- Volume 44, Issue 6, 1996

To evaluate the role of the acidic pH of phagosomes on the invasive ability and fate of Listeria monocytogenes within host cells, entry and replication of this gram-positive bacterium in a human enterocyte-like cell line (Caco-2) were investigated by a combination of biochemical and ultrastructural approaches. The effects of inhibitors of vacuolar acidification – the lipophilic weak base ammonium chloride, the carboxylic ionophore monensin and the vacuolar proton ATPase inhibitor bafilomycin A1 – on the bacterial invasion pathway were analysed. These agents, which raise the intracellular vesicle acidic pH of living cells by different mechanisms, affected L. monocytogenes replication in Caco-2 cells. Bacteria internalised by bafilomycin-treated cells were unable to escape from phagosomes, as demonstrated by electronmicroscopy. The results provide evidence that low pH is required for efficient intracellular growth of L. monocytogenes.

This study investigated how pH and the presence of urea affect the survival and growth of Helicobacter pylori and whether these factors affect susceptibility to antibiotics in vitro. The viability of a wild-type strain and a urease-deficient mutant of H. pylori was studied after incubation for 1 h in buffers at different pH values at 37°C under microaerophilic conditions. Viable counts were not affected at pH 5 and pH 7. In buffer at pH 3, there were no viable organisms, but urea (6.25 mm) protected the wild-type strain, which survived well. At pH 9, urea further reduced the viability of H. pylori and flurofamide almost abolished the effect of urea on the wild-type strain. Neither urea nor flurofamide affected the viability of the urease-deficient mutant under the same conditions. Growth was also pH dependent and was enhanced in shake-cultures. At pH 5, urea supported growth of the wild-type strain, but at pH 7 a toxic effect on the bacteria was observed. Growth of H. pylori at pH 5.9 was poor, and susceptibility to amoxycillin, erythromycin and clarithromycin was markedly less than at pH 7.2 and 7.9. The bactericidal activities of metronidazole and tetracycline were similar at the different pH values studied. At neutral pH the killing rates of amoxycillin and clarithromycin were growth rate dependent. Susceptibility to metronidazole was enhanced in stationary cultures. The interaction obtained between the proton pump inhibitor, omeprazole, and amoxycillin at pH 7 was of additive type. These results suggest that pH and growth conditions may be important in the antibacterial efficacy of different antibiotics in vivo and also provide a possible explanation for the potentiating effect of omeprazole with antibiotics in the treatment of H. pylori infections.

One hundred and eight strains of Aeromonas from clinical and environmental samples were speciated. Seven species were identified, the most prevalent of which was A. hydrophila. Experimental studies in an animal model with 36 representative strains of different species revealed that all strains could cause significant fluid accumulation in rabbit ileal loops. Of 107 strains showing single or multiple antimicrobial resistance, the highest incidence of resistance was shown for β-lactam antibiotics other than cefotaxime. Transferable resistance plasmids, encoding resistance to ampicillin, cephalexin, cefoxitin, erythromycin and furazolidone, either alone or in combination, were detected in 35 strains. A further proportion of strains could be cured of one or more resistance markers, including resistance to nalidixic acid, and this was accompanied by the loss of plasmid DNA. The plasmids ranged in size between 85.6 and >150 kb.

