About the Project

   We selected for sequencing the Legionella pneumophila subsp. pneumophila (serogroup 1), Philadelphia 1 strain, derived from the original 1976 isolate (D.W. Fraser et al., J. Med. 297, 1189 (1977)). The sequence was assembled using a combined approach. Approximately 8-fold coverage was obtained from whole-genome shotgun sequences and 5-fold coverage from BAC clone-based shotgun sequences. A detailed description of the methods is available at Science online.

The complete circular Legionella pneumophila genome consists of 3,397,754 base pairs (bp).Additional global details are available by clicking on “current statistics”.  A 45 kbp portion of the sequence containing the lvr/lvh type IV secretion system gene region and genes potentially involved in recombination, possesses a higher than average GC content, implying that it was acquired by lateral transfer.  In addition to its maintenance as part of the chromosome, this region exists also an episomal plasmid-like element (pLP45).

We have predicted approximately 3000 genes, belonging to 181 pathways as assessed by PathoLogic (link will be available shortly). Of the genes with predicted functions, the largest percentages appear to be involved in metabolic processes, information transfer or transport.  Nonetheless, there are large numbers of putative proteins with regulatory, protective, signal transduction and virulence functions (a bar graph showing these percentages is available as Figure S1 at Science online), which also provides additional details on the gene complement, along with tables of previously known and new candidate virulence factors.

The genome has a substantial number of expanded gene families, mainly transporters.  A complete list of gene family expansions is available via the “genome annotation” button.  Comparative analysis with completely sequenced genomes from other g-proteobacteria and with other intracellular pathogens, suggests that gene repertoires may be determined as much by bacterial lifestyle as by evolutionary relatedness.  The most closely related completely sequenced genome is that of Coxiella burnetii, a highly infectious pathogen that causes the disease Q fever and a potential bioterror agent.  Like Legionella, it is taken up by lung macrophages, but unlike Legionella which at least temporarily evades lysosomes, it survives and replicates within this acidic organelle, perhaps aided by its apparent possession of a high pI protein complement (R. Seshadri et al., Proc. Natl. Acad. Sci. U.S.A. 100, 5455 (2003)) .  A high percentage of Legionella genes are found as orthologs in Coxiella, and a few genes such as EnhA, EnhB, and the icm/dot complex, seem to be unique to these two bacteria.

The presence of a large repertoire, including expanded families, of genes encoding drug and heavy metal effluxers, may explain both the ability of the bacteria to survive within protozoa that are known to concentrate heavy metals, and to some extent its ability to survive in biocide-treated plumbing systems.  With further studies of the genes and proteins of L. pneumophila and other Legionellae, some may eventually emerge as targets for drug or vaccine therapy.