UBC Faculty Research and Publications

Comparative genomics of parasitic silkworm microsporidia reveal an association between genome expansion… Pan, Guoqing; Xu, Jinshan; Li, Tian; Xia, Qingyou; Liu, Shao-Lun; Zhang, Guojie; Li, Songgang; Li, Chunfeng; Liu, Handeng; Yang, Liu; Liu, Tie; Zhang, Xi; Wu, Zhengli; Fan, Wei; Dang, Xiaoqun; Xiang, Heng; Tao, Meilin; Li, Yanhong; Hu, Junhua; Li, Zhi; Lin, Lipeng; Luo, Jie; Geng, Lina; Wang, LinLing; Long, Mengxian; Wan, Yongji; He, Ningjia; Zhang, Ze; Lu, Cheng; Keeling, Patrick J; Wang, Jun; Xiang, Zhonghuai; Zhou, Zeyang Mar 16, 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12864_2012_Article_4862.pdf [ 1.21MB ]
JSON: 52383-1.0215898.json
JSON-LD: 52383-1.0215898-ld.json
RDF/XML (Pretty): 52383-1.0215898-rdf.xml
RDF/JSON: 52383-1.0215898-rdf.json
Turtle: 52383-1.0215898-turtle.txt
N-Triples: 52383-1.0215898-rdf-ntriples.txt
Original Record: 52383-1.0215898-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessComparative genomics of parasitic silkwormmicrosporidia reveal an association betweengenome expansion and host adaptationGuoqing Pan1†, Jinshan Xu2,3†, Tian Li1†, Qingyou Xia1, Shao-Lun Liu3, Guojie Zhang4, Songgang Li4, Chunfeng Li1,Handeng Liu1, Liu Yang1, Tie Liu1, Xi Zhang2, Zhengli Wu1, Wei Fan4, Xiaoqun Dang1, Heng Xiang1, Meilin Tao1,Yanhong Li1, Junhua Hu1, Zhi Li1,2, Lipeng Lin1, Jie Luo1, Lina Geng1, LinLing Wang2, Mengxian Long1,Yongji Wan1, Ningjia He1, Ze Zhang1, Cheng Lu1, Patrick J Keeling3, Jun Wang4, Zhonghuai Xiang1and Zeyang Zhou1,2*AbstractBackground: Microsporidian Nosema bombycis has received much attention because the pébrine disease ofdomesticated silkworms results in great economic losses in the silkworm industry. So far, no effective treatmentcould be found for pébrine. Compared to other known Nosema parasites, N. bombycis can unusually parasitize abroad range of hosts. To gain some insights into the underlying genetic mechanism of pathological ability andhost range expansion in this parasite, a comparative genomic approach is conducted. The genome of two Nosemaparasites, N. bombycis and N. antheraeae (an obligatory parasite to undomesticated silkworms Antheraea pernyi), weresequenced and compared with their distantly related species, N. ceranae (an obligatory parasite to honey bees).Results: Our comparative genomics analysis show that the N. bombycis genome has greatly expanded due to thefollowing three molecular mechanisms: 1) the proliferation of host-derived transposable elements, 2) the acquisition ofmany horizontally transferred genes from bacteria, and 3) the production of abundnant gene duplications. To ourknowledge, duplicated genes derived not only from small-scale events (e.g., tandem duplications) but also from large-scale events (e.g., segmental duplications) have never been seen so abundant in any reported microsporidia genomes.Our relative dating analysis further indicated that these duplication events have arisen recently over very shortevolutionary time. Furthermore, several duplicated genes involving in the cytotoxic metabolic pathway were found toundergo positive selection, suggestive of the role of duplicated genes on the adaptive evolution of pathogenic ability.Conclusions: Genome expansion is rarely considered as the evolutionary outcome acting on those highly reduced andcompact parasitic microsporidian genomes. This study, for the first time, demonstrates that the parasitic genomes canexpand, instead of shrink, through several common molecular mechanisms such as gene duplication, horizontal genetransfer, and transposable element expansion. We also showed that the duplicated genes can serve as raw materials forevolutionary innovations possibly contributing to the increase of pathologenic ability. Based on our research, wepropose that duplicated genes of N. bombycis should be treated as primary targets for treatment designs againstpébrine.Keywords: Gene duplication, Horizontal gene transfer, Host-derived transposable element, Host adaptation,Microsporidian, Silkworms* Correspondence: zyzhou@swu.edu.cn†Equal contributors1State Key Laboratory of Silkworm Genome Biology, Southwest University,Chongqing 400715, China2College of Life Sciences, Chongqing Normal University, Chongqing 400047,ChinaFull list of author information is available at the end of the article© 2013 Pan et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Pan et al. BMC Genomics 2013, 14:186http://www.biomedcentral.com/1471-2164/14/186BackgroundMicrosporidia are obligate intracellular parasitic fungi thatcan infect a wide variety of organisms including vertebrateand invertebrate (particularly insects). Some species leadto severe syndromes in immunocompetent hosts andcause opportunistic infections in Acquired Immunodefi-ciency Syndrome (AIDS) patients [1,2]. More than 1200microsporidia species that belong to 150 genera have beenreported thus far [3]. Among them, the genus Nosemais the most diverse one. The domesticated silkworm,Bombyx mori, has long been considered as the primarysource for the silk production worldwide. A highly mortaldisease referred to as pébrine is currently the major threatto the silk production. Pébrine is caused by the infectionof the microsporidian parasite, Nosema bombycis. Thisdisease was first recognized during the destruction of theEuropean silk industry in 1857 [4]. N. bombycis infectssilkworms through vertical transmission from the motherhost to their progenitive eggs, and chronically damagesthe entire body of the worm (including intestines, silkglands, muscles, and Malpighian tubules). After infections,the silkworm larvae are inactive and slow in development.Later, black spots, a disease symptom called pébrine [5],will appear throughout their bodies and lead to death.Since no effective treatment methods have been developedup to this point, the infections by N. bombycis inevitablycause devastating economic losses in the silkworm indus-try. Apart from the domesticated silkworms, N. bombyciscan also infect various lepidopteran insects [6-8], indica-tive of their broad hosts range.So far, the underlying genetic mechanisms of the highlyinfectious ability and the broad host range of N. bombycisremain unknown. To this end, we conducted a compa-rative genomic approach, from which we might learn agreat deal about the genetic basis as to why and howN. bombycis can be so infectious across various hosts.In this study, we sequenced the genome of two micros-poridian parasites: N. bombycis and N. antheraeae (an ob-ligatory parasite to undomesticated silkworms Antheraeapernyi [9]). By comparing their genomes with a publisheddistantly related Nosema genome, N. ceranae [10] (servingas outgroup), we show that the N. bombycis genomesurprisingly expands due to the production of duplica-ted genes, the proliferation of host-derived transposableelements, and the acquisitions of many horizontally trans-ferred genes from bacteria. Some duplicated genes associ-ated with the cytotoxic pathway have experienced positiveselection, implying that this adaptive evolution might en-hance the infectious ability of N. bombycis, as well as theexpansion of its host range. Considering that all reportedmicrosporidian genomes are highly reduced and compact[11,12], our data, for the first time, reveal a