- Volume 45, Issue 5, 1996

Previous investigations have demonstrated the presence of both superoxide dismutase and catalase enzymes in several intracellular pathogens, including a number of mycobacterial species. These enzymes are believed to be involved in the protection of the pathogen from the bactericidal products of oxidative metabolism. Superoxide dismutase and catalase were identified in crude extracts of Mycobacterium ulcerans by polyacrylamide gel electrophoresis. Inhibition experiments showed that the superoxide dismutase probably contained manganese as the metal cofactor. Other mycobacterial species examined for comparison produced bands of superoxide dismutase activity with a different mobility to that of M. ulcerans, suggesting possible structural differences between the enzymes.

The significance of multiplicity of infection (moi) for invasiveness and intracellular multiplication of Listeria monocytogenes in Caco-2 and HeLa cell monolayers was investigated. A low moi (1:1) resulted in recovery of significantly more L. monocytogenes when these bacteria were used to infect either cell line. At high moi (100:1), the percentage recovery of bacteria was comparatively low, even after extensive invasion and intracellular multiplication. Microscopic analysis of Giemsa- and immunofluorescent-stained infected monolayers revealed extensive cell disruption and exposure of the internalised bacteria to the bactericidal effect of gentamicin. By contrast, a low moi resulted in minimal cytopathic effects and evidence of cell to cell spread by L. monocytogenes was consistently observed in HeLa and J774, but not in Caco-2 cell lines. Nevertheless, the use of HeLa and Caco-2 cell monolayers enabled a clear distinction to be made between invasive (L. monocytogenes) and non-invasive Listeria spp. (L. innocua, L. ivanovii, L. seeligeri, L. grayi, L. welshimeri and L. monocytogenes LLO19). The use of a low moi with HeLa cell monolayers provided a reliable tissue-culture model of infection for L. monocytogenes.

The binding of fibronectin, vitronectin, collagen and sialoprotein to 65 anaerobic strains was investigated by means of latex agglutination tests. The binding of fibronectin, collagen and lactoferrin to the same strains was also tested by means of 125I-labelled proteins. The strains were isolated from abdominal infections (55%), from the faeces of healthy subjects (29%) or from the depths of tonsils removed at tonsillectomy (16%). The binding of fibronectin and collagen to Bacteroides fragilis strains, tested by the latex agglutination assay, was stronger than their binding to other species. The vitronectin binding of the strains was less common, but was always accompanied by fibronectin binding. Binding to fibronectin-coated beads was inhibited by pre-incubation of the bacterial cells with soluble fibronectin and by heat or protease treatment of the bacterial suspension. No inhibition of the binding was observed with carbohydrates. None of the 65 strains exhibited any binding to fetuin or asialofetuin; 8% of the strains had a binding site for mucin. The binding to mucin-coated beads was inhibited by pre-incubation of the cells with mucin. The radiolabelling method indicated a low binding to 125I-fibronectin. The binding of 125I-collagen-I and 125I-lactoferrin was higher for the anaerobic strains tested.

The purpose of the present study was to determine the relative distribution of plasmids in 87 clinical isolates of Nocardia, belonging to the five major pathogenic species. A correlation between plasmid content and the site of infection within the host, resistance to antibiotics and enzymic profiles was also investigated. The plasmid extraction procedure of Kado and Liu was used. Electrophoretic analysis revealed one-to-four plasmid bands, ranging in size from < 8 to >50 kb, in 27 strains (31%). Based on the number of isolates tested, the incidence of plasmid-bearing strains was significantly higher among N. farcinica than N. asteroides strains. Within N. farcinica, the incidence of plasmids was higher among strains isolated in the Paris area than in strains isolated elsewhere, such as in the French provinces or outside France. A statistically significant correlation was demonstrated between the cutaneous localisation of infections and the incidence of plasmid-bearing strains. The presence of plasmids in nocardiae could not be associated with specific phenotypic traits such as resistance to antibiotics or enzymic activity. The fact that the majority of Nocardia clinical isolates (60 of 87) did not contain plasmids suggests that plasmids are not involved directly in virulence and that there is no selective pressure for plasmid acquisition.

A total of 3065 strains of Escherichia coli isolated between 1958 and 1992 from patients with diarrhoea in different countries were examined for virulence factors by hybridisation with biotinylated DNA probes for genes that coded for production of heat-labile and heat-stable enterotoxins, enteroinvasiveness, production of verotoxins and attaching-and-effacing factor and were serotyped. Of the 3065 strains, 1998 were placed into one of four pathogenic categories by their virulence factors: 1057 enterotoxigenic E. coli (ETEC) comprising 30 O-groups, 73 serovars and 137 untypable strains; 132 enteroinvasive E. coli (EIEC) comprising 11 O-groups and 13 serovars; 64 verotoxin-producing E. coli (VTEC) comprising 11 O-groups, 17 serovars and 13 untypable strains; and 745 enteropathogenic E. coli (EPEC) comprising 34 O-groups, 92 serovars and 91 untypable strains. The remaining 1067 strains did not hybridise with any of the DNA probes used. About half the number of O-groups recognised were not restricted to a single pathogenic category, although the combinations of O- and H-antigens were different in each category.

Between July 1994 and March 1995, 64 isolates of Pseudomonas aeruginosa were implicated in bacteraemia in 25 cancer patients in five wards of two hospitals. These, together with 24 environmental isolates and one isolate from a bacteraemia in a non-cancer patient were examined by three PCR-based DNA fingerprinting methods: random amplified polymorphic DNA (RAPD), enterobacterial-repetitive intergenic consensus (ERIC)-PCR, and 16S-23S spacer region-based RAPD. These methods were reproducible, discriminatory and showed close agreement; all indicated that 47 isolates that had caused bacteraemia in 19 cancer patients were indistinguishable. Seventeen other isolates that had caused bacteraemia in 10 cancer patients were discriminated into eight further groups, and the 24 environmental and non-cancer patient isolates into further distinct groups. No environmental source of the epidemic strain was found, but it was suspected that the outbreak was related to infusion implants.

The importance of Ureaplasma urealyticum as a pathogen in premature neonates and patients with a profound defect in humoral immunity has, over the last few years, become well recognised. U. urealyticum is unique amongst the Mycoplasmataceae for its use of urea metabolism as an essential source of energy. The urease enzyme responsible for this is, therefore, of prime importance and any variability in expression of this enzyme may play a role in virulence of the organism. U. urealyticum is divided into 14 serovars comprising two biovars-the parvo-biovar and T960-biovar. In this study monoclonal antibodies (MAbs) were produced against the urease enzyme. Two distinct epitopes of the 72-kDa α-subunit were recognised by three different MAbs. Under denaturing conditions both epitopes were shown to be specific for the parvobiovar.