considerably ( see [2] , [18] , [19] , [22] ) . For example , Schizosaccharomyces pombe chromosomes are highly polarised , with all telomeres confined to a small part of the nuclear volume , where as telomeres in other organisms are restricted to a broader region of the nuclear membrane [22] , [23] . Telomere clustering is generally tighter in organisms with a micro-tubule organising centre to which the telomeres are closely associated [2] . The significance of the Rabl and bouquet configurations is not fully understood . The Rabl organisation could be a relic of the preceding mitotic anaphase [6] . In S . cerevisiae meiosis it is possible that the Rabl configuration contributes to early centromere pairing [24] . In other organisms the structure is weak or absent in premeiotic cells [6] , [25] . Over several years many genes have been identified in various organisms as being required for relocalisation of telomeres to the nuclear envelope ( reviewed in [19] , [26] , [27] ) . Using S . cerevisiae , several laboratories have examined mutants in these genes to determine the significance of the Rabl to bouquet transition . Chromosome movement into the bouquet structure is brought about by telomere interaction , via adapter proteins , with the perinuclear cytoskeleton [28] , [29] , [30] . Location and movement of telomeres on the nuclear periphery requires a complicated and diverse array of functionally conserved proteins that interact directly with telomeres ( e . g . Ndj1/Tam1 [31] , [32] , [33] and yKU70/80 [30] , [34] , [35] ) , link telomere bound proteins to the nuclear membrane ( SUN domain proteins e . g . Mps3 [33] , [36] ) and link trans membrane proteins to the cytoskeleton ( KASH or KASH-domain proteins e . g . Csm4 [37] ) . The timing of meiotic chromosome organisation overlaps with that of recombination . The functional interrelationship between these two meiotic activities is not well understood , and while they influence each other they are not interdependent [33] , [38] , [39] , [40] . For example , deletion of S . cerevisiae NDJ1/TAM1 causes telomeres to become internalised and less mobile , with a change in crossover frequency and distribution , and delayed first meiotic division [31] , [32] , [33] , [37] , [41] , [42] . Similarly deletion of yKu70 ( HDF1 ) disrupts telomeres attachment and bouquet formation [35] . Chromosome pairing in such strains is delayed and the rate of segregation errors ( non-disjunction ) increases in some , though not all , reports [31] , [32] , [35] , [37] . The major increase in nondisjunction in ndj1/tam1 mutants is for nonrecombinant chromosomes reliant the distributive segregation system [31] , [32] , [33] . These observations support the view that the bouquet contributes to chromosome pairing but it cannot be essential . While chromosome movement is probably universally important , the contribution of the bouquet varies widely between species . In some organisms chromosome pairing precedes bouquet formation [20] , and in S . pombe the role of bouquet genes extends beyond pairing , to the proper regulation of the spindle pole body [43] . Recent live cell studies have revealed that chromosome movement in budding yeast meiosis extends well beyond a simple movement of centromeres out of the Rabl configuration , and telomeres into the bouquet formation . From early prophase I , at least until synapsis is complete at pachytene , there are continuous rapid shifts in chromosome position . These can separate whole chromosomes from the main chromosome mass , causing shape changes in the nuclear membrane [33] , [41] , [44] . The movements in yeast are telomere led and unequal throughout the length of chromosomes , with the centromeres sometimes being appreciably less motile , and whole chromosomes transitioning from being rapid movers to stationary [33] , [44] . This activity is dependent on actin , ATP and various proteins also needed for telomere location on the nuclear periphery and bouquet formation [29] , [33] , [41] , [44] . Similarly , studies and modeling from plants have also shown that chromosome movements leading to telomere clustering require microtubules and are directional [45] , [46] , [47] . The chromosome movement is a highly organised and regulated feature of the meiotic programme , not Brownian motion . As prophase I in S . cerevisiae proceeds and chromosomes become more paired and then synapsed , the speed and tendency for movements reduces [33] . That chromosome movement is most vigorous in early prophase I has led to a widely held view that it stirs unpaired chromosomes , both to help homologues to find each other and to break up unwanted prexisting ( ectopic ) interactions [26] , [33] . While chromosome pairing has the more obvious role of bringing homologues close enough to each other to recombine , it probably also has an important role in separating unrelated chromosomes or chromosome regions . All genomes contain a degree of DNA sequence repetition due to the presence of transposons , pseudo genes and gene families evolved from a single locus . Repeated sequences can be dispersed on many chromosomes , or they can be positioned at nonallelic loci on homologous chromosomes . Early in meiosis there is a chance that dispersed repeated sequences will make contact , and compete effectively with allelic sequences for pairing . When this happens there is an opportunity for so called ectopic recombination between nonallelic sequences . Ectopic recombination between diverged repeated sequences is largely repressed by chromatin structure and the mismatch repair system [48] , [49] , [50] , but it occurs at measurable frequencies in various organisms , including budding yeast [51] , [52] , [53] , [54] , [55] , [56] , [57] . Avoiding ectopic crossovers is important because they alter chromosome structure creating translocations . These disrupt chromosome segregation and increase the risk of infertility or abnormal offspring ( e . g . [53] , [56] , [57] , [58] ) . In S . cerevisiae , genetic experiments have used the efficiency of ectopic recombination events ( as compared to allelic recombination ) as a measure both homologue and heterologue chromosome juxtaposition [14] , [58] , [59] , [60] , [61] . Among conclusions drawn from these experiments is the notion that the chance of ectopic interactions is related to telomeres in two ways . Firstly , the distance of interacting loci from telomeres influences their chances of interaction [58] . Secondly , the location of telomeres on the nuclear periphery influences the efficiency of ectopic recombination [59] , [60] . Thus , the telomere led movements that contribute to the bouquet structure and further pairing may be important for disruption of unwanted contacts between repeated sequences [26] , [33] , [59] , [60] . One difficulty with genetic experiments is that pleiotropic effects are very difficult to separate from direct effects . For example , mutating a gene that modifies both chromosome movement and recombination makes it difficult to determine which of these aspects ( either , or both ) is directly responsible for an observed chromosome pairing defect . To augment what has been learnt from genetic and cell biology studies , we set out to develop an in silico test for the possible significance of chromosome tethering to the nuclear envelope , with or without clustering forces . We developed polymer statistic models of chromosome behaviour that can incorporate a diverse range of centromere and telomere clustering , reminiscent of the Rabl and bouquet structures . The model has been used to investigate the potential roles of chromosome tethering and clustering forces upon the likelihood of loci becoming physically close to each other . While the modeling process can be used for any organism , we have set parameters to model the widely used experimental Eukaryotic microbe S . cerevisiae . Our main goal was to determine whether or not simply moving to the bouquet formation could increase the chances of close homologue juxtaposition , and reduce the chances of unwanted ectopic contacts . The model supports the view that telomere led movements into the bouquet structure can be an aid to chromosome pairing . We found that chromosome length and persistence length ( chromosome flexibility ) have a measurable difference on how beneficial the bouquet structure might be to chromosome pairing .
We first investigated how the different physical properties would affect the layout of individual chromosomes in the nucleus . For a particular choice of chromosome parameters ( defining chromosome architecture ) , an average of all pair wise intra-chromosome distances could be calculated from sample trajectories to yield a matrix of average distances . This matrix is represented in a heat map , referred to here as the intrahomologue locus distance map ( LDM; Fig . 3 A ) . The colour code chart indicates the relative distances as a proportion of nuclear diameter ( ND ) . The gradation is from deep red ( zero distance ) through yellow and green to deep blue ( maximum distance equal to the nuclear diameter; Fig . 3 A ) . By locating different loci on the horizontal and vertical axes , the colour at the intersection provides the mean distance between two loci . For example , the telomere-to-telomere distance is given by the region indicated γ in Fig . 3 A . The layout of each of the three chromosomes was tested under two flexibility regimes , as defined by the chromosomes' persistence length . The persistence lengths used were informed by previous measurements of S . cerevisiae chromosomes , with the most flexible value in line with interphase measurements , 0 . 2 µm [62] ( but slightly larger than outside estimates from 3C modelling [68] ) . The more rigid case , 2 . 0 µm , is between the flexible values and high values inferred from pachytene chromosomes [26] . Results from both persistence lengths are displayed in Figs . 3 B , 4 and 5 for short , medium and long chromosomes , respectively . We also used a range of position restricting conditions . These were , No Tether ( spherically confined only ) , Centromeres Tethered to the nuclear envelope , Telomeres Tethered to the nuclear envelope and Telomeres Tethered plus Clustering forces . For all chromosomes tested , and in all four restraining conditions , increasing the persistence length ( and therefore rigidity ) had the expected effect of increasing the mean distance between loci . Thus , the intrachromosomal LDMs for rigid chromosomes in Figs . 3 to 5 are cooler than for flexible chromosomes . Another way to view the data is by rank scoring the distances . For comparison later with the 300 interallelic rank scores , the 90 , 000 mean distances ( of 177 , 000 samples ) per chromosome were ranked and then binned into 300 mean scores . Comparing the flexible and rigid rank score graphs ( Figs . 3 to 5 ) shows that a wider range of mean distances is adopted when the chromosomes are more rigid . For flexible chromosomes , the centromere tether changed the distribution of distances for each chromosome . Pericentromeric regions were more likely to be close to each other , thus on the intrachromosomal LDMs there is more red/orange around the centromeres ( Figs . 3 to 5; Centromeres Tethered region α ) . This change was accompanied by an increase in mean distances between the centromeres and distant loci , causing the intrachromosomal LDMs to become more yellow/green in regions indicated β . The rank scores indicated that centromere tethering caused a reduction in the range of mean distances for the flexible chromosome I ( Fig . 3; Centromeres Tethered ) , but an increase for chromosomes XVI and IV ( Figs . 4 and 5 , Centromeres Tethered ) . While by eye these differences may seem small and to affect a small proportion of the chromosomes' length , it is noteworthy that the mean curves for No Tether and Centromeres Tethered are statistically significantly different from each other ( p<0 . 01 , using both a Kolmogorov-Smirnov and Wilcoxon rank-sum test to compare the distribution of the means for all 90 , 000 intrachromosomal distances , where each mean is averaged across the population of ∼177 , 000 cells for each chromosome and condition ) . Thus , the overall impact on tethering is chromosome size dependent . An influence from chromosome size is also apparent when looking at the data for Telomeres Tethered without clustering force . When telomeres of the same chromosome are randomly attached to the nuclear periphery they become relatively dispersed compared to when they were free to lie anywhere with the nucleus . This is seen in region γ of the intrachromosomal LDMs , which become cooler when compared to No Tether ( Figs . 3 to 5 ) . For all flexible chromosomes this causes a widening of the range of mean intrachromosomal distances that is statistically significant ( p<0 . 01 , using Kolmogorov-Smirnov and Wilcoxon rank-sum tests ) . The longer chromosomes are more spread out in the condition Telomeres Tethered than chromosome I , widening the mean gap further , presumably because the telomeres can be located at more distant sites on the nuclear membrane ( Figs . 3 to 5; Telomeres Tethered , compare rank scores ) . These observations are consistent with those seen in previous polymer-statistics models , in which tethering of centromeres or telomeres to the nuclear periphery increased the average distance between opposite telomeres [64] . We also tested the effect of tethering telomeres to the nuclear periphery with a strong chance of being located close to each other , as would be seen in a bouquet structure ( ν = 50 , see Materials and Methods; Figs . 3 to 5 , Telomeres Tethered LDMs are warm in region γ ) . For chromosome I the clustering of telomeres caused a significant reduction in mean distances for a large proportion of loci ( Fig . 3; Telomeres Tethered plus Clustering , rank scores ) . By effectively pulling the short chromosome into a U-shape , interstitial regions on opposite arms become closer than in any other condition tested . This is shown by the deepening red in region δ , and on the rank score graph the distribution of means is lower than that in other conditions . The effect is similar but less pronounced on chromosome XVI in region δ ( Fig . 4 Telomeres Tethered plus Clustering ) as more loci will be further away from the joint tether site . Also important is the observation that the mean distances increased in region β compared to untethered chromosomes . For chromosome IV the increase in distance in the region β was sufficient to increase the overall range of mean distances compared to the No Tether condition ( Fig . 5; Telomeres Tethered plus Clustering ) . At distance from the tether , order imposed by clustering gives way to changes in chromosome trajectory and the effect of the clustering on the mean distance between loci wanes . This observation has implications , predicting that any influence of the bouquet on chromosome pairing will be chromosome length limited . The different conditions effected rigid chromosomes in similar ways to flexible chromosomes , particularly for the small chromosome I ( Figs . 3 to 5 ) . For the two longer chromosomes the combination of rigidity ( fewer turns in trajectory ) and length means that collision with and deflection from the boundary is more likely ( see e . g . [69] ) . The deflection of chromosome ends away from the nuclear periphery boundary can increase the chances of loci on distant chromosome regions coming close to each other in the nuclear volume . On the intrachromosomal LDMs this caused a striated pattern of alternating warmer and cooler colours ( Figs . 4 and 5; rigid LDMs ) . This effect also reduces or reverses the impact of tethering and clustering of either centromeres or telomeres ( Figs . 4 and 5; compare rank scores flexible versus rigid ) . We set out to determine how similar parameters would impact on the proximity of homologous chromosomes in an otherwise empty nucleus . For each condition we measured the proximity of 300 allelic loci along homologous chromosome pairs I , XVI and IV . A sample distribution of telomeres is indicated for each flexible chromosome during the Rabl to Tight Bouquet time course ( Figs . 6 to 8 ) . In the first condition no special constraints were given to the locations of centromeres or telomeres , this provided a baseline of interhomologue juxtapostion to compare with more restrained conditions ( Figs . 6 to 8; No Tether ) . Even though there is a 6-fold difference in length between chromosomes I and IV , for all three flexible chromosomes the distance between alleles on homologous chromosomes ranged only from approximately 0 . 40 to 0 . 45-times ND . By definition , rigid chromosomes are less likely to make a turn in direction , and therefore they are more spread out as indicated by the measurements of intrachromosomal distances . This has an impact on the position of chromosomes in the nucleus and , therefore , homologue juxtaposition . Increasing rigidity causes an increase in the mean distances between alleles compared to the flexible chromosomes ( Figs . 6 to 8; No Tether , compare Rigid with Flexible ) . It is noteworthy that as chromosome size increases , there is a fluctuation in pairing contacts due to boundary effects ( Figs . 7 and 8; No Tether , Rigid ) . Next , five stationary conditions were used to mimic time course sampling of a continuous process with chromosomes moving from a Rabl configuration to a tight bouquet formation . The Rabl configuration was created by localising centromeres to the nuclear periphery with strong clustering forces ( ν = 50; see Materials and Methods ) . For Telomeres Tethered , telomeres were tethered to random sites on the nuclear periphery . We then utilised three levels of increasing clustering tendencies for tethered telomeres . These are referred to as Early Bouquet , Loose Bouquet and Tight Bouquet . Respectively , the three bouquets had v values of 5 , 10 and 50 ( Materials and Methods ) . For the flexible chromosomes , when centromeres are in the Rabl configuration alleles closest to the centromere were separated by less than 0 . 1-times ND ( Figs . 6 to 8; Rabl , Flexible ) . This compares with a separation of ∼0 . 40-times ND for pericentromeric alleles in the No Tether condition ( Figs . 6 to 8; No Tether , Flexible ) . Moving away from the clustered centromeres leads to a gradual increase in the mean distance between alleles ( Figs . 6 to 8; Rabl Telomere-Telomere graphs , yellow lines tend towards white line moving away from the centromere ) . Chromosome I is not long enough for the influence of the Rabl configuration to completely wane near the telomeres . For chromosomes XVI and IV the mean distance between alleles converges with that for untethered chromosomes . Thus , as chromosomes become longer a decreasing proportion of the total length of homologues will be influenced by the Rabl configuration . Increasing chromosome rigidity had the effect of causing a more rapid drop off of the clustering influence for all chromosomes ( Figs . 6 to 8; Rabl , compare Flexible and Rigid ) . This is shown for chromosome I by the near convergence of the data for Rabl with the data for No Tether on the left arm furthest from the centromere . However , mean interhomologue measurements are still closer to each other than for No Tether ( in all cases , p<0 . 01 , using Kolmogorov-Smirnov and Wilcoxon rank-sum tests ) . The same trends are apparent for the longer chromosomes , but there is some periodicity to the homologue juxtaposition due to boundary effects ( Figs . 4 and 5; Rabl , rigid ) . For all three chromosomes used , tethering telomeres tended to reduce the chance of interhomologue proximity compared to the No Tether condition ( Figs . 6 to 8 , Telomeres Tethered ) . This is most apparent at the tether sites , as without clustering forces tethered homologous telomeres could be constrained to distant sites on the nuclear envelope . At distance from the tethered telomeres the proximity of homologues tends towards that seen for untethered chromosomes . The shortest chromosome is not long enough for any loci to escape the relative disruption to homologue juxtaposition created by Telomeres Tethered without clustering forces . Therefore , in this condition pairing short homologues might be more difficult than pairing long chromosomes , which for a portion of their length are as close as in the No Tether condition . Overall , the mean distance between alleles is increased by loss of the Rabl configuration . Thus , our model could explain why just prior to meiosis , yeast chromosomes appear to be paired and this paring is lost on entry into meiosis until meiotic chromosome pairing is established [70] , [71] , [72] . Early Bouquet formation was modelled by creating a small chance of telomere clustering ( ν = 5 ) . For flexible short chromosomes in Early Bouquet , the range of distances between alleles was 0 . 27- to 0 . 34-times ND ( Fig . 6; Early Bouquet , Flexible ) . This compares to a range of 0 . 40- to 0 . 45-times ND for No Tether and a range of 0 . 47- to 0 . 67-times ND for Tethered Telomeres without clustering . Thus , a relatively small chance of telomeres being close to each other creates a measurable improvement in homologue juxtaposition over many Kb . Increasing the clustering forces to create the Loose and Tight Bouquets brought telomeres even closer together ( respectively , to within 0 . 20- and 0 . 12-times ND ) . As seen for the Rabl configuration , the influence of these clustering forces reduced moving away from the cluster site . This caused a convergence towards the mean distances between homologues established for the No Tether condition . However , chromosome I is sufficiently short that even at its mid point ( where it bows towards the central nuclear volume ∼120 Kb from each telomere ) , the chance of close juxtaposition is higher compared to chromosomes with No Tether and Telomeres Tethered ( Fig . 6 ) . For flexible chromosome XVI the tight bouquet also brought the entire length of the chromosome pair into closer proximity . At the mid point of the chromosomes , the distance between alleles was ∼0 . 37 ND , where it converged on the distances recorded for No Tether and Telomeres Tethered with no clustering ( Fig . 7 ) . The longer chromosome IV pair gained close juxtaposition over a similar length to the chromosome XVI pair . Thus , up to ∼400 Kb from each telomere the distance between alleles was closer than for the same chromosome with No Tether ( Fig . 8; Loose and Tight Bouquet , Flexible compare yellow and white lines on graphs ) . While we do not know what would be a critical distance between alleles on homologues to define them as paired or not in meiosis , the implication is that longer chromosomes as a whole might benefit less from the bouquet formation than short chromosomes . Although as shown below , the measure of benefit in vivo would be dependent on the true persistence length of chromosomes . Our modeling of intrachromosomal contacts illustrates the importance of chromosome rigidity in defining trajectory through the nuclear volume . In meiosis it is thought that chromatin cycles through rounds of expansion and contraction , which presumably change their flexibility or contour length [3] . Such oscillatory changes could have an impact on chromosome pairing . Here we have considered two stable states of chromosome rigidity and analysed their impact on homologue juxtaposition . We found that making the chromosomes more rigid by increasing their persistence length caused an increase in the average distance between loci for all tethered chromosomes . The impact of increasing rigidity was more modest on the small chromosome I pair than the largest chromosome IV . Considering the Tight Bouquet , for chromosome I the rigid condition increased the range of separation between alleles from 0 . 12- to 0 . 23-times ND for the flexible chromosomes to 0 . 16- to 0 . 29-times ND ( Fig . 9; Tight Bouquet compare rank scores , chromosome I flexible and rigid ) . For the chromosome IV pair the range of distances between alleles increased from 0 . 12- to 0 . 38-times ND for the flexible chromosome to 0 . 12- to 0 . 52-times ND for the rigid chromosome ( Fig . 9; Tight Bouquet compare rank scores , chromosome IV flexible and rigid ) . Even thought the mean distances for the rigid chromosome IV are wider than for the flexible chromosome IV , they are significantly lower than in the No Tether condition and Telomeres Tethered without clustering ( Fig . 9; p<0 . 01 , both a Kolmogorov-Smirnov and Wilcoxon rank-sum test ) . A potential benefit to pairing long chromosomes is the distribution of closer juxtaposition periodically along the length of the chromosomes ( Figs . 7 and 8 ) . We suggest this is due to boundary effects created by the combination of rigidity and length . Such periodicity could help pairing at distance from the tethered telomeres . This point illustrates the importance in accurate information about chromosome persistence length as it influences chromosome trajectory and the potential impact of the bouquet structure . The clustering forces we have used for centromeres in the Rabl configuration and telomeres in Tight Bouquet were equal . As modelled so far both the Rabl and Tight Bouquet configurations increase homologue juxtaposition relative to the hypothetical state of No Tether , and the Tethered Telomeres with no clustering forces ( Figs . 6 to 8 ) . On average , however , the bouquet can reduce further the overall distribution of mean distances between alleles ( Fig . 9 ) . We do not know if in vivo the degree of clustering at centromeres in Rabl or telomeres in bouquet are similar or significantly different . If in vivo centromeres are less tightly clustered in Rabl than telomeres in the bouquet , then the bouquet would produce even more advantage to chromosome pairing than this model suggests . On the other hand , if clustering in vivo is tighter at centromeres and this is not lost in movement to the bouquet then the bouquet may be more dispensable . It will be interesting to determine the in vivo relative clustering tendencies in these two polarised arrangements for a range of organisms . All eukaryotic genomes contain a degree of repetition of genomic DNA sequence and this is a potential source of problems during meiosis . Genetic studies show that chromosome pairing has the unwanted effect of increasing interhomologue ectopic contacts [58] , [59] , [60] . It therefore makes sense that there should be a counter pairing process to discourage ectopic interactions . We have illustrated the distances between nominal ectopic loci on chromosome I and IV homologues by plotting interhomologue LDMs ( Figs . 10 and 11 ) . The colour code chart indicates the relative distances between interhomologue sites as a proportion of nuclear diameter ( Fig . 10A ) . The diagonal on the LDMs represents distances between alleles , with all off diagonal colour representing ectopic distance between nonallelic loci . The mean distances between alleles were rank scored for comparison with the mean ectopic distances , which were rank scored and then grouped into 300 bins . A few landmark examples of LDM areas representing potential interhomologue ectopic interactions are indicated ( EI , EII , EIII ) . In the various conditions used , and for both flexible chromosomes I and IV , the trends in change in proximity of nonallelic loci on homologues mirrors that seen for allelic loci ( Figs . 10 B and 11 ) . In the condition of No Tether the colour range in the interhomologue LDMs along the diagonal is similar to that off diagonal . Viewing the mean distances as rank scores shows a high degree of overlap for allelic and ectopic distances . The Rabl configuration causes the homologous chromosomes to be more aligned increasing the register between allelic loci . Thus the off diagonal colours on the interhomologue LDMs are on average cooler than on diagonal . As the centromeres have a strong tendency to be anchored and clustered , the furthest distances are between the centromere and the long arm telomere ( Figs . 10 B and 11; Rabl , Flexible , EIII left border on LDM ) . The associated rank score graphs reveal the overall wider mean distances between ectopic loci compared to the mean distance between allelic loci . In the condition of Telomeres Tethered all four telomeres can be widely separated on the nuclear envelope , and therefore there is little linear register between homologues . This causes the mean distances between ectopic loci and allelic loci to be more similar than in the Rabl configuration . As the bouquet becomes progressively tighter , the difference between mean ectopic and mean allelic distances increases . While by definition telomeres are close in Tight Bouquet ( Figs . 10 B and 11; LDMs for Tight Bouquet are warm at position EI ) , other ectopic regions are cooler than the diagonal ( regions EII and EIII ) . The rank score graphs for Tight Bouquet indicate that interhomologue ectopic mean distances are overall wider than allelic mean distances . As chromosome IV is longer than chromosome I , the potential gap between nonallelic loci on chromosome IV homologues is wider . As the telomeres are fixed on the nuclear periphery , it is not surprising that the greatest separation occurs between telomeres and interstitial regions ( Fig . 10 B and 11; Tight Bouquet region EII ) . This is consistent with genetic data from yeast that indicates interhomologue ectopic recombination between telomeres is more likely than ectopic recombination between telomeres and distant interstitial loci [58] , [59] . When the chromosomes are more rigid both allelic and ectopic mean distances increase compared to flexible chromosomes . This is demonstrated by the general change to cooler colours in the rigid interhomologue LDMs ( Figs . 10 B and 11 ) . The increase in distance associated with making chromosomes more rigid is greater for nonalleic loci , thus in Tight Bouquet ( rigid ) there is more yellow/blue in regions EII and EIII . For chromosome I the maximum of mean distances between allelic sites increased from 0 . 23-times ND for Flexible to 0 . 29-times ND for Rigid . The maximum of mean distances between nonallelic sites increased from 0 . 25-times ND for Flexible to 0 . 38-times ND for Rigid . This trend is clearly demonstrated by the rank score graphs in which the gap between allelic and ectopic scores is wider for rigid chromosomes . This phenomenon is also more exaggerated for the larger chromosome IV . The maximum of mean distances between allelic sites increased from 0 . 38-times ND for Flexible to 0 . 52-times ND for Rigid . The maximum of mean distances between nonallelic sites increased from 0 . 52-times ND for Flexible to 0 . 74-times ND for Rigid . Thus while the improvement in close homologue juxtaposition caused by the bouquet is less for the longest versus the shortest chromosome , the longest chromosome may benefit more from the wider differential between interallelic and ectopic distances , particularly when rigid . We next tested the degree to which chromosome tethering and the tendency for clustering forces impacts on the competition between allelic and ectopic interactions between heterologous chromosomes . Related dispersed sequences among heterologous chromosomes have the potential to compete for chromosome interactions , which should be limited to between alleles . Avoidance of physical proximity between heterologues at the pairing stage would contribute to reducing the risk of deleterious interheterologue ectopic recombination . The highly polarised bouquet and rapid telomere led chromosome movement of S . pombe chromosomes have long been proposed as a size sorting mechanism [73] . Our in silico model supports the view that the bouquet acts as a size sorter . We measured the pair wise distances between all 300 notional loci on each of our shortest and longest chromosomes ( i . e . 90 , 000 measurements ) . The distances have been plotted in interheterologue LDMs in Fig . 12 , for the shortest ( flexible ) and longest ( rigid ) persistence lengths used . For the flexible chromosomes in the condition of No Tether , the mean distances between heterologues are very similar to that between homologues ( Figs . 12 B; No Tether , compare ectopic and allelic rank scores ) . As the Rabl configuration increases the chances of all centromeres being close to each other , pericentric regions of heterologous chromosomes I and IV are more likely to be close in the conditions that model Rabl ( Fig . 12 B; Rabl , LDM is warmer around the centromeres ) . With increasing distance from the centromeres the distance between chromosomes I and IV increases more than the interallelic distances ( Figs . 12 B; Rabl , LDM is cooler moving away from the centromeres; in rank score graphs the distances are higher for interheterologue ectopic ) . This Rabl induced size sorting becomes lost in the Telomeres Tethered condition . Heterologues are further apart in Telomere Tethered than in No Tether ( Fig . 12 B; Telomere Tether LDM is cooler than No Tether LDM ) , but the mean distances between chromosomes I and IV are intermediate between the mean distances between homologues ( Fig . 12 B; rank score graphs ) . This supports the view that only tethering telomeres to the nuclear periphery might hinder the requirement to bias the proximity of homologues over the proximity of heterologues . Introducing an increased chance for telomeres to be close to each other as the bouquet develops re-establishes size sorting . In the Tight Bouquet heterologous telomeres will by definition have a tendency to be close to each other ( Fig . 12 B; Early to Tight Bouquet , LDMs are warmer near telomeres ) . Moving away from telomeres the distance between heterologues increases more rapidly than the mean distances between homologues , producing a greater separation ( Fig . 12 B; Early to Tight Bouquet , compare rank scores ) . Chromosome flexibility influences the degree to which the gap increases between heterologues , compared to the mean distances between homologues . In the Tight Bouquet , for flexible chromosomes the maximum distance between chromosome IV homologues is 0 . 39- compared to 0 . 53-times ND between heterologues . But , with rigid chromosomes the maximum for chromosome IV homologues is 0 . 53-times ND compared to 0 . 77-times ND for heterologues . This suggests that when the chromosomes are rigid the bouquet may be less effective at bringing longer chromosomes into juxtaposition , but it is more effective at separating them from short chromosomes . Taken together , the observations on large chromosomes suggest that for them the bouquet may be as important for disrupting unwanted ectopic interactions as fostering allelic interactions , particularly if they are relatively rigid . Chromosome flexibility is thought to fluctuate during prophase I , due to changes in chromatin compaction [3] . In particular , it is suggested that chromosomes would be more rigid during Leptotene than Zygotene [3] . If correct , this change in persistence length around the time of bouquet formation may be important to alternately separate heterologues and juxtapose homologues . The idea that the bouquet discourages interaction between heterologues is supported by genetic experiments in mutants unable to form the bouquet , as they show a 2-fold increase in ectopic recombination [59] , [60] . The model presented here argues that the attachment of telomeres to the nuclear envelope and a tendency to cluster them ( forming bouquet structure ) increases the chances of homologous chromosomes lying close to each other . The influence and potential contribution of the bouquet appears to be different for short versus long chromosomes . We suggest the bouquet could be a major contributing factor and possibly sufficient for pairing small chromosomes . Another important physical attribute of chromosomes that could influence their juxtaposition is rigidity . The model suggests that increasing rigidity has more of an effect on large chromosomes possibly helping to separate them from unwanted ectopic interactions . Importantly , increasing rigidity also reduces the chances of interheterologue ectopic interactions . The differential importance of the bouquet structure to short and long chromosomes is consistent with modeling of yeast interphase chromosomes , which showed chromosome length influences positioning in the nucleus [65] . This difference may explain why on the one hand the bouquet is important and well conserved , while not being absolutely essential to chromosome pairing . Another issue worth considering is that telomere attachment to the nuclear envelope may have a function independent of chromosome pairing . As telomere attachment reduces homologue juxtaposition , the movements and bouquet structure may be there to counter this effect . This model is the first one we are aware of that uses both telomere or centromere tethering to the nuclear periphery combined with a directional force applied to the tethered site , representing the effect of a microfilament network . While this work represents a significant improvement over current models available , it nonetheless has some notable limitations . In particular we have not considered excluded-volume interactions arising from other chromosomes or subnuclear structures such as the nucleolus [17] , [65] , [74] . Accounting for such excluded-volumes represents an obvious extension to this work , in particular by jointly modeling the entire genome ( see e . g . [63] ) . Additional extensions to the model could be to allow simultaneous tethering and clustering of centromeres and telomeres , as the Rabl configuration may not be entirely lost when the bouquet forms [24] , [65] . It will also be important to incorporate the rapid prophase chromosome movements [29] , [33] , [41] , [44] , the function of which is probably not restricted to bringing about bouquet formation . Introducing homology comparisons to bias homologue interactions will also be important to creating a more accurate model [46] . Including all of these additional factors however , will require a very significant increase in computer processing power . Further measurements to define better the nuclear and chromosome size changes that take place in meiosis are important to inform the modeling process . For example persistence length measurements vary considerably in the literature , and probably along the chromosome length [44] , [62] , [68] . There is some evidence in genetic data to suggest short chromosomes are more susceptible than long chromosomes to nondisjunction in mutants lacking telomere tethering [37] . We are keen to test more directly , the prediction that small chromosomes are more susceptible to loss of the bouquet than large chromosomes . With further refinements of the model we hope to determine if the physical constraints on chromosomes during pairing impact on which loci are more likely to recombine , and perhaps influence the genetic map and therefore evolution .
The behaviour of chromosomes has previously been investigated in terms of flexible or semiflexible polymers [75] , [76] , [77] . Unconfined worm-like chain ( WLC ) statistics , wherein the chromosome is modelled as a continuous polymer with parameterised stiffness , can also be used to model the statistical behaviour of chromosomes far from physical boundaries . Since unconfined WLC models admit closed-form solutions to many statistical measures of interest , including the expected distance between any two loci [78] , [79] , [80] , [81] , [82] , the WLC has become a standard model for investigating chromosome behaviour in silico and for inferring chromosome properties from in vivo observations [26] , [62] . In cell conditions , however , chromosomes are confined to move within the nucleus and by various structures contained therein . Additionally , at various stages throughout the cell-cycle and meiosis , chromosomes are observed tethered at , or close to , the inner nuclear surface rendering WLC treatments analytically intractable . Despite this intractability , useful properties may still be estimated for confined WLCs by discretising the chromosome into a series of loci , , connected via inextensible rods , and adopting a sample-based approach . In these coarse-grained representations the chromosome is described in terms of the three-dimensional positions of its N loci , and by the inextensible rods that connect them , , which may be related via: ( 1 ) The constraint , , ensures the inextensibility of the ith rod , and may be set using , with denoting the compaction-factor and the fully-extended contour-length of the chromosome . The statistical behaviour under steady-state conditions is calculated by considering the energy associated with a particular configuration , , and assuming a Boltzmann distribution: ( 2 ) where denotes the thermodynamic beta and any free parameters used to define the energy . For spherically confined and/or tethered chromosomes this energy is calculated as: ( 3 ) where the first term represents the bending-energy associated with a particular configuration , represents the bending-modulus of the chromosome ( where represents the persistence length ) , the unit vector of the ith rod and the dot-product between vectors and . The second term in Equation ( 3 ) represents the confining potential imposed upon each locus by the nucleus ( and other large nuclear structures ) and is typically assumed to correspond to hard-core confinement . The vector , , therefore contains the quintuple of parameters and . Additional terms may be included in Equation ( 3 ) to represent the fact that two loci cannot approach with a certain distance of one another ( excluded-volume interaction ) or approach within a certain distance of various nuclear structures . The inclusion of such terms , however , requires significantly more computation , and consequently were not included in our models . Notable coarse-grained models include studies in which interphase chromosomes are modelled as spherically-confined worm-like chains with excluded-volume effects [63] , and studies by [64] , who used Markov chain Monte Carlo ( MCMC ) procedures to investigate the statistical behaviour of spherically-confined interphase chromosomes when either the centromere or telomeres were tethered at , or close to , the nuclear periphery . Other approaches infer chromosome structure from G1-phase measurement of cross-linking by assuming that chromosomes correspond to flexible polymers with no excluded-volume interactions [68] , [83] . As far as we are aware no studies exist using coarse-grained , sample-based modeling of semiflexible chromosomes during meiosis , although some notable studies based upon scaling arguments exist [84] , [85] Besides chromosome tethering , a noticeable feature of nuclear architecture in vivo , particularly during meiosis , is the polarisation of chromosomes within the nucleus , wherein centromeres or telomeres are located to a limited region of the nuclear periphery . The polarisation of centromeres during early meiosis appears to require a degree of microfilament control [86] . Similarly , the ( transient ) polarisation of telomeres during the bouquet stage of meiosis appears to involve the directed motion of telomeres rather than random diffusion [46] , with further studies identifying a nuclear-hugging microfilament network as the likely source of this biased motion [26] . Taken together these results suggest that clustering of centromere/telomeres over the nuclear periphery is actively enforced rather than an emergent property of confined and tethered polymers , and must therefore be explicitly incorporated into models of chromosome behaviour . This may be achieved by including additional terms in the systems energy , representing the force imposed upon centromere/telomeres by a microfilament network . For the case in which centromeres are tethered and experience polarising forces we write for the system energy: ( 4 ) where denotes the position of the centromere in space . The functional form of in Equation ( 4 ) is parameterised as ( 5 ) which corresponds to the von Mises-Fisher ( vMF ) distribution with mean-vector , μ , and angular-variance ( or spread of the distribution ) ν . The von Mises-Fisher distribution represents a distribution over the surface of a sphere . When the angular variance ν = 0 , samples from a vMF distribution will be uniformly distributed over the surface of the sphere , whilst increasingly positive values for ν will result in samples increasingly clustered on the surface about a mean vector , μ . Here is chosen to correspond to the nuclear radius , and the constraint , ensures the centromere always lies on the surface of the nuclear periphery . The effect of substituting Equation ( 5 ) into Equation ( 4 ) is , therefore , to cluster centromeres ( over the nuclear periphery ) about a mean vector , μ , with angular variance that depends upon both ν and emergent properties of the first two terms in Equation ( 4 ) . Similarly , when telomeres are tethered to the nuclear periphery and clustered , the system's energy is calculated as: ( 6 ) where and denote the positions of the first and second telomeres respectively . The functional form of is chosen to correspond to a product of two independent von Mises-Fisher distributions: ( 7 ) It is important to note that within this model the vMF distributions ( with positive , nonzero ν ) only induce a clustering force upon either telomere and are not , in themselves , necessarily sufficient to induce clustering . The statistical behaviour of chromosome with polarising forces depends upon and emergent properties of the other terms in Equation ( 6 ) . For short , rigid chromosomes , for example , the clustering forces will tend to want to induce a folding of the chromosome , whilst the internal rigidity of the chromosome will want to promote a straight trajectory , with the overall behaviour of the chromosome depending upon the relative magnitudes of these two effects . The probability density associated with a particular configuration , , may be calculated by substitution of Equations ( 4 ) or ( 6 ) into ( 2 ) . Whilst these distributions are analytically intractable , it is possible to sample representative trajectories by adopting an MCMC procedure similar to that used in [64] . The polymer-statistic models outlined above have been implemented in a Matlab toolbox Markov chain Monte Carlo ( MCMC ) for meiotic chromosomes ( 3MC ) and used to investigate the influence of chromosome architecture upon locus proximity within S . cerevisiae . The 3MC package including front end graphical user interface ( GUI ) is available for download at http://wsbc . warwick . ac . uk/software/3MC/3MC . zip In all subsequent models , the nuclear diameter was set to 2 µm in accordance with previous observations of nuclear diameter in S . cerevisiae [66] , [67] , with the spindle pole body ( SPB ) aligned along the positive z-axis ( [0 , 0 , 1000] nm ) . Within the model the SPB has no physical influence on chromosome trajectories , but is used to set the direction of the centromere/telomere clustering i . e . , the mean parameter , μ , in the von Mises-Fisher distribution ( s ) are aligned to the SPB . Currently , four different levels of clustering have been implemented , as defined by the angular variance parameter , ν . These values were set by eye , by observing the level of clustering induced on independent samples from the corresponding von Mises Fisher distribution . Values ranged from ν = 0 , representing chromosomes in which centromere/telomeres are uniformly distributed over the entire nuclear surface [64] , through weak ( ν = 5 ) , intermediate ( ν = 20 ) , and strong ( ν = 50 ) , representing the case in which centromeres/telomeres experience a strong force acting to cluster them about the SPB . Three difference chromosome sizes were simulated , with the shortest modelled on yeast chromosome I ( ∼240 Kb ) , chromosome XVI ( ∼950 Kb ) and chromosome IV ( ∼1530 Kb ) . The three chromosomes were assumed to correspond to the 30-nm fibre [62] , [83] , resulting in an approximately 40-fold reduction in contour length compared to dsDNA [87] . Sample chromosome trajectories were generated using 3MC , with chromosomes represented as 300-locus WLCs with centromeres located at an appropriate distance for the chromosome lengths tested . The choice to discretise into 300 beads represented a trade-off between the ideal number ( ) and computational time . Specifically , in the ideal case the chromosome would be divided into an infinite number of segments at which point the continuous WLC and discrete models become equivalent . In practice , however , the statistical behaviour of the model was found to converge very rapidly as the number of links increased , and the choice of 300 loci was found to be a good approximation to for all chromosome sizes tested ( e . g . Fig . S8 ) . Additionally , the choice of 300 segments meant that , for 40-fold compaction , even the longest chromosome would be divided into approximately 40-nm segments ( close to the width of the 30-nm fibre ) whilst being sufficiently small enough to allow calculations to be performed on desktop computers in a reasonable time . The persistence length was varied over 2 increments in the range , capturing the behaviour of flexible and semiflexible regimes . The above range of values was chosen to cover the range of previous values inferred from experimental measurements , with the lesser value corresponding to that for interphase chromosomes [62] , and the more rigid value lying somewhere between this value and that observed for pachytene chromosomes [44] . In total , 10 million sample chromosome trajectories were generated for each model condition using MCMC procedures ( the 3MC package ) , with the first 3 million samples discarded for burn-in . The remaining 7 million samples were thinned by a factor of 40 , with the remaining samples used to empirically calculate the desired statistics , including the physical distance between allelic loci and between different loci on the same chromosome or on heterologous chromosomes . In order to allow investigation of a large range of chromosome architectures as well as a large range of model parameters , a number of simplifying assumptions were made . Specifically , the chromosomes were treated as line-like objects by ignoring the effects of excluded-volume interactions due to chromosome width/volume . In tests the influence of excluded volume ( intrachromosomal only ) was found to have negligible influence on the chromosome trajectories modeled in isolation ( Fig . S9 ) . This would not be the case if persistence lengths were very short ( e . g 30 nm; data not shown ) . In vivo the volume of unmodeled chromosomes and subnuclear structures creating prohibited areas [17] , [65] , [74] would have an influence on the measured trajectories . Consequently , our results represent a first look into the effect of tethering and clustering during meiosis , and provide a good foundation for future studies that will include excluded volumes from other chromosomes . The influence of external volume has previously been included in coarse-grained models of interphase chromosomes [63] . With well defined parameters , excluded volumes can be incorporated in future models using additional terms in Equations ( 3 ) or ( 4 ) . | Organisms store their genetic material in the form of chromosomes that must be replicated and shared out during cell division . In sexual reproduction the cell division , called meiosis , halves the number of chromosomes to form gametes . This halving requires a complex reorganisation of chromosomes . Each gamete receives one maternal or one paternal copy of every chromosome . This requires a pairing process between the maternal and paternal chromosomes of each type . Once paired the two chromosomes are organised in space to bias subsequent movement in opposite directions when the nucleus divides . How chromosomes pair is of great importance to understanding fertility , and manipulating chromosomes in crops species , for which it is desirable to breed in new genes to improve hardiness or yield . We have modelled chromosomes in 3-dimensions based on the experimental organism Saccharomyces cerevisiae . We used our model to ask if various physical features of chromosomes might influence their ability to pair . We found that binding chromosome ends to the nuclear wall and pushing those ends together helps to encourage pairing along the length of chromosomes . It has long been known this special chromosome organisation occurs in live cells , but the significance of it has been difficult to determine . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"computer",
"science",
"computer",
"modeling",
"genetics",
"biology",
"computational",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Modeling Meiotic Chromosomes Indicates a Size Dependent Contribution of Telomere Clustering and Chromosome Rigidity to Homologue Juxtaposition |
We used a multi-round , two-party exchange game in which a healthy subject played a subject diagnosed with a DSM-IV ( Diagnostic and Statistics Manual-IV ) disorder , and applied a Bayesian clustering approach to the behavior exhibited by the healthy subject . The goal was to characterize quantitatively the style of play elicited in the healthy subject ( the proposer ) by their DSM-diagnosed partner ( the responder ) . The approach exploits the dynamics of the behavior elicited in the healthy proposer as a biosensor for cognitive features that characterize the psychopathology group at the other side of the interaction . Using a large cohort of subjects ( n = 574 ) , we found statistically significant clustering of proposers' behavior overlapping with a range of DSM-IV disorders including autism spectrum disorder , borderline personality disorder , attention deficit hyperactivity disorder , and major depressive disorder . To further validate these results , we developed a computer agent to replace the human subject in the proposer role ( the biosensor ) and show that it can also detect these same four DSM-defined disorders . These results suggest that the highly developed social sensitivities that humans bring to a two-party social exchange can be exploited and automated to detect important psychopathologies , using an interpersonal behavioral probe not directly related to the defining diagnostic criteria .
Social interactions among humans reflect the execution of some of the most important and complex behavioral software with which humans are endowed . Consequently , we should expect the computations involved in human social exchange to be subtle and perhaps even difficult to expose and study in controlled settings . However , exposing these computations is crucial if we are to improve our characterization and understanding of normal human cognitive function and dysfunction . In recent years , the components of social exchange in healthy subjects have been probed using interactive economic exchange games [1]–[8] . These games typically involve two subjects interacting for one or multiple rounds through the exchange of monetary gestures to one another . For our purposes here , these games require three classes of computation be intact and functioning in the minds of the interacting subjects . They require that each subject can ( 1 ) compute norms for what is fair in each exchange , ( 2 ) detect deviations in monetary gestures that deviate from these norms , and ( 3 ) choose actions predicated on such deviations [9]–[15] . These experimental probes have been used previously in the area of behavioral economics and neuroeconomics , but here we show that the behavioral gestures elicited in the context of economic exchange games can be used to classify certain psychopathologies . The twist in our effort here is that we use a data-driven approach examining the reactions of the healthy partner as a kind of biosensor while playing an exchange game with a subject possessing a psychopathology . In this paper , we used a multi-round fairness game played by pairs ( “dyads” ) of interacting humans to extract behavioral phenotypes defined by the dynamics of play exhibited over the 10 rounds of a complete game [6] , [7] , [16] . The game we employ is called a trust game [17]–[19]; see Figure 1A . In the 10-round trust game , one player ( called the investor or the proposer ) is endowed with 20 monetary units and chooses to send some fraction to their partner ( called the trustee or the responder ) . The amount sent is tripled to on the way to the trustee . The trustee decides which fraction to return in response to the investor , thus each round is represented by two numbers: the investment fraction and the repayment fraction . All the rules are transparent to both players . The game is played for 10 rounds and the repeated exchanges allow the players to build models of what to expect from their partner providing that their capacity to sense , model , and respond to their partner's decisions is intact . In most of the dyads , the subjects were given no information about their partner and did not meet or speak to the partner before , during , or after the task . Following [16] , we also included “personal” dyads , in which the partners met before the task , were instructed together , and saw a picture of their partner during each round . The basic approach of this paper derives from our prior work showing that this same game elicits unique behavioral phenotypes when a game is played between a healthy investor and a trustee diagnosed with a range of DSM-defined disorders – Autism Spectrum Disorder ( ASD ) [20] , Borderline Personality Disorder ( BPD ) [6] , Major Depressive Disorder ( MDD ) , and Attention Deficit Hyperactivity Disorder ( ADHD ) [20] . In all these studies , we noticed that the behavioral differences affect not only the trustee , but also a healthy investor who plays with this trustee . A similar conclusion that a healthy subject is sensing the psychological nature of the opponent during play was obtained in a recent paper [21] , where it was shown that a subject can gauge the strategic sophistication of the opponent in repeated play of a complex stag hunt game . These observations suggested the hypothesis that the healthy ( or typical ) investor's behavior might be used to ‘read out’ features that could characterize the psychopathology group playing in the trustee role . This possibility was also suggested by the nature of the interpersonal interaction enforced by the game . In any multi-round interaction with another human , a player's choices are rather dramatically entangled with those of her partner . In addition , although the game is characterized by two numbers per exchange ( investment and repayment ratio ) , it does require players to have several cognitive capacities intact to accomplish a ‘normal’ exchange . These include short-term and working memory , sufficiently accurate models of what to expect from another human in this exchange , appropriate sensitivity to positive and negative social signals , and intact capacity to respond to such signals . Collectively , these observations support the basic hypothesis that humans bring highly developed social sensitivities to two-party interactions that might be profitably exploited as a biological “sensor” ( biosensor ) – first using a human proposer ( investor ) and later capturing this behavior in a computer agent .
We analyze the results of 287 dyads , in which healthy participants play against healthy trustees , as well as against the trustees that have four different disorders: ASD , BPD , MDD , and ADHD . Each subject played only one game . With the exception of some patients with BPD , participants with disorders were not medicated . A detailed description of the data is given in Table S1 . We sought to classify the dynamics using only the numbers exchanged in the game between players ( investment and repayment ratios ) , the number of “types” or styles of play ( number of clusters ) , and the functional dependence of the next investment on preceding investment and repayment ratios . In short , we sought a heavily data-driven approach . We extended a previously published method [22] , [23] to cluster available trust game data . This method uses a regression approach to the functional dependence that clusters individuals based on coefficients of the regression . This method has advantages over traditional clustering approaches: ( i ) the number of types in our population is estimated directly from the data , and ( ii ) classification uncertainty is captured by probabilities rather than categorical cluster assignments . An investor is not classified as either within or not within a cluster , but instead a probability of being in a cluster is computed . This allows us to identify clusters where a style of behavior ( a type ) is over-represented ( in comparison with what is expected by chance ) , under-represented , or neither ( see below for details of this calculation ) . The basic model is determined directly from the numbers exchanged by the two players during the game . We model the healthy proposer's investment at time as a function of preceding investment and repayment ratios . In this “black-box” , regression approach [22]–[23] , we assume that we can capture meaningful variations in types of investor play by using a regression model based only on previous investment and return ratios , in contrast to other approaches [21] , [24] which commit to more explicit models of how these values are used in mental processes to generate behavior . It is known that an arbitrary continuous function can be approximated , with any given accuracy , by a polynomial of an appropriate order . As a result , a widely used approach to describe such functions is to try polynomial dependence of increasing order . For a first order dependence of the current investment on previous investments and repayments the model is: ( 1 ) where indexes the subject and indexes the current round of the game . For a second order dependence of the current proposer investment on previous investments and repayments , this expression would accrue all possible second order terms in lagged investments and repayments , including terms of the type that describe interaction between investments and repayments . Such terms acknowledge that the current choice by the investor is entangled with their previous interactions with their partner . Although expression 1 depicts a first order dependence on previous investments and repayments extending back two rounds of the game , in this paper , we do not pre-commit to the exact functional dependence for the current proposer investment nor to the number of exchanges into the past that best predict the person's current decision . Instead , we assume a general polynomial dependence of the current investment ratio on previous investments and repayments , and determine the order of this polynomial dependence directly from the data . Similarly , we determine the number of rounds into the past required to predict optimally the person's current investment ratio from previous investment and repayment ratios . The details of this general approach follow . Formally , we model the behavior as a mixture of regressions . For a fixed order of polynomial dependence , a fixed look-back window , and a fixed number of clusters , we assume a single investor's data is given byHere , is an investor's ( 8-dimensional ) vector of investments ( we consider models looking back as many as two rounds; to make the models comparable we only consider 8 investments ) , is the model matrix of independent variables defining the regression ( all less than or equal to -degree monomials in lagged investments and repayments going back rounds ) , is the -th cluster's regression coefficients , is the variance of the error term in the -th cluster , is the weight assigned to the -th cluster , the multivariate normal , and is the identity matrix . This behavior model is applied to the data from the whole group , with the data itself determining both the appropriate subdivision into clusters and the regression coefficients within each cluster . We use the data augmentation approach [25] , defining latent variables which assign investors to clusters , to form the complete data . We then get the joint posterior of the parameters and the latent variables by combining the complete data likelihood with priors over the parameters . We choose for priorswhere Dir denotes the Dirichlet distribution and the inverse gamma distribution [26] . These are the same priors that were used by Houser-Keane-McCabe in their work [23] . As our independent variables all lie on the interval , we chose the prior variance of the coefficients to be proportional to this range . For the above model , we use a two-stage Gibbs sampling algorithm to estimate the parameters [27]: Start with initial parameters then repeat: Step 1: Sample allocations given : , where Mult is a multinomial distribution , and . Step 2: Sample given 's:for to Here , is the pooled investment data over cluster , is the pooled model matrix over cluster , and is the normal multivariate density with mean and covariance . The sequences of samples can then be used to estimate parameters . To avoid possible adverse effect of potential outliers on this Gaussian-based ( hence outlier-sensitive ) method , we check that the empirical distribution of the differences between the observed and predicted values is indeed consistent with the normality hypothesis . Finally , the optimal number of clusters , polynomial order , and look-back window can be determined by computing the marginal likelihood of each model ( see the Methods section for details ) and selecting the model with the largest value . The method described above identified 4 clusters . In terms of the relevant parameters , two rounds were found to be the optimal number of previous moves for predicting the influence of past investments and repayment ratios on the current investment ratio made by the investor . To connect our clusters to the DSM-IV phenomenology , we determined which groups of subjects defined by DSM-specific criteria were over- or under-represented in each cluster and the number of standard deviations by which they were over- or under-represented . The results of the clustering are shown in Figure 2 ( see Table S2 , Table S3 and Table S4 for a detailed description ) . Cluster 1 over-represents individuals with ADHD . Although 54% of these individuals would be expected to fall into this cluster by chance , 89% of them end up in this cluster . Cluster 2 significantly over-represents individuals with Autism Spectrum Disorder . By chance , 23% of these individuals should fall in this cluster; however , we see 44% of them in the cluster . In Cluster 3 , medicated and non-medicated individuals with Borderline Personality Disorder are over-represented . By chance , 15% of individuals from each group should fall into this cluster . However , 36% of medicated and 27% of non-medicated Borderline Personality Disorder individuals belong to this cluster . Cluster 4 should by chance represent 8% of individuals with MDD , but 20% of them fall into this cluster . The chi-square analysis confirms the statistical significance of this over-representation ( see Methods section ) . For two disorders , there are known scores describing its severity: for ASD , there is a score on the Autism Diagnostic Interview-Revised [28] Repetitive behavior subscale , and for BPD , there is a score on the Interpersonal Trust Scale [29] . In both cases , we found a statistically significant correlation between these scores and the probability of belonging to the corresponding cluster ( = percent match of the dyad in this cluster from 30 , 000 draws from the posterior ) : and for ASD ( Figure 3 ) and and for BPD ( Figure 4 ) . With the clusters defined as described , we sought to characterize the kinds of social gestures ( signals sent across rounds and between players ) that define them . In Figure 5 , we summarize the across-round social gestures for each cluster in terms of the regression coefficients for the investment and repayment ratios and the constant term ( see also Figure S1 and Figure S2 ) . We discuss the potential importance of these findings below , but here we summarize in Figure 5 the average social gesture of each cluster by plotting the average regression coefficients for each restricting the number of rounds back to two – the optimal number that predicts the investors next investment ratio ( Figure 5B ) . Notice that in Cluster 4 , the dependence is dominated by the constant term; this term reflects universally high investments . In Cluster 4 , investors playing subjects with major depressive disorder are over-represented . The other over-represented group in Cluster 4 are investors playing trustees that they meet before the game and whose pictures they see each round of the exchange . It is interesting to note that investors playing subjects with ASD end up over-represented in the same cluster ( Cluster 2 ) as investors playing subjects in an impersonal version of the game – where subjects do not meet nor see each other . The above results provide evidence that examining investor-side behavior provides a new kind of ‘readout’ for some important psychopathology groups studied under the probe of the multi-round trust game . The game itself , although simple ( in each round only two numbers are exchanged ) , requires a number of intact cognitive functions including working memory , short-term memory , the capacity to model and predict the partner's likely response , the capacity to sense deviations from these expectations , good a priori models of human trade instincts ( reflected by round one offers and responses ) , and so on . One value of this approach is that it utilizes a probe that is not directly related to the symptom lists that define DSM classifications , and therefore provides a possible alternative method of classifying some psychopathologies – or at least identifying or isolating some of their malfunctioning computations . To verify the robustness of the clustering algorithm we employed a previously described computer agent designed to play the trustee role . The possibility to design agents of this type was shown in our previous work [6] . The corresponding “-nearest neighbor” agents use the database containing the results of all the rounds of all the dyads . A healthy trustee agent , to describe how much to repay , looks at the vector of 6 previous choices ( last 3 investments and last 3 repayments ) and finds , of all the records with healthy trustees , situations in which corresponding previous choices were the closest ( in the Euclidean distance ) . Out of the 6 recorded outcomes of these closest situations , the agent selects one with equal probability . A BPD trustee agent similarly selects from dyads with a BPD trustee . These trustee agents were validated in [6]: in interaction with healthy human investors , the BPD agent was shown to reproduce accurately ruptures in cooperation normally observed when a healthy investor plays a BPD trustee . Such ruptures were not observed in healthy investors playing a healthy computer trustee . In our case , we need to supplement these agents with a similar investor agent that select the investment value based on the 6 closest dyads . Our hypothesis is that the same correlation with disorders can be detected by players playing against the investor agent . Since it was already shown that the trustee agents adequately describe the trustee behavior , we had healthy investor agent play either the healthy or BPD trustee agent in the trust task for ten rounds ( Figure 6A ) 1 , 000 times . These interactions were then assigned to the previously determined clusters using the posterior distribution of parameters generated from the analysis of the human dyads ( see details in the Methods section ) . Notably , interactions between the BPD trustee and healthy investor agent were statistically significantly over-represented by 7 . 19 standard deviations in Cluster 3 – the same cluster in which investors playing both medicated and non-medicated individuals with Borderline Personality Disorder are over-represented . On the other hand , interactions between the healthy investor and healthy trustee agents were not statistically significantly over-represented in this same cluster; see Figure 7 and Figure S3 . Thus , for BPD , the same correlation between the statistical clustering and disorders can indeed be achieved by using the investor agents ( For the ASD group , there were insufficient data ( ) to develop an analogous trustee agent and so no validation along this psychopathology was possible at this time ) .
Intuitively , one might expect the investment on the next round to be an interactive function of both previous investment and the repayment the investor received , rather than independent effects of each . However , our analysis shows that the optimal clustering corresponds to polynomials of order , i . e . , to the linear dependence ( 1 ) . This means that , contrary to this intuition , the second-order terms – in particular , interaction terms between investments and repayments ( such as ) – do not lead to a statistically significant improvement of the model's explanatory power . For patients diagnosed with a DSM-IV disorder , medication is an important potential confound . In our study , only some BPD patients were medicated . According to Figure 2 , both medicated and non-medicated BPD patients were statistically significantly over-represented in the corresponding Cluster 3 . Thus , the presence or absence of medication does not affect our classification . In this paper , we use a purely data-driven approach to data analysis . This approach is important from the foundational viewpoint , since it enables us , in particular , to further confirm the objective nature of the existing psychopathology classification . From the practical viewpoint , once this classification is established , we can improve the diagnostic efficiency if we explicitly use the known diagnoses in classification and regression analysis . For example , this may make it possible to find the markers that identify healthy subjects with superior discriminatory power .
Informed consent was obtained for all research involving human participants , and all clinical investigation were conducted according to the principles expressed in the Declaration of Helsinki . All procedures were approved by the Baylor College of Medicine Institutional Review Board . The game is described in the previous section . Healthy participants were invited to the Human Neuroimaging Laboratory at Baylor College of Medicine . Prior to playing the game , each participant was instructed they would earn between $20 and $40 , scaled by number of monetary units ( MU ) each player individually accrued . Following the game , each participant was compensated as follows: <68 MU = $20 , 68–133 MU = $25 , 134–200 MU = $30 , 201–300 MU = $35 , and >300 MU = $40 . We discarded 1 , 000 draws as burn-in , sampled 30 , 000 draws from the posterior , and assessed convergence using the Raftery-Lewis test [35] . We used the R Bayesian Output Analysis program to perform these calculations [36] . We repeated our analyses using 8 , 000 cycles total as per Houser , Keane , and McCabe [23] and 1 , 000; 3 , 000; and 5 , 000 cycles as burn-in and arrived at similar over-representation results . To check that the empirical distribution of the differences between the observed and predicted values is indeed consistent with the normality hypothesis , we normalize each difference by subtracting the sample mean of the differences from the corresponding cluster and then divide by the sample standard deviation of these differences . We then compute the sample skewness and the sample kurtosis of the collection of all these differences , and use Matlab's Jarque-Berra test to check normality . Normality has been confirmed with . Since the null hypothesis of normality is rejected when , our value of indicates a strong empirical support for the normality hypothesis . We used the Laplace-Metropolis estimator of the marginal likelihood [37] , as described in Houser , Keane , and McCabe [23] , to compare models with different values of the number K of clusters , order P of the polynomials , and the number D of past rounds on which the model depends . We did not include any results in which 2 of 3 samplers arrived at at least one empty type in the mode of the last 5 , 000 of 8 , 000 draws from the posterior . To maximize marginal likelihood ( i . e . , to find a posterior mode ) , we used component-wise optimization ( also known as conditional maximization or step-wise ascent; see , e . g . , p . 312 of [38] ) , the use of which is well-established for Bayesian problems such as maximizing the posterior mode , and arrived at the same answer when comparing the maximum log marginal likelihoods for different models . As a result , we concluded that the optimal model has clusters , a first order polynomial , and a dependence of ratios of investment on ratios of investment and return rounds into the past . We found that , in contrast to the simpler case described in [23] , our marginal likelihood values are sometimes fairly close to one another in many cases and thus , the results of comparing these values can potentially change if we repeat the same computational experiment . To makes sure that our selection of 4 clusters does not change , we supplemented the conditional maximization by the exhaustive analysis of all possible triples with up to 10 clusters , polynomials of order 1 to 3 , and a time dependence of 1 or 2 rounds into the past . For each such model , we used several samplers and got several values of marginal likelihoods; when we compare two models , we select the simpler one ( the one with fewest overall parameters ) unless the other one has a statistically significantly larger mean . Since for the same model , the distribution of marginal likelihood values is sometimes not Gaussian ( see Figure S6B ) , we could not use the usual t-test . Instead , we used the Wilcoxon rank-sum test at the 5% significance level [39] . The results ( shown on Figure S6A and detailed in Table S4 ) confirm that the model with = ( 4 , 1 , 2 ) is optimal . To check whether the observed over-representation of participants with disorders in different clusters is statistically significant , we apply the chi square test corresponding to a null hypothesis that the participants of different disorders are randomly distributed in different clusters . Let be the number of elements in the -th cluster , the number of elements of -th group in this cluster , the cluster corresponding to the group , and the ratio of group in the population as a whole . Under the null hypothesis , due to the central limit theorem , the value is asymptotically normally distributed , with mean and variance . Thus , the ratio is normally distributed with mean and variance . Thus , to test the null hypothesis , we can form the test statistic , where is a relative over- ( under- ) representation of the group in the cluster . For , the null hypothesis is rejected with when . Thus , when each of the four terms in the sum satisfies the inequality , the null-hypothesis is rejected . We therefore consider the groups which are over-represented at the level . Please note that when , already the over-representation of the group in cluster is statistically significant with . Our clustering is based on iterations of Gibbs sampling . Every additional vector ( e . g . , of agents playing ) is then classified as follows . For each recorded iteration of the Gibbs sampling ( after the burn-in ) , based on the recorded values of and , we compute the probabilities of belonging to different clusters ( we use the same formula as in the subsection “Estimating the parameter” ) . Then , we select a cluster with the probability . After all these selections , we assign the dyad characterized by the vector to the cluster to which , among all the iterations , this vector was assigned the largest number of times . | Human social interaction is exquisitely complex , and perturbed social interaction is a hallmark of psychological pathogy . When someone has a psychological disorder the focus is generally on their behavior , but this behavior is rarely something displayed in isolation and typically induces profound changes in the people interacting with the disturbed individual . In this work we asked if the behavior of one person in a simple two-person economic exchange game is sensitive to features that could classify the pathology of their partner . We analyzed a large group of previously recorded interactions involving healthy persons and people diagnosed with a variety of psychological disorders , and found that a healthy person's behavior is indeed quantitatively and systematically influenced by their partner's pathology . These results could ultimately lead to a different way of understanding and diagnosing psychological disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"mathematics/statistics",
"neuroscience",
"neurological",
"disorders/neuroimaging",
"neurological",
"disorders/neuropsychiatric",
"disorders"
] | 2010 | Biosensor Approach to Psychopathology Classification |
Adult-onset hearing loss is very common , but we know little about the underlying molecular pathogenesis impeding the development of therapies . We took a genetic approach to identify new molecules involved in hearing loss by screening a large cohort of newly generated mouse mutants using a sensitive electrophysiological test , the auditory brainstem response ( ABR ) . We review here the findings from this screen . Thirty-eight unexpected genes associated with raised thresholds were detected from our unbiased sample of 1 , 211 genes tested , suggesting extreme genetic heterogeneity . A wide range of auditory pathophysiologies was found , and some mutant lines showed normal development followed by deterioration of responses , revealing new molecular pathways involved in progressive hearing loss . Several of the genes were associated with the range of hearing thresholds in the human population and one , SPNS2 , was involved in childhood deafness . The new pathways required for maintenance of hearing discovered by this screen present new therapeutic opportunities .
Hearing loss is a very common disorder with a significant social impact , including delayed speech and language development , reduced academic achievement , increased social isolation , and risk of depression , and has recently been reported to be a major risk factor for dementia [1] , adding new impetus to the need to develop therapies . Approximately 1 in 850 children are born with permanent hearing impairment in the United Kingdom [2] , and the number of people affected by adult-onset hearing loss increases with each decade of life , with 60% of people in their 70s having a hearing loss of 25 dB or worse [3] . Environmental factors including noise or drug exposure play an important role in its etiology , but there is also a strong genetic contribution . Over 360 genes are known to be involved in human or mouse deafness , but ascertainment bias has led to many of these having early developmental effects , and little is known about the genetic contribution to adult-onset hearing loss . We set out to identify further genes underlying deafness , including those with mild effects , using a physiological screen based on the auditory brainstem response ( ABR ) in a large cohort of newly generated targeted mouse mutants . In this report , we review the findings from this screen and present the new data; both mice and ABR waveform data are available for further analysis . From the unbiased sample of 1 , 211 genes tested , we found 38 unexpected genes to be involved in hearing impairment . This indicates that around 600 additional genes remain to be found ( see later ) , making deafness an extremely heterogeneous condition , with around 1 , 000 genes that may contribute . The observed impairments ranged from mild to profound , including several with progressive hearing loss , and with a wide range of underlying pathological mechanisms . The 38 genes represent a range of functions from transcription factors and a microRNA to enzymes involved in lipid metabolism . Eleven were found to be significantly associated with auditory function in the human population , and one gene , SPNS2 , was associated with childhood deafness , emphasising the value of the mouse for identifying genes and mechanisms underlying complex processes like hearing .
We used a rapid ( 15-minute ) , noninvasive electrophysiological test , the ABR [4] ( S1 Fig; S1 Data ) in anaesthetised mice aged 14 weeks old as part of an extensive pipeline of phenotyping tests on a set of new mouse mutants generated from targeted embryonic stem ( ES ) cells [5 , 6 , 7] . The allele design was mostly the knockout first , conditional-ready ( tm1a; targeted mutation , first allele with design type a [6] ) allele , which reduced or eliminated expression of the targeted gene by inclusion of a large cassette designed to interfere with transcription , but a few were the derived tm1b allele with an exon deleted or were edited alleles ( S1 and S2 Tables ) . A total of 1 , 211 genes were tested . Of these , 38 genes with no prior association with deafness had raised thresholds for detecting a response to sounds ( Fig 1; a small number of these have been published recently , after their discovery in the screen ) . Using objective criteria ( see Materials and methods ) we classified these 38 genes into five main groups based on thresholds: 5 showed severe or profound deafness , 10 had raised thresholds at high frequencies only , 2 showed raised thresholds at low frequencies only , 7 had moderately raised thresholds across frequencies , and 14 had a mild hearing impairment ( Fig 1 ) . In addition to these 38 unexpected genes , 10 known deaf mutant lines were tested as positive controls ( S2 Fig A-J; S2 Data ) , and 9 genes known to be involved in mouse ( Srrm4 ) or human deafness ( MYO7A , MYO15 , USH1C , WHRN , ILDR1 , ESPN , CEP250 , CLPP ) were tested in newly generated alleles; all showed raised thresholds ( S1 Table; S2K–S2T Fig; S2 Data ) . Of the set of 1 , 211 genes , 3 . 14% were new associations with raised auditory thresholds . As the genes targeted were an unbiased set showing no significant enrichment for any functional class compared with the total set of mouse genes ( see Materials and methods ) , we can extrapolate to estimate that over 600 further genes required for normal auditory thresholds remain to be found . Added to the 362 human and mouse genes already known and 38 reported here , this indicates that there may be as many as 1 , 000 genes involved in deafness , a very high level of genetic heterogeneity . There were several targeted genes screened that we expected to show raised thresholds because they had previously been reported to underlie deafness in either humans ( GSDME/DFNA5 , MYH14 , MYH9 , PNPT1 , PRPS1 , CHD7 ) or mice ( Barhl1 , Fzd6 , Hmx3 , Nfkb1 , Sgms1 , Sms , Synj2 ) . However , the new alleles had normal ABR thresholds ( S3 Table ) . The lack of raised thresholds could be due to incomplete knockdown of targeted gene expression in the tm1a allele ( e . g . , raised thresholds of Selk mutants were seen only in the tm1b allele , not in tm1a ) ; onset of hearing loss after 14 weeks , when we carried out the screening; screening was carried out on heterozygotes due to reduced homozygote viability; the genetic background may have influenced phenotype expression ( e . g . , Chd7 , where the new allele was not viable on the C57BL/6N background ) ; or the original deafness may have been the result of a specific effect of the mutation on the protein rather than a consequence of reduced expression ( e . g . , GSDME ) ( S3 Table ) . Alternatively , the original allele might have led to deafness via a long-range cis effect on a nearby gene , as in the Slc25a21tm1a ( KOMP ) Wtsi targeted mutation , which causes deafness by reducing expression of Pax9 [8] . Thus , we probably missed additional genes involved in hearing loss , so our calculation of 600 more genes awaiting association with deafness may be an underestimate . All mutant mice reported here are available via public mouse repositories for further investigation to explore these alternative explanations , and the unprocessed ABR data are available at the Dryad repository: http://dx . doi . org/10 . 5061/dryad . cv803rv [9] . Of note , all of the mutant lines that we have studied further following the initial screening results have shown raised ABR thresholds , even those in the mild class , suggesting that the screen has produced robust , reproducible calls ( mutant lines studied further: Spns2 , Zfp719 , Ocm , Klhl18 , Wbp2 , Pex3 , Acsl4 , Gpr152 , Mcph1 , Slc25a21/Pax9 , Ywhae , Lrig1 , Klc2 , Usp42 , Srsf7 ) [8 , 10 , 11 , 12] . As mouse and human inner ears are very similar in structure and function ( e . g . , [13] ) , the newly identified mouse genes represent good candidates for involvement in human deafness . A child from a United States clinic detected through clinical whole exome sequencing inherited a frameshift mutation of SPNS2 from her father ( c . 1066_1067delCCinsT: p . Pro356Cysfs*35; CADD phred score 26 ) and an in-frame deletion of a serine codon in SPNS2 from her mother ( c . 955_957delTCC: p . Ser319del; CADD phred score 20 . 9 ) . Neither variant has been reported in the gnomAD database ( Genome Aggregation Database; accessed January 2019 ) . Visual reinforcement audiometry at two years old revealed moderate to moderately severe hearing loss between 250 Hz and 4 kHz with no response at 8 kHz in the right ear , and severe hearing loss sloping to profound deafness from 500 Hz to 8 kHz with no response at 4 and 8 kHz in the left ear . Bone conduction testing indicated moderate-severe hearing loss at 2 kHz in the right ear , suggesting a sensorineural ( not conductive ) impairment , and acoustic reflexes were absent . However , the child had surprisingly good sound localisation performance . The severe level of hearing impairment associated with predicted damaging SPNS2 variants is similar to our findings in the mouse Spns2 mutant ( Fig 1 in [11] ) . Furthermore , we have previously reported that deaf children from two families in a Chinese cohort carried recessive WBP2 mutations [10] . We asked if these new candidates had any role in hearing ability in the general population by a candidate gene association analysis . We tested genomic markers within 0 . 1 Mb up- and downstream of each gene for association with auditory thresholds measured at age 44–45 in 6 , 099 individuals born during one week in the UK 1958 British Birth Cohort , using genetic data imputed to the 1 , 000 Genomes dataset [14] . Eleven of the thirty-seven candidate genes tested ( including SPNS2 ) showed a significant association of markers with threshold at either 1 or 4 kHz or both frequencies ( Table 1 ) , indicating that these 11 genes may play a role in normal variation of hearing ability in the human population . These findings emphasise the value of the mouse in resolving complex human diseases , because very few significant markers have been reported to be linked to hearing through genome-wide association studies of adult-onset hearing loss directly in humans: GRM7 [15]; PCDH20 and SLC28A3 [16]; ISG20 or ACAN and TRIOBP [17] . Several of the new genes that we found to be involved in deafness in the mouse had human orthologues close to or within unidentified non-syndromic deafness loci ( S4 Table , column J ) and so are good candidates for further exploration . These were USP42 within the DFNB90 interval , BRD2 close to the DFNA21 and DFNA31 loci , CAMSAP3 close to the DFNA57 region , and MCPH1 very close to the DFNM2 marker reported . The 38 genes newly associated with hearing are involved in a broad range of functions , including transcriptional and translational regulation , chromatin modification , splicing factors , cytoskeletal proteins , membrane trafficking , calcium buffering , peroxisome biosynthesis , thyroid hormone generation , ubiquitination and deubiquitination , kinases , signaling molecules ( including Wnt signaling ) , and proteins with no known or predicted function ( S4 Table ) . A microRNA gene , Mir122 , was one of the new genes underlying hearing impairment . Seven of the gene products have a role in lipid metabolism: Fads3 is a fatty acid desaturase; Agap1 and Zcchc14 bind phospholipids; Klc2 transports phosphatidylinositol 3-kinase , which is required for phospholipid processing; Pex3 is involved in biosynthesis of peroxisomes , which are involved in lipid processing; Acsl4 is a long-chain fatty acid coenzyme A ligase converting free long-chain fatty acids into fatty acyl-CoA esters; and Spns2 is a transporter of sphingosine-1-phosphate , a key intermediate in sphingolipid metabolism with a role in signalling . We carried out a GOSlim ( high-level version of Gene Ontology ) analysis to ask if the genes newly associated with hearing impairment ( n = 38 ) showed a similar distribution of gene ontology features to the group of genes previously known to be involved in deafness ( n = 362 ) , and to compare them with all genes tested in this screen ( n = 1 , 211 ) and a group of genes associated with ABR waveform defects ( n = 27 , described later ) . The novel genes showed nothing notably different to the full set of genes screened ( Fig 2 ) . However , nearly 70% of previously known genes had a Gene Ontology ( GO ) annotation for developmental processes , which was much higher than in the other groups analysed , suggesting an ascertainment bias for deafness due to early developmental defects in human and mouse . This finding suggests that our unbiased screen for new genes involved in hearing impairment at all levels of severity has revealed a fundamentally different class of genes compared with previously known genes underlying deafness . Some of the 38 genes had links to existing pathways involved in deafness . For example , Duoxa2 is required for maturation and transport from the endoplasmic reticulum ( ER ) to the plasma membrane of Duox2 , also known to underlie deafness through its role in hypothyroidism , leading to retarded cochlear development and impaired hearing [18] . Spns2 is a sphingosine-1-phosphate ( S1P ) transporter , and our discovery of its involvement in deafness supports the role of the S1P signaling pathway in hearing loss , alongside reports of S1PR2 and Sgms1 mutations causing deafness [19 , 20 , 21 , 22 , 23] . In contrast , many of the other genes discovered in this screen , such as A730017C20Rik ( Minar2 ) , have no demonstrated role in a biological process and no a priori reason to predict they might be involved in deafness . Although some of the 38 new genes identified by this screen showed strong expression in cochlear hair cells ( S4 Table , column Q; https://gear . igs . umaryland . edu ) , in general expression levels did not show a strong correlation with our ABR findings . As our goal in carrying out the screen was to identify new genes involved in adult-onset hearing loss , we carried out recurrent ABR recordings ( usually at 4 , 8 , and 14 weeks old , plus shortly after the normal onset of hearing at 2 and 3 weeks old if thresholds were raised at 4 weeks ) on some of the mutant lines . Remarkably , several of the mutants we studied showed relatively normal early development of ABRs followed by progressive increase in thresholds . These included Srsf7 heterozygotes ( encoding a splicing factor; Fig 3A–3C ) , Gpr152 homozygotes ( a G-protein–coupled receptor; Fig 3D–3F ) , and Klc2 ( Fig 4A–4E ) , plus Spns2 [11] , Wbp2 [10] , Acsl4 , Zfp719 , Ocm , and Klhl18 homozygotes . The finding of multiple new genes underlying progressive hearing loss and/or impairment of responses to high frequencies ( Figs 1 and 3 ) is important because progressive hearing loss in humans is very common and often affects high frequencies first , yet we have few clues to the pathological molecular processes involved . The mouse alleles studied here are relatively severe in their effect on protein expression , but variants in the human population may have milder effects on protein function and lead to later onset of hearing loss . Importantly , the finding of genes involved in normal development but later deterioration of hearing identifies molecular pathways likely to underlie adult-onset progressive hearing loss in humans . We analysed further a subset of mutant lines and revealed a wide range of pathological conditions underlying hearing impairment . Two examples of contrasting phenotypes are the Klc2 and the Ywhae mutant lines . Klc2 mutants showed a progressive increase in ABR thresholds with age , mostly affecting low frequencies ( Fig 4A–4F ) , with a sensorineural ( not conductive ) pathology . Klc2 encodes kinesin light chain 2 , which , together with kinesin heavy chains ( encoded by Kif5 ) , forms the kinesin-1 motor complex , a microtubule-associated anterograde transporter . The allele design ( Fig 4Ki ) allowed us to use the gene encoding β-galactosidase ( LacZ ) as a reporter system , showing that Klc2 was expressed in the epithelial cells lining the cochlear duct and strongly in the spiral ganglion ( Fig 4Kii-iii ) . The middle ear and gross structure of the inner ear appeared normal . The endocochlear potential ( EP ) was maintained at a normal level even up to 6 months of age , but the anoxia potential in scala media was significantly less negative in these mutants , consistent with loss of hair cell conductance ( Fig 4I ) . At one month old , there was extensive loss of outer hair cell ( OHC ) hair bundles ( Fig 4Gi-vi ) , DAPI-stained OHC nuclei , and CtBP2-labelled presynaptic ribbons of OHCs primarily in the region that normally responds best to 12 kHz ( 60%–70% of the distance along the cochlear duct from the base ) , corresponding to the worst ABR thresholds ( Fig 4Hi-v ) . There were few signs of inner hair cell ( IHC ) degeneration , but the increase in threshold was larger than would be expected if only OHCs were affected , suggesting IHC dysfunction . Klc2 is involved in anterograde transport of PI3K , which mediates insertion of AMPA receptors at synaptic membranes [24] and GluR1/2-containing vesicles to axon terminals [25 , 26] . We found no abnormality in GluR2-labelled postsynaptic densities below mutant IHCs ( Fig 4Ji-vii ) , suggesting other transport systems must move this AMPA receptor to the membrane . Klc2 also interacts with Kcnma1 , the calcium-activated potassium channel ( BK channel ) that underlies the IK , f current required for very rapid responses of IHCs and contributes to protective efferent suppression of OHCs [27 , 28 , 29 , 30] . We found that labelling of Kcnma1 in IHCs was less extensive in mutants compared with wild-type controls ( at 12kHz; Fig 4Ji-vi ) , implicating Kcnma1 in the pathological mechanism; however , knockout of Kcnma1 leads to less severe loss of thresholds [31] , so this alone cannot explain the extent of dysfunction in the Klc2 mutants . Finally , kinesin-1 has been implicated in maintenance of the hair cell nucleus in its correct position by interacting with Nesp4 in the outer nuclear membrane [32 , 33] , and Nesp4 mutations lead to location of the OHC nucleus at the top of the cell and subsequent degeneration [34] . We did not find mislocalisation of OHC or IHC nuclei ( Fig 4Li-ii ) , suggesting this was not the mechanism underlying hearing loss in the Klc2 mutants , and that redundancy between kinesin light chains may compensate for loss of Klc2 in nuclear localisation . No disease-associated loss-of-function mutations of human KLC2 have been reported yet to compare with the Klc2 mutant mice , which show a complete lack of Klc2 mRNA ( Fig 4M ) . However , a human gain-of-function KLC2 mutation ( 216-bp deletion upstream of the coding region ) leading to increased KLC2 expression causes spastic paraplegia , optic atrophy and neuropathy ( SPOAN ) , a neurodegenerative disorder involving progressive axonal neuropathy [35] . In contrast , the Ywhae mutants showed increased thresholds across all frequencies associated with variable amounts of accumulated fluid and exudate containing inflammatory cells in the middle ear , suggesting predisposition to otitis media ( Fig 5A–5M ) . The middle ear mucosa appeared thickened with granulation tissue in sections ( Fig 5L ) , and scanning electron microscopy of the luminal surface showed an open Eustachian tube in mutants , but abundant clusters of goblet cells ( presumed to produce mucus ) with fewer ciliated epithelial cells , which would normally contribute to the clearing of excess mucus ( Fig 5M and 5N ) . The variability in thresholds between individual Ywhae mutants , relatively flat increase across all frequencies , and near-normal ABR waveform support a conductive hearing loss ( Fig 5A–5H ) . The surface of the organ of Corti looked normal ( Fig 5O and 5P ) but we cannot exclude a sensorineural component in some of the more severely affected mutants , possibly due to the impact of 129S5-derived alleles in the mixed genetic background , or an effect of persistent inflammation of the middle ear [36] . Ywhae , also known as 14-3-3ε , is a member of the highly conserved 14-3-3 phosphoserine/threonine binding family , which have many interacting partners and are thought to provide a scaffold allowing coordination of intracellular signaling [37 , 38] . Ywhae is normally widely and strongly expressed throughout the body [39] and within the cochlea [40] , and the mutation led to an absence of detectable Ywhae protein in homozygotes ( Fig 5R ) . Not surprisingly , Ywhae mutants showed a number of other abnormal phenotypes similar to those reported in another Ywhae mutant [41] ( see S1 Table ) , including reduced viability of heterozygotes on a C57BL/6N background ( 20 heterozygotes out of 192 offspring from heterozygote × wild-type matings at weaning , chi- squared p < 0 . 0001 ) and homozygotes on a mixed C57BL/6N and 129S5/SvEvBrd/Wtsi background ( 53 homozygotes at 2 weeks old from 685 offspring from heterozygous intercrosses , chi-squared p < 0 . 0001 ) . Ywhae homozygotes also showed reduced growth ( Fig 5S ) and a shortened skull ( Fig 5T–5V ) . Craniofacial malformations may affect Eustachian tube structure and function , leading to otitis media , but there are many other possible pathological mechanisms not yet explored . Similar fluid-filled middle ear and conductive hearing loss phenotypes were found in the Mcph1 mutant [12] and Slc25a21 mutants with reduced Pax9 expression [8] . A third distinct pathology found was a reduction in EP . Normally , a high resting potential in the cochlear endolymph is generated by the stria vascularis . This is necessary for normal sensory hair cell function . Progressive disorganisation of the stria vascularis accompanies the reduced EP in Spns2 mutants [11] . A fourth example is the Wbp2 mutant , in which abnormal structure of synapses between IHCs and cochlear neurons and swelling of nerve terminals leads to progressive increase in ABR thresholds [10] . The finding of a wide range of primary pathological processes in these mouse mutants as outlined above suggests that the pathogenesis of hearing loss in the human population may be equally heterogeneous . The limited information gleaned from human temporal bone studies supports the suggestion of heterogeneous pathophysiology underlying progressive hearing loss [43] . Many mouse mutants with deafness were originally detected , because they also had a balance defect leading to circling and/or head bobbing; thus , many of the earliest genes to be identified were those involved in early developmental problems such as gross inner ear malformations or sensory hair cell developmental abnormalities affecting both the cochlea and vestibular part of the inner ear . It is notable that none of the 38 new mutant genes we report here showed any sign of leading to a balance defect ( S1 Table ) . Nine lines had reduced viability assessed at postnatal day 14 , with three of these lines producing so few homozygotes that heterozygotes were passed through the phenotyping pipelines instead ( Brd2 , Srsf7 , and Setd5 ) . Six lines had either male or female infertility ( Pex3 , Mkrn2 , Herc1 , Camsap3 , Mcph1 , Usp42 ) , which is higher than the expected 5% based on larger panels of mutant alleles . Corneal or lens defects were observed in five lines ( Spns2 , Pex3 , Agap1 , Mcph1 , Usp42 ) . Occurrence of anomalous features in other systems tested were generally scattered across mutant lines and phenotypes , with Duoxa2 and Ywhae mutants showing the largest number of other abnormalities ( S1 Table ) . By analysis of click-evoked ABR waveforms , we identified 27 additional mutant lines with normal hearing sensitivity , but which had abnormal patterns of neural responses , such as smaller ABR wave amplitudes or prolonged latencies , determined using objective criteria ( S1 Table; Fig 6; S3 Fig; S3 Data ) . Ten further mutant lines from the 38 with ABR threshold elevation also exhibited abnormal ABR waveform shapes ( S1 Table; S4 Fig; S4 Data ) . The ABR waveform is a complex mixture of voltage changes reflecting the sum of excitatory and inhibitory activity at different times after stimulus onset and different physical locations within the brain relative to the position of the recording electrodes . Wave 1 reflects auditory nerve activity , and later waves reflect activity higher up the central auditory pathways . Some mutant lines had reduced wave 1 amplitudes ( Fig 6; S3 and S4 Fig; S4 Data; S1 Table ) , which may result from desynchronisation of the onset of firing in auditory nerve fibres [44] , or reduced numbers of auditory nerve fibres contributing to the ABR , or a selective loss of high spontaneous rate/low threshold neurons , which have maximal discharge rate at stimulus onset , or inefficient synaptic recovery during the short gap ( 23 . 5 ms ) between stimuli . Other mutants showed abnormal amplitudes or latencies of later waves , suggesting auditory processing anomalies in the central auditory system ( Fig 6; S3 and S4 Figs; S3 and S4 Data ) . These changes could reflect abnormal inherent excitability of auditory neurons or an alteration in the balance of excitatory and inhibitory inputs onto these neurons , resulting in increased or decreased discharge or synchrony . Mutants showing changes in latency are most easily explained by changes in neural conduction speed , alterations in synaptic delays , or changes in the relative contributions of different components in this complex neuronal pathway . Finally , in the most extreme example ( Bai1 , also known as Adgrb1 ) , the mice exhibited clear auditory-evoked responses and measurable thresholds , but the ABRs were so abnormal that it was not possible to determine the equivalent peaks to quantify and compare with control mice ( Fig 6 ) . This set of 27 mutants with waveform anomalies will be an interesting group to analyse further , because central auditory function is critical for normal sound perception . Such deficits may translate in humans to altered performance in sound localisation , ability to follow salient acoustic stimuli in background noise , discrimination of specific speech features , and other auditory processing disorders ( e . g . , [45] ) . We used a rapid ABR protocol to carry out a high-throughput screen of 1 , 211 new mouse mutants and revealed a new spectrum of functional deficits in hearing that would not have been detected using simpler screens , such as the startle response . These include mild-moderate degrees of hearing impairment , frequency-specific impairments ( low or high frequencies ) , and a group with abnormal ABR waveforms that likely have deficits in central auditory pathways . In a subset of the new mutant lines , we have examined other ages to establish the time course of hearing loss and investigate the pathophysiological mechanisms underlying the raised ABR thresholds . A broad range of pathologies was found , and many mutants showed normal development followed by progressive hearing loss . We have shown that some of the genes highlighted by this study play a role in human hearing , including 2 genes with mutations that can account for recessive deafness in families and 11 genes that are associated with variation in auditory thresholds in the UK 1958 British Birth Cohort cross-sectional population . Thus , mouse mutants can be an effective means to identify candidate genes for human deafness . This project has provided insights into the wide range of pathological processes involved in hearing impairment and has revealed a surprising number of unexpected genes involved in deafness , suggesting extreme genetic heterogeneity . For this reason , it is likely that therapies will need to be directed at common molecular pathways involved in deafness rather than individual genes or mutations . Each new gene identified gives insight into the metabolic pathways and regulatory processes involved in hearing and thus provides a rich source of targets for development of therapies for the restoration of hearing .
Mouse studies were carried out in accordance with UK Home Office regulations and the UK Animals ( Scientific Procedures ) Act of 1986 ( ASPA ) under UK Home Office licences , and the study was approved by the King’s College London and Wellcome Trust Sanger Institute Ethical Review Committees . Mice were culled using methods approved under these licences to minimise any possibility of suffering . For human studies , informed consent was obtained from the adult participants and the parents or guardians of children prior to participation , and the experiments conformed to the principles set out in the WMA Declaration of Helsinki and the Department of Health and Human Services Belmont Report . The US patient provided consent for clinical whole exome analysis and written consent for inclusion as a case report . Testing was conducted during the routine clinical care of a patient in the US; thus , in accordance with US law , this study is exempt from Institutional Research Board approval . Mutant mouse lines were generated using targeted mutations in mouse ES cells [5 , 6] . The viability of new mutants was determined by genotype distribution at weaning . When possible , mice homozygous for the targeted mutation were used for screening . If a mutation proved to be embryonic lethal or had significantly reduced viability at weaning , heterozygous mice were used instead ( 356 genes out of 1 , 211; 29 . 4% ) . In most cases , the knockout first conditional-ready ( tm1a ) allele for each gene was used , but a subset of the genes were tested using the derived tm1b allele , which had deletion of a critical exon ( s ) , or other mutations ( S2 Table , column B ) . The tm1a allele is designed to knock down transcription by introducing a large cassette into the gene , but not all genes were completely inactivated ( see S2 Table and column X in [5] , for some examples ) . Most mutant mice were screened on a C57BL/6 genetic background . This included some lines on a mixed C57BL/6Brd-Tyrc-Brd; C57BL/6N line and others on a pure C57BL/6N line ( S1 and S2 Tables , column D ) . Mutant data were compared with a large set of wild-type data on the same genetic background . When mice were screened on mixed genetic backgrounds ( for example , Ywhaetm1e ) , age and strain-matched wild-type mice were used alongside the mutants . Positive control lines known to have a hearing impairment were compared with their littermate controls on the same , varied genetic backgrounds ( S1 Table ) . At the age screened , 14 weeks , we found no apparent effect of the known Cdh23ahl allele carried by C57BL/6 mice on ABR thresholds [46] , and mutant responses were compared with mice of the identical genetic background . The Cdh23ahl allele may have interacted with any of the new mutations to exacerbate their effect , such that the phenotype was easier to detect at 14 weeks , so this screen could be regarded as a sensitised screen [47] . Details of the full phenotyping pipelines in the Mouse Genetics Project ( MGP ) have been reported elsewhere [5] . Two assays required mice to be immobilised and so to minimise stress to the mice , ABR testing was performed under anaesthesia immediately before the DEXA/Faxitron assays . Balance was assessed by observation of gait , head bobbing or circling , the rotarod test , or contact righting test . ABR testing was performed on three phenotyping pipelines used over successive time periods , termed MGP Pipeline 2 , Mouse GP , and MGP Select , respectively . ABRs were recorded in mice aged 13 weeks ( ±3 days ) on MGP Pipeline 2 , and at 14 weeks ( ±3 days ) on Mouse GP and MGP Select pipelines . Mice were maintained on a normal lab chow diet on MGP Pipeline 2 and MGP Select pipelines , but on a high-fat lab chow for the Mouse GP pipeline . ABRs were recorded using the methods described in detail in [4] . Mice were anaesthetised using intraperitoneal ketamine ( 100 mg/kg , Ketaset , Fort Dodge Animal Health , KS ) and xylazine ( 10 mg/kg , Rompun , Bayer Animal Health ) , or a 10% greater dose for the MGP Select pipeline . Mice were placed on a heating blanket inside a sound-attenuating booth . Subcutaneous needle electrodes were inserted in the skin on the vertex ( active ) and overlying the ventral region of the left ( reference ) and right ( ground ) bullae . Stimuli were presented as free-field sounds from a loudspeaker whose leading edge was 20 cm in front of the mouse’s interaural axis . The sound delivery system was calibrated using an ACO Pacific 7017 microphone . For threshold determination , custom software and Tucker Davis Technologies hardware were used to deliver click ( 0 . 01-ms duration ) and tone pip ( 6 , 18 , 24 , and 30 kHz of 5-ms duration , 1-ms rise/fall time ) stimuli over a range of intensity levels from 0 to 95 dB sound pressure level ( SPL , re . 5 μPa ) in 5-dB steps . Averaged responses to 256 stimuli , presented at 42 . 6/s , were analysed and thresholds established as the lowest sound intensity giving a visually detectable ABR response ( S1 Fig; S1 Data ) . Following completion of recording , mice were injected with intraperitoneal atipamezole ( 1 mg/kg , Antisedan , Pfizer ) to promote recovery . A fixed recording protocol was followed: 1 . A series of click-evoked ABRs were recorded , ranging from 0 to 85 dB SPL in 5-dB intervals . 2 . Tone-evoked ABRs were recorded for a fixed set of frequencies ( 6 , 12 , 18 , 24 , and 32 kHz ) over sound levels ranging from 0 to 85 dB SPL in 5-db intervals . Different SPL ranges were recorded for different test frequencies to improve the time efficiency of the recording process ( 6 kHz , 20 to 85 dB; 12 kHz , 0 to 70 dB; 18 kHz , 0 to 70 dB; 24 kHz , 10 to 70 dB; 30 kHz , 20 to 85 dB ) . Responses were recorded in an array , beginning with the lowest stimulus level , in decreasing frequency order before stepping up to the next ( 5 dB higher ) stimulus level . If mice appeared to have hearing impairment , the upper limit of SPLs was extended to 95 dB for each test frequency and for clicks ( representing the upper limit of the linear range of our sound system at these frequencies ) . For the ABR screen , we aimed to test a minimum of four mutant mice per line ( of either sex ) . For other tests on the pipeline , 14 mutant mice ( 7 males and 7 females ) were required . Phenotyping cohorts were issued as mice became available , such that several partial cohorts were issued , to achieve the required number of 14 mice for each single line . This allowed the ABR assay to pick up further mice from any lines that exhibited any features of interest to extend the number tested beyond the target of four . In addition to mutant mice , at least four wild-type mice from the same matings used to generate the mutants were tested each week from each core genetic background of the mutants tested . These wild-type results formed a local control group for comparison with the mutant lines and also contributed to a large reference range of control data that were used to determine if ABR results from a particular mutant line were significantly abnormal . We compared ABR thresholds measured in 1 , 142 wild-type mice ( female , n = 583; male , n = 559 ) from the pure C57BL/6N or mixed C57BL/6J and 6N genetic backgrounds . No sex differences were noted for mice of either genetic background ( S2 Fig U-X; S2 Data ) . We compared mutant data to a fixed population of control mice of the same genetic background ( Pipeline 2 , C57BL/6J and 6N mixed n = 201; Mouse GP , C57BL/6J and 6N mixed n = 951; MGP Select , pure C57BL/6N n = 742 ) . The reference range for each parameter was defined as a 95% range of control values , from the 2 . 5 percentile to the 97 . 5 percentile of the control population , shown as a distribution around the median of the control data . ABR responses were considered in three phases of analysis as follows: 1 . ABR threshold and hearing sensitivity . For each stimulus used ( click and five tone frequencies ) , ABRs recorded over the range of sound levels tested were plotted as a stack , ordered by increasing dB SPL ( S1 Fig; S1 Data ) . Threshold ( dB SPL ) was estimated by visual inspection of the stacked ABR traces as the lowest sound level at which any component of the ABR waveform was recognisable and consistent with responses recorded at higher sound levels , taking into account the characteristic lengthening of peak latency as threshold is approached . Thresholds for each stimulus were plotted to give a profile of the hearing sensitivity of each mouse . 2 . Waveform shape comparisons . Through consistent , reproducible electrode placements , it was possible to compare , qualitatively and quantitatively , the waveform shapes of click-evoked ABRs . Wave 1 is understood to reflect auditory nerve activity , but as the responses represent a complex mixture of responses detected at a single point , there is some uncertainty in ascribing specific brain locations to specific features of the ABR waveform . The free-field binaural stimulation conditions we used complicates interpretation further , because there are binaural interactions even within the cochlear nucleus , e . g . , [48] . The ABR represents the summed electrical vectors detected by the electrodes as synchronised action potential volleys ( particularly from onset-responding neurons ) traverse the central auditory pathways . As these pathways can be both excitatory and inhibitory , as well as both ascending and descending , and are distributed in a 3D volume , interpretation is complex . Waveforms recorded to clicks at 20 dB and 50 dB above threshold ( sensation level [SL] ) were plotted for mutant and control mice , along with an average of the ABR amplitude over time across mice for each genotype . In these responses , we could identify four waves ( positive to negative deflections; S1 Fig; S1 Data ) . We first performed a qualitative comparison of click-evoked waveforms recorded at 50-dB SL ( Fig 6; S4 Fig; S4 Data ) . The averaged mutant waveforms were compared with both the averaged control waveform and a 95% reference range of waveform amplitudes . We also compared individual mutant responses with the reference range . If both comparisons were in agreement between at least two of three experienced observers , a quantitative analysis was carried out of the peak amplitude , latencies , and intervals of these waveforms , by measurement of input-output functions ( IOFs ) . 3 . IOFs . Using click-evoked ABRs , waveforms were analysed in detail to determine the amplitude and latency of positive and negative peaks of the waveform at each stimulus level recorded ( S1 Fig; S1 Data ) . This was performed using software routines developed by Brad Buran and kindly donated for our use by M . C . Liberman ( Harvard University ) . We found wave 1 and wave 3 were highly consistent in control mice . However , whilst wave 2 was clearly present as a single peak at low sound levels , it often split into two components at higher sound levels , making analysis complicated , so we did not include it . From these measures , we calculated the peak-peak amplitude of waves 1 , 3 , and 4 , the amplitude of the N2-P3 component , and the intervals from P1 to P3 and N1 to N3 . IOF curves were plotted relative to click threshold for each mouse ( i . e . , parameter plotted against dB SL ) . IOFs of individual mice ( mutants and local controls ) were plotted , together with 95% reference range generated for each parameter for controls . Wild-type control mice of the same genetic background tested in the same week as mutants were used as a local control population . As mutant mice were often tested in separate weeks to obtain the required numbers , the local control population for each mutant line varied in numbers . We compared each mutant population of results with a 95% reference range obtained from a fixed large number of wild-type controls of the same genetic background . Control populations for ABR thresholds of C57BL/6N or mixed C57BL/6N and 6J mice were not normally distributed ( Shapiro-Wilk Normality test , p < 0 . 001 ) . Furthermore , the large disparity in population sizes of the groups invalidates the use of traditional statistical tests giving p-values ( e . g . , t tests or analysis of variance ) [42] . Thus , we used the following criteria to define parameters considered to be abnormal compared with controls . 1 . ABR thresholds . A mutant line was considered to have abnormal ABR thresholds if one of two criteria were met: ( 1 ) at least 60% of mutant mice had thresholds for any stimulus outside the reference range , or ( 2 ) if the mean mutant threshold for any stimulus was at least 20 dB different from the median of the reference range for that stimulus . Thus , thresholds could be considered abnormal if they were elevated above controls ( lower sensitivity ) or reduced below controls ( enhanced sensitivity ) . We did not find any mutants with enhanced sensitivity . 2 . Waveform shape . Click-evoked ABR waveform shapes were compared at 20 dB and 50 dB SL . These comparisons were used as a subjective triage step in the assessment of whether waveform shapes were normal or perturbed in responses from mutant mice . A dataset was considered potentially interesting if two experienced observers considered the waveforms to be perturbed . In these cases , peak amplitudes and latencies were determined and IOFs plotted . 3 . IOFs . IOFs were plotted for peak amplitude , latency , and also for wave 1–3 inter-peak interval as a function of dB SL . Due to the dependency of amplitude and latency on SPL , it is important to plot IOFs relative to stimulus threshold , so that any changes seen are not a result of variation in response threshold . Parameters for a mutant line were considered significant if the mutant mean value was outside of the reference range for at least 40% of the SLs measured ( i . e . , for at least 5/12 SLs , when comparing over a 60-dB suprathreshold range ) . Patterns of raised thresholds for ABRs were classified according to the following criteria: Severe-Profound , if no responses were detected ( up to 95 dB SPL ) for at least two adjacent frequency stimuli , for all mice of that genotype; High-Frequency , if thresholds were elevated at 30 kHz ( by >30 dB ) and thresholds were not elevated for at least one of the lower-frequency stimuli; Low-Frequency , if thresholds were elevated for 6 and 12 kHz and were normal for at least one of the higher-frequency stimuli ( with a minimum mean threshold elevation <15 dB ) ; Moderate , when thresholds were significantly elevated for at least four of the six stimuli tested ( with a minimum mean threshold elevation >15 dB ) ; Mild , when mean thresholds were elevated by 30 dB or less for up to three stimuli tested; Normal Hearing , when no stimuli produced altered thresholds . Methods used for histology , immunolabelling and confocal analysis , EP recording , and associated statistical tests used have been published elsewhere [10 , 11 , 12 , 20 , 42] . ABR thresholds were compared using the Mann-Whitney test . GO term enrichment was analysed using FuncAssociate v3 . 0 ( [53] http://llama . mshri . on . ca/funcassociate/ ) , based on 22 , 644 genes ( genespace ) , with 19 , 064 GO attributes , downloaded on 22 December 2015 . FuncAssociate was configured to exclude computationally predicted GO annotations ( Inferred from Electronic Annotation [IEA] evidence code ) and run against the 1 , 211 genes tested by ABR . Revigo was used to reduce the redundancy in GO terms classed as over- or under-represented ( by FuncAssociate ) by clustering significant terms into more representative subsets ( [54] http://revigo . irb . hr/ ) . Only 0 . 42% GO terms ( n = 80 ) out of 19 , 064 were overrepresented in our list of 1 , 211 genes , and 0 . 09% ( n = 18 ) were under-represented . The over- and under-represented GO attributes were not significantly clustered into distinct subsets ( visualised as TreeMaps ) . We also analysed this list of genes with the Reactome pathway database using Reactome V58 ( www . Reactome . org ) , released October 2016; 10 , 168 human reactions were organised in 2 , 069 pathways involving 10 , 212 proteins and 10 , 214 complexes . Of the 1 , 211 genes tested , 564 were not listed in Reactome ( 46 . 6% ) . Of the 38 new genes underlying increased ABR thresholds , only 12 were included in Reactome . Of the 27 genes associated with abnormal ABR waveforms , only 17 were included in Reactome . Of 362 deafness genes already known , 233 were found in the Reactome databases . The 647 ( 53 . 4% ) genes represented in the Reactome database out of the initial 1 , 211 did not produce any overrepresentation of any particular pathways ( false discovery rate probability , FDR > 0 . 839 ) . Thus , taking the GO term and Reactome analyses together , the 1 , 211 mouse genes targeted can be considered to be representative of the entire mouse genome . We used the GOSlim analysis tool on the MGI website ( http://www . informatics . jax . org/gotools/MGI_GO_Slim_Chart . html ) to compare the proportions of high-level GO terms associated with genes in each of five lists: new genes associated with raised ABR thresholds ( n = 38 ) ; previously known genes associated with raised thresholds , human and mouse ( n = 362 ) ; new genes associated with ABR waveform abnormalities ( n = 27 ) ; all genes screened by ABR ( n = 1 , 211 ) ; and all genes in MGI ( n = 33 , 395 ) . We excluded evidence code IEA . The proportions of genes labelled with each high-level GO term in each group are plotted in Fig 2 . The hearing-impaired child was ascertained through the Kaiser Permanente clinic , testing was performed at Ambry Genetics , and the mutations were identified in SPNS2 by sequencing of candidate genes as part of clinical whole exome analysis . Variant pathogenicity prediction was carried out as previously described [55 , 56] . The link to the mouse study was established through GeneMatcher [57] . The 1958 British Birth Cohort and the collection of hearing data and analysis have been described previously [58 , 59 , 60] . Participants were drawn from 17 , 638 individuals born in England , Scotland , and Wales in one week of March 1958 . Of the original cohort , 9 , 377 members were revisited by a research nurse for a biomedical follow-up in 2002–2004 . Hearing measures consisted of pure tone audiometry at 1 kHz and 4 kHz at age 44–45 years and were adjusted for sex , nuisance variables ( noise at test , nurse performing test , audiometer used in test ) , conductive loss , and hearing loss in childhood . DNA was collected from 6 , 099 individuals and genotyped on various Illumina and Affymetrix SNP chips ( for detail , see http://www2 . le . ac . uk/projects/birthcohort/1958bc/available-resources/genetic ) . These data were then imputed to the 1 , 000 Genomes haplotypes ( released March 2012 ) using MACH and Minimac . Measured SNPs with >95% call rate and Hardy–Weinberg p-value >0 . 0001 were included as the input set . In subsequent analysis , imputed SNPs with low imputation quality ( r2-hat < 0 . 3 or MAF < 1% ) were omitted . For the association analysis , hearing thresholds at 1 kHz and 4 kHz were log transformed and adjusted for sex , nuisance variables , and conductive hearing loss in childhood; they were analysed by performing a 1-df ‘per allele’ significance test for association between mean hearing threshold and number of minor alleles ( 0 , 1 , or 2 ) as described previously [60] . The Bonferroni-corrected p-value for this candidate gene analysis is p < 6 . 76 × 10−4 , based on testing 37 genes against two threshold frequencies ( 1 kHz and 4 kHz ) at a significance p-value of 0 . 05 , so all p-values listed are significant after correction for multiple testing . A p-value of p < 6 . 76 × 10−4 represents the top 0 . 06% and 0 . 08% associations of all approximately 9 million imputed variants for 1 kHz and 4 kHz , respectively . | Progressive hearing loss with age is extremely common in the population , leading to difficulties in understanding speech , increased social isolation , and associated depression . We know it has a significant heritability , but so far we know very little about the molecular pathways leading to hearing loss , hampering the development of treatments . Here , we describe a large-scale screen of 1 , 211 new targeted mouse mutant lines , resulting in the identification of 38 genes underlying hearing loss that were not previously suspected of involvement in hearing . Some of these genes reveal molecular pathways that may be useful targets for drug development . Our further analysis of the genes identified and the varied pathological mechanisms within the ear resulting from the mutations suggests that hearing loss is an extremely heterogeneous disorder and may have as many as 1 , 000 genes involved . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"ears",
"brain",
"alleles",
"middle",
"ear",
"otology",
"inner",
"ear",
"genome",
"analysis",
"hearing",
"disorders",
"genomics",
"head",
"otorhinolaryngology",
"brainstem",
"gene",
"ontologies",
"genetic",
"loci",
"deafness",
... | 2019 | Mouse screen reveals multiple new genes underlying mouse and human hearing loss |
Phylogenetic studies have largely contributed to better understand the emergence , spread and evolution of highly pathogenic avian influenza during epidemics , but sampling of genetic data has never been detailed enough to allow mapping of the spatiotemporal spread of avian influenza viruses during a single epidemic . Here , we present genetic data of H7N7 viruses produced from 72% of the poultry farms infected during the 2003 epidemic in the Netherlands . We use phylogenetic analyses to unravel the pathways of virus transmission between farms and between infected areas . In addition , we investigated the evolutionary processes shaping viral genetic diversity , and assess how they could have affected our phylogenetic analyses . Our results show that the H7N7 virus was characterized by a high level of genetic diversity driven mainly by a high neutral substitution rate , purifying selection and limited positive selection . We also identified potential reassortment in the three genes that we have tested , but they had only a limited effect on the resolution of the inter-farm transmission network . Clonal sequencing analyses performed on six farm samples showed that at least one farm sample presented very complex virus diversity and was probably at the origin of chronological anomalies in the transmission network . However , most virus sequences could be grouped within clearly defined and chronologically sound clusters of infection and some likely transmission events between farms located 0 . 8–13 Km apart were identified . In addition , three farms were found as most likely source of virus introduction in distantly located new areas . These long distance transmission events were likely facilitated by human-mediated transport , underlining the need for strict enforcement of biosafety measures during outbreaks . This study shows that in-depth genetic analysis of virus outbreaks at multiple scales can provide critical information on virus transmission dynamics and can be used to increase our capacity to efficiently control epidemics .
Highly pathogenic avian influenza ( HPAI ) viruses represent a major concern for public health and global economy , as outbreaks in the last decades resulted in vast socioeconomic damages and numerous human infections . Thanks to increasing availability of avian influenza virus sequence data and the development of new computational and statistical methods of analysis , phylogenetic studies have largely contributed to a better understanding of the emergence , spread and evolution of HPAI epidemics [1]–[3] . However , sampling of genetic data has never been used or dense enough to allow detailed studies of a single outbreak [4] . The rapid evolutionary dynamics of avian influenza viruses suggest that sufficient genetic diversity may be produced during an outbreak in poultry to permit the reconstruction of the inter-flock transmission network , providing important insights for the implementation of efficient control measures . Notably , such detailed genetic data could be used in combination with epidemiological data to study the dynamics of epidemic spread , as has been done for the 2001 food-and-mouth disease outbreak in the UK [5] . However , much remains to be learned about the way evolutionary processes , such as natural selection or reassortment , shape avian influenza virus diversity during an epidemic and how these processes could affect the inference of virus transmission dynamics [4] . We also expect that successful identification of inter-farm transmission pathways depend on the extent and structure of intra-flock and intra-animal viral genetic variation , but perhaps most notably on the size of the virus population bottleneck in the process of inter-farm transmission [4] . The epidemic of HPAI H7N7 in the Netherlands in 2003 represents a unique opportunity to study the epidemiological and evolutionary processes involved in HPAI transmission dynamics in detail . This epidemic started in the most poultry-dense area of the Netherlands ( Gelderse valley , Gelderland province ) on February 28 , 2003 . Despite implementation of control measures , the outbreak spread across the entire Gelderland area as well as in a contiguous central region with a lower density of poultry farms . New outbreaks were reported in April in the Limburg province , another poultry-dense area in the South of the Netherlands , in Germany and in Belgium [6] . A total of 255 Dutch farms became infected in a 9 weeks period , and more than 30 million birds were culled during the course of the epidemic [6] . The virus was transmitted to 89 people who were directly involved in handling of infected poultry [7] , including one veterinarian who died after developing acute respiratory distress syndrome [8] . Detailed data gathered during the epidemic ( e . g . location , date of suspicion and sampling , type of farm , culling date ) have been used to estimate epidemiological parameters characterizing this epidemic , notably the spatial range over which the virus spread between farms [9] . However , the transmission route between farms could not be resolved , leaving critical questions about the mechanisms of virus transmission and the efficiency of control measures unanswered . The H7N7 virus was sampled from the majority of the 255 farms infected , but , to date , only little genetic data have been published from this epidemic [8] , [10] . In this study , we present virus sequence data from 72% of the farms infected during the 2003 HPAI H7N7 epidemic in the Netherlands . Phylogenetic analyses were used to unravel the pathways of virus transmission between farms and between outbreak areas . In addition , we investigated the evolutionary processes ( substitution rate , selection pressure , reassortment etc . ) that were shaping the H7N7 genetic diversity . We also examined the within-flock viral sequence variation on selected farms using clonal sequencing to assess its impact on our phylogenetic analyses . Finally we discuss the implications of the obtained results on our knowledge of the evolutionary and epidemiological dynamics of avian influenza viruses and consequences for disease control .
Virus RNA was extracted from homogenized trachea tissue samples from dead chickens ( 5 chickens per sample ) obtained from 184 of the 255 farms infected during the H7N7 outbreak ( 72% coverage of the epidemic , Figure 1 ) . We could not process more samples due to logistical constraints , but we considered that this coverage was sufficient to reach the aims of this study . The viral sequence datasets consist of full-length sequences of the H7-hemagglutinin ( HA ) , N7-neuraminidase ( NA ) and basic polymerase 2 ( PB2 ) gene segments; preliminary analysis of five full viral genomes previously obtained from humans and chickens infected at early and late stages of the H7N7 outbreak ( available in public databases ) showed that these three genes contain the highest level of genetic diversity among the 8 gene segments ( data not shown ) . Farms are labelled from F1 to F255 , following the order of sample submission to the laboratory during the outbreak . Samples were selected for sequencing in order to cover the entire timeline and all areas of the epidemic ( Gelderland , Limburg , central area and southwest area; Figure 1 ) . Moreover , all farms infected within 7 days before the first report of infection in the Limburg area ( April 3 , 2003 ) were analysed in an attempt to find the source of this new outbreak . Details of location and date of sample collection , and GISAID accession numbers are listed for each sample in Table S1 . The HA , NA and PB2 sequences of the human fatal case ( A/Netherlands/219/03 , [8] ) were included in the final dataset . A total of 74 substitution sites were recovered in HA , defining 71 sequences among which 50 were unique in the dataset . NA was less polymorphic ( 59 substitution sites ) , but a strand of 52 to 74 nucleotides in the NA stalk region was also found deleted in 13 samples from the Limburg area , with a total of 7 different types of deletions , 3 of which resulted in a frame shift in the NA coding sequence ( Table S1 ) . In total , the complete NA sequence dataset defined 64 different genotypes ( 42 singletons ) . The PB2 sequence data had the highest number of polymorphic sites ( 81 ) , defining 64 different genotypes ( 38 singletons ) . The combination of the genetic data from the three genes permitted us to define farm specific genotypes for 141 out of the 184 farms ( 76% ) . The HA , NA and PB2 sequence datasets were found to be free of homologous recombination using Recombination Detection Program version 2 ( RDP2 ) [11] . Rates of nucleotide substitution and time of most recent common ancestor ( TMRCA ) of the HPAI H7N7 viruses were estimated separately for the three gene datasets using a Bayesian Markov Chain Monte Carlo ( BMCMC ) method [12] as implemented in BEAST [13] , using sampling dates to calibrate the molecular clock ( Table S1 ) . Bayes Factors ( BF ) [14] were used to select among strict and relaxed clock models of evolution [15] , and among demographic models of population growth . The relaxed uncorrelated exponential clock model associated with an exponential growth model fitted better the data ( Table S2 ) . The analyses showed that the mean substitution rate was very high for both HA and NA datasets ( 1 . 18×10−2 and 1 . 02×10−2 substitutions per site per year ( substitutions/site/year ) , respectively; Table 1 ) , whereas the estimated rate for the PB2 dataset was twice lower ( 0 . 54×10−2 substitutions/site/year ) . These estimates were associated with large 95% highest posterior density intervals ( HPD; Table 1 ) . TMRCA estimations showed that the origin of the HPAI H7N7 virus dated back to mid-January 2003 according to the HA dataset , and as far back as late December and late October 2002 for the NA and PB2 datasets , respectively ( Table 1 ) . Again , estimations from the NA and PB2 datasets were affected by large HPD intervals . Similar estimations of substitution rates and TMRCA were obtained with other sub-optimal clock and demographic models ( Table S2 ) , showing that these results are robust and not artefacts of the priors used in the Bayesian analyses . Phylogenetic trees of the HPAI H7N7 virus sequences were reconstructed for the three separate HA , NA and PB2 sequence datasets using Bayesian Inference and Maximum Likelihood methods ( Figure 2A–C , Figure S1A–C ) . For each gene phylogeny , multiple sequences could be grouped in clusters that were well supported statistically . Notably , we could identify 4 clusters of sequences present in all gene phylogenies ( Cluster I–IV , Figure 2A–C ) . Three of these clusters regrouped virus samples from farms infected in the Gelderland area only ( Cluster I , II , IV; in yellow in Figure 2A–C , Table S1 ) , whereas Cluster III included all samples from the outbreaks in the Central area ( blue labelling ) , the fatal human case , a sample from the outbreak in the southwest of the Netherlands ( green labelling; F238; Figure 1 ) , the sample from the most northern outbreak in the Limburg area ( red labelling , F222 ) , and some samples from outbreaks in the Gelderland area ( Figure 2A–C , Table S1 ) . To assess the inter-farm transmission network , we manually concatenated the HA , NA and PB2 sequences for all virus samples , and used this single alignment to construct a Median Joining phylogenetic network [16] with the program NETWORK [17] ( Figure 3 ) . The network obtained included all the most parsimonious trees , thus represented all the plausible evolutionary pathways linking the farm samples . The network showed that most virus sequences were grouped in multiple clusters of infection , including the 4 transmission clusters identified with the gene-specific phylogenetic analyses ( Figure 3 ) . Sequences within these 4 clusters were separated in average by 3–4 nucleotide differences , whereas 11–20 differences were observed between clusters . All clusters were connected at the base of the network by complex reticulations that rendered the relationships between the clusters hard to determine . In most cases , one virus sequence identified in multiple farms was at the origin of an infection cluster . All Limburg samples ( apart from F222 ) were grouped with Gelderland samples in 2 clusters that were separated by one single mutation step . These 2 clusters presented some chronological anomalies ( Figure S2 ) . Notably , the network showed that a virus strain from a farm ( F18 ) that had been culled a month before the first infection in Limburg was the closest ancestor to a group of 3 Limburg samples ( F192 , F204 and F206; Figure 3 , Table S1 , Figure S2 ) . Also , according to the network , 4 virus strains that emerged in Gelderland during the first 3 weeks of the epidemic ( F40 , F57 , F103 and F107 ) would have originated from a virus strain that infected farms in Limburg after the 5th week of the epidemic . We identified 15 pairs of farm samples that uniquely shared identical sequence genotypes , representing likely transmission events ( Table 2 , Figure 3 ) . Furthermore , we could identify 13 pairs of samples that were unambiguously connected in the phylogenetic network . Of these 28 likely inter-farm transmission events , 25 involved farms located in the same infected area , with a distance of 0 . 8–13 . 6 Km separating them ( Table 2 ) . The three remaining transmission events linked farms separated by much larger distances ( 31 . 3–84 . 4 Km ) . The two longest transmission events corresponded to a transmission from a farm in Gelderland to the index farm of Limburg ( F167–F191 , distance: 84 . 4 Km ) , and from a farm in the central area to a farm located is the southwest of the Netherlands ( F236–F238 , distance: 65 . 9 Km; Table 2 , Figure 1 ) . We assessed the selection pressures acting on the three genes by estimating the ratio of non-synonymous to synonymous nucleotide substitutions ( ω = dN/dS ) in the different datasets using in CODEML [18] . When averaged over all sites , all three genes were predominantly affected by neutral or purifying selection ( ω <1 ) , with PB2 under the strongest negative natural selection ( ω = 0 . 313; Table 3 ) . Additionally , likelihood ratio tests revealed that a model allowing site-specific positive selection pressure ( M2a in CODEML ) fitted significantly better than a model of nearly neutral selection ( M1a ) for the HA gene ( p = 0 . 031; Table 3 ) . A Bayes Empirical Bayes analysis [19] identified 7 amino acid sites in HA that were under positive selection ( residues 127 , 129 , 143 , 183 , 188 , 284 , 340 ) , although none of these sites were supported by a significant posterior probability value ( Pr<0 . 95 ) . We further tried to identify sites under positive selection in the three genes using the single-likelihood ancestor counting ( SLAC ) , the random effect likelihood ( REL ) , and the fixed effect likelihood ( FEL ) methods [20] . The SLAC and FEL methods failed to detect positive selection in any of the three genes . The REL method identified the same 7 amino acid sites in HA already detected with CODEML as under positive selection with high posterior probability values ( Pr>0 . 95 ) . In addition , it detected 7 amino acid sites under positive selection in the NA dataset ( residues 54 , 64 , 66 , 200 , 247 , 442 , 458 ) . The biological functions of most of these positively selected residues in the HA and NA molecules are not known . Only the A143T substitution , which introduces a new potential N-linked glycosylation site in HA , has previously been identified as being associated with enhanced virulence in avian hosts [21] . Also , three positively selected amino acid changes , A143T in HA , and T442A and P458S in NA , have been also detected in the human fatal case [8] , and have been shown to contribute in enhanced replication efficiency of the HPAI H7N7 virus in mammalian hosts [10] . The A143T substitution was found in virus samples from 27 different farms ( Table S1; Figure 3 ) . The T442A and P458S amino acid changes in NA were present in the majority of the farm samples ( 113 farms ) . Three other amino acid changes identified in the human fatal case and linked to enhanced replication in mammalian hosts ( E627K in PB2; N308S and A346V in NA [10] ) were not found to be positively selected . The N308S and A346V in NA were identified in 12 and 36 farm samples , respectively , as well as in the fatal human case ( Table S1; Figure 3 ) . The E627K in PB2 was not found in chicken samples . We found discrepancies in the phylogenetic relationship between the four identified transmission clusters in the HA , NA and PB2 phylogenies . Cluster III was closely related to Cluster IV in the HA phylogeny , but Cluster III was closely related to Cluster I in the NA and PB2 phylogenies , and Cluster IV closely related to Cluster II in the PB2 phylogeny ( Figure 2A–C ) . These discordances suggest that one or more of the transmission clusters originated from reassortment events . We further investigated the putative reassortant viruses using bootscan analyses [22] on a selected dataset ( n = 50 ) of manually concatenated HA-NA-PB2 sequences ( Figure 4A–B , Figure S3A–D; see methods ) . Results of the bootscan plot showed that Cluster IV was highly similar to Cluster III in the HA segment , but clustered with Cluster II in the NA and PB2 segments ( Figure 4A ) . The graph did not produce a clear-cut breakpoint between the HA and the NA-PB2 segments , probably because of the poor level of genetic diversity in some gene regions . We noticed that sequences grouped in the Cluster III and IV were all characterized by the presence of the A143T amino acid change in their HA gene ( Figure 3 , Table S1 ) . Removing the codon position 143 from the HA dataset resulted in the loss of support for the clustering of these two groups of sequences in the phylogenetic trees and the bootscan analysis , thus for the signal of reassortment ( Figure 4B ) . In addition , we also observed that the placement of the sequence of three Gelderland farm samples differed between the NA phylogeny and the HA and PB2 phylogenies ( F45 , F76 and F143; Figure 2A–C ) . None of the bootscan analyses performed on these three samples showed a significant signal for recombination ( Figure S3A–C ) . Similarly to the potential reassortant event detected for Cluster III and IV , the F45 , F76 and F143 sequences were characterized by the presence of positively selected amino acid changes in NA ( Table S1 , Figure 3 ) . To estimate the viral genetic diversity within hosts and within flocks , we performed clonal sequencing targeting an 850 bp portion of the NA gene ( position 57–908 ) on 6 farm samples ( 5 chickens per sample ) . We chose 4 samples ( F36 , F167 , F191 and F193; Figure 1 ) positioned at the base of the Limburg-Gelderland transmission clusters in the network ( within groups G8 and G9 in Figure 4 ) in order to further assess the origin of the Limburg outbreak and of the chronological anomalies detected in the network . We also performed clonal sequencing on the F26 farm ( Figure 1 ) , because two samples taken three days apart ( March 6 , and March 9 , 2003 ) were available for this farm , allowing us to assess changes in viral genetic diversity within a flock . A total of 50–54 clones with NA inserts were sequenced per sample ( Table 4 , Figure 5 ) . We performed an additional clonal sequencing analysis targeting a 570 bp portion of the HA gene ( bases 81–650 ) , but , due to poor cloning success , only 27 clones with HA inserts could be obtained from F191 and 12 clones from F193 . For all the 4 samples tested , the most abundant NA and HA sequence variant represented 17–76% of the obtained sequences ( Table 4 , Figure 5 ) . With the exception of the NA variants obtained for F191 , this dominant variant was identical to the sequence originally obtained and used in the general HA and NA datasets . The other identified sequence variants were usually present at low frequency ( most of the time only once ) and directly linked to the dominant variant , differing from it by 1–4 nucleotide substitutions ( Figure 5 ) . The clonal diversity of NA in F26 ( March 9 , 2003 ) and F193 , and of HA in F191 were characterised by the presence of another sequence variant at relatively high frequency and at the origin of low-frequency variants . The diversity of NA sequence variants was extremely high in F191 , with all variants but one containing a stalk deletion ( 18 different types of deletion were identified; Table 4; Figure 5 ) . A total of 168 nucleotide substitutions were recorded , among which 116 were non-synonymous substitutions ( Table 4 , Figure 5 ) . Only 5 out of 168 substitutions had already been identified in the HA and NA epidemic datasets . Notably , 14 of the 35 NA variants identified in F191 shared a mutation that had been only found in farm samples F44 , F192 , F204 and F206 , establishing a link that was missing between the index farm of Limburg and the three Limburg farms ( F192 , F204 , F206; red node Figure 5; Figure 3 ) .
Boender et al . [9] have previously performed a spatial analysis of inter-farm transmission using epidemiological data from the HPAI H7N7 epidemic . They showed that risk of transmission decreased with inter-farm distance and they could map higher-risk areas for the spread of the virus . However , the epidemiological data did not permit the resolution of the pathways of transmission between farms . Our results show that the analysis of viral genetic data can complement epidemiological studies , allowing notably the identification of clusters of infections and of specific farm-to-farm transmission events . The geographical position of the farms associated with the transmission clusters identified from the phylogenetic analyses is indicated in Figure 1 . Most of the farms of Cluster I are located geographically close to one another , suggesting that inter-farm virus transmission during the epidemic was at least partially caused by short distance air-borne movements of virus particles [30] . However , farms of cluster III showed a combination of aggregated and dispersed geographical location , whereas Cluster II and Cluster III were more dispersed within the Gelderland area . It is possible that such transmission between farms separated by 5–15 km occurred naturally , as avian influenza viruses could persist for long periods in the environment ( e . g . water or feather ) [31] , [32] . The poultry production system with its professional contacts could also have favoured virus spread during the epidemic . Notably , some operations of farms , such as egg collection , were resumed during the epidemic and may have played a role in virus transmission [33] . We estimated that the H7N7 virus was present in poultry maybe weeks before the first outbreak , so it is possible that the virus spread via a network of contacts formed by normal poultry operations across the Gelderland area before the implementation of the transport ban . Also , it has been recently shown that many humans involved in the control of the epidemic were infected by H7N7 in farms they visited [34] . Thus , H7N7 virus might have been transmitted between farms by infected people or by human-mediated mechanical transport . Analyzing the sequence of virus isolated from infected humans and the movements of people involved in control activities could help to determine whether human-mediated transport played a role in the inter-farm transmissions . Importantly , our results also suggest that a discrete number of long distance transmission events were at the origin of the virus spread into new areas , rather than a slow wave-like movement of the virus towards the south of the country . Notably , it is interesting to note that farm UN167 , a back-yard poultry farm , seems to be at the origin of the outbreak in the Limburg area . Conversely , Bavink et al . [35] showed with epidemiological models that back-yard poultry probably played a marginal role during the outbreak , suggesting that pre-emptive culling of this type of farm may not always be necessary . Our results suggest that these types of poultry farms should still be considered important in control strategies . This result and all other farm-to-farm transmission events identified should be considered with caution because 27% of the farms infected during the epidemic could not be sequenced . However , only few missing farms could still be at the origin of the infection in Limburg and all are located in Gelderland >50 km away from the index farm of Limburg ( Figure 1 ) . It strongly suggests that a long distance transmission event is at the origin of the second important H7N7 outbreak in the Netherlands . Such long distance movements of avian influenza are most likely the results of human-mediated transport of the virus , although airborne spread cannot be totally ruled out . Possible causes of human-mediated virus movements are lack of knowledge or poor compliance of the biosafety measures implemented , such as unauthorized movements of birds or their products . Better enforcement and more widely distributed biosafety instructions and training could substantially decrease the risk of introduction of the virus to new areas . Future studies combining genetic data with available epidemiological data should provide a better resolution of the inter-farm transmission network that shaped the epidemic , and further understanding on the mechanisms involved in H7N7 spread during the epidemic . This study shows that partial viral genomic data ( here 3 genes out of the 8 composing the AI genome ) can provide important insights on the transmission dynamics of HPAI viruses even at the scale of temporally and spatially limited epidemic . Our study also strongly suggests that comprehensive study of the evolutionary processes involved in shaping virus diversity are needed in order to use viral genetic data in such ways .
Samples were collected as part of the diagnosis by veterinarians of the Food and Consumer Product Safety Authority ( the Netherlands ) and submitted to the Central Veterinary Institute for confirmation by virus isolation . The authors of this study were not involved in sample collection . Viral RNA was directly extracted from 184 infected trachea tissue samples from dead chickens ( 5 chickens per sample ) using a High Pure Viral RNA extraction kit ( Roche Diagnostics Indianapolis IN , USA ) . Complementary DNA was synthesized by reverse transcription reaction using SuperScript III ( Invitrogen , Carlsbad CA , USA ) , and gene amplification by PCR was performed using the PCR Expand high fidelity kit ( Roche Diagnostics Indianapolis IN , USA ) and primers specific for the hemagglutinin ( HA ) , neuraminidase ( NA ) and basic polymerase 2 ( PB2 ) gene segments . Sequencing was performed by using the BigDye Terminator v . 1 . 1 . sequencing kit ( Applied Biosystems , Foster City CA , USA ) and an ABI Prism 3130 genetic analyzer ( Applied Biosystems ) . Primers and PCR protocols are detailed in Table S4 . Nucleotide sequences are available from the GISAID database ( EPI_ISL_68268-68352 , and EPI_ISL_82373-82472; Table S1 ) . Sequences were aligned using BIOEDIT [36] . We used likelihood ratio tests , Akaike and Bayesian information criteria as implemented in DataMonkey [37] to select the simplest evolutionary model that best fit the different dataset . For the three genes , a Hasegawa–Kishino–Yano ( HKY ) model with gamma-distributed rates among sites was selected . Rates of nucleotide substitution and time of most recent common ancestors ( TMRCAs ) were estimated for the three genes using a Bayesian Markov Chain Monte Carlo ( BMCMC ) method [12] as implemented in the program BEAST [13] . Isolation dates were used to calibrate the molecular clock . Different combinations of molecular clock ( strict clock or uncorrelated relaxed clocks [15] ) and demographic models were attempted independently and the best-fit clock and demographic models were selected by performing Bayes factor tests [14] . The limited timespan of our samples required the use of a simple model to avoid over-parameterization [38] , so we used a single HKY model over all sites in preference to a codon-partitioned model for these analyses . For each dataset , three independent runs were conducted for 60 million generations , sampling every 2 , 000 generations . Convergences and effective sample sizes of the estimates were checked using TRACER [39] . Trees were summarized in a maximum clade credibility ( MCC ) tree after a 20% burn-in using TREEANNOTATOR [13] . The resulting time-scaled phylogenetic trees were visualised with FIGTREE [40] . Additional methods were used to infer the phylogenetic relationships from the HA , NA and PB2 datasets . A Bayesian MCMC inference method was performed in MRBAYES [41] , with multiple runs of 10 million generations with a 20% burn-in , sampling every 100 generations , and using the default heating parameters . A Maximum Likelihood ( ML ) method implemented in PHYML [42] was also used with a bootstrap analysis of 1 , 000 full bootstrap replicates to test the robustness of tree topologies . The three gene segment alignments were manually concatenated to generate a single alignment that was used to construct a phylogenetic network using the Median Joining method [16] implemented in the program NETWORK [17] . This model-free method uses a parsimony approach , based on pairwise differences , to connect each sequence to its closest neighbour , and allows the creation of internal nodes ( “median vectors” ) , which could be interpreted as unsampled or extinct ancestral genotypes to link the existing genotypes in the most parsimonious way . The parameter epsilon , which controls the level of homoplasy , was set at the same value as the weight of characters used to calculate the genetic distances ( weight value = 10 ) . The average number of nucleotide differences within and between the phylogenetic clusters identified was calculated with MEGA [43] . Homologous recombination within each gene segment was searched using Recombination Detection Program version 2 ( RDP2 ) [11] . Putative reassortant viruses were preliminarily identified by the topological incongruity between transmission clusters identified across the phylogenies of different gene segments ( see results ) . This was further investigated with a subset of virus sequences including samples from transmission cluster I ( n = 9 ) , cluster II ( n = 19 ) , cluster III ( n = 8 ) , and cluster IV ( n = 10 ) , from Limburg area ( n = 10 ) , and four additional samples with incongruent phylogenies ( F45 , F76 , F143 , F210 ) . For each sample , the sequences from the 3 genes were manually concatenated , and the resulting alignment was analyzed using bootscan analyses [22] implemented in SIMPLOT [44] . Selection pressure on the HA , NA and PB2 genes was investigated by estimating the ratio of non-synonymous to synonymous nucleotide substitutions ( ω = dN/dS ) using codon-based phylogenetic methods implemented in CODEML ( available in the PAML package [18] ) . Likelihood ratio tests ( LRTs ) were used to test whether model M1a of neutral evolution ( sites restricted to 0<ω<1 ) or model M2a of positive selection ( allows sites with ω >1 ) was a statistically better fit to the data [45] . If the null model M1a was rejected in preference of M2a , a Bayes Empirical Bayes method was used to identify individual codons under positive selection [19] . In addition , positively selected sites were detected using the single-likelihood ancestor counting ( SLAC ) , the random effect likelihood ( REL ) , and the fixed effect likelihood ( FEL ) methods [20] via the Datamonkey website [37] . PCR amplification targeting a 850 bp portion of the NA gene ( bases 57–908 ) was performed on cDNA obtained from five samples ( F26 , F36 , F167 , F191 and F193 ) using the PCR Expand high fidelity kit ( Roche Diagnostics ) . An additional PCR was used to amplify a 570 bp portion of the HA gene ( bases 81–650 ) on the two samples from Limburg only ( F191 , F193 ) . Primers and PCR protocols are described in Table S4 . PCR products were purified using the High Pure PCR Products Purification kit ( Roche Diagnostics ) , and were cloned using the pGEM-T Easy Vector System ( Promega , Madison WI , USA ) . Clones with inserts of the correct size were identified by agarose gel electrophoresis . A total of 50–56 clones with NA inserts were sequenced per farm sample using the BigDye Terminator sequencing kit , version 1 . 1 . and an ABI Prism 3130 genetic analyzer ( Applied Biosystems ) . Nucleotide sequences are available from the GISAID database ( EPI_ISL_82561-82902 ) . Sequences were aligned to the original HA or NA sequence obtained for each farm samples using BIOEDIT [36] , and nucleotide differences were recorded manually . | Outbreaks of highly pathogenic avian influenza ( HPAI ) viruses have affected poultry worldwide in the last decades , resulting in vast socioeconomic damages and many human infections . It is important to determine the route of transmission between poultry farms to be able to implement efficient control measures . Here , we investigate possible use of sequence data to unravel the route of virus transmission during an HPAI H7N7 epidemic that took place in 2003 in the Netherlands . We obtained virus sequence data from most of the outbreaks during the epidemic , and found a high level of genetic diversity driven by a rapid evolutionary rate of HPAI H7N7 virus . The phylogenetic inference of the inter-farm transmission network turned out to be difficult due to the presence of potential reassortant virus strains , multiple mutations at highly variable sites and within farm virus diversity . However , most virus samples could be grouped within clearly defined and chronologically sound clusters of infection , giving us valuable insights on the diffusion of the virus during the outbreak . We discuss the implications of the results obtained for the evolutionary and epidemiological dynamics of avian influenza viruses and disease control . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"evolutionary",
"ecology",
"infectious",
"disease",
"epidemiology",
"influenza",
"infectious",
"disease",
"control",
"veterinary",
"science",
"infectious",
"diseases",
"veterinary",
"diseases",
"zoonoses",
"veterinary",... | 2011 | Evolutionary Analysis of Inter-Farm Transmission Dynamics in a Highly Pathogenic Avian Influenza Epidemic |
Quiescent long-term somatic stem cells reside in plant and animal stem cell niches . Within the Arabidopsis root stem cell population , the Quiescent Centre ( QC ) , which contains slowly dividing cells , maintains surrounding short-term stem cells and may act as a long-term reservoir for stem cells . The RETINOBLASTOMA-RELATED ( RBR ) protein cell-autonomously reinforces mitotic quiescence in the QC . RBR interacts with the stem cell transcription factor SCARECROW ( SCR ) through an LxCxE motif . Disruption of this interaction by point mutation in SCR or RBR promotes asymmetric divisions in the QC that renew short-term stem cells . Analysis of the in vivo role of quiescence in the root stem cell niche reveals that slow cycling within the QC is not needed for structural integrity of the niche but allows the growing root to cope with DNA damage .
The development of multicellular organisms depends on the ability of stem cells to self-renew and to generate new cellular progeny . Transcription factors play key roles in the maintenance of the stem cell state . In embryonic stem cells , for example , stem cell transcription factors repress lineage-specific differentiation programs while maintaining cell proliferation [1] , [2] . In mammals , stem cell quiescence occurs in multiple tissue contexts , where some cells divide infrequently but can recover multiple lineages after injury to the niche [3]–[5] . It has been proposed that quiescent cells reside alongside active stem cells to ensure longevity and output of stem cell compartments [6] . Although quiescence can be released , for example , in murine hematopoietic stem cells by the silencing of p21 [7] and Retinoblastoma ( Rb ) homologs [8] , or by genetic ablation of active stem cells in the gut [5] , it has been difficult to directly test how quiescent and active stem cells share labor within compartments . Stem cell niches in plants and animals display structural similarities [9] . Active stem cells in the Arabidopsis root stem cell niche , also called initials , surround infrequently dividing quiescent centre ( QC ) cells ( Figure S1A ) . The Arabidopsis QC is required to maintain division and prevent differentiation in the surrounding stem cells through non-cell-autonomous signaling [10] , [11] . The central QC cells in Arabidopsis were initially identified as cells that infrequently enter S-phase measured by 3H-thymidine incorporation [12] . Later studies identified the QC as an organizing centre that signals to the surrounding stem cells to prevent differentiation [13] , [14] . It has been proposed that the Arabidopsis QC also acts as a reservoir to replace short-lived stem cells in the root [13]–[14] . However , more than half a century since the initial description of the QC in maize roots [15] , a thorough description of the division of labor between initials and QC cells , and a further investigation of the significance of QC quiescence , has been lacking . In the Arabidopsis stem cell niche , two genetic pathways are involved in stem cell specification . One is dependent on the plant hormone auxin and the PLETHORA ( PLT ) transcription factor family [16] . The other involves protein movement and activation of the heterodimeric transcriptional regulator SCARECROW-SHORTROOT ( SCR-SHR ) [16]–[19] . In addition , reduction of the RETINOBLASTOMA-RELATED protein ( RBR ) , the single Rb homolog , expands stem cell lineages in roots [20] and leaves [21] . SCR-SHR and RBR interact , and the resulting genetic network is biased by auxin and cell cycle progression in order to specify asymmetric cell division in the ground tissue stem cell [22] . Here , we show in vivo that QC cells , in addition to their role as niche organizer , replenish a distal stem cell pool . Intriguingly , quiescence and asymmetric cell division in the QC are balanced by RBR-SCR interactions , which also control asymmetric cell division in ground tissue stem cells . We provide evidence that the physiological function of quiescence is to control a trade-off between genotoxic stress protection and replacement of short-term stem cells .
Previous clonal analyses revealed that in a WT root the QC divides , although at a low rate , and that the QC could be a source for all stem cells in the Arabidopsis root [23]–[25] . However , due to the low QC division frequency , their exact frequency and division pattern has not been determined . We monitored entrance into S-phase using the nontoxic nucleoside analog F-ara-EdU [26] , which allowed normal root growth for as long as 7 days after transfer ( dat ) ( unpublished data ) . Time course analysis of F-ara-EdU uptake showed different times for entry into S-phase for each cell type ( Figure 1A–D ) . In addition , plants were germinated in F-ara-EdU for 5 d and then transferred into nonlabeled growth medium . In such pulse-chase experiments , loss of the label could be detected from the transit amplifying area at 1 dat , but stem cells and QC could maintain label after 3–7 d ( Figure 1E–H ) . Quantification of the data revealed that QC cells divide with half the frequency of the surrounding stem cells and one-fourth the frequency of transit amplifying cells ( Figure 1I ) . To assess the fate of the QC progeny , we used the BOB clonal analysis system , which allowed us to follow genetically marked QC cells and their progeny over time [27] . Briefly , clones generated by this system express two different fluorescent proteins depending on the recombination events , which allows for finer dissection of the clones . We heat-shock-induced 32 QC clones and followed them from 2 to 16 days after heat shock ( daHS ) . Eighty-four percent ( 27 clones ) of the QC clones ( Figure 1J–L ) divided during the experiment , and the rootward daughter always contributed to the columella ( Figure 1M–O ) . We did not observe a contribution of any QC clones to the vascular and ground tissues . Our data indicate that , under physiological conditions , the QC undergoes infrequent divisions to populate the columella region . To relax quiescence in the stem cell niche , we developed a cell-type-specific RBR silencing tool based on artificial microRNA ( amiRNA ) , termed artificial microRNA for Gene-silencing Overcome ( amiGO ) ( Figure 2A ) , which allows for cell autonomous silencing and easy complementation ( Text S1 ) . When expressed ubiquitously , amiGORBR caused supernumerary divisions in stem cells , producing extra columella and Lateral Root Cap ( LRC ) layers that increased over time ( Figure 2B–D ) , and phenocopying previously described roots with reduced RBR function [20] , [27] . amiRNA accumulation was correlated with a reduction in RBR mRNA levels and decrease in protein levels ( Figure 2E–G ) , and the degradation of the target was spatially constrained when amiRNA was driven from tissue-specific promoters ( Figure 2H–J ) . The phenotypes in the stem cell region were similar to those observed upon clonal deletion of RBR [27] , indicating that the level of silencing was sufficient to deplete RBR function in these cell types . Promoters from the ground-tissue- and QC-expressed gene SCR and the QC-specific WOX5 gene ( Figure S1B–C ) allowed us to investigate the role of RBR in specific cell types . In pSCR::amiGORBR roots , extra periclinal cell divisions occurred in the endodermis , consistent with the RBR role in this asymmetric cell division ( Figure 2M , arrowhead ) [22] , and QC cells divided , while no extra LRC layers were produced ( Figure 2K–M , n = 15 ) . pWOX5::amiGORBR roots displayed extra QC divisions , shown by the presence of pWOX5::GFP marker in newly divided cells . In addition , the number of cell layers in the columella increased ( Figure 2N , asterisks; n = 15 ) . The LRC and ground tissue were not affected , consistent with a cell-autonomous role for RBR in QC maintenance . WT plants had a maximum of two undifferentiated columella layers , but p35S::amiGORBR roots displayed up to four layers as revealed by starch granule staining . Quantification of the number of columella and LRC layers revealed that the increase in columella layers in p35S::amiGORBR roots was caused by extra divisions in both QC and columella stem cells , with each of the divisions creating one extra layer ( Figure S3 ) . These observations indicated that the rootward daughters of QC divisions contributed to the columella root cap . To analyze the effect of RBR loss by a different strategy , we next induced and followed QC clones that lost at least one genomic copy of RBR . Homozygous BOB-RBR seedlings ( rbr-3/rbr-3;BOB-RBR+/+ , WOX5::CRE:GR ) [27] were germinated on dexamethasone to induce RBR deletion clones in the QC . QC clones were selected prior to QC division ( Figure S5A–C ) , and followed through division and differentiation . The rootward-most cells ( Figure S5D–I ) acquired starch granules characteristic of differentiated columella cells , demonstrating that QC cells with reduced RBR activity , as in the WT , contribute to the columella . To address whether QC cell divisions were symmetric or asymmetric , we first confirmed the expression of pWOX5::GFP ( ER fluorescence ) and pSCR::SCR:YFP ( nuclear fluorescence ) in the undivided QC of WT ( Figure 3A ) . After a QC cell divided in the pWOX5::amiGORBR background , both daughters expressed pWOX5::GFP ( Figure 3B ) . However , the rootward daughter lost pWOX5::GFP signal over time ( Figure S6A–C ) . pSCR::SCR:YFP was more rapidly lost in the rootward daughter but retained in the shootward daughter ( Figure 3B ) , which , based on these markers , retained QC fate . To determine the fate of the rootward cell , we introgressed two columella markers , pSMB::SMB:GFP and pACR4::ACR4:GFP , marking one differentiated columella layer [28] , and the plasma membrane of columella stem cells and daughters [29] , respectively ( Figure 3C–F ) . In pWOX5::amiGORBR roots , SMB-GFP was expressed in the cell bellow the divided QC cell ( Figure 3C–D ) , and ACR4-GFP was expressed in the rootward daughter and two additional layers of columella ( Figure 3E–F ) , indicating columella identity of the rootward cell . Time lapse analysis of dividing QC cells from 4 to 8 dpg using a brighter nuclear-localized pACR4::H2B:YFP reporter in the pWOX5::amiGORBR background confirmed the progressive acquisition of pACR4 promoter activity in the rootward daughter of the divided QC cell ( Figure S6D–F ) . Together , our results reveal that reduction of RBR activity triggers more frequent asymmetric cell division ( ACD ) in the QC . This ACD generates one shootward daughter cell with QC fate expressing SCR and WOX5 and a rootward columella stem cell expressing ACR4 . Binding of mammalian Rb to LxCxE-motif–containing proteins has been implicated in the maintenance of quiescence in animal cells [30] , [31] . We therefore mutated RBR in residue 849 ( RBRN849F ) , which has been implicated in the interactions between Rb and LxCxE motif proteins in animals [32] . Accordingly , in a yeast two-hybrid assay , the RBRN849F mutant lost the capacity to interact with the strong LxCxE-dependent interactor Histone Acetyl-transferase 2 ( HAT2 ) , while it retained the capacity to bind E2Fa , which does not contain LxCxE motif ( Figure S8 ) . We fused this RBRN849F mutant and the WT RBR cDNAs to the vYFP CDS , under the control of the WOX5 promoter ( pWOX5::RBRN849F:vYFP , pWOX5::RBR:vYFP ) . Because the amiGO system does not target cDNA variants lacking the 3′-UTR sequence , these constructs could be tested for complementation in the pWOX5::amiGORBR background . pWOX5::amiGORBR;pWOX5::RBR-vYFP roots fully complemented the QC division phenotype , assessed both by absence of QC division ( Figure 3G ) and by number of columella layers ( Figure 3K–N and Table 1 ) . pWOX5::RBRN849F:vYFP , however , failed to complement the QC division phenotype of pWOX5::amiGORBR ( Figure 3H , N and Table 1 ) , suggesting that the role of RBR in controlling QC division is dependent on its capacity to bind LxCxE-domain–containing proteins and not through E2F repression . It has recently been shown that RBR represses SCR activity via LxCxE binding to inhibit asymmetric cell division in the mature endodermis tissue [22] . As SCR is involved in establishing and maintaining QC identity , we analyzed the role of the RBR-SCR interaction in QC division , by studying pSCR::SCRAxCxA:YFP; scr-4 mutants , where the RBR-SCR interaction is abolished [22] . Notably , pSCR::SCRAxCxA:YFP; scr-4 mutants displayed extra QC divisions and one extra columella layer ( Figure 3I–J , O ) , indicating that the regulation of QC cell division by RBR is likely due to its interaction with SCR . In the endodermis , SCR and SHR activate CycD6;1 expression , which in turn phosphorylates RBR and inhibits RBR-mediated inactivation of the SHR-SCR complex [22] . However , CycD6;1 was very weakly expressed in the QC of pSCR::SCRAxCxA:YFP; scr-4 plants and pSCR::amiGORBR plants ( Figure 4A , B ) . There was no QC division in the pSCR::SCRAxCxA:YFP; scr-4 , shr-2 background ( Figure 4C ) , demonstrating that , as in ground tissue , this division depends on SHR . Together , these data indicate that QC division is restrained by the SHR-SCR-RBR network . Induction of SHR activated CycD6;1 expression in the ground tissue but not in the QC ( Figure 4D ) . Auxin , either endogenously produced in the QC and ground tissue layer ( Figure 4E ) or exogenously applied ( Figure 4F ) , was also not able to induce CycD6;1 expression in the QC at levels comparable to the ground tissue , nor could we detect more QC divisions upon increased auxin activity . Thus , a QC factor normally inhibits CycD6;1 transcription by the SHR-SCR circuit , and the QC division is not dependent on high CycD6;1 activity . As our data showed a correlation between QC division rates and the number of columella layers , we analyzed the columella in mutants for RBR-SCR network components . Higher SCR-SHR activity led to extra layers ( Figure 3Q ) , whereas lack of activity led to fewer layers in scr-4 and shr-2 mutant backgrounds ( Figure 3R–T ) . Together , our data indicate that division of the QC is regulated by components of the network that regulates ground tissue stem cell ACD with the exception of CycD6;1 . In the QC context , this network variant is used to regulate replenishment of the columella stem cell pool . To address the significance of quiescence , we analyzed the effect of a faster dividing QC in root development . Root growth rates of pWOX5::amiGORBR plants that have continuously dividing QC cells are similar to those of WT at 10 and 25 dpg ( unpublished data ) . Moreover , there were no evident effects on stem cell niche activity and organization in 25 dpg pWOX5::amiGORBR or pSCR::SCRAxCxA:YFP; scr-4 roots ( Figure 5A–C ) . These observations demonstrate that quiescence of the organizing cells is not strictly necessary for function or structural integrity of the niche over this time span , and that the shootward daughter after ACD retains full QC function in its continued ability to maintain the stem cell niche . Replication stress is known to induce division in quiescent stem cells [33] , [34] . We tested whether , in analogy , the induction of QC division might be induced by the ribonucleotide reductase inhibitor hydroxyurea ( HU ) , which is known to delay S phase entry [35]–[37] . Indeed , treatment with 1 µM HU significantly increased the frequency of QC division in Col-0 , but did not further increase QC division in pWOX5::amiGORBR and in pSCR::SCRAxCxA:YFP; scr-4 plants . ( Figure 5D ) . We concluded that replication stress enhances QC divisions through the SCR-RBR pathway . HU causes cell death at higher concentrations [38] , [39] . In addition , columella and vascular tissue stem cells in Arabidopsis roots , but not QC cells , undergo cell death after treatments with drugs that induce DNA damage [40] . In line with classical ideas on the function of mammalian stem cell quiescence , we hypothesized that the reduced mitotic activity of QC cells might enable them to escape cell death and thus ensure the permanence of the organizing centre in the long term and , as a consequence , the maintenance of the root stem cell niche . Therefore we asked whether actively dividing QC cells in pWOX5::amiGORBR roots respond as sensitively as short-term stem cells to DNA damage . We first analyzed the response of WT , pWOX5::amiGORBR and SCR::SCRAxCxA:YFP; scr-4 roots when grown in the presence of the DNA-damaging agent Zeocin . In line with previous observations [40] , WT roots accumulate propidium iodide ( PI ) as a sign of cell death in the vasculature and columella stem cells ( Figure S7A ) after 14 hours postzeocin ( hpz ) , and none of the plants analyzed showed QC death ( n = 27 ) . However , in 63% ( 21/33 ) of pWOX5::amiGORBR roots , at least one of the QC cells accumulated PI , in addition to vascular cells and columella stem cells ( Figure 5G ) . We next pulsed plants with Zeocin to analyze the effects of cell death on the proliferative capacity and growth potential of the roots . At 24 hpz , WT roots with intact QC decreased growth ( Figure 5E–F ) . Importantly , pWOX5::amiGORBR roots with similar stem cell loss but additional QC cell death revealed an exaggerated root growth reduction as shown by the proximity of the root differentiation zone to the meristem and the induction of root hair formation and elongation near the root tip ( Figure S7G–H ) , indicating cell division arrest and progressive differentiation . At 72 hpz , WT roots showed a disorganized , but still active , meristem , while pWOX5::amiGORBR roots underwent rapid differentiation of the transit-amplifying cells followed by differentiation of the root meristem ( Figure S7E–F ) , which is reflected in the reduction of primary root growth ( Figure 5H ) . Together , these data indicate that division of the QC is induced to replenish columella cells and restrained to create a differential stress response within the long-term stem cells that allows the root to cope with DNA stress .
Here we demonstrate that reduction of RBR levels leads to asymmetric cell divisions in the central QC cells of the stem cell niche , thus regenerating short-term stem cells . In addition , our data indicate that RBR acts in a cell-autonomous manner to maintain near-quiescence within the QC . The QC was initially identified as a group of cells with a relatively low mitotic activity at the position where different cell files that form the root converge . Subsequently the QC has been shown to form the structural and functional core of the root meristem in diverse plant species [41] . In Arabidopsis , laser ablation of the QC leads to differentiation of surrounding stem cells [13] . This maintenance function cell-autonomously requires SCR [18] . However , the Arabidopsis QC can perform infrequent divisions , which become more abundant at elevated temperature [23] . These divisions are promoted by the plant hormones ethylene and brassinolide , and do not interfere with stem cell maintenance [24] , [42] . Our results fit with these observations , since extra ACDs caused by the silencing of RBR in the QC do not affect maintenance of the stem cell niche and QC gene expression patterns . It was previously proposed that QC divisions facilitate replacement of the stem cell pool , and therefore that the QC can be a source of stem cells of all lineages in the Arabidopsis root meristem [23] , [42] . In contrast , our results show that after RBR silencing in the QC under normal growth conditions , using either the amiGO or BOB systems , QC divisions only produce columella stem cells . These observations suggest that the shared embryonic lineage of QC and columella may bias QC cells to acquire , upon SCR-mediated asymmetric cell division , columella fate rather than ground tissue fate . Upon stem cell niche injury , however , positional and hormonal cues are dominant over lineage cues in respecifying QC and stem cells from actively dividing cells [14] , [43] . It is surprising to note that the SHR-SCR network , hitherto strictly correlated with ground tissue ( cortex and endodermis ) cell fates , controls two asymmetric cell divisions resulting in different cell identities , demonstrating that this pathway can be deployed for ACDs giving rise to different cell fates . We can at this point not exclude that , alternatively , the SCR-RBR complex represses a QC cell division specified by other factors . It is thought that the quiescence of stem cells in animals is pivotal to ensure tissue maintenance and to protect the stem cell pool from exhaustion under diverse stresses ( reviewed in [44] ) . In mammalian cells , cell cycle exit ( also referred to as quiescence ) is a key step during differentiation . Rb mutants disrupted in LxCxE binding in mammalian cells exhibit defects in quiescence , both during differentiation of myocytes mediated through HDAC1 [32] and during stress-induced senescence mediated by chromatin remodeling proteins such as RBP1 , Sin3 , CtBP , HDAC1 , HDAC2 , and RbAp46 , [45] . Our results suggest that interaction of Rb with LxCxE-containing proteins may represent an evolutionarily conserved mechanism for modulating quiescence . Under natural growth conditions the plant root niche faces multiple biotic and abiotic stresses . Several of these stresses have been shown to affect stem cell niche maintenance [46] , [47] . Additionally , environmental challenges such as hypoxia , temperature , or ozone stress can cause DNA damage , leading to cell death [48] . Similar pathways control cell cycle progression and the cell cycle window where cell differentiation and apoptosis can be initiated , and they seem to converge on RB function in mammalian cells during early G1 [49] . DNA stress caused by hydroxyurea or 5-fluorouracil induces division and differentiation in hematopoietic stem cells , leading to a premature loss of the stem cell niche . Exhaustion of the stem cell niche also occurs after loss of p21 , which inhibits cyclin/cdk complexes that inactivate Rb , suggesting that Rb-related pathways also control quiescence and DNA damage . Loss of p21 leads to cell death upon treatment with hydroxyurea . In plants , where there is no p21 , RBR inactivation leads to QC division , and the same thing happens upon hydroxyurea treatment . However , in plants the pathways do not seem to be additive as in mammalian cells , because hydroxyurea treatment of pWOX5::amiGORBR has no effect in QC division frequency . This suggests that the same pathway that activates QC division by hydroxyurea treatment is regulated by the RB-SCR network . Intriguingly , Arabidopsis columella and vascular tissue stem cells are more sensitive than the QC to zeocin-induced DNA damage [40] . In a similar way , stem cell niches in epithelia contain two stem cell populations , of which the slow dividing population is able to replenish multiple lineages after injury [5] . Our data indicate that RBR-dependent quiescence of the QC plays a crucial physiological role in the maintenance of the niche , and maintains the QC cells as a stem cell reservoir . Quiescence does not need to be absolute in order to protect cells from DNA damage , but rather modest changes in cell cycle frequency are sufficient to bestow protection . Accordingly , shoot apical meristems of plants do not contain distinct QCs but rather a central zone undergoing slower cycling rates . While it is clear that activation of the QC division potential in the root can be triggered by stem cell damage , for example , by laser ablation [14] or stress signals [42] , future work should reveal how exactly plants control the balance between protective quiescence and replacement of short-term stem cells .
Seeds were fume sterilized in a sealed container with 100 ml bleach supplemented by 3 ml of 37% hydrochloric acid for 2–5 h , then suspended in 0 . 1% agarose , and plated on a growth medium consisting of half-strength Murashige Skoog salts , 1% sucrose , 0 . 8% plant agar , MES ( pH 5 . 8 ) , 50 mg/ml ampicillin , and 1–5 mM dexamethasone ( optional for CRE∶GR induced clones ) , stratified for 2 d in 4°C dark room , and grown vertically in long day conditions ( 16 h light followed by 8 h of dark ) . For HS induction , plates with 2–3 days postgermination ( dpg ) seedlings were placed in a 37°C incubator for 1 h and analyzed 2 d later . The 21 mer amiGORBR ( 5′-UACAGAUGCUAUAACUGAGGA-3′ ) and amiGORBR* ( 5′-UACUCAGUUAU ACCAUCUGUA-3′ ) were cloned into Arabidopsis endogenous miR319a precursor via overlapping PCR . The final precursor for amiGORBR was amplifying using modified AttB1 and AttB2 primers and the PCR product was recombined by a BP Single Gateway reaction ( Invitrogen ) in a pGEM-T easy 221 vector for further use in Multisite Gateway Cloning ( Invitrogen ) . The amiGORBR is antisense to the 21 nt sequence located at 3′UTR of RBR mRNA ( 5′-UCUUCAGUUAUAG CAUCUGUA-3′ ) . We loaded 20 µg of the small RNA-enriched fraction per lane , and 5′-end-labeled oligonucleotide complementary to the mature amiGORBR was used as probe . The experiment was performed as described [50] . Total RNA of Col-0 , p35S::amiGORBR ( Col-0 ) , pRB::gRB:GFP ( Col-0 ) , and p35S::amiGORBR;pRB::gRB:GFP seedlings at 5 dpg were obtained using Spectrum Plant total RNA Kit ( Sigma ) . The cDNA was synthesized from 1 µg total RNA using odT18VN primer ( Biolegio ) and RevertAid M-MuLV reverse transcriptase ( Biolegio ) . For the PCR reaction , a 2 µl cDNA sample was used to amplify in a total volume of 20 µl . The relative expression levels of RBR and RB:GFP mRNAs were determined by using primers ( Biolegio ) : RBR FW ( 5′-GATCAAAGATGGATGCTC-3′ ) and RBR RV ( 5′-TACAGATGCTATAACTGAAGA-3′ ) for RBR; RBR FW ( 5′-GATCAAAGATGGATGCTC-3′ ) and GFP RV ( 5′-GAATTGGGACAACTCCAG-3′ ) for RB:GFP . ACTIN1 expression was determined as an internal control using primers Actin FW ( 5′-GCCGATGAAGCTC AATCCAAA-3′ ) and Actin RV ( 5′-GGTCACGACCAGCAAGATCAA-3′ ) . For analysis of protein expression in planta , plants were grown for 12 d under long day conditions , and 0 . 5 g of roots were grinded and extracted in Extraction Buffer ( 100 mM Tris-HCl PH 7 . 5 , 150 mM NaCl , 0 . 5% Nonidet P-40 , 1 mM phenylmethylsulfonyl fluoride [PMSF] , 2× Protease inhibitor cocktail , 100 µM MG132 ) . Equal amounts of protein extracts were loaded in a gel and transfered to a Hybond-ECL membrane ( GE Healthcare ) and inmunodetected with anti-RB antibody ( provided by Dr . L . Bako ) 1/7 , 500 and goat-anti-chicken 1/20 , 000 ( ab97131 Abcam ) and developed with Amersham Western Blotting Detection Reagent ( GE Healthcare ) . Constructs and plant lines used are listed in Table S1 . All different constructs using the amiGORBR expression ( p35S::amiGORBR , pRCH1::amiGORBR , pSCR::amiGORBR , and pWOX5::amiGORBR ) were generated using Multisite Gateway technology ( Invitrogen ) . CaMV 35S-driven amiGORBR construct was generated using a pGII229 binary vector , while other promoter-specific versions were recombined into a pGII226 binary vector . To generate an amiGORBR sensor line , a version of Venus YFP ( vYFP ) containing the amiGORBR target sequence at its 3′ was first obtained by nested PCR and recombined into a pGEM-T easy 221 entry vector . The vYFPamiRBRtarget fragment was then recombined into a pB7m34GW binary vector under the CaMV 35S promoter . Transformation was performed on Columbia ecotype Col-0 and transgenic pWOX5::amiGORBR ( Col-0 ) plants according to the floral dip method [51] . The description of all constructs and lines generated for this study is listed in Text S1 and Table S1 . Whole-mount visualization of roots and starch granule staining were previously described [52] . Starch granules in the columella root cap were stained with 1% lugol solution for 30 s before the visualization . Confocal laser scanning microscopy ( CLSM ) images were performed on a Leica SP2 inverted laser-scanning microscope . Analysis of BOB clones was performed as described [27] . Construction and use of the BOB deletion system is described in Wachsmann et al . 2011 [27] . Seedlings harboring red or cyan clones were preselected under Leica MZ16F fluorescence stereoscope and further analyzed by confocal microscopy . To excite and collect red , cyan , and yellow fluorescences in a Leica SP2 confocal microscope , we performed sequential scanning as follows: the CyPetER and the vYFPNLS were excited together using 458 and 514 nm laser , respectively , and emission was collected at 465–506 nm for the CyPetER and 523–566 nm for the vYFPNLS . Propidium iodide , which marks cell walls ( 3 mg/ml , final concentration ) , and TagRFPER were visualized by exciting at 488 nm and 543 nm , respectively , and emission collected at 502–522 and 561–633 nm . Interactions between RBR and HAT were analyzed by yeast two-hybrid using the ProQuest Two Hybrid System ( Invitrogen Life Technologies ) . RBRwt , RBRN849F , E2Fa , and HAT sequences were cloned in pDONR221 and recombined into pDEST32 BD ( former two ) and pDEST22 AD ( latter two ) . Yeast two-hybrid analysis was performed by duplicate as previously described [22] . MS plates containing 0 . 5% phytagel were supplemented with 20 µg/ml zeocin ( Duchefa Z0186 ) . Plants were grown after transference for a minimum of 14 and a maximum of 24 h . Plants were analyzed using PI-staining and CLSM . For primary root growth analyses after zeocin , data shown are the results of two biological duplicates , with a minimum of 20 seedlings per line in each duplicate . F-ara-EdU treatment was performed in MS plates containing 0 . 5% phytagel and supplemented with 2 µM F-ara-EdU , which was synthesized as described [26] . Incorporation treatments were performed by transfering 4 dpg seedlings to F-ara-EdU–containing plates and growing the plants for further 1–4 dat . Pulse and chase experiments were performed by germinating the seeds in F-ara-EdU–containing plates for 5 dpg and then transferring them into MS plates for further 1–4 dat . Plants were then fixed in 1% formaldehyde , 0 . 1% Triton X-100 in PBS , and Click-iT EdU staining kit ( C10338 , Invitrogen ) was used for signal development before image analysis by confocal microscopy as previously described [53] . Hydroxyurea ( HU ) treatment was performed in MS-agar plates supplemented with 1 µM HU ( SIGMA , H-8627 ) . We treated 4 dpg seedlings for 24 and 48 h , and the root apical meristem of treated plants was analyzed by confocal imaging . QC divisions were scored as QC cells with a newly formed cell wall . Frequency analysis was performed from 20 roots in duplicate experiments . Statistical differences between treatments , as well as between genotypes , were assessed using pairwise student's t tests . | In the plant Arabidposis thaliana , root meristems ( in the growing tip of the root ) contain slowly dividing cells that act as an organizing center for the root stem cells that surround them . This centre is called the quiescent centre ( QC ) . In this study , we show that the slow rate of division in the QC is regulated by the interaction between two proteins: Retinoblastoma homolog ( RBR ) and SCARECROW ( SCR ) , a transcription factor that controls stem cell maintenance . RBR and SCR regulate quiescence in the QC by repressing an asymmetric cell division that generates short-term stem cells . Here we genetically manipulate the cells in the QC to alter their quiescence by regulating the RBR/SCR interaction to demonstrate that quiescence is not needed for the organizing capacity of the QC but instead provides cells with a higher resistance to genotoxic stress , allowing stem cells in the QC to survive even if more rapidly cycling stem cells are damaged . A role for mitotic quiescence has been reported in animal stem cells , in which Rb has been implicated . These findings indicate that it might serve a similar role in plant stem cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | A SCARECROW-RETINOBLASTOMA Protein Network Controls Protective Quiescence in the Arabidopsis Root Stem Cell Organizer |
In order to identify genetic factors related to thyroid cancer susceptibility , we adopted a candidate gene approach . We studied tag- and putative functional SNPs in genes involved in thyroid cell differentiation and proliferation , and in genes found to be differentially expressed in thyroid carcinoma . A total of 768 SNPs in 97 genes were genotyped in a Spanish series of 615 cases and 525 controls , the former comprising the largest collection of patients with this pathology from a single population studied to date . SNPs in an LD block spanning the entire FOXE1 gene showed the strongest evidence of association with papillary thyroid carcinoma susceptibility . This association was validated in a second stage of the study that included an independent Italian series of 482 patients and 532 controls . The strongest association results were observed for rs1867277 ( OR[per-allele] = 1 . 49; 95%CI = 1 . 30–1 . 70; P = 5 . 9×10−9 ) . Functional assays of rs1867277 ( NM_004473 . 3:c . −283G>A ) within the FOXE1 5′ UTR suggested that this variant affects FOXE1 transcription . DNA-binding assays demonstrated that , exclusively , the sequence containing the A allele recruited the USF1/USF2 transcription factors , while both alleles formed a complex in which DREAM/CREB/αCREM participated . Transfection studies showed an allele-dependent transcriptional regulation of FOXE1 . We propose a FOXE1 regulation model dependent on the rs1867277 genotype , indicating that this SNP is a causal variant in thyroid cancer susceptibility . Our results constitute the first functional explanation for an association identified by a GWAS and thereby elucidate a mechanism of thyroid cancer susceptibility . They also attest to the efficacy of candidate gene approaches in the GWAS era .
Thyroid cancer is the most common endocrine malignancy , and accounts for 1% of all neoplasias [1] . Among them , papillary thyroid carcinoma ( PTC , 80–85 % of cases ) , and follicular thyroid carcinoma ( FTC , 5–10 % ) are the most frequent [2] . The etiology of PTC , both sporadic ( 95 % of cases ) and familial ( about 5 % ) , seems to be rather complex . Exposure to ionizing radiation and deficiency in iodine intake have been suggested as environmental risk factors related to PTC and FTC , respectively [3] . Different genetic alterations involving the RET/PTC-RAS-BRAF signalling pathway have been described as causal somatic changes in PTC and FTC [4]–[9] . In addition , PTC has a strong genetic component , since it shows one of the highest relative risks ( FRR = 8 . 60–10 . 30 ) in first degree relatives of probands among cancers not displaying Mendelian inheritance [10] , [11] . Several putative loci associated with familial forms of PTC have been suggested by linkage analysis [12]–[15] , although no high penetrance gene has been convincingly described , even within the putative loci , probably due to the heterogeneity of the disease . Finally , microRNAs ( miRs ) have also been suggested to be involved in the disease [16] , [17] , although their specific role remains unclear . Therefore , it is expected that thyroid cancer is the result of multiple low- to moderate-penetrance genes ( LPGs ) interacting with each other and with the environment , thus modulating individual susceptibility [10] , [18] . In this scenario , linkage analysis does not have the power to identify these LPGs [19] , [20] . Thus , GWAS or carefully designed candidate gene approaches may be more appropriate strategies to define genetic risk factors . We performed a candidate gene association study in thyroid cancer , showing that FOXE1 , formerly called TTF2 ( Thyroid Transcription Factor 2 ) , exhibits the strongest association with PTC susceptibility . FOXE1 itself is a good candidate LPG because it is the centre of a regulatory network of transcription factors and cofactors that initiate thyroid differentiation [21] and whose function is essential for thyroid gland formation and migration , as well as for the maintenance of the thyroid differentiated state in adults [22] . Our study , which involves the largest collection of patients with this pathology from a single population to be studied to date , also identifies a causal variant within FOXE1 as well as the underlying molecular mechanism involved . The variant rs1867277 ( NM_004473 . 3:c . −283G>A ) within the FOXE1 5′ UTR affects gene transcription through differential recruitment of USF1/USF2 transcription factors only when the −283A allele is present . By contrast , a protein complex in which DRE- and CRE-binding proteins participate , binds to both alleles . Recently , a genome-wide association study ( GWAS ) identified two SNPs located at 9q22 . 23 and 14q13 . 3 that are strongly associated with an increased risk of PTC and FTC [23] . The closest gene to the top marker at 9q22 . 33 , rs965513 ( OR = 1 . 75; P = 1 . 7×10−27 ) is FOXE1 . Overall , it is noteworthy that , in this particular case , both strategies identify the same gene , although the study of carefully selected candidate genes remains a more direct , practical and efficient approach to reveal functional variants within LPGs .
Patients diagnosed with thyroid cancer were recruited from the Spanish hospital network . A total number of 615 cases were available for the study , representing , to our knowledge , the largest thyroid cancer series from a single population . Our series included the main thyroid follicular-cell derived carcinomas: 520 PTC , represented by their main subtypes ‘classic PTC’ ( cPTC; n = 304 ) and ‘follicular variant PTC’ ( FVPTC; n = 146 ) , as well as 69 Follicular Thyroid Carcinomas ( FTC ) . Medullary Thyroid Carcinomas ( MTC ) were not included in the study , since we previously performed a similar study of these cases [24] . Clinicians fulfilled a detailed clinical questionnaire for all patients , which included questions regarding both personal and clinical data , such as tumour subtype and stage , surgery option , treatment details in terms of 131I doses , and development of metastasis during the follow-up . Diagnoses were assessed by pathologists from the different institutions that participated in the study . A series of 525 healthy controls , free of cancer and representative of the Spanish population , were selected as the reference group . These subjects came from the same geographical regions as covered by the hospitals involved in the study . Informed consent was obtained from all subjects included in the study . Median age and sex ratio ( female∶male ) were 46 years and 4 . 6 , respectively , in both cases and controls . Mean age and gender distribution were similar in controls and cases ( Mann-Whitney's U and Kruskal-Wallis associated Ps>0 . 05 ) . An additional 78 Spanish cases , obtained over the last year , were genotyped for the significant SNP analyzed in most detail , in order to increase the power of the test . These additional patients were recruited from the same hospitals as the original Spanish series . The distributions of thyroid cancer subtype , age and gender were also similar to those of the first series ( P>0 . 05 ) . A second stage of the study consisted of an independent Italian series , used as a validation set , and was composed of 482 thyroid cancer patients and 532 representative controls . Cases included 412 individuals with PTC and 44 with FTC . Their median age at diagnosis and female∶male sex ratio were 48 . 5 years and 4 . 9 , respectively . Italian controls were recruited from the same two geographical regions as cases , and had a similar sex ratio and age distribution as cases . The design of the study is summarized in Figure S1 . Blood ( n = 878 ) or saliva ( n = 262 ) samples were obtained from Spanish patients and controls . Genomic DNA was extracted from peripheral blood lymphocytes by automated DNA extraction according to the manufacturer's instructions ( Magnapure , Roche ) and from saliva using the Oragene DNA Self-Collection Kit ( DNA Genotek , Ottawa , Canada ) . Genomic DNA was isolated from Italian blood samples ( n = 1014 ) using standard methods [25] . DNA concentration was quantified in all samples prior to genotyping by using Quant-iT PicoGreen dsDNA Reagent ( Invitrogen , Eugene , OR , USA ) . We used a candidate gene approach in this study . Three different criteria were used for selecting loci . First , we chose genes we found to be differentially expressed in primary thyroid tumours and normal tissue [26] , or as described in public databases , such as CGAP-SAGE ( http://cgap . nci . nih . gov/SAGE ) . Second , we picked genes involved in thyroid follicular cell biology and metabolism . Finally , critical metabolic pathways such as the MAP kinase , JAK-STAT and TGF-beta signaling pathways were represented by selecting genes encoding proteins that play key roles in those pathways ( membrane receptors , signal transducers , transcription factors , inhibitors , etc . ) . The latter criterion was applied after an exhaustive review of the information contained in the pathway databases KEGG Pathways ( http://www . genome . ad . jp/kegg/pathway . html ) , Biocarta ( http://www . biocarta . com/genes/index . asp ) , and Pathway Studio 4 . 0 ( evaluation version ) . Candidate non-coding MIRN genes were considered in the initial list using miRBase ( http://microrna . sanger . ac . uk/sequences/ ) . We ranked the loci based on the above criteria and finally selected 97 genes for our association study ( manuscript in preparation ) . The selected genes were represented by Single Nucleotide Polymorphisms ( SNPs ) within the intragenic region and within the regions spanning 10 kilobases upstream ( to cover the hypothetical entire promoter area ) and 2 kilobases downstream of the gene . We chose a total number of 768 SNPs , that can be divided into two main categories: ( i ) 523 ‘tag SNPs’ , used to infer Linkage Disequilibrium ( LD ) blocks according the HapMap project ( http://www . hapmap . org/ ) [27]; and ( ii ) 245 potentially functional SNPs , as predicted by bioinformatic tools PupaSuite ( http://pupasuite . bioinfo . cipf . es/ ) [28] and F-SNP ( http://compbio . cs . queensu . ca/F-SNP/ ) [29] . Predictions of functionality included SNPs that caused an aminoacid change in the protein ( non-synonymous SNPs ) , as well as those variants located within putative transcription factor binding sites ( TFBS ) and exonic splicing enhancers ( ESE ) . SNP codes , locations , and frequencies were obtained from the NCBI SNP database , build 126 ( http://www . ncbi . nlm . nih . gov/projects/SNP/ ) . The initial list of more than 52 , 000 SNPs fulfilling the above criteria was filtered by applying the following additional criteria: ( i ) a threshold minor allele frequency ( MAF ) in the HapMap-CEU population of 0 . 10 for ‘tag SNPs’ and of 0 . 02 for putative functional SNPs; and ( ii ) an ‘Illumina score’ not less than 0 . 6 ( as recommended by the manufacturer ) , to ensure a high genotyping success rate . No variants within the MIRN genes considered were described in the databases . Finally , we selected and genotyped 768 SNPs within 97 loci , thus fulfilling the platform requirements . SNPs were genotyped using the Illumina GoldenGate Genotyping Assay ( San Diego , CA , USA ) system , on a Sentrix Universal-96 Array Matrix multi-sample array format . Genotyping was carried out using 400 nanograms of DNA per reaction following the manufacturer's instructions ( http://www . illumina . com/ ) . Genotyping specificity was assessed by including two DNA duplicates ( an intra-assay and an inter-assay duplicate ) and a negative control in each 96-well plate genotyped , yielding 100% consistent replication results . In addition , cases and control samples were always included in the same run . Validation set genotyping was performed by means of the KASPar SNP Genotyping System ( Kbiosciences , Herts , UK ) . Fifteen nanograms of DNA were used for the genotyping reactions . The 7900HT Sequence Detection System ( Applied Biosystems , Foster City , CA , USA ) was used for fluorescence detection and allele assignment . An additional variant , rs1867277 , not included in the Illumina assay , was selected for its predicted effect on the transcriptional activity of FOXE1 to perform functional assays . This SNP was analysed on a subset of 200 cases and controls by DHPLC on the WAVE HT system ( Transgenomic , Omaha , NE ) using an acetonitrile gradient; it was scrutinised for aberrant profiles with the Navigator software ( Transgenomic , Omaha , NE ) to determine the correlation between this potential functional SNP and the other variants in FOXE1 genotyped by the Illumina platform . Genotyping accuracy of both KASPar and DHPLC technologies was confirmed by direct sequencing of 5% of the samples selected at random . WRO cells derived from a human follicular thyroid carcinoma were cultured as described [30] . The expression vectors used were pcDNA3 . 1-DREAM [31] , pSG5-αCREM [32] , pGal4-CREB [33] , pN3 ( USF1 ) , pN4 ( USF2 ) [34] and pUSF-1 , expressing a dominant negative form of USF1 . The pGl3b-FOXE1 reporter construct contains the 5′ upstream regulatory region from −1934 to +539 bp relative to the transcription start site of human FOXE1 [35] and carries an A allele for the rs1867277 SNP . In this plasmid , the A allele was replaced by the G allele ( pGl3b-FOXE1-283G ) by means of site directed mutagenesis ( QuikChange II XL kit; Stratagene ) following the manufacturer's instructions using the oligonucleotide 5′-cagtcccggtc[g]cgaggccaccgc-3′ . The c . −283A>G substitution and the absence of artefacts were confirmed by direct sequencing . The vector pRL-CMV , containing a cDNA coding for Renilla , was used to monitor transfection efficiency . Nuclear protein extracts from WRO cells were obtained following standard procedures [36] . Specific proteins were synthesized from the USF1/2 expression vectors pN3 and pN4 by in vitro transcription/translation using the TNT coupled reticulocyte lysate system ( Promega ) . Seven µg of protein extracts or 3 µl of TNT pools were incubated with 200 ng of the corresponding dsDNA probes representing rs1867277-A: 5′-gtcccggtcAcgaggccaccg-3′ ( referred to as “Allele A” ) ; rs1867277-G: 5′- gtcccggtcGcgaggccaccg-3′ ( “Allele G” ) ; and the DRE element from the prodynorphin gene: 5′- gaagccggagtcaaggaggcccctg-3′ ( “DRE-pDyn” ) , previously labelled with γ32-ATP by T4 polynucleotide kinase . For competition , a 100-fold excess of the same ( “related” ) or different ( “unrelated”: 5′-ccataatgcaaaaatggaaagaattaaa-3′ ) unlabeled oligonucleotide was used as indicated in each experiment . Additional dsDNA probes used were: the USF consensus sequence ( USF-cons ) 5′-cctgcccacgtgacccggcct-3′; the CRE binding region of the somatostatin gene ( CRE-Cons ) 5′-cctcctagcctgacgtcagagagagagt-3′; and the CRE-like region of the FOXE1 gene ( CRE-FOXE1 ) 5′-accagagtcgagtcccggtcacgaggcca-3′ . When required , specific antibodies recognizing human DREAM , USF1 , or USF2 ( sc-9142 , sc-229 and sc-862 , respectively , from Santa Cruz Biotechnologies ) were incubated together with protein extracts and dsDNAs . EMSA conditions were similar to those described previously [24] . Hela cells were transient transfected using the JetPei reagent ( PolyPlus Transfection ) with different amounts of expression vector ( as indicated in each experiment ) ; 3 µg of reporter construct and 60 ng of pRL-CMV were used . Forty-eight hours after transfection , cells were harvested , lysed , and analyzed for luciferase and Renilla activities . The promoter activity in cells transfected with the expression vector was determined as the ratio between luciferase and Renilla , relative to the ratio obtained in cells transfected with the corresponding empty expression vector . The results shown are the average±SD of three independent experiments , each performed in triplicate . Departure from Hardy-Weinberg equilibrium ( HWE ) for all SNPs was tested in controls using Fisher's exact test . Associations between each SNP and thyroid cancer risk were assessed using Pearson's χ2 test . Genotype frequencies in cases and controls were compared and odds ratios ( OR ) per allele were estimated by applying unconditional logistic regression , using homozygotes of the most frequent allele in controls as the reference group . Significant P values were adjusted for the putative confounding factors age , gender , and population . Haplotypes were inferred using PHASE 2 . 0 , a computational tool based on Bayesian methods . Case-control comparisons of haplotype distributions were carried out by applying the inbuilt permutation test , based on 10 , 000 permutations . Associations between specific haplotypes and risk of thyroid cancer were assessed using the Haplo . Stats package in R ( http://www . R-project . org ) . For transfection assays , statistical significance was determined by t-test analysis ( two-tailed ) , and differences were considered significant with P<0 . 05 . Statistical tests were performed using SPSS for Windows 17 . 0 software .
Five out of six significant SNPs in FOXE1 were located in a single LD block ( Figure 1 ) , and were subsequently used for haplotype analysis of the region spanning from chromosomal coordinates 99 , 648 , 503 to 99 , 668 , 059 ( Figure 1 , Table 2 ) . This approach , used to reinforce results from individual SNP studies , showed a significant difference in haplotype distribution between all PTC cases and controls ( PHASE associated P = 0 . 0147 ) . Again , these differences were stronger when considering the classic PTC subtype ( PHASE associated P = 0 . 0005 , Table S2 ) . Among the ten haplotypes inferred in our population , three represented 95% of individuals . The remaining seven haplotypes showed frequencies lower than 1% , and were not considered for further analyses . A detailed study of the three mentioned FOXE1 haplotypes ( shown in Table 2 ) , allowed us to identify a risk haplotype significantly overrepresented in classic PTC patients ( OR = 1 . 66; P = 0 . 0005 ) . Since none of the six FOXE1-associated variants showed a consistent putative functional role ( according to the bioinformatics tools used ) , a more detailed analysis of the sequence across this LD region was performed . This allowed us to identify a functionally interesting SNP ( rs1867277 , -283G>A ) , not initially included in the Illumina assay . According to bioinformatics predictions , this variant may influence the binding of transcription factors that could regulate FOXE1 transcription . DHPLC results in a subset of 200 individuals allowed us to experimentally confirm a complete LD between this functional variant ( rs1867277 ) and the rs907577 tagSNP ( included in the Illumina assay ) , and thus to impute genotypes of the first variant ( OR[per allele] = 1 . 39; P = 2 . 1×10−4 ) . This correlation is represented in Figure 1 , and specific LD values are provided in Table S3 . Interestingly , the risk haplotype identified ( Table 2 , haplotype #2 ) , included the rs907577 G allele and the functional variant rs1867277 A ( located only 797 bases downstream ) , which has an effect on FOXE1 transcription ( see below ) . An independent Italian population ( phase II ) validated the phase I association results for the rs1867277 FOXE1 promoter variant . This polymorphism was found to be significantly overrepresented in a series of 405 Italian PTC cases versus 525 Italian controls ( OR[per allele] = 1 . 64; 95% CI = 1 . 31–2 . 05; adjusted P = 1 . 3×10−5 ) . Phase I , II , and combined analyses for both series are summarized in Table 3 . Pool analysis for rs1867277 in 984 PTC cases vs . 1028 controls confirmed the involvement of this FOXE1 variant in PTC development ( OR[per allele] = 1 . 49; 95% CI = 1 . 30–1 . 70; adjusted P = 5 . 9×10−9 ) . rs1867277 lies within a sequence with high similarity to the DRE consensus core sequence described for the prodynorphin promoter region [37] ( Figure 2 ) . In a first attempt to evaluate the role of rs1867277 in the transcriptional regulation of FOXE1 , nuclear extracts from WRO cells were tested for their ability to bind both A and G alleles in an EMSA assay . As shown in Figure 3A , protein/DNA complexes were differentially formed when using A and G-allele probes ( lanes 2 and 5 , respectively ) : a lower complex was formed with both alleles , while an upper complex was formed exclusively with the A allele . The complexes were specific , as they were competed by a 100-fold excess of unlabelled related oligonucleotide ( lanes 3 and 6 ) but not by an unrelated one ( lanes 4 and 7 ) . In order to identify the transcription factors that bind to both alleles , we first focused on DREAM ( Downstream Regulatory Element Antagonist Modulator ) since rs1867277 contains a putative DRE consensus sequence ( Figure 2 ) . When the prodynorphin promoter containing the DRE consensus sequence ( DRE-pDYN ) was used as a probe , a specific complex with the same mobility as the A or G-allele lower band was detected ( Figure 3A , lane 8; related and unrelated competition , lanes 9 and 10 ) . Interestingly , this complex was partially competed by A and G-allele oligonucleotides ( lanes 11 and 12 ) . Furthermore , a specific DREAM antibody substantially reduced the intensity of the lower band ( lane 14 ) . These data demonstrate the involvement of endogenous DREAM in the lower shifted complex , although the lack of total competition when oligonucleotides containing the A or G allele were used , suggests that other proteins may be a part of this complex . Intriguingly , a CRE-like sequence was identified close to rs1867277 ( Figure 2 ) ; CRE-binding factors can behave as interacting partners of DREAM [38] . As the upper complex was exclusively formed when the A allele was present , we decided to identify the transcription factor/s that form part of that complex by interrogating databases ( Gene Regulation , http://www . gene-regulation . com ) . The Upstream Stimulatory Factors ( USF1 and USF2 ) were predicted to bind exclusively to the rs1867277 A allele . Thus , EMSAs with the A allele were performed with different oligonucleotides as competitors , including a USF consensus sequence and several oligonucleotides containing CRE sequences , due to the relationship between DRE- and CRE ( Figure 3B ) . When the oligonucleotide containing the G allele was used , only the lower complex was competed , demonstrating the specificity of the upper complex for the A allele ( lane 5 ) . Similarly , CRE-like FOXE1 , the consensus CRE sequence , and consensus DRE-pDyn competed exclusively with the lower band ( lanes 6–8 ) , suggesting that the lower complex contains more than one transcription factor . Interestingly , the upper band was totally competed when the USF consensus sequence was used ( lane 9 ) . Furthermore , the in vitro translated ( TNT ) USF1 or USF2 proteins formed a complex with a similar mobility as the upper complex ( lanes 11–12 ) . When specific antibodies against either transcription factor were added , the upper complex formed over the A allele was supershifted when using nuclear extracts ( Figure 3C , lanes 6 and 14 ) , and also when using in vitro translated USF proteins ( lanes 8 and 16 ) . This unequivocally demonstrates that USF1/USF2 form part of the upper complex . On the other hand , the rs1867277 G allele shift and supershift experiments using the same approach did not show this upper DNA-protein complex ( not shown ) . Altogether , these results strongly indicate that only the rs1867277 A allele is able to form a protein complex that includes the transcription factors USF1 and USF2 . In addition , both rs1867277 A and G form a complex in which DREAM and possibly other CRE-binding related transcription factors participate . To validate the functional significance of the transcription factors identified for FOXE1 gene expression , Hela cells were transiently transfected with expression vectors harbouring the cDNAs of the different transcription factors , together with the FOXE1 reporter construct containing the rs1867277 A allele ( pGl3b-FOXE1-283A ) . While DREAM transfection did not generate significant variations in FOXE1 promoter activity , USF1 and USF2 transfection increased this activity ( P<0 . 01 and P<0 . 05 , respectively ) ( Figure 4A ) . Since USF homo- and heterodimer formation has been described [39] , we cotransfected both expression vectors simultaneously , and observed a large increase ( 8-fold , P<0 . 0001 ) of FOXE1 promoter activity . This effect was specific , since it was abolished when a dominant negative form of USF was cotransfected . Moreover , cotransfection of the DREAM expression vector together with USF1/USF2 clearly reduced the activation of the FOXE1 reporter when compared to the USF1/USF2 condition ( P<0 . 0001 ) , confirming the corepressor function of DREAM [31] , [40] . All these data highlight the relevant regulatory role of USF factors in the expression of FOXE1 , as well as the inhibitory effect of DREAM on USF-dependent FOXE1 activation . To verify that the effect of the USF factors is mediated through the previously mentioned putative USF binding site in the A allele , transient transfection assays were performed using pGl3b-FOXE1-283G , the reporter construct containing the G allele . No effect was found for USF1 and USF2 when used independently , and simultaneous cotransfection of USF1/2 resulted in a transcriptional activation that was clearly less than that observed with the A allele ( P<0 . 0001 ) . In summary , the rs1867277 G allele partially impairs the recruitment of USF1/2 factors to the FOXE1 promoter and alters the expression status of the FOXE1 gene . Transcriptional activation of the FOXE1 gene is regulated by hormonal factors , particularly by TSH via cAMP [41] . The transcription factors αCREM and CREB bind to CRE consensus sequences within the promoters of genes regulated by cAMP . Given the existence of a CRE-like site located near rs1867277 ( Figure 2 ) , we evaluated the role of αCREM and CREB in FOXE1 regulation . Overexpression of the two isoforms induced significant increases of FOXE1 promoter activity , with αCREM being the most potent activating factor ( changes >50 fold ) ( Figure 4B ) . In order to evaluate how this αCREM-dependent activation could be modified by the transcription factors that bind within the FOXE1 promoter at the region containing the rs1867277 risk variant , various amounts of the USF1 , USF2 , CREB and DREAM expression vectors were transfected alone or in combination with fixed amounts of αCREM expression vector and FOXE1 reporter . USF1/USF2 and CREB did not affect αCREM-induced FOXE1 reporter activity , while DREAM overexpression led to an 80% repression ( P<0 . 0001 ) . Considering these data , CRE-binding factors are acting as positive regulators of FOXE1 gene expression , while DREAM displays a negative regulatory effect on αCREM-dependent regulation . Moreover , USF factors do not modulate the transcriptional activity of αCREM , which suggests the existence of two independent regulatory mechanisms that include DREAM and CRE-binding factors or USF proteins .
In the present work , by means of a two-step candidate-gene association study , we have identified FOXE1 as a low penetrance gene ( LPG ) associated with papillary thyroid cancer ( PTC ) . Regression analysis allowed us to pinpoint several SNPs within the FOXE1 locus as highly significant risk-conferring variants for PTC . Haplotype analysis , followed by an exhaustive search within the sequence of the LD block containing the FOXE1 gene , enabled us to identify a promoter SNP ( rs1867277; c . −283 G>A ) that we postulate to be a causal variant due to its predicted effect on a transcription factor binding site . The association results for this variant were validated in an independent population . The combined OR [per allele] was 1 . 49 ( 95% CI = 1 . 30–1 . 70; P = 5 . 9×10−9 ) . Functional assays revealed that the c . −283A allele led to a differential recruitment of USF1 and USF2 transcription factors . Thyroid cancer is believed to be a complex disease , in which common genetic variants located in low penetrance genes may interact with each other and with the environment , determining individual susceptibility . Our association results suggest that FOXE1 is especially important for developing the classic PTC subtype , which represents around half of the PTC cases . This observation provides a basis for the heterogeneity described for this neoplasia , which includes more than 15 histological subtypes . Therefore , we propose that FOXE1 is acting as an LPG related to thyroid cancer , which is in agreement with the increasing evidence of forkhead box ( Fox ) proteins having a crucial role in the development and progression of cancer , and their emerging role as potential biomarkers [42] . The FoxE1 transcription factor belongs to the forkhead family of transcription factors . These factors share a highly conserved winged helix DNA binding domain that is able to interact with nucleosomes and alter chromatin structure , creating a local exposed domain necessary for the action of other transcription factors . This property defined FOXE1 as a pioneer transcription factor [43] , whose action is essential for the development , differentiation , and hormone responsiveness of the thyroid gland [22] . Thus , the control of its expression must be exquisitely regulated . However , few data are available concerning the transcription factors involved in FOXE1 expression in the thyroid , although the fact that TSH controls FOXE1 expression through cAMP and Ca2+ [41] , [44] suggests that CRE- and Ca2+-binding factors may play a key role . In fact , sequence analysis of the FOXE1 promoter , where rs1867277 ( c . −283G>A ) is located , revealed the existence of a DRE site , similar to the one previously found by D'Andrea et al . [45] and that shared a high similarity with a previously defined consensus DRE site in the prodynorphin gene [37] . Unexpectedly , functional assays showed that the Ca2+-dependent transcription factor DREAM bound equally well to both rs1867277 alleles . The most remarkable functional data obtained was the formation of a DNA-protein complex exclusively between the A allele and USF1/USF2 factors . These are ubiquitously expressed proteins , which belong to the basic helix-loop-helix ( HLH ) leucine zipper family of transcription factors . They share a highly conserved C-terminal domain responsible for dimerization and DNA binding , which recognizes the canonical E-box sequence CACGTG [39] , [46] . The involvement of USF factors in FOXE1 regulation was confirmed by a transfection approach . These factors , which mainly act as heterodimers , induced significant increases in FOXE1 transcriptional activity when the rs1867277 A allele was present . These data agree both with the role of USF factors as positive transcriptional regulators of their target genes [47] and with the predominant role of USF1/USF2 heterodimers in comparison with homodimers [39] . Moreover , interactions between N-terminal domains of USF dimers and cell-specific transcription factors have been described to be involved in cooperative transcriptional regulation [48] . Interestingly , we identified a potential CRE site located near the rs1867277 sequence , which could act as a target of CRE-related proteins ( CREB and αCREM ) . Transient transfection assays demonstrated that CREB , and more strongly αCREM , activate the FOXE1 promoter , while DREAM reduces significantly αCREM dependent-transcriptional induction . These data suggest a direct competition between CRE-binding factors and DREAM for binding to the FOXE1 promoter , which will ultimately control the transcriptional status of the FOXE1 gene . In fact , it has been described that Ca2+ and cAMP concentrations modulate a regulatory network which links CRE-binding proteins and DREAM [38] , [49] . In this way , these transcription factors could regulate the expression of FOXE1 in response to TSH in a physiological situation . Given that both CRE proteins and USF factors are associated with an increase in FOXE1 expression levels through closely neighbouring sites , the question arises if these factors also cooperate to regulate FOXE1 transcription . Our results showed no synergistic effect of αCREM and USF1/2 proteins on FOXE1 transcription , which is in agreement with other reports [50] , [51] . It therefore remains to be elucidated which additional proteins , if any , are acting together with USF proteins in modulating FOXE1 expression . Finally , considering the specific binding of USF factors to the disease risk-conferring rs1867277 A allele , an increased expression of FOXE1 in thyroid follicular cell tumours in comparison to normal thyrocytes is to be expected . Few data are available regarding FOXE1 status in thyroid cancer , although Sequeira et al demonstrated that an increased FOXE1 expression paralleled the dedifferentiation process of thyroid carcinomas [52] . Therefore , it is necessary to understand in which manner FOXE1 , a thyroid transcription factor involved in the maintenance of the differentiated adult thyroid phenotype , could be involved in acquiring a malignant status . One possible explanation relies on the results obtained from FOXE1 knockout mouse models . During embryonic development , thyroid cell precursors require FOXE1 transcription initially to allow their own migration from the thyroid bud , and later to constitute a functional endocrine organ [21] , [53] . Considering these data , we hypothesised that increased FOXE1 expression in thyroid carcinomas could be related to a motile advantage of malignant thyroid cells , which would be enhanced by the presence of the rs1867277 A risk predisposing allele . While FOXE1-specific studies are needed to further understanding the role of this gene in thyroid tumour cell migration and invasion , several studies have confirmed a tumoral role of forkhead family of transcriptional factors . In this regard , genes encoding forkhead factors , among others , have been recently identified as a molecular signature for Epithelial to Mesenchymal transition in a human colon cancer [54] and overexpression of FOX factors has been described in several cancers [55]–[62] . This opens an interesting future for understanding the role of FOXE1 in thyroid tumour cell migration and invasion . Taken together , our association study , combined with a functional assessment , allowed us to pinpoint a causal variant within the FOXE1 promoter , and to propose a mechanism by which this causal variant acts as a genetic risk factor specifically related to PTC susceptibility . This finding reveals the importance of considering a complex disease such as thyroid cancer as a heterogeneous entity . It is crucial not to study complex diseases as a collective , but to cluster cases according to homogeneous and well-established clinical features . It seems obvious that just one variant cannot explain the phenotype , and many other signals and mechanisms may be also involved . However , our results are an important step forward in understanding the disease , offering new insights into the genetic mechanism involved in non-medullary thyroid cancer development . It is also important to remember that the complex nature of each locus could be operative through additional variants or mechanisms . In addition , our study validates the results of the first GWAS performed in thyroid cancer [23] , and addresses one of the major limitations of GWAS: the enormous difficulty to provide a link between a significant intergenic tagSNP and the precise variant that has a causal role and provides a biological explanation [63] . The modest power to identify causal variants using GWAS is in part because , in this hypothesis-free approach , approximately 20% of common SNPs are partially tagged , and rare variants are not tagged at all . Overall , our study provides proof that , when reliable biological knowledge or expression data from the tissues of interest are available , candidate gene approaches can be straightforward to identify low penetrance genes and to find putative causal variants within those genes . | Although follicular cell-derived thyroid cancer has an important genetic component , efforts in identifying major susceptibility genes have not been successful . Probably this is due to the complex nature of this disease that involves both genetic and environmental factors , as well as the interaction between them , which could be ultimately modulating the individual susceptibility . In this study , focused on genes carefully selected by their biological relation with the disease , and using more than 1 , 000 cases and 1 , 000 representative controls from two independent Caucasian populations , we demonstrate that FOXE1 is associated with Papillary Thyroid Cancer susceptibility . Functional assays prove that rs1867277 behaves as a genetic causal variant that regulates FOXE1 expression through a complex transcription factor network . This approach constitutes a successful approximation to define thyroid cancer risk genes related to individual susceptibility , and identifies FOXE1 as a key factor for its development . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/cancer",
"genetics"
] | 2009 | The Variant rs1867277 in FOXE1 Gene Confers Thyroid Cancer Susceptibility through the Recruitment of USF1/USF2 Transcription Factors |
Schistosoma blood flukes , which infect over 200 million people globally , co-opt CD4+ T cell-dependent mechanisms to facilitate parasite development and egg excretion . The latter requires Th2 responses , while the mechanism underpinning the former has remained obscure . Using mice that are either defective in T cell receptor ( TCR ) signaling or that lack TCRs that can respond to schistosomes , we show that naïve CD4+ T cells facilitate schistosome development in the absence of T cell receptor signaling . Concurrently , the presence of naïve CD4+ T cells correlates with both steady-state changes in the expression of genes that are critical for the development of monocytes and macrophages and with significant changes in the composition of peripheral mononuclear phagocyte populations . Finally , we show that direct stimulation of the mononuclear phagocyte system restores blood fluke development in the absence of CD4+ T cells . Thus we conclude that schistosomes co-opt innate immune signals to facilitate their development and that the role of CD4+ T cells in this process may be limited to the provision of non-cognate help for mononuclear phagocyte function . Our findings have significance for understanding interactions between schistosomiasis and other co-infections , such as bacterial infections and human immunodeficiency virus infection , which potently stimulate innate responses or interfere with T cell help , respectively . An understanding of immunological factors that either promote or inhibit schistosome development may be valuable in guiding the development of efficacious new therapies and vaccines for schistosomiasis .
Extensive co-evolution of parasitic organisms and their hosts has given rise to complex host-parasite relationships in which exploitation of host responses to infection by parasites is a recurring theme . Nowhere is this complexity in host-parasite relationships better exemplified than in the parasitic helminths , which infect a third of the world's human population [1] by establishing chronic infections that persist for years , often in the face of vigorous immune responses [2] . Schistosoma blood flukes account for a significant proportion of these helminth infections , causing considerable morbidity and mortality . After initiating infection by direct skin penetration , schistosomes migrate in the bloodstream to hepatic pre-sinusoidal venules , where rapid growth and development ensues , culminating in mating of adult worms and the production of eggs . It was previously shown that Schistosoma mansoni co-opts CD4+ T cell-dependent mechanisms to facilitate both parasite development during pre-patent infection and the excretion of parasite eggs after the onset of oviposition [3] , [4] , [5] . The latter requires formation of Th2-dependent granulomas in the bowel wall to allow passage of eggs from the portal vasculature into the intestinal lumen [3] , [4] , [6] . However , the mechanism by which CD4+ T cells facilitate development of schistosome worms has not been elucidated . While homeostatic maintenance of peripheral CD4+ T cells through the action of γc cytokines is required to provide a permissive environment for schistosome development [7] , previous studies did not identify a role for effector Th1/Th2 responses or any single effector cytokine in parasite development [5] , [8] , [9] . In this study we tested whether activation of CD4+ T cells through the T cell receptor ( TCR ) by schistosome antigens is required for schistosome development to proceed . Unexpectedly , our data indicate that CD4+ T cells that lack specificity for schistosome antigens can facilitate schistosome development in the absence of antigen-mediated T cell activation . Interestingly , the presence of naïve , resting CD4+ T cells also correlates both with steady-state transcriptional changes in the expression of genes that are critical for development of monocytes and macrophages and with changes in the composition of peripheral monocyte populations . Further , direct stimulation of the mononuclear phagocyte system bypasses the requirement for CD4+ T cells in schistosome development , suggesting that innate cells of the mononuclear phagocyte system facilitate schistosome development and that CD4+ T cells influence the parasites indirectly by modulating monocyte/macrophage function . Together , our results demonstrate that blood flukes exploit innate immune signals to facilitate their development and suggest that the role of CD4+ T cells in this process may be limited to the provision of non-cognate help for innate mononuclear cell function .