usual genomeevolution process showing that the genome of parasitescould expand. Those expanded genetic gears might haveinfluenced the infectivity and the survivorship of parasitesas we report herein.ResultsGenomic architecture of N. bombycis and N. antheraeaeBy using various sequencing platforms, 6.7X, 10X, and28X physical coverage of whole genome sequence ofN. bombycis were obtained from the Sanger sequencingmethod (plasmids with 2Kb inserts), the miniBAC endsequencing method, and the Illumina short-read sequen-cing method respectively (Additional file 1). Our sequen-cing efforts resulted in 1,605 scaffolds built from 3,551contigs. The total assembly genome size is 15.7 Mb (theN50 of the scaffolds = 57.4 Kb and the maximum scaf-fold size = 571.1 Kb) (Additional file 2). A total of 4,458protein coding sequences were identified (Table 1). Theassembled genomic size (~15.7 Mb) is close toprevious estimation by pulse-field gel electrophoresis(~15.3 Mb) [13], indicating that the coverage of the as-sembled genome is nearly complete. To sequence thegenome of N. antheraeae, a total of 657 Mb of Illuminareads were obtained after filtering ambiguous reads. Ouranalysis results in 6,215 scaffolds and a total of 6.6 Mbof unique sequence. Totally, 3,413 protein coding se-quences were identified (Table 1). The assembled gen-ome size of N. antheraeae was estimated using thefollowing equation: genome size = number of 15-mersper kilo-bases/depth of 15-mers per kilo-bases. The assem-bled genome size of N. antheraeae (~7.4 Mb) (Additionalfile 3) is close to our previous estimation by pulse-field gelelectrophoresis [14], indicating that approximately 90% ofthe N.antheraeae genome was captured. A comparison ofgenome features among the generated Nosema genomesand published microsporidian genomes (one Nosema spe-cies and two Encephalitozoon species) is listed in Table 1.To gain some insights into the variations of gene contentamong N. antheraeae, N. bombycis, and N. ceranae, thenumber of orthologous genes were compared. Most genesare shared among all three species, but 8% of genes wereN. antheraeae-specific, 15.7% were N. bombycis-specific,and 30.5% were N. ceranae-specific (Figure 1). Gene ontol-ogy analysis revealed that no distinct differenceswere found among different gene functional categories(Additional file 4). Collectively, these three Nosemaspecies lack genes for tricarboxylic acid cycle, oxidativephosphorylation, and fatty acid β-oxidation, consistent withprevious observations [15-19]. Our observations furthersupport that microsporidia do not possess tricarboxylic acidcycle and oxidative metabolism, and microsporidia parasiteshave often experienced on-going genome streamlining viathe relaxation of purifying selection.To determine if the Nosema proteins were more com-pact than other microsporidian parasites, the length ofthose silkworm Nosema proteins with assigned functionsPan et al. BMC Genomics 2013, 14:186 Page 2 of 14http://www.biomedcentral.com/1471-2164/14/186was compared to homologs of two published micro-sporidian parasitic Encephalitozoon species, E. cuniculiand E.intestinalis (Additional file 5). Our results showthat the average length of total homologous genes fromN. antheraeae and N. bombycis is shorter than thatfrom E. cuniculi and E. intestinalis, indicating thatproteins in Nosema were more compact than those inEncephalitozoon.Overall, our comparative genomics analysis showed thatN. bombycis possesses a much larger genome size thanother two Nosema species (Table 1). Considering that N.bombycis has wide host range, the genome expansionmight facilitate the host adaption in N. bombycis. Thus,for the subsequent analyses, we aim to seek for the under-lying genetic mechanisms as to why and how N. bombycisgenome expands. Furthermore, we seek for the putativegenetic components that contribute to the infectious abi-lity of N. bombycis in a hope that our analyses couldprovide some clues on the development of treatmentstrategies of pébrine.Proliferation of host-derived transposable elements inN. bombycisAfter obtaining the genomes of the two Nosema species,we seek for the potential molecular mechanisms under-lying the genome expansion of N. bombycis. Consideringthat the proliferation of transposable elements often con-tributes to the genome size variation in many eukaryotes[20], it was considered as the first molecular mechanismsfor us to check. Although the genomes of several humanpathogenic microsporidians have been shown to lacktransposable elements, transposable elements have beendetected in the genomes of other non-human pathogenicmicrosporidians [21-25]. To understand what degree thosetransposable element shape the genomic architectures inNosema, we searched for transposons in N. bombycis andN. antheraeae (for details, see Materials and Methods).Two different approaches were implemented in this study.Because most transposable elements comprise internalprotein-coding genes (e.g., transposase or reverse trans-criptase) that are necessary for their transposition, we firstidentified those putative transposable elements bysearching for their internal protein-coding sequences. Inmany cases, the internal protein-coding sequences arehighly generated but recognizable. Second, for those thatdo not possess readily identifiable internal protein-codingsequences, other features such as inverted repeats orinsertion sites were used to recognize the transposableelements.Overall, our results show evidence that a larger gen-ome size of N. bombycis is partly due to both the acqui-sition of new transposons and expansion of existingtransposable elements (Table 2). Among all identifiedtransposable elements, the Ty3/gypsy retrotransposons[22] constitute the largest part of known classes oftransposable elements in N. bombycis. A broad samplingfrom GenBank shows that these transposable elementsalso reside in other microsporidian groups includingSpraguea lophii, Edhazardia aedis, and Brachiolaalgerae, indicating that this transposable element familyexists back to the common ancestor of most micro-sporidian species and further expand in N. bombycis.Majority of transposable elements among Nosema ge-nomes are common across three Nosema species, whereasPiggybac transposons were only found in N. antheraeaeand N. bombycis except for N. ceranae (Additional file 6).To test whether Piggybac was lost during the evolution ofN. ceranae or was gained in the most recent common an-cestor of N. antheraeae and N. bombycis, the phylogeny ofPiggybac was reconstructed from Nosema, domesticatedsilkworms, and other insects. Our analyses show that theNosema Piggybac sequences fall into four well-supportedgroups, and three out of them are closely related toPiggybac elements from domesticated silkworms (Figure 2).Although the exact relationships of these Piggybac ele-ments between Nosema and Bombyx is complicated, ourTable 1 A comparison of genome features among three Nosema species (N. bombycis, N. antheraeae, and N. ceranae)and two Encephalitozoon species (E. cuniculi and E. bieneusi)Genomic features N. bombycis N. antheraeae N. ceranae [10] E. cuniculi [24] E. bieneusi [26]Chromosomes(bands) 18 ~12 ND 11 6Assembly(Mbp) 15.7 6.6 7.9 2.9 3.86Genomic coverage 100% ~89% 90% 86% ~64%Scaffold Num 1,605 6,215 5,465 11 