Penicillin binding proteins of Vibrio cholerae

Eleven penicillin binding proteins (PBPs) of Vibrio cholerae have been identified using [125I] labelled p-hydroxybenzyl penicillin (PenX). These proteins are localised in the inner membrane and have molecular weights ranging from 97,000 to 22,000. Neutral hydroxylamine released the labelled PenX from the PBPs and pretreatment with cold benzyl penicillin inhibited labelling completely. The PBP 4 is the most sensitive target for cephaloridine and aztreonam. Cephaloridine also binds to three other high molecular weight PBPs, 1, 2 and 3. Aztreonam, in addition to PBP 4, has affinity for another low molecular weight PBP, PBP 7. Mecillinam has affinity for PBPs 1, 4 and 11.     

Lactoferrin binding properties of Vibrio cholerae.

The lactoferrin binding properties of Vibrio cholerae, a non-invasive pathogen were investigated. Screening of fifty V. cholerae strains of different serogroups and serotypes, showed that 10% of the V. cholerae strains bound to 125I-labelled lactoferrin, and 40% of the 125I-labelled lactoferrin bound to V. cholerae strain 623 could be displaced by unlabelled lactoferrin. Other iron-binding glycoproteins and ferroproteins like ferritin, transferrin, haemoglobin, and myoglobin inhibited the binding of 125I-lactoferrin to a lesser degree. Monosaccharides (GalNac, Man, Gal, and Fuc), and other glycoproteins such as fetuin and orosomucoid also inhibited the binding to a lesser extent. V. cholerae 623 showed a cell surface associated-proteolytic activity which cleaved off the cell-bound 125I-labelled lactoferrin. The generation of cryptotopes on the V. cholerae cell surface by proteolytic digestion favoured the binding of ferritin, transferrin, haemoglobin, and haemin, as well as Congo red, to cells of V. cholerae 623.     

Characterization of non-membrane-damaging cytotoxin of non-toxigenic Vibrio cholerae O1 and its relevance to disease

The non-membrane-damaging cytotoxin which causes dramatic cell rounding of cultured HeLa cells was purified to homogeneity from a clinical strain (WO5) of non-toxigenic Vibrio cholerae O1 Inaba belonging to the El Tor biotype. The purified protein has a denatured molecular weight of 35 kDa and a native molecular weight of approximately 37 kDa indicating the monomeric nature of the protein. The 15 N-terminal amino acid sequence of non-membrane-damaging cytotoxin showed complete homology to the hemagglutinin protease previously purified and characterized from V. cholerae O1. Purified non-membrane-damaging cytotoxin from V. cholerae O1 was immunologically and biochemically identical to that previously purified from V. cholerae O26. Non-membrane-damaging cytotoxin was found to be enterotoxic in rabbit ileal loop assay inducing accumulation of non-hemorrhagic fluid at 100 μg and elicited a concentration dependent increase in short circuit current and tissue conductance of rabbit ileal mucosa mounted on Ussing chambers. A significant serum immunoglobulin G response against non-membrane-damaging cytotoxin was elicited by patients infected with V. cholerae O139 but not with V. cholerae O1. These properties make non-membrane-damaging cytotoxin a potential virulence factor of V. cholerae which should be taken into consideration while making live, attenuated recombinant vaccine strains against cholera.     

Carbohydrate specific binding of fibronectin to Vibrio cholerae cells

Cells of 10 strains of Vibrio cholerae were grown on Trypticase Soy Broth and were tested for different surface porperties such as expression of surface haemagglutinins, cell-surface hydrophobicity and binding to 3 connective tissue proteins: fibronectin, type II collagen and fibrinogen.All strains bound fibronectin and one selected strain was shown to bind in a time-dependent and saturable manner.The binding of ¹²⁵I-labelled fibronectin could be completely inhibited by unlabelled fibronectin, and also partly by some other glycoproteins. Mannose inhibited binding of fibronectin up to 60%. The data indicate that carbohydrate structures within the 40 kDa (gelatin binding) and 105 kDa (cell binding) fragments of fibronectin are recognized by lectins on V. cholerae. The binding of collagen or fibrinogen was low or negligible.     