To evaluate the role of T cell activation in schistosome development , we first examined schistosome development in Bcl10 ( Unigene accession no . Mm . 239141 ) -deficient and protein kinase C θ ( PKCθ; Mm . 329993 ) –deficient mice , where impairment of NF-κB activation in response to TCR ligation renders T cells unresponsive to antigen [10] , [11] . Unexpectedly , S . mansoni recovered from Bcl10-/- and wild type mice were indistinguishable in size ( Figure 1A ) and deposited comparable numbers of eggs ( Figure 1B ) , unlike schistosomes from recombination activating gene 1 ( RAG-1; Mm . 828 ) -deficient mice that exhibit a severely stunted phenotype and greatly reduced rates of egg production ( [5] and data not shown; see also Figure 1D and E ) . As expected , CD4+ T cell responses to S . mansoni were impaired by Bcl10 deletion , as measured by acquisition of an activated phenotype ( Figures S1A and B ) and cytokine production ( Figure S1C ) , suggesting that activation of CD4+ T cells through the TCR and subsequent CD4+ T cell responses are dispensable for normal schistosome development . Identical results were obtained with PKCθ-/- mice ( data not shown ) . To directly test whether Bcl10-/- CD4+ T cells can facilitate schistosome development , parasite development was examined in RAG-1-/- mice that were reconstituted with Bcl10-/- CD4+ T cells . Adoptive transfer of Bcl10-/- CD4+ T cells prior to infection partially restored worm growth ( Figures 1C and D ) , resulting in a phenotype that was intermediate between RAG-1-/- recipients of wild type CD4+ T cells and non-reconstituted RAG-1-/- mice . Likewise , adoptive transfer of Bcl10-/- CD4+ T cells produced an intermediate egg production phenotype that was not significantly different from RAG-1-/- recipients of wild type CD4+ T cells or non-reconstituted RAG-1-/- mice , even though egg production in these latter two groups was significantly different from each other ( Figure 1E ) . Identical results were obtained when adoptive transfers were performed with PKCθ-/- CD4+ T cells ( data not shown ) . Because TCR signals are required for T cell homeostasis [12] , Bcl10-/- CD4+ T cells exhibited variable levels of engraftment that were consistently lower than those of wild type cells ( Figure 1F ) . However , there was a significant positive correlation between the number of Bcl10-/- CD4+ T cells and the mean length of the worms recovered from each recipient ( Figure 1G ) . These data suggest that CD4+ T cells do not require intact antigen receptor signaling to facilitate parasite development . Further , our data suggest there is a simple requirement for sufficient numbers of CD4+ T cells to allow parasite development to proceed and predict a minimum of approximately 2×106 CD4+ T cells as the number required to restore parasite growth to wild type levels ( Figure 1G ) . To test whether TCR specificity for schistosome antigens is necessary for CD4+ T cells to facilitate schistosome development , we examined S . mansoni development in TCR-transgenic RAG-/- mice that transgenically express previously rearranged MHC class II-restricted TCRs specific for irrelevant antigens . In RAG-/- mice possessing only chicken ovalbumin ( OVA ) -specific ( OT-II/RAG-1-/- C57BL/6 , DO11 . 10/RAG-2-/- BALB/c ) or pigeon cytochrome C ( PCC ) -specific ( Cyt5-CC7/RAG-2-/- B10 . A ) CD4+ T cells , schistosome development was enhanced relative to RAG-/- controls that lack CD4+ T cells , as determined by assessment of parasite size ( Figures 2A and B ) and egg production ( Figure 2C ) . The extent to which parasite development was rescued varied between the different strains , with DO11 . 10/RAG-2-/- mice supporting parasite development that was comparable to wild type mice and parasites from OT-II/RAG-1-/- and Cyt5-CC7/RAG-2-/- exhibiting intermediate levels of development . As expected , OVA- and PCC-specific CD4+ T cells from S . mansoni-infected RAG-/- mice were completely unresponsive to schistosome antigens ( Figures S2A and S2B ) , though these cells retained the ability to respond to the appropriate antigen ( Figure S2B ) . These data demonstrate that CD4+ T cells do not require specificity for schistosome antigens and do not need antigen stimulation in order to facilitate schistosome development . Furthermore , the MHCII molecule by which the TCR is restricted is also likely irrelevant , as the three transgenic TCRs used recognize different murine MHCII molecules . Taken together , the data in Figures 1 and 2 suggest that CD4+ T cells influence the outcome of schistosome infection by mechanisms that are independent of antigen receptor specificity and signaling . Regardless of antigen specificity , naïve CD4+ T cells are maintained by active homeostatic processes [12] , requiring γc cytokines and interactions with professional APCs that express MHCII [13] , [14] , [15] , [16] , [17] , [18] . Because the effect of CD4+ T cells on schistosome development is independent of antigen receptor specificity and antigen-mediated activation ( Figures 1 and 2 ) , we hypothesized that steady-state homeostatic interactions between naïve CD4+ T cells and MHCII+ APCs in host tissues may play a role in promoting schistosome development . To identify transcriptional changes induced by the steady-state homeostatic interactions of naïve CD4+ T cells , we used a whole genome microarray to compare transcript levels in liver tissue from non-infected RAG-1-/- and OT-II/RAG-1-/- mice ( NCBI GEO series accession number GSE8340 . ) Using this approach , we identified 165 genes that were differentially expressed only in the presence of naïve CD4+ T cells ( Figure 3A and Table S1 ) . None of these genes are specifically expressed by CD4+ T cells , suggesting that OT-II CD4+ T cells contribute relatively little to the total hepatic RNA , being relatively few in number and transcriptionally quiescent . Supervised average-linkage hierarchical clustering of the expression data revealed that the transcriptional profile of hepatic tissue from OT-II/RAG-1-/- mice was more similar to that of wild type mice than RAG-1-/- mice , with the single largest grouping of genes comprising those that were expressed at higher levels in wild type and OT-II/RAG-1-/- mice when compared to RAG-1-/- mice ( Figure 3A ) . To identify biological functions associated with the differentially expressed genes , we employed a structured , knowledge-based approach to identify networks of differentially expressed genes with related functions [19] . Ingenuity Pathways Analysis software ( Ingenuity Systems , www . ingenuity . com ) was used to overlay differentially expressed genes onto a global molecular network developed from information contained in the Ingenuity Pathways Knowledge Base . Networks of up to 35 differentially expressed genes were then algorithmically generated based on the functional connectivity between the genes . Numerical network scores were calculated to rank networks according to their degree of relevance to the molecules in the dataset . Using this approach , five high-scoring networks ( score>20 ) of differentially expressed genes were identified , a representative of which ( score = 45 ) is illustrated in Figure 3B . Biological functions significantly associated with the networks , together with the genes associated with each function , are displayed in Table S2 . Biological functions with the most significant associations included Cell Morphology ( P = 1 . 37×10−4−4 . 80×10−2 ) , Cellular Compromise ( P = 1 . 37×10−4−4 . 31×10−2 ) , Cell Cycle ( P = 2 . 28×10−4−4 . 58×10−2 ) , Cell Death ( P = 6 . 30×10−4−4 . 31×10−2 ) and Cell-to-Cell Signaling and Interaction ( P = 7 . 51×10−4−4 . 31×10−2 ) , suggesting significant changes in cellular development within the liver in the presence of naïve CD4+ T cells . Interestingly , genes significantly up-regulated in the presence of naïve CD4+ T cells included Csf1/macrophage colony-stimulating factor ( M-CSF ) ( UniGene accession no . Mm . 795 ) , which is an essential growth factor for monocytes and monocyte-derived macrophages [20] , and other genes involved in myeloid cell development and function ( Myd116 ( Mm . 4048 ) [21] , Cd2 ( Mm . 22842 ) [22] , Gli1 ( Mm . 391450 ) [23] and Plld ( Mm . 29933 ) [24]; highlighted in Figure 3A ) , suggesting specific alterations in mononuclear phagocyte development . Other genes that were up-regulated in the presence of naïve CD4+ T cells are associated with the cellular response to stress ( highlighted in Figure 3A ) , including genes involved in the response to mis-folded proteins [25] , e . g . several chaperones ( Hsp90aa1 ( UniGene Accession No . Mm . 341186 ) [26] , Hsp90ab1 ( Mm . 2180 ) [26] , Hsp110 ( Mm . 270681 ) [27] , p23 ( Mm . 305816 ) [28] , Dnajb9 ( Mm . 27432 ) [29] ) , enzymes involved in protein folding ( Sep15 ( Mm . 29812 ) [30] , a disulfide isomerase ( Mm . 33692 ) [31] , a peptidyl-prolyl cis-trans isomerase ( Mm . 32842 ) ) , enzymes involved in protein degradation ( Ubxd2 ( Mm . 29812 ) [32] , the Syvn1 ubiquitin ligase ( Mm . 149870 ) [33] , Usp43 ( Mm . 158885 ) ) , and DNA-binding proteins involved in the transcriptional response to stress ( an XBP-1-like transcription factor ( Mm . 187453 ) [34] , the ER transmembrane transcription factor Creb3l2 ( Mm . 391651 ) [35] , H2afx ( Mm . 245931 ) [36] , and Junb ( Mm . 1167 ) [37] ) . Junb in particular is a central regulator of the cellular response to stress and is targeted for phosphorylation by stress-activated protein kinases [38] . Presumably because of the role of these cells in phagocytosis and in antigen processing and presentation , stress responses are constitutively active in immature APCs and are required for APC development and survival [39] , Indeed , the stress response gene Junb is also a critical regulator of myeloid cell differentiation and is classified as a myeloid differentiation primary response ( MyD ) gene [40] . Also up-regulated in the presence of naïve CD4+ T cells were genes involved in another critical aspect of phagocyte and APC function , namely vesicle formation and transport , e . g . Rabgap1l ( Mm . 25833 ) [41] , Arhgap21 ( Mm . 28507 ) [42] , Arfip2 ( Mm . 41637 ) [43] , Kif13b ( Mm . 23611 ) [44] , and Diap1 ( Mm . 195916 ) , which is also implicated in receptor-mediated phagocytosis in myeloid cells [45] . Thus , transcriptional changes consistent with alterations in mononuclear phagocyte and/or APC development and function correlate with the presence of naïve CD4+ T cells in OT-II/RAG-1-/- mice . To explore the possibility that mononuclear phagocyte development was altered by the presence of CD4+ T cells , leukocytes from the spleen and liver of non-infected RAG-1-/- and OT-II/RAG-1-/- mice were compared by flow cytometry . After gating to exclude granulocytes and OT-II T cells , mononuclear phagocytes were identified by expression of the myeloid marker CD11b ( Mm . 262106 ) [46] . In both mouse strains , CD11b+ cells segregated into two populations based on relative CD11b expression levels and forward scattering ( FSC ) properties – CD11bhi FSChi and CD11blo FSClo ( Figure 3C ) . However , the relative proportions of the two populations differed markedly in the two strains , with RAG-1-/- mice possessing relatively more CD11blo cells than OT-II/RAG-1-/- mice ( Figure 3C ) . Further analysis revealed that the two CD11b+ populations also differed significantly in expression of CD115 ( Mm . 22574 , the receptor for M-CSF ) , with CD11hi cells exhibiting higher CD115 expression than CD11lo cells ( Figure 3D; P = 0 . 0016 ) . Both CD11b+ populations also expressed Ly6C ( Mm . 1583 , a monocyte marker ) , although again , CD11bhi cells exhibited higher levels than CD11blo cells ( Figure 3D ) . Together , the greater forward scatter and elevated expression of CD115 by CD11bhi cells strongly suggest these are functionally more mature mononuclear phagocytes than CD11blo cells , as increased size and expression of CD11b and CD115 are all implicated in monocyte activation and maturation [47] , [48] , [49] . Finally , systematic analysis of CD11bhi and CD11blo populations in age-matched groups of non-infected OT-II/RAG-1-/- and RAG-1-/- mice revealed that cells of the functionally more mature CD11bhi phenotype were significantly more abundant in OT-II/RAG-1-/- mice than RAG-1-/- mice ( Figure 3E ) . Thus , our data suggest that , compared to OT-II/RAG-1-/- mice , RAG-1-/- mice exhibit a defect in mononuclear phagocyte development , with accumulation of cells that possess an immature CD11blo phenotype . Furthermore , this alteration in myeloid cell development correlates with reduced expression of genes that are required for mononuclear phagocyte development and function ( Figure 3A , Figure 3B , Tables S1 and S2 ) . As the only difference between RAG-1-/- and OT-II/RAG-1-/- mice is that , in the latter , transgenic expression of a previously rearranged MHC II-restricted TCR allows for development of a monospecific population of CD4+ T cells with specificity for chicken ovalbumin , our data suggest that steady-state MHCII-TCR-mediated interactions between mononuclear cells and naïve CD4+ T cells enhances developmental progression of mononuclear phagocytes to a functionally more mature state . Since enhanced parasite development and egg production in OT-II/RAG-1-/- mice correlated with enhanced steady-state maturation of monocytes in these animals , we hypothesized that non-responsive OT-II CD4+ T cells facilitate schistosome development indirectly , through steady-state interactions with mononuclear cells that promote monocyte maturation . To test this hypothesis , we tested whether direct stimulation of mononuclear phagocyte maturation , in the absence of any CD4+ T cells , could substitute for CD4+ T cells in restoring parasite development in RAG-/- mice . As the Toll-like receptor ( TLR ) -4 ( Mm . 38049 ) ligand lipopolysaccharide ( LPS ) has previously been shown to stimulate the in vivo maturation of monocytes [50] , [51] , we administered ultrapure LPS to schistosome-infected RAG-/- mice during pre-patent infection and assessed the effect on parasite growth and egg production at 6 weeks post infection . Consistent with other reports [50] , we found that LPS administration caused a significant decrease in the relative numbers of CD11b+ monocytes ( Figure 4A ) , indicative of the terminal maturation and subsequent apoptosis of these cells in response to TLR-4 ligation . Treatment of both BALB/c RAG-2-/- ( Figures 4B-D ) and C57BL/6 RAG-1-/- ( Figures 4E and 4F ) mice with LPS led to significant increases in both parasite length ( Figures 4B , 4C and 4E ) and egg production ( Figure 4D and 4F ) . Indeed , parasite growth ( Figure 4E ) and egg production ( Figure 4F ) were responsive to increasing doses of LPS . Together these data indicate that innate immune signals alone are sufficient to support schistosome development and suggest that blood flukes exploit the role of CD4+ T cells in providing help for mononuclear phagocyte maturation and function .
Parasites of the genus Schistosoma likely arose over 70 million years ago [52] and have undergone complex co-evolution with their definitive hosts , resulting in parasite adaptations that both evade and exploit host immune functions . Previous studies showed that schistosomes require host CD4+ T cells for normal development [5] , [53] , [54] and to mediate egress of eggs from the body across the bowel wall [3] , [55] . While egg excretion requires induction of an effector Th2 response to egg antigens [6] , the CD4+ T cell effector functions that facilitate blood fluke development have not been elucidated . Because previous studies failed to identify a specific role for either Th1 or Th2 responses in schistosome development [5] , [8] , the question of whether any CD4+ T effector functions are required to promote schistosome development remained unresolved . Therefore , the purpose of this study was to test whether antigen receptor signaling and subsequent activation of CD4+ T cells are necessary for normal parasite development to proceed . Our results show that , in contrast to the requirement for effector T cell responses to facilitate egg excretion , neither recognition of schistosome antigens nor TCR-mediated activation of CD4+ T cells are required for normal parasite development . Our findings suggest that none of the effector functions typically associated with CD4+ T cell responses are directly implicated in facilitating schistosome development , and may explain why previous attempts to identify a single T cell factor that modulates schistosome development , using knockout mice deficient in individual cytokines or their receptors [8] , have been unsuccessful . The schistosome requirement for CD4+ T cells , but the lack of necessity for traditional T cell effector functions , suggests that the steady-state homeostatic activities of naïve CD4+ T cells make the host environment more conducive to schistosome development . While the role of homeostatic interactions with MHCII+ cells in maintaining the peripheral CD4+ T cell pool is well established , the necessity of these same interactions for efficient APC maturation is being increasingly recognized [18] , [56] , [57] . In T cell-deficient mice , APC development and function are compromised but can be restored by reconstitution of peripheral T cell populations [18] . Furthermore , it was recently demonstrated that T cell conditioning of APCs for efficient maturation occurs under steady-state conditions before the initiation of T cell responses , in the absence of cognate antigen , and is mediated , at least in part , by the co-stimulatory molecule B7-H1 expressed by naïve T cells [57] . We propose that similar steady-state interactions between naïve OT-II T cells and MHCII+ cells in OT-II/RAG-1-/- mice account for the baseline increases in APC-related gene expression and the alterations in mononuclear cell maturation we observed in these mice . Given the ability of naïve T cells to prime for efficient APC maturation , we hypothesized that exploitation of APC function by schistosomes , rather than of CD4+ T cells directly , accounts for the enhanced parasite development we observed in TCR-transgenic RAG-/- mice . While we cannot exclude the possibility that unknown factors elaborated by resting , naïve T cells directly influence schistosome development , our hypothesis provides a parsimonious explanation for why naïve CD4+ T cells that do not respond to schistosome infection can still influence parasite development . That schistosome development can be restored in the complete absence of CD4+ T cells , through direct maturation of APCs , supports this hypothesis by demonstrating that CD4+ T cells are not directly required for parasite development . As the sentinels of the immune system , circulating monocytes and tissue macrophages and dendritic cells , also known as the mononuclear phagocyte system , specifically express pattern recognition receptors ( PRRs ) , including TLRs , which allow for the detection of invading pathogens [58] . Indeed , monocytes express high levels of TLRs and are the predominant producers of proinflammatory cytokines during endotoxic shock [59] . Thus while other cells in RAG-/- mice express PRRs , the monocytes , macrophages and dendritic cells are the predominant responders to pathogen-associated molecular patterns ( PAMPs ) such as LPS and are the likely mediators of the effect of LPS on schistosome development . Monocytes , macrophages and dendritic cells are ontologically related , as monocytes are the macrophage and dendritic cell precursors that migrate into the tissues , both during steady-state conditions and during infection . In response to infection , PRR ligation stimulates increased monocyte flux to meet the elevated demand for antigen-presenting and effector cells to combat infection [60] . Our demonstration that LPS administration restores schistosome development is therefore further evidence in support of the hypothesis we propose above , that schistosome development is influenced by innate mononuclear cell function . Indeed , our data point to mononuclear cell function as the common mechanism underpinning the enhancement of schistosome development by both naïve CD4+ T cells and LPS . This hypothesis is under further investigation . Several hypotheses can be proposed to explain how mononuclear phagocytes might influence shistosome development . Recruited to sites of infection and tissue damage , mononuclear cells produce cytokines and chemokines in response to activation that could act as cues for developing schistosomes [51] . Alternatively , the localized release of proinflammatory cytokines by mononuclear cells would be predicted to increase blood flow and vascular permeability , perhaps increasing the supply of host factors required for normal parasite development . At sites of inflammation , monocyte-derived cells also contribute to blood vessel remodeling and angiogenesis , by secreting angiogenic factors and trans-differentiating into endothelial cells [61] . As schistosomes are intravascular parasites and early parasite development occurs within portal venules , vessel remodeling may be required to allow for parasite growth . Yet another possibility is that mononuclear phagocytes may directly damage schistosomes , therefore requiring their local depletion for schistosome development to proceed normally . Activated macrophages and dendritic cells can produce nitric oxide [62] , a molecule that is toxic to developing schistosomes [63] and mediates immunity to schistosome infection in animals vaccinated with irradiated cercariae [64] . Priming of APC maturation by CD4+ T cells or TLR ligands might allow for more rapid progression of these cells to apoptosis in response to activation . In RAG-/- mice , developmental impairment of APCs might allow for their persistence , constituting a persistent source of nitric oxide and/or other molecules that impair schistosome development . If this is the case , analysis of innate responses to developing schistosomes in RAG-/- mice may identify innate effector mechanisms that can be harnessed to enhance immunity to schistosome infection . These hypotheses are currently being tested . Our data demonstrate that blood flukes do not respond to CD4+ T cells directly , but rather respond to signals that originate from the innate immune system . These findings corroborate previous studies where tumor necrosis factor ( TNF; Mm . 1293 ) , an innate proinflammatory cytokine , was shown to stimulate parasite egg laying in CD4+ T cell-deficient Prkdcscid/scid mice [65] . While TNF does not appear to stimulate parasite development directly [5] , [8] , the overlap between the TNF receptor and TLR signaling pathways could account for the ability of both ligands to enhance schistosome development [66] , [67] . Our findings also have implications for understanding epidemiological associations between schistosomiasis and other infections [68] , such as salmonellosis . While schistosome infection has been implicated in persistence of Salmonella infection [69] , our data also suggest that proinflammatory stimuli produced in response to bacterial LPS could play a role in exacerbating schistosome infection by supporting parasite development in co-infected individuals . These putative immunological and epidemiological associations are currently under investigation . The possible requirement of schistosomes for proinflammatory signals to support normal development is intriguing , as schistosomes , and helminths in general , are largely unable to stimulate such responses themselves due to a lack of potent TLR ligands [70] . It is therefore tempting to speculate that , under evolutionary pressure to avoid immune detection , schistosomes have lost the ability to stimulate the inflammatory feedback required for their successful development and now rely on other mechanisms to generate these essential inflammatory signals . In summary , our investigation of the mechanism by which CD4+ T cells facilitate schistosome development has revealed that blood flukes require neither CD4+ T cell responses nor associated effector functions . Furthermore , our data show that schistosomes do not respond directly to CD4+ T cells , as their requirement for these cells can be bypassed completely by direct stimulation of innate immune responses . Indeed , our data suggest that the role of CD4+ T cells in facilitating schistosome development may be limited to the provision of non-cognate T cell help for the maturation of MHCII+ APCs . We provide two lines of evidence in support of this hypothesis . First , non-responsive , naïve CD4+ T cells , which also condition immature APCs to undergo maturation , support improved parasite development . Second , direct stimulation of APC maturation in the absence of CD4+ T cells restores schistosome development . These data , together with previous findings that macrophages mediate vaccine-induced immunity to schistosome infection [64] , implicate mononuclear cells as central host determinants of the outcome of schistosome infection in the definitive host . A detailed understanding of the interactions between blood flukes and the mononuclear phagocyte system could therefore identify opportunities to modulate mononuclear cell function in ways that impair or prevent the establishment of schistosome infections .
All animal studies were performed in accordance with protocols approved by the USUHS Institutional Animal Care and Use Committee . RAG-1-/- mice were purchased from Jackson Laboratory ( Bar Harbor , ME ) and bred in-house to generate sufficient numbers for experiments . Bcl10-/- mice [11] were kindly provided by Dr . Tak Mak . OT-II mice [71] were the kind gift of Dr . Francis Carbone . Bcl10-/- and OT-II mice , originally with a mixed 129/C57BL/6 background , were backcrossed to the C57BL/6 background . OT-II mice were then bred with RAG-1-/- in house , to generate OTII/RAG-1-/- mice . C57BL/6 ( National Cancer Institute , Frederick , MD ) and 129× C57BL/6 F1 hybrid wild type mice ( Taconic ) were used as positive controls , although no differences in parasitological parameters were found in parasites recovered from either wild type strain ( data not shown ) . BALB/cTac-TgN ( DO11 . 10 ) -Rag2tm1 ( DO11 . 10/RAG-2-/- ) mice [72] and B10 . A/AiTac-[Tg]TCRCyt5CC7-I-[KO]-Rag2tm1 ( Cyt/RAG-2-/- ) mice [73] were kindly provided by Dr . Dragana Jankovic . RAG-2-/- and wild type mice on the BALB/c and B10 . A2 backgrounds ( Taconic ) were used as controls , respectively . Mice were infected percutaneously via the tail skin with 150 S . mansoni cercariae ( Puerto Rican strain ) shed from infected Biomphalaria glabrata snails [74] . All animal studies were performed in accordance with protocols approved by the USUHS Institutional Animal Care and Use Committee . Parasites were recovered from the portal system by perfusion [74] 42 days post-infection , immediately fixed in 4% neutral-buffered formaldehyde and photographed using a Nikon Coolpix 4500 4 . 0 megapixel digital camera connected to a Vistavision trinocular dissecting microscope at 20× magnification . Length of male parasites was determined from digital images using ImageJ software ( http://rsb . info . nih . gov/ij ) . Quantitative analysis of parasite length was performed on male worms as male schistosomes always outnumber females in experimental infections and female growth is significantly influenced by pairing with males [75] . Liver tissue was digested in 0 . 7% trypsin ( 50 ml ) in phosphate-buffered saline ( PBS ) for 2–3 hours at 37°C , and eggs were counted under a dissecting microscope . Egg production per schistosome pair was determined by dividing the total number of eggs calculated for each mouse liver by the number of parasite pairs recovered . Lymph nodes and spleens from wild type C57BL/6 mice or Bcl10-/- mice were dispersed through a 70-µm nylon strainer . Single cell suspensions were washed and red blood cells were lysed using ACK lysing buffer ( Quality Biological , Inc . ) Cells were incubated with anti-CD4 ( Mm . 2209 ) coated microbeads ( Miltenyi Biosciences ) and separated using Midi-Macs magnetic columns ( Miltenyi Biosciences ) . Flow cytometric analysis of isolated CD4+ T cells routinely demonstrated a purity of ∼99% . 3×106 cells suspended in PBS were transferred into RAG-1-/- mice by intravenous injection into a lateral tail vein . Control recipients received PBS alone . All tissue culture reagents used were free of endotoxin as determined by routine testing . Recipient animals were then infected with cercariae 24 hours post transfer , as described above . To verify the efficacy of adoptive transfers at necropsy , splenocytes from reconstituted RAG-1-/- mice were surface labeled with APC-Cy7-conjugated antibodies to CD4 , FITC-conjugated antibodies to CD8 ( Mm . 1858 ) , APC-conjugated antibodies to TCRβ ( Mm . 333026 ) , PE-conjugated antibodies to NK1 . 1 ( Mm . 6180 ) and PerCp-Cy5 . 5-conjugated antibodies to CD19 ( Mm . 4360 ) ( BD Biosciences ) and analyzed using a LSR II Optical Bench flow cytometer with FACSDiva and Winlist software , version 5 . 0 ( Verity Software House ) . The total number of T cells in each recipient after adoptive transfer was estimated by multiplying the percent of CD4+TCRβ+ cells by the total number of splenocytes for each mouse . CD11c+ cells were isolated from wild type spleens through incubation of cells suspensions with anti-CD11c ( Mm . 22378 ) coated microbeads ( Miltenyi Biosciences and separated using Midi-Macs magnetic columns ( Miltenyi Biosciences ) . CD4+ cells were isolated from spleens and livers of wild type , Bcl10-/- and OTII/RAG-1-/- mice as described above . CD4+ T cells and CD11c+ cells , pulsed with 50 µg/ml schistosome worm antigen preparation ( SWAP ) [76] , [77] , 5 µg/ml OVA or 1 µg/µl anti-CD3 ( Mm . 210361 ) ( BD Bioscience ) , were co-cultured for 72 hours in 96 well plates at a ratio of 5×105 CD4+ cells to 5×104 CD11c+ cells . Culture supernatants were assessed for IFNγ ( Mm . 240327 ) and IL-10 ( Mm . 874 ) by ELISA using BD Opt EIA Mouse IFNγ and IL-10 antibody pairs and ELISA reagents ( BD Bioscience ) and analyzed using a Spectramax M2 Plate reader ( Molecular Devices ) . For analysis of T cell activation in cells recovered from spleen and liver , expression of CD44 ( Mm . 423621 ) , CD62L ( Mm . 1461 ) , CD69 ( Mm . 74745 ) , and CD25 ( Mm . 915 ) was examined by flow cytometry , after gating on CD4+TCRβ+NK1 . 1- cells . Cells isolated from spleens and livers were surface labeled with FITC-conjugated antibodies to CD44 or CD69 , PE-conjugated antibodies to TCRβ or NK1 . 1 , PerCp-Cy5 . 5-conjugated antibodies to NK1 . 1 or CD4 , APC-conjugated antibodies to CD62L , and APC-Cy7-conjugated antibodies to CD4 or CD25 , ( BD Biosciences ) . For analysis of mononuclear populations , cells were stained with PE-conjugated anti-CD11b ( Mm . 262106 ) , PerCP-Cy5 . 5-conjugated anti-Ly6C ( Mm . 1583 ) and APC-conjugated anti-CD115 ( Mm . 22574 ) . Aqua dead cell stain ( Invitrogen ) was used to discriminate live and dead cells and granulocytes were excluded based on their high side scatter . All samples were analyzed using a LSR II Optical Bench flow cytometer with FACSDiva ( BD Biosciences ) and Winlist software , version 5 . 0 ( Verity Software House ) . Livers from naïve female wild type C57BL/6 , RAG-1-/- , and OT-II/RAG-1-/- ( N = 9 for each genotype ) mice were pooled into groups of three and prepared for cDNA microarray analysis . Briefly , RNA was isolated via RNAzol ( Tel-Test , Friendswood , TX ) and the RNeasy protocol ( Qiagen ) and analyzed for purity and concentration on a NanoDrop® ND-1000 Spectrophotometer ( Wilmington , DE ) . cDNA was prepared from two 30 µg aliquots of each pooled sample and labeled with either Cy3 or Cy5 fluorescent probes . One 30 µg aliquot from each pool was used to create a background control pool , while the second aliquot was used as the comparative sample . For further isolation and labeling protocol details please refer to http://www . niaid . nih . gov/dir/services/rtb/microarray/protocols . asp . Samples were hybridized as described by Schaupp [78] to Mmbe custom arrays manufactured by the NIAID Microarray Facility . Further information about these arrays can be found at http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GPL1057 . Hybridizations were performed in triplicate . Images were scanned by GenePix4000B Scanner ( Axon Instruments/Molecular Devices , Sunnyvale , CA ) and analyzed using the mAdb program ( http://madb . niaid . nih . gov/ ) . Signals were calculated as mean intensity – median background . Data were analyzed using Significance Analysis of Microarrays ( SAM ) , version 2 . 11 ( Stanford University ) and student's T tests ( EXCEL , with P values <0 . 05 considered significant ) to identify only those genes that exhibited differential expression in RAG-1-/- liver tissue compared to both OT-II/RAG-1-/- and wild type liver tissue ( i . e . genes differentially expressed in RAG-1-/- liver tissue compared to either OT-II/RAG-1-/- or wild type alone were not considered specific to the presence of naïve CD4+ T cells ) . We applied supervised average-linkage hierarchical clustering on differentially regulated genes , as implemented in the program Cluster version 2 . 11 ( M . Eisen; http://www . microarrays . org/software ) , separately to both the genes and arrays . The results were analyzed , and figures generated , using TreeView version 1 . 60 ( http://www . microarrays . org/software ) . Initial gene annotation was performed through GoMiner ( http://discover . nci . nih . gov/gominer/ ) . Functional analyses and networks of differentially expressed genes were then generated through the use of Ingenuity Pathways Analysis ( Ingenuity® Systems , www . ingenuity . com ) . To identify networks of differentially expressed genes that were functionally related , Ingenuity Pathways Analysis software was used to overlay differentially expressed genes onto a global molecular network developed from information contained in the Ingenuity Pathways Knowledge Base . To facilitate analysis , networks were limited to a size of up to 35 differentially expressed genes . A numerical network score was then calculated for each network , as described under Statistical Analysis below , to determine the probability of obtaining the same network by random chance . In RAG-/- mice , APC maturation was induced by biweekly intraperitoneal injection of low doses of ultrapure LPS , E . coli 0111:B4 ( InvivoGen , San Diego ) , at doses of 2 µg or 20 µg LPS/mouse . Monocyte numbers were monitored by flow cytometry , as described above . Because unequal variances were observed among some of the groups analyzed in this study , stringent non-parametric tests were used throughout to test the significance of differences between experimental groups . For two groups , significance of differences between experimental groups was tested using Mann-Whitney tests , and for three groups the significance of differences was tested using Kruskal-Wallis tests followed by Dunns' multiple comparison tests . Statistical analyses were performed with GraphPad Prism Version 4 . 0 software ( GraphPad Software , Inc . , San Diego , CA ) . P values of less than 0 . 05 were considered significant . Experiments were repeated at least twice , with 4–5 animals per group . For microarray data , Functional Analysis of differentially expressed genes was performed using Ingenuity Pathways Analysis ( Ingenuity® Systems , www . ingenuity . com ) to identify networks of differentially expressed genes that were functionally related . For putative networks , a numerical score was calculated from the hypergeometric distribution of the network using the right-tailed Fisher's Exact Test . The network score is the negative log of the Fisher's Exact Test P value . The probability that each biological function assigned to the network is due to chance alone was also tested using Fischer's Exact Test . P values of less than 0 . 05 were considered significant . The gene expression data discussed in this manuscript have been deposited in the NCBI's GEO database and are available under GEO series accession number GSE8340 . | Schistosomes , or blood flukes , cause a debilitating illness in millions of people worldwide , which manifests when inflammation develops in response to parasite eggs that become trapped in the liver and other organs . Paradoxically , schistosomes require signals from the host's immune system in order to develop fully into egg-producing adults . Previously , we showed that CD4+ T cells facilitate schistosome development . Here , we show that the mere presence of CD4+ T cells is sufficient for schistosome development to proceed . There is no requirement for these cells to respond to the parasite , or to exhibit any typical “effector” response . Two pieces of data suggest this effect on parasite development is mediated by antigen-presenting cells of the innate immune system such as monocytes and macrophages , which interact with CD4+ T cells by expressing MHC class II molecules . First , the presence of naïve CD4+ T cells correlates with baseline changes in the development of monocyte/macrophage populations . Second , direct stimulation of the monocyte-macrophage system restores parasite development , bypassing the requirement for CD4+ T cells in schistosome development . Understanding the mechanisms that promote or inhibit blood fluke infection may facilitate the development of new treatments and vaccines for schistosomiasis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"immunology/immune",
"response",
"immunology/innate",
"immunity",
"developmental",
"biology/microbial",
"growth",
"and",
"development",
"infectious",
"diseases/helminth",
"infections",
"immunology/leukocyte",
"signaling",
... | 2010 | Blood Fluke Exploitation of Non-Cognate CD4+ T Cell Help to Facilitate Parasite Development |
Chlamydia trachomatis is globally the predominant infectious cause of blindness and one of the most common bacterial causes of sexually transmitted infection . Infections of the conjunctiva cause the blinding disease trachoma , an immuno-pathological disease that is characterised by chronic conjunctival inflammation and fibrosis . The polymorphic Killer-cell Immunoglobulin-like Receptors ( KIR ) are found on Natural Killer cells and have co-evolved with the Human Leucocyte Antigen ( HLA ) class I system . Certain genetic constellations of KIR and HLA class I polymorphisms are associated with a number of diseases in which modulation of the innate responses to viral and intracellular bacterial pathogens is central . A sample of 134 Gambian pedigrees selected to contain at least one individual with conjunctival scarring in the F1 generation was used . Individuals ( n = 830 ) were genotyped for HLA class I and KIR gene families . Family Based Association Tests and Case Pseudo-control tests were used to extend tests for transmission disequilibrium to take full advantage of the family design , genetic model and phenotype . We found that the odds of trachomatous scarring increased with the number of genome copies of HLA-C2 ( C1/C2 OR = 2 . 29 BHP-value = 0 . 006; C2/C2 OR = 3 . 97 BHP-value = 0 . 0004 ) and further increased when both KIR2DL2 and KIR2DL3 ( C2/C2 OR = 5 . 95 BHP-value = 0 . 006 ) were present . To explain the observations in the context of chlamydial infection and trachoma we propose a two-stage model of response and disease that balances the cytolytic response of KIR expressing NK cells with the ability to secrete interferon gamma , a combination that may cause pathology . The data presented indicate that HLA-C genotypes are important determinants of conjunctival scarring in trachoma and that KIR2DL2/KIR2DL3 heterozygosity further increases risk of conjunctival scarring in individuals carrying HLA-C2 .
Chlamydia trachomatis ( Ct ) is an obligate intracellular bacterium [1] which causes significant morbidity as the causative factor of around 106 million new sexually transmitted infections per annum [2] . As the cause of trachoma , the same bacterium is the most common infectious cause of blindness [3] . Ct serovars exhibit highly specific tissue tropism , with serovars A–C being limited to the mucosal epithelium of the ocular conjunctiva . The remaining serovars are sexually transmitted , but whilst serovars D–K are limited to the mucosal epithelia of the genitourinary tract and rectum , the strains L1–L3 are able to invade other tissues including the lymph nodes . Ocular infection in trachoma is spread among young persons through exposure to secretions from the infected eye via direct physical contact , on fomites or by eye-seeking flies [4] . Repeated and prolonged cycles of infection and inflammation have been identified as the main factors that lead to the progressive formation of fibrotic scars on the tarsal conjunctiva , which ultimately becomes deformed . This can cause entropion and trachomatous trichiasis ( TT ) , a condition where the eyelashes turn inwards and irreversibly damage the cornea by scratching the globe of the eye . If left unchecked , TT causes corneal opacity , visual impairment and blindness . Active trachoma is frequently found in the absence of detectable Ct infection and both tissue damage and scarring are thought to be the result of a chronic immuno-pathological reaction [5] . Human conjunctival transcriptome studies in trachoma suggest that in addition to T cell and innate responses of epithelial cells , the activation and cytotoxic responses of natural killer ( NK ) cells is an important determinant of the severity of active trachoma [6] , [7] . NK cells are a rich source of multiple chemokines and cytokines , including interferon gamma ( IFNγ ) , a cytokine that is central to the control of chlamydial intracellular development and growth . IFNγ also has anti-fibrotic properties that can counteract the effects of TGF-β and inhibit fibroblast proliferation and collagen synthesis [8] , but when inappropriately expressed may cause immunopathology . NK cells in mucosal-associated lymphoid tissues are known to be important in the maintenance of epithelial cell integrity via production of the cytokine IL-22 [9] . NK cells therefore have the potential to fulfil multiple roles that encompass tissue homeostasis , tissue re-modelling and immunity . Early studies in murine chlamydial model infections found that NK cell depletion exacerbated disease , delayed clearance and limited the development of specific T cell responses [10] , [11] . Subsequent studies have confirmed that in response to chlamydial stimulation , NK cells are promoters of T cell immunity and a major source of IFNγ [10] , [12] but their role as lytic effector cells is less clear . Although Ct infected cell lines are lysed in vitro , NK cells purified from the peripheral blood of individuals with current chlamydial infection had diminished lytic activity ( and reduced IFNγ ) compared with uninfected controls [13] . Population diversity in the highly polymorphic genes that encode the variable NK receptors and their ligands [14] along with functional heterogeneity in the NK cell repertoire may account for these findings [15] . Trachoma is a complex inflammatory fibrotic disease in which host polymorphism in immune response genes plays a significant role [16]–[18] . The conjunctival epithelial surface is compromised in trachoma [5] as a result of the host response to the causative bacterium , which occupies an intracellular niche . Therefore the mechanisms used by NK cells in the control of other intracellular infections such as Hepatitis B [19] , Hepatitis C [20] and HIV [21]–[23] might also be effective against intracellular Ct . NK cells become activated when they are released from inhibition that is normally bound by interaction of specific HLA class I ligands with inhibitory Killer-cell Immunoglobulin-like Receptors ( KIRs ) [24] . The ligands of several inhibitory KIR have been described including HLA-A3 and HLA-A11 alleles , which are ligands of the KIR3DL2 receptor [25] , [26] and the HLA-Bw4 public epitope which is the ligand of KIR3DL1 [27] , [28] . HLA-C alleles can be classified ( according to a functional dimorphism at amino acid position 80 ) as carrying one of two KIR binding epitopes , which are known as HLA-C1 and HLA-C2 [29] . The HLA-C2 group of alleles ( HLA-C*02/04/05/06… ) are ligands of the inhibitory receptor KIR2DL1 [30]–[32] and its activating counterpart KIR2DS1 [33] . The HLA-C1 group alleles ( HLA-Cw*01/03/07/08… ) are ligands of both KIR2DL2 and KIR2DL3 [30]–[32] , however , the latter KIR are both able to cross-react ( with differing avidities ) with a small number of HLA-C2 and HLA-B allotypes [34] . Although germ-line encoded , the KIR gene system is highly polymorphic and exhibits extensive diversity both between individuals and between populations [35]–[38] . KIRs exhibit haplotype diversity such that different individuals possess variable gene contents . Since KIR and HLA are also found on different chromosomes , individuals can possess a KIR for which they have no cognate ligand , or vice versa . The extensive polymorphism in the KIR system culminates in a repertoire of NK cells within an individual that is more or less sensitive to release from inhibition under appropriate physiological conditions [39] . The strength of the signals mediated by interactions between specific HLA and KIR alleles is also highly variable [29] , [40] , [41] and this further limits overall NK cell responsiveness [42] . In part the responsiveness might be predicted by the presence of type ‘A’ and type ‘B’ KIR haplotypes . Type A haplotypes carry genes encoding predominantly inhibitory KIRs . B haplotypes contain some or all of the same genes found on A haplotypes , but additionally may carry the inhibitory KIR2DL2 and KIR2DL5 genes and numerous activating KIRs [36] . KIR haplotypes can be separated in to two variable regions , defined by their orientation towards the centromeric ( Cen ) or telomeric ( Tel ) regions of the chromosome [43] . The KIR A and B haplotypes are present in all populations studied to date and are thought to be maintained by the balancing selection pressures of infection , immunopathology and healthy reproduction [44]–[47] . In recent human history , a wide range of infectious diseases may have reduced the balancing effects in African populations , leading to more directional selection and a unique pattern of HLA and KIR diversity in this region [38] , [47] . We therefore assessed the extent to which host genotypes at the HLA and KIR loci were associated with trachomatous scarring in a trachoma endemic population from The Gambia .
The study was conducted in accordance with the tenets of the Declaration of Helsinki . The Ethics Committee of the Gambian Government/Medical Research Council Unit , and the ethics committee of the London School of Hygiene and Tropical Medicine approved the study ( MRC SCC1177 ) . Individual written informed consent was obtained from all adult participants . Written informed consent was obtained from a parent/guardian on behalf of those subjects aged <18 years who wished to take part in the study . All samples were anonymised . We selected a family study design and identified probands at a relatively early age for clinical signs of conjunctival scarring . This maximised statistical power whilst controlling for population stratification through the use of related control samples . The study population came from multiple regions of The Gambia and included multiplex families of mixed ethnic background . Families were ascertained through the identification of probands in which there were signs of trachomatous scarring at an early age ( age ≤30 years ) . This approach maximised the extent to which genetic rather than environmental factors could be expected to have contributed significantly to the probands' phenotypes . We recruited first-degree relations of the probands . In most cases this meant that we sampled both biological parents of the probands and all their ( self-described ) full siblings . Samples for DNA analysis were collected from buccal mucosae using sterile cyto-brushes ( Part Number F-440151 , SLS , Nottingham , UK ) . After collection , brushes were returned to their original packaging and stored dry at room temperature for up to 6 months [48] before DNA extraction was performed using a salting out procedure . An average of 4 offspring per family was assumed with a population prevalence of scarring in those <30 years of age in The Gambia of ∼2% [49] . The Pedigree Based Association Test ( PBAT ) v3 . 6 program [50] was used to calculate the power of the study to detect with 95% confidence ( α<0 . 05 ) a genetic association with odds ratios 1 . 5 , 2 and 3 when the hypothetical disease allele had a frequency between 0 . 01 and 0 . 50 . Figure S1 shows the estimated power of this study to detect genetic associations with trachomatous scarring at a range of allele frequencies and effect magnitudes , given the sample size . We had >90% power to detect an effect size greater than an odds ratio ( OR ) = 3 when the allele frequency was ≥0 . 05 and similar power to detect an effect size of OR = 2 when the allele frequency was ≥0 . 19 . Trachoma was graded in the field using the WHO simplified grading system by field supervisors certified for trachoma grading with regular performance checks as described by the PRET clinical trial manual of operations [51] . Left and right tarsal conjunctivae of all subjects were photographed as described by Derrick et al . [52] . Photographs were subsequently reviewed by two ophthalmologists with experience of grading trachoma and a final grade agreed . Subjects were assigned to the ‘scarred’ group if there were any signs of trachomatous scarring , in either eye . Individuals where phenotypes could not be confirmed for reasons of poor quality photography ( n = 5 ) did not contribute to the statistical tests of association . KIR genotyping for the presence or absence of 17 KIR genes was performed by PCR using the set of sequence specific primers described by Vilches et . al . [53] . The genotyping method was validated by participation in the UCLA Immunogenetics Center KIR exchange programme ( http://www . hla . ucla . edu/cellDna . htm ) . Medium resolution HLA-A , -B and -C genotyping was performed using LABtype sequence specific oligonucleotide probes ( OneLambda , Canoga Park , CA . USA ) on a Luminex platform ( Luminexcorp , Austin , TX . USA ) . Medium resolution HLA typing data generates strings of possible allele combinations . Information from the HLA genotypes of family members was used to reduce the length of the strings of possible allele pairs and to eliminate alleles that were not compatible with Mendelian inheritance within a given pedigree . Strings were further shortened where possible to include only common and well-defined alleles [54] . In order to maximise statistical power , highly sequence similar HLA alleles were combined in to groups ( table S1 ) before FBAT . KIR ligands of HLA ( HLA-A*03/11/Bw4 , HLA-B-Bw4 , HLA-C1/C2 ) were inferred from the full HLA genotypes of individual specimens rather than the reduced strings . The HLA-C*16:01 ( HLA-C1 ) and HLA-C*16:02 ( HLA-C2 ) alleles were frequently ambiguous and where this was the case alleles were assigned to HLA-C*16:01 because the HLA-C*16:02 allele has not been observed in other West African populations whilst HLA-C*16:01 is very common ( data from allelefrequencies . net ) . HLA types were used to identify cases of parental mis-assignment and inconsistent parent-offspring genotypes . KIR phenotypes ( presence/absence ) were tested for Mendelian inconsistencies . KIR2DL5 , KIR2DS3 and KIR2DS5 were not included in the association tests as they can segregate to both Cen-B and Tel-B regions and confound haplotype assignments . KIR gene frequencies were compared to those of other world populations using data from allelefrequencies . net and PCA using R . Family based tests of HLA association were carried out using FBAT v . 2 . 0 . 3 [55] performing a series of bi-allelic tests ( i . e . association of an index allele against all other alleles ) under an additive genetic model and the null hypothesis of no linkage and no association of any factor of the HLA system with trachomatous scarring . This approach is robust to effects of population structure [55] , [56] and is applicable to a data set with samples originating in mixed ethnic backgrounds . We tested for associations between scarring and all HLA alleles with a sample frequency greater than 0 . 05 with an offset value of 0 . 02 ( population prevalence of scarring in persons ≤30 years of age ) to allow the unaffected siblings to contribute to the test statistic . All FBAT p-values were adjusted using a conservative Bonferroni correction . Significant associations were tested again using a case/pseudo-control conditional logistic regression ( CLR ) [57] , which generated estimates of odds ratios and associated p-values . To test for independence between the disease-associated alleles , we included all alleles that had a corrected p<0 . 05 in a multivariate CLR model . To establish whether significant HLA associations were restricted to F1 subjects with specific KIR genotypes we tested the full data set under a genotype model [58] , [59] , using CLR , in different subsets of the F1 data where the population was limited by the KIR genotype . Because of the high linkage disequilibrium between factors of the KIR system , these tests were not considered to be independent and test statistics were corrected using the Benjamini-Hochberg method .
We sampled 830 individuals from 134 pedigrees and 146 nuclear families in which scarring trachoma had been identified in the first filial ( F1 ) generation . The self-described ethnic background of the parental ( P0 ) population ( n = 260 ) was approximately 40% Mandinka , 23% Fula , 15% Jola , 15% Wolof , 5% Bambara and 2% other minority ethnic groups . There were 570 persons in the F1 generation , where the gender distribution was 52% ( n = 296 ) male and 48% ( n = 274 ) female . The median number of offspring per pedigree was 4 ( range 1–11 ) . Eight families had one missing parent . There were 180 ( ∼32% ) cases of trachomatous scarring in the F1 generation and of these , 72 ( 40% ) were female and 108 ( 60% ) were male . Three hundred and eighty six ( ∼67 . 8% ) F1 individuals were unaffected and phenotypic status could not be confirmed for 4 ( <1% ) . Table 1 gives a detailed description of the phenotype distribution in the families . Detailed examination of photographs revealed that 12 probands did not have sufficient signs of trachomatous scarring . One proband could not be graded . In all the families where there was no photography confirmed scarred proband , at least one sibling was identified who was under 30 years of age and had signs of scarring . HLA genotyping identified paternal misassignment in 63 F1 individuals ( 11% ) who were reassigned to an unknown father but were otherwise retained for analysis . Table 2 shows the Family Based Association Test ( FBAT ) estimates of the HLA allele and KIR epitope frequencies in the sample population . Figure 1 describes the 64 unique KIR genotypes that were observed in the P0 generation . Thirty-eight additional KIR genotypes were revealed by re-assortment of the parental haplotypes in the F1 generation ( Figure 2 ) . All observed genotypes were assigned as either the ‘AA’ or ‘Bx’ genotypes ( where Bx includes both AB and BB genotypes ) for the full KIR region and where possible , for each of the Cen and Tel regions . A number of unusual genotypes were identified in this population , most notably , 10 . 4% of P0 individuals ( n = 27/260 ) possessed KIR2DL2 but not KIR2DS2 . Pairwise linkage disequilibrium data ( LD ) for the KIR genes were calculated ( figure S2 ) . Contrary to data from other studied human populations [58] , [60] , [61] and consistent with other findings within Africa [47] , we observed reduced LD between KIR genes . We did not identify any pairs of KIR genes that were in perfect LD ( r2 = 1 : only two of the four possible haplotypes observed ) , although a number of KIR genes were found to be in complete LD ( D′ = 1 : only three of the four possible haplotypes observed ) . The extent of LD was insufficient for high confidence imputation of missing KIR genotypes for use in FBAT [58] . Any HLA alleles and KIR epitopes with estimated frequencies above 0 . 05 were included in the FBAT . Three sets of HLA alleles were significantly associated with trachomatous scarring ( Table 2 ) . These were HLA-B*08:01 ( Z = −3 . 548 , p = 0 . 0004 , corrected p = 0 . 01 ) , HLA-C*03:04 ( Z = −3 . 201 , p = 0 . 0014 , corrected p = 0 . 04 ) and the KIR epitope HLA-C1/C2 ( Z = 3 . 622 , p = 0 . 0003 , corrected p = 0 . 008 ) . Only HLA-C1/C2 remained significant ( HLA-C2 , OR = 1 . 684 p = 0 . 0033 ) in a multivariate case/pseudo-control , additive model that included all three factors ( Table 3 ) , indicating that the HLA-C1/C2 epitope was the only significant independent factor of the HLA system that was associated with trachomatous scarring . In line with previous study designs and analyses we divided the data into several subsets [58] , [59] . We identified that in the majority of subsets , as with the unselected sample , the relative risk of scarring increased with the number of genomic copies of the HLA-C2 epitope in an additive manner ( Table 4 ) . The association of the HLA-C2 homozygote genotype with trachomatous scarring was restricted to the subsets of offspring who were KIR2DL2+ and KIR2DL3+ ( Cen-AB ) ( OR = 5 . 95 , p = 0 . 0025 , BH corrected p = 0 . 006 ) and to those who were KIR3DL1+ KIR3DS1− and KIR2DS1− ( Tel-AA ) ( HLA-C2 homozygote OR = 4 . 89 , p = 0 . 00006 , BH corrected p 0 . 0004 ) . Elevated odds ratios were observed in sensitivity analyses ( Table 4 ) in F1 samples where the case definition was restricted to those with moderate or severe ( WHO FPC grade C2 or C3 ) rather than evidence of any ( C1 , C2 or C3 ) scarring . We used Principle Components Analysis ( PCA ) to compare the KIR gene frequencies observed in the P0 generation of the Gambian trachoma families to those observed in other populations where data was available ( allelefrequencies . net database , ( Figure 3 ) ) . The proportions of the total variance explained by the first three principle components were 0 . 42 ( σ = 2 . 05 ) , 0 . 28 ( σ = 1 . 69 ) and 0 . 11 ( σ = 1 . 03 ) . The P0 specimens clustered with other populations of African descent , which could be recognised by the observation of high frequencies of the genes defining the Cen-B ( KIR2DS2 , KIR2DL2 ) and Tel-A ( KIR3DL1 and KIR2DS4 ) haplotypes .
The chlamydial protease , CPAF has been reported to interfere with the surface presentation of HLA class I molecules [14] , [64]–[66] , but recently this has been called in to question by Chen et al . [67] . Kägebein et al . [68] then demonstrated that Ct infection does not lead to alteration in normal MHC Class-I expression , maturation or surface presentation . This implies that Ct infected cells are unlikely to be targets for missing-self reactions mediated by NK cells which selectively monitor down-regulation or loss of self-type MHC class I on target cells . Instead it is more likely that cytotoxic NK responses in chlamydial infections are controlled by dynamic changes in the expression levels of activating NK receptors . These changes may occur as a result of infection and other environmental triggers [69] , [70] and might overwhelm the inhibitory effects of the more strictly expression-regulated [69] NK inhibitory pathways . HLA-C1:KIR2DL3 inhibited NK cells have weaker inhibitory signals than other HLA-C inhibited cells [29] and may have a lower threshold for activation . Khakoo et al . [40] reported that the HLA-C1/C1 KIR2DL3/2DL3 genotype constellation increased probability of clearance of early stage Hepatitis C Virus ( HCV ) infections . Ahlenstiel et al . [42] provided evidence that HLA-C1 homozygotes might be better able to challenge early infections by showing that the proportion of the total NK cell repertoire that is educated and inhibited by HLA-C is ∼50% greater in this group than that in HLA-C2 homozygotes [42] . The same study showed that HLA-C1 inhibited NK cells are better able to mount rapid , intense responses to infection through degranulation and IFNγ secretion [42] . In Ct infections , HLA-C1/C1 individuals may be able to limit chronicity by controlling the early stages of Ct infections with an NK response that is easily activated , and involves a more substantial component of the NK repertoire than in HLA-C2/C2 individuals . This may also be true of HLA-C2+ individuals who possess only weakly responsive KIR2DL1 alleles , such as those alleles that are found on the commonest B haplotypes in Caucasian populations [41] . However , in the Ga-Adangbe population of Ghana , there was a great diversity B haplotypes , none of which were found at high frequency and many of which carried non-attentuated KIR2DL1 alleles [38] . Any assumption about how the presence of Cen-B might indicate reduced cellular inhibition in Gambians should therefore be made with some caution . The role of KIR in mediating NK cytotoxic responses is well studied , but it is now clear that KIR expressing NK cells are also a major source of IFNγ [71] . The ability of NK cells to produce IFNγ in response to microbial stimuli is related to the density of NCAM-1 ( CD56 ) expressed on their surface , their KIR genotype and the degree of stimulus by accessory cells . An indication of the strength of regulation imposed by the KIR genotype can be estimated as a ratio , known as the ‘DIM factor’ , between the response of the CD56dim ( KIR-HLA dependent ) and CD56bright ( KIR-HLA independent ) IFNγ responding populations [71] . The majority of human NK cells in the periphery are CD56dim , express KIR and are susceptible to inhibition through KIR-HLA interaction . KIR genotype directly influences the DIM factor , but the exact genotypic conformation that defines this has yet to be elucidated . It has been proposed that the NK cell IFNγ response will be higher in individuals with more KIR educated NK cells , a situation found when there is a greater diversity of within-person inhibitory KIR genes . Experimentally , IFNγ production in CD56dim NK cells showed least inhibition ( and the highest DIM factor ) in KIR AB heterozygotes [71] . In HLA-C2 homozygotes , we observed a significant KIR2DL2/L3 heterozygote ( Cen-AB ) disadvantage ( Table 4 ) and an increased relative risk in those with the Tel-AA genotype . The number of persons with Tel-B genotypes was very low in this study , which reflects the low diversity in the Tel region that was reported in another West African population [38] . The high phenotypic frequency of KIR3DL1 ( Tel-A ) in this Gambian population ( ∼99 . 6% ) indicates that most individuals with the Cen-AB genotype possess at least one Tel-A haplotype . The Cen-AB , Tel-A+ genotype represents a full complement of the known MHC specific inhibitory KIRs ( KIR2DL1 , KIR2DL2 , KIR2DL3 , KIR3DL1 and KIR3DL2 ) and this genotype might define a high DIM factor [71] . NK cell clones with a Cen-AB genotype would therefore be relatively resistant to inhibition ( DIM factor >1 ) and would retain the potential for high IFNγ production . The KIR system exhibits extensive diversity in African populations [47] , [72] , [73] possibly driven by a high burden of life threatening infectious diseases , that have exerted strong ( diversifying ) selective pressures on each population [46] , [47] , [74] . The high prevalence of Ct STIs in some African populations has been implicated as a contributory factor to the high incidences of infection related infertility that are observed in Africa [75] . It is therefore surprising that Ct disease associated KIR and HLA genotypes are enriched in Africa . One explanation is that opposing selection pressures from other infectious diseases negate selection by Ct . Our sample was selected based on disease phenotype and we found KIR gene frequencies similar to other African populations ( Figure 3 ) . The Gambian samples are clearly separated from those in other geographical regions by high frequencies of the genes that define the Cen-B and Tel-A haplotypes ( Figure 3 ) [72] . The frequency of HLA-C2 epitopes is reported to be higher in African populations than in other populations [46] , [76] , [77] and the HLA-C epitope frequencies that we observed are similar to those previously described [77] . Yindom et al . [78] reported that the proportion of persons with the constellation HLA-C1 and KIR2DL2/KIR2DS2 ( Cen-B ) is higher in cases of malaria than in population matched , cord-blood controls [78] . In a study of a South-East Asian population , Hirayasu et al . [79] reported that natural selection may have reduced the frequency of the HLA-C1 and KIR2DL3 ( Cen-A ) because this genotype associates with cases of cerebral malaria . Both studies identify HLA-C1 in association with malarial disease , but they implicate different KIR Cen haplotypes . In the Gambian trachoma families , we observed that many Cen-B haplotypes lacked KIR2DS2 , whilst maintaining KIR2DL2 . This genotype has previously been identified in an African population [73] and its presence could be explained if KIR2DS2 , rather than KIR2DL2 , were mediating the Cen-B risk effect . The combined evidence of several TB studies shows that KIR Cen-A [80]–[82] and Tel-B [82] , [83] haplotypes associate with TB cases . We therefore suggest that the HLA-C2 homozygous , Cen-AB , Tel-A+ population are more resistant to the complications of both malaria and TB , but more susceptible to trachomatous scarring and that trachomatous scarring ( and possibly reduced fertility ) is the penalty of increased survival . We identified KIR-HLA interactions as an important contributory factor to risk of scarring . The HLA-C2 homozygous , KIR2DL2+ , KIR2DL3+ genotype associates with high relative risk of scarring . We suggest a model that may explain the data in which HLA-C2 may favour chronic infection , whilst KIR2DL2/L3 heterozygosity favours chronic inflammation ( Figure 4 ) . In some aspects this is similar to the model put forward by Hollenbach et al . to explain the observation of HLA-C1/C1 , KIR2DL2/L3 heterozygote risk in Crohn's disease [84] , which like trachoma is characterised by chronic inflammation and fibrotic immunopathology . It is possible that the high burden of trachomatous scarring , TT and infection related infertility in observed in Africa can be explained in part by unusually high frequencies of HLA-C2 and KIR2DL2/L3 heterozygosity and the effects of NK cell responsiveness . The therapeutic consequences of such a theory would impact on vaccine immune-therapies and we would expect that current efforts in the development of chlamydial vaccines , adjuvants and immunisation schedules would additionally monitor the boosting or modulating effects on the NK cell compartment . As early as the 1990 it was shown that vaccination with influenza virus was able to elicit NK cell responses [85] . More recent work has demonstrated that many vaccination regimes against viruses boost not only adaptive T and B cell response but also lead to repetitive expansions of NK cells [86]–[88] . Some immunologists have termed these “memory-like NK cell responses” and have now begun to consider the role of these responses in vaccine induced immunity [89] . The effectiveness of NK cells as targets of vaccine immuno-therapy has been described in oncology [90] . Efforts are now required to investigate the role of NK cells in immunity following vaccination with a wider spectrum of bacterial vectors and in natural immunity to infectious diseases such as trachoma . | Chlamydia trachomatis is a pathogen that causes sexually transmitted infections ( STIs ) and the blinding disease trachoma . Natural Killer ( NK ) cells are part of the host immune system's first line of defence against infection . NK cell functions are genetically encoded and differences between individuals mean that some people are better able to respond to infections than others . We found that in certain combinations , specific variants of the gene HLA-C ( Human Leucocyte Antigen , C ) and of a complex set of genes called the Killer-cell Immunoglobulin-like Receptors ( KIR ) were associated with a six-fold increase in the relative risk of scarring tissue damage resulting from ocular C . trachomatis infection ( trachoma ) . This combination of genetic variants may reduce the host's ability to effectively resolve infections and result in a harmful immune response that ultimately leads to tissue damage and scarring . KIR+ NK cells are potential cellular mediators of the damaging immune response . Previous studies have identified that the same HLA-KIR genetic constellation that associates with trachoma is actually protective against infectious diseases such as malaria and tuberculosis . The high frequency of the trachoma-associated constellation in African populations may therefore be explained by the evolutionary benefits of protection from the complications of severe disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"haplotypes",
"immune",
"cells",
"genetics",
"of",
"the",
"immune",
"system",
"major",
"histocompatibility",
"complex",
"nk",
"cells",
"genetic",
"association",
"studies",
"natural",
"selection",
"immunity",
"innate",
"immunity",
"genetics",
"immunity",
"to",
"infecti... | 2014 | Conjunctival Scarring in Trachoma Is Associated with the HLA-C Ligand of KIR and Is Exacerbated by Heterozygosity at KIR2DL2/KIR2DL3 |
cis-regulatory modules ( CRMs ) generate precise expression patterns by integrating numerous transcription factors ( TFs ) . Surprisingly , CRMs that control essential gene patterns can differ greatly in conservation , suggesting distinct constraints on TF binding sites . Here , we show that a highly conserved Distal-less regulatory element ( DCRE ) that controls gene expression in leg precursor cells recruits multiple Hox , Extradenticle ( Exd ) and Homothorax ( Hth ) complexes to mediate dual outputs: thoracic activation and abdominal repression . Using reporter assays , we found that abdominal repression is particularly robust , as neither individual binding site mutations nor a DNA binding deficient Hth protein abolished cooperative DNA binding and in vivo repression . Moreover , a re-engineered DCRE containing a distinct configuration of Hox , Exd , and Hth sites also mediated abdominal Hox repression . However , the re-engineered DCRE failed to perform additional segment-specific functions such as thoracic activation . These findings are consistent with two emerging concepts in gene regulation: First , the abdominal Hox/Exd/Hth factors utilize protein-protein and protein-DNA interactions to form repression complexes on flexible combinations of sites , consistent with the TF collective model of CRM organization . Second , the conserved DCRE mediates multiple cell-type specific outputs , consistent with recent findings that pleiotropic CRMs are associated with conserved TF binding and added evolutionary constraints .
The generation of cell-specific gene expression patterns during development is critical for proper morphogenesis . Gene expression at the transcriptional level is controlled by cis-regulatory modules ( CRMs ) , which recruit transcription factor ( TF ) complexes that alter RNA polymerase activity [1–4] . In general , CRMs are relatively short genomic regions containing clustered binding sites for numerous sequence-specific TFs . CRM activity is determined by which TFs are expressed in each cell and the ability of these TFs to form active transcription complexes on CRM sequences [2 , 5] . Recently , large-scale genomic studies have identified thousands of CRMs [6–11] . Furthermore , human studies have increasingly found disease-associated single-nucleotide polymorphisms ( SNPs ) within putative CRMs [6–11] . Hence , understanding how CRMs integrate the appropriate combination of TFs to yield cell-specific transcriptional outcomes is fundamental to understanding both normal development and disease . Two aspects of TF biology make it hard to predict CRM activity based on primary sequence . First , most TFs bind short degenerate DNA sequences present in high copy numbers throughout the genome [12] . Hence the number of potential genomic binding sites for a TF can exceed the number of TF molecules within a nucleus [13] . Second , the number of TFs encoded in the metazoan genome ( >1000 in the human genome ) makes predicting which specific TFs bind and regulate a CRM difficult [12] . For example , most TFs are members of large protein families that bind similar DNA sequences , yet CRMs are typically regulated by only one or a small subset of factors from each TF family [12] . Thus , the challenge lies in predicting which particular TFs will functionally bind which of the multitude of potential TF binding sites . To better understand this problem , three models have been proposed for how CRMs integrate transcriptional inputs: the enhanceosome , the billboard , and the TF collective [5 , 14 , 15] . All three models require clustered TF binding sites , but they differ in both sequence conservation and modes of TF recruitment . Enhanceosomes are highly conserved , and recruit a highly cooperative TF complex . Known enhanceosomes have rigid constraints on the order , spacing , and orientation of binding sites , and point mutations in single sites disrupt both complex formation and transcriptional output . The best-characterized enhanceosome is the interferon-β enhancer that coordinates the stepwise recruitment of a series of TFs to mediate high levels of transcriptional activation following viral infection [15 , 16] . In contrast , billboard CRMs are characterized by flexible orientations/spacing of binding sites that recruit TFs independently and are thereby under less evolutionary constraint [14 , 17] . The rapid evolution and rearrangement of binding sites within the even-skipped ( eve ) stripe 2 enhancer in dipterans supports the flexible billboard model [18–21] . The TF collective model proposes that groups of TFs form cooperative complexes on CRMs via a combination of protein-DNA and protein-protein interactions [5] . Unlike the enhanceosome , however , the TF collective posits that protein-protein interactions provide flexibility that eases binding site constraints . For example a TF can be recruited to CRMs lacking its binding site as long as there are sufficient sites for the other TFs of the collective . A collective of five TFs form transcription complexes on numerous CRMs containing various combinations of TF binding sites to regulate gene expression in the Drosophila heart [22] . The differing requirements for how TF sites are organized between the enhanceosome , billboard , and TF collective models may help explain the varying degree of sequence conservation between CRMs . Genomic sequencing of related species revealed that only a subset of CRMs involved in regulating developmentally important genes are highly conserved [23] . For example , the Drosophila vestigial boundary enhancer contains blocks of high sequence conservation , while the eve stripe 2 enhancer is not highly conserved at the sequence level [19 , 24 , 25] . This raises an interesting question; why are only some developmentally important CRMs highly conserved ? While the answer is currently unclear , one reason may be the different ways CRMs integrate TFs . The enhanceosome model requires tight constraints on TF binding sites consistent with high sequence conservation . In contrast the billboard and TF collective models relax constraints on binding sites , consistent with rapid sequence turnover . Unfortunately , few highly conserved CRMs have been thoroughly dissected and thus , we lack an understanding of which models best explain CRM function and conservation . The DMX is a conserved CRM that activates the Distal-less ( Dll ) appendage selector gene in thoracic segments to initiate leg development [26–28] . While activators that can stimulate the DMX are also present in the abdomen , DMX activity is restricted to the thorax via a highly conserved sequence ( the Distal-less conserved regulatory element , DCRE ) [27 , 28] . Previous studies demonstrated that the DCRE represses transcription by recruiting TF complexes containing abdominal Hox factors ( either Ultrabithorax ( Ubx ) or Abdominal-A ( Abd-A ) ) , Extradenticle ( Exd ) , Homothorax ( Hth ) , Engrailed ( En ) , and the FoxG Sloppy-paired TF ( Slp1 and Slp2 , referred to here as Slp ) [28 , 29] ( Fig 1A ) . Like the Hox factors , Exd ( vertebrate Pbx ) and Hth ( vertebrate Meis ) are conserved homeodomain TFs that regulate segment identity and cell fates along the anterior-posterior axis of metazoans [30–32] . Exd and Hth form cooperative TF complexes with Hox factors on DNA via several protein-protein interactions , and the DCRE recruits an abdominal Hox/Exd/Hth/Hox complex via two Hox binding sites ( Hox1 and Hox2 ) that are coupled to either adjacent Exd ( Exd1 ) or Hth sites ( Fig 1A ) . DCRE-mediated repression also requires compartment-specific inputs with an En site needed for posterior-compartment repression , and FoxG ( Slp ) sites are required for anterior-compartment repression ( Fig 1A and [28] ) . Based on the presence of high sequence conservation , one may reasonably predict that a highly conserved CRM such as the DCRE indicates constrained interactions between TFs as in the enhanceosome model . Here we provide evidence that despite high sequence conservation , the DCRE is most consistent with the TF collective model of CRM function . First , we used quantitative transgenic reporter and DNA binding assays to show that the DCRE contains an additional Exd/Hox site ( Exd0/Hox0 , Fig 1A ) , and that multiple combinations/configurations of linked Hox/cofactor binding sites can mediate robust transcriptional repression . Unlike the independent TF binding of the billboard model , however , we found that abdominal Hox , Exd , and Hth factors mediate cooperative TF complex formation on the DCRE . Moreover , cooperative complex formation and transcriptional repression can tolerate both individual DNA binding site mutations as well as deletion of the Hth DNA binding domain . These findings are consistent with the TF collective model of CRM function . However , we also found that the linked Hox/cofactor sites in the DCRE enhance thoracic Dll expression in a Hox-dependent manner , and that the re-configured Hox/cofactor binding sites failed to perform all DCRE-dependent functions . Taken together , these findings suggest that the pleiotropic functions of the DCRE ( thoracic activation and abdominal repression ) add constraints that limit sequence variation , thus providing a potential mechanistic understanding for why some CRMs are highly conserved .
Thoracic Distal-less ( Dll ) expression is essential for the specification of leg precursor cells of the Drosophila embryo [26 , 33] . Previous studies identified a conserved Dll CRM , the DMX , which mediates early thoracic leg expression [26] . DMX contains two distinct regions: the DMEact ( bp 1–661 ) , which activates gene expression in thoracic and abdominal segments , and an abdominal repression element ( Fig 1A ) [26 , 28 , 34] . The repression element has been defined several times based on different criteria including restriction enzyme sites ( “NRE-BX” bp 681–877 [26] ) , functional studies ( “DllR” bp 681–713 [27] ) , and genomic conservation ( “DMXR” bp 675–731 [28] ) . In this study , we use conservation across 21 Drosophila species to define the repression element as the Distal-less Conserved Regulatory Element ( DCRE ) , bp 662–731 , ( S1 Fig ) . This conserved sequence contains six previously characterized TF binding sites , including the linked Hox1/Exd1 and Hth/Hox2 sites that recruit a cooperative abdominal Hox complex as well as FoxG ( Slp ) and En binding sites , all of which are required for complete abdominal segment repression ( Fig 1A and [26 , 28 , 35] ) . Our current understanding of DMX function suggests the DMEact ( 1–661 ) mediates equal activation in all body segments ( thorax and abdomen ) and the DCRE ( 662–731 ) mediates abdominal repression to restrict expression to the thorax . To test these ideas , we integrated DMEact-lacZ ( DCRE-lacking ) and DMX-lacZ ( DCRE-containing ) into the same genomic locus and measured β-gal expression normalized to thoracic Dll expression in age-matched embryos . If the DCRE only contributes to abdominal repression , then DMEact-lacZ and DMX-lacZ embryos should have equal levels of thoracic expression . However , the DCRE-lacking DMEact-lacZ embryos express β-gal in significantly fewer thoracic cells , and those that do , express β-gal at lower levels when compared to DMX-lacZ embryos ( Fig 1B–1E ) . Next , we determined if the DMEact is capable of equal activation in thoracic and abdominal segments in the absence of the DCRE by comparing thoracic versus abdominal gene expression in DMEact-lacZ embryos . We found significantly fewer abdominal cells express β-gal and those that do have reduced levels compared to thoracic cells ( Fig 1C–1E ) . Taken together , these findings show that the DMEact and DCRE each contribute to thoracic and abdominal gene regulation , and together yield expression differences between the thorax and abdomen . Because thoracic and abdominal DMEact-lacZ levels differ , we hypothesized that abdominal Hox factors repress the DMX in a DCRE-independent manner . To test this idea , we mis-expressed Abd-A or Ubx using Paired-Gal4 ( PrdG4 ) and measured DMX-lacZ and DMEact-lacZ activity in the thorax . PrdG4 is active in every other segment , which allows for direct comparisons between experimental ( T2 ) and wild type segments ( T1/T3 ) . Care was taken to use conditions that express near physiological levels of Ubx and Abd-A ( see Materials and Methods ) . As expected , either Ubx or Abd-A repressed approximately 80% of DMX-lacZ activity in experimental ( T2 ) segments relative to control T3 segments ( Fig 1F and 1H and 1J ) . Importantly , either also repressed DMEact-lacZ , though to a lesser extent than DMX-lacZ ( ~40% , Fig 1G and 1I and 1J ) , indicating that Hox factors repress the DMEact either through direct binding or indirectly through the repression of thoracic activators . Thus , abdominal Hox factors repress the DMX through DCRE-dependent and DCRE-independent mechanisms . To better characterize Hox-mediated regulation of the DCRE , we generated two synthetic transgenic reporter assays to isolate the DCRE from other DMX regulatory sequences . First , we created an abdominal repression assay by placing lacZ under the control of three copies of the Grainyhead-binding element 1 ( GBE-lacZ ) ( Fig 2A ) . Embryos containing GBE-lacZ exhibit strong uniform epidermal expression during stage 15 [36] ( Fig 2B and 2C ) . Incorporating the DCRE ( GD-lacZ ) resulted in a pronounced decrease in β-gal expression within a subset of abdominal cells compared to GBE-lacZ embryos ( Fig 2B–2E ) . Previous studies showed that the DCRE mediates repression in a compartment-specific manner within the context of the DMX enhancer [28] . In the posterior compartment , abdominal Hox factors repress with Engrailed ( En ) , whereas in the anterior compartment they repress with the FoxG factors , Sloppy-paired ( Slp1 and Slp2 ) . In the GD-lacZ assay , the DCRE is sufficient to repress transcription within abdominal cells that express Slp ( Fig 2E ) . However , the DCRE is not sufficient for posterior compartment repression , suggesting that En and Hox repression through the DCRE requires additional sites within the DMEact . Quantification of β-gal levels in Slp2+ cells of GD-lacZ embryos revealed a 70% decrease in abdominal segments relative to thoracic segments , whereas β-gal levels were equivalent between Slp2+ thoracic cells and Slp2-negative thoracic and abdominal cells ( Fig 2F ) . Importantly , repression in Slp2+ cells is DCRE-dependent as no difference in β-gal was observed between thoracic and abdominal Slp2+ cells in GBE-lacZ embryos ( Fig 2G ) . Thus , GD-lacZ is a quantifiable assay to study the mechanisms of DCRE-mediated abdominal repression in Slp+ cells . The second synthetic reporter assay consists of lacZ under control of two copies of the upstream activation sequence ( UAS ) that can be activated by Gal4 ( 2xUAS-lacZ ) ( Fig 3A ) . When 2xUAS-lacZ is crossed to ubiquitous Gal4 drivers such as armadillo-Gal4 ( ArmG4 ) , relatively weak , stochastic expression is observed in stage 11 embryos ( Fig 3B ) . Incorporating the DCRE into the 2xUAS reporter ( 2xUD-lacZ ) and crossing to ArmG4 surprisingly did not reveal abdominal repression , suggesting the DCRE cannot repress Gal4-mediated activation ( Fig 3C and 3D and 3E ) . However , consistent with the DCRE enhancing thoracic expression in the context of the DMX , analysis of 2xUD-lacZ activity in the thorax revealed a 2 to 3 fold increase in β-gal levels relative to control 2XUAS-lacZ embryos ( Fig 3B and 3C and 3E ) . Note , we also observed enhanced thoracic expression relative to abdominal segments in early GD-lacZ embryos , but this difference is lost in older embryos due to the uniform increase in strength of the grainy-head activator ( compare thoracic reporter activity in Slp2+ and Slp2- cells in Fig 2F ) . To better quantify the effect the DCRE has on thoracic gene expression in the UAS assay , we incorporated a control 2xUAS-GFP reporter and found that while 2xUAS-GFP and 2xUAS-lacZ are both expressed stochastically , the relative levels of the two reporters are equal between the thorax and abdomen ( Fig 3B–3D ) . In contrast , β-gal expression from 2xUD-lacZ is significantly increased relative to 2xUAS-GFP expression in thoracic but not abdominal cells ( Fig 3D and 3E ) . A similar induction was observed using different drivers ( Tubulin-Gal4 , Daughterless-Gal4 ) yet no expression was observed in 2XUD-lacZ embryos lacking a Gal4 driver ( S2 Fig ) . Hence , the DCRE is insufficient to initiate gene expression on its own , but it can selectively enhance transcription in thoracic segments . Like abdominal repression in the GD-lacZ assay , enhanced thoracic activation of 2xUD-lacZ was observed in only a subset of cells , even though ArmG4 is active throughout these segments as shown by 2xUAS-GFP expression ( Fig 3C and 3D ) . Co-stains revealed that enhanced β-gal largely overlaps with Dll+ cells and a group of Vestigial ( Vg ) -positive cells that arise from the Dll+ leg primordia ( Fig 3F ) [33 , 37] . These results are consistent with the finding that the DCRE-containing DMX-lacZ expresses significantly higher β-gal in Dll+ cells of the thorax than the DCRE-lacking DMEact-lacZ ( Fig 1D and 1E ) . Altogether , these results support a model whereby the DCRE mediates multiple cell-specific transcriptional outputs: In the abdomen , the DCRE is sufficient to repress transcription in a cell-specific manner ( Slp+ cells ) in the anterior compartment . In addition , the DCRE contributes to abdominal repression in the posterior compartment in the context of the DMX [28] , but the DCRE is not sufficient to perform this function in isolation from the other DMX sequences . In the thorax , the DCRE functions as a conditional activation element that does not initiate expression but can increase transcription of both endogenous ( DMEact ) and heterologous ( 2xUAS ) enhancers in the leg primordia . Thus , the GD-lacZ and UD-lacZ assays provide tools that can be used to study the role of Hox , Exd , and Hth factors in regulating a subset of DCRE-mediated activities in isolation from the other DMX regulatory sequences . The published model of DCRE-mediated repression in the anterior compartment requires an abdominal Hox factor ( Ubx or Abd-A ) , the Exd and Hth cofactors , and a FoxG Slp factor [26 , 28] . However , genetic removal of hth , exd , or Slp results in severe embryonic defects , including the loss of wingless ( wg ) expression , which is required for DMX activation [33 , 38 , 39] . Since GD-lacZ does not require Wg for activation , it provides a useful tool for genetic tests of these factors . While a deletion removing both Slp1 and Slp2 ( Slp∆34b ) results in gross morphological abnormalities due to segmentation defects [29] , GD-lacZ expression is equal in the thorax and abdomen of Slp mutant embryos ( Fig 4A and 4B ) . Thus , Slp factors are required to mediate DCRE repression . To assess the roles of Hth and Exd , we took advantage of the finding that hth and exd are co-dependent for proper function; genetic removal of hth results in exclusion of Exd protein from the nucleus [30 , 40 , 41] . Hence , we assayed GD-lacZ activity in a severe hypomorph of hth ( hthP2 ) and found abdominal repression is abolished ( Fig 4C ) . Since abdominal Hox factors are expressed in both Slp and hth mutant embryos [40 , 42] , these findings demonstrate abdominal Hox factors are insufficient to mediate DCRE repression . However , at least one abdominal Hox factor is required for repression . GD-lacZ activity in single Ubx1 and Abd-AM1 , and double Ubx1Abd-AMX1 null embryos revealed that either abdominal Hox factor mediates DCRE-repression whereas removal of both abolishes repression ( S3 Fig ) . Together , these data support the model that the DCRE integrates abdominal Hox/Exd/Hth complexes with the Slp FoxG factors to repress abdominal gene expression . While a role for abdominal Hox factors in repressing Dll was previously established [26] , no prior studies revealed a role for a thoracic Hox factor in activating Dll . The best candidate for a potential positive regulator of Dll is the Antennapedia ( Antp ) Hox factor , as Antp and nuclear Exd/Hth are co-expressed with Dll in thoracic cells that activate 2XUD-lacZ ( Fig 5A and 5C and 5E ) . Moreover , the enhanced thoracic β-gal expression of 2XUD-lacZ is nearly eliminated in Antp25 null embryos as well as in HthP2 embryos that lack both Hth and nuclear Exd ( Fig 5B and 5D ) . These data suggest Antp directly contributes to thoracic Dll expression through the DCRE . To test this idea , we quantified Dll expression in Antp25 null mutants and heterozygous siblings and found a significant reduction of Dll levels ( ~40% , Fig 5E and 5F and 5K ) . In addition , we analyzed expression of DMX-lacZ and DMEact-lacZ in Antp25 mutants and found that the DCRE-containing DMX reporter lost over 50% of its thoracic activity in Antp25 null embryos whereas the DCRE-lacking DMEact reporter was not substantially different from heterozygous siblings ( Fig 5G–5J and 5L ) . These data are consistent with Antp increasing DMX-lacZ expression levels in a DCRE-dependent manner . The behavior of the DCRE in the GD-lacZ , UD-lacZ and DMX-lacZ reporters supports the idea that the DCRE conveys multiple transcriptional outcomes: thoracic activation versus abdominal repression . Moreover , genetic analysis revealed that both activities are Hox-dependent; Antp for activation and abdominal Hox factors for repression . To assess Hox factor binding to the DCRE , we performed comparative electromobility shift assays ( EMSAs ) using equimolar concentrations of Antp or Abd-A in the absence and presence of Exd/Hth . We found that Abd-A or Antp weakly bound the DCRE in the absence of Exd/Hth , whereas inclusion of Exd/Hth resulted in highly cooperative complex formation with either Hox factor ( Fig 6A and 6B and 6D and 6E ) . However , the Abd-A complex bound DCRE more strongly than Antp , and Abd-A formed a third , slower migrating complex not seen with Antp ( arrow in Fig 6E ) . Since previous studies had identified only two Hox sites , we scanned the DCRE and found a conserved region containing another potential Hox site preceded by a possible Exd site ( TTATG , the ‘Hox0’ site and GAAT , the Exd0 site , see Fig 1A ) . Interestingly , this region coincides with the ‘BX0’ site that was footprinted by an abdominal Hox factor [26] . To assess the nature of the Abd-A and Antp Hox complexes on the DCRE , we assayed complex formation on a series of probes containing one or two linked Hox/cofactor binding sites ( S4 Fig ) as well as on DCRE probes containing point mutations in one , two , or all three Hox sites ( S5 Fig ) . Neither Abd-A nor Antp formed strong complexes with Exd/Hth on probes containing individual Hox/cofactor sites . However , binding was increased cooperatively on probes containing two or more Hox/cofactor sites , and the number of molecular species observed increased according to the number of Hox/cofactor sites . These findings indicate that nearby Hox/cofactor binding sites contribute to cooperative DNA binding , even if the Hox/cofactor sites are suboptimal ( the Exd0 sequence differs from the consensus sequence and the Exd1/Hox1 site contains an unfavorable nucleotide between the sites ) . To assess the role of each Hox site in mediating DCRE-dependent repression and activation , we utilized site-selective mutagenesis in the GD-lacZ and 2XUD-lacZ assays and quantified gene expression . Though the DCRE mediates both thoracic activation and abdominal repression in the context of the DMX , our assays effectively separate the two processes , allowing us to compare and quantify embryos as follows: 1 ) GD-lacZ assay: By stage 15 of embryogenesis no difference in β-gal levels was measured between cells across the thoracic segment ( compare Slp2+ versus Slp2-negative thoracic cells in Fig 2F ) , indicating that localized DCRE-mediated thoracic activation is not observed at this stage of embryogenesis in the GD-lacZ assay . In addition , like the GBE-lacZ , no differences in levels were observed between Slp2-negative thoracic and abdominal cells in GD-lacZ embryos ( see Fig 2F ) . Thus , thoracic DCRE-mediated activation was negligible in the GD-lacZ assay of stage 15 embryos , and we made direct comparisons between the T3 segment and the remaining thoracic and abdominal segments . 2 ) UD-lacZ assay: Our data indicates that the DCRE does not mediate significant abdominal repression in the UD-lacZ assay . In fact , quantification of β-gal intensity relative to Dll intensity in 2xUAS-lacZ and 2xUD-lacZ embryos reveals the DCRE significantly alters thoracic but not abdominal expression ( Fig 3E ) . Thus , we normalized thoracic 2xUD-lacZ β-gal levels to the A1 segment for each construct . To assess the dependence of DCRE abdominal repression on Hox/Hox cofactor sites , we first generated mutations in each Hox site or Hox cofactor site in the GD-lacZ assay . In each case , we found a significant decrease in DCRE-mediated repression in Slp+ abdominal cells indicating that all sites are required for optimal repression ( Fig 7A and 7B and S6 Fig ) . However , no single point mutation abolished repression whereas double and triple Hox site mutations resulted in a complete loss of abdominal repression ( Fig 7B and S6 Fig ) . These findings are consistent with previous mutation analysis on the DMX , which revealed double site mutations were required to yield full de-repression [28] . Taken together with the Hox DNA binding assays , these results indicate that the multiple linked Hox/cofactor sites in the DCRE can mediate robust Abd-A/Exd/Hth complex formation capable of abdominal transcriptional repression . To assess whether thoracic activation by Antp/Exd/Hth complexes on the DCRE was also dependent upon the Hox binding sites , we analyzed the effect of single point mutations within each Hox site or Hox cofactor site using the 2XUD-lacZ assay . We found that thoracic activation was dependent upon both the Hox0 and Hox1 and their associated cofactor sites ( Exd0 and Exd1 , respectively ) but not the Hox2 or its associated Hth site ( Fig 7C and S7 Fig ) . Hence , unlike abdominal repression , thoracic activation in the 2XUD-lacZ assay is abolished by individual mutations in a subset of the Hox/cofactor binding sites . Of the three major models of CRM function ( billboard , enhanceosome , TF collective ) , our results are most congruent with Hox factors , especially Abd-A , functioning as a TF collective with Exd and Hth on the DCRE . First , unlike the all or none activity predicted by the enhanceosome model , the DCRE mediates significant repression even when individual TF binding sites are mutated in both the GD-lacZ and DMX assays ( Fig 7B and S6 Fig and [28] ) . Second , we found that unlike the independent binding of TFs predicted by the billboard model , Abd-A/Exd/Hth forms multiple cooperative complexes using several distinct binding sites , and can even do so with individual binding sites mutated ( S4 Fig and S5 Fig ) . An additional postulate of the TF collective is that not all TFs of the collective are required to directly bind DNA to contribute to transcriptional activity . Indeed , while individual point mutations within the sole Hth binding site decreased DCRE-mediated abdominal repression in the GD-lacZ assay , significant repression was still observed in this assay as well as in the context of the full DMX ( Fig 7 and [28] ) . As a further test of this idea , we used a hth point mutation ( allele hth100 . 1 ) that inserts a premature stop codon to generate homeodomain-less Hth proteins [43] . Importantly , this allele mimics a naturally occurring alternative splice isoform of Hth ( as well as the vertebrate Meis proteins ) , and while these Hth∆HD proteins fail to directly bind DNA , they still interact with and translocate Exd into the nucleus [44] . As expected , we found that 2XUD-lacZ activated thoracic expression in ArmG4;hth100 . 1 embryos to a level similar to wild type embryos , demonstrating that Hth DNA binding is not required for this activity ( Fig 8A and 8B ) . We also analyzed GDZ activity in hth100 . 1 embryos , and found significant repression in abdominal Slp2+ cells , albeit , the level of repression was reduced in hth100 . 1 embryos compared to wild type embryos ( 45% versus 70% repression , Fig 8C–8E ) . By comparison , repression is abolished in hthP2 null embryos ( Figs 8C and 4C ) . This data is consistent with a previous study that reported normal Dll and DMX expression in hth100 . 1 embryos [44] . We confirmed this finding by quantifying DMX-lacZ expression in wild type and hth100 . 1 embryos and found no significant difference in abdominal repression ( S8 Fig ) . We also tested Hox point mutant-carrying GD-lacZ reporters in the context of hth100 . 1 embryos . As expected , point mutations within the Hox2 site , which is linked to the adjacent Hth site , did not further decrease GD-lacZ dependent repression in hth100 . 1 embryos ( S6 Fig ) . In contrast , Hox1 point mutations in this genetic background lost all repression activity , a result that is consistent with the fact that multiple Hox/cofactor sites need to be mutated to abolish DCRE-mediate repression ( Fig 7B and S6 Fig ) . Next , we assessed whether the homeodomain-less Hth protein can contribute to cooperative Abd-A DNA binding on the DCRE . We also tested the role of Exd DNA binding on complex formation using an Exd protein containing a point homeodomain mutation ( N51A ) that disrupts DNA binding . Importantly , purified Exd/Hth∆HD ( Fig 6F ) and Exd51A/Hth ( Fig 6G ) heterodimers did not significantly bind the DCRE in the absence of Abd-A , even when added at a concentration three times higher than the wild type heterodimer ( compare second column of each EMSA to wild type Exd/Hth binding in Fig 6E ) . Inclusion of Abd-A , however , revealed that either DNA binding deficient heterodimer ( Exd/Hth∆HD or Exd51A/Hth ) stimulated significant cooperative Hox complex formation on the DCRE ( Fig 6F and 6G ) . To determine the independent role of Hth and Exd protein in complex formation , we performed EMSAs using Abd-A with only purified Exd or Hth ( Fig 6H and 6I ) . In contrast to the DNA binding deficient heterodimers , the addition of equimolar concentrations of Exd or Hth alone with Abd-A did not yield significant complex formation on the DCRE ( Fig 6H and 6I ) . These findings are consistent with the TF collective model of CRM function in which protein-protein interactions between Exd and Hth contribute to cooperative TF complex formation with Abd-A on the DCRE . To determine if different configurations of Hox/Exd/Hth sites could confer similar transcriptional outcomes , we replaced a subset of the Hox/cofactor sites within the DCRE with a distinct set of sites from another Hox-regulated CRM . Previous studies revealed that a rhomboid CRM ( RhoBAD ) mediates transcriptional activation in sensory organ precursors by integrating an Abd-A/Hth/Exd complex with the Pax2 TF [45 , 46] . The RhoBAD CRM contains separable binding sites for Pax2 and Abd-A/Hth/Exd ( Fig 9A ) . To determine if the Hox/Hth/Exd sites found in RhoBAD can function in transcriptional repression in the DCRE , we replaced the Hox1/Exd1-Hox2/Hth sites of the DCRE with the Hox/Hth/Exd sites from RhoBAD ( DCRE-RhoA , Fig 9A ) . This fusion transgene lacks the RhoBAD Pax2 site necessary for activation but contains the DCRE FoxG ( Slp ) sites as well as the Exd0/Hox0 sites that contribute to , but are not sufficient , for mediating repression . We found that the GD-RhoA-lacZ was able to substantially repress gene expression in Slp+ abdominal cells , although not as strongly as the wild type DCRE ( Fig 9B–9D ) . To determine if this modified element was sufficient to repress the DMX enhancer in the abdomen , we compared the activity of DMX-lacZ and DMX-RhoA-lacZ transgenes . Since the DCRE-RhoA element lacks the En site required for posterior compartment repression , significant de-repression in En+ cells was expected and observed in DMX-RhoA-lacZ ( Fig 9F ) . In contrast , repression of the DMX-RhoA-lacZ was comparable to that of DMX-lacZ in Slp+ abdominal cells ( Fig 9E–9G ) . However , similar to DMEact-lacZ , the DMX-RhoA-lacZ configuration of sites expressed decreased levels of β-gal in the thorax compared to the wild type DMX-lacZ ( Fig 9E–9G ) . Altogether , these findings demonstrate that while the DMX-RhoA configuration of Exd/Hth/Hox sites can mediate significant abdominal repression in Slp+ cells , this configuration of sites failed to perform two other DCRE-dependent activities ( posterior compartment repression in the abdomen and conditional activation in the thorax ) .
In spite thirty years of study , we lack a general understanding of how Hox factors gain sufficient specificity to differentially regulate cell fates along the anterior-posterior axis of metazoans . As monomers , Hox factors bind highly similar DNA sequences in vitro [47 , 48] . The discovery of two general Hox cofactors that also encode TFs , Exd ( vertebrate Pbx ) and Hth ( vertebrate Meis ) , suggested that the formation of TF complexes enhances Hox DNA binding affinity and specificity [32 , 49 , 50] . Consistent with this idea , the biochemical characterization of Exd/Hox binding sequences using SELEX-seq revealed DNA binding preferences between Hox factors are enhanced by Exd ( termed latent specificity ) [51] . The Forkhead ( Fkh ) CRM , for example , contains a unique Hox/Exd site that is specifically bound and regulated by a Sex combs reduced ( Scr ) /Exd complex [52 , 53] . More recent studies revealed that Exd also enhances Hox specificity by binding several low affinity sites . Crocker et al . found two CRMs from the shavenbaby ( svb ) locus that are activated in the abdomen by either Ubx/Exd or Abd-A/Exd complexes via low affinity sites [54] . Altering these sequences to high affinity Hox/Exd sites resulted in a loss of Hox specificity and transcriptional activation by anterior Hox factors . These findings suggest high affinity Hox/Exd sites are more likely to be pan-Hox target sequences regulated by numerous Hox factors whereas low affinity Hox/Exd sites provide specificity . In this study , we show that the DCRE mediates two opposing transcriptional outcomes using three linked Hox-cofactor binding sites . In the thorax , an Antp/Exd/Hth complex activates largely via two Hox/Exd sites , whereas the linked Hox/Hth sites are less important for DCRE-mediated activation . In the abdomen , all three Hox sites contribute to repression via the recruitment of several Abd-A/Exd/Hth complexes . Hence , the most specific Hox site within the DCRE is the linked Hth/Hox site that mainly contributes to abdominal repression by binding Abd-A and Ubx ( Fig 10 ) . In fact , directly linked Hth/Hox sites may be preferentially regulated by posterior Hox factors as the Abd-A specific target gene rhomboid ( rho ) contains a CRM that is activated via a linked Hth/Hox site [45 , 46]Additionally , biochemical studies using vertebrate Hox factors revealed that only posterior Hox factors form direct complexes with the Meis factor on DNA [55] . In contrast , both Exd/Hox sites within the DCRE are regulated by both thoracic Hox factors ( activation ) and abdominal Hox factors ( repression ) ( Fig 10 ) . Sequence analysis reveals that neither DCRE Exd/Hox site is optimal as an extra nucleotide is inserted between the Hox1 and Exd1 site whereas the Exd0 site has several mismatches to its consensus sequence ( S1 Fig ) . Moreover , DNA probes containing isolated Exd/Hox sites from the DCRE are poorly bound by Hox/Exd proteins , whereas combining these suboptimal sites resulted in the formation of Hox complexes that contribute to gene regulation . Thus , the DCRE uses multiple Hox/Hox cofactor sites to recruit distinct complexes that mediate two opposing transcriptional outcomes along the anterior-posterior axis . While a repression function for the DCRE was expected based on previous studies , the DCRE also contributes to Hox-mediated activation in the thorax . We termed the DCRE a ‘conditional’ activator in the thorax because it fails to initiate transcription , but when coupled to a ubiquitous activation element the DCRE enhances transcription in a subset of thoracic cells . Importantly , the cells that activate the DCRE derive from the endogenous Dll expression domain , and the DCRE contributes to activation of the DMX leg enhancer in an Antp-dependent manner . These data support the model that Antp and Exd/Hth are required for the conditional activation function of the DCRE . However , it is currently unclear why this activity is restricted to the Dll+ leg/wing primordium since Antp and Exd/Hth are broadly expressed throughout the thorax . One possibility is that , much like in the abdomen , an additional factor ( s ) interacts with the DCRE to provide position-specificity . How CRMs integrate transcription factor complexes to mediate cell-specific outputs remains an active area of study . The two best-known CRM models are the enhanceosome and the billboard . These models can be seen as extreme opposite ends of the spectrum of rigidity and constraints ( enhanceosome ) versus flexibility and adaptability ( billboard ) , with most CRMs likely to contain aspects of both models . Since many TFs use protein-protein interactions to promote cooperative complex formation on DNA , these interactions often place constraints on the order , orientation , and spacing of TF binding sites within CRMs . Hence , cooperative DNA binding has often been seen as evidence consistent with an enhanceosome model of CRM function . Dimerization between TFs such as the basic Helix-Loop-Helix ( bHLH ) proteins and retinoic acid receptors , for example , results in the formation of TF complexes that bind palindromic sequences with restrictions on distances between individual binding sites . In 2012 , Junion et al proposed an alternative role for protein-protein interactions between TFs [22] . Using a series of chromatin immunoprecipitation experiments , the Furlong lab found that a group of five TFs regulate a set of cardiac CRMs in the Drosophila embryo . Sequence analysis of co-regulated CRMs revealed combinatorial binding of these TFs does not require specific motif organization , a finding that is also consistent with the billboard model of CRM function . However , unlike the billboard , the TF collective does not require individual DNA binding sites for every TF to mediate appropriate functional outputs . Instead , a TF collective uses a combination of clustered DNA binding sites and protein-protein interactions to recruit large-scale TF complexes containing all the members of the collective . Although the biochemical basis of TF interactions between the five TFs was not explored , previous studies did find that a subset of these TFs form direct protein-protein interactions . Thus , Junion et al proposed the TF collective model of CRM function that predicts a common group of TFs can form many different cooperative complexes via multiple interactions between TFs , which results in greater CRM flexibility rather than rigidity in DNA binding site organization [22] . In this study , we provide evidence consistent with a Hox TF collective regulating early Dll expression in the Drosophila embryo . First , we show that the DCRE uses at least three distinct Hox sites that are each linked to an adjacent Exd or Hth binding site to recruit functional Hox complexes . Focusing on abdominal Hox-mediated repression , we used DNA binding assays and a synthetic reporter system ( GD-lacZ ) to reveal the following correlations between DNA binding affinity and transcriptional repression: 1 ) The wild type DCRE containing all three Hox sites yielded the strongest Abd-A/Exd/Hth binding and transcriptional repression in abdominal Slp+ cells . 2 ) Individual point mutations within any one Hox site partially compromised complex formation and repression . However , significant repression was still observed in the GD-lacZ assay , and in the DMX-lacZ assay single point mutations were still able to mediate abdominal repression in Slp+ cells in the DMX reporter [28] . 3 ) Mutations that compromise any two Hox sites or two Hox co-factor sites further decreased Abd-A/Exd/Hth complex formation , and abolished GD-lacZ-mediated abdominal repression . Consistent with the TF collective model , we found that Abd-A could still form robust complex formation on the DCRE even in the presence of DNA binding deficient Exd or Hth proteins , and genetic studies revealed that the DNA binding activity of one of the factors ( Hth∆HD ) is not required to mediate significant abdominal repression or thoracic activation . Moreover , we replaced the Hox1/Exd1-Hth/Hox2 sites with a distinct configuration of Exd/Hth/Hox sites from a different Abd-A regulated CRM and observed significant repression in both the GD-lacZ and DMX-lacZ assays ( Fig 10 ) . In total , these data demonstrate that , in the anterior compartment of the abdomen , multiple Hox/Exd/Hth binding site configurations can recruit a Hox TF collective capable of mediating robust transcriptional outputs . Interestingly , other Hox CRMs also contain characteristics consistent with TF collective enhancers . For example , congruent with variable binding of TFs in a collective , comparison of five mouse hindbrain enhancers controlled by HoxA1 and HoxB1 along with the Exd/Hth homologs , Pbx and Meis demonstrated that the presence , orientation , location , and sequence of the Meis sites are highly variable [56–61] . Additionally , the Hth homeodomainless protein is functional on other Hox-regulated CRMs , including the Fkh250 and Lab550 CRMs in Drosophila embryos [44] . Together , these results suggest that the DCRE is not unique among Hox CRMs in fitting the TF collective model . An unanswered question emerges from these studies: if interactions between members of the Hox TF collective permit added flexibility in binding site configurations , why is the DCRE so highly conserved across Drosophilid species ? One possible reason is that the DCRE mediates multiple opposing Hox-dependent outputs , which places added constraints on sequence conservation . For example , while replacing the Hox1/Exd1-Hth/Hox2 sites with the Exd/Hth/Hox configuration from the RhoBAD CRM can mediate strong repression in Slp+ anterior compartment cells , this configuration fails to repress gene expression in the posterior compartment due to the lack of an En binding site . Similarly , DCRE reporters containing this configuration of Hox/Hox cofactor sites also yielded lower levels of β-gal expression in the thorax , consistent with the idea that Antp fails to regulate linked Hth/Hox sites . Hence , we propose that the dual repression mechanisms of the DCRE in the anterior and posterior compartments of the abdomen as well as its conditional activation function in the thorax requires numerous TF sites , which thereby places evolutionary pressure to maintain sequence conservation . Several different hypotheses have been proposed for why some CRMs are highly conserved , including pleiotropic functions of CRMs placing added constraints on conservation [62–65] . Moreover , a recent vertebrate study comparing TF binding to syntenic regions of mouse and human genomes revealed that the most highly conserved TF binding activities were found on CRMs with pleiotropic functions in multiple cell types [66] . This study also noted that pleiotropic CRMs enrich for the co-association of many TFs . While this study did not score each CRM for nucleotide identity , their findings are consistent with our functional study on the DCRE and suggest that pleiotropy places added constraints on CRM sequence conservation .
The DMX [28] , DMEact ( basepairs 1–661 of DMX ) , 3xGrainyHead binding element1 ( 3xGBE ) [67] , and 2xUAS elements were generated by PCR ( sequences available upon request ) . DCRE-containing plasmids were created by ligating annealed complementary oligonucleotides containing restriction enzyme overhangs into the 3xGBE , or 2xUAS plasmids . Sequences of DCRE mutants are located in the figures . All enhancers were subcloned into the placZAttB plasmid . UAS-Abd-A was generated by PCR and subcloned into the pUAST-AttB plasmid . All plasmids were confirmed by DNA sequencing . Transgenic fly lines were generated by ΦC31 integration into the 51C insertion site [68] ( Injections by Rainbow Transgenics ) . The following fly lines were used: Antp25 , Ubx1 , hthP2 , PrdG4 , ArmG4 ( Bloomington Stock Center ) ; Abd-AMX1 , UbxMx12Abd-AM1 , Slp∆34b , UAS-Ubx ( Richard Mann , Columbia University , NY , USA ) ; hth100 . 1 ( Kurant et al . , 2001 ) ; UAS-Abd-A ( this work ) . Embryos were collected , fixed and stained using standard procedures at 25°C except for PrdG4;UAS-Abd-A and PrdG4:UAS-Ubx experiments which were performed at 18°C to lower Gal4 activity . The following primary antibodies were used: En ( mouse 1:10 ) ( Developmental Studies Hybridoma Bank , DSHB ) , Antp ( mouse 1:50 ) ( DSHB ) , Abd-A ( guinea pig 1:500 ) ( Li-Kroeger et al . , 2008 ) , Ubx ( mouse 1:20 ) ( Richard Mann ) ; Vestigial ( rabbit 1:25 ) ( Sean Carroll , University of Wisconsin-Madison , WI , USA ) ; and β-gal ( chicken 1:1000 ) ( Abcam ) . Antibodies were generated against Slp2 ( amino acids 1–275 ) and Dll ( full-length ) using purified His-tagged proteins injected into rats ( Cocalico Biologicals ) . Both the Slp2 and Dll sera were used at 1:500 . All immunostains were detected using fluorescent secondary antibodies ( Jackson Immunoresearch Inc . or Alexa Fluor , Molecular Probes ) . For quantitative analysis of gene expression , sets of embryos were harvested , fixed , and imaged under identical conditions at the same time . When possible , age-matched siblings were analyzed . For GD-lacZ and UD-lacZ assays , images used for quantification were taken using a single exposure time and normalized to segment T3 , A1 , or Dll expression levels within the same embryo as indicated . Pixel intensities and areas were measured using NIH-ImageJ software . The following proteins were purified from BL21 cells as previously described [27]: His-tagged Abd-A [69]; Antp [27]; his-Hth [70] and untagged Exd heterodimers [27]; his-Hth∆HD/Exd heterodimers [51]; his-Exd51A/Hth heterodimers [56]; his-Hth and his-Exd . Purified proteins were confirmed using SDS-PAGE and Coomassie blue staining and concentrations measured by Bradford assay . EMSAs were performed as previously described using native polyacrylimide gel electrophoresis [56] . Probes were used at 0 . 36 μM , and protein concentrations are noted in figure legends . The dried acrylamide gels were exposed to a phosphor screen for imaging using a StormScanner ( GE Healthcare ) . Densitometry was performed using ImageQuant 5 . 1 software . All EMSA experiments were performed in triplicate . | Enhancers are regulatory elements that interact with transcription factor proteins to control cell-specific gene expression during development . Surprisingly , only a subset of enhancers are highly conserved at the sequence level , even though the expression patterns they control are often conserved and essential for proper development . Why some enhancer sequences are highly conserved whereas others are not is not well understood . In this study , we characterize a highly conserved enhancer that regulates gene expression in leg precursor cells . We find that this enhancer has dual regulatory activities that include gene activation in thoracic segments and gene repression in abdominal segments . Surprisingly , we show that the conserved enhancer can tolerate numerous sequence changes yet mediate robust transcription factor binding and abdominal repression . These findings are consistent with abdominal transcription factors binding numerous different configurations of binding sites . So , why is this enhancer highly conserved ? We found that overlapping sequences within the enhancer also contribute to thoracic activation , suggesting the enhancer sequences are under added functional constraints . Altogether , our results provide new insights into why some enhancers are highly conserved at the sequence level while others can tolerate sequence changes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"abdomen",
"dna-binding",
"proteins",
"dna",
"transcription",
"developmental",
"biology",
"mutation",
"embryos",
"embryology",
"proteins",
"gene",
"expression",
"homeobox",
"biochemistry",
"point",
"mutation",
"anatomy",
"thorax",
... | 2016 | A Hox Transcription Factor Collective Binds a Highly Conserved Distal-less cis-Regulatory Module to Generate Robust Transcriptional Outcomes |
Alpha-methylacyl-coenzyme A racemase ( AMACR ) regulates peroxisomal β-oxidation of phytol-derived , branched-chain fatty acids from red meat and dairy products — suspected risk factors for colon carcinoma ( CCa ) . AMACR was first found overexpressed in prostate cancer but not in benign glands and is now an established diagnostic marker for prostate cancer . Aberrant expression of AMACR was recently reported in Cca; however , little is known about how this gene is abnormally activated in cancer . By using a panel of immunostained-laser-capture-microdissected clinical samples comprising the entire colon adenoma–carcinoma sequence , we show that deregulation of AMACR during colon carcinogenesis involves two nonrandom events , resulting in the mutually exclusive existence of double-deletion at CG3 and CG10 and deletion of CG12-16 in a newly identified CpG island within the core promoter of AMACR . The double-deletion at CG3 and CG10 was found to be a somatic lesion . It existed in histologically normal colonic glands and tubular adenomas with low AMACR expression and was absent in villous adenomas and all CCas expressing variable levels of AMACR . In contrast , deletion of CG12-16 was shown to be a constitutional allele with a frequency of 43% in a general population . Its prevalence reached 89% in moderately differentiated CCas strongly expressing AMACR but only existed at 14% in poorly differentiated CCas expressing little or no AMACR . The DNA sequences housing these deletions were found to be putative cis-regulatory elements for Sp1 at CG3 and CG10 , and ZNF202 at CG12-16 . Chromatin immunoprecipitation , siRNA knockdown , gel shift assay , ectopic expression , and promoter analyses supported the regulation by Sp1 and ZNF202 of AMACR gene expression in an opposite manner . Our findings identified key in vivo events and novel transcription factors responsible for AMACR regulation in CCas and suggested these AMACR deletions may have diagnostic/prognostic value for colon carcinogenesis .
Alpha-methylacyl-CoA racemase ( AMACR ) is a peroxisomal and mitochondrial enzyme that is indispensable in the catabolism of phytol-derived , 2-methyl-branched-chain fatty acids and the synthesis of bile acids [1] . In hepatocytes , AMACR catalyzes the conversion of pristanoyl-CoA and C27-bile acyl-CoAs from R- to S-stereoisomers , which are the only stereoisomers that can undergo β-oxidation . Bile acid intermediates undergo one round of β-oxidation in the peroxisomes and are secreted . In contrast , branched-chain fatty acid derivatives are transported to mitochondria , where they are further degraded to generate biological energy . Since most malignancies increase fatty acid utilization as an energy source to fuel growth [2] , it has been suggested that increased β-oxidation of branched-chain fatty acids provides transformed cells with a unique metabolic advantage [3] . This idea is supported by recent findings that knockdown of AMACR transcripts or inhibition of the racemase activity effectively blocked growth of prostate cancer ( PCa ) cells [4] , [5] . In humans , the major sources of phytol-derived , 2-methyl-branched fatty acids are dietary ruminant fats , meat , and dairy products . Increased consumption of these foods are known risk factors for prostate and colon carcinoma ( CCa ) [6] , [7] . Aberrant expression of AMACR was first reported in PCa and high-grade prostatic intraepithelial neoplasia but not in benign hyperplastic lesions or normal epithelia [8] , [9] . These findings quickly led to the establishment of AMACR as a reliable diagnostic marker for PCa [10]–[13] . More recently , overexpression of AMACR also was reported in CCa [14]–[18] , with a prevalence between 45% and 75% [19]–[21] . However , the relationship between levels of AMACR expression and the sequence of adenoma-carcinoma progression in the colon [22] has not been fully characterized . Except for a report that identified a non-canonical CCAAT enhancer element in the AMACR promoter [5] and a lack of regulation of this gene by androgen [16] , [23] , no information is available regarding how the AMACR gene is regulated . Furthermore , although recent studies have identified a few AMACR gene variants to be associated with PCa [24] , [25] or CCa [26] risks , a sequence polymorphism in the promoter region of AMACR has not been reported . Given the potential significance of AMACR in CCa , our objectives in this study were to determine the mechanisms of AMACR gene regulation in vivo during neoplastic transformation of the colon epithelium . Through the use of a comprehensive panel of immunostained-laser-capture-microdissected ( iLCM ) clinical samples comprising the entire colon adenoma-carcinoma sequence , we now report that the deregulation of AMACR during colon carcinogenesis involves non-random events , resulting in a double-deletion at CG3 and CG10 , and alterations in the frequencies of deletion of CG12-16 in a newly identified CpG island ( CGI ) located within the core promoter of AMACR . We also identified deletion of CG12-16 as a putative regulatory polymorphism and the double-deletion at CG3 and 10 as a somatic lesion . The DNA sequences housing these deletions were indicated to be a cis-regulatory element for Sp1 and a putative ZNF202-binding site , respectively , and to exert opposite effects on AMACR transcription .
We first provided a detailed description of the relationship between AMACR expression levels and the sequence of adenoma-carcinoma progression in the colon . The levels of AMACR in 55 foci representing seven normal , premalignant and malignant histological entities in 35 colon specimens were semiquantified in immunostained slides ( Figure 1A to 1H ) . These foci were subsequently microdissected for AMACR promoter studies . In general , AMACR immunostaining was negative to weak in normal cryptal ( Figure 1A ) and apical ( Figure 1B ) epithelia , as well as in tubular adenomas ( TAs ) with mild dysplasia ( Figure 1C ) . In contrast , villous adenomas ( VAs ) ( Figure 1D ) , well- ( Figure 1E and 1F ) and moderately ( Figure 1G ) differentiated adenocarcinomas expressed high levels of AMACR . AMACR immunostaining was almost absent to negligible in poorly differentiated carcinomas ( Figure 1H ) . Compared with the expression in normal crypt , levels of AMACR expression , represented as a score of 0 to 4 , were significantly increased in VAs and in well- and moderately differentiated carcinoma but not in normal apex , TAs , and poorly differentiated carcinoma ( Figure 1I ) . Because virtually no information is available on how AMACR is regulated in vivo , we initially were interested in determining if changes in DNA methylation status of the AMACR 5′ flanking promoter region play a role in gene regulation . In silico analysis revealed that AMACR transcripts share the same first exon with an 88-bp 5′ untranslated region ( 5′ UTR ) , suggesting that the gene is controlled by one promoter . Two CGIs were identified flanking the transcription start site ( Figure 2A ) . The first is a novel CGI located upstream of the ATG site ( −230 to −60; the position of the translation start site was set as +1 ) with 18 CG dinucleotides , whereas the second CGI downstream of the ATG site ( 48 to 357 , not shown in Figure 2A ) has been reported and shown to not be involved in gene regulation in PCa cells [5] . In concordance , our pilot studies indicated that the downstream CGI exhibited no differences in methylation/deletion/mutation status among the histological entities of the colon ( data not shown ) . Hence , subsequent studies were focused on analyses of the previously not reported proximal CGI in the AMACR promoter region ( the AMACR promoter CGI ) . Our bisulfite sequencing data did not support the involvement of DNA methylation of this newly identified CGI in AMACR gene regulation in vivo , since the promoter is largely unmethylated in all 55 iLCM samples ( next section ) . However , in silico analyses identified two putative Sp1 binding sites at CG3 and CG10 and a non-canonical ZNF202 [27] cis-element at CG12-16 of this CGI ( Figure 2B ) . Variable frequencies of deletions were found at these sites and later shown to be involved in gene regulation ( next section ) . A previously reported non-canonical CCAAT enhancer element [5] was aligned to CG5 . Two direct repeat sequences , 7 bp in length , were noted to flank the transcription start site . We later proposed that these two repeated sequences are involved in the generation of the CG12-16 deletion ( dotted lines; see Discussion below ) . A 222-bp region encompassing all 18 CG sites in the newly identified AMACR promoter CGI ( Figure 2 ) was analyzed for methylation , deletion , and mutation changes using DNA obtained from LCM samples and CCa cell lines . Bisulfite sequencing analyses of 239 alleles from 55 foci and regular DNA sequencing of 37 alleles from 9 foci as the control ( also see next section ) showed that most of the CG sites were unmethylated ( Table 1 ) . However , variable frequencies of deletions , methylation , and mutations were found to occur almost invariably at CG3 , CG10 , and CG12-16 , with deletions as the predominant lesion among all aberrations . The sequences of these deletion and mutation variants were deposited to Genbank with the accession number from EF636492 to EF636496 . Cluster analyses demonstrated that deletion of CG12-16 was the most common co-occurrence , followed by deletion at CG3 and CG10 ( double-deletion at CG3 and CG10 ) ( Figure 3A ) . Cluster analyses data for methylation ( Figure 3A ) , mutations ( Figure S1 ) , and all aberrations ( Figure S1 ) were also obtained . The number of deleted nucleotides ( nts ) was 2 to 8 nts at CG3 and 2 nts at CG10 ( Figure 3B ) . Deletion at CG12-16 was found to be precisely 20 nts . Among the four CCa cell lines examined , CG12-16 deletions were found in SW480 and SW620; no double-deletion of CG3 and 10 was detected in any of these cell lines . Thus , while methylation of this novel CGI does not appear to play a role in gene regulation , deletions of specific sequences or deletion hotspots within this sequence were identified and might play critical roles in the regulation of gene expression and/or the adenoma-carcinoma progression . As our focus on promoter assay will be based on the above sequencing results , we herein provide several pieces of data to ensure that the deletions were not artifacts of bisulfite-treatment of the DNA , PCR or sequencing . First , bisulfite modification reduced the GC content to ∼41% in the 222-bp AMACR promoter CGI , which made the sequencing easier to read; second , visual examination of sequencing chromatogram files showed clean and discrete peaks in the CGI region , indicating that the deletions we observed in bisulfite sequencing were not due to a GC compression artifact ( Figure S2A ) . In addition , as an internal control to ensure complete bisulfite modification , we routinely examined and found that almost 100% of the non-CpG cytosines in this region were converted to T , indicating complete bisulfite modification . To demonstrate that the CG12-16 deletion was not due to the PCR artifact , we used sequencing-verified plasmids with or without CG12-16 deletion as PCR templates , the PCR products showed expected size with different positions in 3% agarose gel ( Figure S2B , left panel ) . Additionally , we used unmodified ( not shown ) and bisulfite-treated genomic DNA , with or without the CG12-16 deletion , as templates and performed multiple PCRs on the same two samples ( Figure S2B , right panel ) . Results demonstrated the sizes of the amplicons derived from wild-type and deletion-variant templates were consistent , indicating that the deletion of CG12-16 was neither a PCR artifact nor a result of bisulfite-treatment . Blast searches provide additional evidence that the CG12-16 deletion exists in the human genome , as of the two genome sequences , one is the reference assembly that corresponds to the sequence ( NT_006576 . 15 ) without CG12-16 deletion and the other is the Celera assembly ( NW_922562 . 1 ) exhibiting the deletion , which exactly matches what we discovered in the AMACR promoter in clinical samples ( Figure S2C ) . Finally , we conducted parallel bisulfite and regular sequencing on DNA isolated from LCM-captured normal or malignant colon epithelial cells from 9 colon specimens . Identical sequence results were obtained with the two methods ( data not shown ) . Thus , in conclusion , these control experiments and in silico analyses demonstrate that the observed deletion hotspots in this CGI exist in colon tissues and are not results of artifacts generated from bisulfite-treatment , PCR or sequencing . We then investigated the relationship between deletion patterns in the AMACR promoter CGI and levels of AMACR expression ( Table 2 , left ) in 55 iLCM samples to gain insight into how these deletions might affect gene expression in vivo . CG3-only deletions were rather common ( 13–41% ) but were not correlated with AMACR expression , and CG10-only deletions were rare ( 0–5% ) . However , double CG3 and 10 deletions occurred at higher frequencies and invariably only in foci with no or little AMACR expression ( 17–28% , scores 0 and 1 ) . In contrast , CG12-16 deletions were common ( 21–67% ) and showed a positive correlation with the AMACR expression score . In total , foci with moderate and high AMACR expression ( scores 2–4 ) had a high frequency of CG12-16 deletions ( 53–67% ) and no double CG3 and 10 deletions . Next , we examined the type of deletions found in the six histological entities ( Table 2 , right ) to determine their relationship to the adenoma-carcinoma progression paradigm . CG3-only deletions were commonly found among normal and CCa foci . In most cases , GC10 deletions occurred as double CG3 and 10 deletions found in normal epithelium and TA ( 24–25% ) . In contrast , the double-deletion was not identified in VA or in CCa of any grade . CG12-16 deletions were found in all six histological entities; however , their frequency markedly increased in well- ( 56% ) and moderately ( 89% ) differentiated cancers and correlated with high AMACR expression in these lesions ( mean expression score ∼3; Figure 1E–G ) . It is of interest that the frequency of deletions of all kinds was low in poorly differentiated cancers; 72% of these foci have no lesions in the AMACR promoter CGI . Like all other CCas , they lack the deletion of CG3 and 10; the frequency of CG12-16 deletion in these CCas was low ( 14% ) , which correlates with negligible to low levels of AMACR expression in these lesions ( mean expression score ∼0; Figure 1H ) . Compared with the CG12-16 deletion in the moderately differentiated group that has the highest deletion rate , statistic analysis indicated the deletion was significantly changed in the normal , TA , VA and poorly differentiated groups but not in the well differentiated group . Together these data showed an intriguing in vivo phenomenon . Consistently , in all the samples analyzed , deletion of CG12-16 are not co-existed with double CG3 and 10 deletions ( frequency = 0; Table 2 ) . Additionally , double-deletions at CG3 and 10 are found only in normal epithelium and TA and are not observed in VA and CCa of all grades . In contrast , CG12-16 deletions are associated with moderate and well differentiated CCa that express high AMACR but not in poorly differentiated cancers that show negligible AMACR expression . These findings provide the impetus for a study of the effects of these deletions on AMACR transcription an in vitro system ( the HCT 116; see below ) . To better understand the relevance of these deletions to colon carcinogenesis , we must ask if these deletions are results of genetic events occurring in somatic cells of the colon or are constitutional alleles exist in the general population . Before this study , the only information available is that a sequence ( NW_922562 . 1 ) harboring the CG12-16 deletion in the Celera assembly ( Figure S2C ) . No AMACR sequences with deletion at CG3 or CG10 , or at both sites have been reported in genomic databases . We used randomly sampled genomic DNA isolated from whole blood of 96 individuals ( 48 males and 48 females ) from a relatively homogeneous Caucasian population of northern German for our study [28] . A 173 bp region encompassing all 18 CG sites within the AMACR promoter CGI were analyzed by regular and bisulfite sequencing ( Figure 2B ) . The CG12-16 deletion was found to be a sequence variant with an allele frequency of 43% in the population ( Table 3 and 4 ) . The observed genotype frequencies conform to the expectations of Hardy-Weinberg proportions ( Table 3 , p>0 . 05 ) . Between male and female samples , chi-square test for the genotype difference and allele frequency differences are not statistically significant ( p>0 . 05 ) . In contrast , in these blood DNA samples , no other deletions or mutations were found at any of the other CG sites in this region of the AMACR CGI , including CG3 and CG10 . Interestingly , although deletions/mutations at CG3 and/or CG10 were not found by normal sequencing , bisulfite sequencing demonstrated that the two CG sites are methylation hotspots in blood DNA samples , exhibiting a prevalence of 16 . 7% and 11 . 1% , respectively ( Table 4 ) . These frequencies were higher than those observed in tissue samples in which deletion is the predominant type of lesion at these two sites ( Table 1 ) . The fact that both single and double deletions at CG3 and CG10 are completely absent in blood samples but occur at frequencies between 13–30% in colon tissue DNA indicates that they are somatic lesions . To determine whether the in vivo deletions affect AMACR gene transcription , we first established that the human CCa cell line HCT 116 is a suitable model for AMACR promoter study in vitro . These cells express AMACR transcripts , have an intact promoter sequence with an unmethylated CGI ( data not shown ) , and therefore should have an intact “transcriptional machinery , ” including transcription factors for AMACR expression . Real-time RT-PCR showed that this cell line expresses both Sp1 and ZNF202 at significant levels . We cloned a long ( 1 , 818 bp; −1821/−4 ) and a short ( 599 bp; −602/−4 ) 5′ AMACR promoter sequence , both containing the newly identified CGI , into pGL3b reporter vector ( Figure 4A ) . The two sequences showed comparable promoter activities in HCT 116 cells . These data suggest the localization of core promoter elements within the 599-bp sequence ( AMACR599 ) , which was used to derive all other mutants in this study . To directly demonstrate that the deletion hotspots affect gene transcription , we generated deletion and/or mutation mutants of AMACR599 by targeting single or multiple sites ( Figure 4B ) . Since a previous study reported gene-regulatory activity of the CCAAT enhancer aligned to CG5 [5] , we also included deletion mutants targeting this sequence in our study . Reporter assays performed in HCT 116 cells showed that deletion of the CCAAT enhancer sequence at CG5 led to a marked reduction in promoter activity ( ∼60% ) regardless the integrity of CG3 , CG10 , or CG12-16 ( Figure 4C ) . However , in the presence of an intact CCAAT enhancer , deletion of CG12-16 , in the absence or presence of CG3 , CG10 , or double CG3 and 10 deletions , resulted in augmentation of promoter activity ( ∼100% ) . In contrast , deletion of CG3 and 10 , but not a single deletion of either CG3 or CG10 , caused a significant loss of promoter activity ( ∼60% ) . These findings indicate that deletion of CG12-16 and double-deletion of CG3 and 10 exert opposite actions on AMACR transcription . To demonstrate that these regulatory mechanisms are not limited to CCas , we transfected these mutants into two PCa cell lines ( PC-3 and LNCaP ) and similar data were obtained ( data not shown ) . We next sought to understand how these deletions affect AMACR gene transcription . In silico analyses suggest the localization of Sp1 binding sites at CG3 and CG10 and a non-canonical ZNF202 binding site within the CG12-16 region ( Figure 2B ) . However , it should be noted that in silico-based prediction requires experimental confirmation since recent ChIP-chip results have demonstrated a weak match between many consensus sequences and in vivo binding sites for specific transcription factors ( TFs ) [29] , [30] . Poor correlations could be due a high degree of degeneracy for some motifs and/or the participation of other proteins at the binding sites . A series of confirmation studies were therefore performed to support our in silico-based predictions . We predict that deletion at CG3 or CG10 affects one of the two putative Sp1 binding sites , and deletion at CG12-16 impede occupancy of a ZNF202 protein to its cis-element located between CG12-16 ( Figure 2B ) . Using nuclear extracts from HCT116 cells , chromatin immunoprecipitation ( ChIP ) experiments were performed . Sp1 was found binding to a 174-bp sequence ( −234/−60 ) that contains the two putative Sp1-sites at CG3 and CG 10 ( Figure 5A , upper panel ) but not to a 169-bp sequence ( 19553/19721 ) located in the last exon of AMACR ( Figure 5B , lower panel ) . Small interfering ( si ) RNA-mediated Sp1 knockdown decreased AMACR mRNA expression at the second-round of transfection ( Figure 5B ) but did not reduce transcript levels of glucuronidase β ( GUSB ) or cyclophilin A ( PPIA ) , two unrelated genes ( data not shown ) , in HCT 116 cells . Since there is no commercially available ZNF202 antibody for ChIP , gel shift assays were performed to assess HCT 116 nuclear protein binding to the putative ZNF202 binding site located within CG12-16 of the AMACR CGI . As can be seen in Figure 6A , one specific protein–DNA complex ( arrow ) was formed on the 45-bp 32P-labeled double-stranded oligonucleotide ( ODN ) encompassing CG12-16 and its flanking sequences ( Probe WT ) . The formation of this complex could be impeded by 100-fold excess of unlabeled WT or a 26-bp ZNF202 consensus sequence ( GnT; [27] ) . However , it is resistant to competition by 100-fold excess of a 45-bp mutant with the ZNF202 core sequence [31] mutated ( Mut ) or a 32-bp ODN devoid of CG12-16 ( Del ) . Interestingly , protein-DNA complex formation patterns on labeled WT and Del were different with notable absence of the lower band that could be competed off by excess cold WT or GnT ( Figure 6B ) . Finally , ectopic expression of ZNF202 induced a dose-dependent reduction of AMACR599 promoter activity and concordant lower levels of AMACR mRNA ( Figure 6C ) . In sum , these findings provide evidence in support of CG3 and CG10 as Sp1 binding sites and CG12-16 as a ZNF202 cis-element . Sp1 and ZNF202 appear to regulate AMACR expression in an opposite manner .
The main objective of this study was to elucidate the regulatory mechanism underpinning AMACR gene expression in relation to CCa development . We identified a novel CGI upstream the translation start site in the proximal core promoter of AMACR . Although aberrant methylation of promoter CGIs is a common cause of transcriptional deregulation of genes involved in tumorigenesis [32] , we found that AMACR activation did not occur by this mechanism during colon carcinogenesis . Instead , we found that two non-random , mutually exclusive in vivo events , involving a double-deletion at CG3 and 10 and the deletion of CG12-16 , play essential but opposite roles in the process . Additionally , we discovered the differential “origins” of these two in vivo deletions by comparing sequencing data from blood DNA in a general population and those from LCM-microdissected colon samples . The deletion of CG12-16 in the AMACR 5′ CGI was found to be a constitutional allele with a frequency of 43% in a general population . In contrast , deletions at CG3 and/or CG10 were not observed in the blood samples indicating that these are genetic events occurring in somatic cells of the colon . We observed a strong positive correlation between AMACR expression and the sequence of adenoma-carcinoma progression , suggesting a promotional function of AMACR in colon carcinogenesis . This postulate agrees with recent studies reporting that siRNA-mediated knockdown of AMACR mRNA or inhibition of the enzyme activity effectively curbed the growth of PCa cells [3] , [4] . Intriguingly , both gene expression and CCa progression were closely correlated with the status of two mutually exclusive deletions found in the iLCM samples . Specifically , the double CG3 and 10 deletion was found only in histologically normal colonic glands and TAs that had negligible to absent AMACR expression and was absent in VA or CCas of all grades that had variable levels of AMACR expression . More important , the simultaneous deletion of these two sites effectively negated AMACR transactivation in HCT 116 . We therefore propose that deletion at CG3 and 10 may effectively obviate colon carcinogenesis , possibly by impeding AMACR expression in vivo . In this regard , adenomas harboring double-deletions of CG3 and 10 might have a low likelihood of development to CCas; its potential diagnostic value merit further study . In contrast to CG3 and 10 double-deletions , deletions of CG12-16 were highly prevalent in well- and moderately differentiated CCas that strongly expressed AMACR . This finding is impressive because it stands in stark contrast with the classical view that deletions cause functional inactivation of genes . In this instance , the CG12-16 deletion in the AMACR promoter CGI behaves like a “gain-of-function” deletion . When viewed in this context , the CG12-16 deletion may be part of the sequential genetic changes that occur in tumor suppressors , DNA repair genes , and oncogenes during the development of CCas from adenomatous lesions [33] . Unlike the better differentiated CCas , most poorly differentiated cancers had a low percentage of deletions at CG12-16 and lacked AMACR expression . These cancers also had very few other aberrations including the double-deletion at CG3 and 10 , in their AMACR promoter CGI . Because the gene is not silenced by DNA methylation , or by irreversible genetic events such as deletions , we have to consider the possibility that these cancers may have a clonal origin different from that of the better differentiated carcinomas . Alternatively , during their evolution , these cancers may acquire a metabolic phenotype that is independent of AMACR overexpression . Much could be learned about the relationship between the CG12-16 deletion polymorphism and CCa risk by comparing the allelic frequency of this sequence variant in blood samples ( Table 3 & 4 ) to those found in LCM-captured histological entities of the colon ( Table 1 ) . The overall allelic frequency of this deletion amongst the 239 alleles from the various histological entities of the colon was found to be ∼42% ( Table 1 , row 1 ) , which matches the allele frequency observed in the blood samples is ∼43% . This suggests that there may not be additional somatic events altering the frequency of this constitutional sequence in the colon . Yet , the prevalence of this lesion reaches 89% in the moderately differentiated CCas , which showed significant difference ( p<0 . 05 ) compared to the normal , TA , VA and poorly differentiated but not the well differentiated samples . Collectively , these data adds up to the hypothesis that individuals with the CG12-16 deletion variant are more likely to develop CCas that are well or moderately differentiated . Conversely , those carrying the wild type variant may be more prone to develop poorly differentiated CCas . Clearly , such a provocative hypothesis would have to await a well-designed population study for confirmation . Contributing significantly to our understanding of how AMACR is regulated , we here provide the first evidence that deletion hotspots in the AMACR promoter CGI correspond to cis-elements for Sp1 and that this transcription factor regulates AMACR expression . Our data support the regulation of AMACR by Sp1 , as ChIP assays showed Sp1 binding to a region of the AMACR promoter CGI containing the predicted sites and siRNA-mediated Sp1 knockdown decreased AMACR mRNA levels in HCT 116 . Reporter assays revealed that single deletion at either CG3 or CG10 did not affect AMACR transcription , whereas the double-deletion significantly abrogated the promoter activity , suggesting that the integrity of one site at either CG3 or CG10 is sufficient to maintain the promoter activity . In contrast , deletion of CG12-16 enhanced AMACR transcription , signifying the likely presence of a repressor binding site in this region . We did not perform a ChIP assay for ZNF202 because no antibody was commercially available for the immunoprecipitation . However , results from gel shift assays were highly suggestive of the existence of a non-canonical ZNF202 binding site within this sequence . Our finding that ectopic overexpression of ZNF202 reduced AMACR promoter activity lends credence to this notion . However , we are aware of the fact that these data did not provide the definitive evidence that the CG12-16 sequence contains a ZNF202 cis-element , which still awaits a formal demonstration in future investigations . It is always possible that some unknown transcription factors other than ZNF202 could be involved in this regulation . Intriguingly , ZNF202 is a transcriptional repressor for genes affecting the vascular endothelium as well as lipid metabolism . We have examined the promoters of five other ZNF202 target genes [27] , [34] , [35] ( apoA4 , apoE , lecithin cholesterol acyltransferase , lipoprotein lipase , and phospholipid transfer protein ) and did not find deletions or other aberrations in their ZNF202 cis-element in colon and prostate cancer cells ( unpublished data ) . Thus , the activation of AMACR via deletion of a ZNF202 cis-element would be a phenomenon unique to AMACR gene regulation , if the CG12-16 sequence was shown to house this element . Several epidemiologic and animal studies have observed associations between the risk of metabolic syndromes/coronary heart diseases and the prevalence of colon adenomas/carcinomas [36]–[38] . Perhaps the loss of ZNF202-mediated repression of specific target genes , including AMACR , is a common cause of these diseases . Apropos of this view , carcinogenesis is being recognized increasingly as a metabolic disorder characterized by a shift from glycolysis to fatty acid utilization as the energy source fueling cell growth [2] . Finally , deletion of the CCAAT enhancer resulted in the loss of promoter activity regardless of the status of other elements , indicating that CCAAT enhancers are part of the basal transcriptional complex for AMACR . However , we did not find alterations in this cis-element in the iLCM samples , suggesting that such alterations do not contribute to aberrant expression of AMACR during colon carcinogenesis . At present , it is unclear how the deletions at in the AMACR promoter arise . However , first , we noticed that deletion hotspots at CG3 and CG10 are also methylation hotspots ( Table 1 and Table 4 ) . It has been reported that methylated CG sites are mutation hotspots [39] as suggested in Figure S3A . Second , scrutiny of the CGI sequence revealed two 7 nt direct repeats ( Figure 2B ) . We postulated that forward slipped-strand mispairing [40] , [41] of the repeats , may result in the CG12-16 deletion during DNA replication ( Figure S3B ) . If this mispairing happens , such slippage will cause the exact 20 bp deletion found in AMACR promoter . These proposed mechanisms speculated to be responsible for these deletions will of course have to await future experiments for corroboration . Collectively , we identified two major types of in vivo deletions in the AMACR promoter that appear to modulate gene expression and may play contrasting roles in carcinogenesis . In essence , a double-deletion at CG3 and 10 prevents AMACR overexpression and may impede colon carcinogenesis . In contrast , carriers of sequence variants with or without the CG12-16 deletion may have different propensity to develop well/moderately differentiated CCas versus the poorly differentiated cancers . Finally , our data suggest that these deletion hotspots are cis-elements for Sp1 at CG3 or CG10 and for ZNF202 at CG12-16 . The proposed mechanisms for AMACR promoter regulation and the deletion hotspots provided important platforms for the further study of AMACR gene deregulation during carcinogenesis .
Archival specimens were obtained from the Department of Pathology at the University of Massachusetts Medical School . Specimens from 35 cases were immunostained and microdissected to obtain the 55 iLCM samples: 11 TAs with mild dysplasia , 8 VAs , 6 well differentiated carcinomas , 6 moderately differentiated carcinomas , 7 poorly differentiated carcinomas , and 17 histologically normal colon tissues with 9 normal crypt and 8 apical surface epithelial samples . For the TAs , pronounced dysplastic changes , which often linked to positive AMACR , were uncommon . Most of the foci had mild dysplastic changes , and we focused our study of TAs on this type of sample . These samples were used for bisulfite sequencing analysis . Specimens for nine additional cases were obtained from the Pathology Department of the University of Cincinnati Medical Center and used to obtain nine LCM samples of normal epithelial , adenomatous , and carcinomatous cells for a regular DNA sequencing for comparison with bisulfite sequencing . Blood samples for polymorphism assay were from a relatively homogeneous Caucasian population of northern German [28] . The use of these samples was reviewed and approved by the respective institutional review boards at the two institutions . Multiple sections were cut from each case specimen . One section was stained with hematoxylin and eosin ( H&E ) and used for identification of histologic entities . The others were immunostained for AMACR with the P504S antibody ( Dako Cytomation , Carpinteria , CA ) and lightly counterstained with hematoxylin as previously described [8] , [42] . Areas representative of the histologic features and the overall intensity of AMACR expression found in a given case were identified in immunostained sections . These areas were then located in the replicate . The coverslips were then removed , H&E-stained , and microdissected as previously described [43] . Each of microdissected foci was given a score ( 0–4 ) reflective of the level of AMACR expression . When uniformly intense immunostaining was observed in at least 95% of cells in the section , the level of AMACR expression was designated as very strong ( score = 4 ) . If staining was less intense , not uniform throughout the section , and in fewer than 95% of the cells , the level of expression was designated as strong ( score = 3 ) . If the intensity of stain was weak , not uniform , and in 50% or fewer the cells , the section was graded as medium ( score = 2 ) or weak ( score = 1 ) . Cases were scored as negative ( score = 0 ) when the section showed no staining . Genomic DNA was extracted from the LCM samples by DNeasy Blood & Tissue Kit ( Qiagen , Valencia , CA ) with 20 µg of yeast tRNA added as a carrier . DNA was bisulfite-modified with the CGenome DNA Modification Kit ( Millipore , Billerica , MA ) . Sequencing service was provided by Macrogen ( Seoul , Korea ) with BigDye terminator used in a 96-capillary 3730xl DNA analyzer . Bisulfite-sequencing PCR-targeting AMACR promoter CGI was conducted by nested PCR . Primers AM-bisF1/AM-bisR1 and AM-bisF2/AM-bisR2 ( Table 5 and Figure 2 ) were used in the first round and nested PCR , respectively . The targeting region was from −276 to −55 , with the translation start site designated as +1 . PCR was performed with platinum Taq ( ABI/Invitrogen , Carlsbad , CA ) for 38 cycles with the annealing temperature at 56°C and 57°C in the first and nested PCR , respectively . Amplified fragments were purified in 1% agarose gel , TA-cloned , and about five colonies were picked from each sample for sequencing . Regular sequencing of the same CGI flanking region was performed in parallel using unmodified DNA samples and the regular primers AM-F1/AM-R1 and AM-F2/AM-R2 ( Table 5 ) . Proper controls were included in all experiments to ensure that the findings were not confounded by incomplete bisulfite modification , PCR artifact , or sequencing errors . Blood genomic DNA for the polymorphism study was extracted by DNeasy Blood & Tissue Kit . Using 50 ng genomic DNA as template , the PCR was performed for 40 cycles in the presence of 5% DMSO by platinum Taq with PolyF/PolyR as the primers ( Table 5 and Figure 2 ) . The annealing temperature was set at 58°C . The expected PCR product encompassing CpG sites 1–18 without CG12-16 deletion is 173 bp in length . After gel purification , the PCR products were TA cloned and the plasmids in colonies were directly amplified for sequencing by the Rolling Circle Amplification Kit ( GE Health Care , Piscataway , NJ ) . PCR products from alleles with the deletion of CG12-16 could also be visualized by a size difference from amplicons derived from wild type alleles in a 3% agarose gel . To determine the prevalence of methylation in this region of the AMACR promoter , aliquots of the extracted genomic DNA was subjected to bisulfite sequencing . The AMACR promoter region immediately upstream of the translation start site was amplified from genomic DNA of HCT116 cells . With forward primer pAM-F1 ( Table 5 , XhoI site underlined ) and reverse primer pAM-R0 ( HindIII site underlined ) used in PCR , the resulting 1818-bp AMACR promoter ( from −1821 to −4 ) was cloned into luciferase reporter vector pGL3b ( Promega , Madison , WI ) and designated as AMACR1818 . The promoter sequence was verified by sequencing . A 5′ truncated promoter ( designated as AMACR599 , from −602 to −4 ) was generated by nested PCR with PAM-F2/PAM-R0 as the primers . Promoter site-specific deletion variants were obtained by using the Genetailor site-directed mutagenesis kit ( Invitrogen ) . After sequencing , the promoter variants were released from the cloning vector and recloned into pGL3b . All reagents used for cell cultures , including heat-inactivated FBS , were obtained from Invitrogen . Human CCa cell lines HCT 116 , SW480 , SW620 , and DLD-1 were obtained from the American Type Culture Collection ( Manassas , VA ) . The cells were maintained in the same condition as HCT 116 cells , which are cultured according to the provider's recommendations . Unless specified , 6×104 HCT 116 cells were plated one day before transfection in each well of the 24-well plate . The cells were transfected with a total of 0 . 2 µg of DNA , including 10 ng of cotransfected CMV promoter-driven LacZ gene ( CMV-LacZ ) as the internal control . Plasmids for transfection were purified with the EndoFree Plasmid Maxi Kit from Qiagen . Two microliters of Plus and 1 µl of Lipofectamine ( Invitrogen ) were used in the transfection according to the protocol . The promoter activity was analyzed as previously described [44] . The ChIP assay was performed with the EZ ChIP Kit from Millipore according to the manufacture's instruction . A total of 7 . 5 µg of anti-Sp1 rabbit polyclonal IgG ( cat . no . 07-645 , Upstate/Millipore ) was used in each IP . Primers Sp1-IPf/Sp1-IPr targeting −234 to −60 CGI ( Table 5 ) were used in PCR with platinum Taq in the presence of 5% DMSO with an initial denaturation at 94°C for 1 min , followed by 36 cycles of 94°C for 30 sec , 58°C for 30 sec , and 72°C for 15 sec . As a negative control for DNA IP , primers ChIPnegF/ChIPnegR targeting the gene's last exon were used in PCR ( Table 5 ) . RNA extraction , reverse transcription , and real-time PCR , together with the primers for GAPDH and 18S rRNA , were described previously [44] . The tested primers used to detect AMACR and Sp1 transcripts were AMf/AMr and Sp1f/Sp1r , respectively ( Table 5 ) . As the siRNA control , primers for GUSB and PP1A gene were used in the real-time RT-PCR and listed in the Table 5 . The relative level of gene expression was calculated by the 2−ΔΔCt method as described in detail in our previous studies [44] , [45] . 1 . 5×105 HCT 116 cells were seeded at day -1 before transfection in each well of the 6-well plate . At day 0 , transfection was performed with 5 µl of Lipofectamine 2000 ( Invitrogen/ABI ) and 7 . 5 µl of 20 µM siRNA per well according to the protocol . siSp1 ( ON-TARGETplus SMARTpool , cat . no . L-026959-00 , Dharmacon , Lafayette , CO ) was used to knockdown Sp1 expression with Non-Targeting siRNA ( cat . No . D-001210-01-05 ) as the control . At day 3 , the cells either were collected for real-time RT-PCR analysis or were split at 1 . 5×105 cells per well . The second round of siRNA was performed on day 4 and analyzed on day 7 . To demonstrate the specificity of siRNA knockdown effects , in parallel , the expression of two unrelated genes of GUSB and PP1A were analyzed . Full-length coding sequence of ZNF202 m1 transcript [27] was amplified by primers NotIZ202 and Z202ApaI ( Table 5 ) from LNCaP cDNA . The sequencing-verified fragment was subcloned into pcDNA4/His/Max A expression vector ( Invitrogen ) . For real-time RT-PCR , the expression plasmid was transfected into HCT 116 in the 6-well plate with the Nucleofector Kit and Nucleofector II device from Amaxa ( Gaithersburg , MD ) . Probe sequences were shown in Table 5 . Complementary single-strand DNA oligos were annealed in 1×PCR buffer ( 20 mM Tris-HCl , 50 mM KCl , pH 8 . 4 ) in a water-filled heat block . The annealing mixture was heated at 95°C for 3 min and cooled to below 30°C in 1 hr to generate 50 µM double-strand oligo . The double-strand oligos showed a single and stronger band in 3% agarose gel , and located at a different position than the single-strand oligos ( photos not shown ) . HCT 116 cells nuclear extract was prepared by Nuclear Extract Kit ( Active Motif , Carlsbad , CA ) according to the manufacturer's instructions . Three µg of nuclear extract ( 1 µl ) was used in each binding assay at 18°C in 10 µl . The assays were carried out according to the protocol described in the Gel Shift Assay System ( Promega ) with the following modifications: Probe labeling was performed with 10 U T4 polynucleotide kinase ( New England Biolabs , Ipswich , MA ) and 2 µl [γ-32P]ATP ( 3 , 000 Ci/mmol at 10 mCi/ml , Perkin Elmer , Waltham , MA ) in a total volume of 10 µl at 37°C for 20 min . Electrophoresis of DNA-protein complexes was resolved in 6% DNA Retardation gel ( Invitrogen ) using 4°C 0 . 5× TBE buffer at 250 V for ∼35 min . Dried gels were exposed to X-ray film at −80°C for ∼1 hr and the images were captured by a digital camera . Five newly identified sequences of AMACR promoter variants with deletion/mutation at CpG hotspots were deposited into the Genbank ( http://www . ncbi . nlm . nih . gov/Genbank/ ) . The accession numbers for these variants are from EF636492 to EF636496 , which represent a CG3 deletion , a CG3 mutation , a CG10 deletion , CG3 and 10 double-deletions , and a CG12-16 deletion , respectively . In addition , the accession number for the AMACR promoter from the Genbank reference assembly and Celera assembly are NT_006576 . 15 and NW_922562 . 1 , respectively . The transcript reference sequences are NM_014324 . 4 and NM_203382 . 1 for AMACR , NM_003455 for ZNF202 m1 , NM_138473 . 2 for Sp1 , NM_000181 . 2 for GUSB , and NM_021130 . 3 for PP1A . Extensive gene analyses were carried out with GeneCards ( www . genecards . org ) . BLAST ( www . ncbi . nlm . nih . gov/BLAST/ ) was used to compare the sequence against Genbank . CGI was identified by MethPrimer at http://www . urogene . org/methprimer/index . html . Gene exon and intron information was obtained from Blat ( http://genome . ucsc . edu/ ) . PCR primers , except for real-time RT-PCR negative control ( Real Time Primers , Elkins Park , PA ) and bisulfite PCR , were designed by Primer3 [46] at http://frodo . wi . mit . edu/cgi-bin/primer3/primer3_www . cgi . The sequencing data were analyzed by ClustalW at http://www . ebi . ac . uk/clustalw . Putative TF binding sites in AMACR promoter deletion hotspots were scanned by MatInspector [31] . MatInspector utilizes transcription factor knowledge base to locate putative TF binding sites in sequence and minimize the number of false positive hits , but requires further confirmation through wet-bench works . It defines the “core sequence” ( Table 5 and Figure 2B ) of a putative binding site as the consecutive highest conserved positions in the DNA binding site . The hierarchical cluster analysis was based on the average linkage principle , and the absolute number of co-occurrences of different CG deletions was based on the similarity measure . The differences in AMACR expression ( Figure 1 ) in the microdissected foci were compared among the different histologic categories using a one-way analysis of variance ( nonparametric ) , followed by Tukey's HSD post hoc test for comparisons of all classes of lesions against normal cryptal cells . The analysis of CG12-16 deletion among the different histologic categories ( Table 2 ) was carried out by SAS Proc Genmod software that assuming a log link and robust standard error estimation . The program estimates and tests differences between groups with respect to the proportion of deletion . A generalized linear model of binomial proportions was analyzed to detect differences . In other experiments , a two-tailed , unpaired t-test was performed between two groups . Except else where mentioned , the columns with error bars in the figures represent mean±95% confidence interval . For the CG12-16 deletion polymorphism study , Hardy-Weinberg equilibrium was used to test if specific disturbing influences are introduced to the samples , and chi-square test was used to exam genotypic and allelic differences between male and female . In all the analyses in this paper , unless otherwise stated , p<0 . 05 was considered as statistically significant . | Men consuming high amounts of red meat and dairy products are at a higher risk of developing colon and prostate cancer . Alpha-methylacyl-coenzyme A racemase ( AMACR ) is an enzyme that helps to break down fat from these foods to produce energy . An increase in the utilization of energy from fat is a hallmark of many cancers including colon and prostate cancers . Indeed , the AMACR gene was first found to be abnormally active in prostate cancers , and its abnormal expression has become a diagnostic marker for the cancer . However , little is known about how AMACR becomes activated in cancer cells . Here , we show that AMACR is also highly expressed in certain stages of colon cancer , though not all stages . A close examination of the AMACR gene in a panel of normal and progressively malignant colon tissues reveals that deletions of specific sequences in the AMACR gene may trigger its abnormal expression during the evolution of colon cancer . We also identify unique proteins known as “transcription factors” that normally bind to these deleted sequences to maintain normal expression of the gene . Finally , we report a new deletion variant of the AMACR gene in the general population that may influence the course of colon carcinogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"oncology",
"genetics",
"and",
"genomics/gene",
"expression",
"pathology/immunology",
"molecular",
"biology",
"cell",
"biology/gene",
"expression"
] | 2009 | Deletion Hotspots in AMACR Promoter CpG Island Are cis-Regulatory Elements Controlling the Gene Expression in the Colon |
Regulatory T ( Treg ) cells dampen an exaggerated immune response to viral infections in order to avoid immunopathology . Cytomegaloviruses ( CMVs ) are herpesviruses usually causing asymptomatic infection in immunocompetent hosts and induce strong cellular immunity which provides protection against CMV disease . It remains unclear how these persistent viruses manage to avoid induction of immunopathology not only during the acute infection but also during life-long persistence and virus reactivation . This may be due to numerous viral immunoevasion strategies used to specifically modulate immune responses but also induction of Treg cells by CMV infection . Here we demonstrate that liver Treg cells are strongly induced in mice infected with murine CMV ( MCMV ) . The depletion of Treg cells results in severe hepatitis and liver damage without alterations in the virus load . Moreover , liver Treg cells show a high expression of ST2 , a cellular receptor for tissue alarmin IL-33 , which is strongly upregulated in the liver of infected mice . We demonstrated that IL-33 signaling is crucial for Treg cell accumulation after MCMV infection and ST2-deficient mice show a more pronounced liver pathology and higher mortality compared to infected control mice . These results illustrate the importance of IL-33 in the suppressive function of liver Treg cells during CMV infection .
Regulatory CD4+Foxp3+ T ( Treg ) cells play an essential role in maintaining immune homeostasis and suppressing an overwhelming immune response in several diseased conditions including viral infections and cancer . The transcription factor Foxp3 is essential for Treg cell differentiation and function , thus a mutation in the Foxp3 gene results in an immune-mediated disorder affecting multiple organs in both mice and humans [1] . Beside the naturally occurring Treg cells ( nTreg ) which mature in the thymus , a variety of induced Treg cells ( iTreg ) arise from naive CD4+Foxp3− T cells in the periphery , under influence of tissue microenvironment and cytokines [2] . Treg cells employ various immunoregulatory mechanisms including the inhibition of antigen presenting cell function , a direct killing of effector cells , the consumption of IL-2 and the production of immunosuppressive cytokines such as IL-10 , TGFβ and IL-35 or amphiregulin [3–5] . However , the phenotype of Treg cells and their suppressive mechanisms differ depending on particular tissue and disease settings [3] . For example , certain subsets of Treg cells , specifically those in adipose tissue and intestines , express high amounts of the IL-33 receptor ST2 , and require IL-33 for their maintenance and suppressive function [6] . Tissue alarmin IL-33 has been associated with the differentiation and function of various lymphocytes including Treg cells . In addition to T helper 2 ( Th2 ) cells , Treg cells constitutively express high amounts of ST2 , unlike other CD4+ and CD8+ T cell subsets [7] . Several studies have described the involvement of Treg cells in the immune response to viral infections [8] . For instance , Treg cells can modulate early T-cell trafficking to infected nonlymphoid sites and facilitate protective responses against herpes simplex virus ( HSV ) , lymphocytic choriomeningitis virus ( LCMV ) and respiratory syncytial virus ( RSV ) infection [9 , 10] . On the other hand , Treg cells can reduce the effector T-cell response and inhibit anti-viral cytokine production [8] . Although the suppression of an excessive immune response is beneficial for the host since it limits immunopathology , the suppression of an early response could adversely affect virus control . Thus , some viruses are able to expand activity and number of Treg cells as a mechanism to escape from an effective immune response [11–13] . This has been suggested also for cytomegalovirus ( CMV ) which is well known for developing different immune evasion strategies aimed at avoiding immune cell recognition [14 , 15] . The support for this idea came from several previous studies , which demonstrated that murine cytomegalovirus ( MCMV ) infection induces both nTreg and iTreg cells [16–18] . The depletion of Treg cells leads to an increased T cell response , major players in controlling an early MCMV infection [16 , 17] . In addition , Treg depletion results in reduced viral titers in salivary glands of BALB/c mice [16 , 17] highlighting these cells as a target of immune evasion . However , recent studies link human Treg expansion to a decreased vascular pathology in CMV infected elderly individuals [19] and the depletion of mouse Treg cells in MCMV infected brain augmented chronic gliosis [18] . Thus , it remains unclear whether Foxp3+ Treg cells function in a positive way to limit an exaggerated immune activation and consequent CMV-induced immunopathology . Here we aimed to determine if the host benefits from an early induction of Foxp3+ Treg cells upon MCMV infection . Particularly , we were interested to determine whether these cells can counteract MCMV-induced liver damage . Our data demonstrate that MCMV infection induces both splenic and liver Treg cells . However , the activation and proliferation of Treg cells is much more pronounced in the liver compared to the spleen . In addition , Treg deficiency results in severe liver pathology in infected mice . Similar results were obtained in mice lacking the IL-33 receptor , with an impaired accumulation of Treg cells in the liver following infection . Together , our data unveil the importance of IL-33-dependent Treg cells in preventing MCMV-induced liver damage .
To assess the impact of acute MCMV infection on the Treg cell response we have characterized in detail the kinetics and phenotype of Treg cells in the spleen and liver , two major target organs for MCMV replication . For this , we have infected naive BALB/c mice intravenously ( i . v . ) with WT MCMV and analyzed Treg cell responses in both spleen and liver at day 1 . 5 , 4 , 7 , 14 and 21 post infection ( p . i . ) . We have observed a significant increase in the absolute numbers of these cells in the acute phase of infection which peaked at day 7 p . i . in both organs ( Fig 1A ) . A similar trend has been observed when percentages of Treg cells were assessed , particularly in the liver ( S1A Fig ) . The expansion was followed by a contraction phase in the liver similar to MCMV specific non-inflationary CD8+ and CD4+ T cell responses [20] , whereas splenic Treg cells were maintained at high numbers even 3 weeks after infection . To determine the proliferative capacity of naive Treg cells from either the spleen or the liver , we have measured the expression of proliferation-associated nuclear antigen Ki-67 . A significantly higher percentage of liver Treg cells expressed Ki-67 with a higher median fluorescence intensity ( MFI ) compared to splenic Treg cells ( Fig 1B–1D ) . In line with a higher Ki-67 expression , liver Treg cells exhibited a significantly higher incorporation of BrdU than splenic Treg cells in naive but also MCMV infected mice confirming their enhanced proliferation ( Fig 1E ) . In contrast , splenic Treg cells expressed significantly higher levels of the anti-apoptotic protein Bcl-2 than did liver Treg cells ( Fig 1F ) . Notably , an inverse correlation between expression of Bcl-2 and Ki-67 by liver and splenic Treg cells was observed . The large majority of Ki-67 positive Treg cells were also positive for Helios , i . e . belong to nTreg subset ( S1B Fig ) . We have further characterized splenic and liver Treg cells in MCMV infected mice and assessed the expression of several cell surface molecules . Based on an elevated expression of activation markers on Treg cells , such as CD69 , GITR , CTLA-4 and OX-40 , Treg cells from 7-days infected mice showed an activated phenotype ( Fig 1G ) . Interestingly , the difference in activation between naive and MCMV-induced Treg cells was even more pronounced in the liver than in the spleen of the same mice , with a corresponding very low expression of the homing receptor L-selectin ( CD62L ) . The secretion of immunosuppressive cytokines and cytolytic granzyme B is a well described mechanism by which Treg cells exert their suppressive activity [21] . To characterize their production by Treg cells in MCMV infected mice , we have determined the frequency of Treg cells expressing IL-10 , the latency associated protein ( LAP ) , a part of TGFβ precursor , its cell surface receptor GARP , and granzyme B in spleen and liver . The expression of IL-10 , LAP/TGFβ , GARP and cytolytic granyzme B by Treg cells was induced after MCMV infection . While IL-10 production was detected only in Foxp3 negative CD4+ cells , LAP/TGFβ and GARP were predominantly expressed by CD4+Foxp3+ cells ( S1C Fig ) . Unlike IL-10 and TGFβ , granzyme B was expressed in a proportion of both subsets . Altogether , MCMV infection results in expansion of Treg cells with a more activated phenotype in the liver compared to the spleen . Previous studies have shown the importance of IL-33 in Treg cell induction and tissue-dependent expression of IL-33 receptor ST2 by Treg cells [6] . Using flow-cytometric analysis of splenic and liver Treg cells from BALB/c and ST2-/- mice we have shown an enrichment of ST2-expressing Treg cells in the liver compared to the spleen of BALB/c mice , measured as either the fraction of ST2+ cells or ST2 MFI ( Fig 1H–1J ) . Next , we have determined the kinetics of ST2+ Treg cells after MCMV infection and showed an increase of ST2+ Treg cells upon infection in both organs ( S1D Fig ) . To determine ST2-related effects on Treg cell proliferation , we have measured Ki-67 expression and the incorporation of BrdU by ST2+ and ST2- Treg cells . The proliferation of ST2+ Treg cells was clearly upregulated compared to ST2- Treg cells , with a reduced Bcl-2 expression ( S2A and S2B Fig ) . The majority of ST2+ Treg cells were thymus derived and expressed Helios and Neuropilin-1 ( S2C Fig ) . Next , we further compared the phenotypic characteristics of ST2+ and ST2- Treg cells ( S2D Fig ) from the liver of naive BALB/c mice and this revealed that ST2+ Treg cells , while sharing the expression of classical Treg markers , are distinct from their ST2- counterparts . Namely , ST2+ Treg cells were mostly CD62Llo and exhibited an elevated expression of CD103 , KLRG1 , CTLA-4 , GITR and PD-1 compared to their ST2 negative counterparts , resembling a previously described phenotype of splenic ST2+ Treg cells [22 , 23] . A differential expression of these markers on ST2+ Treg cells was also observed ( maintained ) during MCMV infection ( S2E Fig ) . Our data point to a constitutively higher frequency of Treg cells expressing ST2 in the liver with a distinct phenotype from ST2 negative Treg cells . Treg cell activation and expansion has been proposed as a viral immune evasion strategy to avoid immune cell control [13] . However , Treg cells are also crucial for the maintenance of peripheral tolerance and prevention of autoimmunity; therefore , their role during viral infection is also beneficial for the host . To investigate the possible beneficial role of Treg cells during the acute phase of CMV infection , we have assessed the functional and histological parameters of liver damage following infection . Liver is one of the organs strongly affected with CMV infection where human cytomegalovirus ( HCMV ) can cause clinically relevant damage , leading to chronic hepatitis , cirrhosis and possible death in neonates or immunocompromised patients [24 , 25] . MCMV causes liver damage in mice which is comparable to the damage in HCMV infected individuals including elevated liver enzymes , hepatitis and hepatocellular necrosis [26] . To assess the role of Treg cells in the liver , we have carried out the depletion of Treg cells in BALB/c mice by administration of anti-CD25 antibody 2 days prior to infection . This administration of anti-CD25 resulted in a significant elevation of aspartate aminotransferase ( AST ) and a modest increase in alanine aminotransferase ( ALT ) levels in the serum of treated mice compared to untreated , with no difference in virus titers ( S3A and S3B Fig ) . However , the effects of anti-CD25 treatment can be misleading because despite the depletion of CD25high cells a significant number of Foxp3+ cells remains [27] . In addition , activated effector T cells may transiently express CD25 and are thus potential targets for anti-CD25 antibodies [28] . Therefore , we have used BALB/c DEREG mice [29] , which express the diphtheria toxin receptor under control of the Foxp3 locus , allowing for a selective depletion of Foxp3+ Treg cells by Diphtheria toxin ( DT ) administration . Mice were infected with MCMV and treated with DT on the day of infection and 1 day later , or left untreated , and analyzed 5 days p . i . Littermate controls treated with DT were used to exclude possible toxic and unspecific DT effects . Analysis of liver enzymes revealed significantly higher levels of AST and ALT in the serum of Treg-depleted mice compared to untreated mice , indicating a more severe liver damage in the absence of Treg cells in MCMV infected mice ( Fig 2A ) . The uninfected DEREG mice did not develop any signs of liver pathology resulting from Treg ablation ( S3C Fig ) . In addition , changes in body weight were monitored daily in DT-treated and control mice . Following MCMV infection , all groups of mice displayed a marked reduction in body weight , with highest loss observed in mice depleted of Treg cells ( Fig 2B ) . Furthermore , histopathological analysis revealed a markedly increased severity and extent of overall tissue damage in the absence of Treg cells characterized by confluent areas of hepatocellular necrosis with mononuclear infiltrates and extravasation of erythrocytes ( Fig 2C and 2D ) . To test whether this Treg effect is mediated by TGFβ secretion , we have injected BALB/c mice with TGFβ neutralizing antibodies prior to infection . Neutralization of TGFβ resulted in a 1 . 5-2-fold increase in liver enzymes at day 5 p . i . ( Fig 2G ) and histological analysis revealed a strongly induced liver pathology in treated mice ( Fig 2E and 2F ) . Taken together , these data demonstrate that Treg cells and TGFβ can inhibit the development of severe liver damage during the acute phase of MCMV infection . To assess whether TGFβ and Treg cells have overlapping influence on virus-induced liver pathology we have treated DEREG mice with TGFβ neutralizing antibodies prior to infection and DT injections . As expected , liver enzyme levels and weight loss on day 4 p . i . were similar in both groups of mice , indicating a Treg/TGFβ protecting axis during MCMV infection ( S4A and S4B Fig ) . In the same experiment , we have treated mice with CD8 and CD4 depleting antibodies to show their putative role in liver damage . Mice depleted of either CD4+ or CD8+ T cells had similar levels of liver enzymes and weight loss as Treg-sufficient mice , although this was slightly more pronounced in animals depleted of CD8+ T cells . We have also measured the viral titers and CD8+ T cell response after re-stimulation with immunodominant viral peptides IE1 ( immediate-early 1 ) and m164 [30 , 31] in the liver and spleen of Treg depleted and nondepleted DEREG mice . Depletion of Treg cells resulted in a minor or no difference in viral loads ( Fig 3B ) but higher frequencies of IFNγ and granzyme B producing CD8+ T cells on day 5 p . i . in both organs ( Fig 3A ) . Thus , Treg cells inhibit the MCMV-specific CD8+ T cell response in both the liver and spleen . To further determine the contribution of conventional T and Treg cells to liver damage , we have isolated splenocytes from naive BALB/c mice , pretreated with anti-CD25 antibody , and transferred them alone or together with purified Treg cells , to infected SCID mice . Mice that received only conventional T cells had 2-fold higher ALT and AST serum levels than untreated SCID mice . However , co-transfer of Treg cells rescued liver from MCMV induced damage to the same levels as in untreated SCID mice ( Fig 3C ) . We have also adoptively transferred BALB/c CD8+ T cells alone or together with Treg cells to infected SCID mice . Transfer of only CD8+ T cells has induced a significant release of ALT , which was not the case in mice which have received Treg cells in combination with CD8+ T cells ( S4C Fig ) . Altogether , these data demonstrate a protective role of Treg cells against T cell mediated MCMV-induced liver pathology . It has been well established that some viruses alter the expression of IL-33 during infection [32 , 33] . Our goal was to assess the impact of MCMV infection on IL-33 expression and the consequent regulation of Treg cells . First , we have measured the expression of IL-33 by three different cell lines: B12 ( SV40-transformed BALB/c fibroblasts ) , RAW 264 . 7 ( A-MuLV-transformed BALB/c macrophages ) and BALB/c LSEC ( SV40-transformed liver sinusoidal endothelial cells ) . Cells were infected with virus lacking the m138 gene in order to avoid unspecific interactions of the viral Fc receptor ( encoded by m138 gene ) with the Fc portion of immunoglobulins . The expression of IL-33 was detected under normal conditions in all of the tested cell lines and increased 24 h post MCMV infection with the highest difference between uninfected and infected cells observed in LSECs ( S5 Fig ) . In order to dissect the in vivo impact of MCMV infection on IL-33 expression , we have measured IL-33 mRNA levels at day 1 . 5 , 4 , 7 and 10 post infection in the liver of BALB/c mice . IL-33 mRNA expression peaked at day 4 post infection in the liver ( Fig 4A ) and correlated well with the kinetics of viral replication in this organ and subsequent hepatic inflammation [34 , 35] , confirming its role as an alarmin . In an attempt to characterize the cellular source of IL-33 we performed immunohistological co-staining of IL-33 and MCMV IE1 in liver tissue from uninfected and infected BALB/c mice . In livers of uninfected mice , IL-33 has been demonstrated in a small number of sinusoidal endothelial cells , but not in hepatocytes . In MCMV infected livers , IL-33 was found to be concentrated in a large number of cells that form focal infiltrates by selectively surrounding the foci of infection ( Fig 4B ) . To identify cell type that produces IL-33 in these inflammatory foci of liver tissue , 2-color immunohistochemical ( 2C-IHC ) staining was performed using antibodies to CD31 for endothelial cells ( EC ) , CD3ε for T and NKT cells , or F4/80 ( Ly71 ) for macrophages ( Mø ) . As shown in Fig 5 , only cells expressing F4/80 co-localized with IL-33 ( Fig 5A ) and indeed co-expressed IL-33 ( Fig 5B ) , thus demonstrating that macrophages are the major IL-33 producing cells in the liver of MCMV infected mice . To further test whether IL-33 is important for the Treg cell response during MCMV infection we have used ST2-deficient ( ST2-/- ) mice . WT and ST2-/- mice were infected with MCMV and analyzed on day 7 p . i . The accumulation of Treg cells in the liver of ST2-/- mice was significantly impaired during the peak of the Treg cell response , compared to their WT counterparts ( Fig 6A ) . In line with a reduction in the accumulation of liver Treg cells in ST2-/- mice , we have detected 2-3-fold higher AST and ALT levels in ST2-/- mice compared to WT mice infected with MCMV ( Fig 6B ) . Similar to finding in Treg depleted MCMV infected mice , a number of distinct presentations of liver pathology were observed histologically in liver of ST2-/- mice characterized by confluent areas of hepatocellular necrosis , mononuclear infiltrates and extravasation of erythrocytes ( Fig 6C and 6D ) . In Con A induced hepatitis , IL-33 has been reported to suppress the expression of active caspase-3 in the liver parenchyma leading to cell apoptosis [36] . We have also investigated whether IL-33 affects the level of expression of caspase-3 in the liver of MCMV infected mice . The number of caspase-3 positive cells was remarkably higher in the liver of MCMV infected ST2-/- mice compared to WT mice , indicating a protective role of IL-33/ST2 signaling in MCMV induced apoptosis and liver injury ( Fig 6E ) . Next , we have studied whether the differences in liver damage could influence the survival rate of ST2-/- mice after the infection with a highly virulent salivary gland-derived MCMV ( SGV ) . All of the BALB/c mice survived infection with 2 . 5x104 PFU of the SGV and resisted a dose of 5x104 PFU of the SGV better than ST2-/- mice ( Fig 6F ) . Thus , the absence of IL-33 signaling leads to a reduced accumulation of Treg cells in the liver and consequently a more severe liver pathology in MCMV infected mice . We next assessed the impact of ST2 deficiency on CD8+ T cell responses to MCMV . We analyzed these effector cells in the spleen and liver of WT and ST2-/- mice . In agreement with a previously published study [32] the absence of ST2 affected the antiviral CD8+ T cell response . Specifically , ST2-/- mice showed a reduced frequency of CD8+ T cells directed to the immunodominant MCMV epitopes IE1/m123 and m164 in spleen . However , ST2-/- mice had a comparable frequency of virus-specific CD8+ T cells in the liver as their WT counterparts , suggesting that the accumulation of these cells in the liver is taking place in spite of their lower frequency in lymphoid organs ( Fig 7A ) . Thus , the impaired CD8+ T cell response is likely an intrinsic function of ST2 deficiency rather than effect of this pathway on Treg cells . Notably , the difference in CD8+ T cell response in the spleen did not affect the virus control , as ST2-deficient mice were able to control MCMV with the same efficiency as WT mice ( Fig 7B ) . Similarly to the above , infection of mice with SGV resulted in no difference between WT and ST2-deficient mice ( S6 Fig ) . This is in line with our earlier studies which have shown that even a complete lack of CD8+ T cells did not impair the kinetics of virus clearance during primary infection [37] . To test whether the requirement for ST2 expression in vivo was Treg cell intrinsic we approached the model of mixed bone marrow chimeras generated from congenically marked CD45 . 1+ WT and CD45 . 2+ ST2-/- hematopoietic cells . Chimeras were infected with Δm157 MCMV to avoid dominant NK cell recognition via Ly49H/m157 axis in mice on C57BL/6 genetic background . Analysis of splenic and liver Treg cells showed a reduction in the proportion of ST2-deficient Treg cells in both naive and infected chimeric mice ( Fig 8A ) . However , the ratio between WT and ST2-/- Treg cells was significantly higher in the liver compared to the spleen after the MCMV infection ( Fig 8B ) . The Treg cell intrinsic effect of ST2 signaling was also supported by a lower percentage of Ki-67 expression in ST2-deficient Treg cells compared to WT Treg cells ( Fig 8C ) . Next , we have analyzed whether the treatment of infected mice with recombinant IL-33 could further boost liver Treg cells in MCMV infected mice . BALB/c mice were injected with IL-33 or vehicle ( phosphate-buffered saline; PBS ) on the day of the infection with MCMV and 2 days later . On day 5 p . i . the percentage of liver Treg cells ( Fig 9A ) , their expression of Ki-67 ( Fig 9B ) , ST2 ( Fig 9C ) and Helios ( Fig 9D ) were all significantly increased following treatment with IL-33 compared to PBS-treated mice . Moreover , the percentage of overall CD8+ T cell compartment was significantly decreased in the liver after IL-33 treatment ( Fig 9E ) . No difference was observed in the frequency of CD8+ T cells in spleen . These data demonstrate a strong impact of IL-33 on the Treg cell population in the liver upon MCMV infection .
Treg cells control immune responses under both physiological and pathological conditions . Although the role of Treg cells in homeostasis of immune response and the prevention of immunopathology is well established in different disease models , less is known about their role in the prevention of immunopathology during various viral infections . Many viruses , particularly herpes viruses , encode numerous genes aimed at suppressing the immune response; by doing so they may also prevent immunopathology [38] . CMV is widely spread among mammalian hosts and is usually well controlled by the immune system . However , CMVs establish lifelong persistent ( latent ) infection from which reactivation can occur whenever the immune response is compromised . Here we show that Treg cells are essential in limiting liver damage during the immune response to primary CMV infection . Liver Treg cells expand 2-3-fold and upregulate their activation markers upon MCMV infection . Ablation of Treg cells in MCMV infected mice led to a significant increase in liver pathology and consequent body weight loss . Liver pathology in Treg depleted mice correlated with an enhanced CD8+ T cell response but not with virus load . Similar to previous studies , no significant difference of virus titers was observed between Treg depleted and non-depleted mice or mice in which Treg cells were selectively expanded to avoid graft-versus-host disease [16 , 39] . We have shown that CD8+ T cells readily accumulate in liver of CMV infected mice and are heavily activated , and that a significant proportion of them express granzyme B and IFNγ . It has been well established that the T cell response to viral infections contributes not only to limit viral replication but can also cause immunopathology if the response is not properly regulated [40] . Primary infection with non-cytolytic viruses such as hepatitis C virus ( HCV ) and hepatitis B virus ( HBV ) , results in T cell mediated immunopathology in the liver [8 , 41] . Being a cytolytic virus , CMV is not a prototype of virus that induces immunopathology in the immunocompetent host . Yet , recently evidence has been gathered pointing out that part of liver pathology after MCMV infection is a consequence of the immune response rather than the virus infection itself [42 , 43] . The absence of liver pathology in MCMV infected SCID mice suggests the role of T cell immunity in liver damage . Here we confirm these observations and provide evidence that Treg cells dampen an exaggerated immune response and are crucial for the prevention of CMV-induced immunopathology in the liver . This is further supported by our observation of a strong proliferative capacity and activation status of liver Treg cells . Viruses use different mechanisms to induce Treg cells . Among these , the most well known mechanisms are the ones induced via the ligation of various pathogen associated molecular pattern ( PAMP ) receptors which are expressed on Treg cells [44] . There are many other factors involved in triggering and expansion of Treg cells including various cytokines such as IL-2 , TNFα , TGFβ and galectin molecules [8] . Since CMV infection upregulates most of the factors that can induce Treg cells , it is not unexpected that these cells are strongly expanded after infection . CMV is an ubiquitous virus which can infect almost all cell types and essentially all tissues including mucosa and glands . One can predict that Treg cells play an important role in homeostasis of the local immune response and prevention of immunopathology as has been demonstrated here for the liver . In addition to induction by pathogens , Treg cells can also be triggered by factors released by tissue damage such as IL-33 . Several studies have demonstrated that administration of IL-33 both in vivo and in vitro can induce Treg cell expansion and have highlighted the importance of IL-33 signaling in Treg cells during inflammatory conditions such as obesity-induced insulin resistance , chronic colitis , acutely injured muscles or graft-versus-host disease [22 , 23 , 45–48] . Here we show that liver Treg cells have a constitutively higher expression of IL-33 receptor ST2 compared to splenic Treg cells . Moreover , ST2+ Treg cells exhibited an increased expression of Ki-67 and incorporation of BrdU compared to ST2- Treg subset , suggesting their high proliferative capacity . Given that IL-33 functions as an endogenous danger signal or alarmin in response to tissue damage [49] , the upregulation of IL-33 expression after MCMV infection indicates its importance in limiting virus-induced immunopathological liver damage . A study by Nabekura et al . reported the upregulation of IL-33 mRNA in the splenic fibroblastic reticular ( FRC ) and lymphatic endothelial cells ( LEC ) but not blood endothelial cells ( BEC ) after MCMV infection [50] . Our results demonstrate upregulation of IL-33 mRNA in the liver tissue upon MCMV infection . Moreover , here we demonstrated that the vast majority of IL-33 producing cells in the liver of MCMV infected mice are contained within the inflammatory foci and represent F4/80+ macrophages . In line with this , the infection of ST2-deficient mice resulted in a reduced accumulation of Treg cells in the liver , and the liver damage was similar to the one after Treg cell depletion . Therefore , our results describe a previously unknown link between IL-33 expression and Treg cell function in the context of a viral infection . In addition to their role in the modulation of conventional immune responses , Treg cells are important in nonimmunologic disease as tissue protective modulators of the inflammatory response after tissue damage . For instance , endogenous Treg cells strongly impact the outcome of ischemic stroke [51] . Their absence dramatically accelerates postischemic activation of inflammatory cells , and their secretion of TNFα and IFNγ is key pathogenetic factor in disease progression . Treg cells may act to dampen ischemic injury in other organs including the liver [52] . So far , there is no strong evidence that Treg depletion can influence the outcome of ischemic liver injury [53] . However , an increase in the proportion of intrahepatic Treg cells was observed in rats after treatment with some immunosuppressive agents [54] . CMV infection is frequently associated with ischemic injury most likely as a consequence of a cytokine storm [35] . This is particularly obvious in case of mice infected with a highly virulent MCMV isolate derived from salivary glands . In this and previous studies we have noticed that many lesions in liver of CMV infected mice are located in the proximity of the central vein , resembling the lesions caused by ischemic injury . Thus , it is likely that the Treg response in liver of CMV-infected mice is not only induced by infection itself but also by tissue lesions that are indirect to infection . In summary , our study demonstrates that Treg cells inhibit a severe MCMV-induced immunopathological hepatitis . This function of Treg cells is largely dependent on IL-33-induced accumulation of Treg cells . Overall , The IL-33/Treg axis seems to be a promising route for the development of future therapies in viral infections .
BALB/c , DEREG BALB/c [29] , ST2-/- BALB/c [55] , ST2-/- C57BL/6 [55] kindly provided by Daniel Pinschewer , CBySmn . CB17-Prkdcscid/J ( SCID ) , C57BL/6 CD45 . 1+ and C57BL/6 CD45 . 1+CD45 . 2+ mice were housed and bred under specific-pathogen-free conditions at the Central Animal Facility of the Medical Faculty , University of Rijeka . For depletion of Foxp3+ Treg cells in DEREG mice , 25 ng/g body weight of DT ( Merck ) were injected i . p . on day of the MCMV infection and the following one . Gender- and aged- matched littermates were used as controls . To deplete Treg cells with antibodies , BALB/c mice were injected i . p . with 150 μg of anti-CD25 antibody ( PC-61 . 5 . 3; BioXCell ) two days before infection . TGFβ blockade was performed by i . p . injection of 500 μg of anti-TGFβ antibody ( 1D11 . 16 . 8; BioXCell ) on the day of infection . Depletion of CD4+ and CD8+ T cells was performed by i . p . injection of 150 μg of anti-CD8 antibody ( YTS 169 . 4 ) and anti-CD4 antibody ( YTS 191 . 1 ) , respectively , on the day of infection . IL-33 treatments were performed by i . p . injection of recombinant mouse IL-33 ( Biolegend ) . Mice received 2 μg IL-33 at day 0 and 2 after infection . Mice were inoculated i . v . with 106 plaque-forming units ( PFU ) of tissue culture ( TC ) -grown virus , except for flow cytometric and 2C-IHC analysis where the inoculum was reduced to 2x105 PFU and 5x105 PFU , respectively . MCMV strain MW97 . 01 [56] and pSM3fr-MCK-2fl clone 3 . 3 [57] are referred to as WT MCMV . In addition to tissue culture grown virus , for some experiments we have used salivary gland derived virus ( SGV ) , and mutant viruses lacking m157 [58] or m138 gene [59] . The viral titers in organs were quantified by the standard plaque assay , as described previously [60] . Eight- to 12-week-old mice were used in all experiments . All animal experiments described in this paper were performed in accordance with the guidelines contained in the International Guiding Principles for Biomedical Research Involving Animals and approved by the Animal Welfare Committee at the University of Rijeka . Single-cell suspensions of spleen and liver were prepared according to standard protocols . Flow cytometric analysis were performed by using anti-mouse CD4 ( GK1 . 5 ) , Foxp3 ( FJK-16a ) , CD25 ( PC61 . 5 ) , ST2 ( RMST2-2 ) , Ki-67 ( SolA15 ) , Bcl-2 ( 10C4 ) , CD69 ( H1 . 2F3 ) , GITR ( DTA-1 ) , CTLA-4 ( UC10-4F10-11 ) , OX40 ( OX-86 ) , CD62L ( MEL-14 ) , CD103 ( 2E7 ) , LAP ( TW7-16B4 ) , GARP ( YGIC86 ) , IL-10 ( JES5-16E3 ) , granzyme B ( NGZB ) , CD8 ( 53–6 . 7 ) , IFNγ ( XMG1 . 2 ) , KLRG1 ( 2F1 ) , PD-1 ( J43 ) , Helios ( 22F6 ) , Neuropilin-1 ( 3DS304M ) , IL-33 ( 396118 ) , CD45 . 1 ( A20 ) and CD45 . 2 ( 104 ) purchased from eBioscience or R&D Systems . Intracellular staining for Foxp3 , IL-10 , IFNγ , IL-33 , Ki-67 , Bcl-2 and granzyme B was performed with the Foxp3 staining kit from eBioscience according to the manufacturer’s recommendations . For the cell proliferation assay , mice were provided for 6 days with 0 . 8 mg/ml BrdU in the drinking water starting on the day of infection ( BrdU; Sigma ) To detect incorporated BrdU , cells were stained according to the manufacturer’s protocol ( BrdU flow kit; BD Pharmingen ) . Fixable Viability Dye from eBioscience was used to stain dead cells . For intracellular cytokine staining , cells were re-stimulated with plate-bound anti-CD3 ( 5 μg/ml per well; 145-2C11; eBioscience ) and anti-CD28 ( 2 μg/ml per well; 37 . 51; eBioscience ) for 4h in the presence of Brefeldin A ( 10 μg/ml; eBioscience ) . For IFNγ staining , cells were re-stimulated with 1 μg of peptides IE1/m123 ( 168–176 YPHFMPTNL176 ) or m164 ( 257–265 AGPPRYSRI265 ) for 4h in the presence of Brefeldin A ( 10 μg/ml; eBioscience ) . H-2L ( d ) /IE-1 ( 168–176 YPHFMPTNL ) and H-2D ( d ) /m164 ( 257–265 AGPPRYSRI ) tetramers were provided by NIH tetramer core facility . All data were acquired using FACSAria or FACSVerse ( BD Biosciences ) and analyzed using FlowJo software ( Tree Star ) . Single-cell suspensions of spleen used in adoptive transfers were prepared according to standard protocols . Splenic Treg cells were isolated using CD4+CD25+ Regulatory T Cell Isolation Kit ( Miltenyi Biotec ) . Splenic CD8+ T cells were isolated using CD8a+ T cell isolation kit ( Miltenyi Biotec ) . A total of 1x107 conventional T cells or 2x106 CD8+ T cells and 1x106 Treg cells in a total volume of 500μl of DMEM were injected into tail veins of SCID mice . Mice were injected with MCMV 8 hours before the adoptive transfer . Formalin-fixed , paraffin-embedded liver sections were used for hematoxylin and eosin staining and Caspase-3 ( Asp175; Cell Signaling Technology ) immunohistochemical staining . For the identification of the IL33-expressing mononuclear infiltrate cell type that surrounds infected hepatocytes , thus forming foci , 2-color immunohistochemical ( 2C-IHC ) analyses were performed on paraffin-embedded liver tissue specimens . Consecutive serial 1-μm sections were prepared for combining intranuclear IE1-specific IHC labeling of infected cells [61 , 62] with the detection of cell type-specific markers , which were alternatively CD31 for endothelial cells ( EC ) , CD3ε for T and NKT cells , and F4/80 ( Ly71 ) for macrophages ( Mø ) [63] , as well as with the detection of IL33 . For demasking antigens , the method of heat-induced epitope retrieval ( HIER ) [61] was employed for CD31 and CD3ε IHCs with EDTA buffer ( 10 mM; pH 8 . 0 ) or for the IL33 IHC with tri-sodium-citrate-dihydrate buffer ( 10 mM; pH 6 . 0 ) . F4/80 IHC does not require HIER . In the first step , specific labeling was performed alternatively with antibodies directed against CD31 ( dianova , clone SZ31 ) , CD3ε ( BioRad , clone CD3-12 ) , F4/80 ( eBioscience , clone BM8 ) , or IL33 ( Bioss Antibodies , polyclonal rabbit antiserum ) . For CD31 , CD3ε , and F4/80 IHCs , biotin-conjugated polyclonal anti-rat Ig antibody ( BD Biosciences ) was used as the secondary antibody , and black staining was achieved with peroxidase-coupled avidin biotin complex ( Vectastain Elite ABC Kit ) using DAB as the substrate and ammonium nickel sulfate hexahydrate for color enhancement [61] . In the case of IL33 IHC , the secondary antibody was biotin-conjugated polyclonal goat aniserum anti-rabbit-IgG ( Sigma-Aldrich ) , and turquoise-green staining was achieved with peroxidase-coupled avidin biotin complex ( Vectastain Elite ABC Kit ) as described above , except that substrate and chromogen were provided by the HRP-Green Solution Set ( 42 life sciences ) . Alternatively , black staining was achieved with DAB as the substrate followed by color-enhancement as described above . Finally , viral IE1 protein in the nuclei of infected cells , ( of infected hepatocytes in the specific case ) has been labeled with monoclonal antibody CROMA 101 [35] , and red staining was achieved with alkaline phosphatase-conjugated polyclonal goat anti-mouse IgG ( BioRad ) and the Fuchsin+ substrate-chromogen system ( Dako ) . For 2C-IHC of macrophages coexpressing F4/80 and IL33 , HIER was performed for demasking IL33 epitopes , followed by turquoise-green staining as described above . After this , the macrophage marker F4/80 was stained in red after specific labeling with monoclonal rat antibody anti-F4/80 ( eBioscience , clone BM8 ) , followed by alkaline phosphatase-conjugated anti-rat Ig antibodies ( BioRad , polyclonal goat serum ) and the Fuchsin+ substrate-chromogen system . Slides were analyzed on a Zeiss Axiophot 1 or Olympus BX51 microscope , and digital images were acquired by the VisiCAM-100 Imaging System ( Visitron ) using a CCD camera ( Basler ) or Olympus camera ( DP71 ) . Scores of cumulative liver pathology for apoptosis , intranuclear inclusion bodies ( INIBs ) , inflammation , and necrosis were determined using the following scoring system: 0 , normal ( no pathology ) ; 1 , mild ( 1–3 abnormal areas ) ; 2 , moderate ( 3 to 5 abnormal areas ) ; 3 , severe ( >5 abnormal areas ) [42 , 64] . Histological samples were blinded prior to evaluation . The presence of aspartate aminotransferase ( AST ) and alanine aminotransferase ( ALT ) in previously frozen serum samples was determined by standard enzymological methods in the Clinical Institute of Laboratory Diagnostics , Rijeka Clinical Hospital Center or LABOKLIN GmbH&Co . KG , Linz . B12 and RAW 264 . 7 cells were grown in DMEM supplemented with 10% FCS . BALB/c LSECs were grown in RPMI supplemented with 10% FCS . Cell were infected with 3 PFU/cell of Δm138 MCMV , harvested 24 hours later and stained for intracellular IL-33 expression . Total RNA was isolated from liver tissue with High Pure RNA Tissue Kit ( Roche ) and reversely transcribed to cDNA with High-Capacity cDNA Reverse Transcription Kit ( Applied Biosystems ) . Real-time PCR was performed using TaqMan assay ( Mm00505403_m1; Applied Biosystems ) . Values were normalized to mouse GAPDH ( Mm99999915_g1 ) . Mixed bone marrow chimeras were generated by lethal irradiation ( 9 . 5 Gy ) of C57BL/6 CD45 . 1+CD45 . 2+ mice followed by i . v . injection of 5x106 wild-type ( WT; CD45 . 1+ ) and 5x106 ST2-/- ( CD45 . 2+ ) bone marrow cells . Mice were allowed 8 weeks to reconstitute . After reconstitution , mixed chimeras were i . v . injected with Δm157 MCMV . The analysis was performed 7 days later . Unless otherwise noted , data are presented as mean ± SEM . Statistical significance was determined by either two-tailed unpaired Student’s t test or one-way ANOVA with Bonferroni correction . Differences in viral titers between experimental groups were determined by the unpaired two-tailed Mann-Whitney u test using GraphPad Prism 5 . A value of p>0 . 05 was deemed not statistically significant ( ns ) ; *p<0 . 05 , **p<0 . 01 and ***p<0 . 001 . The study has been approved by the Animal Welfare Committee at the University of Rijeka . | Treg cells are crucial for immune homeostasis and for dampening immune response to several diseased conditions , including viral infections . Murine cytomegalovirus ( MCMV ) is a herpesvirus with pathogenic potential , so that early immune mechanisms are essential in controlling virus and protecting from virus-induced pathology . Studies on Foxp3+ Treg cells have revealed their inhibitory role on the early T cell response to MCMV infection and have suggested Treg cells as a target of MCMV’s immunoevasion mechanisms . Here we demonstrate that the number and activation status of liver Treg cells is strongly induced upon MCMV infection in order to protect the host from severe liver damage . They constitutively express high amounts of IL-33 receptor ST2 and their accumulation depends on IL-33 , which is released as a tissue alarmin after the cell damage . For the first time , we show an immunoregulatory role of IL-33-dependent Treg cells in the liver of MCMV infected mice and their suppression of MCMV-induced immunopathology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"liver",
"immune",
"cells",
"immune",
"physiology",
"spleen",
"immunology",
"cytomegalovirus",
"infection",
"cloning",
"cytotoxic",
"t",
"cells",
"molecular",
"biology",
"techniques",
"research",
"and",
"analy... | 2017 | IL-33/ST2 pathway drives regulatory T cell dependent suppression of liver damage upon cytomegalovirus infection |
Selective brain responses to objects arise within a few hundreds of milliseconds of neural processing , suggesting that visual object recognition is mediated by rapid feed-forward activations . Yet disruption of neural responses in early visual cortex beyond feed-forward processing stages affects object recognition performance . Here , we unite these discrepant findings by reporting that object recognition involves enhanced feedback activity ( recurrent processing within early visual cortex ) when target objects are embedded in natural scenes that are characterized by high complexity . Human participants performed an animal target detection task on natural scenes with low , medium or high complexity as determined by a computational model of low-level contrast statistics . Three converging lines of evidence indicate that feedback was selectively enhanced for high complexity scenes . First , functional magnetic resonance imaging ( fMRI ) activity in early visual cortex ( V1 ) was enhanced for target objects in scenes with high , but not low or medium complexity . Second , event-related potentials ( ERPs ) evoked by target objects were selectively enhanced at feedback stages of visual processing ( from ~220 ms onwards ) for high complexity scenes only . Third , behavioral performance for high complexity scenes deteriorated when participants were pressed for time and thus less able to incorporate the feedback activity . Modeling of the reaction time distributions using drift diffusion revealed that object information accumulated more slowly for high complexity scenes , with evidence accumulation being coupled to trial-to-trial variation in the EEG feedback response . Together , these results suggest that while feed-forward activity may suffice to recognize isolated objects , the brain employs recurrent processing more adaptively in naturalistic settings , using minimal feedback for simple scenes and increasing feedback for complex scenes .
Object recognition is often regarded as a task that is solved in the first wave of visual processing [1] . The human brain indeed recognizes objects at astonishing speed , with single neurons exhibiting object-selectivity from 100 ms after stimulus onset [2] , and global brain signals diverging within 100–200 ms [3 , 4] . Furthermore , hierarchical feed-forward models can emulate human object recognition performance [5 , 6] , and neural representations in human and non-human primate brains match those in feed-forward neural networks [7–9] . However , the visual system is not a strict feed-forward hierarchy: it also contains an abundance of feedback connections [10–13] . Visual response modulations that occur within visual cortex after the initial feed-forward sweep has passed ( i . e . beyond ~150 ms after stimulus onset ) are thought to reflect recurrent interactions ( ‘feedback’ ) between e . g . V1-IT that aid segmentation of figures from backgrounds [14–19] and perceptual grouping [20] . One way in which feedback is thought to facilitate object recognition is through visual routines such as curve tracing and texture segmentation , to integrate line segments and other low-level features encoded in early visual areas [21–23] . Supporting this model of visual processing , transcranial stimulation evidence shows that detection of target objects in natural scenes deteriorates when neural activity in early visual cortex is disrupted not just at feed-forward processing stages ( e . g . , 100 ms after stimulus onset ) , but also at feedback stages ( e . g . , 220 ms after stimulus onset [24 , 25] ) . How can we reconcile the speed of object recognition with an important role for feedback ? Two lines of evidence suggest that feedback may be employed adaptively depending on the complexity of the visual input . First , computer simulations indicate that disrupting feedback activity has stronger effects for occluded or degraded target objects [26] . Second , backward masking , which interrupts recurrent processing [27–29] , has weaker effects for scenes with target objects that are “easily segregated” compared to “more demanding backgrounds” , as assessed through behavioral ratings of independent observers [30] . But how does the visual system determine the complexity of a visual input scene ? Computer vision and scene perception research shows that computational summary statistics of low-level image features are diagnostic of scene complexity [31–33] . For example , contrast distributions can be summarized by two statistics that reflect a scene’s contrast energy ( CE , average contrast ) and spatial coherence ( SC , variability in contrast across the scene ) [34 , 35] . Computing these statistics for a large set of scenes results in a two-dimensional space in which sparse scenes with just a few scene elements separate from complex scenes with a lot of clutter and a high degree of fragmentation ( Fig 1A ) . Since CE and SC appear to provide information about the ‘segmentability’ of a scene , we hypothesized that the visual system computes these scene statistics as a measure of overall scene complexity , with the goal to determine a need for enhanced visual processing mediated by feedback . Importantly , for this hypothesis to be biologically realistic , CE and SC need to be a ) plausibly computable in the visual system and b ) available early in visual processing . The contrast distribution of a scene can in theory be deduced from the population response of contrast-sensitive neurons , e . g . neurons in early visual areas such as LGN and V1 , which respond to local scene elements ( edges ) . Since these responses constitute the first stage of the feedforward processing cascade , this information can be made available very early on in visual processing through a ‘read-out’ of the early population response across the visual scene within the feed-forward sweep . Consistent with this idea , CE and SC values computed from simulated early visual contrast responses [35 , 36] have been shown to modulate the magnitude of single-trial evoked responses to natural scenes as early as 100 ms after stimulus onset [35 , 37–42] . Here , we tested whether scene complexity predicts the degree of feedback activity by measuring brain responses while participants performed a target object detection task in scenes that were systematically sampled to contain either low , medium or high CE and SC values ( see Fig 1B and 1C ) . First , we measured whole-brain fMRI responses to target objects in scenes with low , medium or high complexity ( Experiment 1 ) . Next , we measured EEG responses to the same stimuli to examine the time-course of visually evoked activity ( Experiment 2 ) . Importantly , complexity was matched for target and non-target scenes within each complexity condition , allowing us to disambiguate any feedforward response differences due to scene complexity from subsequent feedback modulations by examining differential responses to target and non-targets . In addition , scene complexity was varied on a trial-by-trial basis , such that participants could not form an expectation of scene complexity beforehand , allowing us to measure responses with unbiased feed-forward processing ( i . e . without a difference in top-down task set or attentional state ) . Together , these experiments show that successful detection of target objects in high , but not low or medium complexity scenes is associated with enhanced activity in early visual areas ( in fMRI ) which emerges at feedback time-points in visual processing ( in EEG ) . Moreover , behavioral performance for high complexity scenes was associated with decreased accuracy and slower response times , reflecting a slower rate of evidence accumulation as formalized by the drift rate parameter in the drift-diffusion model [43–45] . In addition , trial-by-trial variations in drift rate within the high complexity condition were coupled to EEG feedback responses on those trials . Together , these results demonstrate a contribution of feedback to object detection in complex natural scenes .
This interpretation is consistent with the global-to-local processing frameworks which suggest that detailed scene analysis takes place via reentrant processing [23 , 50–52] . These frameworks propose that the feed-forward sweep provides visual cortex with a base representation . If the visual task can be solved based on this representation alone , no further operations are necessary and action selection can be initiated . If it is not sufficiently informative , elemental operations are applied such as contour grouping and texture segmentation , which integrate individual line segments and other features encoded in low-level areas via incremental grouping [21 , 22] . We believe that the differential neural response observed in our fMRI and EEG data for targets relative to non-target scenes reflects the implementation of these elemental operations . A psychological correlate of this feedback-driven process could be termed ‘attentive processing’ while the feed-forward stage could be considered pre-attentive [53]; see [21] for a detailed discussion . Our observation of an increased fMRI response to targets is clearly consistent with numerous observations that attentional modulations are reflected in response enhancements in visual cortex [54] . Importantly , however , our results suggest that this response enhancement is selectively applied on a trial-by-trial basis to scenes which have high complexity , as determined by scene statistics . Our behavioral observations support the presence of selectively increased feedback for complex scenes in two ways . First , in Experiment 1 we observed that scene complexity influenced response selection , but not response suppression [55–58] . Specifically , we observed that the decision time was prolonged for complex scenes , suggesting a slower rate of information accumulation . In the Experiment 2 we examined this hypothesis through the manipulation of time restrictions . Behaviorally , we showed that while RTs were always increased for highly complex scenes , accuracy only declined when the instructions emphasized a speeded response . The drift diffusion model [43 , 44 , 59] was then used to show that drift rate ( the latent variable capturing the rate of information accumulation ) was indeed slower for highly complex scenes [60–62] . In addition , we observed that drift rate was related to single-trial feedback-driven ERP responses for high , but not low or medium complexity scenes . This reliance on feedback provides an explanation for the difference in behavioral performance when participants emphasized speed above accuracy ( and thus were not able to incorporate the information provided by the feedback response ) . Moreover , this ERP-drift rate relation is consistent with reports of a ‘discriminating component’ around 300 ms reflecting the amount of sensory evidence for perceptual decisions on objects in phase noise [61 , 63] . Taken together , these observations underline the importance of feedback activity for processing complex scenes to optimize behavioral performance . We propose that the complexity of the visual input can be estimated using image statistics derived from contrast distributions , in particular contrast energy and spatial coherence . These statistics are potentially suitable computational substrates for such a representation because they can be computed directly from local contrast responses in e . g . LGN [35] . These parameters strongly affect the amplitude of evoked EEG activity early in time in visual processing [35 , 37 , 64] suggesting they are indeed available at early stages of visual computation and could therefore serve as ‘markers’ to determine whether further visual operations are necessary . This idea is reminiscent of previous proposals suggesting that scenes are summarized as coarse ‘blobs’ that precede detailed analysis at smaller spatial scales depending on their diagnostic value [65 , 66] , or via low spatial frequencies that are used to direct top-down facilitation [67] . It is also consistent with results showing that feedback is necessary to model categorization of degraded scenes [26] and with results reported by [30] , who found that masking effects were stronger for scenes that were ‘less easily segmented’ . Here , rather than filtering , degrading or otherwise manipulating the scenes , we used scene statistics to sample variation in scene complexity , providing a quantitative computational approach to estimate this property . We sampled scenes with low , intermediate or high CE and SC values , assuming a linear relation between these values and scene complexity . We note , however , that several of our results , in particular the fMRI responses in higher-order regions and behavioral accuracy in Experiment 1 , appeared to show a U-shaped , rather than linear modulation across conditions , suggesting that CE and SC are not a straightforward parametric index of scene complexity . In particular , while the high condition consistently deviated from the medium condition in terms of both behavioral ( higher RT , lower accuracy ) and neural ( increased differential V1 fMRI and late ERP responses ) effects , the low condition sometimes also showed a decrease in performance , but without a clear neural correlate . Given the clear neural effects for the high condition , the main focus of this paper is on that condition . However , a separate investigation of the U-shaped effects based on an additional set of behavioral studies [68] shows that behavioral performance is indeed often optimal for scenes with intermediate , rather than low , scene complexity . One potential explanation for this benefit is that feed-forward processing for intermediate CE/SC scenes is faster or more efficient due to , for example , increased familiarity with such scenes . However , intermediate scenes could also contain more or better contextual cues regarding the presence of target objects , resulting in increased interaction between scene and object processing pathways [69–71] . In sum , our results do not provide a clear explanation for this pattern , and future research is needed to determine the neural underpinnings of the improved performance for intermediate relative to low complexity images as defined by the CE/SC parameterization . A question that remains unresolved in this study is which , if any , brain areas might be involved in computing scene complexity , i . e . which region ‘reads out’ the scene statistics from the population response in early visual regions . Consistent with previous studies [72 , 73] , we found a strong distinction between animal- and non-animal images in object-selective LOC and face-selective FFA . However , this differential activity was not strongly modulated by whether the object was embedded in a complex scene or a simple scene . We did observe , however , an effect of scene complexity in PPA . While the scene-selectivity of PPA is commonly attributed to coding of 3D spatial layout [74–76] it is also sensitive to object information [77–79] , as well as low-level features such as spatial frequency , contrast , rectilinearity and texture [79–84] , suggesting that PPA may be a suitable region to demonstrate an influence of the broader scene context on object-related activity [69] . Importantly , PPA is biased towards the visual periphery [85 , 86] , containing relatively large receptive fields [87 , 88] making it suited for computation of larger-scale summary statistics of the input [89] . Consistently , a recent study found that PPA was sensitive to difference in scene complexity as defined by various image-computable computational measures such as image compression and self-similarity [90] . Based on this prior literature , we speculate that the enhanced recurrent processing in early visual regions is initiated based on a feed-forward , summary statistics based computation of scene complexity in PPA . However , future research using for example time-resolved measurements of PPA activity will be necessary to confirm or deny the presence of such a representation in PPA during object detection in natural scenes . Perceiving real-world scenes involves more than detection of objects: for example , we can recognize a scene as a specific place , or determine its navigability [91] . It is unclear whether feedback is ‘intrinsically’ enhanced upon perceiving a scene of high complexity regardless of observer goal , or whether it is selectively enhanced when a participant is actively searching for a target object in a complex scene . One way to test this is to compare the current results to a situation in which participants see scenes of varying levels of complexity while performing a task without any object detection requirement ( e . g . an orthogonal visual task at fixation , or detecting a concurrent auditory signal ) . Such top-down task manipulations have been shown to affect ERP responses to natural scenes in recurrent processing time windows [64] , and feedback is thought to be involved in the application of attentional templates [92 , 93] , which may differ between different tasks [94] or the level of detail necessary to solve the task [66] and therefore in the amount of feedback required . While we varied scene complexity on a trial-by-trial basis , making it difficult for participants to use a top-down strategy to ‘predict’ how much attention they would need to direct to solve that trial , they still needed to apply a search target in order to solve the task . This search target was the same throughout the entire experiment ( animal ) , but our whole-brain fMRI analysis ( Fig 3 ) did indicate some target-related activity differences for high complexity scenes in posterior parietal regions , which have been associated both with representing attentional templates as well as outcomes of attentional selection [92] . A deeper understanding of the interaction between scene complexity and top-down task requirements and their respective representation in cortical regions requires future experimental study . So far , however , our results suggest that although object recognition based on feed-forward information [5] may be possible for simple stimuli , detection of objects in complex real-world scenes additionally involves feedback processing . | How much neural processing is required to detect objects of interest in natural scenes ? The speed and efficiency of object recognition suggests that fast feed-forward buildup of perceptual activity is sufficient . However , there is also evidence that disruption of visual processing beyond feed-forward stages leads to decreased object detection performance . Our study resolves this discrepancy by identifying scene complexity as a driver of recurrent activity . We show that recurrent activity is enhanced for complex , cluttered scenes compared to simple , well-organized scenes . Moreover , recurrent activity reflects the amount of accumulated evidence for target object presence . These findings elucidate the neural processes underlying perceptual decision-making by demonstrating that the brain dynamically directs neural resources based on the complexity of real-world visual inputs . | [
"Abstract",
"Introduction",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"functional",
"magnetic",
"resonance",
"imaging",
"reaction",
"time",
"visual",
"object",
"recognition",
"brain",
"electrophysiology",
"social",
"sciences",
"electrophysiology",
"brain",
"neuroscience",
"... | 2018 | Scene complexity modulates degree of feedback activity during object detection in natural scenes |
Genomics has posed the challenge of determination of protein function from sequence and/or 3-D structure . Functional assignment from sequence relationships can be misleading , and structural similarity does not necessarily imply functional similarity . Proteins in the DJ-1 family , many of which are of unknown function , are examples of proteins with both sequence and fold similarity that span multiple functional classes . THEMATICS ( theoretical microscopic titration curves ) , an electrostatics-based computational approach to functional site prediction , is used to sort proteins in the DJ-1 family into different functional classes . Active site residues are predicted for the eight distinct DJ-1 proteins with available 3-D structures . Placement of the predicted residues onto a structural alignment for six of these proteins reveals three distinct types of active sites . Each type overlaps only partially with the others , with only one residue in common across all six sets of predicted residues . Human DJ-1 and YajL from Escherichia coli have very similar predicted active sites and belong to the same probable functional group . Protease I , a known cysteine protease from Pyrococcus horikoshii , and PfpI/YhbO from E . coli , a hypothetical protein of unknown function , belong to a separate class . THEMATICS predicts a set of residues that is typical of a cysteine protease for Protease I; the prediction for PfpI/YhbO bears some similarity . YDR533Cp from Saccharomyces cerevisiae , of unknown function , and the known chaperone Hsp31 from E . coli constitute a third group with nearly identical predicted active sites . While the first four proteins have predicted active sites at dimer interfaces , YDR533Cp and Hsp31 both have predicted sites contained within each subunit . Although YDR533Cp and Hsp31 form different dimers with different orientations between the subunits , the predicted active sites are superimposable within the monomer structures . Thus , the three predicted functional classes form four different types of quaternary structures . The computational prediction of the functional sites for protein structures of unknown function provides valuable clues for functional classification .
Structural biology in the post-genome era faces the challenge of determination of function from 3-D structure , the critical next step toward the realization of the promises of genomics . On the order of 103 protein structures in the Protein Data Bank ( PDB ) are annotated as “hypothetical” or of “unknown function , ” and this number is increasing dramatically as structural genomics initiatives deposit large numbers of structures in the PDB . Functional annotation is usually dependent on sequence similarity to identify proteins that are expected to be similar in structure and therefore may be similar in function . Even when sequence comparison fails to find a closely related protein , the overall structural fold still may be similar to one that is already known . Such structural relationships , however , still do not necessarily identify a functional relationship . The reason for the discrepancy is that currently there is not adequate understanding of the relationship between macromolecular structure and function for most proteins . Thus , structural similarity in many cases has proved to be a poor guide to function . Many proteins with similar and recognizable folds have completely different functions , even sometimes when there is sufficient sequence similarity to consider them “homologous . ” The best examples of this principle are the enzymes having the TIM ( triosephosphate isomerase ) barrel fold . The types of reactions catalyzed by proteins having this fold are numerous and varied . Conversely , two proteins may have completely different folds , but catalyze the same reaction , with the same residues and configurations in the active site . A good example is the set of pyridoxal phosphate–dependent transaminases of fold types I and II . These proteins catalyze the same reaction , with active sites that are practically identical , but the two folds are completely different . In addition , the important residues in an enzyme active site may not be obvious . Many reactions in biology may be characterized by the steps required to bring about any chemical transformation . The catalytic entities involved in each step , such as acids or bases , can be inferred from the known chemistry . Residues that can play these roles are well-defined; however , it is not so easy to determine which particular residues in a given protein are actually playing these roles . Ideally , a structure with substrate bound would resolve the question , but such structures are rarely available for proteins of unknown function . Therefore , another method is needed to identify residues involved in catalysis and molecular recognition . In this paper we demonstrate how a computational predictive tool can aid in the identification of the functionally important residues in proteins of unknown function . We have previously reported on THEMATICS ( theoretical microscopic titration curves ) , a simple and fast computational tool for the prediction of catalytic and recognition sites in proteins that requires only the 3-D structure of the query protein as input [1–7] . THEMATICS is based on Poisson–Boltzmann calculations of the electrical potential for the protein structure , calculation of the theoretical titration curves ( average charge as a function of pH ) for all of the ionizable residues , and then statistical analysis of the computed titration curves to identify the ones that deviate the most from typical Henderson–Hasselbalch behavior . Clusters in coordinate space of two or more residues with deviant theoretical titration behavior are considered predictive and indeed predict localized interaction sites in proteins with high recall ( 91% ) and high precision , as measured by the low filtration ratio ( the fraction of ionizable residues selected ) , of about 8% . Here we report on how these predictive tools can be used to aid the experimental study of proteins of unknown function . In the present paper we focus on a family of structurally similar proteins of biomedical importance that apparently have different biochemical functions , the DJ-1 superfamily . Human DJ-1 is a protein of unclear function that apparently plays a neuroprotective role and is involved in the cellular response to oxidative stress [8] . Mutations of DJ-1 have been associated with certain forms of early onset Parkinson disease , and DJ-1 has been independently identified as a ras-dependent oncogene . Members of the DJ-1 superfamily have been annotated as proteases because of similarity to a bacterial protease . However , recent experimental evidence suggests that DJ-1 and some other family members are not proteases . The purpose of the present paper is to sort these structurally similar proteins into functional classes , based on theoretical predictions of active site residues and the spatial arrangements of these residues . We compare THEMATICS predictions with the experimental evidence that is currently available and argue that these structurally similar proteins fall into at least three distinct functional classes . The catalytic power of an enzyme relies not only on the nature of the residues that aid catalysis , but also on their position relative to the substrate . The method that identifies residues in the active site of a structure therefore also locates their relative positions and defines the type of chemistry that is possible , and potentially the substrate that can be recognized . Here we show that the arrangements in space of the residues predicted by our method form structural motifs from which one can deduce important clues about functionality . We illustrate the principle with a set of structurally similar proteins with different probable functions . Our predictions enable the similar structures to be sorted into distinct functional categories .