1,646N50(bp) 57,394 1,883 2,902 ND 2,349Largest scaffold length(bp) 571,060 53,183 65,607 209,983 204,069G + C content (%) 31 28 27 48 26No .of CDS 4,458 3,413 2,614 1,997 3,632Mean CDS length (bp) 741 775 904 1,077 995GenBank No. NA30919 NA183977 NA32971 NA155 NA21011Pan et al. BMC Genomics 2013, 14:186 Page 3 of 14http://www.biomedcentral.com/1471-2164/14/186phylogenetic analysis suggests that Piggybac was acquiredin the most common ancestor of N. antheraeae and N.bombycis through horizontal transfer events from possiblyhost silkworms. These transfers likely independently tookplace three times, leading to the three major subgroups(labeled as HGT in Figure 2). In addition, the NosemapBac3,4,5 are closely related to the elements from otherinsects (in the middle part of tree; Figure 2), suggestingthat these Nosema Piggybac elements might originate fromthe insects. To rule out the possibility of the host contami-nations during the genome assembly, we amplified regionsflanking Piggybac elements using Nosema-specific primers,and confirmed the existence of Piggybac in Nosemagenomes (data not shown).Since the host-derived Piggybac elements are so abun-dant in N. bombycis, can those host-derived Piggybacelements serve as the vector of capturing host-derivedgenes? To answer this question, we checked the host-derived Piggybac elements that do not have any readilyidentifiable internal protein-coding sequences. Becausethey are usually hard to be identified due to the lack ofthe internal readily recognizable protein-coding genes,the terminal inverted repeat (ITR) and the insertion siteTable 2 Classification of repetitive families in N. bombycisgenomeType Subtype Length (bp) Percent (%) of genomeDNA hAT 1,011,459 6.45Merlin 470,573 3.00PiggyBac 441,876 2.82TcMar 786,733 5.02MuDR 109,681 0.70others 58,866 0.38LTR Gypsy 577,653 3.68others 33,635 0.21LINE Dong-R4 162,622 1.04others 59,703 0.38Rolling-circle Helitron 102,334 0.65SINE — 28,669 0.18Unknown — 2,204,497 14.06Total — 6,048,301 38.57Figure 1 Venn diagram showing the number of homologous genes and lineage-specific genes amongst three Nosema species,N. bombycis, N. antheraeae, and N. ceranae. The arabic numbers followed by characters represent the number of homologous genes in eachNosema species (‘a’ denotes N. antheraeae, ‘b’ denotes N. bombycis , and ‘c’ denotes N. ceranae). For instance, 2455b:1746a:1459c means that 2455genes of N. bombycis, 1746 genes of N. antheraeae, and 1459 genes of N. ceranae are homologous to each other.Pan et al. BMC Genomics 2013, 14:186 Page 4 of 14http://www.biomedcentral.com/1471-2164/14/186(TTAA) of the Piggybac elements were used as thecriteria for our search. In other words, we searched forthe N. bombycis genomic regions that are flankedby the ITR and the insertion site (TTAA) of thePiggybac elements and comprise “extrinsic” sequences.After identification of those Piggybac elements, weexamine whether the “extrinsic” sequences were re-cently transferred by the transposition of the Piggybacelements that are specific to N. bombycis via compa-ring the colinearity of these regions with those inN. antheraeae and N. ceranae. A total of 17 Piggybacelements with an internal “extrinsic” sequence wereidentified (Additional file 7). Among them, only onecase might be recently gained in N. bombycis basedon the colinearity (Additional file 8 and Additionalfile 9). When we blasted the internal “extrinsic” se-quence of the 17 Piggybac elements in GenBank usingthe “nr” database by the blastn function, no detect-able similarity with any known sequences was found.Our analysis thus far suggests that the host-derivedPiggybac elements might not be able to serve as thevector of capturing genes from hosts to N. bombycis.Figure 2 A maximum-likelihood phylogenetic tree of host-derived Piggybac transposase sequences. Arrows show the putative recenthorizontal gene transfer (HGT) events of host-derived transposable elements. Several transposons are closely related to those from insects. Blackboxes indicate elements from the two silkworm-infecting Nosema species, while white boxes indicate elements from the domesticated silkwormB. mori. Numbers in parentheses indicate the total copy numbers for each transposable element.Pan et al. BMC Genomics 2013, 14:186 Page 5 of 14http://www.biomedcentral.com/1471-2164/14/186Horizontally transferred protein-coding genes is anothersource of genetic expansion in N. bombycisSince N. bombycis has experienced the proliferation ofnative and host-derived transposons, we sought to deter-mine if horizotally transferred protein coding genes fromother organisms can also facilitate the genome expan-sion in N. bombycis, as was recently found inEncephalitozoon [26,27]. To maximize the likelihood ofdetecting horizontal protein coding gene transfer (HGT)events in N. bombycis, we implemented two different ap-proaches: a genome-wide prediction method based onorthologous sequences using the software Darkhorse,and a phylogenetic method where we screened the puta-tive HGTs in a total of 4458 gene family phylogenies.Overall, these two different approaches identify 50 and53 different HGTs in N. bombycis, respectively. Amongthem, 48 genes are shared between these two approaches(Figure 3A), resulting in a set of 55 union HGT genesbetween two different dataset. By investigating the taxo-nomic origin of these 55 unions HGT genes in a phylo-genetic framework, all of them were transferred fromprokaryotes (Figure 3B). No host-derived genes werefound in our analysis, further suggesting that only host-derived transposable elements can be transferred intothe N. bombycis genome instead of host-derived protein-coding genes. Using the clusters of orthologous groupdatabase, we found that 21 HGT candidates are unknownin functions, and 34 are predicted to fall into diverse genefunctions (Additional file 10). Among 34 HGT genes, fivegenes are involved in nucleotide metabolism and two genesare involved in sugar metabolism. Interestingly, one HGTgene that was annotated as phosphomevalonate kinase(EC2.7.4.2) is shown to be an important player in themevalonate pathway of N.bombycis (Additional file 11). Inthe mevalonate pathway, phosphomevalonate kinase is akey enzyme to catalyze the rate-limiting step for the pro-duction of isopentenyl pyrophosphate (IPP). IPP is import-ant for various molecular functions such as terpenoidsynthesis, protein prenylation, cell membrane maintenance,protein anchoring, and N-glycosylation. Overall, our obser-vations lead us to hypothesize that some of HGTs mightplay a pivotal role on the adaptation or survivorship ofN. bombycis over the course of evolution. Alternatively,many HGTs might be merely neutral without any immedi-ate adaptive consequences after their transfers. Furtherhypothesis testing will be necessary.Recent gene duplication events contribute to the genomeexpansion in N. bombycisAlthough our previous analyses showed that the prolifera-tion of host-derived transposable elements and horizon-tally transferred genes could contribute to the genomesize expansion in N. bombycis, their contributions are notsufficient to explain the much larger genome size ofN. bombycis than other two small-genome Nosema species(N. antheraeae and N. ceranae). Considering that gene du-plication is a common