Search

Vibrio cholerae and V. anguillarum are recognized as aquatic-borne human and fish pathogens, respectively. Based upon analyses of several genes and the presence of novel genetic elements it seems that these two species are very closely related. Studies in this laboratory have identified an association of IS1358 with rfb and capsule loci in these two species. The most recent findings suggest that IS1358 is associated with the rfb region in V. cholerae O1 and O139 and in V. anguillarum O1 and O2. In addition, the rfb region in both V. cholerae serogroups and in V. anguillarum O1 is limited at one end by gmhD. These features make it feasible to envisage a mechanism by which the evolution of new rfb genes is taking place involving IS1358 and the region around gmhD. Furthermore, it is possible to envisage that there is or has been an exchange of genetic material between these species leading to new rfb/capsule regions. This review examines the genetics and biosynthesis of the O-antigen and capsule of V. cholerae O1 and O139, as well as the V. anguillarum serogroup O1 and the role of IS1358. Throughout this review we have used the new nomenclature for rfb genes proposed by.

The first four genes (rfbA,B,D,E) of the rfb region of Vibrio cholerae O1 are predicted to encode the enzymes required for the biosynthesis of perosamine, which constitutes the backbone structure of the O-antigen of the lipopolysaccharide. Based on homology to known proteins/protein families, the following functions are predicted: RfbA, phosphomannose isomerase-guanosine diphosphomannose pyrophosphorylase; RfbB, phosphomanno-mutase; RfbD, oxido reductase and RfbE, perosamine synthetase (amino-transferase). Thus, perosamine is synthesized from fructose 6-phosphate via the intermediates mannose 6-phosphate by RfbA, to mannose 1-phosphate by RfbB, to GDP-mannose by RfbA, to GDP-4-keto-6-dideoxymannose by RfbD and to GDP-perosamine by RfbE. This final product would then serve as the substrate for the addition of the tetronate, which could then be polymerized into the O-antigen for transfer to the lipid A plus core oligosaccharide and export to the cell surface. The organization of these genes are such that one would expect them to be translationally coupled as part of the rfb operon. However, the absence of readily detectable promoter sequences suggests low levels of transcription, in line with other studies. The nucleotide sequence of these genes is absolutely conserved in the two isolates 569B (classical, Inaba) and O17 (El Tor, Ogawa) which were expected to show maximal sequence variation. This suggests very tight constraints on the micro-evolution within these sequences.

The nucleotide sequence of a region of the rfb genes, encoding biosynthesis of the Vibrio cholerae (Vc) O1 O-antigen, was determined. Analysis of the open reading frames (ORFs) within this region has revealed similarities with a number of different classes of biosynthetic proteins and enzymes. The ORFs have been designated RfbK, RfbL, RfbM, RfbN and RfbO. RfbK is a small, acidic protein which has similarity to the family of proteins known as acyl-carrier proteins (ACP). The RfbL protein has similarity to a super-family of enzymes which adenylate their substrates as a part of their reaction mechanism. Included in these are several acetyl-CoA ligases. Alignment of RfbL with these proteins reveals a highly conserved domain containing the motif GlyXaaXaaGlyXaaPro. This resembles the ATP-binding site motif and may represent a variant of the usual motif, except that Pro replaces Gly. The VcRfbM protein has similarity with a family of long-chain, iron-containing alcohol dehydrogenases, of which the Escherichia coli K-12 fucO and adhE gene products are also members. The RfbN protein has sequence homology with LuxE and LuxC of Vibrio harveyi (Vh) and other bioluminescent bacterial species. The latter are two components of the enzyme complex which synthesizes the long-chain aldehyde used in the V. harveyi bioluminescence system. Finally, the VcRfbO protein has sequence similarity with acetyl-CoA transferases. We were able to identify a number of the gene products using a T7 expression system, confirming several of the allocated ORFs. A biosynthetic pathway for the Vc O-antigen component 3-deoxy-L-glycero-tetronic acid, based on the enzymatic functions predicted for the RfbK, RfbL, RfbM, RfbN and RfbO proteins, is presented.

