Open Collections

UBC Faculty Research and Publications

Genome sequence surveys of Brachiola algerae and Edhazardia aedis reveal microsporidia with low gene… Williams, Bryony A; Lee, Renny C; Becnel, James J; Weiss, Louis M; Fast, Naomi M; Keeling, Patrick J Apr 29, 2008

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

Item Metadata


52383-12864_2007_Article_1393.pdf [ 570.51kB ]
JSON: 52383-1.0221616.json
JSON-LD: 52383-1.0221616-ld.json
RDF/XML (Pretty): 52383-1.0221616-rdf.xml
RDF/JSON: 52383-1.0221616-rdf.json
Turtle: 52383-1.0221616-turtle.txt
N-Triples: 52383-1.0221616-rdf-ntriples.txt
Original Record: 52383-1.0221616-source.json
Full Text

Full Text

ralssBioMed CentBMC GenomicsOpen AcceResearch articleGenome sequence surveys of Brachiola algerae and Edhazardia aedis reveal microsporidia with low gene densitiesBryony AP Williams†1, Renny CH Lee†1, James J Becnel2, Louis M Weiss3, Naomi M Fast1 and Patrick J Keeling*1Address: 1Canadian Institute for Advanced Research, Department of Botany, University of British Columbia, 3529-6270 University Boulevard, Vancouver, BC, V6T 1Z4, Canada, 2Center for Medical, Agricultural and Veterinary Entomology, USDA/ARS, Gainesville, FL 32608, USA and 3Department of Pathology, Division of Tropical Medicine and Parasitology, Albert Einstein College of Medicine, Bronx, New York 10461, USAEmail: Bryony AP Williams -; Renny CH Lee -; James J Becnel -; Louis M Weiss -; Naomi M Fast -; Patrick J Keeling* -* Corresponding author    †Equal contributorsAbstractBackground: Microsporidia are well known models of extreme nuclear genome reduction andcompaction. The smallest microsporidian genomes have received the most attention, but genomesof different species range in size from 2.3 Mb to 19.5 Mb and the nature of the larger genomesremains unknown.Results: Here we have undertaken genome sequence surveys of two diverse microsporidia,Brachiola algerae and Edhazardia aedis. In both species we find very large intergenic regions, manytransposable elements, and a low gene-density, all in contrast to the small, model microsporidiangenomes. We also find no recognizable genes that are not also found in other surveyed orsequenced microsporidian genomes.Conclusion: Our results demonstrate that microsporidian genome architecture varies greatlybetween microsporidia. Much of the genome size difference could be accounted for by non-codingmaterial, such as intergenic spaces and retrotransposons, and this suggests that the forces dictatinggenome size may vary across the phylum.BackgroundMicrosporidia are obligate intracellular eukaryotic para-sites that have been found to infect members of all majoranimal lineages [1]. The many apparently "primitive" fea-tures of microsporidian cells led evolutionary biologiststo suggest that they were an early-branching lineage of theeukaryotes [2,3], but molecular phylogeny has sinceshown that they are instead a derived relative of fungihave been re-evaluated as products of reduction and adap-tation to life inside another cell [4-8].One such feature that has attracted considerable attentionis their highly reduced genomes. The genome size hasbeen determined for numerous microsporidian species,and they range from 19.5 Mbp in Glugea atherinae to just2.3 Mbp in Encephalitozoon intestinalis, the smallestPublished: 29 April 2008BMC Genomics 2008, 9:200 doi:10.1186/1471-2164-9-200Received: 12 September 2007Accepted: 29 April 2008This article is available from:© 2008 Williams et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 9(page number not for citation purposes)[4,5]. In light of this, their seemingly primitive features eukaryotic genome known [9]. The two best-studiedexamples are the vertebrate parasite EncephalitozoonBMC Genomics 2008, 9:200 (2.9 Mbp), the genome of which has been com-pletely sequenced and encodes 1,997 protein-codinggenes [10], and the insect parasite Antonospora locustae(5.4 Mbp) for which two sequence surveys are available[11,12]. These two species revealed just how microsporid-ian genomes had become so small compared with thoseof other eukaryotes. There has been a severe reduction inthe number of genes in the genome, most likely a reflec-tion of the fact that microsporidia are dependent on theirhosts for many metabolic processes and import manycompounds from their host. Furthermore the genes thatremain are packed together very densely: intergenic spacesare minimal. In E. cuniculi there are no selfish elementsand just 15 small introns. Genes in E. cuniculi are alsoshorter than their homologues in Saccharomyces cerevisiae,which is hypothesized to result from the small number ofproteins within the cell, and a correspondingly smallerinteraction network [10,13]. This extreme compactionappears to have resulted in a high level of gene order con-servation between different species of microsporidia[12,14-16] and an unusually high level of overlappingtranscription between adjacent genes [17,18].Because the smallest microsporidian genomes are so unu-sual, they have garnered the greatest attention, and to dateno large-scale survey of