The severe diarrhoeal disease cholera is caused by the gram-negative bacterium Vibrio cholerae and continues to be a major cause of morbidity and mortality. Like many Vibrio species V. cholerae inhabits an aquatic ecosystem, and most V. cholerae isolates do not possess the ability to cause cholera. Of more than 200 known serotypes, only O1 and O139 serogroups are highly pathogenic and acknowledged to cause epidemic disease.1 The O1 serogroup can be divided into two groups: classical and El Tor, with the first cholera pandemic, beginning in Asia in 1817, and the subsequent five pandemics probably caused by the classical biotype. The seventh and present pandemic began in 1961 caused by the El Tor biotype.2 In 1992 a novel O-serogroup, O-139, emerged to cause epidemic cholera.3 All V. cholerae O1 and O139 serogroups encode the major virulence factors cholera toxin and toxin coregulated pilus, the latter being encoded on a pathogenicity island, named Vibrio Pathogenicity Island-I (VPI-1).4 Several other genomic regions have been identified that occur mainly among epidemic O1 and O139 serogroup isolates, including the so-called Vibrio seventh pandemic island-I (VSP-I) encoding ORFs VC0175 to VC0185, VSP-II encoding ORFs VC0490 to VC0516 and VPI-2 encoding ORFs VC1758 to VC1809.5–7 A microarray analysis of V. cholerae El Tor isolates was used to identify VSP-I and VSP-II encompassing genes that are possibly responsible for the unique characteristics of the seventh pandemic (El Tor) strains.5 An evolutionary genetic analysis of clinical and environmental isolates suggested that pandemic strains arose from a common O1 serogroup progenitor through the successive acquisition of new virulence regions.8 VPI-2 is a 57.3 kb region which consists of 52 ORFs present in all toxigenic O1 and O139 serogroup isolates, but lacking in non-O1 and non-O139 nontoxigenic isolates.6 VPI-2 encodes a type-I restriction modification system, a nan-nag region of genes involved in sialic acid metabolism that may play a nutritional role,9 a sialidase/neuraminidase known to convert intestinal higher-order gangliosides to GM1,10 and a region with homology to Mu phage. In addition, VPI-2 contains several genes that code for hypothetical proteins. One of these proteins, VC1805, is a 148 amino acid protein with no function revealed so far through sequence analysis. VC1805 is also encoded in the reduced 20 kb VPI-2 region present among most O139 serogroup isolates that are missing ORFs VC1761 to VC1788.6 A paralogue of VC1805 exists within VSP-II, VC0508 a 147 residue protein that shares 59% amino acid sequence identity with VC1805. A search of the sequence database using PSI-BLAST reveals homologues of VC1805 in several Vibrio species: V. vulnificus, V. splendidus, V. alginolyticus and V. fischeri, and orthologues in Altermonas macleodii, Aeromonas hydrophilia, and some Shewanella species. In addition, PSI-BLAST reveals that the adjacent hypothetical protein VC1804, with 104 residues, is a homologue of VC1805 sharing 26% sequence identity. Similarly the hypothetical protein VC0509 is a homologue of the adjacent protein VC0508. The four proteins VC0508, VC0509, VC1804, and VC1805 are therefore likely to share the same protein fold and be functionally related. As part of a structural genomics approach to understanding the function of hypothetical proteins within V. cholerae genomic islands we have determined the crystal structure of VC1805 to a resolution of 2.1 A using heavy atom isomorphous replacement. The structure reveals a similarity to the human mitochondrial protein p32 that is known to have several binding partners, including the human complement system protein C1q. This study shows that VC1805 does bind C1q, suggesting potential biological roles for the protein and its homologues.
Read full abstract