The nucleotide sequence of that part of the Vibrio cholerae (Vc) O1 rfb region encompassing rfbG, rfbH and rfbI is presented. Expression of these genes has enabled the products for rfbG and rfbI to be confirmed, but the rfbH product has not been detected. Comparisons with the sequences of known proteins reveals that RfbH and RfbI are likely to be involved in the export of lipopolysaccharide (LPS). RfbH shows considerable homology to a number of integral membrane proteins, some of which have been identified as possibly having a role as an export channel for capsular polysaccharides. RfbI corresponds to an ATP-binding protein usually found linked to the membrane protein and is thought to be required for energizing this export process. Thus, we propose that RfbH and RfbI form a complex for the export of Vc O1 LPS. The function of RfbG is unknown, but it would appear to be a relatively hydrophilic protein and we can only speculate that it may be either a specific transferase or possibly the O-antigen polymerase.

The rfaD gene of Escherichia coli encodes ADP-L-glycero-D-mannoheptose-6- epimerase, an enzyme required for the biosynthesis of the lipopolysaccharide (LPS) precursor ADP-L-glycero-D- mannoheptose, associated with production of the core oligosaccharide. We have identified an rfaD homologue in Vibrio cholerae O1. This gene maps adjacent to the rfb region encoding O-antigen biosynthesis, but is transcribed divergently. The complete nucleotide sequence of rfaD and the flanking DNA has been determined, and rfaD would appear to be the only gene homologous to known LPS core biosynthesis genes in this region. Comparison with the E. coli rfaD shows many similar structural features such as the ADP-binding beta alpha beta fold at the N terminus, as well as a high degree of homology of both the nucleotide and amino-acid sequences. Based on homology, rfaD of V. cholerae may be transcribed using both sigma 70- and sigma 54-dependent promoters.

We have cloned and sequenced the enterohemolysin (ehl)-associated region from a temperate bacteriophage isolated from an Escherichia coli O26:H11 strain. Phage phi C3208 was isolated together with other temperate bacteriophages which transduce the enterohemolytic phenotype to non-hemolytic E. coli O26 strains. The nucleotide sequence of the 1245-bp phi C3208 DNA insert in plasmid pEO39, which mediates Ehly2 production in E. coli K-12, was determined and was found to be partially homologous to DNA of bacteriophage lambda but is completely unrelated to DNA sequences encoding the synthesis of Ehly1 [Stroeher et al. Gene 132 (1993), 89-94]. It was shown that part of this region can be used as an Hly2-associated specific DNA probe.

The nucleotide sequence of the 2.1-kb EcoRI-AccI fragment of the enterohemolysin (Ehly)-associated plasmid, pEO21, has been determined. A third of this sequence encodes a 29.6-kDa protein, and coincides with the location of Tn1725 insertions which inactivate the production of Ehly. A protein of similar size (33-35 kDa) was found to be produced in large amounts by JM83[pEO21] and found to be immunologically related to the approximately 65-kDa protein made by the parental strain, C3888. DNA sequences coding for the 29.6-kDa protein of phi C3888, now designated Ehly 1, were found only in some enterohemolytic Escherichia coli indicating the existence of multiple, genetically different Ehlys.

The toxin co-regulated pilus (Tcp) of Vibrio cholerae appears to be a major protective antigen. By cosmid cloning we have isolated a number of clones capable of converting Tcp- El Tor strains of V. cholerae to Tcp+. A synthetic oligodeoxyribonucleotide probe based upon the N-terminal amino acid sequence of TcpA, has been used to localize the structural gene within the cosmid clones. Using suitable subclones, the nucleotide sequence of the tcpA gene has been determined. The gene encodes a 23.3-kDa pre-protein which in its mature form has a size of 20.3 kDa. The N-terminal leader peptide or signal sequence is atypical and does not conform with the usual rules of such sequences. The TcpA protein shows some similarities to the major pilins of the methylated phenylalanine type or type-4 pili from other bacteria; however, it is sufficiently different that it may represent a new class.