any larger genomes is available.This is unfortunate, because the form and content of theselarger microsporidian genomes could differ from thesmaller ones in many potentially interesting ways. On onehand they may contain a great many more genes, andcould therefore reflect a greater cellular or metabolic com-plexity than the microsporidian parasites we presentlyknow best. On the other hand these genomes may encodea great deal more non-coding DNA, which would haveinteresting implications for genome evolution within thegroup, and for why the smaller genomes are so compact.In other eukaryotes, it appears that variation in genomesize on a relatively short evolutionary time scale is due toincreased or decreased proportions of transposable ele-ments in a genome [19]. There are hints that this may alsobe at least partially true in some microsporidian genomeswhere selfish elements have been found [20,21]. Mostinterestingly, a number of TY3/Gypsy retrotransposonshave recently been described from the 15.3 Mbp genomeof Nosema bombycis [22]. These elements are apparentlyflanked by areas of compacted genes that share a highlevel of synteny with E. cuniculi, perhaps suggesting aninvasion of transposable elements into a compactedgenome. More recently, the previously unsequenced sub-telomeric areas of E. cuniculi have been investigated andfound to contain a large family of proteins with at least 30distinct members. This family of duplicated proteins isalso found within other human infecting microsporidia[23], demonstrating that the expansion of gene familiesmight be more common than previously thought.Here, we describe low-redundancy genome sequence sur-veys of two distantly related microsporidia, Brachiola alge-rae (recently proposed to be renamed Anncaliia algerae[24]) and Edhazardia aedis. Both of these species havemosquitoes as their type hosts with the former having avery broad host range, including man, and the latterrestricted to mosquitoes [25]. The genome size of E. aedisis not known, whereas the genome size of B. algerae hasbeen estimated to be 15–20 Mbp [23]. In both species, weshow that gene density is very low compared to that of thebetter-studied species of microsporidia. Specifically, wehave found a considerable proportion of repetitive ele-ments in both genomes, large stretches of non-codingDNA, and some evidence that the gene density may varyover the genome. These surveys open the possibility thatmicrosporidian genomes are not universally compacted,large genomes do not necessarily encode significantlymore genes than do the smaller genomes, and that densegenomes may sometimes revert to a gene-sparse state.Results and discussionGeneral features of the sequence dataTwo short-insert genomic libraries were constructed fromB. algerae and a total of 219 clones fully or partiallysequenced to yield 203,748 bp of non-overlappingsequence. A single E. aedis library was constructed and 290sequence reads from 182 clones yielded 233,509 bp ofnon-overlapping sequence. Comparing these with theTable 1: Summary of the data compared to the genome of E. cuniculiE. cuniculi [10] B. algerae E. aedisbp % bp % bp %Total sequenced 2,497,519 203,748 233,509coding 2,180,498 87 41153(67) 20 33,617 (46) 14hits 1,320,216 61 22260 (34) 54 19,099 (25) 57ORFS 860,274 39 18893 (33) 46 14,518 (21) 43Non-coding 317,029 13 162595 80 199,892 86Page 2 of 9(page number not for citation purposes)Numbers in brackets indicates the number of genes or gene fragments in each sample. The percentages refer to the proportion of each sequenced class relative to the total amount sequenced.BMC Genomics 2008, 9:200, gene-rich genome of E. cuniculi (Table 1) reveals asharp contrast in the overall nature of the genomes. Fromthe B. algerae survey, 34 genes with identifiable homo-logues in other organisms were identified, and a further33 potential ORFs greater than 100 codons but with norecognizable similarity to any other gene were found,resulting in a protein-coding content of 11% identifiablecoding sequence and 20% including putative ORFs. In E.aedis, only 25 identifiable protein-coding genes and 21ORFs were found, pushing the range of coding sequencestill lower, to 8% identifiable coding sequence or 14% ifputative ORFs are included (Table 1). In contrast, 52% ofthe E. cuniculi genome consists of protein-codingsequences with recognizable similarity to other genes, andthe proportion of coding sequence is 87% when ORFs areincluded [10]. The gene density of A. locustae is similar tothat of E. cuniculi [12], as are small regions of other micro-sporidian genomes that have been sampled [20]. Overall,the gene-densities of B. algerae and E. aedis are about anorder of magnitude lower than other microsporidia thathave been examined to date.The overall GC content for both B. algerae (24%) and E.aedis (25%) is also significantly lower than that of E.cuniculi (47%). Not surprisingly, the GC content in thecoding regions is slightly higher: 28% for B. algerae and31% for E. aedis. A smaller sequence survey from Spraguealophii with a genome size of 6.2 Mb has revealed a bias