A search [9] for structures similar to DJ-1 was performed , and 11 structures with a Dali Z score of 15 or higher and an RMSD of 2 . 3 or less were chosen . The next closest proteins had significantly lower Z scores ( 7 . 6 or lower ) and higher RMSD ( 3 . 0 or higher ) . The structures included in the analysis are now described . Unlike some other members of the DJ-1 superfamily ( PfpI family ) , human DJ-1 does not exhibit any significant protease activity . Another family member , the YajL ( formerly labeled ThiJ ) protein from E . coli , is of unknown function [10] . Protease I from P . horikoshii is a known cysteine protease [11] , from which many other proteins in this group have been annotated in sequence databases . PfpI/YhbO from E . coli is a hypothetical protein of unknown function . YDR533Cp from S . cerevisiae is of unknown function [12] . The chaperone Hsp31 from E . coli is a known chaperone with some reported peptidase activity [13] . APC35852 from Bacillus stearothermophilus is a structural genomics protein of unknown function . Two E . coli structures with PDB IDs 1VHQ and 1OY1 are of the identical protein , with the sequences differing only at the C-terminal His tag . Both of these are structural genomics proteins , and the structures were determined by two different groups . 1VHQ is annotated as an enhancing lycopene biosynthesis protein , and 1OY1 is annotated as a putative sigma cross-reacting protein . All of these proteins are members of the DJ-1 superfamily and share closely related 3-D structures in their core fold . These 3-D structures are distinguished from one another by variable insertions into the core fold and by different quaternary structures . Table 1 summarizes the annotations for these proteins given in the databases of Pfam ( http://www . sanger . ac . uk/Software/Pfam ) , Gene Ontology ( http://www . geneontology . org/index . shtml ) , and the PDB ( http://www . rcsb . org/pdb/home/home . do ) . Two additional structures , Catalase I from Neurospora crassa and Catalase II from E . Coli , both have a domain of similar structure to DJ-1 , but the catalytic sites are located in a different domain . For the two catalases , THEMATICS correctly predicts the catalytic sites and predicts nothing in the domains with structural similarity to DJ-1 . There is no experimental evidence of any functional activity in the DJ-1 domain of these catalases . Therefore , these two catalases are excluded from the present analysis of functional classification of the DJ-1 superfamily members . The different types of quaternary structures in the DJ-1 family are illustrated in Figure 1 , showing ribbon diagrams of the dimer structures of the first six of the above DJ-1 family members plus the putative enhancing lycopene biosynthesis protein , with the two subunits colored red and blue in each structure . For all of the structures , the red subunits are oriented so that they are superimposable on each other without rotation . DJ-1 and YajL form similar dimer structures . Protease I and YhbO likewise are similar to each other , with dimer interfaces at a surface different from that of DJ-1/YajL . On the other hand , YDR533Cp and Hsp31 form quaternary structures that are different from each other , with the blue subunit attaching at a common face on the red subunit but at different orientation . The DJ-1 family proteins illustrate the difficulty of functional annotation from sequence [14] . The sequence alignments for this set of proteins ( ranging from 16%–35% identity ) might mislead one into concluding that their functions are similar . Especially the presence of a cysteine in similar positions within each sequence was considered highly suggestive of function . Thus , originally DJ-1 was presumed to be a cysteine protease because of its sequence similarity to the known protease . Later Bandyopadhyay and Cookson [14] studied 311 sequence homologues and analyzed their alignments and phylogenetic trees . These authors report that this set of sequences may be sorted into distinct subgroups; proteins with similar annotations appear to cluster together into distinct clades . The subgroup closest to that of human DJ-1 is the bacterial YajL/ThiJ group , suggesting that DJ-1 may have evolved from bacterial thiamine synthesis genes that have assumed some other function in eukaryotes . We have analyzed the DJ-1 sequence using the Joined Assembly of Function Annotations ( JAFA ) server ( http://jafa . burnham . org ) [15] . JAFA attempts to find consensus among five different sequence-based function annotation methods: GOFigure [16] , GOblet [17] , InterProScan [18] , GOtcha [19] , and PhydBac [20] . For human DJ-1 , three of the five servers were unable to annotate the sequence and returned no predictions . GOFigure reported possible thiamin pyridinylase activity and possible peptidase activity , with the higher score given to the former annotation . The PhydBac analysis gave the highest score to iron ion binding , the next-highest score to heme binding , and the third highest to catalase activity; it also indicated possible biological roles in response to oxidative stress and in response to biotic stimulus . No consensus could be found among the five methods , and thus this sequence analysis is inconclusive . A structural alignment of the monomers of the first six proteins indicates clearly that there are differences in residue arrangements that the sequence alignment cannot reveal . The residues identified as functionally important by THEMATICS are a subset of the structurally aligned residues . These predicted residues show spatial patterns that allow the different proteins to be sorted into groups . THEMATICS predictions , expressed as 3-D constellations of potentially important residues , strongly suggest probable functional groupings . Table 2 shows the THEMATICS predicted clusters for six proteins in the DJ-1 structural family . Structurally aligned residues are aligned in columns in Table 2 . When the conserved cysteine residue is not predicted by THEMATICS , it is shown in Table 2 in parentheses . Note that one residue in a structurally conserved position is predicted to be important for all six proteins; this is a glutamate corresponding to the active site E15 of Protease I . Table 2 suggests that there are three different types of functional sites for the first six proteins . For Protease I from P . horikoshii , THEMATICS predicts a cluster at the protease active site that includes the catalytic triad members C100 , H101 , and E74′; this triad is characteristic of cysteine proteases . Note that the prime indicates a residue from another subunit . A site similar but not identical to that of Protease I is predicted for PfpI/YhbO . Protease I and PfpI/YhbO have similar quaternary structures and similar interfaces . Their THEMATICS predicted sites are located at the interface . For human DJ-1 , THEMATICS finds a distinctly different cluster consisting of E15 , E16 , E18 and D24′ , located adjacent to , but not coinciding with , the corresponding triad site . The sites predicted for DJ-1 and YajL are very similar . The predicted sites consist of four structurally aligned acidic residues . There is one residue difference between the two predictions , in that for YajL R27′ is also predicted . Again , the quaternary structures are similar to each other with similar interfaces . For yeast YDR533Cp , THEMATICS predicts E30 , D57 , H108 , H139 , and E170 , a cluster that overlaps with the corresponding triad site and also contains some additional residues that are not selected either for Protease I or for DJ-1 . H139 is located in a position corresponding to that of the catalytic His101 of the Protease I triad , whereas E30 in the predicted YDR533Cp cluster is structurally aligned with E18 of human DJ-1 . H108 and E170 in the predicted YDR533Cp cluster are not predicted for DJ-1 or Protease I , but the corresponding residues are predicted for the chaperone Hsp31 . For YDR533Cp and Hsp31 , the predicted sites are each contained within a given monomer . Indeed , the similarities in the predicted sites are apparent for the structurally aligned monomers of YDR533Cp and Hsp31 , but these two proteins form dimers with different orientations between the subunits . Figure 2 shows a side-by-side comparison of the predicted active site residues in the dimer structures of Human DJ-1 ( Figure 2A ) and YajL ( Figure 2B ) , Protease I ( Figure 2C ) and PfpI/YhbO ( Figure 2D ) , and YDR533Cp ( Figure 2E ) and Hsp31 chaperone ( Figure 2F ) , plus the prediction for the dimer structure of Enhancing lycopene biosynthesis protein ( 1VHQ; Figure 2G ) and for the monomer structure of APC35852 ( 1U9C , Figure 2H ) . Ribbon diagrams are shown with the backbone of the “a” subunit of the dimer in green and the side chains of the THEMATICS predicted residues from the “a” chain in red; the backbone of the “b” subunit is shown in yellow with the side chains of the THEMATICS predictions from the “b” chain in blue . Note the similar spatial arrangements and locations of the predicted sites for DJ-1 and YajL . Predictions for Protease I and PfpI/YhbO are also similar in spatial arrangement in their relative positions in the structures . YDR533Cp and Hsp31 have predicted clusters located within each subunit , removed from the dimer interface , unlike the first four structures . Note also that the way in which the monomers of YDR533Cp and of Hsp31 come together to form the dimer is different , although the monomers and the predicted sites within them are similar . The two structural genomics protein structures 1VHQ ( Figure 2G ) and 1U9C ( Figure 2H ) have predicted sites quite different from those of the first three pairs of structures . Figure 3 shows superpositions of the THEMATICS-predicted active site residues in magenta and green . Note the similarities in the predicted sites for Figure 3A , DJ-1 ( magenta ) and YajL ( green ) ; Figure 3B , Protease I ( magenta ) and YhbO ( green ) ; and Figure 3C , YDR533Cp ( magenta ) and Hsp31 ( green ) . The yellow and red residues are conserved cysteine residues that are not THEMATICS positives . They are shown in the picture for comparison purposes . This conserved cysteine is shown in Figure 3A , YajL ( yellow ) and DJ-1 ( red ) ; Figure 3B , YhbO ( yellow ) ; Figure 3C , Hsp31 ( yellow ) and YDR533Cp ( red ) ; and Figure 3D , APC35852 ( yellow ) and YDR533Cp ( red ) . The conserved cysteine in Protease I is a THEMATICS positive residue and is shown in Figure 3B in magenta . Even though YDR533Cp and Hsp31 have different quaternary structures , their THEMATICS-predicted active sites are the same except that Hsp31 has one additional histidine residue , H74 . Superposition of their monomers yields nearly identical active site predictions for the remaining five residues . Figure 3D shows a superposition of the predicted residues of APC35852 with those of YDR533Cp . While the analysis illustrated in Figure 3 suggests three different functional classes for those six structures with a common fold , there are probably additional functions for this 3-D structure . For instance , one domain of Catalase-1 ( PDB ID 1SY7 ) [21] is structurally aligned with DJ-1; its catalase active site is in a different domain and is correctly predicted by THEMATICS; nothing is predicted in its DJ-1 domain , consistent with available experimental information . The structural genomics protein 1VHQ , annotated as Enhancing Lycopene Biosynthesis Protein , has a predicted site that somewhat resembles that of DJ-1 but is not clearly coincident with any of the structures studied . The structural genomics protein APC35852 ( PDB ID 1U9C ) is a monomeric protein , and THEMATICS predicts the site [E27 , H96 , H127 , D154 , E156 , E157] . This prediction is closest to those for YDR533Cp and Hsp31 . The E27 is structurally aligned with the glutamate that is common to the THEMATICS predictions for the structures of all of the first six proteins , the H96 and H127 are structurally aligned with two predicted histidines in YDR533Cp ( H108 and H139 ) , as shown in Figure 3D , and in Hsp31 ( H155 and H186 ) , while the E156 and E157 are not structurally aligned with any of the predicted residues for the above six proteins .
It has been shown previously that a relatively small group ( of about five to seven members ) of functionally important residues constitutes a 3-D signature that can be used to identify proteins in a superfamily [22] . Given that the different functional classes within the superfamily have evolved to affect different chemical transformations and to recognize different substrate molecules , it is likely that the full list of residues involved in catalysis and/or in recognition in each structure will contain not just signature residues of the superfamily but also residues characteristic of the particular functional class within the superfamily . THEMATICS is designed to identify exactly those characteristic residues involved in catalytic activity and in substrate specificity [1 , 2 , 4 , 5] . The predicted THEMATICS spatial clusters for the selected members of the DJ-1 family enable us to sort them into groups with similar predicted active sites and hence presumably similar function . In particular , the spatial arrangements of the THEMATICS predicted residues for DJ-1 and YajL are similar and form one such group . The predictions for these two structures are different by one residue , R27′ , but this residue is close to the threshold between positive ( predicted ) and negative ( not predicted ) . The difference between the two predicted sites is small enough to indicate a likely common function for the two structures . Predictions for Protease I and PfpI/YhbO form similar , but not identical , spatial motifs and may constitute a distinct probable functional class . While the present analysis suggests that Protease I is the closest functional relative of YhbO , the two predicted sites do show some differences , and therefore one cannot conclude that YhbO is a cysteine protease . Indeed , Abdallah et al . recently reported [23] that YhbO exhibits neither protease nor chaperone activity . Ydr533c and the chaperone Hsp31 form yet a third probable functional class . The predicted sites for these two latter proteins are contained within each subunit , and the two proteins exhibit different quaternary structures . Thus , in spite of sequence similarity , it is likely that these six proteins belong to at least three different functional classes . Note that the six proteins have similar primary , secondary , and tertiary structures , yet the three predicted functional classes have different quaternary structures and different predicted functional sites . The three predicted functional classes are consistent with the positions of these proteins in the cladogram of Bandyopadhyay and Cookson [14] . The phylogenetic tree and the present method provide very different but complementary types of information . The cladogram indicates which proteins are the closest neighbors in the evolutionary history , based on sequence , while the present method identifies important functional residues and active site structural motifs , based on the 3-D structure . For the DJ-1 superfamily , the two methods support similar conclusions about the likely functional subclasses . Recently we have shown [7] that THEMATICS can make correct site predictions for comparative model structures . The question then arises , can the present method be used to annotate the members of the superfamily whose structures are not known ? This depends on the quality of the model structures and is the subject of further investigation . The facile identification of binding and recognition sites in proteins with a simple calculation provides important and time-saving clues in the determination of a protein's function .
THEMATICS analysis was performed on the protein structures according to the procedures described by Ko et al . [1] , using a Z score cutoff value of 0 . 99 in the statistical analysis and using a distance cutoff of 9 . 0 Å to form the clusters . Structural alignments were performed using a Combinatorial Extension method and the 3-D Protein Structure Comparison and Alignment Server ( http://cl . sdsc . edu ) [24] . Structures were rendered using the graphical programs PyMol ( http://www . pymol . org ) and Yasara ( http://www . yasara . org/index . html ) .
The accession numbers from the Protein Data Bank ( http://www . rcsb . org/pdb/home/home . do ) used in this paper are: human DJ-1 ( 1SOA ) , Catalase-1 ( 1SY7 ) , Protease I from P . horikoshii ( 1G2I ) , PfpI/YhbO from E . coli ( 1OI4 ) , YDR533Cp from S . cerevisiae ( 1RW7 ) , chaperone Hsp31 from E . coli ( 1N57 ) , Catalase II from E . Coli ( 1GGE ) , YajL ( formerly labeled ThiJ ) protein from E . coli ( 2AB0 ) , Enhancing lycopene biosynthesis protein ( 1VHQ ) , putative sigma cross-reacting protein ( 1OY1 ) , and structural genomics protein APC35852 ( 1U9C ) . | Genome sequencing has led to the discovery of many new gene products , proteins . These discoveries hold tremendous potential for totally new approaches to the diagnosis and treatment of disease . To realize this potential , one important step is to understand the function of the thousands of proteins whose function is currently unknown . One of these proteins of unknown function is human DJ-1 , a protein that appears to play a protective role against Parkinson and other neurodegenerative diseases . Here we present a computational approach to the classification by function of DJ-1 and its family members . Eight DJ-1 family members , all with similar 3-D structure , are analyzed . Three different probable functional classes emerge from this analysis on six of the family members , all with a simple calculation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"neuroscience",
"neurological",
"disorders",
"computational",
"biology"
] | 2007 | Identification of Functional Subclasses in the DJ-1 Superfamily Proteins |
Gag , as the major structural protein of HIV-1 , is necessary for the assembly of the HIV-1 sphere shell . An in-depth understanding of its trafficking and polymerization is important for gaining further insights into the mechanisms of HIV-1 replication and the design of antiviral drugs . We developed a mathematical model to simulate two biophysical processes , specifically Gag monomer and dimer transport in the cytoplasm and the polymerization of monomers to form a hexamer underneath the plasma membrane . Using experimental data , an optimization approach was utilized to identify the model parameters , and the identifiability and sensitivity of these parameters were then analyzed . Using our model , we analyzed the weight of the pathways involved in the polymerization reactions and concluded that the predominant pathways for the formation of a hexamer might be the polymerization of two monomers to form a dimer , the polymerization of a dimer and a monomer to form a trimer , and the polymerization of two trimers to form a hexamer . We then deduced that the dimer and trimer intermediates might be crucial in hexamer formation . We also explored four theoretical combined methods for Gag suppression , and hypothesized that the N-terminal glycine residue of the MA domain of Gag might be a promising drug target . This work serves as a guide for future theoretical and experimental efforts aiming to understand HIV-1 Gag trafficking and polymerization , and might help accelerate the efficiency of anti-AIDS drug design .
Gag protein ( Gag ) is the major structural polyprotein of HIV-1 and is synthesized in large amounts in the cytoplasm . Gag diffuses freely within the cytoplasm , hijacks the molecular motors , and moves along microtubules to the cytosolic side of the plasma membrane ( PM ) domain [1 , 2] . Underneath the PM , the immature HIV-1 Gag shell assembles in a radial arrangement . Gag is composed of six constitutive components: the N-terminal matrix ( MA ) domain , the capsid ( CA ) domain , the first spacer peptide ( SP1 ) , the nucleocapsid ( NC ) domain , a second spacer peptide ( SP2 ) and the p6 ( p6 ) domain . During the phase of HIV-1 maturation , Gag disconnects from the MA and reassembles to form the cone-shaped viral core . Therefore , Gag is necessary for HIV-1 replication , interfering with the trafficking and assembly of Gag has been a focus of research [3 , 4] . In the field of theoretical research , Liu et al . [5] first proposed a convection-diffusion equation model to explore the transport of Gag monomers in the cytoplasm . Based on the experimental finding of a monomer-dimer equilibrium in solution under certain biochemical conditions [6] , Wang et al . [7] presented a model studying the transport of Gag monomers and trimers in the cytoplasm . These researchers analyzed the relationship between the timing of the initial appearance of HIV-1 capsid on the PM and the various model parameters . Sadre-Marandi et al . [8] and Liu et al . [9] simulated HIV-1 viral capsid assembly through dynamical systems . A recent study on the events initiating HIV-1 Gag assembly was conducted by Kutluay et al . [10] . These researchers presented quantitative descriptions of monomers and multimers in the cytoplasm and PM , respectively , and demonstrated that only monomer and low-order multimers ( e . g . , dimer ) of Gag were found in the cytoplasm , and that high-order multimers were formed only underneath the PM . In addition , these researchers studied two mutations of Gag: a mutated version of Gag-GFP that lacked the CTD of the CA of Gag ( Gag-dCTD ) and a mutation of the N-terminal glycine residue of the MA to alanine ( Gag-G2A ) . To the best of our knowledge , the reported models [5 , 7 , 11–13] focus only on Gag trafficking in the cytoplasm and do not simultaneously consider the polymerization of Gag underneath the PM . Therefore , we developed a reaction-advection-diffusion equation model to describe the trafficking of two particles and the polymerization of various particles . In our model , we focus on the following aspects: We first estimated the parameters of the model based on experiment data [10] and assessed the robustness of the model . We subsequently applied this model to two mutation cases: Gag-dCTD and Gag-G2A . We also predicted the budding and release time of HIV-1 virus-like particles ( VLPs ) . Using our model , we then analyzed the weight of the pathways involved in the polymerization reactions and deduced the key intermediates in hexamer formation . Moreover , we explored four theoretical combined methods for suppressing the Gag concentration and identified a promising drug target . This work will lead to a better understanding of the dynamics of Gag-Gag interaction and Gag trafficking , which are important in the emergence of HIV , and it might provide theoretical guidance for the design of antiretroviral drugs .
This work aimed to assess the Gag trafficking in the cytoplasm and Gag polymerization underneath the PM . The schematic diagram used to develop the mathematical model is shown in Fig 1 . Several assumptions were made to simplify the model: A1 The cytoplasm is an annulus [5 , 14] . A2 In the cytoplasm , monomers can aggregate into dimers , but can not form higher-order polymers [10 , 15–18] . A3 Monomers and dimers are transported to the PM along microtubules by molecular motors ( e . g . , kinesin and dynein ) , and can also diffuse freely in the cytoplasm [5 , 11 , 13 , 14] . A4 Gag is synthesized by ribosomes attached to the endoplasmic reticulum ( ER ) . A large amount of ER is located in the perinuclear region , and a slight amount of ER is found underneath the PM . We approximated that the density of ER decreased exponentially from the perinuclear region to the PM . In addition , newly synthesized Gag monomers are distributed throughout the cytoplasm , but concentrated in the perinuclear region [19] . Based on these assumptions , we assumed that the synthesis rate of Gag decreased ( roughly ) exponentially from the perinuclear region to the PM . A5 Gag is observed on the PM within 5-10 minutes post-synthesis [20 , 21] . Typically , a period of 5-6 minutes is required to complete the assembly of a single VLP [22] . The budding and release time of a VLP is approximately 6 hours [22] . Furthermore , some researchers [23 , 24] have concluded that the hexamer is the building block of HIV-1 . Therefore , we hypothesized that Gag monomers can only aggregate into dimers , trimers , tetramers , pentamers , and hexamers during the first 30 minuntes , and this assumption was mainly used to build the mathematical model ( Eq ( 6 ) ) in the boundary ( PM ) . A6 Gag polymers degrade with different degradation rates [21] . A7 Some molecular motors ( e . g . , kinesin ) move unidirectionally from the microtubule-organizing center to the cell periphery , whereas others ( e . g . , dynein ) move toward the cell nucleus [25 , 26] . During the process of egress , the difference between the outward and inward speeds is denoted the velocity of egress [5 , 14] . In the cytoplasm ( rN < r < rC ) , the chemical reactions involving monomers and dimers can be described as follows: 2 G a g ⇌ k 1 ′ k 1 G a g 2 , ϕ → g 1 ( r ) G a g , G a g → d 1 ϕ , G a g 2 → d 2 ϕ ( 1 ) where Gag is a monomer and Gag2 is a dimer . Definitions of the variables and symbols are provided in Table 1 . In rectangular coordinates , the transport velocity of monomer is v1 = ( v1x , v1y ) , where v1x and v1y are the velocities along the x and y directions , respectively . For simplicity , we switched the problem to polar coordinates . Thus , the velocity of monomer along the radial direction is denoted by s1 , and the angle between s1 and the polar axis is denoted by θ1 . Therefore , v1 = ( s1 cos θ1 , s1 sin θ1 ) . The total flux of monomer transportation includes both convective and diffusive transport: ∇ ⋅ ( v1P1 − D1∇P1 ) . In polar coordinates , the above equation yields: ∇ · ( v 1 P 1 - D 1 ∇ P 1 ) = 1 r ∂ ∂ r ( s 1 r P 1 - D 1 r ∂ P 1 ∂ r ) The equation for total dimer flux has the same form . Based on the mass conservation law and mass action law [27] , we obtain the following reaction-diffusion-transport equations: { ∂ P 1 ∂ t = 1 r ∂ ∂ r ( D 1 r ∂ P 1 ∂ r − s 1 r P 1 ) + 2 k 1 ′ P 2 − 2 k 1 P 1 2 + g 1 ( r ) − d 1 P 1 ∂ P 2 ∂ t = 1 r ∂ ∂ r ( D 2 r ∂ P 2 ∂ r − s 2 r P 2 ) + k 1 P 1 2 − k 1 ′ P 2 − d 2 P 2 ( 2 ) where t ∈ ( 0 , T ) , r ∈ ( rN , rC ) . At the outer membranes of the nucleus ( r = rN ) , impermeable wall boundary conditions are considered as follows: { D 1 r ∂ P 1 ∂ r - s 1 r P 1 = 0 D 2 r ∂ P 2 ∂ r - s 2 r P 2 = 0 , r = r N ( 3 ) Gag proteins gathered at the “Gag hotspots” underneath the PM , which has a thickness of approximately 20 nm [14] . This domain of “Gag hotspots” is considered a volume , and we set it as the boundary of our model , similarly to the strategy used in a previous study [28] . Therefore , the concentrations of polymers on the boundary reflect all the volume concentrations , which have the same units at P1 and P2 in the cytoplasm . At the PM ( r = rC ) , monomer transport includes both convection and diffusion , and the same is true for dimer transport . However , underneath the PM , a myristoyl group of Gag can attach to the PM , resulting a weaker free diffusion of Gag compared with that in the cytoplasm . Therefore , we reduced the diffusion coefficient in the cytoplasm by kD to obtain the diffusion coefficient underneath the PM . Gag proteins are transported by molecular motors along microtubules , which are found throughout the cytoplasm , but are relatively rare underneath the PM . Therefore , the velocity of Gag loading to the PM is relatively small . We thus also reduced the velocity in the cytoplasm by ks to obtain the transport coefficient underneath the PM . Therefore , the monomer and dimer fluxes can be computed as follows: k D D 1 r ∂ P 1 ∂ r - k s s 1 r P 1 and k D D 2 r ∂ P 2 ∂ r - k s s 2 r P 2 , r = r C ( 4 ) When the termination time T is approximately 30 minutes , Gag monomers can only aggregate into dimers , trimers , tetramers , pentamers , and hexamers based on assumption A5 . Thus , the interactions among monomers , dimers , trimers , tetramers , pentamers , and hexamers underneath the PM ( r = rC ) were studied , and all possible chemical reactions based on the step-growth polymerization [29] are the following: G a g i + G a g j ⇌ k i j ′ k i j G a g i + j , i ≤ j , i + j ≤ 6 , i , j = 1 , 2 , ⋯ , 6 , G a g n → d n ϕ , n = 1 , 2 , ⋯ , 6 , ( 5 ) where Gagi is a polymer with i monomers , kij is the on-rate constant , k i j ′ is the off-rate constant and dn is the degradation rate of n-mers . By combining with the above mentioned chemical reactions and Eq ( 4 ) , the following boundary conditions at the PM are obtained: { ∂ P 1 ∂ t = 1 r ∂ ∂ r ( k D D 1 r ∂ P 1 ∂ r - k s s 1 r P 1 ) + 2 k ′ 11 P 2 - 2 k 11 P 1 2 + k ′ 12 P 3 - k 12 P 1 P 2 + k ′ 13 P 4 - k 13 P 1 P 3 + k ′ 14 P 5 - k 14 P 1 P 4 + k ′ 15 P 6 - k 15 P 1 P 5 - d 1 P 1 ∂ P 2 ∂ t = 1 r ∂ ∂ r ( k D D 2 r ∂ P 2 ∂ r - k s s 2 r P 2 ) + k 11 P 1 2 - k ′ 11 P 2 + k ′ 12 P 3 - k 12 P 1 P 2 + 2 k ′ 22 P 4 - 2 k 22 P 2 2 + k ′ 23 P 5 - k 23 P 2 P 3 + k ′ 24 P 6 - k 24 P 2 P 4 - d 2 P 2 ∂ P 3 ∂ t = k 12 P 1 P 2 - k ′ 12 P 3 + k ′ 13 P 4 - k 13 P 1 P 3 + k ′ 23 P 5 - k 23 P 2 P 3 + 2 k ′ 33 P 6 - 2 k 33 P 3 2 - d 3 P 3 ∂ P 4 ∂ t = k 22 P 2 2 - k ′ 22 P 4 + k 13 P 1 P 3 - k ′ 13 P 4 + k ′ 14 P 5 - k 14 P 1 P 4 + k ′ 24 P 6 - k 24 P 2 P 4 - d 4 P 4 ∂ P 5 ∂ t = k 23 P 2 P 3 - k ′ 23 P 5 + k 14 P 1 P 4 - k ′ 14 P 5 + k ′ 15 P 6 - k 15 P 1 P 5 - d 5 P 5 ∂ P 6 ∂ t = k 33 P 3 2 - k ′ 33 P 6 + k 24 P 2 P 4 - k ′ 24 P 6 + k 15 P 1 P 5 - k ′ 15 P 6 - d 6 P 6 ( 6 ) The initial conditions are Pi = 0 , i = 1 , 2 , ⋯ , 6 , which are based on the experimental data [10] . The nine polymerization reactions ( 5 ) underneath the PM and one polymerization reaction ( 1 ) in the cytoplasm have 20 parameters , including ki , j , k i , j ′ , i ≤ j , i+j ≤ 6 , i , j = 1 , 2 , ⋯ , 6 . To decrease the number of these parameters , we adopted the strategy described by Zlotnick et al . [30–32] in their study of the assembly kinetics of virus capsids . Zlotnick et al . used a system of equations to simulate the sequential aggregation of free building blocks into virus capsids . To reduce the number of parameters , these researchers developed a formula [30 , 31] that mapped the on-rate constant to the off-rate constant . In our study , Gag proteins are aggregated to form a hexamer , and this process has a lot in common with virus capsid assembly . For example , the virus capsid and the subunit in the work conducted by Zlotnick et al . correspond to the hexamer and the low-order polymer serving as one of the two reactants in each polymerization reaction in our work , respectively . For the polymerization reactions ( 5 ) , the association constant Ki+j of Gagi+j can be separated into two statistical components SIi , j and Si , j , and a non-statistical association constant K i + j ′ . These are related by the following function: K i + j = k i , j k i , j ′ = S I i , j S i , j K i + j ′ ( 7 ) where the statistical factor SIi , j describes the degeneracy of the incoming subunit . The second statistical factor Si , j can be treated as the ratio of two factors: the number of pathways for the formation of Gagi+j from Gagi and Gagj and the number of pathways for the dissociation of Gagi+j to Gagi and Gagj . K i + j ′ is a function of the number of contacts formed , i . e . , K ′ i + j = e - c i , j Δ G / R T ( 8 ) where ci , j is the number of contacts of Gagi and Gagj , ΔG is the free energy associated with the formation of a contact , -2 . 72 kcal mol−1 , R is the gas constant , 1 . 987 cal deg−1mol−1 , and T is the temperature in Kelvin , 298K . As determined by substituting Eq ( 8 ) into Eq ( 7 ) , k′i , j can be given by the parameter ki , j based on the following function: k ′ i , j = k i , j S I i , j S i , j e - c i , j Δ G / R T ( 9 ) As an example , the evaluation of k 1 , 5 ′ is described . Owen et al . [23] found that Gag monomers could create a hexameric ring , which was believed to serve as the building block of HIV-1 , thus the hexamer can be considered a hexagon from the perspective geometry . Then , i edges next to each other are removed from the hexagon , and we used these as the geometry of the i-mers of Gag . Fig 2 shows the reaction in which a monomer and a pentamer aggregate to form a hexamer . There is only one way to add a monomer to a pentamer to form a hexamer , and there are six ways to dissociate a hexamer to a monomer and a pentamer , thus S1 , 5 = 1/6 . The incoming subunit is the monomer , thus SI1 , 5 = 1 . Two contacts ( ci , j = 2 ) are made in forming of Gag6 , resulting in a free energy of 2ΔG . Thus , k ′ 1 , 5 = k 1 , 5 1 × 1 / 6 × e - 2 Δ G / R T . For the polymerization reactions ( 5 ) , SIi , j , Si , j and ci , j are counted and these are listed in Table 2 . We only need to optimize the values of ki , j , because k i , j ′ can be obtained by the above-mentioned function ( 9 ) . Therefore , the 20 parameters for nine polymerization reactions ( 5 ) underneath the PM and the single polymerization reaction ( 1 ) in the cytoplasm are cut by half . Combined with two proportionality coefficients kD , ks and the velocity s1 of Gag-G2A , this results in 13 parameters that needed to be optimized . Thus , the number of parameters to be optimized was substantially decreased . The radius of the cell nucleus is ∼5 μm [5 , 14] , and the radius of the cell is ∼10 μm [5 , 14] . The diffusion coefficient for an “average” ( 3 − 6 nm diameter ) protein in the cytoplasm is 5 − 15 μm2/s [33] . The Gag monomer is a highly extended rod with a length of ∼20 nm and a width of 2 − 3 nm [5] , resulting in an average mean of 11 . 25 nm between the length and width . According to the Stokes-Einstein equation , the diffusion coefficient is inversely proportional to the diameter . Therefore , we estimated the diffusion coefficient of Gag as ∼4 μm2/s . Similarly , the diffusion coefficient for a Gag dimer is approximately half of the corresponding value for a Gag monomer . The velocities of the active transport of a monomer and a dimer are approximately equal to the velocity of the molecular motor ( ∼1 μm/s [33] ) in the cytoplasm . Tritel et al . [34] found that 80% of Gag disappeared within 2 hours after synthesis . Therefore , we estimated that the degradation rate of a monomer was ~ 2 . 236 × 10−4/s , and the degradation rate of i-mers was thus ∼2 . 236 × 10−4/i /s . The above-described parameter values estimated by the experimentally measured data are listed in Table 3 . We adopted the Crank-Nicolson method for discretizing the convection-diffusion-reaction equations to form nonlinear equations , and then used Newton’s method to solve them . First , we discretized the system of convection-diffusion-reaction equations using the Crank-Nicolson method . Let the time step and grid size of the radius be Δt and Δr , respectively . Then , the i − mers concentration is denoted by P i , n k = P i ( r N + n Δ r , k Δ t ) . The derivatives of P i , n k with respect to t and r are discretized as follows: ∂ P i ∂ t = P i , n k + 1 - P i , n k Δ t ∂ P i ∂ r = 1 2 ( P i , n + 1 k + 1 - P i , n - 1 k + 1 2 Δ r + P i , n + 1 k - P i , n - 1 k 2 Δ r ) ∂ 2 P i ∂ r 2 = 1 2 ( P i , n + 1 k + 1 - 2 P i , n k + 1 + P i , n - 1 k + 1 ( Δ r ) 2 + P i , n + 1 k - 2 P i , n k + P i , n - 1 k ( Δ r ) 2 ) ( 10 ) The system of convection-diffusion-reaction equations ( Eqs 2 , 3 and 6 ) were discretized according to the above-mentioned rules ( 10 ) , and the resulting equation can be rewritten using vectors as A ( Δ t , Δ r , r ) X k + 1 + F ( r , X k + 1 ) = B ( Δ t , Δ r , r ) X k ( 11 ) where N is the grid number of the radius , X k = ( P 1 , 0 k , P 1 , 1 k , ⋯ , P 1 , N k , P 2 , 0 k , P 2 , 1 k , ⋯ , P 2 , N k , ⋯ , P 6 , 0 k , P 6 , 1 k , ⋯ , P 6 , N k ) T , A and B are all ( 6N + 6 ) × ( 6N + 6 ) matrices that depend on the time step Δt , grid size Δr and radius r . F is the nonlinear part , which depends on r and Xk+1 . At each step , Eq ( 11 ) is a nonlinear algebraic equation that can be solved using Newton’s method . The following process is repeated F 1 ( Δ t , Δ r , r , X k , X k - 1 ) = A ( Δ t , Δ r , r ) X k + F ( r , X k ) - B ( Δ t , Δ r , r ) X k - 1 Δ X k = ( ∇ F 1 ( Δ t , Δ r , r , X k , X k - 1 ) ) - 1 ( F 1 ( Δ t , Δ r , r , X k , X k - 1 ) ) X k + 1 = X k - Δ X k ( 12 ) until a sufficiently accurate value is reached . The numerical algorithms were implemented in MATLAB 2009b on a personal computer . To ensure numerical accuracy , a small time step Δt = 1 s and grid size Δr = 0 . 025 μm were used . The numerical solutions converged for Δt in the range from 0 . 5 to 36 s and Δr in the range from 0 . 0063 to 0 . 05 μm , respectively . Some parameters were determined from a variety of sources , as illustrated in Table 3 , and the others needed to be obtained using an optimization method . For WT Gag , 12 free parameters needed to be optimized . In contrast , for Gag-G2A , only one parameter s needed to be optimized , and the other parameter values are equal to the corresponding values for WT Gag . Therefore , 13 parameters needed to be optimized by fitting to 16 experiment data points ( eight data points for WT Gag and eight data points for Gag-G2A [10] ) . The flow chart of this process is shown in Fig 3 . The advantage of the sequential scheme in Fig 3 is that the second object function will not be run until the first one meets the error criterion , and this process can reduce the program running time on a personal computer . If this program runs in a supercomputer with thousands of computers , other schemes for parallel computing , such as the weighted multi-objective scheme [35 , 36] , would be more efficient . Thirteen parameters need to be optimized using 16 experiment data points ( eight data points for WT Gag and eight data points for Gag-G2A [10] ) : this is the inverse problem . In addition , the measured data is always inevitable mixed with noise . In numerical computation , this problem is often ill-posed . To decrease over-parametrization and guarantee numerical stability of this optimization problem , a regularization term using the Tikhonov regularization method is generally added ( Eq 13 ) [37–39] . The idea of regularization is to add preference to a particular solution with desirable properties [38 , 40–42] . In many cases , the solution is given preference with smaller norms , and this process is known as L2 regulation . This regulation improves the conditioning of the problem , enabling a direct numerical solution . The form of regularization is given as: min θ 1 ≤ θ ≤ θ 2 J ( θ ) = ‖ Y ( θ ) − Y ( exp ) ‖ 2 + λ ‖ θ ‖ 2 ( 13 ) where θ is the parameter vector , θ1 and θ2 are the lower and upper bounds of θ , respectively . Y ( θ ) and Y ( exp ) are the calculation and the experimental data , respectively . ‖ ⋅ ‖2 is the Euclidean norms . λ‖θ‖2 is the regularization term , and λ is the weight coefficient , which is generally small and is set to 0 . 001 . Because the model and the boundary conditions are nonlinear , intelligent optimization algorithms , such as Differential Evolution ( DE ) and Particle Swarm Optimization ( PSO ) , are commonly used to obtain the parameter values . Here , we use the diversity-maintained differential evolution based on a gradient local search ( DMGBDE ) method proposed by Xie et al . [43] , which might have improved local search ability . The DMGBDE procedure can be described as follows .
Kutluay et al . [10] used a chemical crosslinking approach to analyze the initiating events in HIV-1 assembly and genome packaging . In their experiment , 293T cells coexpressing WT Gag and HIV-1 RNA were crosslinked by treatment with EGS , a membrane-permeable crosslinker . After 30 minutes of incubation at room temperature , crosslinking was prevented by the addition of Tris-Cl . The cells were then analyzed through membrane flotation assays . Proteins from the PM and cytoplasmic fractions , including monomers , dimers , trimers , tetramers , pentamers , and hexamers , were precipitated , and their relative concentrations were obtained by western blotting . The values of the parameters were optimized and are shown in Table 4 . The simulated absolute concentrations and experimentally measured relative concentrations of the polymers were normalized by dividing by the concentration of Gag monomer in the cytoplasm . As shown in Fig 4 , the simulation results are consistent with the experimental data . MA comprises the N-terminus of the Gag polyprotein , and it is responsible for targeting the Gag polyprotein to the PM . Therefore , mutation of the N-terminal glycine residue of MA to alanine ( G2A ) can reduce the attachment of a myristoyl group to Gag and impede its recruitment to the PM [44 , 45] . In our model , the speed of Gag-G2A transport is slower compared with that of Gag WT . In addition , we assumed that kD and ks in Eq ( 6 ) for Gag-G2A were equal to the corresponding values for WT Gag . Therefore , we adjusted only one parameter s1 to fit the experimental data [10] . The value of this parameter is 2 . 20 × 10−11 μm/s , and its 95% confidence interval is [0 , 0 . 09] . Because s1 is close to zero , we concluded that Gag-G2A might fail to hijack the molecular motor . The comparison between the simulation values and the experimental data , which is shown in Fig 5 , clearly demonstrates that the simulation and experimental results are similar . In our study , we used eight data points for WT Gag and eight data points for Gag-G2A [10] to fit 13 parameters , including 10 polymerization coefficients ki , j , two proportionality coefficients kD and ks and the transfer speed s1 of Gag-G2A . The constraints for the 13 parameters are as follows: After the parameter values are optimized based on experimental data [10] , we evaluated how well the model parameters were determined by these data . In 2009 , Raue et al . [46] proposed an approach to analyze the structural and practical identifiability of dynamical models by exploiting the profile likelihood , and this method has subsequently been widely applied in many fields , particularly the computational systems biology [47–50] . In this work , we used this technology [46] to analyze the identifiability of the parameters . First , finite sample confidence intervals for the parameters were estimated , and these are listed in Table 4 . As shown , the confidence intervals of all of the parameters are within the bounds . The profile likelihoods of all the parameters are shown in Fig 6 . Specifically , the profile likelihoods for k11 , k12 , k23 , kD and ks show a steep concave shape , indicating that the optimization route can rapidly reach the minimum . The profile likelihoods for k1 , k15 and k24 also show a concave shape , however , the curves on the right side of the vertical dashed lines decrease slowly , indicating that their optimization routes might reach the minimum slowly . The profile likelihoods for k13 , k14 , k22 , k33 and G2A − s1 have several local minima , hence , more iterations might be needed for the optimization route to jump out of and not get stuck at these local minima . The CA is one of the four major domains of Gag and plays an important role in Gag multimerization and assembly at the PM . Furthermore , Gag: RNA binding is mediated by the CTD of the CA , which participates in Gag-Gag interactions . Thus , Gag-dCTD will show decreased on-rate constants . Furthermore , because the CA and MA of Gag are bound to each other , Gag-dCTD might show slightly impaired CA function . The damaged CA domain will slightly decrease the transport velocity of Gag , thus , the transport coefficient of Gag-dCTD might be slightly slower than that for WT Gag . This assumption is also supported by experimental data [10] for Gag-dCTD . Taken together , these assumptions indicate that the parameters s1 , s2 , ki , j are decreased compared with the corresponding values for WT Gag . These limiting conditions were included in the process of parameter optimization . The values of these parameters are listed in Table 5 , and the comparison between the simulation values and experimental data is shown in Fig 7 . As shown , the numerical results agree with the experimental data . According to various references [21 , 22 , 51 , 52] , a VLP buds and releases after ∼6 hours . In 2004 , Briggs et al . [3] reported that the diameter of a VLP was ∼145 nm , and Carlson et al . [53] then found that a VLP was released with ∼2400±700 Gag proteins . Jouvenet et al . [22] observed cells over a period of 30-60 minutes starting 5-6 hours after transfection and found that 50-150 puncta per cell typically appeared during this period . The behavior of these puncta could result in their classification into two discrete classes: slowly appearing puncta and rapidly appearing/disappearing puncta . The slowly appearing puncta represented the majority of events ( 74% ) observed at 5-6 hours after transfection , and were indistinguishable from areas of the PM . The rapidly appearing/disappearing puncta were indistinguishable from endosomes . Therefore , Jouvenet et al . believed that the slowly appearing puncta might represent genuine VLPs assembly events . Nermut et al . showed some pictures of VLPs in the budding state [21] , and these images showed that most Gag proteins gathered to VLPs . Thus , the total Gag protein in the budding state at the PM can be estimated by the number of VLPs and Gag proteins per VLP . Taken together , these findings indicate that the threshold surface density of Gag at the PM can be computed as follows: N G 6 . 023 × 10 23 × 10 24 × N p × p 4 π × R 2 ( u n i t : y m o l / u m 2 ) ( 14 ) where NG is the number of Gag proteins in a VLP , Np is the total puncta per cell in the budding state , p is the ratio of slowly appearing puncta , and R is the radius of a cell . When NG = 2400 , Np = 50 , p = 74% and R = 10 , the threshold surface density is 1 . 17 × 102 ymol/um2 . When Np is changed to 100 and 150 , the threshold surface densities are 2 . 34 × 102 ymol/um2 and 3 . 51 × 102 ymol/um2 , respectively . In our work , the surface density of Gag at the PM can be obtained by the following formula: S G a g = ( ∑ i = 1 6 i × C G a g i ) * H ( u n i t : y m o l / u m 2 ) where SGag denotes the surface concentration of Gag , CGagi denotes the volume concentration of Gagi underneath the PM , and H is the thickness of the concentrated domain of Gag underneath the PM , which is set to 0 . 13 um [5] . When the surface density of Gag is equal to the threshold surface density , the cumulative time is estimated as the budding and release time . Times of 3 . 51 , 8 . 53 and 21 . 34 hours are predicted for threshold surface densities 1 . 17 × 102 , 2 . 34 × 102 and 3 . 51 × 102 ymol/um2 , respectively . The predicted budding and release times agree with earlier findings to some degree [21 , 51 , 52] , indicating that the total concentration of Gag underneath the PM is reasonable . In 2011 , Tavener et al . [54] defined the sensitivity of the ith model output variable Oi ( P , T ) with respect to the jth parameter Pj at the time T , Si , j ( T ) as S i , j ( T ) = ∂ O i ( P , T ) ∂ P j These researchers also defined the elasticity of the ith output variable Oi ( P , T ) with respect to the parameter Pj , Ei , j ( T ) as E i , j ( T ) = P j O i ( P , T ) ∂ O i ( P , T ) ∂ P j Elasticity is defined in terms of relative sensitivity and can describe the rate of change in the relative size of the output variables with respect to the relative size of the parameters . An elasticity analysis would thus yield more reliable results . This definition has an extraordinarily wide range of applications [8] . Based on the literatures [55 , 56] , the elasticity function Ei , j ( T ) was estimated as follows: E i , j ( T ) = P j O i ( P , T ) ∂ O i ( P , T ) ∂ P j ≈ | O i ( P j + Δ P j , T ) - O i ( P j - Δ P j , T ) | / O i ( P j , T ) ( 2 Δ P j / P j ) where ΔPj is a small perturbation of the parameter Pj . The elasticity values for all parameters corresponding to eight outputs , specifically the monomer and dimer concentrations in the cytoplasm and the concentrations of each of the polymers at the PM , are shown in Fig 8 . As shown in Fig 8 , the elasticity values for the polymerization coefficients k11 and k12 are greater than those for the other polymerization coefficients . This findings indicates that perturbations of these two parameters can lead to relatively large changes in Gag polymer concentrations . Therefore , we can conclude that the corresponding two reactions , which involve the polymerization of two monomers to form a dimer and the polymerization of a monomer and a dimer to form a trimer , might be key reactions . If future drugs can decrease the values of k11 and k12 , concentrations of Gag high-order polymers will be significantly reduced . Among all of parameters , the highest elasticity value is found for ks , which measures the ability of Gag to land on the PM by active transport . Therefore , we can conclude that this process has a very significant impact on Gag polymers on the PM . As a result , suppression of the Gag concentration on the PM by reducing the ability of Gag to stay on the PM might be a good strategy . The elasticity value of kD is found to be very small , which illustrates that the landing of Gag on the PM by diffusion has little impact on changes in the concentrations of Gag polymers on the PM . In addition , the elasticity value for the transport speed s1 of Gag-G2A is very small with a value close to zero . This finding further supports the hypothesis that Gag-G2A might hardly be able to hijack molecular motors to move to the PM , and as a result , Gag-G2A might be an important drug target . The global elasticity function was defined as follows: G S A = 1 N ∑ k = 1 N | O ( P + Δ P k , T ) − O ( P , T ) | O ( P , T ) ‖ Δ P k P ‖ ∞ where N is the total number of perturbations , and ΔPk is the simultaneous perturbations of all parameters during the k-th perturbation . The global elasticity function values for 10% perturbations of all parameters are shown in Fig 9 . As shown in the figure , perturbations of all the parameters are not sensitive to the output , which indicates that the proposed model is reasonable and robust . The patterns underlying the formation of polymers constitute a very interesting topic [57 , 58] . Using our model , we analyzed the weights of the pathways for the tetramer , pentamer , and hexamer formation . Hexamer formation consists of three pathways: ( 1 ) 2 G a g 3 ⇌ k 33 ′ k 33 G a g 6 ( 2 ) G a g 2 + G a g 4 ⇌ k 24 ′ k 24 G a g 6 ( 3 ) G a g 1 + G a g 5 ⇌ k 15 ′ k 15 G a g 6 The rate equation for the hexamer concentration can be described as follows: d P 6 d t = k 33 P 3 2 - k 33 ′ P 6 + k 24 P 2 P 4 - k 24 ′ P 6 + k 15 P 1 P 5 - k 15 ′ P 6 - d 6 P 6 The three pathways increase the hexamer concentration based on the following rates: k 33 P 3 2 - k ′ 33 P 6 , k24 P2P4 − k′24P6 and k15P1 P5 − k′15P6 . The largest value among these rates corresponds to the predominant pathway . The values for these three pathways during the first 30 minutes are shown in Fig 10 , and the results clearly show that the first pathway is the most important after 12 minutes . Therefore , the predominant pathway is 2Gag3 ⇌ Gag6 . We also explored the predominant pathways in pentamer and tetramer formation . As illustrated in Fig 11 , the predominant pathway in pentamer formation is Gag2 + Gag3 ⇌ Gag5 , and as shown in Fig 12 , the predominant pathway in tetramer formation is Gag + Gag3 ⇌ Gag4 . We compared these three most important pathways and found that the trimer intermediate was needed in all three predominant pathways . Therefore , we conclude that the Gag trimer might be a key intermediate in hexamer formation . Kutluay et al . [10] revealed that a Gag trimer could not be formed in the cytoplasm and that a Gag trimer on the PM could not return to the cytoplasm . Therefore , the formation of a trimer indicates that the Gag protein complex can now stay on the PM . As shown in Figs 4 , 5 and 7 , the relative concentrations of trimers , tetramers , pentamers and hexamers are similar to each other . Therefore , we could infer that the concentrations of these high-order polymers depend heavily on the tirmer concentration . In addition , we reduced the trimer concentration to the corresponding value for Gag-G2A by increasing its degradation . The concentrations of the various Gag polymers are listed in Table 6 , and as shown , tetramer , pentamer and hexamer concentrations decrease markedly . However , the same finding was not obtained for reductions in the tetramer and pentamer concentrations . Therefore , we can conclude that trimer formation , namely Gag1 + Gag2 ⇌ Gag3 might be a key pathway . As shown in Fig 10 , the predominant pathway in direct hexmer formation is 2Gag3 ⇌ Gag6 . Taken together , the key pathways for the formation of a hexamer from a monomer might be 2Gag1 ⇌ Gag2 , Gag1 + Gag2 ⇌ Gag3 and 2Gag3 ⇌ Gag6 , and the key intermediates in hexamer foramtion might be the dimer and trimer polymers . In the wake of developments in basic science , many of the most promising HIV drugs in clinical development do not target specific retroviral enzymes but rather act by interrupting the assembly of viral factors with host proteins [59] . For example , some agents that disrupt protein-protein interactions during the entry of HIV-1 are showing great clinical potential [59] . However , according to AIDSinfo , no clinically available drug can inhibit Gag transport and assembly ( https://aidsinfo . nih . gov/drugs/Search/a-z/all ) , and the development of these agents is a daunting challenge . Because the current treatments for HIV-1 normally include the use of multiple drugs in an attempt to control this virus , we proposed and analyzed four theoretical combined methods for inhibiting Gag transport and assembly based on our model . These analyses might be helpful to the design of new anti-AIDS drug . We took three approaches , specifically the degradation rates for Gag polymers , Gag-G2A and Gag-dCTD , into account and designed the following four theoretical combined methods: For these four theoretical combined methods , we computed the concentrations of Gag polymers in the cytoplasm and PM during the first 30 minutes , and the corresponding results are shown in Fig 13 . As shown in the figures , the concentrations of high-order polymers ( e . g . , tetramer , pentamer and hexamer ) are relatively lower with methods C1 , C3 and C4 . We also found that these three methods involved the Gag-G2A mutation . Therefore , we speculate that the N-terminal glycine residue of the MA of Gag might be a promising drug target . Gag-G2A significantly decreases the concentrations of Gag higher-order polymers and thus might be a potential key drug target . To explore the mechanisms responsible for decreasing the concentrations of higher-order polymers of Gag-G2A , we first reduced the trimer concentration of WT Gag to 0 . 37 μm3/ymol by increasing the trimer degradation rate to yield the the corresponding concentration of Gag-G2A trimers . The concentrations of Gag polymers are shown in the fourth line in Table 6 . Simelarly , we decreased the tetramer , pentamer and hexamer concentrations to the corresponding low concentrations for Gag-G2A , respectively , and the results are listed in Table 6 . After reducing the trimer concentration in Table 6 , we compared the tetramer , pentamer and hexamer concentrations with those of WT Gag . The concentrations of these higher-order polymers were all markedly decreased , particularly the hexamer concentration . As shown by the results , the tetramer , pentamer and hexamer concentrations are all very dependent on the trimer formation . However , the same does not hold true for the tetramer and pentamer polymers . Therefore , these results further support the conclusion that the Gag trimer might be a key intermediate and that trimer formation might be a key pathway . In addition , we compared the tetramer , pentamer and hexamer concentrations obtained with Gag-G2A after reducing the trimer concentration listed in the fourth line in Table 6 . The resulting concentrations were markedly higher than those found for Gag-G2A . Thus , Gag-G2A does not use decrease the polymerization coefficients to decrease the polymer concentrations . Compared with Gag-G2A , the monomer and dimer concentrations for WT Gag in the cytoplasm are reduced by approximately a quarter and a half , respectively , and the monomer , dimer , tetramer , pentamer and hexamer concentrations for WT Gag on the PM are reduced by approximately 68% , 88% , 93% , 98% , 99% and 99% , respectively . On the PM , Gag-G2A reduces the monomer and dimer concentrations by reducing their active transport speeds , which results in decreases in the concentrations of higher-order polymers on the PM . In the cytoplasm , the low transport speeds of monomers and dimers for Gag-G2A and their very weak diffusions lead to the high monomer and dimer concentrations , and most of these monomers and dimers are gathered near the perinuclear area . However , due to the high degradation rates , these high concentrations near the perinuclear area show rapid reductions . Therefore , in the entire cytoplasm , these monomer and dimer concentrations are ultimately lower than those found for WT Gag .
In this study , we developed a model to simulate the intracellular trafficking and polymerization of HIV-1 Gag protein . The model parameters were fitted using published experimental data [10] . The profile likelihoods of these parameters were used to show their identifiability , and an elasticity analysis of these parameters was used to show the robustness of this model . The model was able to predict the budding and release time of a VLP , and the results were in agreement with the findings of some previous studies [21 , 51 , 52] . Moreover , the model could also be applied to two mutated versions: Gag-dCTD and Gag-G2A . Using our model , we analyzed the weight of the pathways involved in the polymerization reactions , and concluded that the Gag dimer and trimer might be two key intermediates in hexamer formation . Moreover , we inferred that the three key pathways in the formation of a hexamer from a monomer might be the polymerization of two monomers to form a dimer , the polymerization of a monomer and a dimer to form a trimer , the polymerzation of two trimers to form a hexamer . We also explored four theoretical combined methods for suppressing the Gag concentration and concluded that the N-terminal glycine residue of the MA of Gag might be a proming drug target . There is no denying that the presented modeling approach is merely an approximation to reality . However , it successfully provides a consistent and quantitative description of the transport and polymerization of Gag and lays a broad foundation for further developments . Future experimental and theoretical research is required to support the various assumptions employed in the model . A number of important questions have not been fully addressed and need for further examination . For instance , there are two types of motor proteins: one conveys cargo to the nucleus , and the other conveys cargo to the PM . We consider only the average velocity of the transport of cargo to simplify the model . Thus , it is important to address the transport processes of these two types of motor proteins in future studies . In addition , a dynamical analysis [60] of HIV-1 trafficking process and a multilayer networks analysis [61] related to HIV-1 will also be investigated in future work . | The human immunodeficiency virus ( HIV-1 ) is a retrovirus that causes acquired immunodeficiency syndrome ( AIDS ) , an infectious disease with high annual mortality . Gag protein is the major structural protein of HIV-1 and can self-assemble into the HIV-1 sphere shell . Therefore , an in-depth understanding of Gag protein trafficking and polymerization is important for gaining further insights into the mechanisms of HIV-1 replication and the design of antiviral drugs . Through mathematical modeling , optimization and quantitative analysis , we hypothesized the budding and release time of virus-like particles and revealed that the dimer and trimer intermediates might be crucial in hexamer formation . We also concluded that the predominant pathways in hexamer formation might be the polymerization of two monomers to form a dimer , the polymerization of a dimer and a monomer to form a trimer , and the polymerization of two trimers to form a hexamer . Our analysis also suggested that the N-terminal glycine residue of the MA domain of Gag might be a promising drug target . These results serve as a guide for further theoretical and experimental efforts aiming to understand HIV-1 Gag trafficking and polymerization and could aid anti-AIDS drug design . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"dimers",
"(chemical",
"physics)",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"molecular",
"motors",
"rna",
"viruses",
"materials",
"science"... | 2017 | Mathematical modeling and quantitative analysis of HIV-1 Gag trafficking and polymerization |
Maps of genetic interactions can dissect functional redundancies in cellular networks . Gene expression profiles as high-dimensional molecular readouts of combinatorial perturbations provide a detailed view of genetic interactions , but can be hard to interpret if different gene sets respond in different ways ( called mixed epistasis ) . Here we test the hypothesis that mixed epistasis between a gene pair can be explained by the action of a third gene that modulates the interaction . We have extended the framework of Nested Effects Models ( NEMs ) , a type of graphical model specifically tailored to analyze high-dimensional gene perturbation data , to incorporate logical functions that describe interactions between regulators on downstream genes and proteins . We benchmark our approach in the controlled setting of a simulation study and show high accuracy in inferring the correct model . In an application to data from deletion mutants of kinases and phosphatases in S . cerevisiae we show that epistatic NEMs can point to modulators of genetic interactions . Our approach is implemented in the R-package ‘epiNEM’ available from https://github . com/cbg-ethz/epiNEM and https://bioconductor . org/packages/epiNEM/ .
For two fixed knock-out mutations , we denote by 00 the wild type , by 10 and 01 the two single mutants , and by 11 the double mutant . Their effect on the expression of a given gene i is denoted by Ei , 00 , Ei , 01 , Ei , 10 , and Ei , 11 . Gene expression is reported as the log-fold change relative to the wild type 00 , hence Ei , 00 = 0 . Epistasis between the two mutations is defined as εi=Ei , 00+Ei , 11-Ei , 01-Ei , 10=Ei , 11-Ei , 01-Ei , 10 . Van Wageningen et al . consider this quantity over all effect genes i and define different types of epistasis for the multivariate gene expression phenotype ( E1 , … , Em ) . Complete redundancy is the situation in which , for most genes i , Ei , 01 = Ei , 10 = 0 and hence εi = Ei , 11 . Depending on the sign of Ei , 11 , epistasis may be positive or negative for each individual gene i . Mixed epistasis is defined by Ei , 01 , Ei , 10 ≠ Ei , 11 for some genes and some of those not following redundancy . It is mixed in the sense that the single mutants can have any effect , positive or negative , in any combination ( see Fig 1 ) . In this paper we test the hypothesis that mixed epistasis between a gene pair can be explained by the action of a third gene that mediates between the functional interaction and the transcriptional readout . To test this hypothesis , we extend the framework of Nested Effects Models ( NEMs ) , which has been specifically tailored to analyze high-dimensional gene perturbation data [13] . The extended framework , called Epistatic NEMs ( for short epiNEMs ) , incorporates logical functions that describe interactions between regulators . Our method is general and can be applied to all datasets that measure multi-parametric phenotypes for combinatorial perturbations . We benchmark the accuracy of epiNEMs in the controlled setting of a simulation study . In an application to the data of van Wageningen et al [10] and Sameith et al [11] we show that epiNEMs can point to mediators of genetic interactions . There exist many different pathway reconstruction methods [14 , 15] . Biological databases like BioGrid [16] construct their interaction networks by directly linking genes or proteins with known regulatory relationships , e . g . kinases and their substrates . Data-driven statistical measures like correlation [17] or mutual information [18] can be used to define edges between pairs of genes . Other probabilistic approaches for network inference are based on candidate graphs being evaluated according to the underlying data [14 , 19] . Main representatives of this group are Bayesian and Boolean networks . Boolean networks have a long tradition in biology [20] and were used to model signaling pathways [21] and reconstruct them from perturbations [22] . They model regulatory networks by allowing the nodes/genes to take on one out of two possible values ( yes/no , on/off , expressed/not expressed ) . The choice of value depends on the states of the previous nodes/genes in the network . Boolean variables are dependent on conditional or logical statements and might change according to their input . Those statements are represented by a Boolean function that takes several Boolean variables as input , connects them with logical operators and results in one Boolean output value . In the context of mixed epistasis , van Wageningen et al . [10] used Boolean modeling in order to evaluate all possible combinations of connections between two nodes . This approach is fixed on two regulators with two corresponding gene sets and does not aim at network structure learning . Bayesian networks have been used on multi-parametric readouts of gene perturbations [23 , 24] and are flexible enough to capture complex interactions between regulators [25] . However , they require that most perturbation effects are measured directly at other pathway members , while in our setting the transcriptional effects are all measured downstream of the pathway of interest . This limitation motivated the development of Nested Effects Models ( NEMs ) to indirectly reconstruct signaling networks from observations of downstream genes whose expression levels are affected by perturbations of signaling proteins [26] . The name “Nested Effects Models” derives from the fact that NEMs infer directed relations between signaling proteins by the nested structure of subset relations between their perturbation effects ( See Fig 2A ) . Since their introduction NEMs have been applied and extended in several case studies [27–33] . NEMs have also been extended to model pathway dynamics and re-wiring [34–37] as well as unobserved pathway activation [38] and confounders [39] . The key contribution in this paper is to extend NEMs by introducing logical functions modeling the effects of combinatorial perturbations . The fact that NEMs can easily be extended in this way shows their advantage over subset-based methods that are only defined on pairs of variables [40 , 41] . The idea of incorporating logical functions was already introduced in Boolean NEMs ( B-NEMs ) [42] and is also widely used outside the NEM literature [43] . Our approach differs , however , in several important aspects . B-NEMs aim at learning large signaling pathways and achieve this by incorporating prior knowledge , which excludes full network reconstruction . B-NEMs are generalized to model any Boolean function with an arbitrary number of parents . Thus , without prior knowledge , B-NEMs have to tackle a large search space even for a relatively small number of signaling genes . This can impede the inference and the identifiability of the underlying network , which is modeled as a hyper-graph [42] . epiNEMs on the other hand are a straightforward extension of NEMs and model the pathway as a normal graph . If a signaling gene has two parents the incoming edges are annotated with one of five different logical functions . This aspect makes epiNEMs much more practical for handling the special case of epistasis , especially for large knock-out screens , where we test a multitude of single knock-outs ( modulators ) for several double knock-outs .
In total there are five logic gates that represent different biological relationships ( see Fig 2C ) . The AND gate accounts for functional overlap of two genes . The pathway can compensate the loss or knock-down of one gene and only if both parents are off at the same time , the signal flow will be cut off . NOT-A and NOT-B stand for masking or inhibiting effects . The XOR gate can be interpreted as both parent genes preventing each other from acting on the third gene . The OR gate is identical to how two interactors are treated in the classic NEM approach: no interaction . All other theoretically possible logical combinations can be expressed in a simpler graph structure and are therefore disregarded ( see Fig 2D ) . Adding logics extends the S-gene graph into a Boolean network . In general , Boolean networks are dynamical systems , which can exhibit different attractors and steady states [44] . Our implementation covers this general case and uses the R package ‘BoolNet’ [44] to compute attractors and steady states for each single and each double knock-out in a synchronous manner . However , the assumptions we can make for the specific application of identifying modulators of genetic interactions guarantee a single steady state per network . First of all , we assume that effects on signaling genes due to direct or upstream perturbations are irreversible , which prevents feedback loops . Secondly , in our screens for modulators we only evaluate acyclic networks of three genes . Thus , each set of perturbations corresponds to a unique pattern of activation states of pathway genes and we can summarize the expected effects on pathway genes in a row-vector ϕ . Concatenation of these vectors for all perturbations yields a design matrix Φ , in which the rows indicate expected effects for each perturbation . Given the states of all signaling genes Si , we calculate the likelihood of each model hypothesis in the same way as in standard NEMs [26] . Let us first assume that the complete model , i . e . the signaling graph Φ and the effect attachments Θ = {θ1 , … , θm} , is given . With these parameters , the expected effects can be compared to the observed effects to obtain the likelihood P ( D | Φ , Θ ) = ∏ i = 1 m ∏ k = 1 l P ( e i k | Φ , θ i ) , where m denotes the number of effects and l stands for the number of replicate experiments . For the effects , we have eik = 1 if we observe an effect and eik = 0 if we do not observe any effect . Experimental data , however , will always be noisy and therefore the probability P ( eik|Φ , θi ) will be dependent on the false positive rate α and false negative rate β of the experiment . In almost all applications , however , it is not known which effect is directly linked to which signaling gene . Therefore , the marginal likelihood for each silencing scheme is computed by averaging over the effects attachments Θ . This is achieved by summing over all attachment probabilities: P ( D | Φ ) = 1 n m ∏ i = 1 m ∑ j = 1 n ∏ k = 1 l P ( e i k | Φ , θ i = j ) , where n denotes the number of signaling genes Si . The optimal pathway is the one resulting in the highest likelihood . For small networks like the ones we use here , exhaustive search over all network topologies is possible . For faster inference or feasibility for networks with more than five genes , a greedy hill climbing method is provided in the package ‘epiNEM’ . Interpretation of the network inferred from data is not always straightforward . As in the original NEM approach we have to consider the degree of identifiability of the network . Two networks belong to the same equivalence class if they have the same likelihood given the data . In the case of the original NEMs two networks are equivalent , if they have the same transitive closure . Due to our extension of the method , additional equivalences between network hypotheses occur in the case of epiNEMs . Let Φ be a network with two parents regulating their child by one of epiNEMs‘ five logics or one of the three other types of relations ( Fig 2C and 2D ) . If the parents are independent of each other , all eight networks result in different effect matrices and subsequently different data . However , if the parents are not independent , i . e . , one parent is regulating the other parent , a knock-out of the upstream parent is equivalent to the double knock-out of both parents . Thus , networks which are only distinguishable by the effects of the single knock-out of the upstream parent become equivalent and produce the same data . Another major challenge in pathway inference methods are hidden players [39] . In the case of NEM , if two parents have a hidden common child , the data shows all possible pairwise effects , i . e . , effect reporters which react exclusively to one parent’s knock-out and effect reporters which react to both knock-outs . EpiNEMs are designed to use large knock-out screens to identify those hidden signaling genes as modulators of the signal and explanation of the corresponding data .
In a simulation study , we compare epiNEM results to networks reconstructed by NEM without logics as well as B-NEM , ARACNE [18] ( a method based on mutual information ) , and the PC algorithm [45] ( a method based on partial correlation ) . We generated data sets of 4-node networks with 100 effects , Ei , being randomly attached to the 4 signaling genes , Si . In each network , two of the four signaling genes were randomly connected by one of the five possible logic gates . These networks were translated into adjacency matrices with knock-outs in the rows and observed signal disruptions in the columns . For every Ei we check the behavior upon perturbation from the adjacency matrix . We kept the false positive rate α at 0 . 1 and varied the false negative rates β over a wide range of values: β ∈ {0 . 01 , 0 . 025 , 0 . 05 , 0 . 1 , 0 . 2 , 0 . 3 , 0 . 4 , 0 . 5} . In total , we generated data from 100 random networks , for each false negative rate . We compared the five competing methods by running time and accuracy of the predicted edges . In the case of the PC algorithm and ARACNE , we did not consider the edge direction , because they only infer partially directed and undirected networks , respectively . For B-NEM and epiNEM , we additionally scored the accuracy of the inferred logical gates and their expected data generated by the inferred network , which is similar to the truth table of a Boolean network . ARACNE , the PC algorithm , and NEMs are by far the fastest methods . However , they do not infer any logical gates , and the first two report no or only partial edge directions , respectively . B-NEM is almost a magnitude slower than epiNEM . Additionally , epiNEM achieves the highest accuracy for the inferred edges , closely followed by B-NEM and with some distance the other methods . Due to B-NEM’s larger search space , it cannot identify the correct epistatic signaling , even though the accuracy for the expected data is high . EpiNEM on the other hand , while only achieving little higher accuracy for the expected data , has median accuracy of 100% for the logic gates and false negative rates up to 20% ( see Fig 3 ) . We applied epiNEMs to the studies of yeast knock-out screens of van Wageningen et al . [10] and Sameith et al . [11] . Both data sets consist of measurements of gene expression changes from double and single gene knock-out experiments in S . cerevisiae . Our goal is to identify signal modulators that help explaining the mixed epistasis patterns observed under single and double knock-outs of signaling genes . Van Wageningen et al . identified three buffering relationships: quantitative redundancy , complete redundancy , and mixed epistasis [10] . The last case is the most prevalent and defined by two genes interacting in different epistatic ways for different downstream gene sets . Mixed epistasis suggests that genes may only partially overlap in function or be influenced by an additional regulatory module that controls different processes according to condition and environment . In this paper , we have developed a method to address a central question of molecular cell biology: how to characterise the mechanisms underlying the functional redundancies visible in genetic interactions . We hypothesized that mixed epistatic effects found in high-dimensional readouts can be explained by the action of a third gene that mediates between the genetic interaction and the transcriptional response . To explore this hypothesis we extended Nested Effects Models , an established methodology to infer signaling pathways , with logical functions . The resulting method , called epiNEMs , is a general approach to infer pathways including combinatorial regulation from perturbation effects . In particular , it allowed us to screen for modulators of genetic interactions in S . cerevisiae . We were able to identify such modulators and to computationally reproduce the experimental results from van Wageningen et al . in most cases . In a second data set from Sameith et al . consisting of roughly five times more double knock-outs , we calculated the global distribution of epistatic signaling logics of all modulators . Most of them are identified as masking or complete redundancy . Additionally , we thoroughly investigated a previously by Sameith et al . , 2015 identified triplet of growth inducing and repressing factors gln3 , gat1 and gzf3 and found evidence for a more complex signaling network . Furthermore , we globally visualized our findings by gene ontology enrichment analysis ( KEGG pathways ) to support the validity of epiNEMs . Our approach has several limitations . First , extending NEMs with logics increases the size of the model space and makes exhaustive enumeration unfeasible . Second , we only consider logics between pairs of regulators , which helps to limit model space and is very well suited for our application to genetic interactions , but might be an oversimplification in other applications . In the future , the model could therefore be improved by allowing logic gates for more than two parents . This will result in more complex logics but will also allow for capturing more interactions . Also , until now it is only possible to distinguish between complete redundancy and mixed epistasis , while quantitative redundancy cannot be captured . To improve this situation , we plan to extend the model to use quantitative effects rather than binary data . In summary , we presented a general framework to understand mediators of complex phenotypes of genetic interactions . Our case studies on transcriptional phenotypes in yeast showed very promising results and there are potentially many other applications in other organisms using either combinatorial RNAi [47] or pooled CRISPR screens [48] together with multi-parametric phenotyping [49] or single-cell RNA-seq [50–53] .
All our analyses were done in the statistical computing environment R [54] . Our approach is implemented in the R-package ‘epiNEM’ available from https://github . com/cbg-ethz/epiNEM and https://bioconductor . org/packages/epiNEM/ [55] . We used 160 microarray gene expression profiles of single and double mutants from [10] available at ArrayExpress under the IDs E-TABM-907 ( mutants ) and E-TABM-773 ( 200 wild-type replicates ) . Additional we used 154 profiles from [11] with respective ID E-MTAB-1385 . For our analyses , we directly downloaded the flat files from the following locations: http://www . holstegelab . nl/publications/sv/signaling_redundancy/ http://www . holstegelab . nl/publications/GSTF_geneticinteractions/ . All analysis steps including data preprocessing are documented in the vignette of the R-package ‘epiNEM’ . | Genes do not act in isolation , but rather in tight interaction networks . Maps of genetic interactions between pairs of genes are a powerful way to dissect these relationships . Genetic interactions are mostly defined by quantifying individual phenotypes like growth or survival . However , when high-dimensional phenotypes are observed , genetic interactions can become very hard to interpret . Here we test the hypothesis that complex relationships between a gene pair can be explained by the action of a third gene that modulates the interaction . Our approach to test this hypothesis builds on Nested Effects Models ( NEMs ) , a probabilistic model tailored to inferring networks from gene perturbation data . We have extended NEMs with logical functions to model gene interactions and show in simulations and case studies that our approach can successfully infer modulators of genetic interactions and thus lead to a better understanding of an important feature of cellular organisation . | [
"Abstract",
"Introduction",
"Model",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"genetic",
"networks",
"engineering",
"and",
"technology",
"signaling",
"networks",
"logic",
"circuits",
"epistasis",
"regulator",
"genes",
"fungi",
"model",
"organisms",
"network",
"analysis",
"experimental",
"organism",
"systems",
"gene",
"types",
"genetic",
"interact... | 2017 | Inferring modulators of genetic interactions with epistatic nested effects models |
Cancer is considered an outcome of decades-long clonal evolution fueled by acquisition of somatic genomic abnormalities ( SGAs ) . Non-steroidal anti-inflammatory drugs ( NSAIDs ) have been shown to reduce cancer risk , including risk of progression from Barrett's esophagus ( BE ) to esophageal adenocarcinoma ( EA ) . However , the cancer chemopreventive mechanisms of NSAIDs are not fully understood . We hypothesized that NSAIDs modulate clonal evolution by reducing SGA acquisition rate . We evaluated thirteen individuals with BE . Eleven had not used NSAIDs for 6 . 2±3 . 5 ( mean±standard deviation ) years and then began using NSAIDs for 5 . 6±2 . 7 years , whereas two had used NSAIDs for 3 . 3±1 . 4 years and then discontinued use for 7 . 9±0 . 7 years . 161 BE biopsies , collected at 5–8 time points over 6 . 4–19 years , were analyzed using 1Million-SNP arrays to detect SGAs . Even in the earliest biopsies there were many SGAs ( 284±246 in 10/13 and 1442±560 in 3/13 individuals ) and in most individuals the number of SGAs changed little over time , with both increases and decreases in SGAs detected . The estimated SGA rate was 7 . 8 per genome per year ( 95% support interval [SI] , 7 . 1–8 . 6 ) off-NSAIDs and 0 . 6 ( 95% SI 0 . 3–1 . 5 ) on-NSAIDs . Twelve individuals did not progress to EA . In ten we detected 279±86 SGAs affecting 53±30 Mb of the genome per biopsy per time point and in two we detected 1 , 463±375 SGAs affecting 180±100 Mb . In one individual who progressed to EA we detected a clone having 2 , 291±78 SGAs affecting 588±18 Mb of the genome at three time points in the last three of 11 . 4 years of follow-up . NSAIDs were associated with reduced rate of acquisition of SGAs in eleven of thirteen individuals . Barrett's cells maintained relative equilibrium level of SGAs over time with occasional punctuations by expansion of clones having massive amount of SGAs .
Clonal evolution is a theory that explains the phenomenon of the progressive morphological and genetic change of somatic cell populations from normal homeostatic cell division and death within tissues to abnormal neoplastic growth and cancerous spatial expansion within and across tissues [1]–[3] . Clonal evolution is the Darwinian evolution by natural selection of asexually ( mitotically ) dividing somatic cells . Somatic genomic abnormalities ( SGA ) , such as copy number alterations and loss of heterozygosity ( LOH ) , can be used as polymorphic DNA markers for identifying evolving clones . Strictly defined , a clone is a genetically identical subpopulation of cells within the cell population of a tissue , that descends from a most recent common ancestor ( MRCA ) cell and therefore all of the clone's cells inherit the SGAs that were originally present in the MRCA cell . However , a commonly used , relaxed definition of a clone is descent with modification from a MRCA cell , which allows for accumulation of additional SGA heterogeneity among the cells of the clone . A clone ideally represents the shared cell lineage history of a subpopulation of cells . The acquisition of SGA variability ( SGA polymorphism ) over the course of cell division allows for classification of cell subpopulations into clones . In the remainder of this study , we use clone in its relaxed definition and we estimate phylogenetic trees from acquired SGA variability to qualitatively describe relatedness among evolving clones . In other words , we call a clone a set of biopsies that share a large number of SGA features by descent , or from a phylogenetic tree point of view , a set of tips ( taxa ) of the phylogenetic tree , which are more related than others ( i . e . cluster together as a clade ) , but which are not necessarily identical ( identical tips will have interconnecting branches with zero lengths ) . The generation of new clones is stochastic and the change in clones' frequencies in the population is determined by clones' relative fitness as well as stochastic effects ( genetic drift ) . New adaptive and new neutral clones can arise stochastically over time [4] with every newly acquired SGA that does or does not affect fitness , respectively . Though adaptive mutations are thought to drive clonal expansions , it is an open question if adaptive clones tend to expand to fill much or all of the BE segment , or if they tend to remain relatively localized [5] . In order to prevent progression to cancer , mechanisms that modulate clonal evolution by either preventing or managing SGA acquisition and/or the spread of SGA-containing clones need to be elucidated . Barrett's esophagus ( BE ) is a condition of the distal esophagus in which the normal stratified squamous epithelium is replaced by columnar epithelium with intestinal metaplasia [6] . BE is thought to develop as a complication of chronic gastroesophageal reflux disease ( GERD ) and individuals with BE are at increased risk of progression to esophageal adenocarcinoma ( EA ) : 1–7 persons with BE progress to EA per 1000 person-years [7] , [8] . Strategies for early detection and prevention of esophageal adenocarcinoma have focused on all aspects of the GERD-BE-EA sequence: acid suppression medications , anti-reflux surgery , esophagectomy , ablation of BE , endoscopic biopsy surveillance of BE , and chemoprevention using aspirin or other non-steroidal anti-inflammatory drugs ( NSAIDs ) [6] , [9] . BE is a condition in which clonal evolution can be studied in vivo , since a standard of care is periodic endoscopic surveillance , allowing studies of clonal evolutionary dynamics over time . Genomic instability is a common feature of solid cancers [1] , [10]–[13] . In a recent study , Beroukhim et al . evaluated 3131 cancer specimens from 26 histologic types and 1480 normal tissue specimens and found that copy number gains and losses affected 17% and 16% of the genome in a typical cancer specimen and only 0 . 35% and 0 . 1% of the genome in a typical normal tissue specimen [14] . Despite the recent massive accumulation of data on genomic alterations in cancers from the Cancer Genome Atlas and the International Cancer Genome Consortium initiatives , as well as phylogenetic reconstruction of lineages within tumors [15]–[18] , theoretical modeling of the generative process ( clonal evolution ) producing the observed SGA patterns and underlying neoplastic progression has remained limited [2] , [3] , [15] , [19] . BE is associated with genomic instability and acquired SGA [20]–[23] allowing analysis of the acquisition of SGA over time . This provides data for estimating SGA acquisition rate that is a key parameter of clonal evolution . NSAID use significantly reduces the incidence and mortality rates of many types of cancer , including esophageal adenocarcinoma [24]–[29] . Rothwell et al . showed that the hazard ratio for cancer incidence of NSAID users vs . NSAID non-users was 0 . 66 ( 95% CI 0 . 50–0 . 87 ) ; however a robust NSAID cancer preventive effect manifests significantly only after ≥5 years of regular use [24] . The majority of epidemiological studies in BE suggest that NSAID use in individuals with BE reduces risk of developing EA [25]–[28] . Specifically , Vaughan et al . evaluated 350 individuals followed up for a median of 5 . 4 years ( range 0 . 2–8 . 9 ) and showed that the 5-year cumulative incidence of EA was 14 . 3% ( 95% CI 9 . 3–21 . 6 ) for NSAID never users compared to 6 . 6% ( 3 . 1–13 . 6 ) for current NSAID users and that the hazard ratio for EA incidence of NSAID users vs . NSAID non-users was 0 . 20 ( 95% CI 0 . 10–0 . 41 ) [27] . Galipeau et al . showed that NSAID use reduced the 10-year cumulative incidence of esophageal adenocarcinoma from 79% to 30% in individuals with BE who had one or more somatic genomic abnormalities detected at baseline endoscopy , which included DNA content tetraploidy and/or aneuploidy , assayed by DNA content flow cytometry , or genetic abnormalities , such as loss of heterozygosity ( LOH ) on chromosomes 9p and 17p , assayed by PCR of small tandem repeat ( STR ) loci [29] . NSAID use for chemoprevention is attractive due to the widespread use and low toxicity and side effects of that class of drugs; however the molecular mechanisms underlying the NSAID cancer preventive effect are not fully understood . In this study , our aim was to evaluate the effect of NSAIDs on the accumulation of somatic genomic abnormalities by evaluating the entire genome ( 1 Million SNP loci ) for SGA . We hypothesized that NSAID use modulates clonal evolution by reducing the prevalence of SGA by either reducing the incidence of SGA over time ( SGA rate: number of SGAs acquired per genome per year ) or interfering with the expansion of lineages bearing newly acquired SGAs over time . To test this hypothesis , we used a prospective observational crossover study design: a longitudinal study in which the sequence of NSAID use was recorded for each individual during the follow-up period ( Figure 1A ) . We selected thirteen individuals with BE from our cohort , who had endoscopic follow-up of mean 11 . 8±3 years ( range: 6 . 4–19 ) and who began or discontinued NSAID use exactly once during follow-up . All thirteen individuals had to have at least two consecutive time points ( ≥6 biopsies ) off NSAIDs and at least two consecutive time points ( additional ≥6 biopsies ) on NSAIDs ( time and locations of all the biopsies are shown in Figure 1B ) . To estimate SGA prevalence in biopsies on and off NSAIDs we used summary statistics of observed patterns of SGA; to estimate SGA rates on and off NSAIDs we used an evolutionary analysis of observed SGA patterns to take into account SGA phylogenetic identity by descent . Drummond et al . showed that mutation rates can be estimated from longitudinal samples in virus populations using coalescent and phylogenetic methods within a Bayesian Markov Chain Monte Carlo framework for sampling model parameter space ( BEAST package , Bayesian Evolutionary Analysis Sampling Trees ) [30] , [31] . This takes into account the fact that samples at later time points are neither direct descendants of samples from earlier time points , nor independent samples , but rather share common ancestors in a phylogeny that represents their evolutionary relationships . BEAST simultaneously estimates the mutation rates , population sizes , and the forest of most likely phylogenetic trees within an individual's Barrett's segment . We adapted BEAST to separately estimate SGA acquisition rates on and off NSAIDs . Thus , the crossover study design provided 13 independent tests of the hypothesis of NSAID-associated reduction in SGA acquisition rate since every individual had both on and off NSAID periods and SGA acquisition during those periods .
We evaluated the dynamics of detected SGAs over time . The mean number of SGAs and the proportion of the genome they affected did not obviously increase over time , for as many as 19 years ( e . g . , Figure 2 , individual a ) . Individuals b , f , and j , shown in red in Figure 2A and 2B , showed much greater variation in detected SGA per biopsy , per time point , compared to the rest of the individuals , shown in black . Progression to EA was not part of our study inclusion criteria , and individual j was the only individual who progressed . Individual f did not progress to EA , but rather opted for esophagectomy for high-grade dysplasia after 6 . 4 years of follow-up and subsequently died of a different cancer 11 . 9 years later . In individuals b , f , and j , the mean ( ± standard deviation ) number of SGAs per biopsy per time point was 1 , 082±177 , 1 , 844±573 , and 1 , 154±746 , and the amount of genome affected by SGAs was 119±79 Mb , 242±121 Mb , and 227±222 Mb , respectively . In the rest of the individuals , the mean number of SGAs per genome per time point was 279±86 and the amount of genome affected was 53±30 Mb . Assuming a human genome length of 3 , 164 Mb ( Human genome GCRh37 . p5 assembly ) , individuals b , f , and j had 3 . 8±2 . 5% , 7 . 6±3 . 8% , and 7 . 2±7% altered somatic genome per time point , compared to 1 . 7±0 . 9% altered somatic genome in the rest of the individuals . The number of events and total sizes for each type of lesion , as well as the detected presence of within-biopsy heterogeneity , in each biopsy are shown in Table S1 and Figures S1 , S2 . In 10 out of 13 individuals ( everyone except b , f , and j ) the number of SGAs remained relatively constant over time . Different biopsies from these individuals displayed different SGA lesions , leading to upward and downward fluctuations in mean number or genome amount of SGA . In some cases a biopsy from an earlier time point had more genomic lesions than a biopsy at a later time point , suggesting that we sampled a persistent but more ancestral clone at the later time point . For example , the first biopsy in individuals i and l had the highest number of SGAs compared to biopsies at later time points ( Figure S1 and Table S1 ) . Overall , the dynamics of SGAs in BE segments appear more consistent with equilibrium over time rather than with accumulation of SGAs affecting ever greater portions of the genome over time . All biopsies sampled from an individual are related due to common ancestry arising from , or prior to , the time of origination of the Barrett's segment through the process of cell and crypt division . Therefore , we expect biopsies to share some SGAs and we accounted for the statistical non-independence of observed SGA across biopsies , due to common ancestry , by estimating maximum parsimony phylogenetic trees for each individual using SGAs as characters and biopsies as taxa . We estimated phylogenetic trees using PAUP and found that for the majority of individuals the maximum number of SGA events per lineage occurred in branches when the individual was off NSAIDs as opposed to in branches during periods when they were taking NSAIDs ( Figure 3 ) . For example , PAUP analysis showed massive SGA on a single off-NSAID lineage in individual j , which resulted in subsequent SGA-heavy descendant on-NSAID lineages ( phylogeny of individual j , Figure 4H , I ) . While the natural history of clonal evolution is different in each individual , some common patterns can be discerned . The majority of SGAs are present in the first time point , with little accumulation of SGAs afterwards ( Figure S3 ) . In the earliest biopsies , taken at baseline endoscopy , there were many SGAs ( 1442±560 in individuals b , f , and j and 284±246 in the rest ) . This can also be seen in the radial spokes apparent in the Circos plots ( Figures 4B , 4C , 5B , 5C , and S4 ) , and also in the lesions that were detected in all biopsies of an individual ( some of which are the most common lesions in BE [21] , [23] , on chromosomes 9p ( CDKN2A ) and 3p ( FHIT ) ( Figure S5 ) . Multiple clones appear to co-exist over the entire period of follow-up . This can be seen in the strong spatial divergence between biopsies at the same time point in individuals a , f . g and I ( Figure S6 ) . However , within a given level ( ±1 cm ) , there was no significant increase in genetic divergence with time . While it is often assumed that the evolutionary history of a cancer involves multiple selective sweeps by new , selectively advantageous genotypes , we found only one such case of a clone that grew to stably dominate the Barrett's segment ( individual h , Figures S8 and S9 ) over at total of 153 patient-years . Due to sampling limitations , we cannot be sure that clone drove all other clones extinct ( went to fixation ) . Genetic divergence , based on number of SGA events , only significantly increased during follow-up in individuals b and j ( Figure S6 ) ( individuals d and f also showed increasing divergence based on amount of SGA; Figure S7 ) but decreased in the one individual ( h ) with a large clonal expansion ( Figure S6 ) . The absence of selective sweeps can also be seen in the consensus trees generated by BEAST ( Figures 4D , 4G , 5D , 5G , and S8 ) . Even in the one individual who progressed to EA , individual j , the clone with massive SGAs remained spatially localized ( Figure 4 ) . This clone is defined as the set of biopsies 8 , 10 , 11 , and 13 , which had a combined 2 , 291±78 SGAs affecting 588±18 Mb or 19% of the genome , and which were sampled at levels 41 , 39 , 41 , and 40 cm in a segment that spanned levels 35–44 cm from SCJ to GEJ ( Figure 4 C , G–I ) . Interestingly , a precursor of that clone had been detected nine years prior to its emergence ( biopsy 2 in Figures 4 G–I ) . We show clonal evolution in individuals b , j , and f in higher detail in Figures 4 and 5 since these individuals had a higher than average number of SGA events and amount of genome affected by SGA ( Figure 2 ) . We show clonal evolution in individual l ( Figure 5 ) in higher detail to show clonal evolution during an on-off NSAID use pattern . In addition , the SGA amount in individual l is close to the mean SGA amount in all individuals , except b , f , and j , while SGA amount in individual f is higher than the mean ( Figure 2 ) and using Circos plots side-by-side contrasts qualitatively SGA amount and SGA chromosomal location between individuals l and f ( Figure 5 B , C ) . In summary , the majority of individuals showed no dramatic accumulation of new SGAs consistent with long-term evolutionary stasis during follow-up ( Circos plots in Figures 4 , 5 , S1 , S2 ) , and the one progressor to EA , individual j , showed that evolutionary stasis can be punctuated by the expansion of a clone with massive amount of SGAs ( Figure 4 C , G–I ) . The maximum parsimony phylogenetic analysis revealed the shared common ancestry of biopsies within an individual based on SGA homology . Inferred PAUP phylogenetic trees , which had branches scaled by the estimated number of shared SGA events in Figures 4E , H , 5E , H , and Figure S9 , showed significantly imbalanced tree shapes ( Table S2 ) for all individuals , except individual f and j . When we rescaled the branch lengths of the same phylogenetic trees by the amount of genome affected in Figures 4F , I , 5F , I , and Figure S10 , the trees showed that within an individual the majority of biopsies are closely related and only few biopsies or lineages diverge dramatically from the majority cluster , which is indicative of SGA bursts . Phylogenetic methods of analysis are required to account for the complex dependency structures in samples that are related by common ancestry . Bayesian methods allow for more detailed models of evolution than parsimony methods . We tested our hypothesis that NSAID use reduces SGA acquisition rate in BE by estimating SGA acquisition rate during off-NSAID and on-NSAID periods using a custom modified version of BEAST [31] . This reconstructs the set of most likely phylogenies that relate the samples within the Barrett's segment and simultaneously estimates the SGA rates along the branches of those phylogenies during the on and off NSAIDs periods . This method is based on the coalescent in which lineages may disappear either because they are not sampled or because they go extinct . We added a new evolutionary model of SGA into BEAST in order to estimate SGA rate using Bayesian MCMC sampling ( see Methods and Text S1: Equations S3–4 ) . We excluded SGAs detected in the first time point and only measured SGAs that were detected during follow-up , in order to reduce the influence of clonal evolution that occurred prior to surveillance . For the two individuals who were already on NSAIDs when we started surveying them , and later went off NSAIDs , individual l showed a lower SGA rate on NSAIDs than off NSAIDs , but the 95% support intervals for the two rates overlap ( Figure 6 ) . In contrast , individual m showed a higher SGA rate on NSAIDs than off NSAIDs . In individuals a–k , the SGA rate on NSAIDs was approximately an order of magnitude lower than the SGA rate off NSAIDs , with non-overlapping 95% support intervals ( Figure 6 ) , which is consistent with the hypothesis that NSAID use reduces SGA acquisition rate ( on average 7 . 8 SGAs per genome per year off-NSAID vs . on average 0 . 6 SGAs per genome per year on-NSAID in individuals a–k ) .
While it is clear that NSAIDs prevent many forms of cancer [24] , particularly esophageal adenocarcinoma [25]–[27] , the mechanism of that preventive effect is unknown . Because neoplastic progression is a process of somatic evolution , NSAIDs must affect somatic evolution in order to prevent cancer . We hypothesized that NSAIDs slow the rate of somatic evolution by lowering the mutation rate , specifically , the rate of acquisition of copy number alterations and loss of heterozygosity ( SGAs ) . By adapting a tool from evolutionary biology ( BEAST ) we were able to estimate the SGA rates in vivo both on and off NSAIDs within the same individuals . Our data shows that overall NSAID use is associated with an approximately 10-fold reduction in the rate of acquisition of SGAs and expansion of those lineages to detectable levels , from 7 . 8 SGAs per genome per year ( 95% support interval [SI] , 7 . 1–8 . 6 ) to 0 . 6 SGAs per genome per year ( Figure 6 ) . However , this was only clear in 11 of our 13 individuals ( with non-overlapping 95% support intervals in 8 of those 11 ) . In the two individuals who stopped NSAID use during follow-up , the data does not show a significant reduction in SGA rate by NSAIDs . This may be due to the fact that these individuals were originally off NSAIDs for some unknown period of time prior to surveillance , and the lesions acquired during that time are being lumped into the on-NSAIDs SGA rate estimation . It is likely that NSAID use will not affect all individuals in the same way , and that their effects may be modulated by other factors that vary across individuals . A future larger cohort study will be required to determine if this effect generalizes to most individuals with BE . The BEAST phylogenetic analysis assumes two constant SGA rates , one on-NSAIDs and one off-NSAIDs . To relax this assumption , we added a maximum parsimony analysis ( PAUP ) of the biopsy SGA data that allows for differences in SGA load on estimated lineages . In 10/13 individuals we observed the maximum SGA load occurring on an off-NSAID lineage ( Figure 3 ) . In individual j , we observed a single lineage arise during the period off NSAIDs that carried a massive SGA load , and spawned descendant lineages also with heavy SGA loads ( individual j in Figure 3 and Figure 4H , I ) . This suggests that NSAIDs may prevent the occurrence of massive numbers of SGA on single lineages ( branches of the phylogeny ) or limit the clonal expansion of such lineages . This intensive longitudinal study with 5–8 time points per individual , with 10–20 biopsies per individual , over 6 . 4–19 years of follow-up , revealed a number of additional surprises . There was long-term evolutionary stasis in most individuals ( Figure 2 ) , only one large clonal expansion ( in individual h; Figures S8 , S9 ) , and the sudden appearance of a massively altered clone in individual j ( though a precursor of that clone was detected in biopsy 2 , nine years earlier; Figure 4 ) . The term evolutionary stasis has been used in evolutionary biology to describe a lack of phenotypic change in a species over a given timeframe [32] . Models of this stasis are based on an evolving population diffusing across a fitness plateau due to evolutionarily neutral mutations , until they find an adaptive genotype [33]–[35] . This definition of stasis predated the advent of technologies that allowed characterization of genomic alterations , large and small , that can occur in a lineage that does not lead to measurable phenotypic changes . We are using the term stasis in this study to describe genomes that maintain in equilibrium a relatively constant level of genomic alterations over time and space , in contrast to those individuals in which large scale genomic alterations develop . Our results suggest that while clones may come and go , along with the SGAs they carry , the overall population of Barrett's cells maintains an equilibrium level of SGA events , that only grows slowly if at all ( Figures 2 , 6 ) . The only exception was the one individual ( j ) who progressed to cancer , in whom we detected large-scale SGAs just before they started using NSAIDs . The observation of long-term evolutionary stasis is consistent with the observation that BE rarely progresses to EA [7] , [8] and the hypothesis that BE can function as a benign and perhaps protective evolutionary adaptation of epithelial tissue to duodenal gastroesophageal reflux [36] . Our selection criteria , both for individuals that have used NSAIDs , and for at least 4 time points over at least 4 . 5 years of follow-up may have also led to selection bias for individuals with evolutionary stasis in their Barrett's epithelium . Apparent evolutionary stasis at the level of analyses of biopsies may miss ongoing accumulation of SGAs within single crypts . If those clones never grow larger than a few crypts , they would not be detected by our assays . Further work will be necessary to determine if the stasis seen at the biopsy level is a result of the lack of accumulation of SGAs in crypts or the lack of clonal expansions of those SGAs to detectable sizes . In addition , SNP arrays do not reveal point mutations , small indels , and some structural rearrangements that would be revealed by genomic sequencing . Additional analyses will be required to test if there is significant accumulation of these other genomic alterations during progression . The dominant model of neoplastic progression , with sequential selective sweeps , is not supported by our data . There are enough novel lesions in each biopsy that if a clonal expansion was driven by a point mutation , epigenetic change , or other structural alteration not assayed by a SNP array , we would still be able to detect the expansion in the hitchhiker LOH and copy number alterations . In only one case ( individual h ) did we observe a clone taking over the entire Barrett's segment during follow-up , and in that individual it only happened once . Rather , we observed that after the initial expansion of the Barrett's epithelium , there is long-term coexistence of Barrett's cell lineages ( Figures S8 , S9 , S10 ) . In fact , there is little genetic divergence within the same level of the BE segment over time , though biopsies at different levels tend to be genetically divergent ( Figures S6 , S7 ) . We used a relatively simple model of the likelihood of SGA events in BEAST . Future work should improve on this with better models of genomic lesions as well as the inclusion of natural selection in the inferred dynamics . In addition , assaying single cells [37] , [38] , or single crypts , would avoid potential confusion generated by mixed clonal populations within a sample . In summary , NSAID use in BE is associated with approximately an order of magnitude reduction in the rate of acquisition of SGA in 11 of the 13 individuals , suggesting that the pathway whereby NSAIDs exert their protective effect involves the reduction in number of SGAs or the inhibition of spread of SGA-containing clones . Our results also suggest that most genetic lesions occur prior to baseline detection in the clinic , but that during clinical management the Barrett's cells remain in equilibrium at the genome level . Measurement of mutation rates ( i . e . SGA rates ) in vivo might be used in the clinic to reduce overdiagnosis and unwarranted treatment and detection of high mutation rates or massive bursts of SGA might be used to better identify patients needing more aggressive surveillance and therapy .
Individuals were selected from the Seattle Barrett's Esophagus Study , a research cohort founded in 1983 . Surveillance endoscopies were performed and biopsies were taken using a standardized four quadrant sampling protocol [39] . At endoscopy , anatomical landmarks including the gastroesophageal junction ( GEJ ) and squamocolumnar junction ( SCJ ) were noted , which define the lower ( distal ) and upper ( proximal ) boundaries , respectively , of the Barrett's segment . During an endoscopy , biopsies were taken every one or two cm along the length of the Barrett's segment . At each level , four biopsies were taken approximately at 0° , 90° , 180° , and 270° around the circumference of the esophagus for histologic evaluation . Endoscopic biopsies for molecular studies were collected in Minimal Essential Media ( MEM ) with 10% DMSO ( Sigma #D-5879 ) , 5% heat inactivated fetal calf serum , 5 mM Hepes buffer on ice and frozen at −70°C . In 1995 the research protocol added an epidemiologic interview in which individuals were questioned about NSAID use , as previously described [27] . In addition , the protocol added blood collection at the time of endoscopy for use as a control , since blood DNA represents putatively unaltered germline genotype . Individuals were selected in the cohort who had at least a 3 cm-long BE segment at baseline . Individuals were further selected based on NSAID use status changing exactly once during prospective follow-up and based on having at least two endoscopic procedures while using NSAIDs and at least two while not using NSAIDs . At least five years of follow-up was also required in order to observe evolution over time . Thirteen individuals met these inclusion criteria ( Figure 1A ) . The history of NSAID use at each endoscopy was evaluated with a questionnaire that was also used in a US collaborative case-control study of esophageal adenocarcinoma [40] . As part of the questionnaire , individuals are shown cards ( i . e . , typed lists of drugs with trade names and generic names ) to facilitate recall . Individuals were also asked about indications for taking NSAIDs , and reasons for stopping in those who were no longer regular users . The criterion for regular NSAID use at an endoscopy was taking an NSAID at least once per week for the last 6 months . Regular NSAID use over multiple endoscopies defines a time interval on-NSAIDs and absence of NSAID use over multiple endoscopies defines a time interval off-NSAIDs . We approximated the transition point between NSAID use and non-use by taking the middle time point equidistant between the two endoscopies when the NSAID use changed ( Figure 1A , white-gray boundary ) . Eleven individuals ( a–k ) were not on NSAIDs at the start of surveillance and then went on NSAIDs ( had an “off–on NSAIDs” pattern during surveillance ) , and two individuals ( l , m ) had the opposite , starting surveillance on NSAIDs and then stopping their use ( an “on–off NSAIDs” pattern ) . The median follow-up surveillance time per individual was 11 . 6 years ( range 6 . 3–19 ) . A total of 74 endoscopies and 161 biopsies were selected ( Figure 1B ) as well as one blood sample for each of the thirteen individuals to serve as normal constitutive genotype control . The 161 frozen biopsies were thawed and rinsed in Hanks buffered salt solution without divalent cations ( HBSS , Gibco/BRL ) . Biopsies were incubated 60 minutes at room temperature in 30 mM EDTA in HBSS preheated to 37°C . Barrett's epithelium was isolated by gently peeling it away from the stroma with microdissection needles under a dissecting microscope [41] . The 13 frozen blood samples were processed the same way as the biopsies , except for the epithelial isolation step . DNA was extracted using Puregene DNA Isolation Kit as recommended by the manufacturer ( Gentra Systems , Inc . Minneapolis , MN ) . Samples were quantitated using the Picogreen method ( Quant-iT dsDNA Assay , Invitrogen ) . A total of 200 ng of DNA at 50 ng/ul concentration was analyzed using Illumina Omni-Quad 1M SNP arrays according to manufacturer's protocol . All samples were evaluated at the Fred Hutchinson Cancer Research Center Genomics Core Laboratory . All raw intensity files were loaded in Illumina's GenomeStudio v3 , normalized and clustered using the SNP manifest and cluster files for build37 of the human genome . In all our analyses we used the total signal intensity R for each SNP , which is the sum of the normalized X ( “A” allele , Cy5 red ) and Y ( “B” allele , Cy3 green ) intensities . We also used the B allele frequency ( BAF ) , which is a modified version of the allelic intensity ratio theta ( θ = 2/p*arctan ( Y/X ) ) , to reduce SNP-to-SNP variation in theta using the canonical clusters . Each individual's BE DNA samples were paired to the individual's control sample ( DNA from blood from the same individual ) , which always appeared normal , i . e . lacking any chromosomal alterations ( none of the control samples had any split in BAF over the entire genome , Figure S11 ) . For each individual , we first excluded the 0 . 2% of SNPs with the lowest R values in the control sample , to remove SNP probes that either perform poorly or fall within germline copy number variant ( CNV ) regions . We corrected for dye bias ( higher fluorescence of the B allele , Cy3 green ) by re-centering the BAF of heterozygous and homozygous SNPs of all samples from observing that the median BAF of heterozygous SNPs was ∼0 . 53 , instead of 0 . 5 . Then , for each individual , we identified the set of heterozygous SNPs; i . e . , SNPs having a BAF in control sample between 0 . 33 and 0 . 66 . Finally , we separated the data into three signal profiles: log2 ( R of BE sample/R of control sample ) for heterozygous SNPs only , log2 ( R of BE sample/R of reference ) for homozygous SNPs only , and reflected and scaled BAF of BE sample , ( mBAF = abs ( BAF of BE sample – 0 . 5 ) *2 ) for heterozygous ( informative for LOH ) SNPs only . We performed separate wavelet-based segmentation on these three signal profiles using the HaarSeg algorithm [42] from the GLAD [43] package ( using parameters haarStart = 3 , haarEnd = 9 , fdrQ = 0 . 0001 , onlySmoothing = T ) . For each individual , we combined all break points of the segmented three signal profiles from each biopsy to create the set of all the observed breakpoints . We defined the events as the segments between each pair of consecutive breakpoints in this set . Thus , events had to have the same exact breakpoints in order to be considered the same event . For every new segment , we used thresholds to call allelic imbalance based on the smoothed mBAF profile , and to call single or double copy gain or loss , based on the homozygous and heterozygous log2R profiles . Thus every new segment meeting the thresholds received one of eight molecular state calls: AB ( normal ) , AA ( copy neutral LOH ) , A ( single copy loss ) , 0 ( double copy loss ) , AAB ( single copy gain ) , AAA ( LOH plus subsequent single copy gain ) , AAAA ( LOH plus subsequent double copy gain ) , AABB ( double copy gain ) . In summary , Table S3 shows all calling thresholds used and Dataset S1 shows raw data segmentation and SGA calls for all 161 biopsies of individuals a–m . The GLAD segmentation detects break points of SGA for each sample individually . For each individual , we ran a segment merging procedure that merged two adjacent , neighboring segments if they had the same molecular state call across all samples of that individual . Thus , the number of segments per individual can vary . IMPUTE2 [44] and a reference dataset of 566 CEU haplotypes , part of the 1000 Genomes Project [45] , was used to phase each individual's blood control sample . Having haplotype assignments for the A and B alleles of every SNP , we developed an algorithm to assign a haplotype state for every segment of allelic imbalance . This results in conversion of AA , A , AAB , AAA , AAAA calls to BB , B , BBA , BBB , BBBB calls for segments having lost or gained the opposite allele . For simplicity of all subsequent analyses , all segments having AB molecular states were assigned an “absence of SGA” call , and all segments having other molecular states were assigned a “presence of SGA” call . The final results are individual-specific phylogenetic matrices having samples as taxa , chromosomal segments as characters , and binary molecular states ( SGA absence/presence , or 0/1 ) as character states . We found evidence of a mixture of clones in 24 of the 161 biopsies ( 15% , Table S1 ) , based on differential split in BAF on at least two locations within the same chromosome or on at least two locations in two distinct chromosomes . We have done technical replicates comparing two independent blood samples for individuals m and f ( Table S5 ) . All four blood DNA samples were assayed with the Illumina OmniQuad platform on different dates and our results show high concordance between blood samples within individuals ( Pearson r>0 . 99 for B allele frequency and r>0 . 94 for intensity; Figure S11 ) . We detected only 6 and 18 SGA events in blood samples of individuals f and m , respectively , when using the same SGA detection pipeline and SGA events calling thresholds ( Table S6 ) . Overall , the SGA calls are based on many SNPs and are robust to technical noise in single SNPs . To measure mutation rate change associated with NSAID use , we used a two epoch model in BEAST [31] , where the transition time between the first and second sampling periods is the time of change in NSAID use . We ran BEAST for 10 million Bayesian MCMC iterations that sample the space of genealogies and population parameters to obtain posterior distributions for model parameters that best fit the data . We used uniform prior distributions for SGA rate with lower and upper bounds of 10−5 and 104 SGAs per biopsy genome per year , respectively , for the duration of any of the on-NSAID and off-NSAID periods and estimated SGA rate separately for the first and second sampling periods , where each SGA rate adheres to the molecular clock hypothesis ( SGA occur at constant rate for all evolving lineages ) for the period duration . We added a 0/1 mutation model in BEAST for the SGA absence/presence character states ( Text S1: Equations S3–4 ) and this model assumed that SGAs do not revert to the normal type , i . e . , 1→0 transition is impossible . An SGA character was defined by the breakpoints , which had to occur at the exact same SNPs between samples to be considered the same character . We also modified BEAST's likelihood calculation algorithm to consider a last universal common ancestor ( LUCA ) that has an unaltered genomic state ( zeros for all sites ) , and that connects to the most recent common ancestor ( MRCA ) , at the root of the tree , creating an extra LUCA-MRCA branch . Thus , the final likelihood of the tree is the product of the likelihood of the tree at the root , calculated with Felsenstein's pruning algorithm [46] , multiplied by the probability of the LUCA-MRCA branch length . We used crypt density results ( Table S4 ) and a logistic growth model ( Text S1 ) to estimate the age of the root of the tree and to constrain the internal node coalescence times in the two epoch ( off- on- NSAID ) BEAST analysis . Maximum parsimony trees were estimated using Wagner parsimony with delayed transformation ( DELTRAN ) on the individual-specific phylogenetic matrix with 0/1 SGA states using the PAUP program [47] . For PAUP analyses , we also used a character transition matrix that assumes infinite cost for 1→0 transitions , i . e . SGAs do not revert to normal type . For individuals with off-on NSAID regimens ( a–k ) all lineages ( i . e . phylogeny branches ) having any off-NSAID descendant lineages ( i . e . leading to off-NSAID biopsies ) were considered to have evolved during off-NSAID usage period . Similarly , for individuals l and m , all lineages having any on-NSAID descendant lineages were considered to have evolved during on-NSAID usage period . In all individuals ( a–m ) , lineages having descendant lineages leading to all on-NSAID or all off-NSAID biopsies were considered to have evolved during on-NSAID and off-NSAID usage periods , respectively . We calculated genetic divergence between biopsies separated by time and space ( Text S2 , Figures S6 , S7 ) . We also performed additional analyses of SGA lesions new appearances and regressions during on- and off- NSAID periods that do not use phylogenies ( Text S3 , Figures S12 , S13 and Text S4 , Figure S14 ) . The spatial distribution of the number of SGA events detected in biopsies is summarized with a violin plot for each individual ( Figure S15 ) . All code used for the above analyses is licensed under the GNU Lesser GPL ( http://www . gnu . org/licenses/lgpl . html ) and made publicly available at https://github . com/rkostadi/BEClonalEvolutionNSAID . The Seattle Barrett's Esophagus Study has been approved by the University of Washington Human Subjects Review Committee since 1983 with reciprocity from the Fred Hutchinson Cancer Research Center Institutional Review Board since 1994 . . Informed consent had been collected from the research participants . Seventy seven percent of the participants are male and 23% are female . The ethnic distribution of the study cohort is 85% White , 0 . 4% Black , 1 . 3% Hispanic , 0 . 9% Asian , and 0 . 9% Native American and 11% unknown . This gender and ethnic distribution reflects the known demographics of Barrett's esophagus and esophageal adenocarcinoma in the United States , which is predominantly a disease of white , middle aged and elderly men . | Cancer is a disease that develops over decades as result of accumulation of abnormalities in the genomes of otherwise normal cells . Cells in tumors compete for space and resources . Those cells able to survive the Darwinian struggle for existence within tissues progressively evolve uncontrolled growth and in some cases this results in cancer . Aspirin and other non-steroidal anti-inflammatory drugs ( NSAIDs ) reduce death rate from multiple types of cancer by about 20% . However , the mechanisms by which NSAIDs act to prevent cancer are not fully understood . By examining thirteen individuals with Barrett's esophagus over time , we showed that the rate of accumulation of genomic abnormalities decreased when most individuals started taking NSAIDs . We also observed that , surprisingly , the number of abnormalities in the Barrett's tissues did not increase much over decades . However , in one individual who progressed to esophageal cancer , we observed massive genomic abnormalities affecting 19% of the genome . These findings suggest that NSAIDs may prevent cancer by reducing the accumulation of genomic abnormalities over time and that detection of stable versus unstable genomes may be used in the clinic to help manage treatment options in Barrett's esophagus . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"cancer",
"detection",
"and",
"diagnosis",
"medicine",
"genome",
"evolution",
"population",
"genetics",
"cancer",
"risk",
"factors",
"cancers",
"and",
"neoplasms",
"chemoprevention",
"gastrointestinal",
"tumors",
"predisposing",
"conditions",
"and",
"syndromes",
"oncology... | 2013 | NSAIDs Modulate Clonal Evolution in Barrett's Esophagus |
Chikungunya virus ( CHIKV ) is a re-emerging arthropod-borne ( arbo ) virus that causes chikungunya fever in humans and is predominantly transmitted by Aedes aegypti mosquitoes . The CHIKV replication machinery consists of four non-structural proteins ( nsP1-4 ) that additionally require the presence of a number of host proteins for replication of the viral RNA . NsP3 is essential for CHIKV replication and has a conserved macro , central and C-terminal hypervariable domain ( HVD ) . The HVD is intrinsically disordered and interacts with various host proteins via conserved short peptide motifs: A proline-rich ( P-rich ) motif that has affinity for SH3-domain containing proteins and duplicate FGDF motifs with affinity for G3BP and its mosquito homologue Rasputin . The importance of these motifs for infection of mammalian cells has previously been implicated . However , their role during CHIKV infection of mosquito cells and transmission by mosquitoes remains unclear . Here , we show that in-frame deletion of the P-rich motif is lethal for CHIKV replication in both mosquito and mammalian cells . However , while mutagenesis of the P-rich motif negatively affects replication both in mammalian and mosquito cells , it did not compromise the infection and transmission of CHIKV by Ae . aegypti mosquitoes . Mutagenesis of both FGDF motifs together completely inactivated CHIKV replication in both mammalian and mosquito cells . Importantly , mutation of a single FGDF motif attenuated CHIKV replication in mammalian cells , while replication in mosquito cells was similar to wild type . Surprisingly , CHIKV mutants containing only a single FGDF motif were efficiently transmitted by Ae . aegypti . The P-rich motif in CHIKV nsP3 is dispensable for transmission by mosquitoes . A single FGDF motif is sufficient for infection and dissemination in mosquitoes , but duplicate FGDF motifs are required for the efficient infection from the mosquito saliva to a vertebrate host . These results contribute to understanding the dynamics of the alphavirus transmission cycle and may help the development of arboviral intervention strategies .
Alphaviruses ( family Togaviridae ) , such as chikungunya virus ( CHIKV ) and Mayaro virus , are ( re- ) emerging arthropod-borne ( arbo ) viruses that impose a global threat on human health [1] . Mayaro virus recently emerged in South America , where a large outbreak was reported in Brazil [1 , 2] . CHIKV is responsible for several large outbreaks in the Americas [3] and over 150 . 000 cases of infection were reported in the Americas in 2016 alone [4] . CHIKV disease is characterized by acute fever combined with severe arthralgia , myalgia or rash , and symptoms can last from weeks to several months [3] . Transmission of CHIKV occurs predominantly by Aedes aegypti mosquitoes , although some CHIKV strains have incorporated an amino acid substitution in the envelope protein ( A226V ) that allows transmission by Ae . albopictus [5] . Alphaviruses have a single-stranded positive-sense RNA genome that contains two open reading frames , which encode the non-structural proteins ( nsP ) 1-4 and the structural proteins , respectively . NsP4 is the viral RNA-dependent RNA polymerase that together with nsP1-3 forms the viral replication complex . While the roles of nsP1 and nsP2 in viral replication have been extensively documented and reviewed [6–8] , the exact functions of nsP3 are less clear [9 , 10] . During infection , nsP3 localizes partially to viral replication complexes [11 , 12] , in accordance with the necessity of nsP3 for virus replication , as well as cytoplasmic foci [13–16] . Additionally , nsP3 interacts with several host-proteins in mammalian and mosquito cells [14 , 17] . Unravelling the significance and mechanisms of the interactions between nsP3 and these host-proteins in mammalian cells has been the focus of several recent studies [17–23] , while the role of these interactions during infection of mosquitoes remains to be assessed . NsP3 consists of three domains: an N-terminal , conserved macro domain , which possesses phosphatase and RNA-binding activity [24] , a conserved zinc-binding central domain [25] , and a highly phosphorylated [15 , 26 , 27] , intrinsically disordered [23] , hypervariable domain ( HVD ) at the C-terminus [6] . Despite the lack of conservation in the HVD between the nsP3 of alphaviruses [28] , it contains several conserved motifs , of which the Proline-rich ( P-rich ) PxxPPR or PxPxPR , and the duplicate FGDF motifs have been studied most extensively ( reviewed in [9 , 10] ) . In mammalian cells , the P-rich motifs of Sindbis virus ( SINV ) , Semliki Forest virus ( SFV ) and CHIKV nsP3 interact with the SH3 domain of the membrane modulating proteins amphiphysin-1/2 , and this interaction has been shown to be important for efficient virus replication [18 , 19] . CHIKV nsP3 has two FGDF motifs that can bind to the NTF2-like domain of the stress granule components G3BP1 and G3BP2 in mammalian cells [17 , 20 , 29] . This interaction inhibits the formation of cellular stress granules that would otherwise stall host and viral translation , and as a consequence may facilitate virus replication [13 , 21 , 30] . In mammalian cells , mutations in the alphavirus nsP3 HVD that disrupt both of the FGDF motifs result in complete inactivation of CHIKV and also attenuates SFV and SINV [17 , 20 , 21 , 31] . Furthermore , deletion of a single FGDF motif in CHIKV or SFV nsP3 decreases the affinity for G3BP and attenuates the virus , suggesting that alphaviruses require duplicate FGDF motifs for their successful replication in mammalian cells [17 , 20] . While the importance of the conserved motifs in the nsP3 HVD for alphavirus replication in mammalian cells has been well-established , little is known about their role during alphavirus transmission by the mosquito vector [32] . Recently , nsP3 and its mosquito interaction partner Rasputin ( Rin ) , the mosquito homologue of G3BP , have been identified as important determinants for alphavirus transmission [33 , 34] . Chimeric CHIKV equipped with the nsP3 from the Anopheles-transmitted alphavirus ONNV showed an increase in infection rate in the CHIKV-refractory vector An . gambiae from 0 to 64% [33] . Interestingly , CHIKV chimeras that contained just the C-terminal region of ONNV nsP3 , including the HVD , already reached an infection rate of 9–18% [33] , suggesting that the nsP3 HVD is an important determinant of mosquito vector specificity . Furthermore , CHIKV nsP3 was shown to co-localize with Rin [34] , and Rin co-precipitated with SINV nsP3 [14] , indicative of a direct interaction between Rin and nsP3 . Interestingly , mutating both FGDF motifs in CHIKV nsP3 together results in complete loss of co-localization of transiently expressed nsP3 and Rin [34] , similar to the interaction with G3BP in mammalian cells . However , co-localisation was maintained when only one of the FDGF motifs was mutated [34] . Importantly , knock-down of Rin severely decreased the infection and transmission rates as well as the viral titers of CHIKV in the heads of Ae . albopictus after an infectious bloodmeal [34] . Thus , the conserved motifs in the nsP3 HVD may play a crucial role in infection of the mosquito vector , potentially through their interaction with mosquito host proteins , and ultimately enable the virus to establish a disseminated infection . Here , we investigated the role of the P-rich and FGDF motifs in the HVD of CHIKV nsP3 during virus replication in mosquito cells and infection of and transmission by in the natural CHIKV vector Ae . aegypti . Infectious cDNA clones were used to generate CHIKV variants with mutations in either the P-rich or FGDF motif ( s ) . The mutant viruses were assessed for replication kinetics in mammalian ( Vero ) and Ae . aegypti ( Aag2 ) cells . We investigated the infection and transmission of the CHIKV mutants in Ae . aegypti mosquitoes through infectious bloodmeal experiments . These experiments elucidate whether the evolutionary basis for conservation of the P-rich and duplicate FGDF motifs in alphavirus nsP3 lies solely in infection of the mammalian host , or whether these motifs are also required for alphavirus transmission by the mosquito vector .
African green monkey kidney Vero E6 cells ( ATCC CRl-1586 ) were cultured at 37°C with 5% CO2 in Dulbecco’s Modified Eagle medium ( DMEM; Gibco ) supplemented with 10% fetal bovine serum ( FBS; Gibco ) , penicillin ( 100 U/ml; Sigma-Aldrich ) and streptomycin ( 100 μg/ml; Sigma-Aldrich ) . Ae . albopictus C6/36 ( ATCC CRL-1660 ) cells were cultured at 28°C in Leibovitz L-15 medium ( Gibco ) supplemented with 10% FBS , 2% tryptose phosphate broth ( Gibco ) and 1% nonessential amino acids ( Gibco ) . Ae . aegypti Aag2 cells were cultured at 28°C in Schneider’s Drosophila medium ( Lonza ) supplemented with 10% FBS . During experiments gentamicin ( 50 μg/ml; Life technologies ) was added to the culture medium of C6/36 and Aag2 cells . Previously reported infectious clone derived chikungunya virus 37997 strain ( CHIKV37997 ) [35] was used in all experiments . Gateway-compatible pIB vectors with CHIKV nsP3 sequences containing the mutations P398A , PPR401AAA , FG479AA , FG497AA and FG479AA/FG497AA were described previously [34] . To insert an internal fluorescent reporter gene into nsP3 , mCherry was PCR amplified from CHIKrep-nsP3mC [13] with primers 1 , 2 ( Primer sequences are listed in Table 1 ) containing BspEI overhangs and cloned as BspEI fragment into the pIB-nsP3 vectors . This introduces the mCherry sequence between amino acids 367–368 of the CHIKV nsP3 gene . The Firefly luciferase gene was amplified from the previously described pCHIKrep-FlucEGFP [13] with primers 7 , 8 and used to replace the FlucEGFP sequence in the same pCHIKrep-FlucEGFP to generate pCHIKrep . Mutant nsP3 containing mutations P398A and PPR401AAA in the P-rich motif were cloned from the corresponding pIB vectors as AgeI/NheI fragment into pCHIKrep-Fluc to create pCHIKrep-nsP3mCP398A and pCHIKrep-nsP3mCPPR401AAA . NsP3 containing mutations in the FGDF motifs were PCR amplified using primers 3 , 4 , nsP4 was PCR amplified using primers 5 , 6 and the two PCR products were fused by overlapping-PCR using primers 3 , 6 . Subsequently , the nsP3-nsP4 fusion PCR products were cloned as AgeI/AscI fragment into pCHIKrep-Fluc to create pCHIKrep-nsP3mC-FGN ( FG479AA ) , pCHIKrep-nsP3mC-FGC ( FG497AA ) and pCHIKrep-nsP3mC-FGNC ( FG479AA & FG497AA ) . The Fluc sequence was replaced with the CHIKV37997 structural cassette from pCHIKIC through AscI/EcoRI cloning to generate pCHIKIC-nsP3mC infectious clones of all mutants . The internal mCherry was removed by BspEI digestion and vector self-ligation . An overview of the recombinant CHIKV encoding plasmids used in this study is shown in Fig 1A . Recombinant viruses were generated from the pCHIKIC infectious clones as described previously [35] . Briefly , pCHIKIC plasmids were linearized by PacI digestion to serve as template for an SP6 RNA polymerase ( NEB ) transcription reaction and the resulting in vitro transcribed RNA was used to transfect pre-seeded Vero cell monolayers with Lipofectamine 2000 ( Invitrogen ) to generate the P0 stock . Subsequent passages were performed on C6/36 cells . Gateway-compatible and OpIE-2 promoter driven pIB insect-cell expression plasmids expressing EGFP , nsP3EGFP , nsP3EGFP-FGN , nsP3EGFP-FGC and nsP3EGFP-FGNC were reported previously [34] . The Ae . aegypti poly-ubiquitin ( PUB ) promoter was amplified from pPUB-Fluc [36] using primers 9 , 10 ( Table 1 ) containing BspHI and SacI restriction sites and cloned as BspHI/SacI fragment into pIB to generate the Gateway-compatible pPUB vector . Gateway-cloning ( Invitrogen ) was used to generate pPUB-EGFP , pPUB-nsP3EGFP , pPUB-nsP3EGFP-FGN , pPUB-nsP3EGFP-FGC and pPUB-nsP3EGFP-FGNC . Plasmid pIB-Rin-mC , expressing the mosquito Rin fused C-terminally to an mCherry red fluorescent reporter gene , was reported previously [34] . EGFP was PCR amplified from pEGFP-N1 with primers 13 , 14 ( Table 1 ) containing KpnI and SacII overhangs and cloned as KpnI/SacII fragment into pIB-Rin-mC to generate pIB-Rin-EGFP . The PUB-promoter was amplified from pPUB-EGFP with primers 11 , 12 ( Table 1 ) containing EcoRI overhangs and cloned as EcoRI/EcoRI fragment into pIB-Rin-mC and pIB-Rin-EGFP to generate pPUB-Rin-mC and pPUB-Rin-EGFP respectively . Vero or Aag2 cell monolayers were seeded one day prior to infection in a 6-wells plate . The amount of inoculum for the specific multiplicity of infection ( MOI ) was calculated based on end-point dilution assay ( EPDA ) on either Vero or Aag2 cells . Cell monolayers were infected with 1 ml of inoculum in either DMEM- 4- ( 2-hydroxyethyl ) -1-piperazineethanesulfonic acid ( HEPES ) ( Gibco ) supplemented with 10% FBS , streptomycin and penicillin ( DMEM-supplemented ) or Schneider’s Drosophila medium supplemented with 10% FBS and gentamycin ( 50 μg/ml ) ( Schneider’s-supplemented ) . After 2 h the inoculum was aspirated , monolayers were washed with phosphate-buffered saline ( PBS ) and 2 ml of fresh culture medium was added . Vero cells were incubated at 37°C with 5% CO2 and Aag2 cells were incubated at 28°C . At the indicated time points post infection 30 μl supernatant samples were taken and after one freeze-thaw cycle ( -80°C ) the infectious virus titre was determined by EPDA on either Vero or Aag2 cells . Vero cells were detached by incubation with Trypsin-EDTA ( Gibco ) and diluted 1:4 in DMEM-supplemented , while Aag2 cells were detached by force and diluted 1:8 with Schneider’s supplemented . Samples were thawed , vortexed and 10-fold serial dilutions were made in either DMEM-supplemented or Schneider’s-supplemented . Cell suspensions were added to the dilutions 1:1 and 10 μl of inoculated dilution was plated 6-fold in micro-titer plates ( Nunc ) . When samples from mosquitoes were titrated , Fungizone ( 50 μg/ml; Invitrogen ) and gentamycin ( 50 μg/ml; Life Technologies ) , was added to the medium and cell fractions prior to titrations ( DMEM-complete , Schneider’s-complete ) . Titrations on Vero cells were scored by CHIKV-induced CPE , while titrations on Aag2 cells were scored by α-CHIKV-E2 based immunofluorescence assay . Vero or Aag2 cells were fixed by >10 min incubation in 4% paraformaldehyde in PBS . Cells were permeabilized by >10 min incubation in 0 . 1% sodium dodecyl sulphate in PBS , washed three times with PBS and blocked by incubation in 5% FBS in PBS for 30 min . Primary antibody binding was performed with α-CHIKV-E2 ( rabbit polyclonal; 1:5000 [37] ) or α-G3BP1 ( 1:500; G6046 Sigma Aldrich ) diluted in 5% FBS in PBS for 1 h at room temperature ( RT ) . Cells were washed three times with PBS and incubated with the secondary antibody goat-α-rabbit-Alexa Fluor 488 ( 1:1000; Invitrogen ) in 5% FBS in PBS for 1 h at 37°C . Cells were washed three times with PBS and visualized using an Axio Observer Z1m inverted microscope ( Zeiss ) in combination with an X-Cite 120 series lamp . Ae . aegypti mosquitoes ( Rockefeller strain , obtained from Bayer AG , ) were reared as described previously [35] and transported to the Biological Safety level 3 facility in buckets ( diameter: 12 . 2 cm , height: 12 . 2 cm; Jokey ) for virus infection assays . Mosquitoes were starved for one day and then offered an infectious bloodmeal from a 1:1 mixture of virus and human blood ( Sanquin ) in a total volume of 1 ml via a Hemotek feeder covered with a Parafilm membrane . Mosquitoes were allowed to feed for 1 h under light conditions at 24°C and 70% relative humidity ( RH ) . Mosquitoes were anesthetized with CO2 and fully engorged females were selected and maintained in buckets at 28°C , 12:12 light:dark cycle , 70% relative humidity and provided with a cotton pad soaked in 6% glucose solution . After 7 or 14 days , mosquitoes were anesthetized with 100% CO2 , and immobilized by removing their legs and wings . The proboscis of each stripped mosquito was inserted into a 200 μl pipet tip containing 5 μl of a 1:1 mixture of 50% glucose and FBS . Mosquitoes were allowed to salivate for >45 min , after which the bodies were individually transferred to 1 . 5 ml Safe-Seal micro tubes ( Sarstedt ) that contained a small scoop of 0 . 5 mm zirconium beads ( Next Advance ) . Saliva samples were added to 1 . 5 ml micro tubes ( Sarstedt ) that contained 60 μl DMEM-complete . All samples were stored at -80°C until further processing . For most experiments , transmission was determined at 7 days post infection ( dpi ) , as wild type CHIKV reaches relatively high transmission rates at this time-point . This therefore presents a good resolution for the detection of a putative attenuation of a mutant virus . Frozen mosquito bodies were homogenized for 2 min at maximum speed in a Bullet Blender Storm ( Next Advance ) . Homogenized bodies were centrifuged for 1 min at 14 . 500 rpm , after which 100 μl of DMEM-complete was added and the homogenization was repeated . Finally the mosquito debris was pelleted by 1 min centrifugation at 14 . 500 rpm . Thirty microliters of mosquito body supernatant or saliva samples were used to inoculate either Vero or Aag2 cell monolayers in 96-wells plates . After 2–3 h incubation the inoculum was removed and replaced with fresh DMEM-complete or Schneider’s-complete . Infectivity was scored at 3 dpi through virus-induced CPE in Vero cells or by CHIKV-E2-based immunofluorescence in Aag2 cells , as described above . Mosquitoes with CHIKV-positive saliva were selected from mosquitoes inoculated with CHIKIC , CHIKICP398A , CHIKICPPR401AAA , CHIKIC-FGN or CHIKIC-FGC . A selection of mosquitoes was made to represent each replicate experiment . Thirty μl of mosquito sample supernatant was used to inoculate a well of a 96-wells plate pre-seeded with C6/36 cells . At 3 days post infection RNA was isolated with TRIzol reagent ( Invitrogen ) and 300 ng total RNA was subjected to one-step RT-PCR with primers 3 , 4 using SuperScript III one-step RT-PCR system with Platinum Taq DNA polymerase ( Invitrogen ) following the manufacturer’s protocol . RT-PCR products were observed on agarose gel to confirm the infection of mosquitoes with CHIKV . To investigate the presence and preservation of the expected mutations , RT-PCR products were subjected to digestion with SacII ( NEB ) , NotI ( NEB ) or ApeKI ( NEB ) according to the manufacturer’s protocol . Digestions were observed on agarose gel . Transfections of mosquito cells with DNA plasmids were performed with Fugene-HD ( Promega ) in serum-free media following the manufacturer’s protocol . Transfections of Vero cells with DNA plasmids and transfections of C6/36 , Aag2 and Vero cells with in vitro transcribed RNA were performed using Lipofectamine 2000 ( Invitrogen ) in Opti-MEM ( Gibco ) . At 4 hours post transfection ( hpt ) with Lipofectamine 2000 the transfection mix was replaced with fresh culture media . The cell culture volume was aspirated and cells were lysed by 20 min incubation in 1X passive lysis buffer ( Promega ) at RT . Ninety μl of cell lysate was pipetted into an opaque 96-wells plate and 90 μl of Steady-Glo reagent ( Promega ) was added . After 10 min the luciferase activity was measured using a FLUOstar Optima microplate reader ( BMG Labtech ) . Pre-seeded Vero or Aag2 cells were transfected with plasmids expressing nsP3 or Rin or both . One day post-transfection cells were detached , washed once with PBS and lysed by 30 min incubation in lysis buffer ( 10 mM Tris/Cl pH 7 . 5; 150 mM NaCl; 0 . 5 mM EDTA; 0 . 5% NP-40 ) supplemented with Complete protease inhibitors ( Roche ) and 1 mM phenylmethylsulfonyl fluoride ( PMSF ) . Lysates were cleared by 10 min 20 . 000 g centrifugation at 4°C and diluted 5:2 with dilution buffer ( 10 mM Tris/Cl pH 7 . 5; 150 mM NaCl; 0 . 5 mM EDTA ) . Twenty-five μl GFP-Trap_A beads ( Chromotek ) were added to 500 μl dilution buffer and pelleted by 2 min 2 , 500 g centrifugation at 4°C . Washing was repeated twice . Cell lysates were added to the equilibrated beads and incubated for 1–2 h at 4°C with overhead rotation . Beads were washed thrice by resuspension in dilution buffer followed by 2 min 2 , 500 g centrifugation at 4°C . Bound protein complexes were eluted by resuspension in 2X SDS-loading buffer containing 10% β-mercaptoethanol followed by 10 min incubation at 95°C . Beads were removed by 5 min 3 , 000 g centrifugation and eluates were subjected to western blotting . Briefly , proteins were size-separated by SDS-PAGE and semi-dry blotted onto Immobilon-P membranes ( Merck Millipore ) . Membranes were blocked overnight at 4°C by incubation in 1% milk powder/PBS-0 . 05%-Tween-20 ( PBST ) . Blocked membranes were probed for 1 h at RT with primary antibodies α-G3BP ( 1:1000; G6046 Sigma Aldrich ) , α-GFP ( 1:2000; A6455 Molecular Probes ) , α-mCherry ( 1:1000; ab183628 Abcam ) diluted in 1% milk powder/PBST . Membranes were washed thrice for 5 min with PBST and probed for 1 h at RT with alkaline phosphatase ( AP ) conjugated secondary antibody goat-α-rabbit-AP ( 1:2500; D0487 Dako ) diluted in PBST . Membranes were washed thrice for 5 min in PBS-T and developed with nitroblue tetrazolium ( NBT ) /BCIP ( 5-bromo-4-chloro-3-indolylphosphate ) ( Roche ) until the desired signal was achieved . Statistical analysis were performed in SPSS Statistics 23 . Virus titers in mosquito bodies and supernatants from growth curve experiments were checked for normality by Kolmogorov-Smirnov test . Data that did not follow a normal distribution was Log10 transformed and confirmed for normality . Means were compared by one-way ANOVA with Tukey post-hoc test ( α = 0 . 05 ) . Data that did not follow a normal distribution , despite Log10 transformation was analysed by Kruskal-Wallis test with Dunn’s post-hoc test ( α = 0 . 05 ) . Two-tailed Fisher’s exact tests were used to compare infection and transmission rates and performed with GraphPad QuickCalcs ( α = 0 . 05 ) . Ns non-significant; * P < 0 . 05; ** P ≤0 . 01; *** P ≤0 . 001 .
For SINV , SFV and CHIKV it has been shown that deletion or mutation of the conserved P-rich motif ( Fig 1B ) results in decreased replication in mammalian cells [13 , 18] . To investigate whether the P-rich motif is similarly indispensable for alphavirus replication in mosquito cells luciferase assays were performed on cells transfected with in vitro transcribed RNA of a CHIKV replicon ( CHIKrep ) with an in-frame deletion of the P-rich motif ( ΔPVA ) ( Fig 1A , Fig 2A ) . The mutant replicon CHIKrepΔPVA was unable to replicate in either mammalian Vero or Aag2 Ae . aegypti mosquito cells , whereas wild type CHIKrep was able to replicate in both cell types . We also transfected cells with a CHIKV infectious clone containing the same deletion of the P-rich motif ( CHIKICΔPVA ) , but again we were unable to produce infectious virus in either mammalian and mosquito cells . Together , these results clearly indicate that the P-rich motif is essential for CHIKV replication in both mammalian and mosquito cells . To investigate whether the entire , intact P-rich motif is required for virus production in mammalian or mosquito cells , amino acid substitutions were made in the CHIKV infectious clone ( CHIKIC ) to produce CHIKICP398A and CHIKICPPR401AAA ( Fig 1A ) . These specific mutations are designed to disrupt the residues in the P-rich motif that are conserved among old-world alphaviruses ( PxxPPR; Fig 1B red boxes ) . Viral growth curves were determined in duplicate using an MOI of 0 . 01 and viral titers were determined by EPDA on Vero cells . The results show that both mutants displayed significantly delayed growth kinetics compared to the wild type virus in both Vero ( Fig 2B ) and Aag2 cells ( Fig 2C ) . Even when virus infections were performed at a higher MOI of 5 this attenuating effect was observed at 24 hpi , although both CHIKICP398A and CHIKICPPR401AAA still reached a titer of ~1 . 0 × 106 TCID50/ml ( Fig 2D ) . Together , these results indicate that an intact P-rich motif in CHIKV nsP3 is important , but not essential , for virus replication in mammalian and mosquito cells . The P-rich motif is conserved in most if not all mosquito transmitted alphaviruses and also in the insect-specific alphaviruses Tai Forest Alphavirus ( TALV ) and Eilat virus ( EILV ) ( Fig 1B ) . As TALV and EILV are proposedly unable to replicate in a vertebrate host [38 , 39] , conservation of a P-rich motif for these viruses suggests its importance for alphavirus infection in the invertebrate vector . To investigate whether the P-rich motif of CHIKV nsP3 is involved in virus transmission by mosquitoes , female Ae . aegypti were offered an infectious bloodmeal containing either CHIKIC , CHIKICP398A or CHIKICPPR401AAA ( Fig 3A ) . After 14 days , the infection and transmission rates were determined by infectivity assay on Vero cells ( Fig 3B ) . For all three viruses , similar infection ( 88–96%; P > 0 . 37 ) and transmission ( 6–7 . 7%; P = 1 . 00 ) rates were obtained , indicating that the P-rich motif is not required for virus transmission by mosquitoes . To investigate whether putative differences could be observed earlier after infection of the mosquito vector , we also determined the infection and transmission rates after 7 days ( Fig 3C ) . Again , infection ( 72–76%; P > 0 . 61 ) and transmission ( 18–23%; P > 0 . 47 ) rates were obtained that were not significantly different when compared to the wild type virus . In order to determine whether the mutations in CHIKICP398A and CHIKICPPR401AAA were preserved during infection of the mosquito , the nsP3 sequences were amplified by RT-PCR for a selection of mosquitoes with CHIKV-positive saliva . Presence of the expected mutations in nsP3 was confirmed , indicating that the mutant viruses did not revert to wild type during replication in the mosquito vector ( S1 Fig ) . To assess whether the titers in the mosquito bodies would perhaps be decreased due to the mutations in the P-rich motif , CHIKV titers were determined in the bodies of mosquitoes with virus-positive saliva ( Fig 3D ) . However , no difference in viral titer between CHIKIC , CHIKICP398A and CHIKICPPR401AAA ( P > 0 . 99 ) was observed . Together , these results indicate that the CHIKV nsP3 P-rich motif is not required for the transmission by Ae . aegypti mosquitoes . Research has indicated that mutating both FGDF motifs together significantly attenuates SFV and renders CHIKV unable to replicate in mammalian cells [17 , 20] . Here , we investigated whether mutations in one ( nsP3-FGN , nsP3-FGC ) or both ( nsP3-FGNC ) FGDF motifs would interfere with the replication of CHIKV in mosquito cells . These amino acid mutations were previously shown to disrupt the interaction of nsP3 with G3BP [20 , 40] , and the co-localization of nsP3 with Rin [34] . We tested the effect of mutating the FGDF motifs on CHIKV replication using an infectious clone that contains an in-frame insertion of the mCherry reporter gene in nsP3 ( CHIKICnsP3mC; Fig 1A ) . In vitro transcribed , capped RNA from CHIKICnsP3mC , CHIKICnsP3mC-FGN , CHIKICnsP3mC-FGC or CHIKICnsP3mC-FGNC was used to transfect Vero and C6/36 mosquito cells . At 36 hpt , the cells were fixed and analysed for mCherry fluorescence as a marker for virus replication ( Fig 4A ) . Red fluorescence was observed in cells transfected with RNA from the infectious clones with mutations in a single FGDF motif ( CHIKICnsP3mC-FGN and CHIKICnsP3mC-FGC ) and wild type CHIKV . However , mammalian and mosquito cells transfected with RNA from CHIKICnsP3mC-FGNC did not show any red fluorescence , indicating that a minimum of one functional FGDF motif is essential for CHIKV replication in both mammalian and mosquito cells . To assess whether the RNA of CHIKICnsP3mC-FGNC was of sufficient quality , we co-transfected replication-competent RNA from CHIKIC , which does not express mCherry . In this experiment , CHIKIC RNA is able to complement the CHIKICnsP3mC-FGNC RNA , leading to mCherry expression . Indeed , upon co-transfection red fluorescence could be observed in some cells ( Fig 4B ) , indicating that the quality of CHIKICnsP3mC-FGNC RNA was fine but that the RNA was replication-incompetent due to the mutation of both FGDF motifs . To confirm this result using a more sensitive assay , Vero and C6/36 cells were transfected with CHIKV replicons expressing firefly luciferase CHIKrep or CHIKrep-FGNC ( Fig 1A ) and firefly luciferase activity was measured at 24 hpt ( Fig 4C ) . In both Vero and C6/36 cells , only the wild type replicon CHIKrep produced above-background luciferase levels , indicating that indeed mutation of both FGDF motifs together results in complete inactivation of CHIKV in mammalian and mosquito cells . To assess the effect of single FGDF mutations on virus production in mammalian and mosquito cells , we performed viral growth curves on Vero and Aag2 cells at an MOI of 0 . 01 using CHIKIC , CHIKIC-FGN and CHIKIC-FGC ( Fig 4D and 4E ) . These viruses do not contain the mCherry reporter gene ( Fig 1A ) . It is important to note that we determined the viral titers for the inoculum and the samples for the mammalian growth curves by EPDA on Vero cells . For the growth curves on mosquito cells , the titrations were performed by EPDA on Aag2 cells . As expected , CHIKIC-FGN and CHIKIC-FGC were severely attenuated in Vero cells ( Fig 4D ) . However , the result was very different in Aag2 cells; CHIKIC-FGN and CHIKICFGC displayed remarkably similar growth rates as compared to CHIKIC ( Fig 4E ) . In summary , duplicate FGDF motifs are very important for CHIKV replication in mammalian cells , while a single FGDF motif is sufficient , but required , for efficient CHIKV replication in mosquito cells . While duplicate FGDF motifs do not appear to be required for virus replication in mosquito cells ( Fig 4E ) , they might be important to sustain transmission of CHIKV by mosquitoes . Therefore , we infected Ae . aegypti mosquitoes via an infectious bloodmeal with 2 . 8 × 105 TCID50/ml CHIKIC , CHIKIC-FGN or CHIKIC-FGC and determined the infection and transmission rates at 7 dpi by infectivity assay on Vero and Aag2 cells ( Fig 5A ) . When scored on Vero cells the mosquito infection rate of CHIKIC-FGC ( 54 . 8%; P <0 . 001 ) was significantly lower as compared to CHIKIC ( 73 . 7% ) , while the infection rate of CHIKIC-FGN ( 72 . 6%; P = 0 . 89 ) was similar to wild type ( Fig 5B ) . Next , the saliva of infected mosquitoes was tested for the presence of virus . This showed that the transmission rates of both CHIKIC-FGN ( 7 . 4%; P <0 . 001 ) and CHIKIC-FGC ( 9 . 6% P <0 . 001 ) were significantly lower as compared to CHIKIC ( 25 . 7% ) , indicating that duplicate FGDF motifs are important for CHIKV transmission by Ae . aegypti to the vertebrate host . However , given that CHIKIC-FGN and CHIKIC-FGC were severely attenuated on Vero cells ( Fig 4D ) , we also determined the infection and transmission rates by infectivity assay on Aag2 cells , using exactly the same mosquito bodies and salivas ( Fig 5C ) . The infection rate did not change much , and CHIKIC-FGC ( 57%; P <0 . 01 ) still reached a lower infection rate than CHIKIC ( 74 . 7% ) or CHIKIC-FGN ( 78 . 5%; P = 0 . 49 ) . Interestingly , however , when scored on Aag2 cells the transmission rates were very similar for CHIKIC ( 19 . 3% ) , CHIKIC-FGN ( 16 . 3%; P = 0 . 54 ) and CHIKIC-FGC ( 13 . 7%; P = 0 . 21 ) . To directly compare the infectivity of the mosquito bodies and salivas between Vero and Aag2 cells , the ratio between the number of positive samples on Vero and Aag2 cells was determined for each replicate and normalized to the ratio for mosquitoes inoculated with wild type CHIKIC ( Fig 5D ) . Indeed , the ratio of Vero/Aag2 positive samples was lower for CHIKIC-FGN and CHIKIC-FGC as compared to wild type CHIKIC ( FGN 0 . 18; FGC 0 . 33 ) . To determine whether the mutations in CHIKIC-FGN and CHIKIC-FGC were preserved during replication in the mosquito , the nsP3 sequences of a selection of mosquitoes with CHIKV-positive saliva were amplified by RT-PCR . Presence of the expected mutations in nsP3 was confirmed , indicating that the mutant viruses did not revert to wild type ( S1 Fig ) . Together , these results imply that , while CHIKIC-FGN and CHIKIC-FGC are perfectly able to infect the mosquito body and disseminate to ultimately reach the mosquito saliva , their virions are less able to replicate in vertebrate cells . These results are in agreement with the observed differences between the growth curves on Vero and Aag2 cells ( compare Fig 4D with Fig 4E ) . To assess the effect of mutating one FGDF motif on the viral load in the mosquito body , the viral titers in mosquito bodies of which the saliva was virus positive were determined by EPDA on Aag2 cells ( Fig 5E ) . Viral loads of mosquitoes infected with CHIKIC , CHIKIC-FGN or CHIKIC-FGC were not significantly different ( 2 . 7–5 . 0 × 106 TCID50/ml; P > 0 . 65 ) . Thus , a single FGDF motif in nsP3 is sufficient for the infection and dissemination of CHIKV in Ae . aegypti mosquitoes , but duplicate FGDF motifs are required for efficient infection of the vertebrate host by mosquito bite . We have previously shown that mutations in a single FGDF motif do not abrogate the co-localization of CHIKV nsP3 and the mosquito homolog of G3BP , Rin , in insect cells [34] . However , in mammalian cells it has been suggested that mutation of just the N-terminal FGDF motif of nsP3 already dismantles the interaction with G3BP [17 , 20] . We therefore investigated the effect of mutating either of the two FGDF motifs on the co-localization of CHIKV nsP3 with G3BP in mammalian cells and Rin in mosquito cells . Because there is no anti-Rin antibody available , Aag2 cells were first transfected with a plasmid expressing an EGFP-tagged Rin ( pPUB-Rin-EGFP , Fig 6A ) . Vero ( Fig 6B ) and Aag2 ( Fig 6C ) cells were infected with CHIKICnsP3mC , CHIKICnsP3mC-FGN or CHIKICnsP3mC-FGC . At 24 hpi , cells were fixed and permeabilized and Vero cells were immuno-stained for G3BP-1 . In both mammalian and mosquito cells mutation of either the N-terminal or the C-terminal FGDF motif did not abrogate the co-localization of nsP3 with G3BP and Rin . As co-localization observed by fluorescent microscopy does not prove that there is a physical protein-protein interaction we investigated the importance of the FGDF motifs for interaction of nsP3 with G3BP in mammalian cells and Rin in mosquito cells by co-immunoprecipitation ( co-IP ) . First , Vero cells were transfected with CMV-driven plasmids expressing the nsP3EGFP fusion proteins nsP3 , nsP3-FGN , nsP3-FGC , nsP3-FGNC or just EGFP as a control ( Fig 6A ) . At 24 hpt cells were lysed and lysates were subjected to co-IP with α-GFP beads followed by western blot detection of nsP3EGFP and G3BP in both the cell lysates and co-IP samples ( Fig 6D ) . The expression of the nsP3 fusion proteins was similar for nsP3-FGN , nsP3-FGC and nsP3-FGNC while expression of wild type nsP3 was lower . Yet after co-IP , G3BP was only enriched for wild type nsP3 , despite the presence of less nsP3 in the co-IP fraction . G3BP was not enriched for nsP3-FGN , nsP3-FGC or nsP3-FGNC , suggesting that nsP3 requires duplicate FGDF motifs to establish high-affinity for G3BP in mammalian cells . Next , Aag2 cells were transfected with PUB-driven plasmids expressing the nsP3EGFP fusion proteins nsP3 , nsP3-FGC , nsP3-FGN or nsP3-FGNC and co-transfected with pPUB-Rin-mC , encoding a Rin-mCherry fusion protein ( Fig 6A ) . Cells were lysed , subjected to co-IP with α-GFP beads followed by detection of nsP3EGFP and Rin-mC in the cell lysates and precipitated samples ( Fig 6E ) . Expression of the nsP3EGFP fusion proteins was similar for nsP3 , nsP3-FGN , nsP3-FGC , nsP3-FGNC . Expression of Rin-mC was similar for nsP3 , nsP3-FGN , nsP3-FGC transfected samples , while Rin-mC expression was highest in cells transfected with nsP3-FGNC . Similar amounts of nsP3 , nsP3-FGN and nsP3-FGC but more nsP3-FGNC was precipitated . Interestingly , however , Rin-mC was only enriched for nsP3 , nsP3-FGN and nsP3-FGC , but not for nsP3-FGNC . This demonstrates that nsP3 requires only one FGDF motif for interaction with Rin . These results suggest that differences in the affinities of FGDF motifs for G3BP and Rin may underlie the difference for the required number of FGDF motifs between mammalian and mosquito cells .
The delicacy of alphavirus-vector interactions has been well-illustrated by the discovery that a single A→V mutation in the CHIKV E1 envelope protein could change the viral vector specificity from Ae . aegypti to Ae . albopictus [5] . Additionally , for VEEV a single amino acid substitution in the E2 protein could determine its potential to infect Ochlerotatus taeniorhynchus mosquitoes [41] . Here , we investigated the importance of the conserved P-rich and FGDF motifs in the HVD of CHIKV nsP3 for infection of mosquito cells and transmission by Ae . aegypti mosquitoes . Deletion of the P-rich motif in the nsP3 HVD renders CHIKV unable to replicate in both mammalian and mosquito cells , while amino acid substitutions of the core residues in the P-rich motif resulted in a significant decrease of replication in both cell types . This is similar to previous findings with SINV where deletions in the SINV nsP3 HVD that disrupt the P-rich motif were reported to decrease growth kinetics and plaque size in C7/10 mosquito cells [31] , an indication for reduced replication potency or cytopathicity , or both . However , mutations of the P-rich motif did not affect CHIKV infection or transmission by Ae . aegypti mosquitoes . The nsP3 P-rich motif has high affinity for SH3-domain containing proteins and is known to interact with amphiphysin-I/II [18 , 23] in mammalian cells , and proposedly also with the Drosophila homolog of amphiphysin [18] , although no evidence has so far been reported for this interaction in insect cells . We indeed demonstrate that mutation of the P-rich motif results in decreased virus production in mosquito cells , which suggests that nsP3 interacts with mosquito SH3-domain containing proteins , although these interactions do not appear to be essential for transmission by the mosquito vector . It was previously shown that deletions of the P-rich motif of SFV and SINV result in decreased affinity for amphiphysins and lowered replication efficiencies in mammalian cells [18] . Therefore , the decrease in replication of CHIKICP398A and CHIKICPPR401AAA in mammalian cells may be due to reduced affinity for amphiphysins or other cellular proteins that contain an SH3 domain . Of note , both the P398A and PPR401AAA mutations do not mutate all prolines in the CHIKV PxxPPR motif , thus the remaining prolines may still partially interact with host proteins . This may underlie the strong attenuation of CHIKICP398A and CHIKICPPR401AAA after low MOI infection , but not after high MOI infection where the remaining proline ( s ) may be sufficient to establish successful replication . Complete inactivation of CHIKV after deletion of the entire P-rich motif implies that such interactions are crucial to establish successful virus infection . It has been demonstrated that the arginine residues downstream of the P-rich motif of CHIKV , which are not conserved in SFV and SINV nsP3 , partially determine the affinity for amphiphsyin-2 [23] . This might explain why deletion of the whole P-rich motif , including the C-terminal arginines downstream of PxxPPR completely abrogates CHIKV replication , while mutating just the core residues in the P-rich motif results in a less dramatic effect . Mutation of both FGDF motifs together fully abrogated CHIKV replication in both mammalian and mosquito cells . This indicates that the FGDF motifs not only play an important role during infection of mammalian cells , as previously reported [17 , 19–22 , 30] , but that they also determine the successful infection of mosquito cells . Furthermore , we show for CHIKV that a single FGDF motif is sufficient to infect mammalian cells , although the replication and infection efficiencies are dramatically decreased compared to wild type CHIKV . In contrast , in mosquito cells , CHIKIC-FGN and CHIKIC-FGC replicated similarly to the wild type CHIKIC and were efficiently transmitted by Ae . aegypti mosquitoes after an infectious bloodmeal . This indicates that CHIKV requires only a single FGDF motif for infection of the mosquito vector . Thus , duplicate FGDF motifs appear to be required for the efficient infection from the mosquito’s saliva to a subsequent vertebrate host , which supports the conservation of duplicate FGDF motifs for most mosquito-borne alphaviruses . The mosquito protein Rin , a homolog of mammalian G3BP , was previously identified as an important regulator of CHIKV transmission , as knock-down of Rin severely decreased the transmission of CHIKV by Ae . albopictus mosquitoes [34] . In our study , we confirmed that , during CHIKV infection , a single FGDF motif is sufficient for the co-localization of nsP3 with Rin , similar to previous experiments with transiently expressed nsP3 [34] . Mutation of both FGDF motifs in nsP3 together abolished its ability to co-localize and interact with Rin , suggesting that the inability of CHIKVIC-FGNC to replicate in mosquito cells is due to absence of the nsP3-Rin interaction . Furthermore , we demonstrate that only one FGDF motif is required for the interaction between nsP3 and Rin , in contrast to mammalian cells where the interaction between nsP3 and G3BP was only observed for wild type nsP3 and not for nsP3 with one mutated FGDF motif . For SFV nsP3 it has been shown that mutation of just the phenylalanine of FGN largely abrogates the interaction with G3BP while mutating the phenylalanine of FGC only decreases the affinity of nsP3 for G3BP [20] . We mutated not only the first phenylalanine of FGC but also the glycine to alanine , both of which have been shown to be important for G3BP binding [40] . This may explain why we could not observe an interaction between G3BP and either nsP3-FGN or nsP3-FGC . G3BP can be present in monomeric , dimeric and polymeric forms [42] and the two FGDF motifs of alphavirus nsP3 each bind one NTF2-like domain of two separate G3BP monomers [20 , 40] . As a result , mutation of one FGDF motif results in decreased affinity for G3BP and attenuates alphavirus replication [17 , 20 , 21] . It is uncertain whether mosquito Rin forms a dimer , although the crystal structure of the Drosophila Rin NTF2-like domain was demonstrated in dimeric form [43] . Thus , a single FGDF motif might be sufficient for the interaction with monomeric Rin to support CHIKV replication in mosquito cells and in Ae . aegypti mosquitoes . Alternatively , the FGDF motifs of nsP3 might have a higher affinity for Rin as compared to G3BP , resulting in sufficient recruitment of Rin to replication complexes in mosquito cells with only a single FGDF motif . As Vero and Aag2 cells were cultured at 37°C and 28°C respectively , lowered affinity of the FGDF motifs for G3BP as compared to Rin may partially be explained by the higher incubation temperature . Potentially , the higher temperature at which mammalian Vero cells are cultured decreases the stability of the nsP3-G3BP interaction , thus resulting in lower affinity of nsP3 for G3BP as compared to Rin . The exact reason for recruitment of Rin/G3BP by nsP3 is still speculative , although several hypotheses such as recruitment of viral gRNA followed by shielding of the viral replication complex from degrading enzymes [17] and prevention of antiviral stress granule formation [13 , 21] have been brought forward . Possibly , the interaction with G3BP/Rin is particularly crucial early after infection and a certain level of G3BP/Rin has to be bound by nsP3 to successfully form viral replication complexes . Interestingly , while arthritogenic-alphaviruses recruit G3BP via their FGDF motifs , neurotropic alphaviruses lack FGDF motifs and instead recruit members of the Fragile X syndrome ( FXR ) family via a different motif in the nsp3 HVD [17 , 19] . Proteins of the FXR family are abundantly expressed in neuronal cells [44] , which may therefore be more susceptible to infection by alphaviruses that use FXR family proteins for their replication . Whether a direct correlation exists between the use of FXR or G3BP for replication and the symptoms caused by alphavirus infection should be the focus of future research . Mosquitoes possess a homolog of one FXR family protein , called Fragile X syndrome-related protein 1 ( FMR1 ) . Based on our results , we hypothesize that the nsP3 of neurotropic alphaviruses may interact with mosquito FMR1 , similarly to the CHIKV nsP3-Rin interaction . Future research should investigate the possibility of an interaction between nsP3 and mosquito FMR1 and its importance for the transmission of neurotropic alphaviruses by mosquitoes . Concluding , our results increase the understanding of the functionality of conserved motifs in the alphavirus nsP3 HVD in the context of the arboviral dual-host life cycle . Understanding the underlying mechanisms of G3BP/Rin recruitment by alphavirus nsP3 remains the focus of future research and may well contribute to the development of antiviral drugs and compounds that target this crucial interaction . | Chikungunya virus ( CHIKV ) is a re-emerging arthropod-borne virus that is transmitted predominantly by Aedes aegypti mosquitoes . In 2016 alone CHIKV caused over 100 . 000 infections in South-America , exemplifying the impact of CHIKV disease . Previous research has suggested that the CHIKV non-structural protein 3 ( nsP3 ) may determine the infection of mosquitoes . NsP3 is known to interact with several host proteins through a conserved proline ( P ) -rich and duplicate FGDF motifs that are present in its C-terminal domain . Here we investigated the importance of these conserved motifs for the infection and replication of CHIKV in both Aedes mosquito cells and mammalian cells . Furthermore , we assessed the role of these motifs for the transmission by Ae . aegypti mosquitoes via infectious bloodmeal experiments . We show that mutation of the P-rich motif negatively affects the replication of CHIKV in both mammalian and mosquito cells . In contrast , mutating the P-rich motif did not affect the transmission by Ae . aegypti . Mutation of both FGDF motifs together completely inactivated CHIKV in mammalian and mosquito cells , while mutation of a single FGDF motif negatively affected replication only in mammalian cells . Importantly , CHIKV containing only a single FGDF motif was still efficiently transmitted by Ae . aegypti mosquitoes . These results contribute to understanding the key interactions between alphaviruses and their mosquito vector . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"invertebrates",
"vero",
"cells",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"chikungunya",
"infection",
"pathogens",
"biological",
"cultures",
"tropical",
"diseases",
"microbiology",
"saliva",
... | 2018 | Conserved motifs in the hypervariable domain of chikungunya virus nsP3 required for transmission by Aedes aegypti mosquitoes |
DNA sequence variation causes changes in gene expression , which in turn has profound effects on cellular states . These variations affect tissue development and may ultimately lead to pathological phenotypes . A genetic locus containing a sequence variation that affects gene expression is called an “expression quantitative trait locus” ( eQTL ) . Whereas the impact of cellular context on expression levels in general is well established , a lot less is known about the cell-state specificity of eQTL . Previous studies differed with respect to how “dynamic eQTL” were defined . Here , we propose a unified framework distinguishing static , conditional and dynamic eQTL and suggest strategies for mapping these eQTL classes . Further , we introduce a new approach to simultaneously infer eQTL from different cell types . By using murine mRNA expression data from four stages of hematopoiesis and 14 related cellular traits , we demonstrate that static , conditional and dynamic eQTL , although derived from the same expression data , represent functionally distinct types of eQTL . While static eQTL affect generic cellular processes , non-static eQTL are more often involved in hematopoiesis and immune response . Our analysis revealed substantial effects of individual genetic variation on cell type-specific expression regulation . Among a total number of 3 , 941 eQTL we detected 2 , 729 static eQTL , 1 , 187 eQTL were conditionally active in one or several cell types , and 70 eQTL affected expression changes during cell type transitions . We also found evidence for feedback control mechanisms reverting the effect of an eQTL specifically in certain cell types . Loci correlated with hematological traits were enriched for conditional eQTL , thus , demonstrating the importance of conditional eQTL for understanding molecular mechanisms underlying physiological trait variation . The classification proposed here has the potential to streamline and unify future analysis of conditional and dynamic eQTL as well as many other kinds of QTL data .
Natural genetic variation affects gene expression levels and thereby impacts on molecular and physiological phenotypes such as protein levels , cell morphology or disease phenotypes . In this respect , gene expression has proven instrumental as an intermediate phenotype from which conclusions about the emergence of high level traits can be drawn . A genetic locus containing a sequence variant that affects transcript levels of a gene is called an expression quantitative trait locus ( eQTL ) . Studying eQTL has demonstrated its value for revealing the molecular mechanisms underlying disease associated SNPs , that were previously identified e . g . through genome wide association studies ( GWAS ) [1] , [2] . Moreover , it has been shown that eQTL SNPs are more likely to be disease causing than random genetic loci [3] and can thus be used to prioritize genetic markers in GWAS . Differences in mRNA expression levels caused by natural genetic variation can manifest themselves between individuals , populations , environments and , very importantly , between cell types and tissues ( see [4] , [5] and references therein ) . Since cells forming different tissues must have very different morphology , organization and function , distinct patterns of gene expression are required for each cell type . This variation of gene expression between cell types is under the influence of natural genetic variation . A number of studies ( summarized in Table 1 ) compared eQTL across different cell types and tissues in mouse and human samples and report that of the eQTL are cell type-specific . Potential reasons for the seemingly divergent outcomes of these studies are the different levels of relatedness of tissues under study and the different sample sizes of the studies . The last point is especially important in that cell type specificity is probably over-estimated due to low power of eQTL studies [4] , [6] . Nevertheless , there is clear evidence for cross-tissue differences in genetic variation influencing transcript levels . This raises the question whether conclusions drawn from an eQTL study in one cell type or even a cell line translate to other cell types . The answer to this question is obviously relevant for explaining disease mechanisms with eQTL studies that are conducted in tissues other than the disease tissue or when several cell types are involved in the disease etiology [7] . Most diseases are caused by a limited set of highly specialized cells , but cell- and tissue interactions are crucial for their etiology . Understanding the tissue and cell type-specificity of molecular traits is therefore essential for revealing the molecular mechanisms underlying disease phenotypes . Another layer of complexity is added when considering dynamic processes such as cellular differentiation or responses to internal or external stimuli . These changes go along with drastic alterations of the cell's morphology or molecular state being induced through the adaptation of gene expression patterns . Therefore , it is important to not only compare eQTL observed in individual cell types ( at steady state ) , but to additionally map the expression changes measured during cell state transitions . Intriguingly , the concepts of cell type-specific and differential eQTL have rarely been investigated together [8] . Hence , the main goals of the present study are to bring together and consolidate the different varieties of eQTL that have been proposed in the context of comparative eQTL mapping; to provide a thorough and functional classification of these eQTL classes reflecting the spectrum of genetic contributions to gene expression variation over a range of dynamically changing cell states; to show that these classes represent different sets of eQTL corresponding to different modes of expression variation and to demonstrate that their distinction facilitates the biological interpretation . A well-studied model for a dynamic process , being accompanied by substantial gene expression changes , is the differentiation of hematopoietic stem cells ( HSC ) into the different lineages of mature blood cells [9] . We decided to use this system to investigate eQTL based on three different categories of expression-based traits: ( i ) eQTL that are observed across all cellular states ( static eQTL ) , ( ii ) eQTL being specific to one or a subset of cell states ( conditional eQTL ) and ( iii ) eQTL affecting changes of transcript levels during differentiation ( dynamic eQTL ) . We propose strategies to map eQTL in the different classes and we demonstrate that eQTL from the above three classes , although based on the analysis of the same set of expression and genotype data , comprise different sets of regulatory loci having to be inferred from separate mappings . The choice of the eQTL mapping procedure has considerable influence on the outcome of the study . In particular , we show that basic cellular processes and state and differentiation specific functions are regulated by different eQTL categories . Although our scheme can serve to classify eQTL across any set of cell states , we will use the term cell type in the remainder of this paper , referring to the application to hematopoietic cell types .
We distinguish static , conditional and dynamic eQTL ( Figure 1 , Table 2 ) . A static eQTL affects a gene's expression in all conditions under consideration ( Figure 1A ) . It is independent of the cell type and will thus be detected in all cell types . In contrast , a conditional eQTL can be found in one or a subset of the conditions under consideration ( Figure 1B ) . In rare cases , a conditional eQTL might even be present in all four cell types . The difference between a static eQTL and a conditional eQTL active in all ( i . e . four ) cell types is the following: the static eQTL has the same effect throughout all cell types , whereas the conditional eQTL , although being active in all cell types , has effects dependent on the cell type . For example , the magnitude of the effect may differ between cell types or even the direction of the effect may change , i . e . , the major allele may yield higher expression levels of the target gene in one cell type and lower expression levels in another cell type . A third reason for the cell type dependence of conditional eQTL is that the effect may be dependent on different co-factors , i . e . there might be different epistatic interactions with other markers dependent on the cell type . Both static and conditional eQTL impact the absolute expression levels of their target genes in the given cell types . As opposed to that , dynamic eQTL drive changes in mRNA levels during the transition from one cell type to another and thus act on expression differences between cell types ( Figure 1C ) . Thus , the trait value used for mapping dynamic eQTL is the differential expression between two states or conditions ( in other words , we use the fold-change between two conditions as a trait value ) . In this respect our definition of dynamic eQTL differs from definitions used in the literature . For example , Gerrits et al . [10] define a dynamic eQTL as an eQTL that is present in one condition but not in another . We refer to those eQTL as conditional . A concept very similar to dynamic eQTL has been introduced in the context of studying transcriptional regulation in different growth conditions in yeast [8] . The authors define eQTL affecting expression changes between conditions as gene-environment interaction eQTL ( gxeQTL ) . A similar study has been conducted on differential expression in two different temperatures in worms [11] . Despite their application by several groups , the three different eQTL classes have never been mapped and compared in one single study . Different computational means can be used to detect the three eQTL types defined above . Dynamic eQTL require mapping of the expression changes ( fold changes , slopes ) observed at the transition from one type to another ( [8] , [11] , Table 2 , Methods Section “Dynamic eQTL mapping” ) . Conditional eQTL may be detected through independently mapping eQTL in the various cell types and then identifying such eQTL that were found in some , but not all conditions . Such an approach requires defining two thresholds: first a significance threshold ( e . g . maximum p-value ) for calling eQTL that are active in one cell type and second , an insignificance threshold ( e . g . minimum p-value ) for deciding that the same eQTL is not active in other cell types . Note that both thresholds are required and that they have to be sufficiently different . Using just one threshold would lead to a situation where all eQTL that are just above the threshold in one cell type and just below the threshold in other cell types would be called “conditional” although the eQTL scores are very similar across all conditions . Here we propose a different approach that we termed “simultaneous mapping” , because it simultaneously identifies static and conditional eQTL and because it simultaneously uses the expression data from all conditions ( Table 2 ) . The goal of simultaneous eQTL mapping is to infer eQTL that are specific for each of the cell types under study ( conditional eQTL ) as well as static eQTL in one single analysis . Static eQTL should lead to expression patterns that are similar across conditions . Combining expression data from all conditions in a single mapping therefore drastically increases the statistical power for detecting static eQTL . To this end , we combined gene expressions over all cell types into one trait vector ( Figure 2 ) . This resulted in a single matrix containing the expression values of all genes and individuals across all conditions . In order to get a matching genotype matrix , we replicated the genotype matrix as many times as there are cell types . Because not all individuals ( mouse lines ) were measured under all conditions , we had to subset the genotype matrices to the samples for which gene expression data was available . The resulting matrices were concatenated in order to obtain a predictor matrix matching . Finally , we added one new predictor for each cell type indicating whether a given sample belongs to the respective cell type . These additional variables allow to relate eQTL to the cell types in which they are active . Next , the eQTL mapping was conducted using Random Forests ( RF ) [12] , a multivariate machine learning technique that has been successfully tested on and applied to a number of QTL studies before [13]–[23] and that has been shown to outperform traditional univariate mapping approaches on simulated and real data [24]–[28] . RF learns decision trees based on bootstrap samples of the data . Genetic markers are used as predictor variables and RF will select markers if they are predictive for the expression of a given gene . Thus , the selection frequency can be used as a measure for the strength of an eQTL [25] . In case of static eQTL , a marker will be predictive of expression irrespective of the cell type . Thus , it will be predictive across the whole vector . In the case of conditional eQTL , the marker will be predictive on only a subset of the samples , namely those corresponding to the cell type ( s ) in which the eQTL is present . Because the cell type indicator variables are part of the predictor matrix , RF can “split” the samples on such indicator variable and subsequently identify markers that are predictive for expression in the respective cell type . In both cases ( static and conditional ) such markers will have high selection frequencies , allowing them to be detected through appropriate permutation tests ( Methods ) . In order to determine if a significant eQTL is static or conditional we exploited interactions between markers and cell type indicators . Using ANOVA we tested if the predictive value of a marker depends on the cell type variable:In this model , denotes the genotype vector of marker , denotes the cell type factor variable with as many levels as there are types , denotes their interaction and is a vector of normally distributed errors . The interaction term reflects the dependence of the eQTL on the respective cell type . A static eQTL should not interact with the cell type variable since its activity is ubiquitous and does not depend on the cell type of the sample . On the other hand , conditional eQTL are active in one or a subset of the measured conditions and thus will show a significant interaction with the cell type in which they affect their targets . In this case , the model including the interaction term should explain the gene expression significantly better than a reduced model containing only main effects . If this is the case , i . e . if the False Discovery Rate ( FDR ) of the ANOVA is , we call the eQTL “conditional” . Subsequent testing of contrasts can then identify the relevant cell types ( Methods ) . Overall , simultaneous eQTL mapping allows to discover static and conditional eQTL in one single analysis , thus reducing the multiple hypothesis testing problem as well as the computation time and rendering the choice for an “insignificance” threshold unnecessary . The approach of combining data over cell types also increases the power to detect static eQTL . Dynamic eQTL cannot be inferred with this approach since they are associated with a different trait , namely relative expression changes between cell types . Therefore , we analyzed dynamic eQTL in a separate mapping of gene expression differences using the same RF framework . Hematopoietic stem cell ( HSC ) differentiation is a prominent example of a dynamic process that is heavily genetically regulated [9] , [29]–[32] . This has been shown , among others , by analyzing natural genetic variation between mouse recombinant inbred lines exhibiting very different hematopoietic phenotypes [33] , [34] . One of the best studied examples is the panel of BXD recombinant inbred lines that were derived from crossing the C57BL/6 and DBA/2 lines . We are using genome-wide mRNA expression levels measured in 25 BXD strains in four cell types of HSC differentiation with varying degrees of lineage commitment: multipotent HSC with the potential for self-renewal , lineage restricted erythroid-myeloid progenitor cells , and lineage committed erythroid as well as myeloid cells ( cf . scheme in upper right corner of Figure 3 and [10] ) . We applied the above eQTL classification scheme to systematically search for genetic regions affecting gene expression dynamics during hematopoiesis as well as the static and conditional variation of expression in the different cell types . Using the data from [10] , we focused on three cell type transitions during HSC differentiation: from stem to progenitor cells ( S-P ) , from progenitor to erythroid cells ( P-E ) as well as from progenitor to myeloid cells ( P-M ) . Prior to the analysis , we summarized the mRNA expression measurements to the gene level by calculating the median expression profiles across probes . After preprocessing ( Methods ) we selected 849 markers and expression data of 14 , 724 genes in 22 to 24 BXD strains per cell type . Our simultaneous eQTL mapping detected 3 , 916 significant eQTL target gene pairs at an FDR of 0 . 1 . Among those , 2 , 729 eQTL did not show a significant interaction with the cell type indicator and thus constitute the class of static eQTL . We also found 1 , 187 conditional eQTL . These eQTL have to fulfill two conditions: ( i ) simultaneous mapping and ( ii ) FDR for interaction between marker and cell type indicator . The majority of conditional eQTL was active ( significant ) in only one cell type ( Figure 3 ) . However , we also observed conditional eQTL being active in two , three , or even four cell types . eQTL with four significant cell type interactions arise if an eQTL is active in all cell types , but with changing effect sizes . Hence , a conditional eQTL active in four cell types is distinct from a static eQTL . Most of the eQTL that are conditional in exactly one cell type ( “cell type-specific” ) occur in the more committed lineages ( 218 in erythroid cells , 206 in myeloid cells , Figure 3 ) . We find less eQTL in the multipotent stem cells ( 176 ) and the smallest number of eQTL ( 43 ) in progenitor cells , an observation that is consistent with the original presentation of the data [10] . Likely , this reduced number of eQTL is due to increased levels of noise in the data , which in turn might be caused by different effects . First of all , purification of the cell types using FACS is imperfect . Thus , the observed expression levels actually reflect expression in a heterogeneous mix of cells . Increased impurity would then increase the level of noise and thus likely decrease the number of eQTL being detected . Another explanation comes from the fact that the progenitor cells are in a transient state . I . e . , the dynamic nature of these cells might induce additional heterogeneity , which then also increases the noise and decreases the power to detect eQTL . In contrast to the large number of static and conditional eQTL , we detected very few dynamic eQTL . At an FDR of 0 . 1 there were six eQTL driving gene expression changes during the transition from progenitor to erythroid cells and 66 eQTL for the transition from progenitor to myeloid cells . Two of the eQTL in these two groups are identical , i . e the same loci ( both in cis ) affect the same target genes during both , the P-E and the P-M transition . These targets are Gadd45gip1 and Lrrc51 . We were not able to find any dynamic eQTL in the transition from stem to progenitor cells . Obviously , dynamic eQTL might overlap with conditional eQTL ( “overlapping” means that the eQTL link the same locus-target gene pair , Figure 4 ) . To facilitate comparison of conditional eQTL obtained with different mapping approaches ( see Discussion ) , eQTL that are detected in exactly one cell type ( i . e . cell type-specific eQTL ) are shown as a subgroup of conditional eQTL . By definition , there is no overlap between conditional and static eQTL . As expected , none of the 70 dynamic eQTL overlap with static eQTL , while 45 coincide with conditional eQTL . Intriguingly , 25 loci that influence the dynamics of gene expression during the transition from one cell type to another ( of all dynamic eQTL ) could not have been detected by the simultaneous mapping , i . e . these eQTL did not overlap with eQTL from any other class [8] . An eQTL can either act locally ( in cis ) or on a distant gene ( in trans ) . That is , the target gene of a cis-eQTL is encoded in the eQTL-region . A trans-eQTL refers to eQTL affecting a gene encoded elsewhere in the genome . Such influence can only be explained by trans-acting factors . Around ( 244 ) of the static eQTL are cis-eQTL ( left-hand side of Figure 5 ) . It is noteworthy that the number of static and conditional cis-eQTL is relatively similar , whereas we find substantially more static than conditional trans-eQTL ( Figure 5 ) . The statistical power for detecting static eQTL is much higher than the power for detecting conditional eQTL in the framework of simultaneous eQTL mapping . This is because additional statistical power is needed for detecting significant differences between the cell types . References [10] and [35] , among others , have shown that cis-eQTL are linked very strongly with their target genes while the effects of trans-eQTL are often weaker and several trans-eQTL are needed to explain the expression variation of a distant target gene . Hence , the increased power in case of static eQTL leads to an increased number of detectable trans-eQTL , whereas cis-eQTL seem to be “saturated” already at lower power . We confirmed this interpretation by varying the number of samples considered in the analysis , which showed that increasing the number of samples increased the number of detectable trans-eQTL more than the number of detectable cis-eQTL ( Figure S1 ) . This observation has two implications: first , the total number of cis-eQTL seems to be limited in this mouse population and second , it is possible to detect most cis-eQTL with a relatively small number of strains . Dynamic eQTL comprise a much larger fraction of cis-eQTL compared to simultaneous eQTL ( , Figure 5 ) . This is not surprising considering the fact that dynamic eQTL depend on gene expression measurements in two cell types at a time . They are thus more vulnerable to noise , but at the same time they have to be inferred from only one fourth of the samples available for the simultaneous mapping . Hence , we might only catch the strongest effects here , which are often found in cis [35] . A comparison of the results of our analysis with the original results from [10] reveals considerable differences between both studies ( Figure 6 ) , which are caused by the different mapping approaches . First of all , the simultaneous mapping in combination with RF is able to capture many more ( probably small effect ) eQTL than a linear model [27] , [32] , [33] . However , since [10] based their results on the number of probes having at least one significant eQTL and we are reporting significant eQTL-target gene pairs , and since the number of significant eQTL depends on the chosen p-value or FDR thresholds , we decided to compare fractions of eQTL classes instead of absolute numbers . We restricted the comparison to the static and conditional eQTL classes , since there is no equivalent to dynamic eQTL according to our definition in the original paper . Our study detected a much larger fraction of static eQTL than the original paper ( ) owing to the larger power of simultaneous mapping to capture this class of eQTL . Note that such ratios will always depend on the power to map eQTL in the corresponding classes with a given approach . Therefore , all ratios that have been reported so far ( including our own ) suffer from statistical biases . We cannot claim that any of them reflects “biological truth” . Furthermore , the fraction of trans-eQTL is larger in our study compared to [10] ( , Figure 6 , center ) . This can again be explained by the ability of the simultaneous mapping with RF to detect more small effect eQTL than a linear model . In contrast , the fraction of cell type-specific eQTL from the four hematopoietic cell types is rather consistent between the two studies ( Figure 6 , rightmost bars ) . Interestingly , both studies detect only very few regulatory loci in progenitor cells , pointing to a general problem to detect specific regulatory relationships within this cell type . As mentioned before , this might be due to issues with the cell purification and the transient nature of this cell population . In order to show the conditionality of certain regulatory regions , we selected loci containing a larger number of eQTL-target pairs and tested their enrichment for conditional eQTL of a specific ( subset of ) cell types . This analysis is independent of the fact whether the given region has significantly more target genes than expected by chance as long as there are enough targets to be tested for conditionality . Therefore , we refer to these regions as “eQTL-rich regions” . The visualization of all cell type-specific and static eQTL in an eQTL map ( Figure 7 ) reveals putative cell type-specific eQTL hotspots . A Friedman test for differences in the distributions of contrast test p-values of the targets of such eQTL-rich regions uncovered some eQTL that have an effect on many genes in specific cell types . An example of such a hotspot is a locus on chromosome 19 ( 52 . 3–55 . 2 Mb ) affecting 31 stem cell-specific and 59 static target genes . Even though only one third of the eQTL in this locus meet the significance threshold of a stem cell-specific eQTL , there is a clear tendency towards stem cell specificity for most of them ( Figure S3A , ) . The eQTL contains the gene Shoc2 for which we also find a cis-eQTL . We have previously shown that trans effects are often caused by genes being themselves affected through a cis-eQTL [36] , which makes Shoc2 a putative causal gene in the region . The protein encoded by this gene is a scaffold for a Ras/Raf interaction [37] . The Ras pathway is important for hematopoietic differentiation processes and frequently activated in hematopoietic malignancies [38] . However , we did not find any direct links between Shoc2 and its putative target genes . We found a second cell type-specific eQTL-rich region on chromosome 2 ( 168 . 3–169 . 7 Mb , ) , whose eQTL - target gene pairs are enriched for myeloid as well as stem cell-specific eQTL ( Figure S3B ) . One possible regulator gene in this locus is Nfatc2 ( nuclear factor of activated T cells ) , which is gradually down-regulated at certain stages during the differentiation from myeloid progenitors to megakaryocytes and neutrophils [39] . Several of the eQTL target genes are predicted to be functionally related to Nfatc2 [40] and many of them ( e . g . Ccdc99 , Cdk2 , Cdca8 , Birc5 ) are involved in cell cycle control . Indeed , it is known that Nfatc2 negatively regulates the expression of Cdk4 , which controls the entry and progression of a cell in the cell cycle [41] . In line with that , Cdk4 links Nfatc2 and its target genes in the STRING network . Although it has been shown that Nfatc2 is not required to block cell cycle entry , it is likely that it prevents HSCs from differentiation into neutrophils and megakaryocytes via an effect on their proliferation [39] , [42] . The importance of Nfatc2 for both the HSC and the myeloid cells is reflected by the lower cell type specificity p-values of its targets in both types ( Figure S3B ) and corresponds well to Nfatc2 expression levels that have been found to be high at the beginning of myeloid differentiation , go down during differentiation and finally increase again [39] . Static eQTL affect a gene's expression in all cell types . Therefore , we expect their target genes to have different , broader biological functions than genes affected by non-static eQTL . An example of such a static eQTL is an eQTL impacting on the expression of Peroxiredoxin-2 ( Prdx2 ) ( Figure 8A ) , a gene involved in the response to and protection of erythrocytes against oxidative stress [43] . It is one of the most abundant proteins in erythrocytes [44] , which is reflected in the elevated expression levels in erythroid cells compared to the other cell types . However , due to the severe impact of damage from oxidative stress on hematopoietic cell homeostasis in every cell type [45] , Prdx2 expression levels need to be controlled across all cell states . Since Prdx2 is encoded at the same locus as the eQTL itself , the expression differences between the eQTL alleles are probably due to a mutation in the gene itself or in a cis-regulatory region . Figure 8B shows the deacetylase Sirt2 as an example of a gene being target of a conditional eQTL . The expression of Sirt2 is strongly correlated with the alleles at the eQTL in erythroid cells , but not the other cell types . We found the expression of Sirt2 to be correlated with hematocrit levels in female mice ( data not shown ) . Thus , the eQTL indirectly affects hematocrit levels in mice through the regulation of Sirt2 . In line with its elevated expression levels in myeloid cells , there is first evidence that Sirt2 might be involved in myeloid differentiation [46] . Il12rb2 is an example of a gene being affected by a dynamic eQTL . The gene encodes for a transmembrane protein constituting one subunit of the Interleukin 12 receptor complex . Together with other colony-stimulating factors Interleukin 12 is involved in myelo- as well as erythropoiesis [47] , [48] . We find a dynamic eQTL for Il12rb2 in the differentiation from progenitor to myeloid cells , which is characterized by almost constant expression levels for strains carrying the D allele at the eQTL while mRNA levels increase for individuals carrying the B allele . The expression profiles of Il12rb2 in progenitor and myeloid cells indicate that the eQTL might actually be conditional in both cell types with very small and opposite effects . Hence , such switching allelic effects are an example of a situation where dynamic eQTL mapping has increased power compared to conditional mapping . Intuitively , one expects that a significant allele-dependent expression change from one to another cell type ( i . e . a dynamic eQTL ) will coincide with significant , allele-dependent expression in at least one of the two cell types involved in the transition ( i . e . a conditional eQTL ) . We often observed such coincidence ( Figure 4 ) and the cell cycle inhibitor Gadd45gip1 [49] is a particularly interesting example ( Figure 8D ) . Gadd45gip1 is one of only two genes for which we found a dynamic eQTL affecting the transition to both , erythroid and myeloid cells . The protein encoded by this gene physically interacts with Gadd45b , which is involved in cell growth arrest during myeloid cell differentiation [49] , [50] . Gadd45gip1 might support this function and arrest cell cycle in a particular phase in myeloid precursor cells , which is a prerequisite for differentiation [51] . Gadd45gip1 is up-regulated in stem and progenitor cells in samples carrying the D allele at the eQTL locus ( Figure 8D ) . The eQTL is in cis , suggesting that a mutation in the Gadd45gip1 gene itself or in its promoter region leads to decreased expression of the gene in individuals carrying the B allele . Accordingly , down-regulation of Gadd45gip1 during the transition to myeloid cells only occurs in samples carrying the D allele . This leads to a dynamic eQTL from progenitor to myeloid cells . Interestingly , individuals having high Gadd45gip1 levels in progenitor cells show a down-regulation of its expression during the transition to erythroid cells , while the gene is up-regulated in individuals with low Gadd45gip1 levels in progenitor cells . This leads to an expression equilibration in erythroid cells and to a dynamic eQTL . Thus , ( i ) compensatory feedback mechanisms can “revert” the effect of an eQTL and ( ii ) there seems to be a need to tightly control Gadd45gip1 expression in erythroid cells . In order to test more systematically whether cell type independent ( i . e . static ) eQTL impact on different cellular functions than conditional and dynamic eQTL , we assessed the enrichment of functional categories among genes causing eQTL and among genes being affected by eQTL using gene annotations obtained from Gene Ontology ( GO ) Biological Process [52] . Such GO enrichment analysis is non trivial for genetic regions causing eQTL , because these regions typically contain multiple genes and it is usually unknown which of them is the true causal gene [53] . Therefore , we decided to annotate each region with the GO terms of all associated genes ( see Methods ) . This rigorous solution avoids false positive GO enrichment due to local clusters of functionally related genes . The enrichment testing was conducted with the R package topGO [54] , which corrects for the nested structure of GO . Since we found only six significant dynamic eQTL for the differentiation towards erythroid cells , we did not perform GO enrichment for this subset of eQTL . The top 10 most significantly enriched GO terms for each eQTL mapping are reported in Tables S1 , S2 , S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 , S11 , S12 . Figure 9 shows exemplary results of the enrichment distinguishing cell type-specific , dynamic and static eQTL . Static eQTL are enriched for very generic functional categories such as translation , transcription and cell cycle regulation . As opposed to that , conditional eQTL are enriched for hematopoiesis-related functions: For example , stem cell eQTL targets are enriched for the term “cell migration involved in sprouting angiogenesis” , in which HSCs play an important role [55] . Myeloid progenitor cell eQTL are enriched for the generic immune term “myeloid leukocyte mediated immunity” , while conditional eQTL in myeloid cells are enriched for very specific immune response terms like “defense response to Gram-negative bacterium” . We found several GO terms related to MAP kinases enriched among eQTL in erythroid and myeloid cells . This family of serine/threonine kinases plays a crucial role in maintenance and differentiation of HSC , especially during erythropoiesis [56] . Dynamic progenitor-myeloid eQTL are specifically enriched for categories related to T cell selection . This could be an indirect effect related to the role of macrophages and dendritic cells , which belong to the myeloid lineage , in adaptive immunity . These cells are involved in presenting antigens bound to the major histocompatibility complex ( MHC ) to naive T cells in order to activate or suppress these cells [57] . Accordingly , we find MHC coding genes among the dynamic eQTL targets . This analysis shows that static , conditional , and dynamic eQTL affect functionally distinct classes of genes and it therefore underlines the need to distinguish these types of eQTL . It has been suggested that eQTL might help better understanding the molecular mechanisms underlying the variation of physiological traits ( i . e . causing “physiological QTL” ) . This notion is based on the observation that expression variation is underlying the variation of many physiological traits [7] . Indeed , eQTL studies have already demonstrated their value for the prioritization of disease associated SNPs [3] , [58] , [59] . Moreover , some of these studies have shown that there exists an association between the disease and the tissue in which the eQTL was found [5] . These findings suggest that knowledge about the eQTL class and ( in the case of conditional and dynamic eQTL ) the tissues in which it is detected might further improve our understanding of the molecular mechanisms causing the disease symptoms . In order to investigate the impact of eQTL conditionality on physiological trait QTL , we analyzed the representation of different eQTL classes among QTL affecting hematological phenotypes ( downloaded from www . genenetwork . org [60] ) . Out of 91 hematological phenotypes available , we selected 13 traits for which we found at least one significant QTL ( ) based on at least 15 BXD strains . In total , we found 17 QTL associated with those 13 physiological traits and further investigated all 15 of those with at least one significant eQTL linking to the corresponding QTL region ( Tables S13 , S14 , S15 , S16 , S17 , S18 , S19 , S20 , S21 , S22 , S23 , S24 , S25 , S26 , S27 ) . One QTL affected two very similar traits ( “transferrin saturation of males and females” and “transferrin saturation of females” ) . Therefore , we counted it as one QTL in all subsequent analyses . We found that the eQTL linking to these regions were enriched for cis-eQTL ( of the QTL and of all regions contain a cis-eQTL ) , which was associated with an increased number of conditional eQTL ( of all eQTL - target gene pairs in these loci were conditional , compared to overall ) . The cell types in which these cis-eQTL were active , were often related to the respective cellular phenotype , suggesting that indeed these cis-eQTL are underlying the physiological changes . For example , we found five cis-eQTL in a region affecting hemoglobin levels in female mice ( Table S13 ) . Based on their known function , only two of the respective genes were plausible candidates for actually affecting hemoglobin levels: E2f1 and Asxl1 [61] , [62] , where the latter apparently has only very mild effects . Consistent with this , E2f1 was the only gene among those five having a specific , conditional cis-eQTL in erythroid precursors , the cell type most closely related to the hemoglobin phenotype . Thus , the consideration of cell-type specific eQTL facilitates the identification of plausible candidate genes .
The difference between static and non-static eQTL was very striking in our analysis . Due to the increased statistical power resulting from the simultaneous mapping we could identify substantially more static than non-static eQTL . Further , static and non-static eQTL differed substantially with respect to the functions of the involved genes , regarding both regulators ( i . e . loci ) and target genes . Whereas static eQTL involve mostly genes with generic , unspecific functions , non-static eQTL affect more cell type-specific pathways . We found relatively few dynamic eQTL , ranging from zero ( stem to progenitor cells ) to 66 ( progenitor to myeloid cells ) per cell type transition . This is not very surprising given the fact that expression differences are prone to increased noise since they “inherit” the independent errors of expression experiments in two different conditions [63] . We would also expect a large overlap between conditional and dynamic eQTL . If there is a dependency between gene expression levels and genotype in one but not another cell type , then the magnitude of expression change between these cell types ( i . e . the slope ) should be genotype-dependent as well . However , we only find 45 eQTL as belonging to both , the conditional and the dynamic class , while 1 , 142 and 25 eQTL are exclusively conditional and dynamic , respectively . One reason for this observation is the reduced power of the dynamic mapping leading to a failure to replicate conditional eQTL . Intriguingly , we also detect dynamic eQTL that we do not find among the conditional eQTL . Thus , there are modes of expression variation that are detectable with higher power when mapping expression differences instead of absolute expression levels . For example , we find eQTL with swapping effects on transcript levels ( such as Il12rb2 , Figure 8C ) among 10 out of the 25 eQTL-target gene pairs that are unique in the dynamic class . This emphasizes the need to include different expression traits ( like expression differences ) into a comprehensive eQTL analysis in order to detect a wide spectrum of eQTL . Another notable feature of dynamic eQTL mapping is its ability to mitigate systematic measurement errors affecting all cell types in a similar way . In this respect , a score for relative expression change can still be meaningful even though the absolute expression levels were not [63] . The approach we proposed for mapping different classes of eQTL is only one of a palette of possible strategies . Since the focus of the present work was on the introduction of a comprehensive , coherent and functional eQTL classification , in particular the discussion of each classes' characteristics and its implications on biological interpretation of eQTL results , we did not comprehensively compare different approaches for eQTL mapping . However , we still tested several variants , in particular the aggregation of static and conditional eQTL from separate mappings in every condition , which is the most widely used approach for comparative eQTL studies in the literature ( see references in Table 1 ) . Comparative eQTL studies have so far mostly mapped eQTL separately in each cell type , subsequently classifying eQTL as “static” if they are significant in all mappings , otherwise as “cell type-specific” ( Table 1 ) . This approach leads to a situation very different from our simultaneous mapping: in separate mappings an eQTL has to be significant independently in each cell type in order to be classified as static . In other words , large power is needed to detect static eQTL . As opposed to that , in our approach the eQTL has to be significantly dependent on the cell type in order to be classified as conditional . Therefore , simultaneous mapping is more conservative with respect to calling conditional eQTL . As a consequence , eQTL obtained with these two mapping strategies overlap only partially ( Figure S2 ) , which is mostly owed to the fact that simultaneous eQTL mapping detects many more significant eQTL , the largest fraction of which are static . The advantage of the simultaneous mapping with Random Forests ( combined with an ANOVA to disentangle conditional eQTL ) instead of doing an ANOVA only is its ability to detect non-linear relationships . Therefore , the simultaneous mapping is able to detect a larger range of regulatory genetic variation than the simple linear model . The strategy we followed for mapping dynamic eQTL has an obvious counterpart for static eQTL , namely the mapping of mean expression levels over all conditions . However , when applying this approach to the four hematopoietic cell types , we noticed that a large fraction of the resulting static eQTL were in fact conditional eQTL in one or several types . The erroneous classification resulted from the fact that a strong cell type-specific effect may dominate mean expression levels . Thus , this approach is prone to detect false positive static eQTL and in our opinion is not well suited to classify static eQTL . The fact that we find of all simultaneous eQTL to be conditional for one or several cell types emphasizes the condition specificity of many regulatory relationships , even if the conditions under study are very related . Note that simultaneous mapping is conservative for calling conditional eQTL and the true fraction of conditional eQTL is most likely even higher . In addition , we find that the number of conditional eQTL differs between cell types , partly due to differences in sample size and tissue impurity , but maybe also due to functional differences . The particular importance of conditional eQTL for cell type-specific molecular traits was further demonstrated by a GO enrichment analysis of eQTL and their targets in different eQTL classes . Moreover , an integration of eQTL results with QTL affecting hematological phenotypes revealed that a large fraction of these physiological QTL conditionally affect the expression of genes in phenotype-related cell types and are enriched for cis-eQTL . It has previously been shown that eQTL causing variation of disease traits are often cis-eQTL [59] . Moreover , we and others have demonstrated that genes causing a trans-eQTL , i . e . affecting the expression of a distant target gene , often also exhibit a cis-eQTL affecting their own expression [9] , [36] . Our analysis of the BXD mice confirms that genes with cis-eQTL are more likely causal . Beyond that , our results underscore the biomedical relevance of the distinction of different eQTL classes that we propose here , especially the impact of conditional eQTL on cell type-specific molecular and physiological phenotypes [58] . Since genetic variation affecting physiological phenotypes is often linked to conditional eQTL , the detection of the molecular mechanism underlying the QTL association critically relies on the cell type in which the eQTL study is conducted . These findings call for due caution when drawing conclusions about regulatory mechanisms in one condition based on results from another condition [58] , although other groups have claimed the innocuousness of such an approach [59] . A typical example for such a propagation of results would be the use of molecular mechanisms derived from eQTL studies in blood samples to explain disease mechanisms in other tissues like the brain [64] . The use of eQTL results for the elucidation of disease etiology is further complicated by the fact that the onset of complex diseases often involves pathways in several tissues . Increasing statistical power by simultaneous mapping and distinguishing static , conditional and dynamic eQTL are important steps towards accounting for tissue and cell-type specificity , which is key for elucidating the molecular alterations underlying changes of complex physiological and disease traits [7] . The classification of regulatory genetic variation is of course not limited to expression phenotypes . Almost all traits under genetic control ( such as protein abundance , phosphorylation , alternative splicing and disease phenotypes , to name but a few ) are dynamically regulated and depend on the specific context of the cell . Therefore , our classification scheme will be readily applicable to many other QTL studies and has the potential to unravel the dynamics underlying many biological processes . The simultaneous mapping will be beneficial to investigate different kinds of QTL across conditions and might even be extended ( after appropriate data normalization ) to comparative analyses across different datasets in the same organism .
Preprocessed gene expression data of [10] were downloaded from GeneNetwork [60] ( http://www . genenetwork . org , accession numbers GN144–151 ) . The preprocessing comprised the transformation and subsequent joint quantile normalization of expression data from all four cell types ( HSCs , myeloid progenitors , erythroid and myeloid cells ) as well as a batch correction . We mapped Illumina probe IDs to Ensembl gene IDs using mapping information from GeneNetwork and the R [65] biomaRt package [66] and summarized expression measurements for each gene by calculating the median expression profile over all its probes . Finally , we discarded all genes with a standard deviation of less than 0 . 1 in all four cell types , resulting in expression measurements of 14 , 724 genes on 22 to 24 BXD strains , depending on the cell type . Genotype information of the BXD strains was also downloaded from GeneNetwork . Since we had expression information on only 25 strains , some neighboring genetic markers in the genotype matrix contained identical information ( i . e . they were perfectly correlated ) . It is impossible to distinguish these markers with respect to their association to gene expression traits in the eQTL mapping . Therefore , we merged neighboring markers with identical genotype profiles across strains , which resulted in genotype information on 849 distinct markers or marker intervals across the mouse genome with a median interval size of 1 . 5 Mb ( min: 4 . 6 kb , max: 32 . 1 Mb ) . To carry out eQTL mapping in all cell types simultaneously , each gene's expression vectors from all conditions are concatenated to form a new trait vector ( Figure 2 ) . Note that this vector might contain several entries for the same sample , each from a different cell type . Accordingly , genotype vectors belonging to each of the samples in each cell type are combined into a predictor matrix . Since we would like to distinguish static and conditional eQTL , we need to add additional predictors indicating whether a sample was measured in a certain cell type or not . Therefore , is extended by as many dummy variables as there are cell types . We use Random Forests ( RF ) [12] for mapping eQTL . RF is a machine learning approach based on an ensemble of decision trees , each predicting gene expression for a different bootstrap sample of the data by testing different subsets of predictors at each split . We use the selection frequency of each predictor across the forest as a measure of its importance for predicting mRNA levels . A marker that is used more often than expected by chance is an eQTL of the corresponding gene . p-values are calculated using a permutation approach , see Subsection “p-value calculation” . For each significant eQTL - target gene pair ( ) , we fit two linear models to the gene expression: a full model containing the eQTL genotype , a cell type factor variable with as many levels as there are cell types and an interaction term between the two variables; and a reduced model containing only the two main effects without their interaction . If the full model explains the gene expression significantly better than the reduced one ( ) , we call the eQTL “conditional” . The cell types in which the eQTL is active are found with post-hoc Wald tests ( [67] , chapter 1 . 3 . 3 ) . The resulting p-values are corrected for multiple hypothesis testing using the stringent Bonferroni correction [68] . In principle , the second step of the simultaneous eQTL mapping , the distinction between conditional and static eQTL , could be directly resolved in the primary eQTL mapping step . The RF framework allows to extract epistatic interactions between predictors directly from the trees [16] , [69]–[71] . However , this requires a large enough sample size in order to grow deep trees where different combinations of variables will be used for splitting in the same branch . When trying this line of action on the hematopoiesis data , it became clear that the small sample size ( 22 to 24 samples per cell type ) is prohibitive for this step , leading to rather unstable results . Hence , we used the remedy of applying an ANOVA to filter the conditional eQTL out of the set of simultaneous eQTL . We believe that with the improvements made on costs and quality of large sequencing studies and the further increase in computing power this approach will become feasible very soon . For mapping genetic loci driving expression dynamics between two cell types , we create a new trait vector containing the sample-wise expression differences of a given gene between the pair of cell types . The predictor matrix is made up of the marker genotype vectors of each sample for which expression changes could be inferred . We then conduct the eQTL mapping using RF as described for simultaneous eQTL in Subsection “Simultaneous eQTL mapping” . We use the RF selection frequency ( SF ) as a measure of the impact of each genetic locus on gene expression . We have previously shown that this importance measure outperforms classic measures like the permutation importance in eQTL mapping [25] . However , the raw SF itself is not an absolute indicator of the importance of each predictor since the SF is biased for certain markers even under the null hypothesis [25] . A simple solution to this problem is the calculation of p-values based on a permutation approach: The expression vector is permuted many times . For each permutation , the eQTL mapping with the calculation of SFs is repeated . We assume that under the null hypothesis of no correlation between a given marker and a gene's expression , the distribution of SFs of that marker is the same for all genes . Hence , we pool SFs of each marker over all genes and all permutations in order to obtain a specific null distribution of SFs for each marker . Finally , the p-values of an eQTL - target gene pair are calculated as the fraction of permutation SFs exceeding the observed SF . The bottleneck of this approach is the run-time of the RF , strongly restricting the number of permutations , which in turn results in a rather low resolution of the eQTL p-values , even after pooling SFs over genes . In order to overcome this problem , we decided to combine the permutation procedure with an analytical p-value calculation . After pooling SFs over a small number of permutations ( 10 in all our eQTL mappings ) , we fit an exponential function to the top one percent of the SF distribution . Consequently , we can calculate more precise p-values for the tail of the observed SF distribution , which contains the interesting eQTL - target gene pairs . The remaining of the p-values are still obtained from the empirical SF distribution as described before . FDR is calculated with the procedure of Benjamini and Hochberg [72] . We tested for the enrichment of certain biological functions among eQTL regions and target genes . We used Gene Ontology ( GO ) Biological Process [52] gene annotation , which we retrieved from the Ensembl database release 66 ( www . ensembl . org ) via the biomaRt [66] interface of R . eQTL loci were annotated with the functions of all genes encoded in the locus or being closer to this locus than to any other ( if not more than 1 cM away from it ) . This approach ensures a conservative evaluation of functional enrichment and prevents a bias in the results due to clusters of functionally related genes within a locus . The GO enrichment testing was conducted using topGO [73] with the “weight” algorithm ( R package topGO [54] ) . Although topGO already accounts to some extent for multiple hypothesis testing , we further calculated an empirical FDR for each term based on a shuffled gene/eQTL region to GO term assignment , preserving the number of terms assigned to each gene/region . We call all terms with an significant . | Complex physiological traits are affected through subtle changes of molecular traits like gene expression in the relevant tissues , which in turn are caused by genetic variation . A genetic locus containing a sequence variation affecting gene expression is called an expression quantitative trait locus ( eQTL ) . Understanding the tissue and cell type specificity of eQTL effects is essential for revealing the molecular mechanisms underlying disease phenotypes . However , so far the cell-state dependence of eQTL is poorly understood . In order to systematically assess the importance of cell state-specific eQTL , we propose to distinguish static , conditional and dynamic eQTL and suggest strategies for mapping these eQTL classes . We applied our framework to mouse gene expression data from four hematopoietic stages and related cellular traits . The different eQTL classes , although derived from the same expression data , represent functionally distinct types of eQTL . Importantly , conditional eQTL are well correlated with relevant hematological traits . These findings emphasize the condition specificity of many regulatory relationships , even if the conditions under study are related . This calls for due caution when transferring conclusions about regulatory mechanisms across cell types or tissues . The proposed classification will also help to unravel dynamic behaviors in many other kinds of QTL data . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"dna",
"modification",
"gene",
"expression",
"genetics",
"population",
"genetics",
"biology",
"computational",
"biology",
"gene",
"networks",
"dna",
"transcription"
] | 2013 | Impact of Natural Genetic Variation on Gene Expression Dynamics |
The malaria parasite Plasmodium falciparum invades , replicates within and destroys red blood cells in an asexual blood stage life cycle that is responsible for clinical disease and crucial for parasite propagation . Invasive malaria merozoites possess a characteristic apical complex of secretory organelles that are discharged in a tightly controlled and highly regulated order during merozoite egress and host cell invasion . The most prominent of these organelles , the rhoptries , are twinned , club-shaped structures with a body or bulb region that tapers to a narrow neck as it meets the apical prominence of the merozoite . Different protein populations localise to the rhoptry bulb and neck , but the function of many of these proteins and how they are spatially segregated within the rhoptries is unknown . Using conditional disruption of the gene encoding the only known glycolipid-anchored malarial rhoptry bulb protein , rhoptry-associated membrane antigen ( RAMA ) , we demonstrate that RAMA is indispensable for blood stage parasite survival . Contrary to previous suggestions , RAMA is not required for trafficking of all rhoptry bulb proteins . Instead , RAMA-null parasites display selective mislocalisation of a subset of rhoptry bulb and neck proteins ( RONs ) and produce dysmorphic rhoptries that lack a distinct neck region . The mutant parasites undergo normal intracellular development and egress but display a fatal defect in invasion and do not induce echinocytosis in target red blood cells . Our results indicate that distinct pathways regulate biogenesis of the two main rhoptry sub-compartments in the malaria parasite .
Malaria is a devastating disease of tropical and subtropical regions . Requiring a mammalian host and a mosquito vector for transmission , at least six species of the genus Plasmodium cause disease in humans , with Plasmodium falciparum being responsible for the great majority of mortality . All the manifestations of clinical disease result from repeated cycles of invasion , replication within and lytic egress from red blood cells ( RBC ) . Invasion is an orchestrated process , comprising several steps including merozoite attachment , deformation of the RBC membrane , merozoite reorientation , formation of a high affinity interaction between the apical zone of the merozoite and the RBC surface , active entry , and finally sealing of the RBC membrane behind the intracellular parasite [1–3] . Invasion is generally immediately followed by a period of transient RBC echinocytosis , a morphological transformation of the RBC surface into an undulated or ‘spiky’ appearance , although this can also be induced under certain conditions even in the absence of successful invasion [3–5] . Entry into the host cell occurs concomitantly with formation of a membrane-bound parasitophorous vacuole ( PV ) within which the invading parasite comes to rest . The parasite then transforms within minutes into a ‘ring’ stage form before initiating intracellular development , progressing through a mononuclear trophozoite stage to a multinucleated schizont which undergoes segmentation to form a new generation of daughter merozoites . Parasite-induced rupture of the PV membrane ( PVM ) and host RBC membrane eventually enables egress of the merozoites to initiate a fresh erythrocytic cycle . Egress and invasion involve the regulated discharge of at least four classes of secretory organelles–rhoptries , micronemes , exonemes and dense granules—that are unique to apicomplexan parasites and that are positioned within the apical end of the merozoite [6] . The largest of these organelles , the rhoptries , are twinned club or pear-shaped structures that are formed de novo during each blood-stage life cycle by progressive fusion of post-Golgi vesicles , with the possible involvement of an endosome-like pathway [7–9] . Rhoptries are characterised by a relatively wide bulb region which narrows to a neck or duct towards the apical prominence of the merozoite; rhoptry discharge takes place via the duct . Despite the bulb and neck regions not being separated by a discernible membranous boundary , examination by transmission electron microscopy ( TEM ) has shown a distinct sub-compartmentalisation of each rhoptry , with the bulb and neck regions displaying different staining characteristics in both Plasmodium [7] and the related apicomplexan parasite Toxoplasma gondii [10] . Based on the detailed examination of the fate of a subset of rhoptry neck proteins ( RONs; [11] ) and bulb proteins , it has been suggested that the contents of the different rhoptry sub-compartments are released at different points in the invasion pathway , with the RONs being discharged before rhoptry bulb proteins [12] . Consistent with this , several RON proteins have been extensively implicated in the early stages of host cell entry , notably in formation of the moving junction or tight junction ( TJ ) , a short-lived doughnut-shaped structure through which the parasite passes as it enters the host cell ( e . g . [13–17] ) Some members of the P . falciparum reticulocyte binding protein homologue ( RH ) family of RBC adhesins , which also appear to localise to both the rhoptry bulb and neck [18] , also play central roles in early steps in invasion ( e . g . [3 , 19 , 20] ) . However , the association between function , timing of discharge and compartmentalisation within the rhoptries may not be a strict one , since the Plasmodium-specific protein RON12 is predominantly released post-invasion , into the nascent PV [21] . Similarly , a rhoptry bub location does not preclude roles during invasion , since the rhoptry bulb protein RhopH3 , a component of the high molecular weight ( HMW ) RhopH complex , is important for RBC entry [22] . In contrast , RhopH3 and its partner proteins RhopH1/Clag and RhopH2 have additionally been shown to be involved in nutrient uptake during development of the intracellular parasite [22–25] , whilst the rhoptry bulb low molecular weight ( LMW ) RAP1/RAP2 complex has been implicated in PVM formation [26] . In Toxoplasma ( but not in Plasmodium ) several rhoptry bulb proteins are enzymes , including proteases , phosphatases and kinases or pseudokinases , the latter two groups of which dramatically modulate host cell STAT signalling and the immunity-related GTPase ( IRG ) pathway involved in controlling parasite replication , and play key roles in virulence [27 , 28] . Collectively , the current evidence points to multiple diverse roles for rhoptry proteins , with a general theme being that rhoptry neck proteins are often but not exclusively involved in host cell entry whilst rhoptry bulb proteins usually fulfil subsequent roles in the life cycle ( see [29–31] for excellent reviews of this subject ) . Despite these many insights , of the ~60 known and putative Plasmodium rhoptry proteins identified through a combination of numerous antibody studies and proteomic analyses of isolated apicomplexan rhoptries [11 , 32–34] , the molecular functions of very few Plasmodium rhoptry proteins have been characterised . Crucially , how the different subclasses of Plasmodium rhoptry proteins are selectively delivered to their distinct sub-compartments within the organelle is unknown . Rhoptry-associated membrane antigen ( RAMA ) was initially identified in P . falciparum as a rhoptry bulb-resident protein which is expressed relatively early in the asexual blood stage cycle , before the de novo formation of nascent rhoptries [35 , 36] . Like many other Plasmodium rhoptry bulb proteins , RAMA has no obvious orthologue in Toxoplasma or other coccidian Apicomplexa . Unique amongst known Plasmodium rhoptry proteins , RAMA is associated with the inner ( lumenal ) face of the rhoptry bulb membrane via a glycosyl phosphatidylinositol ( GPI ) membrane anchor [36] . Trafficking of RAMA to the rhoptries shares several features of other rhoptry proteins , including transport via the Golgi apparatus and sensitivity to brefeldin A . The ~170 kDa RAMA precursor then undergoes proteolytic processing in the rhoptries or nascent rhoptries , involving removal of an N-terminal segment to generate a mature ~60 kDa form ( p60 ) . To gain insights into its function , Topolska and colleagues [36] and Richard et al . ( 2009 ) [37] used fluorescent resonance energy transfer ( FRET ) and immunoprecipitation to demonstrate that RAMA appears to interact with both RhopH3 and RAP1 , proteins which are enriched in lipid rafts but that lack lipid anchors . The authors hypothesised that RAMA plays an escorter role in recruiting and trafficking these proteins to the developing rhoptries , and that proteolytic cleavage of RAMA then facilitates their dissociation from the escorter complex . RAMA p60 ( but not the RAMA precursor ) was detectable in free extracellular merozoites and evidence was also presented that the protein is released from rhoptries at invasion , supported by the observation that a recombinant form of the extreme C-terminal region of RAMA could bind the RBC membrane . However , in newly invaded ‘ring’ stage parasites , RAMA p60 was found associated with the PVM , indicating that at least some of the protein was transported into the host cell with the invading parasite [36] . Very recent work has suggested that in fact the RAMA-RAP1 complex may itself be cargo for the Plasmodium orthologue of sortilin , an integral membrane protein that in mammalian cells acts as an escorter to transport proteins to the lysosomes , endocytic pathway and plasma membrane [38] . This is consistent with a role in rhoptry formation , since in Toxoplasma sortilin is required for biogenesis of both rhoptries and a second class of secretory organelle called micronemes [39] , and the P . falciparum sortilin orthologue has been localised to the Golgi [9] . Although restricted to the Plasmodium genus , RAMA orthologues were identified in all Plasmodium species examined [40] , and this was subsequently confirmed by the now extensive genome sequence data from numerous additional Plasmodium species ( see PlasmoDB RAMA gene ID: PF3D7_0707300; https://plasmodb . org/plasmo/ ) . More recent evidence from targeted or global reverse genetics studies indicates that disruption of the P . falciparum or P . berghei RAMA gene is deleterious or lethal [41–43] , consistent with an important role for RAMA in the asexual blood stage parasite life cycle . However , the molecular function of RAMA has remained obscure . Here we have used a robust conditional mutagenesis approach to examine the function of RAMA in the P . falciparum asexual blood stage life cycle . In contrast to previous suggestions , we found that disruption of RAMA expression does not affect trafficking of the rhoptry bulb proteins RhopH3 and RAP1 , nor biogenesis of the rhoptry bulb . Instead , our results indicate a selective role for RAMA in formation of the rhoptry neck structure . Perhaps as a result , the RAMA mutants show mislocalisation of several RON proteins and a lethal defect in host RBC invasion .