Structural requirements of cholesterol for binding to Vibrio cholerae hemolysin

Cholesterol is necessary for the conversion of Vibrio cholerae hemolysin (VCH) monomers into oligomers in liposome membranes. Using different sterols, we determined the stereochemical structures of the VCH-binding active groups present in cholesterol. The VCH monomers are bound to cholesterol, diosgenin, campesterol, and ergosterol, which have a hydroxyl group at position C-3 (3betaOH) in the A ring and a C-C double bond between positions C-5 and C-6 (C-C Delta(5)) in the B ring. They are not bound to epicholesterol and dihydrocholesterol, which form a covalent link with a 3alphaOH group and a C-C single bond between positions C-5 and C-6, respectively. This result suggests that the 3betaOH group and the C-CDelta(5) bond in cholesterol are required for VCH monomer binding. We further examined VCH oligomer binding to cholesterol. However, this oligomer did not bind to cholesterol, suggesting that the disappearance of the cholesterol-binding potential of the VCH oligomer might be a result of the conformational change caused by the conversion of the monomer into the oligomer. VCH oligomer formation was observed in liposomes containing sterols with the 3betaOH group and the C-C Delta(5) bond, and it correlated with the binding affinity of the monomer to each sterol. Therefore, it seems likely that monomer binding to membrane sterol leads to the assembly of the monomer. However, since oligomer formation was induced by liposomes containing either epicholesterol or dihydrocholesterol, the 3betaOH group and the C-C Delta(5) bond were not essential for conversion into the oligomer.     

Testing the biological activity of Vibrio cholerae on cell subcultures

Testing the supernatants of ctx(+) strains of V. cholerae eltor and V. cholerae O139 on cell subcultures confirmed the possibility of the synthesis of hemolysin by V. cholerae under the condition of growing them in tripton medium lacking FeCl3. At the same time ctx(+) strains of V. cholerae of both serogroups retained, simultaneously with hemolysin production, their capacity for the synthesis of cholera toxin.     

Morphological and cytoskeletal changes caused by non-membrane damaging cytotoxin of Vibrio cholerae on Int 407 and HeLa cells

End-group mediated conjugation of bacterial polysaccharides (PSs) to carrier proteins containing T-helper cell epitopes renders such polysaccharides immunogenic also in young infants. Optimal construction of such conjugate vaccines requires fragmentation of the PS prior to the coupling reaction. In the present study a general simple and inexpensive method for the fragmentation of PSs is presented. It is based on the irradiation of isolated PSs in an electron beam accelerator. Exposure of isolated pneumococcal capsular polysaccharides (PnPSs) to ionizing radiation resulted in their partial depolymerization in a radiation dose-dependent manner. Radiation, unlike sonication, generated PnPS fragments of molecular size lower than 50 kDa and as small as 1.5 kDa when high radiation doses were used. These PnPS fragments have terminal reducing groups that can be easily used for chemical activation and subsequent coupling to any chosen carrier protein. The radiation-produced PnPS fragments retained their antigenic epitopes, when compared to native, full-size PnPSs as determined by enzyme-linked immunoassay.     

Cloning of the Vibrio cholerae recA gene and construction of a Vibrio cholerae recA mutant

A recombinant plasmid carrying the recA gene of Vibrio cholerae was isolated from a V. cholerae genomic library, using complementation in Escherichia coli. The plasmid complements a recA mutation in E. coli for both resistance to the DNA-damaging agent methyl methanesulfonate and recombinational activity in bacteriophage P1 transductions. After determining the approximate location of the recA gene on the cloned DNA fragment, we constructed a defined recA mutation by filling in an XbaI site located within the gene. The 4-base pair insertion resulted in a truncated RecA protein as determined by minicell analysis. The mutation was spontaneously recombined onto the chromosome of a derivative of V. cholerae strain P27459 by screening for methyl methanesulfonate-sensitive variants. Southern blot analysis confirmed the presence of the inactivated XbaI site in the chromosome of DNA isolated from one of these methyl methanesulfonate-sensitive colonies. The recA V. cholerae strain was considerably more sensitive to UV light than its parent, was impaired in homologous recombination, and was deficient in induction of a temperate vibriophage upon exposure to UV light. We conclude that the V. cholerae RecA protein has activities which are analogous to those described for the RecA protein of E. coli.     