of28% [20,26]. There is therefore no obvious correlationbetween genome size and drift towards low GC content inmicrosporidia, and similarly no pronounced lineage-spe-cific bias.Presence of transposable elementsIn contrast to the E. cuniculi genome, which does notencode any selfish genetic elements, fragments of reversetranscriptase or complete retrotransposons have beenreported in the genomes of N. bombycis, V. corneae and S.lophii [20-22], and repeated sequences suggested to bemobile were reported in N. bombycis and N. costelytrae[27]. The V. corneae reverse transcriptase is closely relatedto a human LINE sequence, and both the N. bombycis andS. lophii retrotransposons have sequence similarity to eachother and to Ty3/gypsy retrotransposons.In both B. algerae and E. aedis surveys we found extensiveevidence of numerous transposable elements (Table 2).The E. aedis fragments all share high similarity to Ty3/gypsy retrotransposons from the N. bombycis and S. lophii.Nine of the seventeen fragments of putatively selfish ele-ments identified in B. algerae are also members of thesame family, and once more also share a high degree ofsimilarity to the N. bombycis and S. lophii elements. Theremaining fragments from B. algerae were similar to hypo-thetical proteins resembling transposases.Table 2: List of hits to suspected transposable elements:Top Blast hit annotation Accession no.Brachiola algeraePol polyprotein Nosema bombycis 91176517Pol polyprotein Nosema bombycis 91176521Pol polyprotein Nosema bombycis 91176521Pol polyprotein Nosema bombycis 91176521Pol polyprotein Nosema bombycis 91176521Pol polyprotein Nosema bombycis 91176523Pol polyprotein Nosema bombycis 91176523Pol polyprotein Nosema bombycis 91176525Pol polyprotein Nosema bombycis 91176525Transposase, putative Acaryochloris marina 158337326Predicted protein, Nematostella vectensis 156394155Conserved hypothetical protein Akkermansia muciniphila 166832600Neisseria meningitidis IS1016 transposase 161869234Caenorhabditis briggsae hypothetical protein CBG18017 157775203Caenorhabditis briggsae hypothetical protein CBG18017 157775203Caenorhabditis briggsae Hypothetical protein CBG00277 157771110Caenorhabditis briggsae Hypothetical protein CBG21915 157749299Edhazardia aedisPol polyprotein Nosema bombycis 911765256 different Pol polyproteins Nosema bombycis 91176519Page 3 of 9(page number not for citation purposes)Pol polyprotein Nosema bombycis 91176525BMC Genomics 2008, 9:200 level of similarity between the Ty3/gypsy elementsfrom N. bombycis, S. lophii, E. aedis and B. algerae and thefact that the host groups for these four species are notclosely related, is strongly suggestive that an ancestralfamily of retrotransposons existed in the common ances-tor of these microsporidia. In molecular phylogenies ofmicrosporidia, the true Nosema-group is consistentlyfound to be a sister-lineage of the Encephalitozoon-group tothe exclusion of lineages that include E. aedis, B. algeraeand S. lophii [28-30]. If the Ty3/gypsy retroelements iden-tified here are ancestral to the genomes where they havebeen found, it means that it was also ancestral to E.cuniculi and must have been completely purged from itsgenome. This raises some curious questions about the N.bombycis genome. Here, the Ty3/gypsy retrotransposonswere found to be nested within blocks of compacted genesthat were often conserved in order with homologues fromother microsporidian genomes [22]. Based on this it wassuggested that the elements could have invaded a compactgenome, and perhaps later facilitated some genomic rear-rangements [22]. Reconciling the ancient presence ofthese elements with the nature of the N. bombycis genomeis complicated. It is possible that the ancestral genomecontained many such elements and had a low gene den-sity. This genome could have subsequently compacted inseveral lineages, some of which lost the retroelements aspart of the compaction process (e.g., E. cuniculi), whileothers kept them and compacted the genome aroundthem (e.g., N. bombycis). It is also possible that compac-tion happened in an earlier common ancestor of some ofthese lineages and that certain genomes have 're-expanded'. In either event, the retention of large numbersof selfish elements in an otherwise compact genome is ofinterest, as one might expect that compaction would bestrongly inclined to lead to the loss of non-coding mate-rial such as selfish elements. It serves to illustrate the waycompaction affects different aspects of the genome in dif-ferent lineages, another possible example being the differ-ential loss or retention of introns in relict nucleomorphgenomes [31,32].Gene density, order, and sizeThe small number of genes identified and the large con-tinuous stretches of non-coding sequence in both surveyslead to the obvious conclusion that the gene-density ofthese genomes is much lower than those of E. cuniculi orA. locustae. The average intergenic distances in thesegenomes cannot readily be determined since few havebeen completely sequenced. In B. algerae four clonesencoded two genes and the distances