Previous attempts to directly disrupt the P . falciparum RAMA gene using conventional targeted genetic techniques were unsuccessful [41] , suggesting an important role for RAMA in the haploid asexual blood stages . To gain insights into the molecular function of RAMA we therefore used a conditional gene disruption approach based on the rapamycin-inducible DiCre conditional recombinase system . To enable this , we first modified the RAMA gene in the DiCre-expressing 1G5DC P . falciparum line [44] to introduce ‘silent’ loxP motifs suitable to act as sites for DiCre-mediated site-specific recombination . The RAMA gene encodes an 861 amino acid residue protein and comprises 4 small exons plus a single much larger exon ( exon 2 ) . The complexity and relatively large size ( ~3 . 2 kb ) of the RAMA locus precluded facile floxing of the entire coding sequence in a single manipulation as previously performed with other P . falciparum genes [4] , so we took an alternative approach in which we inserted into exon 2 a synthetic construct combining two heterologous loxPint elements ( i . e . synthetic introns each containing a loxP site; [45] ) , flanking 176 bp of intervening recodonised exon sequence . The inserted sequence was precisely integrated in a marker-free manner in a single Cas9-promoted homologous recombination step ( Fig 1A ) . The resulting modified RAMA gene ( referred to as RAMAloxP ) thus effectively comprised 7 exons with an internal recodonised exon flanked by loxPint sequences . DiCre-mediated recombination between the introduced loxP sites was predicted to excise this exon . Importantly , excision was also expected to create a frame-shift in the coding sequence immediately downstream of the single chimeric loxPint site remaining after excision , leading to a severely truncated RAMA coding sequence encoding only the N-terminal 220 residues , referred to below as RAMAΔE2 . Limiting dilution cloning of the transfected parasite population resulted in the isolation of 2 parasite clones called RAMAloxP-9C10 and RAMAloxP-12C7 , each independently generated using different sgRNAs . The expected modification of the native RAMA locus was confirmed in both parasite clones by diagnostic PCR ( Fig 1B ) and Southern blot ( Fig 1C ) and the clones were therefore used for all subsequent experiments . Growth assays comparing replication rates of RAMAloxP-9C10 and RAMAloxP-12C7 parasites with the parental 1G5DC line revealed no significant differences , indicating that the modifications made to the RAMA gene did not detectably impair parasite growth ( S1 Fig ) . To examine the efficiency of conditional excision of the floxed RAMA sequence , tightly synchronised newly-invaded ring stage cultures of both the RAMAloxP clones were divided into two and treated for just 4 h with either rapamycin ( RAP; 100 nM final ) or DMSO vehicle control ( 1% v/v , mock-treated ) . Following washing and further incubation for ~44 h to allow maturation to mature schizont stage , parasite genomic DNA ( gDNA ) was extracted and examined by analytical PCR . This revealed highly efficient RAP-induced excision of the floxed RAMA sequence ( Fig 1D ) . To determine the impact of excision on RAMA expression , the schizonts were analysed using a rabbit polyclonal antibody called anti-RAMA-D , specific for a C-terminal segment ( amino acid residues 482 to 758 ) of RAMA [36] . As shown in Fig 1E and Fig 1F , little or no signal was detected by the anti-RAMA-D antibodies in schizonts of the RAP-treated cultures , either by Western blot or by indirect immunofluorescence analysis ( IFA ) . These results confirmed the PCR data , demonstrating in two independent RAMAloxP parasite clones essentially complete conditional disruption of RAMA expression within a single erythrocytic cycle . The IFA images in Fig 1F indicated that , despite efficient disruption of the RAMA gene in RAP-treated RAMAloxP parasites , the parasites were able to form multinucleated schizonts towards the end of the erythrocytic growth cycle in which they were RAP-treated ( henceforth referred to as cycle 0 ) . To examine the effects of gene disruption on longer-term parasite viability , the mock- and RAP-treated RAMAloxP clones were maintained in culture and parasite replication monitored by flow cytometry over the ensuing erythrocytic cycles . Mock- and RAP-treated parental 1G5DC parasites were also included in these experiments to control for any off-target effects of RAP treatment . As shown in Fig 2A , a dramatic reduction in parasite replication rate was quickly evident in the RAP-treated RAMAloxP parasites , indicating a defect in parasite proliferation . This was confirmed by diagnostic PCR ( Fig 2B ) which revealed that upon passage of the cultures for up 9 erythrocytic cycles , the excised locus quickly became undetectable in RAP-treated cultures ( indicating disappearance of the RAMAΔE2 mutants ) whilst in contrast the initially very minor fraction of non-excised parasites gradually expanded to take over the cultures . This was consistent with an acute fitness defect in the RAMAΔE2 mutants . A further quantitative assessment of the impact of RAMA disruption upon parasite viability was obtained by comparing the capacity of mock- and RAP-treated RAMAloxP parasites to form zones of erythrocyte lysis , or plaques [46] , in static cultures in flat-bottomed microplates ( Fig 2C ) . This showed that RAP-treatment of both the RAMAloxP-9C10 and RAMAloxP-12C7 clones produced a highly significant reduction in plaque formation; the mean relative plaque forming capacity ( RPFC ) of RAP-treated parasites was 3 . 39% ± 0 . 833 ( meaning that for every 10 , 000 plaques produced by the control mock-treated cultures , only 339 were formed in the same number of RAP-treated culture-containing wells ) ( p = 3 . 32 x 10−8 , t = 115 . 9 , d . f . = 4 ) . Together , these data confirmed that RAMA plays a critical role in P . falciparum asexual blood stage viability . Given the established role of rhoptries in invasion , we considered that the replication defect shown by the RAMAΔE2 parasites might most likely be explained by a reduced capacity to invade host cells . To investigate this , mature RAMAloxP-9C10 and RAMAloxP-12C7 cycle 0 schizonts were isolated from highly synchronous mock- or RAP-treated cultures then added to fresh RBCs to achieve a similar starting parasitaemia . Following further incubation for 4 h to allow schizont rupture , the cultures were stained with Hoechst 33342 and the proportions of newly-invaded ring-stage parasites determined by flow cytometry . This consistently showed that the RAP-treated parasites displayed substantially reduced ring formation , corresponding to only ~10% of that of mock-treated parasites ( Fig 3A ) . As egress and invasion are intimately linked , we explored whether the lack of ring formation in these assays was indicative of a defect in egress from the host erythrocyte . To do this the capacity of RAMAΔE2 parasites to undergo egress was examined by time-lapse differential interference contrast ( DIC ) microscopy . As shown in Fig 3B , no detectable differences in the efficiency , kinetics or morphology of merozoite egress from RAP-treated RAMAloxP parasites was evident compared to their mock-treated counterparts . It was concluded that disruption of RAMA does not affect egress , but results in a severe defect in productive invasion of new host erythrocytes . To characterize this invasion defect in further detail , the behaviour of naturally-released merozoites and their interactions with host RBCs was examined by time-lapse DIC video microscopy . As shown in S1 Movie and S2 Movie , and quantified in Fig 3C , merozoites released from RAP-treated ( RAMAΔE2 ) cycle 0 schizonts interacted initially with neighbouring host RBCs in a similar manner to their mock-treated RAMAloxP counterparts , inducing repeated and often strong deformation of the host cells . However , unlike the mock-treated merozoites , in no case ( out of a total of 20 similar egress events examined ) did we observe successful invasion by the RAMAΔE2 merozoites . Of particular additional interest , no RBC echinocytosis was ever observed in the case of the RAMAΔE2 merozoite-RBC interactions . Echinocytosis generally takes place shortly following invasion ( see S1 Movie ) , but under certain conditions can also occur in the absence of invasion , where it is thought to indicate discharge of rhoptry components [3–5] . It was concluded that loss of RAMA selectively results in a fatal block in invasion , likely due to a defect in rhoptry discharge . As mentioned above , morphological characterisation of mature cycle 0 RAMAΔE2 schizonts by light microscopy ( IFA ) indicated apparently normal schizont development ( Fig 1F ) . However , given the clear invasion phenotype displayed by the RAMAΔE2 mutants and the known rhoptry bulb localisation of RAMA [36] , we decided to examine the structure of the mutant parasites in greater detail to assess the possible impact of RAMA disruption on the makeup and morphology of the rhoptries . This was considered particularly important given previous claims that RAMA interacts with RhopH3 and RAP1 and plays a role in the trafficking of both proteins to the rhoptries [36 , 37] . To determine the effects of RAMA disruption on the expression and trafficking of other rhoptry proteins , mature cycle 0 schizonts from mock-treated and RAP-treated RAMAloxP cultures were probed by IFA using a suite of antibodies specific for a range of rhoptry bulb proteins . As shown in Fig 4A , this revealed that the subcellular localisation and staining intensity of the LMW rhoptry complex protein RAP2 , the HMW rhoptry complex proteins RhopH1/Clag3 . 1 , RhopH2 and RhopH3 , the invasion ligand Rh5 , and the rhoptry-associated protein ARO ( thought to localise to the cytosolic face of the rhoptry bulb and neck [47–49] ) were unaffected by RAMA disruption; all retained their characteristic apical punctate rhoptry signal in the RAP-treated RAMAloxP schizonts . Note that similar findings were noted above for RAP1 ( Fig 1F ) . It was concluded that truncation of RAMA does not discernibly affect the trafficking of a number of established rhoptry bulb proteins . To examine the effects of RAMA truncation upon the subcellular localisation of RON proteins , mock- and RAP-treated RAMAloxP schizonts were next probed with antibodies specific for RON2 [50] , RON3 , or RON4 [51] . As shown in Fig 4B and Fig 4C , in all cases this revealed evidence for RON protein mislocalisation in the RAMAΔE2 mutants , although the precise phenotype appeared to depend upon the extent of parasite maturity . Specifically , for all three RON proteins , in immature schizonts a much reduced signal intensity was detectable which lacked the punctate , apically-disposed pattern typical of rhoptries . In highly mature segmented schizonts , IFA signals for RON2 , RON3 and RON4 were often completely absent . These findings were particularly intriguing since , whilst RON2 and RON4 are established rhoptry neck proteins that play roles in TJ formation , RON3 has been localised by immuno electron microscopy to the rhoptry bulb in Plasmodium [52] , rather than to the rhoptry neck as originally determined for the Toxoplasma orthologue [11] . Work in Toxoplasma also supports a rhoptry bulb location ( see ToxoDB annotation for the TgRON3 gene at https://toxodb . org/toxo/app/record/gene/TGME49_223920 ) . The current nomenclature for RON3 is therefore somewhat misleading , and these results therefore suggested that disruption of RAMA expression can modulate the localisation of rhoptry neck proteins as well as at least one rhoptry bulb protein . To investigate this issue further using additional rhoptry markers , we examined RAMAloxP schizonts by IFA using antibodies to two proteins that have been previously localised to either the rhoptry neck ( RON12; [21] ) or both the rhoptry body and the neck ( Rh2b; [18] ) in P . falciparum merozoites . As shown in S2 Fig and in contrast to the effects on RON2 , RON3 and RON4 , disruption of RAMA expression had no discernible effect on expression or localisation of either RON12 or Rh2b . As expected , disruption of RAMA did not impact on the localisation of the major merozoite surface protein MSP1 or the microneme protein apical membrane antigen 1 ( AMA1 ) ( Fig 4D ) . AMA1 is often observed to translocate onto the merozoite surface in highly mature schizonts , and this phenomenon too was observed at similar frequency in both mock-treated and RAP-treated RAMAloxP schizonts ( Fig 4D , lower panels ) . Collectively , these results indicated that disruption of RAMA results in selective mislocalisation and/or loss of rhoptry neck and certain rhoptry bulb proteins , whilst having no detectable impact on the expression and trafficking of other rhoptry proteins examined . Previous studies of RAMA suggested a role in rhoptry biogenesis [36] . Our discovery that some rhoptry proteins become mislocalised upon RAMA disruption prompted us to next examine the impact of RAMA disruption on rhoptry formation and morphology . To do this , mature segmented cycle 0 mock- or RAP-treated RAMAloxP schizonts were examined by TEM . To quantify rhoptry formation , the number of rhoptry profiles detectable in individual intracellular merozoites in 70 nm thin sections was recorded . This revealed no statistically significant differences in the total number of visible rhoptry profiles between the control and RAMAΔE2 parasites ( glm fit with Poisson model , predicting rhoptry count from status , non-zero status coefficient test p-value = 0 . 232 ) ( Fig 5A ) . This result was in complete accord with the IFA results obtained with the rhoptry bulb-specific antibodies described in Fig 4 , suggesting that the overall number of rhoptry organelles assembled in wild-type and RAMAΔE2 parasites was similar . During the above analysis , we noticed that whilst other intracellular features of the control and RAP-treated parasites ( including microneme morphology ) were similar , there was a relative absence of typical club-shaped rhoptry profiles in the RAMAΔE2 TEM micrographs . To attempt to quantify these apparent differences in rhoptry morphology , the TEM rhoptry images were categorised as being either circular or club-shaped based on blinded visual examination ( Fig 5B ) . As shown in Fig 5C , whilst rhoptries of both shapes were observed in merozoites of mock-treated and RAP-treated RAMAloxP parasites , a significantly lower proportion of club-shaped profiles was observed in the RAMAΔE2 samples . Moreover , in those RAMAΔE2 rhoptries that visibly possessed an apical projection , these appeared qualitatively different from wild type , being relatively short and stubby compared to the elongated projections typical of the latter ( Fig 5B ) . This difference in shape was confirmed by quantifying the aspect ratio ( the ratio of the longest to shortest diameter ) in the electron micrographs of all non-circular rhoptries from both control and RAP-treated parasites . The resulting data ( Fig 5D ) clearly indicated that the length of the rhoptry neck projections in control parasites ( mean aspect ratio = 1 . 73 ) was significantly greater than those of RAP-treated parasites ( mean aspect ratio = 1 . 21 ) ( p <0 . 0001 , t = 5 . 29 , d . f . = 58 ) . Since these profiles represent random section planes through rhoptries , it was considered unlikely that the differences noted were a chance result of random variation in the section plane obtained during sample preparation . However , to finally examine this point , electron tomography was performed on serial 200 nm thick sections of a single mock-treated RAMAloxP schizont and a RAP-treated RAMAloxP schizont ( S3 Movie and S4 Movie ) . This allowed computational three-dimensional modelling of much of the rhoptry content of each schizont by segmentation and rendering of the tomograms . The resulting models ( Fig 5E ) supported the thin section TEM data in showing a clear difference between the elongated shape of rhoptries in the DMSO-treated RAMAloxP schizonts and the strikingly more spherical nature of the RAMAΔE2 rhoptries . Collectively , our EM data indicated that disruption of RAMA does not affect the number of rhoptries formed , but has a major impact on their morphology , leading to the generation of relatively spherical or ‘neckless’ rhoptries . This may explain why certain rhoptry neck proteins mislocalise in the RAMA mutants .
The involvement of rhoptry proteins in host cell entry has long been proposed , supported by early observations that rhoptry discharge coincides temporally with invasion in several apicomplexan parasites , including Toxoplasma [53–56] and P . falciparum [17] . Those studies have been burgeoned by recent demonstrations that correct positioning of rhoptries at the apical pole of the Toxoplasma tachyzoite is a prerequisite for invasion ( but not egress ) [48] whilst in both P . falciparum merozoites and Toxoplasma tachyzoites rhoptries appear to undergo fusion with each other during invasion [30 , 57] . The discovery of the role of several RON proteins in TJ formation [15 , 16] was consistent with this overall model for rhoptry function , but as discussed in the Introduction it is also now clear that rhoptry proteins play roles following completion of invasion in such diverse functions as PVM generation , subversion of host cell signalling , and nutrient acquisition . Here we have established a new role for a Plasmodium rhoptry protein in biogenesis of these important organelles . We draw three major conclusions from our study . First , we have shown that RAMA is essential for parasite survival , and that disruption of RAMA expression results in a selective , fatal defect in host cell invasion with no obvious effect on intracellular parasite replication or egress . This finding is reminiscent of the results of conditional ablation in Toxoplasma of ARO , a palmitoylated protein localised to the cytoplasmic rhoptry surface which is essential for correct tethering of the rhoptries within the tachyzoite apex [47–49] . Loss of TgARO results in rhoptries that are dispersed throughout the cytoplasm of the parasite , and the mutant tachyzoites undergo normal egress but cannot invade host cells . As well as not being able to penetrate targeted RBCs , we found that RAMAΔE2 merozoites showed no capacity to induce echinocytosis in interacting RBCs . This provides important insights into the underlying defect . Evidence that rhoptry discharge can take place upon contact with the host cell even in the absence of productive invasion was first demonstrated through observations of the appearance of small nascent intraerythrocytic channels or vacuoles at the site of apical attachment of P . knowlesi merozoites to the RBC surface; this occurred both under normal , invasion-permissive conditions but also in cultures containing cytochalasin B , an inhibitor of actin polymerisation that blocks invasion downstream of TJ formation [58–60] . Subsequent work in P . falciparum combining electron microscopy with super-resolution immunofluorescence microscopy showed that these vacuoles contain rhoptry proteins [13] . Similar structures ( termed ‘evacuoles’ when formed in the presence of cytochalasin ) were observed upon interaction of Toxoplasma tachyzoites with host cells , and these were also found to contain large amounts of rhoptry bulb proteins [53 , 61] . More recent comprehensive microscopic examination of the interactions between free P . falciparum merozoites and host RBCs provided evidence that rhoptry discharge is the primary inducer of echinocytosis in target RBCs [3] , and consistent with the above observations this can occur even in the absence of productive invasion . Very recent work has added further support to this model , showing that merozoites genetically deficient in key components of the cyclic nucleotide signalling pathway or the molecular motor that drives invasion can induce echinocytosis in target RBCs , despite being unable to penetrate the cells [4 , 5] . It has thus become clear that host RBC echinocytosis—irrespective of whether it is followed by successful invasion—provides a useful reporter for rhoptry function in Plasmodium . In our current study , the complete absence of echinocytosis upon interactions between RAMAΔE2 merozoites and target RBCs is therefore completely consistent with a defect in rhoptry discharge and/or the absence of the ( currently unidentified ) rhoptry component ( s ) that induce echinocytosis . RAMA was previously proposed to act as an invasion ligand based on indications that anti-RAMA antibodies inhibit merozoite invasion of RBCs [36] and that recombinant RAMA-based peptides could apparently bind to RBCs and inhibit invasion [62] . Our study did not directly address this proposed function for RAMA . The second clear conclusion from our study is that RAMA is not required for correct trafficking of RAP1 or RhopH3 , as previously proposed based on protein-protein interaction and co-localisation studies [36–38] , nor for trafficking of several other rhoptry bulb proteins examined . Our results are therefore not consistent with a rhoptry bulb-specific protein escorter role for RAMA . However , our findings may not necessarily be in conflict with the interaction model; RAMA may associate with RhopH3 and RAP1 , for example , in order to facilitate formation of the respective HMW and LMW complexes , whilst other factors may regulate localisation of these protein complexes to the rhoptry bulb . It is important to note that our RAMA disruption strategy did not remove the entire gene and could in principle result in the expression of the N-terminal 220 residues of the protein as a truncated protein product ( Fig 1 ) . We did not have the antibody tools to detect whether such a truncated product was expressed in our mutants . However , were it to be expressed , the truncated protein would lack the C-terminal sequences between residues 315–840 previously shown to interact with RAP1 and sortilin , and required for trafficking to the rhoptries [36–38] . We are therefore confident that our strategy generated a loss-of-function mutant , and the phenotype supports that . Third , our results show that RAMA is instead required for correct localisation of a subset of RON and rhoptry bulb proteins and for formation of the rhoptry neck , providing the first genetic evidence that RAMA plays an essential role in rhoptry biogenesis . The causal relationship between these phenotypes is unclear , in that we cannot discriminate between whether a defective rhoptry morphology leads to mislocalisation of the RON proteins or alternatively whether correct trafficking of RON proteins is required for rhoptry neck morphogenesis . RAMA may interact with RON proteins to ensure their correct trafficking , although no associations between RAMA and RON proteins were identified in the earlier studies [36 , 37] . The involvement of a rhoptry bulb protein in trafficking of rhoptry neck proteins appears somewhat counterintuitive . Could RAMA direct these proteins to the bulb then regulate their onward trafficking to the nascent neck domain ? It is interesting to note that RON12 , a rhoptry neck protein localisation of which was unaffected by RAMA disruption in this study , has very recently been confirmed as a rhoptry neck protein in merozoites of the rodent malaria parasite Plasmodium berghei , but a rhoptry bulb protein in P . berghei sporozoites , a developmental stage of the parasite generated in the mosquito vector [63] . This suggests that the trafficking characteristics of rhoptry proteins may not be an intrinsic function of their structure per se , but dependent upon the developmental context of their expression . Both RON2 and RON4 are key for formation of the TJ , so their mis-trafficking may explain in part the invasion defect in the RAMAΔE2 mutants . However , we propose that the most dramatic element of the phenotype–the defect in rhoptry neck formation–is most likely primarily responsible for the loss of invasive capacity in the mutants . Absence of the neck structure would presumably prevent docking and fusion of the rhoptries with the parasite plasma membrane at the merozoite apical prominence , reducing or ablating the capacity for rhoptry discharge . It might be expected that loss of the rhoptry neck structure would also result in a phenotype similar to that of loss of TgARO [48 , 49] , with the rhoptries becoming dispersed throughout the merozoite cytoplasm rather than apically located; however the relatively small dimensions of the P . falciparum merozoite relative to the Toxoplasma tachyzoite likely limits cytoplasmic movement of untethered rhoptries , rendering this phenotype less obvious . In conclusion , the selective impact of RAMA disruption on the correct localisation of some but not all rhoptry proteins supports the existence of subsets of rhoptry proteins , some of which require RAMA for correct trafficking whilst others do not . We have shown that RAMA disruption is lethal , likely through a direct impact upon rhoptry biogenesis and a resulting indirect impact upon invasion but not egress . Given that there is no obvious orthologue of RAMA in Toxoplasma and other apicomplexan parasites outside of the Plasmodium genus , our findings raise obvious questions as to how rhoptry neck formation is regulated in other Apicomplexa . In this regard is interesting to note that disruption of the gene encoding a GPI-anchored Toxoplasma rhoptry bulb protein called TgCA_RP results in dysmorphic , fragmented rhoptries [64] , although the phenotype is quite different from that of the RAMAΔE2 P . falciparum mutants and the morphologies of rhoptries are also distinct between the two apicomplexan genera . TgCA_RP and RAMA share no homology , but it is conceivable that the roles of the two proteins in rhoptry biogenesis may be similar .
This work used human blood that was provided anonymously through the UK Blood Transfusion Service . Oligonucleotide primers ( see S1 Table ) were purchased from Sigma-Aldrich ( Gillingham , UK ) . Rapamycin was also obtained from Sigma-Aldrich and stored frozen as a stock solution ( 10 μM ) in DMSO . The antifolate drug WR99210 ( Jacobus Pharmaceuticals , Princeton , NJ ) was stored as a 20 μM stock solution in DMSO . The anti-RAMA-D , anti-RhopH3 and anti-KAHRP antibodies were kind gifts of Ross Coppel ( Monash University , Australia ) whilst Osamu Kaneko ( Nagasaki University , Japan ) kindly provided antibodies against RhopH1/Clag3 . 1 . The polyclonal anti-AMA1 antibody has previously been described [65] , as has the RhopH2-specific mAb 61 . 3 [66] . The anti-MSP1 mAb 89 . 1 , the anti-RON3 mAb 1H1 , the anti-RAP1 mAb 4F3 , a rabbit anti-ARO polyclonal antibody , and rabbit antibodies against RON12 were kind gifts of Tony Holder ( Francis Crick Institute , UK ) , and the antibodies to Rh2b were generously provided by Julian Rayner ( University of Cambridge , UK ) . Other antibodies were generously provided by John Vakonakis , University of Oxford , UK ( anti-MAHRP1 antibody ) , Simon Draper , University of Oxford , UK ( anti-Rh5 antibody ) , Takafumi Tsuboi , Ehime University Japan ( anti-RON2 antibody ) and Jean-François Dubremetz , Université Montpellier 2 , France ( anti-RON4 mAb 24C6 ) . RAMAloxP mutant P . falciparum lines were generated by transfecting parasites with 60 μg of a repair plasmid alongside 20 μg of a plasmid for expression of Cas9 . Repair plasmid pESS_RAMA_E2INT_loxP ( synthesised by ThermoFisher Scientific ) comprised synthetic heterologous SERA2:loxP and SUB2:loxP introns which were inserted at AG-AT sites 776 bp and 952 bp respectively downstream of the RAMA gene ATG start codon . The intervening 174 bp was codon-optimised for expression in Escherichia coli . Sequence comprising 568 bp of exon 2 both 5’ and 3’ to the site of the heterologous intron insertions was included to act as flanking regions for homologous recombination . EuPaGDT software was used to identify 20 bp protospacer sequences ( flanked by a 5’-NGG-3’ protospacer adjacent motif ) in order to specifically target Cas9-mediated cleavage to the RAMA gene locus . pRAMAsgRNA1 and pRAMAsgRNA2 were created by modifying the Cas9-sgRNA-encoding pDC2 vector ( a kind gift of Marcus Lee , Sanger Institute , UK ) by introduction of sgRNAs 1 and 2 generated by annealing complementary primer sequences RAMA_sgRNA_E2_F to RAMA_sgRNA_E2_R ( sgRNA1 ) , and RAMA_sgRNA_E2S_F to RAMA_sgRNA_E2S_R ( sgRNA2 ) . All parasite work described used the 3D7-derived DiCre-expressing 1G5DiCre ( 1G5DC ) P . falciparum clone [44] . Parasites were maintained in culture and synchronised using standard procedures [67 , 68] . DNA constructs were introduced by electroporation into mature schizont-stage 1G5DC parasites using previously described methods [69] . Approximately 24 h post-transfection the cultures were supplemented with WR99210 ( 2 . 5 nM ) to allow selection of parasites carrying the human dhfr selectable marker encoded by the pRAMAsgRNA plasmids . Parasites were maintained under drug selection for 2 erythrocytic cycles before being transferred to drug-free culture medium . Following confirmation of integration in the resulting parasite populations by diagnostic PCR , parasites were cloned by limiting dilution [46] . Conditional excision of floxed genomic sequences in the transgenic parasite clones was induced as previously described [44] by treating tightly synchronized ring stage parasite cultures with 100 nM RAP in 1% ( v/v ) DMSO ( or the same concentration of DMSO only as a vehicle control ) for 4 h at 37°C . 5ʹ and 3ʹ integration of pESS_RAMA_E2INT_loxP into the genome was detected by PCR using the integration specific primers RAMA_exon1_F plus RAMA_exon2recod_R , and RAMA_exon2recod_F plus RAMA_exon2_R . No amplification was expected from the wild type specific primer set RAMA_exon1_F plus RAMA_exon2WT_R . Diagnostic PCR was also used to confirm that the integrated DiCre expression cassette remained unmodified in isolated clones , using primers SERA5_DiCre_F and DiCre_integrated . Reversion to the wild type SERA5 locus by spontaneous excision of the integrated DiCre cassette ( which can occur at low levels in the 1G5DC line; [4] ) was indicated by the amplification of a 1700 bp amplicon using primers specific to the unmodified SERA5 locus ( SERA5_DiCre_F and SERA5_DiCre_R ) . To further confirm the genotype of potential RAMAloxP integrant clones , a 400 bp digoxigenin ( DIG ) -labelled probe corresponding to the 5’ 400 bp of the region of exon 2 used as the second homology arm was generated by PCR amplification using primers RAMA_SB_homology2_F and RAMA_SB_homology2_R , performed in the presence of DIG-labelled nucleotides provided by the PCR DIG Labelling Mix Plus ( Roche ) . Approximately 5 μg of gDNA from putative integrant cultures was digested with an excess of XmnI restriction enzyme , then fractionated by agarose gel electrophoresis , transferred overnight to Amersham Hybond N+ nylon membranes ( GE Healthcare Life Sciences ) , and cross-linked with UV light ( 1200 μjoules/cm2 ) . After pre-hybridisation , hybridisation , washing and blocking steps , hybridised signal was detected by incubation with polyclonal alkaline phosphate-conjugated anti-DIG antibodies ( Roche ) . The membrane was incubated with detection buffer and CDP-star ( Roche ) , before being exposed to X-ray film . RAP-induced recombination between genomic loxP sites was detected by PCR analysis of schizont-stage gDNA ( harvested 42–44 h following mock- or RAP-treatment ) using primers RAMA_Harm1_F and RAMA_Harm2_R . For immunoblot analysis , SDS-PAGE fractionated proteins were transferred to nitrocellulose membranes before being blocked , incubated with primary antibody diluted to the appropriate concentration , washed and then probed with a HRP-conjugated secondary antibody as previously described [65] . After final washes , membranes were incubated with Immobilon Western Chemiluminescent HRP Substrate ( Merck ) and signals visualised by exposure to X-ray film . Thin blood films of mock- or RAP-treated Percoll-enriched RAMAloxP schizonts were air-dried and stored desiccated at -80°C . As required , samples were thawed at 37°C and parasites fixed in 4% ( w/v ) paraformaldehyde and permeabilised in 0 . 1% ( v/v ) Triton X-100 . Fixed slides were blocked then probed with the relevant antibody diluted as follows: rabbit anti-RAMA-D , 1:1000; mouse mAb 4F3 ( anti-RAP1 ) , 1:100; rabbit anti-RhopH1/Clag3 . 1 , 1:100; mouse mAb MRA876 ( anti-RAP2 ) , 1:200; rabbit anti-Rh5 , 1:10 , 000; mouse mAb 61 . 3 ( anti-RhopH2 ) , 1:100; rabbit anti-ARO , 1:500; rabbit anti-RhopH3 ( anti-Ag44 ) , 1:2000; rabbit anti-RON2 , 1:250; mouse mAb 1H1 ( anti-RON3 ) , 1:100; mouse mAb 24C6 ( anti-RON4 ) , 1:500; rabbit anti-AMA1 , 1:500; mouse mAb 89 . 1 ( anti-MSP1 ) , 1:1000; rabbit anti-RON12 , 1:5000; rabbit anti-Rh2b , 1:500 . After incubation and washing , slides were probed with the required Alexa Fluor 488 or 594-conjugated secondary antibody diluted 1:10 , 000 . Slides were then mounted in ProLong Gold Antifade Mountant containing DAPI ( ThermoFisher Scientific ) , sealed with Cytoseal-60 ( Thermofisher Scientific ) and images collected using a Nikon Eclipse Ni-E wide field microscope with a Hamamatsu C11440 digital camera and 100x/1 . 45NA oil immersion objective . Identical exposure conditions were used at each wavelength for each pair of mock- and RAP-treated samples under comparison . Images were processed using Fiji software . To determine the growth capability of mutant parasites over multiple cycles , DMSO and RAP-treated cultures were adjusted to a parasitaemia of 0 . 1% , with samples for flow cytometry being sampled for fixation every 48 h for up to 6 erythrocytic growth cycles . Pelleted parasites were fixed in 4% paraformaldehyde ( Electron Microscopy Sciences ) and 0 . 02% glutaraldehyde ( Sigma-Aldrich ) in PBS for 1 h at 37°C , before being diluted into PBS . When ready for flow cytometry , samples were stained with a 1:10 , 000 dilution of Hoechst 33342 ( ThermoFisher Scientific ) in PBS for 30 min at 37°C . The invasive capacity of parasites was determined by adding RBCs to highly synchronous Percoll-purified mature schizonts to obtain a parasitaemia of 1% . After further incubation for 4 h , cultures were fixed and the percentage of newly infected RBCs was determined . Parasitaemia and DNA replication were indicated by Hoechst-staining intensity determined using a LSRFortessa ( BD Biosciences ) flow cytometer , collecting a minimum of 105 events per sample . Data were analysed using FlowJo software . The relative plaque-forming capacity of parasites was determined as previously described [46 , 70] . All experiments were performed in triplicate using blood from different donors , and statistical analysis was performed using GraphPad Prism . To visualise merozoite egress , schizont-stage mock- or RAP-treated parasites were imaged as previously described [4 , 69] . Invasion videos were similarly performed using schizonts purified from mock- or RAP- treated cultures mixed with fresh uninfected RBCs . DIC images were acquired every 150 ms and Nikon NIS Elements AR analysis software was used to produce the resulting time-lapse videos . Schizonts of RAMAloxP parasites were harvested ~44 h following mock- or RAP-treatment , before being fixed in 2 . 5% glutaraldehyde and 4% formaldehyde in 0 . 1 M phosphate buffer . After fixation , cells were washed in 0 . 1 M phosphate buffer , embedded in 2% agarose and processed as described previously [4] . For each sample , three sections were taken from each of three of the 1 mm3 blocks of agarose-embedded schizonts . Twenty schizonts from each block were quantified , scoring the number of rhoptries within each intracellular merozoite . Each rhoptry was also visually classified as being circular or club-shaped ( i . e . possessing distinct neck and bulb regions ) . Statistical analyses of the data were performed using R 3 . 3 . 1 ( https://www . R-project . org/ ) ( R Core Team , 2013 ) . Generalised linear models were fit using the glm ( ) function , with the first model ( rhoptry_count~DMSO_or_Rapa_status , family = Poisson ) being used to test for a difference in rhoptry counts . For comparison of ‘club-shaped’ counts on the non-zero rhoptry data , the model used was club_count~DMSO_or_Rapa_status + rhoptry_count , family = Poisson . Analysis of Deviance was performed using the anova ( ) function on the resultant glm object , with test = ‘Chisq’ . Tilt series images of a schizont from each experimental condition were collected from +60° to -60° with 1° increments using the Tecnai User Interface software ( FEI ) . The 1024 x 1024 pixel images were recorded with an Ultrascan charge coupled device camera ( Gatan Inc . ) with a pixel size of 5 . 42 nm . The IMOD package [71] was used to reconstruct individual tomograms , with patch tracking used to create a fiducial model . For each schizont , tomograms from three serial 200 nm sections were flattened and joined together in z to obtain a continuous volume . The 3dmod programme of IMOD was used to manually segment rhoptry membranes , from which three-dimensional surface models were generated . Models were left open where rhoptries extended outside the joined volume , and where the top or bottom surface of a rhoptry was lost between adjacent tomograms . | Despite improved control measures over recent decades , malaria is still a considerable health burden across much of the globe . The disease is caused by a single-celled parasite that invades and replicates within host cells . During invasion , the parasite discharges a set of flask-shaped secretory organelles called rhoptries , the contents of which are crucial for invasion as well as for modifications to the host cell that are important for parasite survival . Rhoptry discharge occurs through fusion of the relatively elongated rhoptry neck to the apical surface of the parasite . Different proteins reside within the bulbous rhoptry body and the neck regions , but how these proteins are selectively sent to their correct sub-compartments within the rhoptries and how the rhoptries are formed , is poorly understood . Here we show that a malaria parasite rhoptry bulb protein called rhoptry-associated membrane antigen ( RAMA ) plays an essential role in rhoptry neck formation and correct trafficking of certain rhoptry neck and bulb proteins . Parasites deficient in RAMA produce malformed rhoptries and–probably as a result—cannot invade host red blood cells . Our work sheds new light on how rhoptries are formed and reveals insights into the mechanism by which the correct sorting of proteins to distinct regions of the rhoptry is regulated . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"parasite",
"groups",
"parasite",
"replication",
"plasmodium",
"viral",
"transmission",
"and",
"infection",
"microbiology",
"cloning",
"parasitic",
"protozoans",
"parasitology",
"apicomplexa",
"protozoans",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"technique... | 2019 | The Plasmodium falciparum rhoptry bulb protein RAMA plays an essential role in rhoptry neck morphogenesis and host red blood cell invasion |
Stimulus dimensionality-reduction methods in neuroscience seek to identify a low-dimensional space of stimulus features that affect a neuron’s probability of spiking . One popular method , known as maximally informative dimensions ( MID ) , uses an information-theoretic quantity known as “single-spike information” to identify this space . Here we examine MID from a model-based perspective . We show that MID is a maximum-likelihood estimator for the parameters of a linear-nonlinear-Poisson ( LNP ) model , and that the empirical single-spike information corresponds to the normalized log-likelihood under a Poisson model . This equivalence implies that MID does not necessarily find maximally informative stimulus dimensions when spiking is not well described as Poisson . We provide several examples to illustrate this shortcoming , and derive a lower bound on the information lost when spiking is Bernoulli in discrete time bins . To overcome this limitation , we introduce model-based dimensionality reduction methods for neurons with non-Poisson firing statistics , and show that they can be framed equivalently in likelihood-based or information-theoretic terms . Finally , we show how to overcome practical limitations on the number of stimulus dimensions that MID can estimate by constraining the form of the non-parametric nonlinearity in an LNP model . We illustrate these methods with simulations and data from primate visual cortex .
The neural coding problem , an important topic in systems and computational neuroscience , concerns the probabilistic relationship between environmental stimuli and neural spike responses . Characterizing this relationship is difficult in general because of the high dimensionality of natural signals . A substantial literature therefore has focused on dimensionality reduction methods for identifying which stimuli affect a neuron’s probability of firing . The basic idea is that many neurons compute their responses in a low dimensional subspace , spanned by a small number of stimulus features . By identifying this subspace , we can more easily characterize the nonlinear mapping from stimulus features to spike responses [1–5] . Neural dimensionality-reduction methods can be coarsely divided into three classes: ( 1 ) moment-based estimators , such as spike-triggered average ( STA ) and covariance ( STC ) [1 , 5–8]; ( 2 ) model-based estimators , which rely on explicit forward encoding models [9–16]; and ( 3 ) information and divergence-based estimators , which seek to reduce dimensionality using an information-theoretic cost function [17–22] . For all such methods , the goal is to find a set of linear filters , specified by the columns of a matrix K , such that the probability of response r given a stimulus s depends only on the linear projection of s onto these filters , i . e . , p ( r|s ) ≈ p ( r|K⊤s ) . Existing methods differ in computational complexity , modeling assumptions , and stimulus requirements . Typically , moment-based estimators have low computational cost but succeed only for restricted classes of stimulus distributions , whereas information-theoretic and likelihood-based estimators allow for arbitrary stimuli but have high computational cost . Previous work has established theoretical connections between moment-based and likelihood-based estimators [11 , 14 , 17 , 19 , 23] , and between some classes of likelihood-based and information-theoretic estimators [14 , 20 , 21 , 24] . Here we focus on maximally informative dimensions ( MID ) , a well-known information-theoretic estimator introduced by Sharpee , Rust & Bialek [18] . We show that this estimator is formally identical to the maximum likelihood ( ML ) estimator for the parameters of a linear-nonlinear-Poisson ( LNP ) encoding model . Although previous work has demonstrated an asymptotic equivalence between these methods [20 , 24 , 25] , we show that the correspondence is exact , regardless of time bin size or the amount of data . This equivalence follows from the fact that the plugin estimate for the single-spike information [26] , the quantity that MID optimizes , is equal to a normalized Poisson log-likelihood . The connection between the MID estimator and the LNP model makes clear that MID does not incorporate information carried by non-Poisson statistics of the response . We illustrate this shortcoming by showing that MID can fail to find information-maximizing filters for simulated neurons with binary or other non-Poisson spike count distributions . To overcome this limitation , we introduce new dimensionality-reduction estimators based on non-Poisson noise models , and show that they can be framed equivalently in information-theoretic or likelihood-based terms . Finally , we show that a model-based perspective leads to strategies for overcoming a limitation of traditional MID , that it cannot tractably estimate more than two or three filters . The difficulty arises from the intractability of using histograms to estimate densities in high-dimensional subspaces . However , the single-spike information depends only on the ratio of densities , which is proportional to the nonlinearity in the LNP model . We show that by restricting the parametrization of this nonlinearity so that the number of parameters does not grow exponentially with the number of dimensions , we can obtain flexible yet computationally tractable estimators for models with many filters or dimensions .
Previous work has shown that MID converges asymptotically to the maximum-likelihood ( ML ) estimator for an LNP model in the limit of small time bins [20 , 24] . Here we present a stronger result , showing that the equivalence is not merely asymptotic . We show that standard MID , using histogram-based estimators for raw and spike-triggered stimulus densities p ( s ) and p ( s|spike ) , is exactly the ML estimator for the parameters of an LNP model , regardless of spike rate , the time bins used to count spikes , or the amount of data . The standard implementation of MID [18 , 20] uses histograms to estimate the projected stimulus densities p ( K⊤s ) and p ( K⊤s|spike ) . These density estimates are then used to compute Îss ( K ) , the plug-in estimate of single-spike information in a subspace defined by K ( Equation 4 ) . We will now unpack the details of this estimate in order to show its relationship to the LNP model log-likelihood . Let {B1 , … , Bm} denote a group of sets ( “histogram bins” ) that partition the range of the projected stimuli K⊤s . In the one-dimensional case , we typically choose these sets to be intervals Bi = [bi−1 , bi ) , defined by bin edges {b0 , … , bm} , where b0 = −∞ and bm = +∞ . Then let p ̂ = ( p ̂ 1 , … , p ̂ m ) and q ̂ = ( q ̂ 1 , … q ̂ m ) denote histogram-based estimates of p ( K⊤s ) and p ( K⊤s|spike ) , respectively , given by: p ^ i = # stimuli inB i # stimuli = 1 N ∑ t = 1 N 1 B i ( x t ) q ^ i = # stimuli in B i | spike # spikes = 1 n s p ∑ t = 1 N 1 B i ( x t ) r t , ( 6 ) where xt = K⊤st denotes the linear projection of the stimulus st , n s p = ∑ t = 1 N r t is the total number of spikes , and 1Bi ( ⋅ ) is the indicator function for the set Bi , defined as: 1 B i ( x ) = 1 , x ∈ B i 0 , x ∉ B i ( 7 ) The estimates p ̂ and q ̂ are also known as the “plug-in” estimates , and correspond to maximum likelihood estimates for the densities in question . These estimates give us a plug-in estimate for projected single-spike information: I ^ s s = ∑ i = 1 m q ^ i log q ^ i p ^ i = 1 n s p ∑ i = 1 m ∑ t = 1 N 1 B i ( x t ) r t log q ^ i p ^ i = 1 n s p ∑ t = 1 N r t log g ^ ( x t ) ( 8 ) where the function ĝ ( x ) denotes the ratio of density estimates: g ^ ( x ) ≜ ∑ i = 1 m 1 B i ( x ) q ^ i p ^ i . ( 9 ) Note that ĝ ( x ) is a piece-wise constant function that takes the value q̂i/p̂i over the ith histogram bin Bi . Now , consider an LNP model in which the nonlinearity f is parametrized as a piece-wise constant function , taking the value fi over histogram bin Bi . Given a projection matrix K , the ML estimate for the parameter vector α = ( f1 , … , fm ) is the average number of spikes per stimulus in each histogram bin , divided by time bin width Δ , that is: f ^ i = 1 Δ · ∑ t = 1 N 1 B i ( x t ) r t ∑ t = 1 N 1 B i ( x t ) = n s p N Δ q ^ i p ^ i . ( 10 ) Note that functions f ^ and g ^ are related by f ̂ ( x ) = ( n s p N Δ ) g ̂ ( x ) and that the sum ∑ t = 1 N f ̂ ( x t ) Δ = n s p . We can therefore rewrite the LNP model log-likelihood ( Equation 2 ) : ℒ l n p ( θ ; D ) = ∑ t = 1 N r t log n s p N g ^ ( x t ) - n s p - ∑ t = 1 N log r t ! = ∑ t = 1 N r t log g ^ ( x t ) + n s p log n s p N - 1 - ∑ t = 1 N log r t ! ( 11 ) This allows us to directly relate the empirical single-spike information ( Equation 8 ) with the LNP model log-likelihood , normalized by the spike count as follows: I ^ s s ( K ) = 1 n s p ℒ l n p ( θ ; D ) - 1 n s p n s p log n s p N - n s p - ∑ log r t ! ( 12 ) = 1 n s p ℒ l n p ( θ ; D ) - 1 n s p ℒ l n p ( θ 0 , D ) ( 13 ) where ℒlnp ( θ0 , D ) denotes the Poisson log-likelihood under a “null” model in which spike rate does not depend on the stimulus , but takes constant rate λ 0 = n s p N Δ across the entire stimulus space . In fact , the quantity −ℒlnp ( θ0 , D ) can be considered an estimate for the marginal entropy of the response distribution , H ( r ) =−∑ p ( r ) logp ( r ) , since it is the average log-probability of the response under a Poisson model , independent of the stimulus . This makes it clear that the single-spike information Iss can be equally regarded as “LNP information” . Empirical single-spike information is therefore equal to LNP model log-likelihood per spike , plus a constant that does not depend on model parameters . This equality holds independent of time bin size Δ , the number of samples N and the number of spikes nsp . From this relationship , it is clear that the linear projection K that maximizes Îss also maximizes the LNP log-likelihood ℒlnp ( θ;D ) , meaning that the MID estimate is the same as an ML estimate for the filters in an LNP model: K ^ M I D = K ^ M L . ( 14 ) Moreover , the histogram-based estimates of the raw and spike-triggered stimulus densities p ̂ and q ̂ , which are used for computing the empirical single-spike information Îss , correspond to a particular parametrization of the LNP model nonlinearity f as a piece-wise constant function over histogram bins . The ratio of these plug-in estimates gives rise to the ML estimate for f . MID is thus formally equivalent to an ML estimator for both the linear filters and the nonlinearity of an LNP model . Previous literature has not emphasized that the MID estimator implicitly provides an estimate of the LNP model nonlinearity , or that the number of histogram bins corresponds to the number of parameters governing the nonlinearity . Selecting the number of parameters for the nonlinearity is important both for accurately estimating single-spike information from finite data and for successfully finding the most informative filter or filters . Fig . 3 illustrates this point using data from a simulated neuron with a single filter in a two-dimensional stimulus space . For small datasets , the MID estimate computed with many histogram bins ( i . e . , many parameters for the nonlinearity ) substantially overestimates the true Iss and yields large errors in the filter estimate . Even with 1000 stimuli and 200 spikes , a 20-bin histogram gives substantial upward bias in the estimate of single-spike information ( Fig . 3D ) . Parametrization of the nonlinearity is therefore an important problem that should be addressed explicitly when using MID , e . g . , by cross-validation or other model selection methods . Under the discrete-time inhomogeneous Poisson model considered above , spikes are modeled as conditionally independent given the stimulus , and the spike count in a discrete time bin has a Poisson distribution . However , real spike trains may exhibit more or less variability than a Poisson process [27] . In particular , the Poisson assumption breaks down when the time bin in which the data are analyzed approaches the length of the refractory period , since in that case each bin can contain at most one spike . In that case , a Bernoulli model provides a more accurate description of neural data , since it allows only 0 or 1 spike per bin . The Bernoulli and discrete-time Poisson models approach the same limiting Poisson process as the bin size ( and single-bin spike probability ) approaches zero while the average spike rate remains constant . However , as long as single-bin spike probabilities are above zero , the two models differ . Here we show that the standard “Poisson” MID estimator does not necessarily maximize information between stimulus and response when spiking is non-Poisson . That is , if the spike count r given stimulus s is not a Poisson random variable , then MID does not necessarily find the subspace preserving maximal information between stimulus and response . To show this , we derive the mutual information between the stimulus and a Bernoulli distributed spike count , and show that this quantity is closely related to the log-likelihood under a linear-nonlinear-Bernoulli encoding model . For neural responses binned at the stimulus refresh rate ( e . g . , 100 Hz ) , it is not uncommon to observe multiple spikes in a single bin . For the general case , then , we must consider an arbitrary distribution over counts conditioned on a stimulus . As we will see , maximizing the mutual information based on histogram estimators is once again equivalent to maximizing the likelihood of an LN model with piece-wise constant mappings from the linear stimulus projection to count probabilities . A significant drawback to standard MID is that it does not scale tractably to high-dimensional subspaces; that is , to the simultaneous estimation of many filters . MID has usually been limited to estimation of only one or two filters , and we are unaware of a practical setting in which it has been used to recover more than three . This stands in contrast to methods like spike-triggered covariance ( STC ) [1 , 7] , information-theoretic spike-triggered average and covariance ( iSTAC ) [19] , projection-pursuit regression [28] , Bayesian spike-triggered covariance [14] , and quadratic variants of MID [21 , 22] , all of which can tractably estimate ten or more filters . This capability may be important , given that V1 neurons exhibit sensitivity to as many as 15 dimensions [29] , and many canonical neural computations ( e . g . , motion estimation ) require a large number of stimulus dimensions [22 , 30] . Before we continue , it is helpful to consider why MID is impractical for high-dimensional feature spaces . The problem isn’t the number of filter parameters: these scale linearly with dimensionality , since a p-filter model with D-dimensional stimuli requires only Dp parameters , or indeed only ( D − 1 ) p − 1 2 p ( p − 1 ) parameters to specify the subspace naively after accounting for degeneracies . The problem is instead the number of parameters needed to specify the densities p ( x ) and p ( x|spike ) . For histogram-based density estimators , the number of parameters grows exponentially with dimension: a histogram with m bins along each of p filter axes requires mp parameters , a phenomenon sometimes called the “curse of dimensionality” . A variety of neural dimensionality reduction methods have been proposed previously . Here , we consider the relationship of the methods described in this study to these earlier approaches . Rapela et al [28] introduced a technique known as extended Projection Pursuit Regression ( ePPR ) , where the high-dimensional estimation problem is reduced to a sequence of simpler low-dimensional ones . The approach is iterative . A one-dimensional model is found first , and the dimensionality is then progressively increased to optimize a cost function , but with the search for filters restricted to dimensions orthogonal to all the filters already identified . From a theoretical perspective this assumes that the spiking probability can be defined as a sum of functions of the different stimulus components; that is , p ( s p i k e | s ) = g 1 ( k 1 ⊤ s ) + g 2 ( k 2 ⊤ s ) + ⋯ g N ( k N ⊤ s ) . ( 43 ) Rowekamp et al [43] compared such an approach to the joint optimization more common in MID analysis ( as in [18] ) , and derived the bias that results from sequential optimization and its implicit additivity . By contrast , we have focused here on parametrization rather than sequential optimization . In all cases , we optimized the log-likelihood simultaneously over all filter dimensions . For high-dimensional models , we advocate parametrization of the nonlinearity so as to avoid the curse of dimensionality . However , the CBF form we have introduced is more flexible than that of ePPR , both in that two- or more dimensional components are easily included , and in that the outputs of the components can be combined non-linearly . Other proposals can be seen as assuming specific quadratic-based parametrizations for the nonlinearity , that are more restrictive than the CBF form . The iSTAC estimator , introduced by Pillow & Simoncelli [19] , is based on maximization of the KL divergence between Gaussian approximations to the spike-triggered and stimulus ensembles—thus finding the feature space that maximizes the single-spike information under a Gaussian model of both the spike-triggered and stimulus ensembles . Park & Pillow [44] showed its relationship to an LNP model with an exponentiated quadratic spike rate , which takes the form: p ( s p i k e | s ) = exp ( a + K ⊤ s + s ⊤ C s ) . ( 44 ) Such a nonlinearity readily yields maximum likelihood estimators based on the STA and STC . Moreover , they proposed a new model , known as “elliptical LNP” , which allowed estimation of a non-parametric nonlinearity around the quadratic function ( instead of assuming an exponential form ) . Rajan et al . [24] considered a similar model within an information-theoretic framework and proposed extending it to nonlinear combinations of outputs from multiple quadratic functions . In a similar vein , Sharpee et al[45 , 46] used p ( s p i k e | s ) = 1 1 + exp ( a + K s + s ⊤ C s ) . ( 45 ) This model corresponds to quadratic logistic regression , and thus assumes Bernoulli output noise ( and a binary response ) . The authors also proposed a “nonlinear MID” in which the standard MID estimator is extended to by setting the firing rate to be a quadratic function of the form f ( k⊤s+s⊤C s ) . This method is one-dimensional in a quadratic stimulus space ( unlike multidimensional linear MID ) and therefore avoids the curse of dimensionality . Other work has used independent component analysis to find directions in stimulus space in which the spike-triggered distribution has maximal deviations from Gaussianity [8] .
We have studied the estimator known as maximally informative dimensions ( MID ) , [18] a popular approach for estimating informative dimensions of stimulus space from spike-train data . Although the MID estimator was originally described in information-theoretic language , we have shown that , when used with plug-in estimators for information-theoretic quantities , it is mathematically identical to the maximum likelihood estimator for a linear-nonlinear-Poisson ( LNP ) encoding model . This equivalence holds irrespective of spike rate , the amount of data , or the size of time bins used to count spikes . We have shown that this follows from the fact that the plug-in estimate for single-spike information is equal ( up to an additive constant ) to the log-likelihood per spike of the data under an LNP model . Estimators defined by the optima of information-theoretic functionals have attractive theoretical properties , including that they provide well-defined and ( theoretically ) distribution-agnostic characterizations of data . In practice , however , such agnosticism can be difficult to achieve , as the need to estimate information-theoretic quantities from data requires the choice of a particular estimator . MID has the virtue of using a non-parametric estimator for raw and spike-triggered stimulus densities , meaning that the number of parameters ( i . e . , the number of histogram bins ) can grow flexibly with the amount of data . This allows it to converge for arbitrary densities , in the limit of infinite data . However , for a finite dataset , the choice of number of bins is critical for obtaining an accurate estimate . As we show in Fig . 3 , a poor choice can lead to a systematic under- or over-estimate of the single-spike information , and in turn , a poor estimate of the most informative stimulus dimensions . Determining the number of histogram bins should therefore be considered a model selection problem , validated with a statistical procedure such as cross-validation . A second kind of distributional assumption arises from MID’s reliance on single-spike information , which is tantamount to an assumption of Poisson spiking . To be clear , the single-spike information represents a valid information-theoretic quantity that does not explicitly assume any model . As noted in [26] , it is simply the information carried by a single spike time , independent of all other spike times . However , conditionally independent spiking is also the fundamental assumption underlying the Poisson model and , as we have shown , the standard MID estimator ( based on the KL-divergence between histograms ) is mathematically identical to the maximum likelihood estimator for an LNP model with piece-wise constant nonlinearity . Thus , MID achieves no more and no less than a maximum likelihood estimator for a Poisson response model . As we illustrate in Fig . 4 , MID does not maximize the mutual information between the projected stimulus and the spike response when the distribution of spikes conditioned on stimuli is not Poisson; it is an inconsistent estimator for the relevant stimulus subspace in such cases . The distributional-dependence of MID should therefore be considered when interpreting its estimates of filters and nonlinearities . MID makes different , but not necessarily fewer , assumptions when compared to other LN estimators . For instance , although the maximum-likelihood estimator for a generalized linear model assumes a less-flexible model for the neural nonlinearity than does MID , it readily permits estimation of certain forms of spike-interdependence that MID neglects . In particular , MID-derived estimates are subject to concerns regarding model mismatch that arise whenever the true generative family is unknown [47] . In light of the danger that these distributional assumptions be obscured by the information-theoretic framing of MID , our belief is that the safer approach is to specify a model explicitly and adopt a likelihood-based estimation framework . Where the information theoretic and likelihood-based estimators are identical , nothing is lost by this approach . However , besides making assumptions explicit , the likelihood-based framework also readily facilitates the introduction of suitable priors for regularization of suitable priors , or hierarchical models [48 , 49] , or more structured models of the type discussed here . Having clarified the relationship between MID and LNP model the , we introduced two generalizations designed to recover a maximally informative stimulus projection when neural response variability is not well described as Poisson . From a model-based perspective , the generalizations correspond to maximum likelihood estimators for a linear-nonlinear-Bernoulli ( LNB ) model ( for binary spike counts ) , and the linear-nonlinear-Count ( LNC ) model ( for arbitrary discrete spike counts ) . For both models , we obtained an equivalent relationship between log-likelihood and an estimate of mutual information between stimulus and response . This correspondence extends previous work that showed only approximate or asymptotic relationships between between information-theoretic and maximum-likelihood estimators [20 , 24 , 25] . The LNC model is the most general of the models we have considered . It requires the fewest assumptions , since it allows for arbitrary distributions over spike count given the stimulus . It includes both LNB and LNP as special cases ( i . e . , when the count distribution is Bernoulli or Poisson , respectively ) . We could analogously define arbitrary “LNX” models , where X stands in for any probability distribution over the neural response ( analog or discrete ) , and perform dimensionality reduction by maximizing likelihood for the filter parameters under this model . The log-likelihood under any such model can be associated with an information-theoretic quantity , analogous to single-spike , Bernoulli , and count information , using the difference of log-likelihoods ( see also [35] ) : I l n x ≜ ∑ r , s p ( s ) p x ( r | s , θ ) log p x ( r | s , θ ) - ∑ r p x ( r | θ 0 ) log p x ( r | θ 0 ) , ( 46 ) where px ( r|s , θ ) denotes the conditional response distribution associated with the LNX model with parameters θ , and px ( r|θ0 ) describes the marginal distribution over r under the stimulus distribution p ( s ) . The empirical or plug-in estimate of this information is equal to the LNX model log-likelihood plus the estimated marginal entropy: I ^ l n x ( θ ) = 1 n ℒ l n x ( θ ; D ) - ℒ l n x ( θ 0 ; D ) , ( 47 ) where n denotes the number of samples and θ0 depends only on the marginal response distribution . The maximum likelihood estimate is therefore equally a maximal-information estimate . Note that all of the dimensionality-reduction methods we have discussed treat neural responses as conditionally independent given the stimulus , meaning that they do not capture dependencies between spike counts in different time bins ( e . g . , due to refractoriness , bursting , adaptation , etc . ) . Spike-history dependencies can influence the single-bin spike count distribution—for example , a Bernoulli model is more accurate than a Poisson model when the bin size is smaller than or equal to the refractory period , since the Poisson model assigns positive probability to the event of having ≥ 2 two spikes in a single bin . The models we have considered can all be extended to capture spike history dependencies by augmenting the stimulus with a vector representation of spike history , as in both conditional renewal models and generalized linear models [10 , 12 , 27 , 50–52] . Lastly , we have shown that viewing MID from a model-based perspective provides insight into how to overcome practical limitations on the number of filters that can be estimated . Standard implementations of MID employ histogram-based density estimators for p ( K⊤s ) and p ( K⊤s|spike ) . However , dimensionality and parameter count can be a crippling issue given limited data , and density estimation becomes intractable in dimensionalities > 3 . Furthermore , the dependence of the information on the logarithm of the ratio of these densities amplifies sensitivity to errors in these estimates . The LNP-likelihood view suggests direct estimation of the nonlinearity f , rather than of the densities . Such estimates are naturally more robust , and are more sensibly regularized based on expectations about neuronal responses without reference to any regularities in the stimulus distribution . We have proposed a flexible yet tractable form for the nonlinearity in terms of linear combinations of basis functions cascaded with a second output nonlinearity . This approach yielded a flexible , computationally efficient , constrained version of MID that is able to estimate high-dimensional feature spaces . It is also general in the sense that it encompasses standard MID , generalized linear and quadratic models , and other constrained models that scale tractably to high-dimensional subspaces . Future work might seek to extend this flexible likelihood-based approach further , for example by including priors over the weights with which basis functions are combined to improve regularization , or perhaps by adjusting hyperparameters in a hierarchical model as has been successful with linear approaches [48 , 49] . In recent years , the ability to successfully characterize low-dimensional neural feature spaces using MID has proved useful to address questions relating to multidimensional feature selectivity [53–56] . In all of these examples however , issues with dimensionality have prevented the estimation of feature spaces with more than two dimensions . The methods presented within this paper will help to overcome these issues , opening access to further important questions regarding the relationship between stimuli and their neural representation .
Here we present a derivation of the lower bound on the fraction of total information carried by silences for a Bernoulli neuron , in the limit of rare spiking . For notational convenience , let ρ = p ( r = 1 ) denote the marginal probability of a spike , so the probability of silence is p ( r = 0 ) = 1−ρ . Let Q1 = p ( s|r = 1 ) and Q0 = p ( s|r = 0 ) denote the spike-triggered and silence-triggered stimulus distributions , respectively . Let Ps = p ( s ) denote the raw stimulus distribution . Note that we have the Ps = ρQ1+ ( 1−ρ ) Q0 . The mutual information between the stimulus and one bin of the response ( Equation 18 ) can then be written I ( s , r ) = ρ D K L Q 1 | | P s + ( 1 - ρ ) D K L Q 0 | | P s . ( 48 ) Note that this is a generalized form of the Jensen-Shannon ( JS ) divergence; the standard JS-divergence between Q0 and Q1 is obtained when ρ = 1 2 . In the limit of small ρ ( i . e . , the Poisson limit ) , the mutual information is dominated by the first ( Q1 ) term . Here we wish to show a bound on the fraction of information carried by the Q0 term . We can do this by computing a second-order Taylor expansion of ( 1−ρ ) DKL ( Qo|Ps ) and I ( s , r ) around ρ = 0 , and show that their ratio is bounded below by ρ/2 . Expanding in ρ , we have ( 1 - ρ ) D K L Q o | | P s = 1 2 ρ 2 V ( Q 1 , Q 0 ) + O ( ρ 3 ) , and ( 49 ) I ( s , r ) = ρ D K L Q 1 | | Q 0 - 1 2 ρ 2 V ( Q 1 , Q 0 ) + O ( ρ 3 ) , ( 50 ) where V ( Q 1 , Q 0 ) = ∫ Ω Q 1 ( Q 1 Q 0 - 1 ) d s , ( 51 ) which is a an upper bound on the KL-divergence: V ( Q 1 , Q 0 ) ≥ D K L ( Q 1 | | Q 0 ) , since ( z−1 ) ≥ log ( z ) . We therefore have ( 1 - ρ ) D K L Q o | | P s I ( s , r ) = 1 2 ρ 2 V ( Q 1 , Q 0 ) + O ( ρ 3 ) ρ D K L Q 1 | | Q 0 - 1 2 ρ 2 V ( Q 1 , Q 0 ) + O ( ρ 3 ) ≥ ρ V ( Q 1 , Q 0 ) 2 D K L Q 1 | | Q 0 ≥ ρ 2 ( 52 ) in the limit ρ → 0 . We conjecture that the bound holds for all values of ρ . For the case of ρ = 1 2 , this corresponds to an assertion about the relative contribution of each of the two terms in the JS divergence , that is: D K L Q 1 | | 1 2 ( Q 0 + Q 1 ) D K L Q 1 | | 1 2 ( Q 0 + Q 1 ) + D K L Q 1 | | 1 2 ( Q 0 + Q 1 ) ≥ 1 4 ( 53 ) for any choice of distributions Q0 and Q1 . We have been unable to find any counter-examples to this ( or to the more general conjecture ) , but have so far been unable to find a general proof . An important general corollary to the equivalence between MID and an LNP maximum likelihood estimate is that the standard single-spike information estimate Îss based on a PSTH measured in response to repeated stimuli is also a Poisson log-likelihood per spike ( plus a constant ) . Specifically , the empirical single-spike information is equal to the log-likelihood ratio between an inhomogeneous and homogeneous Poisson model of the repeat data ( normalized by spike count ) : I ^ s s = 1 n s p ℒ ( λ ^ M L ; r ) - ℒ ( λ ¯ ; r ) , ( 54 ) where λ^ML denotes the maximum-likelihood or plug-in estimate of the time-varying spike rate ( i . e . , the PSTH itself ) , λ^ is the mean spike rate across time , and ℒ ( λ;r ) denotes the log-likelihood of the repeat data r under a Poisson model with time-varying rate λ . We can derive this equivalence as follows . Let {rjt} denote spike counts collected during a “frozen noise” experiment , with repeat index j ∈ {1 , … , nrpt} and index t ∈ {1 , … , nt} over time bins of width Δ . Then T = ntΔ is the duration of the stimulus and N = nt nrpt is the total number of time bins in the entire experiment . The single-spike information can be estimated with a discrete version of the formula for single-spike information provided in [26] ( see eq . 2 . 5 ) : I ^ s s = 1 n t ∑ t = 1 n t λ ^ ( t ) λ ¯ log λ ^ ( t ) λ ¯ , ( 55 ) where λ ̂ ( t ) = 1 Δ n r p t ∑ j = 1 n r p t r j t is an estimate of the spike rate in the t’th time bin in response to the stimulus sequence s , and λ ‾ = ( ∑ t = 1 n t λ ̂ ( t ) ) / n t is the mean spike rate across the experiment . Note that this formulation assumes ( as in [26] ) that T is long enough that an average over stimulus sequences is well approximated by the average across time . The plug-in ( ML ) estimator for spike rate can be read off from the peri-stimulus time histogram ( PSTH ) . It results from averaging the response across repeats for each time bin: λ ^ ( t ) = 1 n r p t Δ ∑ j = 1 n r p t r j t . ( 56 ) Clearly , λ ‾ = n s p N Δ , where nsp = ∑j , t rjt is the total spike count . This allows us to rewrite single-spike information ( Equation 55 ) as: I ^ s s = n r p t Δ n s p ∑ t = 1 n t λ ^ ( t ) log λ ^ ( t ) - log n s p N Δ . ( 57 ) Now , consider the Poisson log-likelihood ℒ evaluated at the ML estimate λ^= ( λ^ ( 1 ) , … , λ^ ( nt ) ) , i . e . , the conditional probability of the response data r = {rjt} given rate vector λ^ . This is given by: ℒ ( λ ^ ; r ) = ∑ t = 1 n t ∑ j = 1 n r p t r j t log λ ^ ( t ) Δ - λ ^ ( t ) Δ - log r j t ! = ∑ t = 1 n t ∑ j = 1 n r p t r j t log λ ^ ( t ) - n s p + n s p log Δ - ∑ t , j log r j t ! = n r p t Δ ∑ t = 1 n t λ ^ ( t ) log λ ^ ( t ) - n s p + n s p log Δ - ∑ t , j log r j t ! = n s p I ^ s s + n s p log n s p N - n s p - ∑ t , j log r j t ! = n s p I ^ s s + ℒ ( λ ¯ ; r ) , ( 58 ) which is identical to relationship between single-spike information and Poisson log-likelihood expressed in Equation 13 . Thus , even when estimated from raster data , Iss is equal to the difference between Poisson log-likelihoods under an inhomogeneous ( rate-varying ) and a homogeneous ( constant rate ) Poisson model , divided by spike count ( see also [57] ) . These normalized log-likelihoods can be conceived as entropy estimates , with −1nspℒ ( λ¯;r ) providing an estimate for prior entropy , measuring the prior uncertainty about spike times given the mean rate , and −1nspℒ ( λ^;r ) corresponding to posterior entropy , measuring the posterior uncertainty once we know the time-varying spike rate . A similar quantity has been used to report the cross-validation performance of conditionally Poisson models , including the GLM [13 , 58] . To penalize over-fitting , the empirical single-spike information is evaluated using the rate estimate λ^ obtained with parameters fit to training data and responses r from unseen test data . This results in the “cross-validated” single-spike information: I ^ s s [ x v ] = 1 n s p [ t e s t ] ℒ ( λ ^ [ t r a i n ] ; r [ t e s t ] ) - ℒ ( λ ¯ [ t e s t ] ; r [ t e s t ] . ( 59 ) This can be interpreted as the predictive information ( in bits-per-spike ) that the model captures about test data , above and beyond that captured by a homogeneous Poisson model with correct mean rate . Note that this quantity can be negative in cases of extremely poor model fit , that is , when the model prediction on test data is worse than of the best constant-spike-rate Poisson model . Cross-validated single-spike information provides a useful measure for comparing models with different numbers of parameters ( e . g . , a 1-filter vs . 2-filter LNP model ) , since units of “bits” are more interpretable than raw log-likelihood of test data . Generally , I ̂ s s [ x v ] can be considered to a lower bound on the model’s true predictive power , due to stochasticity in both training and test data . By contrast , the empirical Iss evaluated on training data tends to over-estimate information due to over-fitting . To gain intuition for the different information measures we have considered ( Poisson , Bernoulli , and categorical or “count” ) , it is useful to consider how they differ for a simple idealized example . Consider a world with two stimuli , ‘A’ and ‘B’ , and two possible discrete stimulus sequences , s1 = AB and s2 = BA , each of which occurs with equal probability , so p ( s1 ) = p ( s2 ) = 0 . 5 . Assume each sequence lasts T = 2s , so the natural time bin size for considering the spike response is Δ = 1s . Suppose that stimulus A always elicits 3 spikes , while B always elicits 1 spike . Thus , when sequence s1 is presented , we observe 3 spikes in the first time interval and 1 spike in the second interval; when s2 is presented , we observe 1 spike in the first time interval and 3 spikes in the second . Single-spike information can be computed exactly from λ1 ( t ) and λ2 ( t ) , the spike rate in response to stimulus sequence s1 and s2 , respectively . For this example , λ1 ( t ) , takes the value 3 during ( 0 , 1] and 1 during ( 1 , 2] , while λ2 ( t ) takes values 1 and 3 during the corresponding intervals . The mean spike rate for both stimuli is λ̄ = 2 sp/s . Plugging these into Equation 54 gives single-spike information of Iss = 0 . 19 bits/spike . This result is slightly easier to grasp using an equivalent definition of single-spike information as the mutual information between the stimulus s and a single spike time τ ( see [26] ) . If one were told that a spike , sampled at random from the four spikes present during every trial , occurred during [0 , 1] , then the posterior p ( s|τ = 1 ) attaches 3/4 probability to s = s1 and 1/4 to s = s2 . The posterior entropy is therefore −0 . 25 log 0 . 25−0 . 75 log 0 . 75 = 0 . 81 bits . We obtain the same entropy if the spike occurs in the second interval , so H ( s|τ ) = 0 . 81 . The prior entropy is H ( s ) = 1 bit , so once again we have Iss = 1−0 . 81 = 0 . 19 bits/spike . The Bernoulli information , by contrast , is undefined , since r takes values outside the set {0 , 1} , and therefore cannot have a Bernoulli distribution . To make Bernoulli information well defined , we would need to either truncate spike counts above 1 ( e . g . , [59] ) , or else use smaller bin size so that no bin contains more than one spike . In the latter case , we would need to provide more information about the distribution of spike times within these finer bins . If , for example , the three spikes elicited by A are evenly spaced within the interval and we use bins equal to 1/3s , then the Bernoulli information will clearly exceed single-spike information , since the time of a no-spike response ( r = 0 , a term neglected by single-spike information ) provides perfect information about the stimulus , since it occurs only in response to B . Lastly , the count information is easy to compute from the fact that count r carries perfect information about the stimulus , so the mutual information between stimulus ( A or B ) and r is 1 bit . We defined Icount to be the mutual information normalized by the mean spike count per bin ( Equation 35 ) . Thus , Icount = 0 . 5 bits/spike , which is more than double the single-spike information . Here we provide formulas useful for fitting the the many-filter LNP model with cylindrical basis function ( CBF ) nonlinearity . We performed joint optimization of filter parameters K and basis function weights {αi} using MATLAB’s fminunc function . We found this approach to converge much more rapidly than alternating coordinate ascent . We used analytically computed gradient and Hessian of the joint-likelihood to speed up performance , which we provide here . Given a dataset { ( st , rt ) }t=1nt , define r = ( r1 , … , rnt ) ⊤ and λ = ( f ( K⊤s1 ) , … , f ( K⊤snt ) ) ⊤ , where nonlinearity f = g ( ∑αi φi ) depends on basis function Φ = {φi} and weights α = {αi} ( Equation 39 ) . We can write the log-likelihood for the many-filter LNP model ( from Equations 38–40 ) as: ℒ ( θ ) = r ⊤ log λ - ( Δ ) 1 ⊤ λ ( 60 ) where θ = {K , α} are the model parameters , Δ is the time bin size , and 1 denotes a vector of ones . The first and second derivatives of the log-likelihood are given by ∂ ℒ ∂ θ i = ∂ λ ∂ θ i ⊤ r λ - Δ 1 ( 61 ) ∂ 2 ℒ ∂ θ i ∂ θ j = ∂ 2 λ ∂ θ i ∂ θ j ⊤ r λ - Δ 1 + ∂ λ ∂ θ i ∂ λ ∂ θ j ⊤ r λ 2 , ( 62 ) where multiplication , division , and exponentiation operations on vector quantities indicate component-wise operations . Let k1 , … , km denote the linear filters , i . e . , the m columns of K . Then the required gradients of λ with respect to the model parameters can be written: ∂ λ ∂ k i = S ⊤ ( λ ' ∘ Φ ( i ) α ) ( 63 ) ∂ λ ∂ α = Φ ⊤ λ ' ( 64 ) where S denotes the ( nt×D ) stimulus design matrix , Φ denotes the ( nt×nφ ) matrix whose ( t , j ) ’th entry is φj ( K⊤st ) , and Φ ( i ) denotes a matrix of the same size , formed by the point-wise derivative of Φ with respect to its i’th input component , evaluated at each projected stimulus K⊤st . Finally , λ′ = g′ ( Φα ) is a ( nt×1 ) vector composed of the point-wise derivatives of the inverse-link function g at its input , and ‘∘’ denotes Hadamard or component-wise vector product . Lastly , second derivative blocks , which can be plugged into Equation 62 to form the Hessian , are given by ∂ 2 λ ∂ k i ∂ k j = S ⊤ diag λ ' ' ∘ ( Φ ( i ) α ) ∘ ( Φ ( j ) α ) + λ ' ∘ Φ ( i , j ) α S ( 65 ) ∂ 2 λ α 2 = Φ ⊤ diag λ ' ' Φ ( 66 ) ∂ 2 λ ∂ k i ∂ α = S ⊤ diag λ ' ' ∘ ( Φ ( i ) α ) Φ + diag λ ' Φ ( i ) , ( 67 ) where λ′′ = g′′ ( Φα ) and Φ ( i , j ) is a matrix of point-wise second-derivatives of Φ with respect to i’th and j’th inputs , evaluated for each projected stimulus K⊤st . To examine performance in recovering high-dimensional subspaces , we analyzed data from macaque V1 cells , driven by 1D binary white noise “flickering bars” stimulus , presented at a frame rate of 100 Hz ( data published in [29] ) . The spatiotemporal stimulus had between 8 and 32 spatial bars and we considered 10 time bins for the temporal integration window . This made for a stimulus space with dimensionality ranging from 80 to 320 . The cbf-LNP model was implemented with a cylindrical basis function ( CBF ) nonlinearity using three first-order CBFs per filter . For a k-filter model , this resulted in 3k parameters for the nonlinearity , and ( 240+3 ) k parameters in total for a stimulus with 24 bars . The traditional MID estimator ( rbf-LNP ) was implemented using radial basis functions ( RBFs ) to represent the nonlinearity . Unlike the histogram-based parametrization discussed in the manuscript ( which produces a piece-wise constant nonlinearity ) , this results in a smooth nonlinearity and , more importantly , a smooth log-likelihood with tractable analytic gradients . We defined a grid of RBFs with three grid points per dimension , so that CBF and RBF models were identical for a 1-filter model . For a k-filter model , this resulted in 3k parameters for the nonlinearity , and 240k+3k parameters in total . For both models , the basis function responses were combined linearly and transformed by a “soft-rectification” function: g ( ⋅ ) = log ( 1+exp ( ⋅ ) ) , to ensure positive spike rates . We also evaluated the performance of an exponential function , g ( ⋅ ) = exp ( ⋅ ) , which yielded slightly worse performance ( reducing single-spike information by ∼ 0 . 02 bits/spike ) . The cbf- and rbf-LNP models were both fit by maximizing the likelihood for the model parameters θ = {K , α} . Both models were fit incrementally , with the N+1 dimensional model being initialized with the parameters of the N dimensional model , plus one additional filter ( initialized with the iSTAC filter that provided the greatest increase in log-likelihood ) . The joint likelihood in K and α was ascended using MATLAB’s fminunc optimization function , which exploits analytic gradients and Hessians . The models were fit to 80% of the data , with the remaining 20% used for validation . In order to calculate information contributed by excitatory filters under the cbf-LNP model ( Fig . 8F ) , we removed each filter from the model and refit the nonlinearity ( using the training data ) using just the other filters . We quantified the information contributed by each filter as the difference between log-likelihood of the full model and log-likelihood of the reduced model ( on test data ) . We sorted the filters by informativeness and computed the cumulative sum of information loss to obtain the trace shown in ( Fig . 8F ) . Measurements of computation time ( Fig . 8D ) were averaged over 100 repetitions using different random seeds . For each cell , four segments of activity were chosen randomly with fixed lengths of 5 , 10 , 20 and 30 minutes , which contained between about 22000 and 173000 spikes . Even with 30 minutes of data , 8 filters could be identified within about 4 hours on a desktop computer , making the approach tractable even for large numbers of filters . Code will be provided at http://pillowlab . princeton . edu/code . html . | A popular approach to the neural coding problem is to identify a low-dimensional linear projection of the stimulus space that preserves the aspects of the stimulus that affect a neuron’s probability of spiking . Previous work has focused on both information-theoretic and likelihood-based estimators for finding such projections . Here , we show that these two approaches are in fact equivalent . We show that maximally informative dimensions ( MID ) , a popular information-theoretic method for dimensionality reduction , is identical to the maximum-likelihood estimator for a particular linear-nonlinear encoding model with Poisson spiking . One implication of this equivalence is that MID may not find the information-theoretically optimal stimulus projection when spiking is non-Poisson , which we illustrate with a few simple examples . Using these insights , we propose novel dimensionality-reduction methods that incorporate non-Poisson spiking , and suggest new parametrizations that allow for tractable estimation of high-dimensional subspaces . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | The Equivalence of Information-Theoretic and Likelihood-Based Methods for Neural Dimensionality Reduction |
IFN-γ activates cells to restrict intracellular pathogens by upregulating cellular effectors including the p65 family of guanylate-binding proteins ( GBPs ) . Here we test the role of Gbp1 in the IFN-γ-dependent control of T . gondii in the mouse model . Virulent strains of T . gondii avoided recruitment of Gbp1 to the parasitophorous vacuole in a strain-dependent manner that was mediated by the parasite virulence factors ROP18 , an active serine/threonine kinase , and the pseudokinase ROP5 . Increased recruitment of Gbp1 to Δrop18 or Δrop5 parasites was associated with clearance in IFN-γ-activated macrophages in vitro , a process dependent on the autophagy protein Atg5 . The increased susceptibility of Δrop18 mutants in IFN-γ-activated macrophages was reverted in Gbp1−/− cells , and decreased virulence of this mutant was compensated in Gbp1−/− mice , which were also more susceptible to challenge with type II strain parasites of intermediate virulence . These findings demonstrate that Gbp1 plays an important role in the IFN-γ-dependent , cell-autonomous control of toxoplasmosis and predict a broader role for this protein in host defense .
Toxoplasma gondii is an apicomplexan protozoan parasite with a broad host range that is capable of causing significant disease in humans and animals [1] . Many wild or domestic animals serve as intermediate hosts , becoming infected either by ingestion of oocysts shed by cats [1] , or by carnivorous/omnivorous feeding that facilitates transmission [2] . Human toxoplasmosis is therefore zoonotic , with infection caused by ingestion of tissue cysts in undercooked meat or oocysts that may contaminate food or water [3] , [4] . Given the central role of the mouse in the completion of the life cycle of T . gondii , understanding the mechanisms of immune control in the mouse are relevant to human infection and may also identify pathways important in human resistance . Within North America and Europe , strains of T . gondii are largely comprised of one of three highly clonal genotypes , referred to as type I , II , and III [5] . These genotypes have highly different phenotypes in the laboratory mice , with type I strains being acutely virulent , type II strains having intermediate virulence , while type III strains are essentially avirulent [5] . Previous genetic crosses have revealed that these differences are due to a small number of polymorphic serine/threonine kinases that are secreted from the rhoptries ( ROPs ) into the host cell [6] . Among these ROP18 was identified based on the large genetic contribution it makes to differences in acute virulence between highly virulent type I , intermediate virulence type II , and avirulent type III strains [7] , [8] . A second locus that contributes more substantially to acute virulence differences between these strains types encodes a polymorphic family of pseudokinases called ROP5 [9] , [10] . Collectively , these two loci account for the major strain differences in virulence in the murine model , although other loci have also been implicated in pathogenesis [6] . Resistance to infection with T . gondii is largely mediated by IL-12 [11] driving expression of IFN-γ , which activates both toxoplasmastatic and toxoplasmacidal mechanisms [12] , in both hematopoietic and non-hematopoietic cells [13] . The primary mechanism of cell-autonomous killing in the mouse is due to IFN-γ induced expression of immunity-related GTPases ( IRGs ) [14] , which are essential for control of infection in macrophages in vitro [15] , and during in vivo infection with type II strains of T . gondii [16] , [17] . IRG-mediated clearance involves the cooperative recruitment and loading of the GTP-bound IRGs onto the parasitophorous vacuolar membrane ( PVM ) surrounding the parasite , with subsequent vesiculation and rupture of the vacuole , and destruction of the parasite [18] , [19] , [20] . A subset of IRG proteins , known as IRGM proteins modulate the activation state of the effector IRG proteins [21]: absence of these IRGM proteins causes spontaneous activation of IRG effectors that form aggregates , compromising their ability to combat pathogens [14] . Similarly , cells lacking Atg5 also show disruption in the function of IRG proteins , which accumulate as GTP-bound forms in cytoplasmic aggregates , and hence fail to clear susceptible strains of T . gondii [22] , [23] . The accumulation of IRGs on the PVM and the clearance of parasites in macrophages are blocked by ROP18 , which phosphorylates several IRG proteins , thus preventing their association with the PVM [24] , [25] . Although IRGs are expanded in rodents , they are absent or numerically reduced in many vertebrates including humans , while all vertebrate groups express members of another interferon-inducible gene family , the p65 guanylate-binding proteins ( GBPs ) [14] , [26] . GBPs are structurally related to the dynamins and another known antiviral protein family , the Mx proteins . GBPs range in size from 65–73-kDa and account for over 20% of the proteins induced after IFN-γ treatment [26] , [27] . The human genome encodes seven GBPs , while the mouse contains 13 GBPs including two alternative splice isoforms [28] , [29] . It has been reported that type I T . gondii parasites do not accumulate Gbp1 or Gbp2 on their PVM , while a large percent of both type II and type III parasites show accumulation of these proteins [30] , [31] , although the molecular basis for this is unknown . Recent work has shown that a deletion of a cluster of GBPs on chromosome 3 , including Gbp1 , 2 , 3 , 5 , 7 and the splice variant Gbp2ps , increases susceptibility to type II parasites both in vivo and in vitro [32] . Loss of Gbp1 or Gbp5 reduced the ability of mice to resist Listeria and Mycobacteria infection [28] , [33] , and loss of Gbp2 leads to susceptibility to T . gondii [34]; however , the role of other individual GBPs in the control of T . gondii has not been explored in vivo . Here we explored the role of Gbp1 in cell-autonomous resistance to T . gondii and probed the interaction between known parasite virulence factors , ROP5 and ROP18 , and the GBP pathway . We also investigate the role for the autophagy protein Atg5 in the homeostasis and function of GBPs and their interdependence on the IRG system in controlling resistance to infection with T . gondii .
To determine whether the recruitment of GBPs to the PVM surrounding intracellular parasite is blocked in a ROP18-dependent manner similar to IRGs , we localized Gbp1 in IFN-γ-activated bone marrow derived macrophages ( BMM ) . The influence of ROP18 on GBP recruitment was examined using a previously described transgenic parasites that express ROP18I in the type III background [8] . Type I parasites ( i . e . GT-1 strain ) largely prevented recruitment of Gbp1 ( Fig . 1A , B ) . Similarly , type III parasites expressing a kinase-active version of ROP18 ( i . e . CTG+ROP18 ) avoided accumulation of Gbp1 over the first two hr post infection ( Fig . 1A , B ) . In contrast , a significantly higher percent of type III vacuoles ( i . e . CTG strain ) were positively stained for Gbp1 on the PVM ( Fig . 1A , B ) . The kinase activity of ROP18 was required to prevent Gbp1 recruitment , as expression of a kinase dead ROP18 ( i . e . CTG+ROP18 D/A ) did not prevent recruitment to the PVM ( Fig . 1A , B ) . In addition to the active kinase ROP18 , it has recently been shown that the pseudokinase ROP5 is important for acute virulence of T . gondii [10] , [35] . Therefore , we examined the role of these two rhoptry proteins in the prevention of Gbp1 recruitment to the PVM using a loss-of-function approach . Recruitment of Gbp1 was monitored in IFN-γ-activated BMM infected with type I parasites lacking ROP18 ( RHΔku80Δrop18 ) or ROP5 ( RHΔku80Δrop5 ) , vs . the respective complemented strains ( RHΔku80Δrop18/ROP18 and RHΔku80Δrop5/ROP5 ) and the wild type strain ( RHΔku80 ) ( Fig . 1C , D ) . Wild type or complemented parasites expressing ROP18 and ROP5 essentially prevented Gbp1 accumulation on the PVM , while Δrop18 parasites showed significantly higher Gbp1 accumulation . Interestingly , Δrop5 parasites showed the highest level of Gbp1 recruitment ( Fig . 1C , D ) . Collectively these data indicate that known T . gondii virulence factors ROP18 and ROP5 are necessary to prevent Gbp1 accumulation on the PVM surrounding parasites in IFN-γ-activated macrophages . IRG proteins have previously been shown to localize to the PVM surrounding susceptible T . gondii parasites [18] , [19] , [24] , although similar findings have not been reported for GBPs . Cryo-immuno electron microscopy ( EM ) of IFN-γ-activated RAW 264 . 7 macrophages revealed that Gbp1 was localized diffusely in the cytosol of cells infected with wild type parasites , and not on the PVM ( Fig . 2A , B ) . In contrast , Gbp1 strongly localized in the vicinity of the PV in cells infected with Δrop18 parasites ( Fig . 2C , D ) . Although limited staining was observed on the PVM , Gbp1 was associated with nearby membrane vesicles that clustered around the vacuole ( Fig . 2E , F ) . The prominent localization of Gbp1 to vesicles surrounding the PVM suggests that it may be involved in delivery of other components to the compartment , or disposal of membrane following vesiculation and vacuole membrane rupture . Gbp1 positive vesicles also collect around phagosomes containing mycobacteria in IFN-γ-activated macrophages [28] , suggesting that Gbp1 recruitment could serve as a common protective mechanism against different pathogen classes . We also examined the morphological features of the PVM surrounding susceptible parasites by conventional EM ( Fig . 3 A–F ) . Vacuoles containing ROP5-deficient ( Fig . 3B ) or ROP18-deficient ( Fig . 3D ) parasites in IFN-γ-activated BMM from wild type mice were slightly distended with an enlarged lumen and showed marked vesiculation , membrane blebbing , and accumulation of small vesicles around the PV . In addition , both ROP5 and ROP18 deficient parasites showed evidence of vacuole rupture , leaving the parasite free in the cytosol , where it often underwent degradation ( Fig . S2 ) . In Gbp1−/− BMM cells activated with IFN-γ , vacuoles containing ROP5-deficient parasites also showed frequent vesiculation accompanied by an enlarged lumen ( Fig . 3A ) , while ROP18-deficient ( Fig . 3C ) parasites were found in more a closely-fitting vacuole surrounded by host cell mitochondria and which had a smooth circumference , characteristic of an intact PV . Enlargement of the PVM revealed that membrane blebbing around both ROP5 or ROP18-deficient parasites in wild type cells occurred with a marked curvature and regular scalloped pattern ( Fig . 3 E , F ) . These features are highly reminiscent of the previously described vesiculation of the PV membrane that accompanies IRG-recruitment to the PVM and vacuole destruction [18] , [23] , [24] . Notably , this process is interrupted the Gbp1−/− cells , at least in the terms of the fate of ROP18-deficient parasites , a result consistent with the survival vs . clearance of ROP5 or ROP18-deficient parasites in IFN-γ-activated BMM from wild type vs . Gbp1−/− mice , as described below . The autophagy protein Atg5 is required for resistance to type II parasite infection in mice [23] , in part due to the fact that in its absence , members of the IRG family are mislocalized into large aggregates in the host cell and hence are not recruited to parasite-containing vacuoles [22] , [23] . To examine the role for Atg5 in the localization of Gbp1 , we stained IFN-γ-activated Atg5-deficient ( Atg5 KO ) and wild type ( Atg5 WT ) mouse embryonic fibroblasts ( MEFs ) for Gbp1 and Irga6 . In wild type cells , Irga6 showed a diffuse cytoplasmic distribution ( Fig . 4A ) and these cells exhibited a normal homogenous cytosol where Gbp1 was distributed in small clusters , as shown by cryo-immuno EM ( Fig . 4B , C ) . In contrast , in the absence of Atg5 , Gbp1 was localized to large aggregates within the cell , and these structures partially co-localized with aggregates of Irga6 that was in the GTP-bound state ( as detected with a conformation specific antibody ) ( Fig . 4D , E ) . Transmission electron microscopy showed large accumulations of vesicles in the cytoplasm of Atg5-deficient cells ( Fig . 4F , G ) , which were absent in the wild type cells ( Fig . 3B ) . Cryo-immuno EM revealed an accumulation of Gbp1 in clusters of vesicles that accumulated in the Atg5-deficient cells ( Fig . 4H ) . Gbp1 positive vesicles were distinct from the endoplasmic reticulum ( ER ) , although they were interspersed with ER membranes , as indicated by staining for protein disulfide isomerase ( PDI ) ( Fig . 4H , I ) . To further assess the composition of the Gbp1 aggregates in the Atg5-deficient ( Atg5 KO ) MEFs , we examined their colocalization with ubiquitin and p62 , which were first reported to localize with Gbp1 in uninfected and mycobacterium-infected macrophages [28] , and also implicated in aggregates of Gbp2 that form in the absence of IRGM proteins [36] . Immunofluorescence labeling demonstrated that Gbp1 aggregates largely co-localize with p62 and ubiquitin ( Fig . 4J , K ) ; however , the aggregates were LAMP1 negative ( Fig . 4L ) . In comparison , the rare aggregates that also normally form in wild type cells were typically surrounded by LAMP1 positive vesicles , suggesting they fuse with lysosomes and are degraded ( Fig . S1 ) . Consistent with this model , the frequency of IRG-GBP aggregates was substantially increased by treatment with bafilomycin , which blocks lysosome fusion ( Fig . S1 ) . In bafilomycin treated cells ∼20% of cells showed aggregates of Gbp1 and Irga6 compared to ∼1% of DMSO treated cells . The altered distribution of Gbp1 and Irga6 in Atg5-deficient MEFs suggested that recruitment of other IRG members , such as Irgb6 , to the PVM of susceptible parasites might be impaired . To examine the role of Atg5 in recruitment of Irgb6 and Gbp1 in BMM , we took advantage of the previously described conditional deletion strain Atg5flox/flox+LysMcre , in which Atg5 is specifically ablated in myeloid cells [23] . In IFN-γ-activated BMM with functional Atg5 ( Atg5flox/flox ) , Δrop18 or Δrop5 parasites showed increased recruitment of Irgb6 ( Fig . 5A ) and Gbp1 ( Fig . 5B ) compared to the type I wild type or the ROP18 and ROP5 complemented strains . In contrast , BMM lacking functional Atg5 ( Atg5flox/flox+LysMcre ) showed significantly reduced Irgb6 ( Fig . 4A ) and Gbp1 ( Fig . 5B ) recruitment to all strains . This reversal in accumulation was particularly evident for Δrop18 and Δrop5 parasites , which normally show elevated accumulation of Irgb6 [24] and Gbp1 ( Fig . 1B ) . Infection of IFN-γ-activated macrophages revealed that both Δrop18 and Δrop5 parasites underwent enhanced clearance in wild type cells ( Atg5flox/flox ) ( Fig . 5C ) . This decrease in survival was reverted to normal in Atg5-deficient macrophages ( Atg5flox/flox+LysMcre ) ( Fig . 5D ) . Together , these data indicate that Atg5 plays an important role in homeostasis of Irgb6 and Gbp1 and in its absence , recruitment of these effectors to the PVM surrounding susceptible parasites is compromised , preventing parasite clearance in IFN-γ-activated macrophages . To examine the role of Gbp1 in resistance to T . gondii , we first tested the ability of IFN-γ-activated BMM from Gbp1-deficient mice ( Gbp1−/− ) to clear parasites following overnight infection in vitro . We infected BMM from C57BL/6 and Gbp1−/− mice with wild type parasites , which largely resist clearance in activated macrophages , and Δrop18 parasites , which are cleared by ∼50% in activated macrophages [24] . As expected , C57BL/6 BMM cleared Δrop18 parasites to about 50% of the initial infection ( Fig . 6A ) . Strikingly , Gbp1−/− BMM showed a reduced ability to clear susceptible Δrop18 parasites following overnight incubation , restoring survival to ∼80% , a level that was similar to wild type parasites in either strain of mice ( Fig . 6A ) . In contrast , the increased clearance of Δrop5 parasites seen in C57BL/6 BMM was not reversed in Gbp1−/− deficient BMM ( Fig . 6A ) . Recent studies on the deletion of the Gbpchr3 locus in the mouse indicated that the ability of IFN-γ-activated macrophages to prevent replication of intracellular parasites , referred to as stasis , is also compromised in the absence of this locus [32] . Therefore we examined the ability of activated wild type ( C57BL/6 ) or Gbp1−/− BMM to restrict the intracellular replication of T . gondii at 20 hr post infection . Under the activation conditions used here , the majority of intracellular T . gondii in wild type cells were found in vacuoles containing 1–2 parasites , with a few having replicated to clusters of 4 ( Fig . 6B ) . In contrast , the majority of intracellular T . gondii were found in rosettes of 8 , with the remainder largely being found in clusters of 4 . Although the level of stasis achieved in wild type cells was less than that reported previously [32] , the ability of Gbp1−/− BMM to control replication was significantly impaired ( Fig . 6B ) . To examine the kinetics of IRG vs . GBP recruitment , we examined the percentage of PVs containing ROP18 deficient parasites that became visibly positive over the first 2 hr post-infection in BMM . Irgb6 positive vacuoles were elevated at 30 min post-infection and they remained at similar levels during the first 120 min ( Fig . 6C ) . In contrast , accumulation of Gbp1 was delayed: the percentage of positive vacuoles was initially lower at 30 min and only plateaued at 90–120 min ( Fig . 6C ) . Over this time course , the majority of vacuoles became positive for both Irgb6 and Gbp1 ( 64 . 7±4 . 2% ) , while most of the remaining vacuoles stained only with Gbp1 ( 33 . 8±5 . 8% ) , and only a minority being Irgb6 positive only ( 2±2% ) . Combined with the differences in kinetics , these findings suggest that Gbp1 is recruited after Irgb6 , and that it remains on the vacuole after Irgb6 is recycled . Alternatively , Gbp1 may have additional IRG-independent mechanisms to target the PV , as discussed below . Although the percentage of Irgb6 positive vacuoles remained the same in Gbp1−/− cells at 30 min , it was significantly reduced at 2 hr when compared to wild type cells ( C57BL/6 ) ( Fig . 6D ) . In contrast , Irga6 positive PV increased slightly between 30 min and 2 hr in wild type and did not change in Gbp1−/− BMM ( Fig . 6D ) . These findings indicate that Gbp1 influences the recruitment and/or retention of some IRG proteins onto the PVM surrounding susceptible parasites . To assess the requirement of IRG proteins in the recruitment of Gbp1 to susceptible parasites , we examined accumulation of Gbp1 , Irgb6 , and Irga6 to Δrop5 parasites in Irgm3−/− vs . wild type ( C57BL/6 ) BMM at 2 hr post infection . Similar to previous reports showing that Irga6 and Gbp2 partially form aggregates in the cytosol of Irgm3−/− cells [36] , we observed that Irga6 , Irgb6 and Gbp1 showed focal clusters of staining in Irgm3−/− cells , although a majority of the proteins were still homogenously dispersed when examined by immunofluorescence microscopy ( data not shown ) . Despite the formation of some aggregates , Gbp1 accumulation on the PV was not significantly different in Irgm3−/− vs . wild type BMM , while Irga6 positive PV increased in Gbp1−/− knockout cells ( Fig . 6E ) . In contrast , the percent of Irgb6 positive PV was significantly lower in the Irgm3−/− BMM . These data support previous findings that Irgm3 is required for efficient Irgb6 loading onto the PV of susceptible parasites [9] , but reveal that is not required for either Irga6 or Gbp1 recruitment . Indirectly this implies that Irgb6 is also not required for Gbp1 recruitment to the PVM . Collectively , these results demonstrate that Gbp1 is critical for the control and clearance of T . gondii in IFN-γ-activated BMM in vitro , and suggests that it while it works cooperatively with the IRGs , it may not depend on them for recruitment . Next we wanted to examine how Gbp1 deficiency would affect in vivo challenge with parasites . We challenged C57BL/6 and Gbp1−/− mice with highly virulent type I strain RH vs . ROP18-deficient parasites and monitored their survival . Δrop18 parasites showed a significantly delayed time to death compared with wild type parasites in wild type mice ( Fig . 7A ) . The deficiency of Δrop18 parasites was partially reversed in Gbp1−/− mice , which succumbed to challenge 4 days earlier than C57BL/6 mice ( Fig . 7A ) . When challenged with the highly virulent wild type parasites , Gbp1−/− mice showed almost equivalent survival to C57BL/6 mice , although this is not unexpected given the high virulence of type I strains ( Fig . 7A ) . Additionally , the highly attenuated phenotype of the Δrop5 parasites in mice was not reversed in Gbp1−/− mice ( data not shown ) , consistent with the enhanced clearance of this parasite mutant in Gbp1−/− cells in vitro , both of which reflect much more severe defect in this mutant , as described previously [9] . To further explore the defect in resistance , we challenged Gbp1−/− and C57BL/6 mice with a moderately virulent type II parasite strain at two different doses and monitored their survival for 60 days . C57BL/6 mice were largely resistant to infection with type II parasites , with a single mouse succumbing to infection at day 13 post infection with 1 , 000 parasites ( Fig . 7B ) . Gbp1−/− mice were more susceptible to type II parasites with animals starting to succumb to challenge 3 days earlier than C57BL/6 mice . Additionally , significantly fewer Gbp1−/− mice survived challenge with either 500 or 1 , 000 parasites ( Fig . 7B ) . The delayed death phenotype of Gbp1−/− mice is reminiscent of type II infection in Nos2−/− mice , which succumb due to encephalitis [37] . Therefore , we examined the brains of surviving C57BL/6 and Gbp1−/− animals at day 60 days post infection for signs of encephalitis by H&E staining . Sections from the brain of an infected C57BL/6 mouse showed mild perivascular cuffing , minimal focal accumulation of lymphocytes , and a single tissue cyst ( Fig . 7C , D ) . Sections from the brain of an infected Gbp1−/− mouse showed more severe pathological changes including moderate focal gliosis , multifocal perivascular cuffing , moderate thickening of the meninges , and multiple tissue cysts ( Fig . 7 E , F ) . Collectively , these results demonstrate that Gbp1 plays an important role in the control of infection in vivo as Gbp1-deficent animals show increased susceptibility both during acute and chronic infection .
IFN-γ plays a crucial role in activating cells to control proliferation and destroy intracellular parasites . Here we demonstrate that Gbp1 plays an important role in this cell-autonomous control in vitro and in resistance to T . gondii infection in vivo . GBPs may work cooperatively with IRGs , which have previously been implicated in resistance to T . gondii , and both families of effectors rely on Atg5 for homeostasis . Our work underscores the importance of the GBPs in resistance in the mouse , and suggests that they may have similar roles in other hosts , including humans . Previous studies have shown that Gbp1 , Gbp2 , Gbp3 , Gbp6 , Gbp7 , and Gbp9 are recruited to PV containing T . gondii in IFN-γ-activated MEFs or RAW 264 . 7 macrophages and that the type I BK strain avoids recruitment of a subset of these ( i . e . Gbp1 , Gbp2 , Gbp3 , and Gbp6 ) [30] . Similar findings were reported using MEFs transfected with an epitope-tagged version of Gbp1 that was recruited to PV containing type II ( Pru strain ) and type III ( CEP strain ) , but not type I ( RH strain ) parasites [31] . This later study also made use of transgenic type III strain parasites that express the type 1 allele of ROP18 , which enhances virulence , and found that this resulted in decreased recruitment of Gbp1 [31] . Here , we confirm and extend these findings by showing that the virulent type I GT-1 strain blocks recruitment of Gbp1 , while PV containing avirulent type III ( CTG ) parasites accumulated this effector protein . We also show that the ability of ROP18 to confer protection on the normally susceptible type III strain CTG depends on the kinase activity , as parasites expressing a kinase dead form of ROP18 ( CTG+ROP18 D/A ) accumulated Gbp1 at levels slightly higher than CTG alone . Further , type I mutants lacking either ROP18 or ROP5 were also susceptible to recruitment of Gbp1 , a phenotype that was fully complemented by re-expression of the respective virulence factors . ROP18 has previously been shown to affect survival in IFN-γ-activated macrophages , a process that occurs due to selective phosphorylation of IRG proteins , thereby blocking GTPase activity and disrupting vacuole recruitment [24] , [25] . In contrast , ROP5 functions by disrupting oligomerization of Irga6 [38] , [39] , and/or by directly enhancing the catalytic activity of ROP18 [9] . The mechanisms by which the virulent alleles of ROP5 and ROP18 prevent GBP accumulation are presently uncertain but may involve similar functions . Gbp1 and Gbp2 have been shown to form tetramers in response to GTP binding , leading to cooperative activation of GTPase activity [28] , [40] . Mutants in the GTP binding domain of Gbp2 that block GTPase activity and alter formation of multimers , disrupts loading onto PV containing susceptible T . gondii [40] , while similar mutations in Gbp1 impairs control of both Listeria and Mycobacteria [28] . Whether GBPs are a direct target of ROP18 phosphorylation is presently unknown , although they contain regions similar to the described motif of ROP18 [24] . Further studies designed to elucidate the mechanisms by which ROP5 and ROP18 disrupt GBP recruitment are currently underway . Previous kinetic studies of IRG recruitment to PV containing susceptible parasites shows that Irgb6 and Irgb10 arrive early , followed by other IRG proteins such as Irga6 [22] . Given their previously established importance in the clearance process , Irga6 [41] and Irgb6 [24] were used here as sentinels of this pathway . Previous studies have shown that PV containing susceptible parasites are rapidly stripped of their membrane and the parasite within the vacuole is digested [22]; hence , the percentage of IRG positive vacuoles at any given time point reflects only a portion of the parasites that ultimately are destroyed in the overnight clearance assay . Our findings are consistent with this model in that we observed efficient loading of Irgb6 at early time points ( 30 min ) and the level increased slightly over the first 2 hr , consistent with previous reports [23] , [24] . Others have reported that Irgb6 accumulation to vacuoles containing susceptible parasites ( i . e . CTG ) is decreased in cells lacking the Gbpchr3 locus [32] . Here , we show that the initial recruitment of Irgb6 is normal at 30 min , but the retention of this marker is impaired with fewer PV remaining positive at 2 hr . Combined with the finding that Irgb6 co-precipitates with Gbp1–5 [32] , this suggests that accumulation of GBPs is associated with stabilization of IRGs and their retention on the vacuole membrane . Whether GBPs contribute directly to membrane blebbing and physical disruption of the PVM , or modulate the function of other effectors at this interface , remains to be demonstrated by further studies . Together these data suggests that destruction of the PV in response to IFN-γ requires a sequential recruitment of both IRG and GBP proteins , ultimately resulting in membrane disruption . Autophagy proteins also participate in cell autonomous clearance of parasites , although these may be due to indirect effects on the homeostasis of effectors such as IRGs and GBPs . Utilizing Atg5-deficient MEFs , we examined the localization of Gbp1 in IFN-γ-activated cells . Immunofluorescence and electron microscopy showed that Atg5-deficient cells have large membrane-associated aggregates of Gbp1 in their cytoplasm and that these co-localize with the active GTP-bound form of Irga6 , ubiquitin , and the autophagy adaptor protein p62 . Although , they do not co-localize with LAMP1 in Atg5−/− cells , they do proceed to autophagolysosomes in wild type cells and Gbp1 co-immunoprecipitates p62 from cell lysates in a manner that does not require the p62 ubiquitin-binding domain [28] . Similarly , previous studies have also shown that in the absence of IRGM proteins , both IRG and GBP proteins form co-aggregates that colocalize with the adaptor p62 and the autophagy protein LC3 [36] . Collectively , these results suggest that autophagy is necessary for removal of aggregates that spontaneously form in wild type cells , and which are degraded by a classical autophagy pathway that terminates in lysosomal fusion . The aberrant localization of Gbp1 in Atg5-deficient cells provided an opportunity to examine the functional consequences of this disruption on parasite clearance in macrophages that were selectively deleted for Atg5 using LysMcre recombination . Recruitment of both Irgb6 and Gbp1 to susceptible parasites ( RHΔku80Δrop18 and RHΔku80Δrop5 ) was significantly reduced in BMM lacking Atg5 . Additionally , clearance of these susceptible parasite strains was almost completely reversed in Atg5-deficient BMM . These effects are most likely due to the aberrant localization of IRGs and GBPs that occurred in the absence of Atg5 . Improper localization of IRGs and GBPs may be due to a requirement for autophagy to remove misfolded aggregates that otherwise drive inappropriate activation . Interestingly , we have observed a similar requirement for Atg7 and Atg16 in cell-autonomous control of T . gondii in IFN-γ-activated BMM in vitro , although this requirement may not extend to all mediators of the autophagy pathway ( S . Hwang and H . W . Virgin , unpublished ) . Alternatively , the requirement for Atg5 may not reflect a classical autophagy degradation pathway , but rather a role in the delivery of effectors to pathogen containing vacuoles , as suggested by the membranous accumulations that occur in the proximity of PV that are targeted for destruction ( Figure 2 , 3 present study , and [23] , [28] ) . Consistent with this , GBPs are localized to membrane vesicles within the cytoplasm of host cells , and several contain a C-terminal CaaX box allowing for isoprenylation that facilitates interaction with host cell membranes [28] , [30] . Regardless of the exact role of Atg proteins , this illustrates the critical balance of control of these two families of effectors , which are maintained in an inactive state in order to avoid damage to host membranes , yet need to be readily mobilized for pathogen clearance . Previous work has shown that a genetic knockout of the Gbpchr3 locus increases susceptibility to a type II strain T . gondii , which has an intermediate level of virulence in the mouse [32] . Analyzing the phenotype of this mutant is complicated due to the simultaneous disruption of five intact genes as well as a truncated form of Gbp2 . Characterization of this mouse revealed that it has normal expansion of CD4+ and CD8+ T cells , and produces normal levels of IL-12 and IFN-γ following infection with T . gondii [32] . Macrophages from these Gbpchr3−/− mice also produce normal levels of O2− and NO in response to IFN-γ [32] . Despite having intact responses , IFN-γ-activated BMM from these animals show specific defects in the control of T . gondii replication ( stasis ) , as well as ability to clear parasites ( clearance ) in vitro . This former phenotype suggests that GBPs play a role in the NO-mediated stasis that impairs parasite replication , while the later defect is due to decreased recruitment of Irgb6 and an inability to vesiculate and destroy PV [32] . Reconstitution experiments suggested that Gbp 1 , 5 and 7 contribute to this clearance defect in vitro , although the over-expression strategy used in this study complicates the interpretation of these findings . In the present work , we have examined a single knockout of Gbp1 to address the role of this protein in resistance to T . gondii . We observed that both the ability to induce stasis and the capacity for clearance of susceptible parasites in vitro was greatly reduced in IFN-γ-activated macrophages from Gbp1−/− vs . wild type mice . Hence , the phenotype of BMM from the Gbp1−/− mouse partially recapitulates the two major defects seen in cells from the Gbpchr3 deletion mouse when tested in vitro . Greater susceptibility was also seen in vivo , with Gbp1−/− mice succumbing to infection earlier with Δrop18 parasites , which are partially attenuated in virulence towards wild type mice . Furthermore , Gbp1 was important for resistance to challenge with an intermediate virulence , type II strain of T . gondii , with the majority of mice succumbing after the initial acute infection . Surviving Gbp1−/− mice showed elevated CNS pathology consistent with encephalitis as the cause of death . Recent studies using a single gene deletion of Gbp2−/− also reported loss of the ability to control replication in vitro and increased susceptibility to type II strain challenge during the chronic phase [34] . The increased chronic susceptibility of Gbp1−/− or Gbp2−/− mice is similar to that previously described for Nos2−/− mice , which lack inducible nitric oxide synthase and hence produce lower levels of NO [37] . However , Gbpchr3−/− mice produce normal levels of NO when stimulated with IFN-γ , despite losing the capacity to control parasite replication in vitro . Hence , the increased susceptibility in Gbp1−/− or Gbp2−/− mice to chronic infection may reflect reduced NO production locally within the CNS , or result from impaired clearance via the IRG or GBP pathways , leading to higher chronic burdens of infection . GBPs have recently been implicated in resistance to several pathogens in the mouse and yet they likely play overlapping yet different roles with respect to individual pathogens [26] . Gbp1−/− and Gbp5−/− mice were first shown to be important for control of Listeria and Mycobacteria infections [28] , [33] . Individual deletion of Gbp2 [34] or Gbp1 ( present report ) , renders mice susceptible to T . gondii , while additional genes within the Gbpchr3 locus may likewise contribute since the phenotype of this deletion is more severe than any of the single mutants [32] . In contrast , individual deletion of Gbp2ps , an alternatively spliced variant , or of Gbp5 does not affect susceptibility to i . p . challenge with type II ME49 parasites [32] . The diversity of GBPs may be an adaptation to target different intracellular pathogens: availability of additional gene disruptants will facilitate further testing of the role of individual GBPs in resistance to a variety of pathogens . As well , the diverse nature of pathogens affected by GBPs suggest that they play indirect roles in affecting delivery of other effectors , as reported previously for control of intracellular bacterial pathogens [28] , or the retention of Irgb6 to T . gondii vacuoles as shown here , and reported previously [32] . Although studies on GBPs conducted to date have focused on their role in resistance in the mouse , this family of proteins is highly conserved in vertebrates [42] and even some protochordates [33] . Several studies have examined their participation in resistance to infection in humans . Human GBP1 has been shown to contribute to anti-viral activity against vesicular stomatitis virus and encephalomyocarditis virus [43] and hepatitis C [44] , and recent studies implicate hGBP3 in resistance to influenza [45] . Previous work examining the role of GBPs in human cells show that GBP1 is recruited to the inclusion membrane of Chlamydia and over-expression contributes to smaller inclusion size [46] . The conservation of this family of proteins in humans and other higher order mammals [26] suggests that the GBPs play a more widespread role in resistance to infection .
Type I ( GT-1 ) , type III transfection control ( CTG Ble ) , type III parasites expressing ROP18 clone V1 ( CTG+ROP18 ) and type III parasites expressing a kinase-dead ROP18 clone L1 ( CTG+ROP18 D/A ) were described previously [8] . Type I RHΔku80ΔHx ( RHΔku80 ) parasites described previously [47] , were used here as wild type . Type I parasites lacking ROP18 ( RHΔku80Δrop18 ) [24] and the complemented strain ( RHΔku80Δrop18/ROP18 ) [35] , were described previously . A type I parasite strain lacking ROP5 , ( RHΔku80Δrop5 ) and the complemented strain ( RHΔku80Δrop5/ROP5 ) were described previously [35] . Luciferase expressing parasites were generated by transfection of parasites with pClickluc , as described previously [48] and isolation of single cell clones . Type II Prugnaud strain parasites expressing firefly luciferase and GFP ( PRU-Luc-GFP ) were provided by John Boothroyd ( Stanford University ) . Parasites were cultured in human foreskin fibroblasts grown in DMEM supplemented with 10% fetal bovine serum ( HyClone , Thermo Scientific , Rockford , IL ) , 2 mM glutamine , 10 mM HEPES pH 7 . 5 and 20 µg/ml gentamicin at 37°C under 5% CO2 . For all experiments , parasites were allowed to naturally egress and harvested shortly thereafter as described previously [49] . All animal experiments were conducted according to the U . S . A . Public Health Service Policy on Humane Care and Use of Laboratory Animals . Animals were maintained in an AAALAC-approved facility and all protocols were approved by the Institutional Care Committee ( School of Medicine , Washington University in St . Louis ) . CD-1 and C57BL/6 mice were obtained from Charles River Laboratory ( Wilmington , MA ) . Gbp1−/− mice , originally derived on a 129 background and backcrossed 6–8 times to C57BL/6 , were provided by the MacMicking laboratory as described previously [28] , and bred locally . Atg5flox/flox ( control ) mice and Atg5flox/flox+LysMcre mice ( 8–12 weeks old ) provided by the Virgin laboratory were bred locally and genotyped as described [23] . Irgm3−/− mice on a C57/BL6 background were provided by Greg Taylor ( Taylor et al . , 2000 ) , and bred locally . For in vivo challenges with type I strains , 8–12 week old female C57BL/6 and Gbp1−/− mice were infected s . c . with 100 freshly egressed parasites and survival was monitored for 30 days as described [8] , [49] . For in vivo challenges with PRU-Luc-GFP parasites , 8–12 week old ( male and female , matched per group ) C57BL/6 and Gbp1−/− mice were infected i . p . with either 500 or 1 , 000 freshly egressed parasites and survival was monitored for 60 days . Bone marrow-derived macrophages ( BMMs ) and RAW 264 . 7 macrophages were cultured as described previously [24] . Immortalized MEFs were cultured in DMEM supplemented with L-glutamine and 10% FBS . Where indicated , cells were activated by treatment with murine recombinant IFN-γ ( R&D Systems , Minneapolis , MN ) and LPS ( E . coli O55:B5 ) ( Sigma-Aldrich , St . Louis , MO ) for 18–24 hr before use . The PVM was localized with mouse mAb anti-GRA5 Tg17–113 [50] , or rabbit polyclonal sera to GRA7 [51] . Intracellular T . gondii parasites were localized with either mouse mAb anti-SAG1 DG52 or rabbit polyclonal sera to RH strain tachyzoites . Gbp1 and was localized with rabbit polyclonal sera raised against peptides specific to each protein [30] . Irgb6 was localized with rabbit anti-Irgb6 [52] or goat anti-TGTP ( Santa Cruz Biotechnology , Dallas , TX ) as indicated . Irga6 was localized with mouse mAb 10D7 , which recognizes the GTP bound form . p62 was localized with guinea pig polyclonal antibody specific to the C-terminus ( Progen , Heidelberg , Germany ) . Ubiquitin was localized with mouse mAb FK2 ( EMD Millipore Corporation , Billerica , MA ) . LAMP1 was localized with rat mAb ID4B , obtained from the Developmental Studies Hybridoma Bank ( http://dshb . biology . uiowa . edu ) . Secondary antibodies conjugated to Alexa Fluor 488 or 594 ( Invitrogen , Grand Island , NY ) were used for detection by immunofluorescence . Cells for immunofluorescence were fixed in 4% formaldehyde , permeabilized with 0 . 05% saponin , and stained using primary and secondary antibodies as described previously [24] . Samples were visualized using a Zeiss Axioskop 2 MOT Plus microscope equipped for epifluorescence and using a 63× PlanApochromat lens , N . A . 1 . 40 ( Carl Zeiss , Inc . , Thornwood , NY ) . Images were acquired with an AxioCam MRm camera ( Carl Zeiss , Inc . ) using Axiovision v4 . 6 , and processed using similar linear adjustments for all samples in Photoshop CS4 v9 . For deconvolution , images were acquired as above using automatic Z-stack acquisition in Axiovision and deconvolved using the nearest neighbor algorithm . To examine the distribution of host cellular proteins , samples were fixed between 30 min to 2 hr post infection ( see legends for specific details ) . After staining with appropriate primary and secondary antibodies , recruitment was determined by first visualizing the parasitophorous vacuole or parasite marker , then assessing whether there was an accumulation of host protein around each parasite . The percentage of positive parasites was determined from 10 representative fields that were examined at 63× . To examine intracellular clearance , IFN-γ-activated macrophages were infected with freshly egressed parasites , and then either fixed at 30 min post infection , or returned to culture in complete medium for 20 hr . Clearance was assessed by comparing the percentage of cells infected at 30 min vs . those remaining after 20 hr , as described previously [24] . For 3 biological replicates , 10 fields were counted on each of three coverslips that were examined at 63× . To examine intracellular replication , BMM were activated with IFN-γ and infected with wild type parasites for 30 min , washed to remove extracellular parasites , and recultured overnight in complete medium . At 20 hr post infection , cells were fixed , permeabilized as above , and visualized using mAb to SAG1 ( DG52 ) directly conjugated to Alexa Fluor 594 . The number of parasites per vacuole was determined from counting 100 vacuoles per sample from three separate coverslips . In general , 15–30% of cells were singly infected at 30 min . In order to combine experiments with different initial infection rates , the data were normalized by expressing the infection rate at 20 hr as a percentage of the infection rate at 30 min . Samples for EM were activated with IFN-γ and LPS for 18–24 hr . Where indicated , cells were infected with freshly egressed parasites for 30 min , washed three times with PBS then fixed at 2 and 6 hr post infection . For ultrastructural analysis , cells were fixed in 2% paraformaldehyde/2 . 5% glutaraldehyde ( Polysciences Inc . , Warrington , PA ) in 100 mM phosphate buffer , pH 7 . 2 for 1 hr at room temperature , processed and examined as described previously [23] , [24] . For immuno-EM , cells were fixed in 4% paraformaldehyde/0 . 05% glutaraldehyde ( Polysciences Inc . , ) in 100 mM PIPES/0 . 5 mM MgCl2 , pH 7 . 2 for 1 hr at 4°C , and processed as described previously [24] . Sections were stained with mouse anti-protein disulfide isomerase ( Enzo Life Sciences , Inc . Farmington , NY ) and rabbit anti-Gbp1 antibodies for 1 hr at room temperature , followed by gold-conjugated secondary antibodies ( Jackson ImmunoResearch Laboratories , Inc . , West Grove PA ) . Sections were stained and viewed with a JEOL 1200EX transmission electron microscope ( JEOL USA Inc . , Peabody , MA ) , as described previously [24] . Parallel controls omitting primary antibodies were consistently negative at the concentration of colloidal gold conjugated secondary antibodies used in these studies . Animals were sacrificed at 60 days post infection; the brain was removed and fixed in 10% neutral- buffered formalin . Tissues were dehydrated in ethanol and embedded in paraffin , and 5 micron sections were stained with hematoxylin and eosin ( H&E ) . Sections were evaluated by veterinary pathologist in the Department of Comparative Animal Medicine , Washington University . Statistical analyses were conducted using Microsoft Excel and PRISM . Excel results were compared using the Student's t tests performed under the assumption of equal variance and with a two-tailed test where P≤0 . 05 was considered significant . Survival statistics were compared using log-rank and Gehan-Breslow-Wilcoxin tests in PRISM . Data were graphed as means ± standard deviation of the population ( S . D . P . ) , or as standard deviation ( S . D . ) , as noted . | Emerging evidence suggests that the p65 family of guanylate-binding proteins ( GBPs ) , which is upregulated by interferon gamma , play an important role in host defense against intracellular pathogens . We demonstrate that the ability of virulent strains of Toxoplasma gondii to avoid recruitment of mouse Gbp1 is mediated by two parasite virulence factors; the serine threonine kinase ROP18 and the pseudokinase ROP5 , which controls its activity . GBP proteins required the autophagy protein Atg5 for proper cellular trafficking , recruitment to parasite-containing vacuoles , and pathogen control , strengthening the link between innate immunity and autophagy . The attenuation of mutants lacking ROP18 , which show increased susceptibility to clearance by macrophages and decreased virulence in mice , was reverted by deletion of Gbp1 , indicating this host factor is needed for resistance to T . gondii . Collectively , these findings demonstrate a key molecular interaction between host defenses mediated by GBPs and parasite virulence factors that thwart innate immunity . As GBPs are phylogenetically conserved among vertebrates , including humans , they likely play a broader role in host resistance . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology",
"biology",
"microbiology"
] | 2013 | Guanylate-binding Protein 1 (Gbp1) Contributes to Cell-autonomous Immunity against Toxoplasma gondii |
Schistosomiasis and STH are among the list of neglected tropical diseases considered for control by the WHO . Although both diseases are endemic in Zimbabwe , no nationwide control interventions have been implemented . For this reason in 2009 the Zimbabwe Ministry of Health and Child Care included the two diseases in the 2009–2013 National Health Strategy highlighting the importance of understanding the distribution and burden of the diseases as a prerequisite for elimination interventions . It is against this background that a national survey was conducted . A countrywide cross-sectional survey was carried out in 280 primary schools in 68 districts between September 2010 and August 2011 . Schistosoma haematobium was diagnosed using the urine filtration technique . Schistosoma mansoni and STH ( hookworms , Trichuris trichiura , Ascaris lumbricoides ) were diagnosed using both the Kato Katz and formol ether concentration techniques . Schistosomiasis was more prevalent country-wide ( 22 . 7% ) than STH ( 5 . 5% ) . The prevalence of S . haematobium was 18 . 0% while that of S . mansoni was 7 . 2% . Hookworms were the most common STH with a prevalence of 3 . 2% followed by A . lumbricoides and T . trichiura with prevalence of 2 . 5% and 0 . 1% , respectively . The prevalence of heavy infection intensity as defined by WHO for any schistosome species was 5 . 8% ( range 0%–18 . 3% in districts ) . Only light to moderate infection intensities were observed for STH species . The distribution of schistosomiasis and STH varied significantly between provinces , districts and schools ( p<0 . 001 ) . Overall , the prevalence of co-infection with schistosomiasis and STH was 1 . 5% . The actual co-endemicity of schistosomiasis and STH was observed in 43 ( 63 . 2% ) of the 68 districts screened . This study provided comprehensive baseline data on the distribution of schistosomiasis and STH that formed the basis for initiating a national control and elimination programme for these two neglected tropical diseases in Zimbabwe .
Schistosomiasis and soil transmitted helminthiasis ( STH ) are among the most widely distributed neglected tropical diseases ( NTDs ) that affect people living in vulnerable communities with poor and limited access to safe water , sanitary facilities and inadequate health facilities [1] , [2] , [3] . Worldwide , over 200 million people are infected with schistosomiasis [1] and more than 1 billion are infected with STH [2] . Morbidity due to schistosomiasis and STH include impairment of cognitive development in young children resulting in poor educational outcome [4] , [5] , [6] , [7] . Soil transmitted helminths also cause anaemia due to worm induced blood loss and compromised nutrition [8] , [9] , intestinal obstruction as well as reduced absorption of vitamin A , impacting on growth [10] , [11] . Of major concern , schistosomiasis has been shown to be a pre-disposing factor for HIV infection [12] . Furthermore , iron deficiency anaemia is exacerbated by co-infection with schistosomiasis , STH and Plasmodium malaria [13] . Schistosomiasis and STH may cause reduced birth weight , worker productivity and poor socio-economic development [9] . Thus , control of the two diseases will provide an ancillary contribution towards achieving some Millennium Development Goals; namely , poverty alleviation , universal primary education , reduced child mortality and improved maternal health [14] . A renewed interest for the integrated control of schistosomiasis , STH and other NTDs emerged following the 54 . 19th World Health Assembly resolution that specified the need to give treatment to at least 75% or all primary school age children at risk of morbidity due to schistosomiasis and STHs by the year 2010 [9] , [15] , [16] , [17] . In 2006 the World Health Organisation introduced the preventive chemotherapy ( PCT ) strategy encouraging integrated control of overlapping neglected tropical diseases in co-endemic communities . For that to be achieved , data on the extent of overlap in distribution and morbidity levels of the targeted NTDs is essential as it justifies appropriate intervention . Pursuant to the World Health Organisation's launch of the 2020 Roadmap for implementation of NTD intervention strategies with specific goals for control and elimination , the NTD control initiative received the first ever collaborative donor commitment . Merck KGaA committed to donating 250 million praziquantel tablets annually for an indefinite period towards the treatment of African school children for schistosomiasis . GlaxoSmithKline committed itself to donating up to 400 million tablets per year for STH treatment of school-age children world-wide [14] . The London declaration on NTDs of 30 January 2012 resulted in the largest coordinated action to address NTDs in which partners pledged new levels of collaboration including funding for mapping , tracking and reporting of progress on NTD control ( the London Declaration 2012 ) . Member states have expressed political support to the initiative through ( i ) appointment of steering committees for NTD control and elimination ( ii ) appointment of the National Focal Persons responsible for spearheading NTD mapping and integrated control in endemic countries [18] . In Zimbabwe , two major national surveys were conducted between 1985 and 1992 [19] , [20] . In addition numerous small-scale studies have shown distribution of schistosomiasis in Zimbabwe [21] , . In 2012 , Chimbari highlighted major pilot studies that were aimed at controlling schistosomiasis in various parts of Zimbabwe . In contrast studies demonstrating the existence of STH were conducted in few loci , namely Burma Valley farming areas and Kariba [23] , [24] , [25] , [26] , [27] . Given that Zimbabwe experienced major environmental and socio-economic changes in the past decade and that previous national surveys focused on schistosomiasis in rural areas , the need for a survey mapping both schistosomiasis and STH throughout the country is necessary in order to plan appropriate integrated control strategies [28] . We conducted a national survey to map the current distribution of schistosomiasis and for the first time , to map the distribution of STH in Zimbabwe . Furthermore , we determined current morbidity levels for both infections in order to provide information required for an integrated control program .
A school based cross sectional survey was conducted nationwide in rural based provinces between September-October 2010; in metropolitan provinces ( Harare and Bulawayo ) and Chitungwiza town from July - August 2011 . The extension of the national survey from 2010 to 2011 was due to limited financial resources for the project in 2010 . The survey was a joint collaboration between the Ministry of Health and Child Care ( MOHCC ) and the Ministry of Education Sport Arts and Culture ( MOESC ) . Zimbabwe is divided into 58 administrative districts many of which are rural . Although districts in metropolitan provinces are not recognized by the MOHCC , the MOESC recognises 6 districts in Harare and 5 districts in Bulawayo . For the purpose of this study districts in Harare and Bulawayo were considered , and Chitungwiza , a third largest urban area was divided into two districts ( Seke-Chitungwiza and Zengeza ) giving a total of 71 districts . Of the 71 districts , the national survey was conducted in 68 ( 95 . 8% ) as well as peri-urban areas surrounding Harare and Bulawayo . Only three districts ( Gweru , Kwekwe and Bindura ) were left out due to limited resources . Ten teams , each made up of two laboratory technicians , one technical assistant , a District Community Nurse , one District Education Officer , the District Environmental Health Officer and a driver conducted data collection . One of the two laboratory technicians was drawn from National Institute of Health Research to lead the team with the overall responsibilities of organising and managing field data collection , filing and safe keeping of research results . This technician was also responsible for performing the urine filtration and the Kato Katz techniques . The second technician drawn from the province was responsible for performing the formol ether concentration technique as well as assisting the team leader in executing other duties . The technical assistant helped in processing specimens as well as cleaning filters , templates and sieves for re-use . The community nurse was responsible for treatment of study participants whilst the District Environmental Officer was responsible for ensuring clean environment in school and at every stage during sample collection as well as feeding children during treatment . The District Education Officer was responsible for locating primary schools randomly selected for the national survey , introducing the research team to the school authorities and for the random selection of primary school children enrolling into the study ( participants ) . Prior to sample collection , data collection teams were trained on data collection methodologies for 5 days to ensure standardization of data collection procedures . Each of the 10 core teams was given a manual for field data collection and was assigned to collect data in one province for 30 days . Prior to the deployment of the data collection teams , the Secretary for Health and Child Care wrote to all provinces and districts informing Provincial and District Health Executives about the school based national schistosomiasis and STH survey as well as requesting for support including transport and treatment logistics in schools . The Secretary for Education Sport Arts and Culture wrote to all provinces and districts informing the Provincial and District Education Executives about the school based national schistosomiasis and STH survey as well as seeking support for the national data collection teams . The proposal to conduct the national schistosomiasis and STH survey was approved by the national ethical review board , the Medical Research Council of Zimbabwe . The ethical approval number for the study MRCZ/A/1207 dated 11th March 2010 . The Secretary for Education Sport Arts and Culture also approved the study . Written informed consent was sought from the parents/guardian of the study participants . UNICEF delivered Parental/guardian informed consent forms addressed to each school by the Secretary for Education Sport Arts and Culture throughout the country in advance to allow school heads sufficient time to liaise with parents/guardians for their consent . On the day of sample collection , only the assenting children whose consent forms were signed by their parents/guardians participated in the study . Enrollment into the study was voluntary . Participants were free to withdraw from the study at any time . Primary school children aged 10–15 years were targeted for the study as they constitute the high-risk age group for schistosomiasis and STH in the community [1] and hence are a proxy of the burden of schistosomiasis and STH in the population [29] , [30] . There were exceptions in which non-targeted younger and older children were included in the study population to fulfil the required sample size per school ( n = 50 ) . The national sample size was based on the total enrollment of primary school children , n = 2 490 568 ( MOESC 2005 ) . A sample of 15 818 children was calculated using EPI Info 6 statistical package ( Epi Info version 6 , Centers for Disease Control and Prevention , Atlanta , GA 30333 ) using 37% as the assumed mean prevalence of schistosomiasis [20] and the error margin of 0 . 75% . The number of schools selected per district was determined by dividing the district sample size calculated proportionally from the national sample size by the number of children that would be screened per school ( n = 50 ) . Simple random sampling was used to select schools per district using the lottery method [29] . This involved listing names of all primary schools in each district on small slips of paper , then after a thorough mixing of the names , five schools per district were selected one by one . At each school 50 children equally distributed by gender were randomly selected using the lottery method [1] , [31] . While the primary school children aged 10–15 years constituted the desired sampling frame , children aged 6–9 ( n = 598 ) and some aged >15 years ( n = 6 ) were included in some schools where the number of children aged 10–15 years was less than 50 . Single urine , about 100 ml and whole stool specimens were collected in 100 ml screw cap plastic specimen bottles from each child between 1000 and 1400 hours , a period when peak egg excretion is expected [32] , [33] . S . mansoni and STH were diagnosed using a combination of two diagnostic techniques , ( i ) the Kato Katz technique [34] and ( ii ) the formol ether concentration technique [35] in order to improve sensitivity for intestinal helminths diagnosis [36] . A single Kato Katz thick smear was prepared from stool given by each individual for the diagnosis of STHs and S . mansoni . This involved straining stool through a Kato Katz sieve with a mesh size of 250 µm . The fine stool was filled in a Kato Katz template producing 41 . 7 mg . Using the measured fine stool , a Kato Katz thick smear was prepared on a slide and this was covered with the cellophane coverslips soaked in 50% glycerine–malachite green . The slide was examined within 60 minutes of preparation in order to detect and count hookworm eggs before they clear . The slide was left to clear for at least 24 hours after which it was re-examined for S . mansoni ova . Using a pear-sized stool ( about 1 g ) from the remaining stool specimen , the formol ether concentration technique was performed [35] . Results from the Kato Katz and the formal ether concentration techniques for each individual were combined as follows: a person was considered negative for each STH species and S . mansoni if no ova of these parasites were detected using both techniques . A person was considered positive for STH species or S . mansoni if ova were detected by either of the two or both techniques . Urinary schistosomiasis ( S . haematobium ) was diagnosed using the urine filtration technique [37] . The technique involves filtration of 10 ml of a thoroughly mixed urine specimen , through a Nytrile filter ( 12–14 µm pore size ) . Since the formol ether concentration technique is not quantitative infection intensities for S . mansoni and STH were estimated from results obtained using the Kato Katz technique only whilst infection intensity for S . haematobium was estimated using the urine filtration techniques In order to estimate infection intensities for intestinal worms , slides were examined from the 41 . 7 mg stool and all eggs counted . The number of eggs counted was multiplied by 24 to obtain the no of eggs per gram of stool . Schistosoma haematobium egg intensity was expressed as the number of eggs per 10 ml urine . The infection intensities of schistosomes and STH species were classified as light , moderate or heavy according to the World Health Organisation thresholds [1] . Using the stratified infection intensities , the prevalence of heavy infection by any schistosome species ( morbidity ) was expressed as the number of subjects heavily infected with any schistosome species divided by the number of subjects investigated [38] . Sixty-eight districts included in the national survey were classified into different classes of morbidity . Schistosomiasis intervention strategies for districts in each class were proposed based on WHO guidelines for the elimination of schistosomiasis [38] . Due to the low prevalence and light to moderate infection intensities observed , the prevalence of heavy infection with any STH species was not calculated in this study . In Zimbabwe , a district is an implementation unit . The observed combined prevalence of schistosomiasis was used to classify implementation units ( districts ) into treatment strategies based on the World Health Organisation prevalence thresholds as follows: high risk area ( prevalence ≥50% ) ; moderate risk area ( prevalence ≥10% and <50% ) and low risk area ( prevalence >1% and <10% ) [38] . The combined prevalence of STH was also used to stratify implementation units into treatment strategies as follows: high-risk area ( prevalence ≥50% ) and low risk area ( <15% ) . Due to low prevalence of STH in Zimbabwe the threshold for the low prevalence was reduced from <20% recommended by the World Health Organisation [38] to 15% . The prevalence of schistosomiasis -STH co-infections was also calculated . In this study individual schistosomiasis-STHs co-infection was considered for individuals who were infected with at least one schistosome species and at least one soil transmitted helminths species . District S . haematobium and S . mansoni co-endemicity was defined as the co-existence of S . haematobium and S . mansoni in the same implementation unit whether due to individuals being co-infected or due to different individuals being infected with different schistosome species District Schistosomiasis -STH co-endemicity was defined as the existence of at least any schistosome species and at least one STH species in the same district whether due to individuals being co-infected or due to different individuals being infected with different helminths . Following submission of stool and urine samples all participants received orange juice and bread to eat after which they simultaneously received a single dose of praziquantel and albendazole in tablet form at the recommended doses of 40 mg/kg body weight and 400 mg respectively regardless of their infection status since both drugs are considered safe [30] . Data collected from the field was entered and analysed using the Statistical Package for Social Scientists ( SPSS ) version 8 . 0 ( SPSS Inc , Chicago , IL , USA ) . Differences in prevalence of infection among different groups were tested for statistical significance using the Chi-square test . The student's t-test was used to determine the difference in mean age between males and females . The significance level was set at a p-value of 0 . 05 . Geographical positions of study schools were used to produce maps using geographical information system ( GIS ) software . The GIS data codification and cleaning was carried out using Microsoft Excel . The clean and coded data were imported into a Microsoft Access database . Further analysis was conducted with Microsoft Access to integrate the different datasets into a single dataset . Then the integrated table was re-imported into Excel sheet before its conversion into the GIS data format . MapInfo ( Pitney Bowes Software Inc . ) release 6 . 5 was used to process the data . A GIS data table was created for the schools using their geographic coordinates , which were converted into decimal format from their initial GPS reading format . Out of 280 included in the national survey , 256 ( 91 . 4% ) had geographic coordinates and were successfully integrated in the school GIS database . The prevalence class breaks are the recommended categorisation by World Health Organization [30] .
Of the 280 schools included in the national survey , schistosomiasis was observed in 237 ( 84 . 6% ) schools . In schools where schistosomiasis was detected , 116 ( 48 . 9% ) had moderate prevalence ( ≥10% but <50% ) and 46 ( 19 . 5% ) had high prevalence ( ≥50% ) . Table 1 shows the overall combined prevalence of schistosomiasis: S . haematobium and S . mansoni; and the overall combined prevalence of STH species: hookworms , A . lumbricoides and T . trichiura stratified by gender and province . The overall combined prevalence of schistosomiasis was 22 . 7% ranging from 3 . 3% to 39 . 3% in provinces , 0% to 62% in districts and from 0% to 83 . 7% in schools respectively . Significant variations in prevalence of schistosomiasis between provinces , districts and schools were observed ( p<0 . 0001 ) . The prevalence of combined schistosomiasis was significantly higher in males ( 25 . 4% ) than in females ( 20 . 0% ) , ( p<0 . 0001 ) . Schistosomiasis was predominantly distributed in the north , north-east and eastern regions extending through the central plateaux and the eastern highlands to the south east . The prevalence of schistosomiasis was low in the entire western region of the country . Of the 12 252 participants screened for any STH , the overall combined prevalence of STH was 5 . 5% , ranging from 0% to 18 . 3% in provinces , 0% to 45% in districts and 0% to 78 . 7% in schools . There was no significant difference in prevalence of STH between males ( 7 . 5% ) and females ( 6 . 9% ) , p = 0 . 231 . The prevalence of STH was highest in Binga district ( 45 . 5% , 95%CI = 38 . 46–52 . 67 ) followed by Mutoko ( 43 . 5% , 95%CI = 35 . 55–51 . 72 ) and Murehwa district ( 40 . 6% , 95%CI = 34 . 07–47 . 46 ) . There was significant variation in prevalence of STH between provinces , districts and schools ( p<0 . 0001 ) . Overall , STH were predominantly distributed in the north , north east; eastern region and scantly distributed in the western region of Zimbabwe . Of the 13165 children screened for schistosomiasis ( either S . haematobium or S . mansoni ) , 10 389 ( 78 . 9% ) lived in rural areas , 886 ( 6 . 7% ) lived in peri-urban areas , 12 . 2% ( 1 061 ) lived in urban high-density areas and 2 . 2% ( 289 ) lived in urban-low density areas . Figure 1 describes the prevalence of schistosomiasis by type of settlement . There was a significant variation in prevalence of schistosomiasis by settlement type: rural area ( 26 . 5% , 95%CI = 25 . 60–27 . 31 ) ; peri-urban area ( 9 . 9% , 95%CI = 8 . 04–12 . 09 ) , urban high-density areas ( 8 . 9% , 95%CI = 7 . 52–10 . 37 ) and urban-low density areas ( 1 . 7% , 95%CI = 0 . 56–3 . 99 ) , p<0 . 0001 . Of the 12 252 children screened for any STH , 9 828 ( 80 . 2% ) lived in the rural areas , 830 ( 6 . 8% ) lived in peri-urban areas , 1 380 ( 11 . 3% ) lived in urban high-density areas and 214 ( 1 . 8% ) lived in urban low-density areas . There was a significant variation in prevalence of STH between settlement types ( p<0 . 0001 ) . The prevalence of STH was 6 . 5% ( 95%CI = 6 . 00–6 . 99 ) in rural areas; ( 1 . 7% , 95%CI = 0 . 92–2 . 81 ) in peri-urban areas; ( 1 . 6% , 95% = 1 . 00–2 . 40 ) urban high-density areas and ( 1 . 9 , 95%CI = 0 . 50–4 . 67 ) urban-low density areas , p<0 . 0001 . Table 2 describes the prevalence of different species of schistosomes and STH . Overall , the prevalence of S . haematobium was 18 . 0% ( 95%CI = 17 . 38–18 . 71 ) , ranging from 3 . 2% to 30 . 5% in the provinces , 0% to 55 . 9% in districts and 0% to 76 . 0% in schools . The prevalence of S . haematobium was significantly higher in males ( 20 . 8% ) than in females ( 15 . 4% ) , p<0 . 0001 . There was a significant variation in prevalence of S . haematobium by settlement type: 20 . 6% ( 95%CI = 19 . 87–21 . 44 ) in rural areas; 9 . 8% ( 95%CI = 7 . 93–12 . 02 ) in peri-urban areas , 8 . 6% ( 95%CI = 7 . 27–10 . 09 ) in urban –high density areas and 1 . 7% ( 95%CI = 0 . 57–4 . 00 ) in urban-low density areas , p<0 . 0001 . Figure 2a describes the point prevalence of S . haematobium indicating its predominance in the country compared with S . mansoni that is discontinuously distributed . Schistosoma haematobium is highly endemic in the central plateaux , north , north east and the eastern region extending to the south and south east . The overall prevalence of S . mansoni was 7 . 2% ( 95%CI = 6 . 74–7 . 66 ) ranging from 0% to 20 . 4% in provinces , 0% to 43 . 7% in districts and 0% to 73 . 6% in schools . Chiredzi district in Masvingo province registered the highest prevalence of S . mansoni ( 43 . 7% ) followed by Hwedza district in Mashonaland East and Nyanga in Manicaland province that had the prevalence of 32 . 3% and 31 . 5% respectively . There was no significant difference in prevalence of S . mansoni by gender ( p>0 . 231 ) . There was a significant difference in S . mansoni prevalence between provinces , districts and schools , p = <0 . 0001 . Schistosoma mansoni prevalence varied significantly between settlement types: 8 . 8% ( 95%CI = 8 . 27–9 . 40 ) in rural areas; 0 . 7% ( 95%CI = 0 . 35–1 . 33 ) in urban high-density areas , 0 . 5% ( 95%CI = 0 . 13–1 . 23 ) in peri-urban areas and 0% in urban-low density areas , p>0 . 0001 . Figure 2b describes the point prevalence of S . mansoni indicating common occurrence in some parts of the northern region of Zimbabwe ( Mashonaland Central province ) , eastern region ( Manicaland province ) and in the south east of the country ( Masvingo province ) . It was almost non-existent in the western region of Zimbabwe . Of the 12 252 participants screened for any STH , the prevalence of hookworm , A . lumbricoides and T . trichiura was 3 . 2%; 2 . 5% and 0 . 1% respectively ( Table 2 ) . Hookworm was the most common STH species occurring in all provinces in Zimbabwe except for Matabeleland South ( Table 2 and Figure 2c ) . Whilst hookworm accounted for the highest prevalence of STH observed in Binga ( 45 . 5% ) , A . lumbricoides was predominant in the north-eastern region of the country ( Mashonaland East Province ) and the Eastern Highlands ( Manicaland Province ) . The point prevalence of A . lumbricoides is described in Figure 2d which shows that A . lumbricoides was scantly distributed in Zimbabwe and occurred in the north , north-east and Eastern Highlands . Trichuris trichiura occurrence is insignificant in Zimbabwe . It was observed in few points in the north-eastern region and the Eastern Highlands of Zimbabwe . The overall arithmetic mean egg intensity for S . haematobium and standard error ( SE ) was 15 . 0 eggs/10 ml ( 0 . 84 ) , range ( 0–4 000 eggs/10 ml ) and that for S . mansoni was 7 . 65 epg ( 0 . 67 ) , range ( 0–4 152 epg ) . Table 3 describes schistosome infection intensities stratified from light to heavy infection according to the WHO guidelines [1] . The overall prevalence of S . haematobium light and heavy infection intensity was 12 . 4% and 5 . 6% respectively . The overall prevalence of light , moderate and heavy infection intensity for S . mansoni was 3 . 6%; 1 . 4% and 0 . 3% respectively . In order to assess the level of morbidity due to schistosomiasis , the prevalence of heavy infection with any schistosome species was determined and stratified by gender and province ( Table 3 ) . Overall , the prevalence of individuals heavily infected with any schistosome species was 5 . 8% . More males ( 7 . 1% ) were heavily infected than females ( 4 . 7% ) , p = 0 . 001 . The highest prevalence of heavy infection intensity with any schistosomes species was observed in Masvingo province ( 9 . 7% ) followed by Midlands ( 9 . 6% ) and Mashonaland Central province ( 9 . 1% ) respectively Table 4 classifies the Zimbabwean districts according to the prevalence of heavy infection intensities with any schistosome species ( morbidity ) . Of the 68 districts included in the national schistosomiasis and STH survey , 32 districts ( 50% ) had prevalence of heavy infection with any schistosome species ≥5% . Seventeen districts ( 25% ) had prevalence of heavy infection with any schistosome species ≥1% but <5% . Four districts had the prevalence of heavy infection with any schistosome species >0% but <1% and 15 ( 22 . 1% ) districts had no detectable morbidity . The overall arithmetic mean egg intensity and standard error ( SE ) for hookworm was 2 . 99 epg ( 0 . 67 ) , range ( 0–4 488 epg ) ; for A . lumbricoides it was 10 . 97 epg ( 4 . 07 ) , range ( 0–33 912 epg ) and 0 . 25 epg ( 0 . 17 ) , range ( 0–1 416 epg ) for T . trichiura . The overall prevalence of light infection intensity for hookworm was 0 . 9% , moderate infection intensity was 0 . 04% . Only 1 person had heavy infection intensity . The prevalence of light infection intensity for A . lumbricoides was 1 . 0% , moderate infection intensity was 0 . 03% . Heavy infection intensity with A . lumbricoides was not observed . The prevalence of light infection intensity for T . trichiura was 0 . 1% and that for moderate infection intensity was 0 . 02% . Heavy infection intensity for T . trichiura was not observed . Table 5 describes the prevalence of schistosomiasis and STH co-infections by province . Only those participants diagnosed for any or all of the schistosome species and any or all of the STH species ( n = 12 257 ) were considered in this analysis . Overall , 1 . 5% of the participants had schistosomiasis - STH co-infections , 21 . 6% had single infection from schistosomiasis and 4 . 0% had single infection from STH . Schistosomiasis - STH co-infections were observed in 43 ( 62 . 3% ) of the 68 districts . There was a significant difference in prevalence of schistosomiasis -STH co-infections between provinces and districts with the highest prevalence occurring in Murehwa ( 19 . 6% ) followed by Mutoko ( 18 . 2% ) and Masvingo ( 5 . 9% ) districts respectively , p<0 . 0001 . There was also a significant difference in the prevalence of schistosomiasis-STH co-infections between settlement types with the highest prevalence occurring in the rural areas ( 1 . 9% ) followed by the peri-urban areas ( 0 . 5% ) , low density urban areas ( 0 . 4% ) and high density urban areas ( 0 . 1% ) respectively , p<0 . 0001 . Overall , S . haematobium – S . mansoni co-infections were observed in 252 ( 2 . 1% ) of the 12140 participants screened for both schistosomes species . The overlap of S . haematobium and S . mansoni was predominant in the northern region of the country ( Mashonaland central province ) in which the prevalence of co-infection was 7 . 0% . The prevalence of S . haematobium - S . mansoni co-infection was 12 . 8% in two districts , Hwedza which is located in the eastern region of the country ( in Mashonaland East province ) and Mt . Darwin located to the north of Zimbabwe ( in Mashonaland Central province ) . Chiredzi district is located to the south of Zimbabwe ( Masvingo province ) had a prevalence of S . haematobium and S . mansoni co-infection of 10 . 5% . Table 6 shows the classification of Zimbabwean districts according to the prevalence and overlap of schistosomiasis and STH observed in this study . The prevalence categories are recommended by WHO and form the basis for which different intervention strategies are specified for the control of schistosomiasis and STH [30] . Of the 68 districts included in the national survey , the prevalence of schistosomiasis ≥50% was observed in 5 rural based districts . Prevalence of schistosomiasis ≥10% but <50% was observed in 39 of the 68 districts ( 57 . 4% ) . Eighteen districts ( 26 . 5% ) had the prevalence of schistosomiasis >0% but <10% . Seven rural and urban districts had prevalence of 0% whilst peri-urban areas around Harare and Bulawayo had an average schistosomiasis prevalence of 10 . 7% and 1 . 9% respectively . Eight districts had the prevalence of STH ≥15% whilst 35 ( 51 . 5% ) districts had the prevalence of STH >0 . 0% but <15% . Schistosomiasis was not detected in 9 of the 25 districts that were non-endemic to STH . Schistosomiasis–STH co-endemicity was observed in 41 of the 68 districts ( 60 . 3% ) . Peculiar was the observed high prevalence of both schistosomiasis and STH in Murehwa: 40 . 6% and 43 . 5% respectively , and Mutoko: 47 . 4% and 36 . 1% districts respectively . Figure 3 shows the recommended Preventive Chemotherapy ( PCT ) strategies based on prevalence and schistosomiasis-STH co-infections according to WHO guidelines . On the basis of district level prevalence results ( Table 6 ) , annual mass drug administration for the control of schstosomiasiss was predicted in 5 districts red colour ) whilst annual mass drug administration ( MDA ) for the control of STH was predicted in 7 districts . Biennial MDA for schistosomiasis control was predicted in 33 districts ( Figure 3 , blue and brown colours ) .
The present study provides comprehensive baseline data showing geographic distribution of schistosomiasis and STH , their co-endemicity and the prevalence of heavy infection with any schistosome species in Zimbabwe . Intervention strategies intended to direct the country in controlling morbidity , eliminate and interrupt schistosomiasis and STH transmission are suggested . Uninterrupted annual MDA is recommended in districts where the prevalence of schistosomiasis and STH is ≥50% and ≥15% respectively . Similarly , annual MDA is recommended in districts where the prevalence of heavy infection by any schistosome species is ≥10% . Complementary strategies including health education , provision of safe water and adequate sanitary facilities should concurrently be implemented . The intervals for the co-administration of different medicines directed at different PCT neglected tropical diseases should be dependent on the NTD with the highest prevalence in any implementation unity . | Schistosomiasis ( S . haematobium and S . mansoni ) and soil transmitted helminthiasis ( STH ) are endemic in Zimbabwe but there has not been any study conducted to systematically map out the disease for control . Due to the public health significance of schistosomiasis and STH in Zimbabwe , a nationwide cross-sectional survey was conducted to map schistosomiasis and STH in 2010 and 2011 , among 13 , 195 primary school children in 280 primary schools from 68 rural and urban districts in Zimbabwe . Urine filtration technique was used to determine the prevalence and intensity of S . haematobium . Both the Kato Katz and formol ether concentration techniques were used to determine the intensity and prevalence of S . mansoni and STH ( hookworm , A . lumbricoides and Trichuris trichiura ) . Overall , the prevalence of schistosomiasis was 22 . 7% ( ranging from 0%–62% in districts ) and that of STH was 5 . 5% , ranging from 0%–45% in districts . There was an overlap of schistosomiasis and STH in space ( 43 of the 68 districts ) . Based on the results of this study that confirm nationwide predominance in distribution of schistosomiasis and a lesser extent of STH , Preventive Chemotherapy strategies were determined and recommended for the elimination of these two specific NTD in Zimbabwe . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"public",
"and",
"occupational",
"health",
"research",
"design",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"clinical",
"medicine",
"diagnostic",
"medicine",
"primary",
"care",
"clinical",
"research",
"design",
"global",
"health",
"biology",
"a... | 2014 | Distribution of Schistosomiasis and Soil Transmitted Helminthiasis in Zimbabwe: Towards a National Plan of Action for Control and Elimination |
Ticks ( Family Ixodidae ) transmit a variety of disease causing agents to humans and animals . The tick-borne flaviviruses ( TBFs; family Flaviviridae ) are a complex of viruses , many of which cause encephalitis and hemorrhagic fever , and represent global threats to human health and biosecurity . Pathogenesis has been well studied in human and animal disease models . Equivalent analyses of tick-flavivirus interactions are limited and represent an area of study that could reveal novel approaches for TBF control . High resolution LC-MS/MS was used to analyze the proteome of Ixodes scapularis ( Lyme disease tick ) embryonic ISE6 cells following infection with Langat virus ( LGTV ) and identify proteins associated with viral infection and replication . Maximal LGTV infection of cells and determination of peak release of infectious virus , was observed at 36 hours post infection ( hpi ) . Proteins were extracted from ISE6 cells treated with LGTV and non-infectious ( UV inactivated ) LGTV at 36 hpi and analyzed by mass spectrometry . The Omics Discovery Pipeline ( ODP ) identified thousands of MS peaks . Protein homology searches against the I . scapularis IscaW1 genome assembly identified a total of 486 proteins that were subsequently assigned to putative functional pathways using searches against the Kyoto Encyclopedia of Genes and Genomes ( KEGG ) database . 266 proteins were differentially expressed following LGTV infection relative to non-infected ( mock ) cells . Of these , 68 proteins exhibited increased expression and 198 proteins had decreased expression . The majority of the former were classified in the KEGG pathways: “translation” , “amino acid metabolism” , and “protein folding/sorting/degradation” . Finally , Trichostatin A and Oligomycin A increased and decreased LGTV replication in vitro in ISE6 cells , respectively . Proteomic analyses revealed ISE6 proteins that were differentially expressed at the peak of LGTV replication . Proteins with increased expression following infection were associated with cellular metabolic pathways and glutaminolysis . In vitro assays using small molecules implicate malate dehydrogenase ( MDH2 ) , the citrate cycle , cellular acetylation , and electron transport chain processes in viral replication . Proteins were identified that may be required for TBF infection of ISE6 cells . These proteins are candidates for functional studies and targets for the development of transmission-blocking vaccines and drugs .
Tick-borne flaviviruses ( TBFs; family Flaviviridae ) are a complex of positive , single-stranded RNA viruses , many of which cause hemorrhagic fever and encephalitis in humans and are associated with high morbidity and mortality [1 , 2] . Humans are incidental hosts for TBFs that are transmitted by an infected tick ( subphylum Chelicerata , subclass Acari; superfamily Ixodida ) during blood feeding . Tick-borne encephalitis virus ( TBEV ) is the most prevalent TBF worldwide and is responsible for over 10 , 000 confirmed cases of encephalitis globally per annum [3 , 4] . Several TBFs associated with hemorrhagic disease are identified on the Centers for Disease Control and Prevention ( CDC ) “Select Biological Agents and Toxins” list ( http://www . selectagents . gov/ ) due to their high virulence ( biosafety level 3 and 4 ) , anticipated ability to establish zoonotic transmission cycles , and their potential use in bioterrorism . Of these , Kyasanur Forest Disease virus ( KFDV ) is responsible for an estimated 400–500 human cases per year in India [5–7] while Omsk hemorrhagic fever virus ( OHFV ) is estimated to cause an average of 24 human cases per year ( 1946–2000 ) [8] . In the U . S . , The increasing incidence of human cases of Powassan virus ( POWV ) and the corresponding genotype virus , Deer Tick virus ( DTV ) [9 , 10] in the northeast and upper mid-west of the U . S . , has refocused attention on TBFs in North America . Langat virus ( LGTV ) was discovered in Southeast Asia in the 1950s [11] . LGTV exhibits low levels of virulence to humans , is classified as biosafety level 2 ( BSL2 ) and employed routinely as a model for more virulent TBFs such as TBEV , KFDV , OHFV , and POWV/DTV . Other than for TBEV [5 , 12] , there are no vaccines or therapeutics available to prevent or treat infection with these virulent TBFs . Globally , there is an urgent need to identify novel prophylactics and therapeutics against TBFs . The NIH-funded Ixodes scapularis ( Lyme disease tick ) Genome Project represents the first genome assembly for a tick and an important resource to understand the molecular processes in ticks [13] . The IscaW1 . 2 annotation comprises 20 , 450 gene models predicted via a combination of ab initio methods and manual curation . These models are a source of new targets [14] for the identification of novel chemistries [15] and vaccines [16–18] for control of ticks and tick-borne diseases . Research has shown that proteins and metabolites produced by human [19 , 20] and mosquito [21–24] cells ( i . e . , “host-cell factors” ) may facilitate or play essential roles in flaviviral infection [25–28] . The mechanisms by which these molecules contribute to the pathogenesis of the Flaviviridae , are not well understood . Proteomics has been used to investigate interactions between ticks and bacterial pathogens [29–31] . Studies have also investigated global changes in the transcriptome of I . scapularis and in tick cells following LGTV infection [32] , although there is little known about how these responses correlate to changes at the protein level . Tick proteins that facilitate viral infection and replication in the arthropod vector are logical targets for interventions aimed at disrupting transmission of TBF . Here we developed an in vitro assay using the I . scapularis ISE6 embryonic cell line [33–35] and LGTV ( TP21 wildtype strain ) . We performed high-resolution LC-MS/MS analyses to evaluate global changes in the proteome of tick cells following flavivirus infection and identified proteins that displayed increased and decreased expression . We describe the cellular response to infection and employ small molecule functional assays to evaluate the involvement of several tick proteins in the infection and replication of LGTV in ISE6 cells .
Ixodes scapularis embryonic ISE6 cells ( provided by T . Kurtti , University of Minnesota , Minneapolis , MN ) were cultured at 34°C in L15B-300 medium in the absence of CO2 [36 , 37] . Baby hamster kidney 15 ( BHK15; ATCC cell provider ) cells , used for plaque assay and immunofluorescent focus assay ( IFA ) , were cultured at 37°C in Minimum Essential Medium ( MEM ) supplemented with L-glutamine , non-essential amino acids ( NEAA ) , and 10% heat-inactivated fetal calf serum ( FCS ) with 5% CO2 . Green African monkey kidney ( Vero; ATCC cell provider ) cells , used to create LGTV stock and for IFA to determine LGTV stock titer , were cultured at 37°C in MEM supplemented with L-glutamine , NEAA and 10% heat-inactivated FCS with 5% CO2 . LGTV TP21 wildtype strain , passage 2 ( obtained from A . Pletnev , NIH-NAID , Bethesda , MD [38] ) stock was amplified in Vero cells ( multiplicity of infection 0 . 01 ) [39] and grown as described above , except with 2 . 5% heat-inactivated FCS , up to passage 4 ( p4 ) to provide a working stock for experimental infections . Serial IFAs were conducted in parallel as previously described [40] in 96-well cell culture plates to determine LGTV stock titers . To create non-infectious LGTV ( UV-LGTV ) , LGTV p4 stock medium was placed in 48 well cell culture plates and treated with UV radiation at a distance of 11 cm from a standard ( 12 . 4 watt ) UV lamp in a biological safety cabinet ( Nuaire Labgard ES , Plymouth , MN ) for 30 second intervals over a five minute period . LGTV inactivation was confirmed by blind passage of UV-LGTV on ~2 x 107 ISE6 cells and ~80% confluent BHK15 cells , followed by immunofluorescent and plaque assay as described by Perera et al . [41] to demonstrate lack of infectivity . IFAs were used to assess the level of LGTV infection in ISE6 cell populations . Detection of the LGTV non-structural protein 3 ( NS3 ) was performed using YP-conjugated chicken anti-LGTV NS3 ( provided by S . Best , NIH-NAID , Hamilton , MT ) as primary antibody and IgG-conjugated goat anti-chicken , Alexa Fluor 488 ( Invitrogen , Grand Island , NY; A11039 ) as secondary antibody . Cell nuclei were labeled with 4' , 6-diamidino-2-phenylindole ( DAPI; Life Technologies , Grand Island , NY; D1306 ) . Glass coverslips were used to culture and infect cells for the IFAs and were placed onto microscope slides , which were viewed on an Olympus model IX81F-3 microscope and images were collected using an Olympus U-CMAD3 camera . Fluorescence excitation was provided by the EXFO X-Cite Series 120PC and Olympus IX2-UCB . Image overlays were produced with Metamorph Basic v7 . 6 . 5 . 0 software . To establish an MOI and time-point corresponding to optimal LGTV replication in ISE6 cells , three concentrations ( MOIs of 7 , 13 , and 26 ) of LGTV were used to infect cells . For each , cells were fixed at 3 , 9 , 24 , and 48 hpi with five technical replicates that were imaged under 20x magnification . On the basis of complete infection ( >96% ) of ISE6 cell populations between two MOIs ( 7 and 13 ) and time points ( 24 and 48 ) , an MOI of 10 was selected for subsequent experiments for maximum infection . Separately , an assessment of the cumulative virus release was carried out in LGTV-infected ISE6 cells at a MOI of 10 . Medium from these LGTV-infected ISE6 cells was harvested at 12 hour intervals for up to 120 hours , and subjected to plaque assays to measure replication . Peptides with homology to I . scapularis , IscaW1 . 2 gene models were assigned to putative functional class by searching accession numbers against the KEGG orthology database ( http://www . genome . jp/kegg/ko . html ) and the KEGG pathway database ( http://www . genome . jp/kegg/pathway . html ) . ISE6 proteins with orthology to KEGG entries were populated within KEGG pathways that also included mammalian and arthropod orthologs . The concentration of cells in each sample ( cells/ml ) was estimated by counting cell number on a Scepter 2 . 0 Automated Cell Counter with 40 μM Scepter sensors ( EMD Millipore; PHCC20040 ) in order to equalize cell numbers between biological replicates and between treatment groups prior to protein extraction . For cell population and growth analyses , initial cell counts ( cells/mL ) were determined manually using a hemocytometer and subsequently verified by sample analysis on the Scepter 2 . 0 Automated Cell Counter . Trichostatin A ( Sigma-Aldrich; T8552 ) and Oligomycin A ( Sigma-Aldrich; 75351 ) were separately re-suspended in DMSO to a final concentration of 10 mM . 96-well plates , pre-treated with 0 . 01% Poly-L-Lysine ( Sigma Aldrich; P4832 ) , were separately seeded with ISE6 and Vero cells and incubated for 24 hours to final cell density of ~1 x105 cells/96 well . ISE6 and Vero cells were infected with LGTV ( passage 4 , MOI of 10 ) and ( passage 4 , MOI of 3 ) , respectively . Following adsorption , compounds diluted in DMSO , were added to cells to a final concentration of 0 . 01 , 0 . 1 , 1 , and 10 μM ( 1% of total overlay medium ) and cells were incubated at 37°C . Culture supernatant was collected at 36 hpi and used to quantify LGTV replication by plaque assay . To assess cell viability , cells were treated with alamarBlue reagent ( AbD Serotec; BUF012A ) diluted 1:10 with fresh medium for 12 and 2 hours , respectively . Fluorescence ( excitation at 560nm , emission at 590 nm ) was measured at 48 and 38 hpi using a Molecular Devices SpectraMax M5 plate reader coupled with SoftMax Pro v4 . 8 software . Control was solvent only . Five technical replicates were performed for each concentration with biological replicates ( n = 2 ) . Trypan blue cell exclusion assay was used to assess mortality of ISE6 cells following LGTV infection . Poly-L-Lysine-treated 96-well plates were seeded with ISE6 cells for 48 hours to a cell density of ~9 x 104 cells/well . Cells were treated with LGTV infection ( MOI 10; p4 LGTV stock ) or condition medium as described above . Cells were harvested at 12 , 24 , 36 , and 48 hpi , centrifuged at 1 , 510 g for 5 min , medium was removed and the cell pellet was re-suspended in 1X PBS . Subsequently , a 1:1 0 . 4% trypan blue:cell suspension , was prepared , incubated for ~3 min at RT , the cells were immediately counted using a hemocytometer [54] and the percentage of stained ISE6 cells was determined for LGTV and mock treatments . Three technical replicates were collected per treatment with two biological replicates ( n = 2 ) .
IFA and plaque assays were used in time course experiments to assess levels of LGTV in ISE6 cells and to confirm UV inactivation of LGTV ( Fig 1 ) . Under the assay conditions described herein , IFA revealed that the maximum level of LGTV infection of the ISE6 cell population ( >96% ) corresponded to an MOI of 10 as determined by percentage of cells labeled with the LGTV NS3 protein ( Fig 1A and 1B ) , and plaque assays revealed that the peak of LGTV release from ISE6 cells occurred at 36 hpi ( Fig 1C ) . These conditions were selected for subsequent proteomic analyses . Plaque assays revealed that UV radiation for ≥120 sec was sufficient to achieve 100% inactivation of LGTV as determined by the lack of plaque formation ( Fig 1D ) . The minimum time required for lack of plaque formation was 3 . 5 minutes . UV-LGTV used for proteomic analyses and subsequent assays was inactivated for five minutes . ISE6 cell viability was reduced during the acute stage of infection with LGTV ( i . e . , ≤48 hpi ) as measured based on presence of cellular reducing agents ( FMNH2 , FADH2 , NADH , NADPH , and cytochromes ) . No change in cell growth or mortality was observed , as measured by counting cell population numbers and utilizing the trypan blue cell exclusion assay for LGTV-infected and mock-treated groups ( S1 Fig ) . Completion of the virus lifecycle as determined by release of infectious virus particles ( Fig 1C ) was observed in ISE6 cells infected with LGTV . In comparison , in cells treated with UV-inactivated virus ( UV-LGTV ) we observed no release of infectious virus particles ( Fig 1D ) . Comparative proteomics analyses were used to identify proteins expressed throughout the process of cell infection ( LGTV ) versus those associated only with viral attachment and entry of the host cell ( UV-LGTV ) . The sequence of proteomic analyses performed using the three treatments ( LGTV , UV-LGTV , and mock ) is shown in S2 Fig and S1 Table . LC-MS data were compared for LGTV , UV-LGTV and mock samples ( Fig 2 ) . The expression pattern of LC-MS peaks for LGTV samples was more similar to that of UV-LGTV samples than to that of mock samples . The t-test and ANOVA ( four separate statistical analyses ) were used to identify proteins that exhibited differential expression ( p < 0 . 05 ) between LGTV and UV-LGTV samples as compared to the mock samples ( Fig 3A ) . In total , 486 ISE6 proteins ( S2 Table ) were identified based on homology to NCBI/VectorBase accessions . Of these , 266 and 248 proteins were identified as differentially expressed in the LGTV and UV-LGTV samples , respectively compared to mock samples . Sixty-eight proteins had increased expression , while 198 proteins showed decreased expression in the LGTV samples as compared to mock samples . Additionally , 82 and 166 proteins showed increased and decreased expression ( Fig 3B ) , respectively in the UV-LGTV samples in comparison to mock samples . Overall , 243 proteins ( 50% ) exhibited decreased expression while 120 ( 24 . 7% ) showed increased expression in LGTV and UV-LGTV samples as compared to the mock treatment ( Fig 4A ) . Of the 486 ISE6 proteins identified in this study , 265 ( 54 . 5% ) mapped to orthologous proteins in the KEGG database , while 221 proteins had no match ( KEGG; genome . jp/keg/ko ) . Of the 265 proteins , 176 ( 36 . 2% ) mapped to 66 KEGG pathways and 16 KEGG modules ( S3A Fig and S3 Table ) . The KEGG pathways identified in the present study were categorized into five cellular functions: “metabolism” , “genetic information processing” , “environmental information processing” , “cellular processes” , and “organismal systems” . The majority of proteins ( 52% ) were identified to the functional category “genetic information processing” , followed by “metabolic” ( 38 . 7% ) and “cellular” ( 6 . 3% ) , “environmental information processing” ( 2% ) , and “organismal systems” ( 1% ) ( S3B Fig ) . LGTV samples exhibited the highest number of proteins ( 53 ) identified to the KEGG pathway “genetic information processing” ( Fig 4B–4D ) . Within this group , eight proteins exhibited increased expression and were classified in the pathway , “translation” ( Fig 4B ) . For UV-LGTV samples , the majority of ISE6 proteins ( 57 ) were also classified in the pathway , “genetic information processing” . The majority of proteins exhibiting increased expression ( 17 ) were classified in the protein processing pathways of “folding , sorting , and degradation” ( 7 proteins; 41 . 2% ) , followed by “translation” ( 6 proteins; 35 . 3% ) and “transcription” ( 4 proteins; 23 . 5% ) . Proteins from the LGTV and UV-LGTV samples that lacked a match to KEGG database entries , also displayed differential expression . Of these , 30 proteins had increased expression and 91 had decreased expression in LGTV samples in comparison to mock samples ( S4 Fig and S2 Table ) . Additionally , 38 and 85 proteins were identified with increased and decreased expression , respectively , in the UV-LGTV samples as compared to mock samples . Proteins that showed an increase in expression in LGTV samples were mapped onto the KEGG functional categories of cell signaling ( CYC , STK3 , RPS6 ) , proteolysis ( UCHL3 , PSMA , UBE2N ) , carbon-nitrogen hydrolase activity ( DDAH , VNN ) , replication and mRNA processing ( PARP , TRA2 , CUTL , H2A , CSTF2 ) , translation ( RPS6 , RPL17 , AARS , NARS ) , glutamate metabolism/glutaminolysis ( prostate-specific transglutaminase , putative ISCW011739; Fig 5 and S4 Table ) , pyruvate metabolism and energy association ( MDH2; Fig 6 ) . Proteins that exhibited decreased expression were associated with the functional categories of glycolysis ( GAPDH; Fig 6 ) , energy processes ( ATP5H , ATP5A1 ) , and mRNA surveillance ( PABPC , PELO , MSI , THOC4 ) . Proteins exhibiting increased expression in UV-LGTV samples were mapped onto the KEGG functional categories of signaling ( RHOGDI , RAB35 , SIP , LAMC1 ) , cytoskeletal components , ( ACTN , TUBA ) , unfolded protein response and ER-associated degradation ( HSPA1_8 , RAB7A ) , lysosomal functions ( PSAP ) , and phagosome functions ( RAB7A ) . Proteins that exhibited a decrease in expression were associated with transport ( BAP31 ) , cell survival ( BAP31 , HYOU1 , DERL1 , GROEL ) , cell growth ( SUMO , NOP10 , MAD1L ) , translation ( NOP10 ) , and protein folding ( GROEL ) . Responses common to LGTV and UV-LGTV samples included proteins exhibiting increased expression and associated with signaling ( ITGB , MO25 ) , cytoskeletal structure perturbation ( TLN ) , amino acid metabolism ( ACAT , DP5CD , GLUD1 , CARP , FAH ) , glutamate metabolism/glutaminolysis ( DP5CD , GLUD1 , membrane protein , putative ISCW001521; Fig 5 and S4 Table ) , RNA interference ( AUB ) , and energy-production ( ACAT ) . Proteins with decreased expression and common to both treatment groups were classified to KEGG functions of glycolysis ( ALDOA/B/C , ALDH2/1B1/3A2; Fig 6 ) , energy association ( ATP5D , ATP5B ) , RNA interference ( VIP ) , and structural manipulation ( ACTB_G1 , TUBB ) . 185 of the 265 ISE6 proteins with orthology to KEGG entries ( 70% ) were also identified in a proteomics study of HCV infection of HUH7 . 5 cells [19] ( S5 Fig ) . Sixteen ISE6 proteins ( 6% ) matched orthologs identified in a study of West Nile virus ( WNV ) infection of Vero cells [55] , 16 proteins matched orthologs in a yeast two-hybrid study of flavivirus-host interactions [56] , and 15 proteins ( 5% ) matched orthologs identified in Aedes aegypti infected with dengue virus ( DENV ) [28] . A subset of proteins that exhibited increased expression following LGTV infection and/or UV-LGTV treatment and matched proteins in the studies above , were associated with protein synthesis and proteolysis ( Fig 7 and S5 Table ) . Of the remaining 66 proteins ( 24 . 9% ) , those that exhibited increased expression in LGTV samples were classified in the KEGG functional categories of proteolysis ( PMSA , CARP ) , ATP association/interaction ( PSMA , ANMK ) , cell and matrix adhesion ( VNN , ITGB ) , and as well as oxidative stress and redox homeostasis ( VNN and conserved hypothetical protein ISCW020127-PA ) . Additionally , the cellular function of hydrolase activity was suggested by increased expression of PSMA and VNN ( S5 Table ) . In order to manipulate metabolic functions and subsequent LGTV infection , small molecule assays were completed . In cellular assays , Trichostatin A ( TSA ) , a compound known to inhibit histone deacetylase ( HDAC ) and to activate enzymes involved in intermediate metabolism , including MDH2 , decreased viability of Vero cells ( with and without LGTV infection ) and LGTV replication ( as measured by a decrease in release of infectious virus particles ) at increasing concentrations ( Fig 8A ) . Conversely , an increase in TSA concentration was associated with an increase in the viability of LGTV-infected ISE6 cells and an increase ( ~0 . 5 log pfu ) in LGTV replication ( Fig 8A ) . Oligomycin A ( OligoA ) , a small molecule inhibitor of the mitochondrial H+ ATPase pump , known to inhibit terminal processes of the electron transport chain by reducing ATP production , was associated with a decrease in the viability of Vero cells ( ~20% reduction ) and ISE6 cells ( ~60% ) at increasing concentrations . Significant reduction of LGTV in the mammalian ( ~1 . 5 log reduction in pfu in Vero cells ) and tick ( ~2 log reduction in pfu in ISE6 cells ) system was observed with increasing concentrations of OligoA ( Fig 8B ) .
Several proteins were identified were included in notch and mTOR signaling pathways . The putative histone deacetylase 1 , 2 , 3 ( ISCW007830-PA ) exhibited decreased expression in LGTV and UV-LGTV samples . Several studies [58 , 59] suggest a link between herpesvirus infection and gene regulation through with the binding of viral proteins to histone deacetylases [59] . We hypothesize that LGTV infection may impact the regulation of ISE6 genes via effects on histone deacetylase . In other systems , it has been shown that histone deacetylase can act as a co-repressor in the notch signaling pathway . TSA traditionally binds and inhibits histone deacetylases and treatment of ISE6 cells with TSA during LGTV infection increased LGTV replication , suggesting that LGTV infection impacts gene regulation through histone deacetylases . The putative 40S ribosomal protein S6 ( ISCW024315-PA ) and Mo25 ( ISCW004710-PA ) exhibited increased expression in LGTV cells . These proteins are members of the mTOR signaling pathway which has been implicated in human cytomegalovirus ( HCMV ) infection of mammalian cells [60 , 61] and DENV infection of A . aegypti mosquitoes [62] . Increased expression of Mo25 may reflect a cellular stress response while increased expression of S6 may reflect an increase in translation to maintain growth of the infected cell or facilitate LGTV replication . Manipulation of mTOR signaling has been noted with WNV infection in mammalian systems [63] . The putative calcyclin-binding protein CacyBP ( ISCW013691-PA ) known to function in the Wnt signaling pathway in other systems , had increased expression in UV-LGTV-treated cells and decreased expression in LGTV-infected ISE6 samples . Our observation suggests an increase in proteolysis following virus treatment since the Wnt pathway is associated with the Ca2+-dependent , ubiquitin-mediated proteolysis pathway . Future investigations regarding roles of post-translational modifications in regulating signaling pathways following tick-borne flavivirus infection is necessary . Recently , the piwi-interacting RNA ( piRNA ) pathway has been implicated in the antiviral response of mosquitoes [64] and tick I . scapularis IDE8 cells [65] . Esther et al . 2014 identified three paralogs ( ISCW015916 , ISCW0021130 , and ISCW011768 ) of the tick I . scapularis argonaute ( aubergine ) protein as antiviral factors to LGTV infection . The I . scapularis aubergine protein possesses the paz and piwi domains [66] associated with RNA binding . Homologs of these proteins were not identified in this study , although a homolog of argonaute ( AUB; ISCW011373-PA ) was identified that exhibited increased expression in both LGTV and UV-LGTV ISE6 cells and may play an antiviral role in LGTV infection . Histone ( H2A ) is involved in DNA binding and chromatin packing of DNA , and therefore likely has a role in gene regulation and downstream host protein translation that may be important for homeostasis . The I . scapularis H2A ( ISCW004478-PA ) exhibited increased expression in LGTV-infected ISE6 cells . H2A also had increased expression during DENV infection of HUH7 liver cells and binds with the capsid protein to inhibit nucleosome formation in these human cells [67] . This protein has also been found to bind antisense RNA [68] , also suggesting a possible anti-pathogen role as a result of changes in gene regulation . The proteasome subunit alpha type protein ( ISCW021572-PA ) exhibited increased expression in LGTV samples and the 20S proteasome , regulatory subunit alpha type PSMA7/PRE6 ( ISCW007139-PA ) had increased expression in both LGTV and UV-LGTV samples . These proteins are subunits of the proteasome-associated 20S core particle and may exert antiviral roles through proteolysis and transcriptional regulation . Protein subunits of the proteasome have been shown to play a role in HCV internal ribosome entry site ( IRES ) -mediated translation [69] and may also interact with the HIV protein TAT and HBV protein HBX [70 , 71] . Decreased expression of actin was observed in both LGTV-infected and UV-LGTV-treated samples . Cofilin ( CFN; ISCW006326-PA ) , a putative actin-depolymerizing factor , exhibited decreased expression in these samples . CFN was also identified in a proteomic study of HCV-infected HUH7 . 5 cells [19] . Actin polymerization is involved with formation of actin stress fibers , a process that may facilitate vacuole formation [72] and mammalian neuronal cell entry of Japanese encephalitis virus [73] . UV-LGTV-treated cells exhibited increased expression of the signaling and structural proteins RHOGDI , and ACTN and TUBA , respectively . RHOGDI has been implicated in actin depolarization [74] and showed increased expression in HCV-infected HUH7 . 5 cells [19] at an early ( 12 hpi ) infection time point . ACTN showed increased expression in HUH7 . 5 cells at early ( 24 hpi ) and intermediate ( 48 hpi ) time points post HCV infection and increased expression in UV-HCV-treated cells at a late ( 72 hr ) time point . In the present study , we observed increased expression of ACTN in UV-LGTV samples at the 36 hpi time point . In addition to crosslinking actin fibers and facilitating filament assembly , ACTN been shown to bind the HCV nonstructural proteins NS3 and NS5 [56 , 75] . We hypothesize that this protein may assist LGTV cell entry in ISE6 cells . The proteins acetyl-CoA acetyltransferase ( ACAT1; ISCW016117 ) and aldehyde dehydrogenase 4A1 ( DP5CD; ISCW015982 ) exhibited increased expression in LGTV-infected cells . These enzymes operate upstream of the TCA cycle and are associated with the production of acetoacetyl-CoA and pyruvate , respectively during cellular metabolism ( Fig 6 ) . This result suggests an increase in acetyl-CoA production following viral infection . Interestingly , citrate synthase ( CS; ISCW009586 ) showed decreased expression following LGTV infection and may reflect a reduction of TCA protein activity late in LGTV infection . We observed a decrease in expression of fumarate hydratase ( FH; ISCW020593 ) that may also similarly reflect reduction of TCA protein activity late in LGTV infection . The increased expression of MDH2 ( ISCW003528 ) , a protein involved in the final steps of the TCA cycle , may produce an increase in oxaloacetate , S-malate , and NADH in ISE6 cells . Moreover , increased expression of fumarylacetoacetase ( FAH; ISCW020196 ) may increase fumarate , also involved in the final steps of the TCA cycle . ACAT1 , DP5CD , MDH2 , and FAH may aid in maintaining the TCA cycle late in LGTV infection . In parallel , these observations suggest an impact of LGTV on the TCA cycle at 36 hours post infection that may be linked to successful replication of the virus . Our observation of a decrease in expression of fructose-bisphosphate aldolase ( ALDOA ) , glyceraldehyde-3-phosphate dehydrogenase ( GADH ) , aldehyde dehydrogenase family 7 member A1 ( ALDH3A2 ) , and pyruvate kinase ( PKLR ) in LGTV-infected and UV-LGTV-treated cells , suggests an impact of LGTV on glycolytic processes . This finding is at odds with that of Patramool et al 2011 , who observed that DENV-infected C6/36 A . albopictus cells [27] exhibit increased glycolysis . The in vivo study of Tchankouo-Nguetcheu et al 2010 highlighted an increased expression of glycolytic proteins in the midgut tissues of DENV-infected A . aegypti [28] . Diamond et al 2010 also identified members of the glycolysis pathway that exhibited increased expression at early to intermediate time points ( i . e . , prior to peak release of infectious virus ) following HCV infection , but not at the late ( during and following peak release of virus from the cell ) time point [19] . Although ticks are exclusive blood feeders and mosquitoes regularly take sugar meals between blood meals , these data suggest the possible increase in glycolysis at early to intermediate time points post flaviviral infection , but a decrease in glycolysis at later time points . Our in vitro studies have shown that the compounds TSA and OligoA can affect levels of LGTV replication , presumably through impacts on a variety of cellular metabolic processes . TSA is thought to inhibit histone deacetylases and stimulate the acetylation of histones and metabolic enzymes , while OligoA may inhibit oxidative phosphorylation and electron transport . OligoA may activate AMPK activity [76] , inhibit ATP production , and affect cellular energy levels . Clearly , further studies are required to determine the mode of action of TSA and OligoA in the LGTV-ISE6 system . Glutaminolysis can produce an alternative energy source for the cell by generating ATP during the conversion of glutamine to α-ketoglutarate . Although tick medium has relatively large amounts of glutamine , glutamic acid , and α-ketoglutarate , increased expression of proteins associated with glutaminolysis ( Fig 5 and S4 Table ) suggest that LGTV infection of ISE6 cells may stimulate glutaminolysis and the production of α-ketoglutarate , a key intermediate in the TCA cycle . Studies suggest that glutaminolysis is manipulated during infection of human cells by both HCMV [77 , 78] and HCV [19] . Thus , glutaminolysis and α-ketoglutarate are likely critical not only for maintaining the TCA cycle , but also supporting oxidative phosphorylation and ATP production in the infected cell . Additionally , stimulation of α-ketoglutarate has been shown to increase mTOR activity [79 , 80] which operates in parallel with glutaminolysis . In ongoing studies , we are assessing viral manipulation of glutamate dehydrogenase ( GDH ) activity using inhibitory compounds with the goal of disrupting flaviviral infection . To contribute to an improved understanding of flavivirus-I . scapularis interactions , we developed an in vitro system to identify changes in ISE6 protein expression following infection with the TBF , LGTV . We present the first study to identify ISE6 proteins that are differentially-expressed following LGTV infection . In total , 486 proteins were identified with 66/198 showing increased/decreased expression following LGTV infection and 82/166 showing increased/decreased expression following UV-LGTV treatment . We identified proteins associated with the cellular functions of genetic information processing ( GIP ) , metabolism , cellular processes , environmental information processing , and organismal systems . The majority of proteins populate GIP-specific pathways followed by metabolism-specific pathways . The identifications of these proteins provide a critical resource to improve understanding of the I . scapularis proteome , improve gene annotations , and facilitate further studies in the tick cell culture system . Further understanding of protein function can also be achieved using approaches such as IFA , targeted mass spectrometry , small molecule in vitro assays , and RNAi . The present study is an important first step toward identifying tick proteins tied to LGTV replication as candidates for anti-tick vaccines and/or as targets for therapeutic screening to disrupt tick-borne flavivirus transmission . | High-throughput proteomics offers an approach to evaluate changes in cell protein levels following arboviral infection . Research to understand the molecular basis of human-flavivirus interactions has advanced significantly over the past decade , but comparatively little is known regarding interactions between ticks and tick-borne flaviviruses ( TBFs ) . Here , we employed a proteomics approach using an I . scapularis ISE6 cell line infected with the TBF Langat virus ( LGTV ) to identify proteins and biochemical pathways affected by viral infection . An LC-MS/MS approach was used to identify proteins that were subsequently assigned to putative cellular pathways based on orthology to proteins in the KEGG database . Biochemical pathways common among arthropods in response to infection with flavivirus and possibly unique to tick-flavivirus interactions , were identified . In vitro cellular assays using small molecules suggest the involvement of the ISE6 proteins , malate dehydrogenase ( MDH2 ) , and mitochondria in viral replication . These analyses provide a basis for further studies to identify tick proteins associated with viral replication that could be targeted to disrupt TBF transmission . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"cell",
"physiology",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"protein",
"metabolism",
"ixodes",
"microbiology",
"database",
"searching",
"cell",
"metabolism",
"animals",
"protein",
"expression",
"molecular",
"biology",
"techniques",
"ticks",
"research",... | 2016 | Changes in the Proteome of Langat-Infected Ixodes scapularis ISE6 Cells: Metabolic Pathways Associated with Flavivirus Infection |
Repair of programmed DNA double-strand breaks ( DSBs ) by meiotic recombination relies on the generation of flanking 3′ single-stranded DNA overhangs and their interaction with a homologous double-stranded DNA template . In various common model organisms , the ubiquitous strand exchange protein Rad51 and its meiosis-specific homologue Dmc1 have been implicated in the joint promotion of DNA–strand exchange at meiotic recombination sites . However , the division of labor between these two recombinases is still a puzzle . Using RNAi and gene-disruption experiments , we have studied their roles in meiotic recombination and chromosome pairing in the ciliated protist Tetrahymena as an evolutionarily distant meiotic model . Cytological and electrophoresis-based assays for DSBs revealed that , without Rad51p , DSBs were not repaired . However , in the absence of Dmc1p , efficient Rad51p-dependent repair took place , but crossing over was suppressed . Immunostaining and protein tagging demonstrated that only Dmc1p formed strong DSB–dependent foci on meiotic chromatin , whereas the distribution of Rad51p was diffuse within nuclei . This suggests that meiotic nucleoprotein filaments consist primarily of Dmc1p . Moreover , a proximity ligation assay confirmed that little if any Rad51p forms mixed nucleoprotein filaments with Dmc1p . Dmc1p focus formation was independent of the presence of Rad51p . The absence of Dmc1p did not result in compensatory assembly of Rad51p repair foci , and even artificial DNA damage by UV failed to induce Rad51p foci in meiotic nuclei , while it did so in somatic nuclei within one and the same cell . The observed interhomologue repair deficit in dmc1Δ meiosis is consistent with a requirement for Dmc1p in promoting the homologue as the preferred recombination partner . We propose that relatively short and/or transient Rad51p nucleoprotein filaments are sufficient for intrachromosomal recombination , whereas long nucleoprotein filaments consisting primarily of Dmc1p are required for interhomolog recombination .
Meiosis is a pivotal process in the sexual reproduction cycle . It is a nuclear division that reduces the diploid somatic to the haploid gametic chromosome set . Successful disjunction of the two genomes requires that every chromosome pairs with its corresponding partner before they eventually separate . In most organisms , chromosome pairing relies on the base pair matching of single-stranded DNA molecules that are generated in the wake of self-inflicted DNA double-strand breaks ( DSBs ) . Repair of DSBs occurs by using an intact DNA strand as a template , and may lead to the exchange of the broken and the template strand , thereby causing meiotic genetic recombination . It is essential that DNA from the homologous chromosome is preferred as a template for recombinational repair over DNA from the sister chromatid , since only the former situation will generate crossovers that ensure the orderly segregation of homologous chromosomes during the reductional division . Strand exchange at meiotic recombination sites is accomplished by the RecA homologs Rad51 and Dmc1 . They have similar and partially overlapping roles in DNA heteroduplex formation [1] , but whereas Rad51 is indispensable both for mitotic and meiotic recombination , Dmc1 is meiosis-specific . It has been suggested that this is due to its role in interhomolog rather than intersister recombination [2] . Consequently , in budding yeast , where Rad51-dependent repair via the sister is suppressed in meiosis ( see [3] and lit cit . therein ) , meiotic DSB repair is dramatically reduced and DSBs acquire long single-stranded tails in the absence of Dmc1 [4] . However , upon overexpression of Rad51 or its stimulating partner Rad54 , and in certain mutant strains , high levels of interhomolog recombination can occur in the absence of Dmc1 [5]–[7] . Thus , it seems that Rad51 is not , in principle , unable to support interhomolog recombination , but that it is held in check during meiosis . It may be impeded in order to promote a Dmc1-dependent pathway , which may perform better in meiotic crossover , especially in recombination partner choice . Dmc1 is also required for interhomolog recombination in mammals [8] , [9] . In other organisms , this division of labor is not as strict , and the dmc1 mutant phenotype is less severe . In Arabidopsis dmc1 mutants , chromosome fragmentation was not observed , and chromosomes segregated randomly at meiosis I . This indicated that DSBs were repaired by a mechanism that did not produce crossovers [10] , [11] . In the fission yeast , Schizosaccharomyces pombe , deletion of DMC1 caused only a moderate reduction in crossing over and a slight reduction in fertility , suggesting that Rad51 can partially substitute for Dmc1 in interhomolog recombination [12] . Drosophila , Caenorhabditis elegans , and Neurospora lack Dmc1 orthologues , altogether ( see [1] ) . Because of the variability between model organisms with respect to their dependence on Rad51 and Dmc1 for meiotic crossing over , we decided to study the roles of Rad51p and Dmc1p in the unconventional meiosis of the evolutionarily distant model system Tetrahymena . Tetrahymena thermophila is a unicellular ciliated protist that possesses two nuclei , the polyploid somatic macronucleus and the diploid generative micronucleus . While the former is transcriptionally active and divides by an amitotic process , the latter constitutes the germline , is transcriptionally silent , and divides mitotically . Only the micronucleus undergoes meiosis . Both the rough alignment and the precise matching of homologous chromosomes during meiotic prophase depend on Spo11-induced meiotic DSBs and their signaling through an ATR-dependent pathway [13] . Promotion of Dmc1 over Rad51 usage in meiosis seems to be , in part , related to the presence of a synaptonemal complex ( SC ) ( see [7] and lit cit . therein ) , and Tetrahymena does not feature a canonical SC . Therefore , we wondered whether dependency on Dmc1 would be low . To test this , we designed experiments to elucidate the contributions of Rad51 and Dmc1 to meiotic recombination in Tetrahymena .
The differing importance of RAD51 and DMC1 for meiosis in a variety of organisms led us to study in detail , the role of these two genes in the meiosis of the protist Tetrahymena . The situation in this evolutionarily distant model system may help to clarify the primordial division of labor between the two proteins . Tetrahymena possesses two RecA homologs , TTHERM_00142330 ( RAD51 ) and TTHERM_00459230 ( DMC1 ) [14] . RAD51 is expressed during the mitotic cell cycle with a particularly high abundance of RAD51 mRNA during S phase of the somatic nucleus , and during meiotic prophase [15] . It is also strongly induced following DNA damage [16] , [17] . Consistent with its somatic expression pattern , it has been reported that Rad51p is essential for vegetative growth [15] . Also , small amounts of Rad51p were found in the somatic nucleus ( [18] and Figure S1A ) . DMC1 , on the other hand , showed meiosis-specific expression in mRNA microarray hybridizations [19] . Here , we confirmed the exclusive meiotic expression by RT-PCR and showed that maximal expression occurs approximately 3 h after induction of meiosis , which is similar to that of SPO11 ( Figure 1A ) . To investigate the functions of Rad51p and Dmc1p , we produced strains lacking one or the other . dmc1Δ strains were constructed by knocking out the ∼45 transcriptionally active copies of DMC1 in the polyploid somatic nucleus by gene replacement . Southern hybridization demonstrated the removal of the wild-type copies ( Figure 1B ) , and the complete lack of Dmc1p was confirmed by Western detection ( Figure 1C ) . The requirement of Rad51p for vegetative growth precluded gene knockout . Therefore , we created a strain carrying a RAD51 hairpin construct ( rad51hp ) , which expressed double-stranded RNA hairpins under a Cd2+-inducible promoter [20] . This allowed the knockdown of RAD51 by RNA interference ( RNAi ) specifically during meiosis . High efficiency of RNAi was confirmed by Western analysis ( Figure 1C ) , where the protein was estimated to be reduced to ∼1% of the wild-type level ( Figure S1 ) . Also , rad51 knockdown ( rad51i ) cells were unable to express Rad51p upon UV-induced DNA-damage ( Figure S1 ) . We triggered RNAi by the exposure of cells to CdCl2 in stationary cultures before the initiation of meiosis ( see Materials and Methods ) . This regime avoids depletion of Rad51p needed for vegetative propagation , yet allows for abundant hairpin expression to knock down RAD51 during meiosis . First , we tested the ability of Rad51p- and Dmc1p-depleted cells to undergo meiosis and sexual reproduction ( Table 1; explanations in Table 1 ) . While rad51i cells did not produce any sexual progeny , the viability of sexual progeny from dmc1 KO crosses was found to be 3 . 4% ( as compared to 98% of wild-type crosses - Table 1 ) . Thus , a low proportion of dmc1Δ meioses can give rise to viable zygotes . To determine the basis of the different meiotic success rates of Rad51- and Dmc1-depleted cells , we studied meiotic progression cytologically in rad51i and dmc1Δ strains . In Tetrahymena , nutrient-starvation makes cells competent for conjugation ( = cell mating ) . Upon mixing starved cells of complementing mating types , they will form pairs which initiate synchronous meioses of their generative nuclei ( see [21] , [22] ) . Meiotic prophase is characterized by an enormous elongation of the nuclei ( Figure 2A ) . The appearance of DSBs and increasing chromosome pairing during the elongation process permit the attribution of the elongating meiotic nucleus to the leptotene-pachytene stages ( see [13] ) . During a stage corresponding to diplotene , the nucleus shortens again and chromosomes condense . Five bivalents emerge in diakinesis-metaphase I and finally undergo two meiotic divisions ( Figure 2A ) . It was previously found in a rad51 knockout strain that meiotic nuclei did elongate , but that cells arrested prior to chromosome condensation and rarely progressed to anaphase I [15] . However , due to the vegetative function of Rad51p , the observed phenotypes could have been caused by accumulative pre-meiotic damage rather than genuine meiotic defects [15] . To prevent mitotic defects from obscuring the meiotic phenotype of Rad51p loss , we induced RNAi knockdown of RAD51 only at the onset of meiosis [20] . Cytological inspection of conjugating RNAi cells ( rad51i ) revealed that stages up to diplotene were indistinguishable from wild-type ( Figure 2B ) . However , diakinesis-metaphase I showed signs of chromosome fragmentation ( Figure 2B , 3A ) with 98% of the nuclei at that stage ( n = 100 ) displaying granular or diffuse chromatin and only 2% featuring compact entities . Moreover , normal meiotic divisions did not take place ( Figure 2B ) . In a similar experiment , a rad51i strain was conjugated to a wild-type strain . In this case , both partners were affected due to the trans-activity of RNAi ( see Materials and Methods ) . As a control , a rad51hp strain without stimulation of RNAi was mated to a wild type strain , confirming that the mere presence of the construct did not notably interfere with the progress of meiosis ( Figure S2 ) . Thus , the fragmentation of metaphase I chromosomes and the inability to perform normal meiotic divisions upon induced RNAi are genuine consequences of the depletion of Rad51p and not side-effects of the experimental system . The fragmented metaphase I chromosomes strongly indicated that DSBs were not properly repaired . To confirm the persistence of DSBs , we performed an assay for the detection of DSB-dependent chromosome fragmentation by pulsed-field gel electrophoresis ( PFGE ) [13] . Under our standard PFGE conditions , intact chromosomes of the generative nucleus do not enter the gel , whereas DNA fragments of different size migrate as a distinct band [13] . In the wild type , DSB-dependent fragments appeared from ∼2 h–5 h after induction of meiosis ( Figure 3B ) . They were missing in a spo11Δ strain , which is unable to generate meiotic DSBs ( Figure 3B ) . In rad51i meiosis , the band diagnostic for DSBs did not disappear within 6 h from induction of meiosis ( Figure 3B ) . Thus , meiotic DSBs are not repaired in the absence of Rad51p ( Figure 3B ) . However , in repeated experiments this band consistently became weaker at 6 h post-meiosis induction . We speculate that this reduction in band intensity may result from DSB-dependent fragments being converted to other DNA species with different migration . This could result from the hyper resectioning of DSB ends ( see [2] , [23] ) . Clarification of the nature of this putative intermediate must await its analysis at defined DSB hotspots , which we have yet to detect . As an additional test for DSB behavior , we analyzed the localization of phosphorylated histone variant H2A . X . Phosphorylation of H2A . X ( γ-H2A . X ) is a DSB marker ( see [24] and lit . cit . therein ) . In the wild type , γ-H2A . X was most prevalent in elongated meiotic nuclei , but was virtually absent in metaphases ( Figure 3C; see also [25] , [26] ) . Consistent with the persistence of DSBs , γ-H2A . X was present into aberrant metaphase I in rad51i ( Figure 3C ) . In the absence of Dmc1p , elongation of the generative nucleus occurred as in the wild type ( Figure 2C ) . However , at diakinesis-metaphase I , univalents appeared instead of bivalents ( Figure 2C , Figure 3A ) . These univalents subsequently went through two meiotic divisions ( Figure 2C ) during which they are presumably randomly segregated , with a small chance of generating genetically balanced viable progeny . Of 164 favorable metaphase I nuclei evaluated , 15 ( 9% ) displayed fewer than the 10 entities expected if they contained only univalents . Therefore , we can not exclude the possibility of rare bivalent formation . The elongation of meiotic nuclei is indicative of the occurrence of DSBs [25] , while intact diakinesis-metaphase I univalents suggest that these DSBs are repaired ( Figure 3A ) . We used the PFGE-DSB assay to test for the transient appearance of DSBs . It confirmed that in dmc1Δ , DSBs appeared about 3 h after induction of meiosis and virtually disappeared by 6 h after induction of meiosis ( Figure 3B ) , which is similar to the wild type . Also , γ-H2AX foci disappeared from dmc1Δ mutant meiotic nuclei after the exit from the elongated state ( Figure 3C ) . Together with the production of some viable progeny ( see above ) , these results suggest that efficient DSB repair takes place in dmc1Δ . However , the ( almost complete ) lack of bivalents suggests that this repair does not take place by homologous crossover recombination . The formation of univalents in the dmc1Δ mutant and the persistence of chromosome fragments in rad51i indicated that in neither case did DSB repair occur via crossing over with the homologous chromosome . Homologous strand invasion by ssDNA not only initiates crossover , but in many organisms also confers precise meiotic pairing [27] , [28] . Therefore , we tested if pairing was affected in the absence of Dmc1p and Rad51p . To this end , we determined the pairing of FISH-labeled homologous loci in fully elongated meiotic nuclei ( Figure 4A ) . We found that pairing was reduced in dmc1 and rad51 deficiencies ( Figure 4B ) , like in mre11 and com1 mutants where homologous strand exchange failed to take place [13] . This is consistent with our previous finding that the parallel , bouquet-like arrangement of chromosomes within the elongated meiotic nucleus is not sufficient to bring about precise homologous alignment , and that strand exchange may be required for homologous recognition and stable pairing [29] . Having established that RAD51 is essential for meiotic DSB repair , whereas DMC1 is required for its homologous recombination outcome , we wanted to study the localization of the two proteins in meiotic nuclei . To this end , we used several antibodies and tags for their detection . First , we used a commercial antibody ( see Materials and Methods ) , which detects Rad51p in somatic nuclei ( see above ) . When we applied this antibody to Western blots of meiotically expressed proteins , we found that it recognizes not only Rad51p but also Dmc1p ( Figure 1C ) , which shares a 32% similarity with the former over a length of 328 amino acids . In cytology , this antibody labeled meiotic nuclei of wild-type cells ( Figure 5A ) . When we applied the anti-Rad51/Dmc1 antibody to rad51i cells , it stained the meiotic nucleus ( Figure 5B ) . Since RNAi efficiently depleted Rad51p ( Figure 1 and Figure S1 ) , this labeling must be due to Dmc1p . Conversely , labeling was found in dmc1Δ meiotic nuclei as well ( Figure 5C ) , in which case it must be due to the exclusive presence of Rad51p . It was noticed that staining of Dmc1p produced a granular pattern ( Figure 5B ) , whereas Rad51p staining of the meiotic nuclei was more uniform ( Figure 5C ) . Notably , the anti-Rad51/Dmc1 antibody labeled meiotic nuclei of all stages from the beginning of their elongation ( ∼ leptotene ) to well beyond anaphase I , and it highlighted chromatin and chromatin-free regions ( Figure 5D ) . To find out if this staining pattern reflects the actual distribution and temporal appearance of Rad51p and/or Dmc1p , we generated a specific anti-Rad51 antibody ( data on the specificity of the antibody are summarized in Figure S3 ) , and we tagged Rad51p with HA . As with the commercial antibody , we observed ubiquitous staining of meiotic nuclei during prophase and beyond , and sporadic spots in somatic nuclei ( Figure 5E , 5F ) . Moreover , both antibodies labeled meiotic nuclei of spo11Δ cells ( which are lacking DSBs ) ( Figure 5G ) . This confirms our previous observation [25] that Rad51p localizes to meiotic nuclei independently of the presence of DSBs . To specifically study Dmc1p localization , we tagged this protein with mCherry . Dmc1-mCherry localized to chromatin and chromatin-free regions of meiotic nuclei but not to somatic nuclei ( Figure 5H ) . Thus , Rad51p and Dmc1p are expressed and can be detected in meiotic nuclei even if they do not assemble near DSBs . To detect if , in addition to their ubiquitous presence in meiotic nuclei , a subset of Rad51p and Dmc1p would localize to DSBs , we applied a detergent spreading method to remove soluble protein . Upon staining with the Rad51/Dmc1 antibody , spread wild-type meiotic nuclei displayed numerous foci ( Figure 6A ) . On the other hand , signals were lost in spread spo11Δ meiotic nuclei ( Figure 6B ) , demonstrating the removal of protein that was not bound to chromatin ( compare with the unspread nuclei in Figure 5G ) . Unlike in conventionally prepared nuclei , no staining was detected at metaphase I or later in spread wild-type meiotic nuclei ( data not shown ) . Together , this suggests that Rad51p and/or Dmc1p are associated with chromatin only in the presence of DSBs . Therefore , to eliminate background staining from recombination proteins that were present all over the nucleus , the following studies on the participation of Rad51p and Dmc1p in recombination foci were performed on spread nuclei . To test the extent of Dmc1p contribution to the foci , we constructed cells expressing Dmc1-mCherry . Dmc1p localized to spread meiotic chromatin in numerous strong foci ( Figure 6C ) . Foci were present from the beginning of elongation to shortly before diakinesis-metaphase . 100% of fully elongated nuclei ( n = 100 nuclei longer than the cell ) displayed strong foci . Since it is possible , although unlikely , that Dmc1p focus formation was caused by the mCherry tag , we independently localized Dmc1p by applying the Rad51/Dmc1 antibody to rad51-RNAi cells , where it only highlights Dmc1p . Also in this experiment , Dmc1p was found assembled into numerous foci ( Figure 6D ) , with 100% of fully elongated meiotic nuclei ( n = 100 ) displaying foci . Moreover , it also demonstrated that Dmc1p forms strong foci in the absence of Rad51p . To further confirm the independence of Dmc1p localization on Rad51p , we performed a rad51i × DMC1-mCherry crossing ( where both conjugating cells are depleted of Rad51p and incorporate mCherry-tagged Dmc1p ) . Meiotic nuclei displayed strong Dmc1p foci in this case as well ( Figure 6E ) . We quantified the brightness of Dmc1p foci by measuring gray values of foci on images ( see Materials and Methods ) . They were 183±18 . 5 ( DMC1-mCherry × rad51i , n = 110 foci from 10 nuclei ) vs . 177±21 . 1 ( DMC1-mCherry × wild type , n = 110 foci from 10 nuclei ) , hence Dmc1p signal intensity was found to be not reduced in the absence of Rad51p . It is likely , yet unproven , that nuclear foci of recombination proteins localize to DSB sites ( for discussion see [1] , [30] ) . To corroborate this theory , we attempted to colocalize Dmc1p foci with γ-H2A . X , which occurs in chromatin flanking a DSB ( see [24] ) . In spread nuclei , γ-H2A . X formed patches . Double staining with Dmc1p revealed a high degree of colocalization: of 431 Dmc1p foci scored in four different meiotic nuclei , 394 ( 91% ) localized to a γ-H2A . X patch ( Figure 6F ) . To obtain an estimate of the number of DSBs that occur in Tetrahymena meiotic nuclei , we investigated the number of Dmc1p foci in spreads of wild-type and rad51i matings . There were , on average , 174 Dmc1p recombination foci per meiotic nucleus ( Figure 6G ) . The number of Dmc1p immunostained foci in the wild type was not significantly different from the number of Dmc1-mCherry foci in rad51i ( Figure 5D ) , which is additional evidence for the independence of Dmc1p localization on Rad51p . To detect if Rad51p associates with meiotic chromatin , we stained spread cells with the specific anti-Rad51 antibody . In contrast to Dmc1p , Rad51p foci were virtually absent from meiotic nuclei . However , the presence of Rad51p signals in the somatic nuclei of the same cells indicated that immunostaining was working and that the spreading procedure did not remove Rad51p altogether ( Figure 7A ) . In quantitative terms , none of 200 evaluated fully elongated ( i . e . , longer than the cell ) , meiotic nuclei displayed any foci . Similarly , we did not observe Rad51p foci in meiotic nuclei of dmc1Δ cells stained with the Rad51p/Dmc1p antibody ( which , in this case , exclusively highlights Rad51p ) ( Figure 7B ) or of strains expressing HA-tagged Rad51p ( Figure 7C ) . In all these cases , foci were present only in somatic nuclei . Double staining of tagged Rad51p and Dmc1p revealed the distinct localization of these two proteins in one and the same cell ( Figure 7D ) . The failure to detect meiotic Rad51p foci by three different cytological approaches suggests that Rad51p is much less abundant at DSBs than Dmc1p . Only occasionally did we observe very weak Rad51p foci in barrel- or spindle-shaped meiotic nuclei of an uncertain stage ( Figure 7A ) . At t = 2 . 5 h , 5% ( n = 100 ) and at t = 4 h , 16% ( n = 100 ) of barrel- or spindle-shaped meiotic nuclei displayed foci . The greater abundance of such nuclei at the later timepoint ( when most nuclei have progressed beyond the maximal elongation stage - [13]; Figure S4 and compare Figure 2 ) suggests that a subset of meiotic nuclei develop Rad51p foci after maximal elongation and hence much later than the first appearance of Dmc1p foci . The sporadic occurrence of these cells could indicate that a small amount of Rad51p is associated with meiotic DSBs during a short period late in meiotic prophase and/or in a subset of cells that fail to undergo normal meiosis . The weakness of these foci , at the limit of detectability , precluded their quantification . As mentioned before , efficient DSB repair takes place in dmc1Δ meiosis , while at the same time prominent Rad51p foci are absent ( Figure 7C ) . Thus , there does not seem to occur a compensatory Rad51p localization to DSBs . Therefore , it might be argued that the absence of Dmc1p triggers an alternative Rad51p-independent repair process such as single-strand annealing [31] or non-homologous end-joining . To exclude this possibility , we created a dmc1i strain and mated it to rad51i , creating a situation where meiotic cells are depleted of both proteins . In these double-deficient meioses , diakinesis-metaphase I chromosomes were fragmented just as in rad51i ( Figure 3D ) , indicating that DSB repair did not take place and that Dmc1p-independent repair requires Rad51p . If there was an undetectably low amount of Rad51p associated with meiotic DSBs , it might be possible to increase it above the threshold of visibility by inducing additional DNA damage . Therefore , we exposed cells to 25 J/m2 UV-C prior to meiosis and tested for Rad51p focus formation . We found that UV irradiation induced strong foci ( detected by the specific anti-Rad51 antibody ) in somatic nuclei , whereas it failed to do so in meiotic nuclei ( Figure 7E , top ) . Likewise , the Rad51/Dmc1 antibody highlighted foci in the somatic and the meiotic nuclei of the wild type ( Figure 7E , bottom ) , but only in the somatic nuclei of dmc1Δ cells ( Figure 7F ) . This , again , confirms that the foci detected in meiotic nuclei of the wild type consisted of Dmc1p only . Conversely , no foci were produced by Rad51/Dmc1 immunostaining in the somatic nuclei of irradiated rad51i cells , indicating that RNAi was efficient , and that UV does not induce Dmc1p in the somatic nucleus ( Figure 7G ) . The exclusive expression of UV-induced Rad51p in the somatic nucleus and of Dmc1p in the meiotic nucleus was also seen by the simultaneous detection of HA-tagged Rad51p and mCherry-tagged Dmc1p ( Figure 7H ) . Altogether , UV irradiation induced strong Rad51p foci in 100% of the somatic nuclei ( n = 100 ) , whereas only 5% of meiotic nuclei ( n = 100 ) displayed very faint Rad51p staining . Similarly , bleomycin ( 50 µg/ml ) , which we previously had shown to produce DSBs in the generative nucleus [13] , induced Rad51p foci in somatic nuclei but not in meiotic nuclei ( data not shown ) . Thus , DNA damage seems to trigger different processing of lesions in nonmeiotic and meiotic nuclei , with less Rad51p involvement in the latter . The absence of Rad51p foci from spread meiotic nuclei strongly suggests that very little Rad51p is associated with DNA at DSBs . However , there remains the unlikely alternative that spreading removes chromatin-associated Rad51p but not Dmc1p , and that this loss affects meiotic nuclei , but not somatic nuclei . Therefore , we wanted to confirm the low abundance of Rad51p associated with meiotic chromatin by an independent approach , the proximity ligation method . In this method , oligonucleotides attached to antibodies against two target proteins ligate and can be amplified to multiple DNA circles when bound in close proximity , i . e . , separated by ∼40 nm or less . These polymers are visualized by the incorporation of fluorescence-labeled nucleotides ( for the principle of the method see [32] and Figure S5 ) . We performed the assay on conjugating wild-type cells carrying both Rad51-HA and Dmc1-mCherry and conventionally prepared for microscopy . To detect Rad51p-Dmc1p colocalization , cells were labeled with anti-Dmc1-mCherry and anti-Rad51-HA ( Figure 8 , Table 2 ) . For control experiments , antibody pairs were used which detected the same protein: anti-Dmc1-mCherry and anti-Rad51p/Dmc1p to mark Dmc1p , anti-Rad51-HA and anti-Rad51p to mark Rad51p ( Figure 8 , Table 2 ) . Signals produced by the proximity ligation assay were less abundant than those from immunostaining ( Figure 8 ) , presumably because of a threshold of local antigen concentration required to support the reaction . Although Rad51p was found to be abundant in unspread meiotic nuclei by immunostaining ( see above ) , the proximity ligation assay detected significantly fewer Rad51p-dependent signals ( Figure 8A ) than Dmc1p-dependent signals ( Figure 8B , Table 2 ) . This is consistent with the possibility that non-chromatin-associated Rad51 molecules are dispersed in the nucleus , but their local concentration is not sufficient to develop a signal . Of the few Rad51p-dependent signals detected , a large proportion localized to the somatic nucleus ( Figure 8A ) . This is consistent with the detection of Rad51p in a subpopulation of somatic nuclei by immunostaining ( see above ) . In contrast to Rad51p , numerous Dmc1p-dependent signals were produced in meiotic nuclei ( Figure 8B ) , suggesting that a substantial amount of the protein forms clusters , such as would be expected from nucleoprotein filaments . Only a few signals were produced after combining antibodies detecting Dmc1p and Rad51p ( Figure 8C ) . The scarcity of reaction products confirms the conclusion from immunostaining that Dmc1p and Rad51p form few , if any , mixed nucleoprotein filaments .