molecular mechanism mediatingthe expansion of genome size in many eukaryotes [28], wethen seek for the evidence if gene duplications also play arole on the genome expansion in N. bombycis. We firstperformed a syntenic analysis to identify possible segmen-tal duplication events in each Nosema species. Amongthree species, we found that N. bombycis contained 942pairs of segmental duplications throughout its genome(Figure 4A, Additional file 12). In contrast, almost nosegmental duplication could be detected in either N.antheraeae or N. ceranae. Because the assemblies of allthese genomes are fragmented, it is not possible to con-clude whether these segmental duplications are large innumber and spread throughout the genome, or arise dueto multiple whole chromosome duplication events or anancient whole genome duplication event. Nevertheless,we have identified a region where it appears that a sin-gle large-scale duplication event explain the data betterthan several independent large-scale duplication events(Figure 5). To date these duplication events, we estimatedsynonymous substitution rate (dS) for paralogous genesfrom segmental duplications in N. bombycis, and com-pared them with the dS derived from orthologs be-tween N. antheraeae and N. bombycis. The dS values arecommonly used as the proxy of age of gene duplicationbecause the synonymous substitutions evolve in a neutralfashion [28]. On average, the dS values of paralogs fromsegmental duplications in N. bombycis are generally lowerthan that of orthologs between N. antheraeae and N.bombycis (Figure 4B), suggesting that these duplicationevents took place after the separation of N. antheraeaeand N. bombycis. In addition to the detection of segmentalduplications, we identified numerous tandem duplicationevents among three Nosema species. We detected a higherrate of tandem duplications in N. bombycis compared toother two Nosema species, and in some cases multipleevents could be mapped at a single locus (Figure 4C). Onaverage, the dS values of these paralogs are also muchlower than that of orthologus genes between N. bombycisand N. antheraeae (Figure 4D), indicating that mosttandem paralogs in N. bombycis also arose relativelyrecent after the separation between N. bombycis andN. antheraeae. In short, the N. bombycis genome hasexpanded in size largely due to many large-scale andsmall-scale gene duplication events.Adaptive evolution of duplicated genes might enhancethe pathogenic ability in N. bombycisParalogs often provide raw materials for evolutionary in-novations, including the survival of parasites in theirhosts [29]. We therefore sought to identify possible in-stances of adaptive changes associated with thePan et al. BMC Genomics 2013, 14:186 Page 6 of 14http://www.biomedcentral.com/1471-2164/14/186pathogenic ability of N. bombycis among those dupli-cated genes derived from large-scale duplication eventsin N. bombycis. First, we examine if paralogs of N.bombycis contribute to the adaptive evolution moreoften than orthologs among all Nosema species. Clustersof homologous genes in N. bombycis were classified tofour different groups: 1) clusters of orthologus genes(COGs) of 1:1:1 trios of N. bombycis, N. antheraeae, andN. ceranae, 2) COGs of 1:1 gene pairs of N. bombycisand N. antheraeae, 3) COGs of 1:1 gene pairs of N.bombycis and N. ceranae, and 4) clusters of paralogousgenes (CPGs) in N. bombycis. Pairwise dN/dS ratio analysesfor these four different clusters of homologous genes werecomputed and their cumulative dN/dS ratio curve werecompared (see Materials and Methods for details). Com-pared to COGs, a higher proportion of CPGs in N.bombycis showed higher value of dN/dS ratio, suggestingthat CPGs are evolving at a faster rate than COGs at theamino acid level (Additional file 13). In most cases, this islikely due to the relaxation of purifying selection. However,we observed that a higher proportion of CPGs showed dN/dS ration greater than 1, indicative of positive selection.Overall, our observations support the view that CPGs con-tributed more to adaptive evolution than COGs in N.bombycis.To examine if any particular codons of CPGs ofN. bombycis have undergone positive selection, we ap-plied a site model approach with maximum likelihoodusing the software PAML (see Materials and Methodsfor details). The results show that 24 out of 240 CPGshave experienced positive selection in N. bombycis (Table 3),and 62% (37/60) genes in 24 CPGs have the support of ESTtags (Additional file 14). The estimated parameters andpositively selected sites for those positively selected CPGsare shown in Additional file 15. Although the majority ofpositively selected CPGs are hypothetical proteins withFigure 3 Horizontal gene transfers of protein-coding genes in N. bombycis. (A) Venn diagram showing the numbers of HGT genes betweentwo different dataset that were identified using two different methods, the Darkhorse method and the phylogenetic method. The total numberof the union of HGT genes between two dataset is 55. (B) The diagram showing the origination of those 55 HGT genes. All of them originatedfrom prokaryotes.Pan et al. BMC Genomics 2013, 14:186 Page 7 of 14http://www.biomedcentral.com/1471-2164/14/186unknown functions, a handful of them are not. For ex-ample, CPG844 is related to LPXTG motif cell wall an-chor domain protein and CPG1776 is related to surfaceadhesion protein. Positive selection acting on these twoCPGs might play an important role in host recognitionand interaction since they are involved in surface adhe-sion. Other examples are two positively selected CPGsthat are related to serine protease inhibitor (SPN106).The functions of serine protease inhibitor have beenshown to decrease the immune responses in hosts[30-32]. In the melanization pathway of B. mori, serineprotease cascade is one of the most important biochem-ical reactions to inhibit the propagation of pathogens[33]. We scanned the gene expression pattern of B. moripost infection of N. bombycis by microarray analysisand found that the key gene PPO in melanization path-way was significantly suppressed. From our gene chipexpression analysis (unpublished data), we found thatthe gene expression of two upstream regulators of hostPPO melanization pathway, β-GRP2 and β-GRP4, in B.Figure 4 Gene duplications and the dS distribution of paralogs and orthologs among three Nosema species. Abbreviation: Na,N. antheraeae; Nb, N. bombycis; Nc, N. ceranae. (A) A circos map showing the comparative genomics among three different Nosema species basedon all available scaffolds. Each line represents the homologous syntenic regions between any two species or between any given twochromosome positions of single species. Many lines across different scaffolds of N. bombysis indicates higher rate of segmental syntenicduplications. (B) The dS distribution of segmental paralogs of Nb and the orthologs between Nb and Na showing a higher dS values in orthologsin general. Notably, a higher peak (arrow) seen in Nb suggests the possibility of a burst of paralogs recently over the Nb evolution after theseparation of Na and Nb. (C) An example of syntenic comparisons among three Nosema species showing a cluster of tandem paralogs. Thecorresponding genetic position and names of identified element are