between them are108, 206, 276, and 552 bp. In E. aedis a single cloneencoded two adjacent genes, and the intergenic spacesbetween them is 1,324 bp. At the same time, the largestmeric region next to an SSU gene) and 2,068 bp in B. alge-rae and E. aedis, respectively. The average distancebetween genes in E. cuniculi and A. locustae samples is 129and 211 bp [10,12]. From the existing data it seems likelythat the average distance between genes in B. algerae andE. aedis is much larger than that of either of the well stud-ied microsporidian genomes, and that the density acrossat least the B. algerae genome may be more heterogene-ous.Of the pairs of adjacent genes we identified, one B. algeraepair is also adjacent in both A. locustae and E. cuniculi (Fig-ure 1) (the pair separated by 206 bp in B. algerae). It haspreviously been shown that the order of gene pairs in A.locustae and E. cuniculi is highly conserved, and this hasbeen hypothesized to be related to the compaction of thegenome [12]. The conservation of one of four B. algeraegene pairs suggests that areas of this genome may beunder similar constraints. If the conservation of genomeorder is related to compaction, this also suggests that thecompacted state may have existed in the ancestor of B.algerae, E. cuniculi and A. locustae, which is consistent withArea of conserved gene synteny between three species of microsporidiaFigure 1Area of conserved gene synteny between three spe-cies of microsporidia. A fragment of the B. algerae genome aligned to corresponding regions from E. cunculi and A. locus-tae. The gene order, but not orientation, is conserved. Arrowheads indicate gene orientation and dashed white line indicates incomplete gene sequences. Intergenic space 08_1830 08_184008_184008_183008_1830 08_184031 bp206 bp61 bpE. cuniculiA. locustaeB. algeraePage 4 of 9(page number not for citation purposes)continuous stretches of sequence from which we couldidentify no genes were 2,412 (or 2,943 in a likely subtelo-lengths are indicated.BMC Genomics 2008, 9:200 that suggest some relationship between B.algerae and A. locustae [[25] and unpublished data].In addition to being densely packed, E. cuniculi genes havealso been shown to be shorter on average than homo-logues in the S. cerevisiae genome [10]. This has been dis-cussed in the context of genome compaction, but alsosuggested to be the result of a reduction in the number ofproteins in the cell, which leads to smaller interaction net-works, which in turn allows proteins to reduce theirnumber or complexity of interacting domains [13]. Inyeast, it has been shown that there is a correlationbetween protein size and connectivity, with larger pro-teins displaying a greater number of interactions [33]. Weexamined the only five full-length genes identified in theB. algerae survey with homologues in yeast and found thatall five were shorter than S. cerevisiae homologues, andmore surprisingly most were also shorter than the E.cuniculi homologues (Figure 2). Similarly, from the E.aedis survey, only five full-length genes were found (Figure2), and four of these were shorter than S. cerevisiae homo-logues, and comparable in size with the E. cuniculi homo-logues. The sole E. aedis protein predicted to be larger thanthe yeast counterpart encodes a ribosomal protein (Figure2). Given that these genomes are not compacted, it sug-gests that proteins are either shorter due to a reduced pro-teome complexity, or that they had an ancestor with acompacted genome, or both.Gene content and coding capacityThere is no experimental estimate of the E. aedis genomesize, but the B. algerae genome has been estimated to bebetween 15–20 Mbp by pulsed-field gel electrophoresis[23]. This is much larger than the genomes of either E.cuniculi (2.9 Mbp) or A. locustae (5.4 Mbp). As we show,much of the genome size difference can be attributed tothe significantly lower gene-densities of B. algerae and E.aedis. However, it is still possible that one or both of thesegenomes is also larger because it contains more genes thanE. cuniculi.The E. cuniculi genome clearly under went a massive geneloss relative to other eukaryotes, but this gene loss mayhave been ancestral to microsporidia. If this were the casewe would expect to find few genes in E. aedis and B. algeraethat are present and conserved in other eukaryotes that arenot also present in E. cuniculi. That is to say that the B.algerae and E. aedis genomes would not have more con-served genes than the pool remaining in the ancestralmicrosporidian after this gene loss event.Our sampling of E. aedis and B. algerae genomes showsthat this scenario is quite possible. Of the protein-coding(25 cases), every one is also present in E. cuniculi (Table3). Given the sample size, it is likely that either genomecould contain some genes found in other organisms butnot E. cuniculi, but it is unlikely that they are abundant.This lack of excess conserved gene homologues is of inter-est because it implies that the large-scale gene loss charac-teristic of E. cuniculi took place relatively early inmicrosporidian evolution, in