Links to databases

coliBASE Schema

A simplified representation of the coliBASE relational database schema is shown below. In the real database the Alignment table is separated into individual tables for each pairwise alignment, to minimise the database access time. Similarly, data from the cds and swissprot tables are consolidated into an additional annotation table to improve the response speed of the initial search page. The schema is generic, and can be applied to other bacterial species (as has already been done for Campylobacter and Clostridium).

Type III Secretion System genes in E. coli K12

Previous authors have noted the existence of K12 homologues of genes from the Spi-1 and Spi-3 pathogenicity islands from Salmonella enterica, and in particular remarked on the presence in this commensal strain of genes usually associated with type III secretion (Hueck, 1998; Blanc-Potard et al., 1999). Why these genes occur in K12 remains a mystery until genomic comparisons with several pathogenic strains are performed.

One of the novel features revealed by the complete genome sequencing of enterohaemorrhagic E. coli O157:H7 (Hayashi et al., 2001; Perna et al., 2001) was the presence of a second type three secretion system (TTSS), which was named ETT2 (E. coli Type Three secretion system 2). ETT2 is present in O-island 122, a region of the O157 genome that appears to be an insertion relative to the K12 genome. O-island 122 and its flanking regions from the E. coli O157:H7 Sakai strain are shown below, aligned to the equivalent regions from E. coli K12 MG1655 (this image was generated using the coliBASE Alignment Viewer). Inspection of the genes flanking O-island 122 shows the presence of genes that, although present in both O157 and K12, are homologues of known TTSS genes. For example the gene ygeH (ECs3709) is a homologue of the Salmonella typhimurium TTSS regulator gene hilA, and is annotated as encoding a putative invasion protein.

The observation of K12 homologues of type III secretion system genes upstream of the position of O-island 122 suggests that a remnant of ETT2 is present in the K12 genome, a hypothesis supported by further comparison with the uropathogenic E. coli CFT073 genome (Welch et al., 2002). This comparison (shown below) reveals that ygeH is present within an insertion relative to CFT073.

The most parsimonious explanation for these observations is that a complete ETT2 system was inserted into a common ancestor of O157 and K12 at some point after their divergence from CFT073. This insertion was followed by a large internal deletion within the system in the K12 lineage. Had this partial TTSS been identified solely in K12 it would have been puzzling, but comparative analyses such as those available through coliBASE enable such observations to be placed in their true evolutionary context.