provided in Additional file 8. The number of all identified tandem paralogsfor each Nosema genome is summarized on the right side. (D) The dS distribution of tandem paralogs of Nb and orthologs between Na and Nbshowing that majorities of tandem paralogs arose after the separation of Na and Nb because the dS values of those tandem paralogs are smallerthan that of orthologs.Pan et al. BMC Genomics 2013, 14:186 Page 8 of 14http://www.biomedcentral.com/1471-2164/14/186mori are up-regulated during the infection of N.bombycis (Figure 6). The upregulation of β-GRP2 and β-GRP4 subsequently suppress the production of PPOagainst the infection of N. bombycis. To seek for thetreatment method of pébrine, the interplay betweenSPN106 and β-GRP should be treated as our priorityin the future studies. These observations lead us tohypothesize that adaptive evolution of serine proteaseinhibitor in N. bombycis might deter the melanizationpathway by blocking the serine protease cascade in do-mesticated silkworms (Figure 6).Discussion and conclusionPreventing the infection of Nosema bombycis is one of theprime concerns in the domesticated silkworm industry.However, attempts to manage these pathogens have beenhindered by our poor knowledge of the underlying mo-lecular mechanisms contributing to the highly infectiousability of N. bombycis. In the absences of transformation,genetics, and axenic cultivation, the comparative genomicsapproach is one of the few tools available to tackle theseissues. In this study, we compared the genome of themajor commercial silkworm pathogen, N. bombycis, tothose of N. antheraeae and N. ceranae. Our study showedthat the large genome size in silkworm Nosema genome isdue to the proliferation of host-derived transposable ele-ments, horizontally transferred genes from prokaryotes,and the production of segmental and tandem duplicates.Previous studies on the characterization of microsporidiangenomic architectures have focused more on the gen-ome reduction aspect [15-19,34-36]. Although few stud-ies assumed the possibility of genome expansion in themicrosporidia [23,37], the direct evidence is lacking. Fromthe genome streamlining perspective, it is evident thatmany metabolic essential genes (e.g., the tricarboxylic acidcycle, fatty acid β-oxidation, respiratory electron-transportchain always) are missing in the microsporidian genome. Instark contrast, our study provides the first solid evidenceshowing that the microsporidian genome can expand. Geneduplications and proliferation of host-derived transposableelements are the two predominant molecular mechanismscontributing to the genome expansion in N. bombycis.Recently, two studies have reported that some genes inthe microsporidia Encephalitozoon romaleae were derivedfrom an ancestral host [38,39], but we did not find any evi-dence of host-derived genes in N. bombycis. Instead, somegenes in the N. bombycis genome were apparently derivedfrom prokaryotes or viruses by horizontal gene transfer,similar to other microsporidian genomes [26,27]. Surpris-ingly, these prokaryote-transferred genes could comple-ment some important metabolic pathways in Nosema,indicative of its essentiality over the course of Nosemaevolution. The mobility of transposable elements has beenshown to be associated with the frequency of horizontalgene transfer [40,41]. Although the N. bombycis genome iscomposed of ~38% repetivitive elements, only 55 geneswere found to be horizontally transferred. Such observa-tion indicated that a great number of transposons will notlead to higher rate of HGTs in N. bombycis. One explan-ation is that most transposons of N. bombycis have losttheir activities such that the rate of HGTs is not enhanced.Alternatively, those repetitive elements do not possess theability of capturing genes to facilitate the rate of HGTs.The other major source of novel genetic materials inN. bombycis is the numerous paralogs by large-scale andsmall-scale duplication events. Some of them showFigure 5 An example from the syntenic analysis showing that N. bombycis often consist of two homologous regions, instead of onesuch as N. antheraeae. When summarized the number of syntenic regions in both N. antheraeae and N. bombycis, the number of paralogoussyntenic region of N. bombycis is often twice more than that of orthologous syntenic region between N. antheraeae and N. bombycis, indicatingthat large segmental duplication events have occurred over the evolution of N. bombycis.Pan et al. BMC Genomics 2013, 14:186 Page 9 of 14http://www.biomedcentral.com/1471-2164/14/186evidence of accelerated changes through the relaxationof purifying selection, whereas others show evidence ofpositive selection. In either case, these paralogs seem tohave provided raw materials for functional innovationsas we showed in this study. Among them, the serine pro-tease inhibitor family stands out as potential targets tostudy the higher infectious rates in N. bombycis.MethodsExtraction of DNA, library construction, genomeassembly, and annotationAbout 1 × 109 spores of N. bombycis CQ1, were isolatedfrom infected silkworms in Chongqing. Using lysis buffercontaining SDS and proteinase K, the N. bombycis gen-omic DNA was extracted from the germinated sporesfor each library construction. N. antheraeae YY isolatewere collected from a farm in Henan of China and pre-served in our lab, and then DNA was extracted withcetyl trimethylammonium bromide (CTAB) method. De-tailed methods for extracting genomic DNAs can see theonline supplementary materials (Additional file 16).After the library construction of plasmid and miniBAC, allsequence reads were obtained by Sanger and Illuminasequencing. To assemble the N. bombycis genome, theillumina reads were assembled by BGI’s de novo assemblysoftware [42], which assembled unique and frequency re-peat areas of genome. Next, we assembled further fromthe mixed data of illumina scaffolds and Sanger reads byusing Phrap program.To ensure the quality of gene annotation, we enrichedthe ESTs data by constructing two cDNA phage librariesand two Illumina cDNA libraries. 11,155 high qualityreads were obtained by Sanger sequencing and 307,900reads were obtained by illumina sequencing. Finally, 1517unique ESTs were obtained with average length of 430 bp.Next, N. bombycis protein-coding genes were annotatedusing the following three different software: (1) Glimmer(version 3.0) with low eukaryote parameters [43], (2)GeneMarkS (version 4.6) with low eukaryote parameters[44], and (3) Augustus (version 2.0) with defaultTable 3 Site test of adaptive evolution for N. bombycis paralogous genesDescription CPG Members DuplicatetypeLRT statistics dN/dSM1 vs. M2 M7 vs. M8 M8 vs. M8aHypothetical protein CPG199 4 DD 17.30*** 17.68*** 17.07*** 6.55472Hypothetical protein CPG293 4 SD,DD 7.34* 7.58* 6.68* 67.43545Hypothetical protein CPG446 3 TD 15.34*** 16.15*** 16.52*** 96.64162Hypothetical protein CPG767 3 TD,DD 13.80** 13.92*** 13.26** 12.30571Surface adhesion protein CPG844 3 SD, DD 21.04*** 21.03*** 22.81*** 6.72603Serine protease inhibitor 106 CPG945 3 TD,DD 5.85 5.92 5.8 19.13308Serine protease inhibitor 106 CPG1974 2 DD 15.81*** 16.31** 15.81*** 52.1588LPXTG-motif cell wall anchor domain protein+ CPG1776 2 SD 14.38*** 14.39** 14.38*** 20.02006DnaJ homolog subfamily C