the ancestor of E. cuniculi, A.locustae, E. aedis and B. algerae.However, the B. algerae survey consisted of 20% codingsequence, so taking into account the range or estimatedgenome sizes for B. algerae (15–20 Mbp), this suggestsbetween 2,786 and 3,714 genes in the Brachiola genome(assuming an average gene length of 1,077 as in E.cuniculi). The discrepancy between this predicted codingcapacity of B. algerae and the observation that all the rec-ognizable genes we sampled are shared with E. cuniculicould be explained in many ways. First, our sample maybe biased to gene-encoding regions, and this would leadComparison of microsporidian and yeast protein lengthsFigure 2Comparison of microsporidian and yeast protein lengths. The number of codons for all full-length proteins found within B. algerae (Ba) and E. aedis (Ea) sequence sur-veys compared to homologues from S. cerevisiae (Sc) (or Schizosaccahromyces pombe (Sp) in cases where S. cerevisiae does not encode a homologue).0100200300400500Transcription factor GCN5Methyltransferase40S Ribosomal protein 26S Proteasome B-type subunit U6 snRNA-associatedsmall nucleoprotein RAD31Ubiquitin conjugating enzyme E2-24KDRho GTPaseRRP40Cell division kinaseBaEcScBaEcScBaEcScBaEcScBaEcSpEaEcScEaEcScEaEcScEaEcScEaEcScLength in amino acidsPage 5 of 9(page number not for citation purposes)genes with identifiable homologues in some othergenome that we found in B. algerae (34 cases) and E. aedisto an overestimate of the gene-density. Second, a largenumber of lineage-specific ORFs could skew the estimate,BMC Genomics 2008, 9:200 the proportion of ORFs we found (47% and 43% forB. algerae and E. aedis, respectively) is similar to that foundin E. cuniculi (39%). This issue is also complicated by thefact that we identified several E. cuniculi "ORFs" in oursample, and therefore the proportions of putative ORFs ischanging. Third, the genome size estimates may be wrong.Lastly, it is possible that there are many more than 2,000genes in these organisms, but that the excess is mostly dueto recent duplications. We did not sample any duplicatesin either genome, though we did find areas of repeatsamongst the non-coding areas in both B. algerae and E.aedis. If gene duplications are common, the genome couldEvolution of genome compaction in microsporidiaThough the phylogenetic relationships of major micro-sporidian lineages are not well resolved, phylogenies ofrRNA [28] and concatenated tubulin genes (unpublisheddata) suggest a relationship between B. algerae and A.locustae to the exclusion of E. aedis or E. cuniculi (Figure 3).This raises interesting questions about whether micro-sporidian genomes have compacted more than once dur-ing the diversification of the phylum, or if some have re-expanded from a compacted state.An obvious factor in the dynamics of genome size is trans-Table 3: List of genes identified by BLAST searchBrachiola algerae Edhazardia aedisHit E. cuniculi locus Hit E. cuniculi locusCell Division Kinase 08_0230 14-3-3 Protein 1 03_1010Coatomer coat delta 08_0340 26S proteasome beta-type subunit 05_0290DNA directed RNA pol 01_0600 40S ribosomal protein S28 09_1275DNA ligase 02_1220 60S ribosomal protein L8 01_0310DNA mismatch repair 11_1260 Aldose Reductase 01_0970Dnm1 01_1210 DNA Mismatch Repair Protein 05_0300E. cuniculi hypothetical protein 01_0390 Endochitinase 09_1320E. cuniculi hypothetical protein 04_0270 E. cuniculi hypothetical protein 03_0870E. cuniculi hypothetical protein 05_1460 E. cuniculi hypothetical protein 05_1000E. cuniculi hypothetical protein 06_0970 E. cuniculi hypothetical protein 05_1080E. cuniculi hypothetical protein 07_0810 E. cuniculi hypothetical protein 09_1690E. cuniculi hypothetical protein 08_1830 Myosin heavy chain 09_1970E. cuniculi hypothetical protein 08_1840 NIFS-like protein 11_1770E. cuniculi hypothetical protein 09_0300 Phospholipid-transporting ATPase 09_1440E. cuniculi hypothetical protein 09_1240 Putative methyltransferase 05_0950E. cuniculi hypothetical protein 10_1360 RING-finger-containing ubiquitin ligase 07_0330E. cuniculi hypothetical protein 11_0260 Similarity to oxidoreductase 11_1070GPI Anchor Biosynthesis 09_1210 SSU geneHsp70 02_0100 Topoisomerase 1 06_1520Isopentyl pyrophosphate δ isomerase 02_0230 TPR domain hypothetical protein 09_1180LSU gene Transcriptional activator 10_1430Pelota protein 03_1380 Translation initiation factor IF-2P 09_0070Phenylalanine tRNA synthase 07_1660 Trehalose-6-phosphate synthase 01_0870RAD31 DNA damage tolerance 08_0460 U6 snRNA-associated small RNP 05_1310RAS-like GTP binding protein 10_0350 UTP glucose-1-phosphate uridyltransferase 03_0280Septin 09_0820 Vacuolar protein sorting-associated protein 03_0900SER/THR protein kinase 08_1620Signal Recognition Particle 04_0980SSU geneSyntaxin 05_0820TFIID 111 KDa 01_0760TFIID 72/90 KDa 11_1750TFIID 150 kDa 09_0090TFIID I 04_1440U5 Associated snRNP 11_0870Ubiquitin Conjugating Enzyme E2 08_0860Page 6 of 9(page number not for citation purposes)contain more genes without an increased complexity inthe proteome.posable elements. One could imagine a