In all organisms where the localization of Rad51 or Dmc1 recombinases has been studied so far , they appear as nuclear foci , and impressive circumstantial evidence suggests that these foci reflect the association of protein complexes with DSBs ( see [1] , [30] ) . Here , we found that in Tetrahymena , there is abundant expression of Dmc1p and Rad51p in meiotic nuclei . The bulk of the two proteins localized throughout the nuclei including chromatin-free regions , and their presence was independent of DSBs . However , when we applied preparation conditions under which free nuclear proteins are removed , numerous DSB-dependent Dmc1p foci , but virtually no Rad51p , remained associated with meiotic chromatin . On the other hand , Rad51p foci were readily detectable in the somatic nuclei within the same cells ( for a graphical summary see Figure 9 ) . Independent experiments using HA-tagged Rad51p , a specific antibody against Rad51p , or a Rad51p/Dmc1p antibody ( applied to dmc1Δ ) all failed to detect meiotic Rad51p foci . This was not due to deficits in the reporter systems since all three approaches detected Rad51p clusters in the somatic nucleus . In contrast to Rad51p , strong meiotic Dmc1p foci were observed by Dmc1p tagging and with the antibody against Rad51p/Dmc1p . Similarly , a proximity ligation assay produced abundant signals associated with meiotic nuclei only for Dmc1p , whereas it detected Rad51p in somatic nuclei . Together , this suggests that Dmc1p is much more abundant than Rad51p at meiotic DSB sites . Despite the well-established correlation between cytological foci and DSBs , it is not clear if cytological foci represent recombination proteins that form nucleoprotein complexes with ssDNA at DSBs ( see [1] , [33] ) . Some support for the equivalence of nucleoprotein filaments and foci comes from the observation that focus formation was strongly reduced in mre11Δ and com1Δ mutants [13] where ssDNA resection and single-strand exposure are believed to be reduced ( see [13] ) . Moreover , here we showed that Dmc1p foci localize to patches of γ-H2A . X , which in turn appears around DSBs [24] . Thus , if the presence and strength of foci reflected the amount of protein involved in filament formation , the proportion of Rad51p in nucleoprotein filaments must be small as compared to Dmc1p . The alternative explanation , that Dmc1p foci might be more resistant to preparation- related loss than Rad51p foci , is rendered less likely due to the similar physical properties of Dmc1- and Rad51-ssDNA filaments in vitro [34] . Strikingly , Rad51p foci were not detected on meiotic chromatin even after artificial induction of DNA lesions , whereas Rad51p formed numerous strong foci on somatic chromatin under these conditions . This suggests that in the meiotic nucleus , a specific regime prevails that prohibits the association of large amounts of Rad51p with damaged DNA , perhaps in the interest of promoting homologous over sister repair . In meiotic nuclei of budding yeast , Shinohara et al . [35] found foci that contained Rad51 and Dmc1 side-by-side , leading them to propose that Rad51 and Dmc1 each form homo-oligomers rather than mixed complexes at recombination sites . Our failure to observe meiotic Rad51p foci suggests that long , pure Rad51p nucleoprotein filaments do not form in Tetrahymena , and is consistent with short ( and perhaps transient ) Rad51p filaments or mixed filaments with Dmc1p predominating . Moreover , the proximity linkage assay for Dmc1-mCherry and Rad51p produced very little signal , which suggests that little if any Rad51p colocalizes with Dmc1p , and argues against the abundant presence of Rad51p in mixed nucleoprotein filaments . Unlike in the budding yeast and in Arabidopsis [36] , [37] , strong Dmc1p foci did form even in the absence of Rad51p . This suggests that Dmc1p does not require Rad51p for its polymerization along ssDNA . However , abundant loading of Dmc1p is not sufficient for the repair of meiotic DSBs in the absence of Rad51p . Despite the fact that Rad51p does not assemble in cytologically visible recombination foci , suggesting that it is not present at notable amounts at DSBs , it is indispensable for meiotic DSB repair and recombination . In the absence of Dmc1p , Rad51p is sufficient to allow efficient repair , however , a compensatory accumulation of Rad51p does not occur . This is in contrast to budding yeast dmc1Δ strains , where bright Rad51p foci are formed [38] . It may be assumed that if Dmc1p is missing , repair takes place via the sister chromatid and requires only minimal Rad51p nucleoprotein filament formation , below cytological detectability . In the wild-type situation , when Dmc1p is present , Rad51p is necessary to support interhomolog crossover recombination . It is conceivable that this is achieved by Rad51p somehow activating Dmc1p at DSBs , without being incorporated in nucleoprotein filaments . Alternatively , a small number of Rad51p molecules , below cytological detection , might be part of these nucleoprotein filaments . In the absence of Dmc1p , we observed intact univalents during diakinesis to metaphase I . Accordingly , 3 . 4% viable sexual progeny were produced by dmc1Δ meiosis , which is consistent with the random segregation of univalent chromosomes , which may provide a low percentage of progeny cells with balanced chromosome sets . Therefore , our observations provide strong support of efficient DSB repair in the absence of Dmc1p . While we cannot exclude the rare occurrence of bivalents in dmc1Δ meiosis , sporadic interhomologue recombination can not account for the complete disappearance of DSBs observed in the PFGE assay . Therefore , it is likely that this Dmc1-independent repair , like in budding yeast [2] , [39] , preferentially takes place via the sister with the help of Rad51p , yet with only little Rad51p actually localizing to DSBs ( see above ) . Only in the presence of Dmc1p are bivalents regularly formed . Thus , Dmc1p is needed for recombination with the homologue . The formation of strong Dmc1p foci during meiotic prophase invites the simple interpretation that long Dmc1p-containing nucleoprotein filaments are formed at DSBs . It is tempting to speculate that short Rad51p nucleoprotein filaments are sufficient for intersister recombination whereas long nucleoprotein filaments consisting primarily of Dmc1p are advantageous or even indispensible ( but not sufficient ) for interhomolog recombination . Both in the presence or absence of Rad51p , ca . 170 Dmc1p foci were counted per meiotic nucleus ( see Results ) . Data from budding yeast , where the maximum number of Dmc1-Rad51 foci was about 2 . 5-fold less than the average recombination frequency ( CO + NCO ) , suggest that , because of their transient nature , recombination foci may underestimate recombination events [38] , [40] . Thus , it can be estimated that in Tetrahymena there occur 200 or more DSBs per meiosis , whereas the rod- or ring-shape of bivalents suggests that only a fraction are converted to chiasmata . In this respect , Tetrahymena resembles higher plants and animals such as maize , the lily , and the mouse , where a considerable excess of DSBs over crossovers/chiasmata was found [41]–[43] . For the success of meiosis it is essential that strand exchange takes place between homologous chromosomes rather than sisters , which would be the more readily available option . In the budding yeast , two mechanisms for the promotion of interhomolog recombination have been identified ( see [3] ) . In one , the homolog-over-sister recombination preference is conferred by a not yet understood activity of Dmc1 [2] . In the other , an activation/phosphorylation cascade involving the Red1 , Hop1 , and Mek1 proteins at the axial elements of the SC impedes intersister exchange by Rad51 through phosphorylation of its binding partner , Rad54 ( see [3] , [7] , [44]–[47] ) . Similarly , axial element proteins may be involved in recombination partner choice in the fission yeast [48] and in C . elegans [49] . Such an axial element-dependent barrier or impediment to unwanted intersister recombination is not universal , however . In Arabidopsis dmc1 mutants , DSBs are readily repaired by a noncrossover pathway ( possibly via the sister chromatid ) , even when the SC is in place [10] , [11] . In Tetrahymena , homologous chromatids are closely conjoined in the tubular meiotic prophase nucleus , with homologous loci in opposing positions [29] . Nevertheless , the cohesion of sister chromatids may cause their more intimate contact and an intrinsic preference for intersister recombination . Homologues of axial element proteins and SC-related structures have not been detected ( see [25] ) , thus a mechanism involving axial elements to overcome the intersister bias may be lacking . Our data suggest that Tetrahymena shares with the budding yeast a Dmc1-dependent mechanism to promote a sufficient rate of interhomolog recombination . A simple physical model would pose that Dmc1p-containing nucleoprotein filaments are longer than those consisting of Rad51p and therefore are able to bridge the larger interhomolog distances .
Tetrahymena thermophila strains B2086 and CU428 served as wild types and as the source material for the construction of gene knockout and knockdown strains . spo11 knockout ( spo11Δ ) lines were described previously [25] . Cells were grown in standard medium at 30°C [50] . For induction of conjugation and meiosis , cells of complementing mating types were starved in 10 mM Tris-Cl ( pH 7 . 4 ) for at least 16 h and mixed in equal amounts ( ∼2×105 cells/ml ) . To create the dmc1 knockout constructs , ∼500 bp-fragments of genomic Tetrahymena DNA upstream and downstream of the DMC1 open reading frame ( ORF ) were amplified using the following primers: DMC1KO5FW ( 5′-cag aag ttg cta gaa gc-3′ ) , DMC1KO5RV ( 5′-gtc tat cga att cct gca gcc cgc ttt tca gtg cag cta g-3′ ) , DMC1KO3FW ( 5′-ctg gaa aaa tgc agc ccg cct tct act ggt tga ttt-3′ ) , and DMC1KO3RV ( 5′-gct gat aga tct aaa tga aat taa g-3′ ) . These fragments were then joined to each end of the neo4 selection cassette using overlapping PCR [51] . The knockout construct was introduced into B2086 and CU428 cells by biolistic transformation as described previously [52] . The NEO4 resistance gene is expressed under the Cd2+-inducible MTT1 metallothionein promoter [53] . Transformants were selected in media containing 0 . 1 – 1 µg/ml CdCl2 and increasingly higher concentrations of paromomycin ( from 120 µg/ml to 10 mg/ml ) until the wild-type chromosomes were completely replaced by the knockout chromosomes in the somatic nucleus . To create the rad51 RNAi construct , a ∼500 bp-fragment of the RAD51 ORF was amplified from genomic DNA using PCR primers to add appropriate restriction sites for cloning into the RNAi hairpin vector ( PmeRad51FW 5′-cgt tta aac gaa aca ggc tct ctc act g-3′ , ApaRad51FW 5′-cgg gcc cga aac agg ctc tct cac tg-3′ , SmaRad51RV 5′-gcc cgg gcc gaa ttc gtc agc aag tc-3′ , XhoRad51RV 5′-gct cga gcc gaa ttc gtc agc aag tc-3′ ) . These fragments were then used to replace the SERH3 fragments in the RNAi vector construct described previously [20] . The finished hairpin construct was introduced into mating cells by biolistic transformation at 10 hrs post-mixing to allow cells to process the rDNA vector [52] , [54] . Transformants were selected initially in media containing 120 µg/ml paromomycin , and then were transferred to increasingly higher concentrations , up to 600 µg/ml . Expression of dsRNA under the MTT1 promoter was induced by the addition of CdCl2 ( final concentration: 0 . 1 µg/ml ) to cells carrying the hairpin construct ( rad51hp ) . For RAD51 knockdown in meiosis , CdCl2 was added ca . 3 h after the beginning of starvation , i . e . when mitotic divisions had ceased . Since CdCl2 was found to impair conjugation , the rad51hp cells were washed twice and resuspended in 10 mM Tris-Cl ( pH 7 . 4 ) before mixing . For some experiments , two rad51hp lines of different mating types were mixed . For others , a rad51hp line was mated to lines not carrying the rad51 hairpin construct . Because RNAi causes a systemic response in Tetrahymena [20] , the construct-free partner also displayed the RNAi phenotype . Cells were harvested at appropriate times in meiosis . For controls , meiosis was induced in the wild type under the same regime of CdCl2 treatment , and in rad51hp cells without CdCl2 treatment . Meiosis was normal in CdCl2-treated wild type . In rad51hp cells without CdCl2 , bivalents appeared somewhat less condensed , but meiotic divisions were not notably affected . To induce Rad51p overexpression , cells were exposed to UV radiation ( 254 nm UV-C; 20 Joule/m2 ) using a Stratalinker UV crosslinker [29] . To study the rad51-dmc1 double deficiency phenotype , a dmc1hp strain was created and mated to a rad51hp strain . The dmc1 interfering RNA was constructed in the same way as the rad51hp . The primers used were: PmeDmc1FW 5′-cgt tta aac gag ttt gtt ctc ggt act ac-3′ , ApaDmc1FW , 5′-cgg gcc cga gtt tgt tct cgg tac tac-3′ , SmaDmc1RV 5′-gcc cgg gca gcc att ctt tat aat ctg ctc-3′ , XhoDmc1RV 5′-gct cga gca gcc att ctt tat aat ctg ctc-3′ . Expression of dsRNA was induced as described above . To detect meiotic DSBs , DNA was separated by PFGE under conditions where intact chromosomes do not enter the gel , whereas DSB-dependent chromosome fragments of different sizes appear as a distinct band [13] . Chromosome-sized DNA was prepared in agarose plugs as described elsewhere [13] . Gels were run on a CHEF apparatus . Chromosomes of the generative nucleus were distinguished from the background of somatic minichromosomes ( which are distributed throughout the gel ) by Southern hybridization with a micronucleus-specific probe [13] . DMC1-mCherry was created by fusing the mCherry red fluorescent protein gene to the C-terminus of the DMC1 ORF . To create the tagging construct , the last ∼500 bp of the DMC1 gene ( excluding the stop codon ) , and ∼500 bp of DNA downstream of the DMC1 gene were amplified from genomic Tetrahymena DNA using the following primers: 5AmDmc1FW ( 5′-gct gat ggc gat gaa tga aca ctg gct cta cta gtt gtt gat tca ata atg gc-3′ ) , DMC1mCheRV ( 5′-gtt atc ttc ttc tcc ttt tga aac cat gga tcc acc agt aga agg ctt ttt atc aca ttc aac-3′ ) , Neo4-Dmc1FW ( 5′-ccc ggg gga tct gaa ttc gat atc aag ctt gaa tat tct ttg aga aag tta gtt aaa tga-3′ ) and 3AmDmc1RV ( 5′-gcg agc aca gaa tta ata cga ctg ctg ata gat cta aat gaa att aag aat ga-3′ ) . These fragments were then joined to each end of the mCherry-neo4 cassette ( amplified from the pmCherry-neo4 plasmid , gift of Kazufumi Mochizuki ) using overlapping PCR . Tagged mCherry was at least partially functional , since matings of Dmc1-mCherry cells to dmc1Δ cells produced viable sexual progeny , although with a reduced frequency ( Table 1 ) . Strains expressing Rad51-HA were created in a similar manner using the following primers and the pHA-Neo4 plasmid ( a gift from Kazufumi Mochizuki ) : 5′FW ( 5′-gct gca tgc gat gaa tga aca ctg ttc agc cac tgc tct tta c-3′ ) , 5′RV ( 5′-aag ttc ttc acc ctt aga aac cat gga tcc ctc gtt gaa gtc ttc aat acc-3′ ) , 3′FW ( 5′-ccc ggg gga tct gaa ttc gat atc aag ctt gct aaa aga taa taa gat aaa att c-3′ ) and 3′RV ( 5′-gcg gtc gac gaa tta ata cga cta tat tat att ggt ata aca tta ttt tat ag-3′ ) . Transformations and selections were performed as for the knockout strains described above . Antiserum was produced in rabbits against the peptide sequence AIYAIGKGGIEDFNE from the C-terminus of the Rad51 protein . This region has only low similarity to the related Dmc1p . The serum was immunopurified against the polypeptide ( Eurogentec , Seraing , Belgium ) , and its specificity was confirmed by its failure to label nuclei of rad51i cells ( Figure S3 ) . Protein extracts were prepared from 5 ml of conjugating cells by trichloroacetic acid precipitation . 20 µl of extracts were run on 10% SDS-PAGE gels and blotted to Hybond-P PVDF membrane ( Amersham Biosciences ) . Proteins of interest were detected by incubating the blot for 2 h with anti-Rad51/Dmc1 antibody ( 1∶200 mouse monoclonal , Clone 51RAD01 , NeoMarkers , Fremont , CA ) in TBST ( 20 mM Tris pH 7 . 5 , 140 mM NaCl , 0 . 05% Tween 20 ) +1% dry milk , washing , incubating for 1 h with HRP-conjugated anti-mouse antibody ( 1∶100 . 000 ) , washing , then incubating with chemiluminescent reagent ( Thermo Scientific ) , and exposing to X-ray film . Following [22] and [18] , 5 ml of a suspension of conjugating cells were fixed by the addition of formaldehyde and Triton X-100 ( final concentrations of 4% and of 0 . 5% , respectively ) . After careful mixing , the cells were left for 30 min at room temperature , then centrifuged , and the pellet was resuspended in 500 µl of a solution of 4% paraformaldehyde and 3 . 4% sucrose in water . A drop of this mixture was spread on a clean slide and air-dried . These slides were used for nuclear staining with DAPI ( 4′ , 6-diamidino-2-phenylindole ) or for immunostaining . For the cytological detection of chromatin-associated Dmc1p and Rad51p foci , a protocol for enhanced detergent spreading of cells was applied [13] . In short , a mixture of 450 µl of 10% Triton X-100 and 50 µl of 37% formaldehyde solution was quickly added to a tube with 5 ml of conjugating cells . The liquids were mixed by inverting the tube , and after 25 min on ice another 450 µl of formaldehyde solution were added . After 5 min the cells were pelleted and resuspended in 500 µl of a solution of 4% paraformaldehyde and 3 . 4% sucrose in water . Eighty microliters of this suspension were spread on a slide and allowed to dry . For immunostaining , slides prepared by either of the above methods were washed with 1× PBS and 1× PBS +0 . 05% Triton X-100 , incubated with primary antibody for 3 h at room temperature or over night at 4°C , washed as above , incubated with Cy3- or FITC-labeled secondary antibody for 1 . 5 h–3 h at room temperature , washed again and mounted under a coverslip in Vectashield anti-fading agent ( Vector Laboratories Inc . , Burlingame , CA , ) supplemented with 0 . 5 µg/ml DAPI as a DNA-specific counterstain . The primary antibodies were: anti-Rad51/Dmc1 ( 1∶50 mouse monoclonal , Clone 51RAD01 , NeoMarkers , Fremont , CA ) , anti-Rad51 ( 1∶100 rabbit polyclonal ) , anti-DsRed fluorescent protein/mCherry ( 1∶50 rabbit polyclonal , Clontech , Mountain View , CA ) , anti-HA ( 1∶200 mouse monoclonal , Roche Diagnostics ) , anti-phosphorylated H2A/H2A . X ( 1∶200 rabbit polyclonal 07-745 , Upstate Biotechnology , Charlottesville , VA ) and anti-phosphorylated H2A . X ( 1∶200 mouse monoclonal , BioLegend , San Diego , CA ) . To test if two proteins occupy adjacent positions within the cell , we applied a proximity ligation assay [32] . Conventional cell preparations were first incubated with primary antibodies generated in the rabbit and in the mouse , respectively , which recognize the proteins of interest . Next , secondary anti-rabbit and anti-mouse antibodies coupled with short complementing DNA strands ( Olink Bioscience , Uppsala , SWE ) were applied , using the immunostaining protocol from above . Reactions for the ligation of DNA strands to a circularized oligo and the subsequent rolling circle amplification incorporating labeled nucleotides were performed using the Duolink II kit ( Olink Bioscience , Uppsala , SWE ) according to the instructions of the manufacturer . Slides were washed and mounted under a coverslip in Vectashield plus DAPI as above . Fluorescence indicating the interaction between oligonucleotides attached to neighboring antigens was evaluated under the fluorescence microscope . Cells from 5 ml of conjugating cell suspension were pelleted and fixed in 1 ml of Carnoy's fixative ( methanol-chloroform-acetic acid , 6∶3∶2 ) . After 1 h at room temperature , cells were pelleted and resuspended in 500 µl of 70% ethanol . A few drops of this suspension were applied to a slide and air-dried . A FISH probe was produced by pooling PCR-amplified sequences corresponding to a 22 . 1 kb intercalary chromosomal region [29] . The purified PCR products were labeled with Cy3 by nick translation . The probe and chromosomal DNA were denatured by hot formamide and hybridized for 36 h at 37°C . Fluorescent signals generated by DAPI , immunostaining , proximity ligation or FISH were visualized by appropriate filter combinations in a fluorescence microscope and recorded with a cooled CCD camera . Of thick DAPI- and immuno-stained nuclei , z stacks were taken using MetaView software ( Universal Imaging , Downingtown , PA ) , deconvolved using AutoDeblur ( AutoQuant Imaging , Watervliet , NY ) and projected with ImageJ ( Wayne Rasband , N . I . H . ; http://rsb . info . nih . gov/ij/ ) software . Images from different color channels were colorized and merged using Photoshop software . For counting recombination foci and for colocalizing Dmc1p and γ-H2A . X signals in spread cells , image stacks were taken at 100× magnification and projected . In Photoshop , resolution was enhanced to 150 pixel/inch and brightness and contrast were adjusted to give optimal differentiation from background staining . A dot was counted as a recombination focus if its size and/or brightness was higher than the average background signal . Stretched foci were counted as single if they were oval or as two or more if they had constrictions . In all cases , the fluorescence was Cy3 . For evaluating Dmc1-mCherry × rad51i and Dmc1-mCherry × wild type mating cells , the mCherry tag was enhanced with anti-dsRed and Cy3-coupled secondary antibody . For meioses of wild type × wild type matings , Dmc1p was detected using anti-Rad51/Dmc1 antiserum and Cy3-coupled secondary antibody . To determine the brightness of fluorescence foci , gray values ( within a range from 0–255 ) were measured on 8-bit images using ImageJ . For this , images of stained nuclei were adjusted to the same level of background signal intensity and the mean gray values of foci within a mask were calculated for the strongest ten foci for each evaluated nucleus . Evaluation of colocalizing Dmc1p and γ-H2A . X signals was done on well-spread and flat regions of nuclei . | Sexual reproduction relies on meiosis , the specialized cell division that allows diploid organisms to halve their chromosome content , resulting in the production of gametes containing one copy of each chromosome . In humans , defects in meiosis cause infertility , stillbirths , and congenital diseases . Homologous recombination is a key step in the meiotic program and is essential for maintaining the fidelity of segregation and for creating genetic diversity . Meiotic recombination begins with self-inflicted DNA breaks that are repaired using the homologous chromosome as a template , in a process that depends upon the universal repair protein Rad51 and its meiosis-specific homologue , Dmc1 . The relative contributions of Rad51 and Dmc1 to homologous recombination differ among yeasts , plants , and mammals . We have undertaken a study of these proteins in the evolutionarily distant model organism Tetrahymena thermophila with the hope of clarifying the specialization of these recombinases throughout eukaryotic evolution . We show that , while Rad51 is required for DNA repair , only Dmc1 localizes prominently to meiotic DNA break sites . Also , repair via the homologous chromosome depends on Dmc1 . These results suggest that nucleoprotein filaments consisting of primarily Dmc1p allow efficient interhomologue repair , while shorter Rad51 filaments may suffice for repair via the sister chromatid . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/nuclear",
"structure",
"and",
"function",
"molecular",
"biology/recombination",
"cell",
"biology/cell",
"growth",
"and",
"division",
"molecular",
"biology/chromosome",
"structure",
"genetics",
"and",
"genomics/chromosome",
"biology",
"molecular",
"biology/dna... | 2011 | The Recombinases Rad51 and Dmc1 Play Distinct Roles in DNA Break Repair and Recombination Partner Choice in the Meiosis of Tetrahymena |
In molecular recognition , it is often the case that ligand binding is coupled to conformational change in one or both of the binding partners . Two hypotheses describe the limiting cases involved; the first is the induced fit and the second is the conformational selection model . The conformational selection model requires that the protein adopts conformations that are similar to the ligand-bound conformation in the absence of ligand , whilst the induced-fit model predicts that the ligand-bound conformation of the protein is only accessible when the ligand is actually bound . The flexibility of the apo protein clearly plays a major role in these interpretations . For many proteins involved in signaling pathways there is the added complication that they are often promiscuous in that they are capable of binding to different ligand partners . The relationship between protein flexibility and promiscuity is an area of active research and is perhaps best exemplified by the PDZ domain family of proteins . In this study we use molecular dynamics simulations to examine the relationship between flexibility and promiscuity in five PDZ domains: the human Dvl2 ( Dishevelled-2 ) PDZ domain , the human Erbin PDZ domain , the PDZ1 domain of InaD ( inactivation no after-potential D protein ) from fruit fly , the PDZ7 domain of GRIP1 ( glutamate receptor interacting protein 1 ) from rat and the PDZ2 domain of PTP-BL ( protein tyrosine phosphatase ) from mouse . We show that despite their high structural similarity , the PDZ binding sites have significantly different dynamics . Importantly , the degree of binding pocket flexibility was found to be closely related to the various characteristics of peptide binding specificity and promiscuity of the five PDZ domains . Our findings suggest that the intrinsic motions of the apo structures play a key role in distinguishing functional properties of different PDZ domains and allow us to make predictions that can be experimentally tested .
A number of structural studies comparing holo and apo forms of proteins have demonstrated that ligand binding is often coupled to conformational changes of the interacting partners [1]–[3] . The real challenge is , however , to uncover the exact sequence of events resulting in the observed structural changes . Two main models , the induced fit ( Koshland ) and the conformational selection ( or population shift ) hypothesis ( see [4] for a review ) , have been introduced to describe the limiting cases of the complex process of molecular recognition [5]–[8] . According to the induced fit model , ligand binding happens first and the formation of a ‘weak complex’ is followed by the conformational rearrangement of the protein that results in stronger binding [9] . By contrast , in the conformational selection model , the intrinsic dynamics of the protein lead it to spontaneously transition between a stable unbound and a less stable ‘bound conformation’ . As the apo protein actually visits the bound state with significant probability , the ligand can bind directly to this conformation shifting the distribution of conformers towards the bound population . As recently reviewed [4] , it seems likely that the induced fit and conformational selection mechanisms often act together in the ligand recognition process . Furthermore , in terms of protein-protein interactions , it is increasingly clear that many proteins display functional promiscuity which requires them to be able to interact with multiple partners [10] . If the conformational selection mechanism is involved in promiscuous ligand binding , this assumes that the protein needs to visit multiple ( often dissimilar ) binding conformers capable of binding the different ligands . An example of structural evidence of such multi-specificity can be found in the X-ray crystallography study of the SPE7 antibody ( a monoclonal immunoglobulin E raised against a 2 , 4-dinitrophenyl hapten ) that has been shown to adopt different binding conformers and is consequently able to bind to multiple unrelated antigens [11] . Another example is the NMR study of apo ubiquitin which has found that this protein exists in an ensemble of conformers that are almost identical to the conformations of ubiquitin in complex with 46 different binding partner proteins [12] . One of the best known examples of a promiscuous enzyme is cytochrome P450 which has been shown to adopt a great variety of active site conformations and is able to bind and transform diverse substrates [13] . As shown by these examples , the intrinsic dynamics of promiscuous proteins let them visit multiple unrelated binding conformers and the property of multispecificity seems to be related to conformational flexibility . Promiscuous proteins that are able to bind to multiple partners through conformational selection need to explore a larger conformational space than those that bind to only a single partner . More rigid binding sites therefore may have restricted specificity with the benefit of higher binding affinity . Indeed , a study of human cytochrome P450 enzymes has found that while a relatively rigid member of the family ( CYP2A6 ) has narrow substrate specificity , the most flexible member ( CYP3A4 ) is also the most promiscuous one [14] . Functionally promiscuous proteins could be of key importance for the emergence of new functions in protein evolution . Recent research about the relationship between binding promiscuity , conformational flexibility and evolvability of proteins has been reviewed by Tokuriki et al . [15] , [16] . As discussed in these reviews , these studies suggest that for proteins that exist in equilibrium between a highly populated native state ( interacting with a native ligand ) and less populated conformers ( binding to alternative partners ) , mutations can gradually shift the equilibrium towards a promiscuous conformer . This can eventually lead to a new dominant primary function . While mutations may be neutral with regards to the original function ( i . e . hardly change the relative occupancy of the native conformer ) , they may cause significant increase in the occupancy of the alternative conformer . On the other hand , point mutations that reduce the occupancy of promiscuous conformers may result in a decreased flexibility ( rigidification ) but increased specificity ( and higher affinity ) for the native ligand as for example observed in the process of antibody maturation [17] . Promiscuity may therefore be a common feature of highly evolvable proteins . Despite their highly conserved overall fold and binding sites , PDZ ( PSD-95 , Dlg , ZO-1 ) domains have been found to have surprisingly diverse binding specificities [18] . PDZ domains bind peptidic ligands , usually located at the C-terminus of partner proteins . A recent study at the genome level confirmed that this location is dominant [19] , but other modes of interaction have also been reported [20]–[22] . Although a series of different classification systems have been proposed aiming to organize PDZ domains based on their preference towards peptide ligands there is no consensus on the best way of classification [23] , [24] although some progress has been made towards mapping determinants of specificity [25] . PDZ specificity turned out to be unexpectedly complex as many PDZ domains are able to bind to multiple ligands that belong to different classes of peptide motifs . This property is often referred to as degenerate specificity , multivalent specificity or most commonly , binding promiscuity [10] . In addition , single peptides have been shown to bind to multiple PDZ domains . The complex picture of PDZ-peptide interactions therefore makes it rather difficult to develop a simple specificity-based classification scheme . In addition , very little is known about what determines the specificity and promiscuity of PDZ domains . To address this question , Stiffler et al . [26] have used protein microarrays and quantitative fluorescence polarization to study the binding specificity of 157 mouse PDZ domains and found only a weak correlation between the pairwise sequence divergence of PDZ domains and their divergence in selectivity space . The fact that overall sequence similarity proved to be a poor predictor of PDZ domain function indicates that the majority of sequence variation in the PDZ family is neutral with regards to peptide-binding selectivity . This also suggests that binding specificity is mostly determined by only a subset of residues that are likely to be located in the binding pocket of the domain [26] . In order to study the sequence determinants of specific ligand recognition , Tonikian et al . [25] performed mutagenesis at ten binding site positions in the Erbin PDZ domain . As a result , they identified several mutations that altered binding specificity . Since not all of these critical residues were in direct contact with the ligand , Tonikian et al . concluded that both direct interactions and cooperative , long-range effects may play important roles in determining the specificity of PDZ domains [25] . In a recent study , using a combinatorial peptide library and site-directed mutagenesis , Shepherd et al . [27] have found that only four point mutations were enough to switch between the distinct binding specificities of the Tiam1 ( T-cell lymphoma invasion and metastasis 1 ) PDZ and Tiam2 PDZ domains . Gee et al . [28] have come to similar conclusions after performing in-vitro mutagenesis studying the PDZ domains of PSD-95 ( postsynaptic density protein 95 ) and α1-syntrophin . By identifying a few critical sequence positions , they have found that single-amino acid substitutions can alter specificity and affinity of PDZ domains for their ligands . The fact that ligand specificity relies on minor sequence modifications , while the chemistry of the binding pocket and the overall fold are well conserved , suggests a very favorable flexibility property of the PDZ domain fold [29] . PDZ domains are both versatile and robust because mutations frequently change their specificities without a loss of function [25] . Similar robustness under high mutational pressure has also been observed for other peptide-binding domains , for example the WW [30] and SH3 domains [31] . On the other hand , a number of experimental and computational studies ( outlined below ) have shown that the conformational dynamics of PDZ domains may also play a crucial role in determining binding specificity . These results suggest that the intrinsic fluctuations of PDZ structures are also likely to be related to the selectivity for peptide ligands . Recently , Gerek et al . [32] used a modified coarse-grained elastic network model to find characteristic residue fluctuation patterns for PDZ domains belonging to different specificity classes . By clustering these residue fluctuation profiles , they have identified common motion characteristics of Class I and Class II type PDZ domain interactions [32] . Basdevant et al . performed 20–25 ns molecular dynamics simulations of 12 PDZ domain complexes and used the MM/PBSA ( Molecular Mechanics/Poisson-Boltzmann Surface Area ) method to analyze electrostatic , nonpolar and configurational entropy contributions to the binding free energies [33] . Their results show that the degree to which the dynamics of the peptide ligands are coupled to those of the PDZ domains varies highly . They concluded that complex-specific dynamical or entropic responses may form the basis of the selective recognition of peptides . It is important to note that different flexible docking strategies have already been proposed to be able to incorporate the effect of binding site flexibility in structure-based drug design studies targeting PDZ domains [34] , [35] . Another aspect that has been investigated is the role of temperature on binding behaviour . Staneva and Wallin [36] applied an all-atom Monte Carlo based approach to analyze various aspects of the process of peptide binding to PDZ domains . They found that the probability that peptide ligands can occupy the correct bound state in the simulations increased sharply with the decrease of temperature . In another study , Cecconi et al . [37] have analyzed the temperature-dependence of the unbinding of peptide ligands from PDZ domains . They have found that the free-energy landscape determining the kinetics of ligand escape is sensitively dependent on the temperature . However , PDZ-peptide complexes are stabilized within a physiologically relevant temperature interval . Given all of the above , we were interested in the role of conformational dynamics in determining the ligand binding specificity of PDZ domains . In particular , given the possible relationship between flexibility and promiscuity , we wanted to examine how well the property of multi-specificity of these domains is correlated with the flexibility of their binding pockets . We were also interested to examine to what extent PDZ domains obey the conformational selection versus induced fit mechanism . We thus selected five , well-characterized , PDZ domains: Dvl2 PDZ capable of binding both C-terminal and internal ( i . e . not at the terminus of a protein ) peptides and shows large conformational changes between binding modes , Erbin ( ERBB2 interacting protein ) PDZ which binds both class I and class II ligands , but comparison with the apo structure reveals very little conformational change , InaD PDZ1 for which it is known that peptides bind in different modes , but structural information is thus far only available for one mode , PTP-BL PDZ2 for which induced fit has been predicted to be important in the binding process and GRIP1 PDZ7 for which structural studies suggest that the binding cleft is not capable of binding peptides in the expected manner for PDZ domains . All five of the aforementioned PDZ domains are of clinical interest due to their central role in disease pathways . Four of these PDZ domains ( Dvl2 PDZ , Erbin PDZ , InaD PDZ1 and PTP-BL PDZ2 ) are promiscuous in the sense that they are able to interact with multiple partners . However , while for example , Dvl2 PDZ is capable of interacting with peptides using different binding modes ( binding both classical C-terminal and non-classical internal peptides ) , Erbin PDZ is able to interact only with very similar peptide binding modes . On the basis of this , one can formulate a definition of strong promiscuity , which is the ability to interact with multiple ligands that require the binding pocket to adopt significantly different conformations . In this sense , Dvl2 PDZ is promiscuous and Erbin PDZ is not . If conformational selection plays a role in the recognition of peptides , the above-defined property of promiscuity must correlate with intrinsic conformational flexibility since the binding pocket needs to visit all different conformations required for binding multiple ligands . In this paper we explore the relationship between the dynamics , promiscuity and flexibility of PDZ domains . The results have implications for many protein-protein interaction pathways .
To compare the inherent flexibility of the five PDZ binding pockets , we used a measure of the overall fluctuation , Θ , which reflects the mean pairwise distance variance of binding pocket residues ( See Methods for details ) . This approach has the added advantage that it is not superposition dependent as it only depends on distances rather than coordinates . The overall fluctuation was calculated for the five conformational ensembles of the 200 ns MD simulation trajectories ( 40000 snapshots for each PDZ domain ) . We assessed the convergence of the trajectories via calculation of the root mean square inner product ( RMSIP ) and obtained values between 0 . 59 and 0 . 69 for the binding pocket residues ( and high overlaps for the full proteins as well ) from the simulations which according to Lagerge and Yonetani [38] suggests adequate convergence ( see Supporting Information , Text S1 , for more details ) . The Θ fluctuation values of the five binding pockets ( i . e . the five sets of binding site residues defined by the multiple sequence alignment ) are summarized in Table 3 . As discussed in Methods , the Θ measure shows the size of conformational space the binding pocket explores in the simulation . Table 3 shows that despite the high structural similarity of the five binding sites ( Table 2 ) , one can see large differences in the extent of their intrinsic fluctuation . The InaD PDZ1 and Dvl2 PDZ binding sites have the most flexible binding pockets , while the binding site of Erbin PDZ is the most rigid of these five PDZ domains . The Θ value of Dvl2 PDZ is almost twice as large as that of Erbin PDZ . These results are in good agreements with the conclusions of experimental studies [22] , [39]–[42] that have found that Erbin PDZ binding site shows little structural variability while the Dvl2 PDZ binding site is flexible showing large structural variation . The results suggest that the rigidity/flexibility of these binding sites demonstrated in other studies by comparison of apo and holo crystal structures can be explained by the intrinsic dynamics of the apo proteins . The flexibility of the binding pocket of the Dvl2 PDZ domain has been discussed in the literature before [22] . Therefore we decided to compare the dynamics between Dvl2 PDZ and Erbin PDZ domains . The difference in the overall fluctuation of the two binding pockets can also be seen in their fluctuation matrices ( Figure 2A , B ) , defined as the matrix of variance of pairwise residue distances . We also define “flexibility” as a measure of the maximum range any pairwise residue distance can exhibit ( see Methods ) . The flexibility matrices , which essentially capture extreme movements , reveal that , as expected , there are regions of high flexibility for Dvl2 PDZ . They also reveal , unexpectedly that although the fluctuation matrices suggested that Erbin PDZ is quite rigid , they also highlight that there is flexibility in terms of the distance between K396 ( located at the C-terminal of the α1 helix ) and the β2 strand and in particular S335 ( see Supporting Information Figure S1 ) . Taking the result of the fluctuation and flexibility matrices together suggests that a section of the binding site can open up considerably , but that these extremes in conformation are infrequent and essentially the Erbin PDZ binding site behaves as a rigid structure . To better understand the role that intrinsic dynamics might play in ligand binding to the Dvl2 PDZ domain , we performed the fluctuation and flexibility analysis on an experimentally derived ensemble . We took the structure of the apo Dvl2 PDZ domain ( PDB code: 2rey ) and four crystal structures of different ligand-bound conformations ( PDB codes: 3cbx , 3cby , 3cbz and 3cc0 which are also referred to in the literature as the pep-C1 , pep-N1 , pep-N2 and pep-N3 complexes [22] ) . The pep-C1 structure exemplifies C-terminal ligand binding , whereas the other three illustrate internal ligand binding . The flexibility matrix was computed for this ensemble and is shown in Figure 3 . The matrix shows us which binding pocket residue pairs have the largest relative displacement between the apo and ligand-bound structures . The experimentally derived flexibility matrix has remarkable similarity to the simulation-based fluctuation and flexibility pattern ( Figure 2A and C ) with a correlation of 0 . 74 and 0 . 68 respectively . The largest displacements seen experimentally are for residues I14 and S15 with respect to the α2 helix , which is the same as that observed in the simulations . This suggests that the Dvl2 PDZ domain is capable of visiting conformations that are consistent with the ligand-bound conformations even in the absence of ligands . We were interested to know how the snapshots of the MD simulations were distributed in conformational space . To that end we performed multi-dimensional scaling on snapshots taken every 50 ps ( the first ns of the trajectories were excluded ) . The input was thus a 3981×3981 matrix containing the pairwise dRMSD dissimilarity values of 3981 conformers . Groups of similar conformers were identified with the k-means cluster analysis and clustering was validated with the silhouette index measure ( see Methods ) . The optimal number of clusters corresponding to the maximal overall average silhouette index ( SOVER = 0 . 411 ) was found to be 2 . Figure 4A shows the results of multi-dimensional scaling ( MDS ) where conformations are represented by dots on a 2D-map and similar conformers are adjacent . The map suggests that the conformational space can be split into two distinct contiguous clusters . When the ligand-bound structures were included in this MDS analysis ( Figure 4B ) it was found that one ( pep-N2 ) resides within cluster two , whilst the other three sit within cluster one . Examination of ligand-bound conformations with , superposed on , the medoid conformations ( ie . those conformations that are most representative of the ensemble ) from the two clusters ( Figure 4C , and D ) shows that the key difference lies in the motion of I14 and S15 ( at the N-terminal of the β2 strand ) relative to the α2 helix . Thus the intrinsic fluctuations observed for the Dvl2 PDZ domain allow access to two distinct conformational states that have been captured in ligand-bound structures . In contrast , MDS performed on the data for the Erbin PDZ domain shows just one cluster and the two experimental structures lie within this cluster ( Figure 4E ) . The complex with the class I peptide is located close to the cluster center , while the complex with the class II peptide is placed closer to the edge of the cluster . The Erbin PDZ domain is promiscuous in the sense that it binds multiple peptides , but the experimental structures and this analysis shows that these peptides are essentially binding to the same conformational state and thus it does not satisfy the “strong” definition of promiscuity defined here . The binding pocket of the InaD PDZ1 domain has the largest overall fluctuation of the five PDZ binding pockets examined here ( Table 3 ) . The fluctuation matrix ( Figure 5A ) shows that the part of the InaD PDZ1 domain that fluctuates the most is the three residues at the C-terminal end of the α2-helix ( I91 , K92 , and E93 ) with regards to the entire β2-strand . However , the flexibility matrix ( Figure 5B ) shows that this PDZ domain can undergo even larger distortions within the binding cleft . MDS analysis of the conformational ensemble identified two main clusters ( Figure 5C ) with the known experimental structure of the InaD PDZ1 domain in complex with the NorpA peptide ( PDB code: 1ihj ) belonging to cluster one . The overall average silhouette index , SOVER was 0 . 43 and as can be seen the division between clusters is not as distinct as for the Dvl2 PDZ . The presence of two distinct clusters for InaD is intriguing and raises the question of whether the second cluster has biological relevance . Besides the NorpA peptide , the InaD PDZ1 domain has been shown to bind to the unconventional myosin NinaC . Intriguingly the experimental results [43] suggest that InaD PDZ1 may interact with NinaC in a different mode than it does with NorpA [44] . If conformational selection plays a role in this interaction , then one would expect the InaD PDZ1 domain to be relatively flexible in the apo state and able to visit distinct regions of conformational space , which is exactly what is observed here . Conformations in Cluster 1 are likely to be relevant for NorpA binding whilst excursions into Cluster 2 may be essential for NinaC peptide binding . Taken together these data suggest that the InaD PDZ1 domain is likely to satisfy the ‘strong’ definition of promiscuity as it probably binds to different partners using considerably different binding modes . Thus we would predict from this that any structure of the InaD PDZ1-NinaC complex would be placed into Cluster 2 of the MDS analysis . These results are in support of the earlier anticipation of Kimple et al [44] and Wes et al [43] that InaD PDZ1 binds to NorpA and NinaC using different binding modes . Based on our results we would predict that the difference is expected to be in the shift of the C-terminal end of the α2-helix ( I91 , K92 and E93 ) with respect to the β2-strand . Although InaD PDZ1 ( and also Dvl2 PDZ ) has distinct conformation clusters defined by k-means clustering , it is perhaps also useful to define states in terms of kinetics . We performed a temporal analysis to ascertain whether our geometrically defined states are supported by a kinetic definition simply defined by asking is the intra-cluster relaxation time faster than the inter-cluster transition time ( see Supporting Information , Text S1 , for details ) . For InaD PDZ1 , the average inter-cluster transition time was 14 . 1 ns whilst the intra-cluster relaxation time was 100 ps . Similarly for Dvl2 PDZ , the average inter-cluster transition time was 7 . 03 ns whereas the intra-cluster relaxation time was again 100 ps . Thus , this analysis suggests that the conformational clusters defined by the dRMSD similarity measure correspond to kinetically separated , metastable states of the protein . The fluctuation pattern of PTP-BL PDZ2 ( Figure 6A ) shows that this domain [45] has a considerably rigid binding site , similar to the Erbin PDZ domain . However , the flexibility pattern ( Figure 6B ) reveals that the N-terminal end of the α2 helix is flexible with regards to the β2 strand . MDS analysis ( Figure 6C ) shows that the majority of conformations appear to be distributed within a single compact cluster , which also has a large number of outliers . The results here place the experimental ligand-bound conformation in the main conformational cluster , but that does not rule out the possibility that induced-fit plays an essential role in the binding process . Indeed for PTP-BL PDZ2 , the binding to the Adenomatous Polyposis Coli-protein ( APC ) peptide has been proposed to occur through induced fit [46] . In order to investigate this , the structural differences between the APC-bound conformation and the most similar ( neighboring ) conformations sampled in the apo MD simulation were characterized using Q values ( see Methods ) which is introduced as a quantitative measure of similarity . Table 4 summarizes the results of this analysis for the PDZ domains where there is a complex reported . It can be seen that the complex of PTP-BL PDZ2 domain with the APC peptide is the least similar to the apo MD simulation ensemble . It has the highest Q ( 1 ) and Q ( 10 ) values ( 0 . 37 Å and 0 . 39 Å ) which represent the average dRMSD dissimilarity between the ligand-bound conformer and the most similar and ten most similar stimulation snapshots , respectively . By contrast , the complex of InaD PDZ1 with the NorpA peptide has significantly lower Q ( 1 ) and Q ( 10 ) values ( 0 . 15 Å and 0 . 18 Å ) indicating that this ligand-bound binding site conformation is more closely approached in these simulations . We also performed an additional simulation of the PTP-BL PDZ2 domain in complex with the APC peptide to examine whether the presence of the peptide kept conformational space closer to the ligand-bound crystal structure . As expected , the Q ( 1 ) and Q ( 10 ) values are lower ( see Supporting Information , Text S1 ) for the simulation with the peptide bound compared to apo ( Table 4 ) , lending further support to the induced fit mechanism . Although due to sampling limitations , we are unable to tell if the apo structures get any closer to the peptide-bound conformations in reality , the data presented here suggest that , out of the five PDZ domains studied here , PTP-BL PDZ2 is the most likely to involve an induced fit mechanism when binding to the APC peptide . Figure 6D shows the mean absolute difference distance matrix ( Δ ) pattern calculated between the peptide-bound structure and the 100 most similar snapshots . We can see that the largest deviations are found in the distances between S28 and L85 and between V29 and R86 . The Δ pattern suggests that these two inter-residue distances are altered the largest extent upon binding to the APC peptide . Visual inspection of the PTP-BL PDZ2 trajectory shows some subtle rearrangements of the protein from the starting crystal structure . The movement between residues S28 and L85 along with V29 and R86 appears to be facilitated by re-arrangement of the “pre-β2” loop and the “post-α2” loop and the movement of K10 , the side-chain of which appears to act as a helix cap for the α2 helix most of the time . As these movements occur , water molecules penetrate deeper into the cleft but are then expelled as the cleft returns to conformations more similar to the starting structure . However , the whole structure appears to be further stabilized by the formation of salt-bridge between residues D22 and K50 which is not initially present in the crystal structure ( see Supporting Information , Figure S2 ) . The overall change in shape of the pocket in this extreme is similar in nature to opening of the Erbin PDZ domain ( which occurs infrequently – see Supporting Information , Figure S1 ) . The solution structure of the GRIP1 PDZ7 domain [47] suggests that the α2/β2 binding pocket adopts a “closed conformation” and has a significantly smaller carboxyl peptide-binding site than other PDZ domains which would restrict its ability to interact with peptides . However , it is the case that other PDZ domains with similar closed pockets appear to be able to open up in order to incorporate a peptide ligand such as LARG PDZ domain [40] . Thus we examined the conformational dynamics of the GRIP1 PDZ7 domain to see if it would open up to a conformation capable of peptide binding . The fluctuation and flexibility patterns of the GRIP1 PDZ7 binding pocket ( Figure 7A and B ) show that the N-terminal end of the β2-strand has notable fluctuation with regards to the C-terminal end of the α2-helix . On the other hand , the patterns also show that the C- terminal end of the β2-strand has little mobility with regards to the N-terminal end of the α2-helix . Since the bottom of the binding pocket is located between the C-terminal end of the β2-strand and the N-terminal end of the α2-helix , their low relative fluctuation suggests that the base of the binding site does not open significantly . In order to examine this in more detail , and in a comparative way to the other PDZ domains studied here , the distance between the C-terminal residue of the β2-strand and the N-terminal residue of the α2-helix was used to characterize to what extent the base part of the binding pockets in all the PDZ domains is open . Figure 7C shows these distance distributions for each PDZ domain . The distributions for Erbin PDZ and PTP-BL PDZ2 are almost identical ( both are approximately Gaussian functions with a mean of 6 . 28 Å and 6 . 4 Å , and standard deviation of 0 . 29 Å and 0 . 49 Å , respectively ) indicating that the base parts of the binding groove of these two PDZ domains behave in a very similar fashion . The distribution of the InaD PDZ1 domain , however , has larger spread ( a standard deviation of 0 . 58 Å ) , but the mean distance is about the same ( 6 . 35 Å ) as for Erbin PDZ and PTP-BL PDZ2 . Interestingly , the distance distribution of Dvl2 PDZ is a superposition of two Gaussian distributions ( with a mean of 6 . 0 Å and a standard deviation of 0 . 58 Å ) . However , the location of one of the two superposed Gaussian curves agrees well with the distributions observed for Erbin PDZ and PTP-BL PDZ2 . Most importantly , the distance distribution of GRIP1 PDZ7 ( which can be approximated well as a single Gaussian distribution ) is significantly shifted relative to the other four distributions . It has a mean of only 5 . 68 Å , and a standard deviation of 0 . 39 Å . The probability that the base part of the binding pocket is open with an extent larger than 0 . 6 Å is considerably lower in the case of the GRIP1 PDZ7 domain but is high in the four other PDZ domains . These results show that the bottom of the binding groove of the GRIP1 PDZ7 domain is closed and it remains closed in the course of the 200 ns MD simulation unlike in other PDZ binding sites . This unique property of GRIP1 PDZ7 is probably the reason why this PDZ domain has been found to be unable to bind to carboxyl peptides . The intrinsic dynamics of the binding sites of five PDZ domains have been compared in this paper , based on 200 ns all-atom molecular dynamics simulations of the apo structures . Despite the remarkable structural similarity of the five PDZ folds and binding sites , their fluctuation and flexibility properties have been found to be surprisingly different . Furthermore , the differences of their mobility correlate well with differences of their functional properties suggesting that intrinsic dynamics is an important determinant of function . The binding sites of InaD PDZ1 and Dvl2 PDZ are the most flexible of those of the five PDZ domains and this high degree of flexibility is likely to be necessary for them to be able to interact with multiple partners using significantly different binding modes , a property referred to as “strong promiscuity” . The Erbin PDZ domain , by contrast , has a rigid binding site and while it is also promiscuous , it interacts with very similar peptides using very similar binding modes . We do not count interactions with proteins at distal sites such as that reported for the Erbin-Smad3 MH2 interaction [48] which appears to be well away from the classical PDZ interaction groove . The results presented here are consistent with the proposed link between binding site flexibility and promiscuity discussed in other studies [14] , [16] . Currently there is no experimental structure available of the complex of InaD PDZ1 with the NinaC peptide . Based on the results presented in this study , we predict that InaD PDZ1 interacts with NinaC in a significantly different binding mode than it does with NorpA , a conclusion also made by Kimple et al [44] and Wes et al [43] . This hypothesis should be readily testable via structural characterization experiments . The results for PTP-BL PDZ2 have revealed that the conformational space explored by the apo protein is the most different from the APC peptide-bound conformation compared to the other PDZ-peptide complexes . These results , in accordance with experimental data , suggest that the induced fit mechanism may be crucially involved in the binding of PTP-BL PDZ2 to the APC peptide and play a larger role in the recognition mechanism compared to other PDZ domains . Overall it seems likely that conformational selection and induced fit both appear to play roles in binding of PDZ domains to their peptides . One can formulate the two mechanisms into distinct roles; Firstly , conformational selection seems to be an essential mechanism for PDZ domains to visit regions of the conformational space that are close to different ligand-bound states . Visiting these regions is probably necessary for the formation of weak ( initial ) complexes . Once a weak complex is formed , the induced fit mechanism , as a fine-tuning step , could lead to minor changes in the shape of the binding pocket stabilizing the PDZ-peptide complex . The extent to which these mechanisms are required is likely variable across the PDZ domain family . The MD simulations confirm that GRIP1 PDZ7 has a closed canonical binding site which is consequently unable to accommodate carboxyl peptides . The binding pocket does not appear to undergo a transition from its closed state to an open state in the course of the 200 ns trajectory . These results agree with the experimental observations that GRIP1 PDZ7 cannot interact with carboxyl ligands . The results highlight how one fold can exhibit quite different dynamics . For PDZ domains this issue should be borne in mind when considering structure-based drug-design [49] . Considering conformational selection in the docking strategies of virtual screening is a promising new paradigm recently reviewed by Amaro and Li [50] . Furthermore , describing binding site flexibility was suggested to be crucial for designing compounds of high selectivity for a given drug target [51] . As the dynamics of the PDZ binding pocket seems to be a key factor determining the ability to interact with different peptides , the flexibility of the binding site should also be taken into account alongside steric and electrostatic effects [52] in rational drug design . From this work we would anticipate that intrinsic dynamics would play a role in other systems ranging from influencing large domain movements through to allosteric transitions . As simulation times approach experimental timescale , particularly for NMR , it will become possible to assess how well these observations fit into solvable models for conformational selection and induced fit such as the one proposed by Zhou [53] .
All-atom 200 ns MD simulations were performed for the five apo PDZ domains summarized in Table 1 with the GROMACS software package [54] , [55] using the OPLS force field [56] in an NPT ensemble . Pressure coupling was performed using the Berendsen barostat with a time constant , tau , of 1 . 0 ps . The systems were coupled using a Berendsen heat bath [57] with a tau value of 0 . 1 ps . Electrostatics were treated with a Particle Mesh Ewald scheme with a real-space cut-off of 10 Å . The neighbour list cut-off was also set to 10 Å and was updated every 10 steps . The proteins were solvated in explicit SPC water [58] and Na+ and Cl− ions added to make up a neutral solution of 150 mM . A short steepest descents minimization of 225000 steps was performed , followed by a short restrained run of 200 ps whereby the Cα atoms of the protein were restrained by a harmonic potential with a force constant of 1000 kJmol−1 mn−2 . Snapshots from the trajectories were saved every 5 ps for analysis . Convergence was assessed via root mean square inner product ( RMSIP ) between sections of the trajectories ( see Supporting Information , Text S1 , for more details ) . Let A and B denote two proteins that consist of NA and NB residues , respectively . In this study , residues are represented by their α-carbon atoms . An alignment between the two structures defines a mapping between the two sets of residues . Let N denote the number of aligned residue pairs ( after removing positions aligned to gaps ) . The two sets of aligned residues are described by the NxN distance matrices of their α-carbon atoms denoted by dA and dB: i . e . the matrix entry is the distance of α-carbon atoms of aligned residues i and j in structure A . The difference distance matrix δ between structure A and B is defined as: ( 1 ) Positive entries in this matrix indicate pairs of atoms of larger distance in structure A than in structure B . This matrix can be used to characterize the location and extent of structural differences between two different proteins or two conformations of the same protein . The dRMSD ( distance root mean square deviation ) measure of dissimilarity between the two structures is defined as: ( 2 ) We use this measure instead of the standard RMSD dissimilarity because dRMSD is not dependent on structural superposition . Let S = {S1 , S2 , … , SK} denote an ensemble of conformations of a protein represented by its α-carbon atoms . Let the number of its residues be N . We define an NxN matrix as the F fluctuation matrix , which describes the extent of the pairwise fluctuation of α-carbon atoms . Matrix F contains the variances of the distance of each α-carbon pair , where the variance is calculated over the whole ensemble . It is precisely defined as: ( 3 ) where is the mean distance of α-carbon atoms i and j in the ensemble . We have previously described the use of a similar matrix where standard deviation rather than variance of the distances was used [59] . Although variance describes the spread of a distance distribution characterizing the relative fluctuation of two atoms , it is not always informative about how much the distance between two atoms can change . Even if the distance of two atoms significantly deviates from their mean distance in some conformations , the variance may still be low provided that most of the variation is around the mean . To measure the pairwise flexibility of two atoms ( i . e . the maximal difference of their distance in the ensemble ) , the flexibility matrix denoted as X is introduced . Matrix X describes the range of distance distribution for each pair of atoms: ( 4 ) Note that the above definitions of F and X matrices allow that two pairs of atoms that have equal pairwise fluctuation can have considerably different pairwise flexibility . While the F matrix contains pairwise atomic fluctuation values , a measure of the overall fluctuation of the whole structure ( or a subset of residues ) was also introduced . This overall fluctuation measure denoted by Θ was defined as the root mean square of dRMSD dissimilarity of each structure with regards the mean distance matrix calculated for the whole S ensemble . In other words , Θ is a measure for the size of conformational space the protein explores in the ensemble . It is easy to see that the above definition is equivalent to the root mean of the entries of F fluctuation matrix calculated for the same conformational ensemble . The precise definition of overall fluctuation is therefore ( 5 ) where is the mean distance matrix of the ensemble . Equivalent binding site residues were defined on the basis of a multiple sequence alignment ( MSA ) ( Figure 1B ) . The binding groove of PDZ domains is located between the β2 strand and the α2 helix . Two sequence regions were therefore selected in the MSA that correspond to the conserved structural elements of the β2 strand and α2 helix ( or α1 helix , in Erbin PDZ ) . The binding sites were characterized by 5×10 submatrices of the δ , F and X matrices describing the relative structural difference , fluctuation and flexibility of the α-helix and the β-strand . MD simulation trajectory snapshots were clustered with k-mean cluster analysis , a simple unsupervised learning algorithm [60] , [61] . The method can be used for partitioning N data points ( here , protein conformations ) into k disjoint subsets ( or clusters ) denoted by C1 , C2 , … , Ck . The parameter k is fixed a priori . The goal of the algorithm is to find the optimal partitioning of conformations to minimize the within-cluster sum of squares ( WCSS ) : ( 6 ) where the dRMSD measure is used to capture the similarity of conformations and Ci is the mean distance matrix of cluster i . Since k is an arbitrary parameter , the goodness of clustering results was estimated using the Silhouette Index cluster validity measure ( see below ) [62] . The optimal k-value that provided the highest overall average Silhouette Index was selected . Once the conformational ensemble is clustered , the following Silhouette Index measure is calculated for each conformation: ( 7 ) where a ( i ) is the average dRMSD dissimilarity of conformation i to all other conformations in the same cluster and b ( i ) is the minimum of average dRMSD dissimilarities of conformation i to all other clusters . The silhouette index is between −1 and 1: if S ( i ) it is close to 1 , it means , the conformation is well-clustered; if S ( i ) is close to 0 , it means the conformation could be assigned to another cluster as well; if S ( i ) is close to −1 , it means the conformation is misclassified . The goodness of clustering result was then measured by the overall average silhouette index SOVER which is simply the average of S ( i ) for all conformations in the ensemble: ( 8 ) Multidimensional scaling ( MDS ) ( also known as Principal Coordinates Analysis ) is a dimensionality reduction method often used to visualize high-dimensional data on a two-dimensional map [63] . The input of the method is a dissimilarity matrix that contains distances ( dissimilarities ) between pairs of objects calculated in a high-dimensional space . The output is a configuration of points embedded into lower ( ideally , two or three ) -dimensions . In Classical MDS ( CMDS ) ( also referred to as Torgerson-Gower scaling ) [64] used in this study , the goal is that the Euclidean distances between the outputted points should approximately reproduce the original dissimilarity matrix . In order to study the difference between induced fit and conformational selection binding , a simple definition is introduced to measure how similar conformations are sampled in an apo simulation to a given experimental ligand-bound structure . Let S ( k ) denote the set of k most similar conformations ( neighboring conformers ) with regards to a reference experimental structure E ( ranked based on the dRMSD dissimilarity measure ) . The following Q ( k ) value is defined as the average dRMSD dissimilarity of conformations in S ( k ) with regards to structure E: ( 9 ) In this study the quantities Q ( 1 ) , Q ( 10 ) , Q ( 100 ) and Q ( 200 ) were used to characterize the similarity of the most similar , 10 most similar , 100 most similar and 200 most similar conformations to an experimental ligand-bound structure of interest . | Proteins that are capable of binding to many different ligands are said to have broad specificity . This is sometimes also referred to as promiscuity . Whether a protein is promiscuous or not can sometimes be readily explained by the structure of the protein and the ligand in terms of electrostatic and steric effects . Sometimes however , this simple interpretation can struggle to explain the experimentally observed data . A prominent case in point is the PDZ domains . These small protein domains bind to unstructured regions of other proteins and are involved in many signaling pathways . Some PDZ domains appear to be more promiscuous than others , but this has been difficult to explain purely on the basis of the composition of residues in the binding groove . In this work we examine the dynamics and conformational flexibility of five key PDZ domains and demonstrate that despite similar folds , these proteins can exhibit quite different dynamics . Furthermore the difference in the dynamic behavior appears to correlate with the observed promiscuity . Our findings suggest that knowledge of the dynamic behavior of the PDZs can be used to rationalize the extent of expected promiscuity . Such knowledge will be critical for drug design against PDZ domains . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"biomacromolecule-ligand",
"interactions",
"protein",
"interactions",
"molecular",
"dynamics",
"chemical",
"biology",
"protein",
"structure",
"biophysics",
"simulations",
"biochemistry",
"simulations",
"proteins",
"chemistry",
"biology",
"biophysics",
"drug",
"discovery",
"bi... | 2012 | The Role of Flexibility and Conformational Selection in the Binding Promiscuity of PDZ Domains |
Pattern formation in developing tissues is driven by the interaction of extrinsic signals with intrinsic transcriptional networks that together establish spatially and temporally restricted profiles of gene expression . How this process is orchestrated at the molecular level by genomic cis-regulatory modules is one of the central questions in developmental biology . Here we have addressed this by analysing the regulation of Pax3 expression in the context of the developing spinal cord . Pax3 is induced early during neural development in progenitors of the dorsal spinal cord and is maintained as pattern is subsequently elaborated , resulting in the segregation of the tissue into dorsal and ventral subdivisions . We used a combination of comparative genomics and transgenic assays to define and dissect several functional cis-regulatory modules associated with the Pax3 locus . We provide evidence that the coordinated activity of two modules establishes and refines Pax3 expression during neural tube development . Mutational analyses of the initiating element revealed that in addition to Wnt signaling , Nkx family homeodomain repressors restrict Pax3 transcription to the presumptive dorsal neural tube . Subsequently , a second module mediates direct positive autoregulation and feedback to maintain Pax3 expression . Together , these data indicate a mechanism by which transient external signals are converted into a sustained expression domain by the activities of distinct regulatory elements . This transcriptional logic differs from the cross-repression that is responsible for the spatiotemporal patterns of gene expression in the ventral neural tube , suggesting that a variety of circuits are deployed within the neural tube regulatory network to establish and elaborate pattern formation .
Embryonic development relies on the coordinated and dynamic control of gene expression . This is achieved , in the main , by interactions between transcription factors ( TFs ) and the genomic cis-regulatory modules ( CRMs ) associated with regulated genes [1] , [2] . The aggregate of these interactions produces a gene regulatory network ( GRN ) that is responsible for imparting distinct molecular identities and consequently the pattern of cell fate in a tissue . Within these large GRNs , sub-circuits can be discerned that confer specific behaviors and responses [1] , [3] . Thus , elucidating functional interactions between TFs and CRMs provides insight into the mechanism and regulatory logic of the transcriptional networks responsible for tissue patterning . The specification of progenitor identity in the vertebrate neural tube is a well studied example of developmental patterning [4] . Motor neurons and several classes of associated interneurons are generated in ventral regions of the neural tube in response to the morphogen Sonic hedgehog ( Shh ) . Secretion of Shh from the notochord and floor plate establishes a gradient of intracellular signaling activity that regulates the expression of TFs specifying the ventral progenitor domains [5]–[8] . Key phylogenetically conserved CRMs associated with many of these TFs have been identified and shown to integrate the activity of Shh signaling with general neural TFs and Shh regulated TFs [9] , [10] . Within this network selective cross-repressive interactions between TFs operating downstream of Shh signaling appear critical , both to establish and maintain the distinct spatial domains of progenitors [5] , [6] , [11]–[14] . By contrast , less is known regarding the specification of sensory interneurons within the dorsal spinal cord [15] . One key TF involved in this process is the paired homeodomain protein Pax3 , which is amongst the first to delineate the dorsal neural tube and then later , together with its paralog Pax7 , identifies the 6 progenitor domains that comprise dorsal progenitors [16] , [17] . Both bone morphogenetic protein and Wnt signaling have been implicated in the induction of Pax3 transcription and the establishment of dorsal progenitors [18] , [19] . Conversely , Shh mediated repression of Pax3 has been suggested to eliminate expression in the ventral neural tube 20 , 21 . Several studies indicate that a genomic interval immediately upstream of the mouse Pax3 promoter is sufficient to direct expression to the neural tube , however this region is not required for Pax3 expression [22]–[25] . A further two CRMs have been identified within the 4th intron that recapitulate elements of Pax3 expression in the central nervous system ( CNS ) [22] , [26] . The activity of one region , termed ECR2 [22] or IR1 [26] , is dependent on Tcf/Lef binding sites , consistent with a role for Wnt signaling in the initiation of Pax3 transcription . Nevertheless , how the spatial domain of Pax3 expression is determined and maintained during the elaboration of neural tube patterning has not been explained . Here we take advantage of lineage tracing analyses in mice and transgenic assays in chick and zebrafish embryos to dissect the molecular mechanism and regulatory logic of Pax3 expression in dorsal neural progenitors . We show that Pax3 expression is refined during neural tube patterning by the temporal activity of distinct regulatory elements . We provide evidence that in addition to Wnt signaling , Nkx family homeodomain ( HD ) containing repressors are critical for establishing the restricted expression of Pax3 . Moreover , we demonstrate that autoregulation and positive feedback is required to maintain Pax3 expression in the neural tube .
The establishment of the Pax3 expression domain in the neural tube distinguishes the progenitors of sensory neurons from those fated to give rise to motor neurons and associated ventral interneurons , however the cellular and molecular mechanisms that regulate this key patterning event remain poorly understood . In order to gain insight into this process , we employed a lineage tracing approach to assay the spatiotemporal dynamics of Pax3 expression in the neural tube . Transgenic mice in which Cre recombinase was targeted to the first exon of the Pax3 locus [27] were crossed with either Rosa26-YFP or Rosa26-Tomato/GFP reporter strains ( Figure 1A–E′ and data not shown ) . The resulting embryos were analysed between embryonic days ( E ) 8 . 5 and E11 . 75 in transverse sections . From E8 . 5 to E9 . 5 , all cells marked by transgene expression also express Pax3 , demonstrating that this transgenic line accurately reports the Pax3 lineage ( Figure 1A , A′ and data not shown ) . At these early stages the Pax3 expression domain is not well defined and isolated cells expressing both GFP and Pax3 can be detected within the intermediate region of neural tube ( arrows in Figure 1A′ ) . From E9 . 5 onwards transgene labelled cells were observed beyond the ventral boundary of Pax3 , indicating that the position of the Pax3 domain was refined during early stages of neural tube patterning ( Figure 1B , B′ and C ) . Two distinct populations of transgene labelled progenitors that no longer express Pax3 protein were present within the ventral neural tube of each embryo . The first comprised isolated clusters of cells ( asterisks in Figure 1B ) . The dispersal of these cells within the ventral neural tube was highly variable , both between stage-matched siblings and along the anterior-posterior ( AP ) axis of individual embryos . We attributed this to the induction of Pax3 transcription within the neural plate and cell mixing within the neuroepithelium at early stages of development [28]–[30] . By contrast , the second population of transgene expressing cells was a continuous domain that spanned 3–4 cell diameters adjacent to the ventral boundary of Pax3 ( Figure 1B′ and arrows in B and C ) . The extent of transgene expression encompassed , but was not limited to , the Evx1 expression domain across the AP axis of embryos assessed at E11 . 75 ( Figure 1D ) . These data indicated that cells fated to become ventral interneurons extinguish Pax3 expression during early CNS patterning [31] . In agreement with this observation , the Pax3 lineage apposed Nkx6 . 1 expression at E9 . 75 , which labels the 3 most ventral progenitor domains of the neural tube ( Figure 1E and E′ ) . Together , these data demonstrated that the initial domain of Pax3 expression was refined by a switch in progenitor identity and subsequently maintained at this DV position . We next sought to investigate the molecular basis of Pax3 expression during CNS development by employing comparative genomics to identify functional CRMs associated with the gene . Conservation of the Pax3 locus and the surrounding intergenic regions across the human , mouse , zebrafish and fugu genomes revealed 6 conserved non-coding elements ( CNEs ) , all of which were located within the 4th intron of the gene ( Figure 1F and Figure S1 ) . Candidate CRMs were assayed in zebrafish embryos , which have been previously shown to exhibit similar spatio-temporal profiles of pax3a mRNA expression to that observed in chick and mouse [16] , [20] , [32] . Accordingly , pax3a expression was first observed in the posterior neural plate at 10 hours post fertilisation ( hpf ) ( Figure 1G ) before becoming restricted to the lateral limit of the tissue at 12 hpf ( Figure 1H ) . Maximum expression within the dorsal neural tube was observed at 24 hpf ( Figure 1I and I′ ) , after which transcription rapidly decreased to undetectable levels in the spinal cord by 48 hpf ( Figure 1J and data not shown ) . Pax3 and Pax7 ( Pax3/7 ) proteins were expressed in the lateral regions of the posterior neural plate at 10 hpf ( Figure 1K ) and progenitors within the dorsal spinal cord at 24 hpf ( Figure 1L ) , in agreement with pax3a transcription . Strikingly , each CNE assayed in transient transgenic zebrafish exhibited a tissue specific enhancer activity at 24 hpf , the majority of which recapitulated elements of pax3a expression ( Figure S1 ) . However , only CNE1 and CNE3 reproducibly labelled the developing CNS ( Figure 1M–R′ ) . The genomic interval corresponding to CNE3 is a highly conserved region of a previously defined CNS specific Pax3 CRM , termed both ECR2 [22] and IR1 [26] . CNE3 transgenic embryos assessed between 10 and 24 hpf exhibited reporter expression in the posterior neural plate and neural rod between 10 and 18 hpf ( Figure 1P , Q and data not shown ) . The developing midbrain and hindbrain were labeled at 24 hpf , however transgene expression within dorsal spinal cord progenitors was markedly reduced by this stage ( Figure 1R , R′ , 2E , E′ and Figure S1 ) . Together , these data suggested that the profile of CNE3 activity correlated with the induction of Pax3 transcription and the establishment of this expression domain . However , the down regulation of CNE3 activity in spinal cord progenitors indicated that a separate CRM was required to maintain high levels of Pax3 at later developmental stages . In addition to CNE3 , our functional assays identified CNE1 as a CNS specific Pax3 enhancer . CNE1 transient transgenic zebrafish exhibited Citrine expression in presumptive dorsal progenitors at 10 hpf ( Figure 1M ) and 12 hpf ( Figure 1N ) . Robust labeling of the midbrain , hindbrain and dorsal spinal cord progenitors was observed at 24 hpf ( Figure 1O , O′ and Figure S1 ) . These data were supported by the creation of 3 independent CNE1 stable lines , which revealed that the activity of this CRM recapitulated Pax3 expression during the first 24 hours of CNS development ( Figure S2A–B′ ) . Moreover , CNE1 activity was restricted to the Pax3/7 domain of the zebrafish neural tube at 24 hpf ( Figure 1O′ and Figure S2B′ ) . These findings were consistent with reports documenting the activity of an approximately 2 . 5 kilobase genomic interval containing this enhancer , named IR2 , in zebrafish embryos [26] . These data suggested that CNE1 might act in concert with CNE3 , initially to define the Pax3 domain in the neural plate and later function independently to maintain expression in the neural tube . This hypothesis was supported by examining the binding profile of the transcriptional coactivator p300 in E11 . 5 mouse tissue , which suggested that CNE1 was the only active enhancer within the Pax3 locus at this comparatively late stage of CNS patterning ( Figure S3 ) [33] . We sought to investigate the molecular basis of CNE3 activity in order to gain insight into the regulation of Pax3 expression in the CNS . The conservation of sequence across 12 vertebrate genomes was used to define 5 statistically enriched 15 bp motifs , predicted to contain the transcription factor binding sites ( TFBS ) that mediate enhancer activity ( Table S1 and Figure S4 ) . Annotation of TFBS within Motif4 and Motif5 of CNE3 suggested that these regions contained highly conserved Tcf/Lef binding sites ( Table S1 ) , consistent with reports demonstrating the requirement of Tcf/Lef sites within this CRM [22] , [26] . Furthermore , the Wnt pathway effector Tcf3 has been shown to bind this enhancer in ChIP-Seq experiments performed in mouse embryonic stem cells [34] ( Figure S5 ) . These data suggested that CNE3 received positive transcriptional input from the Wnt pathway , which has been shown to be both necessary and sufficient for the induction of Pax3 transcription [18] . We used TCFSiam transgenic zebrafish [35] to assay the transcriptional activity of the Wnt pathway at 10 hpf , the stage at which CNE3 activity and Pax3 expression was first detected ( Figure 2A , A′ ) . This revealed activated Wnt signaling throughout the medio-lateral axis of the posterior neural plate ( Figure 2A and A′ ) , whilst Pax3 expression was restricted laterally ( Figure 2A′ ) . This profile of activated Wnt signaling was consistent with that described in the neural plate of independent zebrafish [36] and mouse transgenic reporters [37]–[39] . These data supported the described role of the Wnt pathway in the initiation of Pax3 transcription , but also suggested that it could not provide sufficient positional information to establish the domain of Pax3 expression in the neural plate . TCFSiam transgenic embryos assessed between 12 hpf and 18 hpf exhibited activated Wnt signaling within the tailbud and neural tube , however reporter expression was markedly decreased in the spinal cord of 24 hpf embryos ( Figure 2B , C and data not shown ) . This temporal profile of activated Wnt signaling correlated with the activity of CNE3 during embryogenesis ( Figure 1P–R′ and data not shown ) , supporting a positive transcriptional role for Wnt pathway effectors upon CNE3 . We next searched for conserved TFBS that could facilitate the binding of putative repressors to CNE3 , which might act to prohibit Pax3 transcription in the presumptive ventral neural tube . We focused upon Motif3 and Motif5 , as these matrices exhibited homology to sites bound by HD and Fox family transcription factors ( Figure 2D and Table S1 ) . We were particularly intrigued by the identification of highly conserved Nkx binding sites within CNE3 as members of this gene family are key fate determinants within the ventral neural tube , functioning as repressors via recruitment of Groucho/Tle proteins [40] . We deleted either Motif3 or Motif5 from CNE3 and assayed enhancer activity in chick and zebrafish . Deletion of Motif3 resulted in an increase in the number of progenitors with reporter activity within the zebrafish spinal cord , however this effect was not observed in chick embryos ( data not shown ) . By contrast , deletion of Motif5 resulted in a significant increase in CNE3 activity along the entire D–V axis of the zebrafish spinal cord ( compare Figure 2G , G′ to Figure 2E , E′ ) . Ectopic activation of the zebrafish enhancer sequence was also seen within the intermediate neural tube of chick embryos , in a domain adjacent to the Pax3 ventral boundary ( Figure 2H ) , compared to the wildtype sequence ( Figure 2F ) . Targeted substitutions were then engineered into CNE3 that mutated the HD binding site within Motif5 ( Figure 2I ) . This mutation resulted in an increase of CNE3 activity in neural progenitors within the zebrafish spinal cord at 24 hpf ( Figure 2J , J′ ) and the ectopic activation of CNE3 in the chick neural tube at E3 ( Figure 2K ) , recapitulating the effect of Motif5 deletion . To investigate the binding of putative repressors to CNE3 , we performed electrophoretic mobility shift assays ( EMSAs ) using nuclear extracts from dissected E3 chick spinal cords and a 48 bp DNA probe which spanned Motif3 , 4 and 5 . Supershift reactions were performed by the addition of antibodies raised against β-catenin , several candidate HD containing ventral fate determinants and FoxA2 ( Figure 2L and data not shown ) . Assays containing labeled probe and nuclear extracts led to the formation of 3 distinct complexes , compared to reactions in which the nuclear extract was omitted ( compare the Control and Probe lanes in Figure 2L ) . Addition of Nkx6 . 2 , Dbx1 , Dbx2 and Pax6 antibodies resulted in some reduction in the motility of the second complex ( Figure 2L and data not shown ) , however this effect was substantially more pronounced in reactions containing Nkx6 . 1 antibody . Addition of β-catenin antibody did not affect the motility of complexes . Furthermore , the addition of an antibody raised against FoxA2 completely blocked the formation of the 3 DNA/protein complexes ( Figure 2L ) . These assays indicated that both Nkx6 . 1 and FoxA2 were present within DNA bound complexes , consistent with the annotation of conserved binding sites within the motifs that comprise this region ( Figure 2D and Table S1 ) . Together , these data suggested that Nkx6 . 1 , or a protein with a similar binding specificity , interacted with Motif5 of CNE3 to mediate transcriptional repression of Pax3 within the developing spinal cord . This is consistent with the observation that the ventral domain of the Pax3 lineage was mutually exclusive with Nkx6 . 1 expression in transgenic mice assessed at E9 . 75 ( Figure 1E and E′ ) . Furthermore , overexpression of Nkx6 . 1 in chick embryos repressed endogenous Pax3 protein expression in the dorsal neural tube at E3 ( Figure 2M and M′ ) . Thus , these data support a model in which CNE3 functions to establish the Pax3 domain in the posterior neural plate by integrating a positive input from the Wnt pathway and Nkx family transcriptional repressors . We next sought to investigate the molecular basis of CNE1 activity by identifying statistically over represented conserved 15 bp motifs within this ∼170 bp enhancer ( Figure 3A , Table S2 and Figure S6 ) . Control experiments demonstrated that CNE1 activity was restricted to the Pax3/7 domain of the dorsal spinal cord in both zebrafish embryos at 24 hpf ( Figure 3B , B′ ) and chick at E3 ( Figure 3C ) , consistent with Pax3 transcription at these developmental stages . Deletion assays revealed that Motif1 ( Figure 3D ) was specifically required for CNE1 mediated transcription in the spinal cord of both zebrafish ( Figure 3E , E′ ) and chick embryos ( Figure 3F ) . By contrast , deletion of Motif3 ( Figure 3J ) reduced transgene expression throughout the AP axis of the zebrafish CNS ( Figure 3K , K′ ) and extinguished enhancer activity in chick spinal cord ( Figure 3L ) . Motif2 ( Figure 3G ) and Motif4 ( Figure 3M ) appeared to be dispensable in the context of these experiments ( Figs . 3H–I and N–O ) . Amongst the TFs that potentially interact with Motif3 ( Table S2 ) , the SoxB family represented the best candidates to promote the activity of CNE1 across the AP axis of the CNS . In agreement with several recent studies demonstrating the essential role of this family of TFs to promote enhancer activity in neural lineages [9] , [10] , point mutations within the putative HMG binding site of Motif3 reduced CNE1 activity in zebrafish and chick embryos ( Figure S7 ) . Moreover , examination of ChIP-Seq datasets produced in stem cell derived neuronal progenitors demonstrated that Sox3 and Sox11 directly bind CNE1 , supporting a general positive input of SoxB proteins upon Pax3 expression in the CNS ( data not shown ) [41] . However , the input of this family of transcription factors is unlikely to explain the spatial restriction of CNE1 activity to the dorsal neural tube . Examination of the matrix represented by Motif1 revealed a 14 bp sequence similar to a Pax6 paired domain ( PD ) binding site ( Table S2 ) . We were particularly intrigued by this , as members of the Pax gene family have been shown to participate in selective auto- and inter-regulatory interactions [42] . Comparison of Motif1 consensus sequence with the defined Pax6 [43] , Pax5 [44] and paired [45] PD binding sites revealed a high degree of homology towards the 5′ of the alignment , but poor consensus in the 3′ region ( Figure 4A ) . Given these data and the described requirement for Motif1 in the dorsal spinal cord , we hypothesized this region could represent a PD binding site that exhibited specificity for Pax3 and its paralog Pax7 . We assessed the ability of both Pax3 and Pax7 to bind Motif1 by EMSA , using a 33 bp DNA probe and in vitro synthesized proteins . These assays revealed that both TFs interacted with this sequence , verifying it as a functional PD binding site ( Figure 4B ) Furthermore , supershift EMSAs using nuclear extracts from chick spinal cord and antibodies raised against selected Pax family members demonstrated the preferential occupation of Motif1 by the PD coded by Pax3/7 class genes ( Figure 4C ) . Previous binding and structural studies have shown that the PD is comprised of two helix-turn-helix subdomains , commonly termed PAI and RED , that interact with the 5′ and 3′ region of the binding site , respectively 46–48 . The sequence of the PAI domain is largely conserved across the Pax gene family , whereas the variant RED domains have been proposed to underlie target site specificity [48]–[50] . We assessed the requirement of PAI and RED domain mediated Pax3/7 binding to Motif1 for CNE1 activity by targeting mutations within either half site , guided by the degree of conservation across this matrix ( Figure 4D ) . A 5 bp substitution within the putative PAI half site resulted in a loss of CNE1 activity in the spinal cord progenitors of both zebrafish ( Figure 4E , E′ ) and chick embryos ( Figure 4F ) . Similarly , a 2 bp substitution within the putative RED domain half site extinguished the activity of this CRM in the dorsal neural tube of both model organisms ( Figure 4G–H ) . In agreement with these in vivo observations , recombinant Pax3 protein was unable to bind Motif1 DNA probes carrying either the PAI or RED mutations in vitro ( Figure 4I ) . Furthermore , competition EMSAs revealed the reduced ability of the RED mutant sequence to compete for Pax7 binding versus a wildtype probe ( Figure 4J ) . We next sought to assess the ability of Pax3 and Pax7 to induce CNE1 activity in the neural tube . Electroporation of a dominant active protein consisting of the DNA binding domain of human PAX3 fused to the transactivation domain of FOXO1A ( PAX3FOXO1A ) was sufficient to activate CNE1 mediated transcription in the ventral neural tube of chick embryos at E3 ( Figure 4K ) . Moreover , PAX3FOXO1A was able to induce ectopic Pax3 protein expression ( Figure 4K ) . Similarly , Pax7 electroporation was sufficient to induce ectopic Pax3 expression and CNE1 mediated transcription ( Figure 4L ) . By contrast , electroporation of dominant-negative isoforms , constructed by fusion of the engrailed repressor domain to either Pax3 or Pax7 [51] , reduced Pax3 expression within its endogenous domain at E3 ( Figure 4M , N ) . Taken together , these findings support a model in which CNE1 functions to maintain Pax3 expression in the spinal cord by facilitating PD mediated autoregulation and positive feedback via Motif1 .
In this study we provide evidence that the dynamic expression profile of Pax3 within the developing neural tube is achieved by the coordinated action of two distinct regulatory mechanisms , acting through separate CRMs . The combined activity of these enhancers converts transient inductive cues into a sustained domain of gene expression . CNE3 integrates inductive Wnt signaling and the repressive activity of HD transcription factors to initiate expression of Pax3 within the prospective dorsal neural tube . Subsequently , an autoregulatory loop acting via CNE1 is established that maintains Pax3 expression in the absence of activated Wnt signaling and HD mediated repression . Previously , the zebrafish Pax3-GFPI150 BAC stable line has been shown to recapitulate the expression profile of pax3a in the spinal cord [52] . CNE1 and CNE3 are the only conserved enhancers contained within this genomic interval that exhibit specificity for the developing neural tube , suggesting that they are sufficient to induce and maintain Pax3 expression in this tissue . It is notable that the Pax3 locus contains an additional CRM located upstream of the promoter in higher vertebrate genomes that directs activity in the neural tube . However , this element is not phylogenetically conserved and is not required for gene expression in mice [22]–[25] . Thus , these data suggest that CNE1 and CNE3 represent the core regulatory circuit governing Pax3 expression in the CNS that has subsequently been further elaborated during vertebrate evolution . The initiation of Pax3 transcription in response to Wnt signaling appears to be mediated through CNE3 , a small highly conserved region of the genomic interval that has previously been identified as ECR2 and IR1 [22] , [26] . This conclusion is supported by the observations that Tcf3 is bound to CNE3 in mouse embryonic stem cells and the requirement for Tcf/Lef sites for enhancer activity in zebrafish embryos [22] , [26] , [34] . However , the wide distribution of activated Wnt signaling within the posterior neural plate is inconsistent with the spatial restriction of Pax3 induction ( Figure 2A′ ) [36]–[39] . Our analysis of CNE3 provides evidence that repression by HD proteins is essential to restrict the induction of Pax3 to the prospective dorsal neural tube . Nkx6 . 1 is able to directly bind CNE3 and mutation of a conserved HD binding motif results in ectopic enhancer activity . Moreover , the dorsal limit of Nkx6 . 1 expression coincides with the ventral limit of cells derived from the Pax3 lineage and gain-of-function experiments indicate that Nkx6 . 1 is sufficient to repress endogenous Pax3 protein expression . It should be noted that Nkx6 . 2 exhibits similar binding specificity to Nkx6 . 1 [53] , [54] and is also expressed in the intermediate region of the neural plate and latterly the neural tube [14] , [55] . Thus , a combination of these Nkx class repressors is likely to contribute to the establishment of the Pax3 expression domain during early CNS patterning . The transcriptional activity of the Wnt pathway decreases in the neural tube as development progresses , as indicated by a downregulation of Wnt reporter transgene expression prior to the peak of Pax3 transcription in progenitors [36] , [38] , [39] , [56] . Consistent with this , the activity of CNE3 diminished within the spinal cord over time . These data suggested that a separate enhancer is responsible for maintaining Pax3 expression in the neural tube and our analysis indicates that CNE1 is likely to fulfill this role . In support of this , CNE1 is the only Pax3 enhancer associated with p300 binding in CNS derived tissues at E11 . 5 ( Figure S5 ) [33] and functional assays indicated that CNE1 remains active in the zebrafish and chick spinal cord after CNE3 activity has decreased . Motif3 within CNE1 , comprising Fox and Sox TFBS , is required for the activity of this enhancer across the AP axis of the CNS . The reduction of CNE1 activity in constructs carrying mutations in the HMG box binding site of Motif3 , together with the identification of Sox11 and Sox3 binding at CNE1 in neural progenitors , supports a positive role for SoxB proteins on Pax3 expression and might account for the neural specificity of this enhancer [9] , [10] , [41] . More importantly , we provide evidence that direct autoregulation and positive feedback via a PD binding site is necessary for CNE1 activity in the spinal cord and that Pax3/7 bind this site in vitro and in neural cells . Furthermore , the activity of CNE1 and endogenous Pax3 expression is altered by misexpression or blockade of Pax3/7 . Thus PD mediated autoregulation and positive feedback via CNE1 is likely to explain the maintenance of Pax3 expression in the neural tube . Together , these data provide a molecular framework that describes the regulatory logic of Pax3 expression in the developing spinal cord . Expression is initiated by Wnt induction acting through CNE3 , however this induction is limited to prospective dorsal regions of neural tissue by the activity of ventrally induced Nkx class proteins ( Figure 5A ) . Pax3 protein expression then triggers a neural specific autoregulatory loop acting through CNE1 that secures transcription and removes the requirement for continued Wnt signaling and Nkx mediated ventral repression ( Figure 5B and C ) . In addition , the induction of Pax7 expression at later developmental stages provides a means to augment and reinforce this maintenance loop ( Figure 5C ) . Moreover , SoxB family proteins via a HMG box binding site in Motif3 ( Figure 5 B and C ) may limit CNE1 activity to neural tissue . It is notable that at early developmental time points , both CNE1 and CNE3 appear to be active in Pax3 expressing cells . We propose they act in a cooperative manner to establish the Pax3 expression domain , thereby offering increased robustness and precision . For example , CNE1 mediated autoregulation could function not only to increase output from the Pax3 promoter but may also buffer fluctuations in CNE3 mediated transcription . This mechanism is reminiscent of the increased robustness and precision of gene expression achieved by the synergistic activity of multiple CRMs during Drosophila development [57]–[60] . In agreement with this hypothesis , a construct harboring genomic intervals containing both CNE1 and CNE3 has been shown to be more resistant to signaling pathway manipulation that either enhancer alone [26] . The acquisition of unique molecular identities within defined progenitor populations requires the translation of transient inductive cues into discrete expression domains . In the ventral neural tube this process is achieved using a transcriptional circuit involving cross-repression between TFs downstream of the ventral morphogen , Shh . This mechanism functions both to establish and maintain expression domains , as well as confer robustness to fluctuating levels of intracellular signaling [5] . In the case of Pax3 , repression is used in combination with the inductive cue to establish the expression domain , but in contrast to the ventral neural tube , a direct autoregulatory loop plays a major role in maintaining expression . These two distinct functions are segregated between separate genomic elements . Thus , our findings expand the motifs employed within the neural tube GRN and highlight how differing regulatory mechanisms are manifest in the genome .
The genomic interval representing the Pax3 locus in each species was defined in the UCSC genome browser ( http://www . genome . ucsc . edu/ ) and uploaded to the Mulan alignment suite [61] , with appropriate repeat masking . Regions conserved with at least 65% identity over at least 40 bases in each input genome were selected for further study . The oligonucleotides listed in Table S3 were used to amplify CNEs from freshly prepared DKEY-20F20 BAC DNA ( Genbank accession BX085193 ) . Amplicons were subsequently inserted between the HindIII and SbfI sites upstream of a minimal thymidine kinase promoter in the MiniTol2 TKProm Gal4-5×UAS Citrine expression vector ( Genbank accession KF545600 ) , which exhibits no specific activity in control experiments ( data not shown ) . The Human sequence representing each CNE was used as the query sequence for cross-species BLAST searches , performed within the Ensembl genome browser ( http://www . ensembl . org/ ) . The alignment of enhancer sequences across vertebrate genomes was produced and analysed using ClustalW2 [62] and phylogenetically conserved motifs were defined using MEME [63] . TFBS within each motif were annotated using the TomTom tool of the MEME suite [64] . Targeted mutation of motifs and putative TFBS was performed using the Quickchange II XL site-directed mutagenesis kit ( Stratagene ) and the oligonucleotides listed in Table S3 . Zebrafish embryos were collected within 15 minutes of laying according to established procedures and injected with injected with a mixture of plasmid DNA ( 20 ng/µl ) and Tol2 transposase mRNA ( 14 ng/µl ) . Embryos were maintained at 28 . 5°C and sorted on the basis of Citrine expression before fixation with 4% paraformaldehyde ( PFA ) at the desired stage . Chick assays were performed in Hamburger and Hamilton stage 11–13 embryos [65] by electroporation of reporter plasmid ( 500 ng/µl ) , Tol2 mRNA ( 14 ng/µl ) and pCAGGS LacZ ( 1 . 5 µg/µl ) , according to described protocols [6] . DNA constructs contained within either the pCAGGS or RCAS expression vectors were electroporated at a concentration of 1 . 5–4 µg/µl . The PAX3FOXO1A-RCAS plasmid was created by subcloning the insert from the pCAGGS vector [66] using the oligonucleotides listed in Table S3 . Embryos were maintained at 37°C and fixed with 4% PFA at the appropriate stage . Lineage tracing studies were performed by crossing Pax3Cre transgenic mice [27] with reporter strains that expressed YFP or both Tomato and GFP from the Rosa26 locus [67] , the resulting embryos were fixed at the desired stage in 4% PFA . Wholemount in situ hybridizations for pax3a ( gift from Simon Hughes ) were performed as described [68] before 14 µm transverse sections were prepared , when required . Analysis was carried out using a Zeiss Axiophot2 and Adobe Photoshop CS3 . Antibody stainings of transverse sections of chick and mouse embryos were performed as previously described [6] , [11] . The antibodies used for immunohistochemistry were rabbit anti-β-Galactosidase ( ABD Serotec ) , mouse anti-Evx1 ( DSHB ) , rabbit anti-GFP ( Molecular Probes ) , sheep anti-GFP ( Biogenesis ) , mouse anti-Nkx6 . 1 ( DSHB ) , mouse anti-Pax3 ( DSHB ) , and mouse anti-Pax7 ( DSHB ) . Zebrafish Pax3/7 protein was visualized with DP312 ( Gift from Nipam Patel ) , which was raised against the conserved homeodomain of paired [69] . DP312 has previously been shown to recognize both Pax3 and Pax7 in zebrafish embryos [70] , [71] . Wholemount images were acquired using a Leica M205FA stereo-microscope and transverse sections were imaged using a Leica TCS SP2 confocal microscope . All images were processed with Adobe Photoshop CS3 . Zebrafish experiments were analysed by determining the number of embryos labeled by Citrine expression in spinal cord progenitors as a proportion of the total transgenic population . The activity of enhancer constructs in the chick neural tube was assayed against the presence of β-Galactosidase antibody staining , which was used as an internal control of transgenesis . The statistical significance of transgenic assays was determined using two-tailed Fisher's exact tests . Radiolabelled DNA probes were produced as described in [72] using the oligonucleotides listed in Table S3 . In vitro synthesized proteins were produced using the TnT coupled rabbit reticulocyte lysate system ( Promega ) . Chick spinal cord nuclear extracts were prepared by manual tissue dissection , cell lysis in a buffer containing 100 mM HEPES , 15 mM MgCl2 , 100 mM KCl and 1 M DTT and protein extraction in a buffer of 20 mM HEPES , 1 . 5 mM MgCl2 , 0 . 42 M NaCl , 0 . 2 mM EDTA and 25% glycerol . Binding reactions were performed in a buffer of 4% Ficoll , 20 mM HEPES , 30 mM KCl , 1 mM DTT and 0 . 1 mM EDTA . Supershift reactions were performed using antibodies validated for use in chick tissue as described in [72] . | The complex organization of tissues is established precisely and reproducibly during development . In the vertebrate neural tube , as in many other tissues , the interplay between extrinsic morphogens and intrinsic transcription factors produces spatial patterns of gene expression that delineate precursors for specific cell types . One such transcription factor , Pax3 , defines the precursors of all sensory neuron subtypes and distinguishes them from precursors fated to give rise to the motor circuits . To gain insight into the molecular mechanisms by which the spinal cord is segregated into these two functional domains , we analysed the genomic regulatory sequences responsible for controlling Pax3 activity . We identified two regions of the genome , the coordinated activity of which establishes and refines Pax3 activity . We showed that the combination of activating signals from secreted Wnt factors together with Nkx family homeodomain repressors restrict Pax3 activity to the presumptive sensory region of the neural tissue . Subsequently , Pax3 acts to directly potentiate its own transcription and this autoregulation sustains Pax3 expression at later developmental stages . Together , our study reveals the way in which intrinsic and extrinsic signals are integrated by cells and converted into a sustained pattern of gene activity in the developing nervous system . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Distinct Regulatory Mechanisms Act to Establish and Maintain Pax3 Expression in the Developing Neural Tube |
Morbidity and mortality from acute diarrheal disease remains high , particularly in developing countries and in cases of natural or man-made disasters . Previous work has shown that the small molecule clotrimazole inhibits intestinal Cl- secretion by blocking both cyclic nucleotide- and Ca2+-gated K+ channels , implicating its use in the treatment of diarrhea of diverse etiologies . Clotrimazole , however , might also inhibit transporters that mediate the inwardly directed electrochemical potential for Na+-dependent solute absorption , which would undermine its clinical application . Here we test this possibility by examining the effects of clotrimazole on Na+-coupled glucose uptake . Short-circuit currents ( Isc ) following administration of glucose and secretagogues were studied in clotrimazole-treated jejunal sections of mouse intestine mounted in Ussing chambers . Treatment of small intestinal tissue with clotrimazole inhibited the Cl- secretory currents that resulted from challenge with the cAMP-agonist vasoactive intestinal peptide ( VIP ) or Ca2+-agonist carbachol in a dose-dependent fashion . A dose of 30 μM was effective in significantly reducing the Isc response to VIP and carbachol by 50% and 72% , respectively . At this dose , uptake of glucose was only marginally affected ( decreased by 14% , p = 0 . 37 ) . There was no measurable effect on SGLT1-mediated sugar transport , as uptake of SGLT1-restricted 3-O-methyl glucose was equivalent between clotrimazole-treated and untreated tissue ( 98% vs . 100% , p = 0 . 90 ) . Treatment of intestinal tissue with clotrimazole significantly reduced secretory responses caused by both cAMP- and Ca2+-dependent agonists as expected , but did not affect Na+-coupled glucose absorption . Clotrimazole could thus be used in conjunction with oral rehydration solution as a low-cost , auxiliary treatment of acute secretory diarrheas .
The discovery of oral rehydration solution ( ORS ) for the treatment of acute diarrheal disease is considered one of the greatest medical advances of the 20th century [1] . Since its introduction in the late 1960’s , the worldwide promotion and distribution of ORS has led to a significant reduction in diarrheal disease-associated mortality [2] . Still , despite the high efficacy of ORS in replenishing isotonic fluid losses and preventing dehydration , diarrheal disease continues to be a leading cause of world-wide mortality , particularly in developing countries , in vulnerable populations such as children and the elderly , and especially in cases of natural ( or man-made ) disasters such as occurred following the recent earthquakes in Haiti and Nepal [2] . Furthermore , recurrent or prolonged infectious diarrhea in childhood is often associated with significant morbidity that impacts long-term growth and development [3 , 4] . As such , there remains a need for new and effective pharmacological therapies for the treatment of acute diarrheal disease [5 , 6] . In many cases of acute infectious diarrhea , the active secretion of salt and water contributes to the clinical disease . This depends primarily on the secretion of Cl- anions into the intestinal lumen by enterocytes lining the intestinal crypt . The resulting increase in the lumenal concentration of Cl- ions mediates a transepithelial electrochemical gradient , which drives the paracellular flux of Na+ and water into the intestinal lumen and thus accounts for the net loss of salt and fluid in diarrhea [7] . In many cases , inhibition of Na+ ( re ) absorption by blockade of the apical membrane Na+/H+-exchanger ( NHE3 ) on absorptive enterocytes in intestinal villi contributes to a further loss of salts and water in the stool [8] . Intestinal Cl- channels are operative under physiological conditions , but become abnormally activated in many diarrheal diseases by agonists that increase levels of intracellular cAMP , cGMP , or Ca2+ . In the intestinal crypt , these second messengers open apical membrane Cl- channels and basolateral membrane K+ channels , allowing efflux of Cl- from enterocytes . These enterocytes are Cl- ion-loaded through the action of the basolateral Na+-coupled Na+/K+/2Cl- co-transporter NKCC1 . Transport by NKCC1 of Cl- into the cell is driven by the inwardly directed electrochemical Na+ gradient produced by the Na+/K+ATPase [7] . In all cases , the driving force for Cl- uptake and secretion is maintained by the concomitant activation of basolateral membrane K+ channels that balance membrane potential and enable the Na+/K+/2Cl- co-transporter to continue operating [7] . Without concomitant K+ efflux , the Cl- secretory pathway shuts down . The imidazole antifungal clotrimazole ( CLT ) blocks conductive K+ transport mediated by both cyclic nucleotide- and Ca2+-gated basolateral K+ channels [9] . This abolishes the driving force for Cl- secretion in the intestine and , therefore , has therapeutic potential to reduce intestinal fluid loss . Indeed , proof-of-principle experiments in mouse and rabbit intestine , as well as human intestinal cell lines suggest the utility of CLT for treatment of secretory diarrheas [9 , 10] . Because of the high efficacy and widespread implementation of ORS in the treatment of diarrheal disease , it is paramount that novel anti-diarrheal agents show efficacy without disrupting rehydration therapy . Fundamental to the working of ORS is intestinal Na+-coupled glucose co-transport by members of the SGLT family [11] . These transporters , most prominently represented by SGLT1 , are directly and functionally dependent on the electrochemical Na+-gradient maintained by the basolateral Na+/K+ATPase and basolateral K+ efflux to maintain cell potential during the absorptive process . There are many K+ channels on the basolateral membranes of enterocytes that allow for Na+/K+ATPase function and that are not responsive to cAMP and Ca2+ . It has been shown in vitro that CLT can affect the Na+/K+ATPase [12] and in principle it might also inhibit other K+ channels affecting membrane potential in the absorptive state , thus inhibiting Na+-dependent solute absorption fundamental to ORS . To address this concern , we tested whether CLT inhibits Na+-dependent glucose uptake in the mouse intestine .
All animals used in these experiments were kept and euthanized in accordance with Institutional Animal Care and Use Committee approved protocol number 13-06-2415R ( IACUC , Boston Children’s Hospital ) , and in adherence to the National Research Council’s ‘Guide to the care and Use of Laboratory Animals’ . Boston Children’s Hospital is accredited by AAALAC International and maintains appropriate Assurance of Compliance with the Office for Protection of Research Risk of the National Institutes of Health . Boston Children’s Hospital assurance number is A3303-01 . Adult Balb/C mice ( 6–8 weeks of age ) were bred in house and euthanized by CO2 . Clotrimazole ( C6019 ) , glucose ( G8270 ) , mannitol ( M4125 ) , phloridzin dihydrate ( P3449 ) , 3-O-Methyl-D-glucopyranose ( M4879 ) , vasoactive intestinal peptide , and carbachol ( C4382 ) were all obtained from Sigma-Aldrich ( St . Louis , MO ) . Krebs-Henseleit buffer ( KHB ) was prepared at a pH of 7 . 6 with NaCl ( 118 mM ) , KCl ( 4 . 7 mM ) , CaCl2 ( 2 . 5 mM ) , MgSO4 ( 1 . 6 mM ) , NaHCO3 ( 24 . 9 mM ) , KH2PO4 ( 1 . 2 mM ) , and sodium pyruvate ( 2 mM ) . Following euthanasia with CO2 , the small intestine of mice was resected and transferred to KHB on ice . At 6 cm from the pylorus , the jejunum was incised and 4 adjacent sections of approximately 2 cm were cut , avoiding Peyer’s patches , and gently flushed with KHB . Sections were opened longitudinally and mechanically stripped of tunica serosa and tunica muscularis . Opened sections were transferred onto Ussing chamber inserts with a 0 . 3 cm2 open surface area ( Physiologic Instruments ) and kept in KHB on ice while adjacent sections were prepared . Up to 3 sections of jejunum were used per mouse . Experiments consisted of the simultaneous analysis of up to 6 individual samples . Inserts with jejunal tissues were mounted in Ussing chambers ( Physiologic Instruments ) and equilibrated in 10 ml pre-warmed , pre-oxygenated ( 95% O2 / 5% CO2 ) KHB , containing 10 mM glucose in the serosal chamber and 10 mM mannitol in the lumenal chamber . Electrophysiological measurements were performed with Ag-AgCl electrodes connected to salt bridges that were prepared from 3% agarose in 3 M KCl solution in plastic electrode tips ( P2023-100 , Physiologic Instruments ) . After 15 minutes of equilibration , transepithelial resistance was measured to assess tissue viability and complete stripping of muscle layer . Vital intestinal tissue was treated by the addition of CLT or ethanol vehicle to both serosal and lumenal chambers for 20 minutes , after which 10 mM glucose was added to the lumenal chamber . The secretagogues VIP and carbachol were added to the serosal chambers . Because current convention identifies the serosal ( basolateral ) bath as ground [13] , both the absorption of positively charged Na+ ions , as well as the lumenal secretion of negatively charged Cl- ions result in an increase in negative Isc . During the experiments , Isc measurements were collected every minute using Acquire software ( Physiologic Instruments ) . ΔIsc was defined as the absolute change in Isc following challenge with the respective compound as follows: ΔIsc = Isct = x−Isct = 0 , in which t = 0 represents the Isc immediately before each challenge and x varies with the investigated compounds as indicated per experiment . In some experiments , the maximal recorded change from baseline ( Max ΔIsc ) was shown for individual tissue samples in addition to the pooled Isc registration . Samples were compared with a two-sided student t test or ANOVA with Bonferroni’s correction for multiple testing where indicated at an alpha level of 0 . 05 . Analyses were performed with Prism 5 . 0 ( GraphPad Software , La Jolla , CA ) .
To confirm the previously described inhibitory effect of CLT on intestinal Cl- secretion induced by cAMP- or Ca2+-dependent agonists [9 , 10] , we compared the short-circuit current ( Isc ) observed before and after challenge with vasoactive intestinal peptide ( VIP ) or carbachol of mouse jejunal tissues treated with 0 , 5 , 30 , or 150 μM CLT in Ussing chambers . CLT treatment resulted in a dose-dependent decrease in Isc induced by both secretagogues , consistent with reduced secretion of Cl- anions by enterocytes ( Fig 1A and 1B ) . With this limited sample number , we did not observe a difference in VIP-induced responses between treatment with 5 or 30 μM CLT , but a dose of 30 μM was found to be effective in significantly reducing the Isc response to both VIP and carbachol by 50% and 72% , respectively . Based on these studies , together with our well delineated dose dependency studies of CLT in human intestinal cell lines [9] , this dose was used in all subsequent studies . To test if CLT interferes with Na+-coupled glucose uptake in the small intestine , we determined the effect of CLT on changes in Isc caused by the addition of glucose to the lumenal ( apical ) chamber of mouse jejunum mounted in Ussing chambers . In both CLT and vehicle-treated tissue , the addition of lumenal glucose resulted in a rapid increase in negative Isc , signifying an increase in the uptake of Na+ cations ( Fig 2A and 2B , and individual tracings are shown in supplemental S1 Fig ) . Fifteen minutes after the addition of glucose to the lumenal reservoir , the intestinal tissues were challenged basolaterally with the cAMP-dependent agonist VIP and Ca2+-dependent agonist carbachol to confirm that CLT inhibited cAMP and Ca2+-induced Cl- secretion under these conditions ( Fig 2C and 2D ) . Though well within normal variation , we did note a small reduction in glucose-induced ΔIsc in CLT-treated tissues when compared to untreated jejunum ( decreased 14% , p = 0 . 37 , Fig 2B ) . Thus , to specifically test the effect of CLT on the function of SGLT1 , the dominant Na+-coupled glucose transported in the intestine [11] , we repeated the experiment depicted in Fig 2 with the SGLT1-specific ligand 3-O-methyl glucose . High-dose ( 30 mM 3-O-methyl glucose ) lumenal challenge resulted in a rapid and predictably more pronounced solute-induced ΔIsc than observed with 10 mM glucose . Even with this greater signal size , no difference was observed between CLT and vehicle-treated tissue ( Fig 3A , 98% vs . 100% , p = 0 . 90 ) ; while secretory responses to VIP and carbachol again remained significantly inhibited under these conditions ( S2 Fig ) . These studies indicate that CLT has no detectable effect on Na-coupled glucose absorption through SGLT-1 . To further confirm that the changes in Isc after glucose or 3-O-methyl glucose administration were mediated by SGLT-1 ( or related family members ) , we performed separate experiments in which tissues were treated with phloridzin , which blocks SGLT1 and to lesser extent other SGLT1 isoforms [14] . In the case of intestinal tissues exposed to 3-O-methyl glucose , the solute-induced ΔIsc was almost completely inhibited by pre-treatment with phloridzin ( Fig 3B and 3C ) . Phloridzin-treated intestinal tissues continued to respond normally with VIP ( cyclic nucleotide ) -mediated Cl- secretion , showing that phloridzin treatment did not affect the viability of intestinal tissue sections and that SGLT1 therefore was the source of the 3-O-methyl glucose-induced ΔIsc ( Fig 3B and 3C ) . In the case of intestinal tissues exposed to glucose ( Fig 3D ) , the solute-induced ΔIsc was also inhibited , and to approximately the same extent as in tissues exposed to 3-O-methyl glucose ( about 30 μAmp ) , but the total ΔIsc induced by glucose was greater and inhibition by phlorizin was incomplete . Thus , SGLT1 explains a large fraction of Na+-coupled solute absorption ( ΔIsc ) induced by glucose; but the results also implicate transport through other SGLT family members known to be expressed in the intestine [11] that are less responsive to phloridzin , such as SGLT4 [14 , 15] . Combined , these studies confirm that Isc changes in response to glucose stimulation reflect Na+-coupled glucose transport , and demonstrate that SGLT1 is unaffected by CLT treatment , further strengthening the rationale for clinical applications of CLT in treatment of secretory diarrhea .
In this report we have assessed whether the K+-channel blocker CLT affected Na+-coupled glucose uptake in the intestine , which could be raised as an important concern for use of this drug in the treatment of secretory diarrheas . Our results demonstrate that Na+-coupled glucose uptake is not affected by CLT when administered in doses that effectively reduced both cAMP- and Ca2+-dependent chloride secretion . This strengthens the indications for use of CLT as an adjunct treatment in the management of diarrheal diseases of diverse etiologies . The experiments presented here directly addressed the concern that the blocking effect of CLT on cAMP- and Ca2+-activated basolateral K+ channels in the enterocyte might disrupt Na+-coupled glucose uptake either by interfering with the maintenance of the intracellular Na+ electrochemical gradient established by the Na+/K+ ATPase , or by direct inhibition of SGLT1 . Inhibition of Na+/K+ ATPase has previously been found in vitro using membrane fragments at a half-inhibiting concentration of 24–30 μM [12] . Similarly , direct inhibition by CLT of gastric H+/K+ ATPase has been documented [16] . Inhibition of these ATPases , however , is partial , even at high doses of CLT ( 100 μM ) [12 , 16] . Furthermore , CLT inhibition of H+/K+ ATPase was pH-dependent [16] . In contrast to these studies , we observed that treatment of small intestinal tissue ex vivo with 30 μM CLT did not reduce Na+-coupled glucose uptake , suggesting that any effects of CLT on Na+/K+ ATPase in intact intestine ex vivo were not sufficient to interfere with effective glucose uptake . As such , our results are in line with previous in vivo studies in both mice and humans using prolonged oral treatment with CLT that showed normal growth and no evidence of toxicity [10 , 17] , suggesting the effects of CLT on Na+/K+ ATPase are unlikely to be physiologically relevant . Our experiments have furthermore ruled out that CLT directly inhibits SGLT1 . Even at doses of 150 μM , which is unlikely to be attained in patients due to the relative insolubility of CLT , there was only a modest apparent decrease ( 39% , n = 4 ) of Na+-glucose absorption while cAMP- and Ca2+-dependent Cl- secretion was nearly completely abrogated . With respect to potential clinical applications , we note that current FDA approved formulations of CLT are restricted to 1% topical/vaginal creams or solutions , and 10 mg lozenges for the treatment of oropharyngeal candidiasis . Previous studies , however , demonstrated that oral CLT is adequately absorbed , reaches peak serum values after 3 hours , and is metabolized primarily in the liver and excreted mostly in the feces [18] . In a trial with adult patients suffering from sickle cell anemia , a daily dose of 20 mg/kg in two tablets resulted in plasma levels <1 μM and was well-tolerated [19] . A small pediatric study reported no adverse events after CLT treatment with 100 mg/kg/day in 4 oral doses for 14–63 days [20] . We propose the use of CLT in treatment of acute diarrheal disease as an additive to currently available ORS solutions , though other forms of oral administration ( e . g . tablets or capsules ) are feasible as well . Our target concentration of 30 μM can be achieved by dissolving 10 . 3 mg CLT per liter of ORS . When administered in this way , the daily dose is unlikely to exceed 10 mg/kg bodyweight and thus remains far below daily doses that have previously been established to be safe [19 , 20] . Furthermore , direct addition of CLT to ORS establishes a very straightforward dosing schedule , since patients with the most severe diarrhea will require more ORS and will thus receive a higher daily dose of CLT . Conversely , CLT treatment will be tapered in parallel with reduced ORS requirement during the recovery phase . In the design of such formulations , it is important to recognize that CLT has poor solubility in water , measured by some to be as low 5 . 6 μg/ml ( 16 μM ) [21] . Although our data demonstrated therapeutic efficacy with concentrations three times below this threshold , solubility and bioavailability of CLT could be enhanced by inclusion with β-cyclodextrin [22 , 23] or solid dispersions [21] . Lastly , it should be recognized that CLT has the potential to interact with other drugs that are metabolized through CYP3A [24] . In summary , our studies provide evidence that CLT can reduce cAMP- and Ca2+-dependent Cl- secretion without affecting the ability of the enterocyte to absorb glucose , salts , and water . CLT is already an FDA-approved drug in widespread use as an anti-fungal , with considerable pharmacokinetic and toxicological data available . It is an ideal candidate for development as a novel , inexpensive anti-diarrheal agent . We suggest that combination treatment of CLT along with ORS could provide further gains in reducing mortality and morbidity caused by secretory diarrheas . | In acute infectious diarrhea , the active secretion of Cl- ions contributes to the secondary loss of Na+ and water from the intestine . Apical Cl- secretion from intestinal epithelial cells is dependent upon cyclic nucleotide- and Ca2+-dependent intracellular signals and requires the concomitant transport of K+ through basolateral K+ channels for maintenance of an electroneutral state . Hence , when efflux of K+ in enterocytes is blocked , Cl- secretion necessarily shuts down . The FDA-approved antifungal drug clotrimazole has been demonstrated to be a potent blocker of basolateral cAMP- and Ca2+-gated K+ channels in enterocytes , and therefore likely has therapeutic efficacy for secretory diarrheas . One important concern that could compromise its clinical applicability as a novel anti-diarrheal drug , however , is that clotrimazole might affect intestinal Na+-coupled glucose absorption , which constitutes the physiological basis of oral rehydration therapies and is thus critical for the efficacy of the current golden standard treatment for acute infectious diarrheal diseases . In this work , we demonstrate that clotrimazole effectively blocks Cl- secretion in mouse intestine after stimulation with secretory stimuli , without affecting the capacity to take up Na+ and glucose . These results pave the way towards further clinical development of clotrimazole as a new pharmacologic strategy for acute diarrheal disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Electrophysiological Studies into the Safety of the Anti-diarrheal Drug Clotrimazole during Oral Rehydration Therapy |
Nutritional immunity describes the host-driven manipulation of essential micronutrients , including iron , zinc and manganese . To withstand nutritional immunity and proliferate within their hosts , pathogenic microbes must express efficient micronutrient uptake and homeostatic systems . Here we have elucidated the pathway of cellular zinc assimilation in the major human fungal pathogen Candida albicans . Bioinformatics analysis identified nine putative zinc transporters: four cytoplasmic-import Zip proteins ( Zrt1 , Zrt2 , Zrt3 and orf19 . 5428 ) and five cytoplasmic-export ZnT proteins ( orf19 . 1536/Zrc1 , orf19 . 3874 , orf19 . 3769 , orf19 . 3132 and orf19 . 52 ) . Only Zrt1 and Zrt2 are predicted to localise to the plasma membrane and here we demonstrate that Zrt2 is essential for C . albicans zinc uptake and growth at acidic pH . In contrast , ZRT1 expression was found to be highly pH-dependent and could support growth of the ZRT2-null strain at pH 7 and above . This regulatory paradigm is analogous to the distantly related pathogenic mould , Aspergillus fumigatus , suggesting that pH-adaptation of zinc transport may be conserved in fungi and we propose that environmental pH has shaped the evolution of zinc import systems in fungi . Deletion of C . albicans ZRT2 reduced kidney fungal burden in wild type , but not in mice lacking the zinc-chelating antimicrobial protein calprotectin . Inhibition of zrt2Δ growth by neutrophil extracellular traps was calprotectin-dependent . This suggests that , within the kidney , C . albicans growth is determined by pathogen-Zrt2 and host-calprotectin . As well as serving as an essential micronutrient , zinc can also be highly toxic and we show that C . albicans deals with this potential threat by rapidly compartmentalising zinc within vesicular stores called zincosomes . In order to understand mechanistically how this process occurs , we created deletion mutants of all five ZnT-type transporters in C . albicans . Here we show that , unlike in Saccharomyces cerevisiae , C . albicans Zrc1 mediates zinc tolerance via zincosomal zinc compartmentalisation . This novel transporter was also essential for virulence and liver colonisation in vivo . In summary , we show that zinc homeostasis in a major human fungal pathogen is a multi-stage process initiated by Zrt1/Zrt2-cellular import , followed by Zrc1-dependent intracellular compartmentalisation .
Certain trace metals such as iron and zinc ( collectively termed micronutrients ) are essential for cellular life , and at least a third of all proteins interact with a metal cofactor [1] . Zinc is particularly important for eukaryotes as around 9% of their proteomes require this metal for function [2] . However , these essential metals can also be highly toxic to cells , and precise metal ion homeostasis is critical for survival . Pathogenic microorganisms face a complicated relationship with micronutrients as the mammalian host uses both high antimicrobial concentrations of metals , as well as metal sequestration to kill microbes or inhibit their growth . Collectively , these processes are known as nutritional immunity ( 4 ) . The “battle for iron” is an established paradigm in host-pathogen interactions [3] and , more recently , important roles for manganese , copper and zinc have emerged within the framework of nutritional immunity [4] . Zinc in particular represents a double-edged sword for potentially invasive species . Botella et al . established that phagocytosed Mycobacterium tuberculosis cells experience acute zinc toxicity within macrophages , and that intracellular survival is reliant on heavy metal efflux P-type-ATPase activity [5] . In other host niches , zinc availability is extremely limited due to systemic zincaemia or locally produced zinc-chelating agents such as calprotectin . In these environments , efficient zinc uptake is crucial for pathogenicity , and a number of recent studies have demonstrated the importance of the znuABC high affinity zinc importer for bacterial virulence [4 , 6] . Fungi do not appear to encode ABC transport systems for zinc acquisition . Instead , eukaryotic zinc transport can be mediated by members of two protein families: the Zip and ZnT transporters , which transport zinc into and out of the cytoplasm , respectively [7] . In the model yeast S . cerevisiae , Zip family members have been shown to assimilate zinc from the environment or to export zinc from intracellular organelles such as the vacuole [8–10] . In contrast , ZnT proteins play roles in organellar zinc accumulation . In S . cerevisiae , the major target for excess zinc is the vacuole [11] , as well as small vesicular zinc storage compartments called zincosomes [12] . Whilst vacuolar zinc import is mediated by the ZnT-type transporter Zrc1 [13] , the mechanism of fungal intracellular zincosomal zinc compartmentalisation is not known [12] . Predicted plasma membrane Zip transporters have now been characterised in the major human fungal pathogens Aspergillus fumigatus , H . capsulatum , Cryptococcus neoformans and C . gattii [14–17] , and Zip transporter mutants for all four species exhibit attenuated virulence , underscoring the importance of zinc uptake for fungal , as well as bacterial pathogenicity . Candida albicans is a normal commensal member of the human gastrointestinal microbiota and other mucosal surfaces , a common cause of mucosal infections , and a serious invasive pathogen in certain patient groups [18] . In fact , invasive candidiasis , predominantly caused by C . albicans , affects more than a quarter of a million individuals each year and is responsible for at least 50 , 000 deaths annually [19] . We have previously shown that this fungus can scavenge zinc via the secreted protein Pra1 and that this “zincophore” system is important for host cell damage in tissue culture infection models [20] . However , a pra1Δ mutant is hyper-virulent in a mouse model of infection as it also serves as a ligand for neutrophil alphaMbeta2 [21 , 22] . In this study we have functionally dissected zinc transport in C . albicans . We identified nine putative zinc transporters including two predicted plasma membrane Zip proteins , Zrt1 and Zrt2 , as well as five ZnT proteins . Regulatory and functional analysis demonstrates that pH-dependent adaptation to zinc limitation may be conserved in fungi , but that distinct transporter subclasses differentially contribute to growth in vivo for different human pathogenic species . Moreover , for the first time , we define a molecular mechanism of zincosomal zinc accumulation in a human fungal pathogen .
Zinc transport in eukaryotes can be mediated by members of the Zip and ZnT protein families , which transport their substrate to or from the cytoplasm , respectively [7] . In order to determine how C . albicans acquires zinc from its environment , we first focussed on Zip transporters . Using the FungiDB [23] InterPro domain-finder ( PFAM: PF02535; http://fungidb . org/fungidb/ ) we identified four Zip-type C . albicans proteins . Only Zrt1 and Zrt2 are predicted plasma membrane transporters . In contrast , Zrt3 and orf19 . 5428 share similarity with S . cerevisiae Zrt3 ( vacuolar zinc ) and Atx2 ( Golgi manganese ) transporters . We had previously generated a C . albicans zrt1Δ mutant as part of our efforts to characterise the fungal zincophore , Pra1 [20] . In this previous study we found that Zrt1 was essential for the reassociation of soluble Pra1 to the fungal cell surface , indicating that Zrt1 is likely cell surface-localised . ZRT2 , on the other hand , was ( at time of writing ) annotated in the Candida Genome Database ( www . candidagenome . org/ ) [24] as a possibly essential gene ( Aaron Mitchell , personal communication to the CGD ) . Indeed , our initial attempts to delete the second allele of ZRT2 were unsuccessful ( the first 104 second-round clones retained their second allele of ZRT2 ) . Supplementation of the transformation selection medium with 1 mM ZnSO4 permitted the successful isolation of a C . albicans zrt2Δ homozygous mutant , suggesting that ZRT2 is conditionally essential . Subsequent attempts to culture C . albicans zrt2Δ in SD ( YNB + glucose ) medium ( the minimal yeast growth medium , routinely used for the selection of transformants ) failed , indicating that ZRT2 is indeed essential for growth under this laboratory condition . Consistent with conditional essentiality , growth of zrt2Δ was restored to wild type levels via zinc supplementation or by genetic complementation with a single copy of ZRT2 ( Fig 1A ) . Deletion of ZRT1 did not impact growth in SD medium . We also tested growth in liquid and agar hyphae-inducing medium and under biofilm conditions , but observed no difference between wild type and zrt1Δ or zrt2Δ strains ( S1 Fig and S2 Fig ) . In the pathogenic mould A . fumigatus , Zrt1 and Zrt2 orthologues ( ZrfC and ZrfB ) are required for zinc uptake at neutral/alkaline and acidic pH , respectively [25–27] . Although pH-dependent zinc transport has not been reported in the more closely related yeast , S . cerevisiae , a previous study has indicated that C . albicans ZRT1 and ZRT2 are also pH-regulated [28] . As SD minimal medium has a native pH of ~4 . 8 , we tested the effect of neutralising the growth medium . Buffering the medium to pH 7 . 4 restored growth of zrt2Δ , and had no adverse effect on zrt1Δ , which again grew to wild type levels ( Fig 1A ) . Similar pH dependent growth patterns and zinc rescue effects were observed in synthetic limited zinc medium ( S3 Fig ) . As C . albicans encodes only two predicted plasma membrane zinc importers , these data indicated that , in laboratory medium , Zrt1 can support growth at neutral-alkaline pH , whilst Zrt2 is essential for growth at acidic pH . Based on these growth patterns , we hypothesised that ZRT1 is specifically expressed at neutral/alkaline pH , whilst ZRT2 expression is pH-independent . We note that this regulatory and functional model aligns more closely with that of the pathogenic mould A . fumigatus [25–27] . To test this hypothesis we constructed C . albicans reporter strains with GFP [29] expression driven from either the ZRT1 or ZRT2 promoters . Fig 1B shows the expression profiles of PZRT1 and PZRT2 in low-zinc medium at a range of environmental pH values . GFP fluorescence driven by PZRT1 activity was low at pH 4 . 6 . However , as the media was neutralised , fluorescence increased . Expression was 40-fold higher at pH 7 . 5 than at pH 4 . 6 . At pH 6 . 5 and above , PZRT1-GFP expression was significantly higher than at pH 4 . 6 . In contrast , PZRT2-GFP expression was not as strongly affected by the pH of the surrounding media , with expression at pH 4 . 6 being only 2-fold higher than at pH 6 . 5 . These data are in agreement with the previous study of Bensen et al . who reported alkaline- and acidic- induction of ZRT1 and ZRT2 , respectively [28] . However , from our own observations , we conclude that expression of ZRT1 is more strongly influenced by environmental pH than ZRT2 . These expression data support our hypothesis that Zrt2 is essential in acidic environments , whilst either Zrt1 or Zrt2 can support growth at neutral pH . To test this directly we created a zrt1Δ/zrt2Δ double mutant and performed more detailed growth kinetics analysis . Fig 1C shows that zrt2Δ again grew at neutral , but not acidic pH , whilst zrt1Δ/zrt2Δ failed to grow at both pH values . Growth was fully restored in the revertant strain ( Fig 1C ) . The above growth and expression assays indicated that Zrt2 is the dominant cellular zinc transporter in C . albicans and the only functional importer at acidic pH . To test this , wild type , zrt2Δ and zrt2Δ+ZRT2 were cultured in low zinc medium ( SD zinc-dropout , acidic ) , provided with 25 μM Zn++ and zinc uptake from the medium measured . Wild type C . albicans sequestered all measurable zinc within 60 minutes . Zinc uptake was virtually abolished in the zrt2Δ mutant and ZRT2 complementation restored uptake to 68% ( Fig 2A ) . Therefore Zrt2 is essential for zinc acquisition by yeast cells in SD minimal medium . We next assessed the relative impact of Zrt1 and Zrt2 on zinc uptake at neutral pH in RPMI medium ( pH 8 . 2 ) at 37°C in tissue culture plates . Under these conditions the wild type took up 74% of zinc from the medium by 180 min ( Fig 2B ) . In line with our observations that ZRT1 and ZRT2 are expressed at neutral pH , both zrt1Δ and zrt2Δ mutants acquired zinc from the medium , but this was reduced by approximately 50% compared to the wild type . Simultaneous deletion of both ZRT1 and ZRT2 abolished zinc uptake . Respective complementation with ZRT1 and/or ZRT2 increased zinc uptake to 55–63% . Therefore , both Zrt1 and Zrt2 contribute to zinc acquisition in RPMI . In summary , Zrt2 is the major zinc importer in C . albicans whilst Zrt1 can support zinc uptake and growth specifically at neutral/alkaline pH . Both transporters are members of the Zip ( Zrt/Irt protein ) family , which also include iron transporters . Therefore , to assess the metal specificity of ZRT1 and ZRT2 regulation , we tested their expression in response to zinc and three other physiologically relevant trace metals—iron , manganese and copper . The reporter strains were incubated in low zinc media , buffered to pH 5 or to pH 7 . 5 , and supplemented with zinc , iron , manganese or copper at 100 μM . At pH 5 , PZRT1 activity was again very low and supplementation with the different metals had no appreciable effect on expression ( Fig 3A ) . At pH 7 . 5 , PZRT1 was 13 . 4-fold induced compared to pH 5 ( Fig 3A vs . 3B ) . The addition of zinc to the medium resulted in 40-fold repression of PZRT1 whilst iron , manganese and copper supplementation had no effect ( Fig 3B ) . PZRT2 was again active in both acidic and neutral/alkaline media . At pH 5 and pH 7 . 5 , zinc supplementation resulted in 6 . 5- and 3 . 2- fold repression , respectively . Supplementation with iron , manganese or copper had no effect ( Fig 3C & 3D ) . From these data , we conclude that the metallo-regulation of ZRT1 and ZRT2 is zinc-specific in C . albicans . In order to functionally assess metal specificity , wild type , zrt2Δ and zrt1Δ/zrt2Δ cells were again cultured in minimal media , supplemented with zinc , iron , manganese , or copper . Fig 4 shows that zinc , but not iron , manganese or copper supplementation restored growth , indicating that the growth defect of these mutants is due to an inability to acquire zinc in minimal media . From these in vitro assays , it would appear that zinc transport in C . albicans is actually more similar to A . fumigatus than to S . cerevisiae . Baker’s yeast encodes two plasma membrane importers: the high affinity Zrt1 and low affinity Zrt2 , neither of which are known to be subject to pH-regulation [8 , 9] . In contrast , A . fumigatus encodes three zinc importers: ZrfA and ZrfB , which are expressed in acidic environments , and ZrfC , which is expressed at neutral/alkaline [25 , 27] . As A . fumigatus ZrfB and ZrfC are respective orthologues of C . albicans Zrt2 and Zrt1 [30] , this suggests that zinc transporter pH-dependence may be conserved in multiple fungal species . In [30] and in supplementary data S4 Fig we propose an evolutionary framework of how pH adaptation may have shaped the evolution of fungal zinc transporters . The role of zinc uptake in C . albicans virulence remains largely unexplored . Here we used a murine model of disseminated candidiasis to directly assess the role of Zrt1 and Zrt2 in C . albicans fitness in vivo . Mice were infected intravenously and kidney fungal burden assessed at day one and day three post-infection . Fig 5A shows that by day one post-infection , all strains exhibited similar levels of kidney fungal burden , indicating that neither Zrt1 nor Zrt2 are required for initial kidney colonisation . However , by day three post-infection , C . albicans wild type kidney fungal burden had increased significantly by 6 . 5-fold ( P = 0 . 034 ) , indicating that cells had proliferated in this organ . In contrast , deletion of ZRT2 precluded an increase in kidney fungal burden between day one and day three post-infection ( P = 0 . 597 ) . Complementation of zrt2Δ with a single copy of ZRT2 restored kidney colonisation at day three ( 4 . 5-fold higher than at day one , P = 0 . 004 ) . In contrast , deletion of ZRT1 did not inhibit fungal proliferation in the kidney . These data indicate that Zrt1 and Zrt2 are dispensable for initial kidney colonisation ( day one post-infection ) and that Zrt2 is important for systemic candidiasis at later stages . These data are in agreement with previous studies . Xu and co-workers identified a transcription factor , Sut1 , which governs the expression of zinc assimilation genes during invasive candidiasis [31] . Deletion of SUT1 attenuated C . albicans virulence , however , sut1Δ virulence was restored to wild type levels via ZRT2 overexpression , indicating that defective in vivo expression of ZRT2 was responsible for the attenuated virulence of sut1Δ [31] . Several other studies have analysed the C . albicans transcriptome during kidney colonisation . Walker and co-workers reported that only two genes were transcriptionally upregulated during both rabbit [32] and mouse [33] kidney colonisation: ADR1 and ZRT2 . ZRT2 is also upregulated during in vitro incubation with macrophages [34] . Combined with the zrt2Δ in vivo growth defect reported here , these expression studies suggest an important role for Zrt2 in zinc uptake during invasive candidiasis . We next addressed the role of host-driven nutritional immunity on fungal growth in vivo . Calprotectin plays a key role in mediating zinc nutritional immunity . Calprotectin is a heterodimeric protein composed of S100A8 and S100A9 subunits which has potent antifungal activity via zinc sequestration [35 , 36] . Calprotectin expression in C . albicans-infected murine kidney tissue has been reported to be upregulated between day one and day three post-infection in two independent studies [36 , 37] . As Zrt2 is important for growth under zinc limitation in vitro and exhibited impaired growth in vivo , we examined the impact of calprotectin on C . albicans kidney colonisation . Surprisingly , all five tested C . albicans strains exhibited lower kidney fungal burdens in calprotectin-deficient mice than in wild type animals at both day one and day three post-infection ( Fig 5B ) . At day three post-infection , the fungal burden of calprotectin-deficient mouse kidneys infected with wild type C . albicans was significantly lower ( p < 0 . 05 ) than wild type mice infected with the same strain . This was unexpected , as calprotectin-deficient mice have been previously shown to succumb earlier to C . albicans infections [35] , however a recent study has also reported lower kidney fungal burden in calprotectin deficient mice compared to wild type [36] . In addition to its anti-fungal activity via zinc sequestration , calprotectin plays additional roles in immunity . Indeed , as well as its role in nutritional immunity , calprotectin has been implicated as an inflammatory mediator and has been shown to exacerbate disease in other models of candidiasis [38] . These additional immune properties may explain the decreased fungal burden observed in calprotectin-deficient mice . Nevertheless , in calprotectin-deficient mice , the zrt2Δ mutant did not exhibit a notable difference in kidney colonisation compared to wild type C . albicans . This indicates that , in the absence of a host calprotectin response , fungal Zrt2 is dispensable for kidney colonisation by C . albicans . Calprotectin constitutes around half the cytoplasmic protein content of neutrophils and is a major component of neutrophil extracellular traps ( NETs ) , from which it elicits its antifungal activity via zinc sequestration [35 , 39] . In order to explore the host-pathogen relationship between pathogen Zrt2 and host calprotectin in greater detail , we next compared the antifungal properties of wild type and calprotectin-deficient NETs . Calprotectin-decoration of NETs and associated antifungal activity via zinc sequestration has been well defined [35] . In line with this , S100A9-/- NETs exhibited highly attenuated antifungal activity compared to NETs from wild type neutrophils ( S6 Fig ) . Deletion of ZRT2 rendered C . albicans sensitive to NET antifungal activity in a calprotectin-dependent manner ( Fig 6 ) , suggesting a role for Zrt2 in growth in the presence of calprotectin+ NETs . In summary , host ( calprotectin ) and pathogen ( Zrt2 ) factors appear to define the struggle for zinc during C . albicans infection: Zrt2 is the major zinc transporter of this important fungal pathogen and is essential for growth in the presence of calprotectin in vivo and ex vivo . From this study , and work from the groups of Mitchell , Calera , Deepe , Staats and Jung , it is becoming increasing clear that zinc acquisition plays a critical role in fungal pathogenesis , as perturbation of zinc transporter function in C . albicans , A . fumigatus , H . capsulatum , C . neoformans and C . gattii attenuates virulence in all five organisms tested thus far [14–17] . Moreover , deletion of the master regulator gene of zinc homeostasis in fungi , ZAP1 , also attenuates virulence in A . fumigatus , C . gattii and C . dubliniensis and decreases in vivo fitness in C . albicans [40–43] . In supplementary information S5 Fig we discuss how different zinc uptake genes are differentially required for virulence in the major fungal pathogens of humans . We next sought to address how the fungal cell deals with zinc following its internalisation . This is an important issue because , as well as serving as an essential micronutrient , zinc can be highly toxic to cells . In order to assess the dynamics of intracellular zinc compartmentalisation , we utilised zinquin . Zinquin is a zinc-specific fluorescent probe which accumulates in storage vesicles known as zincosomes and fluoresces upon zinc binding [44] . Zinc-depleted cells , prepared by growing the cells overnight in low zinc medium , were pulsed with 25 μM zinc , washed and fixed at five minute intervals and stained with zinquin . Fig 7 shows that even with immediate washing and fixation , C . albicans already stained positive with zinquin , indicating that zincosomal zinc compartmentalisation upon exposure to zinc occurs rapidly . By 20 minutes post-pulse , the majority of cells exhibited numerous zincosomes as indicated by zinquin fluorescence . Therefore , C . albicans rapidly compartmentalises zinc within zincosomes in response to changes in environmental zinc . We therefore turned our attention to ZnT-type transporters which , in contrast to Zip transporters ( such as Zrt2 ) , transport their substrate from the cytoplasm to either outside the cell , or into the lumen of intracellular compartments [7] . Five C . albicans ZnT-type ( PF01545 ) transporters were identified using FungiDB with sequence similarity to S . cerevisiae Mmt1/2 ( orf19 . 52 ) , Zrg17 ( orf19 . 3769 ) , Msc2 ( orf19 . 3132 ) , and Cot1/Zrc1 ( orf19 . 1536 ) , as well as a fifth protein encoded by orf19 . 3874 which does not have an orthologue in S . cerevisiae ( Table 1 ) . We therefore created deletion mutants for these five putative zinc transporter genes . For orf19 . 1536 , we propose the common name , Zrc1 . To determine which ZnT-transporter may mediate zincosome compartmentalisation , wild type , zrc1Δ , orf19 . 3874Δ , orf19 . 3769Δ , orf19 . 3132Δ , and orf19 . 52Δ cells were pulsed with zinc for 20 minutes and stained with zinquin . Fig 8A shows that the isogenic wild type exhibited a significant 5 . 6-fold increase in zinquin fluorescence following the zinc pulse Deletion of orf19 . 3874 , orf19 . 3769 or orf19 . 3132 had no effect in this assay . The orf19 . 52Δ mutant exhibited perturbed zincosome generation , but this was not significant under the conditions tested here . Deletion of ZRC1 , on the other hand , strongly inhibited zincosome formation and this was restored to wild type levels by genetic complementation with a single copy of ZRC1 ( Fig 8A ) . This screen indicated that the ZnT-type transporter , Zrc1 , plays a role in zincosome formation . For wild type , zrc1Δ and zrc1Δ+ZRC1 strains , the experiment was repeated and zincosome accumulation determined at 5 , 10 and 20 minutes post-pulse by flow cytometry . Fig 8B shows that both wild type and zrc1Δ+ZRC1 strains exhibited progressive increases in zinquin fluorescence following the zinc pulse , resulting in more than a 10-fold increase by 20 minutes compared to the pre-pulse condition . In contrast , zrc1Δ exhibited only a moderate ( ~3-fold ) increase in fluorescence by 5 minutes , and the signal did not significantly increase at later time points . These data show that the ZnT-type transporter Zrc1 is required for zincosomal zinc accumulation in C . albicans . Interestingly , when we measured actual zinc uptake within this shorter time period , cells took up less than 30% within 20 minutes , suggesting that these very early zincosome formation events ( Figs 7 and 8 ) may be the result of intracellular zinc mobilisation , prior to significant cellular uptake ( Fig 2 ) . Indeed , we have very recently demonstrated that C . albicans undergoes very rapid ( seconds ) remobilisation of intracellular zinc pools upon changes in environmental zinc , in the absence of cellular uptake [45] . The kinetics of zincosome formation in the model yeast S . cerevisiae have been reported to be similar to those described here , occurring within 5–20 minutes exposure of zinc-depleted cells to a zinc pulse [12] . However , the mechanistic basis of zincosomal zinc accumulation appears to be fundamentally different in these two species . S . cerevisiae encodes two orthologues of C . albicans Zrc1: Zrc1 and Cot1 . However , single zrc1Δ , cot1Δ and zrc1Δ/cot1Δ double mutants exhibited wild type zincosome formation , suggesting that neither ScZrc1 nor its paralogue , Cot1 , are involved in zincosome formation in S . cerevisiae [12] . In fact , S . cerevisiae Zrc1 instead plays a clear and important role in vacuolar zinc accumulation [11] . We therefore sought to characterise the relationship between our novel Zrc1-zincosome pathway and vacuolar zinc in C . albicans . Co-staining cells with zinquin and the vacuolar membrane dye FM4-64 [46] revealed that zincosomes are not found within the fungal vacuole in C . albicans but rather , close to the outer leaflet of the vacuolar membrane ( Fig 9A ) . Given this relatively close spatial relationship , we next questioned whether Zrc1-dependent zincosomal zinc compartmentalisation was an upstream component of vacuolar zinc trafficking in C . albicans . We first established that C . albicans can sequester zinc within the vacuole using the fluorescent probe Zinpyr1 ( Fig 9B ) . Interestingly , in our zinc-pulse experiment , zrc1Δ accumulated vacuolar zinc to the same levels as the wild type , even after extended incubation ( Fig 9C ) . Therefore , under the conditions tested , Zrc1 in C . albicans is not essential for vacuolar zinc import . Zrc1 has been reported to localise to the vacuole in S . cerevisiae and C . neoformans [47 , 48] . However , our own analysis indicated that C . albicans Zrc1 is dispensable for vacuolar zinc import under the conditions tested here ( Fig 9C ) . To test whether Zrc1 localises to the C . albicans vacuole , we tagged the protein at its C-terminus with a codon optimised Venus fluorescent protein . Fig 10 shows that C . albicans Zrc1 , unlike its S . cerevisiae and C . neoformans orthologues , does not localise predominantly to the vacuolar membrane , but instead to the internal membrane system , reminiscent of the endoplasmic reticulum . This localisation is more similar to that of Schizosaccharomyces pombe Zhf1 which transports zinc into the endoplasmic reticulum [49] . Given the importance of Zrc1-mediated vacuolar zinc detoxification in the model yeast S . cerevisiae and in the basidiomycete pathogen C . neoformans , we next questioned whether a relationship exists between Zrc1 , zincosomes , and metal tolerance in C . albicans . First , we screened zrc1Δ , as well as all other ZnT-transporter deficient mutants for sensitivity to log10-fold increases in Zn++ , Fe++ , Mn++ and Cu++ alone or in combination . We included the other mutant strains and other metals to test for potential redundancy and transporter promiscuity . We did not observe significant synergistic toxicity of the tested metals , however excess manganese protected cells from zinc toxicity . The mutant lacking orf19 . 3874 exhibited increased sensitivity to excess manganese and all strains exhibited relatively similar levels of iron and copper tolerance . ( S7 Fig ) . Lack of Zrc1 , on the other hand , resulted in approximately 100-fold increased Zn++-sensitivity ( Fig 11A and S7 Fig ) and genetic complementation restored Zn++ tolerance back to wild type levels ( Fig 11A ) . The observed Zn++ sensitivity of C . albicans observed in these experiments is likely due growth inhibition , rather than fungal killing . Indeed , we had to expose cells to molar concentrations of zinc to kill C . albicans . Although zrc1Δ was also hypersensitive to Zn++ killing ( Fig 11B ) , it is unclear whether C . albicans will face such high levels of Zn++ in nature . On the other hand , sub-millimolar to millimolar concentrations are well within the physiological range C . albicans will likely face in its natural environment as a human commensal and pathogen . Therefore , C . albicans Zrc1 plays a crucial role in adaptation to environmental zinc . To examine whether there was a link between Zrc1-dependent zinc tolerance and zincosome formation , we exposed cells to 1 mM Zn++ for 2 h and measured zinquin fluorescence . This was chosen because Zrc1 is essential for growth at this concentration ( S7 Fig and Fig 11A ) and , whilst it is tolerated by wild type cells , is close to toxicity . Wild type C . albicans cells exhibited a considerable ( 31-fold ) increase in zinquin fluorescence in response to challenge with 1 mM Zn++ . This was significantly reduced in zrc1Δ and restored to wild type levels by genetic complementation with ZRC1 ( Fig 12A ) . Fluorescence microscopy revealed that these quantitative measurements reflect zincosome formation in wild type and zrc1Δ+ZRC1 , but not in zrc1Δ cells ( Fig 12B ) . Therefore , Zrc1 plays a crucial role in zincosomal zinc compartmentalisation in response to both relatively minor fluctuations in zinc availability ( Fig 8 ) and potentially toxic levels of heavy metal ( Figs 11A and 12 ) . Together these data suggest that Zrc1-dependent zincosome formation is important for C . albicans adaptation to environmental zinc . The current study is amongst the first detailed reports of intracellular zinc trafficking in a human fungal pathogen . We therefore assessed whether C . albicans Zrc1 plays a role in virulence . For this we chose two different infection models . Insect larvae have been reported to accumulate high levels of zinc [50] . We therefore first performed infection experiments on the commonly used Galleria mellonella larva . The majority of wild type and zrc1Δ+ZRC1 infected larvae succumbed to infection within 2–3 days post infection . Strikingly , only a single zrc1Δ infected larvae died in these experiments , showing that Zrc1 is essential for virulence in this model ( Fig 13 & S8 Fig ) . Although C . albicans is not a known pathogen of insect larvae , this observation is interesting because it suggests that Galleria may possess a form of high-zinc nutritional immunity; a phenomenon which has been reported in mammals [5] and , recently , in plants [51] . In mammals , inflammation and the acute phase response result in zinc is trafficking to the liver in order to induce zincaemia [52] . We therefore assessed the capacity of zrc1Δ to colonise the murine liver . As shown in Fig 14 , zrc1Δ exhibited a clear and significant defect in liver colonisation compared to both wild type and zrc1Δ+ZRC1 . In contrast , zrc1Δ exhibited the same kidney fungal burden as the wild type ( S9 Fig ) . zrc1Δ mice gained 5% body weight between day 1 and day 3 post infection , whilst wild type and zrc1Δ+ZRC1 infected mice lost weight ( 1 . 2–3% ) . This , together with larval survival ( Fig 13 ) and liver colonisation ( Fig 14 ) data indicate that Zrc1 plays an important role in C . albicans virulence . In summary , we have described a novel pathway of zinc import and compartmentalisation in C . albicans and demonstrated the significance of these mechanisms for both microbial physiology and in vivo fitness . Interestingly , the cellular import pathway of this fungus appears to be highly similar to that of A . fumigatus and we have proposed an ecological-evolutionary framework which may explain some of the conservation and divergence that we observe in extant human fungal pathogenic species . We also demonstrate that unlike any previously characterised pathogenic fungi , C . albicans assimilates zinc from environment to zincosomes using a Zrt1 , 2/Zrc1-dependent biphasic mechanism .
C . albicans strains used in this study are listed in S1 Table . The triple-auxotrophic strain BWP17 complemented with plasmid CIp30 served as the isogenic wild type control in all experiments . Homozygous C . albicans mutants were constructed as described previously [53] and the primers used for this are listed in S1 Table . Briefly , forward primers were designed with 104 bp homology to the immediate upstream region of the gene of interest , followed by a 22 bp sequence , with homology to the pFA plasmids , immediately upstream of the respective selective marker . Similarly , reverse primers were designed with 104 bp homology to the immediate downstream region of the gene of interest ( reverse complement ) , followed by 24 bp sequence with homology to the pFA plasmids , downstream of the selective marker . These long primers , together with plasmids pFA-HIS1 and pFA-ARG4 were used to create deletion constructs for each of the zinc transporter encoding genes and the two alleles of each gene sequentially deleted using the improved transformation protocol [54] and selecting for histidine or arginine prototrophy . In each case , correct integration was determined using gene-specific upstream and downstream primers , lying outside the site of homologous recombination to determine absence of wild type copy and presence of::HIS1 and::ARG4 alleles , as well as HIS1 and ARG4 specific internal primers to ensure correct integration of selective markers at both 5’ and 3’ . For double deletion of ZRT1 and ZRT2 , the zrt1Δ uridine auxotrophy was sequentially transformed by the SAT flipper technique to delete both copies of ZRT2 . All these uridine auxotrophs were URA3 complemented with NcoI-linearised CIp10 plasmid [55] . For the double mutant , both ZRT1 and ZRT2 including up- and down- stream sequences were sub-cloned into CIp10 . For ZRT2 and ZRC1 , the wild type alleles , together with the up- and down- stream intergenic regions were amplified from SC5314 gDNA with phusion polymerase and cloned into CIp10 at MluI and SalI sites . Resulting plasmids were linearised with NcoI and used to complement the respective homozygous deletion mutants . For creation of the PZRT1 and PZRT2 GFP reporters , the upstream intergenic regions of ZRT1 and ZRT2 were amplified with phusion polymerase from SC5314 gDNA , cloned into CIp10-GFP [29] at XhoI and MluI sites and verified by sequencing . Resulting plasmids were linearised with NcoI and transformed into CAI4 for integration at the RPS1 locus . In order to localise Zrc1 , the protein was tagged at the C-terminus which is predicted to face the cytoplasm ( Octopus [56] , Phobius [57] , and TMHMM [58] ) , with a Venus yellow fluorescent protein . The Venus sequence was codon optimised for expression in C . albicans and synthesised ( GeneArt ) , flanked by Pfl23II ( 5’ ) and BamHI ( 3’ ) . The gene was subcloned into pFA-HIS1 at these sites generating pFA-HIS1-Venus . Both Venus and the HIS1 cassette were amplified with primers ZRC1Ven-FG and ZRC1Ven-RG . These primers include 30 and 29 base pairs sequence homology to the template plasmid for amplification at the 3’ , and 100 and 99 bp homology to the ZRC1 locus , to replace the ZRC1 stop codon with Venus . The forward primer additionally contained ggtggtggt between locus- and plasmid- specific regions to introduce a 3 × glycine linker between Zrc1 and Venus . The amplified construct was used to replace the remaining stop codon in the zrc1Δ/ZRC1 heterozygote which was then URA3-complemented with CIp10 as above . Resulting zrc1/ZRC1-VENUS strains were successfully cultured in the presence of 250 μM ZnSO4 to ensure functionality of the tagged protein . Strains were maintained on YPD agar [1% yeast extract , 2% myco-peptone , 2% D-glucose , 2% agar] . Liquid overnight cultures were grown in YPD or SD medium in a shaking incubator at 30°C and 200 rpm . Transformants were selected on SD agar supplemented with arginine , histidine and/or uridine ( each 20 μg ml-1 ) as required . For isolation of the zrt2Δ deletion mutant , selection plates were additionally supplemented with 1 mM ZnSO4 . Escherichia coli was grown on LB agar [1% bacto-tryptone , 0 . 5% yeast extract , 1% NaCl , 2% agar] and overnight E . coli cultures were cultivated in a shaking incubator at 37°C and 200 rpm . For selection purposes 50 μg/ml ampicillin were added to solid or liquid LB medium . To elicit severe zinc restriction , cells were precultured in YPD , washed three time in ultra-pure water and inoculated at OD600 ( 0 . 05 ) in 4 ml LZM ( limited zinc medium with the components listed in S1 Table ) in plastic Universal flasks and incubated at 30°C , 200 rpm for three days . For growth experiments in 96 well plates , cells were inoculated to OD600 ( 0 . 001 ) and incubated for seven days . For pH-defined LZM , NaOH was added to alkalinise the medium as required and then the media was buffered with 50 mM Na-tartrate ( pH4 . 5 ) MES ( pH 5–6 . 5 ) or HEPES ( pH 7–8 ) . To determine ZRT1 and ZRT2 promoter activity , CAI4+CIp10 , PZRT1-GFP and PZRT2-GFP strains were cultured overnight in YPD , washed three times with ultra-pure water and inoculated to OD600 ( 1 ) into pH-buffered LZM in black walled , clear-bottomed 96 well plates and incubated for 16 h . Fluorescence was measured at 485/520 nm and background ( CAI4+CIp10 ) fluorescence subtracted . To determine metal toxicity , cells from an SD overnight culture were inoculated into SD medium containing indicated metals ( starting OD600 0 . 05 ) and OD600 determined following 24 h incubation at 30°C . To determine fungal killing , cells were pre-grown in YPD for 24 h , washed twice in 1mM EDTA , twice in ddH20 , then inoculated into fresh SD0 medium to a final OD600 = 0 . 5 for 24 h . After incubation , cells were adjusted to 105 cells/mL in SD0 + 1M ZnSO4 or , as a control , ddH20 for 3 h . Following incubation , cells were washed twice in ddH20 , counted and then diluted to 1000 cells/mL in ddH20 . Subsequently , 100μl of cell suspension ( 100 cells ) was spread on YPD plates and incubated at 30˚C . Following incubation , CFUs were counted and compared to determine % survival . To assess zincosomal zinc compartmentalisation , cells were pregrown in YPD , 30°C , 200 rpm for one day , washed three times with distilled water and inoculated into minimal medium without added zinc “SD0” ( 2% glucose , 0 . 5% NH4SO4 , 1X YNB without zinc [Formedium] ) . Whilst this medium does not contain added zinc , it also lacks a chelator , and thus represents moderate zinc depletion . For microscopy and flow cytometry experiments , cell were inoculated to OD600 = 0 . 05 . For the mutant screen , cells were inoculated to OD600 = 4 . This was to ensure that all strains were at a similar phase of growth , because the zrt2Δ mutant grows poorly in the absence of exogenous zinc . These prestarved cells were then exposed to 25 μM ZnSO4 for various times . Pre-pulsed and zinc-pulsed cells were fixed in Histofix , washed in PBS and stained with 25 μM zinquin ethyl ester ( Sigma ) for 40 minutes . Cells were again washed with PBS and analysed . For microscopy , cells were additionally stained with Concanavalin A Alexafluor 647 to visualise the cell surface and analysed using DeltaVision microscope using appropriate filters ( DAPI and RhTRITC ) . Original microscopy DV files are in S10 Fig . For the mutant screen , stained and unstained cells were added to the wells of a black-walled clear-bottomed 96 well plate and fluorescence measured at 355/475 nm using a FluoStar plate reader . Measurements were normalised by subtracting the background fluorescence of unstained cells from the stained samples . For flow cytometry , approximately 105 cells were measured using a BD LSRFortessa . To localise zincosomes and the fungal vacuole , cells from an overnight YPD culture were washed with 1 mM EDTA and then ddH2O , incubated in SD0 for 2–3 h . Cells were then incubated with 40 μM FM4-64 and 250 μM ZnSO4 for 45 minutes , washed with EDTA then PBS , incubated zinquin for 45 minutes , washed and visualised using a DeltaVision fluorescent microscope . To visualise vacuolar zinc , cells were pre-grown in YPD for 24 h , washed twice in PBS , and then stained with ZinPyr-1 ( 10 μM , 1 h , 37°C , 200 rpm , washed twice in PBS and incubated for a further 1 h ) . Following ZinPyr-1 staining , cells were stained with 40μM FM4-64 in YPD + 1mM ZnSO4 for 40 min at 30˚C with shaking in the dark . Following this , cells were washed twice in YPD + 1mM ZnSO4 and subsequently inoculated into YPD + 1mM ZnSO4 for 90 min without dye at 30˚C with shaking in the dark . Cells were then imaged using confocal microscopy . To determine vacuolar import kinetics , wild type zrc1Δ cells were pre-grown in YPD for 24 h , washed twice in 1mM EDTA , twice in ddH20 , and then inoculated into fresh SD0 medium to a final OD600 = 0 . 5 for 24 h . After incubation , cells were stained with 10μM ZinPyr-1 in PBS for 1 h at 37˚C with shaking in the dark . Cells were then washed twice in PBS and incubated for a further 1 h at 37˚C with shaking in the dark . Following incubation , cells were pulsed with 25μM ZnSO4 in SD0 medium and incubated at 30˚C with shaking in the dark . At indicated time points , 100μl of sample was collected and transferred to a black-bottomed 96 well plate and quantified for ZinPyr-1 fluorescence using a fluorescent microplate reader . Cells were pre-grown in YPD for 24 h , washed twice in 1mM EDTA , twice in ddH20 , then inoculated into fresh SD25 medium to a final OD600 = 0 . 5 for 24 h . After incubation , cells were washed twice in 1mM EDTA , twice in PBS , adjusted to 5 x 106 cells/mL in PBS , and then 20 μl ( 1 x 105 cells/mL ) injected into the abdominal pro-leg of larvae . Survival of the larvae was monitored on a 12 h basis post-infection . Mice were kept in the animal facility Umeå Centre for Compartive Biology ( UCCB ) . All animal experiments in this study were carried out in strict accordance with the recommendations in the guide for the care and use of laboratory animals conformed to Swedish animal protection laws and applicable guidelines ( djurskyddslagen 1988:534; djurskyddsförordningen 1988:539; djurskyddsmyndigheten DFS 2004:4 ) and with the Swedish animal protection law in a protocol approved by the local Ethical Committee ( Umeå djurförsöksetiska nämnd ) permit number A79-14 . For analysis of in vivo fitness and virulence , C57BL/6 wild-type mice and S100A9-/- mice from the same background were infected intravenously with 5 x 105 CFUs per animal from logarithmically growing C . albicans cultures . Male and female mice were included in equal numbers for all infections , the average age of the mice was 12–16 weeks . Mice were sacrificed by cervical dislocation after one or three days of infection . Kidneys and liver were harvested , homogenised and resulting cell suspensions were plated on YPD plates to determine fungal burden . Neutrophils were isolated as described before [59] . Briefly , C57BL/6 mice were sacrificed by cervical dislocation and femurs and tibia of both hind limbs were dissected . Bone marrow was flushed out with RPMI1640 w/o PR supplemented with 100 μg/ml Carbenicillin and 50 μg/ml Kanamycin ( Duchefa , both ) . After red blood cell lysis , neutrophils were purified using a discontinuous Percoll gradient of 52% , 69% and 78% PBS-buffered Percoll ( GE Healtcare ) . Collected neutrophils from the 69%/78% interface were washed , resuspended in HBSS- and kept on ice . Prior to use , neutrophils were counted using a Vi-CELL cell counter ( Beckman Coulter ) and diluted to desired concentration in RPMI1640 w/o PR with antibiotics . All following assays were performed in this medium , if not stated otherwise . Inhibitory capacity of mouse NETs was quantified as explained earlier [60] . 5 x 105 mouse neutrophils were seeded into a 24-well plate . NET formation was induced by 100 nM phorbol myristate acetate in the presence of 1% ( V/V ) DNase-free mouse serum . Incubation occurred for 20–22 h at 37°C with 5% CO2; NET induction was verified microscopically . NET supernatants were gently removed and 500 μl RPMI w/o PR were added containing 5 x 104 Candida cells to reach a multiplicity of infection ( MOI ) of 0 . 1 . Incubation occurred for 20–22 h at 37°C with 5% CO2 . Fungal viability was assessed by metabolic activity [61] . Briefly , 0 . 33 mg/ml XTT ( 2 , 3-bis ( 2-methoxy-4-nitro-5-sulfophenyl ) -5-[ ( phenylamino ) carbonyl]-2H-tetrazolium hydroxide; Invitrogen ) and 27 μg/ml Co-enzyme Q0 ( Sigma-Aldrich ) were added to each well . After an incubation of 15 min at 37°C , the 450 nm absorbance of the supernatants was measured using a Fluostar Omega plate spectrometer ( BMG Labtech ) . Kidney fungal burden was analysed in IBM SPSS Statistics 24 . Normality and homogeneity of variance were first tested , and ANOVA and Kruskal-Wallis tests performed as appropriate for each data set . For growth assays and expression analysis , data were analysed using GraphPad Prism and either Student’s t-test or ANOVA performed as appropriate . For phylogenetic analyses , amino acid sequences were acquired from FungiDB [23] or from the Candida Genome Database [24] . To construct phylogenetic trees Phylogeny . fr One Click was used [39 , 62]: Alignments were performed using MUSCLE , maximum likelihood calculated using PhyML and tree rendering using TreeDyn . | All living organisms must secure certain trace metals such as iron and zinc in their diets . For the microbes that infect us , the source of these micronutrients is the tissues of their host . However , mammals have developed sophisticated mechanisms to manipulate microbial access to trace metals–a process called nutritional immunity . Therefore , successful pathogenic microorganisms must have evolved mechanisms to counteract nutritional immunity and acquire micronutrients in order to grow within their hosts and cause disease . This struggle for micronutrients represents a key host-pathogen battleground . In this study we demonstrate how the major human fungal pathogen , Candida albicans , acquires and stores zinc from its environment . We find that the mechanistic basis of zinc uptake is highly dependent on the acidity of the surrounding environment . Interestingly , this pH-dependence appears conserved in the fungal kingdom and we propose a potential framework for the evolution of zinc uptake in extant fungal species . Moreover , following cellular assimilation , C . albicans shuttles this potentially toxic transition metal into subcellular compartments called zincosomes . We also show that both zinc uptake and compartmentalisation are critical for C . albicans growth , both under laboratory conditions and in experimental models of invasive candidiasis . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"zinc",
"transporters",
"manganese",
"pathogens",
"microbiology",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"biological",
"transport",
"kidneys",
"fungal",
... | 2018 | Biphasic zinc compartmentalisation in a human fungal pathogen |
Epidemic severe leptospirosis was recognized in Nicaragua in 1995 , but unrecognized epidemic and endemic disease remains unstudied . To determine the burden of and risk factors associated with symptomatic leptospirosis in Nicaragua , we prospectively studied patients presenting with fever at a large teaching hospital . Epidemiologic and clinical features were systematically recorded , and paired sera tested by IgM-ELISA to identify patients with probable and possible acute leptospirosis . Microscopic Agglutination Test and PCR were used to confirm acute leptospirosis . Among 704 patients with paired sera tested by MAT , 44 had acute leptospirosis . Patients with acute leptospirosis were more likely to present during rainy months and to report rural residence and fresh water exposure . The sensitivity of clinical impression and acute-phase IgM detected by ELISA were poor . Leptospirosis is a common ( 6 . 3% ) but unrecognized cause of acute febrile illness in Nicaragua . Rapid point-of-care tests to support early diagnosis and treatment as well as tests to support population-based studies to delineate the epidemiology , incidence , and clinical spectrum of leptospirosis , both ideally pathogen-based , are needed .
Leptospirosis is a zoonosis of worldwide distribution caused by pathogenic species of the spirochete Leptospira . The genetic diversity of Leptospira is increasingly recognized . Currently there are 9 species of known pathogenicity to humans or animals , 5 of unclear clinical significance [1] . Humans usually become infected with Leptospira through direct or indirect exposure to the urine of infected wild or domestic animals , which is common in less developed countries wherein humans and animals often live in close proximity and sanitation is poor [2]–[5] . The median incidence of leptospirosis in humans in such countries is as high as 975 per 100 , 000 population per year . Longer survival of leptospires in the environment in warm , humid conditions contributes to the higher incidence of leptospirosis in tropical versus temperate regions and to the peak in leptospirosis infections in the rainy season in tropical countries and during the summer in temperate regions [6] . In ever increasing recognition of its public health importance , leptospirosis has now been classified as an emerging or reemerging infectious disease by the Centers for Disease Control and Prevention and the World Health Organization . In recognition of the disproportionate impact on rural and urban residents living in resource-poor setting , the WHO now classifies leptospirosis as a neglected zoonotic disease . In the past twenty years , large outbreaks of human leptospirosis associated with occupational [7] and recreational exposures [3] , heavy seasonal rainfall [8] , [9] , and natural disasters have been detected in many countries , including in Nicaragua [10]–[13] . These outbreaks are often initially attributed to other pathogens and are recognized only after severe and/or unusual clinical illness prompts an epidemiologic investigation [8] , [11] , [14] , [15] . The occurrence and extent of illness related to unrecognized epidemics or sustained transmission between epidemics often is underappreciated because of limited resources for surveillance and difficulties with diagnosis . However , rigorous prospective study of unselected patients with fever coupled with near complete convalescent follow-up identified leptospirosis as a leading but unsuspected cause of acute febrile illness in Sri Lanka [16] , and leptospirosis could be of similar importance in Nicaragua . The presence of leptospirosis in animal reservoirs in Nicaragua and elsewhere in Central America has been recognized for >50 years [17] , [18] . The annual incidence of leptospirosis in Nicaragua has been estimated to be 23 . 3 per million; however , the actual rate is uncertain , since few cases are identified and confirmed [3] , [6] . To determine the burden of and risk factors associated with unrecognized leptospirosis in Nicaragua , we conducted a prospective study of children and adults presenting acutely with fever to a large teaching hospital from a region and time without a recognized epidemic .
Study doctors verified eligibility and willingness to return for a 2–4 week convalescent follow-up visit and obtained written informed consent from patients ( ≥18 years ) or parents ( <18 years ) , and assent if aged 12–17 years . The institutional review boards of Johns Hopkins University and Duke University Medical Center ( USA ) as well as Universidad Nacional Autonoma de Nicaragua , León ( Nicaragua ) approved the study . We recruited patients in the emergency department and adult and pediatric wards of Hospital Escuela Oscar Danilo Rosales Arguello ( HEODRA ) , the 400-bed primary public teaching hospital of Universidad Nacional Autonoma de Nicaragua in León , Nicaragua . Between August 2008 and May 2009 , we enrolled consecutive febrile ( ≥38°C , tympanic ) patients ≥1 month old without prior ( within 1 week ) trauma or hospitalization who presented during the day or early evening hours Monday through Saturday . Dedicated study doctors verified eligibility and willingness to return for follow-up and obtained written informed consent from patients ( ≥18 years ) or parents ( <18 years ) , and assent if 12–17 years . Study personnel recorded structured epidemiological and clinical data , including duration of illness and clinical provider's presumptive diagnosis , on a standardized form and then obtained specimens for on-site clinician-requested testing and off-site research-related testing . Patients returned for clinical and serologic follow-up 2 to 4 weeks later , or were visited at home if they did not return and could be located . Blood was centrifuged and sera frozen on site at −80°C . Serum and EDTA-anti-coagulated blood samples were stored promptly at −80°C . Samples were shipped on dry ice to and within the United States to diagnose acute leptospirosis infections . Sera were tested for the presence of specific anti-leptospiral IgM antibodies by ELISA ( Institut Viron Serion GmgH , Warburg , Germany ) after removal of rheumatoid factor at Johns Hopkins University per the manufacturer's instructions . The assay provided qualitative results — positive , negative , and equivocal ( borderline positive/negative ) . Using a standard curve and evaluation table provided with the kit , optical density ( OD ) measurements were adjusted for plate-to-plate variation with a correction factor to yield quantitative results that correlated with titers . [19] . We tested serum samples with the MAT to confirm the diagnosis of leptospirosis serologically at Yale University [20] . Twenty-five reference strains were used , which represented 6 pathogenic and one non-pathogenic species ( Table S1 ) . Patoc , a non-pathogenic strain , was used as a marker for possible infections with serovars not included in the panel . The presumptive infecting serogroup was determined based on the serovar against which highest agglutination titers were directed . DNA was prepared from 1 mL of archived EDTA-anticoagulated blood with the automated QIAsymphony SP system ( Qiagen Inc . , Valencia , CA ) at Johns Hopkins . Quantitative real-time PCR for leptospirosis was performed at Yale University using 5′ nuclease ( TaqMan ) assay and primers that amplified a sequence of lipL32 , a gene that is exclusively present in pathogenic Leptospira , as has been described previously [21] . Duplicate samples detected within 40 PCR cycles and Sanger sequenced to confirm amplification of lipL32 gene from Leptospira species were considered positives . We tested paired sera by IgM ELISA to identify a subset of patients with probable and possible acute leptospirosis to be confirmed by MAT and/or PCR ( Figure 1 ) . In those with probable acute leptospirosis ( IgM seroconversion by ELISA and/or the equivalent of a 4-fold rise in IgM titer ) , paired sera were tested by both MAT and PCR to confirm acute leptospirosis . In those with possible acute leptospirosis ( stable , decreasing , or less than 4-fold rise in titer ) , paired sera were tested by MAT only if the convalescent-phase sera screened positive ( titer 200 for a pathogenic serogroup ) by MAT . In this latter group with possible acute leptospirosis , PCR was performed only for MAT-confirmed acute leptospirosis . Finally , PCR was performed in a subset of patients with unlikely acute leptospirosis ( convalescent serum positive by ELISA but negative by MAT ) . We defined a confirmed case of acute leptospirosis according to PCR and MAT criteria as proposed by the Center for Disease Control and Prevention in 2013 ( http://wwwn . cdc . gov/NNDSS/ ) . Briefly , in a patient with probable or possible acute leptospirosis identified by ELISA , confirmation required either a positive leptospirosis PCR result and/or seroconversion by MAT ( defined as a negative acute-phase titer and convalescent-phase titer of ≥200 , a 4-fold rise in titer by MAT between paired samples , or a single acute-phase MAT titer of ≥800 ) . We compared proportions by the Chi-square test or Fisher's exact test and continuous variables by Student's t-test or the Wilcoxon rank sum test if not normally distributed . Confidence intervals for risk ratios were calculated by exact methods . We assessed IgM in the acute-phase sample for seroprevalence and clinical impression versus paired-sera testing for acute leptospirosis . We correlated epidemiologic features , duration of illness , and symptoms and signs with serologic results and performed univariable and multivariable logistic regression . We chose the multivariable model explaining the largest variance and with the lowest Akaike information criterion ( AIC ) . Analyses were performed with Stata IC 11 . 0 ( StataCorp LP , College Station , TX , USA ) .
IgM ELISA testing for leptospirosis was completed for 800 ( 97 . 0% ) of 825 consecutively enrolled patients . Of these 800 , 748 ( 90 . 7% ) had paired sera available , since 52 patients did not return and could not be located for follow-up . The likelihood of a subject returning for convalescent serum sampling and clinical follow-up did not differ by age ( p = 0 . 90 ) , sex ( p = 0 . 93 ) , or self-reported urban vs . rural residence ( p = 0 . 53 ) . The reported median distance from residence to hospital was 2 km ( interquartile range [IQR] 2–20 ) for those who followed up versus 3 km ( IQR 2–30 ) for those who did not ( p = 0 . 08 ) . Among the 748 patients with paired sera , the median age was 9 years ( IQR 3–29 ) . Slightly more were male ( 52 . 5% ) , and males were younger than females ( median age 9 vs . 11 , p = 0 . 007 ) . The median reported duration of fever and of illness at presentation was 2 days and 3 days ( IQR 1–4 and IQR 1–5 , respectively ) . Many ( 30 . 0% ) reported taking an antibiotic before presentation . The median interval between acute and convalescent follow-up was 15 days ( IQR 14–28 ) . Testing paired sera by IgM ELISA identified 37 cases of probable acute leptospirosis , including 25 with seroconversion by ELISA ( Figure 1 ) . Notably , the acute-phase sera from all 25 cases with seroconversion by ELISA were also seronegative by MAT , as were an additional 2 acute-phase sera with equivocal IgM ELISA results from patients with paired ELISA results suggestive of a 4-fold rise in titer . Among the 37 cases of probable acute leptospirosis , 26 ( 70 . 3% ) were confirmed ( 13 by MAT alone , 7 by MAT and PCR , and 6 by PCR alone ) and 11 unconfirmed ( no acute leptospirosis by MAT and PCR ) . Among the 20 ( 54 . 1% ) of 37 cases of probable acute leptospirosis that were confirmed by MAT , 11 were seroconversions by ELISA . Among the 6 cases of probable acute leptospirosis confirmed by PCR alone , 5 were seroconversions by ELISA . Among the 221 cases of possible acute leptospirosis identified by ELISA , 177 had convalescent sera available for testing by MAT . Acute leptospirosis was ruled out by MAT in 159 of the 177 patients , of whom 30 were tested additionally by PCR and all 30 negative . However , acute leptospirosis was confirmed in 18 ( 10 . 2% ) patients with possible acute leptospirosis ( 14 by MAT alone , 4 by MAT and PCR ) . Hence , screening by ELISA with confirmatory MAT and PCR identified a total of 44 ( 6 . 3% ) with confirmed acute leptospirosis and excluded the diagnosis in 660 patients , with only 44 of the original 748 patients with paired sera not tested by MAT . Clinicians suspected leptospirosis in only five ( 11 . 4% ) patients with acute leptospirosis ( sensitivity and specificity of clinical impression 11 . 4% [95% CI 3 . 8—24 . 6] and 99 . 7% [95% CI 98 . 9—100 . 0] , respectively , with receiver operating characteristic curve ( ROC ) area 0 . 55 ( 95% CI 0 . 51–0 . 60 ) . If a positive acute-phase IgM result were used to diagnose acute leptospirosis , 27 true infections would be detected ( sensitivity 62 . 8% , 95% CI 46 . 7—77 . 0 ) but 137 patients would be diagnosed erroneously ( specificity 77 . 0% , 95% CI 73 . 4–80 . 3 , ROC area 0 . 699 [95% CI 0 . 62–0 . 77] ) . Among 20 probable and 18 possible cases of acute leptospirosis confirmed by MAT , only 3 were positive in the acute-phase sera ( 2 associated with a 4-fold rise in titer and 1 with stable high titer ) . In comparison , PCR on acute-phase blood specifically identified 7 of 20 ELISA-probable MAT-confirmed acute leptospirosis infections , an additional 6 ELISA-probable acute leptospirosis infections negative by MAT , and 4 of 18 ELISA-possible MAT-confirmed acute leptospirosis infections . All positives by PCR were confirmed to amplify the lipL32 gene . Epidemiologic and clinical characteristics of patients with and without acute leptospirosis are detailed in Tables 1 and 2 , respectively . Those with and without acute leptospirosis had similar durations of illness ( median 3 . 5 [IQR 2–6] vs . 3 [IQR 1–5] days , p = 0 . 10 ) and times to convalescent follow-up ( 14 days [IQR 14–31] versus 15 days [IQR 14–28] , p = 0 . 73 ) . Patients with acute leptospirosis were older ( median 18 years , IQR 10–37 ) than febrile patients without acute leptospirosis ( median 9 years , IQR 2–27 ) , p = 0 . 0009 ( Table 1 and Figure 2 ) . The proportion of patients with leptospirosis that were male vs . female was similar ( 55% vs . 53% , respectively , p = 0 . 82 ) . Acute leptospirosis was a more common cause of fever in rural than in urban residents ( 12 . 6% vs . 3 . 8% , p<0 . 001 ) . Patients reporting pig exposure were both more likely to report rural residence ( 60% vs . 40% , p<0 . 001 ) and more likely to have confirmed acute leptospirosis ( 45% vs . 23% , p = 0 . 001 ) . However , rural residence was associated with a statistically significant increased risk of acute leptospirosis even in the absence of reported pig exposure ( 10 . 2% vs . 3 . 2% , p = 0 . 003 ) . Patients with acute leptospirosis were more likely to report fresh water exposure than were others ( 36 . 4 vs . 10 . 3% , p<0 . 001 ) , which was predominantly river exposure ( 34 . 1% vs . 9 . 7% , p<0 . 001 ) . Acute leptospirosis occurred throughout the year , but 11 . 3% of the febrile cohort had acute leptospirosis in rainy May to October ( median rainfall 364 . 6 cm , IQR 145 . 0–539 . 8 cm ) versus 4 . 8% of the cohort in drier November to April ( median 0 cm , IQR 0–1 . 2 cm ) , p = 0 . 003 . The largest number of acute leptospirosis cases occurred in October and November , when leptospirosis accounted for 20% ( 13/66 ) and 14% ( 10/70 ) cases of acute febrile illness , respectively ( Figure 3 ) . Fresh water exposure was also associated with acute leptospirosis ( reported in 36 . 4% with leptospirosis vs . 10 . 33% without leptospirosis , p<0 . 001 ) , and fresh water exposure was reported as commonly ( by 11 . 4% of patients ) in the wet season as in the dry season . Clinical features associated with acute leptospirosis are detailed in Table 2 . The duration of reported fever was similar in patients with and without acute leptospirosis ( median 2 days , p = 0 . 09 ) , but chills were more frequently reported in those with acute leptospirosis ( 86 vs . 60% , p<0 . 001 ) . Those with acute leptospirosis were more likely to report headache , joint pain , and muscle pain and less likely to report cough than other patients . Most ( 89 . 2% [33/37] ) patients with acute leptospirosis and headache also reported chills . Nearly all ( 92 . 9% [26/28] ) patients with acute leptospirosis and joint or muscle pain reported both , with joint pain alone and muscle pain alone equally infrequent ( 3 . 6% [1/28] for each ) . In contrast , there was less complete ( 73 . 7% [151/205] ) overlap in patients without acute leptospirosis , with muscle pain alone ( 17 . 6% [36/205] ) more common than joint pain alone ( 8 . 8% [18/205] ) among patients reporting either symptom . Physical findings in patients with acute leptospirosis were largely similar to those without acute leptospirosis , and conjunctival suffusion was not observed . Patients with acute leptospirosis had lower leukocyte counts ( median 9 , 700 vs . 11 , 875/µL , p = 0 . 004 ) with lower lymphocyte counts ( median 1826 , vs . median 2452/µL , p = 0 . 003 ) . Hemoglobin concentrations were similar , but platelet counts lower in those with acute leptospirosis ( median 216 , 000 vs . 277 , 000/µL , p = 0 . 02 ) . In a multivariable logistic regression model in which potentially statistically significant ( p≤0 . 10 ) epidemiologic and clinical features in univariable analyses were eligible for inclusion , rainy season , rural residence , fresh water exposure , headache , joint pain , and absence of cough were independently associated with confirmed acute leptospirosis . In the final model including these variables , the odds of acute leptospirosis in the rainy vs . the dry season was 2 . 10 ( 95% CI 1 . 03 , 4 . 31 , p = 0 . 04 ) , with rural vs . urban residence 2 . 14 ( 95% CI 1 . 05 , 4 . 37 , p = 0 . 04 ) , and with exposure to freshwater vs . no such exposure 2 . 89 ( 95% CI 1 . 36 , 6 . 15 , p = 0 . 006 ) . The presence of headache and joint pain conferred 2 . 86 ( 95% CI 1 . 15 , 7 . 14 , p = 0 . 02 ) higher odds and 2 . 52 ( 95% CI 1 . 22 , 5 . 18 , p = 0 . 01 ) , respectively , than no such exposure . In contrast , the presence of cough was associated with a lower odds ( OR 0 . 23 ) of acute leptospirosis ( 95% CI 0 . 09 , 0 . 57 , p = 0 . 002 ) . Of those with confirmed acute leptospirosis , 12 ( 28% ) reported taking an antibiotic before presentation , including a second generation cephalosporin ( 8 ) , third-generation cephalosporin ( 2 ) ; fluoroquinolone ( 1 ) , and chloramphenicol ( 1 ) . Patients with acute leptospirosis were as likely to be admitted to hospital as others but the length of stay was shorter ( median 2 days , IQR 1 . 5–3 . 5 vs . 4 days , IQR 2–6 , p = 0 . 008 ) . Among 647 patients with data recorded , 10 ( 24 . 4% ) of those with leptospirosis were treated with penicillin or amoxicillin vs . 187 ( 30 . 9% ) of others ( p = 0 . 38 ) . No patient with leptospirosis received a tetracycline ( vs . 1% of others , p = 0 . 60 ) . Nearly all were asymptomatic at convalescent follow-up , similar ( p = 0 . 11 ) to those without leptospirosis ( 80 . 1% ) . No one with leptospirosis died , but 9 of the 10 deaths in the study occurred before follow-up . Among those who died , the acute-phase serum was IgM-negative in 7 , positive in 2 , and equivocal in 1 .
This study identifies leptospirosis as a major ( 6 . 3% ) and clinically unrecognized cause of acute febrile illness in Nicaragua , and provides strong evidence for either sustained high-level transmission of leptospirosis or an unrecognized epidemic . Evidence that leptospirosis is perhaps endemic in the area includes the identification of new ( acute ) infections throughout the 10-month study period during both rainy and dry months . The presence of leptospirosis in domestic and wild animal reservoirs was first documented in Nicaragua in 1962 [17] , [18] , but disease in humans was not evaluated . Agriculture accounts for 20% of Nicaragua's GDP and employs more than 29% of the workforce ( data: USDA foreign agricultural service ) , so frequent human exposure to leptospires would be expected . However , the threat to human health was not recognized until 1995 , when epidemic “hemorrhagic fever” without jaundice or renal manifestations was reported in rural Nicaragua , including the region surrounding our study area , following heavy rains and flooding and a comprehensive investigation implicated leptospirosis . Two months later a cross-sectional sero-survey identified IgM anti-Leptospira antibodies in 85 ( 15% ) of 566 persons studied , but only 25 ( 29 . 4% ) reported febrile illness during the preceding two months . Investigators concluded that the epidemic's attack rate was 15% , and that a large proportion of outbreak-related leptospirosis infections were asymptomatic . However , those sero-positive could have had leptospirosis before the epidemic , since IgM antibodies can persist for a year in 40% [18] . We hypothesized that both the importance of leptospirosis as a cause of acute febrile illness and the clinical spectrum of leptospirosis in Nicaragua were underappreciated . Because of a uniquely high ( 90% ) rate of follow-up , we were able to identify acute infections and reliably distinguish them from past infections by testing paired sera . We identified acute infections in 6 . 3% ( 44 ) of the febrile cohort , which is likely a conservative estimate since most of our acute cases were seroconversions detected at a median of 15 days follow-up and 10% of patients seroconvert ≥30 days after onset of illness [22] . The impact of early treatment with a penicillin , recorded for over 30% of patients without acute leptospirosis , is unknown . Use of antimicrobial agents can alter the clinical course and or serologic response in patients with leptospirosis [23] . Further , we did not test all cases of possible leptospirosis by PCR , or those screen-negative by ELISA using MAT or PCR . However , it is unlikely that many cases were missed . First , the proportion of acute leptospirosis confirmed by MAT was much higher in those with probable vs . possible leptospirosis ( 54 . 1% vs . 10 . 2% ) and in the former group only 6 were MAT-negative but PCR-positive . Second , all 30 MAT-negative ELISA-possible cases tested by PCR were negative . By rigorous diagnosis of acute leptospirosis based on paired serology , we identified headache , chills , muscle pain , and joint pain as frequent and discriminatory symptoms in contrast to physical signs , and headache , joint pain , and absence of cough as independent predictors of acute leptospirosis in our final multivariable model . Headache , chills , and musculoskeletal pain were associated with epidemic acute leptospirosis in Nicaragua in 1995 [12] . Although WHO's recommended case definition includes headache and muscle pain , muscle pain and joint pain were highly correlated in our dataset; joint pain was retained in the final model because it better discriminated between acute leptospirosis versus other acute febrile illness ( not the intended use of the WHO case definition ) . Conjunctival suffusion , often touted as a sensitive and specific feature , was not identified in our patients with acute leptospirosis . However , disease was relatively mild , as evidenced by absence of end-organ involvement ( no oliguria , jaundice , hemorrhage , lung crackles , altered mental status , neck stiffness ) , relatively normal vital signs , and nearly normal complete blood counts . Although dyspnea , cough , and hemorrhage can be observed with acute leptospirosis , these features were infrequent in our patients in contrast with those with leptospirosis-associated pulmonary hemorrhage in Nicaragua in 1995 and more recently in Brazil [12] , [15] . Although anicteric undifferentiated febrile illness with infreqent respiratory symptoms contrasts with the severe disease reported in some recent outbreaks [13] , [24] , it is not unexpected in early ( median 3 days ) leptospirosis in unselected febrile patients . In Peru , seroconversion was also associated with symptomatic but not severe illness [25] . We were also able to prospectively evaluate potential epidemiologic risk factors for infection suggested by retrospective studies , and found that rainy season , rural residence , and fresh water exposure ( related to non-tap water source or swimming , bathing , or wading ) were independent predictors of acute leptospirosis . Heavy seasonal rainfall has been associated in multiple studies with outbreaks of leptospirosis [8] , [9] . Studies in Brazil and Peru found both urban slums and rural residence to be associated with increased risk of leptospirosis [9] , [15] , [25]–[27] . The 1995 Nicaragua epidemic case-control study found walking through creeks , household rodents , and ownership of dogs with titers ≥400 to Leptospira were independently associated with illness [12] . Similar to our finding that fewer patients with leptospirosis reported drinking water from a tap , the post-epidemic 1995 Nicaragua serosurvey identified indoor water source as independently associated with protection from infection [28] . Strengths of this study include the prospective study design , large sample size , inclusion of an unselected population with fever , near complete follow-up , epidemiologic and clinical correlation , and extensive testing to confirm acute leptospirosis . We prospectively assessed epidemiologic and clinical data and used objective criteria to sequentially enroll a large cohort of febrile patients six days a week to prevent recall bias and minimize selection bias . By obtaining both acute and convalescent sera in over 90% , we avoided misclassification of previously exposed individuals as current clinical cases , a frequent occurrence when only acute sera are tested . Further , those from whom paired sera were not available did not differ significantly from the population included . That patients were well on follow-up is consistent with the lack of late complications reported by others [29] . Limitations include the paucity of general laboratory data , such as liver function tests and serum chemistries; however , the disease spectrum suggests that the clinical utility of non-specific laboratory tests would be low . Further , this reflects routine clinical practice in the public sector in Nicaragua . Our estimate of leptospirosis could also be low if a wider array of serogroups and serovars is circulating than are detected by the ELISA and MAT panel used , but the ELISA detects a large number of serogroups and serovars and a broad MAT panel was used . Confirmation of probable and possible cases by MAT allowed us to identify likely circulation of multiple serogroups , including Autumnalis , Ballum , Bataviae , Australis , Canicola , Djasiman , Hebdomadis , Icterohaemorrhagiae , Mini , Pomona , and Pyrogenes , and Sejroe in our population , in contrast to another setting such as an urban slum in Brazil . Confirmation of circulation of these serogroups would require culture for leptospirosis , which was not possible in this study . MAT reactivity against serogroups Ballum and Icterohamorrhagiae suggest exposure to rats ( maintenance hosts for these serogroups and/or mice ) ; rat exposure was very common ( 70% ) in the population studied . MAT reactivity against Canicola was also observed , and might be expected with 60% of patients exposed to dogs . Notable was the significant association between pig exposure and leptospirosis and of reported pig exposure and antibodies against pig-associated serogroups of leptospires , such as Bataviae , Australis , and Pomona . In conclusion , we found that leptospirosis accounted for 6 . 3% of acute febrile illness in Nicaragua and was more common during rainy months , in rural residents , and after exposure to freshwater . Consistent with mild non-specific acute febrile illness , leptospirosis was clinically not suspected . Clinical diagnosis , single acute-phase IgM ELISA , and single acute-phase MAT were poor predictors of confirmed acute leptospirosis with PCR reasonably sensitive and also specific . The lack of agreement between testing acute sera vs . paired sera [16] underscores the need for improved point-of-care , especially pathogen-based , diagnostics . The occurrence of incident ( acute ) leptospirosis throughout the 10 months of the study suggests that leptospirosis is endemic in the region surrounding León or that unrecognized epidemics occur . A population-based incidence study is needed to define the full spectrum of disease . Nonetheless , the high proportion of febrile patients with acute infection underscores the importance of public health interventions to reduce transmission , which should include education about risk factors for leptospirosis [7] , such as exposure to fresh water , and control of animal reservoirs . | Leptospirosis , transmitted by pathogenic species of the bacterium Leptospira , is distributed worldwide but infections due to unrecognized epidemic or endemic disease are under-appreciated . We prospectively studied patients ≥2 years of age who presented with acute febrile illness in Nicaragua and systematically collected detailed information about exposures and features of the illness as well as serum and blood to confirm acute infections . Among 704 patients with paired sera tested by MAT , we found acute leptospirosis in 6 . 3% ( 44/704 ) . Patients with acute leptospirosis were more likely to present during rainy months and to report rural residence and fresh water exposure . Leptospirosis is a common ( 6 . 3% ) but unrecognized cause of acute febrile illness in Nicaragua . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"global",
"health"
] | 2014 | Unsuspected Leptospirosis Is a Cause of Acute Febrile Illness in Nicaragua |
Characterizing the link between small-scale chromatin structure and large-scale chromosome folding during interphase is a prerequisite for understanding transcription . Yet , this link remains poorly investigated . Here , we introduce a simple biophysical model where interphase chromosomes are described in terms of the folding of chromatin sequences composed of alternating blocks of fibers with different thicknesses and flexibilities , and we use it to study the influence of sequence disorder on chromosome behaviors in space and time . By employing extensive computer simulations , we thus demonstrate that chromosomes undergo noticeable conformational changes only on length-scales smaller than 105 basepairs and time-scales shorter than a few seconds , and we suggest there might exist effective upper bounds to the detection of chromosome reorganization in eukaryotes . We prove the relevance of our framework by modeling recent experimental FISH data on murine chromosomes .
Understanding how genomes fold within the crowded environment of the nucleus [1] during interphase represents a necessary step for the comprehension of important cellular processes such as gene expression and regulation [2] . The combined results of high-resolution microscopy [3 , 4] and mathematical and computer modelling [5 , 6] seem to suggest that genomes are organized hierarchically [7 , 8] . Each genome is partitioned into a set of single units , the chromosomes , and each chromosome is made of a single filament of DNA complexed around histone octamers to form a necklace-like fiber ≈ 10 nm thick known as the 10nm chromatin fiber . In in vitro conditions close to the physiological ones , this fiber is observed to fold into a thicker , more compact structure known as the 30nm fiber [1] , whose role and existence in vivo are nonetheless still quite debated [9 , 10] . On larger scales , chromosome conformation capture ( 3C ) techniques [2] have shown that chromosomes appear organized in Topologically Associated Domains ( TADs ) of sizes ranging from ≈ 0 . 1 to ≈ 1 megabasepairs ( Mbp ) . Chromosome loci within TADs interact frequently between themselves , but less frequently across different TADs . Finally , chromosomes do not spread inside the whole nucleus , rather they occupy well localized nuclear regions ( the so-called “chromosome territories” ) which play a crucial role in gene expression and regulation [11] . While much progress on the causal relationship between chromosome structure and function has now been made , many fundamental aspects remain still obscure . One of these crucial issues concerns the link between chromosome ( re ) organization at various length and time-scales and the spread of inhomogeneties in the sequence of the chromatin fiber which may arise from ( 1 ) selective epigenetic marks induced by chemical modifications of the histone tails [12] , ( 2 ) nucleosomes arrangements in discrete nanodomains of different sizes [4] , and ( 3 ) selective stimulation of particular kinds of genes [13] or entire gene families [14] . These events will produce modifications in the local polymer properties of the chromatin fiber ( as its persistence length or the local compaction ratio ) which might affect in turn the whole hierarchical folding of chromosome organization [9 , 10] . Motivated by these considerations , we present here the results of Molecular Dynamics computer simulations of a minimal polymer model for interphase chromosomes in order to quantify to what extent the simultaneous presence of chromatin fibers of heterogeneous composition ( different thicknesses and flexibilities ) is able to generate observable effects on the small- and large-scale structures and motions of the associated chromosome . The proposed model complements previous work by one of the authors [15–18] concerning the description of chromosome folding in terms of fundamental polymer physics . Similarly to other recent works discussing the explicit role of sequence disorder in chromatin [19] and chromosome behaviors [20 , 21] , here we “deviate” from the description of the chromatin filament as a homopolymer and we discuss sequence effects in space and time through the introduction of controlled amounts of disorder in the chromatin sequence . In this way , we provide a quantitative description for many crucial aspects concerning the structure and dynamics of interphase chromosomes which are “spontaneously” driven by the physical properties of the underlying chromatin sequence with a definite copolymer structure . In particular , by considering the two “extreme” cases of chromosomes made of: ( 1 ) short stretches of a thinner , more flexible fiber randomly interspersed in a “sea” of thicker fiber and ( 2 ) chromosomes partitioned into two distinct blocks of thinner and thicker fibers , we show that significative spatial and dynamical rearrangements of chromatin loci appear to be restricted to limited contour lengths ( up to ≈ 105 basepairs ( bp ) ) and times scales ( up to few seconds ) . Interestingly , there exists a limited range ( 104−105 bp ) of chromatin contour lengths where chromosome expansion is accompanied by an increase ( rather , than a decrease ) of the associated contacts between the fibers . We apply this framework to rationalize the outcome of recent experiments which employ fluorescent microscopy to monitor conformational changes of chromosomes that occur upon transcription activation or chromatin decondensing in mouse embryonic stem cells [13] . Finally , we argue that the effects discussed here are not the consequence of the details of the model , but involve more general aspects of polymer physics .
The results of our three case studies are summarized in Fig 1 . It is visible that only length-scales smaller than L ≈ 0 . 1 megabasepairs ( Mbp ) are affected with 〈R2 ( L ) 〉 expanding sensibly more than in the situation where chromosomes are composed only of 30nm fiber ( panels A , C , E ) , while the behaviour at large scales remains unaffected . Moreover , in the case where the unfolded sequences are grouped into a single cluster ( middle and bottom panels ) , there is no dependence on their spatial positioning with respect to the center of mass of the corresponding chromosome territory . Insensitivity of large scales to changes at small ones is also confirmed ( panels B , D , F ) by the analysis of contact frequencies , 〈pc ( L ) 〉 , whose trend remains , in particular , compatible with the experimentally observed power-law 〈pc ( L ) 〉 ∼ L−1 [27] . Interestingly , instead of decreasing as expected from panels A , C and E , corresponding contact frequencies in the limited range [0 . 01 Mbp − 0 . 1 Mbp] also increase as a function of the 10nm fiber content ( see in particular panel B ) . In order to understand this result , we use the mathematical relationship [16] between average contact frequencies 〈pc ( L ) 〉 and distribution function pL ( R ) of internal distances R ( L ) given by: 〈 p c ( L ) 〉 = ∫ 0 r c p L ( R ) 4 π R 2 d R ∫ 0 ∞ p L ( R ) 4 π R 2 d R , ( 1 ) where rc = 60 nm is the cut-off distance for two monomers to form a contact . In particular , eq 1 implies that pL ( R ) should also increase as a function of the 10nm fiber content at given L . The inset of panel B confirms this behavior for L = 0 . 015 Mbp . This specific trend is due to the decreasing of the chromatin “effective persistence length” ( from ≈ 104 to ≈ 102 basepairs for model chromosomes entirely made of 30nm and 10nm fibers , respectively ) following from increasing amounts of 10nm fibers , which allows genomic loci to contact each other with enhanced probability . For a more quantitative view on chromosome reorganization at small scales , the same data for 〈R2 ( L ) 〉 and 〈pc ( L ) 〉 were normalized to corresponding values for chromosome conformations made of 30nm fiber , see S1 Fig . We highlight in particular the pronounced peaks in the left panels corresponding to a maximum volume expansion of ≈ 30 , and we pinpoint again the increasing of contact frequencies in the aforementioned interval [0 . 01 Mbp − 0 . 1 Mbp] . The latter , in particular , constitutes a result with non trivial experimental implications . First , in connection to some recently proposed protocols for the reconstruction of chromosome conformations based on 3C ( reviewed in [28] ) where a monotonous relationship between chromatin distances and contacts is often assumed , our finding demonstrates that some caution is needed in order to avoid systematic bias in the final reconstructed structure . Second , the recent puzzling result [24] where chromatin domains with a high propensity to form 3C contacts seem to undergo rather pronounced decondensation when monitored by using FISH can be understood by considering that the two experimental techniques sample very different intervals of the corresponding pL ( R ) ’s: close to the average value or the median for FISH , around to the lower tail for 3C techniques , as shown in the insets of Fig 1B , 1D and 1F . Our work thus supports the important conclusion of references [24 , 29] , namely that one needs to take into account both kinds of data to reconstruct correctly the shape of chromatin domains . We complete the discussion by considering the full distributions pL ( R ) for L = 0 . 003 , 0 . 015 and 3 Mbp , see Fig 2 . For L = 0 . 003 and L = 0 . 015 Mbp ( panels A-F ) there are quantitative differences between the cases where the 10nm fiber is located at sparse random positions along the chromosome and where they form a single chromatin cluster , while for L = 3 Mbp ( panels G-I ) all distributions show no noticeable difference . In panels A-F , the largest peak corresponds to the most probable value of spatial distances between loci on the 30nm fiber region . Additionally , for random locations of 10nm fiber ( panels A and D ) corresponding pL ( R ) ’s show a broader population of R values , while in the other cases there exist smaller peaks corresponding to the most probable distance between loci on the 10nm fiber region . Interestingly , similar distribution functions for spatial distances between chromatin loci seem to have been reported in yeast ( see panels B , C , D , and E of Fig 1 in reference [30] ) . Although a direct comparison between these experimental results and our data is not possible ( our setup applies to large chromosomes , like mammalian ones ) , we are tempted to speculate that the results reported in reference [30] are a manifestation of the presence of chromatin fibers of different compositions . Alternatively , we may interpret the observed shape of the pL ( R ) ’s in terms of a bias towards large spatial distances: in that respect , we report that a similar feature has been observed in human and mouse chromosomes [14 , 24] ( including the experimental data that will be discussed in this work ) . We have then considered those chromosome configurations with single long sequences of 10nm fiber and we have calculated 〈R2 ( L ) 〉 and 〈pc ( L ) 〉 on these sequences , see Figs 3 and 4 . Our results demonstrate that 〈R2 ( L ) 〉 increases systematically with respect to the analogous quantity for model chromosomes made of 30nm fiber , with no dependence on the specific location of 10nm fiber along the chromosome . This is also shown by the decreased contact probability . This insensitivity to positioning follows from the proposed picture [31] that chromosomes resemble a uniformly dense , semi-dilute solution of branched polymer rings . We expect then chromatin filaments to be similarly constrained or accessible regardless of their position along the genomic sequence . In order to prove that the reported volume increase is not a fortuitous coincidence of the chosen sequence , we have also repeated the analysis on the same genomic region for the model chromosome with no 10nm fiber ( 0% ) and the model chromosome entirely made of 10nm fiber ( 100% ) . Both 〈R2 ( L ) 〉 and 〈pc ( L ) 〉 differ quantitatively from the case in which just one part of the chromosome is allowed to swell . Similarly to the previous section , the same results can be recast in terms of ratios to the corresponding quantities calculated for chromosome conformations with no 10nm fiber , see S2 and S3 Figs , which allows to appreciate to what extent chromosomes may effectively reorganize . Because of the copolymer structure of our model chromosomes , the two types of fibers have different stiffnesses and thicknesses which cause the thinner fiber to move away from the thicker one . Consequently , 〈pc ( L ) 〉 calculated for the 10nm fiber sequence tends to decrease and shows a scaling behavior ∼L−1 . 18±0 . 05 for L in the interval 0 . 01 − 1 Mbp which , being slightly steeper than the one for the entire chromosome , suggests an increase in the volume spanned by the fiber . Qualitatively , volume differences between chromosome regions with different transcription activities have been reported in a recent study on Drosophila [12] . Here the authors have studied three regions of the Drosophila genome , which , according to the histone modifications that they were bearing , were classified as active , inactive and polycomb-repressed . They found that the polycomb-repressed region was the most compact , followed by the inactive region , while the active region was the least compact . These results encourage us to speculate that the difference between the active and inactive regions could be due to different polymer physical properties induced by the histone modifications . Here we discuss the impact of chromatin unfolding on the dynamics of the corresponding genomic loci . Specifically , we have considered the mean-square displacement 〈 δ r 2 ( τ ) 〉 ≡ 〈 ( r → i ( t + τ ) - r → i ( t ) ) 2 〉 at lag time τ , where r → i ( t ) is the spatial position of monomer i at time t and we implicitly assume average over specific monomer positions along the chromatin chain . In fact , this is an observable widely employed in many experiments monitoring the dynamic activity of specific chromatin loci , being especially suitable for comparing genome behavior in response to changes of the environment [32] , or when the cell is targeted with drugs which are able to activate selectively certain types of genes [14] . Fig 5 summarizes our results for 〈δr2 ( τ ) 〉/τ vs . τ for the two cases of random positioning of small filaments of 10nm fiber within the chromatin fiber ( panel A ) and for chromosomes made of two large separated domains with different fiber composition ( panel B ) . In both cases , we notice a general increase of chromatin mobility as larger and larger portions of 30nm fiber unfold , and , at larger times , a trend which does not substantially depend on the small scale details of chromatin fiber . Not surprisingly , data at short times reflect in part the discussed results for chromatin structure: in particular , we notice that differences in chromatin mobility before and after chromatin unfolding in random locations are only visible below time-scales of about 5 seconds ( panel A ) . Slightly larger discrepancies are observed in the other situation where chromosomes are organized as two separate domains ( panel B ) . In this latter case , unfolded chromatin loci move on average more than folded ones , the latter displaying the same motion than in the case of the homogeneously folded chromosome ( compare green vs . blue lines ) . In a recent work [13] , Therizols et al . used FISH to show that , in embryonic stem cells , chromosome decondensation is sufficient to alter nuclear organization . Two sets of experiments were done: first , a viral transactivator was used to activate transcription of three genes ( Ptn1 , Nrp1 and Sox6 ) . Alternatively , repositioning of the same genes was also observed after the treatment with an artificial peptide ( DELQPASIDP ) which decondenses chromatin without inducing transcription . By comparing the measured spatial distances between the two ends of each selected sequence , the authors concluded that nuclear organization is driven mainly by chromatin remodeling rather than transcription . We have tested the predictions of our model by using data from these experiments . We have simulated the unfolding of a chromatin region of genomic length corresponding to the size of the specific gene we wanted to mimic . This situation corresponds to the second case studied in this work , namely the unfolding of a chromatin cluster . For a fair comparison , we have processed the experimental data as follows: we have reconstructed first the probability density distribution function for spatial distances between the ends of each gene , for the control condition ( denoted as eGFP ) and cells in which DELQPASIDP is recruited to the chromosome loci ( denoted as DEL ) . Since FISH distances are recorded as two-dimensional vectors projected on the confocal plane , for the purpose of comparison a ( large ) set of three-dimensional distances with equivalent 2d projections was generated numerically by assuming random orientations of the 3d vectors in relation to the axis orthogonal to the confocal plane . It can be observed in Fig 6 panels A , C and E that , upon chromatin decondensation , the peak and the shape of the distribution change dramatically , in particular the distributions become wider . The same effect is displayed by our simulations ( panels B , D and F ) . This comparison thus validates our result that larger genes tend to expand when unfolded . While our model describes reasonably well the trend of the experimental data , the median values and the part of the distribution covering the larger distances , it seems to perform worse for the part of the distribution describing the small distances . This could be due to several factors which , to maintain the model simple , have been neglected . These include either protein linkers that physically bridge two sites along the genes forming a loop [20 , 33 , 34] , either chemical modifications which could change the charge on some histones , thus inducing an effective electrostatic attraction [21] between regions of the gene that would make it more compact . Moreover , for the experiments using DEL , not all cells receive the exact same quantity of plasmid . That is , the experimental distributions of distances could be biased towards small R values because , in some cells , there was not enough plasmid to induce decondensation . The choice for the physical parameters employed in this work as fibers flexibilities , thicknesses and excluded volume interactions ( see section “Materials and Methods” ) was mainly motivated by reasons of simplicity . Understanding to what extent our results depend on these specific choices requires a deeper analysis . In order to isolate and quantify the effects of one single parameter , we have considered a simpler model chromosome where monomers have fixed nearest neighbor distance along the sequence equal to 30 nm . The polymer was then split in two complementary domains: the fiber in one domain has the same features of the 30nm fiber , while the physical properties of the fiber in the other domain are changed , in turns , to: ( 1 ) completely flexible fiber; ( 2 ) thickness equal to 10 nm; ( 3 ) larger strength for the corresponding excluded volume interaction . Each of these changes was applied , in turns , to a domain occupying 20% , 50% and respectively 100% of the polymer . S4 Fig ( top panels ) shows that , not surprisingly , the persistence length influences 〈R2 ( L ) 〉 and 〈pc ( L ) 〉 only at small scales up to 0 . 1 Mbp , suggesting that one of the main effects of having a heterogeneous chromatin composition consists in a global change of the overall persistence length of the chromatin fiber . By varying instead the range of the excluded volume interactions without modifying the nominal persistence length we produce variations in 〈R2 ( L ) 〉 and 〈pc ( L ) 〉 similar to the ones reported in Figs 3 and 4 ( middle panels ) . Viceversa , when varying the strength of the interaction no change is observed ( bottom panels ) . These observations suggest a leading role for the effective range of the excluded volume interaction: in our model , the thinner 10nm fiber tends to occupy a larger region because its own self-repulsion becomes less important when compared to the repulsion from the thicker 30nm fiber . Quite interestingly , a similar effect concerning the change in the excluded volume was observed in equilibrium simulations of generic block copolymers [35] . Finally , we stress that this effect is less visible in Fig 1 because of the base pair content assigned to the monomers: since the base pair content of the 10nm fiber region is in general low compared to 30nm fiber region , its effect is not visible when averaging over the entire chromosome .
