The lipopolysaccharide outer core of Yersinia enterocolitica serotype O:3 is required for virulence and plays a role in outer membrane integrity

Lipopolysaccharide (LPS) of Yersinia enterocolitica O:3 has an inner core linked to both the O-antigen and to an outer core hexasaccharide that forms a branch. The biological role of the outer core was studied using polar and non-polar mutants of the outer core biosynthetic operon. Analysis of O-antigen- and outer core-deficient strains suggested a critical role for the outer core in outer membrane properties relevant in resistance to antimicrobial peptides and permeability to hydrophobic agents, and indirectly relevant in resistance to killing by normal serum. Wild-type bacteria but not outer core mutants killed intragastrically infected mice, and the intravenous lethal dose was approximately 10(4)-fold higher for outer core mutants. After intragastric infection, outer core mutants colonized Peyer's patches and invaded mesenteric lymph nodes, spleen and liver, and induced protective immunity against wild-type bacteria. In mice co-infected intragastrically with an outer core mutant-wild type mixture, both strains colonized Peyer's patches similarly during the first 2 days, but the mutant was much less efficient in colonizing deeper organs and was cleared faster from Peyer's patches. The results demonstrate that outer core is required for Y. enterocolitica O:3 full virulence, and strongly suggest that it provides resistance against defence mechanisms (most probably those involving bactericidal peptides).

Outer membrane differences between pathogenic and environmental Yersinia enterocolitica biogroups probed with hydrophobic permeants and polycationic peptides

Sensitivities to polycationic peptides and EDTA were compared in Yersinia enterocolitica pathogenic and environmental biogroups. As shown by changes in permeability to the fluorescent hydrophobic probe N-phenylnaphthylamine (NPN), the outer membranes (OMs) of pathogenic and environmental strains grown at 26 degrees C in standard broth were more resistant to poly-L-lysine, poly-L-ornithine, melittin, cecropin P1, polymyxin B, and EDTA than Escherichia coli OMs. At 37 degrees C, OMs of pathogenic biogroups were resistant to EDTA and polycations and OMs of environmental strains were resistant to EDTA whereas E. coli OMs were sensitive to both EDTA and polycations. Similar results were found when testing deoxycholate sensitivity after polycation exposure or when isogenic pairs with or without virulence plasmid pYV were compared. With bacteria grown without Ca++ available, OM permeability to NPN was drastically increased in pathogenic but not in environmental strains or E. coli. Under these conditions, OMs of pYV+ and pYV- cells showed small differences in NPN permeability but differences in polycation sensitivity could not be detected by fluorimetry. O:1,6 (environmental type) lipopolysaccharide (LPS), but not O:3 or O:8 LPS, was markedly rough at 37 degrees C, and this could explain the differences in polycation sensitivity. LPSs from serotypes O:3 and O:8 grown at 37 degrees C were more permeable to NPN than O:1,6 LPS, and O:8 LPS was resistant to polycation-induced permeabilization. These data suggest that LPSs relate to some but not all the OM differences described. It is hypothesized that the different OM properties of environmental and pathogenic biogroups reflect the adaptation of the latter biogroups to pathogenicity.

Yersinia pseudotuberculosis and Yersinia pestis are more resistant to bactericidal cationic peptides than Yersinia enterocolitica

The action of bactericidal polycationic peptides was compared in Yersinia spp. by testing peptide binding to live cells and changes in outer membrane (OM) morphology and permeability. Moreover, polycation interaction with LPS was studied by measuring the dependence of dansylcadaverine displacement and zeta potential on polycation concentration. When growth at 37 degrees C, Yersinia pestis and Yersinia pseudotuberculosis bound less polymyxin B (PMB) than pathogenic or non-pathogenic Yersinia enterocolitica, regardless of virulence plasmid expression. Y. pseudotuberculosis OMs were unharmed by PMB concentrations causing extensive OM blebbing in Y. enterocolitica. The permeability to lysozyme caused by PMB was greater in Y. enterocolitica than in Y. pseudotuberculosis or Y. pestis and differences increased at 37 degrees C. Similar observations were made with other polycations using a polymyxin/novobiocin permeability assay. With LPS of cells grown at 26 degrees C, polycation binding was highest for Y. pseudotuberculosis and lowest for Y. pestis, with Y. enterocolitica yielding intermediate results which were lower for pathogenic than for non-pathogenic strains. With LPS of cells grown at 37 degrees C, polycation binding remained unchanged for Y. pestis and pathogenic Y. enterocolitica, increased for non-pathogenic Y. enterocolitica and decreased for Y. pseudotuberculosis to Y. pestis levels. Polycation binding related in part to differences in charge density (zeta potential) of LPS aggregates, suggesting similar effects at bacterial surfaces. It is suggested that species and temperature differences in polycation resistance relate to infection route, invasiveness and intracellular multiplication of Yersinia spp.