genome expand-ing due to the invasion of such elements, and indeedBMC Genomics 2008, 9:200 such elements have been found in N. bombycis [22]and now B. algerae and E. aedis. However, the majority ofthese elements are closely related members of the Ty3/gypsy family. Therefore genome expansion cannot beentirely due to an invasion since the elements must haveexisted in the common ancestor and been purged from E.cuniculi and possibly other compacted microsporidiangenomes. It also remains to be seen if these genomes aresubstantially heterogeneous. It is possible that manygenes do exist in relatively compact regions while otherregions are dominated by non-coding sequence. Anextreme version of such a situation is seen in the small andcompacted genome of the picoplankton Ostreococcus tauri.Here most chromosomes in the genome show a high genedensity but 2 chromosomes out of 20 contain 77% of thetransposons identified in the genome [34].A second factor that has been hypothesized to affectgenome size is cell size. A correlation between genomesize and cell size has been observed in eukaryotes gener-ally [35] and microsporidia specifically [36]. However inthe microsporidia, variation in cell size in the different lifestages can confound correlations between genome sizelocustae is reported to be 5.4 Mb [37], whereas the sporesof both B. algerae and A. locustae are of comparable sizes[38,39] suggesting, in this case, that cell size is not neces-sarily a factor. A further consideration is whether the com-plexity of the life cycles of different microsporidia isreflected in genome size and gene number. Both A. locus-tae and E. cuniculi have a simple life cycle with monomor-phic spores and are restricted to a narrow host range.Edhazardia aedis has a more complex life cycle with multi-ple spore types and must be adapted to both the larval andadult stage of the mosquito. Brachiola algerae, known tohave a larger genome, has a simple life cycle with mono-morphic spores, but has a broad host range and can infectboth mammals and insects.ConclusionThe E. cuniculi genome is a model for compacted nucleargenomes, but potentially not a good model for micro-sporidian genomes generally. We have shown that thegenomes of B. algerae and E. aedis are structured very dif-ferently: they have large proportions of non-codingsequence and many transposable elements, resulting in avery low and perhaps variable gene-density comparedwith E. cuniculi. The sample of identifiable genes found inthe surveys, and the proportions of these genes sharedwith E. cuniculi both suggest that the complexity of theproteome is not the major factor contributing to genomesize variation. The phylogeny of microsporidia suggestsmultiple events of compaction and/or expansion, whichraises interesting questions about what forces thegenomes to compact so severely, and why such a forcewould then cease to operate on the genome.MethodsMicrosporidia, genomic DNA extraction, and genomic library construction6.1 × 107 uninucleate Edhazardia aedis spores harvestedfrom Aedes aegypti larvae were ruptured by glass bead-beating, and spores were examined for breakage via lightmicroscopy. E. aedis genomic DNA was purified by thestandard phenol-chloroform method and served as tem-plate for whole genome rolling-circle amplification usingGenomiphi (Amersham). 4.5 μg of amplified E. aedisgenomic DNA was sheared, blunt end-repaired, andcloned into pCR4Blunt-TOPO (Invitrogen) according tothe manufacturer's specifications. 182 different E. aedisclones with an average length of 1,283 bp were end-sequenced using ABI Big Dye 3.1 chemistry. Six differentE. aedis library clones containing coding, non-coding, ortransposable segments were checked for chimericsequence by PCR of non-Genomiphi-treated E. aedisgenomic DNA. From this, successful amplification of frag-ments between 250 and 450 bp did not support the ideaSchematic consensus of microsporidian phylogenetic rela-tionshipsFigure 3Schematic consensus of microsporidian phylogenetic relationships. Microsporidian relationships from a consen-sus of published SSU phylogenies [28, 30, 40] and concate-nated tubulin genes (unpublished data). Genome sizes are labeled and the reported presence of Gypsy/Ty transposons is indicated by (Ty).Antonospora locustae 5.4Edhazardia aedis ? TyEncephalitozoon cuniculi 2.9Encephalitozoon intestinalis 2.3Encephalitozoon hellem 2.5Nosema bombycis 15.3 TyNosema pyrausta 10.2Vairimorpha sp. 10.5Nosema furnacalis 10.2Glugea atherinae 19.5 TyVavraia oncoperae (ex. porina)  8.0Vavraia oncoperae (ex grass grub) 10.2Spraguea lophii 6.2 TyBrachiola algerae 15-20 TyTy?Page 7 of 9(page number not for citation purposes)and cell size. As a rough correlation though, the genomesize of B. algerae is estimated at 15–20 Mb and that of A.of chimeras being present in the Genomiphi-created E.aedis genomic library.BMC Genomics 2008, 9:200 was extracted from 2 × 107germinated Brachiola alge-rae spores (a strain originally isolated the mosquitoAnopheles stephensi). Spores were cultivated in vitro inRK13 rabbit kidney cells at 30C in 