Composition and immunochemical properties of the cell surface proteins of Vibrio cholerae

The composition and immunochemical properties of cell surface proteins of Vibrio cholerae belonging to both the biotypes (classical and El Tor) and the serotypes (Ogawa and Inaba) were investigated. Proteins were isolated by extraction with EDTA/NaCl. When the extract was further treated with sodium deoxycholate, a product significantly enriched with the major protein was obtained. The surface localization of these proteins was confirmed by immunoelectron microscopy using protein A-colloidal gold particles as probes. Antisera to these proteins possessed complement-mediated bactericidal activities towards V. cholerae strains belonging to both the biotypes and the serotypes, and upon crossed immunoelectrophoresis produced several immunoprecipitation reactions towards whole-cell sonicates belonging to all types of V. cholerae. These proteins were immunogenic in the rabbit intestine, as antibodies of two classes (IgG and IgA) were detected in the intestinal fluids. The intestinal immune response was greatly enhanced when cell surface proteins were administered with liposomes. These results suggest that cell surface proteins represent common antigens of V. cholerae and can be explored as vaccine candidates against cholera.     

Carbohydrate-dependent binding of the cell-free hemagglutinin of Vibrio cholerae to glycoprotein and glycolipid.

The carbohydrate-binding specificity of the cell-free hemagglutinin (HA) of Vibrio cholerae (K.K. Banerjee, A.N. Ghose, K. Datta-Roy, S.C. Pal, and A.C. Ghose, Infect. Immun.58:3698-3705, 1990) was studied by using glycoconjugates with defined sugar sequences. The HA was not inhibited by simple sugars including glucobiose, galabiose, and their N-acetylated derivatives. The hemagglutination of rabbit erythrocytes by the HA was inhibited moderately by fetuin, calf thyroglobulin, and human alpha 1-acid glycoprotein, all of which contain multiple asparagine-linked complex-type oligosaccharide units alone or in combination with serine/threonine-linked oligosaccharide units. The inhibitory potencies of the glycoproteins increased approximately 10-fold following removal of the terminal sialic acid and were completely destroyed by exhausative proteolysis. The HA agglutinated phosphatidylcholine liposomes containing GM1-ganglioside or its asialo-derivative in the presence of Ca2+ ions. The association constants of the complexes of the HA with asialofetuin, asialothyroglobulin, GM1-ganglioside, and asialo-GM1-ganglioside were determined by an enzyme-linked immunosorbent assay-based assay and found to be 1.7 x 10(7) M-1, 1.5 x 10(7) M-1, 1.8 x 10(7) M-1, and 2.4 x 10(7) M-1, respectively. Studies using chemically modified glycoproteins and plant lectins with defined sugar specificity revealed that the HA recognized the terminal beta 1-galactosyl moiety of these glycoconjugates. There was no evidence for the presence of an extended carbohydrate-binding domain in the HA molecule or a preference of the HA for a complex, branched oligosaccharide structure. Similar to the mechanisms proposed for the binding of cholera toxin and Shiga toxin to glycolipids and neoglycoproteins, the strong interaction of V. cholerae cell-free HA with glycoconjugates appeared to be a consequence of multiple weak binding to terminal beta1-galactosyl moieties of the glycoproteins or glycolipids.     

Toxins of Vibrio cholerae

Surveyed in the paper are published data on properties, biological activity, genetic determinants and action mechanisms of recently known toxins produced by different strains of Vibrio cholerae irrespectively of their capacity for the synthesis of choleric toxin--the main virulence factor. Their possible importance both for the general clinical pattern of cholera provoked by cholerogenic agents and as independent virulence factors causing diarrhea without cholera is elucidated. The sets and levels of expression of additional toxins can differ for different pathogenic clones and they can correspondingly condition degrees of their epidemic and etiological safety.     