Proper expression of the O-antigen of lipopolysaccharide is essential for the virulence of Yersinia enterocolitica O:8 in experimental oral infection of rabbits

The O-antigen of lipopolysaccharide (LPS) is required for virulence in Yersinia enterocolitica serotype O:8. Here we evaluated the importance of controlling the O-antigen biosynthesis using an in vivo rabbit model of infection. Y. enterocolitica O:8 wild-type strain was compared to three mutants differing in the O-antigen phenotype: (i) the rough strain completely devoid of the O-antigen, (ii) the wzy strain that lacks the O-antigen polymerase (Wzy protein) and expresses LPS with only one repeat unit, and (iii) the wzz strain that lacks the O-antigen chain length determinant (Wzz protein) and expresses LPS without modal distribution of O-antigen chain lengths. The most attenuated strain was the wzz mutant. The wzz bacteria were cleared from the tissues by day 30, the blood parameters were least dramatic and histologically only immunomorphological findings were seen. The level of attenuation of the rough and the wzy strain bacteria was between the wild-type and the wzz strain. Wild-type bacteria were highly resistant to killing by polymorphonuclear leukocytes, the wzz strain bacteria were most sensitive and the rough and wzy strain bacteria were intermediate resistant. These results clearly demonstrated that the presence of O-antigen on the bacterial surface is not alone sufficient for full virulence, but also there is a requirement for its controlled chain length.

Virulence characteristics and epidemiology of Yersinia enterocolitica and Yersiniae other than Y-pseudotuberculosis and Y-pestis isolated from water and sewage

Aims: To determine the species, bio-sero-phagetypes, antimicrobial drug resistance and also the pathogenic potential of 144 strains of Yersinia spp. isolated from water sources and sewage in Brazil.Methods and Results: the 144 Yersinia strains were characterized biochemically, serologically and had their antibiotic resistance and phenotypic virulence markers determined by microbiological and serological standard techniques. The Y. enterocolitica strains related to human diseases were also tested for the presence of virulence genes, by the PCR technique. The isolates were classified as Y. enterocolitica, Y. intermedia, Y. frederiksenii, Y. kristensenii and Yersinia biochemically atypical. The 144 isolates belonged to various bio-serogroups. Half of the strains showed resistance to three or more drugs. The Y. enterocolitica strains related to human diseases exhibited phenotypic virulence characteristics and virulence genes.Conclusions: Water from various sources and sewage are contaminated with Yersinia spp. in Brasil. Among these bacteria, virulent strains of Y. enterocolitica were found, with biotypes and serogroups related to human diseases.Significance and Impact of the Study: This is the first documented description of the occurrence of pathogenic Y. enterocolitica in water sources and sewage in Brazil. The occurrence of virulence strains of Y. enterocolitica shows that the environment is a potential source of human infection by this species in this country.

TNF-alpha, H2O2 and NO response of peritoneal macrophages to Yersinia enterocolitica O : 3 derivatives

In this study, the effect of Yersinia derivatives on nitric oxide (NO), hydrogen peroxide (H2O2) and tumor necrosis factor-alpha (TNF-alpha) production by murine peritoneal macrophages was investigated. Addition of lipopolysaccharide (LPS) to the macrophage culture resulted in NO production that was dose dependent. on the other hand, bacterial cellular extract (CE) and Yersinia outer proteins (Yops) had no effect on NO production. The possible inhibitory effect of Yops on macrophage cultures stimulated with LPS was investigated. Yops partially inhibited NO production (67.4%) when compared with aminoguanidine. The effects of Yersinia derivatives on H2O2 production by macrophages were similar to those on NO production. LPS was the only derivative that stimulated H2O2 release in a dose-dependent manner. All Yersinia derivatives provoked the production of TNF-alpha, but LPS had the strongest effect, as observed for NO production. CE and Yops stimulated TNF-alpha production to a lesser extent than LPS. The results indicate the possibility that in vivo Yops may aid the evasion of the bacteria from the host defense mechanism by impairing the secretion of NO by macrophages. (C) 2003 Elsevier SAS. All rights reserved.