5% CO2, purified andgerminated by incubation in 0.3% H2O2 for 16 hours at25°C. The germinated spores were concentrated by cen-trifugation suspended in 100 μg/ml Proteinase K-PBS andincubated for 15 min at 65°C. DNA was extracted usingphenol-chloroform, followed by ethanol precipitation.DNA was then dissolved in TE buffer and stored at -70°Cuntil use. A sample of purified Brachiola algerae DNA wasamplified using Genomphi (Amersham) to produce 10 μgof DNA and another sample of 1.7 μg of purified DNAwas processed directly to make two separate libraries.DNA was sheared, blunt-ended and cloned as describedfor E. aedis above. A respective 64,433 and 140,051 basesfrom the Genomiphi and non-Genomiphi treated librar-ies were sequenced with ABI Big Dye 3.1. This gave a totalof 203,748 non-overlapping bases of sequence in 181contigs with an average length of 1,125 bp.Areas of six representative B. algerae clones from theGenomiphi-amplified DNA library were reamplified byPCR from genomic DNA to confirm that the Genomiphiprocess had not amplified chimeric sequences. Primerswere designed to areas of 6 clones. These fragments werebetween 525 and 1000 base pairs and included non-cod-ing areas, putative transposases, transposons, protein-coding genes, and an SSU gene area.Contigs were analysed by BlastX and BlastN to sequencesin GenBank. Open reading frames were considered signif-icantly similar if E values were less than 0.00001. Contigswere further searched for stretches of nucleotides codingfor sequences of at least 100 amino acids, and these wereconsidered ORFs.New sequences were deposited in GenBank under acces-sion numbers ET437577–ET437812 and ET437979–ET437981 (E. aedis) and ET223031–ET223211 (B. alge-rae).Authors' contributionsBAPW and RCHL constructed libraries, sequenced clones,analysed data and drafted the paper, JJB and LMW con-tributed microsporidian material and to the writing of thepaper. NMF and PJK conceived of the study, analysed dataand drafted the paper.AcknowledgementsThis work was supported by a grant from the Canadian Institutes for Health Research to PJK (MOP-42517) and a grant from the Natural Sciences and Engineering Research Council of Canada to NMF. PJK is a Fellow of the Canadian Institute for Advanced Research and senior investigator awards References1. Didier ES: Microsporidiosis: an emerging and opportunisticinfection in humans and animals.  Acta Trop 2005, 94:61-76.2. Vossbrinck CR, Maddox JV, Friedman S, Debrunner-Vossbrinck BA,Woese CR: Ribosomal RNA sequence suggests microsporidiaare extremely ancient eukaryotes.  Nature 1987, 326:411-414.3. Vossbrinck CR, Woese CR: Eukaryotic ribosomes that lack a5.8S RNA.  Nature 1986, 320:287-288.4. Hirt RP, Logsdon JM Jr., Healy B, Dorey MW, Doolittle WF, EmbleyTM: Microsporidia are related to Fungi: evidence from thelargest subunit of RNA polymerase II and other proteins.Proc Natl Acad Sci USA 1999, 96:580-585.5. Edlind TD, Li J, Visversvara GS, Vodkin MH, McLaughlin GL, KatiyarSK: Phylogenetic analysis of the b-tubulin sequences fromamitochondriate protozoa.  Mol Phylogenet Evol 1996, 5:359-367.6. Keeling PJ, Luker MA, Palmer JD: Evidence from beta-tubulinphylogeny that microsporidia evolved from within the fungi.Mol Biol Evol 2000, 17:23-31.7. Van de Peer Y, Ben Ali A, Meyer A: Microsporidia: accumulatingmolecular evidence that a group of amitochondriate and sus-pectedly primitive eukaryotes are just curious fungi.  Gene2000, 246:1-8.8. Gill EE, Fast NM: Assessing the microsporidia-fungi relation-ship: Combined phylogenetic analysis of eight genes.  Gene2006, 375:103-109.9. Peyretaillade E, Biderre C, Peyret P, Duffieux F, Méténier G, Gouy M,Michot B, Vivarès CP: Microsporidian Encephalitozoon cuniculi,a unicellular eukaryote with an unusual chromosomal dis-persion of ribosomal genes and a LSU rRNA reduced to theuniversal core.  Nucleic Acids Res 1998, 26:3513-3520.10. Katinka MD, Duprat S, Cornillot E, Méténier G, Thomarat F, PrensierG, Barbe V, Peyretaillade E, Brottier P, Wincker P, Delbac F, El AlaouiH, Peyret P, Saurin W, Gouy M, Weissenbach J, Vivarès CP: Genomesequence and gene compaction of the eukaryote parasiteEncephalitozoon cuniculi.  Nature 2001, 414:450-453.11.  [].12. Slamovits CH, Fast NM, Law JS, Keeling PJ: Genome compactionand stability in microsporidian intracellular parasites.  CurrBiol 2004, 14:891-896.13. Zhang J: Protein-length distributions for the three domains oflife.  Trends Genet 2000, 16:107-109.14. Polonais V, Prensier G, Méténier G, Vivarès CP, Delbac F: Micro-sporidian polar tube proteins: highly divergent but closelylinked genes encode PTP1 and PTP2 in members of the evo-lutionarily distant Antonospora and Encephalitozoongroups.  Fung Genet Biol 2005, 42:791-803.15. Slamovits CH, Burri L, Keeling PJ: Characterization of a diver-gent Sec61beta gene in microsporidia.  J Mol Biol 2006,359:1196-1202.16. Corradi N, Akiyoshi DE, Morrison HG, Feng X, Weiss LM, Tzipori S,Keeling PJ: Patterns of genome evolution among the micro-sporidian parasites Encephalitozoon cuniculi, Antonosporalocustae and Enterocytozoon bieneusi.  