In this article , we have presented results of extensive Molecular Dynamics computer simulations of a coarse-grained polymer model for interphase chromosomes that extends previous numerical work [15–18] by introducing a crucial ingredient which was neglected before: namely , the presence of two kinds of chromatin fibers of different thickness and stiffness mutually interacting inside their own chromosome territory . Starting from a chromosome configuration made of a single and homogeneous 30nm fiber like filament , we have monitored chromosome spatial and temporal behaviors when this conformation is altered by the introduction of increasing amounts of 10nm-like fiber through the controlled unfolding of the thicker 30nm fiber . The work shows that there exists detectable chromosome ( re ) organization for spatial scales smaller than 0 . 1 Mbp ( Fig 1 ) and time-scales shorter than just a few seconds ( Fig 5 ) . Quite interestingly , these findings appear systematic and do not depend on the size of the chromosome region affected by the phenomenon of local unfolding or by its location along the chromosome . Interestingly , these results tend to suggest that experimental methodologies like FISH or HiC might be of little or no help in distinguishing between fibers of different compaction , unless they investigate genomic distances smaller than 0 . 1 Mbp . This prediction can be tested , for instance , by employing the recently developed oligonucleotide based FISH probes which seem to provide the necessary fine resolution [36 , 37] . An important conclusion of our work is that , although our model uses parameter values that can be associated to the traditional “10nm/30nm” chromatin fiber paradigm [9 , 10] , our results reflect a generic physical effect [35] which ought to be observable in more general systems of crumpled polymers constituted of fibers with different thickness and/or stiffness ( see section “Polymer physics aspects” and S4 Fig ) . Our simulation protocol compares qualitatively well ( see Fig 6 ) with experimental results on chromosome reorganization in mouse embryos treated with a synthetic transcription factor [13] which produces selective activation of a specific gene . In particular , we predict the observed shifting for distribution functions of spatial distances . Of course , it is quite possible that other mechanisms may explain the experimental results equally well . In fact , intentionally our model tends to neglect other important aspects of chromosome folding which have been highlighted recently by other authors: sequence-specific attractive interactions [21] , protein linkers between chromatin fibers [20] , mechanisms of active regulation [38 , 39] , “loop extrusion” [33 , 34] involved in the reorganization of small chromosome domains , or the anchoring to the nuclear envelope or other nuclear organelles [40–42] . In the lack of a more quantitative analysis , we can only speculate on the fact that the inclusion of these mechanisms into our model might alter significantly the conclusions sketched here . On the other hand , this should represent an important stimulus for investigating further and more quantitatively the delicate relationship between chromosome structure and function .
In this work , the chromatin fiber is modeled as a coarse-grained polymer chain , with monomer-monomer interactions described by analytical expressions similar to the ones used in our previous works [15–17] and suitably adapted to take into account the different sizes of 30nm and 10nm monomers . The full Hamiltonian governing the system , H , consists of three terms: H = ∑ i = 1 N [ U F E N E ( i , i + 1 ) + U b r ( i , i + 1 , i + 2 ) + ∑ j = i + 1 N U L J ( i , j ) ] ( 2 ) where N ( see Table 1 ) is the total number of monomers constituting the ring polymer modeling the chromosome ( see Section “Construction of model chromosome conformation” for details on this point ) and i and j run over the indexes of the monomers . The latter are assumed to be numbered consecutively along the ring from one chosen reference monomer . The modulo-N indexing is implicitly assumed because of the ring periodicity . By taking the nominal monomer diameter of the 30nm chromatin fiber , σ = 30 nm = 3000 bp [16] , as our unit of length , the vector position of the ith monomer , r → i , the pairwise vector distance between monomers i and j , d → i , j = r → j - r → i , and its norm , di , j , the energy terms in eq 2 are given by the following expressions: As in our previous work [15–18] , chromosome dynamics was studied by performing fixed-volume Molecular Dynamics ( MD ) simulations with periodic boundary conditions at near-physiological fixed chromatin density ρ = 0 . 012 bp/nm3 . Note that periodic boundary conditions do not introduce confinement to the simulation box: using properly unfolded coordinates , the model chromatin fibers can extend over arbitrarily large distances [15] . The system dynamics was integrated by using LAMMPS [44] with Langevin thermostat in order to keep the temperature of the system fixed to 1 . 0kB T . Given the unit mass m30 = 1 of the 30nm-bead , we fixed the mass of the 10nm-bead to m 10 = 1 27 . The integration time step was fixed to tint = 0 . 001τMD , where τ M D = σ ( m 30 ϵ ) 1 / 2 is the elementary Lennard-Jones time . γ = 0 . 5/τMD is the friction coefficient [43] which takes into account the corresponding interaction with the background implicit solvent . The total length of each MD simulation run is = 3 ⋅ 105 τMD , with an overall computational effort ranging from a minimum of ≈ 103 to a maximum of ≈ 9⋅104 hours of single CPU for “0%” and “100%” model chromosome conformations , respectively . Single chromosome conformations were sampled every 103 τMD , implying 300 configurations per each run . We have verified that our rings are well equilibrated by considering the mean-square distances , 〈R2 ( L ) 〉 , and mean contact probabilities , 〈pc ( L ) 〉 , calculated on different sets of configurations log-spaced in time . As shown in S5 Fig for a 20% amount of 10nm fiber , all curves give the same results . Similar results are obtained for all simulated systems . Quantities discussed in this work as the mean-square distances , 〈R2 ( L ) 〉 , and the average contact frequencies , 〈pc ( L ) 〉 , between chromosome loci are plotted as a function of the genomic distance , L . In order to avoid unphysical behavior arising from the ring closure condition , we considered contour lengths L ≤ 1/4 of the total contour length of the ring , or L ≤ 30 Mbp . Possible numerical artifacts due to the presence of monomers with different degrees of resolution ( “10nm fiber” monomers vs . “30nm fiber” monomers ) were removed by averaging over all possible pairs of monomers at fixed genomic separations L and spatial resolution of the 10nm fiber: this was achieved by replacing all 30nm fiber monomers not already decondensed by the 27 equivalent 10nm monomers , as shown in Fig 7 ( panels C and D ) . In this way , each chromosome conformation always contributes with the same number of monomers and genomic distances smaller than 3 kbp can be effectively sampled . It is important to stress that what is described here concerns only the final analysis of the data , and it is not implicated in the motion of the monomers during the MD runs which is always performed as described in previous sections . Final values for 〈R2 ( L ) 〉 and 〈pc ( L ) 〉 at large L’s were obtained by averaging further over log-spaced intervals centered at the corresponding L’s . This procedure improves the accuracy by reducing considerably statistical fluctuations . Error bars were calculated accordingly . | A key determining factor in many important cellular processes as DNA transcription , for instance , the specific composition of the chromatin fiber sequence has a major influence on chromosome folding during interphase . Yet , how this is achieved in detail remains largely elusive . In this work , we explore this link by means of a novel quantitative computational polymer model for interphase chromosomes where the associated chromatin filaments are composed of mixtures of fibers with heterogeneous physical properties . Our work suggests a scenario where chromosomes undergo only limited reorganization , namely on length-scales below 105 basepairs and time-scales shorter than a few seconds . Our conclusions are supported by recent FISH data on murine chromosomes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"dynamics",
"chromosome",
"structure",
"and",
"function",
"materials",
"science",
"epigenetics",
"macromolecules",
"chromatin",
"fibers",
"materials",
"by",
"structure",
"polymers",
"polymer",
"chemistry",
"physical",
"properties",
"chromosome",
"biology",
"g... | 2016 | Large Scale Chromosome Folding Is Stable against Local Changes in Chromatin Structure |
Malaria transmission-blocking ( T-B ) interventions are essential for malaria elimination . Small molecules that inhibit the Plasmodium ookinete-to-oocyst transition in the midgut of Anopheles mosquitoes , thereby blocking sporogony , represent one approach to achieving this goal . Chondroitin sulfate glycosaminoglycans ( CS-GAGs ) on the Anopheles gambiae midgut surface are putative ligands for Plasmodium falciparum ookinetes . We hypothesized that our synthetic polysulfonated polymer , VS1 , acting as a decoy molecular mimetic of midgut CS-GAGs confers malaria T-B activity . In our study , VS1 repeatedly reduced midgut oocyst development by as much as 99% ( P<0 . 0001 ) in mosquitoes fed with P . falciparum and Plasmodium berghei . Through direct-binding assays , we observed that VS1 bound to two critical ookinete micronemal proteins , each containing at least one von Willebrand factor A ( vWA ) domain: ( i ) circumsporozoite protein and thrombospondin-related anonymous protein-related protein ( CTRP ) and ( ii ) vWA domain-related protein ( WARP ) . By immunofluorescence microscopy , we observed that VS1 stains permeabilized P . falciparum and P . berghei ookinetes but does not stain P . berghei CTRP knockouts or transgenic parasites lacking the vWA domains of CTRP while retaining the thrombospondin repeat region . We produced structural homology models of the first vWA domain of CTRP and identified , as expected , putative GAG-binding sites on CTRP that align closely with those predicted for the human vWA A1 domain and the Toxoplasma gondii MIC2 adhesin . Importantly , the models also identified patches of electropositive residues that may extend CTRP's GAG-binding motif and thus potentiate VS1 binding . Our molecule binds to a critical , conserved ookinete protein , CTRP , and exhibits potent malaria T-B activity . This study lays the framework for a high-throughput screen of existing libraries of safe compounds to identify those with potent T-B activity . We envision that such compounds when used as partner drugs with current antimalarial regimens and with RTS , S vaccine delivery could prevent the transmission of drug-resistant and vaccine-breakthrough strains .
Each year more than half a million people die from malaria , a disease caused by protozoan parasites in the genus Plasmodium . The life cycle of Plasmodium parasites includes asexual development in the human host and obligatory sporogonic development in the Anopheles mosquito vector with transmission from person to person only made possible through the bite of an infected anopheline . Despite substantial investment in malaria research , it is widely accepted that current interventions are insufficient to achieve the ultimate goal of eradication and that a combination of anti-malaria strategies including those that target parasite transmission give eradication efforts the best chance to succeed [1] . Moreover , the evolutionary capacity of vectors and parasites to overcome chemical- and drug-based interventions emphasizes the need for new weapons in the anti-malaria arsenal . It is in this context that malaria transmission-blocking ( T-B ) interventions ( vaccines and drugs ) have received significant attention [2] . In fact , recent progress has shown that probing the basic biology underlying mosquito-Plasmodium interactions can identify novel intervention targets not only in the parasite , but in the mosquito as well [3]–[6] . Importantly , seminal work by Delves , et al . [7] have brought increased attention to the potential T-B activity of drugs that failed to demonstrate efficacy against Plasmodium asexual stages but have been resurrected as novel T-B candidate compounds . These efforts highlight the need for T-B molecules and open new avenues for the development and/or repurposing of compounds that have direct activity against the parasite soon after ingestion into the mosquito midgut during blood feeding . In the mosquito blood meal , Plasmodium gametocytes differentiate into gametes and fuse to form zygotes , which then develop into motile ookinetes . For parasite development to continue , ookinetes must find and adhere to membrane-associated ligands on the midgut epithelial surface , a pre-requisite for cell invasion . Experimental evidence from Plasmodium berghei and Plasmodium falciparum suggests that ookinete attachment and invasion is mediated by micronemal proteins , including the circumsporozoite protein and thrombospondin-related anonymous protein-related protein ( CTRP ) [8]–[13] and von Willebrand factor A domain-related protein ( WARP ) [10] , [14] . The function of WARP is unclear , while CTRP has a demonstrated role in ookinete motility [9] , [15] . However , both are essential for midgut epithelial cell invasion by Plasmodium ookinetes . Once inside the cell , ookinetes make their way to the midgut basal lamina where they differentiate into oocysts , each giving rise to thousands of sporozoites that are released into the hemocoel upon maturation and rupture . Sporozoites are then swept into the circulating hemolymph and carried to the salivary glands . Following successful salivary gland invasion , sporozoites remain in the lumen of the salivary duct until host delivery during blood feeding . Clearly , negotiating the midgut tissue barrier in the vector is crucial for successful establishment of the parasite in the mosquito and hence , subsequent transmission to human hosts . In this study we exploit knowledge of crucial molecular interactions between Plasmodium ookinetes and the apical microvillar surface of the mosquito midgut to design proof-of-concept small molecules that interfere with ookinete attachment . Previous work demonstrated that sulfated glycosaminoglycans ( GAGs ) are present on both the apical and basal surfaces of the midgut epithelium , with chondroitin sulfate ( CS ) predominant on the apical side ( i . e . , facing the midgut lumen ) and heparan sulfate ( HS ) predominant on the basal side [5] , [16] ( Figure S1A ) . RNAi-mediated knockdown of the Anopheles gambiae peptide-O-xylosyltransferase , an enzyme that catalyzes the first step in CS and HS biosynthesis , resulted in mosquitoes with CS-depleted midgut apical surfaces [5] . When infected with P . falciparum and P . berghei in feeding assays , these mosquitoes demonstrated significantly lower oocyst infection intensities relative to controls . The study also showed binding affinity for two types of CS ( CS-A and CS-E ) in mature ookinetes , consistent with a previous report demonstrating that the ookinete micronemal proteins PfWARP and PfCTRP bind to sulfated GAGs in vitro [10] . Inspired by the idea that these molecular interactions could be disrupted by small molecules that mimic the charged structural elements involved in ligand binding , two short-chain , water-soluble compounds were synthesized for in vitro and in vivo T-B studies based on their potential to interfere with parasite protein-GAG interactions ( Figure 1A ) . Here we describe efforts to test this strategy with the underlying hypothesis that when a mosquito ingests these small molecules in an infectious blood meal , the compounds will interfere with ookinete-GAG interactions , therefore preventing midgut invasion and abrogating subsequent developmental steps in sporogony . Based on key studies in the literature [9]–[15] , we further hypothesize that the molecular basis of ookinete-GAG interactions , and hence those between ookinetes and our GAG-mimetics , involve the Plasmodium micronemal proteins CTRP and/or WARP . Our findings showed that this novel strategy dramatically reduced infection intensity in the mosquito midgut , and multiple lines of evidence suggest that the mechanism underlying the T-B effect involved binding of our GAG-mimetic decoy to one or more von Willebrand factor A ( vWA ) domains found in the protein CTRP .
Synthesis of polysulfonated polymers ( VS1 and VS2-PVP ) proceeded with the addition of 100 mg of potassium persulfate to 5 g ( 0 . 38 mmol ) of vinyl-sulfonic acid ( VS1 ) sodium salt water solution and adjusted to basic pH with sodium hydroxide . The final solution was warmed for 20 hr at 80°C , then cooled to room temperature ( RT ) , diluted with water and ultra-filtered through a membrane with a nominal cut-off of 10 , 000 Da . The fraction retained was freeze-dried and the product obtained was a white powder . Size exclusion chromatography was used following different reaction times to obtain oligomers of different length and molecular mass . These compounds were then purified by ultra-filtration through different cut-off membranes ( 500 Da , 1000 Da , 5000 Da ) , and average molecular weights were measured by size exclusion chromatography and MALDI spectrometry . Transmission-blocking assays for vinyl-sulfonic acid compounds ( including preliminary VS1 and VS2-PVP experiments ) were tested using both in vivo and in vitro systems . In vivo studies were performed using the murine malaria parasite P . berghei ( ANKA 2 . 34 ) following IACUC approved protocols . For each experiment , two to three naïve , donor mice ( Swiss Webster , 20–24 g ) were inoculated ( i . v . ) with blood stage P . berghei and then checked for parasitemia by blood smear five to six days later . Once parasitemia reached ≥10% , donors were sacrificed via cardiac puncture and parasitemic blood was used to inoculate ( i . v . ) eight to ten experimental mice per test compound . Two to three days post-inoculation , experimental mice were smeared and checked for exflagellating gametocytes . Mice demonstrating an average of at least 1 and fewer than 6 exflagellations per 40× field were assigned to a treatment group , weighed and anesthetized . For each mouse , a pre-injection 500 ml cup of Anopheles stephensi mosquitoes ( n = 50 ) were allowed to feed for 15 to 20 min . The mouse was then removed from the mosquito cup , injected with either a vinyl sulfonic acid compound ( 250 µg/ml or 500 µg per 24 g body weight ) , polyvinylpyrrolidone ( PVP , same dose as VS1or VS2-PVP ) , or the carrier only ( PBS ) via tail vein injection ( iv ) and then allowed to recover for 10 to 15 min . Following recovery , a post-injection cup of mosquitoes ( n = 50 ) was allowed to feed as before . Unfed mosquitoes were then removed from both pre- and post-injection cups via mouth aspiration . For each control and test compound , three to five pre- and post-injection sets of mosquitoes were maintained on sucrose and water for 10 days at 19°C , 80% relative humidity . On day 10 , midguts were dissected from all surviving mosquitoes and stained with 0 . 1% mercurochrome for 20 mins . Oocyst number for each midgut was determined by microscopy and at least three independent experiments were performed for each compound . In vitro studies were performed using the human malaria parasite P . falciparum and the old-world malaria vectors Anopheles gambiae and An . stephensi as described [4]–[5] . With the exception of the parasite load experiment , each set of studies consisted of independent experiments in which the age of the gametocyte culture ( 16–17 days ) , the age of mosquitoes ( 4–6 days ) , and the blood-meal gametocytemia ( 0 . 3% ) and hematocrit ( 45% ) were kept consistent . In the parasite-load experiment , all else was the same except for the blood-meal gametocytemia which varied as described in the Results . For each experimental treatment , VS1 or the control compound ( PVP ) were prepared in PBS and diluted 1∶10 to the final experimental dose with infected blood . Full-length PvWARP excluding the signal peptide ( nt 88–867 ) and a fragment of PvCTRP containing the first vWA domain ( nt 79–921 ) were PCR-amplified from genomic DNA of the Salvador I strain with a C-terminal 6×His tag appended to the reverse primer . Fragments were cloned into the EcoRV sites of the vector pEU-E01-MCS ( Cell Free Sciences , Matsuyama , Japan ) . The PvCTRP-vWA1 and PvWARP were expressed in the wheat germ cell-free expression system ( Cell Free Sciences , Matsuyama , Japan ) as described [17] and purified using Ni-affinity chromatography . Biotinylated VS1-NH2 in a volume of 100 µl ( 10 µg/ml ) was applied to each well of a streptavidin-coated ( 2 µg/ml ) 96-well microtiter plate that had been blocked with PBS , 1% BSA ( Thermo Pierce ) and incubated for two hours at RT . During VS1 incubation , 6×His-tagged recombinant PvWARP and PvCTRP ( 5 µg/ml ) were each mixed separately with heparin and CSA ( 100 µg/ml ) in blocking buffer and incubated for 2 hr at RT . The microtiter plate was subsequently washed three times with PBS to remove excess VS1 , and then 100 µl of 6×His-tagged recombinant PvWARP or PvCTRP alone ( 10 or 5 µg/ml ) or of the recombinant protein-GAG mixture was added to each well , with the exception of the no-protein and irrelevant-protein controls . The latter received a 6×His-tagged recombinant glycosyltransferase from An . gambiae . Binding and/or competition with VS1 was allowed to proceed for 2 hr at RT . Following three washes with PBS , anti-His MAb ( Sigma ) was added to each well and incubated for 1 hr at RT . After three washes with PBS + Tween-20 ( 0 . 05% ) , anti-mouse secondary antibodies conjugated to HRP were added and incubated for 1 hr at RT . Following another wash step , TMB ELISA substrate ( Pierce ) was used for detection . VS1 binding was quantified by measuring the OD at 450 nm with a SPECTRA MAX PLUS microplate reader ( Molecular Devices ) . Ookinete samples were fixed with 4% paraformaldehyde and prepared for immunofluorescence microscopy by washing three times with PBS containing 0 . 1 M glycine ( rinsing buffer ) . To permeabilize samples the parasites were incubated with rinsing buffer containing 0 . 2% Triton X-100 for 10 min and then washed as before . After the washes , samples were incubated in rinsing buffer for 30 min and blocked with PBS containing 0 . 05 mM glycine , 0 . 2% fish skin gelatin and 0 . 05% sodium azide for 2 hr . The samples were then incubated with biotinylated-VS1 and anti-Pbs21 ( P . berghei ) or anti-Pfs28 ( P . falciparum ) for 1 hr at RT or overnight at 4°C . Cells were washed as before and incubated with Streptavidin , DyLight 488 conjugated ( Thermo ) , and Goat Anti-Rabbit IgG ( H+L ) , DyLight 594 conjugated , for 1 hr at RT . Following incubation , the cells were washed three times with rinsing buffer , resuspended in PBS , spotted on slides and allowed to air dry . ProLong Gold antifade reagent with DAPI ( Invitrogen ) was added prior to the coverslip and slides were incubated for 24 hr at RT protected from light . Samples were examined with SPOT software using a Nikon Upright E800 microscope . Homology modeling of the CTRP vWA domain was performed using SWISS-MODEL in two different modes of operation [18]–[21] . In the full-automated mode , the Toxoplasma gondii micronemal protein 2 I domain ( 2XGG , chain B , residues 75–212 ) was selected as the optimal template to calculate the CTRP model ( residues 1–148 ) and is based on 22% sequence identity . In the template identification mode and using the InterPro Domain Scan method [22] , the von Willebrand factor A1 domain ( 1AUQ , chain A , residues 1276–1463 ) was selected as the optimal target template ( residues 1–193 ) . The model quality was assessed using the QMEAN server [23] and the Z-score . To determine significance between treatment and control groups in the feeding assays , the nonparametric Mann-Whitney U test was used due to the non-normal distributions typical of oocyst counts . For the in vitro assays , the test was performed comparing the distribution of oocyst counts per midgut for each treatment group to that of the PVP control , followed by a Bonferroni correction of z-scores to adjust for multiple tests . For the in vivo assays , the test was performed comparing oocyst counts per midgut between mosquitoes fed on P . berghei-infected mice pre- and post-injection with VS1 , PVP , or PBS alone . All experimental studies using vertebrate animals ( mice ) were performed in accordance with Johns Hopkins University ( JHU ) ACUC ( Animal Welfare Assurance #A3272-1 ) regulations . The Animal Protocol ( #MO12H232 ) used for these studies was reviewed and approved by the JHU ACUC and are in compliance with the United States Animal Welfare Act regulations and Public Health Service ( PHS ) Policy . No human subject research was performed during this study .
Our primary goal was to design synthetic polymers that can mimic sulfated CS-GAGs that have been shown to bind ookinetes ( Figure 1A , Figure S1A ) . However , the likelihood of CS-GAGs , with C4S and C6S sulfation on the midgut microvillar surface was challenged [16] . Using capillary electrophoresis with laser-induced fluorescence detection , we confirmed the presence of both C4S and C6S chondroitin GAGs on An . gambiae midgut brush border microvilli vesicles ( Figure S1A , Text S1 ) . Based on these data , we hypothesized that polymers with high sulfation densities would be appropriate for our study . As such , two polysulfonated polymers were generated by polymerization of vinyl-sulfonic acid ( VS1 ) and copolymerization of vinyl-sulfonic acid with 1-vinyl-2-pyrrolidone ( VS2-PVP ) . The sulfate groups on the polysulfonated polymers are anionic at physiologic pH and would presumably bind to ookinete proteins that would naturally bind to GAGs on the midgut surface . Note that it was previously shown that the blocking phenomenon is predicted to occur at the apical midgut surface , and that basal lamina GAGs do not influence the ultimate read-out of these T-B studies , which is oocyst prevalence and intensity measurements at 8 or 10 days post-blood feeding , for P . falciparum and P . berghei , respectively [5] . Initial data from malaria T-B studies indicated that VS1 was non-toxic and well tolerated by both mice and mosquitoes . Although VS2-PVP was tolerated by mice , mosquitoes that ingested the compound had poor survivorship in multiple experiments within 24 hrs following blood feeding . This low survivorship prevented a proper comparison of infection intensity between midguts dissected from pre- and post-injection mosquitoes; and consequently , this compound was not pursued further . In preliminary in vitro standard membrane feeding assays ( SMFAs ) , VS1 demonstrated 98 . 5% and 92 . 3% inhibition of P . falciparum oocyst development in An . gambiae and An . stephensi , respectively . In vivo direct feeding assays ( DFA ) with mice infected with P . berghei demonstrated a somewhat lower effect of the compounds on parasite development in An . stephensi , with VS1 reducing oocyst intensity relative to controls by 77% . These preliminary data from both in vitro and in vivo malaria models demonstrated the potential for VS1 to act as a potent T-B compound . Moreover , to exclude the possibility that micro- and macrogamete fertilization events could be adversely affected by VS1 , we tested the effect of the compound on male microgamete exflagellation and noted that the number of exflagellation centers were unaffected ( data not shown ) . We therefore pushed this small molecule forward as the lead compound for further testing , which included ( i ) assays to determine if T-B activity varies according to polymer length , ( ii ) dose-ranging assays to estimate the IC50 of VS1 , ( iii ) ELISA-based binding and competition assays using a candidate gene approach , specifically recombinant versions of the ookinete micronemal proteins CTRP and WARP , and ( iv ) immunofluorescence microscopy to confirm binding of VS1 to wild-type ookinetes from both P . falciparum and P . berghei , as well as ookinetes from knockout lines of P . berghei . To assess the influence of polymer length , the product following VS1 synthesis was fractionated into three molecular-weight categories by size exclusion chromatography , VS1-10 , 000; VS1-3 , 000; and VS1-1 , 000 ( Figure 1A ) . Based on preliminary dose ranging experiments , each new compound was then tested at a concentration of 250 µg/ml using both in vitro ( SMFA ) and in vivo ( DFA ) malaria models . Polyvinylpyrrolidone , a non-sulfated , neutrally charged control ( PVP ) , which represents the unsulfated VS1 backbone , was used as a control . In SMFAs the three VS1 compounds were tested in parallel against P . falciparum in two replicate experiments using two different vector species , An . gambiae and An . stephensi . In all four experiments , each of the VS1 compounds significantly reduced oocyst intensity ( Figure 1B–E ) . VS1-3 , 000 consistently performed the best; inhibiting oocyst development by 86 . 0%–99 . 0% in An . gambiae ( Figure 1 B , C ) and 88 . 0%–93 . 5% in An . stephensi ( Figure 1 D , E ) . Experiments with the in vivo system yielded similar results , as all three VS1 compounds strongly inhibited P . berghei oocyst development in An . stephensi ( Table 1 , Figure 2 , Table S1 ) . Results from two replicate experiments per compound showed that all six VS1 treatment groups experienced a highly significant reduction in median oocyst intensity when comparing mosquitoes fed on pre-injection mice with those fed on mice injected with VS1 ( Table 1 , Table S1 ) . In five of these treatment groups , oocyst development was inhibited by >90% . VS1-3 , 000 had the strongest effect , demonstrating ≥98 . 0% inhibition on average in both experiments ( Table 1 , Table S1B ) . Conversely , pre- and post-injection comparisons of the oocyst burden in mosquitoes fed on mice from PBS and PVP groups demonstrated no consistent effect of either the VS1 carrier or the unsulfated polymer ( Table 1 , Table S1 ) . To confirm that VS1 was available to mosquitoes in blood meals , and hence the likely cause of T-B activity , the presence of VS1 in the mouse bloodstream following injection was confirmed by HPLC ( Figure S2 , Text S1 ) . Due to its consistent T-B activity , VS1-3 , 000 was selected for use in P . falciparum NF54 SMFA experiments to assess the compound's effectiveness across variations in parasite load in the blood meal ( gametocytemia ) and to estimate the IC50 of VS1 in two different anopheline vectors . For the parasite-load experiments , we chose to test VS1 potency at levels of gametocytemia that captured values routinely observed during the conduct of membrane feeding assays in the field [24]–[25] . With the concentration of VS1-3 , 000 set at 250 µg/ml , a SMFA was performed in which four gametocyte concentrations were tested in the presence and absence of VS1 . In this experiment , a day 17 gametocyte culture at 3 . 0% gametocytemia was pelleted and the packed red blood cells ( RBCs ) were diluted with uninfected blood to 0 . 3% and 0 . 1% gametocytemia . Each of these dilutions was in turn diluted 1∶10 with uninfected blood yielding 4 concentrations that ranged from 0 . 01%–0 . 3% gametocytemia ( ∼800–24 , 000 stage V gametocytes per µl of blood ) . The level of gametocytemia commonly used in laboratory-based SMFAs ( 0 . 3% ) is typically much higher than that found in the field to ensure consistent and robust infections in mosquitoes , allowing more rigorous tests of T-B activity [25] . Though a widely accepted approach , a criticism of the SMFA is that the assay better tests the effects of compounds ( or antibodies ) on oocyst intensity than prevalence of infection among mosquitoes . Since the ultimate goal is to reduce the latter to zero , we wanted to perform the SMFA over a range of gametocytemias once we established that VS1 consistently reduces the oocyst burden at the usual gametocytemia of 0 . 3% . In this set of experiments , VS1-3 , 000 once again demonstrated a potent reduction in oocyst intensity at 0 . 3% gametocytemia , reducing the median oocyst number per midgut from 92 . 0 to 8 . 0 ( Figure 3A ) . However , the prevalence of infection was unchanged between carrier-only and VS1 treatments at this level of gametocytemia . Interestingly , as the gametocytemia was reduced from 0 . 3% , the effect of VS1 on infection prevalence increased while the reduction in oocyst intensity remained high ( Figure 3A , B ) . In fact , at the two levels of gametocytemia most relevant to the field ( i . e . , 0 . 03% and 0 . 01% ) , the median oocyst number was reduced to 0 in the VS1 treatments while prevalence was reduced from 89% and 67% in carrier-only treatments to 21% and 13% in VS1 treatments , respectively ( Figure 3B , C ) . In other words , at levels of gametocytemia where untreated mosquitoes averaged fewer than 5 oocysts per midgut and where most mosquitoes were infected , VS1 treatment reduced infection prevalence 4–5 fold and infection intensity by 10 fold . Under these conditions , most VS1-treated mosquitoes were uninfected , while the few that were tended to have a single oocyst . In a set of two dose-ranging experiments with An . gambiae and two with An . stephensi , VS1-3 , 000 was fed to mosquitoes in infectious blood meals using serially diluted concentrations from 400 µg/ml to 12 . 5 µg/ml . The four experiments revealed a consistent pattern of percent inhibition characterized by a linear increase from little to no inhibition at 12 . 5 µg/ml to approximately 80% at 100 µg/ml and then a plateau >90% at concentrations >200 µg/ml ( Figure 3D ) . From these data the IC50 of VS1-3 , 000 was approximated to be 25 µg/ml . Since VS1 is a hypothesized structural mimetic of midgut-microvillar sulfated GAGs , we sought to determine whether VS1 can directly bind to Plasmodium ookinetes . To this end , we used biotinylated VS1-NH2 ( Figure 1A ) to probe non-permeabilized and permeabilized P . berghei ookinetes generated in vitro , as well as ex vivo blood-meal derived P . falciparum ookinetes isolated from dissected mosquito midguts . Only permeabilized Plasmodium ookinetes showed strong binding affinity to VS1 by immunofluorescence microscopy . The VS1 staining pattern suggested that it is not associated entirely with the ookinete surface since it did not consistently bind to non-permeabilized ookinetes nor did it localize with the abundant ookinete surface marker P28 ( also called Pbs21 in P . berghei ) in either P . falciparum ( WTPf ) or P . berghei ( WTPb ) ( Figure 4A , 4B ) . VS1 binding appears to be centrally and apically localized in the cytoplasm , suggestive of interaction with micronemal proteins since these proteins are not constitutively secreted to the ookinete surface and are stored in the micronemes [10] , [26] . In addition to ookinetes , VS1 also bound to P . falciparum retorts ( i . e . , developing ookinetes ) ( data not shown ) , and a portion of permeabilized P . falciparum and P . berghei round cells . These cells are likely to be zygotes or unfertilized macrogametes ( Figure S3E–L ) since they are stained with P28 , a surface marker well described in the literature , which is expressed from macrogametes to early oocyst [27]–[29] ruling out the possibility that these cells are P . berghei or P . falciparum gametocytes . Furthermore , P . falciparum stage IV and V gametocytes are not round but have a distinctive elongated morphology , and these cells did not stain withVS1 in subsequent immunofluorescence experiments targeting gametocytes ( Figure S4A–D , Text S1 ) . We cannot rule out an effect of VS1 in macrogametes and/or zygotes because those stages were not the focus of the current research , but any effect in these stages will potentiate the effect of VS1 against transmission of Plasmodium . We evaluated the effect of VS1 in the ability of microgametes to exflagellate in P . berghei and found that VS1 had no inhibitory effect on exflagellation between pre- and post-injection mice ( Table S2 ) . Furthermore , dissection of mosquitoes 24 hours after feeding with either PBS or VS1 showed presence of ookinetes in the blood meal ( data not shown ) . Nevertheless , more studies are needed to more thoroughly assess the effect of VS1 on macrogametogenesis , fertilization , and ookinete development . Because we are working with a T-B compound , targeting more than one stage of the parasite in the mosquito gut would strengthen the outcome of our final goal , completely blocking transmission of malaria . We used a candidate gene approach to identify potential targets or binding ligand ( s ) of VS1 among the repertoire of Plasmodium ookinete micronemal proteins [30] , focusing particularly on those with established roles in midgut attachment or invasion . The literature indicates that two such proteins , CTRP and WARP , bind to sulfated GAGs [10]; and although both are in the apicomplexan TRAP/MIC2 family of proteins , their domain architectures are quite different [30] . WARP is an approximately 40 KDa protein with a signal peptide and a single vWA domain , and we hypothesize that based on the published data WARP is secreted from the ookinete microneme and can thus work as an extracellular adaptor protein , potentially bridging the parasite surface ( or surface molecules ) with midgut apical membrane ligands . The much larger CTRP is approximately 230 KDa and contains a signal peptide followed by six contiguous vWA domains , seven contiguous thrombospondin ( TS ) domains , a transmembrane domain , and a short acidic cytoplasmic domain at the C-terminus that interacts with the motility actomyosin machinery [30] . The first four vWA domains of CTRP are more similar to one another than to vWA domains 5 and 6 when comparing six species of Plasmodium . Interestingly , a phylogenetic analysis of the vWA domains from TRAP , CTRP , and WARP among these species shows that WARP and CTRP form a single clade and that the vWA domain of WARP most recently shared a common ancestor with the fifth vWA domain of CTRP , suggesting that WARP evolved from CTRP [30] . We emphasize that the domain architectures and amino acid sequences of these two proteins are highly conserved across Plasmodia [10] , [30] and argue that VS1's potency against both rodent and human malaria , as well as its ookinete staining pattern as reported here , suggests that VS1's binding partner ( s ) is likewise highly conserved across Plasmodia . Thus given the above and the aim of identifying the mechanism of action for VS1 , we sought to investigate the binding activity of Plasmodium vivax CTRP and WARP to VS1 , following the argument that VS1 should bind to these two molecules . We produced soluble recombinant WARP and the first vWA domain of CTRP from P . vivax using a cell-free wheat germ system ( Figure S5 ) and evaluated binding affinity via ELISA . As expected , both recombinant PvCTRP and PvWARP bound to VS1 in a dose-dependent manner ( Figure 4C ) . To better delineate binding specificity , competition assays with heparin and chondroitin sulfate A ( CSA ) were performed . If VS1 binds primarily to the putative GAG-binding sites on the vWA domains of CTRP and WARP , then we would expect that heparin , and perhaps to a lesser extent CSA , at a concentration of 100 µg/ml should completely inhibit binding . However , we observed that both heparin and CSA only partially inhibited VS1 binding to PvCTRP and PvWARP ( Figure 4C ) , suggesting that VS1 binds to additional sites not used by heparin on either recombinant protein or that it binds to them with greater affinity . The ELISA results demonstrated that VS1 binds to recombinant WARP and the first vWA domain of CTRP in vitro , so to test binding in vivo we obtained three lines of P . berghei in which CTRP had been either completely [9] , [15] or partially knocked out [15] and compared VS1 staining patterns by immunofluorescence microscopy . One of the partial knockouts , a line known as ΔA6 , expresses CTRP that is missing all six of the vWA domains but contains all seven of the thrombospondin domains . Conversely , CTRP expressed by the other partial knockout line , ΔTS7 , includes the six vWA domains but lacks any of the thrombospondin domains . Since the recombinant CTRP protein used in the ELISA consisted only of the first vWA domain , we predicted a priori that the VS1 binding pattern to ΔTS7 ookinetes would be similar to wild type ookinetes , while the VS1 signal in both the CTRP knockout ( CTRPKO ) and ΔA6 lines would be diminished . However , if VS1 also binds to any of the TS domains in vivo , the VS1 signal would be much lower in the CTRPKO than in either of the partial knockouts . A caveat to this approach is that if VS1 also binds to WARP in vivo , we would expect some portion of the VS1 signal observed in wild type ookinetes to be shared among all of the knockout lines . Immunofluorescence microscopy images from CTRPKO ( Figure 4D ) indicate that VS1 binds to CTRP in permeabilized ookinetes . Furthermore , comparisons of staining patterns from the partial knockouts strongly suggest that VS1 binding involves vWA domains but not the TS domains ( Figure 4E , F ) . Furthermore , the apparent loss of VS1 signal in both the CTRPKO and ΔA6 lines also suggests that VS1 localizes to the micronemes ( as suggested by Figure 4B ) and that CTRP is the primary micronemal target of VS1and not WARP ( Figure 4D , F ) . These data further reconcile the observed cytoplasmic staining of putative zygotes/macrogametes ( Figure S3 ) with the previously reported staining of round forms with CTRP antisera [11] . Without a WARP knockout line we cannot rule out that binding to WARP may occur in vivo or that some of the T-B activity we observed is due to such an interaction . It should be noted that WARP expression/secretion is not well understood and may be temporally regulated or even midgut contact-dependent for different Plasmodium species . Thus , in vitro generated P . berghei ookinetes or P . falciparum ookinetes isolated from the blood-meal bolus may not express detectable levels of WARP . However , the CTRPKO and ΔA6 microscopy data are persuasive and we note that VS1 binding by ELISA is consistently stronger for the first vWA domain of CTRP than for WARP . Moreover , the phylogenetic relationship among CTRP and WARP vWA domains [30] in combination with the ELISA and immunofluorescence microscopy data reported here , suggest that VS1 primarily targets the first four vWA domains of CTRP . The divergence of vWA domains 5 and 6 and their evolutionary relationships with WARP suggest that these domains bind VS1 secondarily or not at all . In the absence of a CTRP crystal structure , we used homology modeling to predict heparin-binding sites on CTRP ( Figure 5A–E ) . The quality of the models was assessed with QMEAN; and models 1 ( PDB: 1AUQ , Figure 5B , C ) and 2 ( PDB: 2XGG , Figure 5 D , E ) had Qmean scores of 0 . 572 ( Z-score = −2 . 892 ) and 0 . 584 ( Z-score = −2 . 314 ) , respectively , indicating comparable model reliability . Model 1 is based on the human von Willebrand factor A1 domain , while Model 2 is based on the vWA-integrin like domain from the Toxoplasma gondii MIC2 protein . The structure of the former has been extensively studied due to its essential role in platelet adhesion [31]–[33] , while the latter is a well-described adhesin involved in host-cell invasion [34] . Despite the selection of markedly different templates , both with low sequence identity to CTRP ( <25% ) , both models had the same overall α/β Rossmann fold . The positively charged residues adopted similar patterns between the two models ( Figure 5 B–E ) , which were also in general agreement with the predicted heparin-binding domains on vWFA1 ( Figure 5A ) . A superposition of Model 1 with vWFA1 ( 1AUQ ) ( Figure S6A–B ) suggests that the two models predominantly differed in their overall length and the conformation of the loops connecting the beta sheet core and flanking alpha helices . A superposition of Models 1 and 2 also demonstrated a difference in overall length as well as the presence of a large alpha helix in Model 1 that is absent in Model 2 ( Figure S6C–D ) . Furthermore , a number of basic residues fall well outside the predicted vWFA1 heparin-binding regions in CTRP ( Figures 5A–E ) , which may represent an extended electropositive surface and additional binding sites for sulfated polymers such as heparin and VS1 .
Host-cell GAGs have been shown to be important mediators of Plasmodium development in its two hosts , including merozoite invasion of RBCs [35]–[36] , infected RBC sequestration to placenta [37] , ookinete invasion of the midgut [5] , and sporozoite invasion of mosquito salivary glands [38] and vertebrate hepatocytes [39] . Here we tested a strategy that exploits this feature of Plasmodium biology and demonstrated that VS1 , a putative GAG-mimetic , reduced midgut oocyst development by as much as 99% in mosquitoes fed with P . falciparum or Plasmodium berghei . Through direct-binding assays , we observed that VS1 bound to two ookinete micronemal proteins necessary for midgut invasion , each containing at least one vWA domain: ( i ) CTRP and ( ii ) WARP . By immunofluorescence microscopy , we observed that VS1 stains permeabilized P . falciparum and P . berghei ookinetes but does not stain P . berghei CTRP knockouts or transgenic parasites lacking the vWA domains of CTRP while retaining the thrombospondin repeat region . Finally , we used structural homology models of the first vWA domain of CTRP to identify residues likely involved in binding GAGs , as well as the VS1 compound . Based on these data , our working model for the mechanism underlying VS1's T-B activity is that it binds to CTRP once the protein is secreted from the micronemes of ookinetes prior to midgut attachment and invasion . CTRP is essential for gliding motility [9] , [15] and contains six vWA domains , which commonly play roles in cell adhesion to GAGs [8] , [10] , [30]–[34] . Our data suggest that VS1 either interrupts the gliding process on the midgut apical microvillar surface or coats the surface of the ookinete through its interaction with CTRP , thus preventing attachment to ligands ( e . g . , chondroitin sulfate [5] ) on the apical surface of the midgut epithelium . Nevertheless , due to observed binding of VS1 to permeabilized round cells , we cannot rule out that VS1 may potentially have an added benefit and affect additional parasite stages found in the blood meal , particularly macrogametes and/or zygotes . Further studies into these beneficial side effects are necessary . When designing the GAG-mimetic strategy , data from the literature suggested that sulfation density per disaccharide unit and the manner of presentation ( i . e . , how the underlying structure of the sugar scaffold influences the 3D projection of sulfated moieties ) are critical factors in inhibiting pathogen-GAG interactions . Boyle et al . [35] , for example , found that heparin and the E . coli-derived K5 polysaccharide inhibits merozoite entry into RBCs and that variations in the average number of sulfate groups/saccharide unit for K5 , which consists of glucoronate as opposed to iduronate , exhibited different inhibitory effects against merozoites , with sulfate densities >3/disaccharide producing the most potent IC50 estimates . Therefore , we sought to determine the minimal T-B polymer length of VS1 with the hope of minimizing the likelihood of diverse structural conformations that can occur with longer polymers , which could in turn affect the presentation/projection of anionic moieties . Although we were able to demonstrate that VS1-3 , 000 was the most effective polymer , we cannot , however , predict its structure . VS1-3 , 000 is unlikely to remain linear in solution or in the midgut after blood feeding . With this caveat in mind , we suspect that binding to the recombinant or native CTRP and WARP molecules may engender a specific VS1 conformation . Regardless , we expect that VS1 binding is largely due to the predicted GAG-binding motifs on the vWA domain ( s ) of CTRP and WARP [10] . However , the concentration of heparin used in our studies , which would otherwise result in the near complete inhibition of high affinity protein-GAG interactions [40]–[42] only reduced VS1 binding by ∼25% . It should be noted , however , that cases exist in the literature where soluble heparin cannot completely outcompete vWA domain-GAG interactions . For example , heparin-BSA binding to the vWA domain of PfTRAP can be competed between 45–66% using 50 µg/ml of soluble heparin and that a 10-fold increase in heparin concentration reduced binding to 9–27% of control [43] . Even more striking is a report that neither a 50- , 100- , or 1 , 000-fold molar excess of soluble heparin could completely inhibit binding between the PfTRAP vWA domain and the surface of HepG2 cells , which was thought to be GAG mediated [44] . In this set of experiments , each concentration reduced binding by approximately 15% , 55% , and 70% , respectively , suggesting that the PfTRAP vWA domain utilizes both GAG and a non-GAG binding sites . In combination , these data suggest that each recombinant protein in our study has either stronger binding affinity for VS1 than for either heparin or CSA , that the predicted GAG-binding regions do not completely explain the interactions of VS1 with PvCTRP or PvWARP , or more likely , a combination of these two scenarios . Tertiary structures of mammalian heparin-binding proteins have also been shown to enhance affinity and specificity [41] . We cannot rule out the possibility of cryptic GAG-binding sites on CTRP and WARP that provide cooperative binding to VS1 , as suggested at least in part by potential basic residue patches identified on two homology models of the first vWA domain of CTRP , which appears to be a primary ligand of VS1 . In terms of sulfation density and propensity to form various non-linear conformations , VS1 is clearly different from Heparin and natural GAGs . In this context , cooperative binding may be conferred by VS1 “wrapping around” CTRP and interacting with basic residues along different faces of the protein . The presence of potential additional binding sites suggests that CTRP ( as opposed to other GAG binding proteins ) can be specifically targeted by the next generation of VS1-based chemical mimetics . Clearly , a crystal structure for CTRP is needed to clarify the hypotheses generated by our two models . To date , the antimalarial pipeline is filled with compounds that act on related biochemical pathways ( e . g . , folate biosynthesis ) , which also increase the likelihood of the development of parasite cross-resistance to these “new” compounds . The need to discover drugs that act on unpredicted or uncharacterized biochemical pathways that are completely different from those associated with current antimalarials is paramount [45] . Our approach fits this mold , as it represents a completely novel mechanism of action compared to those associated with the existing , new , and now “rediscovered” list of antimalarials and T-B compounds [7] . Among the various T-B strategies , drugs offer a distinct advantage over vaccines since the efficacy of the compound is dose dependent and human immune-system independent , the latter being a potentially significant issue given that individuals in malaria endemic regions may suffer from malnourishment and concomitant infections by immune-modulating pathogens such as HIV and helminths . Although we have shown that VS1 is a potent T-B molecule , we emphasize that it cannot be used as a drug in its current form . However , we intend to use the data reported here to establish a high-throughput approach for identifying a next-generation “druggable” malaria T-B compound that would inhibit ookinete invasion of the midgut beyond that observed for VS1 ( i . e . , achieve zero infection prevalence among treated mosquitoes ) . We recognize , however , that if the next generation compound only replicates the T-B activity reported for VS1 and were used alone in the field , it would unlikely reduce infection prevalence in the mosquito population below the threshold necessary for sustained transmission . Nevertheless , it is widely believed that no anti-malarial intervention on its own will lead to regional elimination and eventual eradication [1] , [2] . We envision that in this context , such compounds may be valuable in a range of epidemiologic settings . Potential applications include ( i ) general use in conjunction with existing artemisinin combination therapies , which we emphasize do not kill stage V gametocytes , to prevent recurrent transmission from the treatment-seeking segment of the population , ( ii ) use in regions with unstable malaria ( e . g . , highlands ) to curb transmission during epidemics , ( iii ) use in combination with a T-B vaccine targeting sexual stage parasites to act as a safety net to “mop up” break-through parasites , and ( iv ) at the end game of the malaria eradication effort , as mass distribution of T-B compounds may offer a cost-effective approach to preventing asymptomatic , gametocytemic individuals , who would not otherwise seek treatment , from infecting anopheline mosquitoes , thus preventing resurrection of epidemic malaria transmission . | To achieve malaria elimination , the consensus expert opinion is that new approaches to drug and vaccine design are desperately needed . We have undertaken a novel , comprehensive approach towards the development of a malaria transmission-blocking drug based on the strategy of inhibiting Plasmodium development in the mosquito by interfering with obligate cellular interactions between the parasite and the mosquito-midgut epithelium . We have successfully designed a potent transmission-blocking small molecule ( VS1 ) that mimics the structure of molecules on the mosquito-midgut surface called glycosaminoglycans ( GAG ) , which are thought to serve as ligands for parasite attachment prior to cell invasion . Using assays in which mosquitoes were fed with infectious blood , we tested the effect of VS1 on Plasmodium development in the mosquito and found that the GAG mimic dramatically reduced the intensity of infection in the midgut . Binding experiments and immunofluorescence microscopy indicate that VS1 binds to the circumsporozoite- and TRAP-related protein ( CTRP ) , a micronemal protein expressed by ookinetes essential for midgut invasion . This interaction profoundly inhibits a key step of parasite development , thereby abrogating downstream events necessary for mosquito-to-human transmission . The work described lays the framework for bringing a truly novel transmission-blocking drug to fruition . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | A Small Molecule Glycosaminoglycan Mimetic Blocks Plasmodium Invasion of the Mosquito Midgut |
Mild mutations in BRCA2 ( FANCD1 ) cause Fanconi anemia ( FA ) when homozygous , while severe mutations cause common cancers including breast , ovarian , and prostate cancers when heterozygous . Here we report a zebrafish brca2 insertional mutant that shares phenotypes with human patients and identifies a novel brca2 function in oogenesis . Experiments showed that mutant embryos and mutant cells in culture experienced genome instability , as do cells in FA patients . In wild-type zebrafish , meiotic cells expressed brca2; and , unexpectedly , transcripts in oocytes localized asymmetrically to the animal pole . In juvenile brca2 mutants , oocytes failed to progress through meiosis , leading to female-to-male sex reversal . Adult mutants became sterile males due to the meiotic arrest of spermatocytes , which then died by apoptosis , followed by neoplastic proliferation of gonad somatic cells that was similar to neoplasia observed in ageing dead end ( dnd ) -knockdown males , which lack germ cells . The construction of animals doubly mutant for brca2 and the apoptotic gene tp53 ( p53 ) rescued brca2-dependent sex reversal . Double mutants developed oocytes and became sterile females that produced only aberrant embryos and showed elevated risk for invasive ovarian tumors . Oocytes in double-mutant females showed normal localization of brca2 and pou5f1 transcripts to the animal pole and vasa transcripts to the vegetal pole , but had a polarized rather than symmetrical nucleus with the distribution of nucleoli and chromosomes to opposite nuclear poles; this result revealed a novel role for Brca2 in establishing or maintaining oocyte nuclear architecture . Mutating tp53 did not rescue the infertility phenotype in brca2 mutant males , suggesting that brca2 plays an essential role in zebrafish spermatogenesis . Overall , this work verified zebrafish as a model for the role of Brca2 in human disease and uncovered a novel function of Brca2 in vertebrate oocyte nuclear architecture .
People who are heterozygous for strong mutations in the tumor suppressor gene BRCA2 ( FANCD1 ) have increased susceptibility to breast , ovarian , prostate , and pancreatic cancers [1]-[3] . Breast cancer risk for females heterozygous for germline mutations in BRCA2 is nearly 60% by age 50 [4] and for ovarian cancer is 11% [5] . BRCA2 is expressed in a broad range of mammalian tissues [6] , [7] and null activity alleles are embryonic lethal in mouse and humans but are viable in rats [8]-[11] . Biallelic inheritance of hypomorphic BRCA2 mutations in the germline results in Fanconi anemia ( FA ) , a disease characterized by catastrophic anemia , genome instability , characteristic morphological defects , and enormously elevated risk for leukemia ( 800 fold ) and squamous cell carcinomas ( 2000 fold ) [12]-[14] . The BRCA2 subtype of Fanconi anemia represents complementation group D1 [15] and results in a severe form of the disease with nearly 100% incidence of leukemia and/or solid tumors by 5 years of age [16] , [17] . The role of BRCA2 in tumor suppression and maintenance of genomic integrity is associated with its function in error-free , homology-directed recombination ( HDR ) [18] . HDR helps repair DNA breaks associated with meiosis , and mouse mutants in FA genes have defects in meiotic cells [19] . Zebrafish fancl mutants experience female-to-male sex reversal due to the apoptotic loss of meiotic oocytes at the time of sex determination [20] , consistent with the abnormal activation of the apoptotic pathway in the absence of Fanconi gene activity [21] . The involvement of BRCA2 in HDR , ovarian cancer , hypogonadal phenotypes , and the expression of Brca2 in mouse spermatocytes [22] converge to suggest a role for Brca2 in gonadogenesis . Homozygous Brca2 knockout mice die as embryos [23] , but transgenic mice carrying a BRCA2-containing human BAC that expresses the human gene at high levels everywhere except the gonads survive as sterile males and females [24] . In contrast , rats bearing a premature stop codon survive , but show slow growth and sterility [10] , reflecting conserved and lineage-specific roles of brca2 . To help understand the roles of Brca2 in vertebrates , we characterized zebrafish bearing an insertional mutation in brca2 . We show here that comparative analysis of zebrafish brca2 [25] identifies a few conserved , and hence putatively functional , coding regions and is expressed in proliferating somatic cells and in meiotic oocytes and spermatocytes . Surprisingly , brca2 transcript is asymmetrically localized to the animal pole of the cytoplasm in developing wild-type oocytes . The insertional brca2 null activity allele causes genome instability , slow growth of tissue culture cells , male sterility , testicular neoplasias , and female-to-male sex reversal that is rescued by mutation of the tumor suppressor gene tp53 ( p53 ) . Male and female double mutants are sterile and develop testicular neoplasias and invasive ovarian tumors . Nuclear symmetries are strikingly altered in oocytes of double mutant females , revealing a novel role of Brca2 in establishing or maintaining the architecture of the vertebrate oocyte nucleus . This work reveals that this zebrafish brca2 mutant is a model for unraveling gene functions as well as a valuable tool for small-molecule screens to help discover therapeutic compounds for human patients .
We isolated , cloned , and sequenced a zebrafish brca2 cDNA ( NM_001110394 ) and a BAC clone ( AC149226 ) . Because the zebrafish Brca2 protein shares only 21% identity with human BRCA2 , we confirmed orthology by conserved syntenies [26] . Our meiotic mapping on the HS panel [27] showed that brca2 lies on zebrafish chromosome 15 ( Dre15 , Figure 1A top ) , and sequence data at Ensembl ( http://www . ensembl . org/Danio_rerio/Info/Index ) showed that its genomic neighborhood contains 14 genes with conserved synteny to the orthologous region on human chromosome 13 ( Hsa13 , Figure 1A middle ) , as would be expected if zebrafish brca2 and human BRCA2 are orthologs . The absence of a second copy of brca2 in the co-orthologous region in Dre10 ( Figure 1A bottom ) provides evidence that , like all 13 other zebrafish fanc genes [25] , [28] , brca2 evolved to single copy in the zebrafish lineage after the teleost genome duplication [27] , [29]-[31] . A comparison of our cDNA and BAC sequences revealed that zebrafish brca2 has 26 exons ( numbered 2–27 to follow human nomenclature; Figure 1B ) like its tetrapod ortholog [7] , [32]-[34] . Despite low sequence identity ( 21% ) , zebrafish Brca2 conserves an N-terminal acidic transcriptional activation domain and a C-terminal DNA binding domain ( DBD ) [35]-[37] ( Figure 1C ) . Exon-11 , with 1 , 397 amino acid residues in zebrafish and 1 , 643 residues in human , is one of the longest vertebrate exons , 28 times larger than average [38] . Use of the stickleback brca2 sequence ( Figure S1 ) to help inform alignments showed that exon-11 of zebrafish brca2 contains a central array of BRC repeats conserved in approximate number , relative position , and sequence identity to those in tetrapods [34] ( Figure S2A ) . Phylogenetic analysis of BRC repeats revealed orthology between chicken and human repeats 1 , 5 , 7 , and 8 ( Figure 1D ) but not a one-to-one orthology between zebrafish and tetrapod repeats , suggesting that some individual repeats may have evolved independently by tandem duplication and/or gene conversion . Despite differences in BRC repeat sequences , the correlation of hydrophobicity indexes among repeats revealed great structural similarity ( Figure 1E ) . The DBDs of zebrafish and human Brca2 contain three oligonucleotide binding folds ( OB1-3 ) and a helical domain ( HD ) [37] ( Figure 1C and Figure S3 ) . The mapping of human tumor-derived mutations to these conserved features [37] supports the hypothesis that they are critical for functionally similar molecular interactions across vertebrates . In zebrafish embryos , brca2 has been shown to be expressed maternally and zygotically [25] , and this is confirmed here by RT-PCR ( Figure S4A ) and histological sections that show broad expression that is elevated in rapidly proliferating cells in the embryonic and larval central nervous system , in the proliferating ventricle margins of adult brains , in the blood-forming kidney marrow , and in the proliferative intervillus region of the intestine [39] ( Figure S4B-S4N ) . Germ line cells expressed brca2 in both male and female gonads in transitional stages ( Figure 2A , 2B ) , in immature gonads ( Figure 2C , 2D ) , and in the mature gonads of adults ( Figure 2E , 2F ) . Somatic cells of the zebrafish gonad either do not express brca2 , or do so at a low level like mouse Sertoli cells [40] . Unexpectedly , in stage III and IV oocytes , brca2 transcripts became asymmetrically distributed to a small peripheral patch of the ooplasm ( Figure 2E ) . Comparison with the distribution of pou5f1 ( oct4 ) [41] , [42] revealed that brca2 mRNA transcripts accumulated asymmetrically at the animal pole of the oocyte ( Figure 2G , 2H ) . Expression of brca2 was obvious in spermatocytes ( sc , Figure 2F ) , but not in spermatids and sperm ( sp , Figure 2F ) , as in mouse [40] . Expression of brca2 in meiotic cells is consistent with a role in repairing DNA breaks associated with meiotic HDR . Furthermore , the accumulation of brca2 transcript at the animal pole suggests a role of maternal message in provisioning embryos with Brca2 protein that could help effect DNA repair during the rapid cleavage divisions that occur before the initiation of zygotic transcription at the mid-blastula transition . To understand brca2 function , we studied a zebrafish line with the insertional mutation ZM_00057434 , which disrupts brca2 exon-11 in BRC repeat-z ( Figure 1C , Figure 3A and 3B , Figure S2 ) . Reverse transcriptase-PCR and sequence analysis detected no normal transcript in mutants , but instead identified two aberrant transcripts , one lacking the DBD and the other lacking all BRC repeats ( Figure 3C–3G ) . Because Brca2 protein cannot function without either of these features [37] , we conclude that ZM_00057434 is a null allele . Genome instability is a cardinal characteristic of Fanconi anemia [43] , [44] . Cell cultures from fins of homozygous brca2 zebrafish mutants and wild-type controls revealed normal karyotypes ( 2n = 50 ) with low levels of spontaneous breakage ( Figure 3H ) . After treatment with 10 ng/ml of the DNA-damage agent MMC , however , mutant cells showed many chromosome aberrations , including chromatid and chromosome breaks , radial reunion figures , and acentric chromosome fragments . Of 100 metaphases counted in mutant cells , 66 showed chromosome aberrations , including 32 that showed one or two anomalies , 23 with 3 or 4 abnormal chromosomes , and 11 with more than 5 aberrations ( Figure 3I ) . In contrast , all 24 metaphases from wild-type cells treated with MMC were normal ( Figure 3H ) . These results show that zebrafish cells require brca2 activity to prevent chromosome aberrations . To test genome stability in living animals , we crossed brca2 heterozygotes , stained resulting embryos at 28hpf with acridine orange ( AO , which fluoresces strongly when it intercalates into DNA with double-strand breaks [45] ) , scored the amount of AO staining in individual embryos , and genotyped embryos by PCR . Untreated mutants and wild-type controls had about the same amount of AO staining ( Figure 3J , 3K ) . In contrast , after treatment with the DNA damage agent diepoxybutane ( DEB ) at 4hpf , mutants accumulated substantially more AO-positive cells than wild-type siblings ( Figure 3L-3N ) . Thus , we conclude that Brca2 helps protect zebrafish embryos from DNA damage . To learn the role of brca2 ( fancd1 ) in zebrafish somatic cells , we established tissue cultures from fin biopsies of brca2 mutants and wild types and studied their growth rates . Mutant cultures showed significantly slower growth compared to wild-type cultures ( Figure 3O , p<0 . 001 at day 5 ) . Addition of MMC further delayed culture growth both for brca2 mutants ( p<0 . 5 vs . untreated ) and for wild types , although delay in wild types was not statistically significant ( Figure 3O ) . The poor growth of brca2 mutant cultures was due to high rates of spontaneous apoptosis ( 15% , Figure 3P-3S ) , as evidenced by propidium iodide exclusion and anti-active Caspase-3 staining . In brca2 mutants , addition of MMC increased the non-apoptotic cell death rate from 2 . 9% to 4 . 8% while the proportion of apoptotic cells remained essentially the same ( 15 . 0 and 14 . 1% , respectively , Figure 3P , 3Q ) . Untreated wild-type cultures revealed much less spontaneous apoptosis than mutant cultures ( 2% vs . 15%; Figure 3R ) and showed just a small effect of MMC on the non-apoptotic cell death rate ( 2 . 0% untreated , 3 . 9% treated , Figure 3S ) . We conclude that mutant cultures grow more slowly than wild-type cultures due to high rates of spontaneous apoptosis . To test the viability of zebrafish brca2 mutants , we mated heterozygotes , and among 414 adult offspring , 24 . 9% were homozygous wild types , 44 . 9% were heterozygotes , and 30 . 2% were homozygous mutants , a ratio indistinguishable from the expected 1:2:1 ratio ( X2 test , p = 0 . 37 , df = 2 ) . We conclude that zebrafish brca2 mutants survive about as well as wild types , as in Drosophila [46] . In addition , zebrafish brca2 ( fancd1 ) mutants expressed genes for primitive and definitive hematopoiesis normally ( Figure S5 ) , providing no evidence for the early hematopoietic defects found in human FA patients . Remarkably , however , all homozygous brca2 mutants developed exclusively as males . A series of heterozygote in-crosses gave 199 wild-type homozygotes and heterozygotes , about half of which were females ( 50 . 2%±0 . 1% ( sd ) ) . In contrast , of the 61 homozygous mutants , none were female ( X2 test , p = 0 . 00003 , df = 2 ) . Genotypic ratios following Mendelian principles ruled out female-specific lethality; thus , we conclude that individuals that would otherwise have become females experienced female-to-male sex reversal . Homozygous brca2 mutant males were sterile ( the 35 males tested fertilized no eggs , with an average clutch size of 197 eggs tested per male ) , but wild-type sibling males were all fertile ( the 28 males tested fertilized an average of 80% of the eggs per male with an average clutch size of 210 eggs tested per male ) . These data show that brca2 plays a role in male fertility and is necessary for female development in otherwise wild-type fish . To investigate the developmental basis of sex reversal in brca2 mutants , we analyzed transitional and immature ( but differentiated ) gonads . All juvenile zebrafish , regardless their definitive sex , initially develop oocytes; in females , these oocytes continue to develop but in males , they disappear [47] , [48] . In our experiments , some wild types at 21dpf contained perinucleolar oocytes ( early stage IB ) and other wild types contained a few pyknotic cells and oocytes at earlier stages of development ( early oocytes at stage IA ( leptotene to pachytene ) [49] ) ( Figure 4A , 4B ) . In contrast , all eight homozygous brca2 mutants examined at 21dpf lacked perinucleolar oocytes and contained earlier stage oocytes and large numbers of pyknotic cells ( Figure 4C ) . By 27dpf , wild types contained either ovaries or testes ( Figure 4D , 4E ) . All eight brca2 mutants analyzed , however , showed testis-like gonads that lacked perinucleolar oocytes but retained a few early oocytes and pyknotic cells ( Figure 4F ) . At 32dpf , perinucleolar oocytes in wild-type animals reached late stage IB and entered diplotene , as indicated by the presence of lampbrush chromosomes ( Figure 4G ) , while testes of wild-type males showed all stages of spermatogenesis including sperm ( Figure 4H ) . In contrast , all eight brca2 mutants analyzed at 32dpf had only testes that possessed spermatogonia ( sg ) and spermatocytes ( sc ) but lacked later developmental stages ( spermatids and sperm ) ( Figure 4I ) . In addition , 32dpf mutant gonads showed abnormal clusters of cells with pyknotic nuclei ( pc , outlined by dashed lines in Figure 4I ) and contained tubules abnormally depleted of germ cells ( asterisk , Figure 4I ) . The lack of perinucleolar oocytes in brca2 mutant gonads during the critical period for sex determination is consistent with the finding that gonads lacking oocytes during this period assume a male fate [20] . In addition , results showed that brca2 mutant germ cells became pyknotic , disappeared , and left empty spermatogenic tubules . The presence of pyknotic spermatocytes , lack of spermatids and sperm , and the existence of empty tubules in brca2 mutant testes suggested the hypothesis that spermatocytes did not progress through meiosis and died . To test if the activation of apoptotic pathways is involved in spermatocyte death , we used immunoassays to detect active-Caspase-3 , a marker of apoptosis [50] . In contrast to wild-type gonads , brca2 mutant testes showed clusters of cells with active-Caspase-3 that were clearly pyknotic after hematoxylin and eosin staining ( Figure 4J–4M ) , confirming that brca2 spermatocytes undergo apoptosis . At 47dpf , brca2 ( fancd1 ) mutants already showed hypogonadism ( Figure S6I ) , a characteristic shared by many FA patients . Germ cell distribution as revealed by vasa expression [51] was similar in mutants and wild-type gonads ( Figure S6A , S6E , S6I ) . In mutants , somatic cells expressing the Sertoli cell marker amh ( anti-Müllerian hormone ) [52] , [53] failed to form neat borders surrounding tubules as in wild-type males ( Figure S6F , S6J ) and lacked expression of the female marker cyp19a1a ( aromatase ) [54] ( Figure S6C , S6G , S6K ) . The early meiotic marker sycp3 ( synaptonemal complex protein 3 , [55] ) was expressed by groups of spermatocytes in wild-type and mutant males ( Figure S6H , S6L ) . We conclude that mutant gonads develop a molecular profile similar to wild-type testis accompanied by disorganization of amh-expressing somatic cells that surround testis tubules . To understand the cellular basis of male infertility in brca2 mutants , we compared adult testis histology in wild types ( n = 3 ) and brca2 mutants ( n = 7 ) . Comparison of the anterior part of the testes ( anterior testes ) of wild types and brca2 mutants revealed persistent hypogonadism ( smaller diameter gonads ) in the mutants ( Figure 5A , 5B ) . Strikingly , brca2 mutant testes lacked sperm and showed tubules with central empty cavities ( asterisks ) . Wild-type and mutant testes both contained spermatocytes at the bouquet stage ( late-zygotene/early-pachytene [56] , [57] ) ( sc-b , Figure 5C , 5D ) , but mutants lacked later stages ( spermatids and sperm ) . Mutant testes also contained abnormal clusters of pyknotic cells ( pc , Figure 5D ) . Occasionally , bouquet stage spermatocytes and pyknotic cells occupied the same tubule ( Figure 5E ) , suggesting that spermatocytes blocked in meiosis became pyknotic . Eosinophils ( eos , Figure 5F ) invaded some cavities containing pyknotic cells , suggesting an inflammation-like response in mutant gonads . In the posterior part of the testes ( posterior testes ) of wild types , tubules demarcated by interstitial cells were filled with sperm ( sp , Figure 5G ) , but in the posterior testes of brca2 mutants , tubules were devoid of sperm ( Figure 5H ) . This observation can account for the infertility phenotype of adult brca2 mutant males . In addition to truncated spermatogenesis , brca2 mutant testes displayed abnormal regions of accumulating cells ( Figure 5I-5L ) . Some of these neoplasias contained both spermatogonia ( sg ) and interstitial cells ( ic , Figure 5I , 5K ) and others contained only spermatogonia ( sg , Figure 5J , 5L ) . We conclude that brca2 provides some function that regulates proliferation of spermatogonia and interstitial cells . To help understand the altered morphologies of adult mutant testes , we studied gene expression patterns . In the anterior testis , brca2 mutants contained more clusters of vasa-expressing cells than did wild types ( Figure 6A , 6A′ , 6D , 6D′ ) , reflecting the accumulation of vasa-expressing early germ cell stages ( spermatogonia and spermatocytes , Figure 6A′ , 6D′ ) ) and the depletion of non-vasa expressing late stages ( spermatids and sperm; purple circle , Figure 6A′ ) . In brca2 mutants , amh-expressing cells were less frequent and did not surround tubules normally ( Figure 6B , 6E ) . In addition , sycp3-expressing pachytene spermatocytes accumulated abnormally in mutant testes ( Figure 6C , 6F ) , as expected from the histological data that showed the lack of post-meiotic cells . The posterior testes of adult wild types did not express vasa and amh , consistent with the presence of interstitial cells and late germ cells ( sperm ) , which no longer express vasa , and the absence of Sertoli cells ( Figure 6G–6I ) . In contrast , posterior testes of adult mutants contained empty cavities and abnormally proliferating cells ( Figure 6J–6O ) similar to those observed in histological analyses ( Figure 5I–5L ) . Neoplasias contained either mixtures of vasa-expressing and non-vasa-expressing cells ( Figure 6J–6L ) or possessed only vasa-expressing , early spermatogenic cells ( Figure 6M–6O ) . These results revealed the abnormal presence of early spermatogenic cells in the posterior part of the testes in mutants . Moreover , the presence of scattered amh-expressing cells revealed the abnormal presence of Sertoli cells in mutant posterior testes ( Figure 6K , 6N ) . All zebrafish gonads initially form oocytes but these die in wild-type juvenile males [47] , [48] , and increased germ cell apoptosis leads to oocyte loss and female-to-male sex reversal in fancl mutant zebrafish [20] . Tp53 ( alias p53 ) is an important activator of apoptosis and zebrafish with hypomorphic mutations in tp53 are viable and fertile despite reduced apoptosis [58] , [59] . To learn the role of apoptosis in sex reversal of brca2 mutants , we made double mutants for brca2 and the hypomorphic allele tp53M214K [58] , [59] . Results showed that no brca2−/− mutants with at least one wild-type tp53 allele developed into females ( Figure 7A ) . In contrast , brca2−/− mutants that lacked a normal tp53 allele became males and females with about equal frequency ( Figure 7A ) . This result shows that a tp53 mutation can rescue the female-to-male sex reversal caused by the lack of brca2 activity . Because double mutant females develop ovaries containing oocytes ( Figure 7E ) , we conclude that Tp53-mediated apoptotic cell death is important for sex reversal in brca2 mutants and interpret these results to mean that the survival of oocytes in brca2 mutant gonads can allow individuals to become females . The rescue of sex reversal by Tp53 mutation reveals that brca2 function is required for oocyte survival , which secondarily leads to female gonad fate and ovarian development . In wild-type late stage II to early stage III oocytes , brca2 and pou5f1 transcripts localize to the animal pole and vasa transcripts gradually spread out cortically from the vegetal pole ( Figure 2G , 2H , Figure S7A–S7C , and [41] , [42] , [60]-[62] ) . To test the hypothesis that brca2 function is important for the localization of these transcripts , we examined mRNA distribution in oocytes of brca2;tp53 double mutants . In situ hybridization on adjacent serial sections showed that double mutant females produced oocytes with brca2 and pou5f1 transcripts localized to one pole and vasa transcripts positioned at the opposite pole of the same individual oocytes ( Figure S7D–S7F ) . We conclude that brca2 activity is not necessary to localize the messages tested to their proper location in developing zebrafish oocytes . To learn if tp53 mutation can rescue the infertility phenotype observed in brca2 single mutants , we mated brca2 homozygous mutants that were either wild type , heterozygous , or homozygous for the tp53 mutation to wild-type animals and scored fertility . Results showed that all brca2 mutant males were sterile regardless of their tp53 genotype ( for brca2−/−;tp53+/+ , brca2−/−;tp53+/− , and brca2−/−;tp53−/−: we found 0/200 offspring ( 8 males tested ) , 0/349 offspring ( 13 males tested ) , and 0/148 offspring ( 6 males tested ) , respectively ) . Female brca2;tp53 double mutants were also sterile when mated to wild-type males ( of 549 eggs produced by 9 females , 79 initiated cleavage ( the average double mutant female had 20%±18% fertility compared to doubly heterozygous siblings with 85%±27% fertility ) . Non-developing eggs laid by double mutant females were milky and were of highly variable size . Some double mutant females mated to wild-type males produced doubly heterozygous eggs that completed cleavage and gastrulation , but failed to develop to later stages ( Figure 7B ) . Because doubly heterozygous individuals from doubly heterozygous mothers develop normally but doubly heterozygous embryos from homozygous brca2 mutants are lethal , and because homozygous tp53 mutant females have normal fertility [58] , we conclude that maternal brca2 function is important for proper embryo development . Histological sections of 6mpf ( months post-fertilization ) adult ovaries from wild types ( brca2+/+;tp53+/+; n = 3 ) , tp53 homozygous mutants ( brca2+/+;tp53−/−; n = 3 ) and double mutants ( brca2−/−;tp53−/−; n = 4 ) revealed oocytes at a variety of developmental stages in all three genotypes ( Figure 7C–7E ) . In tp53 mutants and in wild types , late stage IB oocytes ( L-IB ) contained nucleoli distributed uniformly along the nuclear periphery ( Figure 7C , 7D , 7D′ ) , consistent with the normal fertility of homozygous tp53M214K mutants [58] . In contrast , brca2;tp53 double mutants contained degenerating late stage oocytes ( d , Figure 7E ) with a granulosa cell layer ( gc ) that was poorly organized and sometimes separated from the vitelline envelope ( ve , Figure 7E′ ) . Wild-type nuclei of late stage IB to early stage II , stage II , and stage III oocytes were radially symmetrical , containing peripheral nucleoli and central chromosomes ( Figure 7F–7H ) . In contrast , brca2−/−;tp53−/− double mutant oocytes were polarized , showing abnormally enlarged and variably shaped nucleoli that accumulated asymmetrically towards one pole of the nucleus while chromosomes concentrated towards the opposite pole ( Figure 7I–7K ) . We conclude that brca2 activity is essential to establish or to maintain a normal architecture of the oocyte nucleus . In addition to their abnormal location , oocyte chromosomes in double mutants had altered morphology . In wild-type oocytes , chromosomes were distributed independently in the center of the nucleus ( arrows in Figure 7L ) , but chromosomes in oocytes of double mutants were interconnected and formed abnormal loops ( arrows in Figure 7M ) . These chromosome phenotypes were not observed in oocytes of tp53 single mutant females . Aberrant chromosome structure would be expected if recombination-induced chromosome breaks are left unrepaired or are repaired by an error-prone pathway . Our experiments showed that MMC-induced DNA breaks caused chromatid and chromosome damage that led to radial reunion formation in somatic cells ( Figure 3I ) . Likewise , inappropriate repair of recombination-induced DNA breaks could prevent dispersal of oocyte chromosomes . These results suggest a role of brca2 in repairing DNA breaks originating either artificially by MMC or naturally in meiotic recombination . Because humans heterozygous for BRCA2 mutations have elevated risk of tumors , we investigated older brca2:tp53 mutants for abnormal growths . By 6mpf , tumors formed that invaded the ovarian cavity and intercalated between oocytes in two of four brca2−/−;tp53−/− double mutants examined ( asterisks , Figure 7N–7Q ) . Tumors were not detected in wild types ( n = 3 ) or in tp53 homozygous mutants ( n = 3 ) . One tumor appeared to originate at the ovarian cavity membrane ( arrow , Figure 7N ) . Tumor origin in the other female was unclear because it was metastatic and contacted both the ovarian cavity membrane and the swim bladder ( Figure 7P ) . Both tumors involved spindle-shaped cells that invaded the ovary and surrounded the oocytes ( Figure 7N′ , 7O , 7Q ) . The high incidence of ovarian cancer we observed in brca2−/−;tp53−/− double mutants contrasts with the prior finding that animals homozygous for either of two mutant tp53 alleles ( tp53M214K , the mutation used here , or tp53I166T ) are viable and fertile , but at 8 . 5 or 8 . 8mpf ( ten weeks later than our double mutants ) , 1 of 144 and 1 of 417 fish , respectively , began to show tumors and by 16 . 5 or 22mpf , 28% or 100% , respectively , of the tp53 mutants had developed tumors [58] , [59] . Of several hundred tumors previously described in tp53 mutants , none were reported to involve the ovarian cavity membrane [58] , [59] . The early appearance and unique ovarian location of tumors in brca2−/−;tp53−/− double mutants suggest a specific association with brca2 activity , potentiated by impaired tp53 function and Tp53 deficiency is cooperative with Brca2 in tumorigenesis in humans [8] . Future long-term investigations are required that focus on tumor development in a large cohort of brca2−/−;tp53−/− double mutants ( 1/16th of the progeny of double heterozygotes ) compared to their single mutant tp53−/− siblings to more clearly define the role of brca2 in the development of ovarian tumors . Testis development was abnormal in double mutants . Analyses of wild-type ( n = 4 ) and tp53 single mutant ( n = 3 ) testes in 6mpf adults revealed germ cells at all stages of spermatogenesis in all the animals analyzed ( Figure 8A , 8B ) . In contrast , all testes analyzed of brca2−/− single mutants ( brca2−/−;tp53+/+; n = 4 ) and double mutants ( brca2−/−;tp53−/− , n = 6 ) abnormally lacked spermatids and sperm ( Figure 8C , 8D ) . Furthermore , in all brca2−/−;tp53+/+ single mutants and all brca2−/−;tp53−/− double mutants , posterior tubules contained empty cavities lacking germ cells ( asterisks in Figure 8C , 8D ) consistent with our finding that adult double mutants failed to recover fertility . Unexpectedly , some testes in brca2−/−;tp53−/− double mutants ( n = 2 ) , but not in the other genotypes , contained germ cells enlarged up to ten times normal diameter ( Figure 8E ) . Some of these enormous cells we call megalospermatogonia ( ms ) because , like normal spermatogonia [63] , they contained an enlarged central nucleolus ( nc ) . Testes of some brca2−/−;tp53−/− double mutants ( n = 3 ) contained other large cells that showed the peripheral distribution of numerous nucleoli , which constitutes an oocyte-like morphology [63] , so we interpret these as large early oocytes ( eo ) , and additionally we observed the presence of enlarged pyknotic cells ( pc ) ( ms , pc , and eo in Figure 8E , 8F , 8F′ , compare to normal spermatogonia ( sg ) outlined by dashed lines in Figure 8E , 8F ) . The lack of sex-specific markers for early gonial cells precludes a more precise definition of cell type . Somatic cell neoplasias appeared in the posterior testis of all double mutants ( but showing variability on the neoplastic tissue size ) ( Figure 8G , 8G′ ) suggesting the hypothesis that late stage spermatogenic cells negatively regulate the proliferation of somatic cells in testes . To test this hypothesis and to investigate whether the absence of brca2 activity or merely the absence of germ cells allows over-proliferation of the somatic component of the testis , we examined animals depleted of germ cells by dead end-morpholino ( dnd−/−; n = 10; see also [64] ) . Dnd is an RNA binding protein that is essential for germ cell survival in mice and zebrafish [64] , [65] . Our results revealed that testes in 18mpf dnd-morpholino treated animals developed neoplastic somatic proliferation ( Figure 8H , 8H′ ) similar to those observed in brca2 single and brca2;tp53 double mutants ( Figure 5I , 5K , Figure 8G ) . In five of ten dnd-injected animals , neoplasias invaded the intestine and body wall musculature ( Figure 8H ) . Although we did not detect invasive somatic proliferation in brca2 mutants , these dnd-knockdown animals were 12 months older than our brca2 mutants , suggesting that invasive proliferation might arise in brca2 mutants as they age . Overall , these results would be expected if somatic cell neoplasias in brca2 and brca2;tp53 mutants were not due to a direct effect of the lack of brca2 activity , but arose as a secondary effect of germ cell loss in brca2 mutants . A possible mechanism to explain these results is that late stage germ cells in the wild-type spermatogenic pathway exert a negative control over the proliferation of somatic cells in zebrafish testes . Loss of Dead end function in mouse results in germ cell tumors in a strain-specific manner [65] . Because dnd-knockdown zebrafish have no germ cells , the origin of gonad tumors after dnd-knockdown differs between mouse and zebrafish . Long-term investigations are required to examine whether the somatic testicular tumors we observed in zebrafish lacking dnd function are subject to effects of the genetic background .
Genomic and genetic evidence shows that zebrafish has a single-copy of brca2 that is orthologous to its mammalian counterpart . We found that the zebrafish genome has duplicate copies of the human chromosome segment that contains BRCA2 , and that these duplicates arose from the teleost genome duplication ( TGD ) . In one of these duplicated segments , the BRCA2 ortholog disappeared , leaving zebrafish with a single copy of brca2 . About 75% of genes from the TGD event reverted to singletons [27] , but all of the 13 zebrafish orthologs of FA pathway genes are present in single copy [25] , which would happen rarely ( 2 . 4% ) solely by chance . We conclude that evolutionary forces probably acted to reduce brca2 and other FA pathway genes to single copy after the TGD . This finding is predicted by the duplication-degeneration-complementation hypothesis [67] , which suggests that genes with simple tissue- and time-specific regulatory elements would be more likely to revert to singletons than those with complex regulation . In addition , many Fanc proteins join to form molecular machines in a 1:1 stoichiometry , so that if one gene in the network evolves to single copy , the others might follow by natural selection or neutral evolutionary forces . Expression analyses showed that maternal brca2 message accumulated in zebrafish embryos . This message would be available to provide embryos with Brca2 protein that could function to help resolve stalled replication forks [68] during the rapid cleavage divisions that precede the mid-blastula transition , the stage at which zygotic transcription initiates [69] . Our finding that the heterozygous offspring of homozygous brca2 mutant mothers fail to develop much past gastrulation supports this conclusion . Expression of brca2 in meiotic cells of zebrafish , as in mammals [24] , suggests a role in the repair of DNA breaks incurred during meiotic homologous recombination [46] . Zebrafish brca2 ZM_00075660 mutants generate only aberrant transcripts that lack domains essential for Brca2 activity and provide a vertebrate null allele model to unravel the effects of brca2 during embryonic and post-embryonic development . Mutant tissue culture cells and developing embryos show more chromosome damage and excess staining of broken DNA , respectively , than wild-type cells or embryos after exposure to DNA damaging agents . Similarly , loss of BRCA2 function in humans results in hypersensitivity to DNA crosslinking agents [44] , [70] , thus leading to chromosome breaks [43] , [71] , showing that zebrafish and human Brca2 orthologs share functions in maintaining genome stability . Flow cytometry showed that the poor growth of zebrafish brca2 mutant cell cultures results from high rates of spontaneous apoptotic cell death . This finding parallels our finding of increased apoptosis in juvenile mutant gonads that results in oocyte loss and sex reversal and that the inhibition of cell death in brca2;tp53 double mutants rescues sex reversal . Spontaneous apoptosis leading to bone marrow failure is also a problem in hematopoietic stem cells in human FA patients , [21] . Zebrafish therefore appears to be a valid model to study the basic influence of Brca2 deficiency on apoptosis . In contrast , the effect of damage caused by MMC in zebrafish brca2 mutant cell cultures on non-apoptotic cell death rates is surprisingly low compared to humans; the cause of which remains unexplained . Results showed that brca2 mutant zebrafish developed exclusively as males due to female-to-male sex reversal . Sex ratios also appear to be skewed in the offspring of human carriers of BRCA2 mutations , suggesting a possible role in sex determination or differential survival [72] . During the sex determination period , zebrafish mutant gonads contained apoptotic cells and lacked diplotene oocytes , the presence of which is essential to tip gonad fate towards the female pathway in fancl zebrafish mutants [20] . In contrast to fancl mutants , which become fertile males , fancd1 ( brca2 ) mutants become sterile males . The more severe phenotype of brca2 mutants compared to fancl mutants parallels the fact that human homozygotes for FANCD1 ( BRCA2 ) null alleles are lethal as embryos [11] . In the FA-BRCA network , FANCL should act upstream of BRCA2 ( see for review [73] . Because the phenotype of brca2 mutant zebrafish is more severe than that of fancl mutant zebrafish , BRCA2 likely plays roles in addition to its function downstream of fancl . Because FANCD1 ( BRCA2 ) is the only FA complementation group that fails to form RAD51 foci after ionizing radiation and crosslink damage , FANCD1 ( BRCA2 ) , but not FANCL , is required for RAD51-mediated DNA repair [74] , [75] . The inhibition of apoptosis in brca2 mutants by the mutation of tp53 rescued female-to-male sex reversal and led to the development of females , consistent with the idea that brca2 mutant oocytes die by apoptosis when unable to repair DNA breaks associated with meiotic recombination . This conclusion parallels that from zebrafish fancl mutants , in which oocytes die in juveniles followed by female-to-male sex reversal [20] and supports the notion that fanc-related sex reversal acts via Tp53-mediated apoptosis . The brca2 ( fancd1 ) and fancl results combine to support the following model for zebrafish sex determination . ( 1 ) The FA network facilitates DNA repair associated with meiosis and hence the survival of oocytes during the critical period for zebrafish sex determination . ( 2 ) Activation of Tp53-dependent germ cell apoptosis , at least in fanc mutants , alters the total number of germ cells and thus reduces the number of surviving oocytes below the threshold necessary to maintain female fate . ( 3 ) Post-recombinant oocytes release a signal that down-regulates amh and/or maintains cyp19a1a ( aromatase ) expression in somatic cells of the bipotential gonad [20] , [48] , [52] , [64] , [76]-[80]; the fewer the number of post-recombinant oocytes , the less aromatase-maintenance signal . ( 4 ) Aromatase converts testosterone to estrogen , thereby reinforcing ovary development and the female fate . ( 5 ) In normally developing males at the juvenile hermaphrodite stage [47] , unknown genetic factors that may be influenced by the environment stimulate the death of oocytes and hence loss of the aromatase-maintenance signal . According to this model , in the absence of either brca2 ( fancd1 ) or fancl , oocytes do not effectively repair the DNA breaks of meiosis , DNA-damaged oocytes die by apoptosis before they liberate the aromatase-maintenance signal , the gonad becomes a testis , and individuals that otherwise would have become females develop into males . The study of zebrafish brca2 mutants verifies the importance of Brca2 for gonad development and provides a new vertebrate model for the adult roles of Brca2 that is obscured by null mutant lethality in human and mouse . Neither Brca2 knockout rats nor Brca2 knockout mice rescued by a human BRCA2 BAC showed sex reversal [10] , [24] , reflecting lineage-specific sex determining mechanisms . Zebrafish brca2 mutants developed gonads without diplotene oocytes , but rescued mice did develop oocytes , many of which disappeared post-natally , but some of which progressed through meiotic prophase I , became fertilized , and produced embryos [24]; in contrast , Brca2 mutant rats were sterile . A possible explanation for the difference between mouse and rat is that the transgenic mice might not be total null mutants because expression of human BRCA2 was detected in their gonads [24] . Zebrafish brca2 mutants contained spermatocytes arrested in meiosis , as did transgenic mice rescued by human BRCA2 [24] and rats mutant for Brca2−/− [10] . We found that zebrafish brca2 mutant testes ( 1 ) showed hypogonadism like human FA patients; ( 2 ) developed spermatogonia that entered meiosis as shown by sycp3 expression and histological data; ( 3 ) contained bouquet stage spermatocytes that arrested in late zygotene-early pachytene; ( 4 ) lacked post-meiotic spermatogenic stages , including spermatids and sperm; ( 5 ) failed to properly organize Sertoli cells , as shown by amh expression; ( 6 ) contained abnormal pyknotic cells that were positive for the apoptotic marker active-Caspase-3; and ( 7 ) formed tubules that lacked germ cells but contained eosinophils , blood cells involved in inflammation . Together , these results suggest a mechanistic model in which brca2 mutant spermatogenic cells develop rather normally until meiotic recombination ( pachytene ) , fail to repair double strand DNA breaks associated with homologous recombination , then die , leaving empty testis tubules in hypogonadal sterile males . Because spermatogenic cells die even in brca2;tp53 double mutants , we conclude that , after meiotic failure , brca2 spermatogenic cells die by a Tp53-independent pathway in double mutants , or alternatively , that the hypomorphic nature of the tp53M214K allele may allow cells to disappear by a Tp53-dependent pathway . Megalospermatogonia with enormous nuclei appeared in double mutants , possibly due to continued DNA replication in spermatogonia damaged by inadequate DNA repair in the absence of brca2 activity . In brca2 single mutant testes , cells may experience extra rounds of replication but tp53-mediated cell death may delete them . Alternatively , in the absence of brca2 function , tp53 activity might be required to prevent abnormal functions that lead to cell enlargement and megalospermatogonia . The presence of oocytes in testes of double mutant animals might be explained by the alteration of somatic cell-to-germ cell signaling in testes that are developing with abnormal Sertoli cell distribution and lack of spermatids and sperm . Normal Tp53 function might be necessary to induce oocyte apoptosis in brca2 single mutant testes . Six-month old brca2 single mutants accumulated neoplastic growths involving spermatogonia with or without disorganized clumps of interstitial cells . The posterior part of the testis , which completely lacked germ cells in brca2 single and brca2;tp53 double mutants , showed abnormal proliferation of somatic cells in the testes . The investigation of genetically wild-type animals lacking germ cells due to dnd knockdown uncovered somatic neoplasias of the testes similar to those found in brca2 mutant testes . Knockdown of dnd was not previously reported to cause neoplasias , but the oldest dnd-knockdown animals previously reported were 6 months old [64] and our dnd-knockdowns were 18 months old , suggesting that these neoplasias arise with age , although we can't rule out strain-specific effects . We conclude that post-meiotic germ cells , specifically the spermatids or sperm that are lacking from brca2 mutants , control growth of surrounding somatic cells by an as yet unknown mechanism that is secondary to brca2-impaired spermatogenesis . Our gene expression analyses showed unexpectedly that brca2 transcript localizes asymmetrically at the animal pole of wild-type oocytes . The asymmetrical distribution of certain mRNAs in the oocyte cytoplasm helps to promote animal or vegetal pole identity [41] , [42] , [60]-[62] . In brca2;tp53 double mutants , however , the localization of messages for pou5f1 , vasa , and even brca2 itself were normal , suggesting that brca2 activity is not essential to polarize the oocyte cytoplasm . Furthermore , brca2 transcript begins to be localized in stage III oocytes but ccnb1 transcripts begin to be localized earlier [41] , [42] , a timing that is incompatible with the hypothesis that brca2 initiates ooplasm asymmetry . The germinal vesicle ( the oocyte nucleus ) lies in the center of stage I-III oocytes , but moves to the animal pole by stage IV [41] , [49]; thus , the oocyte nucleus -- and hence the resulting zygote and cleavage nuclei -- occupy cytoplasm enriched in brca2 transcript . The translation of this transcript would be available to support the repair of DNA damage incurred in the rapid divisions of cleavage before zygotic transcription initiates at the mid-blastula transition . The observation that brca2;tp53 double mutant females formed doubly heterozygous embryos that cleaved normally , completed gastrulation , but then died by 24 hpf supports the notion that maternally transmitted brca2 transcript is important for normal early development and that zygotic brca2 transcript is too little or too late to rescue the phenotype . brca2;tp53 double mutant females produced oocytes with normally organized ooplasm but aberrantly organized nuclei . Developing oocytes that lacked brca2 activity partitioned nucleoli aberrantly to one side of the nucleus rather than their usual radial location and distributed chromosomes opposite to the nucleoli rather than their normal central position . Although these ovaries were also homozygous mutant for tp53 ( which was necessary to obtain homozygous brca2 mutant females ) , the oocyte nucleus architecture defect is due to brca2 deficiency because homozygous tp53 mutant females form normal oocytes [58] , [59] , as we confirmed . Oocytes in brca2;tp53 females reached the bouquet stage of meiosis in which telomeres cluster at one side of the oocyte . The asymmetric localization of chromosomes in the oocytes of brca2;tp53 females may result from a deficiency in DNA repair that inhibits exit from the bouquet stage . This could result , for example , if Brca2 is necessary for the relaxation of telomere clustering that generates the bouquet stage , a proposition supported by the high rate of recombination in subtelomeric sequences [81]: a high rate of recombination near the telomeres could create interlinked chromosomes like the radial reunion figures we observed in somatic cell chromosomes and these links might not permit normal chromosome dispersal in the nucleoplasm . It is unclear , however , how the abnormal persistence of chromosome clustering would generate the asymmetric distribution and abnormal morphologies of nucleoli that we found in oocytes of mutant females . Drosophila brca2 mutants develop oocytes with an abnormally asymmetric karyosome and dorso-ventral defects [46] , phenotypes that may be functionally related to those we observe in zebrafish . Transgenic female mice rescued with a human BRCA2-containing BAC have abnormal polar bodies in meiosis [24] , a phenotype that may be a consequence of nuclear symmetry problems we demonstrate here in zebrafish . We conclude that brca2 is generally important for the organization of oocyte nuclei both in protostomes and in vertebrates . Tp53 deficiency coupled with diminished Brca2 activity promotes mammalian breast tumors [8] . Likewise , zebrafish brca2;tp53 double mutants develop invasive ovarian tumors , and these appear earlier and more frequently than tumors in animals with lower Tp53 activity alone . We conclude that zebrafish shares with human and rat [10] a requirement for Brca2 activity to help suppress the formation of ovarian cancers . The genome instability we observed in brca2 mutant tissue culture cells may contribute to tumor formation because zebrafish gin mutations identified on the basis of genomic instability have elevated cancer risk [82] . Future studies are necessary to better understand the etiology of brca2-dependent ovarian tumors in zebrafish . FANCD1 is an alias of BRCA2 because homozygous hypomorphic mutations in this gene cause Fanconi Anemia while heterozygous null mutations increase the risk of breast and ovarian tumors and homozygous null mutations cause lethality [11] . In addition to genome instability , bone marrow failure , leukemia , and squamous cell carcinomas , many FA patients experience hypogonadism , impaired gametogenesis , defective meiosis , and sterility [19] , [83] . Thus , zebrafish brca2 ( fancd1 ) shares genome instability and gonad developmental phenotypes with FA patients and ovarian tumors with human heterozygotes for BRCA2 mutations . These findings indicate that zebrafish brca2 mutants provide a suitable model for human BRCA2-related disease . In a result of special significance , the embryonic sensitivity of zebrafish brca2 mutants to a DNA damage agent provides an assay for a small molecule screen to identify compounds that can rescue the DNA damage phenotype and thus has the potential to contribute to the discovery of substances that can ameliorate at least some of the phenotypes observed in human patients .
A partial gene model for zebrafish brca2 was inferred using GenomeScan ( http://genes . mit . edu/genomescan . html ) to match the human BRCA2 protein to a zebrafish genomic BAC clone we identified and sequenced ( Genbank accession #AC149226 ) . Primers for RACE ( Clontech ) were designed using this preliminary gene model . RACE template was 5′ first-strand zebrafish cDNA synthesized from pooled mRNA from embryos at 12 , 24 , and 48 hours post-fertilization ( hpf ) . A BLAST search of the zebrafish EST database ( http://www . ncbi . nlm . nih . gov/genome/seq/BlastGen/BlastGen . cgi ? taxid=7955 ) using the 3′ UTR modeled with GenomeScan recovered EST CT605096 . This EST and our GenomeScan model were used to design primers for amplification of the entire brca2 cDNA ( primers in Table S1 ) using as template second strand cDNA from 60hpf embryos . Amplified fragments were cloned using the TOPO Cloning Kit for Sequencing ( Invitrogen ) . We applied reciprocal best BLAST “hit” ( RBH ) [84] as an initial test for orthology of zebrafish and human BRCA2 genes . We queried the stickleback genome ( http://www . ensembl . org/Gasterosteus_aculeatus/blastview ) using zebrafish Brca2 and found stickleback brca2 on contig 2939 . This contig , together with human BRCA2 , was used to develop a preliminary gene model for stickleback brca2 ( Figure S1 ) . The Synteny Database identified paralogous and orthologous chromosome segments [26] . The insertional brca2 mutant ( ZM_00075660 ) was purchased from Znomics , which randomly inserted derivatives of a Moloney murine leukemia-based retroviral vector into zebrafish [85] . To genotype ZM_00075660 , we used primers F4/R4 to amplify the mutant allele , and primers F1/R1 to amplify the wild-type allele . Table S1 lists primer sequences . To reduce Tp53 activity , the hypomorphic mutation tp53M214K was obtained from ZIRC ( http://zebrafish . org/zirc/home/guide . php ) and was used and genotyped as described [58] . All animals were reared and collected under standard conditions [86] . The University of Oregon Institutional Animal Care and Use Committee approved all animal work ( Animal Welfare Assurance Number A-3009-01 , IACUC protocol #08-13 ) . Genetic nomenclature follows guidelines from ZFIN ( http://zfin . org/zf_info/nomen . html ) , e . g . , human gene , BRCA2; mouse gene , Brca2; zebrafish gene , brca2; human protein BRCA2 and mouse and zebrafish protein , Brca2 . Whole mount in situ hybridizations were performed as described [87] using several individuals for each developmental stage . In situ hybridization experiments on zebrafish cryosections were performed following [52] . Probes for amh and cyp19a1a were made following [52] and probe for vasa was made from its 3′end as described [51] . A brca2 cDNA fragment of 725nt ( nucleotides 2627-3351 of NM_001110394 ) , a pou5f1 cDNA fragment of 778nt ( nucleotides 705-1482 of NM_131112 ) , and a sycp3 cDNA fragment of 620nt ( nucleotides 265-884 of NM_001040350 ) were used to synthesize DIG-labeled riboprobe ( Boehringer Mannheim ) . For gonad histology , paraffin embedded Bouin's fixed tissue was sectioned at 7 microns and stained with hematoxylin and eosin . Animals were fixed at 60dpf in 4% PFA overnight at 4°C , embedded in paraffin , and sectioned at 7 microns . Apoptotic cells were detected by immunofluorescence using anti-active Caspase-3 ( Pharmingen , # 559565 Purified Rabbit anti-active caspase-3 ) following published protocols [20] , [39] . Animals depleted of germ cells were obtained by injecting wild-type zebrafish embryos from the AB strain at the 1–2 cell stage with dead end antisense morpholino oligonucleotide ( Gene Tools ) as described [88] . Injected and non-injected animals were raised to adulthood and collected at 18 months post-fertilization . Embryos were exposed to diepoxybutane ( DEB ) in embryo medium from 7–28hpf , or left untreated . Embryos were stained with acridine orange ( AO ) and mounted following [89] . The initial focal plane was the otolith , and a z-series consisting of seven 10-micron steps was captured on a Bio-Rad Radiance 2100 confocal microscope . Images were merged into a single plane with Velocity 4 . 4 . 0 and the number of AO-positive cells in the central nervous system anterior to the beginning of the yolk extension was quantified using ImageJ . Distal tips of caudal and dorsal fins from mutant and wild-type adult fish were cultured in 1∶1 ( vol/vol ) DMEM ( Gibco ) and Amniopan ( PAN ) media supplemented with 100 µg/ml penicillin and 0 . 1 mg/ml streptomycin in a 5% CO2 atmosphere at 28°C . Adherent fibroblast-like cells grew from primary explants . When culture flasks were confluent , cells were trypsinized and subcultured at split ratios of 1:4 . Cells of the 20th or 21st passage were exposed to 5 or 10 ng/ml mitomycin C for 24 hrs and screened for chromosome morphology . Rates of chromatid and chromosome breaks , radial reunion figures , and other categories of breakage were scored by analyzing 100 cells from each cell line . To visualize mitotic chromosomes , subconfluent cultures were exposed to colcemid for 3 hrs . Accumulated metaphases were prepared following standard methods . Slides were stained with 5% Giemsa solution . Metaphases were screened under a light microscope . Line authenticity was confirmed by PCR genotyping using the primer set Brca2zmwt . F2 5′-GCAGGTTGTGATGAAGCCACC-3′ and Brca2zmwt . R1 5′-GTGGTGTGAGGCCAGAGGTT-3′ for amplification of a 888-bp fragment of the wt brca2 sequence and the primer set 5Fd1ins . F 5′-CTTGCGCACCAAGGCTTCAC-3′ and 5Fd1ins . R 5′-ACCGCATCTGGGGACCATCT-3′ for amplification of a 971-bp fragment of the insert . For cell growth studies , 1×105 cells per line were seeded into six flasks . Following trypsinization , one flask per line was counted daily until day 5 . Resulting numbers were plotted as multiples of the initial cell count . Mitomycin C ( MMC ) was added at given concentrations at time 0 h . For flow cytometric assays , cultures were harvested and cells were washed twice with PBS and fixed in 4% paraformaldehyde at 37°C for 10 min . The reaction was stopped on ice for 2 min , then cells were pelleted and resuspended in 100% methanol at −20°C for permeabilization . For immunostaining , we used the CaspGLOW Fluorescein Active Caspase-3 Staining Kit ( BioVision # K183-25 ) and propidium iodide ( PI ) counterstain at a final concentration of 8 µg/ml . Histograms were recorded on an analytical , triple-laser equipped flow cytometer ( LSRII , Becton Dickinson ) using Sapphire 488 nm solid state laser excitation of 5 ( 6 ) -fluorescein isothiocyanate ( FITC ) and propidium iodide ( PI ) with appropriate filter sets to discern fluorescence intensity of different emission wavelengths . Quantification of cell distributions was by FACSDiva Software , version 6 . 1 . | Women with one strong BRCA2 ( FANCD1 ) mutation have high risks of breast and ovarian cancer . People with two mild BRCA2 ( FANCD1 ) mutations develop Fanconi Anemia , which reduces DNA repair leading to genome instability , small gonads , infertility , and cancer . Humans and mice lacking BRCA2 activity die before birth . We discovered that zebrafish brca2 mutants show chromosome instability and small gonads , and they develop only as sterile adult males . Female-to-male sex reversal is due to oocyte death during sex determination . Normal animals expressed brca2 in developing eggs and sperm that are repairing DNA breaks associated with genetic reshuffling . Normal developing eggs localized brca2 RNA near the nucleus , suggesting a role in protecting rapidly dividing early embryonic cells . Sperm-forming cells died in adult mutant males . Inhibition of cell death rescued sex reversal , but not fertility . Rescued females developed invasive ovarian tumors and formed eggs with abnormal nuclear architecture . The novel role of Brca2 in organizing the vertebrate egg nucleus may provide new insights into the origin of ovarian cancer . These results validate zebrafish as a model for human BRCA2-related diseases and provide a tool for the identification of substances that can rescue zebrafish brca2 mutants and thus become candidates for therapeutic molecules for human disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"developmental",
"biology/germ",
"cells",
"genetics",
"and",
"genomics/animal",
"genetics",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"cell",
"biology/nuclear",
"structure",
"and",
"function",
"cell",
"biology/cell",
"signaling",
"genetics",
"and"... | 2011 | Roles of brca2 (fancd1) in Oocyte Nuclear Architecture, Gametogenesis, Gonad Tumors, and Genome Stability in Zebrafish |
The Gcm/Glide transcription factor is transiently expressed and required in the Drosophila nervous system . Threshold Gcm/Glide levels control the glial versus neuronal fate choice , and its perdurance triggers excessive gliogenesis , showing that its tight and dynamic regulation ensures the proper balance between neurons and glia . Here , we present a genetic screen for potential gcm/glide interactors and identify genes encoding chromatin factors of the Trithorax and of the Polycomb groups . These proteins maintain the heritable epigenetic state , among others , of HOX genes throughout development , but their regulatory role on transiently expressed genes remains elusive . Here we show that Polycomb negatively affects Gcm/Glide autoregulation , a positive feedback loop that allows timely accumulation of Gcm/Glide threshold levels . Such temporal fine-tuning of gene expression tightly controls gliogenesis . This work performed at the levels of individual cells reveals an undescribed mode of Polycomb action in the modulation of transiently expressed fate determinants and hence in the acquisition of specific cell identity in the nervous system .
One of the most challenging issues in biology is to elucidate the mechanisms underlying cell fate determination and maintenance . The Drosophila melanogaster Glial cell missing/Glial cell deficient transcription factor ( Gcm/Glide , referred throughout the text to as Gcm ) is transiently expressed and is key to decide between glial and neuronal fates in the multipotent neural precursors [1]–[6] . Threshold levels of Gcm are necessary and sufficient to induce gliogenesis and the tight regulation of its expression prevents defective/excessive gliogenesis [7]–[11] . These features make Gcm an ideal tool to study cell differentiation and plasticity . Two major classes of proteins that modify the chromatin structure and its condensation state , the Polycomb group ( PcG ) and the Trithorax group ( TrxG ) , are known as critical regulators of HOX transcription factors , which act as molecular switches that are maintained in a silent or in an active state [12] . PcG and TrxG proteins act in large multimeric complexes that bind specific DNA regions called Polycomb ( or Trithorax ) response elements ( respectively PREs and TREs ) [13] . PcG and TrxG complexes trigger posttranslational modification of histone tails that have opposite effects on gene activity , mainly methylation of H3K27 induced by PcG complexes ( negative regulation ) and methylation of H3K4 , H3K36 as well as acetylation of H3K27 by TrxG complexes ( positive regulation ) ( [12] , [14] and references therein ) . PcG proteins enter two main conserved complexes called Polycomb Repressive Complex 1 and 2 ( PRC1 and PRC2 ) . The latter is formed by four core components including , in flies , Enhancer of zeste ( E ( z ) ) , and catalyzes the reaction that leads to di- and tri-methylation of H3K27 . This epigenetic mark is recognized by Polycomb ( Pc ) , which belongs to the PRC1 complex . Recent chromatin immunoprecipitation studies have shown that PcG and TrxG binding is also associated with dynamic transcriptional states modulating different processes including mitogenic pathways and progression from multipotency to differentiation ( [12] , [15]–[19] and references therein ) . Understanding the mode of action of PcG and TrxG proteins in dynamic processes , however , requires analyses at the level of identified cells and times . This is particularly important for developmental genes that are expressed transiently and in specific cell populations . The present in vivo study analyzes the role of Pc in fly gliogenesis . To identify components and regulators of the Gcm pathway , we designed a screen for genetic modifiers of a dominant phenotype due to gcm ectopic expression and identified PcG and TrxG proteins . Importantly , mutations in PcG components and in TrxG members found in chromatin remodeling complexes enhance the gcm dominant phenotype , whereas mutations in TrxG proteins known to specifically counteract PcG function rescue it . This suggests that a balanced action of these chromatin modifiers regulate Gcm function . Moreover , we demonstrate that the gcm regulatory sequences carry a PRE and are bound by Pc . Finally , Pc inhibits the autoregulatory loop ensuring threshold Gcm levels [7] and hence gliogenesis . To our knowledge , this is the first direct evidence that PcG proteins negatively modulate a transiently expressed fate determinant , thereby affecting a specific lineage in the nervous system .
The need of tight Gcm regulation prompted us to screen for interactors using a sensitized background . This approach allows the dissection of molecular cascades when the loss of a gene product is embryonic lethal . The Drosophila thorax ( notum ) carries a stereotyped number of sensory organs called macrochaete or bristles . gcmPyx/+ flies ectopically express gcm in the larval notum , which triggers the differentiation of supernumerary sensory organ precursors ( SOPs ) and bristles ( Figure 1A–1C ) [20] . gcmPyx/+ females show , in average , 18 , 5 bristles instead of the 11/heminotum typical of wild-type ( wt ) animals . Using large overlapping deficiencies ( 67–75% genome coverage , Deficiency kit , Bloomington ) , we performed a primary screen and identified 42 genomic regions that dominantly enhance or suppress the gcmPyx dominant phenotype when deleted ( Figure 1D–1E , Figure S1Aa and S1B ) . These regions were selected for quantitative analyses ( Figure S1Ab ) , which identified weakly and strongly modifying deficiencies . We further analyzed the latter ones ( Figure S1Ab , S1B ) and identified 28 interacting genomic regions . A secondary quantitative screen with smaller deficiencies ( Figure S1Ac , Table 1 ) allowed us to identify those that act as strong modifiers , based on statistical analyses . Single gene loss of function mutations within those deficiencies were then analyzed and the interaction was confirmed for 18 of them ( Figure S1Ad , Figure S2 , Table 1 ) . In sum , the Deficiency kit allowed us to identify large interacting regions and to progressively refine the analysis to single mutations . To evaluate the specificity and the sensitivity of the screen , we asked whether the selected deficiencies eliminate genes expected to interact with gcm . The gcmPyx phenotype correlates with the ectopic formation of proneural territories and precursors of the central ( CNS ) and peripheral ( PNS ) nervous systems , neuroblasts ( NB ) and SOPs , respectively [21] , [22] . Thus , mutations of NB/SOP specific genes should act as gcmPyx suppressors and indeed , the large and the small deficiencies covering three genes – escargot ( esg ) , worniu ( wor ) and snail ( sna ) – expressed in most embryonic NBs act as gcmPyx suppressors ( Table 1 , Figure S1C ) . Testing single gene loss of functions confirmed that sna and esg mutations act as gcmPyx suppressors . Accordingly , esg overexpression triggers the opposite phenotype ( Figure S1C ) . Finally , genes as pimples ( pim ) and crooked legs ( crol ) , identified in a microarray as induced by Gcm [23] , were also identified in our screen ( Figure S1D ) . The fact that known and predicted gcm interactors were identified validates our screen and shows that the dominant bristle phenotype is a reliable and very sensitive readout . A genomic region identified in the screen covers the trxG gene brahma ( brm ) , which encodes a transcriptional coactivator related to yeast SWI/SNF proteins and plays a role in ATP-dependent nucleosomal remodeling [24] . The large and the small deficiencies covering brm , Df ( 3L ) brm11 , Df ( 3L ) th102 and , most importantly , a null brm allele , enhance the gcmPyx phenotype ( Figure 1F ) . To extend our findings , we tested osa , an integral component of the Brahma complex [25] . osa loss of function also enhances the gcmPyx phenotype , whereas osa gain of function ( GOF: hs-Gal4;UAS-osa flies ) suppresses it ( Figure 1G , 1I ) . Thus , osa acts as brm , moreover , double brm/osa heterozygous mutants show an even stronger phenotype . Furthermore , a deficiency covering Enhancer of bithorax ( E ( bx ) ) and the E ( bx ) mutation itself enhance the gcmPyx phenotype ( Table 1 , Figure 1G , 1I ) . Interestingly , E ( bx ) ( also called NURF301 ) encodes a transcription coactivator that belongs to the ISWI chromatin remodeler complex , another TrxG complex , and negatively regulates the JAK-STAT pathway [26] , which is known to interact with gcm [27] . We then tested members of two TrxG complexes that specifically counteract Pc function . Trx is a SET-domain containing protein able to induce H3K4 methylation [28] . It has been purified as a subunit of the Drosophila COMPASS-like complex [29] and of the TAC1 complex that combines histone methylating and acetylating activities ( reviewed in [30] ) . The trx null mutation acts as a suppressor of the gcmPyx phenotype ( Figure 1G , 1I ) . Ash1 is a SET-domain protein reported to have histone methyltransferase activity [30]: its null mutation also suppresses the gcmPyx phenotype ( Figure 1G , 1I ) . Finally , the Drosophila CREBS-binding protein ( dCBP ) encoded by nejire ( nej ) is responsible for H3K27 acetylation [31] and is associated with both TAC1 and ASH1 complexes . The nej null mutation suppresses the gcmPyx phenotype ( Figure 1G , 1I ) . In conclusion , we found that mutations in TrxG proteins known to specifically counteract PcG function [12] act as suppressors of the gcmPyx phenotype , whereas TrxG members found in chromatin remodeling complexes that are involved in more general transcriptional regulation act as enhancers . We therefore tested members of the two PcG complexes , PRC1 ( Pc ) ( three null alleles ) and PRC2 ( ( esc ) , E ( z ) ) , as well as the PcG protein recruiter pipsqueak ( psq ) . Mutations in all four genes enhance the gcmPyx phenotype ( Figure 1H , 1I ) . Thus , PcG mutations act in the same way as mutations in the TrxG genes brm and osa , but have opposing effects compared to mutations in the TrxG genes Ash1 , trx and nej . This suggests that a balanced action of these chromatin modifiers regulate gcm function . In sum , the screen allowed the identification of several chromatin factors as gcm genetic interactors . gcm was identified as a putative Pc target in genome-wide chromatin immunoprecipitation ( ChIP ) studies on Drosophila embryos and different cell lines [32]–[34] , we therefore focused on this chromatin factor . As seen in Figure 2B , a Polycomb Response Element ( PRE ) is present around the transcription start sites ( TSS ) of gcm and gcm2 , which are organized head to head in a 30 kb region [35] . PRC1 binding at the TSS is accompanied by the H3K27 methylation mark ( H3K27me3 ) , the profile of which is much broader , extending throughout the gcm-gcm2 5′ regulatory region . As expected , the profile of H3K4methylation complements that of H3K27me3 ( Figure 2B ) . Pc binding was further validated and quantified by qChIP analysis on specific regions including the TSS region for each gene ( gcm , gcm2 ) , or an adjacent region ( GlacAT ) and a negative control ( Rp49 ) ( Figure 2A , 2B ) . We then asked whether the upstream region of the gcm gene bound by PcG proteins is able to recruit PcG proteins in transgenic assays . For this , we examined PcG binding to a transgene containing the upstream region of the gcm locus on salivary gland chromosomes by Immuno-FISH experiments . Similar to the endogenous gcm locus , which associates with both Pc and Ph proteins ( Figure 2E–2F′″ ) , a transgene carrying a gcm construct including 9 kb from the promoter region ( 9 kb gcm ) induces the recruitment of PcG proteins to an ectopic site ( Figure 2B , 2G–2Hb′″ ) . Interestingly , a transgene carrying a shorter construct ( 2 kb gcm ) is not able to efficiently recruit PcG proteins ( Figure 2B , Figure S3 ) . Importantly , this shorter construct triggers very limited rescue when reintroduced in gcm mutant embryos , whereas the 9 kb gcm construct rescues the embryonic mutant phenotype almost completely [8] , suggesting a correlation between Pc binding and transgene activity . Of note , the transgenes do not contain gcm2 , excluding the requirement of a gene complex for Pc binding . Moreover , gcm2 plays a minor role in gliogenesis and its mutation is viable [35] allowing us to focus on gcm . Finally , we tested the 9 kb construct for pairing sensitive silencing ( PSS ) , as transgenes carrying PREs/TREs in Drosophila have been shown to share this property ( [36] , [37] ) . Transgenic flies carrying the mini-white gene typically have eye colors ranging from yellow to orange in a white mutant background . Normally , flies that are homozygous for such a transgene have a darker eye color than heterozygotes , as the genetic dose of mini-white is doubled . However , with transgenes carrying PRE/TREs , the eye color is similar in homozygotes and heterozygotes or even darker in the latter . This is what we also observed in our transgenic lines ( Figure 2C–2D ) . Altogether , these data indicate that the gcm promoter region contains a PRE and suggest that PcG proteins directly regulate gcm expression . We next scored for Pc gcm interaction in a physiological asset , i . e . , in loss of function conditions for both genes . The gcm-Gal4 line , an insertion in the gcm locus , is a hypomorphic semiviable allele in homozygous conditions and can be used to follow gcm activation and glial cells using a UAS-green fluorescent protein ( GFP ) line [38]–[40] . We analyzed the expression of GFP as well as that of an independent glial marker ( Repo ) and a neuronal marker ( Elav ) in homozygous gcm-Gal4 , UAS-GFP ( referred to as gcm-Gal4 ) animals and in homozygous gcm-Gal4 animals that are also heterozygous for Pc . As a control , we used heterozygous gcm-Gal4 animals . The Drosophila wing contains two major nerves , L1 and L3 , covered by glia that depend on gcm [41] ( Figure 3A , 3B ) . Because of their simple organization , we focused on the L3 glia , which arise from three SOPs called L3-3 , L3-1 and L3-v . Each SOP produces a sensory neuron and a glial precursor ( GP ) that proliferates and produces four to eight glia that are GFP+ ( Figure 3C–3H ) . gcm-Gal4 homozygous flies show the glia to neuron transformation observed in gcm null clones [41] , albeit at much lower penetrance ( Figure 3I–3M ) . To analyze the phenotype at single cell level , we followed glia from a specific lineage , the L3-v , at the time the GP is generated . At this stage , control L3-v lineages contain a GFP+ cell that expresses Repo and a neuron that expresses Elav ( Figure 3S , 3T–3T″ ) . In gcm-Gal4 homozygous animals , the GFP+ cell expresses Elav rather than Repo ( 9% penetrance ) ( Figure 3S , 3U–3U″ ) . By 24 hr after puparium formation ( APF ) , the number of GFP+ and Repo+ cells present in the control animals increases , whereas only one GFP+ cell is present in the transformed lineage , due to lack of proliferation , and this cell is a neuron ( Figure 3C–3M ) . The penetrance of ectopic neurons does not decrease during development ( 16 and 18% by 20 and 24 hr APF , respectively , Figure 3S ) , indicating that low Gcm levels trigger a stable fate conversion; a similar phenotype was observed on L1 glia ( Kumar and Giangrande , unpublished data ) . Based on the genetic data , we then asked whether Pc downregulation rescues the phenotype of homozygous gcm-Gal4 wings . Indeed , no evidence of stable glia to neuron transformation was found in homozygous gcm-Gal4 wings that carry only one Pc functional allele ( Figure 3N–3S ) . The phenotype was verified at early and at late stages of wing development , to exclude the possibility of unstable rescue . These data strongly suggest that Pc affects gcm expression in the gcm-Gal4 line . In order to extend the above findings , we analyzed late gliogenesis upon lowering the dose of Pc . Differentiated gcm-Gal4 homozygous wings carry fewer glia than wt wings in which the three glial precursors have divided more than once in most of the cases ( Figure S4A , 24 hr APF wings ) . Given the low penetrance of the fate transformation phenotype , this suggests an additional , later , effect on the glial cell number . To clarify the nature of the phenotype we counted the Repo+ cells just after the first division of the three L3 GPs in gcm-Gal4 homozygous wings that showed no fate transformation . We could confirm a decreased number of cells ( Figure 4B , 4E , S4B , 20 hr APF wings ) , complementing the finding that sustained gcm expression induces glial overproliferation ( embryo: [11]; wing: Kumar and Giangrande , unpublished data ) . Of note , the gcm-Gal4/+ wings already show a minor but consistent defect as there are cases in which the three GPs have not proliferated yet , which does not occur in wild type wings of the same stage ( Figure 4A and 4E , Figure S4B ) . Moreover , heterozygous wings show a high variance in the number of Repo+ cells . Finally , homozygous gcm-Gal4 wings expressing a single Pc show a higher number of glia compared to those found in homozygous gcm-Gal4 wings ( Figure 4D and 4E , Figure S4B ) , confirming that Pc negatively controls Gcm . This was confirmed by the significant P values obtained with different robust non-parametric tests comparing the homozygous wings with the homozygous wings that carry one dose of Pc ( Mann Whitney test P = 0 , 0127; Wilcoxon test P = 0 , 0122 ) . Moreover , one-way Anova comparison of the three genotypes ( gcm-Gal4/+ , gcm-Gal4 and gcm-Gal4; Pc/+ ) also produces a significant value ( 0 , 0028 ) . These data indicate a partial rescue of the gcm-Gal4 proliferation phenotype by Pc , the moderate difference likely depending on the fact that only one dose of Pc is deleted . To understand the role of Pc in gliogenesis , we also analyzed Pc mutant animals in an otherwise wt background and asked whether the mutation affects the number of glia ( Figure 4F–4H , Figure S4C 24 hr APF wings ) and the frequency of glial dividing cells ( Figure 4I ) . Since removing Pc completely leads to pleiotropic defects , we used heterozygous Pc animals and counted the number of Repo+ cells on the L1 nerve , which shows massive gliogenesis , compared to the sparse glial cells present on the L3 nerve [41] . While the number of Repo+ cells increases very moderately in Pc/+ compared to wt wings ( P = 0 , 03 ) , a stronger , significant , increase is observed in E ( z ) /+ wings ( P = 0 , 0007 ) , which have a compromised PRC2 , and an even stronger phenotype is observed in double heterozygous Pc/E ( z ) animals ( P = 3 , 9×10−6 ) , which display compromised PRC2 and PRC1 ( Figure 4F–4H , Figure S4C ) . Finally , we labeled wings with Repo and phospho-histone H3 ( PH3 ) as a mitotic marker . By 24 hr APF , the Repo/PH3+ cells are very rare in wt wings ( 1 Repo-PH3+ cell in 1/11 wings ) ( Figure 4I ) . E ( z ) /+ or Pc/E ( z ) double heterozygous animals , which show the most significant increase in glial cell number , show a significant increase in the number of wings with proliferating glia , whereas Pc/+ animals , in which the increase in glial cell number very small , do not . Thus , PcG proteins likely synergize and affect both glial differentiation and proliferation . We next analyzed the role of Pc on the gcm expression profile . Positional cues first trigger initiation of transcription , then Gcm positively autoregulates [7] and , as the glial fate is established , gcm expression progressively decreases so that its transcripts are no longer present in mature glia [42] . We analyzed the initiation of gcm transcription in gcm-Gal4/+; Pc/+ wings . Previous analyses showed that the gcm RNA becomes detectable by 8–9 hr APF ( Van de Bor and Giangrande , unpublished data ) . We therefore analyzed 7–8 hr APF wings and found that the GFP appears at the same time as in wt animals ( data not shown ) . Since the binary Gal4 system may not faithfully reproduce the temporal pattern , we analyzed wings carrying one dose of Pc and the P-mediated insertional gcmrA87 allele expressing the LacZ reporter and confirmed that the β-Gal labeling starts as in wt animals ( Figure S5 ) . The finding that Pc does not affect initiation of gcm expression is in line with the wt number of GFP+ cells observed in homozygous gcm-Gal4 wings at early stages , even in cases in which glial cells convert into neurons . We also performed in situ hybridization with a gcm-specific probe in Pc/+ wings . We took advantage of the supernumerary glia phenotype to see whether Pc helps repressing the maintenance of gcm expression . gcm transcripts are well visible on both wt and Pc/+ wings by 19 hr APF , a stage at which the glial precursors have differentiated ( Figure 5A , 5D ) . By 24 hr APF , however , they are absent in wt , but still present in Pc/+ wings ( Figure 5B , 5E ) , which correlates with the slight increase in glial number observed in Pc/+ animals . Interestingly , Pc/+ wings do not show gcm expression at ectopic positions , suggesting that the absence of Pc induces a failure in repressing gcm maintenance rather than a global loss of silencing in whole tissues . We extended the data by analyzing other stages and tissues . In the brain , gcm is expressed in several cell populations: GPC and its glial progeny , lamina neurons , central brain neurons and medulla glia [39] , [40] , [43] , [44] . We focused on gcm expression at the position of lamina glial precursors ( GPCs ) , which produce numerous cells that migrate and form the glia of the lamina visual ganglion ( Figure 5K ) [39] , [40] , [43] . For the sake of simplicity , we analyzed the optic lobes at a stage at which gcm is detectable in the GPC area but just starts being expressed in the other regions . In wt animals , gcm expression fades away as glia differentiate and migrate ( Figure 5G , 5K , 5L ) , whereas in Pc/+ animals gcm is expressed in an expanded area ( Figure 5H , 5L ) . Moreover , gcm is overexpressed in brm/+ brains and this phenotype is suppressed in brm , trx/+ animals . This shows that brm acts similar to Pc on gcm expression , and both act antagonistically to trx , in line with the genetic data ( Figure 5H–5J , 5L ) . All the phenotypes were quantified by comparing the intensity and the area of the gcm signal ( see Text S1 , Figure S4E ) . In the double mutant , the area of labeling resembles that observed in wt animals and the intensity of the signal is even lower than that observed in wt animals . Future analyses will determine whether the increase of gcm expression in the mutant backgrounds reflects longer perdurance in migrating glia , production of more glia or production of more glial progenitor cells in the larval lamina . In some preparations , labeling in other regions is also observed , depending on sample orientation . Even though we cannot formally exclude the possibility that this represents ectopic labeling , these regions correspond to the other positions at which gcm accumulates at slightly later stages in wild type animals , suggesting that in those regions as well Pc negatively controls gcm expression . Finally , we analyzed gcm transcripts in Pc embryos . In wt animals , gcm is expressed at early stages of glial development and transcripts subsequently fade away , first in the ventral cord and then in the brain [42] . The most frequent phenotype of Pc mutant embryos is a persisting gcm expression in the brain , but we also found extreme cases of late gcm expression in the ventral cord ( Figure 5C , 5F ) . The embryonic and the postembryonic brains contain too compact and numerous glia and the perdurance in the ventral cord is a rare event , likely due to the Pc maternal component . Although these tissues/stages do not allow quantitative analyses of glial cells , the expression data and the wing phenotype strongly suggest that Pc represses gcm maintenance . Altogether , our observations highlight the importance of Pc in tightly regulating Gcm levels . To assess whether Pc directly represses gcm , we used in vivo and in vitro assays . Gcm directly and positively autoregulates and alteration of this feedback loop severely affects its gliogenic potential , providing further evidence for the importance of Gcm maintenance at a precise developmental time [7] , [9] . In vivo autoregulation can be documented in gain of function experiments by using the gcmrA87 allele . We asked whether Pc negatively controls Gcm autoregulation by comparing animals that simultaneously overexpress Gcm and Pc to control animals that only overexpress Gcm . Compared to controls , Pc and Gcm cooverexpressing embryos show a drastic reduction in the number of β-Gal+ cells as well in the intensity of β-Gal labeling ( Figure 6A , 6C ) . Accordingly , co-overexpression reduces the number of ectopic glia as assessed by the Repo marker ( Figure 6D , 6F , 6G , 6I ) . Moreover , and in line with these results , overexpressing Gcm in Pc loss of function embryos triggers a significant increase in the number of autoregulating cells compared to that observed in control animals ( Figure 6A , 6B ) . Accordingly , these animals show an increased number of ectopic Repo+ cells ( Figure 6D , 6E , 6G , 6H ) . These data were quantified upon counting the number of β-Gal+ and Repo+ cells ( Figure 6J ) . Loss and gain of function of Pc do not , on their own , alter the expression of the Repo marker ( Figure S6 ) . To evaluate whether the inhibitory effects of Pc in the Gcm pathway are direct , we used transactivation assays in which we transfected S2 cells with a Gcm expression vector and a reporter of its activity in presence or in absence of a Pc expression vector . We first analyzed the repo promoter , a major direct Gcm target that contains several Glide Binding Sites ( GBSs ) [45] ( Figure S7D ) . This promoter is inactive in S2 cells , but Gcm expression is sufficient to activate it . Upon cotransfection with Gcm and Pc expression vectors , however , the transactivation induced by Gcm decreases significantly ( Figure S7C , S7D ) . We repeated the same type of assay using a second , transiently expressed , promoter depending on gcm . The gcm2 2 kb proximal promoter contains four GBSs and was previously shown to be activated by Gcm in transfection assays [35] ( Figure 6K , 6L ) , more robustly than the gcm 2 kb promoter itself , which only contains one GBS . As for repo , the cotransfection with Gcm and Pc reduces the activation of the gcm2 promoter . Thus , Pc represses the expression of Gcm stably and transiently expressed targets . In sum , the above data support the hypothesis that Pc represses gcm autoregulation and Gcm downstream targets , thereby inhibiting glial development .
The genetic screen over a sensitized background proved to be an extremely sensitive tool , as it allowed us to identify several genes that in heterozygous conditions are able to modify the strong dominant gcmPyx phenotype . The screen also provided hints onto the function of the interactors , suppressors or enhancers of a given phenotype . For example , sna and esg act as gcmPyx suppressors , in line with the fact that gcmPyx triggers the expression of NB-specific genes [20] . Identifying an interactor provided an entry point to find members of the same pathway that were initially underscored because located in deficiencies with moderate phenotypes ( perhaps due to the presence of genes with opposite effects ) or in regions that were not covered by the deficiencies . In the first case is Pc , in the second are osa and Ash1 ( Figure S2 ) . The screen also identified members of other signaling pathways ( Table 1 , Figure S2 ) . One of them depends on Notch ( N ) , which controls gcm expression [20] . While the used Deficiency kit does not cover N itself , we identified Suppressor of Hairless ( Su ( H ) ) , which regulates the transcription of the N targets , and Lethal ( 2 ) giant disc 1 , which negatively regulates N receptor trafficking ( [46] and references therein ) . We also tested and validated the genetic interaction with other members of the cascade , including N , its ligand Delta , one of its targets , Enhancer of split , and Groucho , a transcriptional repressor and a partner of Su ( H ) . Future studies will dissect the role of this and of the other pathways on the Gcm cascade . Several TrxG proteins act as genetic modifiers of the gcmPyx phenotype . TrxG proteins were initially identified as positive regulators of HOX genes and considered as PcG counteractors . In recent years , however , it has become evident that they have a much wider role in gene regulation and it is unclear whether they mainly antagonize PcG functions or whether they globally control gene expression [12] . Interestingly , the three TrxG proteins that behave as positive regulators , Trx , Ash1 and dCBP , are found in TAC and ASH1 complexes that contain a histone acetylation activity . The dCBP histone acetyltransferase present in these complexes acetylates H3K27 , a modification that is associated with PcG target genes when they are active [34] . This modification is incompatible with Pc dependent H3K27me3 , as these modifications occur on the same amino acid . Thus , Trx- and Ash1-associated dCBP might be a key player in counteracting PcG-dependent silencing of the gcm gene [31] . Future studies will address the role of dCBP onto the Gcm cascade . osa and brm act as negative regulators of gcm . TrxG proteins can form different complexes that have distinct properties and in some instances repress gene expression . For example , Trx and Brm , which belong to different molecular complexes [30] , act positively on the HOX genes and influence a homeotic transformation phenotype in the same way [47] , however , Brm-containing complexes mediate transcriptional repression of genes other than the HOX genes [48] . The emerging view is that the SWI/SNF TrxG proteins act as transcriptional activators or repressors depending on the temporal and spatial context [49] . Further studies will determine whether the TrxG proteins acting as negative regulators of gcm directly repress its expression or induce a gcm repressor . PcG proteins repress homeotic genes to ensure the maintenance of transcriptional states and provide a cellular memory that is transmitted upon cell division , in contrast , their mode of action in the control of more dynamic processes remains elusive . We show in vivo that members of the PcG negatively regulate the gcm pathway during glial fate establishment and proliferation . At least in the first step , a process based on cell memory can be excluded , as Pc acts prior to the division of the GP , the cell in which gcm starts being expressed [41] . The qChIP assay as well as the expression , the S2 cell transfection and the autoregulation data strongly suggest that Pc directly represses gcm transcription maintenance . In addition , the phenotypes observed upon changing the relative gene dosage indicate that Pc and gcm need to be present in appropriate amounts . The importance of an adequate balance between positive ( Gcm ) and negative ( Pc ) factors in the establishment of the glial fate is also provided by a rare phenotype observed in a gcm-Gal4; Pc/+ background ( 1/17 wings ) in which the GFP+ cell expresses Repo and Elav , indicating an intermediate glial/neuronal state ( Figure 3V–3V″ ) . Thus , Pc acts by finely tuning a transiently expressed fate determinant . We speculate that the role and the mode of action of chromatin factors depend on the target . HOX promoters , which require to stay in an ON or OFF state , may involve strong binding/high accumulation of chromatin regulators and several studies have already shown that HOX activators drastically reduce K27me3 and also PcG protein binding ( Figure 7A ) [34] , [50] , [51] . More dynamically expressed genes may involve less strong binding , a configuration that allows modulation of gene expression . From a mechanistic point of view , as the activator of the transiently expressed genes disappears , PcG proteins may gradually bind and turn these genes OFF ( Figure 7B ) although we cannot formally exclude that PcG proteins may simply provide a constant repressive background as a threshold for activation ( Figure 7C ) . In line with these hypotheses , HOX and Gcm display different behaviors . A fragment of 219bp from Fab7 , the classical PRE described on a HOX promoter , is sufficient to recruit PcG proteins on salivary glands [52] , whereas a 2 kb gcm carrying the PRE seems very inefficient . In addition , the intensity of Pc , Ph and ‘recruiters’ peaks onto the gcm promoter is very low , definitely weaker compared to those found on the classical HOX PRE ( Figure S8 ) . Finally , the heterozygous Pc/+ mutation only temporarily prolongs gcm expression ( Figure S5I ) , whereas it produces a long lasting HOX-dependent phenotype [53] , [54] . Understanding the precise molecular events will require the development of new tools and the in vivo analysis of chromatin organization at the level of specific cell types or in single cells . Our data nevertheless clearly show that Gcm and Pc compete with each other: PcG proteins bind gcm genes as well as repo ( Figure 2 , Figure S7 , Figure S8 ) [33] and counteract Gcm activity . We therefore speculate that Gcm displaces Pc from its target promoters , including itself , which would explain how a general chromatin regulator impinges onto a cell-specific transcriptional program . In mammals as well it has been suggested that cell fate transcription factors play a role in PcG recruitment and displacement and some of them were shown to be PcG targets ( [55] and reference therein ) . Finally , 63 genes are common Pc and Gcm targets , as revealed by analyzing the Pc binding sites in embryos and in cell lines ( from [33] and [34] ) and the genes positively regulated by Gcm identified by microarray ( from [23] ) . Clearly , genome-wide screens for direct Gcm targets will be necessary to support the hypothesis of Pc displacement . These studies will also assess whether the impact of the PRCs on glial proliferation is direct or mediated by Gcm . The rescue of the gcm-dependent phenotype upon Pc downregulation indicates a role for this chromatin factor in glial repression . Interestingly , upregulating or downregulating Pc does not per se produce the opposite fate transformation ( Figure S4D ) , whereas it does modify the number of glia , showing that distinct protein levels are required in different processes . In vertebrates , the PRC2 is also involved in the production of glial cells , which differentiate after a wave of neurogenesis . However , different results were obtained depending on the experimental asset . Livesey and collaborators ( [56] ) deleted Ezh2 constitutively , thereby altering the balance between self-renewal and differentiation , and found precocious astrocyte differentiation . In contrast , Gotoh and collaborators [57] used a conditional Ezh2 knockout and documented a decrease in astrocyte differentiation . In the first case , the authors speculated that the altered timing of neurogenesis and accelerated onset of gliogenesis are secondary to the primary function of PRC2 in cortical progenitor cells . In the second report , it was shown that Ezh2 represses Neurogenin1 , which controls timing during corticogenesis and therefore the relative production of neurons and glia . While these studies indicate the importance of chromatin modifiers in the nervous system , they do not clarify the role of PRCs in gliogenesis . In our study , the combined use of sensitive tools demonstrates that the Pc chromatin factor directly inhibits gliogenesis and identifies gcm as a major target in the pathway . First , we used sensitized backgrounds rather than total knockouts , which makes it possible to score for subtle phenotypes . Second , we analyzed the mutants at the single cell resolution and therefore scored for direct , cell autonomous , effects of the Pc mutation . Third , we analyzed a gene that plays an instructive role rather than simply being permissive for gliogenesis . Fourth , gcm carries a functional PRE and competes with Pc on its targets . Altogether , these findings reinforce the view that distinct chromatin states characterize specific cell fates , as also illustrated by the low levels of histone acetylation observed in both fly and vertebrate glia [6] , [58] .
Flies were grown on standard cornmeal/molasses medium at 25°C . The deficiency kit was obtained from the Bloomington Stock Center ( Bloomington , IN ) , see Supplementary Material and Methods . For the qualitative screen: for each cross ( 180 deficiencies ) , double heterozygous females carrying the gcmPyx allele and a deficiency were scored for the supernumerary bristle phenotype and compared to sibling females carrying the gcmPyx allele and the balancer from the deficiency stock . This allowed us to classify each deficiency as gcmPyx modifier or not modifier ( Figure S1Aa ) . 75 deficiencies covering 42 genomic regions were selected for quantitative analyses ( Figure S1Ab ) ; for each genotype we counted the bristles from 10–80 heminota . The flow chart in Figure S1A shows the details of the screen . Average values +/− SEM were calculated and , for genotype comparisons , the statistical significance was estimated by t-test . To overexpress esg or osa , respectively , w; EP ( 2 ) 0684/CyO or w; P{w[+mC] = UAS-osa}s2/CyO females were crossed with w; gcmPyx/Sp; hs-Gal4/Sb males . A 30 minute heat-shock pulse on 2nd instar larvae was performed at 37°C . qChIP was performed as in [33] . Primers are listed in Figure S3A . these assays were performed as in [41] and [44] . For the antibody list as well as for the protocol of wing and embryo mounting and analysis by confocal microscopy , see Text S1 . Repo and β-Gal positive cells from embryonic VC were subjected to quantification in 3D image using Imaris 7 . 2 software . Masks were generated as a region of interest for three thoracic segments along the z-stack , then volume image was visualized and the “crop 3D” function was applied to isolate the region of interest . Voxels ( volume picture element ) corresponding to cells were identified based on size and intensity . Then automatic voxel ( cell ) counting was performed in the region of interest . t-test was used to quantify the difference between genotypes . For immuno-FISH staining on polytene chromosomes [59] , three consequent probes covering around 3 kb around gcm TSS were used , see Figure S3 . Unless specified , all quantitative analyses used the t-test . The gcm2 promoter construct is pBLCAT6-1 . 96 from [35] . The 4 , 3 kb of the repo promoter [45] was cloned into the pRed H-Stinger vector ( Berzsenyi and Giangrande , unpublished data ) . pPAC-gcm is described in [7] . UAS-gcm is described in [42] . pPAC-Pc and UAS-Pc were obtained by cloning the entire Pc cDNA in backbone vectors . pPAC-lacZ was a gift from T . Cook . Transient transfection of Drosophila S2 cells [60] was performed using Effectene ( Qiagen ) according to the manufacturer's instructions using 3 µg of total DNA . For CAT assay to evaluate the activation of the 2 kb gcm2 reporter construct ( pBLCAT6-1 . 96 ) , cells were harvested 48 hr after transfection and normalized for ß-Gal activity . CAT levels were determined using the CAT ELISA kit ( Roche ) . For repoRFP , images of cells were acquired 48 hr after transfection , and the green ( UAS-GFP ) /red ( repoRFP ) cells , were quantified automatically using the ImageJ software . | Epigenetic mechanisms are essential to define cell identity , and the Polycomb and the Trithorax Group proteins ( PcG and TrxG , respectively ) control the body plan by maintaining the epigenetic state of homeotic genes . PcG and TrxG act by triggering stable chromatin modifications that are “remembered” after cell division and keep gene expression in an OFF or ON state . Recent genome-wide analyses call for additional targets of PcG proteins , but the role of these chromatin factors in dynamic transcriptional states and/or in specific cell fates is difficult to apprehend , mostly because very sensitive readouts are required . This in vivo study performed at the single-cell level shows that PcG proteins affect the levels and the kinetics of the transiently expressed Drosophila glial determinant and transcription factor Gcm/Glide . Thus , PcG proteins also finely tune gene expression , and this is independent of memory mechanisms , suggesting that “transient” promoters may have a different affinity to PcG proteins compared to “stable” promoters . PcG proteins negatively affect Gcm/Glide autoregulation , thereby promoting neurogenesis at the expense of gliogenesis . Thus , PcG genes act in the fate choice between two types of differentiated cells , implying that distinct cell populations have specific requirements for general chromatin modifiers . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"developmental",
"neuroscience",
"molecular",
"development",
"genetics",
"gene",
"expression",
"biology",
"neuroscience",
"genetics",
"and",
"genomics",
"cell",
"fate",
"determination"
] | 2012 | Polycomb Controls Gliogenesis by Regulating the Transient Expression of the Gcm/Glide Fate Determinant |
The Serum and Glucocorticoid-regulated Kinase1 ( SGK1 ) gene is a target of the glucocorticoid receptor ( GR ) and is central to the stress response in many human tissues . Because environmental stress varies across habitats , we hypothesized that natural selection shaped the geographic distribution of genetic variants regulating the level of SGK1 expression following GR activation . By combining population genetics and molecular biology methods , we identified a variant ( rs9493857 ) with marked allele frequency differences between populations of African and European ancestry and with a strong correlation between allele frequency and latitude in worldwide population samples . This SNP is located in a GR-binding region upstream of SGK1 that was identified using a GR ChIP-chip . SNP rs9493857 also lies within a predicted binding site for Oct1 , a transcription factor known to cooperate with the GR in the transactivation of target genes . Using ChIP assays , we show that both GR and Oct1 bind to this region and that the ancestral allele at rs9493857 binds the GR-Oct1 complex more efficiently than the derived allele . Finally , using a reporter gene assay , we demonstrate that the ancestral allele is associated with increased glucocorticoid-dependent gene expression when compared to the derived allele . Our results suggest a novel paradigm in which hormonal responsiveness is modulated by sequence variation in the regulatory regions of nuclear receptor target genes . Identifying such functional variants may shed light on the mechanisms underlying inter-individual variation in response to environmental stressors and to hormonal therapy , as well as in the susceptibility to hormone-dependent diseases .
Substantial genetic and paleontological evidence supports the idea that humans originated in Sub-Saharan Africa and from there expanded across the globe ( [1] and references therein ) . During this dispersal , human populations encountered and settled into new environments that differed in climate , resource availability , pathogen exposure and other features that can challenge human homeostasis . Additional climatic as well as lifestyle changes , e . g . the retreat of the ice sheet and the agricultural transition , further contributed to the environmental diversity that humans adapted to . Many of these critical adaptations likely occurred at the genetic level through Darwinian selection of beneficial genotypes . When selective pressures vary across local environments , the geographic distribution of the advantageous genotypes and the resulting phenotypes are expected to follow distinctive patterns that mirror the presence and intensity of the selective pressure . For example , human skin pigmentation and body mass markedly differ across populations and are correlated with UV radiation and temperature , respectively [2] , [3] . In genome-wide studies , the analysis of allele frequency differences between populations has identified signals of adaptation in genes playing a role in skin pigmentation , host-pathogen interaction , lactase persistence , etc . [4]–[8] . In addition , genes that play a role in cortisol metabolism , sodium homeostasis , and arterial vessel tone were shown to harbor variants that are strongly correlated with latitude [9] , [10] . In these analyses , latitude is considered a proxy for climate; accordingly , these findings were interpreted as evidence for adaptation to heat stress and dehydration . More recently , variation in candidate genes for common metabolic disorders was also shown to be correlated with latitude and a set of climate variables that reflect the impact of cold and heat stress on energy homeostasis [11] . In higher organisms , homeostasis of key physiological processes is achieved through the neuroendocrine response to environmental challenge . This physiological response is in part mediated through the activation of nuclear hormone receptors via a stress-induced ligand ( e . g . , the adrenally secreted hormone cortisol ) and subsequent regulation of target gene expression [12] . Ultimately , nuclear receptors and their associated cofactors , in conjunction with cooperating transcription factors , recognize specific DNA sequences within regulatory regions of genes encoding key physiologic response proteins ( for review see [13] ) . In humans , the stress hormone cortisol mediates gene expression via the GR and , to a lesser extent , the mineralocorticoid receptor ( MR ) . Under conditions of environmental stress , including cold , heat , and dehydration , several stress-associated kinases are activated via rapid post-translational modification , commonly phosphorylation . In contrast , following exposure to physiological stressors , the SGK1 gene is immediately transcriptionally induced via the ligand-bound GR and MR , and its protein product is then constitutively phosphorylated via endogenous PI3-K activity [14] , [15] . The rapid transcriptional induction of SGK1 steady-state levels reflects SGK1's key role in the neuroendocrine response [16] , [17] . For example , SGK1 expression regulates sodium homeostasis in the kidney as well as enhances cell survival following exposure to apoptotic stress such as ultraviolet light and hyperosmolality [14] , [18]–[23] . Consistent with a key role in fundamental stress responses , SGK1 is highly conserved across distantly related species [24]–[26] . At the same time , subtle variation in SGK1's regulatory sequences is hypothesized to alter the threshold of SGK1's hormone-mediated induction , and hence increase or decrease SGK1's ultimate level of activity in response to a given environmental stressor . Under the assumption that the stress response pathway and , in particular , the SGK1 gene were targets of local selective pressures , we searched for genetic variants that influence SGK1 expression in response to stress . To this end , we combined population genetics , comparative genomics and molecular biology approaches to identify variants in candidate regulatory region and then tested them by means of functional assays . We found several variants approximately 30 kb upstream of the transcriptional start site ( TSS ) of SGK1 that show unusually large differences in allele frequencies between populations and that are strongly correlated with both latitude and climate variables . One of these variants lies within a predicted binding site for Oct1 , a transcription factor known to cooperate with the GR [27] . We show by chromatin immunoprecipitation ( ChIP ) assays that the ancestral allele of this variant ( inferred by comparison to the chimpanzee sequence ) results in more efficient binding of the GR-Oct1 complex to this sequence . Furthermore , reporter gene expression assays reveal higher levels of glucocorticoid-dependent transcription from the ancestral allele compared to the derived one ( i . e . , the allele inferred to have been introduced by mutation because it is not present in chimpanzee ) .
We used two approaches to search for signatures of adaptation to local environments in the genomic region surrounding the SGK1 gene . One was to quantify the difference in allele frequency between pairs of populations by means of the FST summary statistic [28] . The HapMap Phase II data were used in this analysis [29] . The other approach was to measure the correlation between allele frequencies in a large set of population samples and an environmental variable ( e . g . , latitude ) , which was assumed to be a good proxy for the selective pressure [11] . In this analysis , we used the Illumina HumanHap 650Y genotype data from the Human Genome Diversity Project ( HGDP ) panel [30] , [31] . At the genome-wide level , the geographic distribution of allele frequencies is mainly determined by the history of migration and population-specific demographic events . Therefore , to distinguish between the effect of population history alone versus that of natural selection , we compared the geographic distribution of genetic variants in the SGK1 region to that of variants from large genome-wide data sets . We calculated the FST value between CEPH Europeans and Yoruba for the SNPs in a region of 105 . 6 kb spanning and upstream of the SGK1 gene . As shown in Figure 1 , the FST values for seven out of 82 SNPs in this region fall in the top 5% of the empirical distribution for the >2 M HapMap SNPs . In particular , only 0 . 2% of the HapMap SNPs have an FST value higher than SNP rs9493857 ( shown in red in Figure 1C ) . These results suggest that the divergence of allele frequency between populations of European and Sub-Saharan African ancestry in the region upstream of SGK1 is greater than expected based on population history alone . To test if allele frequencies in the SGK1 upstream region also correlate with environmental variables , we examined the SNPs genotyped using the Illumina HumanHap 650Y chip in the HGDP panel; in addition , we genotyped two SNPs ( rs9493857 and rs1763502 ) in the same panel . Derived allele frequencies for the 25 SNPs analyzed in the HGDP populations are reported in Table S1 . Following the approach described in Hancock et al . ( 2008 ) [11] , we used two different methods to assess the relationship between allele frequency and environmental variables: Spearman rank correlation and Bayesian geographic analysis . The first one is a non-parametric method that does not assume a linear relationship between the variables . The second one is a model-based method that tests whether a linear relationship between allele frequency and a variable provides a significantly better fit to the data than the null model alone ( where the null model is given by a matrix of the covariance of allele frequencies between populations ) . The environmental variables included latitude and seven climate variables ( see Materials and Methods ) in the summer and winter seasons; because these variables are partially correlated , we reduced the dimensionality of the data by calculating their principal components and used these new variables to test the correlation with allele frequencies [11] . Eight of the 25 SGK1 SNPs genotyped in the HGDP panel are significantly ( p<0 . 05 ) correlated with at least one of the climate principal components or with latitude alone ( Table S3 ) . Among them , SNP rs9493857 is the most strongly correlated with latitude and is also significantly correlated with winter Principal Component 1 ( Figure 2 , Tables S2 and S3 ) . The results of the Bayesian geographic analysis provide more subtle signals , with only three of the 25 SGK1 SNPs showing a significant linear relationship with the climate principal components ( Tables S4 and S5 ) . We also looked for signatures of natural selection using other aspects of genetic variation data , including the haplotype structure and the allele frequency spectrum [5] , [32]–[36] . Unlike the analyses above , these tests did not detect strong signatures of positive selection ( Table S7 ) . This may be due to the fact that these tests are known to have inadequate power under a range of selection scenarios; for example , when natural selection acted on recessive variants or on variants present in the population at appreciable frequencies prior to the onset of selection [37] , [38] . Overall , the analyses of the geographic distribution of allele frequencies in the region upstream of SGK1 point to variants that may have been targets of local adaptation . The fact that these candidate selected variants lie in non-coding sequence and that SGK1 activity is primarily regulated by transcriptional induction [14] , [15] suggests that these variants may modulate SGK1 activity . However , the only established GR response element ( GRE ) in this region is located in the SGK1 promoter [39] , while most of the candidate SNPs are located >30 kb upstream of the TSS . To identify additional regulatory regions beyond the promoter , we performed a bioinformatics analysis to identify additional GREs [40] and Evolutionary Conserved Regions ( ECRs ) between human and dog , mouse , or opossum ( Figure 1 ) [41] . We ultimately considered only regions that contain at least two of the following three features: high FST SNPs , predicted GREs , or ECRs . We thereby narrowed down a region of >100 kb to three candidate regulatory regions ( respectively , 30 kb , 50 kb , and 70 kb upstream ) spanning a total of 10 kb . To further prioritize these three candidate regions ( and possibly identify additional ones ) , we performed a GR ChIP-chip assay in MCF10A-Myc mammary epithelial cells treated with the synthetic glucocorticoid ( GC ) dexamethasone ( 10−6 M ) . The immunoprecipitated DNA was hybridized onto the Affymetrix GeneChip Human Tiling 2 . 0R A Array ( which covers chromosomes 1 and 6 ) . GR binding regions ( GBRs ) were then identified using the MAT software [42] to analyze the data obtained from two independent experiments . By using a p-value cutoff of 10−3 , ChIP-chip identified six GBRs in the SGK1 region . Among these six GBRs , three nonoverlapping GBRs have a MAT p-value<10−5 and also contain two predicted GREs , three ECRs and SNP rs9493857 . Of the remaining GBRs , one is located in a region spanning intron 4 to intron 7 and contains a SNP previously implicated in risk to Type II Diabetes and hypertension [43] , [44] , the second is located 20 kb upstream of the TSS and is close ( <1 kb ) to a predicted GRE , and the third is located 85 kb upstream of the TSS and is 5 kb away from the closest predicted GRE . The GBRs close to a predicted GRE as well as the three candidate regulatory regions defined above ( 30 kb , 50 kb , and 70 kb upstream ) were re-sequenced in a panel of 28 individuals ( 14 Hausa from Cameroon and 14 Italians ) to determine whether additional variants with large allele frequency differences between Africans and Europeans exist ( see Figure 1 ) . We identified 39 SNPs that were not included in the HapMap data and calculated the FST values between Hausa and Italians for all SNPs identified by resequencing [45] . As shown in Table S6 , rs9493857 retained the highest FST value among all the SNPs present in the six surveyed regions . Therefore , we hypothesized that rs9493857 is a target of natural selection due to its effect on the induction of SGK1 expression in response to GR activation . This hypothesis is based on the observation that rs9493857 has both the highest FST value and the strongest correlation with latitude and that it resides in a GBR . To validate the results of the GR ChIP-chip assay , we treated MCF10A-Myc breast epithelial cells with either dexamethasone or vehicle and performed a conventional GR-ChIP assay followed by quantitative real-time PCR of the region containing SNP rs9493857 . Two independent GR-ChIP experiments showed a significant dexamethasone-dependent enrichment for the region containing rs9493857 ( Figure 3A ) . The results of the conventional ChIP assay allowed us to refine the location of the GBR to a 1 kb region spanning rs9493857 . However , this region does not contain a predicted GRE . Therefore , we hypothesized that SNP rs9493857 resides in a binding site for a GR cooperating transcription factor . We used the tool P-Match [46] to search for predicted binding sites for transcription factors and identified a canonical Oct1 binding site that contains rs9493857 . Oct1 is a well-established GR cooperating transcription factor that can enhance the regulation of GR target genes in a GC-dependent manner [47] . To confirm that the region containing rs9493857 is indeed an Oct1 binding site , we performed Oct1 ChIP experiments in MCF10A-Myc cells treated with dexamethasone or with vehicle . As shown in Figure 3B , the anti-Oct1 immunoprecipitated chromatin samples were enriched for the region containing rs9493857 in a GC-dependent manner; this enrichment was significant in all three independent experiments ( p<0 . 003 , unpaired t-test ) . In contrast , no enrichment was detected for a negative control region ( Figure S1 ) . These results , together with the GR-ChIP results , allowed us to formulate a model in which SNP rs9493857 affects GC-dependent Oct1 occupancy of its predicted binding site , thereby modulating GR-dependent SGK1 gene expression . To quantify allele-specific DNA occupancy by the GR-Oct1 complex , we next employed the HaploChIP technique , which allows the direct comparison of two alleles within the same heterozygous sample and the same experiment [48] . Because MCF10A-Myc cells are not heterozygous at SNP rs9493857 , we used six lymphoblastoid cell lines ( LCLs ) from the HapMap project [7] known to be heterozygous at this SNP . Figure 4 shows the results of the GR and Oct1 HaploChIP experiments performed in the presence of dexamethasone . The HaploChIP experiments were performed in duplicate on 6 LCLs for the GR and 3 LCLs for Oct1 . As expected , the amount of input DNA ( starting material ) was equivalent for the two alleles . However , for both GR and Oct1 HaploChIP assays , the amount of immunoprecipitated DNA containing the ancestral allele was significantly greater than that containing the derived allele ( p = 0 . 019 and p = 0 . 016 for GR and Oct1 , respectively ) . These results support the conclusion that the ancestral allele at rs9493857 results in greater DNA occupancy by the GR-Oct1 complex when compared to the derived allele . To test whether the newly identified GR-Oct1 binding site is indeed a GC-dependent enhancer region for which the rs9493857 ancestral versus derived alleles convey differential transcriptional activity , we performed luciferase reporter assays in SK-BR-3 breast cancer cells . A 3 . 8 kb segment encompassing rs9493857 was cloned 5′ to the SV40 promoter driving expression of the luciferase gene ( Figure 5 ) . Because SK-BR-3 cells are known to express endogenous GR at relatively high levels [16] , the reporter gene assay could be performed with endogenous GR . SK-BR-3 cells were transfected with either the DERIVED-enhancer construct or the ANCESTRAL-enhancer construct , which were identical except for the allele at rs9493857 . Upon treatment with dexamethasone for 12 hours , the ancestral allele at rs9493857 resulted in an average of 1 . 5-fold higher luciferase activity compared to the derived allele ( Figure 6 ) ( based on four independent experiments , p = 0 . 002 , one-tailed t-test ) . These results suggest that the region ∼30 kb upstream of SGK1 can in fact act as a GC-dependent enhancer whose activity depends upon the particular allele within the Oct1 binding site at rs9493857 . In summary , rs9493857 is located within a functional GR enhancer; the ancestral allele at rs9493857 demonstrates both increased GR-Oct1 binding and glucocorticoid-driven gene expression .
We have used a combination of population genetics and molecular biology methods to identify regulatory variants of SGK1 , a gene that plays a key role in the stress response and that has been clearly implicated in cell survival , water re-absorption and the insulin response . Because SGK1 expression is induced by the neuroendocrine stress response ( through GR and MR activation ) , we hypothesized that SGK1 regulatory variation was a target of adaptation to environmental stress . Consistent with this hypothesis , we identified a noncoding variant ( rs9493857 ) with marked allele frequency differences between populations and a strong correlation with climate variables . This SNP is located within a binding site for Oct1 , a known GR cooperating transcription factor . Using ChIP-chip , conventional ChIP , HaploChIP and gene reporter assays , we show that the ancestral allele at rs9493857 permits more efficient binding of the GR-Oct1 complex to the enhancer region and induces gene expression at higher levels compared to the derived allele . More broadly , our results show that population genetics approaches may complement computational and traditional molecular biology methods for the identification of regulatory variants in genes involved in the stress response . These variants are expected to contribute to inter-individual differences in hormone ( e . g . , glucocorticoid ) responsiveness , and therefore , could contribute to individual susceptibility to common hormone-dependent diseases , such as cancer and the metabolic syndrome . Variation in gene regulation has long been hypothesized to be a major mechanism in the phenotypic divergence within and between species [49]–[57] . This proposal was recently bolstered by the genome-wide identification of common variants associated with variation in baseline mRNA levels in LCLs [58] , [59] . Consistent with the idea that regulatory variation may contribute to common phenotypes , a large proportion of susceptibility variants for common diseases identified through genome-wide association studies lies in non-coding regions [60] . Moreover , a number of regulatory variants have been shown to be targets of natural selection [61]–[63] . More recently , it was proposed that SNPs showing signals of selection are often associated with variation in baseline expression levels in LCLs , suggesting that selection of gene expression levels plays a key role in human adaptation [64] . However , despite the important role of regulatory variants in health and disease , the identification of such variants continues to present a significant challenge . This is mainly because of the dual challenge in computationally predicting regulatory elements and inferring the functional effects of variation within these elements . To address this problem , two main computational approaches have been developed so far: Prediction of transcription factor binding sites and identification of evolutionarily conserved sequences across distantly related species . Although numerous algorithms have been developed for the computational prediction of transcription factor binding sites , they all suffer from a high false positive discovery rate ( [65] and references therein ) . When overlaying these predictions with sequence conservation , the number of candidate regulatory regions can be narrowed down , but the false positive rate remains too high for experimental follow-up . These two approaches may also be used to predict the effect of genetic variants on gene expression levels , but the accuracy of these predictions remains low . Many studies aimed at the identification of variation in regulatory sequences have focused exclusively on the proximal promoter region of a gene ( generally up to 5–10 kb upstream of the TSS ) . Consistent with the idea that regulatory variation lies at or near the promoter , genome-wide mapping studies of variation in gene expression found that most expression quantitative trait loci ( eQTL ) lie in proximity of the TSS [66] . However , it should be noted that these studies were designed to identify eQTLs with strong effects on baseline expression levels , and only limited information is available about the location of eQTLs in response to a physiological stimulus [59] , [66] , [67] . Therefore , focusing on 5–10 kb upstream of the TSS may miss important regulatory regions , especially for nuclear receptor target genes where long range regulation of gene expression appears to be common [68] . In the present study , we have leveraged both molecular biology and population genetics methods to identify a common variant influencing GR-mediated induction of SGK1 expression . We used computational predictions of GR binding sites and conserved sequence elements to generate a map of candidate regulatory elements in a >100 kb region encompassing the SGK1 gene . This map was compared to one generated by GR ChIP-chip analysis . ChIP-chip mapping tends to identify a smaller number of candidate regulatory elements compared to computational methods; however , these regions are still likely to contain a nontrivial portion of false positives [69] . Using the signature of natural selection to prioritize the regions identified by ChIP-chip , we identified a likely GR enhancer region . Moreover , by combining population genetics information with ChIP-chip , we were able to hone in on a regulatory sequence element that harbors common variation in human populations . Although not all variants in GBRs are expected to carry signals of natural selection , this approach is easily amenable to genome-wide applications and may provide testable hypotheses either by itself or in combination with eQTL mapping . Given the large between-population differences in allele frequency at rs9493857 , our results imply that SGK1 expression levels in response to cortisol could vary greatly across populations with different ancestry . Recent eQTL mapping studies performed on the HapMap LCLs have identified a large fraction of loci with significant differences in mean expression levels among human populations [59] . Although systematic differences between cell lines from different populations may have influenced these results [70] , there is clear evidence for inter-population differences in allele frequencies for variants associated with baseline gene expression levels [71] , [72] . To investigate the contribution of rs9493857 to SGK1 mRNA levels , we have inspected the results of genome-wide eQTL studies in which association data are available for all SNPs examined . We did not find a significant association between rs9493857 genotype and SGK1 mRNA levels [59] , [66] , [73]–[75] . This finding is not entirely surprising considering that the eQTL studies assayed baseline expression levels while our results indicate that rs9493857 influences gene expression in response to a specific stimulus , i . e . glucocorticoid exposure . Overall , our knowledge of inter-individual variation in expression levels induced by the stress response remains poor . The results of our functional studies imply that the ancestral allele at rs9493857 will result in higher SGK1 expression levels in response to physiological stress; this allele is also the most common allele present in populations at lower latitudes . This finding suggests that increased stress-induced SGK1 gene expression may have been advantageous in ancestral , and perhaps current , human populations living in equatorial environments . Increased SGK1 expression is consistent with SGK1's role in mediating sodium retention; however , SGK1 expression is also known to enhance tumor cell survival in breast and prostate cancer cells [14] , [18]–[20] , [76] . Interestingly , these diverse biological processes ( salt retention and breast and prostate cell survival ) underlie disease mechanisms with known inter-population differences in incidence . For example , salt-sensitive hypertension and prostate cancer both have a higher prevalence in African Americans compared to other populations [77]–[79] . Similarly , premenopausal African American women have a higher proportion of the subtype of breast cancer known as “triple negative” , namely negative for estrogen , progesterone and Her2 receptors [80] . The lack of these three receptors suggests that alternative growth signaling pathways drive tumor cell proliferation in this breast cancer subtype [81] . Because the PI3-K/SGK1 pathway represents an alternative to ER- , PR- and Her2-mediated growth signaling , increased SGK1 expression could contribute to susceptibility to triple negative breast cancers [82] . In addition to triple negative breast cancer , African Americans , especially African American women , have a relatively high prevalence of the metabolic syndrome , which includes elevated blood pressure , obesity , and type 2 diabetes [83] . The higher prevalence in African Americans is attributed to both environmental ( e . g . , diet ) and genetic influences . There are many similarities between patients with the metabolic syndrome and those with excessive GC production; however , circulating cortisol levels in the metabolic syndrome are not elevated [84] . This fact suggests that GR signaling may be enhanced in the metabolic syndrome independently of cortisol concentrations . Indeed , SGK1 activity has been associated with hypertension via upregulation of epithelial sodium channel ( ENaC ) activity . Because small increases in the sodium reabsorptive capacity of the renal epithelia can have dramatic consequences on fluid volume regulation , increased SGK1 expression might contribute to the development of hypertension [85] . Furthermore , SGK1 activity has been linked to diabetes through glucocorticoid-mediated inhibition of insulin secretion [86] . Interestingly , SGK1 polymorphisms ( located in both Intron 6 and Exon 8 ) have recently been found to be associated with type 2 diabetes in Romanian and German cohorts [43] . Our finding of a GR-dependent regulatory variant in SGK1 raises the possibility that inter-individual differences in susceptibility to common diseases may be influenced by differential sensitivities to GR signaling . In other words , individuals harboring alleles resulting in increased cortisol-mediated gene expression may , as a result , be at increased risk of some hormone-dependent diseases such as triple negative breast cancer , prostate cancer , and the metabolic syndrome . The recent genome-wide association studies potentially offer an opportunity to assess the contribution of SNP rs9493857 to common disease phenotypes . This SNP is not present in the most widely used genotyping platforms , thus only proxy SNPs could be used to analyze the results of genome-wide association studies . None of the proxy SNPs ( with r2 ranging from 0 . 8 to 0 . 6 in Europeans ) reaches genome-wide significance in the published studies . However , two of the proxy SNPs , rs4896028 ( r2 = 0 . 811 ) and rs1009840 ( r2 = 0 . 616 ) , reach nominal levels of significance for adult BMI ( p = 0 . 027 ) and glycosylated hemoglobin levels ( p = 0 . 038 ) ( data deposited by WTCCC and published on-line from the British 1958 Birth Cohort DNA Collection , http://www . b58cgene . sgul . ac . uk/ ) , attention deficit hyperactivity disorder ( p = 0 . 01−0 . 001 , as reported in dbGAP ) , and systemic lupus erythematosus ( p = 0 . 01−0 . 001 , as reported in dbGAP ) ( Table S8 ) . Further studies are necessary to determine whether SNP rs9493857 indeed influences susceptibility to disease phenotypes . In particular , because this SNP affects glucocorticoid-dependent gene expression , accounting for environmental exposures will be important to determine conclusively if rs9493857 contributes to phenotypic variation related to stress response . Additional human traits and biological processes that show large inter-population differences include skin pigmentation and energy metabolism [2] , [3] . As with the stress response , these processes occur at the interface between the organism and the environment and are important for maintaining homeostasis . Interestingly , in the case of SGK1 , the target of natural selection appears to be the response to a stress-induced hormonal stimulus . This raises the possibility that a signature of local adaptation , and therefore large inter-population differences , may also be found in a global analysis of genes comprising nuclear receptor gene networks . Further studies are necessary to determine whether additional GR target genes show similar inter-population differences in the frequency of ancestral versus derived regulatory alleles .
GREs were computationally predicted by using NUBIScan 2 . 0 [40] , which implements an algorithm that relies on the combination of nucleotide distribution weight matrices of single hexamer halfsites for the prediction of nuclear receptor response elements . The analysis was performed using the default GR matrix and an arrangement consisting of two inverted repeats spaced by three nucleotides . All the GREs with a raw score ≥0 . 6 are reported in Figure 1A . ECRs between human and mouse , dog or opossum were identified using the ECR Browser tool [41] . For each SNP , FST values between pairs of populations were calculated according to [28]; FST can vary between 0 and 1 , with FST = 0 indicating no difference in allele frequencies and FST = 1 indicating that alternative alleles are fixed in the two populations . Spearman rank correlation coefficients between allele frequency and environmental variables were calculated using an in house program . The Bayesian geographic analysis described in Hancock et al . ( 2008 ) [11] was applied to the SGK1 SNPs to assess the evidence for genetic adaptation to varying environments . With both methods , significance was assessed by comparing the value of the test statistic for each SNP to the empirical distribution of the same statistic for the SNPs in the Illumina Infinium HumanHap 650Y chip typed in the HGDP panel [30] . Because a shift in the null distribution was observed for different allele frequency bins and depending upon the genotyping panel used , significance for each SGK1 SNP was assessed against the distribution of the test statistic for SNPs matched by allele frequency and by panel . Neutrality tests and summary statistics of genetic variation for the re-sequenced regions were calculated using the program SLIDER ( http://genapps . uchicago . edu/slider/index . html ) . To estimate the significance of Tajima's D and Fay and Wu's H , we performed 1 , 000 neutral simulations for each population sample separately using the program MS [87] . For the Hausa sample we simulated a simple growth model , while for the Italian sample we simulated a bottleneck model . These demographic scenarios and the corresponding parameter values were chosen based on previous modeling studies showing that they are consistent with patterns of neutral variation in the same population samples [45] . The haplotype test was performed for the re-sequenced candidate regulatory regions as described in [36] . We performed this test separately for each population sample . One thousand replicates were generated under the same demographic scenarios used above . Re-sequencing of candidate regulatory regions . The DNA samples sequenced at the six candidate regulatory regions belong to a panel previously described [45] , [88] . A subset of this panel consisting of 28 unrelated samples ( 14 Hausa from Cameroon and 14 Italians ) was randomly selected . DNA was PCR amplified and the PCR products , after Exo-SAP purification , were sequenced with ABI BigDye Terminator v . 3 . 1 Cycle Sequencing kit . The products were analyzed on an ABI 3730 automated sequencer ( Applied Biosystems ) and the resulting sequences were scored using the software Polyphred version 6 . 11 [89] . The human breast cancer cell line SK-BR-3 was cultured in DMEM supplemented with 10% FBS and 1% Penicillin/Streptomycin . The human breast epithelial cell line MCF10A-Myc was cultured in DMEM-F12 media supplemented with growth factors as described previously [90] . Six HapMap LCLs heterozygous at SNP rs9493857 were cultured in RPMI supplemented with 15% FBS and 0 . 1% Gentamicin . A 3 . 8 kb sequence segment upstream of the SGK1 TSS from an Italian individual bearing the derived allele at SNP rs9493857 was cloned in pGL3-Promoter vector ( Promega ) using the restriction sites KpnI and XhoI . The resulting construct , referred to as “DERIVED , ” was subjected to site-directed mutagenesis at the same SNP in order to obtain a construct , referred to as “ANCESTRAL , ” which is identical to DERIVED except for the nucleotide at rs9493857 . Site-directed mutagenesis was performed according to the manufacturer protocol using the QuikChange II Site-directed Mutagenesis Kit ( Stratagene ) . All constructs were verified by Sanger sequencing and did not contain any artifactual mutations . DNA was prepared using the Qiagen Miniprep and Maxiprep kits and transfected into SK-BR-3 cells using the Polyfect Transfection Reagent ( Qiagen ) . Luciferase and β-galactosidase activity were measured according to standard protocols ( Dual-Luciferase Reporter Assay System and Beta-Galactosidase Enzyme Assay , Promega ) following 6 , 12 , and 24 hours of treatment with either 10−6 M Dexamethasone or vehicle ( ethanol ) alone . Results are given as ratios of luciferase over β-galactosidase activity . Four independent experiments were performed and statistical significance between dexamethasone and ethanol treated samples was evaluated by means of a paired one-tailed t-test . MCF10A-Myc cells ( 4–5×106 ) and LCLs ( ∼20×106 ) were serum starved for 48 hours and then treated with dexamethasone 10−6 M or ethanol for 1 hour . After treatment , cells were cross-linked for 20 minutes with formaldehyde ( 1% final concentration ) followed by addition of glycine to a final concentration of 125 mM for 5 minutes to arrest the cross-linking . ChIP experiments were performed according to a standard protocol ( Upstate Biotechnology , Milipore ) . Rabbit polyclonal anti-GR ( E-20 ) and anti-Oct1 ( C-21 ) antibodies were obtained from Santa Cruz Laboratories . The immunoprecipitated protein/chromatin complexes were either used to perform a Western-Blot or treated to reverse the crosslinks according to the manufacturer's instructions . Specifically , the Western blot was performed on protein/chromatin complexes obtained from MCF10A-Myc cells to confirm that GR and Oct1 were immunoprecipitated only in the presence of their specific antibodies . The DNA obtained after reversing the crosslinks was either analyzed by quantitative RT-PCR to assess for enrichment of the GR and Oct1 binding regions of interest ( DNA from MCF10A-Myc cells ) or used to perform the HaploChIP experiments ( DNA from LCLs ) . MCF10A-Myc cells were serum-starved for 48 hours and then treated with dexamethasone ( 10−6 M ) for 1 hour . Following standard ChIP , the immunoprecipitated and the input DNA were amplified , fragmented , and labeled for hybridization according to the Affymetrix ChIP Protocol . These samples were then hybridized to the Affymetrix Human Tiling Array 2 . 0R A ( chromosome 1 and 6 ) and scanned at the University of Chicago Functional Genomics Core Facility . Probe signals from two independent biological experiments were analyzed using the Model-based Analysis of Tiling-array ( MAT ) software [42] to detect enriched regions of GR binding based on the National Center for Biotechnology Information's build 36 of the human genome . The MAT software identifies ChIP-enriched regions by calculating a MAT score for a given window size . The window size was set to 300 bp based on our observed DNA fragment size after shearing . For each 300 bp sliding window region , a MAT score was calculated by pooling all of the probes across each replicate . To assign a p-value to a window , MAT estimates the non-enriched null distribution of all the MAT scores . To obtain the distribution , MAT uses a non-overlapping sliding window method along the chromosome to calculate MAT scores that cover the array . Assuming MAT scores to be normally distributed , MAT estimates the variance from the windows with MAT scores smaller than the median; the null distribution is then estimated to be symmetric around the median . A threshold of P<10−3 was used to identify regions in chromosome 1 and 6 that are occupied by the GR . The DNA obtained by ChIP performed on LCLs treated with dexamethasone was genotyped by means of quantitative RT-PCR using TaqMan reagents . A custom TaqMan genotyping assay was designed to target SNP rs9493857 , with the fluorochrome VIC identifying the ancestral allele and FAM identifying the derived allele . To account for differences between the two fluorochromes , a standard curve was built for each of the two alleles using serial dilutions of a genomic DNA known to be heterozygous at rs9493857 . The resulting PCR product was quantified for each allele separately in each reaction . The imbalance between the ancestral and the derived alleles was measured as the ratio of the amount of each PCR-product in the immunoprecipitated DNA to that in the corresponding input DNA . Two independent experiments were performed for each cell line and each experiment result was assayed in three RT-PCR technical replicates . Statistical significance was assessed by performing binomial tests on 12 and 6 independent experiments , for the GR and the Oct-1 HaploChIPs , respectively . Rs9493857 and rs1763502 were genotyped in 971 individuals from 52 worldwide human populations from the CEPH Human Genome Diversity Project ( HGDP ) panel [31] , using an Illumina GoldenGate assay at the UCLA Southern California Genotyping Consortium Facility . | Susceptibility to many common human diseases including hypertension , heart disease , and the metabolic syndrome is associated with increased neuroendocrine signaling in response to environmental stressors . A key component of the human stress response involves increased systemic glucocorticoid secretion that in turn leads to glucocorticoid receptor ( GR ) activation . As a result , a variety of GR-expressing cell types undergo gene expression changes , thereby providing an integrated physiological response to stress . The SGK1 gene is a well-established GR target that promotes cellular homeostasis in response to stress . Here , we use a combination of population genetics and molecular biology approaches to identify an SNP ( rs9493857 ) in a distant SGK1 GR-binding region with unusually large differences in allele frequency between populations of European and African ancestry . Furthermore , rs9493857 shows a strong correlation between allele frequency and distance from the equator , a pattern consistent with a varying selective advantage across environments . Indeed , the ancestral allele at rs9493857 results in increased GR-binding and glucocorticoid-regulated gene expression , suggesting that an increased stress response ( i . e . , glucocorticoid responsiveness ) was advantageous in ancestral human populations . We speculate that , in modern times , such variation could favor the negative effects of a heightened glucocorticoid response , potentially predisposing individuals to chronic diseases such as metabolic syndrome and hypertension . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"diabetes",
"and",
"endocrinology/endocrinology",
"genetics",
"and",
"genomics/gene",
"expression",
"evolutionary",
"biology/human",
"evolution",
"genetics",
"and",
"genomics/physiogenomics",
"evolutionary",
"biology/genomics",
"genetics",
"and",
"genomics/genetics",
"of",
"dis... | 2009 | Adaptive Variation Regulates the Expression of the Human SGK1 Gene in Response to Stress |
Dietary restriction extends lifespan in evolutionarily diverse animals . A role for the sensory nervous system in dietary restriction has been established in Drosophila and Caenorhabditis elegans , but little is known about how neuroendocrine signals influence the effects of dietary restriction on longevity . Here , we show that DAF-7/TGFβ , which is secreted from the C . elegans amphid , promotes lifespan extension in response to dietary restriction in C . elegans . DAF-7 produced by the ASI pair of sensory neurons acts on DAF-1/TGFβ receptors expressed on interneurons to inhibit the co-SMAD DAF-3 . We find that increased activity of DAF-3 in the presence of diminished or deleted DAF-7 activity abrogates lifespan extension conferred by dietary restriction . We also observe that DAF-7 expression is dynamic during the lifespan of C . elegans , with a marked decrease in DAF-7 levels as animals age during adulthood . We show that this age-dependent diminished expression contributes to the reduced sensitivity of aging animals to the effects of dietary restriction . DAF-7 signaling is a pivotal regulator of metabolism and food-dependent behavior , and our studies establish a molecular link between the neuroendocrine physiology of C . elegans and the process by which dietary restriction can extend lifespan .
Adult reduction in caloric intake and restriction of feeding periods have been shown to substantially increase lifespan across evolutionarily diverse organisms [1 , 2] . Collectively , such treatments have been referred to as dietary restriction ( DR ) . DR has been shown to be effective even when initiated in later phases of adult life , although the efficacy of the treatment has been observed to diminish with advancing age in Caenorhabditis elegans [3 , 4] . Genetic studies in C . elegans have defined roles for mediators of stress response pathways , such as DAF-16/FoxO , PHA-4/FoxA and SKN-1/Nrf2 , as well as the intracellular energy sensors TOR and AMPK in mediating the effects of DR on longevity [5–8] . Other studies have suggested that external cues are also critical in eliciting a DR response that extends lifespan in C . elegans [9] . In both C . elegans and Drosophila , the efficacy of DR treatment can be abrogated by the addition of food odors , and longevity in Drosophila can be extended by reduction of olfactory function [10 , 11] . Similarly , studies in C . elegans have shown that mutation of genes implicated in sensory systems or ablation of chemosensory neurons results in extended lifespan [12–14] . Specifically , a pair of gustatory neurons in C . elegans , the ASI neuron pair , have been shown to be required for lifespan extension in response to dietary restriction [6] . In the present study , we sought to explore the signaling mechanisms by which perceptions in the nervous system of food availability contribute to the DR response in peripheral tissues . We have focused our attention on the gene daf-7 , which encodes a TGFβ ligand that is secreted from the ASI neurons to control diverse behaviors of C . elegans [15–18] . DAF-7 has previously been implicated in longevity and food sensing; daf-7 mutant animals are reported to be long-lived in a manner that is dependent on food levels and also exhibit defects in adjusting feeding behaviors in response to periods of starvation [19–21] . However , the role of DAF-7 in lifespan extension in response to DR has not been fully investigated . Here , we have focused on expanding understanding of the role that the DAF-7 signaling pathway has in lifespan extension in response to limited nutrient availability . We have determined that DAF-7 is a key neuroendocrine signal required in the ASI neurons for response to dietary restriction . Moreover , we find that age-related changes in daf-7 expression contribute to the reduced sensitivity that older animals have to DR treatment , suggesting that the efficacy of DR interventions that delay aging can be modulated by neuroendocrine signaling .
We investigated the role of the DAF-7/TGFβ pathway in lifespan extension in response to dietary restriction using the bacterial deprivation ( BD ) method , where animals are moved to solid media completely lacking a bacterial food source during adulthood [3 , 4] . Using this protocol ( see Methods for details ) at 25°C and initiating BD treatment at day 3 of adulthood , we observed an average 19 . 5% extension of mean lifespan in wild-type animals , comparable to what has been reported previously when taking into account changes in experimental temperature ( Fig 1B and 1E; S2 Table ) . Using multiple loss-of-function alleles , we observed that mutations in the daf-7 gene , encoding a TGFβ family ligand , or in the daf-1 gene , encoding the Type I TGFβ receptor , abrogated the lifespan extension conferred by BD ( Fig 1C and 1E; S2 Table ) . This is consistent with a prior report which found that daf-7 mutant animals are resistant to longevity fluctuations due to altered food levels [21] . We observed that the strong dependence of lifespan extension conferred by BD on DAF-7 was temperature dependent , as daf-7 mutant animals retained lifespan extension , albeit reduced relative to wild type , when propagated 20°C ( S2 Table ) , as reported previously [22] . Different regimens of dietary restriction have been found to extend lifespan in C . elegans through separate genetic pathways [23] . To ensure the effects we observed were not an outcome specific to the BD method of DR , we also tested daf-7 pathway mutants in a second , distinct protocol for dietary restriction , referred to as solid dietary restriction ( sDR ) , in which adult animals are exposed to a diluted bacterial food source that is refreshed every other day [8] . Using the sDR method , we observed results consistent with our BD data , where mutants in either daf-7 or daf-1 have diminished lifespan extension in response to sDR ( S1 Fig; S3 Table ) . DAF-7 signaling through DAF-1 has been shown to act through inhibition of the co-SMAD DAF-3 ( Fig 1A ) [24 , 25] . We found that daf-3 mutation could suppress the loss of sensitivity to dietary restriction observed in daf-7 and daf-1 mutants ( Fig 1D and 1E , S1 Fig; S2 and S3 Tables ) . Mutations in daf-7 have previously been observed to result in phenotypes such as diminished pumping , increased dauer entry , and increased fat storage [17] . Genetic analysis of the individual phenotypes of daf-7 mutant animals has identified distinct downstream genetic pathways that act to mediate each of these DAF-7-dependent phenotypes [17] , enabling us to determine if any of these pleiotropies might be associated with the diminished ability of DAF-7 pathway mutants to respond to DR . The pumping defect of daf-7 mutants is small in magnitude compared to the decrease in pharyngeal pumping observed in feeding-defective eat mutants that are used as genetic models of DR [26 , 27] . Nonetheless , to test this possibility , we determined the effects of combining a daf-1 mutation with mutations in tbh-1 and tdc-1 , which have been shown to suppress the feeding rate changes in daf-1 and daf-7 mutants [17] . To determine if signaling through pathways promoting dauer formation might be involved in the DR phenotype , we examined a daf-1;daf-12 double mutant . To determine if fat storage might be contributing to the DR defects we observed , we constructed daf-1 mgl-3;mgl-1 mutants , in which fat storage increases arising from diminished DAF-7 signaling are specifically suppressed [17] . None of these secondary mutations were able to suppress the BD defect of daf-1 mutant animals , decoupling these three phenotypes from the DR response that is dependent on DAF-7 signaling ( S2 Fig; S2 Table ) . Prior studies established that daf-7 is expressed principally in the ASI neuron pair , but also in additional sensory neurons when C . elegans is propagated on E . coli bacterial food , and that daf-7 expression is induced in the ASJ neuron pair upon exposure to metabolites of Pseudomonas aeruginosa [15 , 16 , 18] . We found that reintroducing wild-type daf-7 into daf-7 ( ok3125 ) mutants rescued the BD defect of these animals ( Fig 2A and 2B ) . Additionally , daf-7 ( + ) driven by ASI or ASJ specific promoters was also sufficient to rescue the BD defect of daf-7 mutant animals , consistent with the secretory nature of the DAF-7 ligand ( Fig 2C ) . Unlike the expression of the DAF-7 ligand , the DAF-1 receptor is broadly expressed in the C . elegans nervous system [25 , 28] . To determine the functional targets receiving DAF-7 signal , we examined the ability of daf-1 ( m40 ) animals to respond to DR when a wild-type daf-1 transgene had been expressed in different subsets of cell types under heterologous promoters [17] . daf-1 expression in the nervous system was sufficient to restore lifespan extension in response to BD . Furthermore , as has been demonstrated for other daf-7 regulated phenotypes [17] , we observed that the RIM/RIC interneurons are the specific sites of action for the daf-1 receptor for lifespan extension in response to BD treatment ( Fig 2D–2F ) . Given the results of our genetic analysis of the DAF-7 signaling pathway in dietary restriction , we sought to examine how daf-7 expression might change in response to DR intervention . We were unable to detect a change in expression using quantification of the transcriptional reporter , ksIs2[daf-7p::GFP] , in fed versus BD treated animals ( Fig 3A ) . We have previously observed that fluorescent in situ hybridization ( FISH ) provides more precise kinetic resolution of the dynamics of daf-7 transcription than does the ksIs2 GFP reporter [18] . By performing FISH on animals subjected to BD , we were able to detect a slight but consistent upregulation of daf-7 mRNA transcription . Worms exhibited an increase in daf-7 mRNA in ASI neurons in animals fixed 24 hours after BD treatment was initiated , but no detectable difference was found after a period of 5 days had passed ( Fig 3B ) . Of note , we observed that aging adult animals began to exhibit low-level expression in the ASJ neurons , but we did not observe any changes in daf-7 mRNA in the ASJ neurons in response to BD ( Fig 3A and 3C ) . These data suggest that in response to food deprivation , daf-7 transcription is acutely activated in the ASI neuron pair , which promotes lifespan extension mediated by DR ( Fig 3D ) . In response to food cues , neuroendocrine signals originating from chemosensory neurons can influence the activity of DAF-16/FoxO in the intestine [29 , 30] . To determine if DAF-7 signaling contributes to the DR response via DAF-16/FoxO activation , we monitored the localization of the zIs356[daf-16p::daf-16::GFP] transgene in wild-type and daf-7 mutant backgrounds . In response to food deprivation , wild-type animals shift from mostly cytosolic to nuclear localized DAF-16::GFP [29] . A daf-7 loss-of-function mutation abrogated this intestinal DAF-16::GFP translocation in BD conditions compared to wild-type animals ( Fig 4 ) . These data were surprising particularly considering that DAF-16 activation has been implicated in the setting of daf-7 loss-of-function [31] . However , we note that consistent with reports by others [19] , we did observe an increase in nuclear DAF-16::GFP in the daf-7 ( e1372 ) background in other tissues such as the muscle and hypodermis in both fed and BD conditions ( S3 Fig ) . This observation suggests that specifically in response to BD , an increase in daf-7 expression stimulates activation of DAF-16 in the intestine , which helps to promote longevity . This model is fitting with prior reports that have implicated a role for DAF-16/FoxO in mediating lifespan extension in response to various forms of DR [8 , 23] and in food sensing mutants [12] . We measured daf-7 expression as animals aged during adulthood using the ksIs2[daf-7p::GFP] reporter strain . We observed that daf-7 expression is maintained throughout the life of adult animals in the ASI neurons . As noted above , we also observed daf-7 expression in the ASJ neuron pair as animals age , with all animals exhibiting ASJ expression by day 3 of adulthood ( Fig 5A ) . In contrast to the marked induction of daf-7 expression in both ASI and ASJ neurons in response to P . aerugionsa [18] , in aging animals , daf-7 expression in ASJ remained relatively low ( Fig 5B ) . Moreover , we observed that daf-7 expression in the ASI neuron pair significantly decreased with age ( Fig 5B ) . We confirmed these findings by FISH using probes targeted to endogenous daf-7 mRNA to eliminate the possibility that these observations were an artifact of using a transgenic reporter . Our FISH results support our observations of the ksIs2 GFP reporter strain . ASI neurons from aged animals show decreased daf-7 expression; and while there is no detectable daf-7 mRNA in ASJ neurons of young animals , we were able to observe daf-7 mRNA in older adults ( S4 Fig ) . We sought to corroborate these changes in daf-7 expression in these sensory neurons with a measure of how much functional DAF-7 was secreted , so we utilized the cuIs5[C183::GFP] reporter of DAF-3 activity . DAF-3 negatively regulates C183 enhancer activity in vivo , resulting in low GFP fluorescence when DAF-3 is active [32] . The transgenic cuIs5[C183::GFP] reporter provides a measure of DAF-7 signal production by examining the downstream effects on DAF-3 in a neighboring tissue . We found that GFP fluorescence was diminished in an age-related , DAF-7-dependent manner , consistent with less overall DAF-7 signaling in aging worms ( Fig 5C ) . In addition to experiencing declines in healthspan indicators such as feeding rate and mobility , aging worms also become diminished in their ability to respond to dietary restriction treatment to extend lifespan [3 , 33] . We wondered if part of the insensitivity older animals have to dietary restriction treatment could be attributed to diminished levels of DAF-7 that cause an increased amount of DAF-3 activity that blocks responses leading to lifespan extension in response to DR in aging animals . To test this hypothesis , we conducted BD experiments where BD treatment was initiated at multiple time points , beginning on days 1 , 3 , 5 or 7 of adulthood , in wild-type or daf-3 mutant animals . We found that wild-type animals experience a robust lifespan extension when BD was began on days 1 or 3 , but were unable to respond when BD was started on days 5 or 7 ( Fig 5D , S4 Table ) , consistent with prior studies [3] . By contrast , daf-3 mutant animals were able to maintain the ability to respond to BD on day 5 ( Fig 5D ) , suggesting that age related decline in the ability to respond to dietary restriction can be attributed , in part , to increased DAF-3 activation as a result in diminished daf-7 expression . Additionally , animals overexpressing daf-7 retain the ability to respond to BD and extend lifespan late in life at a time when wild type animals no longer exhibit lifespan extension in response to BD ( S5 Fig ) .
DAF-7 is at the nexus of feeding behaviors and fat metabolism [17 , 20] , suggestive of neuroendocrine links between the nervous system and secondary tissues . We have described how neuroendocrine signaling through the DAF-7/TGFβ pathway is required for lifespan extension in response to DR in C . elegans . Whereas canonical energy sensing pathways , such as AMPK and TOR , have been shown to be involved in lifespan extension in response to DR , the role of neural regulation by sensory systems of the DR response is less understood [1 , 10 , 11] . Prior studies have established the ASI neuron pair as a cell non-autonomous regulator of the DR response , identifying the insulin-like peptide INS-6 and the SKN-1/Nrf2 transcription factor as relevant agents in initiating communication to downstream cells and tissues [6 , 29] . We have shown that in response to DR , the ASI pair also secretes the neuroendocrine ligand , DAF-7 , which signals to the RIM/RIC interneurons to suppress the co-SMAD DAF-3 . In the absence of negative regulation by DAF-7 , increased DAF-3 activity blocks the lifespan extension caused by DR ( Fig 6 ) . In the developing animal , the DAF-7 ligand is produced in favorable conditions that promote entry into reproductive development , specifically in the presence rather than the absence of bacterial food [15 , 16] . Our data are suggestive of an acute increase in daf-7 expression in the ASI neuron pair in response to the withdrawal of bacterial food , indicating that the dynamic expression of daf-7 of developing larvae may differ from that of adult animals in response to changing environmental conditions such as DR treatments . Indeed , while bacterial deprivation extends the lifespan of adult animals , the introduction of DR-like treatments in young larvae either prompts entry into the dauer state or has detrimental effects on developing animals that have already surpassed the dauer decision checkpoint [34 , 35] . Whereas a recent study showed that adult animals exposed to diminishing amounts of bacterial food exhibit decreased daf-7 expression in the ASI neurons after a period of four days [21] , our data , recording levels of daf-7 mRNA using FISH-based detection at multiple time points after the complete withdrawal of food , reveal a complex relationship pattern of dynamic daf-7 expression in the ASI neurons of adult animals in response to the withdrawal of bacterial food . We observe an initial increase in daf-7 expression in animals subjected to BD conditions , consistent with our genetic data implicating a requirement for DAF-3 inhibition for lifespan extension in response to BD . We observe that at later times following the withdrawal of bacterial food , daf-7 expression is maintained relative to initial levels of expression , in marked contrast to what has been observed when developing larvae are subjected to starvation conditions [15] . Our study builds upon previous observations that have linked the daf-7 gene with aging and the influence changing food levels has on longevity [19 , 21] . Together , our genetic findings and expression analyses support a model where active DAF-3 is sufficient to disrupt the animals’ sensory abilities and prevent lifespan extension in response to DR ( Fig 6 ) . Because daf-3 mutant worms are capable of responding to DR , the DAF-7 signaling pathway does not seem to have a direct role in altering metabolism in other tissues to extend lifespan in response to limited food levels . Rather , DAF-7 secreted by the chemosensory neurons seems to be a key neuroendocrine signal that allows animals to properly sense reductions in nutrient availability , which eventually results in activation of DAF-16/FoxO in the intestine under food deprivation . Moreover , our data suggest that an age-dependent decline in neuronal daf-7 expression also underlies the diminished sensitivity of aging animals to the lifespan effects of DR , linking a decline in neuroendocrine function to the loss of DR efficacy with advancing age . In human aging , decline in olfactory function is one of the largest predictors of mortality- a stronger independent risk factor for death than causes such as cancer or heart failure [36] . Our study suggests that the modulation of a specific neuroendocrine signaling pathway active in chemosensory neurons involved in the sensation of bacterial food may alter the sensitivity of C . elegans to the effects of DR . We speculate that therapeutic strategies targeting analogous neuroendocrine pathways in mammals may be able to function in concert with dietary modifications to promote longevity .
C . elegans were maintained at 16°C on E . coli OP50 as previously described [37] . For a list of all strains used in this study , see S1 Table . Due to the egg-laying defect of daf-7 pathway mutant animals , synchronized populations were prepared by egg-prep of gravid adult worms in bleach followed by L1 arrest overnight in M9 buffer . L1s were placed on OP50 seeded Nematode Growth Media ( NGM ) plates and raised to the L4 larval stage at 16°C . Upon reaching L4 , worms were transferred onto NGM plates containing 12 μM FUDR ( to prevent matricidal effects of daf-7 pathway mutants as well as progeny production ) and 0 . 01 mg ampicillin seeded with 10X concentrated OP50 from an overnight culture and moved to 25°C ( to avoid AVID [38] as well as enhance daf-7 mutant phenotypes [15 , 39] ) . Unless otherwise noted , on day 3 of adulthood ( where day 0 is defined as L4 stage ) , worms were transferred to either fed or DR conditions on NGM plates made without peptone to prevent bacterial growth and rimmed with 150 μL of 10 mg/mL palmitic acid to prevent worms from crawling off the plates . For BD experiments , fed plates were seeded with 200 μL of 10X concentrated OP50 from an overnight culture and BD plates were unseeded . For sDR experiments , fed plates were seeded with 200 μL of OP50 at a concentration of 2x1010 bacteria/mL and sDR plates with 200 μL of OP50 at 5x108 bacteria/mL . At least 2 plates per condition were used in all experiments . Worms were scored for death ( defined as failure to respond to prodding with a platinum wire ) every 1–3 days beginning around day 4 of adulthood . Animals exhibiting vulval rupture were censored . Worms that crawled off the plate were never considered . Representative experiments are presented here . For lifespan statistics of individual experiments , see S2–S4 Tables . Synchronized populations were prepared as above and treated in the same manner as worms subjected to lifespan analysis ( raised to L4 16°C , then shifted to ampicillin/FUDR plates and placed at 25°C ) . Animals were examined for GFP fluorescence on the indicated days . All images were acquired with an Axioimager Z1 microscope using animals mounted on glass slides , anesthetized by 100mM sodium azide . Quantification of daf-7p::GFP was performed by taking the maximum intensity by FIJI software [40] within the ASI or ASJ neuron at 40X magnification . Quantification of C183::GFP was done by taking the average intensity by FIJI software [40] within the entire pharynx at 20X magnification . All quantifications were normalized by exposure time and background fluorescence ( measured individually for each image ) . Day 3 adult zIs356[daf-16p::daf-16a/b::GFP] strains were examined on a fluorescent dissecting microscope after 4 hours of bacterial deprivation . Representative images were taken at 20X magnification . Two to four replicates were performed for all experiments presented . Synchronized populations were established as above . FISH was performed as previously described [41] . At the indicated times and treatments , animals were washed twice with M9 buffer before fixation with 4% formaldehyde at room temperature , followed by PBS washes and suspension in 70% RNase free ethanol and stored at 4°C . To image , all samples from an individual experiment were incubated overnight with FISH probes designed against daf-7 mRNA ( coupled to Cy5 dye ) [18] in hybridization solution at 30°C . The next day , animals were imaged with a Nikon Eclipse Ti Inverted Microscope outfitted with a Princeton Instruments PIXIS 1024 camera . A GFP marker was used to focus on the neuron of interest and obtain a single image using a Cy5 filter . This method of image acquisition does not allow resolution of single mRNA molecules , thus quantification of daf-7 was done using FIJI software [40] to outline either ASI or ASJ and obtaining the mean intensity and subtracting background fluorescence ( measured by obtaining the mean intensity of a small space immediately adjacent to the neuron being quantified ) . A minimum of 2 replicates was performed for all experiments presented . The log-rank statistical test was used to determine p-values for lifespans . Using Graphpad Prism , an unpaired t-test , one-sample t-test , or one-way ANOVA was used to determine significance in quantification of expression experiments . | Reductions in food intake have long been observed to improve longevity , extending lifespan in many evolutionarily divergent organisms . While great progress has been made in identifying the mechanisms by which nutritional interventions act to delay the aging process , much remains unclear . Particularly , while work in multiple species has found evidence that the sensation of food availability by the nervous system contributes to lifespan extension in response to reduced food levels , little is known about how these contributions are executed . Here , we have characterized how a specific neuroendocrine peptide , expressed in a set of sensory neurons , responds to changes in food conditions to modulate lifespan effects of dietary restriction at the organismal level . We further find that age-related changes in expression of this neuroendocrine signal contribute to the declining efficacy of nutritional interventions as animals get older . This work highlights the importance of neuroendocrine regulation in both the aging process and in treatments aimed at increasing longevity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"invertebrates",
"fluorescence",
"imaging",
"medicine",
"and",
"health",
"sciences",
"molecular",
"probe",
"techniques",
"caenorhabditis",
"neuroscience",
"animals",
"animal",
"models",
"physiological",
"processes",
"developmental",
"biology",
"caenorhabditis",
"elegans",
"... | 2017 | Age-Dependent Neuroendocrine Signaling from Sensory Neurons Modulates the Effect of Dietary Restriction on Longevity of Caenorhabditis elegans |
Thrombocytopenia , bleeding and plasma leakage are cardinal features of severe dengue . Endothelial cell activation with exocytosis of Weibel-Palade bodies ( WPBs ) may play an etiological role in this condition . In a cohort of 73 Indonesian children with dengue hemorrhagic fever ( DHF ) , of which 30 with dengue shock syndrome ( DSS ) , we measured plasma levels of the WPB constituents von Willebrand factor antigen ( VWF:Ag ) , VWF propeptide and osteoprotegerin ( OPG ) , together with activity levels of the VWF-cleaving enzyme ADAMTS-13 and the amount of VWF in a platelet binding conformation ( VWF activation factor ) . Compared with healthy controls ( n = 17 ) , children with DHF/DSS had significantly higher levels of VWF:Ag , VWF propeptide and OPG and decreased ADAMTS-13 activity . The VWF activation factor was also significantly higher in DHF/DSS and highest in children who died . There were significant differences in the kinetics of the various WPB constituents: VWF propeptide and OPG levels decreased toward discharge , while VWF:Ag levels were lower than expected at enrollment with plasma levels increasing toward discharge . Moreover , VWF propeptide levels correlated better with markers of disease severity ( platelet count , liver enzymes , serum albumin and pleural effusion index ) than corresponding VWF levels . Together , these findings suggest that there is consumption of VWF in DHF/DSS . In 4 out of 15 selected children with low ADAMTS-13 levels on admission , we found a remarkable reduction in the large and intermediate VWF multimers in the discharge blood samples , consistent with an acquired von Willebrand disease . These findings suggest that severe dengue is associated with exocytosis of WPBs with increased circulating levels of VWF:Ag , VWF propeptide and OPG . High circulating levels of VWF in its active conformation , together with low ADAMTS-13 activity levels , are likely to contribute to the thrombocytopenia and complications of dengue . During the convalescence phase , qualitative defects in VWF with loss of larger VWF multimers may develop .