member 9 CPG1792 2 SD 13.76** 13.85*** 19.87*** 57.65414NIK and IKK(beta) binding protein CPG1878 2 SD 12.22** 12.59** 14.57*** 36.70677Replication-associated protein CPG2140 2 DD 18.67*** 18.76*** 21.14*** 15.15205Glucan endo-1,6-beta-glucosidase CPG1640 2 DD 5.72 5.72 9.64** 61.53906Integrator complex subunit 4 CPG2120 2 SD 8.46* 8.59* 8.22* 14.58649Hypothetical protein CPG961 3 DD 21.04*** 22.99*** 19.99*** 9.5982Hypothetical protein CPG1520 2 SD 15.28*** 15.37*** 15.13*** 8.84271Hypothetical protein CPG1637 2 SD 18.44*** 18.75*** 20.57*** 14.62068Hypothetical protein CPG987 3 DD 13.65** 13.37** 13.11** 39.08442Hypothetical protein CPG1865 2 SD 10.90** 10.94** 10.70** 23.75418Hypothetical protein CPG1967 2 SD 6.06* 6.81* 5.48 29.25563Hypothetical protein CPG2026 2 DD 8.28* 8.41* 15.98*** 15.9724Hypothetical protein CPG2083 2 DD 29.01*** 29.00*** 31.16*** 61.60344hypothetical protein CPG775 3 SD 14.05*** 14.05*** 12.37** 5.3171Hypothetical protein CPG2128 2 SD 17.79*** 17.85*** 19.27*** 45.49562Hypothetical protein CPG869 3 SD, DD 17.45*** 17.85*** 17.31*** 8.16665Note: SD, segmental; TD, Tandem; DD, Disperse; M1 vs. M2, LRT statistic for model M1 versus M2; M7 vs. M8, LRT for model M7 versus M8; * Significance withP < 0.05; ** Significance with P < 0.01; *** Significance with P < 0.001.Pan et al. BMC Genomics 2013, 14:186 Page 10 of 14http://www.biomedcentral.com/1471-2164/14/186parameters [45]. The details of library construction ofplasmid, miniBAC, DNA and cDNA, as well as the proto-col of genome assembly and annotations, for N. bombycisand N. antheraeae are provided as online supplementarymaterials. All annotated sequences of N. bombycis andN. antheraeae are deposited in Genbank as the follow-ing accession numbers: ACJZ01000001-ACJZ01003558.Identification horizontal gene transfer (HGT)To examine the frequency of host-derived transposableelements, a phylogenetic analysis was conducted usingthe software RAxML [46] with the maximum likelihood(ML) algorithm. The amino acid replacement matrix,the WAG matrix, with gamma distribution was used toreconstruct the phylogenetic tree. Statistical support fornodes was estimated by using the bootstrapping methodwith 100 ML replicates. All other HGT genes of the N.bombycis genome were identified by using both the phylo-genetic method and the Darkhorse methods [47]. For thephylogenetic method, all initial 4,458N. bombycis geneswere clustered to 3609 singletons at the level of ≥ 75%identity over ≥ 90% coverage for cluster members usingBLASTCLUST program. A single randomly chosen repre-sentative of each cluster was used as a seed for BLASTPsearches on nr database, the Bombyx mori genome data-base (http://silkworm.genomics.org.cn/). Sequences withE-value < 1e-5 and > 70% of the protein length) werealigned using clustal W. Bootstrap (100 replicates) consen-sus WAG model was made using RAxML to reconstructNeighbor joining (NJ) trees. For the Darkhorse method, afilter threshold of 20% and two different self-definition key-words (N. bombycis and all species name of Microsporidiaphylum) were used to eliminate the BLASTP matches bycalculating the lineage probability index (LPI) of genes inthe N. bombycis genome. Then, the potential horizontallytransferred genes were retrieved.Identification of segmental and tandem duplicationsTo identify the segmental duplication, we performed all-against-all blast search with a single species to identifyFigure 6 A hypothetical model showing how the SPN protein of N. bombycis suppresses the serine protease cascade of themelanization pathway of the host B. mori. After the suppression of the serine protease cascade, the defensive response, the subsequentformation of melanization will be inhibited in the hosts. Abbreviation: PPO, prophenoloxidase; β-GRP, β-glucan recognition protein.Pan et al. BMC Genomics 2013, 14:186 Page 11 of 14http://www.biomedcentral.com/1471-2164/14/186collinear regions within single genome as segmentalduplicated blocks. A collinear region was defined as onewhere there are at least three homologous pairs withE value < 1E-6 and the distance between genes less than5 kb. Segmental blocks were visualized using the soft-ware Circos-0.55 [48]. To plot duplicated blocks amongN. bombycis, N. antheraeae, and N. ceranae genomes,we ordered the scaffolds as follows: 1) only the scaffoldsthat shared syntenic genes among these three specieswere included; 2) the scaffolds of N. bombycis wereranked from longest to shortest; 3) scaffolds of theother two species were arranged based on synteny toN. bombycis; 4) if N. antheraeae or N. ceranae scaf-folds were syntenic to more than two scaffolds of N.bombycis, we define that scaffold order based on thelongest scaffold of N. bombycis.For the identification of tandem duplicates, we firstclassified gene family using the software MCL with Evalue < 1E-10, and then defined tandem duplicates as fol-lows: 1) belonging to the same gene family, 2) being lo-cated within 5 kb each other, and 3) being separatedby ≤ 3 non-homologous genes.To time the age of paralogs, we first identified collinearregions between N. bombycis and N. antheraeae. Then,genes that lie in the collinear region were classified asorthologs between N. bombycis and N. antheraeae. Syn-onymous substitution rate (dS) of paralogs was estimatedusing the software Codeml in the package PAML 4 [49].Estimation of gene-wide selection and codon-basedselectionThe gene-wide selection and codon-based selection ofgenes in N. bombycis were analyzed following the proced-ure described in [29]. Briefly, clusters of homologousgenes in N. bombycis, N. antheraeae, and N. ceranae iden-tified by MCL (<1E-10) were grouped into four differentcategories: 1) clusters of orthologous genes (COGs) of1:1:1 orthologous trios without any subsequent gene du-plication in any species, 2) COGs of 1:1N. bombycis andN. antheraeae gene pairs, 3) COGs of 1:1N. bombycis andN. ceranae gene pairs, and 4) clusters of paralogous genes(CPGs) in N. bombycis. Prior to the estimation of dN/dS,those clusters with identity < 50% and an area covering<50% of the length of each sequence were filtered. Mul-tiple alignments using MUSCLE [50] were then parsed toremove those poorly aligned regions using the Gblocks al-gorithm [51] with the following criteria: maximum num-ber of contiguous non-conserved positions = 10 andminimum length of a block = 5. The amino acid alignmentwas back-translated into nucleotides sequence alignment.The gene-wide selection was then determined by calculat-ing the median dN/dS value for each cluster. For the de-tection of codon-based positive selection, codon-basedselection analysis was implemented using the Codemlprogram in the PAML package [49]. The site-specificmodel was used to detect positive selection in CPGs of N.bombycis. Two likelihood ratio tests were implemented:M1-M2 and M7-M8. M1 and M7 are the null modelswithout positive selection, while M2 and M8 are the alter-native models with positive selection. For each test, thefirst model (i.e., M1 or M7) is simpler than the secondmodel (i.e., M2 or M8). To test if the second model fitsbetter than the first model, twice difference of logarithmmaximum likelihood estimates between the two comparedmodels was compared against chi-square distribution withtwo degrees of