PLoS ONE 2007,2:e1277.17. Williams BA, Slamovits CH, Patron NJ, Fast NM, Keeling PJ: A highfrequency of overlapping gene expression in compactedeukaryotic genomes.  Proc Natl Acad Sci USA 2005,102:10936-10941.18. Corradi N, Gangaeva A, Keeling PJ: Comparative profiling ofoverlapping transcription in the compacted genomes ofmicrosporidia Antonospora locustae and Encephalitozooncuniculi.  Genomics 2008 in press.19. Bennetzen JL: Mechanisms and rates of genome expansion andcontraction in flowering plants.  Genetica 2002, 115:29-36.20. Hinkle G, Morrison HG, Sogin ML: Genes coding for reversetranscriptase, DNA-directed RNA polymerase, and chitinsynthase from the microsporidian Spraguea lophii.  Biol Bull1997, 193:250-251.21. Mittleider D, Green LC, Mann VH, Michael SF, Didier ES, Brindley PJ:Sequence survey of the genome of the opportunistic micro-sporidian pathogen, Vittaforma corneae.  J Eukaryot Microbiol2002, 49:393-401.22. Xu J, Pan G, Fang L, Li J, Tian X, Li T, Zhou Z, Xiang Z: The varyingPage 8 of 9(page number not for citation purposes)from the Michael Smith Foundation for Health Research.microsporidian genome: existence of long-terminal repeatretrotransposon in domesticated silkworm parasite Nosemabombycis.  Int J Parasitol 2006, 36:1049-1056.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central BMC Genomics 2008, 9:200 Dia N, Lavie L, Méténier G, Toguebaye BS, Vivarès CP, Cornillot E:InterB multigenic family, a gene repertoire associated withsubterminal chromosome regions of Encephalitozooncuniculi and conserved in several human-infecting micro-sporidian species.  Curr Genet 2007, 51:171-186.24. Franzen C, Nassonova ES, Scholmerich J, Issi IV: Transfer of themembers of the genus Brachiola (microsporidia) to thegenus Anncaliia based on ultrastructural and moleculardata.  J Eukaryot Microbiol 2006, 53:26-35.25. Becnel JJ, White SE, Shapiro AM: Review of microsporidia-mos-quito relationships: from the simple to the complex.  Folia Par-asitol (Praha) 2005, 52:41-50.26. Biderre C, Pagès M, Méténier G, David D, Bata J, Prensier G, VivarèsCP: On small genomes in eukaryotic organisms: molecularkaryotypes of two microsporidian species (Protozoa) para-sites of vertebrates.  C R Acad Sci III 1994, 317:399-404.27. Malone LA, McIvor CA: DNA probes for two Microsporidia,Nosema bombycis and Nosema costelytrae.  J Invertebr Pathol1995, 65:269-273.28. Vossbrinck CR, Debrunner-Vossbrinck BA: Molecular phylogenyof the Microsporidia: ecological, ultrastructural and taxo-nomic considerations.  Folia Parasitol (Praha) 2005, 52:131-142.29. Muller A, Trammer T, Chioralia G, Seitz HM, Diehl V, Franzen C:Ribosomal RNA of Nosema algerae and phylogenetic rela-tionship to other microsporidia.  Parasitol Res 2000, 86:18-23.30. Slamovits CH, Williams BA, Keeling PJ: Transfer of Nosema locus-tae (Microsporidia) to Antonospora locustae n. comb. basedon molecular and ultrastructural data.  J Eukaryot Microbiol 2004,51:207-213.31. Gilson PR, Su V, Slamovits CH, Reith ME, Keeling PJ, McFadden GI:Complete nucleotide sequence of the chlorarachniophytenucleomorph: nature's smallest nucleus.  Proc Natl Acad Sci USA2006, 103:9566-9571.32. Douglas S, Zauner S, Fraunholz M, Beaton M, Penny S, Deng LT, WuX, Reith M, Cavalier-Smith T, Maier UG: The highly reducedgenome of an enslaved algal nucleus.  Nature 2001,410:1091-1016.33. Warringer J, Blomberg A: Evolutionary constraints on yeastprotein size.  BMC Evol Biol 2006, 6:61.34. Derelle E, Ferraz C, Rombauts S, Rouze P, Worden AZ, Robbens S,Partensky F, Degroeve S, Echeynie S, Cooke R, Saeys Y, Wuyts J, Jab-bari K, Bowler C, Panaud O, Piegu B, Ball SG, Ral JP, Bouget FY, Piga-neau G, De Baets B, Picard A, Delseny M, Demaille J, Van de Peer Y,Moreau H: Genome analysis of the smallest free-living eukary-ote Ostreococcus tauri unveils many unique features.  ProcNatl Acad Sci USA 2006, 103:11647-11652.35. Cavalier-Smith T: The Evolution of Genome Size.  , John Wileyand Sons, Chichester; 1985. 36. Méténier G, Vivarès CP: Genomics of microbial parasites: Themicrosporidial paradigm.  In Organelles, Genomes and EukaryotePhylogeny Edited by: Hirt RP and Horner DS. London, CRC Press;2004:207-236. 37. Streett DA: Analysis of Nosema locustae (Microsporidia:Nosematidae) chromosomal DNA with pulsed-field gel elec-trophoresis.  J Invert Pathol 1994, 63:301-303.38. Chioralia G, Trammer T, Maier WA, Seitz HM: Morphologicchanges in Nosema algerae (Microspora) during extrusion.Parasitol Res 1998, 84:123-131.39. Sokolova YY, Lange CE: An ultrastructural study of Nosemalocustae Canning (Microsporidia) from three species ofAcrididae (Orthoptera).  Acta Protozool 2002, 41:229-237.40. Terry RS, Smith JE, Sharpe RG, Rigaud T, Littlewood DT, Ironside JE,Rollinson D, Bouchon D, MacNeil C, Dick JT, Dunn AM: Wide-spread vertical transmission and associated host sex-ratiodistortion within the eukaryotic phylum Microspora.  Proc BiolSci 2004, 271:1783-1789.yours — you keep the copyrightSubmit your manuscript here: 9 of 9(page number not for citation purposes)


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