The Zymovars of Vibrio cholerae: multilocus enzyme electrophoresis of Vibrio cholerae

Zymovars analysis also known as multilocus enzyme electrophoresis is applied here to investigate the genetic variation of Vibrio cholerae strains and characterise strains or group of strains of medical and epidemiological interest. Fourteen loci were analyzed in 171 strains of non-O1 non-O139, 32 classical and 61 El Tor from America, Africa, Europe and Asia. The mean genetic diversity was 0.339. It is shown that the same O antigen (both O1 and non-O1) may be present in several genetically diverse (different zymovars) strains. Conversely the same zymovar may contain more than one serogroup. It is confirmed that the South American epidemic strain differs from the 7th pandemic El Tor strain in locus LAP (leucyl leucyl aminopeptidase). Here it is shown that this rare allele is present in 1 V. mimicus and 4 non-O1 V. cholerae. Non toxigenic O1 strains from South India epidemic share zymovar 14A with the epidemic El Tor from the 7th pandemic, while another group have diverse zymovars. The sucrose negative epidemic strains isolated in French Guiana and Brazil have the same zymovar of the current American epidemic V. cholerae.     

PCR-based identification of Vibrio cholerae and the closely related species Vibrio mimicus using the large chromosomal ori sequence of Vibrio cholerae

The bacterial chromosomal replication origin (ori) sequences are a highly conserved essential genetic element. In this study, the large chromosomal replication origin sequence of Vibrio cholerae (oriCIVC) has been targeted for identification of the organism, including the biotypes of serogroup O1. The oriCIVC sequence-based PCR assay specifically amplified an 890 bp fragment from all the V. cholerae strains examined. A point mutation in the oriCIVC sequence of the classical biotype of O1 serogroup led to the loss of a BglII site, which was utilized for differentiation from El Tor vibrios. Interestingly, the PCR assay amplified a similarly sized ori segment, designated as oriCIVM, from V. mimicus strains, but failed to produce any amplicon with other strains. Cloning and sequencing of the oriCIVM revealed high sequence similarity (96%) with oriCIVC. The results indicate that V. mimicus is indeed very closely related to V. cholerae. In addition, the BglII restriction fragment length polymorphism (RFLP) between oriCIVM and oriCIVC sequences allowed us to differentiate the two species. The ori sequence-based PCR-RFLP assay developed in this study appears to be a useful method for rapid identification and differentiation of V. cholerae and V. mimicus strains, as well as for the delineation of classical and El Tor biotypes of V. cholerae O1.     

Covalent binding of flavins to RnfG and RnfD in the Rnf complex from Vibrio cholerae

Enzymes of the Rnf family are believed to be bacterial redox-driven ion pumps, coupling an oxidoreduction process to the translocation of Na+ across the cell membrane. Here we show for the first time that Rnf is a flavoprotein, with FMN covalently bound to threonine-175 in RnfG and a second flavin bound to threonine-187 in RnfD. Rnf subunits D and G are homologous to subunits B and C of Na+-NQR, respectively. Each of these Na+-NQR subunits includes a conserved S(T)GAT motif, with FMN covalently bound to the final threonine. RnfD and RnfG both contain the same motif, suggesting that they bind flavins in a similar way. In order to investigate this, the genes for RnfD and RnfG from Vibrio cholerae were cloned and expressed individually in that organism. In both cases the produced protein fluoresced under UV illumination on an SDS gel, further indicating the presence of flavin. However, analysis of the mutants RnfG-T175L, RnfD-T278L, and RnfD-T187V showed that RnfG-T175 and RnfD-T187 are the likely flavin ligands. This indicates that, in the case of RnfD, the flavin is bound, not to the SGAT sequence but to the final residues of a TMAT sequence, a novel variant of the flavin binding motif. In the case of RnfG, flavin analysis, followed by MALDI-TOF-TOF mass spectrometry, showed that an FMN is covalently attached to threonine-175, the final threonine of the S(T)GAT sequence. Studies by visible, EPR, and ENDOR spectroscopy showed that, upon partial reduction, the isolated RnfG produces a neutral semiquinone intermediate. The semiquinone species disappeared upon full reduction and was not observed in the denatured protein. A topological analysis combining reporter protein fusion and computer predictions indicated that the flavins in RnfG and RnfD are localized in the periplasmic space. In contrast, in NqrC and NqrB the flavins are located in a cytoplasmic loop. This topological analysis suggests that there may be mechanistic differences between the Rnf and Na+-NQR complexes.