Comparative genomic analysis of three Leishmania species that cause diverse human disease.

Leishmania parasites cause a broad spectrum of clinical disease. Here we report the sequencing of the genomes of two species of Leishmania: Leishmania infantum and Leishmania braziliensis. The comparison of these sequences with the published genome of Leishmania major reveals marked conservation of synteny and identifies only approximately 200 genes with a differential distribution between the three species. L. braziliensis, contrary to Leishmania species examined so far, possesses components of a putative RNA-mediated interference pathway, telomere-associated transposable elements and spliced leader-associated SLACS retrotransposons. We show that pseudogene formation and gene loss are the principal forces shaping the different genomes. Genes that are differentially distributed between the species encode proteins implicated in host-pathogen interactions and parasite survival in the macrophage.

Comparative genomics of the fungal pathogens Candida dubliniensis and Candida albicans.

Candida dubliniensis is the closest known relative of Candida albicans, the most pathogenic yeast species in humans. However, despite both species sharing many phenotypic characteristics, including the ability to form true hyphae, C. dubliniensis is a significantly less virulent and less versatile pathogen. Therefore, to identify C. albicans-specific genes that may be responsible for an increased capacity to cause disease, we have sequenced the C. dubliniensis genome and compared it with the known C. albicans genome sequence. Although the two genome sequences are highly similar and synteny is conserved throughout, 168 species-specific genes are identified, including some encoding known hyphal-specific virulence factors, such as the aspartyl proteinases Sap4 and Sap5 and the proposed invasin Als3. Among the 115 pseudogenes confirmed in C. dubliniensis are orthologs of several filamentous growth regulator (FGR) genes that also have suspected roles in pathogenesis. However, the principal differences in genomic repertoire concern expansion of the TLO gene family of putative transcription factors and the IFA family of putative transmembrane proteins in C. albicans, which represent novel candidate virulence-associated factors. The results suggest that the recent evolutionary histories of C. albicans and C. dubliniensis are quite different. While gene families instrumental in pathogenesis have been elaborated in C. albicans, C. dubliniensis has lost genomic capacity and key pathogenic functions. This could explain why C. albicans is a more potent pathogen in humans than C. dubliniensis.

Artemis and ACT: viewing, annotating and comparing sequences stored in a relational database.

MOTIVATION: Artemis and Artemis Comparison Tool (ACT) have become mainstream tools for viewing and annotating sequence data, particularly for microbial genomes. Since its first release, Artemis has been continuously developed and supported with additional functionality for editing and analysing sequences based on feedback from an active user community of laboratory biologists and professional annotators. Nevertheless, its utility has been somewhat restricted by its limitation to reading and writing from flat files. Therefore, a new version of Artemis has been developed, which reads from and writes to a relational database schema, and allows users to annotate more complex, often large and fragmented, genome sequences. RESULTS: Artemis and ACT have now been extended to read and write directly to the Generic Model Organism Database (GMOD, http://www.gmod.org) Chado relational database schema. In addition, a Gene Builder tool has been developed to provide structured forms and tables to edit coordinates of gene models and edit functional annotation, based on standard ontologies, controlled vocabularies and free text. AVAILABILITY: Artemis and ACT are freely available (under a GPL licence) for download (for MacOSX, UNIX and Windows) at the Wellcome Trust Sanger Institute web sites: http://www.sanger.ac.uk/Software/Artemis/ http://www.sanger.ac.uk/Software/ACT/

Gene arrays at Pneumocystis carinii telomeres.

In the fungus Pneumocystis carinii, at least three gene families (PRT1, MSR, and MSG) have the potential to generate high-frequency antigenic variation, which is likely to be a strategy by which this parasitic fungus is able to prolong its survival in the rat lung. Members of these gene families are clustered at chromosome termini, a location that fosters recombination, which has been implicated in selective expression of MSG genes. To gain insight into the architecture, evolution, and regulation of these gene clusters, six telomeric segments of the genome were sequenced. Each of the segments began with one or more unique genes, after which were members of different gene families, arranged in a head-to-tail array. The three-gene repeat PRT1-MSR-MSG was common, suggesting that duplications of these repeats have contributed to expansion of all three families. However, members of a gene family in an array were no more similar to one another than to members in other arrays, indicating rapid divergence after duplication. The intergenic spacers were more conserved than the genes and contained sequence motifs also present in subtelomeres, which in other species have been implicated in gene expression and recombination. Long mononucleotide tracts were present in some MSR genes. These unstable sequences can be expected to suffer frequent frameshift mutations, providing P. carinii with another mechanism to generate antigen variation.

Rapid evolution of virulence and drug resistance in the emerging zoonotic pathogen Streptococcus suis.