Dengue has become a major international public health concern with up to 100 million annual cases worldwide . It usually manifests as a non-specific febrile illness , but its course may become complicated by bleeding and a transient plasma leakage that may ultimately lead to shock and death [1] . Severe dengue with thrombocytopenia , bleeding and plasma leakage is referred to as dengue hemorrhagic fever ( DHF ) . The most severe form of DHF , which is accompanied by hemodynamic instability and shock is referred to as dengue shock syndrome ( DSS ) . DHF/DSS is most frequently seen in children and tends to manifest at the time the fever subsides . The pathogenic mechanisms responsible for the development of DHF/DSS are still poorly understood . A central feature of DHF/DSS is the development of a pronounced thrombocytopenia . The large glycoprotein von Willebrand factor ( VWF ) plays a central role in platelet-vessel wall interaction as it is responsible for mediation of platelet adhesion at sites of endothelial injury . VWF is predominantly synthesized in endothelial cells and , after cleavage of a VWF propeptide , it is either released constitutively or stored in specialized secretory granules , known as Weibel-Palade bodies ( WPBs ) . Injury or activation of the endothelium leads to a rapid secretion of equimolar amounts of stored VWF and VWF propeptide , and both proteins are regarded as markers of endothelial cell activation [2] . Freshly released VWF consists of ultra-large prothrombogenic multimers ( UL-VWF ) . The metalloprotease ADAMTS-13 ( a disintegrin and metalloproteinase with thrombospondin-1-like domains ) functions as a natural regulator that de-activates the prothrombogenic UL-VWF by proteolysis [3] . The importance of ADAMTS-13 is illustrated by the notion that absence of ADAMTS-13 is associated with platelet-rich microthrombi in the microvasculature , a disease known as thrombotic thrombocytopenic purpura ( TTP ) [4] . Under normal circumstances , VWF circulates in the plasma in a knot-like conformation in which the binding sites for platelet glycoprotein ( GP ) receptor Ib is not exposed . A change in conformation into a more elongated ‘activated’ form can be observed under different conditions in which increased platelet-VWF interaction is thought to play a role [5] . Determination of the amount of this ‘active’ VWF in plasma is possible using a llama-derived nanobody ( designated AU/VWFa-11 ) that displays specific binding only to the GPIba-binding conformation of the VWF . Osteoprotegerin ( OPG ) is a member of the tumor necrosis factor receptor superfamily , which is stored in WPBs and in platelets [6] , [7] , similar to VWF . OPG was traditionally known for its role in bone remodeling , but a growing body of literature suggests that OPG also has specific effects on the endothelium , including stimulation of vasculogenesis [8] , expression of adhesion molecules [9] and adhesion of leukocytes [10] . Moreover , OPG is physically associated with the A1 domain of VWF , both in stored and in circulating VWF , suggesting that OPG could interfere with platelet-VWF binding [11] . Children with an acute dengue infection have elevated VWF levels [12] , [13] and one study reported a moderate decrease in ADAMTS13 activity using an indirect assay for ADAMTS13 activity [14] . The aim of this study was to characterize WPB exocytosis and changes in factors involved in VWF-platelet interaction in patients with severe dengue . We determined concentrations of VWF antigen ( VWF:Ag ) , VWF propeptide , OPG and ADAMTS-13 activity in serial obtained plasma samples from a cohort of Indonesian children with DHF/DSS , together with the multimeric pattern of VWF and the amount of VWF in a platelet binding conformation ( VWF activation factor ) .
The Research Ethics Committee of the Faculty of Medicine Diponegoro University , Semarang , Indonesia , approved all legal , ethical and laboratory aspects of the study . Written informed consent was obtained from parents or legal guardians of the patients . This observational study enrolled consecutive children aged 3–14 years who were admitted to the pediatric ward or intensive care unit of the Dr . Kariadi University Hospital in Semarang , Indonesia between July 2005 and July 2006 with a clinical diagnosis of suspected DHF/DSS according to the 1997 WHO criteria [15] . A control group of 17 healthy Indonesian children , aged 6 to 14 years , was also enrolled . Inclusion criteria and clinical characteristics for this group have recently been described [16] . All patients tested positive for a dengue specific IgM in the discharge blood sample . DSS was defined as DHF with evidence of circulatory failure . Blood samples were collected in citrate anti-coagulated blood tubes ( Beckton-Dickinson ) at hospital admission ( enrollment; day 0 ) , on day 1 after admission and on day of discharge . A full blood count was performed daily in all patients as part of routine clinical care until platelet counts had shown a substantial increase . Therefore , no platelet counts or other hematological values were usually available on the day of discharge . A chest X-ray was performed with the patient lying in right lateral decubitus position to detect pleural effusion at enrollment and on day 2 . The pleural effusion index ( PEI ) was calculated by dividing 100 times the maximum width of the pleural effusion by the maximum width of the hemi-thorax . Citrate blood was centrifuged for 20 minutes at 1600 g and plasma was stored at −80°C until further analysis . Plasma levels of VWF:Ag and VWF propeptide were determined by enzyme-linked immunosorbent assay ( ELISA ) as described previously [17] . Active VWF was determined by ELISA using a nanobody ( AU/VWFa-11 ) that specifically recognizes the GP-1bα binding configuration of VWF , as described previously [18] . The term VWF activation factor was used to express the relative amount of VWF that circulates in its active , platelet binding conformation . VWF activation factor of normal pooled plasma was referred to as 1 . The multimeric pattern of VWF was analyzed in a selection of 15 patients , among whom the patients with the lowest ADAMTS-13 activity levels at enrollment or on day 1 using 2% agarose gel electrophoresis , followed by in-gel immunostaining and infrared imaging [19] . ADAMTS-13 activity was determined using the fluorescence resonance energy transfer ( FRETS ) assay for ADAMTS-13 activity ( Peptides International , Inc . , USA ) whereby the ADAMTS-13 activity of normal pooled plasma ( NPP ) of healthy Dutch donors was set at 100% [20] . Values obtained in the study participant samples were expressed as percentage of NPP . Normal values in healthy volunteers in our laboratory were in the range of 60–140% . OPG was measured by an in-house ELISA using microtiter plates coated overnight at 4°C with a mouse anti-human OPG antibody ( MAB8051 , R&D Systems , Minneapolis , MN ) . Diluted plasma samples were incubated for 2 h at room temperature . Detection was done by incubation with a secondary goat polyclonal antibody ( BAF805 , R&D Systems , Minneapolis , MN ) and streptavidin-conjugated HRP ( Sanquin , Amsterdam , the Netherlands ) and SuperSignal West Pico Chemiluminescent Substrate ( Thermo Scientific ) . For statistical analyses , the upper limit of detection for OPG ( 1200 mg/L ) was used in samples with values above this detection limit . Serum SGOT concentrations ( normal value for 7 year old children ≤35 Unit/L ) were measured using a colorimetric method ( Hitachi 7050; Boehringer Ingelheim , Germany ) . Serum albumin concentrations ( normal value ≥3 . 26 g/dl ) were measured using Bromocresol–Green method . Presence of dengue specific IgM and IgG antibodies was determined by capture and indirect ELISA ( Focus Technologies , Cypress , Calif . , USA ) , according to the manufacturer's instructions [21] . A full blood count was performed daily by a standard hematology analyzer . Data were expressed as medians with interquartile ranges ( IQR ) . Differences between groups were assessed by Mann-Whitney tests; changes in laboratory parameters over time within groups were evaluated by Wilcoxon matched-pairs signed rank test . Relationships between continuous variables were examined by Spearman's rank correlation analysis . A p-value of <0 . 05 indicated a significant difference . Statistical analyses were performed with SPSS version 16 . 0 .
A total number of 73 children with severe dengue were enrolled , of whom 43 were classified as having DHF grade I or II and 30 as DSS ( DHF grade III or IV ) . In addition , 17 healthy controls were enrolled . Clinical characteristics of the patients are summarized in table 1 ( adapted from [16] ) . The median duration of fever at enrollment was 4 . 0 days in both groups . All patients showed plasma leakage on day two after enrollment as evidenced by pleural effusion on a lateral chest X-ray . The nadir in platelet counts was observed on day 1 with median ( IQR ) values of 45×109/L ( 27–75×109/L ) and 50×109/L ( 18–74×109/L ) for the DHF and DSS group; by day 3 , median platelet counts had raised to 89×109/L ( 43–111×109/L ) and 80×109/L ( 40–153×109/L ) , respectively . Clinical bleeding occurred in 7 ( 16 . 3% ) patients with DHF ( epistaxis , n = 4; gum bleeding , n = 1; hematemesis , n = 2 ) and in 2 ( 7 . 1% ) patients with DSS ( hematemesis , n = 1; melena , n = 1 ) . Intravenous fluids were administered to all but three children . Free frozen plasma ( FFP ) was administered to 3 ( 7 . 0% ) children with DHF and to 8 ( 28 . 6% ) children with DSS . Two ( 4 . 7% ) children from the DHF group and 4 ( 14 . 3% ) from the DSS group received a platelet transfusion . Six children , all from the DSS group , died during the admission . No plasma sample was available for analysis in two patients from the DHF group and one patient from the DSS group at day 1 and in fifteen and six patients of these respective groups at discharge . Children from the DHF group had the highest VWF:Ag levels at enrollment , followed by children with DSS and healthy controls with median ( IQR ) values of 16 . 6 µg/mL ( 12 . 9–20 . 6 µg/mL ) , 12 . 5 ( 9 . 9–16 . 8 µg/mL ) and 7 . 2 µg/mL ( 5 . 8–9 . 5 µg/mL ) , respectively ( Figure 1A ) . In contrast , the highest VWF propeptide levels were found in the DSS group with a median level of 22 . 7 nM ( 15 . 7–35 . 8 nM ) compared to 21 . 9 nM ( 17 . 7–24 . 9 nM ) in the DHF group and 5 . 7 nM ( 4 . 7–6 . 4 nM ) in healthy controls ( Figure 1B ) . There was also a clear difference in the kinetics of VWF:Ag and VWF propeptide levels , despite the fact that both proteins are released into the plasma in equimolar amounts . While VWF propeptide levels decreased towards discharge , VWF:Ag levels increased significantly . OPG levels followed the same pattern as VWF propeptide: baseline levels were very high with 49% and 70% of the children in the DHF and DSS group , respectively , having a value above the upper detection limit of the assay ( 1200 mg/L ) , followed by a decrease towards discharge ( Figure 1C ) . The median VWF activation factor was about twofold higher in the DHF/DSS patients than in the healthy controls , indicating that a higher amount of the circulating VWF was in a platelet-binding conformation ( Figure 2A ) . This increased activation status of VWF persisted until discharge . A marked reduction in ADAMTS-13 activity levels was also a common finding at the time of enrollment ( Figure 2B ) . At enrollment , 20/43 ( 46% ) of children with DHF and 20/30 ( 67% ) of children with DSS had an ADAMTS-13 activity level of ≤50%; a severe ADAMTS-13 deficiency ( ≤10% ) was found in 1/43 ( 2% ) and in 3/30 ( 10% ) , respectively . ADAMTS-13 activity levels recovered to near normal values at the time of discharge . The multimeric pattern of VWF was determined in 15 patients . These 15 patients included the 7 patients of the cohort with the lowest ADAMTS-13 activity levels at enrollment in whom a blood sample at discharge was available; the remaining 8 patients were randomly selected . Despite the low ADAMTS13 levels , UL-VWF was not observed in any of the blood samples . However , the discharge blood sample of 4 patients with low ADAMTS-13 levels at enrollment showed a reduction in large and intermediate VWF multimers ( Figure 3; patient 1 to 3 , patient 4 not shown ) . None of these 4 patients suffered from clinical bleeding . In the group of children with DSS , there was no significant difference between the 6 children who died compared with those who survived in median values at enrollment for VWF:Ag levels ( 14 . 9 µg/mL vs . 12 . 3 µg/mL; p = 0 . 53 ) , VWF propeptide levels ( 33 . 3 nM vs . 21 . 7 nM; p = 0 . 10 ) , ADAMTS-13 activity ( 30% vs . 49%; p = 0 . 19 ) and platelet count ( 37×109/L vs . 58×109/L; p = 0 . 19 ) . The most outspoken difference between these groups was an almost twofold higher VWF activation factor ( 3 . 2 vs . 1 . 4; p<0 . 01 ) at enrollment in the children who died . No significant differences in these parameters were found between those with and those without clinical bleeding ( data not shown ) . Infusion of blood products had only a minor influence on these median laboratory values because only one of the six children who died received FFP and platelets at the day of enrollment . Correlations of VWF:Ag and VWF-related parameters with clinical and laboratory markers are shown in table 2 . There was no significant correlation between VWF:Ag and markers of dengue severity ( platelets and SGOT levels ) and plasma leakage ( albumin levels and PEI ) . In contrast , VWF propeptide levels had a much stronger correlation with these parameters; VWF propeptide levels were positively associated with severity of plasma leakage ( negative correlation with plasma albumin level and positive correlation with PEI ) and with liver enzyme disturbances and negatively associated with platelet count .
Our study shows that DHF/DSS is associated with acute endothelial cell activation with exocytosis of WPBs and release of VWF:Ag , VWF propeptide and OPG in the circulation , combined with a decrease in ADAMTS-13 activity . The circulating VWF had a higher activation factor , indicating that an increased amount of VWF was in an elongated , ‘active’ conformation enabling spontaneous platelet-VWF binding . The patients who died had a significantly higher amount of VWF in its most active conformation and in some of the patients with low ADAMTS-13 activity at enrollment , qualitative defects in VWF with a pronounced loss of larger VWF multimers was seen in the discharge blood samples . There was a clear difference in the kinetics of VWF:Ag and the other WPB constituents VWF propeptide and OPG: VWF:Ag levels were relatively low at enrollment and increased towards discharge , while VWF propeptide and OPG levels were very high at baseline and decreased upon clinical recovery . In contrast to VWF propeptide levels , VWF:Ag levels did not correlate well with parameters for disease severity . Hence , VWF propeptide levels seem to better reflect endothelial cell activation status and disease severity in DHF/DSS than VWF:Ag levels . We hypothesize that increased VWF consumption due to agglutinating platelets underlies this phenomenon . It is unlikely that VWF would simply leak out of the circulation during plasma leakage , just like albumin , because of the very large size of VWF multimers ( >10 . 000 kDA ) and because plasma levels of VWF propeptide would be expected to leak out even more as it is a smaller molecule . The four- to fivefold shorter half-life of VWF propeptide compared to mature VWF could explain why VWF:Ag levels were still elevated in the discharge samples , while VWF propeptide levels had returned to normal [2] . The VWF activation factor remained elevated across the study period . VWF activation is a measure of the relative amount of VWF that circulates in a platelet binding conformation . This parameter does not take into account the circulating VWF concentration , but total active VWF levels can be approximated by multiplying the VWF activation factor by the VWF:Ag levels . This approximation is hampered in this study by the consumption of VWF . Nonetheless , the high VWF propeptide levels in the early phase after enrollment and the fact that VWF:Ag and VWF propeptide are released in equimolar amounts from the endothelium suggest that total active VWF levels were higher early after enrollment and decreased towards discharge . Our study is the first to report OPG data in dengue . The vascular effects of OPG have received increased attention in recent years . High OPG levels have been related to atherosclerosis and cardiovascular disease in epidemiological studies [22] , [23] and OPG was shown to have specific effects on endothelial cells in vitro , such as prevention of apoptosis and up-regulation of adhesion molecules [8]–[10] . OPG is also closely linked to VWF: both proteins are cohabitants of WPBs and remain associated after release in the circulation . The notion that OPG binds selectively to the VWF A1 domain suggests that it may interfere with the binding of platelets to activated VWF , thereby preventing excessive platelet adhesion and aggregation [11] , [24] . This process may especially be relevant in inflammatory conditions , which are usually accompanied by endothelial cell activation and release of VWF , since inflammatory cytokines up-regulate the synthesis and release of OPG [7] . OPG may also influence dengue pathogenesis in other ways . OPG is a decoy receptor that competes with both tumor necrosis factor-related apoptosis-inducing ligand ( TRAIL ) and receptor activator of nuclear factor kappa-B ( RANK ) ligand ( RANKL ) . Patients with an acute dengue infection have elevated TRAIL serum levels [25] . This may be important in host defense against dengue , because recent work by Warke et al . suggested that TRAIL has dengue antiviral properties and suppresses the production of pro-inflammatory mediators by dengue-virus infected dendritic cells [26] . Moreover , while the RANK/RANKL system is predominantly known for its role in bone remodeling , there is increasing evidence that RANKL , which is among others expressed by activated T-cells , is also involved in regulation of immunity [27] . Hence , the high OPG levels may not only interfere with VWF-platelet interaction , but also with the antiviral and immune effects of TRAIL and the RANK/RANKL system . While OPG levels increased during the acute phase of DHF/DSS , a reduction in ADAMTS-13 activity was common . ADAMTS13 regulates the multimeric size and function of VWF by cleaving VWF within the A2 domain . To our knowledge , no previous study has so far reported data on both OPG and ADAMTS13 in clinical samples . An acute increase in circulating VWF levels in volunteers through administration of desmopressin or endotoxin is followed by a decrease in ADAMTS-13 levels [28] , [29] and lower ADAMTS-13 levels have been observed in several physiologic and pathologic conditions with high VWF levels , e . g . pregnancy and acute inflammatory states [30] . Hence , in our opinion , endothelial cell activation with release of VWF and secondary ADAMTS-13 consumption is the most likely explanation for the low ADAMTS-13 activity levels in these children . Other factors may also be involved: in vitro studies have shown that pro-inflammatory cytokines can reduce ADAMTS-13 synthesis , while thrombin and plasmin can inactivate ADAMTS-13 [31] , [32] . Finally , a case of dengue-associated microangiopathic thrombocytopenia due to an inhibitor of ADAMTS-13 was recently reported [33] . The rapid normalization of ADAMTS-13 levels in the children in our study , however , argues against a role for ADAMTS-13 antibodies . A remarkable observation in our study was the absence of large and intermediate VWF multimers in the discharge blood sample of some of the children . Loss of larger multimers is a prominent feature of acquired type 2A von Willebrand disease ( VWD ) . This condition is associated with a number of different disease states , including hematoproliferative and auto-immune diseases and cardiac abnormalities such as aortic stenosis [34] . One case report described a transient acquired VWD in a child recovering from an acute EBV infection [35] . Different pathophysiologic mechanisms may underly the loss of larger multimers [34] . In essential and reactive thrombocytosis , the increased number of circulating platelets is associated with a reduction of large VWF multimers [36] . In dengue , platelet numbers start to rise abruptly in the convalescent phase and we hypothesize that the combination of rapidly rising platelet numbers , the restoration of ADAMTS-13 activity and ‘exhaustion’ of endothelial cells is responsible for the transient disappearance of large and intermediate VWF multimers . Loss of larger VWF multimers may result in an increased bleeding tendency and , although clinical bleeding in DHF/DSS is usually restricted to the critical phase around defervescence , qualitative abnormalities of VWF should be considered in patients with a persistent bleeding tendency . Is the observed imbalance in the VWF-ADAMTS-13 system clinically relevant ? High circulating levels of VWF in an elongated , active conformation together with reduced VWF proteolysis by ADAMTS-13 may lead to increased platelet adhesion and formation of platelet-rich thrombi . In patients with severe sepsis and severe malaria , a similar imbalance in VWF and ADAMTS-13 was found and this was considered to be related to thrombocytopenia and organ dysfunction [17] , [37]–[40] . What DHF/DSS distinguishes from these other severe infectious diseases is that consumption of VWF;Ag and loss of larger VWF multimers have not been observed in these other diseases [17] , [41] , [42] . The notion that the AB blood group is associated with a higher risk for severe dengue supports a role for VWF in dengue pathogenesis , since blood group AB is associated with higher VWF levels [43] . Hence , even though typical features of thrombotic microangiopathy ( e . g . schizocytes ) are usually not found in dengue and the etiology of dengue-associated thrombocytopenia is multifactorial , we suggest that the observed changes in VWF-ADAMTS-13 play a role in the pathogenesis of DHF/DSS and in the etiology of thrombocytopenia in special . Disturbances in normal platelet-endothelium interaction may especially be relevant for DHF/DSS , since increasing evidence has shown platelet-endothelium interaction to be important for vessel wall stability during inflammation [44] . There is currently no specific treatment for DHF/DSS except careful fluid therapy . Although there is little evidence to support the practice of transfusing fresh-frozen plasma ( FFP ) , this is sometimes done in practice for severe bleeding or for correction of prolonged coagulation tests . An unintended advantage of FFP infusion might be the replenishment of ADAMTS-13 . One small trial in patients with acute dengue from Sri Lanka indeed found a small increase in platelet count in patients treated with 600 ml of FFP compared with isotonic saline [45] . Several limitations to our study should be considered . First , enrollment to our study was restricted to children with suspected DHF/DSS and samples from children with a mild dengue infection and uncomplicated dengue fever were not available for analysis . Whether the reported abnormalities in the VWF-ADAMTS13 system are specific for DHF/DSS or can also be found in less severe dengue infections therefore remains unknown . Second , platelet counts at discharge and in the control group were unavailable . In conclusion , our data show that WPB exocytosis of VWF in its active conformation and consumption of VWF and ADAMTS-13 are prominent phenomena in severe dengue , which may contribute to thrombocytopenia and organ dysfunction . Severe dengue is also associated with very high plasma OPG levels , of which the functional consequences need further study . Finally , a transient reduction in larger VWF multimers may develop during convalescence . | Severe dengue infections are characterized by thrombocytopenia , clinical bleeding and plasma leakage . Activation of the endothelium , the inner lining of blood vessels , leads to the secretion of storage granules called Weibel Palade bodies ( WPBs ) . We demonstrated that severe dengue in Indonesian children is associated with a strong increase in plasma levels of the WPB constituents von Willebrand factor ( VWF ) , VWF propeptide and osteoprotegerin ( OPG ) . An increased amount of the hemostatic protein VWF was in a hyperreactive , platelet binding conformation , and this was most pronounced in the children who died . VWF levels at enrollment were lower than expected from concurrent VWF propeptide and OPG levels and VWF levels did not correlate well with markers of disease severity . Together , this suggests that VWF is being consumed during severe dengue . Circulating levels of the VWF-cleaving enzyme ADAMTS-13 were reduced . VWF is a multimeric protein and a subset of children had a decrease in large and intermediate VWF multimers at discharge . In conclusion , severe dengue is associated with exocytosis of WPBs with consumption of VWF and low ADAMTS-13 activity levels . This may contribute to the thrombocytopenia and complications of dengue . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"coagulation",
"disorders",
"infectious",
"diseases",
"von",
"willebrand",
"disease",
"thrombocytopenia",
"tropical",
"diseases",
"(non-neglected)",
"dengue",
"fever",
"flavivirus",
"platelets",
"neglected",
"tropical",
"diseases",
"travel-associated",
"diseases",... | 2012 | Severe Dengue Is Associated with Consumption of von Willebrand Factor and Its Cleaving Enzyme ADAMTS-13 |
Mechanosensing at focal adhesions regulates vital cellular processes . Here , we present results from molecular dynamics ( MD ) and mechano-biochemical network simulations that suggest a direct role of Focal Adhesion Kinase ( FAK ) as a mechano-sensor . Tensile forces , propagating from the membrane through the PIP2 binding site of the FERM domain and from the cytoskeleton-anchored FAT domain , activate FAK by unlocking its central phosphorylation site ( Tyr576/577 ) from the autoinhibitory FERM domain . Varying loading rates , pulling directions , and membrane PIP2 concentrations corroborate the specific opening of the FERM-kinase domain interface , due to its remarkably lower mechanical stability compared to the individual alpha-helical domains and the PIP2-FERM link . Analyzing downstream signaling networks provides further evidence for an intrinsic mechano-signaling role of FAK in broadcasting force signals through Ras to the nucleus . This distinguishes FAK from hitherto identified focal adhesion mechano-responsive molecules , allowing a new interpretation of cell stretching experiments .
Focal adhesions ( FAs ) act as key cellular locations for mechanosensing by integrating mechanical and biochemical signals between the outside and inside of the cell , thereby regulating processes such as cell proliferation , motility , differentiation , and apoptosis [1–3] . They contain numerous adapter or anchor proteins , which establish the mechanical link of the cytoskeleton with the extracellular matrix [4] . Some of these proteins have been identified as mechano-responsive elements [5–7] . Focal Adhesion Kinase ( FAK ) centrally regulates FAs by establishing adhesive interactions at the cell periphery [8] . Acting as a signaling hub between integrin and multiple proteins participating in downstream signaling pathways , it carries out diverse functions in embryonic development , cell migration , and survival , and its malfunction is associated with cancer progression and cardiovascular diseases [9 , 10] . FAK comprises a central tyrosine kinase domain flanked by two large non-catalytic domains: FERM and FAT ( Fig 1A ) . The N-terminal three-lobed 4 . 1 ezrin radixin moesin ( FERM ) homology domain is connected to the kinase N-lobe through a 50-residue linker . The C-terminal FAT ( focal adhesion targeting ) domain follows a 220-residue long proline-rich and disordered linker , through which it is connected to the kinase C-lobe . The activation of FAK first requires autophosphorylation of Tyr397 , which offers a Src homology 2 ( SH2 ) binding site . Src binding to FAK increases Src kinase activity , inducing the phosphorylation of Tyr576/577 within the kinase domain activation loop [11] . This is needed for maximal FAK-associated activity and leads to the formation of a Src-FAK complex , which triggers subsequent phosphorylations in the FAT domain and binding of downstream signaling proteins [12] . The FERM domain auto-inhibits the kinase domain by blocking the Tyr576/577 phosphorylation site [13 , 14] . Exposure of this site is an essential step to permit its phosphorylation , and thereby render maximum FAK activity . FAK locates at sites of integrin clustering through protein-protein interactions of its FAT domain , which contains binding sites for integrin- and actin-associated proteins [4 , 15] . Integrin signaling and interactions with growth factor receptors were determined as FAK activators [16 , 17] . Recent studies provided evidence that phosphoinositide phosphatidylinsositol-4 , 5-bis-phosphate ( PIP2 ) is critical for efficient FAK activation and autophosphorylation [18 , 19] . PIP2 , a ubiquitous second messenger enriched in the inner leaflet of the plasma membrane and concentrated at FAs , regulates the interaction of cytoskeletal proteins with the membrane [20 , 21] . PIP2 interacts directly with the basic patch ( 216KAKTLR221 ) in the FERM domain [18 , 19] , which induces conformational changes in FAK . The global concentration of PIP2 in the cell membrane is only approximately 1% [22] . However , PIP2-protein interactions [22 , 23] or divalent ions , such as Ca2+ , [24 , 25] can lead to local PIP2 accumulation . Localized increments of Ca2+ were also suggested to increase the residency of FAK at FAs [26] . Evidence for a decisive role of FAK in mechanotransduction is steadily growing [27] . FAK is recruited to the leading edge and phosphorylated in migrating cells under shear stress [28 , 29] . It has also been shown to mediate force-guided cell migration [30 , 31] as well as strain-induced proliferation [32] . Recently , the mechano-sensitivity of FAK has been ascribed to the force-sensing fibronectin-integrin link [33] . However , until now , the available data on mechano-sensing through FAK is indirect . It remains unknown if FAK only lies downstream of mechano-sensing processes such as those involving integrins , or if FAK is also per se exposed to and activated by mechanical force . We here hypothesize that mechanical force acts as a direct stimulus of FAK activity , indications for which are two-fold . First , FAK is tethered between the PIP2-enriched membrane and the cytoskeleton , likely acting as a force-carrying link in FAs . Second , the FERM-kinase structure suggests itself as a mechano-responsive scaffold , in which force could specifically detach the autoinhibitory FERM domain from the active site . FAK would be the first mechanoenzyme of FAs , allowing a direct transduction of a mechanical signal into an enzymatic reaction and downstream events into the nucleus , which would yield a mechanistic explanation of FAK’s mechano-sensing role [10] . Indeed , two analogous cases of mechanically activated enzymes have been previously identified , both of which are kinases and feature force-induced activation by removal of an autoinhibitory domain , namely titin and twitchin kinase in muscle [34–36] . In contrast to the FAT domain [37] , the force response of the autoinhibited FERM-kinase fragment is currently unknown . To test the hypothesis of FAK as a force-sensor , we performed extensive equilibrium molecular dynamics ( MD ) and force-probe molecular dynamics ( FPMD ) simulations of the FERM-kinase fragment of FAK under various conditions . Force propagating onto FAK from a PIP2-enriched membrane and the cytoskeleton specifically opens the hydrophobic FERM-kinase interface , preparing FAK for activation via phosphorylation prior to the unfolding of the kinase domain . Given the low stability of the largely α-helical kinase and FERM domains , this is remarkable . The enforced activation is robust with regard to a large range of pulling velocities , but sensitive to the site of force application . Our force-induced activation pathway suggests a direct mechanoenzymatic function of FAK in FAs .
Equilibrium MD simulations of the FERM-kinase fragment ( FK-FAK ) fragment [14] ( PDB code: 2J0J ) in the apo state , both in the absence and in the presence of a membrane containing PIP2 and POPE lipids , as well as of only the membrane , were performed using the GROMACS package [38] . FPMD simulations [39 , 40] of FK-FAK without a membrane were performed by subjecting the C-terminal C-alpha atom and the center-of-mass of the C-alpha atoms of the basic patch 216KAKTLR221 to harmonic-spring potentials which were moved away from each other with constant velocity . FAT has been suggested to interact with FERM , binding to the same site as PIP2 does [41] . However , PIP2 binding is required for FAK activation [18 , 19] , thus excluding the possibility of FAT-FERM stable interactions for PIP2-mediated FAK activation . Other interactions between FAT and the FERM/Kinase complex are not known or at least suggested to be very dynamic and weak [42] , and thereby easier to break under force conditions . This suggests that under tensile force , FAT is maintained sufficiently far from the complex , and that the force is transduced towards the complex through the fully-stretched 200 amino-acid proline-rich disordered linker . In consequence , in our FPMD simulations neither FAT nor the linker were considered . The force response of membrane-bound FK-FAK was investigated by subjecting its C-terminus to a harmonic potential that was then moved away from the membrane either vertically or diagonally , while keeping the membrane position at its original position . PLS-FMA [43] was used to detect collective motions maximally correlated with the opening of the FERM-kinase interface . The underlying free energy landscape was characterized by analyzing the rupture forces as a function of loading rate , using both the HS model by Hummer & Szabo [44] and the BSK model by Bullerjahn et al . [45] . Kinetic models were based on previous biochemical networks [46–49] and simulated using COPASI [50] . Details of the methods are given in S1 Text .
In vitro , phosphorylation of the activation loop of FAK is enhanced by relieving the autoinhibition through Y180/M183 mutation [14] or PIP2-binding [18] . Catalytic turnover of wild-type FAK , however , requires an additional biochemical stimulus . Here , we ask if mechanical force could promote full domain dissociation of FAK as required for auto- and Src-phosphorylation–analogous to the effect of the Y180/M183 mutation . We examined the effect of tensile force on the autoinhibited FK-FAK using FPMD simulations . Tethering FAK between the membrane and the cytoskeleton results in force transmission from the membrane onto the basic patch of the FERM domain and from the paxillin-interacting FAT domain through the proline-rich linker onto the kinase C-terminus ( Fig 1A ) . Accordingly , in our simulations , a pulling force was applied to the basic patch of FERM and the C-terminus of the kinase domain in opposite directions with 13 different pulling velocities from 6×10−3 nm/ns to 1 nm/ns ( 1 in Fig 1B ) . For each pulling velocity , multiple runs were carried out ( 83 runs in total ) , yielding a concatenated simulated time of about 7 μs , with the slowest pulling simulation covering 1 μs . We observed the autoinhibitory FERM domain to dissociate from the kinase domain in 76 out of 83 FPMD simulations ( more than 90% of the cases ) . Conformational damage of either the FERM F2-lobe or kinase C-lobe occurred in the remaining 7 simulations . Release of Tyr576/577 from FERM occurred always later , i . e . at larger end-to-end distances , than dissociation of the F2-lobe from the C-lobe ( Fig 1B and 1C ) , suggesting the F2-C detachment to be a requirement for mechanical FAK activation . We observed partial unfolding at the kinase C-terminus prior to exposure of Tyr576/577 to an only minor extent and mostly at higher loading rates , comprising at most a 15 nm increase in end-to-end length ( Fig 1B and 1C ) , or no more than 30 residues of the C-terminal α-helix ( S1 Fig ) . As the second half of this helix ( or more ) is typically disordered in other kinases ( e . g . in protein kinase A or Src ) , its partial unfolding under force is likely not to impair FAK enzymatic function . Hence , our data suggest domain-domain dissociation to largely thwart the unfolding of the moderately stable α-helical domain structures . However , more substantial unfolding from the C-terminus of the kinase domain was the dominant pathway when pulling FK-FAK from its N and C-terminus ( S2 Fig ) . Thus , we conclude that force acting specifically between the FERM basic patch and the kinase C-terminus removes the inhibitory FERM domain and thereby facilitates kinase activation , instead of domain unfolding and kinase inactivation . At FAs , the specific interaction of the FERM basic patch with PIP2 is required for the anchoring of FK-FAK to the membrane . Other phospholipids only display background levels of binding [18] . In our previous study , we observed an allosteric change at the FERM-kinase interface upon PIP2 binding to FK-FAK , but no full opening [18] . This raised the question if full domain opening under force , as observed for isolated FK-FAK in solution , also occurs when FK-FAK is anchored to a membrane via PIP2 . This would require both the PIP2-containing membrane as well as the PIP2-FERM link to be mechanically more robust against rupture than the FERM-kinase interaction . To test this , we set up a palmitoyloleoylphosphatidylethanolamine ( POPE ) membrane containing 15% ( mol/mol ) of PIP2 in the inner leaflet of the membrane , which was surrounded by water and neutralized by CaCl2 . Within 100 ns of MD simulations starting from individual PIP2 molecules in the membrane , we observed the formation of small PIP2 clusters involving two or more lipids and Ca2+ ( S3 Fig ) , accompanied by a decrease of area per lipid by ∼ 1 Å2 ( S1 Table ) , in agreement with divalent-cation-mediated PIP2-enrichment in membranes [18 , 24 , 51] . FK-FAK was anchored to the membrane and the dynamics of the resulting complex was monitored over 150 ns of MD . Anchorage further increased clustering . The protein remained stably bound to the membrane through the FERM-PIP2 and additional interactions between the kinase C-lobe and the membrane , independent from the initial orientation of the protein relative to the membrane plane ( Fig 2A left and S3 Fig ) . The same was observed for a membrane with 1% PIP2 , which , however , showed less clustering and provided only a single PIP2 lipid for anchorage of FK-FAK . Next , we monitored the mechanical response of membrane-anchored FK-FAK . In FPMD simulations , we subjected the protein to force by moving a harmonic spring attached to the kinase C-terminus with constant velocity along a direction vertical or diagonal to the membrane , while position restraining the center-of-mass of the membrane bilayer ( Fig 2A ) . At 15% PIP2 concentration , independent of the pulling direction , we observed a loss of contacts of the kinase domain with the membrane and with the FERM domain , while the FERM-membrane interaction remained intact ( Fig 2B ) . While diagonal pulling led to a concurrent dissociation of the kinase from the membrane and the FERM domain , vertical pulling resulted in kinase-membrane dissociation prior to kinase-FERM dissociation . In none of these simulations , we observed kinase unfolding prior to dissociation . Also , for both pulling directions , the membrane and the PIP2-FERM interaction were mechanically more robust than those at the FERM-kinase interface . Thus , the membrane simulations reproduced the process predominantly observed for isolated FK-FAK in solution ( compare Fig 2B with Fig 1B and 1C ) . Namely , they all showed force-induced removal of the autoinhibitory FERM domain and exposure of the activation loop carrying the Tyr576/577 phosphorylation site . When we applied force to FK-FAK anchored to a membrane containing only 1% PIP2 , i . e . to a single PIP2 molecule , detachment of the kinase domain from the membrane was followed by the detachment of also the FERM domain ( Fig 2C ) . Full loss of membrane anchoring naturally stops force transmission and impedes activation . Thus , an interaction of the FERM basic patch with multiple PIP2 , which is likely in PIP2-enriched membranes , is required for mechanical FK-FAK activation . This is in line with the fact that PIP5K overexpression increases and PIP5K knockdown decreases the open FAK conformation [19] . The pulling direction , instead , appears to be less relevant . We next analyzed in further detail the dynamics underlying the force-triggered FK-FAK domain-domain rupture . Applying partial least squares functional mode analysis ( PLS-FMA ) [43] to the simulations of isolated FK-FAK in solution , we obtained a collective opening motion that maximally correlates with the increase in minimal distance between the F2 and C lobes ( S4 Fig ) . This opening motion also strongly correlated with the F2-C lobe distances obtained for the trajectories of membrane-bound FK-FAK , suggesting that it captures the essential opening dynamics of FK-FAK both isolated and bound to the membrane . This implies the simplified system of isolated FK-FAK in solution to follow a FERM-kinase dissociation mechanism , which is highly similar to the one of the more realistic system including the membrane , even though it lacks effects from FERM/kinase-membrane interactions . We then identified the first steps along the opening motion of FK-FAK giving rise to rupture forces . Fig 3A shows typical force profiles and F2/C-lobe interaction areas as a function of the spring locations recovered from the FPMD simulations . For both FK-FAK in isolation and bound to the membrane , and independent of the loading rate , we observed that the interface area between the two lobes was reduced in two steps , both of which coincided with noticeable force peaks . The maximal force was reached when the first decrease in inter-lobe area occurred ( from 3–4 . 5 nm2 to 1 . 5–2 . 8 nm2 ) . This led to a short-lived intermediate , as reflected by a second peak in the distribution of the F2/C-lobe interface area ( Fig 3B ) , before the two lobes fully dissociated . We note that the intermediate becomes less evident for faster pulling velocities . To pinpoint the load-carrying residue-residue interactions across the interface , we calculated the punctual stress of each residue , using time resolved force distribution analysis ( TRFDA ) [52] , and thereby detected the loss of inter-lobe interactions during pulling ( see S5 Fig , S2 Table , and S1 Text ) . Inter-lobe interactions which ruptured reproducibly at one of the two dissociation steps are highlighted in Fig 3C . The first major rupture step required the breakup of a hydrophobic cluster composed of residues Y180 , M183 , N193 , V196 , and F596 , and of an additional salt bridge ( D200-R598 ) . Rupture of these interactions gave rise to the maximal force , thus stressing their critical stabilizing role . Our results are in agreement with the observation that mutations Y180A , M183A , and F596D result in constitutively active FAK with an open FERM-kinase interface [14 , 18] . Residue pairs rupturing at the second step included residues of mostly electrostatic nature ( E182 , R184 , K190 and N595 , N628 , N629 , E636 ) and are located further away from the membrane anchor . This second rupture step is immediately followed by the opening of the remaining FERM-kinase interface established between the F1 and the N-lobe , including the exposure of Tyr576/577 . Thus , the rupture process resembles a zipper-like mechanism , during which the FERM and kinase interface is sequentially opened . Herein , the membrane-proximal hydrophobic patch around F596 represents the most robust mechanical clamp to be opened first . Our PLS-FMA calculations further support this sequential mode of opening ( S4C Fig ) . Are the forces predicted by the simulations to relieve FAK autoinhibition relevant to FAK at FAs ? At thirteen different loading rates , covering two orders of magnitude , we obtained maximal rupture forces for FK-FAK activation between 150 and 450 pN ( Fig 4A ) . This force regime is similar to the one observed for titin kinase ( 400 pN at 0 . 2 pN/ps ) , a kinase known to be mechanically activated by forces present in muscle [35 , 36 , 53] . Our rupture forces are also similar to or slightly higher than those predicted by MD simulations of the focal adhesion proteins talin and vinculin ( 250–400 pN for nanosecond scale activation of talin [54 , 55] and 100 pN for sub-nanosecond activation of vinculin [56] , respectively ) . We then used both the HS model [44] and the BSK model [45] to fit the observed rupture forces as a function of the loading rate . This provided us with a set of compatible model parameters ΔG , D and xb , where ΔG denotes the activation energy , D the effective diffusivity and xb the separation between the inactive state and the transition state . Focusing on those parameter combinations that correspond to a physiologically plausible spontaneous activation rate of no more than k0 = 10−3 Hz , we obtained a number of best-fit estimates from which we derived the force-dependent activation rate k ( F ) used in our kinetic model ( Fig 4A and S6 Fig ) . We note that model parameters corresponding to an unphysiologically high spontaneous dissociation rate can improve our fit to the observed force fluctuations ( S7 and S8 Figs ) . On this basis it might be speculated that there exists a second energy barrier at a larger value of xb that guarantees thermal stability at low forces , but vanishes under the high forces used in our MD simulations . Nevertheless , this does not invalidate our qualitative findings on FK-FAK activation as force sensitivity increases exponentially with the barrier location xb ( see the S1 Text for a more detailed analysis ) .
We here provide computational evidence for a force-induced activation mechanism of FAK , in which tensile force relieves the blockage of its active site , as well as its central Tyr576/577 phosphorylation site , imposed by the autoinhibitory FERM domain . The release of autoinhibition is likely to make the kinase active site accessible for its substrate , Tyr397 of the same or another FAK molecule [42 , 57] , and/or to render Tyr576/577 accessible to Src . It is non-trivial that the exertion of a pulling force at opposite sites of the FERM and kinase domains leads to their dissociation . Other likely scenarios are protein unfolding and PIP2-protein dissociation , both of which would inactivate FAK , because an intact kinase structure and also PIP2 binding [18] are required for FAK activity . In fact , α-helical proteins are known to unfold at forces typically lower than β-sheet proteins [58] , and both the kinase C-lobe and the FERM F2 domain feature mainly α-helical secondary structure . In this regard , force-induced FAK unfolding would be an expected result and was indeed preferred over FERM-kinase dissociation when pulling the FERM F1- or F3-lobe away from the kinase C-terminus . Instead , in the particular –and physiologically relevant– case that force is applied to the PIP2 binding site and the kinase C-terminus , we found domain-domain dissociation to be strongly preferred over unfolding or membrane detachment , over a large range of pulling velocities , and robust with regard to the pulling direction and presence of membrane interactions . Thus , the α-helical regions subjected to the pulling force mostly refrain from unfolding , and they instead transduce the load to the F2/C-lobe interface , which readily opens prior to substantial kinase unfolding . We suggest that it is the zipper-like topology , with the force application sites both located at the membrane-proximal basis of the two domains , that mechanically weakens the domain interface , resulting in efficient FAK opening and activation . Force transduction through the two termini , in contrast , results in shearing the two domains relative to each other , making them less prone to dissociate . The lower mechanical resistance of zipper versus shear-type topologies has been described earlier ( e . g . [53] ) , and FAK force-induced domain-domain rupture and activation appears to be another variation of this theme . Our findings not only decidedly argue for a mechano-sensing function of FAK , but also emphasize the crucial role of the membrane in mechanotransduction . First , membrane binding allows force to propagate to the F2-lobe of FAK . Second , the FAK-membrane interaction withstood the external load only in the case of PIP2-enriched membranes ( 15% PIP2 ) . In contrast , FAK detached from low PIP2-content membranes . This supports the notion of PIP2 clustering as a requirement for FAK activation at FAs [18 , 51] . We note that FAK can be activated in vitro by the sole action of PIP2 and Src , i . e . in the absence of tensile forces acting on membrane-bound FAK at FAs in stretched cells . However , it has become clear that FAK activation can proceed along different routes , depending on the cellular environment , and potentially can also involve pH changes [59] , and/or growth factor receptors [16] . We here propose force to substitute or complement some of these activators shaping the multi-dimensional landscape of FAK activity . Using TRFDA , we recovered the stabilizing role of Y180 , M183 , V196 and F596 , a hydrophobic core previously shown by mutagenesis to stabilize the autoinhibited state [14] , validating our simulation data . In addition , we found D200 and R598 to contribute to the rupture force , and predict their mutation to result in increased FAK activity . The question arises , how force feeds into FAK-mediated Ras GDP/GTP exchange and regulates ERK-dependent gene expression , analogous to the chemical stimulation of growth factor receptor-dependent Ras signaling ( S9 Fig ) [27] . To assess how FAK as a mechanosensor couples mechanical signals into the downstream biochemical network , we defined force-dependent FAK activation as the initial step of a kinetic model for the Ras signaling pathway ( S10 Fig ) [46–49] . Both FAK opening and GDP/GTP exchange in Ras are accelerated by external forces , as expected ( Fig 4B and 4C and S11A Fig ) . Intriguingly , while FAK force-induced activation shows a nearly linear dependency on force on the logarithmic scale ( Fig 4B ) , Ras-GTP production shows a highly non-linear dependency and saturates beyond a critical force ( Fig 4C ) . The reason is that activated FAK in complex with its partners , Grb2 and SOS via c-Src and SHC , acts as an enzyme for Ras activation . As a direct consequence , Ras activation follows mechano-enzymatic kinetics reminiscent of an inhibitory Michaelis-Menten mechanisms ( S11B Fig ) [60] , in which force regulates the enzyme concentration . In conclusion , our computational study provides direct evidence at the molecular level for a mechano-sensory role played by FAK at PIP2-enriched membranes of FAs . Through a specific domain opening mechanism regulated by force , FAK can integrate mechanical and chemical stimuli into downstream signaling to the nucleus . We suggest the mechano-enzymatics of FAK and Ras to provide a cap on the cell’s mechano-response . Our results , on FAK activation and signaling , are directly testable among others by molecular force sensors [61 , 62] and cell stretching experiments . How other putatively mechano-activated kinases , such as the related Src kinase , follow similar mechanisms , at focal adhesions or elsewhere , remains to be shown . | Focal adhesions integrate external mechanical signals into biochemical circuits allowing cellular mechanosensing . Although the zoo of mechanosensing proteins at focal adhesions is steadily growing , force-induced enzymatic mechanisms , as those uncovered for autoinhibited kinases in muscle , remain to be identified for focal adhesion downstream signaling . Here , we provide evidence that focal adhesion kinase ( FAK ) can act as a direct mechano-enzyme at focal adhesions , using molecular dynamics simulations and kinetic modelling . We show that anchorage of FAK to the membrane via PIP-2 is critical for this mechanical activation . Our results suggest similar mechanisms to be at play for other membrane-bound autoinhibited kinases . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Mechanism of Focal Adhesion Kinase Mechanosensing |
Mitochondrial neurogastrointestinal encephalomyopathy ( MNGIE ) is a severe human disease caused by mutations in TYMP , the gene encoding thymidine phosphorylase ( TP ) . It belongs to a broader group of disorders characterized by a pronounced reduction in mitochondrial DNA ( mtDNA ) copy number in one or more tissues . In most cases , these disorders are caused by mutations in genes involved in deoxyribonucleoside triphosphate ( dNTP ) metabolism . It is generally accepted that imbalances in mitochondrial dNTP pools resulting from these mutations interfere with mtDNA replication . Nonetheless , the precise mechanistic details of this effect , in particular , how an excess of a given dNTP ( e . g . , imbalanced dTTP excess observed in TP deficiency ) might lead to mtDNA depletion , remain largely unclear . Using an in organello replication experimental model with isolated murine liver mitochondria , we observed that overloads of dATP , dGTP , or dCTP did not reduce the mtDNA replication rate . In contrast , an excess of dTTP decreased mtDNA synthesis , but this effect was due to secondary dCTP depletion rather than to the dTTP excess in itself . This was confirmed in human cultured cells , demonstrating that our conclusions do not depend on the experimental model . Our results demonstrate that the mtDNA replication rate is unaffected by an excess of any of the 4 separate dNTPs and is limited by the availability of the dNTP present at the lowest concentration . Therefore , the availability of dNTP is the key factor that leads to mtDNA depletion rather than dNTP imbalances . These results provide the first test of the mechanism that accounts for mtDNA depletion in MNGIE and provide evidence that limited dNTP availability is the common cause of mtDNA depletion due to impaired anabolic or catabolic dNTP pathways . Thus , therapy approaches focusing on restoring the deficient substrates should be explored .
Mitochondrial DNA depletion syndrome refers to a group of inherited metabolic disorders that considerably differ in their clinical expression and genetic origin [1] . These are rare , but severe diseases with no effective treatment , and patients die in the early infancy in most cases . Four of the 9 genes known to be associated with this group of disorders to date are directly involved in the metabolism of deoxyribonucleoside triphosphates ( dNTPs ) , required for mtDNA replication [2]–[5] . Two of these genes ( TK2 , Entrez Gene ID 7084 , and DGUOK , Entrez Gene ID 1716 ) encode the mitochondrial enzymes thymidine kinase 2 ( TK2 ) and deoxyguanosine kinase ( dGK ) , which catalyze the first step in mitochondrial salvage of pyrimidine and purine deoxyribonucleosides , respectively . RRM2B ( Entrez Gene ID 50484 ) encodes the p53-inducible small subunit ( p53R2 ) of ribonucleotide reductase ( RNR ) , the key enzyme in de novo synthesis of deoxyribonucleotides . A p53-independent small subunit of RNR is highly expressed in dividing cells , but is degraded and its expression downregulated in quiescent cells; thus , the de novo supply of dNTPs for mitochondrial DNA ( mtDNA ) replication is dependent on p53R2 [6] . Mutations in the fourth gene , TYMP ( Entrez Gene ID 1890 ) , encoding thymidine phosphorylase , cause mitochondrial neurogastrointestinal encephalomyopathy ( MNGIE , OMIM ID #603041 ) [4] . Dysfunction of this cytosolic enzyme causes expansion of the dTTP pool [7] . dNTP supply for mtDNA replication ( Figure 1 ) depends on intramitochondrial salvage of deoxyribonucleosides , or on the mitochondrial import of deoxyribonuceloside phosphates from the cytosol , made de novo from ribonucleotide reduction or from cytosolic deoxyribonucleoside salvage . These pathways are interconnected and the relative contributions of each pathway depend on a number of different factors , most importantly , the cell cycle ( reviewed in [8] ) . It is generally assumed that imbalances in the mitochondrial dNTP pools interfere with the correct mtDNA replication and/or repair [1] , [8] , but there is no experimental evidence on the particular consequences of such imbalances on mtDNA replication . The precise mechanistic details of this effect remain largely unknown; the foreseeable depletion of dNTPs resulting from a loss of function of TK2 , dGK and RNR fits well with mtDNA depletion syndrome , since dNTPs are the substrates of mtDNA replication , but it is unclear how an excess of a given dNTP , specifically the dTTP expansion associated to thymidine phosphorylase dysfunction in MNGIE , might lead to mtDNA depletion . Other groups have reported the effects of thymidine addition on dNTP pools in cell culture . It has long been known that , in cycling cells , millimolar concentrations of thymidine cause pronounced dCTP depletion due to the allosteric effects of an expanded dTTP pool on RNR , which strongly inhibit the nuclear DNA replication rate [9] . More recently , several studies have shown that micromolar thymidine concentrations , similar to those found in MNGIE patients , result in slight to moderate cytosolic and mitochondrial dCTP depletion in cycling and quiescent cultured cells [10]–[12] . Here , we studied the consequences of dNTP imbalances on mtDNA replication in two different experimental models , and found that the limited availability of one dNTP is a common mechanism accounting for mtDNA depletion even under dTTP excess . The replication is limited by the availability of the dNTP present at the lowest concentration and , in the case of dTTP excess , secondary dCTP depletion accounts for the reduction of mtDNA replication rate .
We used an in organello replication approach [13] to investigate whether excess of any of the 4 dNTPs affects mtDNA replication , measured by incorporation of radiolabel from [3H]-dNTP or [α-32P]-dNTP into mtDNA . Through controlled addition of the 4 dNTPs to the reaction , this model enabled us to study the consequences of experimentally designed mitochondrial dNTP imbalances on mtDNA replication . To assess modifications of the intra-mitochondrial dNTP pool resulting from addition of each dNTP to a mitochondrial suspension , mitochondria were extracted from liver of C57BL/6J mice , endogenous dNTPs were measured , and the effect of adding no dNTPs , 1 µM of all 4 dNTPs , and 100 µM of each dNTP was studied ( Table 1 ) . All 4 endogenous dNTPs were partially depleted over 2 hours of in organello reaction . While dCTP was the most abundant endogenous nucleotide in fresh mitochondria , it became the least plentiful at the end . Addition of 1 µM of extramitochondrial dNTPs led to 4-fold to 7-fold increases in intramitochondrial dNTPs after 2-hour incubation . Addition of 100 µM of dTTP , dCTP , or dGTP resulted in a 100-fold to 270-fold increase in the amount of each nucleotide; when 100 µM dATP was added , a much larger increase ( ∼600-fold ) was observed in the intramitochondrial amount of this nucleotide . These results were independently investigated by measuring the radioactivity incorporated into mitochondria when 100 µM of tritium-labeled dNTPs were added to the reaction . We found that 2 to ∼6 . 5 times more dATP ( or its dephosphorylated derivatives ) entered the mitochondria than dTTP , dCTP , or dGTP ( Figure S1 ) . The enhanced transport observed when dATP was added in large excess might be mediated by the abundant ADP/ATP carrier; it has long been known that dADP and dATP can be substrates of this ribonucleotide transporter [14] , [15] . Interestingly , comparison of these results with those of Table 1 shows that most dNTP molecules transported into mitochondria are partially or totally dephosphorylated . Using this model , a 100-fold increase of dATP or dCTP concentrations in the reaction did not produce detectable changes in mtDNA replication , measured by incorporation of 3H or 32P into mtDNA , whereas excess dGTP induced a 45% increase ( Figure 2A , 2B , 2C ) . The same excess of dTTP , however , caused a significant 25% decrease in mtDNA replication . Thus , only dTTP excess had a negative effect on mtDNA replication . In an early study it was reported that addition of thymidine to Chinese hamster ovary cells leads to an increase in the dTTP pool and cessation of nuclear DNA replication , resulting from dCTP depletion [9] . More recently , several reports have shown reductions in cellular or mitochondrial dCTP in parallel to dTTP expansion following thymidine overload in cultured cells [10]–[12] . We then measured the 4 dNTPs in the presence of dTTP overload to determine whether additional mitochondrial dNTP imbalances were produced in our model . A significant decrease in dCTP concentration was observed in association with dTTP excess , with no effect on dATP or dGTP ( Table 1 and Figure 2D ) . The dCTP contraction was not caused by an inhibitory effect of the large dTTP excess on transport of the exogenously provided dCTP because incorporation of [5 , 5′-3H]dCTP label into mitochondria was not affected by increasing amounts of dTTP ( Figure 2E ) . We tested the influence of dTTP overload on mtDNA replication in 2 additional conditions: 1 ) without adding dCTP and 2 ) without adding either dCTP or dGTP ( Figure 2F ) . The inhibitory effect of dTTP on mtDNA replication was also observed in the absence of exogenous dCTP , thus confirming that this effect is not caused by inhibition of dCTP transport by an excess of dTTP . The effect of dTTP overload ( 100 µM added ) on the endogenous dCTP pool at the end of the reaction was tested in 5 independent experiments , which also revealed dCTP contraction ( by 56 . 4%±20 . 6%; P = 0 . 016 , Mann-Whitney U test ) . In contrast , dTTP overload did not influence mtDNA replication when exogenous dCTP and dGTP were both omitted , probably because endogenous dGTP is the limiting substrate of the reaction . This is consistent with the fact that dGTP is present in the lowest concentration in freshly isolated mitochondria ( Table 1 ) . In keeping with this concept , dGTP was the only dNTP that , in excess , stimulated mtDNA replication ( Figure 2A ) . Two alternative mechanisms could account for the negative effect of dTTP excess on mtDNA replication: 1 ) dTTP overload , in itself , may slow down the replication machinery , or 2 ) secondary dCTP depletion may restrict replication because dCTP becomes the limiting substrate . To test these alternatives , we attempted to restore mtDNA replication by supplying additional dCTP to a reaction under dTTP excess . As shown in Figure 3A , 3B , the dose-dependent negative effect of dTTP on mtDNA replication was prevented , also in a dose-dependent manner , by dCTP supplementation . Addition of dATP or dGTP failed to prevent the negative effect of dTTP ( Figure 3C ) , even though dGTP had an independent positive effect , as was mentioned above ( Figure 2A ) . This positive effect was not due to a dGTP-induced dCTP increase , because the addition of 100 µM dTTP induced similar reductions of dCTP levels both in the absence ( 0 . 85±0 . 20 pmol/mg prot , from Table 1 ) and in the presence ( 0 . 71±0 . 20 pmol/mg prot; N = 3 ) of 100 µM dGTP . Monitoring mtDNA replication over 2 hours showed that the dGTP-induced positive effect was very pronounced in the initial phase ( first 30 min ) , and very reduced or negligible during the second hour ( Figure S2 ) , suggesting that , in our model , dGTP concentration is the limiting factor only in the initial phase of the replication . These results demonstrate that the decrease in mtDNA replication occurring under dTTP overload in our model is caused by secondary dCTP depletion , rather than by dTTP excess . We postulate that progression of mtDNA replication in organello is limited by the availability of each dNTP substrate . This notion is supported by 3 observations: 1 ) dGTP , the endogenous nucleotide in lowest concentration , was the only dNTP that , in excess , stimulated mtDNA replication ( Figure 2A ) , 2 ) no positive or negative effect was observed by dTTP excess in the absence of exogenous dGTP ( Figure 2F ) , and 3 ) mtDNA replication was sensitive to removal of each separate exogenous dNTP ( Figure S3 ) . The primary biochemical imbalance in thymidine phosphorylase-deficient MNGIE patients is systemic accumulation of thymidine , which leads to expansion of the dTTP pool in cultured cells [10]–[12] and in vivo [7] . The results reported above indicate that most triphosphate molecules entering the mitochondria lose one or more phosphates . Therefore , we tested whether the effect of thymidine overload in our in organello reaction caused effects similar to those observed with dTTP excess ( Figure 3D , 3E , 3F ) . Thymidine overload slowed down mtDNA replication , with and without addition of 1 µM dCTP ( Figure 3D ) and , again , this effect disappeared when dGTP addition was omitted . Addition of extramitochondrial thymidine increased intramitochondrial dTTP , revealing active in organello phosphorylation through TK2 , and decreased dCTP , paralleling the results obtained with dTTP excess ( Figure 3E ) . dATP and dGTP also showed a slight trend to be reduced , but far from the ∼50% reduction observed for dCTP . Noteworthily , while intramitochondrial dTTP expansion was moderate ( 3-fold ) and far from the nearly 30-fold increase observed when 100 µM of dTTP was added ( Table 1 ) , the extent of the dCTP decrease was very similar , suggesting that dCTP contraction may be caused mainly by thymidine rather than by dTTP . Addition of deoxycytidine or dCTP restored mtDNA replication in the presence of thymidine overload ( Figure 3F ) , thus confirming that thymidine excess , in itself , does not slow down mtDNA replication . Phosphorylation of thymidine and deoxycytidine is catalyzed by the same mitochondrial kinase , TK2 . Studies in the human enzyme have shown that TK2-catalyzed deoxycytidine phosphorylation is competitively inhibited by thymidine , and thymidine phosphorylation is also inhibited ( less effectively ) by deoxycytidine , while both are feedback-inhibited by dTTP and dCTP [16] ( Figure 4 ) . Therefore , we would expect that dCTP excess and the deoxycytidine generated by dCTP dephosphorylation , would also reduce mtDNA replication . In our model , dCTP excess did not affect mtDNA replication in the presence of dTTP ( Figure 2A and Figure 3C ) . However , when we repeated the experiment in the absence of exogenous dTTP ( Figure S4 ) , addition of 100 µM dCTP decreased mtDNA replication by 44 . 4% ( ±4 . 3% ) , supporting the idea that the effects produced by dTTP and dCTP excesses are both TK2-mediated . Secondary dCTP depletion due to TK2 inhibition has been suggested to occur in MNGIE [10] , [17] . The present report provides the first direct evidence that thymidine-induced mitochondrial dCTP contraction , previously observed by others [10]–[12] , delays mtDNA replication in vitro , rather than dTTP excess . We found that ∼40% contraction of the mitochondrial dCTP pool leads to decreased mtDNA replication in organello , while dTTP increases as high 30-fold do not produce observable effects . Previous in vivo observations are consistent with this mechanism . Brain and liver mitochondrial dTTP were both found to be increased in the double Tymp/Upp1 knockout murine model of MNGIE , but mitochondrial dCTP depletion was only seen in brain . Interestingly , brain , but not liver , had mtDNA depletion [7] , supporting the notion that dCTP depletion , and not dTTP excess , is associated with mtDNA copy number decreases . Thymidine overload is reported to cause mtDNA depletion in human cultured cells [11] . Using the same model , we tested whether addition of deoxycytidine prevents the negative effects of thymidine in human cells . We found that mtDNA depletion caused by 40 µM thymidine is prevented by co-treatment with 40 µM deoxycytidine ( Figure 5 ) . Similarly , after 30 days under thymidine overload , mtDNA-depleted cells gradually recovered mtDNA copy number when deoxycytidine was added to the medium . Therefore , our results demonstrate that the negative effect of thymidine on mtDNA in cultured human cells can be prevented by deoxycytidine . Despite the increasing number of studies investigating the molecular mechanisms involved in mtDNA depletion syndrome , many unsolved questions require further effort [1] , [8] , [18] . In the case of MNGIE , other groups have developed models to study the effect of thymidine overload on dNTP pools and on mtDNA replication in cell culture [10]–[12] . Interestingly , all these studies revealed moderate cytosolic and mitochondrial dCTP depletion secondary to thymidine overload , as well as diverse effects on mtDNA . After a long treatment ( 8 months ) of HeLa cells with thymidine , Song et al [12] found multiple deletions in mtDNA , but failed to detect depletion , probably because the study was performed using a highly proliferating cell line . More recently , Pontarin et al [11] tested the effects of thymidine on quiescent lung and skin fibroblasts , and mtDNA depletion , but not deletions , was observed . Here , we have tried to address how dTTP expansion and dCTP contraction contribute to the reduction of mtDNA copy number using isolated murine mitochondria . This model allowed us to study in controlled conditions the changes of the replication rates induced by designed dNTP imbalances . Our results indicate that mitochondrial dNTP availability is the key factor for mtDNA maintenance in mtDNA depletion syndrome caused by mutations in TYMP , TK2 , DGUOK , and RRM2B . These last 3 genes directly compromise the salvage or de novo supply of mitochondrial dNTPs , in agreement with the clinical severity of cases with TK2 , DGUOK , and RRM2B mutations . In MNGIE patients , however , all anabolic steps that provide mitochondrial dNTPs for mtDNA replication are fully functional , and the main factor accounting for the impaired replication is believed to be dTTP excess caused by thymidine phosphorylase deficiency . While this is likely true for the multiple deletions or somatic point mutations generated by a dTTP-driven next nucleotide effect [19] , [20] , the experimental evidence presented here indicates that mitochondrial dCTP depletion secondary to thymidine excess accounts for mtDNA depletion , rather than dTTP excess . Although using isolated mitochondria allowed us to easily study the effect of dNTP imbalances on mtDNA replication , some limitations derived from this simplified system should prompt one to be cautious when interpreting the results . For example , isolated mitochondria lack deoxycytidine kinase ( dCK ) , the cytosolic enzyme that catalyzes the phosphorylation of all deoxyribonucleosides except deoxyuridine and thymidine , which do not inhibit deoxycytidine phosphorylation by dCK . Our in organello system is also lacking de novo dCDP ( and other dNDPs ) synthesis via RNR . In cultured cells and in vivo , these pathways could compensate or reduce the effect of the mitochondrial dCTP decay studied here , because deoxyribonucleotides can be imported from the cytosol to mitochondria [21] . However , as mentioned above , mitochondrial dCTP depletion has been reported in cell culture and in in vivo models of MNGIE [7] , [10]–[12] , indicating that , at least in some cells and tissues , this compensation does not suffice to normalize mitochondrial dCTP levels . Further support came from our recovery experiments in quiescent fibroblasts: cotreatment with deoxycytidine largely prevented thymidine-induced mtDNA , indicating that dCTP contraction should be the cause of the thymidine-induced mtDNA depletion not only in isolated mitochondria but also in cultured cells . Therefore , the common biochemical condition in mitochondrial DNA depletion syndrome due to imbalanced dNTP pools ( including MNGIE ) is a limited availability of one or more substrates for mtDNA replication . Based on this idea , therapy approaches for this group of disorders should focus on supplying the specific substrates that are depleted in each enzyme defect . Previous studies have experimentally examined the potential benefits of biochemically by-passing these enzyme defects . Supplementation with purine deoxyribonucleoside monophosphates prevented mtDNA depletion in dGK-deficient cultured cells [22] , [23] . In the case of MNGIE , the treatment relies on providing thymidine phosphorylase activity through allogeneic hematopoietic stem cell transplantation [24] , but preventing mitochondrial dCTP depletion could be a complementary approach . Deoxycytidine supply would compensate for thymidine competition for TK2-mediated phosphorylation , or the nucleoside would be phosphorylated by dCK resulting in more cytosolic deoxycytidine nucleotides available to be imported by mitochondria . Both mechanisms would increase mitochondrial dCTP levels , thus preventing the negative effects of dCTP depletion on mtDNA replication . However , the optimal way to provide this precursor is not obvious , because of the complicated interconnections between the enzymes of dNTP metabolism . Enzymatic deamination would limit the availability of deoxycytidine , as happens with deoxycytidine analogues [25] . Cytidine deaminase inhibitors , alone or in combination with deoxycytidine , would be an alternative [25] , because they would limit the conversion of endogenous or supplied deoxycytidine to deoxyuridine . The availability of a murine model of the disease [7] will enable us to test the potential benefits of these approaches , and to find the way to avoid possible pitfalls .
Mitochondria were isolated as described [26] with minor modifications . Two to 3-month-old male C57BL/6J mice were killed by cervical dislocation and the liver was rapidly removed and chilled in homogenization buffer ( 4 mL/g tissue , 320 mM sucrose , 1 mM EDTA , 10 mM Tris-HCI , pH 7 . 4 ) . All further operations were carried out at 2–4°C using sterile solutions . The liver was cut in small pieces , homogenized in a Dounce homogenizer ( 4 up-and-down strokes ) , and spun at 1 , 000×g for 5 min . The supernatant was aliquoted in Eppendorf tubes and centrifuged at 13 , 000×g for 2 min . The mitochondrial pellets were washed 3 times in homogenization buffer and once in incubation buffer ( 25 mM sucrose , 75 mM sorbitol , 100 mM KCl , 10 mM K2HPO4 , 0 . 05 mM EDTA , 5 mM MgCl2 , and 10 mM Tris-HCl , pH 7 . 4 ) . Lastly , the mitochondrial pellet was resuspended in 1 mL of incubation buffer and protein concentration was determined ( Coomassie Plus Assay Kit , Thermo Scientific , Rockford IL ) . Samples containing 500 µg of protein were resuspended in 250 µL of incubation buffer supplemented with 1 mM ADP , 10 mM glutamate , 2 . 5 mM malate and 1 mg/mL fatty acid-free bovine serum albumin , as well as dNTPs at the concentrations indicated for each experiment . The radiolabeled nucleotides used were [8-3H]dATP , [8-3H]dGTP , or [5 , 5′-3H]dCTP ( between 12–21 Ci/mmol ) and [α-32P]dATP or [α-32P]dGTP ( 100 Ci/mmol ) , depending on the experiment . In these conditions , mitochondria were incubated at 37°C in a rotary shaker ( in organello reaction ) for 2 h , unless otherwise indicated . Mitochondria were then pelleted ( 13 , 000×g for 1 min ) and washed twice with 10% glycerol , 0 . 15 mM MgCl2 , and 10 mM Tris-HCl ( pH 6 . 8 ) . DNA was extracted from this pellet as described [27] . Briefly , the pellet was lysed with 500 µL of 20 mM Hepes-NaOH , ( pH 7 . 4 ) 75 mM NaCl , 50 mM EDTA , 20 µg/mL proteinase K , and incubated at 4°C for 45 min . Then , 17 µL of 30% lauryl maltoside was added and the mixture was incubated at 50°C for 45 min . Twenty-five µL of the homogenate was used for scintillation counting ( Beckman Coulter LS 6500 , Brea , CA ) to quantify the total label within mitochondria . These data allowed us to rule out that differences in DNA labeling were due to differential amounts of radioactive nucleotide transported into mitochondria . The remaining mitochondrial homogenate was used for phenol-chloroform extraction , and DNA was dissolved in TE buffer ( 1 mM EDTA , 10 mM Tris-HCl; pH 7 . 5 ) and quantified ( Quant-iT PicoGreen dsDNA Reagent , Invitrogen ) . The DNA-incorporated radiolabel was measured by scintillation counting . De novo mtDNA replication was quantified as the apparent fmoles of 3H-labeled nucleotide per ng DNA , calculated from the specific radioactivity of the radiolabeled nucleotide used . In each assay , a reaction with 1 µM of each of the 4 dNTPs was processed . This reference point was considered 100% and the remaining results were expressed as a percentage . When 32P-labeled dATP or dGTP were used , DNA was dissolved in 20 mM Tris-acetate , 50 mM potassium acetate , 10 mM magnesium acetate , and 1 mM dithiothreitol , pH 7 . 9 , and digested with AccI ( New England Biolabs , Ipswich , MA ) . The product was resolved in 1% agarose gel in 40 mM Tris-acetate and 1 mM EDTA , pH 8 . 0 , stained with ethidium bromide to capture the total DNA image , then washed in TE buffer and dried under vacuum . The radiolabeled bands , representing de novo synthesized DNA , were visualized by autoradiography and quantified ( Image J software , NIH ) . Mitochondria ( 500 µg of protein ) were treated with 200 µL of 60% methanol and kept at −20°C for 2 h . The protein was pelleted at 25 , 000×g for 10 min and the supernatant incubated at 100°C for 3 min , dried under speed vacuum and redissolved in 20–100 µL of water , depending on the expected dNTP concentration ( dNTP extract ) . dNTP content was determined using a polymerase-based method [28] with minor modifications . Briefly , 20 µL of reaction mixture contained 5 µL of dNTP extract in 40 mM Tris-HCl , pH 7 . 4 , 10 mM MgCl2 , 5 mM dithiothreitol , 0 . 25 µM oligoprimer , 0 . 75 µM [8-3H]dATP , 12–21 Ci/mmol ( or [methyl-3H]dTTP for the dATP assay ) and 1 . 7 units of Thermo Sequenase DNA Polymerase ( GE Healthcare ) . Reaction mixtures with aqueous dNTP standards were processed in parallel . After incubation at 48°C for 60 min , 18 µL of the mix was spotted on a Whatman DE81 paper and left to dry . The filters were washed 3 times for 10 min with 5% Na2HPO4 , once with water , once with absolute ethanol , and left to dry again . The retained radioactivity was determined by scintillation counting , and dNTP amounts calculated from interpolation on the calibration curves . To ensure the reliability of the results , triplicates of 2 different dilutions of each dNTP extract ( usually undiluted and 1∶3 water-diluted ) were processed in each independent experiment . When dATP was determined in extracts with large amounts of dTTP ( 37 . 5±4 . 5 pmol/mg protein , Table 1 ) , which significantly dilutes the label ( [methyl-3H]dTTP ) used for the assay [28] , the dilution was taken into account and the appropriate correction factor was applied to the results . Human primary skin fibroblasts ( passage 4 or lower ) from 4 healthy controls were seeded in 3 . 5-cm diameter plates ( 100 , 000 cells per plate ) and expanded in Dulbecco's modified Eagle's medium with 4 . 5 g/L glucose , supplemented with 2 mM L-glutamine , 100 U/mL penicillin and streptomycin , and 10% dialyzed FBS ( Invitrogen ) in a humidified incubator at 37°C and 5% CO2 . Four days after tight confluence was reached , FBS was reduced to 0 . 1% . Four days later ( day 0 ) different plates were maintained in the following conditions: 1 ) no added nucleosides , 2 ) 40 µM thymidine , and 3 ) 40 µM thymidine+40 µM deoxycytidine . At day 30 , 2 additional conditions were generated: 4 ) thymidine withdrawal from condition 2 , and 5 ) 40 µM deoxycytidine added to condition 2 . Media was replaced every 3 days to ensure that thymidine and deoxycytidine concentrations were always between 10 and 40 µM , as tested by HPLC . On different days , cells were harvested by trypsinization , washed with phosphate-buffered saline , pelleted and stored at −20°C until DNA isolation . DNA was isolated from cell pellets ( QIAampDNA Mini kit , Qiagen ) , dissolved in TE buffer , and quantified ( NanoDrop Spectrophometer , Thermo Scientific ) . Relative quantitation of mtDNA ( 12S rRNA gene ) versus nuclear DNA ( RNase P single copy gene ) was performed using an ABI PRISM® 7500 real-time PCR system ( Applied Biosystems ) . Custom designed primers and probe for the 12S rRNA gene were used for assessing mtDNA copy number , which was expressed as the ratio over the copy number of the RNase P single copy nuclear gene , as previously described [29] . | Mitochondria are subcellular organelles that constitute the main energy supply within the cell . They contain their own DNA , which should be continuously replicated to ensure the correct mitochondrial function . Several mitochondrial diseases are caused by genetic defects that compromise this replication and result in mitochondrial DNA depletion . In most cases , these genetic defects block the synthesis of dATP , dGTP , dCTP , and dTTP , the 4 nucleotides needed for mitochondrial DNA replication . However , for one of these disorders ( mitochondrial neurogastrointestinal encephalomyopathy , MNGIE ) , the biochemical pathways needed to synthesize them are intact , but degradation of dTTP is genetically blocked , leading to dTTP accumulation . We investigated the biochemical mechanisms through which the dTTP excess leads to mitochondrial DNA depletion in MNGIE , and we found that the delay of mitochondrial DNA replication rate observed when dTTP is in excess is not caused by this excess in itself . Instead , the dTTP overload produces a secondary dCTP depletion that actually delays mitochondrial DNA replication . Therefore , the common factor accounting for mitochondrial DNA depletion in these disorders is the limited availability of one or more nucleotides . This indicates that strategies to provide nucleotides to patients' mitochondria should be explored as a possible treatment for these fatal disorders . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"nucleotides",
"mitochondrial",
"myopathy",
"dna",
"replication",
"dna",
"myoneurogenic",
"gastrointestinal",
"encephalopathy",
"mitochondrial",
"diseases",
"biology",
"biochemistry",
"nucleic",
"acids",
"neurological",
"disorders",
"neurology",
"neuromuscular",
"... | 2011 | Limited dCTP Availability Accounts for Mitochondrial DNA Depletion in Mitochondrial Neurogastrointestinal Encephalomyopathy (MNGIE) |
The protease-resistant prion protein ( PrPres ) of a few natural scrapie isolates identified in sheep , reminiscent of the experimental isolate CH1641 derived from a British natural scrapie case , showed partial molecular similarities to ovine bovine spongiform encephalopathy ( BSE ) . Recent discovery of an atypical form of BSE in cattle , L-type BSE or BASE , suggests that also this form of BSE might have been transmitted to sheep . We studied by Western blot the molecular features of PrPres in four “CH1641-like” natural scrapie isolates after transmission in an ovine transgenic model ( TgOvPrP4 ) , to see if “CH1641-like” isolates might be linked to L-type BSE . We found less diglycosylated PrPres than in classical BSE , but similar glycoform proportions and apparent molecular masses of the usual PrPres form ( PrPres #1 ) to L-type BSE . However , the “CH1641-like” isolates differed from both L-type and classical BSE by an abundant , C-terminally cleaved PrPres product ( PrPres #2 ) specifically recognised by a C-terminal antibody ( SAF84 ) . Differential immunoprecipitation of PrPres #1 and PrPres #2 resulted in enrichment in PrPres #2 , and demonstrated the presence of mono- and diglycosylated PrPres products . PrPres #2 could not be obtained from several experimental scrapie sources ( SSBP1 , 79A , Chandler , C506M3 ) in TgOvPrP4 mice , but was identified in the 87V scrapie strain and , in lower and variable proportions , in 5 of 5 natural scrapie isolates with different molecular features to CH1641 . PrPres #2 identification provides an additional method for the molecular discrimination of prion strains , and demonstrates differences between “CH1641-like” ovine scrapie and bovine L-type BSE transmitted in an ovine transgenic mouse model .
Prion diseases such as Creutzfeldt-Jakob disease ( CJD ) in humans , scrapie in sheep and goats and bovine spongiform encephalopathy ( BSE ) in cattle are tightly associated with the accumulation of an abnormal form of a host-encoded cellular prion protein ( PrP C ) in infected tissues [1] . The biochemical properties of this disease-associated form of the protein ( PrPd ) , which include insolubility in non-denaturing detergents and partial resistance to degradation by proteases , differ from those of the normal form . Whereas the normal protein is fully sensitive to proteases , the abnormal prion protein is only partly degraded ( PrPres ) due to removal of the amino-terminal end . In most cases , a large protease-resistant C-terminal core fragment is identified which has a gel mobility of ∼19–21 kDa in its unglycosylated form . However , in some prion diseases , such as some cases of human Creutzfeldt-Jakob disease [2] or the H-type atypical form of BSE [3] , a much smaller C-terminal PrPres product has also been reported . A typical molecular signature of the BSE agent has been identified by PrPres Western blot analysis , which allows such methods to be used to identify the possible presence of BSE in sheep or goats [4]–[10] . The origin of the BSE agent in cattle is still unknown , and its possible reservoir has not yet been identified . A few isolates of TSEs were described in sheep that showed partial similarities with experimental ovine BSE , with a lower molecular mass of unglycosylated PrPres than in most scrapie cases , as found in ovine BSE . However the very high proportions of diglycosylated PrPres found in ovine BSE were not generally apparent in such isolates . This was first demonstrated in the CH1641 experimental scrapie isolate [11] , [12] , then in a few natural scrapie cases in Great Britain and France [13] , [14] . Bioassays performed in wild-type mice to identify prion strains from TSE isolates were reported to identify the biological signature of the BSE agent [15]–[18] , but the CH1641 source failed to transmit the disease to such mice [11] , [12] . Both CH1641 and “CH1641-like” natural isolates were however transmitted in an ovine transgenic mouse model ( TgOvPrP4 ) , showing similar PrPres molecular features in both transgenic mice and sheep , i . e . a low apparent molecular of unglycosyslated PrPres ( referred as l-type PrPres ) [19] , [20] . In some of the cases this could be not the unique molecular phenotype identified in all scrapie-infected mice , with some of the mice also showing PrPres with a higher apparent molecular mass ( h-type PrPres ) [20] . Deviant phenotypes of BSE have recently been reported in cattle however ( H and L-types , based on the PrPres features in cattle brain ) [21]–[24] . Bioassays in wild-type and transgenic mice showed that these were consistent with the presence of two distinct strains , both differing from the single classical BSE strain involved in the food-borne BSE epidemic [23] , [25]–[27] . Thus , the possible transmission of such forms of BSE in other species such as small ruminants also needs to be considered . Recently the hypothesis has been raised that the classical BSE epidemic might have originated from the recycling of one of these atypical forms of BSE ( L-type BSE ) after a first cross-species transmission , possibly in sheep [27] , [28] . Recent studies of a transmissible mink encephalopathy ( TME ) isolate in TgOvPrP4 mice also showed similar phenotypic features to those of L-type BSE , suggesting a possible cross-species transmission of L-type BSE by oral route [29] . Given their unusual molecular properties , “CH1641-like” or CH1641 isolates might be the result of a transmission of L-type BSE to sheep or might represent similar isolates occurring in sheep . In this study we compared the PrPres molecular features of a series of natural “CH1641-like” and experimental CH1641 scrapie isolates , with those of classical and L-type BSE , after transmission to TgOvPrP4 ovine transgenic mice . We demonstrated the abundance of a C-terminal PrPres fragment ( PrPres #2 ) , which distinguished these ovine scrapie isolates from both bovine classical and L-type BSEs after transmission in a common ovine trangenic mouse model .
We compared the PrPres Western blot profiles , after transmission in TgOvPrP4 ovine transgenic mice , of two recently identified natural sheep TSE isolates ( 05-825 and 06-017 ) that showed PrPres molecular features comparable to the experimental CH1641 scrapie isolate , i . e . , a low apparent molecular mass ( l-type ) close to that found in ovine BSE . When the Bar233 antibody was used to detect the usual form of PrPres ( PrPres #1 ) , the molecular features , i . e . the apparent molecular masses of the three PrPres glycoforms and the glycoforms proportions , were similar in all PrPres positive mice in both experimental groups ( Figure 1A ) . The glycoform proportions in each of the natural 4 “CH1641-like” isolates ( or in CH1641 ) were significantly different from those of classical BSE ( p<0 . 0001 for all 15 tests ) , with essentially lower levels of diglycosylated PrPres than in classical BSE ( Figures 1A and 2A ) . Lane by lane comparisons revealed a slightly lower apparent molecular mass of unglycosylated PrPres in mice infected with scrapie rather than ovine BSE , and also after PNGase deglycosylation ( Figure 1B ) . Nevertheless , these differences ( 0–0 . 3 kDa ) remained within the range of the possible variations of an individual sample in a given Western blot experiment . These molecular features were similar to those found in TgOvPrP4 mice infected with two other previously described “CH1641-like” natural scrapie isolates , at first or second passages ( Figure 1C and 1D ) , in all ( TR316211 isolate ) or in some ( O104 isolate ) of the mice [20] . In contrast , the PrPres #1 molecular features in TgOvPrP4 mice infected with “CH1641-like” isolates did not differ from those found in mice infected with L-type BSE . The low apparent molecular mass of unglycosylated PrPres was similar to that found in mice infected with classical BSE ( Figure 1C ) , and also after PNGase deglycosylation ( Figure 1D ) . Glycoform proportions did not differ significantly between any of the natural “CH1641-like” isolates ( or CH1641 ) and L-type BSE with all cases showing lower levels of diglycosylated PrPres than in classical BSE ( Figures 1C and 2B ) ( p>0 . 30 for all tests except the one comparing the monoglycosylated PrPres band in TR316211-infected mice for which p = 0 . 07 ) . Low apparent molecular masses of PrPres were consistently associated with strongly reduced labeling by P4 monoclonal antibody ( data not shown ) . We then used SAF84 for PrPres detection , that identified an additional band at ∼14 kDa ( PrPres #2 ) in TgOvPrP4 mice infected with the four natural “CH1641-like” isolates and with the experimental CH1641 isolate ( Figure 3A and 3C ) . This was associated with lighter , more diffuse labeling below the well defined ∼19 kDa unglycosylated PrPres #1 band , consistent with the presence of a monoglycosylated form derived from the ∼14 kDa PrPres product . This PrPres #2 fragment was not detected in mice infected with classical BSE transmitted to sheep or with L-type BSE in cattle ( Figure 3A and 3C ) . The existence of two distinct PrPres fragments of ∼19 and ∼14 kDa detected with SAF84 antibody only in the “CH1641-like” isolates was also demonstrated after deglycosylation by PNGase treatment ( Figure 3B and 3D ) . Comparison of PrPres profiles between TgOvPrP4 mice and sheep or cattle ( Figure S1 ) indicate that TgOvPrP4 faithfully reproduced PrPres features of ruminants following transmission of scrapie or BSE , including regarding the presence of PrPres #2 in “CH1641-like” scrapie but not in BSE . The ∼14 kDa band was not detected in TgOvPrP4 mice that had been infected with an isolate from cattle experimentally infected with transmissible mink encephalopathy ( TME ) ( Figure 3 ) , previously demonstrated to have phenotypic features similar to L-type BSE in TgOvPrP4 mice ( PrPres of low apparent molecular mass ) [29] . We analyzed PrPres in TgOvPrP4 mice infected from six different BSE sources including ( i ) 3 natural isolates from cattle , goat , and a cheetah with feline spongiform encephalopathy ( FSE ) and ( ii ) 3 experimental sources from sheep ( homozygous either for A136R154Q71 or A136R154R171 prnp allele ) or from C57Bl/6 wild-type mice . All these BSE sources showed a similar PrPres profile with low apparent molecular mass ( ∼19 kDa ) , close to that found in CH1641-infected mice , but with higher levels of diglycosylated PrPres , after the use of both Bar233 and SAF84 antibodies ( Figures 4A , 4B , and 6 ) . In contrast to the CH1641 source , none of them showed detectable levels of the ∼14 kDa PrPres fragment with SAF84 antibody , even after PNGase deglycosylation ( Figure 4B and 4C ) . However , a ∼14 kDa band was also detected in mice infected with 5 natural scrapie isolates ( Figure 4E ) , otherwise characterized by a higher apparent molecular mass of PrPres #1 compared to CH1641 ( Figure 4D ) or to ovine BSE , although this ∼14 kDa band was clearly less intense than from “CH1641-like” isolates . We then evaluated the presence of the C-terminally cleaved PrPres fragment detected by SAF84 antibody in the experimental sources that had been adapted to TgOvPrP4 mice . ( 1 ) Among the two experimental scrapie isolates , unlike CH1641 , this PrPres fragment was not detected from the SSBP/1 isolate ( Figure 5 ) . ( 2 ) From experimental strains derived from mouse-adapted scrapie or BSE strains , it was only detected in the 87V strain of scrapie which is also characterized by a low apparent molecular mass of the unglycosylated PrPres #1 form , similar to that found in BSE , but not in the three other scrapie strains ( C506M3 , Chandler and 79A ) otherwise characterized by a high molecular mass of unglycosylated PrPres ( Figure 5 ) . In mice infected with natural scrapie isolates and CH641 , the respective proportions of the ∼14 and ∼19 kDa bands , as observed after PNGase deglycosylation , were quantified and the ratios of PrPres #2/PrPres #1 determined ( Figure 6 ) . The mean proportions of PrPres #2 in the “CH1641-like” isolates ( 3–5 mice analysed per experimental group ) represented 22 . 7% to 39 . 3% of the total signal , except for the O104 isolate for which these proportions were smaller ( 12 . 4%–20% ) in 4 of the 5 mice analysed . The proportions of PrPres #2 in mice infected with the other scrapie isolates ( “non CH1641-like” ) ( 1–3 mice analysed per experimental group ) , were below 10% in most mice , except those infected with O111 ( 15 . 9%–19 . 4% ) . Statistical analyses of the data confirmed the significantly higher proportions of PrPres #2 in mice infected with “CH1641-like” isolates ( or CH1641 ) compared with other scrapie isolates ( p<0 . 0001 ) , as well as the significantly higher proportions of PrPres #2 in the O111 isolate within the “non CH1641-like” isolates ( p = 0 . 02 ) . No significant differences in these proportions of PrPres #2 were found between the natural “CH1641-like” isolates and the experimental CH1641 isolate ( p = 0 . 42 ) . Transmission studies of the O104 isolate showed the presence of a mixture of two distinct PrPres phenotypes , with either high ( h-type ) or low ( l-type ) apparent molecular masses of unglycosylated PrPres , in variable proportions in each individual mouse , as shown using Bar233 detection after PNGase treatment ( Figure 7A ) [20] . The l-type PrPres , compared to the h-type , is only faintly labeled by the P4 antibody , but a P4-labelled PrPres sub-population that migrates similarly to the h-type PrPres can be identified in mice with l-type PrPres ( Figure 7B ) . When the SAF84 antibody was used ( Figure 7C ) , the C-terminal PrPres fragment was detected in all the O104 infected mice , but the lowest proportion ( 12 . 4% ) ( Figure 6 ) was found in the sole mouse that showed the most important proportions of h-type PrPres ( lanes 4 in Figure 7 ) . These data are also consistent with a preferential association of PrPres #2 with PrPres #1 of low apparent molecular mass ( l-type ) in this scrapie isolate . Immunoprecipitation experiments were carried out to enrich the PrPres #2 form in the samples and characterize it . Successive rounds of immunoprecipitation on magnetic beads coated with Sha31 N-terminal antibody that only recognizes PrPres #1 , allowed progressive depletion of the PrPres #1 in the samples ( Figure 8 ) . After 7 rounds of immunoprecipitation , PrPres #1 becomes only barely detectable . Immunoprecipitation was then performed using the C-terminal SAF84 antibody that recognizes both PrPres #1 and PrPres #2 . Three bands were detected at ∼22 , 18 , and 14 kDa , showing that PrPres #2 , previously identified as an unglycosylated ∼14 kDa band , is also isolated from the mouse brains in monoglycosylated and diglycosylated forms . The presence of this three band PrPres #2 form was confirmed by differential immunoprecipitation in CH1641 , “CH1641-like” isolates , and 87V but also in the O111 “non-CH1641-like” scrapie isolate . It could not be detected in mice infected with ovine BSE , L-type BSE or cattle experimentally infected with transmissible mink encephalopathy . The neuropathological analyses of the first passage experiments indicated comparable distribution of disease-associated prion protein in the brain of the transgenic mice among the “CH1641-like” sheep scrapie group ( Figure S2 ) . This was particularly clear for the 05-825 and 06-017 isolates that resulted in similar intensity of pathological PrP accumulation ( Figure S2C and S2D ) . Overall , these data were also not dissimilar from those already described for the first passage of L-type BSE [29] . In both “CH1641-like” scrapie and L-type BSE , the florid plaque type of PrPsc deposition reported in this transgenic mouse line infected with classical BSE was never observed . However it is possible to underline some clear distinctive features such as a difference in the cortex targeting that was less intense compared to L-type BSE , even in the most severely affected cases ( 05-825 and 06-017 isolates ) ( Figure S2E ) . Remarkably the types of PrPd deposition were also different; in the “CH1641-like” sheep scrapie group the deposition of pathological PrP was fine granular compared to L-type BSE in which plaque-like deposition were sometimes noticeable . Also , in the mesencephalon ( raphe dorsalis ) , the deposition was intraneuronal for the “CH1641-like” sheep scrapie group but not in the brain of mice infected with L-type BSE . These data thus indicate some differences in the biological features of “CH1641-like” isolates , not only with classical BSE , but also with L-type BSE .
This study describes the molecular analyses of PrPres after transmission into TgOvPrP4 ovine transgenic mice from 4 natural ovine scrapie isolates whose PrPres features in sheep were similar to those previously described for the experimental CH1641 scrapie isolate [12] . Two of these previously unreported isolates ( 05-825 and 06-017 ) behaved as previously described for CH1641 and another natural isolate ( TR316211 ) during the first passage in TgOvPrP4 mice , showing low molecular mass PrPres ( l-type PrPres ) in all mice [19] , [20] . In contrast , all the TgOvPrP4 mice receiving 5 natural scrapie isolates characterized by high PrPres molecular masses ( h-type PrPres ) in the sheep brain , showed PrPres of high molecular mass . Detailed analyses showed , as previously described in the CH1641 isolate in sheep [9] and in TgOvPrP4 mice [19] , a slightly lower PrPres molecular mass in TgOvPrP4 mice from the “CH1641-like” isolates than from ovine BSE , although the resolution of small gels made discrimination difficult . Our results are quite consistent with previous studies of the CH1641 isolate by the immunohistochemical “peptide mapping” method , which revealed that PrPd in the CH1641 isolate was truncated further upstream in the N terminus than from experimental BSE [30] . The biochemical PrPres features of these scrapie isolates differ from BSE mainly in their moderately high proportions of di-glycosylated PrPres ( 50%–60% ) , whereas ovine BSE is characterized by higher proportions of di-glycosylated PrPres [4] , [9] , [12] . Molecular discrimination of strains based on the relative proportions of glycoforms is however less reliable than that of PrPres molecular masses , given the large measurement variations and poor standardization of analytical methods [9] , [10] , [31]–[33] . Furthermore glycoforms proportions of BSE in sheep have only been determined from a very limited number of sources . A recent study of classical BSE in cattle showed large individual variations ( ∼20% ) in the proportions of di-glycosylated PrPres [34] . The question of a possible transmission of BSE in small ruminants now needs to be re-examined considering the recent identification of atypical cases of BSE ( H-type or L-type ) in cattle [21]–[24] . Recent studies have indeed hypothesized that cross-species transmission of such rare atypical cases could be at the origin of the BSE epidemic in cattle [27] , [28] , [35] . The first experimental support for this hypothesis was obtained following the discovery of a BSE-like phenotype in mice following transmission of L-type BSE in wild-type mice ( C57Bl , SJL ) [27] or in an ovine transgenic ( tg338 ) mouse line [28] . However , unlike tg338 , which expressed 8- to 10-fold levels of V136 R154 Q171 ovine PrP , the phenotype of the L-type BSE remained distinct from classical BSE during at least two passages in TgOvPrP4 mice that expressed 2- to 4-fold levels of the A136 R154 Q171 ovine PrP [29] . It is noteworthy that , in cattle , the essential difference between L-type BSE and classical BSE is the slightly lower apparent molecular mass and the lower proportions of diglycosylated PrPres [22]–[24] , reminiscent of the differences between CH1641 and classical BSE experimentally transmitted to sheep [9] , [12] , [30] . The phenotypic features of L-type BSE have not yet been reported in sheep . In this study we showed that the PrPres molecular masses and glycoform proportions between “CH1641-like” scrapie isolates and L-type BSE transmitted into TgOvPrP4 mice were indistinguishable , in addition to survival periods in the same range at second passage . However our study revealed that a highly sensitive C-terminal antibody ( SAF84 ) recognised an abundant PrPres product ( PrPres #2 ) in TgOvPrP4 mice infected with “CH1641-like” isolates , the unglycosylated form of which migrates at ∼14 kDa , in addition to the usual PrPres product ( PrPres #1 ) which migrates at ∼19 kDa in its unglycosylated form . The presence of mono- and di-glycosylated forms derived from this PrPres cleavage product was confirmed by differential immunoprecipitation of PrPres #1 and PrPres #2 . Depletion of PrPres #1 using N-terminal antibodies allowed the samples to be enriched in C-terminally cleaved PrPres #2 , which then appeared in a 3-band pattern between 14 and 22 kDa . Such experiments also confirm that PrPres #2 is only faintly recognized by Sha 31 antibody , which recognizes the 148–155 region of the ovine PrP protein , suggesting that this region is absent from most of the PrPres #2 fragments . PNGase deglycosylation also facilitated the identification of PrPres #2 , and permitted quantification of the respective proportions of PrPres #2 and PrPres #1 . Whereas PrPres #2 was abundant in TgOvPrP4 mice infected with “CH1641-like” isolates , lower levels of PrPres #2 could also be detected from 5 natural isolates with h-type PrPres transmitted into TgOvPrP4 mice . C-terminally cleaved PrPres products have previously been described in sporadic or genetic Creutzfeldt-Jakob disease in humans [2] . Although the presence of low levels of PrPres #2 in BSE and L-type BSE cannot be fully excluded , this PrPres form remained undetected in our experiments with these BSE forms , even after differential immunoprecipitation . This was also the case in classical BSE transmitted in a variety of different species . Interestingly , similar results were obtained in TgOvPrP4 mice infected with an isolate from cattle experimentally infected with transmissible mink encephalopathy ( TME ) , consistent with previous studies showing similarities with L-type BSE [29] . Our results thus reinforce the molecular discrimination of “CH1641-like” scrapie isolates from classical BSE , but also indicate a clear molecular difference with L-type BSE transmitted from cattle to ovine transgenic mice . However , further comparisons including those of biological and histopathological features during serial passages in this mouse model will be required , as well as transmission studies performed from L-type BSE experimentally transmitted to sheep . We have also recently described the identification of a C-terminally cleaved PrPres #2 form in H-type BSE , in cattle and after transmission to C57Bl/6 mice [3] . However , a relationship between “CH1641-like” scrapie isolates and H-type BSE seems unlikely . H-type BSE is indeed characterized by a high PrPres molecular mass comparable to most natural scrapie cases , in contrast to the low PrPres molecular mass , which is the hallmark of “CH1641-like” isolates . Although the transmission of H-type BSE in sheep has not yet been reported , a high PrPres molecular mass was maintained upon transmission in tg338 ovine transgenic mice [26] . Unfortunately , direct comparisons with H-type BSE in TgOvPrP4 mice were not possible since we were unable to transmit the disease from several cattle H-type isolates to these mice , at least at first passage [29] . As these same H-type isolates were transmitted in tg338 expressing higher levels of the V136 R154 Q171 ovine PrP protein [26] , this could suggest a high species and/or strain barrier for H-type BSE in sheep . Conversely , both classical and L-type BSEs were readily transmitted in TgOvPrP4 mice [19] , [29] . The presence of PrPres #2 within the different scrapie sources , was preferentially associated with PrPres #1 of low molecular mass . When several experimental scrapie sources were analysed , PrPres #2 was only detected in the 87V strain , characterized by l-type PrPres , but not in C506M3 , Chandler or 79A strains or in the SSBP/1 isolate with h-type PrPres , still emphasizing the need of further comparisons between 87V and “CH1641-like” isolates [20] . Although PrPres #2 could also be detected after the transmission of natural scrapie isolates with high molecular mass , the levels were consistently lower than in “CH1641-like” isolates . It might be that the presence of low levels of PrPres #2 in scrapie isolates with h-type PrPres indicates a mixture of PrPres phenotypes in these scrapie sources , with the levels of l-type PrPres undetectable . This possibility should be considered in the light of certain observations . ( 1 ) A scrapie case with both h-type and l-type PrPres has recently been described in the UK , each PrPres phenotype originating from two different brain areas [14] . ( 2 ) Our recent transmission studies of two “CH1641-like” isolates ( O100 and O104 ) from the same flock into TgOvPrP4 showed the presence of h-type PrPres in some of the mice suggesting a possible mixture of the two PrPres phenotypes in the initial ovine scrapie isolates; these two PrPres phenotypes might be selected , at least in part , during the second passage in TgOvPrP4 mice [20] . Studies of the initial ovine brain samples by immunohistochemistry indeed revealed the presence of differently cleaved PrPres forms in different brain nuclei [13] . ( 3 ) Transmission of scrapie in cattle from a brain pool ( British source ) with h-type PrPres produced two cows with l-typePrPres [36] . h-type PrPres was detected in a second brain sample from one of the two animals . ( 4 ) Similar results were observed in a bovine transgenic mouse line , the mobility in mice being faster than in the original scrapie isolate ( Irish source ) [37] . All together , these data suggest that l-type PrPres could be present in a number of scrapie sources . The identification of “CH1641-like” isolates might be the fortuitous and rare result of analysing samples in which the l-type PrPres of low molecular mass is more abundant . Further characterization of the biological properties of scrapie sources with l-type PrPres will be required firstly to establish whether these correspond to a single strain of infectious agent or involve a variety of distinct scrapie strains , and secondly to better understand the characteristics of their transmission .
The TSE sheep isolates ( Table 1 ) included the experimental CH1641 scrapie isolate ( kindly provided by N . Hunter , Institute for Animal Health , Edinburgh ) and four natural French “CH1641-like” TSE isolates . Transmission studies and initial data concerning the molecular analyses of CH1641 and of two natural “CH1641-like” isolates ( O104 , TR316211 ) transmitted to TgOvPrP4 ovine transgenic mice , have already been described [19] , [20] . Two other field isolates from A136 R154 Q171 homozygous sheep ( 05-825 and 06-017 ) were now included , that showed a low apparent molecular mass of unglycosylated PrPres ( 0 . 1–0 . 4 kDa lower than in cattle BSE ) , as also described in experimental ovine BSE and reduced PrPres labelling with P4 monoclonal antibody in comparison to most natural scrapie cases that show PrPres of higher molecular mass . Other TSE sources examined in TgOvPrP4 mice included ( i ) 5 natural scrapie isolates identified by clinical surveillance in France , with PrPres of high apparent molecular mass ( “non-CH1641-like” ) ( Table 1 ) ; ( ii ) the SSBP/1 experimental scrapie isolate [19]; ( iii ) experimental scrapie strains , derived from mouse-adapted strains , C506M3 , Chandler , 79A , 87V [19] , [20]; ( iv ) BSE from cattle or obtained after natural transmission ( goat , cheetah ) [6] , [18] , [38] or experimental transmission ( sheep homozygous for the A136R154Q171 or A136R154R171 prnp allele , wild-type mouse ) [19] , [39]–[41]; and ( v ) an experimental bovine isolate of transmissible mink encephalopathy [29] . Breeds and prnp genotypes of sheep and the survival periods observed after transmission in TgOvPrP4 ovine transgenic mice are shown in Table 1 . Four- to six-week-old female TgOvPrP4 ovine transgenic mice [42] were inoculated intra-cerebrally with 10% ( first passage ) or 1% ( second passage ) ( wt/vol ) brain homogenates in 5% glucose in distilled water ( 20 μl per animal ) . The brains were sampled at the terminal stage of the disease or death of the animal due to intercurrent disease or ageing . The guidelines of the French Ethical Committee ( decree 87–848 ) and European Community Directive 86/609/EEC regarding mice were respected . Experiments were performed in the Biohazard prevention area ( A3 ) of the author’s institution with the approval of the Rhône-Alpes Ethical Committee for Animal Experiments . The whole brain of every second mouse was frozen and stored at −80°C before Western Blot analysis . The other brains were fixed in 10% formol-saline solution for histopathological examinations . Post-fixed brain were routinely embedded in paraffin after a 1 hour formic acid ( 98%–100% ) treatment . De-waxed and re-hydrated 5 μm brain sections were then either stained using hematoxylin-eosin in order to study vacuolar lesions or immunostained for PrPsc using SAF84 ( SPI Bio ) and 2G11 ( Pourquier ) monoclonal antibodies with or without an additional step using streptomycin sulfate , following a procedure reported in detail elsewhere [43] . A peroxydase-labeled avidin-biotin complex ( Vectastain Elite ABC , Vector Laboratories ) was used to amplify the signal . Final detection was achieved using a solution of diaminobenzidine intensified with nickel chloride ( Zymed ) , producing black deposits . Finally , slides counterstained with aequous hematoxylin were observed under a microscope coupled to an image analysis workstation ( Morpho Expert software , ExploraNova ) . The lesion profiles were built following referential criteria [44] using a computer-assisted method [45] . PrPres was obtained following concentration by ultra-centrifugation from half of the mouse brains homogenised in glucose 5% in distilled water ( 20% wt/vol ) . A 600 μl volume was made up to 1 . 2 ml in glucose 5% , before incubation with proteinase K ( 10 μg/100 mg brain tissue ) ( Roche ) for 1 h at 37°C . N-lauroyl sarcosyl 30% ( 600 μl; Sigma ) was added . After incubation at room temperature for 15 min , samples were then centrifuged at 100 , 000 rpm for 2 h on a 400 μl 10% sucrose cushion , in a Beckman TL100 ultracentrifuge . Pellets were resuspended and heated for 5 min at 100°C in 50 μl TD4215 denaturing buffer ( SDS 4% , β-mercaptoethanol 2% , glycine 192 mM , Tris 25 mM , sucrose 5% ) . In some experiments , deglycosylation was performed using PNGase F ( kit P07043 , BioLabs ) , as previously described [20] . Differential immunoprecipitation was used to enrich the samples in the C-terminally cleaved form of PrPres ( PrPres #2 ) , by depletion of the usual form of PrPres ( PrPres #1 ) . Superparamagnetic polystyrene beads coated with a monoclonal antibody specific for Fc on all mouse IgG ( Dynabeads® Pan Mouse IgG_DYNAL #110 . 41 ) were used as recommended by the manufacturer . After each step , the beads were recovered using Dynal PMC . 50 μl bead aliquots were washed 3 times in 5 volumes of coating buffer ( PBS with 0 . 1% of BSA ) . Beads ( 50 μl of beads resuspended in 50 μl coating buffer ) collected after the last washing were then coated with IgG mouse monoclonal antibodies Sha 31 or SAF84 ( ascitic fluids; SPI-Bio , France ) for PrPres #1 or PrPres #2 capture , respectively . Sha31 and SAF84 recognise the ovine PrP sequences 148-YEDRYYRE-155 and 167-RPVDQY-172 , respectively . For each cycle of PrPres capture , the sample was incubated with antibody-coated beads for 1 h at room temperature under continuous rotation at 60 rpm . After PrPres ultracentrifugation , the pellets obtained from 2 mg brain tissues were resuspended in 20 μl immunoprecipitation buffer ( phosphate-buffered saline [PBS] at pH 7 . 4 and 0 . 3% of N-lauroyl sarcosyl ) and heated 5 min at 100°C , before addition of a 30 μl suspension of antibody-coated beads . After completing the beads suspension to 1 ml , the sample was enriched in PrPres #2 , by depleting the PrPres #1 in 5 to 7 successive rounds of PrPres #1 capture using Sha31-coated beads . The supernatants collected after each capture cycle were used for the next one . PrPres #2 was then captured by SAF84-coated beads . At the end of each capture cycle , PrPres was removed from the beads by heat denaturation for 5 min at 100°C in 30 μl TD4215 buffer prior to Western blot analyses . Western blot analysis was performed as previously described [20] by 15% SDS-PAGE and electroblotting on nitrocellulose membranes . PrPres was detected with P4 ( 0 . 2 μg/ml ) ( 93-WGQGGSH-99 ovine PrPsequence; R-Biopharm , Germany ) , Bar233 ( 1/5000 ) ( 144-FGNDYEDRYYRE-155 ovine PrP sequence; kindly provided by J . Grassi , C . E . A . -Saclay , France ) , Sha31 ( 1/10 from TeSeE Bio-Rad sheep and goats kit; Bio-Rad , France ) or SAF84 ( SPI-Bio , France ) mouse monoclonal antibodies . Peroxidase-labelled conjugate anti-mouse IgG ( H+L ) ( 1/2500 in PBST; ref 1010-05; Clinisciences , France ) was used to detect P4 , Bar233 , and Sha31 antibodies , whereas SAF84 was used as horseradish peroxidase antibody . Streptavidin ( 5 ng/ml ) ( S5512 ) was added to the conjugate solution . Bound antibodies were then detected by direct capture with the Versa Doc ( Bio-Rad ) analysis system using the ECL chemiluminescent substrate ( Amersham , France ) . Quantitative studies were performed using Quantity One ( Bio-Rad ) software , and the apparent molecular masses were evaluated by comparing the positions of the PrPres bands with a biotinylated marker ( B2787 ) ( Sigma , France ) . The glycoforms proportions of the four “CH1641-like” isolates and the CH1641 isolate were compared with each other and the glycoforms proportions of both classical BSE and L-type BSE were compared with those of each natural “CH1641-like” isolate and with the experimental CH1641 isolate . Comparison of classical BSE alone with each “CH1641-like” isolate and with CH1641 at first passage implies 5 tests for each of the 3 PrPres #1 bands . In view of the high total number of tests ( 19 for each PrPres band ) , paired-sample t tests with Bonferroni adjustment were used to preclude the detection of spurious differences in glycoform proportions . A classical analysis of variance was used for comparisons of PrPres #2 . The statistical analysis was performed with R software ( R version 2–6 . 0 [2007-11-03]: A language and environment for statistical computing . R Foundation for Statistical Computing , Vienna , Austria . ISBN 3-900051-07-0; http://www . R-project . org ) . | The origin of the transmissible agent involved in the food-borne epidemic of bovine spongiform encephalopathy ( BSE ) remains a mystery . It has recently been proposed that this could have been the result of the recycling of an atypical , more probably sporadic , form of BSE ( called bovine amyloidotic spongiform encephalopathy , or L-type BSE ) in an intermediate host , such as sheep . In this study we analyzed the molecular features of the disease-associated protease-resistant prion protein ( PrPres ) found in the brain of transgenic mice overexpressing the ovine prion protein after experimental infection with prions from bovine classical and L-type BSEs or from ovine scrapie . Scrapie cases included rare “CH1641-like” isolates , which share some PrPres molecular features with classical BSE and L-type BSE . Scrapie isolates induced in transgenic mouse brains the production of a C-terminally cleaved form of PrPres , which was particularly abundant from “CH1641-like” cases . In contrast , this C-terminal prion protein product was undetectable in ovine transgenic mice infected with bovine prions from both classical and L-type BSE . These findings add a novel approach for the discrimination of prions that may help to understand their possible changes during cross-species transmissions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/prion",
"diseases"
] | 2008 | A C-Terminal Protease-Resistant Prion Fragment Distinguishes Ovine “CH1641-Like” Scrapie from Bovine Classical and L-Type BSE in Ovine Transgenic Mice |
Visceral leishmaniasis is a neglected parasitic disease with no vaccine available and its pharmacological treatment is reduced to a limited number of unsafe drugs . The scarce readiness of new antileishmanial drugs is even more alarming when relapses appear or the occurrence of hard-to-treat resistant strains is detected . In addition , there is a gap between the initial and late stages of drug development , which greatly delays the selection of leads for subsequent studies . In order to address these issues , we have generated a red-shifted luminescent Leishmania infantum strain that enables long-term monitoring of parasite burden in individual animals with an in vivo limit of detection of 106 intracellular amastigotes 48 h postinfection . For this purpose , we have injected intravenously different infective doses ( 104—5x108 ) of metacyclic parasites in susceptible mouse models and the disease was monitored from initial times to 21 weeks postinfection . The emission of light from the target organs demonstrated the sequential parasite colonization of liver , spleen and bone marrow . When miltefosine was used as proof-of-concept , spleen weight parasite burden and bioluminescence values decreased significantly . In vivo bioimaging using a red-shifted modified Leishmania infantum strain allows the appraisal of acute and chronic stage of infection , being a powerful tool for accelerating drug development against visceral leishmaniasis during both stages and helping to bridge the gap between early discovery process and subsequent drug development .
Leishmaniasis is a complex of neglected parasitic diseases affecting the poorest people in 98 countries , particularly those with weak or non-existent health systems . [1] . There are at least three different forms of clinical presentations; cutaneous , mucocutaneous and visceral leishmaniasis , the latter being fatal if left untreated [2] . Visceral leishmaniasis ( VL ) is estimated to produce 300 . 000 new cases and between 20 . 000–40 . 000 deaths every year . Most of the cases are localized in three geographical regions; South Asia and East Africa where the disease is caused by Leishmania donovani and the transmission is mostly anthroponotic . By its part , in Brazil , where the disease is produced by L . infantum chagasi , the transmission is zoonotic and occurs mainly from infected dogs [3] . Nowadays , therapeutic or prophylactic human vaccines are still lacking , and the cure of patients is based on chemotherapy [4 , 5] . Treatment of VL was mainly based on painful intramuscular injections of pentavalent antimonials , such as sodium stibogluconate ( SSG ) . SSG has been the first-line antileishmanial drug in India , although its clinical efficacy in some areas of North Bihar State has gradually declined , due to the emergence of fully resistant L . donovani strains . SSG is being substituted by liposomal amphotericin B ( AmBisome ) as first-line treatment , despite slow intravenous administration of the drug is needed [6–8] . In East Africa , SSG was the first-line regimen for decades , but due to its toxicity and following WHO recommendations in 2010 , SSG + paromomycin combination therapy became the treatment of choice [9] . However , the administration of this drug combination is painful and requires patient hospitalization , and therefore , more friendly alternatives were implemented . These include single dose of AmBisome plus 10 consecutive days of SSG , single dose of AmBisome plus 10 days of miltefosine or miltefosine alone for 28 days . However , none of these combinations improved the results of the treatment of choice in Phase II clinical trials [10] . Miltefosine is the last drug successfully introduced against VL . It is also the only drug that has a good oral bioavailability . However , an increase in relapse rates has been reported in India and Nepal , probably associated with low drug exposure [11 , 12] . In addition , miltefosine is potentially embryotoxic and fetotoxic in experimental animals and thereby , its administration is not recommended in women during pregnancy [13] . For all these reasons , there is an unmet need to fill the antileishmanial drug discovery pipeline with safer drugs that display new mechanisms of action , likely allowing combination therapy in order to prevent the emergence of resistant strains [14] . During this process , and once compounds have shown high in vitro potency , selectivity , specificity , low toxicity and good predictable pharmacokinetic/pharmacodynamic properties , a proof of concept that undoubtedly shows the in vivo efficacy of lead compounds , is required . Both mice and hamsters are used as models of acute and chronic VL , respectively , during the evaluation of the proof of concept [15 , 16] . The most frequent technique to evaluate the infection course after drug treatment has been microscopic counting of amastigotes in liver , spleen and bone marrow smears stained with Giemsa dye . However , these are labour-intensive techniques that require specific skill training and present low sensitivity when parasite burdens are low after treatment [17] , therefore , they are limited by tissue sampling biases , which require large animal cohorts . In vivo real-time imaging combined with modified parasites expressing bioluminescent or fluorescent reporters may accelerate the initial stage of drug discovery at the preclinical level . Fluorescent reporters in the near infrared wavelength avoid interference with haemoglobin and do not require the addition of substrate . However , and despite the number of near infrared proteins currently available [18 , 19] further reporters with longer emission wavelengths are still required in order to increase sensitivity . In this regard , red-shifted bioluminescent reporters [20] are currently allowing the appraisal of different infections produced by Trypanosomatids in vivo in real-time without the need to kill animals . This tool allows to run longitudinal studies with a reduced number of animals since they are not sacrificed , and in addition each animal is its own control , therefore the variability of experimental outcomes is limited [21–25] . In summary , in vivo real-time imaging allows to develop the proof of concept in a record time , accelerating the drug discovery process . Nowadays , the mouse is used as acute preclinical model of VL , being liver the main affected organ when experimental treatments are initiated at early times postinfection . On the contrary , hamster is a more stringent and relevant model to recreate human VL [26 , 27] . Generally speaking , during chronic infections the persistence of pathogens yields a state of T cell dysfunction known as exhaustion that is characterized by the loss of effector functions , low recall response and suboptimal T cell proliferation [28] . This is a hallmark feature shared by mice , dogs and humans and it is associated with disease progression [29–31] . Here , we describe a chronic murine model of VL that combines in vivo real-time image with stably modified strain expressing red-shifted luciferase ( luc ) aiming to track the presence of parasites in target organs during a long-time course of infection that can be used for preclinical drug-discovery . Miltefosine was used as proof of concept to assess the suitability of this technique during drug discovery .
The animal research described in this manuscript complies with Spanish Act ( RD 53/2013 ) and European Union Legislation ( 2010/63/UE ) . The protocols were approved by the Animal Care Committee of the Centro de Biología Molecular Severo Ochoa ( CBMSO , Madrid , Spain ) , project licence number JMJ/bb . Animals were maintained under specific pathogen-free conditions in individually ventilated cages . They experienced a 12 h light/dark cycle and had access to food and water ad libitum . Seven to eight weeks-old female Balb/c mice were obtained from Janvier Labs ( St Berthevin Cedex , France ) and housed in specific pathogen-free facilities in the P2-facility of CBMSO for this study . L . infantum ( strain MCAN/ES/96/BCN 150 ) promastigotes ( previously obtained from infected dogs ) were a gift from J . M . Requena ( CBMSO , Madrid , Spain ) . Parasites were routinely cultured at 26 °C in M199 medium supplemented with 25 mM HEPES pH 6 . 9 , 10 mM glutamine , 7 . 6 mM hemin , 0 . 1 mM adenosine , 0 . 01 mM folic acid , 1x RPMI 1640 vitamin mix ( Sigma-Aldrich ) , 10% ( v/v ) heat-inactivated fetal calf serum ( FCS ) and antibiotic cocktail ( 50 U/ml penicillin , 50 μg/ml streptomycin ) . The 1647-bp PpyRE9h coding region was amplified by PCR from pGEX-6P-2 HCO RE9h vector , a kind gift from Dr . Bruce Branchini , ( Department of Chemistry , Connecticut College , CT , USA ) . The oligonucleotides used as primers ( RBF919 and RBF920 in Table 1 ) introduced NcoI-NotI as restriction sites for cloning into pLEXSY-PAC vector ( Jena Bioscience ) and XhoI restriction site in the forward primer for cloning into pSK II vector . The 1647-bp PCR amplified fragment containing the PpyRE9h coding region was digested with XhoI and NotI and cloned first into pSK II vector previously cut with the same restriction enzymes to yield pSK-PpyRE9h plasmid . Then , this plasmid was cut with NcoI-NotI and the PpyRE9h ORF was cloned into pLEXSY-PAC vector to yield the pLEXSY-PAC-PpyRE9h construct . Parasites expressing red-shifted luc were obtained after electroporation of L . infantum BCN150 promastigotes with the linear SwaI-targeting fragment obtained from pLEXSY-PAC-PpyRE9h vector . Transfections were performed by electroporation ( Gene Pulser X cell System , Biorad ) using 10 μg of DNA fragments under the following conditions: 25 μF , 1500 v , 9 ms in 4 mm gap cuvettes . Subsequent plating on semisolid media containing 200 μg/mL puromycin as selection antibiotic , allowed the isolation of individual colonies that were subcultured in liquid media under antibiotic pressure . The correct integration of each fragment into the 18S rRNA locus of the resulting clones ( PpyRE9h+L . infantum ) was confirmed by PCR amplification analysis , using appropriate primers ( Table 1 ) . The Luciferase Assay System ( Promega ) was used to assess luc expression in PAC-isolated clones from semisolid in vitro cultures . Briefly , 100×106 parasites were washed off with PBS and then lysed with 1 ml of Cell Culture Lysis Reagent provided by the manufacturer . Cell lysate was serially diluted in 96-well plate and then , ten microlitres of each cell lysate dilution were mixed with 90 μL of luciferin substrate . Luminescence was measured immediately using a Synergy HT microplate reader ( BioTek ) . In order to recover the infectivity of PpyRE9h+L . infantum strain after cloning , 108 metacyclic promastigotes were inoculated intravenously ( IV ) in the tail vein . Mice were sacrificed 4 weeks later and spleens were used to recover infective amastigotes . Amastigotes were isolated from the spleen by passing the tissue through a wire mesh . Then , splenocytes were disrupted by passing sequentially through 27G1/2 and 30G1/2 needles . Finally , cell debris was retained by passing successively through polycarbonate membrane filters with pore sizes of 8 μm , 5 μm and 3 μm ( Isopore , Millipore ) . Released amastigotes ( free of host cells ) , were washed twice with PBS ( 4000 x g for 20 min at 4°C ) and counted by direct microscopy . To assess luciferase expression in intracellular infections , phorbol 12-myristate 13-acetate ( PMA ) -differentiated THP1 human monocytic leukemia cells , were grown on 8-well chambered coverslips ( IBIDI ) and incubated with infective amastigotes recuperated from mice in a ratio of 1:10 for further 4 hours . Extracellular parasites were removed by extensive washing with warm PBS and processed for immunofluorescence analysis 5 days after infection . Briefly , coverslips were fixed with 2% ( v/v ) paraformaldehyde in PBS and incubated with 0 . 1% ( v/v ) Triton X-100 in PBS for 10 min at room temperature in order to permeabilize the cells . Slides were then probed with 1:1000 red firefly luciferase polyclonal antibody ( ThermoFisher ) ; followed by 1:500 DyLight 633-conjugated goat anti-rabbit IgG secondary antibody for 30 min at room temperature . DNA was labelled using 1 μM Hoechst 33342 before mounting with Vectashield mounting medium . Images were acquired on a Zeiss LSM800 microscope with airyscan at 60X magnification . Different infective doses of stationary phase PpyREh9+L . infantum promastigotes ( ranging from 5x104 to 5x108 ) were IV injected to 6–8 weeks-old female Balb/c mice . Every week animals were placed in a Charge-Coupled Device ( CCD ) IVIS 100 Xenogen system ( Caliper Life Science ) for BLI analysis and images were acquired 10–20 min after intraperitoneal D-luciferin injection ( 150 mg/kg ) . Briefly , the animals were lightly anesthetized with 2 . 5% isofluorane ( then reduced to 1 . 5% ) , before being placed on the camera . To standardize image capture and in order to allow comparison between mice , the images presented in the figures correspond to an acquisition time of 1 min duration , taken once luminescence plateaued . To estimate the parasite burden in living mice , Regions Of Interest ( ROIs ) around liver and spleen in ventral and lateral animals’ positions were drawn using Living Image v . 4 . 3 to quantify BLI expressed as radiance ( p/s/cm2/sr ) . The detection threshold for in vivo imaging was estimated using uninfected mice placed in different positions ( ventral and lateral ) , using ROIs of whole animals ( n = 16 ) . At the end of some experiments , animals were euthanized and dissected , to confirm parasite burden by more conventional methods . Briefly , spleens and livers were reimaged ex vivo and used to quantify parasite load by Limit Dilution Assay ( LDA ) . LDA was calculated as the geometric mean of the titer obtained from quadruplicate cultures x reciprocal fraction of the homogenized organ added to the first well . The titer was the reciprocal value of the last dilution in which parasites were observed [32] . Individual animal values were used as the unit of analysis of in vivo and ex vivo experiments . Statistical differences between groups were evaluated using t-student test using SigmaPlot v . 14 . 0 . Differences of P < 0 . 05 were considered significant .
PpyRE9h was stably integrated into the 18S rRNA promoter using pLEXSY vector ( Fig 1A ) . Correct integration of reporter gene into the resulting clones ( PpyRE9h+L . infantum ) was confirmed by PCR amplification analyses using the primers of Table 1 ( Fig 1B ) . The PCR-confirmed clones were screened for luciferase activity and those having higher activity were selected for in vivo experiments . Promastigote cultures of PpyRE9h+L . infantum grew at the same rate as wild-type parasites ( Fig 1C ) . There was a linear relationship between the luciferase activity in vitro and the number of parasites independently of the parasite stage ( logaritmic , metacyclic or freshly isolated amastigotes ) and independently of the instrument for measuring ( luminometer or IVIS camera ) . Fig 1D shows the relationship between luciferase activity and the number of logaritmic promastigotes ( PpyRE9h+L . infantum and wild-type strains ) . Cell lysates from promastigotes were serially diluted into 96-well plate , D-luciferin was added and luciferase activity was measured using a luminometer ( 103–106 cells range , r2 = 0 . 999 ) , being the detection limit of 103 promastigotes/well . The PCR-clone with the highest luciferase activity was selected for recovering infectivity through mouse ( see Materials and Methods ) . Fig 1E shows the outcomes using metacyclic promastigotes and free amastigotes in the IVIS camera . In this experiment 6x106 parasites ( metacyclic and amastigotes ) were serially diluted in a 96-well plate , 100 μl D-luciferin were added and luciferase activity was measured in the IVIS camera ( 4 . 68x104-6x106 cells range , r2 = 0 . 987 for metacyclic and r2 = 0 . 994 for amastigotes ) . The detection limit was 9 . 37x104 metacyclic/well and 7 . 5x105 amastigotes/well . This suggests that more amastigotes than metacyclic promastigotes are required for their detection by the IVIS camera . Spleen-isolated amastigotes were used to infect PMA differentiated THP-1 macrophages . Four hours later the non-phagocytosed parasites were gently washed off with warm PBS and left for further 96 h . The infection was stained using anti-firefly luciferase antibody and its localization was confirmed to be cytosolic by confocal microscopy ( Fig 1F ) . The infectivity of the selected clone was enhanced by passing through Balb/c mice that were successively infected with 108 promastigotes by IV route until spleen weight increased up to 0 . 7–1 g ( 4–5 passages through mice ) . To establish the in vivo sensitivity of the PpyRE9h+L . infantum strain , Balb/c mice were IV injected with different doses of infective metacyclic promastigotes ( 5x104-5x106 ) and photographed 1 h postinfection . At this time , the bioluminescent signal was detected in the liver but only with the highest doses ( 5x105 and 5x106 ) . Forty-eight hours later when most promastigotes have transformed into amastigotes; BLI signal was only detected from mice infected with 5x106 parasites ( Fig 2A ) . The appraisal of the infection showed that BLI signal in the liver peaked ~3 weeks post-infection ( acute phase ) , then disappeared slowly from this organ and increased in the spleen ( chronic phase ) ( Fig 2A ) . To estimate the in vivo limit of detection with PpyRE9h+L . infantum parasites , we used BLI signal expressed as radiance ( p/s/cm2/sr ) . The limit of detection in vivo was estimated to be above 1x106 parasites at 48 h postinfection , when metacyclic parasites have transformed into intracellular amastigotes ( Fig 2B ) . We were interested in developing a chronic model of infection to use as proof of concept for well-established infections in spleen and bone marrow . In order to evaluate the stability of bioluminescent signal through time in chronic infections , 5x108 metacyclic promastigotes were IV injected and animals were photographed in ventral and lateral positions from 5 to 16 weeks post-infection . The spleen infection was detected independently of the animal positions . BLI signal was increasing from week 5 to reach the maximum radiance 12 weeks after infection ( Fig 2C ) . Moreover , bone marrow radiance was detected only in ventral images from 8 to 16 weeks , and the bioluminescent signal was increasing during this time ( Fig 2C ) . To establish a correlation between BLI signal detected in vivo and the parasite burden in liver and spleen , mice ( n = 12 , one animal died before the end of the experiment ) were infected with parasite dose ranging from 5x106-5x108 . Nine weeks postinfection , animals were imaged and the luminescence was recorded in vivo in the regions of interest ( ROI ) previously drawn around the spleen and liver . Animals were euthanized and the liver and spleen processed to determine parasite burden . Both organs showed a good correlation with the in vivo recorded BLI signal ( Fig 2D and 2E ) . PpyRE9h+L . infantum infected mice were treated with miltefosine as a proof-of-concept to validate this model in a long-term follow-up infection . Mice ( n = 30 ) were IV infected with 5x108 metacyclic promastigotes and imaged for BLI after 3 , 7 , 12 and 14 wpi confirming that infection was established ( Fig 3A ) . Animals were divided in groups and half of them were treated with miltefosine 40 mg/kg/day for 5 days by oral gavage . Once miltefosine treatment was ended , animals were imaged and sacrificed at different times ( 48 h , 1 and 6 weeks post-treatment that corresponded to 15 , 16 and 21 wpi ) . Fig 3B ( top panel ) shows that BLI signal in whole animals was almost undetectable after miltefosine treatment ( 48 h post-treatment ) in a chronic infection , and that the BLI reduction persisted for 6 weeks after the end of the treatment . Quantification of the BLI signal revealed that radiance in untreated animals ( 3x105 p/s/cm2/sr ) decreased significantly to 2 . 32x104 and 1 . 61x103; one and six weeks , respectively after the end of the treatment , reaching BLI values similar to non-infected animals ( Fig 3B; bottom panel P<0 . 05 ) . Animals were sacrificed at different times after the end of treatment ( 48h , 1 week and 6 weeks posttreatment ) and the organs ( spleen and liver ) were photographed after injecting D-luciferin ex vivo ( Fig 3C ) . There was a significant marked reduction in the weight of the spleen , which reached values similar to those of the uninfected animals at 6 weeks after the end of the treatment ( Fig 3D P<0 . 001 ) . Ex vivo bioluminescent values recorded from treated and untreated animals over the time were plotted ( Fig 3E ) confirming the BLI reduction seen in vivo . The BLI decrease was significant at all analysed times in both organs ( P<0 . 001 ) with the exception of the liver at 48h posttreatment that was not significant and liver at 6wpi ( P<0 . 01 ) . Both organs showed logarithmic reduction of BLI from the end of the treatment to 6 weeks later . Ex vivo parasite burden was estimated using limiting dilution assay confirming parasite load reductions of 98% , 99 , 9% , and 99 , 9% , at 48h , 1 week and 6 weeks postinfection ( Fig 3F ) .
The introduction of new medicines against VL from the initial concept to public release is a time-consuming and expensive process . Moreover , the clinical recurrences after treatment failure and the emergence of resistances are worsened by the shortage of new clinical entities and the long period needed to release a new medicine [33] . To bridge the gap between early drug identification and in vivo preclinical studies , new bioimaging tools have recently been introduced to accelerate the drug discovery process while drastically reducing the number of animals used . To develop robust preclinical in vivo platforms , several aspects related to the genetic modifications introduced in the pathogen and the suitability of the animal model should be addressed before their validation with a proof of concept [34 , 35] . In such a way , we present here the generation of the strain PpyRE9h+L . infantum and its utility to quantify the parasite load in vivo in infected mice in real time . As the virulence of the modified strain can be lost after genetic manipulation and passage in culture , as soon as the correct integration of the construct was confirmed , the selected clone was passed through mice to recover its infectivity [36 , 37 , 38] . Once D-luciferin was administered , the light detected by CCD camera and transformed to pseudocolor images , enabled parasite traceability in the whole body and the estimation of parasite burden in a murine model of chronic VL , reducing the number of animals to be analysed in longitudinal studies . During in vivo infections amastigotes enter into a semi-quiescent physiological stage in which major energetic processes are specifically repressed [39] , explaining the differences in light emission between metacyclic promastigotes and freshly isolated amastigotes . However , our results show that parasites emitted light enough to provide accurate and rapid radiance that allow the appraisal up to 21 wpi . In this study , light could be detected in Balb/c mice in the liver during the acute phase of infection and later in the spleen and bone marrow during the chronic phase , allowing a continuous and long-term follow-up of the infection . Under these conditions , light detected in vivo–that corresponded to ROIs drawn around liver and spleen—correlated well with parasite burden calculated from LDA , which it would allow to estimate the parasite burden without the sacrifice of animals . The location of parasites ( peripheral or deeper tissues ) within the mammalian host has been pointed as a key factor affecting the limit of parasite detection in vivo [40] . The light emitted by freshly isolated amastigotes from splenic lesions in our system showed a detection threshold similar to the previously reported by other authors [22] . In experimental VL , the hamster is considered the best experimental model since it reproduces many clinicopathological features of the human disease and can be fatal in the absence of treatment [41] . High-dose murine models of VL develop hallmarks of progressive human , primate , and canine disease with loss of gp38 stromal cells [42] , remodelling of splenic marginal zone region [43] , altered migration of DCs [42] and loss of follicular germinal centers [44] . For this reason , Balb/c mice have been proposed as an adequate model of chronic VL [45–46] . In addition , during chronic infections the persistence of pathogens yields a state of T cell dysfunction known as exhaustion that is characterized by the loss of effector functions , low recall response and suboptimal T cell proliferation [28] . In VL , this stage of T cell exhaustion is associated with disease progression in mouse , dogs and human infections [29–31] . For this reason and in order to have a murine BLI model of chronic VL , the inoculum size was increased to 5x108 metacyclic parasites per mouse . In previous studies we have used the same L . infantum/Balb/c model showing hallmarks of progressive infection [47] . PpyRE9h+L . infantum strain allowed a continuous monitoring of parasite load from the beginning of the infection up to animal’s sacrifice , detecting both acute and chronic infections . In view of these results , PpyRE9h+L . infantum constitutes an ideal tool for the appraisal of drug efficacy in in vivo preclinical models . The assay was validated by the treatment with miltefosine , starting 14 weeks post-infection and extended for long-term appraisal ( 6 weeks after drug withdrawal ) . In rodents miltefosine is known to produce significant parasite burden reduction ( 90–99% depending on parasite strain ) in liver and spleen , along with no-sterile cure ( when the treatment is initiated at 7–21 days postinfection [48–50] . These no-sterile curative results were confirmed later in hamster models of chronic VL treated with high-dose miltefosine ( 20 mg/kg/10 days ) , started 40 days post infection , although it resulted in 100% survival measured 20 wpi [23] . In our study , both radiance and parasite burden values dropped immediately after the end of treatment and they remained decreasing during the long-term follow up , although sterile cure was never achieved . A possible disease recurrence due to incomplete parasite suppression was expected . However , and despite the long-term follow up , much beyond the half-life of miltefosine [51] , no recurrence was seen . Several studies with AmBisome and stibogluconate in mice have shown that the infection status influenced treatment outcome , so that treatments were less effective in the chronic infection model than in acute infection models [52 , 53] . The changes that occur in liver and spleen structure and function during early and late stages on infection might be the cause [46] . The mouse model proposed in this work would provide accurate information about potential drugs and their efficacy on the later stages of infection when it has been described that the efficacy of several drugs might be more compromised . In conclusion , the gold standard methods used to evaluate the efficacy of antileishmanial drugs based on LDA or microscopic examination are laborious and time-consuming , have intrinsic variability , require intensive use of animals and cannot be monitored in real time . However , novel in vivo bioimaging models based on bioluminescent L . infantum parasites are highly sensitive , easily traceable , and yet provide statistically valuable outcomes with the use of far fewer animals than traditional methods . This technology is reproducible; less expensive because it reduces the number of animals needed , it is barely distressing for animals and can be easily adapted to different experimental models being thereby suitable to accelerate drug development . | Visceral leishmaniasis is a neglected disease that poses a significant threat to impoverished human populations of low-income countries . Due to the unavailability of vaccines , pharmacological treatment is the only approach to control the disease that otherwise can be lethal . To date , drug management in endemic regions is based on combinations of a handful of mostly unsafe drugs , where the emergence of resistant strains is an additional problem . To accelerate the discovery of new drug entities , several gaps from the early discovery of a compound to its public use , should be filled . One of these gaps is the need of a rapid go/no-go testing system for compounds based on robust preclinical models . Here , we propose a new long-term model of murine visceral leishmaniasis using in vivo bioluminescent imaging . For this purpose , a red-shifted bioluminescent Leishmania infantum strain was engineered . This strain has allowed the appraisal of the disease in individual animals and the monitoring of parasite colonization in liver , spleen and bone marrow . As proof of concept of this platform , mice were infected with the transgenic L . infantum strain treated with a standard schedule of miltefosine , the only oral drug available against Leishmania parasites . Bioluminescence and parasite load in the target organs were compared showing a good correlation . Our findings provide a robust and reproducible tool for drug discovery in a chronic model of murine visceral leishmaniasis . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"liver",
"immune",
"physiology",
"spleen",
"microbiology",
"parasitic",
"diseases",
"protozoan",
"life",
"cycles",
"parasitic",
"protozoans",
"animal",
"models",
"developmental",
"biology",
"model",
"organisms",
"protozoans",
"le... | 2019 | A chronic bioluminescent model of experimental visceral leishmaniasis for accelerating drug discovery |
Diverse plasticity mechanisms are orchestrated to shape the spatiotemporal dynamics underlying brain functions . However , why these plasticity rules emerge and how their dynamics interact with neural activity to give rise to complex neural circuit dynamics remains largely unknown . Here we show that both Hebbian and homeostatic plasticity rules emerge from a functional perspective of neuronal dynamics whereby each neuron learns to encode its own activity in the population activity , so that the activity of the presynaptic neuron can be decoded from the activity of its postsynaptic neurons . We explain how a range of experimentally observed plasticity phenomena with widely separated time scales emerge from learning this encoding function , including STDP and its frequency dependence , and metaplasticity . We show that when implemented in neural circuits , these plasticity rules naturally give rise to essential neural response properties , including variable neural dynamics with balanced excitation and inhibition , and approximately log-normal distributions of synaptic strengths , while simultaneously encoding a complex real-world visual stimulus . These findings establish a novel function-based account of diverse plasticity mechanisms , providing a unifying framework relating plasticity , dynamics and neural computation .
Synaptic plasticity mechanisms shape the spatiotemporal dynamics of neural circuits that implement brain functions [1–4] . These plasticity mechanisms are usually divided into Hebbian and non-Hebbian mechanisms , where Hebbian changes in synaptic strength depend on both presynaptic and postsynaptic activity , but non-Hebbian mechanisms do not [5] . Among the most studied details of plasticity are its temporal characteristics , including dependencies on precise timings of neuronal activities in spike timing dependent plasticity ( STDP ) , for example [6 , 7] . Plasticity mechanisms operate over a diverse range of timescales , where experimentally Hebbian plasticity mechanisms are typically observed to be fast , and non-Hebbian , homeostatic mechanisms are much slower [5 , 8 , 9] . This is despite theoretical arguments suggesting fast acting homeostatic mechanisms are required to prevent runaway changes to synaptic strengths which would lead to either very high firing rates , or all activity dying out [8 , 10 , 11] . Further complicating this are observations of metaplasticity , whereby plasticity is modulated by past activity [12–14] . The interplay of this diverse range of plasticity mechanisms with network dynamics gives rise to complex spatial and temporal neuronal activity in which neurons fire in a very irregular way with variable spike timing and fluctuating firing rates [8 , 15–18] . Experiments indicate that heterogeneous , approximately log-normal distributions of synaptic strengths form the network of physical connections underlying these dynamics [19 , 20] . Often this distribution of connection strengths and spiking activity is arranged in a tight balance whereby excitatory inputs are closely tracked by inhibitory ones [21–23] . However , the fundamental questions of why such diverse plasticity phenomena emerge , and how their dynamics interact with neural circuits to give rise to these complex neural dynamics remain unclear [3 , 5 , 22 , 24] . Some studies have developed models of plasticity by fitting to these detailed experimental results , for example the spike triplet based models [25 , 26] . However , these types of models do not explain why the observed properties of plasticity are required for brain function by deriving them from more general functional principles . Further , these plasticity models are often combined ad-hoc to tune the dynamics in neural circuits . The resulting dynamics are often interpreted as having a potential functional role , for example in memory or decision making , but lack a convincing theoretical basis that would be provided by a deeper connection between the plasticity rules and more general functional principles [23 , 26–30] . To deepen our understanding of synaptic plasticity it is important to unravel the links between these mechanisms and account for them in a unified way . Plasticity mechanisms give neural circuits the ability to reorganise themselves for learning . In theoretical models and applications , effective learning in neural networks requires that plasticity mechanisms are carefully orchestrated by an overarching learning algorithm that is linked to performing a required task [31] . As whole entities , brains perform unsupervised learning in the sense that they do not learn with the aid of an external agent directly supervising the activity of certain target neurons . Instead , brains must learn to produce useful outputs using information obtained from sensory input . Understanding how to perform useful unsupervised learning is an important problem that is subject to ongoing investigation [32] . In machine learning , autoencoders perform unsupervised learning by learning an encoder that maps an input to a representation encoded in a layer of neurons , and a decoder that reconstructs the input from these neurons [31] . However , the usefulness of an autoencoder is not in performing this reconstruction , rather , it is the ability to extract useful features of the input for performing other tasks . Here we show that diverse plasticity phenomena can emerge from a fundamental perspective of learning the function of a neural circuit . Specifically , we consider unsupervised learning in which each neuron learns to encode its own activity in the population activity so that the activity of the presynaptic neuron can be decoded from the activity of the postsynaptic neurons . In doing so , we extend the concept of an autoencoder in machine learning to recurrent networks [31] . Both Hebbian and non-Hebbian plasticity rules are required to account for the spiking statistics underlying this function and we derive a precise relationship between these rules . We then demonstrate that a great variety of experimental observations emerge from this function based model . For synaptic plasticity mechanisms , the model contains the classic exponentially decaying timing dependence of STDP , as well as the frequency dependence of both STDP [33] and non-timing based stimulation experiments [34 , 35] . The derived relation between Hebbian and non-Hebbian plasticity can be interpreted as a form of heterosynaptic metaplasticity . This relationship fluctuates over a range of timescales and resolves the paradox in which only slow homeostatic plasticity has been observed experimentally , but fast acting homeostatic plasticity is required by theoretical arguments to stabilise fast , unstable Hebbian plasticity [10 , 11] . Our plasticity rules predict that classical STDP is a special case of more fundamental plasticity rules , and predict a spike timing dependence of non-Hebbian plasticity . We further illustrate that when our plasticity rules are incorporated into a biologically realistic neural circuit of spiking neurons that are stimulated by a real-world visual stimulus , balanced excitation and inhibition is self organised and related to the stability of the neural dynamics . We show that approximately log-normal distributions of synaptic strengths emerge and are organised into receptive fields , all while the circuit learns to encode the complex real-world visual stimulus . These results thus present a novel , unified account of plasticity , dynamics and neural computation .
Typical approaches to learning in artificial neural networks define a global error function for a chosen set of ‘output’ neurons and compute a gradient of this error with respect to connection strengths for use as plasticity rules [31] . However , in the brain it is not clear what this set of output neurons might be and how they would all communicate their error globally [36] . Instead , we begin by considering the function of neural circuits , focussing on local properties of individual neurons and their spiking activity using the concept of an autoencoder in machine learning [31] . In classical autoencoders , connections only exist between layers . However , neural circuits in brains are highly recurrent , so we consider autoencoding in a recurrent network where the activity of neurons is encoded and decoded across time , instead of distinct layers . This is a continuous process in time but can be visualised discretely by unfolding the recurrent network in time so that each ‘layer’ is the state of the neurons at successive timepoints , as in Fig 1a . We use tied weights meaning that encoding and decoding use the same set of connections . Specifically , each neuron learns to encode and decode its own activity into and from a period of the population activity of its postsynaptic neurons , A ( t ) , as in Fig 1b . The encoding of presynaptic neuron j’s spikes into postsynaptic activity , A ( t ) , is familiar and is described by the neuronal and synaptic dynamics . The non-obvious part is how to describe the decoding of presynaptic neuron j’s activity from its postsynaptic activity , A ( t ) , and ensuring that it approximates the actual activity of neuron j . We define r ^ j ( A ( t ) ) to be the decoded estimate of presynaptic neuron j’s actual instantaneous firing rate , rj ( A ( t ) ) , from the postsynaptic activity A ( t ) . In a network operating in discrete Δt timesteps ( as is the case for simulations ) we can relate this to the conditional probability of presynaptic neuron j spiking in conjunction with the pattern of postsynaptic activity A that happens to occur during the time window from t to t + τA , that is Δ t r j ( A ) = p ( j spiking | A ) . We want learning to ensure that Δ t r ^ j ( A ) ≈ p ( j spiking | A ) . Although we have not yet defined the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , it should be encoded in the network using local properties of the network that neuron j has access to , so that neuron j has the ability to directly encode and decode its own activity . The purpose of learning is to modify the network so that the encoding and decoding processes are able to ( approximately ) invert each other , since in any arbitrary configuration the decoding of presynaptic neuron j’s activity from its postsynaptic activity , A ( t ) , will likely be poor . We continue by applying learning rules directly to the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , not the properties of the network used to encode it . Later , we specify how the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , is decoded from the connectivity , W , and the postsynaptic activity A ( t ) . We use this to translate the learning rules that operate directly on the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , into plasticity rules for modifying connection strengths , W , and demonstrate learning in a spiking neural circuit with a real-world , complex visual stimulus . We do not model the detailed physiology of the synapses represented by W and the mechanisms by which they change , but it should be noted that the resulting neural dynamics can depend on these lower level details , for example , if synaptic changes are presynaptically or postsynaptically expressed [37] . Given a long period of activity in a network , there are two events that need to be counted to estimate the instantaneous presynaptic firing rate , rj ( A ( t ) ) . Firstly , the number of times that postsynaptic activity , A ( t ) , occurs , and secondly , the number of times that presynaptic neuron j spikes ( within an arbitrarily short Δt ) in conjunction with postsynaptic activity , A ( t ) . This second event is a Hebbian event as it consists of both postsynaptic activity , A ( t ) , and a presynaptic j spike , while the first event is non-Hebbian since it only consists of postsynaptic activity , A ( t ) [4 , 5] . Thus , both Hebbian and non-Hebbian learning are required to account for the statistics of the learned activity . We apply Hebbian , RH , and non-Hebbian , RnH , learning rules to the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , in response to each of these events to iteratively learn so that eventually the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , converges to the actual instantaneous presynaptic firing rate , rj ( A ( t ) ) . After a time period in which a particular pattern of postsynaptic activity , A ( t ) , occurs N times , a series of iterated operations of the non-Hebbian , RnH , and Hebbian , RH , rules are applied to the initial value of the decoded instantaneous presynaptic firing rate , r ^ j 0 ( A ( t ) ) ; r ^ j N ( A ( t ) ) = ( R n H ∘ R H ) ∘ . . . ∘ ( R n H ∘ R H ) ∘ r ^ j 0 ( A ( t ) ) , ( 1 ) where ∘ denotes functional composition , and brackets group operations for each occurrence of the postsynaptic activity A ( t ) , noting that the Hebbian rule , RH , is absent if presynaptic neuron j does not spike coincident with postsynaptic activity , A ( t ) . From the decoding target r ^ j N ( A ( t ) ) = r j ( A ( t ) ) , we derive a fundamental relationship between the Hebbian and non-Hebbian learning rules; Δ R n H ( r ^ j ) = - Δ t r ^ j Δ R H ( r ^ j ) , ( 2 ) where ΔRnH and ΔRH are the changes to r ^ due to each application of the non-Hebbian , RnH , and Hebbian , RH , learning rules , respectively . This relationship describes how the Hebbian and non-Hebbian changes should be related so that they cancel each other and no change occurs when r ^ ( A ( t ) ) , the estimate of the instantaneous presynaptic firing rate given some postsynaptic activity A ( t ) , equals r ( A ( t ) ) , the actual presynaptic firing rate given A ( t ) . To ensure convergence , the non-Hebbian , RnH , and Hebbian , RH , rules must slightly decrease and increase the decoded instantaneous presynaptic firing rate , r ^ j , respectively ( for further details of the derivation see S1 Appendix ) . This relationship between the learning rules matches the statistical relationship between the Hebbian and non-Hebbian events , remembering that Δ t r j ( A ) = p ( j spiking | A ) , whereby typically Hebbian events must make larger changes than non-Hebbian events because they occur less frequently . Now that we have established a relationship between Hebbian and non-Hebbian learning rules acting on the decoded instantaneous presynaptic firing rate , r ^ j , we consider how these learning rules may be implemented in a neural circuit as plasticity rules acting on synapses ( Fig 2a ) . In recurrent networks it is necessary to relate learning to the stability of the network to ensure that the network is stable [38] . Indeed , non-Hebbian , homeostatic plasticity is often invoked in models for the purpose of stabilizing unstable Hebbian plasticity . To control the stability of learning in a more principled way , we choose the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , to be such that when presynaptic neuron j spikes at t , r ^ j ( A ( t ) ) is proportional to presynaptic neuron j’s contribution to generating its postsynaptic spikes , as shown schematically in Fig 1b . Thus learning the decoded instantaneous presynaptic firing rate , r ^ j ( A ( t ) ) , also controls the individual gains for each neuron in the network . Further , if the neurons are divided into different populations , then we can compute the decoded instantaneous presynaptic firing rate , r ^ j α β ( t ) , for presynaptic neuron j in population α at time t , decoded from postsynaptic neurons in population β , and therefore control the gains of interactions between different populations of neurons . The change in membrane potential of a postsynaptic neuron i caused by a presynaptic spike in neuron j , decayed according to the leaky dynamics of the postsynaptic neuron , is well approximated a short time after the presynaptic spike occurs by wij ( Vi ( t ) − VE/I ) e−Δt/τ/C , where Δt is the time since the presynaptic spike and VE/I is the excitatory or inhibitory reversal potential of the synapse , τ is the decay constant of the postsynaptic neuron’s membrane potential , wij is the synaptic conductance from presynaptic neuron j to postsynaptic neuron i , and C is the capacitance of the postsynaptic neuron ( see Fig 2b and Methods ) . Thus , the decoded instantaneous presynaptic firing rate , r ^ j α β ( t ) , is given by summing the contributions from the first spike after t in the postsynaptic activity , Aβ ( t ) , for each postsynaptic neuron in population β . Only the first spike is considered since the reset and refractory state mean the presynaptic spike does not contribute to further postsynaptic spikes ( Fig 1b ) , r ^ j α β ( t ) = γ α β ∑ i ∈ A β ( t ) w i j C ( V i ( t ) - V E / I ) e - ( t i - t ) / τ , ( 3 ) where ti is the time at which postsynaptic neuron i first spikes after time t , and γαβ is a constant that scales the summed voltage changes into an instantaneous firing rate . Its value will be described in Sec . Emergent stability and balanced excitation and inhibition . Note that the decoded instantaneous presynaptic firing rate , r ^ j α β ( t ) , can be computed at any t , independent of the activity of presynaptic neuron j , and is a local quantity , only requiring information about postsynaptic neurons to which presynaptic neuron j has a direct connection . This defines the decoding component of the autoencoder , describing how presynaptic activity can be decoded from postsynaptic activity . Given the expression in Eq ( 3 ) for computing the decoded instantaneous presynaptic firing rate , r ^ j ( t ) , we now construct plasticity rules that modify the connection strengths , wij . The collective changes to the connection strengths from all applications of these plasticity rules implement the Hebbian , RH , and non-Hebbian , RnH , learning rule changes to the decoded instantaneous presynaptic firing rate , r ^ j ( t ) . We distinguish Hebbian , RH , and non-Hebbian , RnH , learning rules from their constituent lower level Hebbian , ΔwH , and non-Hebbian , ΔwnH , plasticity rules which are comparable to plasticity experiments . Firstly , when a presynaptic neuron j spikes at time t the Hebbian plasticity rule , Δ w i j H , is applied to the corresponding weight for each postsynaptic neuron i that subsequently spikes within the time period τA that defines the temporal extent of the postsynaptic activity , A ( t ) ( Fig 1b ) . This time period τA is chosen so that the presynaptic spike’s contribution to the postsynaptic neuron’s membrane potential and corresponding contribution to the decoded instantaneous presynaptic firing rate , r ^ j ( t ) , has decayed to be negligible after a delay of τA . The weight change is chosen to depend on the postsynaptic neuron’s contribution to the decoded instantaneous presynaptic firing rate , r ^ j α β ( t ) , in Eq ( 3 ) , see Fig 2b . This choice of credit assignment strengthens synapses to postsynaptic neurons whose spike generation relies strongly on the presynaptic spike from j . This creates competition between postsynaptic neurons and thus diversification of neural responses , including the development of receptive fields ( see Sec . Self-organization of a spatially extended spiking neural network with complex real-world stimulus ) [39] . So , Δ w i j H ( t j + τ A ) = ∓ ϵ w i j ( V i ( t j ) - V E / I ) e - ( t i - t j ) / τ C r ^ j α β ( t j ) , ( 4 ) where ϵ is a small learning rate . The negative and positive signs are for excitatory and inhibitory synapses respectively and ensure that the weight change is ( usually ) positive , thus implementing LTP . The decoded instantaneous presynaptic firing rate , r ^ j ( t ) , requires information about all of presynaptic neuron j’s postsynaptic neurons . We do not describe the biological mechanisms that might implement this . However , it seems plausible that this information would be available to j by integration of signals from its postsynaptic neurons; though this is a weaker locality than is often assumed , where locality is restricted to neurons on either end of a synapse [10] . The inclusion of the decoded instantaneous presynaptic firing rate , r ^ j α β , in Eq ( 4 ) modulates the size of the Hebbian change in response to past activity and can be interpreted as a form of metaplasticity [12–14] . This is because the decoded instantaneous presynaptic firing rate , r ^ , depends both on the connection strengths which change slowly over time due to accumulated small plasticity changes in response to neural activity , and also the immediate postsynaptic activity which evolves rapidly in time . Other plasticity rules can be constructed that together still obey the required relationship between the Hebbian , RH , and non-Hebbian , RnH learning rules described in Eq ( 2 ) . In particular one possibility is to include r ^ in the non-Hebbian plasticity rule instead of the Hebbian plasticity rule . Implementing this version of the non-Hebbian plasticity rule in simulations would be computationally very expensive as it would require computing r ^ at every timestep as part of the integral in Eq ( 6 ) , rather than only at the time of presynaptic spikes when included in the Hebbian rule in Eq ( 4 ) ; however , this may not be a limitation in biology . Based on the functional relationship between Hebbian and non-Hebbian rules in Eq ( 2 ) , we obtain a non-Hebbian plasticity rule from Eq ( 4 ) , d Δ w i j n H ( t ) d t = ± ϵ w i j ( V i ( t ) - V E / I ) e - ( t i - t ) / τ C , ( 5 ) where tk < t < ti . We integrate this over the time period between subsequent postsynaptic i spikes to yield a discrete weight change applied in response to postsynaptic neuron i spiking; Δ w i j n H ( t i ) = ± ϵ w i j C ∫ t k t i ( V i ( t ′ ) - V E / I ) e - ( t i - t ′ ) / τ d t ′ , ( 6 ) where tk is the time of the postsynaptic i spike before ti ( Fig 1b ) , and the positive and negative signs are for excitatory and inhibitory synapses respectively and ensure that the weight change is ( usually ) negative and implements LTD , as illustrated in Fig 2c . This non-Hebbian plasticity rule can be considered a homeostatic mechanism that scales down synaptic strengths [4 , 5 , 23] . This plasticity rule is local , only using information about neurons on either end of a synapse , and predicts a form of non-Hebbian spike timing dependence , whereby the magnitude of the change depends exponentially on the timing between subsequent pairs of postsynaptic spikes . In summary , the form of the combined change from both plasticity rules can be described qualitatively by a Hebbian term ( ϵ / r ^ ) × p r e × p o s t and a non-Hebbian term ϵ × post , Δ w ∼ ϵ r ^ ( p r e - r ^ ) p o s t , ( 7 ) where pre indicates that the Hebbian term requires a presynaptic spike , and post indicates the dependence on postsynaptic membrane potential and spiking as in Eqs ( 4 ) and ( 6 ) . This equation has a stable fixed point at r ^ = p r e . Thus , these two weight changes reach a balance where the size and frequency of each of the weight updates cancel each other on average over time so that the learned quantity , r ^ , corresponding to the decoded instantaneous presynaptic firing rate , approximates the actual presynaptic instantaneous firing rate , r , as shown in Fig 2d . This weight change is similar to that presented in [23] where plasticity takes the form Δw ∼ ϵpre ( post − ρ0 ) . In [23] the fixed point occurs when the postsynaptic activity reaches the fixed target firing rate ρ0 , and so the activity of the network is directed toward this fixed firing rate . The crucial differences to the plasticity rules presented here are that firstly , here the fixed point involves the presynaptic , not the postsynaptic activity . Secondly , the decoded instantaneous presynaptic firing rate , r ^ , varies and does not represent a fixed target firing rate like ρ0 as in [23] . Instead , the actual presynaptic firing rate r can take any value and is a target for the decoded instantaneous presynaptic firing rate , r ^ , which is modifiable and depends on the postsynaptic activity . Thus , neuronal firing rates can fluctuate over time in response to stimuli or other recurrent activity . It is possible to use the same approach we have taken to develop plasticity rules that perform supervised learning in which input connections to a neuron are learned and thus directly control postsynaptic firing rates , with a variable target postsynaptic firing rate provided by an external agent . This supervised learning formulation leads to a weight change of the form Δ w ∼ ( ϵ / r ^ ) p r e ( p o s t - r ^ ) . In this case , presynaptic and postsynaptic activity are swapped compared to Eq ( 7 ) and r ^ becomes the target postsynaptic supervision signal , but the details of this are left for future work . With this in mind , we can then interpret the plasticity rules presented in [23] as implementing supervised learning , but with the postsynaptic supervision signal being a simple fixed firing rate , ρ0 . We next show that these plasticity rules ( Eqs 4 and 6 ) reproduce the characteristics of a variety of experimentally observed plasticity phenomena [1 , 7 , 33–35 , 40] . First we demonstrate these plasticity rules capture an exponential timing dependence in the classic model of STDP , derived from experiments in which pairs of presynaptic and postsynaptic spikes with time delay Δt are repeated at a fixed frequency [6 , 7] . To reproduce the STDP-like plasticity in our model , we use the same spiking protocol with pairs of spikes repeated at 10 Hz as in [7] ( Fig 3a ) , and consider the changes resulting from the plasticity rules in Eqs ( 4 ) and ( 6 ) . To do this we assume that the membrane potential , V ( t ) , and the decoded instantaneous presynaptic firing rate , r ^ ( t ) , are constant . The magnitude of the weight change described by our rules depends on the combination of the membrane potential , V ( t ) , the decoded instantaneous presynaptic firing rate , r ^ , the connection strength , w , the learning rate , ϵ , the capacitance , C , and the neuronal time constant , τ , in Eqs ( 4 ) and ( 6 ) . The combination of these quantities are not fully constrained by available experimental data so the magnitude of the weight changes in Fig 3b and 3c are arbitrary , but the temporal characteristics are not . The temporal characteristics depend on the relative sizes of the pair stimulation period and the neuronal time constant , τ , here we choose a typical value of τ = 20 ms [41] , one fifth of the 100 ms pair stimulation period . The non-Hebbian weight change , ΔwnH , indicated by the red line in Fig 3b , is constant as it depends only on the time between postsynaptic spikes which is fixed by the pair frequency . The Hebbian weight change , ΔwH , decays exponentially with Δt indicated by the green lines in Fig 3b ( see Sec . STDP and triplet comparison and S1 Appendix for details ) . However , when r ^ ≈ 0 . 425 / τ , the combined weight change from both plasticity rules is qualitatively similar to the classical form of STDP ( Fig 3c ) , consisting of LTD for postsynaptic before presynaptic spike timing , increasing in magnitude toward Δt = 0 , and LTP for presynaptic before postsynaptic spike timing , decreasing as Δt becomes large . When r ^ > 0 . 425 / τ , LTP transitions to LTD for large positive Δt . This is consistent with reports of LTP changing to LTD for postsynaptic spikes 20 ms after a presynaptic spike [42] . The relative sizes of τ and the stimulated pair frequency are also important . When the pair frequency is much larger than τ , again LTP transitions to LTD for large positive Δt . These results are consistent with observations in [6] which report that LTP only occurred in STDP experiments for synapses that were initially weak , but LTD did not obviously depend on initial synaptic strength . In the context of these plasticity rules for the LTP regime ( a presynaptic before postsynaptic spike ordering with a small delay ) , a small initial connection strength indicates r ^ is likely to be small . When r ^ is small , the positive Hebbian weight change exceeds the negative non-Hebbian weight change , leading to LTP . However , as the initial connection strength increases , then r ^ will increase , reducing LTP for strong initial connection strengths , consistent with the experiments . This behaviour can be seen in Fig 3b in which the LTP regime is reduced as r ^ increases . For postsynaptic before presynaptic spike ordering and the LTD regime , the Hebbian plasticity change is small due to its exponential decay with the delay between presynaptic and postsynaptic spiking . In this case the non-Hebbian plasticity change dominates , meaning that LTD in this regime does not depend on initial connection strength as observed in the experiments . This behaviour can also be seen in Fig 3b for large Δt where LTD dominates and is less affected by changes to r ^ compared to the LTP regime . STDP experiments in rat visual cortex document a dependence of plasticity on the frequency of spike pairs where LTP is promoted at high frequencies [33 , 40] . Our plasticity model reproduces this same frequency dependent promotion of LTP , as shown in Fig 4a ( see Sec . STDP and triplet comparison , and S1 Appendix for details of the model curves and Table 1 for parameter values ) . This frequency dependence is a prediction of our model strongly attributed to the spike timing dependence of the non-Hebbian plasticity rule , see Eq ( 6 ) . For presynaptic before postsynaptic spike pair ordering with Δt = 10 ms as in [33] ( Fig 4a blue curve ) , the frequency dependence of our plasticity rules arises from the reduction in time between successive postsynaptic spikes as frequency increases . This shorter time period reduces the interval of the integral in Eq ( 6 ) and thus reduces the depressive non-Hebbian plasticity change , ΔwnH . This frequency dependence also applies to postsynaptic before presynaptic spike pair ordering with Δt = −10 ms as in [33] ( Fig 4a red curve ) ; however , in addition , the time between presynaptic and postsynaptic spikes is now dependent on the time between adjacent pairs , not only Δt . At low frequencies there is a large time delay between pairs and therefore the Hebbian plasticity change , ΔwH , is small and the depressive non-Hebbian plasticity change , ΔwnH , dominates . However , as the frequency increases , the time delay between pairs shortens and the Hebbian plasticity change , ΔwH , becomes larger and dominates . Related experiments stimulated rat CA1 presynaptic neurons at different frequencies , inducing a postsynaptic response and plasticity [34] . The resulting plasticity changes also show that LTP is promoted at high stimulation frequencies . These experiments do not report individual spike timings to which we can directly apply our plasticity rules; however , the frequency dependence can be captured by approximating the spiking activity of a pair of presynaptic and postsynaptic neurons as a series of stimulated presynaptic spikes each followed after a Δt delay by an induced postsynaptic spike , akin to an STDP protocol . Then , using the same assumptions for the STDP experiments described above , the model captures the frequency dependence observed in experiment ( see Sec . STDP and triplet comparison , and S1 Appendix for details of the model curves and Table 1 for parameter values ) . Additional experimental results have found this frequency dependence is shifted to lower frequencies when rats or mice are raised in darkness , compared to a normal light/dark cycle [1 , 35] . This has been interpreted as evidence for the sliding threshold mechanism in the Bienenstock , Cooper and Munro ( BCM ) model of plasticity , by associating dark rearing with reduced postsynaptic activity [43] . In our model the frequency at which the transition between LTD and LTP occurs changes with the decoded instantaneous presynaptic firing rate , r ^ , ( Fig 4b and 4c ) because r ^ determines the ratio of the positive Hebbian changes to the negative non-Hebbian changes as described in Eq ( 2 ) . When the decoded instantaneous presynaptic firing rate , r ^ , is low the transition frequency is low , promoting LTP and inducing increased postsynaptic activity . As the decoded instantaneous presynaptic firing rate , r ^ , increases , the transition frequency increases promoting LTD and constraining postsynaptic activity . In the context of light and dark rearing , we predict that the decoded instantaneous presynaptic firing rate , r ^ , values will be higher for a normal light/dark cycle , indicating a better encoding of information than for dark rearing . Thus , we expect the transition between LTD and LTP to occur at lower frequencies in animals raised in darkness , as is found experimentally . Work on STDP has been extended experimentally and in models beyond pairs of spikes to consider triplets of spikes consisting of either two presynaptic spikes and one postsynaptic spike , or one presynaptic spike and two postsynaptic spikes [25–27 , 44] . We use the same triplet protocols as in [44] to compare model predictions with experimental data of synaptic changes resulting from spike triplet stimulation experiments . These consist of two presynaptic spikes and one postsynaptic spike , or one presynaptic spike and two postsynaptic spikes , repeated at a frequency of 1 Hz and with delays between spikes varying for different trials . Using the same assumptions as for the STDP comparison in Sec . Comparison to classical STDP ( see Sec . STDP and triplet comparison , S1 Appendix for details of the model curves and Table 1 for parameter values ) , our model predicts plasticity changes that are consistent with the experimental results , aside from the ( 5 , 15 ) ms presynaptic , postsynaptic , presynaptic triplet , see Fig 5 . This is similar to the triplet model in [25] which also fit well to these same experimental data , aside from this same data point . Our plasticity rules are not triplet based; however , a single application of both the Hebbian and non-Hebbian plasticity rules involves three spikes . Two postsynaptic spikes are involved in the non-Hebbian rule . One of those postsynaptic spikes , plus an additional presynaptic are involved in the Hebbian rule . We now demonstrate that these plasticity rules can be used in recurrent neural populations to encode a high dimensional , rapidly time varying , complex real-world stimulus , approximating the sensory encoding task that real brains must perform . In the process of learning this task , the interplay of our plasticity rules with neural activity gives rise to realistic properties of the circuit dynamics and connectivity . We apply the Hebbian and non-Hebbian plasticity rules in Eqs ( 4 ) and ( 6 ) to a biologically plausible neural circuit composed of leaky integrate and fire ( LIF ) neurons with conductance based synapses arranged in a 2D spatial grid consisting of an excitatory and an inhibitory population [17] . These neural populations receive input from a stimulus population whose spiking activity is given by an event based , complex , real-world visual stimulus collected using a Dynamic Vision Sensor ( DVS ) ( see Fig 6a , Table 2 and Methods ) [45] . Due to the initialisation of strong connections from the stimulus ( see Table 2 ) , the initial spiking activity of the neural populations is characterised by very high firing rates , 〈r〉 = 53 . 3 Hz and 〈r〉 = 53 . 6 Hz for the excitatory and inhibitory populations , respectively ( Fig 6b ) . We choose the initial connectivity to highlight the ability of the plasticity rules to stabilise and balance the network , even from a poor initial configuration . Over time the plasticity rules reduce the initially strong connections from the stimulus to the neural populations and increase the strength of the excitatory and inhibitory recurrent feedback . The initial stimulus driven , high firing rate , low variability spiking activity is replaced by activity driven by recurrent excitation and inhibition as well as the stimulus , so that the spike statistics of the neural populations are variable with moderate firing rates; 〈CVISI〉 = 2 . 0 ( Fig 6c ) and 〈r〉 = 6 . 0 Hz ( Fig 6b ) for the excitatory population , and 〈CVISI〉 = 1 . 9 ( Fig 6c ) and 〈r〉 = 5 . 9 Hz ( Fig 6b ) for the inhibitory population , while the Fano-Factor increases monotonically from about one using a 50 ms window to 1 . 8 using a 500 ms window for all populations ( Fig 6d ) , indicating greater variability than Poisson spiking . Such variable spike timing with fluctuating firing rates have been widely observed in the cortex [17 , 18 , 46] . The fluctuations of the neural populations are related to that of the stimulus driving them ( Fig 7 ) . Locally , the activity of a region of a neural population is partly driven by the activity of the corresponding local region of the stimulus . Thus , the variability of the neural populations tends to reflect that of the stimulus . Comparison of the actual stimulus with the estimates of the stimulus decoded from the activity of the neural populations ( Fig 8 and S1 Video ) illustrates that the circuit has learned its autoencoder function . In the decoded stimulus , the outline of the moving objects in the time varying input are visible and align with the actual stimulus , whereas in the poor initial configuration before learning , the decoded stimulus is not similar to the actual stimulus . The resulting synaptic strengths are distributed approximately log-normally ( Fig 9a ) , as has been indicated experimentally [19 , 20 , 47] . During learning , competition between nearby neurons leads to the development of receptive fields in which the connections from the stimulus to individual neurons adapt so that nearby neurons receive inputs from and respond to different local features of the stimulus ( Fig 9b ) . The development of receptive fields is necessary from a theoretical perspective as these receptive fields indicate that neurons encode features of the input , creating a more useful representation for performing other tasks . Experiments indicate that the excitatory and inhibitory input currents to an individual neuron are close to equal in magnitude or balanced , when averaged over a long time period . Further , tight balance occurs when the currents track each other [16 , 21 , 48] . We now describe how balanced excitatory and inhibitory currents can emerge using these plasticity rules by first considering stability . When a neural population is stable , on average each new spike that occurs will contribute toward the net production of one further spike distributed amongst its postsynaptic neurons . Thus , if we observe a period of activity containing a combined sE , sI , sD excitatory , inhibitory and stimulus spikes respectively , then we require that the net contribution from all of these spikes to further spikes be equal to the change in membrane potential required to produce them . Recalling that r ^ j α β ( t ) is proportional to the summed changes in membrane potential of its spiking postsynaptic neurons ( Eq 3 ) , this condition is described by V s s β = s E ⟨ r ^ E β ⟩ γ E β + s D ⟨ r ^ D β ⟩ γ D β - s I ⟨ r ^ I β ⟩ γ I β , ( 8 ) where Vs = VT − VR is the change in membrane potential from reset to threshold that is required to produce one spike , and 〈 r ^ α β 〉 is the average of the decoded instantaneous presynaptic firing rate , r ^ α β , over each neuron in the α population at their spike times . Each of the terms on the right hand side describe the expected contribution to spikes in β from spikes in each of the inhibitory , excitatory and stimulus populations . The stability condition described by Eq ( 8 ) configures learning to produce balanced excitatory and inhibitory currents if the scaling factors , γαβ , are chosen so that the changes in membrane potential caused by the excitatory and inhibitory currents are large compared to their difference . In S1 Appendix we demonstrate that the learning of this tight balance using our plasticity rules converges because it can be viewed as a gradient descent in a firing rate model . In the demonstration circuit stimulated with the DVS stimulus , the plasticity rules are configured by choosing the scaling factors , γαβ , to satisfy Eq ( 8 ) so that the total amount of excitation from the stimulus population is approximately equal to the total amount of excitation from the excitatory population , and this combined excitation is closely tracked and balanced by inhibition ( Fig 10a ) , with a small 1 . 5 ms delay evident in their cross-correlation ( Fig 10b ) , as found in neural recordings [16] .
There is a growing body of work attempting to describe and understand the diverse range of experimentally observed features of plasticity , for example [3 , 5 , 10 , 11 , 23 , 24 , 49] . Some past studies combine experimentally motivated plasticity rules and apply them to neural circuits . The parameters of these models are then tuned in order to produce behaviour that is then argued to support some kind of function such as memory or decision making [23 , 26–30] . However , this approach does not adequately connect plasticity to the broader , more important picture of why plasticity exists at all; perhaps most importantly , for learning function . In theoretical models , learning implemented via plasticity requires carefully constructed plasticity rules whose properties are governed by an overarching learning algorithm . Thus , an adequate theoretical understanding of plasticity requires a description of its functional role in an overarching learning algorithm . In contrast to previous observation driven models , we have proposed an approach that explicitly links the learning of function to the underlying plasticity phenomena . We started by considering how to learn function and then find that this leads to plasticity rules that have properties similar to that observed experimentally , thus directly explaining and connecting these plasticity phenomena to their functional roles in learning . More specifically , we have investigated a description of plasticity derived from extending the classical autoencoding function used in machine learning to recurrent networks in which individual neurons learn to encode their own activities into the population activity . This necessitates both Hebbian and non-Hebbian plasticity rules as they each account for the spiking statistics required for autoencoding , contrasting with a variety of other plasticity models [5 , 24] where the use of Hebbian and non-Hebbian rules is motivated primarily by stability considerations [23] and to align with experimental observations , for example [25 , 50] . These Hebbian and non-Hebbian rules are related by the presynaptic neuron’s instantaneous firing rate decoded from the population activity . This relationship entails a dependence on past activity . In many past studies , for example [12–14] , the dependence of plasticity rules on past activity has been termed metaplasticity . Thus , in this context the relationship can be considered a form of metaplasticity . However , it should be noted that the relationship is integral to the plasticity rules themselves and is not separate from them , as the term metaplasticity may imply . This extends the existing understanding of metaplasticity by theoretically describing its interaction with plasticity , learning and function , whereas previously metaplasticity has mostly been investigated from an experimental perspective and considered as a mechanism to regulate synaptic strengths and prevent them from saturating [12–14] . The decoded instantaneous firing rate varies at short timescales with changes in postsynaptic activity , but also varies at longer timescales with the gradual , accumulated changes in connection strengths due to plasticity . These variations at multiple timescales explain how fast but unstable Hebbian plasticity can be stabilised by homeostatic plasticity that experimentally appears slow , but also works at faster timescales . This resolves the paradox of the theoretical need for fast homeostatic plasticity to stabilise unstable Hebbian plasticity with experimental observations of considerably slower homeostatic plasticity evolving over hours or days [5 , 10 , 11] . The form of our plasticity rules capture the exponential spike timing dependence of classical STDP as well as the frequency dependence of both STDP and non-timing based stimulation experiments , including a sliding threshold like mechanism in which the decoded instantaneous firing rate controls the frequency at which LTD transitions to LTP . Spike-triplet models have also been investigated as a means of providing a spike timing based account of BCM [51] , and have been investigated for possible functional roles , such as maximising information transmission [52 , 53] , or bounds on differences between target and model distributions [54] . In our model of plasticity , the stability of recurrent neural circuits is linked to their function because the learned quantity for each neuron , its decoded instantaneous firing rate , is proportional to the gain of that neuron . Furthermore , gains between populations can be controlled in this same way by decoding from each population separately . When our plasticity rules are implemented in a neural circuit , balancing these gains between excitatory and inhibitory populations at a network level leads to tightly balanced excitatory and inhibitory currents into individual neurons , which has been observed in experiments [21] . Other models of plasticity have produced this balance [23 , 55] , but have not made the connection to a range of other plasticity phenomena , as described above . This tight balance is thought to explain variability in spiking activity of individual neurons because spikes are generated by an imbalance over short time windows . Consistent with experimental observations of spiking variability , we also find that these plasticity rules are able to shift a network from a deliberately poor initial regime of very strong stimulus input driving high firing rates , to a balanced regime driven by recurrent activity in conjunction with a much weakened stimulus input . In this regime , firing rates are reduced and neurons exhibit variable spiking . The plasticity rules do not enforce a fixed firing rate , unlike other models [23] , rather firing rates fluctuate in response to the stimulus and recurrent dynamics so that the neurons can perform their autoencoding function . The self-organising process in the interplay between the neural circuit and its plasticity mechanisms also accounts for other salient features of neural dynamics including approximately log-normal distributions of connection strengths , as has been indicated experimentally [19 , 20] . Neurons develop receptive fields that extract features optimised for the particular stimulus used in this study . These emergent receptive fields enable the complex , real-world visual stimulus to be accurately decoded from the neural populations , indicating that the original function underpinning these plasticity rules was learned . Our plasticity rules are directly applicable to implementing autoencoder learning in neural circuits stimulated with complex , real-world data . The function of the neural circuit can be directly observed by decoding the circuit’s activity . This is unlike many other models that apply plasticity rules that are not derived from function , or are derived from functional principles but are not linked to a task and therefore have an unclear relationship between their resulting neural dynamics and implementing brain functions [23 , 26–30] . These results provide theoretical motivation for revisiting previous plasticity experiments to more closely examine past interpretations , as expressed elsewhere [24 , 40 , 56] . In particular , these results imply that classical STDP is a special case of a combination of more fundamental plasticity rules , and that non-Hebbian plasticity should also exhibit a spike timing dependence on postsynaptic spikes in individual neurons . This non-Hebbian spike timing dependence can explain the frequency dependence of plasticity , though we are not presently aware of any existing attempts at directly observing non-Hebbian spike timing dependence . To date , many plasticity experiments tend to focus on controlling a single variable , often spike timing , and measuring synaptic changes . However , it is very likely that many neuronal and synaptic variables , not just spike timing , are strongly implicated in brain function and therefore in learning and plasticity . Thus , plasticity rules are likely to depend on a range of variables that are not usually simultaneously reported for individual neurons in experimental work . In the case of our model , since our Hebbian plasticity rule strongly depends on the decoding of the presynaptic activity , to compute this it is necessary to measure the connection strengths , membrane potentials and spiking activity of all postsynaptic neurons to gain a more comprehensive understanding of synaptic plasticity . These properties are likely to vary greatly between individual neurons , thus without such measurements , comparing our model to data requires making assumptions about these unmeasured variables and introducing unconstrained parameters . Further , the potential variability between neurons may mean that averaging across a small number of neurons could be misleading and a potential source of conflicting observations [1 , 6 , 7 , 44 , 57] . More complete experimental characterisations of plasticity will require fewer assumptions to be made when comparing to models . Further to this , fewer experiments have studied plasticity of inhibitory synapses than excitatory ones; however , these experiments show an inconsistent range of different timing based behaviours of inhibitory synapses that differ to those observed for excitatory synapses [58] . Thus , further experiments are required to determine if these differences between excitatory and inhibitory plasticity observations are a result of different experimental designs or differences in the underlying plasticity rules . The plasticity rules we present apply to both excitatory and inhibitory synapses , meaning that broadly they share the same timing and frequency features , but the synaptic reversal potential changes depending on the synapse type , thus excitatory and inhibitory plasticity have a different dependence on postsynaptic membrane potential . Plasticity rules must be local , meaning that modifying a synapse does not require detailed information about distant parts of the network . It is often assumed that plasticity rules can only use information about neurons on either end of a synapse [10]; however , from a functional perspective this assumption is very restrictive as it limits the ability of connections to organise collectively . In contrast , our Hebbian rule involves the decoded instantaneous firing rate which is computed using information about all postsynaptic spiking neurons . Thus , our plasticity rules while still local , have locality relaxed to include information not only about both the presynaptic and postsynaptic neurons on either end of a synapse , but also other neurons that are postsynaptic to the presynaptic neuron . We do not yet have a mechanistic description of how these plasticity rules are implemented in biology . Discovering a mechanism by which postsynaptic signals can be integrated and distributed to synapses would be an important step in demonstrating the biological plausibility of this relaxed locality . Further to this , a detailed description of the underlying biological mechanisms would enable a more informed comparison to experiment . For example , some plasticity experiments isolate target neurons by removing surrounding neurons to which the target neurons are likely connected [44] . At present it is not clear if removing these neurons is equivalent to their connections never being present at all , or if underlying mechanisms are disrupted in this process , leading to spurious results . Further , it is not clear if external control over the activity of neurons disrupts these underlying mechanisms , for example as in STDP experiments . In conclusion , based on the function of neurons learning to encode their activity into the population activity so that neural circuits learn to act as recurrent autoencoders , we present a novel , unified account of synaptic plasticity , dynamics and neural coding that explains a great variety of plasticity features and neural dynamics . This unified account thus represents a significant advance toward understanding the working mechanisms of cortical circuits . Further explanations of how such emergent plasticity rules and their interplay with neural circuits can explain cognitive functions would be an important direction for future studies .
To compare our model to STDP and triplet experiments we apply the plasticity rules in Eqs ( 4 ) and ( 6 ) according to the timings of the spikes , as described in S1 Appendix . The resulting equations describing the expected weight change are as follows . For STDP presynaptic before postsynaptic timing Δ w i j w i j = ∓ ϵ n V i - V E / I C ( e - Δ t / τ r ^ j - τ ( 1 - e - T / τ ) ) . ( 9 ) For STDP postsynaptic before presynaptic timing Δ w i j w i j = ∓ ϵ n V i - V E / I C ( e - ( T - Δ t ) / τ r ^ j - τ ( 1 - e - T / τ ) ) . ( 10 ) For postsynaptic , presynaptic , postsynaptic triplets Δ w i j w i j = ∓ ϵ n V i - V E / I C ( e - Δ t 2 / τ r ^ j - τ ( 2 - e - ( Δ t 1 + Δ t 2 ) / τ ) ) . ( 11 ) For presynaptic , postsynaptic , presynaptic triplets Δ w i j w i j = ∓ ϵ n ( V i − V E / I ) C ( e − Δ t 1 / τ r ^ j − τ ) . ( 12 ) The spiking neural circuit model is based on that used in [17] and is summarised in Tables 2 and 3 . We consider a 2D network of 300 × 300 coupled , conductance-based leaky integrate-and-fire neurons consisting of 80% excitatory and 20% inhibitory neurons , as well as a population of stimulus neurons whose spiking activity is controlled by an external data source ( Fig 6a ) . Both excitatory and inhibitory neurons are evenly spaced , with the spacing between inhibitory neurons twice the spacing between excitatory neurons . Connections exist between all pairs of neurons within 15 grid units of each other . We denote the membrane potential of a neuron i at time t as Vi ( t ) , with dynamics given by the following: C d d t V i ( t ) = - g L [ V i ( t ) - V L ] - g i E ( t ) [ V i ( t ) - V E ] - g i I ( t ) [ V i ( t ) - V I ] , ( 13 ) where the capacitance C = 1 nF , the leak conductance gL = 50 nS , and the reversal potentials are VL = −70 mV , VE = 0 mV , and VI = −80 mV for the leak , excitatory , and inhibitory conductance , respectively [41] . If the membrane potential of a neuron reaches the threshold of VT = −55 mV , a spike is generated and the membrane potential is reset to the reset potential VR = −70 mV for a refractory period τref = 5 ms . The synaptic conductances are as follows: g i E / I = ∑ j w i j E / I ∑ k G E / I ( t - t k ) , ( 14 ) where E and I indicate excitatory and inhibitory respectively , and tk is the time of the k-th spike emitted by the presynaptic neuron j . The time course of the postsynaptic conductance is given by the following: G E / I ( t ) = e - t / τ d E / I - e - t / τ r E / I τ d E / I - τ r E / I ( 15 ) with rise times τ r E = 0 . 5 ms and τ r I = 0 . 5 ms , and decay times τ d E = 2 . 0 ms and τ d I = 7 . 0 ms . The denominator is a normalisation factor such that ∫ 0 ∞ G E / I ( t ) d t = 1 . The network model is simulated using a time step of 0 . 1 ms . The plasticity rules described by Eqs ( 4 ) and ( 6 ) are applied to all connections with the parameters listed in Table 2 . All connections from the stimulus population to both neural populations are initialised to be very strong , while all connections between the neural populations are initialised to be very weak ( Table 2 ) . In this configuration the network activity is driven strongly by the stimulus with little contribution from recurrent connectivity within and between the neural populations . Code for these simulations is available at https://github . com/BrainDynamicsUSYD/SpikeNet . The file /SpikeNet/cases/generate_learning_config . m is a MatLab script that can be used to generate a configuration file that will instruct the simulator to implement these plasticity rules . Further information on installing and running the simulator can be found in the readme /SpikeNet/README . md . In this neural circuit we include a 128 × 128 population of stimulus neurons whose spiking activity is controlled by an external data source . This population connects feedforward to both the excitatory and inhibitory populations . The stimulus dataset used in this work was recorded using a Dynamic Vision Sensor ( DVS ) [45] . The DVS is a retina inspired imaging device that captures events rather than frames as in traditional video cameras , and provides a natural spiking stimulus for SNN models , where each pixel corresponds to one neuron . The DVS used to collect the dataset has a resolution of 128x128 pixels , and 1 μs temporal precision , with a minimum 15μs between sequential events for each pixel . This ability to record events at fast timescales makes it much more suitable as a visual input to a SNN model than conventional frame based video cameras whose frame updates are typically every few tens of milliseconds and are much slower than the neuron dynamics that evolve significantly over millisecond timescales . Each pixel in the DVS records an event when the light intensity on that pixel changes above a threshold value . The DVS is therefore sensitive to movement , changes in lighting or changes in reflective properties of objects . The absence of events means that there has been no change , not that the visual scene is empty . The dataset was obtained from https://www . ini . uzh . ch/∼tobi/dvs/ and is a recording of a person juggling , viewed front on . Dropout is a simple and effective technique used to reduce over-fitting in artificial neural networks by adding noise to the network through probabilistically deactivating neurons . Dropout can be thought of as a form of model averaging and leads to better generalisation [59] . Here , we also inject noise into the learning scheme to reduce over-fitting; however , unlike in dropout we do not probabilistically omit the transmission of spikes between neurons , but only omit spikes from the learning scheme by omitting applications of plasticity rules involving dropped spikes . Because of this no rescaling of weights is required after learning as in dropout , thus this form of regularisation can be implemented online and is biologically plausible . All spikes are communicated between neurons so this does not represent synaptic failure; however , it would be possible to modify the form of noise or synaptic dynamics to investigate effects like synaptic failure or short term depression and how they may enhance or degrade learning . | Many experiments have documented a variety of ways in which the connectivity strengths between neurons change in response to the activity of neurons . These changes are an important part of learning . However , it is not understood how such a diverse range of observations can be understood as consequences of an underlying algorithm used by brains for learning . In order to understand such a learning algorithm it is also necessary to understand the neural computation that is being learned , that is , how the functions of the brain are encoded in the activity of its neurons and its connectivity . In this work we propose a simple way in which information can be encoded and decoded in a network of neurons for operating on real-world stimuli , and how this can be learned using two fundamental plasticity rules that change the strength of connections between neurons in response to neural activity . Surprisingly , many experimental observations result as consequences of this approach , indicating that studying the learning of function provides a novel framework for unifying plasticity , dynamics , and neural computation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"nervous",
"system",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"synaptic",
"plasticity",
"neuronal",
"plasticity",
"developmental",
"neuroscience",
"computer",
"and",
... | 2018 | Functional mechanisms underlie the emergence of a diverse range of plasticity phenomena |
The microbiome shapes diverse facets of human biology and disease , with the importance of fungi only beginning to be appreciated . Microbial communities infiltrate diverse anatomical sites as with the respiratory tract of healthy humans and those with diseases such as cystic fibrosis , where chronic colonization and infection lead to clinical decline . Although fungi are frequently recovered from cystic fibrosis patient sputum samples and have been associated with deterioration of lung function , understanding of species and population dynamics remains in its infancy . Here , we coupled high-throughput sequencing of the ribosomal RNA internal transcribed spacer 1 ( ITS1 ) with phenotypic and genotypic analyses of fungi from 89 sputum samples from 28 cystic fibrosis patients . Fungal communities defined by sequencing were concordant with those defined by culture-based analyses of 1 , 603 isolates from the same samples . Different patients harbored distinct fungal communities . There were detectable trends , however , including colonization with Candida and Aspergillus species , which was not perturbed by clinical exacerbation or treatment . We identified considerable inter- and intra-species phenotypic variation in traits important for host adaptation , including antifungal drug resistance and morphogenesis . While variation in drug resistance was largely between species , striking variation in morphogenesis emerged within Candida species . Filamentation was uncoupled from inducing cues in 28 Candida isolates recovered from six patients . The filamentous isolates were resistant to the filamentation-repressive effects of Pseudomonas aeruginosa , implicating inter-kingdom interactions as the selective force . Genome sequencing revealed that all but one of the filamentous isolates harbored mutations in the transcriptional repressor NRG1; such mutations were necessary and sufficient for the filamentous phenotype . Six independent nrg1 mutations arose in Candida isolates from different patients , providing a poignant example of parallel evolution . Together , this combined clinical-genomic approach provides a high-resolution portrait of the fungal microbiome of cystic fibrosis patient lungs and identifies a genetic basis of pathogen adaptation .
The microbiome has a profound impact on diverse facets of human biology and disease . Anatomical sites such as the respiratory tract that were once thought to be sterile are now recognized to harbor complex microbial communities in healthy individuals as well as those suffering from a multitude of conditions [1] . One respiratory condition for which chronic and complex polymicrobial communities are now appreciated to have a severe impact on prognosis is cystic fibrosis . Patients with this genetic disorder caused by reduced function of the transmembrane conductance regulator CFTR experience thick mucus accumulation in airways , which renders them vulnerable to chronic airway infection and repeated episodes of pulmonary exacerbation [2] . Our understanding of microbiomes of the lung and respiratory tract of cystic fibrosis patients is based largely on studies focused on bacteria [2 , 3] . Traditionally , opportunistic bacterial pathogens such as Pseudomonas aeruginosa , Burkholderia species , and Staphylococcus aureus have been implicated in pulmonary exacerbations [3] . Molecular approaches have provided additional insight into culprits of respiratory infections in cystic fibrosis patients , as with the implication of a bacterial opportunistic pathogen of the Streptococcus milleri group that had been previously overlooked [2 , 4] . The resultant episodes of pulmonary exacerbations cause decline in lung function , ultimately leading to patient death . There is a growing appreciation of the importance of fungi in the lung microbiome . With an excess of 50 , 000 fungal spores/m3 of air in some seasons , the human respiratory tract is exposed to vast quantities of fungi [1] . The lung is often the initial site of colonization from which dissemination can lead to systemic fungal infections , especially in immunocompromised individuals and those treated with antibacterial agents [5] . Fungi are frequently isolated from cystic fibrosis patient sputum samples , where Candida albicans and Aspergillus fumigatus are the most prevalent species , identified in 40–70% of patients [6 , 7] . Fungi can have a profound impact on cystic fibrosis patients by inducing local host inflammatory responses as well as causing direct damage of the respiratory mucosa , thereby contributing to progressive deterioration of lung function . For example , allergic bronchopulmonary aspergillosis ( ABPA ) has been reported in up to 10 . 9% of cystic fibrosis patients , while chronic A . fumigatus infection leads to complications such as asthma , bronchitis , aspergilloma , and invasive pulmonary infection after lung transplants [8 , 9] . Although C . albicans is frequently isolated from cystic fibrosis patient sputum samples , the clinical impact is not yet clear . C . albicans has been reported as the leading cause of allergic bronchopulmonary mycosis ( ABPM ) after Aspergillus species in the general population [10] , and chronic Candida infection has been associated with increased pulmonary exacerbations in cystic fibrosis patients [11] . The few microbiome studies that have focused on fungi to date have revealed that communities in the lungs of healthy people are dominated by fungal species that are ubiquitous in the environment [12] , while communities in cystic fibrosis patients are dominated by species of Candida , Aspergillus , and Malassezia [7 , 13] . Analysis of fungal microbiomes via high-throughput sequencing provides a powerful approach to monitor changes in species dynamics over the course of clinical exacerbations and treatments . Leveraging phenotypic and genotypic characterization of microbial populations in the host provides an important complement to monitoring species level dynamics that is key for understanding microbial ecology and evolution in context . Fungal adaptation in the host has been studied most extensively in Candida species , where genome sequencing has been used to elucidate the genetic basis of the emergence of drug resistance in pathogen populations over the course of patient treatment [14–16] . For C . albicans , a natural member of the human mucosal microbiota and an important opportunistic pathogen , another key trait for which variation has been observed among clinical isolates is morphogenesis [16] . C . albicans can transition between yeast and filamentous morphologies in response to specific environmental cues , which has a profound impact on virulence and host adaptation [17] . C . albicans morphology can also be modulated by inter-kingdom interactions with bacterial pathogens commonly associated with cystic fibrosis . P . aeruginosa inhibits C . albicans filamentation by secreting molecules such as pyocyanin and 3-oxo-C12 homoserine lactone , and selectively kills C . albicans filaments by forming biofilms on their surfaces [18–20] . Burkholderia cenocepacia is also able to inhibit C . albicans filamentation by secreting cis-2-dodecenoic acid [21] . In contrast , S . aureus has been reported to engage in mutualistic interactions with C . albicans , where it preferentially attaches to C . albicans cells in mixed biofilms and enhances disease severity in a mouse co-infection model [22 , 23] . Thus , cystic fibrosis patient lungs provide a clinically important and powerful context for tracking the dynamics and evolution of fungal communities and populations in the host . In this study , we combined high-throughput sequencing of the ribosomal RNA internal transcribed spacer 1 ( ITS1 ) with phenotypic and genotypic analyses to provide the most comprehensive assessment of the cystic fibrosis lung fungal microbiome ( mycobiome ) to date . We found that the fungal communities identified by culture-independent sequencing of 89 sputum samples from 28 cystic fibrosis patients were concordant with those defined by culture-based analyses of 1 , 603 isolates from the same samples . Both methods identified C . albicans as the dominant fungus from most cystic fibrosis patients , although some patients were primarily colonized by Candida parapsilosis , A . fumigatus , or a mixture of different fungal species including Aspergillus flavus , Aspergillus terreus , Candida glabrata , and Candida tropicalis . Our phenotypic characterization revealed extensive inter-species and intra-species diversity in growth characteristics and antifungal drug resistance . For C . albicans , we recovered multiple isolates from different patients over time that displayed filamentous growth in the absence of any inducing cue . Whole genome sequencing identified loss-of-function mutations in NRG1 as the genetic basis of the filamentous phenotype . Mutations in the transcriptional repressor NRG1 were identified in 24 out of 25 filamentous C . albicans isolates that were recovered from six different patients , suggesting that this is a common mechanism of adaptation . The filamentous C . albicans clinical isolates were resistant to the filamentation-repressive effects of the dominant bacterial pathogens present in the patients from which they were recovered , including P . aeruginosa and Burkholderia multivorans , as well as a structural analog of the quorum sensing molecule pyocyanin . Thus , we provide a high-resolution portrait of species and population level dynamics in the fungal microbiome of cystic fibrosis patient lungs , and identify the genetic basis of pathogen adaptation in the host .
We collected a total of 111 sputum samples from 28 adult CF patients from St . Michael’s hospital in Toronto . Of the 28 participants , 20 patients provided at least 2 sputum samples between 3 and 56 months apart for longitudinal analyses ( S1 Table ) . Patients CF020 , CF025 , CF027 , CF060 , CF098 , and CF107 were diagnosed with ABPA , and patients CF014 , CF025 , CF028 , CF060 , CF098 , and CF107 were prescribed fluconazole , itraconazole , or posaconazole during sampling period ( S1 Table ) . Other patients were prescribed antifungal drugs prior or after the sputum collection periods , including CF006 ( nystatin ) , CF011 ( nystatin ) , CF020 ( nystatin ) , and CF027 ( fluconazole ) . According to the clinical microbiology data , P . aeruginosa was the most commonly isolated microorganism from the sputum samples , followed by Aspergillus species , Stenotrophomonas maltophila , and B . cepacia , among other bacterial and fungal species ( S1 Table ) . We used a molecular approach to characterize the cystic fibrosis mycobiome , and extracted DNA from 111 sputum samples for amplification and high-throughput sequencing of ITS1 , which is a commonly used locus for fungal species identification [13] . We were able to amplify ITS1 from 89 of the 111 sputum DNA samples , although all of the 111 sputum samples produced viable fungal isolates . Amplification of ITS1 did not track with the number of fungal isolates from the corresponding sputum sample ( S1 Table and S1 File ) . The amplified products were barcoded , pooled , and sequenced using Illumina short-read sequencing methods , and the taxonomic identities were assigned to each read . In most sputum samples , Candida species were the most abundant fungi ( 51 samples out of the 89 ) , followed by Aspergillus species ( 18 samples ) ( Fig 1 , and S1 Fig ) . We detected pathogens previously reported in other cystic fibrosis studies , such as Exophiala dermatitidis [24] from CF011 , Geosmithia argillacea [25] from CF133 , and Malassezia species [7 , 13] from CF107 and CF117 ( Fig 1 ) . Furthermore , we detected opportunistic fungal pathogens yet to be reported in the context of cystic fibrosis to date , such as Cochliobolus species ( anamorphs Curvularia ) [26] from CF039 and Kluyveromyces marxianus ( anamorph Candida kefyr ) [27] from CF098 ( Fig 1 ) . The fact that we identified additional species in our sputum samples may reflect the more extensive sampling in our study design . Next , we tested for associations between taxonomic abundance and patient characteristics including clinical state , lung disease stage , Pseudomonas load , triazole therapy , lung function index FEV1% predicted value ( based on the volume exhaled during the first second of forced expiration for a population average of similar age ) , and body mass index ( BMI ) , as described in the patient clinical data ( S1 Table ) . We utilized the Simpson’s diversity index , which measures community diversity within the samples and produces a numerical output between 0 to 1 , where 0 indicates no diversity and 1 indicates maximum diversity . There was no significant association between the Simpson’s diversity index from individual samples and patient characteristics , as was the case with Shannon diversity index ( S2–S4 Figs ) . Based on a previous report that chronic C . albicans colonization leads to reduced FEV1 values in cystic fibrosis patients [11] , we also compared the correlation between C . albicans read counts and patient FEV1% predicted values , which revealed a weak , yet significant correlation ( r = -0 . 14 , S5 Fig ) . In order to compare the differences in taxonomic compositions between the sputum samples over time , we selected sputum samples from patients with more than 1 sampling points ( 19 patients ) and utilized the Bray-Curtis dissimilarity index . This measures the dissimilarity between samples and produces a numerical output between 0 to 1 , where 0 indicates no dissimilarity and 1 indicates no similarity . We did not identify clear community structures based on patient status , but most samples clustered based on patient ID ( S6 and S7 Figs ) . The changes observed in taxonomic composition ( Fig 1 ) or Bray-Curtis dissimilarity index ( S6 Fig ) did not correlate with the changes in patient status ( S1 Table ) . We recovered fungal isolates from sputum samples by plating homogenized sputum onto medium containing antibiotics to inhibit bacterial growth ( S1 File ) . We sampled up to 50 colonies per sputum sample , with a mean of 14 colonies and a range of 1 to 50 colonies . For sputum samples with more than 50 fungal colonies , an effort was made to maintain the relative abundance of colony morphologies during sampling . Fungal isolates were initially grouped into two major categories: 182 mold isolates from 10 patients and 1 , 421 yeast isolates from 26 patients ( S1 File ) . In order to assess phenotypic diversity and define phenotypic classes to facilitate species assignments by sequencing , all isolates were then screened for growth rate , antifungal susceptibility , and cellular morphology . For molds , conidia coloration was noted , and yeast isolates were plated on CHROMagar Candida [28] . A minimum of two representative isolates of each phenotypic class were assigned to species by sequencing the internal transcribed spacer 2 ( ITS2 ) region of the ribosomal RNA , which is also commonly used for fungal species identification [29] . In all cases , sequencing confirmed that isolates with concordant phenotypic profiles were the same species . Different patients harbored distinct species of molds . All 182 mold isolates were classified into three species: A . fumigatus , A . flavus , and A . terreus ( Fig 1 ) . Since the ITS2 locus is not able to discriminate between closely related species within Fumigati , which includes A . fumigatus [30] , we selected a minimum of two representative isolates initially identified as A . fumigatus based on ITS2 sequence per patient , and sequenced the β-tubulin locus to confirm the species identity . We recovered 103 A . fumigatus isolates from nine patients , and it was the most frequently recovered mold from the sputum samples . We recovered A . flavus isolates from two patients , with 56 of the isolates from a single patient ( CF028 ) and one from another patient . We also recovered 22 A . terreus isolates from two patients ( Fig 1 ) . As with the molds , different patients harbored distinct species of yeasts . C . albicans was the most frequent species recovered , with 1 , 056 isolates from 23 patients ( Fig 1 ) . Furthermore , C . albicans was the dominant fungal species in 16 of the 23 patients . We recovered 162 C . parapsilosis isolates from nine patients and it was the dominant fungal species in four of the nine patients . We recovered 129 C . glabrata isolates from three patients and it was the dominant fungal species in two of the three patients . We also recovered 24 C . tropicalis isolates from a single patient ( Fig 1 ) . We compared the relative abundances of different fungal species identified from the cultured isolates to the relative abundances of ITS1 reads from the corresponding samples and found that the results were high correlated at both species and genus level ( Fig 1 and S1 Fig , Pearson correlation 0 . 70 and 0 . 78 , respectively ) . However , we were not able to culture some fungi that showed greater than 10% relative abundances based on ITS1 sequencing of specific sputum samples , including Cochliobolus species , E . dermatitidis , A . nidulans , K . marxianus , G . argillacea , and Malassezia species , many of which require specific growth conditions that were not utilized in our study ( Fig 1 ) [31 , 32] . ITS1 sequencing also had limitations relative to the culture-based approach as it did not detect Clavispora lusitaniae ( two isolates out of 25 from CF058 ) and C . tropicalis ( 24 out of 254 from CF170 ) ( S1 File ) . We measured the fluconazole susceptibility of the full set of yeast isolates and the itraconazole susceptibility of the 182 mold isolates . Of the 1 , 421 yeast isolates from 26 patients , 362 isolates from 16 patients showed resistance to a high concentration of fluconazole ( 128 μg/ml ) ; 20 of these resistant isolates were recovered from three of the eight patients who were treated with azoles , but there was no correlation between the recovery of resistant isolates and prior azole treatment ( Fig 2A , S8 Fig , S1 Table , and S1 File ) The majority of C . glabrata and C . tropicalis isolates were resistant to fluconazole , and most C . parapsilosis isolates were susceptible ( S1 File ) . Of the 182 mold isolates from 10 patients , most were susceptible to a fixed concentration of itraconazole ( 0 . 5 μg/ml ) . Furthermore , all A . fumigatus isolates and most A . flavus isolates were susceptible to a fixed concentration of amphotericin B , but most A . terreus isolates were resistant ( Fig 2B and S9 Fig ) . We examined the growth profiles of the 1 , 421 yeast and 182 mold isolates in liquid medium by measuring changes in OD595 over time and calculating area under the curve as the output of growth . When growth of the yeast isolates was compared to the reference C . albicans strain SN95 , most showed similar profiles except for C . parapsilosis isolates , which consistently grew less than other isolates ( S1 File ) . Furthermore , we identified some C . albicans isolates that showed aberrant growth kinetics that was attributable to filamentous growth ( S10 Fig , CF170-P2C11 and S1 File ) . When growth of the mold isolates was compared to the reference A . fumigatus strain Af293 , most showed consistently enhanced growth relative to the reference strain ( S11 Fig and S1 File ) . We examined several additional phenotypes for the mold and yeast species . For the molds we monitored color of conidia to support species assignments . For the yeasts , we monitored cellular morphology of the isolates that showed aberrant growth in rich liquid medium at 30°C , and identified a striking example of intra-species variation in morphology ( S10 Fig CF170-P2C11 and S1 File ) . Of the 1 , 056 C . albicans isolates , 25 isolates from five patients exhibited filamentous growth at 30°C in rich medium ( Fig 2A , S1 File ) : in CF028 , seven out of the 66 C . albicans isolates were filamentous; in CF033 , two out of 172 C . albicans isolates were filamentous; in CF066 , three out of 114 C . albicans isolates were filamentous; in CF170 , seven out of 229 C . albicans isolates were filamentous; and in CF198 , six out of 43 C . albicans isolates were filamentous ( Fig 2A ) . We also identified three C . parapsilosis isolates from patient CF107 showed filamentous growth at 30°C in rich medium ( S12A Fig ) . Given that the morphological transition between yeast and filamentous growth is normally a tightly controlled developmental program that profoundly impacts on host adaptation and virulence [33 , 34] , we asked if there was a genetic basis for the altered regulation of this program in the isolates that showed a filamentous phenotype in the absence of any inducing cue . We performed whole-genome sequencing of three C . albicans isolates from patient CF170 , one with the filamentous phenotype under standard conditions , designated F1 , and two that grew as yeast , designated Y1 and Y2 ( Fig 3A ) . Sequence reads were aligned to the published C . albicans genome ( SC5314 , assembly 21 , with the mean depth of coverage being 50X for all assembled sequences ) [35] . Unique single nucleotide variants present in F1 were identified by comparing the F1 genome assembly to the assemblies of Y1 and Y2 genomes using MuTect [36] ( S2 Table ) . We identified 76 high-confidence single nucleotide variants in F1: 37 were in non-coding regions; 20 were in coding regions and resulted in synonymous heterozygous mutations; 18 were in coding regions and resulted in non-synonymous heterozygous mutations; and one variant manifested as a homozygous non-synonymous mutation in NRG1 [37] . This last mutation is compelling , as NRG1 encodes a transcription factor that is known to repress filamentation ( S2 Table ) . The mutation ( D271N ) was located within the C2H2 zinc finger domain of NRG1 [37] , potentially impairing the ability of this transcriptional repressor of filamentation to bind DNA ( Fig 3B ) . Next , we functionally validated that the mutation in NRG1 conferred the filamentous phenotype . Since we anticipated that the D271N allele would confer a loss of function and would thus be recessive , we deleted one allele of NRG1 in Y1 to generate the Y1 NRG1Y1/nrg1Δ strain; this Y1 NRG1Y1/nrg1Δ strain had the same yeast growth morphology as Y1 ( Fig 3C ) . We then replaced the remaining NRG1Y1 allele with the D271N allele to generate the Y1 NRG1F1/nrg1Δ strain; this Y1 NRG1F1/nrg1Δ strain had a filamentous growth phenotype akin to that observed in F1 . This demonstrates that the NRG1 allele of F1 is sufficient to confer the filamentation phenotype on Y1 . To further confirm that the NRG1 D271 allele is necessary for the filamentation phenotype of F1 , we replaced one allele of NRG1 in strain F1 with the allele from Y1 to generate the F1 NRG1F1/NRG1Y1 strain; this F1 NRG1F1/NRG1Y1 strain lost the filamentation phenotype , confirming that the NRG1 mutation is recessive and that it is necessary for the filamentation phenotype of F1 . The phenotype of the strains harboring only the F1 allele of NRG1 was comparable to that of an nrg1Δ/nrg1Δ homozygous deletion mutant and an NRG1F1/nrg1Δ mutant in independent laboratory strain background ( Fig 3C ) , consistent with the model that the NRG1 mutation identified in F1 causes loss of function of this transcriptional repressor of filamentous growth . To determine if the phenotypes of the remaining 24 filamentous C . albicans isolates that were recovered from five patients were also associated with mutations in NRG1 , we sequenced NRG1 from these isolates . We identified five different homozygous mutations unique to the filamentous isolates ( Fig 4 ) . From patient CF170 , we recovered six additional filamentous C . albicans isolates; five of these six isolates had homozygous D271N mutations , suggesting persistence of this genotype in the patient . The remaining filamentous isolate ( F2 ) did not have a mutation in NRG1 or in the surrounding regions , and whole genome sequencing did not reveal any homozygous mutations or strong candidate mutations that would confer the filamentous phenotype ( S3 Table ) . The seven filamentous isolates from patient CF028 each had a homozygous Q118* nonsense mutation in NRG1 , and both filamentous isolates from CF033 had a homozygous Y138* nonsense mutation in NRG1 . Two different mutations were identified from the three filamentous isolates recovered from patient CF066 , two with a homozygous C233Y mutations in NRG1 , which is within the C2H2 zinc finger domain , and one with a homozygous W260* mutation in NRG1 . Finally , all six filamentous isolates from patient CF198 had a homozygous 247 bp deletion in NRG1 , leading to a frame shift and introduction of a premature stop codon ( Fig 4 ) . Since all of the observed mutations in NRG1 were located before or within its C2H2 zinc finger domain , the filamentous growth phenotypes of these isolates were likely due to loss of function of Nrg1 . Given the prevalence of mutations in NRG1 in the filamentous C . albicans clinical isolates , we asked whether NRG1 might also be central to the filamentous phenotype that we observed in three C . parapsilosis isolates . Indeed , we observed that these isolates each harbored a homozygous mutation in a highly conserved region of the CpNRG1 gene ( R270K ) , suggesting that perturbation of Nrg1 function is a conserved mechanism for pathogenic yeast that allows filamentation in the absence of inducing cues in clinical isolates from cystic fibrosis patients ( S12 Fig ) . P . aeruginosa and B . cenocapacia can repress filamentation and metabolism of C . albicans in co-culture conditions via secretion of quorum sensing molecules [20 , 21] . Given that four out of five patients that produced filamentous C . albicans isolates ( CF028 , CF033 , CF066 , and CF198 ) were heavily colonized with P . aeruginosa and patient CF170 was colonized with Burkholderia multivorans , which is closely related to B . cenocepacia [40] , we tested whether the filamentous C . albicans isolate F1 could still form filaments under the repressing conditions of co-culture with P . aeruginosa or B . multivorans , or in the presence of a structural analog of pyocyanin , phenazine methosulfate ( PMS ) . C . albicans isolate Y1 showed the expected smooth colony morphology under standard conditions that promote yeast growth , and the characteristic wrinkled colony morphology under filament-inducing condition ( Fig 5 ) . Microscopy images of cells collected from the colonies confirmed that smooth colonies were composed of cells with yeast morphology , and wrinkled colonies were composed of cells with filamentous morphology ( Fig 5 ) . In contrast , the F1 filamentous C . albicans isolate from CF170 showed wrinkled colony morphology under both standard and filament-inducing conditions ( Fig 5 ) . Under filament-inducing conditions , the addition of 5 μM PMS or co-culture with P . aeruginosa reference strain PA14 or B . multivorans reference strain ATCC17616 inhibited filamentation of Y1 , but not of F1 ( Fig 5 and S13 Fig ) . This resistance to filamentation-repressing effects of PMS and P . aeruginosa was not specific to F1 , but was a general feature of all filamentous C . albicans isolates with mutation in NRG1 ( S14 Fig ) and an independently generated nrg1Δ/nrg1Δ deletion mutant in a different strain background ( S15 Fig ) . Thus , filamentous C . albicans isolates are resistant to the filamentation-repressing effects of bacterial opportunistic pathogens .