Nine hundred and twenty-five Escherichia coli isolates from cases of diarrhoea in the United Kingdom and belonging to enteropathogenic E. coli (EPEC) O serogroups were examined for virulence properties. The tests included adhesion to HEp-2 cells, the fluorescence actin staining (FAS) test (which correlates with the ability to cause attaching and effacing lesions) and DNA hybridisations with probes to detect sequences for eaeA (E. coli attaching and effacing factor), EAF (EPEC adherence factor), verocytotoxins VT1 and VT2, enteroaggregative E. coli and diffusely adherent E. coli. The O serogroups examined were 18, 26, 44, 55, 86, 111, 114, 119, 125, 126, 127, 128 and 142. Six hundred and sixty strains (71.4%) hybridised with at least one of the DNA probes. Over 80% of strains in O serogroups 26, 55, 119, 125, 127 and 142 and 41% of strains of serogroups 86, 111, 114, 126 and 128 hybridised with the eae probe and most showed localised attachment and were FAS-positive. However, <10% of these eae probepositive strains hybridised with the EAF probe. Eighty-four of 232 strains in O serogroups 44, 86, 111, and 126 were enteroaggregative. VT genes were detected in 57 of 402 strains in O serogroups 26, 55, 111 and 128. Identification of EPEC by serogrouping was shown to be an effective method of identifying strains with pathogenic potential, although the organisms were diverse in their properties.

Escherichia coli MT78, an avian pathogenic strain of serogroup O2, produces a variant form of type 1 fimbriae with distinct antigenic properties and apparent mol. wt of the major subunit. The fim gene cluster of strain MT78 was cloned and its sequence was determined in a region spanning upstream of fimB to the beginning of fimD. Whereas most genes were well conserved relative to fim genes previously described, comparison of the fimA gene from strain MT78 with homologous sequences from other strains of E. coli and Klebsiella pneumoniae revealed that most differences were clustered in four well defined regions. A PCR assay, based upon these variable sequences, allowed amplification of a fragment of gene fimA which is specific for most O2 strains. In addition, the sequence of the previously uncharacterised gene fimI, which is located between genes fimA and fimC, was determined.

Clostridium difficile toxins A and B cause antibiotic-associated colitis. Whereas antigenic determinants specifying neutralisation of toxin A have been partially elucidated, those of toxin B remain unknown. To define antigenic determinants of toxin B, synthetic peptides were prepared for five linear sequences selected by computer analysis for putative T and B epitopes. Peptides spanning the carboxy terminal region (aa 2155-2283) were also selected because this region contains repetitive units thought to bind the toxin to cell receptors. Multiple antigenic peptides were synthesised by linking four peptide copies to a core of four lysine residues (tetraMAP). Outbred mice were given four doses of each tetraMAP by intraperitoneal injection and specific immunoglobulins G and A were measured by enzyme-linked immunosorbent assay (ELISA) in serum, ascitic fluid and faeces. All 14 MAPs induced strong IgG responses against the homologous peptide; peptides representing aa 2155-2179 and 2246-2270 induced the strongest responses, of 592 and 493 ELISA units, respectively-although, to a lower extent, all 14 MAPs induced serum and faecal IgA responses against the homologous peptide. All MAPs induced IgG1 and IgG2b subclasses, documenting their capacity to elicit Th2-dependent mucosal immunity. IgG anti-MAPs were assayed for reaction with native toxins A and B; most anti-MAPs recognised the toxins only weakly or did not recognise them. Antibodies against peptide representing aa 2168-2192 recognised both native toxin B (19 ELISA units) and toxin A (2 ELISA units). None of the antibodies neutralised cytotoxicity of either toxin in cell culture. In contrast, four MAPs (aa 2080-2095, 2168-2192, 2220-2244 and 2233-2257) inhibited cytotoxicity when mixed with toxin B before addition to cells; inhibition was mediated by a direct interaction with toxin B.

A murine IgM monoclonal antibody (MAb H11) was developed against the type polysaccharide capsular antigen of group B streptococcus (GBS), serotype IV, after intraperitoneal immunisation of BALB/c mice with heat-killed bacteria. MAb H11 reacted in immunodiffusion with the purified polysaccharide in both its sialylated and desialylated form, giving a line of identity, and opsonised type IV GBS strains in an invitro assay. When administered at the time of intraperitoneal lethal challenge with homologous GBS, or 4 h earlier, MAb H11 protected 90% of the mice. Protection was still observed when MAb H11 was given 4 h after the challenge. This MAb was strongly effective in preventing septic arthritis induced by type IV GBS.