BACKGROUND: Streptococcus suis is a zoonotic pathogen that infects pigs and can occasionally cause serious infections in humans. S. suis infections occur sporadically in human Europe and North America, but a recent major outbreak has been described in China with high levels of mortality. The mechanisms of S. suis pathogenesis in humans and pigs are poorly understood. METHODOLOGY/PRINCIPAL FINDINGS: The sequencing of whole genomes of S. suis isolates provides opportunities to investigate the genetic basis of infection. Here we describe whole genome sequences of three S. suis strains from the same lineage: one from European pigs, and two from human cases from China and Vietnam. Comparative genomic analysis was used to investigate the variability of these strains. S. suis is phylogenetically distinct from other Streptococcus species for which genome sequences are currently available. Accordingly, approximately 40% of the approximately 2 Mb genome is unique in comparison to other Streptococcus species. Finer genomic comparisons within the species showed a high level of sequence conservation; virtually all of the genome is common to the S. suis strains. The only exceptions are three approximately 90 kb regions, present in the two isolates from humans, composed of integrative conjugative elements and transposons. Carried in these regions are coding sequences associated with drug resistance. In addition, small-scale sequence variation has generated pseudogenes in putative virulence and colonization factors. CONCLUSIONS/SIGNIFICANCE: The genomic inventories of genetically related S. suis strains, isolated from distinct hosts and diseases, exhibit high levels of conservation. However, the genomes provide evidence that horizontal gene transfer has contributed to the evolution of drug resistance.

ACT: the Artemis Comparison Tool.

The Artemis Comparison Tool (ACT) allows an interactive visualisation of comparisons between complete genome sequences and associated annotations. The comparison data can be generated with several different programs; BLASTN, TBLASTX or Mummer comparisons between genomic DNA sequences, or orthologue tables generated by reciprocal FASTA comparison between protein sets. It is possible to identify regions of similarity, insertions and rearrangements at any level from the whole genome to base-pair differences. ACT uses Artemis components to display the sequences and so inherits powerful searching and analysis tools. ACT is part of the Artemis distribution and is similarly open source, written in Java and can run on any Java enabled platform, including UNIX, Macintosh and Windows.

Characterisation of transcripts from the human cytomegalovirus genes TRL7, UL20a, UL36, UL65, UL94, US3 and US34.

The genome of human cytomegalovirus (HCMV) has been studied extensively in some regions, but not others. In this study, transcripts of the genome were further characterised for open reading frames (ORFs) TRL7, UL36, UL65, UL94, US3 and US34, and for the previously unrecognised ORF, UL20a. Reverse transcription-PCR demonstrated the presence of spliced transcripts from the putative glycoprotein gene, UL20a, at early and late times post-infection. US3 full-length and spliced transcripts, including a previously unidentified transcript (US3ii), were described at immediate early times. Sequencing of the complete ORFs of UL20a and US3 from 21 clinical isolates showed that US3 is well conserved in all isolates (97-100% identity), whereas UL20a shows more variation at the nucleotide level, with 90-100% identity. The limits of transcription, and splice donor and acceptor sequences for UL20a and US3 were conserved in all isolates, indicating likely conservation of mRNA splicing patterns. Sequencing a late cDNA library identified the limits of transcription for ORFs TRL7, UL94 and US34 and transcription from the TRL7 ORF was confirmed by northern blotting. Transcripts were found that were congruent with UL36 and UL65, but these differed in the limits previously predicted for these ORFs. These findings show the variation between predicted and actual transcription and indicate the complex nature of transcription from HCMV ORFs.

The mei3 region of the Schizosaccharomyces pombe genome.

Expression of the mei3 gene is sufficient to induce meiosis in the fission yeast Schizosaccharomyces pombe. The mei3 gene is located 0.64 Mb from the telomere of the left arm of Sz. pombe chromosome II. We have sequenced and analysed 107 kb of DNA from the mei3 genomic region. The sequence includes 14 known genes (bag1-B, csh3, dps1, gpt1, mei3, mfm3, pac1, prp31, rpl38-1, rpn3, rti1, spa1, spm1 and ubc4) and 26 other open reading frames (ORFs) longer than 100 codons: a density of one protein-coding gene per 2.7 kb. Twenty-one of the 40 ORFs (53%) have introns. In addition there is one lone Tf1 transposon long terminal repeat (LTR), tRNA(Trp) and tRNA(Ser) genes and a 5S rRNA gene. 14 of the novel ORFs show sequence similarities which suggest functions of their products, including a coatomer alpha-subunit, a catechol O-methyltransferase, protein kinase, asparagine synthetase, zinc metalloprotease, acetyltransferase, phosphatidylinositol 4-kinase, inositol polyphosphate phosphatase, GTPase-activating protein, permease, pre-mRNA splicing factor, 20S proteasome component and a thioredoxin-like protein. One predicted protein has similarity to the human Cockayne syndrome protein CSA and one with human GTPase XPA binding protein XAB1. Three ORFs are likely to code for proteins because they have sequence similarity with hypothetical proteins, three encode predicted coiled-coil proteins and four are sequence orphans.