The beautifully simple but intriguingly complex world of small genomes.

In his review of the Small Genome Meeting in 1997, Eugene Koonin remarked that we have quite a bit of ‘‘... ignorance about the simple cell (Escherichia coli) that has been the primary object of molecular biology for several decades.’’ This message was borne out again this year at the recent conference on Microbial Genomes II: Sequencing, Functional Characterization and Comparative Genomics (The Institute for Genomic Research Genomic Series, Hilton Head, SC, January 31–February 3 1998), where experienced sequencers and microbiologists continued to exchange information about the latest disoveries in genomics and comparative genomics, that is, polygenomics. As Claire Fraser cochair of the meeting, pointed out in her introductory remarks, the emphasis of this year’s meeting was to garner new biological insights from these genomic sequences. With the sequence of seven genomes completed in 1997 and >50 small genome sequencing projects under way worldwide, it is anticipated that at least 25–30 genomes will be completed this year. The small genome sequencing projects reported at the plenary sessions are summarized as supplementary material at www.genome. org or can be obtained at either http:// www.mcs.anl.gov/home/gaasterl/ genomes.html or http://www.tigr.org/ tdb/mdb/mdb.html. With the availability of the relatively large number of sequences available and with many more anticipated, comparative genomics now is a fait accompli. This was exemplified by Fred Blattner’s report comparing sample sequences from the bacterial pathogens E. coli O157:H7, E. coli CFT073, and Yersinia pestis to the completed genomic sequence of E. coli K-12 in an attempt to identify unique regions that might be responsible for pathogenesis. In E. coli O157:H7, 300 unique regions were identified that possibly could define a group of virulence genes (a pathosphere) that interact with the host; 10% are large pathogenicity islands and 90% are smaller segments. Claire Fraser reported a comparative analysis of the single circular spirochetes Treponema pallidum chomosome with the single linear Borrellia burgdorferi chromosome and its 9 circular plasmids and 12 linear plasmids. There was very little sequence similarity between each organism’s individual genes, but both organisms have evolved seemingly independent regions coding for similar function. Like the mycoplasmas, both lack the biosynthetic pathways for amino acid biosynthesis, the TCA cycle, and the electron transport system. They also represent a class of minimal genomes; nearly 50% of the genes are unique, having no biological function yet defined. In his analysis of the B. burgdorferi plasmids, Sherwood Casjens observed a large number (65%) of plasmid-borne paralogous gene families, short repeat tracts, and 12 kb of a 63-bp repeat and rhetorically questioned whether linear plasmids should be dealt with as chromosomes or plasmids. Steven Norris presented the novel aspects of the T. pallidum genome, which evades host defenses by having very few surface proteins. Of note, this organism contains no genes resembling toxins, but it has genes for many surface proteins that could be possible virulence genes. Consistent with the genes of other organisms, almost one-quarter of B. bergdorferi genes are not similar to any known genes and thus have an unknown function. In reporting the sequence of a third mycoplasma genome, Ureaplasma urealyticum, John Glass noted that it is quite different than the two apparently similar Mycoplasma genitalium and Mycoplasma pneumoniae genomes. Not only does the Ureaplasma have a large number of genes without orthologs in either of the other two Mycoplasma species and vice versa, but it also has a significantly rearranged genomic organization, something that is being seen in many supposedly related microorganisms. Just when we thought we understood a biological pathway, such as the aminoacylation of tRNA, Dieter Soll presented results that refute the dogma that aminoacyl-tRNA synthetases are highly conserved and that each amino acid has its own respective synthetase and its cognate tRNA. Comparative analysis of several microbial genomes now reveals the absence of some of the tRNA synthetase homologs. In addition, as new genomic sequence data become available, new biochemical pathways and related enzymes also will begin to emerge, such as a subsidiary pathway (GntII) in Liodinate catabolism reported and experimentally confirmed in E. coli by Tyrrell Conway. The revelation of an additional pathway even in this well-studied model organism suggests that there will be many other surprises in the unique genes yet to be uncovered in additional microorganisms. In phylogenetic and evolutionary studies, Robert Feldman provided evolutionary placement of genes from Aquifex aeolicus, a marine hyperthermophile, with other organisms by grouping them based on their cellular role. Evidence presented suggested that gene trees and species trees do not necessarily match, as species trees can have their own phylogenetic characteristics while gene divergence usually precedes species divergence. John Reeve discussed the phylogenetics of methanogenesis, comparing the methane genes from the two now complete genomes of Methanococcus jannaschii and Methanobacterium thermoautotrophicum. Because both organisms have a common ancestry their methaCorresponding author. E-MAIL [email protected]; FAX (405) 325-4762. Insight/Outlook