Here , we provide a portrait of species and population level dynamics in the lung mycobiome of cystic fibrosis patients , and illuminate the power of complementary high-throughput sequencing coupled with phenotypic and genotypic analyses . Utilizing high-throughput sequencing of the ribosomal RNA internal transcribed spacer ITS1 amplified from longitudinal sputum samples from 28 cystic fibrosis patients ( Fig 1 ) , we were able to identify fungi that were not isolated through culture-based methods , likely due to their specific growth requirements as with lipophilic Malassezia species [31] and slow growing E . dermatitidis [32] . Our culture-based analysis of 1 , 603 fungal isolates from the same sputum samples revealed extensive inter- and intra-species phenotypic diversity in growth rate , drug resistance , and cellular morphology ( Figs 2–4 ) , with profound implications for inter-kingdom microbial interactions and pathogen adaptation to the host ( Fig 5 ) . The strong correlation between our molecular and culture-based methods to define fungal communities ( Fig 1 ) provides strong validation of our approach to study species and population level dynamics in the host . Although studies of the lung mycobiome are in their infancy , there is a growing appreciation that fungi can modulate clinical outcome in the context of chronic respiratory diseases such as cystic fibrosis [1] . We found that different patients harbor distinct fungal communities , although the dominant trend was stable colonization with Candida and Aspergillus species ( Fig 1 , S6 and S7 Figs ) ; this is consistent with previous analyses of the cystic fibrosis mycobiome [6 , 7 , 13] , and distinct from findings with the lung mycobiome of healthy individuals [12] . Neither fungal diversity nor community structure in our patient population was correlated with patient characteristics such as BMI , FEV1% predicted values , clinical status , lung disease stage , or antifungal treatments ( S2–S7 . Figs ) , consistent with a prior study of the bacterial and fungal microbiome of adult cystic fibrosis patients [13] . However , we did detect a weak yet significant negative correlation between relative abundance of C . albicans and patient FEV1% predicted values ( S5 Fig ) , consistent with a previous association between chronic C . albicans colonization and FEV1 decline in some cystic fibrosis patients [11] , reinforcing the relevance of C . albicans colonization in context of cystic fibrosis . Our finding that the lung mycobiome of adult cystic fibrosis patients is relatively stable over longitudinal sampling periods with little perturbation in response to changes in patient physiology or antibiotic therapies resonates with the emerging theme that bacterial diversity remains relatively stable in cystic fibrosis patients over the course of clinical exacerbation and treatment [13 , 41] . It is well established that bacterial communities of cystic fibrosis patient lungs differ significantly between adolescents and adults [3] , with diversification events prevalent in younger patients and community specialization accompanying the deterioration of pulmonary function with age [42 , 43] . Thus , analysis of fungal microbiomes in age-stratified cystic fibrosis patients may reveal more dynamic communities and changes associated with disease progression . Despite relatively stable fungal communities based on taxonomic identities , we identified considerable inter- and intra-species phenotypic variation in traits important for host adaptation . Variation in antifungal drug resistance largely tracked with species identity , and there was no evidence for the evolution of antifungal drug resistance in response to drug treatments ( Fig 2 , S8 Fig ) . This stands in contrast to the rapid emergence of drug resistance in fungal populations in the host that has been observed with Candida species from patients with AIDS or Crohn’s disease [15 , 16] , and with the high frequency of azole-resistant A . fumigatus recovered from cystic fibrosis patients treated with azoles [44 , 45] . This may reflect differences in antifungal treatment regimens , pathogen proliferations rates , or pathogen population sizes in these patients , which can influence the selection pressure for resistance and the probability of accumulating resistance mutations . The most striking phenotype that we identified was the uncoupling of morphogenesis in Candida species from regulation by standard inducing cues . The capacity to transition between yeast and filamentous growth is a key virulence trait for C . albicans , with most mutants that are unable to transition being attenuated in virulence . The current paradigm is that filaments are responsible for tissue invasion and escape from immune cells , while yeasts are critical for dissemination [46] Filaments are further implicated in virulence as they express virulence factors such as adhesins and proteases [47 , 48] . Our discovery of 25 C . albicans isolates that filament in the absence of inducing cues ( Fig 2A ) , suggests that there may be a fitness advantage to enabling altered regulation of morphogenesis . That this trait emerged independently in C . albicans recovered from multiple patients and in C . parapsilosis ( S12 Fig ) but has not been reported in other patient populations , underscores that this may be a prevalent adaptation to the cystic fibrosis lung environment . Inter-kingdom interactions in this environment may provide the selective pressure driving this adaptive change , as dominant bacteria such as P . aeruginosa and Burkholderia species repress C . albicans filamentation [18–21] , and the filamentous isolates we recovered are resistant to the repressive effects on morphogenesis ( Fig 5 , S13 and S14 Figs ) . The intra-species phenotypic diversity in fungal morphology complements the extensive diversity observed in bacterial populations infecting in the cystic fibrosis lung [49–52] , and suggests extensive genetic variation in microbial populations that can enable adaptive evolution in the host . Adaptation of microbial pathogens in response to host selective pressures over the course of chronic infections manifests in genetic signatures in pathogen populations . This is best appreciated in the context of bacterial pathogens , where adaptation of P . aeruginosa and Burkholderia species during chronic pulmonary infections has been accompanied by the acquisition of adaptive mutations [49–54] . It is clear that many mutations remain polymorphic in the bacterial populations , suggesting limited clonal selective sweeps . Our analysis of the genetic basis of the filamentation phenotype of Candida isolates suggests a similar trend in fungal populations . Twenty four of the 25 filamentous C . albicans isolates and all three filamentous C . parapsilosis isolates contained homozygous mutations in the transcriptional repressor of filamentation , NRG1 ( Figs 3 and 4 , S12 Fig ) . The `filamentous isolates recovered from a single patient most often all shared the same nrg1 allele , although there was one patient that harbored C . albicans isolates with distinct homozygous nrg1 mutations ( Fig 4 ) . This suggests that the filamentation phenotype emerged independently within individual patients , and that the loss of function of NRG1 is a common mechanism of adaptation to the cystic fibrosis lung environment . Despite the potential fitness advantage of retaining the capacity to filament in the presence of bacterial pathogens , the frequency of filamentous isolates remained low . This may reflect a bias in culturing the filamentous isolates if they remain embedded in tissues thereby minimizing recovery from sputum samples . Alternatively , it could reflect that the magnitude of fitness benefit is small or that fungal proliferation is limited such that the mutants would not sweep to fixation . Yet another possibility is that loss of function of NRG1 may confer niche-specific fitness advantages , consistent with the observed heterogeneity in bacterial populations that has been attributed to spatial and temporal heterogeneity in the cystic fibrosis lung environment [55–58] . Experimental evolution studies clearly demonstrate that structured physical , nutritional , and cooperative niches can select for the evolution of distinct populations [59–61] . Our finding that the majority of filamentous isolates harbored mutations in NRG1 rather than loss-of-function mutations in other transcriptional repressors such as TUP1 or RFG1 [37 , 62 , 63] , suggests that there may be distinct fitness consequences of mutation of these different transcriptional repressors of filamentation . The complexity of microbial evolution in the host is likely to be exquisitely contingent upon interactions among constituents of the microbiome , and reflects the stunning diversity of adaptive strategies in biological systems .
Protocols for the collection and use of cystic fibrosis patient sputum were approved by the Research Ethics Boards of St . Michael's Hospital and the University Health Network . Informed consent was obtained from each study subject and all sputum specimens were produced voluntarily . Experiments involving patient specimens were conducted in accordance with the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans , of the Canadian Institutes of Health Research ( CIHR ) , the Natural Sciences and Engineering Research Council of Canada ( NSERC ) and the Social Sciences and Humanities Research Council of Canada ( SSHRC ) . The ITS1 region of the 18S-ITS1-5 . 8S-ITS2-28S rRNA complex was amplified from the DNA extracted from sputum using the ITS1F and ITS1R PCR primer set [13] . PCR reactions were performed in triplicate and pooled . The samples were cycled at 95°C for 3 minutes , 30 cycles of 95°C for 15 seconds , 56°C for 15 seconds , 72°C for 15 seconds , followed by a final extension at 72°C for 5 minutes . The amplicons were cleaned using AMPure XP magnetic beads ( Beckman Coulter , Inc . ) and prepared for sequencing using the Nextera XT DNA library preparation kit following the manufacturer instructions ( Illumina , Inc . ) . Samples were pooled and size selected on a 1% TAE ( w/v ) agarose gel . The library was sequenced on the Illumina Miseq using a 150x2 PE sequencing kit . Following the UPARSE pipeline , the sequencing reads were assembled , quality filtered and dereplicated [64] . Operational Taxonomic Units ( OTUs ) were then clustered into groups of ≥97% sequence identity and chimeras were removed . Taxonomic identity was assigned to the OTUs using BLASTn and the FHiTINGS v . 1-2 reference database [65 , 66] . The BLASTn results were input into a modified version of the FHiTINGS program to identify the taxonomy based on BLAST hit frequency , e-value scores and common taxonomic ancestors . The taxonomies and the OTU abundances were converted into an OTU table using the biom-format V1 . 3 . 1 software [67] . OTUs with <0 . 005% relative abundances were removed [68] . For downstream analyses QIIME ( Version 1 . 9 . 0 ) was used [67] . Raw relative abundances were used to generate taxonomic abundance plots then the reads were rarefied at 1 , 000 reads for calculating Simpson’s diversity index and Bray-Curtis dissimilarity index . Principal coordinate plots were generated for sputum samples from patients with more than one sampling time point . Downstream statistical analyses were performed on R ( Version 3 . 1 . 0 ) . Sputa were collected by expectoration and transported to the laboratory on ice . Sputa were solubilized by homogenization with Sputolysin ( EMD Millipore ) and separate aliquots were used for mycobiome and fungal analysis . To culture fungus , duplicate aliquots of each sample were plated directly onto Sabouraud Dextrose Agar ( SDA ) ( Becton Dickinson ) supplemented with 50 μg/mL ampicillin ( Sigma ) and 50 μg/mL kanamycin ( Sigma ) without serial dilution . All cultures were incubated at 37°C for 48 h . Following incubation , colonies were presumptively identified as fungus by visual inspection and up to 50 colonies were selected from each sample for further analysis by morphology . If there were less than 50 fungal colonies from plating a sputum sample , all colonies were included in the sampling . If there were more than 50 fungal colonies from plating a sputum sample , an effort was made to maintain the relative abundance of each colony morphology in the sampling . Fungal isolates were then cryopreserved at -80°C in 25% glycerol ( v/v ) after a single subculture in LB broth ( Wisent Inc . ) . Mold isolates were grown on individual potato dextrose agar ( PDA ) plates containing 100 μg/ml ampicillin ( BioShop ) and 50 μg/ml gentamicin ( BioShop ) and incubated at 37°C for up to 72 hours until robust sporulation was observed . Spores were harvested by gently washing the plates with sterile water and passing through Miracloth filter . After pelleting the spores and re-suspending in sterile water , optical density at 600 nm ( OD600 ) was measured using a spectrophotometer . Spore suspensions were diluted to the final OD600 of 0 . 01 in 100 μl of RPMI1640 or 100 μl of RPMI1640 containing 0 . 5 μg/ml itraconazole ( Sigma ) or 1 μg/ml amphotericin B ( Sigma ) in clear 96-well plate , covered with clear , adhesive seal ( Thermo Scientific ) . Growth was measured by OD595 inside GENios microplate reader ( TECAN ) every 15 min for 48 hours at 37°C . Species profiles were generated by the combination of morphological characterization and ITS2 sequencing of representative isolates . The ITS2 locus was PCR amplified directly from spore suspensions with primers oLC2459 and oLC2460 . Reaction mixtures contained 1x PCR buffer , 0 . 25 mM deoxynucleotide triphosphates ( dNTPs ) , 0 . 5 mM primers , 1 unit of Taq polymerase , 5 μl of spore suspension , and sterile water up to 20 μl . Cycling conditions were 98°C 2 min; 98°C 20 s , 55°C 20 s , and 72°C 20 s for 30 cycles; and 72°C 2 min . PCR products were visually confirmed by gel electrophoresis and purified using PCR cleanup kit ( Sigma ) . Purified products were sent for Sanger sequencing with 100 ng of product and 7 . 14 μM of oLC2459 at TCAG sequencing facility ( Toronto , ON ) . Using BLAST [66] , ITS2 sequences were queried against NCBI nucleotide database and species identity was assigned based on the highest bit score sequence . Once species identity was assigned , all isolates from the same patient with the same growth morphologies were classified as the same species . Yeast isolates were grown on individual yeast extract peptone dextrose ( YPD: 1% yeast extract , 2% bactopeptone , 2% glucose ) plates containing 100 μg/ml ampicillin and 50 μg/ml kanamycin ( BioShop ) and incubated for 24 hours at 30°C . Yeast isolates were subsequently sub-cultured in 200 μl of YPD in clear 96-well plate overnight at 30°C . Overnight cultures were diluted 20 , 000 fold into 100 μl of YPD in clear 96-well plates and covered with clear , adhesive seals . Growth was measured by OD595 every 15 min using a GENios microplate reader with rotation at 800 rpm for 36 hours at 30°C . The same overnight cultures were diluted 20 , 000 fold into 200 μl of YPD and 200 μl YPD containing 128 μg/ml fluconazole ( Sequoia Research Products ) in clear 96-well plates . Final growth was measured by OD600 after 48 h at 30°C using a spectrophotometer ( Molecular Devices ) . Species profiles were generated using both CHROMagar Candida ( BD , 254093 ) and ITS2 sequencing of representative isolates . All isolates were spotted on CHROMagar Candida for identification of the most prevalent Candida species ( C . albicans , C . tropicalis , C . glabrata , C . parapsilosis ) [28] . The ITS2 locus was PCR amplified from isolates representative of each species as with the mold isolate screening method above , but colony suspensions were used instead of spore suspensions . All Candida isolates were typically grown in YPD medium overnight at 30°C in shaking conditions , unless stated otherwise . Solid media plates were supplemented with 1% agar . The effects of PMS ( Sigma ) on C . albicans filamentation were tested as described in Morales et al . , 2013 [20] . The non-inducing condition is YNB ( BioShop ) + 10 mM Glucose ( BioShop ) + 0 . 2% amino acids ( BioShop ) at 30°C and the inducing condition is YNB + 10 mM Glucose + 0 . 2% amino acids 5 mM N-acetylglucosamine ( BioBasic ) at 37°C for 48 hours . Cells were visualized by imaging 5 μl of cultures at specific time points or suspended cells from colonies using DIC microscopy ( Zeiss Axio Imager . MI , Carl Zeiss ) . Colonies were visualized using an M2 Discovery Stereomicroscope ( Carl , Zeiss ) . Sputum DNA was extracted using MasterPure Yeast DNA Purification Kit ( Epicentre ) following the manufacturer’s instructions with a minor modification . To physically disrupt fungal cells , 50 μl of sputum was mixed with 300 μl Yeast Lysis Solution and 200 μl acid-washed glass beads and homogenized in a bead beater for 3 min . Mock preparations were performed in parallel with sputum samples in order to ensure that the reagents were free of contaminating fungal DNA . DNA was quantified using Quant-iT PicoGreen dsDNA Assay Kit ( Life Technologies ) following the manufacturer’s protocol . Fluorescence was measured with excitation wavelength at 480 nm and emission wavelength at 520 nm on the spectrophotometer ( Molecular Devices ) . DNA concentration was calculated based on the standard curve . Cell pellets of clinical isolates were prepared by centrifuging 20 ml of overnight culture at 3 , 000 rpm for 5 min and flash frozing using dry ice and ethanol . Sequencing libraries were prepared using the Nextera XT Kit ( Illumina ) according to the manufacturer’s instructions . Libraries were sequenced on the Illumina MiSeq platform using paired reads ( 150 bp ) . The sequence reads were de-multiplexed and trimmed to remove bases with Phred scores < Q30 . Reads were aligned to SC5314 reference genome using Bowtie2 ( Version 2 . 0 . 7 ) [69] , and the alignment was visualized using Savant Genome Browser [70] . MuTect ( Version 1 . 1 . 4 ) [36] was used to identify unique mutations in filamentous C . albicans isolates . MuTect provides accurate variant detection in diploid genomes based on a model that takes into account the matched normal ( not mutated ) DNA , as well as sequencing errors and allele fractions . The sequence data is publicly available on the NCBI Sequence Read Archive with accession number SRX1084067 . Mutations in NRG1 were further confirmed by PCR amplifying NRG1 from genomic DNA ( gDNA ) of clinical isolates using primers oLC3080 and oLC3282 . Reaction mixtures contained 1x PCR buffer , 0 . 25 mM dNTPs , 0 . 5 mM primers , 1 unit of Taq polymerase , 100 ng of gDNA , and sterile water up to 20 μl . Cycling conditions were 98°C 2 min; 98°C 20 s , 55°C 20 s , 72°C 20 s for 30 cycles; and 72°C 2 min . PCR products were visually confirmed by gel electrophoresis and purified using a PCR cleanup kit ( Sigma ) . Purified products were sent for Sanger sequencing with 100 ng of product and 7 . 14 μM of oLC3080 and oLC3282 at TCAG sequencing facility . E . coli DH5α competent cells were used for plasmid construction . Strains of bacteria and C . albicans that were used engineered in this study are listed in S4 Table and oligonucleotides in S5 Table . To construct a cassette for NRG1 allele replacements , part of the NRG1 open reading frame spanning the polymorphism identified and the NRG1 downstream region were amplified from C . albicans strain Y1 and F1 gDNA , using oLC3093/oLC3155 ( 768 bp ) and oLC3094/oLC3095 ( 544 bp ) . The amplified region of the open reading frame was cloned into pLC49 [71] at KpnI and ApaI , which is upstream of the nourseothricin ( NAT ) resistance marker and FLP recombinase . The NRG1 downstream region was then cloned in at SacI/SacII , which is downstream of the NAT resistance marker and FLP recombinase . The presence of the NRG1 open reading frame and the NRG1 downstream region was tested by PCR with oLC275/M13R ( 970 bp ) and oLC274/M13F ( 736 bp ) , respectively . The final constructs were sequence verified using the same set of primers used to generate the amplicons . The construct to integrate the Y1 NRG1 allele is pLC796 and the construct to integrate the F1 NRG1 allele is pLC798 . Bacteria harbouring the plasmid are propagated with ampicillin ( 100 μg/ml ) and nourseothricin ( 250 μg/ml ) ( NAT , Werner BioAgents ) . Using KpnI and SacI , the cassette was liberated for transformation . For C . albicans transformations , 1 ml of an overnight YPD culture with an OD600 of between 4–8 was used . Cells were pelleted and resuspended with the following transformation mixture: 1 . 5 μg of digested DNA , 40% polyethylene glycol ( PEG ) , 1X Tris-EDTA , 100 mM lithium acetate , pH 7 . 4 , 10 mg/ml of salmon sperm DNA , and 20 mM dithiothreitol ( DTT ) . The mixture was incubated at 30°C for 1 hour and 42°C for 45 min . Cells were washed with 1 ml YPD and re-suspended in 10 ml YPD and allowed to recover at 30°C shaker for 4 hours . Transformants were selected for on YPD plates containing NAT and incubated at 30°C for 48 hours . The FLP recombinase was induced to excise the NAT cassette by growth in yeast nitrogen base bovine serum albumin ( YNB-BSA ) medium at 30°C with shaking for 48 hours . Approximately 100 cells were plated on a YPD plate and incubated at 30°C for 48 h , then replica plated onto YPD + NAT and incubated at 30°C overnight to identify NAT-sensitive colonies . The Y1 NRG1/nrg1Δ mutant was generated by PCR amplification of the NAT-FLP cassette from pLC49 [71] with primers oLC3112/oLC3113 ( 4366 bp ) , which contain homology to precisely replace the NRG1 open reading frame . Upstream and downstream integration of the cassette were tested by PCR using oLC275/oLC3080 ( 635 bp ) and oLC274/3155 ( 357 bp ) , respectively . The NAT marker was excised and the genotype verified by PCR using oLC3080/oLC3155 ( 833 bp deletion allele and 1510 bp native allele ) . | Microbial cells vastly outnumber human cells in our bodies , yet we are only beginning to understand how these microbes influence human health and disease . One disease for which microbial communities are especially important is cystic fibrosis , where persistent lung infections can be lethal . Fungi are associated with poor respiratory function , but how fungal communities change with disease progression or treatment remains enigmatic . Here , we assess the dynamics of fungal communities by combining high-throughput sequencing of sputum samples from 28 patients with detailed analysis of phenotypes and genotypes of 1 , 603 fungal isolates . We found stable communities dominated by Candida and Aspergillus , and diversity in traits important for host adaptation . Antifungal drug resistance varied largely between species , while morphogenesis varied within species . For Candida species , the capacity to transition between yeast and filaments is a key virulence trait that is normally regulated by inducing cues , however , 28 isolates grew as filaments without such cues . Filamentation was due to loss-of-function mutations in the transcriptional regulator NRG1 in most isolates , which conferred resistance to the filament-repressive effects of a common bacterial pathogen . This work provides a portrait of the fungal microbiome associated with a lethal disease , and illuminates a genetic basis of pathogen adaptation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Global Analysis of the Fungal Microbiome in Cystic Fibrosis Patients Reveals Loss of Function of the Transcriptional Repressor Nrg1 as a Mechanism of Pathogen Adaptation |
Chagas disease affects around 18 million people in the American continent . Unfortunately , there is no satisfactory treatment for the disease . The drugs currently used are not specific and exert serious toxic effects . Thus , there is an urgent need for drugs that are effective . Looking for molecules to eliminate the parasite , we have targeted a central enzyme of the glycolytic pathway: triosephosphate isomerase ( TIM ) . The homodimeric enzyme is catalytically active only as a dimer . Because there are significant differences in the interface of the enzymes from the parasite and humans , we searched for small molecules that specifically disrupt contact between the two subunits of the enzyme from Trypanosoma cruzi but not those of TIM from Homo sapiens ( HTIM ) , and tested if they kill the parasite . Dithiodianiline ( DTDA ) at nanomolar concentrations completely inactivates recombinant TIM of T . cruzi ( TcTIM ) . It also inactivated HTIM , but at concentrations around 400 times higher . DTDA was also tested on four TcTIM mutants with each of its four cysteines replaced with either valine or alanine . The sensitivity of the mutants to DTDA was markedly similar to that of the wild type . The crystal structure of the TcTIM soaked in DTDA at 2 . 15 Å resolution , and the data on the mutants showed that inactivation resulted from alterations of the dimer interface . DTDA also prevented the growth of Escherichia coli cells transformed with TcTIM , had no effect on normal E . coli , and also killed T . cruzi epimastigotes in culture . By targeting on the dimer interface of oligomeric enzymes from parasites , it is possible to discover small molecules that selectively thwart the life of the parasite . Also , the conformational changes that DTDA induces in the dimer interface of the trypanosomal enzyme are unique and identify a region of the interface that could be targeted for drug discovery .
Triosephosphate isomerase ( TIM ) is a ubiquitous enzyme that catalyzes the interconversion between glyceraldehyde 3-phosphate and dihydroxyacetone phosphate . In most of the species the enzyme is formed by two identical monomers of approximately 250 amino acids . TIM belongs to the family of α-β barrels proteins , in which 8 central β strands are surrounded by 8 α helices; the strands and helices are joined by loops . It is one of the most thoroughly studied enzymes . Its kinetics are well established [1 , 2] , the crystal structure of the enzyme from 15 different species is available , and significant advances have been made on the dynamics of the enzyme when it is in the resting state and during active catalysis [3–5] . A peculiarity of TIM is that only in its dimeric form the enzyme exhibits high catalytic rates , albeit each monomer has its own catalytic residues [6–8] . Along this line , it has been reported that deletion of some residues of loop3 in TIM from Trypanosoma brucei ( TbTIM ) , which forms an important portion of the interface , yields a monomeric enzyme with drastically reduced catalytic activity [9 , 10] . Likewise , it has been shown that chemical perturbation by thiol reagents of the interfacial Cys15 of TbTIM , and that of TIMs from T . cruzi ( TcTIM ) , Leishmania mexicana ( LmTIM ) [11] , Plasmodium falciparum [12] , and Entamoeba histolytica [13] induces drastic changes in the quaternary and tertiary structure of the respective dimers and abolition of catalytic activity . The latter observations raised the question as to whether agents that interfere with protein-protein interactions in either permanent of transient oligomers , could be exploited for the discovery of molecules with pharmacological potential [14 , 15] . From the point of view of drug discovery for diseases that are caused by parasites , the possibility is particularly attractive . This is because in the course of evolution the catalytic site of enzymes has been largely conserved , whereas the amino acids that form the interfaces of oligomeric enzymes have undergone significant changes [16–18] . For example , TIMs from human and the aforementioned parasites have the same catalytic residues; in contrast , the identity of the approximately 32 interfacial residues of TIM from either of the parasites and human is approximately 52% , whereas the identity of the amino acid residues of the interface of the enzymes from T . cruzi , T . brucei and L . mexicana is approximately 82% [8 , 15] . Therefore , it is theoretically possible to find molecules that exhibit a high specificity for the interface of oligomeric enzymes from parasites . Based on the amino acid composition and structural differences of the interfaces of TIM , we have found agents that by acting on the dimer interface exhibit a high specificity for TIM from parasites [15] . In our attempts to find more effective low molecular weight compounds , we synthesized 3- ( 2-benzothiazolylthio ) -1 propanethioaniline . We found that it induced a strong inhibition of enzyme activity . However , subsequent studies showed that the inhibition was not due to this compound , but to a contaminant that formed during its preparation . This proved to be 2 , 2′-dithiodianiline , ( DTDA ) . Here we describe the action of DTDA on the function and structure of TcTIM . The compound is highly selective for TcTIM . We also found that DTDA inhibited the growth of Escherichia coli that had been transformed with TcTIM , and that in the low micromolar range , the compound also caused the death of T . cruzi epimastigotes . The crystal structure of the enzyme treated with DTDA complex showed that it induced important and rather unique alterations of the dimer interface .
A solution of 1 g of 2-aminothiophenol in 10 ml dichloromethane was stirred at room temperature for 8 h to promote its oxidation . The residual material was washed with hexane and purified by silica gel chromatography with dichloromethane as eluent to give 0 . 97 g of DTDA . The product was characterized by mass spectrometry and NMR analysis . DTDA is spontaneously formed from 2-aminothiophenol and its reactivity is typical of disulfides . Its reaction with sulfides , such as cysteine , yields a mixed disulfide ( Fig . 1 ) . Indeed , as shown in the Results section , we found that Cys118 of wild type TcTIM reacts with DTDA forming a disulfide bond between Cys118 and thioaniline . The reaction would be analogous to the standard opening of disulfide bridges by reaction with an excess of β-mercaptoethanol or dithiothreitol . In all the experiments , a solution of DTDA in dimethylsulfoxide ( DMSO ) was used . The final DMSO concentration in all experiments was 10% ( v/v ) . It is noted that that at this concentration , DMSO did not affect the activity of the enzymes that were used . The indicated TIMs were incubated at pH 7 . 4 at a concentration of 5 µg per ml of 100 mM triethanolamine , 10 mM EDTA , 10% dimethyl sulfoxide ( v/v ) , and DTDA at the concentrations indicated in the Results section for 2 hours at 36°C . At this time , an aliquot was withdrawn for assay of activity . Activity was determined in the direction of glyceraldehyde 3-phosphate to dihydroxyacetone phosphate [11] . The decrease in absorbance at 340 nm was followed in a Hewlett Packard spectrophotometer at 25°C . The reaction mixture ( 1 ml ) contained 100 mM triethanolamine , 10 mM EDTA , 0 . 2 mM NADH , 1 mM glyceraldehyde 3-phosphate , and 0 . 9 units α-glycerolphosphate dehydrogenase ( pH 7 . 4 ) . The reaction was started by the addition of TIM , usually 5 ng . The average specific activity of the various preparations of TcTIM used in this work was 2900±200 µmol/min/mg . E . coli JM103 cells were used as control . E . coli devoid of their endogenous TIM termed VR101 [10] were kindly provided by Dr . Gloria Saab-Rincón; these cells have a kanamicin resistant cassette . The latter cells were transformed with the plasmid pTrc99aTcTIM that had an ampicilin resistant cassette . The strains were grown at 37°C in solid Luria-Bertani medium that had been supplemented with 50 µg of kanamycin and 50 µg of ampicilin per ml . One colony was transferred to M9 medium that had 10% DMSO and 50 µg of each of the latter antibiotics per ml . Growth of the various cells was followed throughout time by measuring the absorbance of the culture at 600 nm . To study the effect of DTDA on the T . cruzi , 106 epimastigotes of the strain ninoa were inoculated into RPMI 1640 media supplemented with 10% fetal bovine serum ( Gibco , BRL , Rockville , Md ) ; the media also had 10% DMSO ( v/v ) ; at this concentration DMSO did not exert a detrimental effect on the growth of E . coli cells , nor on the growth and survival of T . cruzi epimastigotes . Where indicated in the Results section , the media was supplemented with the indicated concentrations of DTDA . The number of cells was recorded at various times for as long as 72 hours . We attempted to co-crystallize TcTIM with DTDA; we tried different concentrations of enzyme and DTDA , but all our attempts were unsuccessful . However , we succeeded in obtaining crystals of the complex by soaking crystals of TcTIM with DTDA . TcTIM was crystallized by the vapor diffusion hanging drop method . TcTIM , 2 . 5 µg in 5 µl of 25 mM triethanolamine ( pH 8 . 0 ) was mixed with 5 µl of reservoir solution ( 0 . 1 M Na-Hepes , pH 7 . 5 , 2% ( v/v ) PEG 400 , and 2 . 0 M ammonium sulfate ) . Crystals appeared after two or three weeks . At this time , 1 µl of 10 mM DTDA was added to the drop ( 1 mM final concentration ) . After 48 hours the crystal was transferred to a cryoprotectant solution ( 30% ( v/v ) glycerol ) and flash frozen . Diffraction data were collected at 113 K with a Rigaku X-ray rotating anode generator and an R-Axis IIC image plate detector . The data was processed and scaled with d*TREK [24] . The 3D structure was solved using molecular replacement with the program MOLREP [25] and the coordinates of native TcTIM ( PDB code 1TCD ) as the search model . Refinement was carried out first with the program CNS [26] , and manual adjustments of the model into electron density maps were done using QUANTA2000 ( Accelrys ) . Five percent of the reflections were set aside for validation . The anisotropic motion of the subunits was described with the TLS parameters as implemented in REFMAC5 [27] . Both the molecular replacement and TLS refinement of the structure were done using the CCP4 program suite version 5 . 2 . 0005 [28] . A summary of the data-collection and refinement statistics is given in Table I . The coordinates of the structure have been deposited in the Protein Data Bank ( PDB code 2OMA ) .
To gain insight into the mechanism through which DTDA affects TcTIM , we attempted to co-crystallize the enzyme in complex with DTDA , however , our efforts were unsuccessful . On the other hand we were able to obtain crystals of the complex by soaking crystals of TcTIM with 1 mM DTDA . The statistics of data collection , reduction and refinement are shown in Table 1 . The general structure of the dimer was not altered by soaking with DTDA ( Fig . 3 , A ) ( Accession Number PDBI code 2OMA ) . The RMS deviation of the Cα traces of native TcTIM and the DTDA treated enzymes was 0 . 39 Å . The data also showed that in the position of Cys118 , which is far from the interface , the two monomers exhibited electron densities that fitted well with a structure in which the sulfur of Cys118 was covalently linked through a disulfide bond to a thioaniline moiety ( Fig . 3 , A and B ) . Thus , in TcTIM crystals , DTDA was able to derivatize Cys118; however , as shown by the data on the Cys118Val mutant , the inhibition of activity by the compound does not depend on the perturbation of Cys118 . Accordingly , and in regard to the inhibition of activity of TcTIM by DTDA , it is relevant that the enzyme soaked with DTDA exhibited significant and rather unique alterations in its dimer interface . One of the most notable was that although loop3 of monomer B exhibited a conformation almost identical to that in native TcTIM , loop3 of monomer A acquired a markedly different conformation ( Fig , 4 , A ) . Loop3 of TcTIM is formed by residues 66–79 ( Q , N , A , I , T , R , S , G , A , F , T , G , E , and V ) . In monomer A of the DTDA treated enzyme , residues 66 to 70 and residues 77 to 79 superpose quite well with those of the “normal” loop , however , the region formed by residues 71 to 76 exhibited a significant displacement ( Fig . 4 , A ) with a hinge at the level of Thr70 and Thr76 . The different conformations that loops3 of monomers A and B adopted in the DTDA treated TcTIM are clearly evident in a superposition of the two ( Fig . 4 , A ) . The change in conformation of loop3 of monomer A was accompanied by alterations in the contacts that it establishes with residues of monomer B . In native TcTIM , the side chain of Cys15 of the two subunits is well defined , and their N and O atoms are respectively , hydrogen bonded to the O and N atoms of Gly73 of the adjacent subunit ( Fig . 4 , B ) . In the DTDA treated enzyme , these H-bonds are missing ( Fig . 4 , C ) . This was consequence of an increase in the distance between Gly73 of monomer A and Cys15 of the other subunit; in the treated enzyme , the distances between N and O atoms of Gly73 and the O and N atoms of Cys15 were 7 . 9 and 9 . 2 Å , respectively . It is noted that the crystallographic data showed a negative electron density in the region of Cys15 of monomer B ( Fig . 3 , C ) . Taken together , the data indicate that the interactions of Cys15 of monomer B with Gly73 of the other subunit are less stable than in the wild type , and thus in all likelihood the inhibition of activity of TcTIM by DTDA is related to the alterations of the dimer interface . Along this line , it is also relevant that the B factors of residues 15 to 21 of monomer B in DTDA treated TcTIM had an average of 42 . 66 Å2 , whereas in monomer A , the average values of the B factors of these same residues were 29 . 48 Å2 . In native TcTIM , the average B factors of monomers A and B were 21 . 77 and 24 . 82 Å2 , respectively . In view of the powerful specific inhibiting effect of DTDA on TcTIM , we considered important to ascertain if the compound is able to cross biological membranes and whether it is detrimental to cells that rely on the presence of TcTIM . Accordingly , we determined the effect of the compound on E . coli cells that possessed their own TIM and in cells that depended on the function of TcTIM . These experiments involved three types of cells: i ) Cells that have their endogenous TIM , ii ) E . coli that lack their TIM , and iii ) E . coli devoid of endogenous TIM that were transformed with TcTIM . Figure 5A shows the growth curve of intact cells in minimal media; the figure also shows that the growth of cells that are devoid of TIM is almost nil . These data therefore , illustrate that in minimal media , TIM is central to cell growth . In this respect it is particularly relevant that the growth of the latter cells was restored when they were transformed with TcTIM ( Fig . 5 . A ) , albeit the lag that precedes logarithmic growth was longer . Thus , the experiments show that the endogenous TIM of E . coli can be successfully replaced by the enzyme from T . cruzi . When the effect of DTDA was assayed on the growth of normal and TcTIM transformed E . coli , it was found that 30 µM of the compound had no effect on cells that worked with their own TIM ( Fig . 5 , B ) . On the other hand , the growth of cells that relied on the function of TcTIM was effectively prevented by 30 µM DTDA ( Fig . 5 , C ) , and that concentrations as low as 4 µM induced an important increase in the lag that precedes logarithmic growth . Clearly , the data indicate that biological membranes are permeable to DTDA and that it can inhibit the activity of intracellular TcTIM . The latter data prompted us to study the effect of DTDA on intact T . cruzi . To this end , 2 ml of RPM media that contained different concentrations of the compound were inoculated with 106 T . cruzi epimastigotes , and incubated for 72 hours . The number of cells in the culture was determined every 24 hours . It was observed that at concentrations higher than 8 µM the compound brought about a significant decrease in the number of cells ( Fig . 6 ) . At lower concentrations ( 4 µM ) , DTDA brought about inhibition of growth . Thus , depending on its concentration , the compound , either prevented cell growth or caused the death of T . cruzi epimastigotes . We would like to point out that there is a difference on the concentrations of DTDA that are effective on the pure enzyme and in whole cells . The former is inhibited by nM concentrations , whereas the adverse effects of the compound on TcTIM transfected E . coli and epimastigotes are observed with concentrations that are about 10 times higher . It is possible that in vivo , the binding of DTDA to the proteins that exist in the intracellular milieu , reduces its effective concentration . Although this phenomenon has been well documented for some pharmacological agents [34] , at the moment it is not possible to offer a precise explanation for the difference in effectiveness of DTDA in vitro and in whole cells .
DTDA is a powerful inhibitor of the activity of TIM from T . cruzi . It is also effective in human TIM , but at concentrations that are nearly 400 times higher . Remarkably , the compound fails to affect the activity of TIM from T . brucei , and L . mexicana , albeit these enzymes are markedly similar to TcTIM in amino acid sequence and three-dimensional structure . An additional salient property of DTDA is that it is able to cross biological membranes as evidenced by the data with E . coli . These experiments also showed that DTDA does not affect the growth of intact E . coli , whereas in cells that depend on the function of TcTIM , low micromolar concentrations induce a strong inhibition of cell growth . These findings thus indicate that cell membranes are permeable to DTDA and that it affects adversely the life of cells that depend on the function of TcTIM . In consonance with these data , it was found that at concentrations of 4–8 µM , the compound induces a significant inhibition of the growth of T . cruzi epimastigotes , and that 10 µM and 15 µM causes death of parasites in cell cultures . Although the overall data suggest that the detrimental effect of DTDA on intact T . cruzi parasites is due to inhibition of the activity of their TIM , the results do not prove unambiguously that death of the parasites is due exclusively to the inhibition of that enzyme . In regard to the properties of DTDA , it is relevant to point out that its effect and that of similar molecules on rodents have been previously reported . For example , it has been reported that the administration of 4 , 4′-diamino disulfide to rats induces hematological alterations [35] , and that DTDA brings about irreversible oxidation of hemoglobin [36]; likewise , it was shown that diaminodiphenyldisulfide causes necrotic changes in the liver , and atrophy and hyperplasia in the kidney [37] . Although some disulfides , such as the ethanol deterrent disulfiram [38 , 39] have been successfully used for a long time as therapeutic agents , the observations on the toxicity of DTDA and similar molecules suggest that per se , this compound has no therapeutic value . However , the specific action of DTDA on TcTIM and the unique alterations that it induces in the structure of the enzyme suggest that it can be used as lead for the discovery of agents with pharmacological potential . The experiments with pure recombinant wild type TcTIM and the mutants in which each of the four Cys was replaced with Val for the case of Cys118 or Ala in the other three Cys showed that regardless of their cysteine content , the four enzymes exhibit similar sensitivities to DTDA . Thus , it would appear that derivatization of cysteines is not the primary event in the inhibition of enzyme activity . On the other hand , the x-ray structure of native TcTIM that had been soaked in DTDA revealed that it induces important structural alterations of the dimer interface; the most obvious was on the interactions of loop3 of monomer A with the neighboring residues of the other subunit . Along this line , it is relevant to point out that because of its importance in catalysis and stability of the dimer , loop3 and its connections with the other subunit have been extensively studied . For example , it has been reported that perturbations of the interface region formed by Cys15 and loop 3 of the other monomer induce inhibition of activity [11–13] . In this connection it is relevant that Brown and Kollman [40] and Aqvist and Fothergill [41] showed by molecular modeling that during catalysis , the hydrogen bond of Thr76 of loop3 with the catalytic His96 of the other subunit shifts to the catalytic Glu . According to the authors , these arrangements are essential for the expression catalytic activity . Thus , it is noteworthy that the change in conformation of loop3 of monomer A had a hinge that localized to Thr76 . In connection to the contribution of loop3 to the stability of TIM dimers , it has been reported that deletions of some of the residues of loop3 , yield enzymes that essentially exist in the monomeric form [9 , 10] . It is also relevant that the residues of loop3 surround the side chain of Cys15 of the other monomer , and that alkylation of the two interface Cys15 of P . falciparum [12] and E . histolytica [13] by thiol reagents induces the formation of stable monomers . In TIMs from T . cruzi and T . brucei , the derivatization of their two interface Cys15 induces aggregation of the enzymes [11] . In addition it is noteworthy that in a hybrid formed by a C15A TcTIM monomer and a monomer of wild type TbTIM , the alkylation of the only interface cysteine yields an enzyme that conserves its dimeric structure , albeit its catalytic properties are reduced by about one-half [42] . Therefore , it is mechanistically important that in the DTDA treated enzyme , loop3 of monomer A acquired a different position . This conformational change was accompanied by alterations of the contacts of Cys15 with the adjacent subunit; specifically , the electron density that corresponds to the β-carbon of Cys15 was not apparent , whereas that of the sulfur atom appeared diffuse . Moreover , DTDA treatment induced the loss of two H-bonds in the Cys15-loop3 interfacial region . The sum of these structural effects most likely accounts for its inhibiting effect of DTDA on the activity of TcTIM . In the crystallographic data , there is another point that merits comment . This concerns the observation that even though in the TIM dimer , there are two equivalent Cys15-loop3 regions , only one of them was altered by DTDA , the other appeared intact . In solution the inhibiting effect of DTDA is accompanied by enzyme aggregation . This is in consonance with previous data [11] that showed that alterations of the dimer interface by chemical modification or site directed mutagenesis led to enzyme aggregation , indicating that perturbation of the interface lead to formation of unstable monomers that subsequently undergo aggregations . Therefore , it is likely that in the crystal , an intermediate of the overall conformational changes induced by DTDA was trapped . In sum , this work shows that DTDA specifically inhibits the activity of TIM from T . cruzi , that it is able to cross biological membranes , and that it is effective in T . cruzi epimastigotes . An additional characteristic of DTDA is that it inhibits TcTIM by perturbing the interactions between its two subunits; thus , this compound is another example of the relatively small number of the so far reported agents that by acting on protein-protein interfaces induce a desired detrimental effect . Since the interfaces of oligomeric proteins would seem to be excellent targets for the discovery of agents that are specific for the enzymes from parasites [14 , 15] , DTDA would seem to be a good model for the discovery of molecules that are less toxic , but that still conserve their effectiveness in the T . cruzi enzyme . | Most of the enzymes of parasites have their counterpart in the host . Throughout evolution , the three-dimensional architecture of enzymes and their catalytic sites are highly conserved . Thus , identifying molecules that act exclusively on the active sites of the enzymes from parasites is a difficult task . However , it is documented that the majority of enzymes consist of various subunits , and that conservation in the interface of the subunits is lower than in the catalytic site . Indeed , we found that there are significant differences in the interface between the two subunits of triosephosphate isomerase from Homo sapiens and Trypanosoma cruzi ( TcTIM ) , which causes Chagas disease in the American continent . In the search for agents that specifically inhibit TcTIM , we found that 2 , 2′-dithioaniline ( DTDA ) is far more effective in inactivating TcTIM than the human enzyme , and that its detrimental effect is due to perturbation of the dimer interface . Remarkably , DTDA prevented the growth of Escherichia coli cells that had TcTIM instead of their own TIM and killed T . cruzi epimastigotes in culture . Thus , this study highlights a new approach base of targeting molecular interfaces of dimers . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biochemistry/biocatalysis",
"biochemistry/experimental",
"biophysical",
"methods",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"biochemistry/protein",
"chemistry",
"biochemistry/biomacromolecule-ligand",
"interactions",
"microbiology/parasitology",
"infectious",
"disease... | 2007 | Perturbation of the Dimer Interface of Triosephosphate Isomerase and its Effect on Trypanosoma cruzi |
The performance of information processing systems , from artificial neural networks to natural neuronal ensembles , depends heavily on the underlying system architecture . In this study , we compare the performance of parallel and layered network architectures during sequential tasks that require both acquisition and retention of information , thereby identifying tradeoffs between learning and memory processes . During the task of supervised , sequential function approximation , networks produce and adapt representations of external information . Performance is evaluated by statistically analyzing the error in these representations while varying the initial network state , the structure of the external information , and the time given to learn the information . We link performance to complexity in network architecture by characterizing local error landscape curvature . We find that variations in error landscape structure give rise to tradeoffs in performance; these include the ability of the network to maximize accuracy versus minimize inaccuracy and produce specific versus generalizable representations of information . Parallel networks generate smooth error landscapes with deep , narrow minima , enabling them to find highly specific representations given sufficient time . While accurate , however , these representations are difficult to generalize . In contrast , layered networks generate rough error landscapes with a variety of local minima , allowing them to quickly find coarse representations . Although less accurate , these representations are easily adaptable . The presence of measurable performance tradeoffs in both layered and parallel networks has implications for understanding the behavior of a wide variety of natural and artificial learning systems .
Learning , the assimilation of new information , and memory , the retention of old information , are competing processes; the first requires flexibility and the second stability in the presence of external stimuli . Varying structural complexity could uncover tradeoffs between flexibility and stability , particularly when comparing the functional performance of structurally distinct learning systems . We use neural networks as model learning systems to explore these tradeoffs in system architectures inspired by both biology and computer science , considering layered structures like those found in cortical lamina [1] and parallel structures such as those used for clustering [2] , image processing [3] , and forecasting [4] . We find inherent tradeoffs in network performance , most notably between acquisition versus retention of information and between the ability of the network to maximize success versus minimize failure during sequential learning and memory tasks . Identifying tradeoffs in performance that arise from complexity in architecture is crucial for understanding the relationship between structure and function in both natural and artificial learning systems . Natural neuronal systems display a complex combination of serial and parallel [5] structural motifs which enable the performance of disparate functions [6]–[9] . For example , layered [1] and hierarchical [10] architectures theoretically important for sustained limited activity [11] have been consistently identified over a range of spatial scales in primate cortical systems [12] . Neurons themselves are organized into layers , or “lamina , ” and both intra-laminar [13] and inter-laminar [14] connectivity differentially impact function . Similarly , information processing systems developed by technological innovation rather than natural evolution have structures designed to match their functionality . For example , the topological complexity of very large integrated circuits scales with the function to be performed [15] . Likewise , the internal structure of artificial neural networks can be carefully constructed [16] to enable these systems to learn a variety of complex relationships . While parallel , rather than serial , structures are appealing in artificial neural networks because of their efficiency and speed , variations in structure may provide additional benefits or drawbacks during the performance of sequential tasks . The dependence of functional performance on structural architecture can be systematically examined within the framework of neural networks , where the complexity of both the network architecture and the external information can be precisely varied . In this study , we evaluate the representations of information produced by feedforward neural networks during supervised , sequential tasks that require both acquisition and retention of information . Our approach is quite different from studies in which large , dense networks are given an extended period of time to produce highly accurate representations of information ( e . g . [17] , [18] ) . Instead , we investigate the links between structure and function by performing a statistical analysis of the error in the representations produced by small networks during short training sessions , thereby identifying mechanisms that underlie tradeoffs in performance . Our work therefore has important implications for understanding the behavior of larger , more complicated systems in which statistical studies of performance would be impossible . In the remainder of the paper , we discuss the extent to which network architectures differ in their ability to both learn and retain information . We first describe the network model and architectures considered in this study . We then quantify the best , worst , and average performance achieved by each network during sequential tasks that vary in both their duration and complexity . We consider the adaptability of these networks to variable initial states , thereby probing the structure of functional error landscapes . Finally , we explore how landscape variations that arise from structural complexity lead to differences in performance .
Our approach differs from traditional machine learning studies in that our goal is not to design the optimal network system for performing a specific task . Rather , we identify tradeoffs in network performance across a range of architectures that share a common algorithmic framework . In this context , the term “architecture” refers specifically to the structural organization of network connections and not , as is found in engineering studies , to the broader set of constraints governing the interactions of network components . In evaluating network performance , we use techniques relevant to both artificial and biological systems . Artificial network systems often favor high accuracy and consistency during a single task , regardless of the time required to achieve such a solution . In biological systems , however , speed and generalizability are often more important that absolute accuracy when dynamically adapting to a variety of tasks . To probe features such as network accuracy , consistency , speed , and adaptability , we examine the representations of information produced by neural networks during competing learning and memory tasks . We choose to study learning and memory within the biologically-motivated framework of feedforward , backpropagation ( FFBP ) artificial neural networks that perform the task of supervised , one-dimensional function approximation . The training process , which consists of adjusting internal connection strengths to minimize the network error on a set of external data points , can be mapped to motion within a continuous error landscape . Within this context , “learning” refers to the ability of the network to successfully navigate this landscape and produce an accurate functional representation of a set of data points , while “memory” refers to the ability to store a representation of previously-learned information . Additional details of this framework are described in the following subsection . To simultaneously study learning and memory processes , information must be presented to the network sequentially . “Catastrophic forgetting , ” in which a network learns new information at the cost of forgetting old information , is a longstanding problem in sequential training of neural networks and has been addressed with several types of rehearsal methods [19]–[21] . Standard rehearsal involves training the network with both the original and new information during sequential training sessions . We use a more biologically motivated approach , the pseudorehearsal method [22] , in which the network trains with a representation of the original information . Pseudorehearsal has been shown to prevent catastrophic forgetting in both feedforward and recurrent networks and does not require extensive storage of examples [22] , [23] . In training FFBP networks , local minima and plateaus within the error landscape can prevent the network from finding a global optimum [24] , [25] . While considered disadvantageous in machine learning studies , the existence of local minima may provide benefits during the training process , particularly in biological systems for which highly accurate global optimums may be unnecessary or undesirable . Additionally , FFBP networks can suffer from overfitting , a problem in which the creation of highly specific representations of information hinders the ability of the network to generalize to new situations [26] . While also considered disadvantageous , failure to generalize has important biological consequences and has been linked to neurological development disorders such as Autism [27] . Instead of attempting to eliminate these sensitivities , we seek to understand the architectural basis for differences in landscape features and examine their impact on representational capabilities such as specificity and generalizability . The construction of our network model is consistent with standard FFBP neural network models [26] . We consider the five distinct architectures shown in Figure 1 ( a ) , all of which obey identical training rules . Each network has 12 hidden nodes arranged into layers of nodes per layer . Nodes in adjacent layers are connected via variable , unidirectional weights . The “fan” and “stacked” networks are both fully connected and have the same total number of connections . The connectivities of the “intermediate” networks , which have slightly greater numbers of connections , were chosen in order to roughly maintain the same total number of adjustable parameters per network , , noted in Figure 1 ( a ) . Each node has a sigmoid transfer function with a variable threshold . The output of each node is a function of the weighted sum of its inputs , given by , where gives the weight of the input connection . Representing the threshold as , where for all nodes , allows us to organize all adjustable parameters into a single , -dimensional weight vector . During training , each network is presented with a training pattern of pairs of input and target values , denoted . We restrict the input space to the range , and the sigmoid transfer function restricts the output space to the range . The set of variable weights is iteratively updated via the Polak-Ribiere conjugate gradient descent method with an adaptive step size [28]–[30] in order to minimize the output error . We use online training , for which is the sum of squared errors between the network output and target output calculated after all points are presented to the network: ( 1 ) Each network shown in Figure 1 ( a ) is trained over two sequential sessions . In describing parameter choices for each training session , we use to denote a continuous uniform probability distribution over the interval . The steps of the sequential training process are shown schematically in Figure 1 ( b ) and are described below:
We train the five networks shown in Figure 1 ( a ) , first considering the differences between the boundary fan ( parallel ) and stacked ( layered ) networks . Given the large number of adjustable parameters relative to the small number of training points , we expect all five networks to fit the points with high accuracy . Instead , the networks show significant differences in performance both within individual training sessions and measured statistically over many sessions . These results , discussed in detail below , show the same qualitative features for larger networks ( Figures S1 and S2 ) and for different sets of original points ( Figures S3 and S4 ) . Both natural and artificial systems can be found in a variety of states when presented with new information . The success in learning this information may depend both on the initial state of the system and on the learning conditions . We explore these possible dependencies by varying both the randomly initialized network state and the training conditions . Given unlimited training time , the distributions in Figure 4 ( a ) mark the error of local minima found within the error landscape of each network . Each minimum can be characterized by the degree of local landscape curvature , where directions of high curvature specify combinations of weight adjustments that produce large changes in error . We adopt the terminology used in previous studies and refer to directions with high and low curvature as stiff and sloppy , respectively [31] , [32] . Stiff and sloppy directions are found by diagonalizing the error Hessian evaluated at the set of weights that produces the local error minimum . For computational efficiency , we use the approximate Levenberg-Marquardt ( LM ) Hessian [33] , defined as: ( 2 ) where is the residual of the original point . The LM Hessian is a good approximation to when the error of local minima , and thus the residual , is small and the additional Hessian term can be neglected . For a given model and data set , the LM Hessian agrees well with the stiffest eigenvectors of and is equivalent to when the model perfectly fits the data . In addition , it has a known number of exactly zero eigenvalues equal to the difference in the number of model parameters and the number of data points [31] , [32] . We diagonalize the LM Hessian about each of the 500 minima with the error values shown in Figure 4 ( a ) . Each error minimum produces a set of eigenvalues and normalized eigenvectors , which give the degrees and directions of stiffness in weight space . As an illustrative example of landscape features observed along these relevant directions , Figures 5 ( a ) and 5 ( b ) show the projection of the error landscape onto the two stiffest eigenvector directions and centered on zero error minima produced by the fan and stacked networks , respectively . The fan landscape shows a single deep basin surrounded by smoothly varying peaks . In contrast , the stacked landscape is rugged , showing a deep valley with several minima separated by small barriers . While these minima appear to be distinct , they may be connected by higher dimensional pathways that cannot be seen in this reduced space .
Given the wealth of structural motifs present in real world systems , it is of interest to first isolate the tradeoffs in performance associated with small parallel and layered network structures which together form the complex architectural landscape of larger systems and thereby constrain their overall performance . Here we found that the deep , narrow basins within the error landscape enabled the fan network to produce very accurate solutions . However , the difficulty of simultaneously adjusting many network connections in order to escape deep basins may have hindered the ability of the fan network to adapt , a result that helps explain the susceptibility of parallel networks to the problems of overfitting and failure to generalize [26] . In contrast , higher variability in the width and depth of local minima enabled the stacked network to quickly find coarse but generalizable solutions through the adjustment of a smaller fraction of weights . In combination , these results support the hypothesis that the number and width of local landscape minima may increase with increasing number of hidden layers [4] , and we suggest that this variability helps explain why layered networks may require fewer computational units and may better generalize than parallel networks [49] , [50] . However , the impact of structural variations on functional tradeoffs , for example between specificity and generalizability , extends beyond artificial network studies and is crucial for understanding the interaction of learning processes in large scale models of the brain [51] . While parallel architectures are often preferred in artificial network studies due to their consistency and accuracy [48] , [50] , our results highlight the advantages of layered architectures when performance criteria favor generalizability and minimization of failure . Building on the intuition gained from the two benchmark extremes – fan and stacked – we further assessed the characteristics of intermediate networks , which can be used to more directly probe the expected behavior of structurally complex composite systems . In particular , our intermediate structures were composed of several adjacent stacked networks and therefore shared principal features of both parallel and layered systems . Additionally , these networks had slightly larger numbers of connections than the fan and stacked networks . Due to these structural differences , the depth of local minima within the intermediate landscapes displayed more variation than fan minima but more continuity than stacked minima . As landscape variability was linked to improved generalization capabilities , a continuous range of basin depths may have enabled the more successful balance between flexible learning and stable memory observed in the intermediate networks . This performance supports the hypothesis that short path lengths ( similar to the serialization [52] ) and low connection densities may facilitate simultaneous performance of information segregation ( memory retention ) and integration ( generalization ) within natural neuronal systems [53] . These competing processes are also maintained in natural neuronal systems and neural circuit models through homeostatic plasticity mechanisms such as synaptic scaling [54] , [55] and redistribution [56] , [57] , in addition to the rehearsal methods employed here [19]–[23] . Even in the absence of such homeostatic plasticity mechanisms , we found that the architectural combination of parallel and layered connectivity helped foster a balance between learning and memory . We extended our analysis from the case of unlimited training time , which revealed information about error landscape structure , to the biologically-motivated case of limited training time . Comparison of these two cases revealed a tradeoff in performance between training speed and solution accuracy . In the absence of temporal constraints , the production of highly accurate representations required longer training times . Similarly , temporal constraints led to larger solution errors . This tradeoff between speed and accuracy has been observed in cortical networks , where emphasis on performance speed during perceptual learning tasks increased the baseline activity but decreased the transient task-related activity of neurons within the decision-making regions of the human brain [58] , [59] . Here we found that network architecture played a significant role in the manifestation of this tradeoff , and the presence of additional hidden layers helped minimize network susceptibility to changes in training time . In particular , the fan network demonstrated the greatest change in performance under temporal constraints , showing a decrease in consistency coupled with occasional catastrophic error values . In contrast , the intermediate and stacked networks improved consistency and minimized inaccuracy once training time was limited . Upon closer inspection , we found that the intermediate networks produced solutions with increased speed given unlimited time and with increased potential for accuracy when time was limited as compared to the fan and stacked extremes . The presence of additional connections may have influenced the number of iterations required to find a solution , or similarly the minimum error found with a fixed number of iterations . While the graph measure of path length is known to influence network efficiency [52] , these results imply that the number of networks connections may additionally enable the network to quickly find an accurate solution . In addition to static variations in connectivity , dynamic structural changes such as synapse formation [60] can facilitate learning and memory processes . The converse case of network degradation , or disruptions to structural connectivity , is also known to have widespread consequences in functional properties of the brain [61]–[63] . A more detailed study of the relationfship between connection number and robustness could provide additional insight into the effects of synapse formation and degradation on functional performance . Our analysis of error landscape features revealed that different architectures showed variable localization properties in the eigenvectors associated with local error minima , and we therefore expect robustness to depend on both the architecture and the location of growth or damage within the network . We found that parallel networks suffered from the creation of excessively detailed representations of information , an “overfitting” problem that is often addressed through the use of cross-validation [64] and weight regularization [65] techniques . As one goal of this study was to uncover the structural basis for differences in representational capabilities , it was crucial to understand network behavior in the absence of task-specific cross-validation schemes . Additionally , as the number of parameters was roughly constant across all network structures ( and identical for the fan and stacked networks ) , we were able to draw comparisons across network architectures in the absence of additional weight regularization constraints . While parallel network models have commonly been used in machine learning studies , multi-layer “deep” networks have recently gained interest due to their potential ability to compactly represent ( using fewer computational units and parameters ) highly variable functions [49] , [50] . The “deep belief” framework has been successful for training large , multi-layered networks , and training methods often couple unsupervised , layer-wise ( greedy ) training with supervised fine-tuning [66] . Recent studies of deep belief networks found that classification performance improved with the addition of layers [48] . In addition , it was suggested that a reduction in the number of hidden layers would require an exponential increase in the number of hidden units in order to achieve similar network performance [50] . These results emphasize the capabilities of layered networks and provide an additional framework in which to explore structure-function tradeoffs . Although biologically-motivated , the FFBP framework includes several simplifying assumptions that could be modified to include additional , realistic complexity . First , we assumed that only the connection weights , analogous to synaptic strengths , were variable . Real neurons also exhibit changes in intrinsic dynamics [67] that interact with network architecture to constrain functionality in the brain [68] . Accounting for such relationships could be particularly relevant , for example , in the study of neuron response profiles within different cortical layers [13] . Second , we assumed that signals passed between nodes had no temporal structure , analogous to representing steady state neuron firing rates . Temporally varying signals could be included to study the dependence of dynamic properties , such as synchronization [68]–[70] and signal propagation [71] , on structural organization [72] . Lastly , we assumed feedforward connectivity . The addition of recurrent connections could be used to study the relationship between recurrent structure and oscillatory functions such as cortical sleep rhythms [73] and oscillation couplings relevant for associative learning and memory [74] . In each of these directions , we anticipate that underlying structural complexity will continue to impact performance through functional tradeoffs . In summary , different network architectures produce error landscapes with distinguishable characteristics , such as the height and width of local minima , which in turn determine performance features such as speed , accuracy , and adaptability . Inherent tradeoffs , observed across a range of architectures , arise as a consequence of the underlying error landscape structure . The presence of local landscape minima enable greater speed , more generalizable solutions , and minimization of catastrophic failure . However , these successes come at the cost of decreased accuracy . Understanding how both the landscape characteristics and the resulting performance features vary across a range of architectures is crucial for both understanding and guiding the design of more complex biological and technical systems . | Information processing systems , such as natural biological networks and artificial computational networks , exhibit a strong interdependence between structural organization and functional performance . However , the extent to which variations in structure impact performance is not well understood , particularly in systems whose functionality must be simultaneously flexible and stable . By statistically analyzing the behavior of network systems during flexible learning and stable memory processes , we quantify the impact of structural variations on the ability of the network to learn , modify , and retain representations of information . Across a range of architectures drawn from both natural and artificial systems , we show that these networks face tradeoffs between the ability to learn and retain information , and the observed behavior varies depending on the initial network state and the time given to process information . Furthermore , we analyze the difficulty with which different network architectures produce accurate versus generalizable representations of information , thereby identifying the structural mechanisms that give rise to functional tradeoffs between learning and memory . | [
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] | [
"statistical",
"mechanics",
"neural",
"networks",
"statistics",
"applied",
"mathematics",
"computerized",
"simulations",
"neuroscience",
"learning",
"and",
"memory",
"mathematics",
"computational",
"neuroscience",
"distribution",
"curves",
"biophysics",
"simulations",
"circui... | 2011 | Learning, Memory, and the Role of Neural Network Architecture |
The self-assembly of proteins into protein quaternary structures is of fundamental importance to many biological processes , and protein misassembly is responsible for a wide range of proteopathic diseases . In recent years , abstract lattice models of protein self-assembly have been used to simulate the evolution and assembly of protein quaternary structure , and to provide a tractable way to study the genotype-phenotype map of such systems . Here we generalize these models by representing the interfaces as mutable binary strings . This simple change enables us to model the evolution of interface strengths , interface symmetry , and deterministic assembly pathways . Using the generalized model we are able to reproduce two important results established for real protein complexes: The first is that protein assembly pathways are under evolutionary selection to minimize misassembly . The second is that the assembly pathway of a complex mirrors its evolutionary history , and that both can be derived from the relative strengths of interfaces . These results demonstrate that the generalized lattice model offers a powerful new idealized framework to facilitate the study of protein self-assembly processes and their evolution .
Any self-assembling system requires two ingredients: assembly subunits with binding sites , and a method for determining the strength of an interaction between two such sites . The arrangement of the sites and their interactions can be described in the form of an assembly graph [13] . From these simple components , structures can be formed through the following stochastic assembly process: If the subunits are square tiles on a lattice , connected sets of tiles are called polyominoes [14] . We can define a genotype that encodes a set of subunit interactions as a sequence , in which each sequence position represents the type of a particular binding site on a subunit . The assembly process maps a given genotype to a single polyomino ( in the case of a deterministic genotype ) or a statistical distribution of several different polyominoes ( for a nondeterministic genotype ) . In either case these polyominoes can be thought of as abstract biological phenotypes . The assembly process is independent of the order in which the subunits are represented in the genotype , and translations , rotations , or reflections of a given polyomino are not considered unique . The implementation of this invariance is outlined in S1 Text . An example of the mapping from genotype to phenotype is shown in Fig 1 , using the integer binding site conventions of existing polyomino models . Certain binding sites are noninteracting ( labeled 0 ) while interactions of equal strength occur between fixed pairs of positive integers . The interacting pairs are 1 ↔ 2 , 3 ↔ 4 , etc . Repeated assemblies of the same genotype do not necessarily produce the same polyomino , a property referred to as nondeterminism . There are many sources of nondeterminism , ranging from unbound aggregations of subunits to branching pathways in the course of the assembly process . A more general insight into nondeterminism in polyomino self-assembly is given by Tesoro , Ahnert , and Leonard [13] . Deterministic genotypes are significantly outnumbered by nondeterministic ones , and the addition of interactions typically increases the fraction of nondeterministic genotypes . In a biological context nondeterministic genotypes can be viewed as less desirable than deterministic ones , as the functions of many proteins strongly rely on the accuracy and reproducibility of their structures . We can therefore use nondeterminism in the polyomino self-assembly model to represent protein misassembly and thereby study the conditions under which proteins may evolve towards more stable and reliable assemblies . In this paper we generalize the standard polyomino self-assembly model as outlined above by introducing interfaces that take the form of binary strings rather than integers . This definition of interfaces gives rise to further definitions of interface strength and symmetry . It also allows for non-transitive interactions between interfaces . The assembly process outlined earlier is unchanged , with only the sites and thus how to determine interactions between them being redefined , as seen in Fig 2 . The number of bits per binding site is given by LI , providing 2 L I unique binding site configurations . Since the subunits are always encoded in a genotype following a common convention ( e . g . clockwise around a tile ) , two adjoined sites have a “head to tail” alignment ( see Fig 2 ) . The interaction strength between two sites relates to the Hamming distance dH between one site and the reversed alignment of the other , normalized by LI . As such , the interaction strength S ^ ∈ [ 0 , 1 ] , and binding can occur if the strength is above some chosen critical strength S ^ ≥ S ^ c . The stochastic assembly process as outlined earlier is now extended to include a binding probability as a function of interaction strengths . Interacting subunits are no longer guaranteed to bind , but binding that does occur remains irreversible . Binding probability can be linked to interaction strength via an abstract temperature T ∈ [0 , ∞ ) . More complex forms may have more physical justification , but a useful form of binding probability is Pr b i n d i n g = H ( S ^ - S ^ c ) S ^ T where H is the Heaviside function , taking H ( 0 ) = 1 . The average number of attempts an interaction will take , effectively the binding time , is the reciprocal of the binding probability . With the choice T > 0 , stronger bonds are expected to assemble more quickly than weaker bonds .
Accessing information on the evolution of real protein binding strengths over sufficiently long time scales is effectively impossible . There are potential proxies , like looking at homologous proteins across an evolutionary tree [15] . Experimental work has suggested a link between ordered assembly pathways and the constraints they place on evolution [11] , but focused on subunits fusing together rather than individual strengths evolving . Here we show how the generalized polyomino model can simulate evolutionary selection for assembly order , such as observed in [11] for real protein complexes . The possibility of nondeterminism in our generalized model , combined with variable binding strengths , give rise to a space in which evolution can optimize binding strengths in order to maximize the probability that critical assembly steps occur in the right order for a desirable phenotype . In the steady-state limit of the evolutionary simulations , mutation and selection effectively eliminate any trace of ancestry in the interface strengths . The steady state properties of interaction strengths depend only on the current phenotype . However , shortly after a new shape has evolved , it is possible to deduce ancestry from interface strengths . In the case of the 12-mer and the 16-mer , where we have one nondeterministic ancestor and one deterministic one , this is obvious as the interface strength distributions of the two ancestors differ considerably . As a result the two alternative ancestries for each of these two polyominoes can be clearly distinguished by bond strengths up to about 50 generations . But even where we have deterministic ancestors , namely for the octomer and the heterotetramer , we notice that at the earliest time points the interface that is also present in the ancestor is stronger than the interface that is absent in the ancestor . This latter observation mirrors results found in real protein complexes , where the ordering of interface strengths often reflects the order of evolution , with the strongest interface as the oldest [10] .
The time ordering of assembly steps in proteins is integral to the correct assembly of the protein structure . This holds true on many length scales of assembly , with cotranslational protein folding able to induce misassembly [16] all the way up to final quaternary structure as examined here . Experimental methods for devising binding strengths are still being developed [17] , with an in silico approach recently introduced focusing on multimeric complexes [18] . One notable result was that given an equal rate of mutation , deterministic and nondeterministic assemblies adapted at different rates . The peak observed rate of binding strength increase in the 12-mer was approximately triple the rate in deterministic assemblies . Such an observation is fairly intuitive , as mutations which alter binding strength correctly or incorrectly are more strongly selected or purified respectively in the nondeterministic assemblies . This is in good agreement with the observation that unstable proteins adapt more quickly [19] . Binding strengths that deviate from neutral expectations do so to optimize determinism , assembling a core of the final structure as quickly as possible before adding further , peripheral elements . This evolutionary selection for a particular assembly pathway parallels real protein complexes , in which gene fusions are a way of cementing particular assembly order under evolutionary selection pressure in order to minimize the risk of misassembly [11] . Generalizing the binding sites from integers to binary strings provides a range of benefits . The number of binding site configurations is now fixed by a physically meaningful parameter and is exponentially large . Previous models frequently had identical binding sites at multiple locations , which is not observed in real proteins , whereas now repeated binding sites are vanishingly rare . Additionally , interaction rules in the integer model have trivial transitivity relations: Maintaining the notation of ↔ for interactions , that is to say for sites A , B , C that ( A ↔ B ) ∧ ( B ↔ C ) → ( A = C ) However , the generalized model does not require the above relation to be true , with knowledge of one interaction having little bearing on other interactions sharing a binding site . That it is to say for sites D , E , F , G that ( D ↔ E ) ∧ ( E ↔ F ) ∧ ( F ↔ G ) ↛ ( D = F ) ∨ ( D ↔ G ) This allows more complex interaction patterns to form , but also allows different binding sites to produce the same interaction behaviour , as seen in Fig 7 . In addition , sites can self-interact , interact with another binding site , or both , like sites D and E supporting the interactions D ↔ E and E ↔ E . Usefully , the generalized interactions are a superset of the integer model , and so any previous results could be trivially recovered by choosing S ^ c = 1 ( up to relabeling binding sites ) . While the generalized model is still a strikingly abstract representation of biological self-assembly , the binary interfaces add physical realism and layered complexity to an already promising model . Phenotype plasticity is another feature that is naturally introduced by the generalized model . By incorporating a dynamic fitness landscape , one that alternatively favors two ( or more ) phenotypes , the interaction strengths can continuously adapt to remain optimal , shown in Fig 8 . The ability to modify a phenotype in a controllable manner , minimizing nondeterminism , is a huge advantage to survival . If a conformational change of a protein , in response to an environmental change or other external conditions , altered its binding strengths , it could quickly shift phenotypes . Since changing interaction strengths can occur much quicker than creating new interactions , this plasticity allows adaptions that would otherwise be potentially too slow to survive . The relationship between conformational changes and their impact on evolution is uncertain , but it has been suggested that this behaviour can impose strong constraints on sequence evolution [20 , 21] . Moreover , adding and removing interactions , rather than just reprioritizing them , exposes the assemblies to intermediate states and greater risk of negative outcomes [22] . Polyomino self-assembly models using integers as binding sites have demonstrated the value of abstract self-assembly models for the study of self-assembly phenomena and genotype-phenotype maps [2–4] . Generalizing the binding interfaces using binary subsites as outlined in this paper retains tractability while expanding applicability to more complex biological research questions . In particular , modeling the evolution of interaction strengths provides qualitative insights beyond the reach of previous polyomino studies . With a few justifiable assumptions , analytic predictions of the interaction strengths in the absence of selection pressures can be found , which show strong agreement with simulations . Significant divergences from this prediction are observed in nondeterministic assemblies where time-ordering is important , and the interaction strengths are therefore under selection . This selection pressure drives these interactions to strengthen or weaken , and thus bind earlier or later in the assemble , to optimize the determinism . Certain interaction strength orderings are more suitable for transitioning to descendant phenotypes , and so can be used to statistically reconstruct evolutionary pathways . Several observations from experimental studies have been recovered by this model , as well as suggesting that nondeterminism in the polyomino model provides an interesting framework for the study of protein misassembly . Many further avenues are imaginable that build on such investigations of nondeterminism , including gene duplication , phenotype plasticity , and more complex genotype-phenotype mappings .
As outlined earlier , evolution was modeled with asexual reproduction of haploids encoding two subunits ( total of 8 binding sites per genotype ) . Binding site lengths were LI = 64 and the critical strength was taken as S ^ c = . 671875 . Genotypes were initialized randomly , with the constraint that there were no interactions . Assembly could begin with either subunit as the seed , although monomers were ignored due to their trivial contribution . A population of 250 individuals evolved for 1000 generations , with each genotype being assembled 25 times . Each binary subsite had a fixed probability to flip , such that the entire genotype had mutations that were binomially distributed with mean μ = 1 . The temperature was set to T = 25 , while the nondeterminism punishment was γ = 5 . An individuals fitness was calculated as ( F ) N I · ϕ γ , where F is the fitness jump between higher order assembly graphs , NI is the number of interactions in an assembly graph , and ϕ is the fraction of assemblies that built the correct phenotype . The fitness jump was F = 5 to balance the strong nondeterminism punishment . So , for example , the fitness for a heterotetramer correctly assembled 20 times out of 25 would be 52 ⋅ . 85 . We restricted fitness allocation to the six stated phenotypes in Fig 3 , assigning fitness 0 to all other phenotypes . The majority of transitions between these other phenotypes did not display any novel dynamics , and so were minimized to present the most concise results . The results presented here were initially observed in a full system , and this restriction was introduced to improve significantly improve the simulation fidelity and computation time required . | Protein complexes assemble by joining individual proteins together through interacting binding sites . Because of the long time scales of biological evolution , it can be difficult to reconstruct how these interactions change over time . We use simplified representations of proteins to simulate the evolution of these complexes on a computer . In some cases the order in which the complex assembles is crucial . We show that biological evolution increases the strength of interactions that must occur earlier , and decreases the strength of later interactions . Similar knowledge of interactions being preferred to be stronger or weaker can also help to predict the evolutionary ancestry of a complex . While these simulations are too idealized to make exact predictions , this general link between ordered pathways in assembly and evolution matches well-established observations that have been made in real protein complexes . This means that our model provides a powerful framework to help study protein complex assembly and evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"protein",
"interactions",
"engineering",
"and",
"technology",
"statistics",
"markov",
"processes",
"decision",
"analysis",
"mathematics",
"management",
"engineering",
"protein",
"structure",
"research",
"and",
"analysis",
"methods",
"decision",
"trees",
"proteins",
"stru... | 2019 | Evolution of interface binding strengths in simplified model of protein quaternary structure |
Hemophagocytic lymphohistiocytosis ( HLH ) is a rare , potentially fatal disorder characterized by fever , pancytopenia , hepatosplenomegaly , and increased serum ferritin . HLH is being increasingly reported as a complication of dengue , a common tropical acute febrile illness . After a cluster of pediatric dengue-associated HLH patients was identified during the 2012–2013 dengue epidemic in Puerto Rico , active surveillance and a case-control investigation was conducted at four referral hospitals to determine the incidence of HLH in children and identify risk factors for HLH following dengue . Patients with dengue-associated HLH ( cases ) were matched by month of illness onset and admission hospital to dengue patients that did not develop HLH ( controls ) . During 2008–2013 , a total of 33 HLH patients were identified , of which 22 ( 67% ) were associated with dengue and 1 died ( dengue-associated HLH case-fatality rate: 4 . 5% ) . Two patients with dengue-associated HLH had illness onset in 2009 , none had illness onset during the 2010 dengue epidemic , and 20 had illness onset during the 2012–2013 epidemic . Frequency of infection with either dengue virus ( DENV ) -1 or DENV-4 did not differ between cases and controls . Cases were younger than controls ( median age: 1 vs . 13 years , p < 0 . 01 ) , were hospitalized longer ( 18 vs . 5 days , p < 0 . 01 ) , and were admitted more frequently to pediatric intensive care units ( 100% vs . 16% , p < 0 . 01 ) . Cases had co-infection ( 18 . 2% vs . 4 . 5% , p = 0 . 04 ) , recent influenza-like illness ( 54 . 5% vs . 25 . 0% , p = 0 . 01 ) , and longer duration of fever ( 7 vs . 5 days; p < 0 . 01 ) . Cases were more likely to have lymphadenopathy , hepatomegaly , splenomegaly , anemia , and elevated liver transaminases ( p ≤ 0 . 02 ) . During this cluster of dengue-associated HLH cases that was temporally associated with the 2012–2013 epidemic , most patients with dengue-associated HLH were infants and had higher morbidity than dengue inpatients . Physicians throughout the tropics should be aware of HLH as a potential complication of dengue , particularly in patients with anemia and severe liver injury .
Hemophagocytic lymphohistiocytosis ( HLH ) is a rare , potentially fatal hematologic disorder characterized by hyperinflammation , uncontrolled proliferation of activated lymphocytes , prolonged fever , pancytopenia , jaundice , and hepatosplenomegaly [1] . The etiology of HLH can be familial ( primary ) or acquired ( secondary ) . Diagnostic testing is available for five genes , PRF1 , MUNC13-4 , STX11 , RAB27A and STXBP2 , which are known to cause familial HLH , and mutations in these genes usually manifest within the first two years of life . The incidence of familial HLH is estimated to be 1 . 2 cases per 100 , 000 population per year [1 , 2] . Approximately 70% of HLH cases present during infancy , although juveniles and adults presenting with familial HLH have also been reported [1] . Treatment for familial HLH consists of hematopoietic stem cell transplantation to replace defective immune effector cells; if untreated , familial HLH is invariably fatal [3] . Acquired HLH is a result of strong immunological activation associated with infection , malignancies , or autoimmune disorders , and is more common than primary HLH [1] . The diagnosis of acquired HLH is based upon completion of the HLH clinical case definition [1] , and treatment consists of administration of high-dose corticosteroids , intravenous immunoglobulin ( IVIG ) , and cyclosporine with or without etoposide to dampen the immune response [1] . Epstein-Barr virus ( EBV ) is the most common infectious agent associated with development of HLH [1 , 4] . Dengue is an acute febrile illness caused by any of four mosquito-borne dengue viruses ( DENV-1–4 ) , and is characterized by fever , myalgia , arthralgia , eye pain , headache , rash , and leukopenia [5] . Dengue is endemic throughout the tropics and subtropics , where an estimated 390 million infections occurred in 2010 [6] . Roughly 5% of clinically-apparent dengue cases will progress to severe dengue , which is characterized by plasma leakage leading to effusions , respiratory difficulty , hypovolemic shock , and hemorrhage [5] . Recent investigations have demonstrated that higher serum ferritin levels are associated with dengue as compared to other acute febrile illnesses , and that degree of elevation of ferritin is associated with disease severity and a pro-inflammatory cytokine profile [7 , 8] . Similar to hemophagocytic syndromes , infection with either DENV or EBV results in a genetic immune response profile that is associated with uncontrolled inflammatory responses [9] , and these immunologic pathways may be common to those that lead to the development of acquired HLH . A total of 74 dengue-associated pediatric and adult HLH cases have been described since 1966 [10–37] , in which the cumulative case-fatality rate ( CFR ) was 9 . 5% . However , risk factors associated with increased risk for dengue patients to develop HLH , such as age , early clinical characteristics , and infecting DENV , have not been previously investigated . Dengue has been endemic in the United States Caribbean territory of Puerto Rico since the late 1960s [38 , 39] , where epidemics occur roughly every 3–5 years . During the most recent epidemic in Puerto Rico during 2012–2013 , a total of 29 , 386 suspected dengue cases were reported via passive surveillance from throughout the island , of which 54% had laboratory evidence of acute DENV infection ( Puerto Rico Department of Health [PRDH] ) . Dengue-associated HLH was first documented in Puerto Rico in 2010 in a 10-month-old patient with severe dengue [19] . However , dengue-associated HLH may be under-recognized in Puerto Rico and other dengue-endemic areas due to overlapping signs and symptoms of HLH and dengue ( e . g . , fever , hepatosplenomegaly , leukopenia , and thrombocytopenia ) . Identification of dengue patients with or at risk of developing HLH may be enabled by detection of markedly elevated serum ferritin [22] . In December 2012 , six cases of HLH were reported to PRDH from Puerto Rico Children’s Hospital . The majority of these cases occurred during a two-month period ( November–December 2012 ) and had initially been admitted due to dengue . PRDH and CDC worked with local clinicians to investigate the cluster of HLH cases associated with dengue to estimate the incidence of HLH in Puerto Rico and identify risk factors associated with developing HLH following dengue .
The protocol for this investigation was reviewed by human subjects’ research advisors at the CDC and all hospitals included in the investigation , and was deemed to be a public health intervention and not research . As such , full IRB review was not required . Puerto Rico comprises 78 municipalities on three populated islands ( 9 , 104 km2 ) , of which the total population in 2010 was 3 , 725 , 789 and the pediatric population was 903 , 295 [40] . Roughly half of the residents of Puerto Rico live in the capital , San Juan , and the surrounding metropolitan area . The population included in this investigation included all patients that presented or were transferred to four referral hospitals in the San Juan metropolitan area ( Puerto Rico Children’s Hospital , San Jorge Children’s Hospital , San Francisco Hospital , and Hospital HIMA-San Pablo Bayamon ) during January 2008 through June 2013 . To identify patients with HLH , only patients aged ≤18 years were included since only pediatric HLH patients had been initially reported and two of the four facilities included in the investigation were pediatric hospitals . Hospital billing and discharge diagnoses at each of the four facilities were queried for patients with the following criteria: 1 ) discharge diagnosis of HLH , hemophagocytic syndrome , macrophage activation syndrome ( MAS ) , pancytopenia , hepatomegaly , splenomegaly , or hepatosplenomegaly; 2 ) quantitated serum ferritin; or 3 ) bone marrow biopsy or aspirate performed . Patients meeting any of these criteria were defined as “potential HLH patients” . Medical records of potential HLH patients were retrieved and reviewed to determine if the patient met the HLH clinical case definition ( i . e . , establishment of at least five out of eight diagnostic criteria including: fever; splenomegaly; cytopenia; hypertriglyceridemia and/or hypofibrinogenemia; hemophagocytosis observed in bone marrow , cerebral spinal fluid , or lymph nodes; decreased or absent NK-cell activity; serum ferritin ≥500 μg/L; and serum IL-2 receptor ≥2 , 400 units/L ) [1] . Patients meeting this case definition were defined as an “HLH patient” . All HLH patients with a diagnosis of leukemia or another cancer of the immune system were excluded from further review . A “familial HLH patient” was defined by a positive genetic HLH test . An “acquired HLH patient” was defined by diagnosis of an infectious disease or exacerbation of a chronic medical condition known to be associated with HLH ( e . g . , juvenile arthritis ) within two weeks of onset of HLH . All identified HLH patients were subsequently queried in the passive dengue surveillance system ( PDSS ) to determine if they also had tested laboratory-positive for dengue during the illness for which they sought medical care . If so , they were defined as a “dengue-associated HLH patient” , including if there was diagnostic evidence of infection with an additional etiologic agent . For all dengue-associated HLH patients , data including sociodemographic characteristics , previous medical history , clinical signs and symptoms , and laboratory test results were abstracted and anonymously recorded from patients’ medical records . Dengue diagnostic testing had previously been completed for all individuals involved in this investigation as part of routine dengue surveillance as previously described [38] . In brief , specimens from suspected dengue cases were: a ) reported to PDSS and tested to detect DENV nucleic acid and/or anti-DENV IgM antibody by real-time reverse transcription polymerase chain reaction ( rRT-PCR ) [41] and/or ELISA ( InBios International , Inc . , Seattle , WA ) , respectively , depending on the day post-illness onset of specimen collection; or b ) sent to a private diagnostic laboratory for testing by anti-DENV IgM ELISA . Specimens that tested positive by rRT-PCR or IgM ELISA were defined as a laboratory-positive dengue case . Clinical dengue case definitions employed followed either 2009 [5] or 1997 [42] World Health Organization ( WHO ) guidelines . In brief , patients with dengue were defined by presence of fever plus two or more of following: nausea , vomiting , rash , aches and pains , tourniquet test positive , leucopenia ( i . e . , a white cell count <5 . 0×109 cells/L ) , or any warning sign . Warning signs included severe abdominal pain , persistent vomiting , mucosal bleed , lethargy or restlessness , liver enlargement ≥2 cm , and concurrent rise in hematocrit with rapid decrease in platelet count to <100 , 000/mm3 . Patients with severe dengue were defined by presence of any of the following: 1 ) plasma leakage leading to shock or fluid accumulation resulting in acute respiratory distress; 2 ) severe bleeding with hemodynamic instability requiring fluid replacement and/or blood transfusion , or an intracranial bleed; or 3 ) severe organ impairment such as acute liver failure , myocarditis , or encephalitis . Patients with dengue hemorrhagic fever ( DHF ) were defined by presence or history of fever , hemorrhagic manifestations , thrombocytopenia ( i . e . , platelet count ≤100 , 000/mm3 ) , and plasma leakage . ELISA performed at CDC was used to determine ferritin and interleukin 2 receptor ( IL-2R ) concentrations in serum specimens from cases and controls . Serum specimens were diluted either serially 4-fold ( 1:16–1:1024 ) prior to testing for human ferritin ( MyBioSource , San Diego , CA ) , or 1:4 and 1:8 prior to testing for IL-2R with the Quantikine ELISA ( R&D Systems , Minneapolis , MN ) . Testing and analysis was performed according to the manufacturer’s instructions . Briefly , specimen concentrations were interpolated from a four-parameter logistics curve of standard concentrations and corrected for the dilution factor . Results were reported as the average of duplicates . Each identified dengue-affiliated HLH patient ( i . e . , cases ) was matched to four laboratory-positive dengue patients ( i . e . , controls ) by month of onset of illness and site of hospitalization . Data were analyzed using R v3 . 2 . 0 . For univariate analyses , disease classification was regressed on a suspected risk factor using exact conditional logistic regression . Age was included in each model as a main effect but not interacted with the suspected risk factor . There were insufficient data to perform multiple regression analysis with more than one risk factor at a time . Univariate confidence intervals and p-values were not adjusted for multiple comparisons .
After reviewing the medical records of 694 potential HLH patients that presented during January 2008 through June 2013 to the four hospitals included in the investigation , a total of 33 ( 4 . 8% ) patients were identified that met the HLH clinical case definition ( 0 . 66 HLH patients per 100 , 000 children per year ) . Of these , two ( 6 . 1% ) had positive genetic testing and were defined as familial HLH patients ( 0 . 04 familial HLH cases per 100 , 000 children per year ) , 28 ( 84 . 8% ) had evidence of acquired HLH ( 0 . 56 acquired HLH cases per 100 , 000 per year ) , and 3 ( 9 . 1% ) had no identified etiology of HLH . Etiologies associated with acquired HLH patients were DENV ( n = 22; 84 . 6% ) , herpes simplex virus ( HSV ) ( n = 2; 7 . 7% ) , systemic-onset juvenile rheumatoid arthritis ( n = 2; 7 . 7% ) , EBV ( n = 2; 3 . 8% ) , respiratory syncytial virus ( RSV ) ( n = 1; 3 . 8% ) , and coxsackievirus ( n = 1; 3 . 8% ) . Five ( 17 . 9% ) acquired HLH cases had evidence of infection with two pathogens , including DENV/HSV ( n = 2 ) , DENV/RSV ( n = 1 ) , DENV/EBV ( n = 1 ) , and EBV/coxsackievirus ( n = 1 ) . Of the 22 identified dengue-associated HLH patients , median age was 1 year ( range: 0 . 2–17 years ) and 12 ( 55% ) were male ( Table 1 ) . Median duration of fever was 8 days ( range: 6–25 ) , and median duration of hospitalization was 18 . 5 days ( range: 8–71 ) . Serum ferritin had been quantitated in all 22 patients , and median maximum value was 18 , 789 μg/L ( range: 754–522 , 000 ) . No significant associations were observed between highest measured ferritin levels and age , sex , length of hospital stay , or total days of fever . The presence of hemophagocytosis was evaluated in bone marrow biopsy or aspirate in 14 of the 22 dengue-associated HLH patients , and was observed in eight ( 57 . 1% ) . Upon discharge , 13 ( 59 . 1% ) dengue-associated HLH patients were diagnosed with “hemophagocytic lymphohistiocytosis” or “hemophagocytic syndrome” , and three ( 13 . 6% ) were diagnosed with “macrophage activation syndrome” . Incidence of dengue-associated HLH during January 2008 through June 2013 was 0 . 44 cases per 100 , 000 children per year , and was nearly 25-fold higher during 2012–2013 ( 1 . 48 ) than 2008–2011 ( 0 . 06 ) . The first two identified dengue-associated HLH patients had illness onset in mid-2009 when incidence of dengue was comparatively low ( Fig 1 ) , and the dominant DENVs in circulation were DENV-1 ( 79% ) , DENV-2 ( 17% ) , and DENV-4 ( 4% ) . No dengue-associated HLH patients were identified during the 2010 dengue epidemic that was caused by DENV-1 ( 69% ) , DENV-4 ( 24% ) , DENV-2 ( 7% ) , and DENV-3 ( <0 . 1% ) [38] . The large majority ( 91% ) of dengue-associated HLH patients had illness onset during the 2012–2013 dengue epidemic , in which the dominant DENVs were DENV-1 ( 75% ) , DENV-4 ( 25% ) , DENV-2 ( 0 . 4% ) , and DENV-3 ( <0 . 1% ) . All but one ( 95% ) dengue-associated HLH patients resided in the San Juan metropolitan area , as did nearly all ( 98% ) dengue patients seen at the same four hospitals over the same time frame ( Fig 2 ) . Small sample size of dengue-associated HLH patients precluded analysis of expected incidence of HLH cases in association with dengue cases by municipality of residence . Of 17 dengue-associated HLH cases in which the infecting DENV was identified by RT-PCR , nine ( 53% ) were DENV-1 and eight ( 47% ) were DENV-4 . A moderate correlation existed between monthly incidence of acquired dengue-associated HLH cases and incidence of dengue cases ( Pearson coefficient = 0 . 62 , p < 0 . 0001 ) . A moderate correlation was also found between acquired HLH and number of monthly dengue cases positive by rRT-PCR for either DENV-1 ( Pearson coefficient = 0 . 41; 95% confidence interval [CI]: 0 . 19–0 . 60 ) or DENV-4 ( Pearson coefficient = 0 . 29; 95% CI 0 . 05–0 . 50 ) , but not for DENV-2 or DENV-3 . Median age of all laboratory-positive dengue cases and most-affected age group was the same during the 2010 and 2012–2013 epidemics ( 18 years and 10–19 year-olds , respectively ) . Dengue-associated HLH patients ( “cases” ) were matched 1:4 with randomly selected , laboratory-positive dengue patients ( “controls” ) by site of hospitalization and month of illness onset . Cases were significantly younger than controls ( 1 vs . 13 years of age; p = 0 . 01 ) , and more frequently had evidence of both co-infection ( 18 . 2% vs . 4 . 5%; p = 0 . 04 ) as well as recent influenza-like illness ( 54 . 5% vs . 25 . 0%; p = 0 . 01 ) ( Table 2 ) . No cases or controls received immunosuppressive therapy prior to admission . The mother of one case had an allergic disease ( i . e . , bronchial asthma ) , the mother of one control was HIV seropositive , and one control had been born premature . Frequency of chronic medical conditions and taking daily medications did not differ between cases and controls , nor did frequency of taking aspirin or acetaminophen . Frequency of infecting DENV did not differ between cases and controls . Cases had significantly longer duration of fever than controls ( 7 vs . 5 days , p < 0 . 01 ) . Cases were hospitalized for longer than controls ( 18 vs . 5 days , p < 0 . 01 ) and were more frequently admitted to the pediatric intensive care unit ( 100% vs . 15 . 9% , p < 0 . 01 ) . One case died ( CFR = 4 . 5% ) , whereas no controls died . Interventions significantly associated with cases included being intubated , and receiving a blood transfusion , corticosteroids , IVIG , etoposide , or chemotherapy ( p < 0 . 01 ) . No cases or controls were given cyclosporine or received a hematopoietic stem cell transplant . Signs and symptoms significantly associated with cases included splenomegaly , hepatomegaly and lymphadenopathy ( p ≤ 0 . 02 ) . Cases more frequently had warning signs of severe dengue and met the case definitions for either DHF or severe dengue ( p < 0 . 01 ) . Clinical laboratory values significantly associated with cases included anemia , and elevated aminotransferases , aspartate transaminase ( AST ) and/or alanine transaminase ( ALT ) ( p ≤ 0 . 01 ) . Maximum serum ferritin levels measured while patients were hospitalized were higher in cases ( median = 17 , 794 μg/L; range = 754–522 , 000 ) than controls ( median = 4 , 139 μg/L; range = 42–30 , 346 ) . Serum ferritin level ≥500 μg/L as were more common in cases than in controls ( p = 0 . 02 ) . Using stored serum specimens that had been sent for dengue diagnostic testing , quantitated serum ferritin was more frequently ≥500 μg/L in cases as compared to controls , whereas IL-2Rc was frequently ≥2 , 400 units/mL in both cases and controls .
During this cluster of dengue-associated HLH cases , infants were most frequently affected and cases were associated with higher morbidity ( 100% ICU admission and longer length of stay ) and mortality ( 4 . 5% ) than hospitalized dengue patients . As compared to dengue patients , the relationship between dengue-associated HLH patients and hepatomegaly , splenomegaly , anemia , and elevated aminotransferases was expected , since these aspects are clinically consistent with HLH but not necessarily dengue . Conversely , thrombocytopenia and neutropenia were not more likely to be associated with dengue-associated HLH patients , which may be reflective of the similarity of these clinical characteristics with dengue . Last , although morbidity was higher in dengue-associated HLH patients than those hospitalized due to dengue alone , this was not unexpected since several clinical characteristics of HLH overlap with those of severe dengue ( e . g . , hemorrhage , fluid accumulation , severe liver impairment ) . Additionally it is important to note that serum ferritin is a marker of macrophage activation in vivo , and is often elevated in dengue patients [7 , 8] . Level of serum ferritin in often higher in pediatric dengue patients as compared to adults [43] . However , elevated serum ferritin is also a diagnostic criterium for HLH . The significance and role for elevated serum ferritin in dengue-associated HLH patients are therefore unclear , as several such patients in this investigation had either severe dengue or warning signs for progression to severe dengue . Several possible explanations for this cluster of dengue-associated HLH were explored . An interesting observation was that the majority of dengue-associated HLH patients were infants . Familial HLH frequently presents in the first few years of life , often in association with an infectious agent . Although not all dengue-associated HLH patients were tested for genetic predisposition to HLH , those that were tested were negative . Thus , this outbreak could not be explained by heredity alone . Infants being more frequently identified as dengue-associated HLH patients could be related to increased viremia , heightened immune response to DENV infection , and/or antibody dependent enhancement of disease , the latter of which is associated with waning maternal anti-DENV IgG antibody [5] . Because sufficient volume of serum specimens was unavailable to define dengue-associated HLH patients’ anti-DENV neutralizing antibody profile , we were unable to determine if development of HLH was associated with infection with a specific DENV on the background of a certain immunologic neutralizing antibody combination . However , since the DENVs with which both infants and older children with dengue-associated HLH were infected was not appreciably different from those infecting controls or the distribution of the DENVs circulating in 2012–2013 , this possibility would seem unlikely . Nonetheless , we were unable to rule out a role for previous DENV infection affecting the likelihood of infant or other dengue patients progressing to develop HLH . Hence , future investigations should explore the possibility of prior DENV infections and the order in which they occurred as playing a role in progression from dengue to HLH . Conversely , patients with dengue-associated HLH were more likely to be currently or recently infected with a pathogen in addition to DENV , which may lend further support to the idea that immunologic over-stimulation is associated with development of HLH . However , this phenomenon was observed in less than one-fifth of dengue-associated HLH cases , and thus cannot entirely explain the increased incidence of HLH during the 2012–2013 dengue epidemic . Though limited in breadth due to small sample size , we also saw no association between development to dengue-associated HLH and chronic medical conditions , concomitant medications , recent vaccination , and municipality of residence . Last , an island wide evaluation of 1500 dengue inpatient records found a significant reduction in corticosteroid use from 2008–2009 to 2011 that was thought to be due , in large part , to a 2010 CDC training initiative on dengue clinical management ( CDC data , in press ) . This reduction in the use of corticosteroids could have unmasked HLH patients that would have been inadvertently treated prior to this initiative . Nonetheless , although individual explanations may explain some of the identified dengue-associated HLH patients , the reason ( s ) for increased incidence of HLH specifically with the 2012–2013 dengue epidemic remains unclear . Few studies have identified the population-specific incidence of HLH , most of which focused on familial HLH in pediatric populations . Such studies have identified 0 . 12–0 . 15 HLH cases per 100 , 000 children per year in Sweden [44] , 1 HLH case per 100 , 000 children per year in Texas [2] , and 1 HLH case per 800 , 000 individuals per year in Japan [45] . Therefore , the incidence of familial HLH identified in this investigation ( 0 . 04 per 100 , 000 children per year ) is several fold lower than that observed in previous studies . This difference may be attributable to lack of diagnostic testing for suspected familial HLH cases , clinical under recognition of familial HLH , or potential differences in genetic predisposition to familial HLH . Importantly , the incidence of familial HLH has not been quantitated in African or Hispanic populations . Of note , the incidence of acquired HLH identified in this investigation ( 0 . 56 cases per 100 , 000 children per year ) was 14-fold higher than the incidence of familial HLH . This difference is on par with previous studies in Germany , wherein acquired HLH was roughly 6-fold more common than familial HLH ( G . Janka , personal communication ) . Although pathogens identified to be associated with acquired HLH in this investigation included EBV and other herpesviruses that have been previously associated with HLH , it was striking that nearly four-fifths of acquired HLH cases were associated with DENV infection . Therefore , the comparatively high incidence of HLH identified in this investigation ( 0 . 66 HLH cases per 100 , 000 children per year ) was attributable in large part to dengue-associated HLH . However , it is interesting to note that the population-specific incidence of HLH has not been reported from dengue-endemic regions of the tropics . Additional investigation of HLH , and particularly acquired HLH , should be conducted in the tropics to better understand the incidence of HLH and the mechanisms by which DENV and other pathogens trigger HLH . A prominent strength of this investigation was identification of HLH cases from pediatric referral hospitals , where HLH cases were likely to have been hospitalized . Therefore , it is unlikely than many additional HLH cases were not identified in this population . Conversely , limitations of this investigation include the inability to compare serum ferritin , IL-2 receptor , and natural killer ( NK ) cell activity–which are well-established markers of HLH–between dengue-associated HLH cases and dengue patients that did not develop HLH . This limitation was attributable to lack of available data for controls ( serum ferritin ) and infrequency of the tests being requested for both cases and controls ( IL-2Rc and NK cell activity ) . Also , because only municipality of residence was available for most dengue-associated HLH and dengue patients , we were unable to perform a more exact association of patients’ location of residence and consequent factors associated with socio-economic status with development of dengue-associated HLH . Thus , potential environmental or residential exposures were unable to be confidently ruled out . Additionally familial HLH exists as multiple genetic subtypes and genetic tests were not completed for all cases , therefore it is possible that some HLH cases in this investigation were misclassified . It remains unclear if the increasing incidence of dengue-associated HLH since it was first documented in 1965 is attributable to improved clinical awareness , environmental , genetic , viral , other changes , or some combination thereof . While additional clinical and epidemiologic investigation is conducted in areas with endemic dengue , clinicians seeing patients at risk for dengue should consider HLH in patients with persistent fever , pancytopenia , and multi-organ dysfunction . Previously established protocols for managements of HLH in patients with dengue should be considered [1] , including administration of high dose corticosteroids , IVIG , and cyclosporine with or without etoposide . | Hemophagocytic lymphohistiocytosis ( HLH ) is a rare , potentially fatal medical condition that can occur after a patient has an infection . While HLH is most commonly associated with Epstein-Barr virus infections , it has been reported as a complication of dengue , a common mosquito-borne , acute febrile illness . After a cluster of pediatric dengue-associated HLH patients was identified in Puerto Rico , active surveillance and a case-control investigation was conducted to determine the rate of HLH in children and identify risk factors for HLH following dengue . During 2008–2013 , a total of 33 HLH patients were identified , of which 22 ( 67% ) were associated with dengue and 1 died ( dengue-associated HLH case-fatality rate: 4 . 5% ) . Most ( 91% ) dengue-associated HLH patients had illness onset during the 2012–2013 epidemic , however , HLH was not found to be associated with a particular type of dengue virus . Dengue-associated HLH cases were younger than dengue inpatient controls , were hospitalized longer , and were admitted more frequently to the pediatric intensive care unit . Cases had longer duration of fever , and were more likely to have anemia , hepatomegaly and elevated liver transaminases than controls . Physicians in the tropics should be aware that HLH may complicate dengue , and they should evaluate dengue patients who develop anemia and severe liver injury . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"geographical",
"locations",
"pediatrics",
"north",
"america",
"signs",
"and",
"symptoms",
"ferritin",
"neglected",
"tropical",
... | 2016 | Incidence and Risk Factors for Developing Dengue-Associated Hemophagocytic Lymphohistiocytosis in Puerto Rico, 2008 - 2013 |
Many fibroblast-secreted proteins promote tumorigenicity , and several factors secreted by cancer cells have in turn been proposed to induce these proteins . It is not clear whether there are single dominant pathways underlying these interactions or whether they involve multiple pathways acting in parallel . Here , we identified 42 fibroblast-secreted factors induced by breast cancer cells using comparative genomic analysis . To determine what fraction was active in promoting tumorigenicity , we chose five representative fibroblast-secreted factors for in vivo analysis . We found that the majority ( three out of five ) played equally major roles in promoting tumorigenicity , and intriguingly , each one had distinct effects on the tumor microenvironment . Specifically , fibroblast-secreted amphiregulin promoted breast cancer cell survival , whereas the chemokine CCL7 stimulated tumor cell proliferation while CCL2 promoted innate immune cell infiltration and angiogenesis . The other two factors tested had minor ( CCL8 ) or minimally ( STC1 ) significant effects on the ability of fibroblasts to promote tumor growth . The importance of parallel interactions between fibroblasts and cancer cells was tested by simultaneously targeting fibroblast-secreted amphiregulin and the CCL7 receptor on cancer cells , and this was significantly more efficacious than blocking either pathway alone . We further explored the concept of parallel interactions by testing the extent to which induction of critical fibroblast-secreted proteins could be achieved by single , previously identified , factors produced by breast cancer cells . We found that although single factors could induce a subset of genes , even combinations of factors failed to induce the full repertoire of functionally important fibroblast-secreted proteins . Together , these results delineate a complex network of tumor-fibroblast interactions that act in parallel to promote tumorigenicity and suggest that effective anti-stromal therapeutic strategies will need to be multi-targeted .
Solid tumors are aberrant tissues where stromal cell types co-develop with and influence cancer cells [1] . Significant epigenetic alterations and gene expression changes occur in stromal cells as tumors progress , and the stromal changes are as strikingly different as those observed in the cancer epithelial compartments [2]–[5] . Many of these stromal cell changes are elicited by factors secreted by cancer cells , such as vascular endothelial growth factor ( VEGF; MIM: 192240 ) , which helps recruit and induce proliferation of endothelial cells [6] . Cancer cells also secrete factors that alter surrounding fibroblasts , such as transforming growth factor ( TGF ) -β ( MIM: 190180 ) which induces fibroblasts to differentiate into myofibroblasts and secrete collagen , thereby contributing to the abundant extracellular matrix often observed in epithelial tumors [7] . In addition to TGF-β factors secreted by cancer cells that influence stromal fibroblasts include platelet derived growth factor ( PDGF; MIM: 173430 ) , interleukin ( IL ) -6 ( MIM:147620 ) , IL1-α ( MIM: 147760 ) , and WNT1 inducible signaling pathway protein ( WISP ) -1 and -2 ( MIM: 603398 and 603399 ) [8] , [9] . Tumor associated fibroblasts have been shown to promote cancer cell proliferation , angiogenesis , extracellular matrix ( ECM ) remodeling , inflammation , invasion and metastasis [10] , [11] . Several fibroblast-secreted or membrane-bound factors that mediate these effects have been identified , including CXCL12 ( MIM: 600835 ) , hepatocyte growth factor ( HGF; MIM:142409 ) , matrix metalloproteinase MMP14 ( MIM: 600754 ) , osteopontin ( MIM: 166490 ) , TGF-β , and CCL2 ( MIM: 158105 ) [12]–[17] . Some basic underlying processes involved in regulating the interactions between the epithelial cancer cells and the stromal fibroblasts have been identified . For example , several fibroblast-secreted factors are inflammatory cytokines whose expression is driven by NF-kappaB-dependent transcription in a process similar to the senescent secretory phenotype observed in aging fibroblasts [18] , [19] . Additionally , a study of how fibroblasts co-evolve with tumor cells determined that fibroblasts gradually implement two signaling loops , involving TGF-β and CXCL12 , which act together through both autocrine and cross-signaling mechanisms [20] . What remains unclear is whether the different factors involved in cancer cell-fibroblast interactions reflects a requirement of a multitude of fibroblast factors acting in parallel to promote tumorigenicity , or whether it reflects the diversity of the approaches and systems used to identity important interactions . Here , we designed a study to explore how the entire repertoire of fibroblast-secreted factors that are induced by human cancer cells function as a whole and compared the factors . This systems-level study was designed to be complimentary to approaches that focus on single genes or single processes . Our study indicates that the majority of induced fibroblast-secreted factors play a role in promoting tumorigenicity and that they do so through diverse effects on the tumor tissue .
We adapted previously used systems of human cancer cells and fibroblasts [21]–[23] , by using fibroblast lines that were amenable for the use of stable RNA interference ( shRNA ) so the relevance of candidate mediators of tumor-stromal interactions could be tested . We determined using co-injection assays that two human fibroblast lines previously shown to promote tumorigenicity in other systems ( HFFF2 and HFF1 ) were able to promote tumorigenicity of two basal breast cancer subtype cell lines , MDA-MB-231 and Cal51 , whereas two other human fibroblast lines ( Wi-38 and CCD1112Sk ) were not ( Figure S1 ) . We used the two tumor-supportive fibroblast lines ( HFFF2 and HFF1 ) as models for patient-derived breast carcinoma associated fibroblasts and the two non-supportive fibroblasts as models for normal breast tissue derived fibroblasts . Using fibroblasts cell lines allowed shRNA transfection , selection and subsequent validation of gene silencing , whereas patient-derived breast fibroblasts cannot be passaged in culture long enough for such manipulations ( data not shown; Ahmet Acar , personal communication ) . To determine at which point the host murine fibroblasts replaced the co-injected human fibroblasts , we tagged the human fibroblasts with green fluorescent protein ( GFP ) so that we could visualize how long they survived ensconced within developing tumors . We found that after three weeks , the number of co-injected GFP-tagged human fibroblasts comprised roughly 20% of the tumor stroma as judged by co-localization of GFP and the fibroblast stromal marker alpha-smooth muscle actin ( α-SMA ) ( Figure S2 ) , but the relative contribution steadily declined from week 3 to week 8 , although the GFP-fibroblasts never completely disappear . The level of α-SMA positive cells within tumor stroma persisted during this same time period ( Figure S2 ) . We performed transcriptional profiling and pathway analysis to determine if exposure of our model tumor-supportive fibroblasts to breast cancer cells in co-culture resembled expression changes seen in patient-derived breast carcinoma fibroblasts . We compared expression profiles of HFF1 and HFFF2 tumor-supportive fibroblasts co-cultured with either Cal51 or MDA-MB-231 ( four combinations total ) to those of Wi-38 and CCD1112Sk non-supportive fibroblasts co-cultured with either Cal51 or MDA-MB-231 ( another four combinations ) . We chose to work with basal subtype breast cancer cell lines due to the greater need to develop new treatments for this poor prognosis subtype . We then compared the pathways selectively enriched in tumor-supportive fibroblasts co-cultured with Cal51 and MDA-MB-231 to pathways enriched by comparing cultures of patient-derived breast carcinoma fibroblasts to their normal counterparts . Strikingly , the top four pathways identified by gene-set enrichment analysis ( GSEA ) that are activated by exposure of tumor-promoting human fibroblasts to breast cancer cells are also amongst the top ten pathways activated in cultures of patient-derived breast carcinoma fibroblasts relative to normal fibroblasts ( Figure 1 ) . These pathways are ECM-receptor interaction , focal adhesion , integrin signaling and integrin cell-surface interactions; interrelated pathways that have been shown to be involved in the activation of cancer associated fibroblasts [21] , [24] , [25] . Additionally , most of the other top ten activated pathways in both systems were activated in the other system at lower ranking but still significant levels . These pathways included cytokine cytokine-receptor interactions , PDGF signaling , and Rho GTPase signaling; which likewise have shown to be involved in activation of cancer associated fibroblasts [7] , [13] , [26] ( Figure 1 ) . On the whole , the overlap in all significantly activated pathways is 44% ( Table S1 ) . This affirmed that this system of tumor-promoting human fibroblasts resembled in vitro cultures of patient-derived breast cancer fibroblasts . We next wanted to test whether this system also reflected changes observed in vivo in human primary breast cancer stroma . To address this , we used gene-set enrichment analysis to determine the pathways that were activated in microdissected human breast stroma relative to normal breast stroma [3] . Despite the fact that human breast stroma contains many cell types in addition to fibroblasts , four of the eight significantly activated pathways in human breast stroma were also significantly activated by exposure of tumor-promoting fibroblasts to breast cancer cells , including cytokine/cytokine-receptor interactions and JAK-STAT signaling ( Figure 1 ) . Additionally , we developed a gene signature based on stimulation of tumor-promoting fibroblasts by breast cancer cells . Based on both clustering and principal component analysis , this gene signature was able to correctly predict whether microdissected human breast stroma was derived from cancerous or normal breast tissue for 98 of 99 samples ( Figure 1 ) . These comparative genomic analyses establish the close correspondence of our system of interaction of breast cancer cells with tumor-promoting fibroblasts to stroma and stromal fibroblasts isolated from primary human breast cancer . We used genome-wide analysis to determine the full repertoire of secreted factors induced by breast cancer cells in our tumor-promoting fibroblasts . As the first step , we compared genes induced in tumor-promoting fibroblast lines to genes induced in fibroblast lines incapable of promoting tumorigenicity . We found 320 genes that were more than 2-fold greater induced in tumor-promoting fibroblasts ( Table S2 ) and within this group were 62 genes encoding secreted proteins . Of these , 42 were also significantly upregulated ( p<0 . 05 ) in stromal cells isolated from primary breast cancer relative to normal breast stroma cells [3] , [4] , [27] . This group contained cytokines ( 15 ) , extracellular matrix proteins ( 7 ) , proteases ( 6 ) , and growth factors and hormones ( 6 ) ( Table 1 ) . Cytokines as a class were significantly enriched in our screen relative to their representation in the set of all secreted proteins ( 36% vs . 12% , p = 5 . 6e-6 ) . This enrichment is consistent with the prior reports that cytokines play key roles in the tumor-supportive function of tumor-associated fibroblasts [19] . In contrast , the other functional classes of secreted proteins were not significantly enriched . The cytokines upregulated in tumor-supportive fibroblasts included CC- and CXC-chemokines ( CCL-2 , -5 , -7 , -8 , -20 , and CXCL-5 , -10; MIM: 158105 , 187011 , 158106 , 602283 , 600324 and 147310 respectively ) , pro-inflammatory interleukins ( IL-1α , -1β , -8 , -11 , -24; MIM: 147760 , 147720 , 146930 , 147681 and 604136 respectively ) , colony stimulating factor ( G-CSF; MIM: 138970 ) , interleukin receptor antagonist ( IL1RN; MIM: 147679 ) , and TNF superfamily member TNSF15 ( MIM: 604052 ) ( Table 1 ) . Systematic literature searches revealed that 16 of the 42 secreted proteins had one or more publication ( s ) implicating them in the tumor-supportive function of fibroblasts while 26 did not ( Table 1 ) . The genes not previously implicated as mediators of the tumor-supportive function of fibroblasts included the majority ( 8/15 ) of cytokines , one-half ( 3/6 ) of the proteases , and the majority ( 4/6 ) of the growth factors and hormones . Examples include CCL8 , encoding a chemokine not previously linked to cancer that is involved in homing of memory T lymphocytes to inflamed skin [28]; pregnancy-associated plasma protein-a ( PPAPA; MIM: 176385 ) , a protease that degrades IGF-binding proteins and acts as a positive modulator of local IGF signaling in skin repair [29]; and EGFL6 ( MIM: 300239 ) , encoding an epidermal growth factor ( EGF ) repeat protein expressed in osteoblastic-like cells and capable of inducing migration of endothelial cells [30] . Some of the fibroblast-secreted candidate tumor-supportive factors fall outside of the known classes of proteins involved in tumor-stromal interactions , including ISG15 ( MIM: 147571 ) , an interferon-inducible , ubiquitin-like protein whose secretion plays a critical role in mediating an effective immune response to mycobacteria [31]; and complement component C3 ( MIM: 120700 ) , which in addition to its role in the complement cascade helps mobilize hematopoietic stem/progenitor cells to wounds [32] . In order to characterize the effects of fibroblasts on tumor progression , we co-injected tumor-promoting fibroblasts and examined their effects on both tumor cells and associated non-tumor cells within the tumor microenvironment . Consistent with a faster tumor growth rate , cancer cells in the co-injected tumors exhibited a two-fold higher proliferative rate based on Ki-67 labeling ( Figure 2 ) . To visualize and quantify the presence of three of the most important features of the tumor microenvironment - inflammation , vascularization and fibroblast activation - we performed immunohistochemical analysis at week six after co-injection . We used antibodies to the 7/4-antigen , which is highly expressed in neutrophils and inflammatory monocytes [33] , the CD31 endothelial antigen , which visualizes blood vessels [34] , and α-SMA , which is expressed by activated fibroblasts [35] . Strikingly , the presence of inflammatory cells , degree of vascularization , and number of activated fibroblasts were all 3 to 4 fold higher in the co-injected tumors ( Figure 2 ) , indicating a profound influence of tumor-supportive fibroblasts on the composition of the tumor microenvironment . We decided to analyze a mix of previously implicated and novel candidate pro-tumorigenic fibroblast-secreted factors . Additionally , we wanted to test whether seemingly redundant factors were indeed functionally redundant . We focused on cytokines , growth factors , and hormones , and chose CCL2 and CCL7 ( both previously validated as functionally important fibroblast secreted factors [17] , [36] ) ; the related factor CCL8 which , like CCL7 , binds CCR1 ( MIM: 601159 ) ; amphiregulin ( AREG; MIM: 104640 ) , which has been implicated in tumor stromal-interactions but as a ligand produced by cancer cells acting on fibroblasts [37]; and stanniocalcin1 ( STC1; MIM: 601185 ) , which has been shown to act as a cancer cell autonomous factor , but not as a stromally produced factor [38] . For each of these five genes , we confirmed by quantitative RT-PCR that there was a significant induction in the co-cultured tumor supportive fibroblasts relative to the co-cultured neutral fibroblasts ( Figure S3 ) . CCL2 , CCL7 , and CCL8 are structurally related chemokines that share a common function of recruitment of monocytes to areas of injury and inflammation [39] . Based on their structural similarity and overlapping functions in inflammation , we wanted to determine if they had redundant roles in the tumor supportive function of co-injected fibroblasts . Remarkably , shRNAs directed against each of these three cytokines suppressed tumorigenicity in the co-injection assay , with the strongest effects observed when silencing CCL2 ( 53% ) or CCL7 ( 66% ) compared to weaker effects exerted by silencing of CCL8 ( 25% ) ( two validated shRNAs per gene , Figure 3 ) . Interestingly , despite their related structure , silencing of each of the three cytokines had a distinct impact on the tumors: silencing of CCL2 suppressed recruitment of innate immune cells and angiogenesis ( Figure 4 ) almost to the same levels as in tumors without co-injected fibroblasts . In contrast , silencing of CCL8 suppressed only the recruitment of innate immune cells , while silencing of CCL7 reduced tumor cell proliferation almost to the levels observed in tumors growing in the absence of co-injected fibroblasts ( Figure 4 and Figure S4 ) . Our data therefore showed that these related chemokines had non-redundant roles in mediating fibroblast-supportive functions . We next tested whether silencing of AREG in fibroblasts affected tumor supportive function . Silencing of AREG had a pronounced effect , with an average reduction in tumor size at six weeks of 55% to 65% ( Figure 5 ) . In contrast to the chemokines , silencing AREG in fibroblasts had no effect on the number of blood vessels or innate immune cells ( Figure 5 ) . However , immunohistochemical analysis of the percentage of cells within the tumor that expressed α-SMA , a marker of mesenchymal-derived cells such as activated fibroblasts , revealed a 78% reduction . This level was comparable to the low levels of α-SMA positive cells in tumors formed in the absence of co-injected fibroblasts ( Figure 5 ) . Although α-SMA is expressed by activated pericytes , a cell type that surrounds endothelial cells of blood vessels , we did not observe a difference in number of α-SMA cells associated with blood vessels ( Figure 5 ) . This suggested that secretion of amphiregulin by the co-injected human fibroblasts plays a major role in establishing a tumor microenvironment that is enriched for activated fibroblasts , and since the co-injected human fibroblasts are almost entirely replaced by mouse fibroblasts at the point of excision ( 6 weeks ) , this effect must be propagated through recruited mouse fibroblasts . Consistent with this , we found that amphiregulin increased proliferation of mouse fibroblasts , as well as human fibroblasts ( Figure 5 ) . We also found that amphiregulin was a chemoattractant for mouse fibroblasts in migration and invasion assays , suggesting a mechanism of recruitment into the tumor , and that amphiregulin directly activated fibroblasts as judged by induction of α-SMA expression ( Figure 6 ) . In addition to the pronounced reduction of activated fibroblasts when AREG was silenced , we also noted changes in activation of the amphiregulin receptor EGFR ( MIM: 131550 ) on tumor cells . Compared to tumor cells co-injected with control tumor-supportive fibroblasts , tumor cells when co-injected with AREG-silenced fibroblasts showed a <2-fold reduction of activated , phospho-EGFR as measured by immunohistochemistry ( Figures 6 ) . This was accompanied by a minor , barely significant ( p = 0 . 06 ) , negative effect on tumor cell proliferative rate in vivo , with no direct effect of amphiregulin on tumor cell proliferation in vitro ( Figure 5 and Figure S5 ) . In contrast , we found a significant increase in necrosis in tumors after silencing of fibroblast-secreted amphiregulin ( Figure 6 ) . Notably , amphiregulin significantly protected the breast cancer cells from cell death induced by detachment from their normal extracellular matrix ( anoikis , Figure 6 ) . Together , these results suggest that fibroblast-secreted amphiregulin has potent effects on tumor progression , with autocrine effects leading to activation of fibroblasts and paracrine effects protecting cancer cells from cell death . Co-culturing also lead to changes in gene expression in the breast cancer cells , which upregulated a shared receptor for CCL2 and CCL7 , the chemokine receptor CCR1 upon co-culturing with tumor-supportive fibroblasts ( Figure 7 ) . We therefore tested whether CCR1 expression by cancer cells was critical for some of the tumor supportive functions of fibroblasts by stably expressing shRNAs targeting CCR1 in Cal51 breast cancer cell line ( knockdown efficiency was quantified by both qRT-PCR and immunoblotting , Figure S6 ) . Silencing of CCR1 had no effect on tumor growth when cancer cells were injected alone . However , in the context of co-injection with fibroblasts , silencing of cancer cell CCR1 resulted in a 3-fold reduction in tumor size , almost eliminating the effect of the fibroblasts ( Figure 7 and Figure S6 ) . We further observed a 40% reduction in the proliferative index of cancer cells and a 2-fold reduction in recruitment of neutrophils and inflammatory monocytes to the tumor , but only minor effects on the number of blood vessels and mesenchymal cells ( Figure 7 and Figure S6 ) . Thus , expression of CCR1 by cancer cells plays a critical role in enabling fibroblasts to exert tumor supportive function , through increased tumor cell proliferation and potentially indirectly through recruitment of leukocytes . We next asked whether blocking two interactions between fibroblasts and cancer cells was more effective than blocking either one alone . We therefore asked whether simultaneously silencing fibroblast-secreted amphiregulin and cancer cell expressed CCR1 was more efficacious than blocking either alone . We found that tumor growth was significantly more reduced when both pathways were targeted ( Figure 7 ) . A number of studies have implicated specific cancer-cell secreted factors in the activation of neighboring fibroblasts , including TGF-β [40] and IL-1β [19] . We wanted to determine whether the repertoire of tumor-promoting fibroblast secreted factors could be induced by single specific inducers , or whether multiple pathways were acting in parallel . To test this , we used quantitative RT-PCR to measure gene expression changes in tumor-promoting fibroblasts of the five factors ( CCL-2 , -7 , -8 , AREG , and STC1 ) that we had determined all had functional relevance , along with two others factors ( NRG1; MIM: 142445 and WISP1 ) that also were induced in the system . Surprisingly , TGF-β , did not induce expression of any of the seven factors ( Table 2 ) , despite its ability to induce activation of fibroblasts to myofibroblasts , which resembles many aspects of carcinoma-associated fibroblasts [10] . AREG also did not induce expression of any of the seven factors , despite its key role in stimulating mammary fibroblasts during normal mammary development [37] . IL-1β , a potent activator of NF-κB signaling , produced an upregulation in chemokines CCL-2 , -7 , -8 similar to that seen by co-culture with breast cancer cells ( Table 2 ) . However , IL-1β did not induce expression of WISP1 , STC1 , AREG , or NRG1 . Interestingly , a combination of IL-1β and AREG lead to significant upregulation of WISP1 in addition to stimulation of CCL-2 , -7 , -8 . In contrast , none of the fibroblast factors were induced when TGF-β and IL-1β were combined ( Table 2 ) . Based on these results , it seems likely that the ability of breast cancer cells to induce the full spectrum of pro-tumorigenic fibroblast-secreted proteins involves a multitude of interacting factors , including some not previously identified . We also found that co-culture of tumor-promoting fibroblasts with a normal , non-malignant breast epithelial cell line , MCF10A , was able to induce fibroblast expression of two out of the seven factors induced by breast cancer cells , namely AREG and WISP2 ( Table 2 ) . For these two factors , it would appear that breast epithelial cells per se , regardless of tumorigenic properties , elicit the same response in fibroblasts . This result is consistent with the key role that stromal AREG plays in normal mammary development [37] .
Numerous studies have used the co-injection assay to identify the factors produced by fibroblasts that are responsible for promoting tumorigenicity . These studies have identified single factors that when inhibited strongly suppress the ability of activated fibroblasts to promote tumorigenicity , and they include scatter factor , SDF-1 , MMP14 , NF-κB , osteopontin , TGF-β , and CCL2 [12]–[17] . However , these studies did not address whether the single factors were unique in their capacity of promoting fibroblast-supported tumorigenicity , nor did they compare different factors . Here , we used comparative genomics to identify 42 candidate mediators of fibroblast-promoted tumorigenicity and we tested the functional impact of a set of five of these factors . Surprisingly , we found that four of the five tested factors promoted tumorigenicity , three of them strongly ( CCL7 , CCL2 , and AREG ) and one of them weakly ( CCL8 ) ( Table 3 ) . Although the fifth factor STC1 significantly affected tumorigenicity by the area under the tumor growth curve test ( Table 3 ) , it failed to show significant effects using t-tests at any single time points ( data not shown ) . Since we only tested five of the 42 fibroblast-secreted-factors that were induced by breast cancer cells , it seems highly likely that an even greater number of fibroblast-secreted factors play a role in promoting tumorigenicity . Thus , our study indicates that even in a single system there are a large number of secreted factors involved in the ability of fibroblasts to promote carcinomas , rather than a single important mediator . Intriguingly , our results also indicate widely diverse mechanisms for fibroblast-secreted factors in the promotion of tumorigenicity . The strong effects of fibroblast CCL7 appeared to be caused by a significant effect on cancer cell proliferation ( Table 3 ) , which was unique among the factors tested . We also found that reducing cancer cell expression of the CCL7 receptor , CCR1 , also reduced fibroblast-induced proliferation . In contrast , the strong tumor promoting effects of fibroblast CCL2 was associated with different effects on the tumor microenvironment , as we found significant decreases in both angiogenesis and recruitment of innate immune cell upon silencing of CCL2 . CCL8 had weaker effects on tumor growth than the other tested chemokines , which likely reflects it inability to affect either tumor proliferation or vascularity ( Table 3 ) . Fibroblast AREG also had very strong effects on tumor growth , and influenced both the total number of activated fibroblasts in the tumors and the survival of the cancer cells , a combination not observed with any other tested factor . Interestingly , secretion of amphiregulin by fibroblasts appeared to potentially act as a chemoattractant to recruit new fibroblasts and induce their proliferation and activation , resulting in a tumor microenvironment that is enriched in activated fibroblasts . Thus , in a single system of carcinoma cells and tumor-supportive fibroblasts , several factors play key roles . Our study also sheds light on how cancer cells modify the stromal cells to enable them to promote tumorigenicity . We tested several cancer cell-secreted factors , previously reported to influence stromal cells , but none of them were able to induce the full panel of verified tumor-promoting fibroblast factors as well as the cancer cells themselves . Even in one simple system , it appears that cancer cells act on fibroblasts through multiple factors , resulting in the secretion of another complex set of factors that influence cancer cells and other components of the tumor microenvironment . One of the key findings of our study was that inhibiting multiple interactions between cancer cells and fibroblasts is more efficacious than blocking individual pathways . This finding is not completely unexpected in light of the complexity of the tumor microenvironment and in fact previous reports have suggested this as a possibility [20] , [41] . Nevertheless , this result highlight that the different interactions are non-redundant and act in parallel , but it also suggests that effective anti-stromal fibroblast therapeutic strategies can be achieved by taking a multi-targeted approach . Future research towards this end will need to employ models that closely resemble the type of tumors to be targeted in human patients . This presents many challenges , including the ability to selectively target endogenous fibroblasts in tumor tissues , along with the ability to monitor the in vivo response of tumor-associated fibroblasts . Despite these challenges , our study shows that there are several potential combinatorial targets for future fibroblast-targeted therapeutic approaches .
All genomic data for this study , including expression analysis of both fibroblasts and breast cancer cells , have been deposited in the Gene Expression Omnibus ( GEO ) repository ( GSE41678 ) . http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? token=bnepxkssyoamgdu&acc=GSE41678 Breast cancer cell lines and human fibroblast strains were obtained from ATCC ( Manassas , VA ) , DSMZ ( Braunschweig , Germany ) or Sigma ( St . Louis , MO ) and grown under standard tissue culture conditions in growth medium recommended by the supplier . For dual-color co-culture experiments , breast cancer cells stably expressing Discosoma sp . red fluorescent protein ( DsRed ) and human fibroblasts stably expressing Zoanthus sp . green fluorescent protein ( ZsGreen ) were generated using pRetroX-IRES-DsRedExpress vector and pRetroX-IRES-ZsGreen1 vector ( Clontech , CA ) respectively via retroviral transduction ( detailed description in Supplemental Experimental Procedures ) . None of the experiments utilized multiple retroviral transfections . For the co-culture experiments , we transfected the original fibroblasts with a fluorescent marker in order to facilitate cell separation before transcriptome analysis ( the breast cancer cells were transfected with a different fluorescent marker ) . For the shRNA experiments , we transfected the original fibroblasts with validated shRNA constructs from the Broad library . Co-culture of fluorescently-tagged breast cancer cells and fibroblasts was initiated by plating 1 . 5 million fibroblasts into 10 cm dishes , and after 18 hours 1 million breast cancer cells were added and then incubated for six days . Monocultures were performed in parallel for the same duration . Following co-culture or mono-culture , cells were trypsinized and resuspended in FACS sorting buffer ( PBS+1% FBS ) for separation into DsRed+ and ZsGreen+ populations using an ARIA II flow cytometer and analyzed using FACS DiVA software ( Becton Dickenson , CA ) . Total RNA was isolated using RNeasy kit ( Qiagen , Netherlands ) and hybridized to Gene 1 . 0 ST arrays ( Affymetrix , CA ) . Data was extracted , background corrected , normalized , and converted from probe values to gene values using the AROMA R package ( www . aroma-project . org ) . All studies utilizing human xenograft experiments were approved by and in accordance with Cold Spring Harbor Laboratory's Institutional Animal Care and Use Committee . Five to six week old female nude mice ( NCR nu/nu; Charles River Inc . , Wilmington , MA ) were irradiated at 400 cGy 24–36 hours prior to injections . One million breast cancer cells were trypsinized , resuspended with or without 1 . 5 million fibroblasts in 100 µl DMEM and injected subcutaneously into both flanks of irradiated , nude mice . Growth was followed over time by taking caliper measurements at indicated time points . Tumor volume was measured as 0 . 52×length×width2 . Tumors were excised six-eight weeks post injections or when one of the measurements reached 2 cm . Immunostaining procedure is described in detail in Supplementary Experimental Procedures . The primary antibodies used for immunostaining are as follows: α-SMA ( 1∶2000; # 1A4; Sigma-Aldrich ) , CD31 ( 1∶100; ab28364; Abcam ) , antigen 7/4 ( 1∶400 , CL8993AP , Cedarlane ) , Ki-67 ( 1∶2000; MIB5; Dako ) , pEGFR ( 1∶100; 1138-1; Epitomics ) and GFP ( 1∶1000; ab290; Abcam ) . Immunostained slides were quantified by counting ( for 7/4 , Ki-67 and CD31 ) , by percentage of stained area ( for α –SMA and pEGFR ) using Image J software ( NIH , Bethesda , MD ) . Additional methods are found in the file Text S1 . | There is increasing interest in developing methods to treat cancer by targeting non-cancer cells that play supportive roles in the tumor microenvironment . One type of non-cancer cell that has received considerable attention along these lines is cancer-associated fibroblasts , which can promote tumor formation and tumor growth . There have been several studies showing that inhibition of individual fibroblast genes or proteins dramatically reduces the tumor supportive function of fibroblasts . From the perspective of developing a therapeutic strategy , what remains unclear is whether the several different important factors discovered to date reflect the requirement of a multitude of fibroblast factors to promote tumorigenicity , or whether it reflects the diversity of the epithelial cancer cells and fibroblasts used in these different studies . Here , we addressed this question directly using a single system of fibroblasts and breast cancer epithelial cells . Importantly , we found that a multitude of fibroblast factors are indeed required to promote tumorigenicity , and that they have different effects on the tumor microenvironment . Furthermore , we found that inhibiting multiple fibroblast-secreted factors is more efficacious than blocking individual factors . These results suggest that fibroblasts and cancer cells act through multiple parallel pathways and that effective anti-stromal therapeutic strategies will need to be multi-targeted . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | System-Wide Analysis Reveals a Complex Network of Tumor-Fibroblast Interactions Involved in Tumorigenicity |
Polycomb group ( PcG ) and trithorax group ( trxG ) proteins are conserved chromatin factors that regulate key developmental genes throughout development . In Drosophila , PcG and trxG factors bind to regulatory DNA elements called PcG and trxG response elements ( PREs and TREs ) . Several DNA binding proteins have been suggested to recruit PcG proteins to PREs , but the DNA sequences necessary and sufficient to define PREs are largely unknown . Here , we used chromatin immunoprecipitation ( ChIP ) on chip assays to map the chromosomal distribution of Drosophila PcG proteins , the N- and C-terminal fragments of the Trithorax ( TRX ) protein and four candidate DNA-binding factors for PcG recruitment . In addition , we mapped histone modifications associated with PcG-dependent silencing and TRX-mediated activation . PcG proteins colocalize in large regions that may be defined as polycomb domains and colocalize with recruiters to form several hundreds of putative PREs . Strikingly , the majority of PcG recruiter binding sites are associated with H3K4me3 and not with PcG binding , suggesting that recruiter proteins have a dual function in activation as well as silencing . One major discriminant between activation and silencing is the strong binding of Pleiohomeotic ( PHO ) to silenced regions , whereas its homolog Pleiohomeotic-like ( PHOL ) binds preferentially to active promoters . In addition , the C-terminal fragment of TRX ( TRX-C ) showed high affinity to PcG binding sites , whereas the N-terminal fragment ( TRX-N ) bound mainly to active promoter regions trimethylated on H3K4 . Our results indicate that DNA binding proteins serve as platforms to assist PcG and trxG binding . Furthermore , several DNA sequence features discriminate between PcG- and TRX-N–bound regions , indicating that underlying DNA sequence contains critical information to drive PREs and TREs towards silencing or activation .
Polycomb group ( PcG ) and trithorax group ( trxG ) proteins are conserved chromatin factors that maintain , respectively , the memory of inactive or active states of homeotic genes throughout development . They also regulate many other target genes ( reviewed in [1] ) and misregulation of PcG and trxG genes leads to loss of cell fates , aberrant cell proliferation and tumorigenesis . Moreover , PcG and trxG factors play an important role in diverse epigenetic processes such as stem cell pluripotency and plasticity , genomic imprinting , and X chromosome inactivation [2] . In Drosophila , PcG and trxG proteins are recruited to chromatin by regulatory DNA elements called PcG and trxG response elements ( PREs and TREs , respectively ) . These elements were shown to drive epigenetic inheritance of silent and active chromatin states throughout development [3 , 4] . Biochemical studies on PcG proteins revealed that they exist in at least three distinct multiprotein complexes ( reviewed in [5] ) . PRC2-type complexes contain the four core components E ( z ) ( Enhancer of zeste ) , Esc ( Extra sex combs ) , Su ( z ) 12 ( Suppressor of zeste 12 ) , and Nurf-55 . The SET domain-containing E ( z ) subunit trimethylates lysine 27 of histone H3 ( H3K27me3 ) . This mark is specifically recognized by the chromo domain of Polycomb ( PC ) , a subunit of the PRC1-type complex [6] . PRC1 contains PC , Polyhomeotic ( PH ) , PSC ( Posterior sex combs ) , and the histone H2A ubiquityltransferase dRing , in addition to several other components , including TBP-associated factors [7] . The PhoRC complexes include the sequence-specific DNA binding proteins Pleiohomeotic ( PHO ) or its homolog Pleiohomeotic-like ( PHOL ) , as well as the dSfmbt protein ( Scm-related gene containing four MBT domains ) . Several trxG complexes have been identified: TAC1 ( Trithorax Acetylation Complex ) with the histone methyltransferse Trithorax ( TRX ) , NURF , SWI/SNF , ASH1 , and ASH2 ( for reviews , see [3 , 8] ) . Interestingly , the human TRX homolog MLL1 has been previously shown to be cleaved at two conserved sites by the Taspase1 enzyme , generating an N-terminal and a C-terminal fragment , which can heterodimerize [9 , 10] . However , it is unknown whether the two moieties can have different functions or chromosomal distributions . Additional PcG/trxG proteins have been identified that are not part of the core of these complexes , but are associated with them and , therefore , can be considered as PcG/trxG-associated proteins [11] . These proteins may exist as individual molecules in the cell , but it is also possible that they are part of other protein complexes that contain additional , as yet unidentified PcG/trxG proteins . PcG and trxG complexes ( except PhoRC ) do not bind their target DNA in a sequence-specific manner in vitro , but are recruited to PRE/TRE sequences in vivo . A simple pathway for PcG protein recruitment based on stepwise recruitment of PRC2 proteins by PhoRC , followed by PRC1 recruitment by the H3K27me3 mark deposited by PRC2 has been suggested [12] . However , PcG recruitment seems to be more complex . PHO interacts with PRC2 as well as with the PC and PH subunits of PRC1 in vitro [13] . PHO/PHOL binding sites alone are insufficient to tether PcG proteins to DNA in vivo [14 , 15] , and most PcG sites are stained normally in polytene chromosomes in pho/phol double mutants despite lack of detectable PHO and PHOL proteins [15] . However , PcG protein binding is lost at the bxd PRE in pho/phol double-mutant wing discs [12] , suggesting that the role of PHO and possibly PHOL is important . Other factors have been shown to be involved in recruitment , such as GAGA factor ( GAF ) , Pipsqueak ( PSQ ) , Dorsal switch protein ( DSP1 ) , Zeste , Grainyhead ( GH ) , and Sp1/KLF ( reviewed in [5] ) . Mutations in the corresponding genes do not have a clear PcG phenotype , and intriguingly , all seem to be involved in activation as well as in repression . In summary , many unresolved questions regarding PcG recruitment still remain , and the current model proposes that a combination of several DNA binding factors , and maybe yet-unknown components , could lead to tethering of PcG proteins to DNA . Recently , the distribution of several core components of PcG members and their associated histone modifications has been analyzed in fly as well as mammalian cells [16–22] . Yet , a comprehensive genome-wide binding map of PcG/trxG recruitment factors and of trxG proteins is still lacking . Here , we have generated high-resolution genome-wide binding maps in Drosophila embryos of two PRC1 components and their associated histone mark H3K27me3 , the N- and the C-terminal part of the TRX protein and their associated histone mark H3K4me3 as well as four sequence-specific DNA binding proteins known to be involved in recruitment of Polycomb proteins . Our results show the complementarity between PcG and trxG protein binding in the genome and suggest that multiple DNA binding proteins participate in setting up this PcG and trxG protein distribution .
Using chromatin immunoprecipitation ( ChIP ) in 4–12-h-old Drosophila melanogaster embryos coupled with genome-wide high-density tiling arrays , we mapped the distribution of the PRC1 components: PC and PH , the N- and the C-terminal part of the Trithorax protein ( TRX-N and TRX-C , respectively ) , and the histone H3K27me3 and H3K4me3 marks . We also determined the genome-wide binding profile of GAF , PHO , PHOL and DSP1 , four DNA binding proteins thought to be involved in PcG recruitment . Reproducibility of biological replicates is shown in Figures S1 and S2 . Figure 1 shows an example of the different profiles along part of chromosome 3R including the HOX gene cluster named ANT-C . The statistics on the number and size of regions significantly enriched for various proteins is shown in Figures S3 and S4 , and in Table S1 . As observed previously , PC and H3K27me3 mark covered over 200 large domains ( >5 kb ) , most of which contain discontinuous subregions with significant p-values for enrichment separated by small intervening subregions that were enriched although their p-values were not significant ( see Text S1 for a precise definition of H3K27me3 and PC domains ) . The number of significantly enriched subregions for PC and H3K27me3 were 2 , 110 and 2 , 480 , respectively . Nearly all PH binding sites fall into PC- and H3K27me3-bound regions ( Figure 2A ) . The sequence-specific DNA binding proteins PHO , PHOL , DSP1 , and GAF are bound to thousands of genomic sites ( Table S1 ) . Surprisingly , whereas PcG binding sites strongly predict the presence of one or more of the DNA binding factors , the converse is not true . In fact , the sequence-specific DNA binding proteins are more frequently bound to sites bound by TRX-N and trimethylated on H3K4 ( see Figure 2B ) . Binding of the N-terminal fragment and the C-terminal fragment of TRX ( TRX-N and TRX-C , respectively ) correlates well at the genome-wide level ( Figure S5 ) , but the relative intensities are very different . TRX-N is significantly bound to 4 , 868 genomic sites , with strong binding correlated to H3K4me3-bound regions ( Figure 2; a total of 4 , 893 regions contained H3K4me3 ) . At most of these sites , TRX-C binding levels are higher than background , but not picked up as significant . Strong binding of TRX-C is only identified at 167 genomic sites , mainly located in PRC1-bound regions ( Figure 2C ) where TRX-N binds weakly if at all . All the profiles are available at an online browser at the address http://purl . oclc . org/NET/polycomb . This browser also contains data from earlier mapping studies [20 , 22] and from transcription profiling of staged embryos [23] . In addition , it contains the annotation of predicted PREs ( M . Rehmsmeier , personal communication [24 , 25] ) , whose genomic location can be visualized along with the significantly enriched regions and with the results from our sequence analysis . Recent analysis of H3K4me3 and H3K27me3 in mouse and human cells revealed the coexistence of these two marks in a large fraction of the H3K27me3 regions [26–29] . These regions encompass most of the H3K27 trimethylated sites in embryonic stem ( ES ) cells and a substantial portion of them in differentiated cells . Although we do frequently observe H3K4me3 occupancy at transcription start sites ( TSSs ) flanking PH sites , this is almost exclusively observed at the boundary of large H3K27me3 domains ( see Text S1 ) . From a total of 4 , 893 H3K4me3 and 2 , 480 H3K27me3 regions , only 161 had an overlap , i . e . , only 6 . 5% of the H3K27me3 regions . Considering that most of the genes identified by these regions of overlap are expressed only in a fraction of the embryonic cells , we believe that most of these cases reflect a mixture of cell populations rather than true bivalency . Moreover , the H3K4me3 profile always showed sharp peaks at promoters within large H3K27me3 regions , in contrast to mammalian cells in which bivalent domains often show similar profiles with H3K4me3 and H3K27me3 spread over regions of several kilobases in size . Thus , our data suggest that H3K4me3 and H3K27me3 are generally exclusive in the fly genome . Nevertheless , individual cases of true bivalency may exist in fly embryos or at other developmental stages . A rigorous demonstration of this point will require sequential ChIP with mononucleosomal chromatin and antibodies directed against the H3K4me3 and H3K27me3 marks . We sought a comprehensive characterization of the joint distribution of PcG and trxG factors and associated marks . Many of the data tracks are highly correlated among themselves ( Figures S5 and S6 ) , and are also tightly associated with other spatial genomic features like TSSs . We therefore developed a new method for dissecting a multivariate genomic profile into a hierarchy of “spatial clusters . ” Briefly , “spatial clustering” can be viewed as the genomic analog of gene clustering , since it dissects the genome into clusters that share a common profile across all experimental tracks ( detailed information is given in the Text S1 ) . Unlike gene clustering , our model takes into account the genomic layout of the data , and organizes clusters spatially to probabilistically describe the typical genomic order among them . We used the clustering results ( Figure 3 ) as a blueprint for our dataset , validating conclusions by running an independent , supervised data analysis . An example of cluster organization is illustrated in Figure S7 . Analysis of the distribution of cluster location with respect to the TSS further demonstrates how the clusters are organized around genes ( Figure 3B , note that TSS data were not used by the algorithm to define clusters ) . As shown in Figure 3 , our data reflect two levels of genomic organization . First , the genome is partitioned into three superclusters . Consistent with the mutually exclusive distribution of H3K27me3 and H3K4me3 , unsupervised spatial clustering identifies a “H3K27me3-marked” supercluster and “H3K4me3-marked” supercluster , in addition to regions with no particular epigenomic enrichment ( “background” supercluster , not shown in Figure 3 ) . Second , each supercluster is subdivided into distinct clusters , and the model identifies the connections between clusters that organize the entire genome ( Figure S18 ) . The H3K27me3 superclusters are anchored around clusters characterized by high levels of PH binding ( labeled as “PH sites” ) . These clusters include also strong PHO enrichment , presence of the recruiter factors GAF and DSP1 and TRX-C occupancy . All of the PH site clusters in the BX-C , the ANT-C , the ph , the hh , and the en genes were previously identified as PREs , suggesting that in general , most of the PH clusters are indeed PREs . The H3K27me3 supercluster also included three clusters with lower levels of PC and a general lack of PH and cofactors . We labeled them as “Strong , ” “Medium , ” and “Weak” PC clusters . Similarly , the H3K4me3-marked supercluster was subdivided by the algorithm into four clusters . These clusters reflect clear organization around annotated TSSs , as identified by their TSS enrichment statistics ( Figure 3B ) and binding preferences ( Figure 3D ) . We denoted the cluster with the most 5′ enrichment as the “K4me3-recruiters” cluster . It is characterized by high levels of GAF , DSP1 , and significant , but weaker levels of PHO and PHOL , as well as medium to weak H3K4me3 levels . Enriched exactly at the TSS is the “K4me3-TSS” cluster with high H3K4me3 levels in combination with high levels of TRX-N , PHO and PHOL . The K4me3 cluster has only high levels of H3K4me3 and represent the region downstream the TSS , whereas the “weak K4me3” cluster shows low , but significant levels of H3K4me3 alone and is more weakly enriched around TSSs . PC and H3K27me3 were bound in large regions , often greater than 5 kb , with the largest ones spanning several hundred kilobases ( see Figures 1 and S4A ) . Globally , H3K27me3 and PC profiles were very well correlated , facilitating the definition of PC domains ( see Text S1 ) , underscoring the significance of the H3K27me3 supercluster ( Figure 3 ) identified by spatial clustering . A similar pattern was observed for PC and H3K27me3 by Schwartz et al . [20] in their genome-wide mapping studies in S2 cells and by Tolhuis et al . who used Kc cells [22] . Nearly all PH peaks were specific to PC and H3K27me3 regions ( the PH sites; Figure 3 ) and were present in all the earlier characterized PREs . The average distribution of H3K27me3 around PH peaks takes a dip at the PH sites ( Figure 4A ) , which may be due to nucleosome depletion at the PREs [20] . The distribution of the domain size , number of PH peaks , and genes in H3K27me3 domains is shown in Figure S4 ( for an identification of candidate PcG target genes , see Text S1 and Table S2 ) . Despite these common features , there are differences in the positions of many of the PcG domains in different biological samples . Although the majority of our 217 H3K27me3 domains also exists in S2 cells , 79 ( 36% ) of them did not overlap any bound regions in S2 cells . These data are corroborated by the analysis of the distribution of the PC protein which , similar to H3K27me3 , forms large domains . In general , H3K27me3 differences between embryos and S2 cells paralleled differences in PC binding . The same was observed in a comparison between ChIP on chip binding of PC from embryos and the PC DamID profile obtained previously in Kc cells [22] . Interestingly , a substantial portion of the PC domains in Kc cells differed from those observed both in embryos and in S2 cells . Thus , many common PC domains are identified in various cell types , but a significant subset of them is cell-type specific rather than constitutive . These data are in agreement with previous studies suggesting that part of the PcG binding is cell-type and developmental-stage specific [19 , 30] . To gain more insight into PRC1 recruitment to chromatin , we examined the distribution of PcG recruitment factors at PH sites that are also bound by PC ( PRC1 sites ) . The combination of different PcG recruitment factors at the PRC1 sites as compared to the genome is listed in Table S3 and shown in Figure 2B . Most PH binding peaks colocalize with the PcG recruitment factor PHO ( 96 . 4% ) ( see Figure 2B and Table 1 ) . DSP1 and GAF were present in about 50% of the PH sites . In contrast , PHOL binding was not common at PH sites , with a frequency ( 21 . 1% ) comparable to that of TRX-N ( 26 . 5% ) . Surprisingly , only a minority of all recruitment factors binding sites ( 3 . 2% to 13 . 5% ) was restricted to PH sites ( Table 1 ) . Comparison with previously published Zeste data [31] showed that a moderate 25% of the Zeste sites colocalized with PH peaks . Together these data suggest a correlation gradient between different recruiters and PREs , with PHO > DSP1/GAF > Zeste/PHOL . The K4me3-recruiter cluster ( including strong GAF and DSP1 and medium to weak H3K4me3 levels ) is located in a position just upstream to the TSS . The K4me3-TSS cluster ( high H3K4me3 levels and strong TRX-N , PHO , and PHOL binding ) is usually following it and is almost exclusively observed over the 2 kb around the TSS . Finally , the K4me3 cluster ( high H3K4me3 levels without TF occupancy ) is enriched 3′ to the TSS . This organization suggests that binding of GAF and DSP1 can promote the activation of a TSS upon binding of TRX-N and the PHO/PHOL factors . Therefore , PH target promoters are strongly bound by PHO and TRX-C and depleted of PHOL and TRX-N ( Figure 4B ) , whereas H3K4me3 promoters are bound by PHO , PHOL , and TRX-N ( Figure 4C ) . Notably , the positions of PHO ( and PHOL ) in the second class of promoters is right at the TSS , whereas at PH-bound promoters , PHO is colocalized with PH upstream to the TSS ( Figure 4D ) . This different architecture may contribute to PH recruitment or to silencing of PH-bound promoters . We further analyzed active promoters and PREs/TREs by analyzing TRX binding . The human TRX homolog MLL1 is cleaved by Taspase1 , generating an N-terminal and a C-terminal fragment , which can heterodimerize in vitro [9 , 10] . Low levels of TRX-N co-occupied PH binding sites in about 26 . 5% of cases ( Figure 4A; Table1 ) . However , TRX-N is present at thousands of other genomic sites , where no PcG binding can be observed . These genomic sites correspond mainly to annotated 5′ ends of genes carrying H3K4me3 peaks slightly offset towards the body of the gene in comparison to TRX-N ( cluster K4me3-TSS; see also Figure 4C ) . Interestingly , although the TRX-C profile overall looks similar to the TRX-N , its relative binding intensities are different . TRX-C is strongly bound at PcG binding sites , whereas low binding is observed at most promoter regions of non-PcG target genes ( Figure 4 ) . These results suggest that whereas the distribution of the N-terminal part of TRX follows a general transcription cofactor role , the C-terminal part is specifically linked to PcG function . PcG proteins might repress transcription by anchoring the C-terminal portion of TRX at PREs . On the other hand , constitutive TRX-C binding at PREs/TREs might allow PcG target genes to switch their state upon strong transcriptional induction . In the case of PHO , PHOL , and GAF , sequence-specific DNA binding in vitro has been shown previously [32 , 33] . By analyzing the collection of statistically significant bound sites for each of these proteins with the Multiple EM for motif Elicitation ( MEME ) algorithm , we detected the expected binding sites ( Figure 5A and 5B , and Tables S6 and S7 ) , whereas for Dsp1 [14 , 34] , the results were not conclusive . The “GAAAA” motif was not strongly enriched among the genomic binding sites for this protein , although a degenerated GAAAA motif was found at DSP1-bound as well as at PHO- and PH-bound regions ( Figures S8–S11 , see Text S1 for a detailed discussion ) . In order to determine whether other sequence features may characterize PREs specifically , we further developed the unsupervised spatial clustering methodology ( Figure 3 ) to allow discovery of sequence motifs that discriminate among clusters or groups of clusters . As shown in Figure 6 , we discovered several known and novel motifs that are either shared among clusters or distinguish them . We visualize these results in terms of the affinities ( or predicted binding energies ) of the inferred position weight matrices ( PWMs ) in and around each our spatial clusters [35] . Two motifs ( GAGA and the CA repeat motif ) are marking clearly the PH sites and the K4me3-recruiter clusters . Three additional motifs are strongly marking the K4me3-TSS cluster and clearly discriminating it from the spatially coupled K4me3-recruiter cluster sites . Two of them are motifs bound by the Myc , Max , and Mad/Mnt proteins [36] and include the DNA replication element ( DRE ) TATCGATA , which is also consensus for several other factors including the TRF2n , Cut , and Beaf-32 . The third motif ( CAGCTG ) is an E-box bound by bHLH proteins [37] which , like DREs , are involved in the regulation of many developmental genes . We note that the detected motif enrichments are specific to the K4me3-TSS cluster and not to general TSSs in the genome since general non–H3K4me3-associated TSSs lack these motifs . Importantly , we also discovered motifs that discriminate between K4me3-recruiters and PH sites . The CAACAACAA motif is enriched around K4me3-recruiters , but not in and around PH sites ( see also Figure S8 ) . On the other hand , the CCGTCGG and the Sp1/KLF-like [38] GGGGTGGG motifs are specific to PH sites and not K4me3-recruiters ( see also Figure S11 ) . These motifs constitute candidates to recruit new DNA-binding factors to PREs . In addition to these motifs , the consensus sites for PHO/PHOL , DSP1 , and GAF are more strongly enriched at the 300-bp core regions around the maximal binding peak of PH than around the other genomic regions bound by the factors without PH ( Table S5–S8 ) . Thus , the density of binding sites is specific to PREs , suggesting that cooperative binding may help recruit PcG proteins . Consistent with this idea , the fold enrichment for each of the factors ( with the exception of PHOL , see below ) is higher at PH-bound regions compared to non–PH-bound regions ( Figures S12 and 4 ) . Of particular interest is the distribution of the PHO motif around PH sites and the K4me3-TSS clusters . Unlike the GAGA ( or CACA ) motif , the frequency of motifs with sequence similarity to consensus PHO motifs is high , but these motifs are not well localized at PH sites . High predicted PHO affinities ( defined by PWMs; see Text S1 ) were also present in the strong PcG clusters surrounding PH sites . This pattern matches perfectly with our ChIP data , which also suggest that PHO levels are regionally high around PH sites . In contrast to this pattern , the K4me3-TSS cluster is characterized by weak , but significant peaks of PHO motifs that were localized right at the TSS . This pattern is again matched by the PHO and PHOL ChIP data at the TSS of H3K4me3 associated promoters ( Figure 4C ) . PHO and PHOL share sequence homology , were shown to bind the same DNA motif in vitro , and have been proposed to play redundant roles in PcG-mediated silencing ( reviewed in [5] ) . Notably , we observed that PHO and PHOL binding patterns do not always overlap in the genome . In particular , PHO binds much stronger than PHOL at PH sites ( Figures 3 , 4 , S13C , and S14 ) , whereas both proteins bind with similar intensities in K4-recruiter and K4-TSS clusters ( Figures 3 and 4 ) . We also noticed that the majority of PHOL sites in the genome colocalized with TRX-N and H3K4me3-bound regions ( Figures 3 and S5; Table S4A ) . To investigate whether PHO and PHOL may fulfill distinct roles in recruitment of PcG and trxG proteins , we computed the genome-wide ratio of PHO/PHOL binding ( see Text S1 ) and plotted it compared to the individual profiles as well as to PH sites . Figure 7A shows that the PHO/PHOL ratio accurately matches the PH distribution profile since the binding of the two proteins at all other sites in the genome cancels out , whereas PHO binding at PREs is much stronger than PHOL . To confirm whether the ratio of PHO/PHOL is linked to the activity state of PRE/TREs , we examined by quantitative ChIP assays the binding levels of PH , PHO , and PHOL at three PcG target genes characterized by ON/OFF expression states in different larval tissues ( Figure 7B–7F ) . Ubx is expressed in haltere/third leg imaginal discs [39] and is repressed in eye imaginal discs ( ED ) . On the contrary , so ( sine oculis ) and toy ( twin of eyeless ) have very low expression in haltere/third leg discs and are highly expressed in eye discs ( Figure S15A ) . For Ubx regulation , we analyzed protein binding levels at the bx PRE , bxd PRE , and the Ubx TSS , and for so and toy , we analyzed their TSS , which overlapped with the PH-bound region ( Figure S15B ) . PH , PHO , and PHOL are bound in all the 5′ regions of the genes that we examined in both the ON and OFF state ( Figures 7 and S16 ) . However , significant differences in binding levels were noticed . In haltere/third leg discs where Ubx is ON , bx PRE , bxd PRE , and Ubx TSS showed a slight decrease in PH binding ( 50% ) as compared to eye discs . Both so and toy TSS showed higher levels of PH binding in haltere/third leg discs , where these genes are silenced ( OFF ) , as compared to eye imaginal discs ( ON ) . At the Ubx TSS and the bx PRE , levels of PHOL were significantly higher in haltere/third leg discs ( ON ) as compared to eye discs ( OFF ) . With regards to PHO , stronger binding was observed at the PREs in eye discs ( OFF state ) , whereas at so and toy stronger binding was observed in haltere/third discs ( OFF state ) compared to eye discs ( ON ) . In summary , a significant decrease in the levels of PH in tissue where target genes are active correlates with a decrease in the PHO/PHOL ratio . On the other hand , increased PH levels at genes that are OFF in a certain tissue correlates well with an increased PHO/PHOL ratio . To further examine the function of the PHO/PHOL ratio in Polycomb-dependent gene silencing , we performed quantitative reverse-transcriptase PCR ( RT-PCR ) on eye , haltere/third leg and wing imaginal discs from wild-type and pho1 homozygous ( null mutant allele of PHO [40] ) third instar larvae . In wild-type eye discs , the Ubx and Antp genes are repressed , and the detection of their transcripts is limited to few copies . In pho1 mutant larval eye discs , Ubx gene becomes derepressed ( 5 . 5-fold ) , and gene activation is even stronger for the Antp gene ( between 10- and 30-fold ) ( Figure 8A ) . These results suggest that the loss per se of PHO has an impact on the level of transcription of Polycomb-silenced target genes , and this underscores its fundamental role in setting up Polycomb-mediated silencing . Binding of PHOL to the same sequence motif in the promoter region of these two genes might partially complement for the loss of PHO . Indeed , we detected increased binding levels of PHOL to chromatin in pho1 mutant imaginal discs ( unpublished data ) . We then analyzed the effect of the pho1 mutation in haltere/third leg discs where the Ubx gene is transcribed and in wing discs where Antp is active . We detected a consistent , yet slight , decrease of their transcripts ( 2-fold and 1 . 5-fold , respectively ) ( Figure 8B ) . These results suggest that PHO may also play a role as an activator of homeotic genes , even if this role is weaker than its silencing function . Because we found a high colocalization of PHO and PHOL with TRX-N at many gene promoters not related to PcG-mediated silencing , we performed quantitative RT-PCR to check the expression of two constitutively transcribed genes such as Chc and Rp49 , which are bound by PHO in wild-type embryos . Again , Chc expression decreased 1 . 6 times in both eye and haltere/third leg discs and Rp49 1 . 3 times in eye discs from pho1 mutant larvae ( Figure 8C ) . In contrast , we could not detect major changes in their expression levels in a phol81A null mutant background ( unpublished data ) , pointing to a redundant role of PHOL in gene activation These results , together with the recent work of Beisel et al . [41] , indicate that PHO is a modulator , not only of PcG-mediated silencing , but also of the active state of many genes .
To date , PREs have been only characterized in Drosophila . These elements are not defined by a conserved sequence , but include several conserved motifs , which are recognized by known DNA binding proteins like GAGA factor ( GAF ) , Pipsqueak ( PSQ ) , Pleiohomeotic and Pleiohomeotic- ( like ) ( PHO and PHOL ) , dorsal switch protein ( DSP1 ) , Zeste , Grainyhead ( GH ) , and SP1/KLF . Our genomic profiles provide a comprehensive view on the potential role of these factors in the establishment of PcG domains . The presence of PHO at all PREs indicates that PHO is a crucial determinant of PcG-mediated silencing , consistent with earlier analysis on one particular PRE [25 , 33 , 42–46] . On the other hand , PHOL and Zeste were bound at a small subset of PREs . Zeste was previously shown to be necessary for maintaining active chromatin states at the Fab-7 ( Frontabdominal-7 ) PRE/TRE [47] . Therefore , Zeste and PHOL may primarily assist transcription rather than PcG-mediated silencing . GAF and DSP1 resemble PHO as they bind to many ( albeit less than PHO ) PREs as well as to active promoters . Supervised DNA motif analysis indicated a higher density of GAF , DSP1 , and PHO binding sites at PREs as compared to other bound regions at non-PH sites . This suggests that cooperative binding of these proteins may provide a platform for PcG protein binding . Moreover , GAF may act by inducing chromatin remodeling [48 , 49] to remove nucleosomes , since the regions bound by PcG proteins show a characteristic dip in H3K27me3 signal that has been attributed to the absence of nucleosomes in those regions [20 , 50 , 51] . These nucleosome depletion sites are the places wherein histone H3 to H3 . 3 replacement takes place [51] . Indeed , several of the Zeste-bound regions and GAGA binding sequences were shown to localize to peaks of H3 . 3 , suggesting the possibility that GAF may recruit PcG components to PHO-site–containing PREs as well as recruit TRX to promoters via nucleosome disruption . In addition to an increased density of motifs for GAF , PHO , and PHOL , unsupervised spatial cluster analysis identified specific motifs that distinguish the PH sites from the K4me3 cluster . Although the identity of the factors binding to these motifs is unknown , this suggests that the DNA sequence of PREs contains much of the information needed to recruit PcG proteins and to define silent or active chromatin states . With this distinction , it may be possible to develop an algorithm to faithfully predict the genomic location of PREs . Earlier attempts to predict PREs in the fly genome have made progress toward this goal , but they are still far from reaching the required sensitivity and specificity [19 , 20 , 22 , 24 , 25] ( see also Tables S9 to S11 ) . The use of a sequence analysis pipeline that is not dependent on prior knowledge was demonstrated here to generate new discriminative motifs with a potential predictive power . The unique genomic organization of PcG domains may suggest that the genome is using , not only local sequence ( high-affinity transcription factor binding sites located at the binding peaks ) information to determine PREs , but also integration of regional sequence information ( stronger affinity on 5 kb surrounding PREs ) . Using such regional information to predict PREs may break the current specificity and sensitivity barriers . Our ChIP on chip data showed that PHO binding comes in two distinct flavors . In one class of target sites , PHO binding coincides with PH sites within PC domains , whereas outside these domains , it is largely colocalized with PHOL , TRX-N , and H3K4me3 ( Table S4 ) . PHOL binding was weaker at PH sites and was mainly present along with marks associated with gene activation . Quantitative ChIP assays ( Figure 7 ) revealed that PH , PHO , and PHOL were bound in PREs/TSS of their target genes in both ON and OFF states , but the ON state was marked by a decrease in PH binding and a corresponding increase in PHOL levels , whereas the OFF state was characterized by an increase in both PH and PHO binding levels . Papp and Muller [39] analyzed chromatin at the Ubx TSS , the bx PRE , and the bxd PRE ( the same primers were used in our study ) by comparing haltere/third leg imaginal discs ( ON state ) with wing imaginal discs ( OFF state ) . They found a 50% reduction of PH binding levels at the bx PRE , a minor decrease at bxd , and no change in the Ubx TSS . Our ChIP experiments demonstrated a 50% decrease in PH levels at bx PRE and at the Ubx TSS and a minor decrease at bxd PRE when comparing haltere/third leg imaginal discs to eye imaginal discs . We also observed a slight decrease in the levels of PHO in haltere/third leg disc ( ON state ) as compared to eye imaginal discs ( OFF state ) at the bx and bxd PRE , whereas Papp and Muller [39] did not see differences in the levels of PHO . The most likely explanation for these discrepancies is that the peripodal membrane cells of the wing imaginal discs express Ubx , whereas all cells silence this gene in eye imaginal discs . In pho1 mutant eye discs , the absence of PHO causes derepression of the homeotic genes Ubx and Antp . However , the expression levels in pho1 mutants are still much weaker compared to tissues where these genes are normally expressed . This low degree of activation could be explained by compensatory binding of PHOL to the PHO sites in order to maintain PcG-mediated silencing , even if the PHOL-dependent rescue function is incomplete as pho1 mutants die as pharate adults . PHO and PHOL have indeed been described as redundant in their role in PcG-mediated silencing since they bind to the same DNA sequence motif in vitro . However , out of the 1 , 757 places wherein both PHO and PHOL were significantly bound , only 807 shared the same local maxima ( 46% ) . Another 559 ( 32% ) peaks were within 250 bp of each other . This suggests that , in vivo , these two proteins prefer slightly different sequences , with PHO more strongly attracted to PREs , whereas PHOL binds better to promoters . Moreover , PHO interacts directly with PC and PH [13] , as well as with the PRC2 components E ( z ) and Esc , whereas PHOL only interacts with Esc in yeast two-hybrid assays [12] . Stronger interactions between PHO and PcG components may stabilize PHO binding at PREs , favoring it over the binding of PHOL . It is thus possible that the primary function of PHOL is as a transcription cofactor , and that its recruitment to PREs is subsidiary to PHO . Here , we report for the first time , to our knowledge , the genome-wide distribution of TRX . This protein has been proposed to counteract PcG-mediated silencing [52] . Petruk et al . [53] demonstrated that TRX colocalizes with Polymerase II and elongation factors in Drosophila polytene chromosomes . They then showed that PcG and TRX proteins bind to a PRE mutually exclusively in salivary gland chromosomes [54] . In contrast , two other studies [39 , 41] found binding of TRX at discrete sites at PREs and promoter regions of HOX genes , and suggested that TRX coexists with PRC1 components at silent genes . We postulated that these differences might be explained by the use of different TRX antibodies , one against the N-terminal domain [53] and one against the C-terminal domain of TRX [39 , 41] . Notably , the TRX protein is proteolytically cleaved into an N-terminal and a C-terminal domain [10] , but the fate of the two moieties after cleavage has never been addressed in vivo . Our genome-wide mapping studies using the same antibody against the N-terminal fragment ( TRX-N ) as used by Petruk et al . [53] , showed that the binding affinity of the N-terminal fragment to PREs is rather weak , whereas TRX-N binds thousands of promoter regions trimethylated on H3K4 , indicating a general role of TRX-N in gene activation . In contrast , ChIP on chip profiling using an antibody against the C-terminal TRX fragment showed high binding levels at PRE/TREs , whereas binding to promoter regions ( where the TRX N-terminal fragment is strongly bound ) is rather weak . The strong quantitative correlation between the binding intensities of PH and TRX-C suggests that TRX-C can indeed bind to silent PcG target genes . These data are confirmed by the colocalization of PH and TRX-C at inactive Hox genes in salivary gland polytene chromosomes and in diploid cell nuclei ( as seen in a combination of DNA fluorescent in situ hybridization ( FISH ) and immunostaining; unpublished data ) . Thus , PcG silencing may involve locking the C-terminal portion of TRX in an inactive state that perturbs transcription activation events . The fact that TRX is recognized by two different antibodies that recognize PREs ( H3K4me3-depleted regions ) or TSSs suggests that these antibodies reflect the activity state of the protein and thus represent a powerful tool to study the switching of genes between silencing and activation . Similar to mapping studies in Drosophila cell lines , H3K27me3 also forms large domains in Drosophila embryos . These large PcG domains could provide the basis of a robust epigenetic memory to maintain gene expression states during mitosis . As previously suggested [55] , stably bound PcG complexes at PREs may loop out and form transient contacts with neighboring chromatin , which become trimethylated on H3K27 . H3K27me3 might then attract the chromodomain of the PC protein , which may be occasionally trapped at these remote sites by cross-linking mediated by the chromodomain of PC . Alternatively , PcG subcomplexes missing some of the subunits might spread from the PRE into flanking genomic regions containing H3K27me3 histones . Although genome-wide PcG profiles in Drosophila embryos correlate well with profiles from Drosophila cell lines , it has recently been shown that PcG protein binding profiles are partially remodeled during development [19 , 30] . Comparison of our PcG target genes ( Figure S19 and Tables S14–S16 ) with Schwartz et al . [20] showed that 40% of our targets were unique ( Figure S17 ) . The fact that a consistent number of targets are only found in one or two of the samples indicates tissue specific PcG occupancy . Thus , although PcG proteins have been often invoked as epigenetic gatekeepers of cellular memory processes , they may be involved as well in dynamic gene regulation during fly development [19 , 56] , similar to their function in mammalian cells .
All antibodies used in this study are listed in Table S12 . ChIP assays were performed on 4–12-h-old embryos of the Oregon-R w1118 line of Drosophila melanogaster . The complete experimental details of the ChIP experiments are available in Text S1 . Briefly , ChIP samples were amplified by ligation-mediated ( LM ) PCR , as described previously [19] , and hybridized to whole-genome tiling arrays manufactured by NimbleGen Systems ( the array design is described in Text S1 ) . A list of all significantly enriched regions ( p-value < 0 . 0001 ) for all profiles are shown in Table S17 . Spatial clustering was performed by training a Hidden Markov Model ( HMM ) to fit the available genomic profiles using a small set of clusters . The HMM represents both the relations between clusters and the joint profile distribution emitted from each cluster . We developed a hierarchical version of the algorithm so that the two layers of genomic organization in the data can be characterized ( for details , see Text S1 ) . We further enhanced the spatial clustering framework to search for motifs that discriminate among clusters . We also used the MEME and Motif Alignment and Search Tool ( MAST ) programs to search for enriched motifs directly [57 , 58] ( a detailed description can be found in Text S1 ) . ChIP assays of imaginal discs were performed as described for embryos with the following modifications: third instar larval eye discs and haltere/third leg discs were dissected in SS M3 insect medium and kept on ice during dissection . A hundred discs were used per immunoprecipitation ( IP ) . Discs were pelleted by centrifugation at 4 . 000 g for 5 min , resuspended in 1 ml of Buffer A1 , and then cross-linked for 15 min in the presence of 1 . 8% formaldehyde by homogenization in a Tenbroeck homogenizer . Chromatin was sonicated using a Bioruptor ( Diagenode ) for 12 min ( settings 30 s on , 30 s off , high power ) . Sheared chromatin had an average length of 500 to 1 , 000 bp . Antibodies used for IP ( PHO , PHOL , and PH ) were diluted 1:100 ( PH and PHO ) or 1:20 ( PHOL ) . Enrichment of specific DNA fragments was analyzed by real-time PCR , using Roche Light Cycler equipment and accessories as described in Comet et al . [59] . Enrichment in specific IPs was determined by normalizing the amount of DNA obtained in each reaction by the amount of a negative control fragment from the robo3 gene . Primer sequences are listed in Table S13 . pho1 homozygous larvae were collected from a stock ey-GAL4/ey-GAL4; pho1/GS15194 kindly provided by R . Paro's lab [41] . Wild-type and pho1/pho1 mutant larvae were dissected in PBS , and 40 eye or haltere/third leg discs were taken for RNA isolation using TRIzol reagent ( Invitrogen ) . RT-PCR was performed using Superscript III First Strand Synthesis Kit from Invitrogen following the manufacturer's instructions . Reverse transcription was primed using hexamer primers . Quantitative polymerase chain reaction ( qPCR ) analysis was done as described for ChIP experiments . The copy number for each investigated gene was normalized to the copy number of the 18S RNA gene . Primer sequences are listed in Table S13 . Experiment , first part ( combined replicates; K27 , PC , PH , PHO , DSP1 , PHOL , GAF , TRX-N , and K4 ) : E-MEXP-1708 . Additionally , all employed microarray designs have their own accessions: PhysicalArrayDesign name: 2005-08-08_Henikoff_Dros_ChIP_1 , ArrayExpress accession: A-MEXP-1251; PhysicalArrayDesign name: 2005-08-08_Henikoff_Dros_ChIP_2 , ArrayExpress accession: A-MEXP-1252; PhysicalArrayDesign name: 2005-08-08_Henikoff_Dros_ChIP_3 , ArrayExpress accession: A-MEXP-1253; PhysicalArrayDesign name: 2007-03-13_Henikoff_Dros_ChIP_1 , ArrayExpress accession: A-MEXP-1254; PhysicalArrayDesign name: 2007-03-13_Henikoff_Dros_ChIP_2 , ArrayExpress accession: A-MEXP-1255; PhysicalArrayDesign name: 2007-03-13_Henikoff_Dros_ChIP_3 , ArrayExpress accession: A-MEXP-1256; and PhysicalArrayDesign name: Cavalli_Dmel_1_tiling , ArrayExpress accession: A-MEXP-1257 . Gene accession numbers: Antp: FBgn0000095; ato: FBgn0010433; cad: FBgn0000251; Chc: FBgn0000319; Dll: FBgn0000157; dsx: FBgn0000504; grn: FBgn0001138; hb: FBgn0001180; robo3: FBgn0041097; Rp49: FBgn0002626; so: FBgn0003460; toy: FBgn0019650; and ubx: FBgn0003944 . | Although all cells of a developing organism have the same DNA , they express different genes and transmit these gene expression patterns to daughter cells through multiple rounds of cell division . This cellular memory for gene expression states is maintained by two groups of proteins: Polycomb-group proteins ( PcG ) , which establish and maintain stable gene silencing , and trithorax group proteins ( trxG ) , which counteract silencing and enable gene activation . It is unknown how this balance works and how exactly these proteins are recruited to their target sequences . By mapping the genome-wide distribution of PcG and trxG factors and proteins known to recruit them to chromatin , we found that putative PcG recruiters are not only colocalized at PcG binding sites , but also bind to many other genomic regions that are actually the binding sites of the Trithorax complex . We identified new DNA sequences important for the recruitment of both PcG and trxG proteins and showed that the differential binding of the recruiters PHO and PHOL may discriminate between active and inactive regions . Finally , we found that the two fragments of the Trithorax protein have different chromosomal distributions , suggesting that they may have distinct nuclear functions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"cell",
"biology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] | 2009 | Functional Anatomy of Polycomb and Trithorax Chromatin Landscapes in Drosophila Embryos |
The use of consumer-grade wearables for purposes beyond fitness tracking has not been comprehensively explored . We generated and analyzed multidimensional data from 233 normal volunteers , integrating wearable data , lifestyle questionnaires , cardiac imaging , sphingolipid profiling , and multiple clinical-grade cardiovascular and metabolic disease markers . We show that subjects can be stratified into distinct clusters based on daily activity patterns and that these clusters are marked by distinct demographic and behavioral patterns . While resting heart rates ( RHRs ) performed better than step counts in being associated with cardiovascular and metabolic disease markers , step counts identified relationships between physical activity and cardiac remodeling , suggesting that wearable data may play a role in reducing overdiagnosis of cardiac hypertrophy or dilatation in active individuals . Wearable-derived activity levels can be used to identify known and novel activity-modulated sphingolipids that are in turn associated with insulin sensitivity . Our findings demonstrate the potential for wearables in biomedical research and personalized health .
Public adoption of consumer-grade wearable activity trackers ( “wearables” ) has been steadily increasing in recent years [1 , 2] , and it is estimated that the global market for wearables will exceed $34 billion US$ by 2020 [3] . Basic activity trackers provide accelerometer-based activity data , whereas more sophisticated models are also capable of monitoring heart rate ( HR ) . Together , these indicators have the potential to provide deep insights into an individual’s cardiovascular health and fitness . For instance , resting heart rate ( RHR ) is an important indicator of cardiovascular health [4–6] , whereas step counts can be used to infer patterns and levels of physical activity . Both metrics play roles in the modulation and prediction of risk of cardiovascular and metabolic disorders ( CVMDs ) [7] . Given the role played by physical activity in determining health outcomes , there has been great interest in the use of wearables in healthcare . Most research on wearables thus far has been focused on their utility in promoting increased physical activity in healthy and diseased populations [8] . Whereas most studies reported increased physical activity after wearable introduction [8] , there is little evidence that this intervention results in clinically significant health outcomes . For instance , a year-long study conducted on corporate employees showed that although wearable introduction increased moderate to vigorous physical activity , it did not improve health outcomes [9] . Furthermore , only around 10% of participants were still using their wearable at study conclusion [9] . More recent studies have also started exploring how wearable data correlate with clinical and biological markers . In one study monitoring wearable data from 43 individuals over an average of 5 months , the authors showed that disease states and physiological differences between individuals ( e . g . , insulin sensitivity and inflammation ) could be discerned from the data [10] . Another study seeking to determine how comprehensive personal data collected from 108 individuals correlated with physiology and disease did not identify any significant correlations with wearable data [11] . There are also studies exploring the use of time series HR data from wearables in the detection of conditions associated with cardiovascular disease such as atrial fibrillation ( AF ) , sleep apnea ( SA ) , and hypertension [12–14] . For example , deep neural networks ( DNNs ) trained on HR and step count data obtained from the Apple Watch ( Apple , www . apple . com ) were able to detect AF , SA , and hypertension at accuracies of 97% , 90% , and 82% , respectively [13 , 14] . Despite these advances , the lack of comprehensive datasets that integrate wearable metrics with other data types means that the utility of consumer-grade wearables to basic , translational , and clinical research , as well as personalized health , remains largely uncharacterized . In this study , our goal was to investigate the utility of consumer-grade wearables in cardiovascular and lipidomics research . To that end , we generated multidimensional data from 233 normal individuals recruited for a longitudinal study ( SingHEART/BioBank; National Heart Centre Singapore [NHCS] , https://www . nhcs . com . sg ) . Subjects were tracked using a consumer-grade wearable activity and HR tracker ( Fitbit Charge HR; Fitbit , www . fitbit . com ) , in addition to comprehensive profiling through lifestyle questionnaires , clinical measurements ( e . g . , weight , height , waist circumference [WC] , blood pressure , etc . ) , lipid panel values , blood glucose test , cardiac magnetic resonance imaging ( CMR ) , and lipidomic profiling . We then performed integrative analysis of the dataset in order to answer three specific questions . First , can wearable metrics obtained from study subjects provide insights into their behavioral and demographic characteristics ? Second , how well do wearable metrics ( both step- and HR-based ) correlate with CVMD risk markers such that they are useful in the areas of clinical and/or translational research and personalized health monitoring ? Finally , can wearable-derived metrics be used to support basic research , particularly in the analysis of cardiac imaging and lipidomic profiling data ?
The cohort of 233 volunteers was tracked for a median duration of 4 days ( range 2–6 days ) per subject . Summary statistics of this cohort are shown in Table 1 ( full details in S1 Data ) . The cohort had a median age of 48 years ( range 21–69 years ) and displayed a female bias ( 137/233 , 58 . 8% ) . The average daily steps ( “DailySteps” ) median was 10 , 395 steps per day , which is consistent with a recent study that compared Fitbit Flex ( Fitbit , www . fitbit . com ) and ActiGraph ( http://actigraphcorp . com/ ) measurements in 104 Singapore-resident individuals ( median steps/day = 10 , 193 ) [15] . There was no significant difference in DailySteps between male and female subjects ( Student t test , p = 0 . 604 ) . In terms of wearable-derived HR metrics , median average day and night HR were 75 bpm and 61 bpm , respectively , with a median RHR of 69 bpm . We compared wearable-derived RHR ( denoted henceforth as “RestingHR” ) with in-clinic measurements obtained from two sources , namely RHR measured by an automatic blood pressure monitor ( ABPM_HR ) and during an electrocardiogram test ( ECG_HR ) . We found that RestingHR correlated better with ECG_HR , which is a generally accepted benchmark ( rs = 0 . 690; p = 2 . 506 × 10−33 ) , as compared with ABPM_HR ( r = 0 . 541; p = 2 . 434 × 10−18 ) . Bland-Altman analysis [16] showed that RestingHR and ABPM_HR were on average higher than ECG_HR ( mean difference = 5 bpm and 9 bpm , respectively ) , with RestingHR having better agreement with ECG_HR ( 95% limits of agreement = −9 bpm , 20 bpm ) compared with ABPM_HR ( 95% limits of agreement = −12 bpm , 30 bpm ) . This suggests that RestingHR is relatively accurate and comparable to clinical measurements . Active [10 , 17] volunteers ( DailySteps > 8 , 000 ) had lower RestingHR compared with their sedentary counterparts , even after accounting for age , gender , and body mass index ( BMI; β = −3 . 185; p = 0 . 001 ) . This observation was also significant when using either continuous step counts ( 0 . 252 bpm less per 1 , 000 additional steps; p = 0 . 021 ) or when using self-reported total activity ( 1 . 723 bpm less per additional unit; p = 4 . 080 × 10−6 ) . Because volunteers comprised males and females of varying age groups ranging from 21 to 69 years , we sought to determine whether there were clusters of volunteers defined by common activity patterns . We obtained per-volunteer daily activity profiles by averaging step counts from multiple days by time of day . These daily profiles were then clustered using unsupervised k-means clustering ( k = 3 ) , with Pearson correlation as a distance measure ( Fig 1A ) . The 3 resulting clusters , of sizes 62 , 63 , and 108 , respectively , showed distinct differences in terms of peak activity periods ( Fig 1A ) . The first ( AM cluster ) showed peak activity in the morning , whereas the second ( PM cluster ) showed peak activity in the evening . A third cluster ( MidDay cluster ) showed a more even distribution of activity , peaking in midday ( Fig 1B ) . To further characterize these activity clusters , we compared the distribution of subject age and gender among the 3 clusters . There was no significant difference in gender composition among the clusters , and the average ages of subjects in the clusters were 50 ( AM ) , 43 ( PM ) , and 46 ( MidDay ) years ( Fig 1C ) . Mean age among the clusters was significantly different ( one-way ANOVA , p = 0 . 002 ) , with the AM cluster having older subjects compared with the PM cluster ( Tukey’s test , p = 0 . 002 ) . We next considered sleep tracking data and characterized sleep and wake times across the clusters . On average , sleep tracking data revealed that our volunteers spent 6 hours and 57 minutes asleep each day , which is consistent with other studies [18] . Average sleep times for the AM , PM , and MidDay clusters were at hours 23:18 , 00:07 , and 23:40 , respectively , whereas average wake times were at hours 06:39 , 08:10 , and 07:43 , respectively ( Fig 1D ) . There was a significant difference in sleep times among the clusters ( one-way ANOVA , p = 1 . 3 × 10−4 ) , with the AM cluster going to bed earlier than the PM cluster ( Tukey’s test , p = 7 . 18 × 10−5 ) and the PM cluster having a later sleep time compared with the MidDay cluster ( Tukey’s test , p = 0 . 019 ) . Similarly , wake times were different between clusters ( one-way ANOVA , p = 8 . 98 × 10−6 ) , with the AM cluster waking up earlier compared with both the PM cluster ( Tukey’s test p = 1 . 47 × 10−5 ) and the MidDay cluster ( Tukey’s test p = 3 . 80 × 10−4 ) . The significant disparities in age and sleep timing between the AM and the PM clusters reflect lifestyle differences that could , in part , be explained by previous reports of an advance in circadian timing with aging [19] . One key aim of this study is to characterize the relationship between wearable metrics and clinical parameters of relevance to CVMD risk . To that end , volunteers had various clinical parameters measured in the clinic upon recruitment , including BMI , WC , systolic blood pressure ( SBP ) , diastolic blood pressure ( DBP ) , as well as fasting levels of total cholesterol ( TotalChol ) , high-density lipoprotein ( HDL ) , low-density lipoprotein ( LDL ) , triglycerides ( TGs ) and fasting blood glucose ( FBG ) . Wearable metrics of interest that were evaluated against these clinical parameters comprised DailySteps and RestingHR . For comparison , we also evaluated clinically measured RHR values ( ECG_HR ) , as well as questionnaire-derived activity scores ( General Practice Physical Activity Questionnaire [GPPAQ] ) . After categorizing our volunteers according to commonly used clinical thresholds ( see Materials and methods ) , we used logistic regression to determine the extent to which wearable metrics are associated with clinical risk markers . For models using DailySteps , adjustments were made for age , BMI , and gender as well as interaction between gender and steps in order to account for known gender-specific differences in metabolism and response to chronic exercise [20 , 21] . For models using RestingHR , age and gender were included as independent covariates . We found that RestingHR is a better predictor compared with DailySteps ( Fig 2 , S1 Table ) . Whereas RestingHR was significantly associated with 7/9 clinical markers , DailySteps was only significantly associated with lower odds of having high BMI , WC , and TG values in more active males . For instance , male subjects benefitted more from taking more steps per day in terms of reduced risk of obesity ( high BMI ) compared with their female counterparts ( odds ratio [OR] 0 . 710; pinteraction = 0 . 002; Fig 2 ) . In contrast , questionnaire-based activity score ( GPPAQ ) was not significantly associated with any clinical parameters ( S1 Table ) . Next , we compared RestingHR to gold standard clinical RHR ( ECG_HR ) . For clinical markers that were significantly associated with either measure , RestingHR achieved more significant p-values than ECG_HR in all clinical markers except for SBP and DBP , probably due to ECG_HR , SBP , and DBP all being measured during the same clinic visit ( S1 Table ) . This suggests that RestingHR is comparable to gold standard ECG_HR in associating with CVMD risk makers . A portion of the volunteers underwent CMR imaging . We therefore sought to determine whether wearable-derived metrics could be used in the analysis of cardiac imaging data . In particular , we were interested in whether wearable-derived physical activity was correlated with cardiac remodeling because recent work done using activity questionnaires had indicated that exercise-induced cardiac remodeling ( EICR; also known as athlete’s heart ) is not exclusive to athletes but can also occur in moderately active individuals [22] . We considered cardiac parameters associated with EICR , namely left ventricular mass ( LVM; n = 202 ) , left ventricular end-diastolic volume ( LVEDV; n = 216 ) , right ventricular end-diastolic volume ( RVEDV; n = 126 ) , and aortic forward flow ( AoF; n = 202 ) . The relationship between wearable-derived activity and cardiac parameters was assessed using multiple linear regression adjusting for age , gender , and SBP ( covariates and indexing methods used are described in the Materials and methods section ) . Because cardiac remodeling is more pronounced at the more extreme end of physical activity levels , we also considered DailySteps binned with cutoffs at the 10th , 50th , and 90th percentiles to produce 4 categories ( Categories I–IV ) . We found that both continuous and categorical step counts were significant predictors for all 4 cardiac parameters ( Fig 3 , S2 Table ) . After adjusting for covariates , DailySteps ( x1 , 000 ) was a significant predictor for LVM ( β = 0 . 353; p = 0 . 012 ) , LVEDV ( β = 0 . 386; p = 0 . 010 ) , RVEDV ( β = 0 . 617; p = 0 . 003 ) , and AoF ( β = 0 . 466; p = 0 . 003 ) . We next determined the risk of having abnormally high LVM ( indexed to body surface area [BSA] ) among our very active volunteers because this is the characteristic most associated with EICR . Those in the upper quartile of DailySteps were more likely to exceed upper population-matched reference limits [23] compared with other volunteers ( OR 3 . 239; CI 1 . 133–9 . 276; p = 0 . 026 ) . We also considered RestingHR as a predictor for these cardiac parameters and found it only significant for LVEDV , RVEDV , and AoF but not for LVM ( S2 Table ) . In summary , the integration of wearable activity metrics and cardiac imaging data can reveal relationships between exercise and cardiac remodeling in normal individuals . A subset of volunteers ( n = 112 ) underwent serum sphingolipid profiling to determine the abundance of various species of circulating sphingolipids , namely ceramides , sphingomyelins , lactosylceramides , and glucosylceramides . Because circulating sphingolipids , especially ceramides , have been shown to be correlated with cardiorespiratory fitness and exercise [24–26] , we sought to determine whether wearable-derived physical activity can contribute to the discovery of relationships between activity and sphingolipid abundance . Using multiple regression to account for age , gender , and BMI , we identified 12 sphingolipids ( Table 2 , Fig 4 ) that were significantly associated with DailySteps ( p < 0 . 05 ) , 8 of which had a false discovery rate ( FDR ) –adjusted p-value less than 0 . 1 . All significant sphingolipids were negatively associated with DailySteps . Of these , the specific sphingolipids most significantly ( p < 0 . 01; q < 0 . 1 ) associated with DailySteps included Cer ( d18:1/18:0 ) , Cer ( d18:1/20:0 ) , and Cer ( d18:1/24:1 ( 15Z ) ) , ceramides previously reported to be negatively correlated with cardiorespiratory fitness as measured by peak oxygen consumption in volunteers [25] . We also identified associations among precursor dihydroceramides Cer ( d18:0/20:0 ) and Cer ( d18:0/24:1 ( 15Z ) ) , which have been linked to obesity [24 , 27] . Together , this suggests that wearable-derived physical activity is capable of identifying relationships between lifestyle and serum ceramide abundance . Apart from ceramides , we identified novel associations among several sphingomyelins ( SM ( 36:0 ) , SM ( 36:1 ) , SM ( 36:2 ) ) and a glucosylceramide ( GlcCer ( d18:1/16:0 ) ) , all of which were lower in more active subjects . We then compared the abundance of these activity-associated sphingolipids with FBG levels and found that 5 of them were also positively associated with FBG ( p < 0 . 05 ) . These included ceramides ( Cer ( d18:1/18:0 ) , Cer ( d18:1/20:0 ) ) and sphingomyelins ( SM ( 36:0 ) , SM ( 36:1 ) , SM ( 36:2 ) ) ( Table 2 ) . In line with our findings , levels of ceramides ( Cer ( d18:1/18:0 ) , Cer ( d18:1/20:0 ) ) —as well as sphingomyelin SM ( 36:1 ) in plasma and skeletal muscle—have been reported to be correlated with insulin resistance [26 , 28–31] . This analysis was also repeated using RestingHR as a metric; however , no significant associations were identified after accounting for multiple testing . These results suggest that activity metrics from consumer-grade wearables are sufficiently accurate to yield biologically relevant insights from lipidomics datasets .
The data we presented above show that even short-duration wearable studies can provide value to biomedical science , particularly in cardiovascular and lipidomics research . Our analysis of time series activity data shows that wearable metrics can stratify a cohort into behavioral clusters with distinct characteristics . This approach can assist researchers seeking to correlate lifestyle physical activity with health outcomes . Our characterization of the relationships between wearable metrics and CVMD markers can also be used to inform future wearable studies . Specifically , we found that RHR metrics are superior to step-based ones in terms of association with CVMD markers . However , this was not the case for the cardiac imaging and lipidomics data . This is likely attributable to RHR being a strong determinant of CVMD risk [5] . Conversely , unlike RHR , the relationship between activity and CVMD risk is less straightforward because it is subject to many confounders not captured in this study , of which diet is likely to be a key factor . Such relationships may only become apparent with larger sample sizes and diverse cohorts [32] . Despite rising adoption of wearables for personal activity and fitness tracking , the translation of wearable metrics into actionable health insights remains a challenge . Our results show that , with respect to CVMD risk , RHR values from wearables are equivalent to—if not better than—clinical RHR measurements . With wearables , users can continuously monitor RHR , enabling early detection of deviations in RHR that herald changes to CVMD risk markers such as weight gain and hypertension [33] . This may improve upon the status quo , whereby individuals are typically only made aware of changes to health status and disease risk during infrequent health checkups . Notably , the correlation between wearable metrics and fasting glucose suggests that wearables may provide an early indication of increased diabetes risk , which is vital given that a large number of Singaporeans have undiagnosed diabetes [34] . From a healthcare research perspective , continuous RHR tracking may become essential in long-term studies tracking lifestyle and health parameters because it could provide a high-resolution view of how changes in lifestyle impact cardiovascular health . Such insights , potentially emergent from large-scale efforts such as Project Baseline ( http://www . projectbaseline . com ) , will also help formulate ways to present wearable data to users in ways that promote compliance with healthy lifestyle behaviors and facilitate early detection of disease states . Our findings on the utility of wearables in analyzing CMR data and identifying individuals at risk of having abnormally high LVM have clinical and research implications . EICR is a benign adaptation to increased cardiovascular load resulting from exercise [35] , although there are conflicting reports of malignant adaptation to excessive exercise [36 , 37] . Initially thought to be exclusive to competitive athletes , a recent population study found that 14% of individuals in the most active category met imaging criteria for left ventricular hypertrophy [22] . Overlapping features between EICR and hypertrophic cardiomyopathy pose a risk of overdiagnosis by clinicians faced with abnormal findings in nonathletic patients , particularly at a time when regular exercise is heavily promoted to the public [35 , 38] . Apart from increased awareness of this phenomenon by clinicians , wearable metrics can play a role in differentiating between pathologic and physiological remodeling . CMR is the preferred imaging modality for population-scale health studies ( e . g . , UK Biobank [n = 100 , 000] [39] , German National Cohort [n = 30 , 000] [40] , Canadian Partnership for Tomorrow [n = 10 , 000] [41] ) because it avoids exposure to ionizing radiation or contrast agents [39] . Our findings hint at a wider role for wearables in population studies aiming to dissect the relationship between lifestyle and cardiac function . The rising number of normal individuals undergoing CMR as part of population-scale studies implies that a non-negligible fraction will be flagged with abnormal CMR findings . Where follow-up of incidental findings is consented and authorized , wearables could thus play a role in reducing the risk of overdiagnosis . Ceramides are sphingolipids involved in cellular stress response and are linked to pathological conditions such as insulin resistance , obesity , and cardiovascular disease [42] . Furthermore , higher levels of cardiorespiratory fitness are associated with lower plasma ceramide abundance , suggesting that physical activity may play a beneficial role in regulating levels of these molecules . Using wearable-derived activity data , we identified specific sphingolipids associated not only with activity but also with insulin resistance . Among the top activity-associated sphingolipids , 3 ceramides—Cer ( d18:1/16:0 ) , Cer ( d18:1/18:0 ) , and Cer ( d18:1/24:1 ( 15Z ) ) —have been previously shown to predict risk of major adverse events due to cardiovascular disease [43 , 44] and have been included in clinical laboratory tests [45] , showing that our approach can identify lifestyle-modifiable sphingolipids linked to health outcomes . Whereas previous studies investigating relationships between cardiorespiratory fitness and circulating ceramides have been interventional [25 , 46] ( i . e . , using exercise training programs or graded treadmill test ) , we found that analysis of baseline wearable-derived activity levels can also provide insights into such relationships . Future studies , including population cohorts , could be conducted using commodity wearable devices , ultimately enabling more data to be collected while reducing study complexity . One limitation of this study is the relatively short duration of the tracking periods , thus compromising power to detect associations between activity and CVMD markers . Our cardiac imaging and lipidomics analyses suggest that longer tracking periods , particularly feasible when volunteers share data from their personal devices , will prove to be even more useful . Additionally , volunteers recruited into this cohort may be enriched for those with a higher level of regard for their health and well-being . There was limited examination of time series ( as opposed to summary ) –wearable data . Activity cluster membership ( with the AM cluster as reference ) was tested as a predictor for all the association analyses in this study; however , apart from LVEDV ( pPM_cluster = 0 . 071; pMidDay_cluster = 0 . 014 ) and AoF ( pPM_cluster = 0 . 467; pMidDay_cluster = 0 . 025 ) in the cardiac imaging data , there were no other significant associations . Further studies on the utility of features derived from the time series data ( e . g . , HR variability , HR recovery , activity intensity ) are therefore warranted . In summary , we have characterized in a sizeable cohort the relationship between wearable metrics and a wide range of volunteer phenotypes including lifestyle patterns , demographics , CVMD clinical markers , cardiac imaging , and serum sphingolipid profiles . Our findings show that apart from fitness tracking , consumer-grade wearables can play a role in both basic and clinical research . Such wearables could also provide a low-cost means for early detection of changes in an individual’s personal CVMD risk profile , potentially resulting in more timely detection and intervention of CVMDs .
The SingHEART/Biobank study was established at the NHCS to characterize normal reference values for various cardiovascular and metabolic disease-related markers in Singaporeans . Normal volunteers were enrolled into this study using a protocol and written informed consent form approved by the SingHealth Centralized Institutional Review Board ( ref: 2015/2601 ) . The volunteers underwent comprehensive profiling in the following areas: ( 1 ) activity tracking using the Fitbit Charge HR wearable sensor , ( 2 ) physical activity and lifestyle questionnaire , ( 3 ) CMR imaging , ( 4 ) serum sphingolipid profiling , ( 5 ) fasting lipid and glucose panel , and ( 6 ) assessment of clinical parameters ( e . g . , HR , blood pressure , BMI ) . A total of 233 volunteers were included in this study after evaluation for completeness of activity tracking data ( details below ) and removal of subjects with extreme outlier activity metrics ( potentially due to improper wearable usage ) . Inclusion criteria were as follows: Volunteers were issued a Fitbit Charge HR wearable activity tracker to be worn over a course of 5 days ( e . g . , typically Monday–Friday ) . However , because the first and last days of the study tended to be partial days , the average yield for each study was 3 days of complete tracking ( defined as ≥20 hours with steps and HR data ) . Data for each subject were downloaded from the Fitbit website using the “fitbitScraper” package ( https://github . com/corynissen/fitbitScraper; NB: This method of data access is now deprecated; the same data can be obtained through the Fitbit API at https://dev . fitbit . com/reference/web-api/quickstart ) . Step counts were available at 2 levels: intraday step counts in 15-minute intervals and daily totals . Intraday HR data were available at 5-minute intervals , along with confidence levels . Intraday sleep tracking data containing details of each sleep session were also retrieved . To determine data completeness , we used HR confidence values as an indicator that the subject is wearing the device . HR data points with confidence value of “−1” were considered to be invalid ( i . e . , device was not worn or was incorrectly worn ) . First , the HR values table was merged with the steps table by their time points . We then counted the number of hours per day that contained HR data with a valid confidence value . Days with ≥20 valid hours were considered to be complete . Furthermore , days with no intraday step data were excluded . Such events typically arose due to delayed syncing of the device resulting in older data being overwritten . To determine RestingHR , we calculated the average HR value for time points that met the following criteria: ( 1 ) had ≤100 steps take place within the 15-minute interval and ( 2 ) had a valid HR value . Day HR was similarly obtained but by restricting to time points between 2 PM and 4 PM , whereas night HR sampled time points between 2 AM and 4 AM . For DailySteps estimation , we obtained the average sum of steps that took place in data-complete days . We also derived estimated daily steps using an alternative method . Briefly , we obtained step-count data points across data-complete days that were matched with valid HR values . We then calculated estimated daily steps by multiplying the average of these step-count values by 96 ( i . e . , the number of 15-minute intervals per day ) . Daily step values derived through both methods were highly correlated ( rs = 0 . 955; p < 2 . 2 × 10−16 ) . We thus used DailySteps as a measure of wearable-derived physical activity for the rest of this study . Sleep tracking data was processed as follows: the amount of sleep for each day was determined by summing the duration of all sleep sessions for that day . An average sleep duration was then obtained from the data-complete days . Sleep hour was determined by calculating the average start time of sleep sessions occurring between 7 PM and 4 AM . Wake hour was determined by averaging the end time of sleep sessions . Only subjects with average sleep duration , sleep hour , and wake hour that were within 2 standard deviations ( SDs ) from mean values were included for statistical analysis in the study ( 216/233 subjects ) . For each volunteer , we obtained a 24-hour daily activity profile in the following manner . First , step-counts were smoothed across sliding windows of 5 time points . Then , step-counts from multiple data-complete days were collapsed into a single average profile by averaging step-counts from the same daily time point . Time points with no valid step counts were filled with zeros , and the daily profile was again smoothed in 5–time point sliding windows . Volunteers were clustered using unsupervised k-means clustering ( k = 3 ) with Pearson correlation as a distance measure . As part of the study , volunteers answered several questionnaires to ascertain their lifestyle . Physical activity was primarily assessed using the GPPAQ ( https://www . gov . uk/government/publications/general-practice-physical-activity-questionnaire-gppaq ) . Briefly , subjects provide information on amount spent in the last week on the following activities: ( 1 ) physical exercise , ( 2 ) cycling , ( 3 ) walking , ( 4 ) housework , and ( 5 ) gardening and/or do-it-yourself activity . Additionally , volunteers were asked of the amount of physical activity involved in their occupations , as well as their walking pace . A physical activity index ( PAI ) was generated based on the amount of physical exercise and/or cycling as well as their occupational activity level . The PAI has four activity levels: Inactive , Moderately Inactive , Moderately Active , and Active . These are treated as numerical scores in this study . In this study , amount of cycling was not factored into PAI derivation in order to facilitate a more direct comparison with wearable-derived activity . The following measurements were performed on the day of volunteer recruitment . First , weight and height were measured using a SECA703 weighing scale ( Seca ) , whereas SBP and DBP were obtained using an Intellivue MX450 patient monitor ( Philips ) . Volunteers fasted for 8 to 10 hours prior to the recruitment appointment , and blood was drawn for the following tests: ( 1 ) lipid and glucose panel and ( 2 ) serum sphingolipid profiling . Additionally , the volunteers also underwent a Pagewriter TC30 16-lead ECG test ( Philips ) . HRs were obtained from two separate sources , the blood pressure monitor ( “ABPM_HR” ) and from the ECG reading ( “ECG_HR” ) . The following clinical markers were considered in the study: BMI , WC , SBP , and DBP as well as fasting levels of TotalChol , HDL , LDL , TG , and FBG . Thresholds used to define risk levels ( S3 Table ) are as follows: High BMI ( >27 . 5 ) , High WC ( >100 cm for males , >90 cm for females ) , High SBP ( >140 mmHg ) , High DBP ( >90 mmHg ) , High TotalChol ( >6 . 2 mmol/l ) , Low HDL ( <1 mmol/l ) , High LDL ( >4 . 1 mmol/l ) , High TG ( >2 . 3 mmol/l ) , and High FBG ( >6 mmol/l ) . Multiple linear regression and logistic regression analyses described in this study were conducted using the GLM ( generalized linear model ) function in R . For multiple linear regression , a Gaussian error distribution was used , whereas a binomial one was used for logistic regression . When gender was considered as a covariate , the female gender was set as the reference level . For logistic regression analysis between wearable metrics and clinical parameters , two models were used depending on the metric . For step-based metrics , the model is Clinical_Marker ~ Age + Gender + Metric + Gender × Metric , whereas for RHR-based metrics , the model is Clinical_Marker ~ Age + Gender + Metric . ORs and p-values reported for step-based metrics are for the Gender × Metric interaction term . To ensure that results from various metrics are comparable , association analyses were conducted on a subset of subjects ( 223/233 ) with valid measurements for all metric types ( i . e . , DailySteps , RestingHR , ABPM_HR , ECG_HR , GPPAQ score ) . For logistic regression between activity clusters ( AM cluster as reference level ) and clinical markers , the following model was used: Clinical_Marker ~ Age + Gender + DailySteps + Gender × DailySteps + ActivityCluster . CMR of the volunteers was performed using either a Magnetom Aera 1 . 5T ( Siemens ) or Ingenia 3T ( Phillips ) scanner under previously described settings [23] . Parameters such as cardiac volumes and mass were analyzed from imaging data using the CMR42 software ( Circle Cardiovascular Imaging ) and standardized protocols [23] . Four cardiac parameters were considered in this study: LVM , LVEDV , RVEDV , and AoF , with the first three being indexed to BSA according to the Dubois formula [47] . When performing multiple linear regression , adjustment was made for age , gender , and SBP . In the case of AoF , additional adjustment was made for weight and height . Only cardiac parameters with values within 2 SDs from their mean value were included for analysis . Numbers of data points analyzed for each cardiac parameter are as follows: LVM ( n = 202 ) , LVEDV ( n = 216 ) , RVEDV ( n = 126 ) , and AoF ( n = 203 ) . For logistic regression analysis of abnormally high BSA-indexed LVM against volunteer physical activity , only those of Chinese ethnicity were considered ( n = 192 ) . Cutoffs for defining abnormal BSA-indexed LVM were obtained from a study of 180 healthy Singaporeans ( 70 g/m2 for males , 50 g/m2 for females ) [23] . Lipid internal standard mix ( Ceramide/Sphingoid Internal Standard Mixture I , Avanti Polar Lipids ) of 500 pmol was added to 100 μl of serum in a microcentrifuge tube . After an equilibration period of 30 s , 1 . 2 ml of HPLC-grade methanol was added to the mixture , followed by vortexing . The mixture was then incubated at 50°C for 10 min , followed by centrifugation to pellet the precipitated protein . The supernatant was then removed and placed in a clean microcentrifuge tube for drying under nitrogen gas . To reconstitute the dried extract , 100 μl of methanol was then used . The reconstituted lipid solution was then separated using a liquid chromatography–mass spectrometry ( LC-MS ) system ( Agilent 1260 ) and a Thermo Scientific Accucore HILIC column ( 100 × 2 . 1 mm; particle size 2 . 6 μm ) . Mobile phase A consisted of acetonitrile/water ( 95:5 ) with 10 mM ammonium acetate , pH 8 . 0 , and mobile phase B consisted of acetonitrile/water ( 50:50 ) with 10 mM ammonium acetate , pH 8 . 0 . For the separation , the column was equilibrated with 100% mobile phase A , increasing to 20% mobile phase B in 5 min , then held for 5 min . The column was then re-equilibrated with 100% mobile phase A for 5 min . Finally , mass spectrometry ( MS ) and data acquisition were performed using an Agilent 6430 triple-quadrupole mass spectrometer . As data were generated in two batches , normalization was performed within each batch on raw values ( measured in pmols ) by performing z-score transformation on a per-sphingolipid basis , prior to combining the data . Odd-chain sphingolipids and sphingolipids with missing values in more than 20% of samples were excluded from analysis . All statistical analyses in this study were performed using the R statistical environment . Unless otherwise stated , correlations described in this study are Spearman correlation coefficients . Adjustment for multiple testing in the lipidomics analysis was done using the Benjamini-Hochberg FDR method [48] . | Little is known about how data from wearable sensors can be used apart from fitness tracking . We comprehensively studied 233 normal volunteers , integrating data from wearable sensors with lifestyle questionnaires , cardiac imaging , sphingolipid profiling , and clinical measurements of various heart and metabolic disease markers . Apart from showing that wearable sensors can be used to identify groups of volunteers with distinct behavioral and demographic characteristics , we showed that resting heart rate ( RHR ) from wearables performed better than step counts in predicting heart and metabolic disease risk markers . Notably , we further demonstrated that wearable data could be used in 2 areas of biomedical research . In the field of cardiac imaging , we showed that activity data from wearables can be used to determine how the size of heart is influenced by physical activity . Wearable data could also identify active individuals that are more likely than others to have enlarged hearts and potentially be misdiagnosed with heart disease . In the field of lipidomics , we showed that wearable data can be used to identify species of sphingolipids that are affected by how active a person is . Some of these compounds are known to be associated with obesity , diabetes , and heart disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"sphingolipids",
"sports",
"and",
"exercise",
"medicine",
"physical",
"activity",
"research",
"design",
"questionnaires",
"research",
"and",
"analysis",
"methods",
"cardiology",
"public",
"and",
"occupational",
"health",
"exercis... | 2018 | Beyond fitness tracking: The use of consumer-grade wearable data from normal volunteers in cardiovascular and lipidomics research |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.