freedom. Only those that showed posteriorprobability > 0.95 in the empirical Bayes method wereconsidered as positively selected sites [52,53].RNA labeling and hybridizationRNA labeling and microarray hybridization was conductedby CapitalBio Corp (Beijing, China). Gene expression ana-lysis was done based on the Affymetrix Silkworm GeneChip kit in accordance following the manufacturer’s in-struction (http://www.capitalbio.com). Briefly, after 5 × 104spores were fed to 3-instar larvae, total RNA was isolatedfrom those 3-instar larvae at day 2, 4, 6, and 8. Then theextracted total RNA was reverse transcribed into cDNA.A dual-dye experiment was conducted. The uninfectedcDNAs were labeled with dye Cy3 and infected cDNAswere labeled with dye Cy5. The labeled cDNA probeswere dissolved in hybridization solution overnight at42°C and then hybridized to the 23 k silkworm genomeoligonucleotide chip (Capital Bio) that consists of22,987 oligonucleotide 70-mer probes [54]. The signalswere scanned with LuxScan 10KA scanner (CapitalBiocorp).Three biological repeats were conducted at eachtime point.Additional filesAdditional file 1: Summary of reads data production in N.bombycis.Additional file 2: Statistics of genome assembly in N. bombycis.Additional file 3: The calculation of genomic size of N. antheraeaebased on the frequency distribution of 15-mers depth of reads.Additional file 4: Gene ontology classification of threemicrosporidian Nosema. Y axis: Log 10 (the proportion of genenumbers in certain sort occupied on the total Go-annotated genenumbers).Additional file 5: Comparison of average length of homologousgene among five different microsporidian species.Additional file 6: List of main transposable elements among threeNosema species.Additional file 7: The potential transposition of DNA sequences byPiggybac element.Additional file 8: Complete list of the annotation genes fromgenomes, and some lists of genetic position/names of identifiedelement in text.Pan et al. BMC Genomics 2013, 14:186 Page 12 of 14http://www.biomedcentral.com/1471-2164/14/186Additional file 9: Diagram showing the Piggybac transposon-mediated exogenous DNA sequence in the collinear region of N.bombycis. TTAA indicates the recognition site of the Piggybactransposon.Additional file 10: Summary of 55 horizontally transferred genes inthe N. bombycis genome.Additional file 11: Figure showing phosphomevalonate kinase thathorizontal transfer from bacteria integrates the mevalonatepathway of N. bombycis.Additional file 12: The location of segmental duplications inNosema bombycis genome.Additional file 13: Cumulative plot and statistics of the dN/dSvalues for CPG and COG genes in N. bombycis.Additional file 14: The numbers of EST tags for positively selectedCPG genes.Additional file 15: Table of parameter estimates for all positivelyselected CPGs.Additional file 16: Supplementary materials for library constructionand genomic assembly.Competing interestsThe author(s) declare that we have no competing interests.Authors’ contributionsConceived and designed the experiments: ZYZ ZHX QYX JW CL ZZ NJH;Performed the experiments: GQP JSX TL CFL HDL LY XZ ZLW XQD MLT YHL;Analyzed the data: GQP JSX TL SLL GJZ SGL TL WF HX; Contributed reagents/materials/analysis tools: JHH ZL LPL JL LNG LLW MXL YJW; Wrote the paper:GQP ZYZ JSX QYX SLL PJK. All authors read and approved the finalmanuscript.AcknowledgementsWe acknowledge the support of Prof. Christian P. Vivarès (Université BlaisePascal, France) for his team’s help on analysis of chromosomes organizationof N. bombycis by PFGE. Thanks for the help of bioinformatics analysis fromChongqing NoeGen Bioinformatics Technology Co., LTD (http://www.noegen.com). This work is supported by the grants from National BasicResearch Program of China (No.2012CB114604), Natural Science Foundation ofChina (No. 30930067, 31001037, 31001036, 31072089, 31272504, 31270138),National High-tech R&D Program (863 Program, No. 2012AA101301-3, No.2013AA102507), Chongqing Science & Technology Commission (No.CSTC2010AA1003), the Program of Introducing Talents of Discipline toUniversities (No.B07045), Key Project of Ministry of Education of China(No.210180).The genome data and annotation information of N. bombycis and N.antheraeae were submitted to GenBank (Accession numbers ACJZ01000001 -ACJZ01003558).These data are also freely available at our public lab website: http://microbe.swu.edu.cn/bio/genome/nosema.Author details1State Key Laboratory of Silkworm Genome Biology, Southwest University,Chongqing 400715, China. 2College of Life Sciences, Chongqing NormalUniversity, Chongqing 400047, China. 3Department of Botany, University ofBritish Columbia, Vancouver, British Columbia V6T 1Z4, Canada. 4BeijingGenomics Institute at Shenzhen, Shenzhen 518000, China.Received: 26 September 2012 Accepted: 26 February 2013Published: 16 March 2013References1. Desportes I, Le Charpentier Y, Galian A, Bernard F, Cochand-Priollet B,Lavergne A, Ravisse P, Modigliani R: Occurrence of a new microsporidian:Enterocytozoon bieneusi n.g., n.sp., in the enterocytes of a human patientwith AIDS. Journal of Protozool 1985, 32:250–254.2. Snowden KF: Zoonotic microsporidia from animals and arthropods witha discussion of human infection. In Opporunistic infections: toxolasma,sarcocystis, and microsporida. Edited by Lindsay DS, Weiss LM. Boston, MA:World Class Parasites; 2004:123–134.3. Wittner M: Historic perspective on the microsporidia: expandinghorizons. In The microsporidia and microsporidiosis. Edited by Wittner M,Weiss LM. Washington, DC: ASM Press; 1999:1–6.4. Nageli K: Uber die neue krankheit der seidenraupe und verwandteorganismen. Bot Z 1857, 15:760–761.5. Pasteur L: Études sur la maladie des vers a soie: moyen pratique assurÉ de lacombattre et d’en prÉvenir le retour. Paris: Imprimeur-Libraire; 1870.6. Kashkarova L, Khakhanov A: Range of the hosts of the causative agent ofpébrine (Nosema bombycis) in the mulberry silkworm. Parazitologiia 1980,14:164.7. Kudo R, DeCoursey J: Experimental infection of hyphantria cunea withnosema bombycis. J Parasitol 1940, 26:123–125.8. Hayasaka S, Yonemura N: Infection and development of Nosema sp. NIS H5(Microsporida: Protozoa) in several lepidopteran insects. JARQ 1999, 33:65–68.9. Xu JS, Zhou ZY: Improving phylogenetic inference of microsporidianNosema antheraeae among Nosema species with RPB1, α-and β-tubulinsequences. Afr J Biotechnol 2010, 9:7900–7904.10. Cornman RS, Chen YP, Schatz MC, Street C, Zhao Y, Desany B, Egholm M,Hutchison S, Pettis JS, Lipkin WI, Evans JD: Genomic analyses of themicrosporidian Nosema ceranae, an emergent pathogen of honey bees.PLoS Pathog 2009, 5:e1000466.11. Keeling PJ, Fast NM: Microsporidia: biology and evolution of highlyreduced intracellular parasites. Annu Rev Microbiol 2002, 56:93–116.12. Corradi N, Slamovits CH: The intriguing nature of microsporidiangenomes. Brief Funct Genomics 2011, 10:115–24.13. Kawakami Y, Inoue T, Ito K, Kitamizu K, Hanawa C, Ando T, Iwano H, IshiharaR: Identification of a chromosome harboring the small subunit ribosomalRNA gene of Nosema bombycis. J Invertebr Pathol 1994, 64:147.14. Xu J, Wang L, Tang F, Huang W, Zhou Z: The nuclear apparatus andchromosomal DNA of the microsporidian nosema antheraeae. J EukaryotMicrobiol 2011, 58:178–180.15. Katinka MD, Duprat S, Cornillot E, Méténier G, Thomarat F, Prensier G, BarbeV, Peyretaillade E, Brottier P, Wincker P, Delbac F, El Alaoui H, Peyret P,Saurin W, Gouy M, Weissenbach J, Vivarès CP: Genome sequence and genecompaction of the eukaryote parasite Encephalitozoon cuniculi. Nature2001, 414:450–453.16. Slamovits CH, Fast NM, Law JS, Keeling PJ: Genome compaction andstability in microsporidian intracellular parasites. Curr Biol 2004,14:891–896.17. Akiyoshi DE, Morrison HG, Lei S, Feng X, Zhang Q, Corradi N, Mayanja H,Tumwine JK, Keeling PJ, Weiss LM, Tzipori S: Genomic survey of the non-cultivatable opportunistic human pathogen. Enterocytozoon bieneusi. PLoSpathogens 2009, 5:e1000261.18. Corradi N, Pombert JF, Farinelli L, Didier ES, Keeling PJ: The completesequence of the smallest known nuclear genome from themicrosporidian Encephalitozoon intestinalis. Nat Commun 2010, 1:77.19. Keeling PJ, Corradi N, Morrison HG, Haag KL, Ebert D, Weiss LM, Akiyoshi DE,Tzipori S: The reduced genome of the parasitic microsporidianenterocytozoon bieneusi lacks genes for core carbon metabolism.Genome Biol Evol 2010, 2:304.20. Feschotte C, Jiang N, Wessler SR: Plant transposable elements wheregenetics meets genomics. Nat Rev Genet 2002, 3:329–41.21. Hinkle G, Morrison H, Sogin M: Genes coding for reverse transcriptase,DNA-directed RNA polymerase, and chitin synthase from themicrosporidian Spraguea lophii. Biol Bull 1997, 193:250.22. Xu J, Pan G, Fang L, Li J, Tian X, Li T, Zhou Z, Xiang Z: The varyingmicrosporidian genome: existence of long-terminal repeatretrotransposon in domesticated silkworm parasite Nosema bombycis.Int J Parasitol 2006, 36:1049–1056.23. Williams BA, Lee RC, Becnel JJ, Weiss LM, Fast NM, Keeling PJ: Genomesequence surveys of Brachiola algerae and Edhazardia aedis revealmicrosporidia with low gene densities. BMC Genomics 2008, 9:200.24. Jinshan X, Jie L, Bettina D-V, Xiaoyan Z, Hangdeng L, Zeyang Z:Characterization of a transcriptionally active Tc1-like transposon in themicrosporidian Nosema bombycis. Acta Parasitologica 2010, 55:8–15.25. Xu J, Wang M, Zhang X, Tang F, Pan G, Zhou Z: Identification of NbMEMITE families: potential molecular markers in the microsporidia nosemabombycis. J Invertebr Pathol 2010, 103:48–52.Pan et al. BMC Genomics 2013, 14:186 Page 13 of 14http://www.biomedcentral.com/1471-2164/14/18626. Fast NM, Law JS, Williams BAP, Keeling PJ: Bacterial catalase in themicrosporidian Nosema locustae: implications for microsporidianmetabolism and genome evolution. Eukaryot Cell 2003, 2:1069–1075.27. Richards TA, Hirt RP, Williams B, Embley TM: Horizontal gene transfer andthe evolution of parasitic protozoa. Protist 2003, 154:17.28. Lynch M: The origins of genome architecture. Sinauer Press; 2007.29. Emes RD, Yang Z: Duplicated paralogous genes subject to positiveselection in the genome of Trypanosoma brucei. PLoS One 2008, 3:e2295.30. Turner PC, Moyer RW: Poxvirus immune modulators: functional insightsfrom animal models. Virus Res 2002, 88:35–53.31. Richardson J, Viswanathan K, Lucas A: Serpins, the vasculature, and viraltherapeutics. Front Biosci 2006, 11:1042–1056.32. Tao M, Pan G, Hu H, Zhou Z: :Identification of a new serpin gene(NbSPN106) from Nosema bombycis. Wei sheng wu xue bao = ActaMicrobiol Sin 2009, 49:726.33. Tanaka H, Yamakawa M: Regulation of the innate immune responses inthe silkworm, Bombyx mori. ISJ 2011, 8:59–69.34. Corradi N, Akiyoshi DE, Morrison HG, Feng X, Weiss LM, Tzipori S, Keeling PJ:Patterns of genome evolution among the microsporidian parasitesencephalitozoon cuniculi. Antonospora locustae and Enterocytozoonbieneusi. PLoS One 2007, 2:e1277.35. Tsaousis AD, Kunji ER, Goldberg AV, Lucocq JM, Hirt RP, Embley TM: A novelroute for ATP acquisition by the remnant mitochondria ofEncephalitozoon cuniculi. Nature 2008, 453:553–556.36. Gill E: Fast NS:tripped-down DNA repair in a highly reduced parasite.BMC Mol Biol 2007, 8:24.37. Corradi N, Haag KL, Pombert JF, Ebert D, Keeling PJ: Draft genomesequence of the Daphnia pathogen Octosporea bayeri: insights into thegene content of a large microsporidian genome and a model for host-parasite interactions. Genome Biol 2009, 10:R106.38. Selman M, Pombert JF, Solter L, Farinelli L, Weiss LM, Keeling P, Corradi N:Acquisition of an animal gene by microsporidian intracellular parasites.Curr Biol 2011, 21:R576.39. Pombert JF, Selman M, Burki F, Bardell FT, Farinelli L, Solter LF, WhitmanDW, Weiss LM, Corradi N, Keeling PJ: Gain and loss of multiple functionallyrelated, horizontally transferred genes in the reduced genomes of twomicrosporidian parasites. Proc Natl Acad Sci USA 2012, 109:38–43.40. Keeling PJ, Palmer JD: Horizontal gene transfer in eukaryotic evolution.Nat Rev Genet 2008, 9:605–618.41. Archibald J, Richards T: Gene transfer: anything goes in plantmitochondria. BMC Biol 2010, 8:147.42. Li R, Zhu H, Ruan J, Qian W, Fang X, Shi Z, Li Y, Li S, Shan G, Kristiansen K, LiS, Yang H, Wang J, Wang J: De novo assembly of human genomes withmassively parallel short read sequencing. Genome Res 2010, 20:265.43. Delcher AL, Harmon D, Kasif S, White O, Salzberg SL: Improved microbialgene identification with GLIMMER. Nucleic Acids Res 1999, 27:4636.44. Lomsadze A, Ter-Hovhannisyan V, Chernoff YO, Borodovsky M: Geneidentification in novel eukaryotic genomes by self-training algorithm.Nucleic Acids Res 2005, 33:6494.45. Stanke M, Diekhans M, Baertsch R, Haussler D: Using native andsyntenically mapped cDNA alignments to improve de novo genefinding. Bioinformatics 2008, 24:637.46. Stamatakis A: RAxML-VI-HPC: maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models. Bioinformatics 2006,22:2688.47. Podell S, Gaasterland T: DarkHorse: a method for genome-wide predictionof horizontal gene transfer. Genome Biol 2007, 8:R16.48. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, Jones SJ,Marra MA: Circos: an information aesthetic for comparative genomics.Genome Res 2009, 19:1639.49. Yang Z: PAML4: phylogenetic analysis by maximum likelihood. Mol BiolEvol 2007, 24:1586–1591.50. Edgar R: MUSCLE: a multiple sequence alignment method with reducedtime and space complexity. BMC Bioinforma 2004, 5:113.51. Castresana J: Selection of conserved blocks from multiple alignments fortheir use in phylogenetic analysis. Mol Biol Evol 2000, 17:540.52. Nielsen R, Yang Z: Likelihood models for detecting positively selectedamino acid sites and applications to the HIV-1 envelope gene. Genetics1998, 148:929.53. Wong WSW, Yang Z, Goldman N, Nielsen R: Accuracy and power ofstatistical methods for detecting adaptive evolution in protein codingsequences and for identifying positively selected sites. Genetics 2004,168:1041.54. Huang L, Cheng T, Xu P, Cheng D, Fang T, Xia Q: A genome-wide surveyfor host response of silkworm. Bombyx mori during pathogen bacillusbombyseptieus infection. PloS one 2009, 4:e8098.doi:10.1186/1471-2164-14-186Cite this article as: Pan et al.: Comparative genomics of parasiticsilkworm microsporidia reveal an association between genomeexpansion and host adaptation. BMC Genomics 2013 14:186.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitPan et al. BMC Genomics 2013, 14:186 Page 14 of 14http://www.biomedcentral.com/1471-2164/14/186


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items