Yersinia enterocolitica O : 3 isolated from patients with or without reactive arthritis induces polyclonal activation of B cells and autoantibody production in vivo

The mechanisms by which arthritis-provoking pathogens such as Yersinia enterocolitica interact with the human immune system to produce inflammatory synovitis are not well known. One of the immunomodulating mechanisms used against these pathogens is the polyclonal activation of lymphocytes. In this study, we investigated the extent of the B-lymphocyte activation induced in mice by a strain of Y. enterocolitica O:3 (FCF 526) isolated from a patient with arthritis, and compared it with two other strains, a virulent one (FCF 397[+]) isolated from a patient without arthritis and its plasmidless isogenic pair (FCF397[-]). Also we investigated the production of autoantibodies in mice infected with these different strains. SPF Swiss mice were infected intravenously with a suspension of Y. enterocolitica . Spleen cells were taken on days 7, 14, 21 and 28 after infection and the number of cells secreting nonspecific and specific antibodies of IgG 1 , IgG 2a , IgG 2b , IgG 3 , IgM and IgA isotypes were determined by the ELISPOT technique. The presence of autoantibodies in mouse serum was investigated by the dot-blot assay. The pattern of infection of the three bacterial strains were almost the same. We observed a general increase in the number of nonspecific Ig-secreting cells with all three strains, and the greatest increases observed were in the IgG 2a and IgG 3 isotypes. Only a small fraction of the immunoglobulins detected were antibacterial, suggesting that the rest resulted from polyclonal B cell activation. The strain isolated from the patient with arthritis (FCF526) induced the greatest production of autoantibodies, coinciding with the period in which the greatest activation of nonspecific B lymphocytes was seen. There were no signs of arthritis or inflammation in the joints of the infected animals. Based on our results, we were unable to determine whether there is an association between the arthritogenic capability of Y. enterocolitica and polyclonal activation of B cells.

Molecular typing and virulence markers of Yersinia enterocolitica strains from human, animal and food origins isolated between 1968 and 2000 in Brazil

Molecular typing and virulence markers were used to evaluate the genetic profiles and virulence potential of 106 Yersinia enterocolitica strains. of these strains, 71 were bio-serotype 4/O: 3, isolated from human and animal clinical material, and 35 were of biotype 1 A or 2 and of diverse serotypes, isolated from food in Brazil between 1968 and 2000. Drug resistance was also investigated. All the strains were resistant to three or more drugs. The isolates showed a virulence-related phenotype in the aesculin, pyrazinamidase and salicin tests, except for the food isolates, only two of which were positive for these tests. For the other phenotypic virulence determinants (autoagglutination, Ca++ dependence and Congo red absorption), the strains showed a diverse behaviour. The inv, ail and ystA genes were detected in all human and animal strains, while all the food isolates were positive for inv, and 3% of them positive for ail and ystA. The presence of virF was variable in the three groups of strains. The strains were better discriminated by PFGE than by enterobacterial repetitive intergenic consensus PCR (ERIC-PCR). A higher genomic similarity was observed among the 4/O: 3 strains, isolated from human and animal isolates, than among the food strains, with the exception of two food strains possessing the virulence genes and grouped close to the 4/O: 3 strains by ERIC-PCR. Unusually, the results revealed the virulence potential of a bio-serotype 1 A/O: 10 strain, suggesting that food contaminated with Y. enterocolitica biotype 1 A may cause infection. This also suggests that ERIC-PCR may be used as a tool to reveal clues about the virulence potential of Y. enterocolitica strains. Furthermore, the results also support the hypothesis that animals may act as reservoirs of Y. enterocolitica for human infections in Brazil, an epidemiological aspect that has not been investigated in this country, confirming data from other parts of the world.

EVALUATION OF DIFFERENT TECHNIQUES FOR THE DIFFERENTIATION OF PATHOGENIC AND NON PATHOGENIC STRAINS OF YERSINIA-ENTEROCOLITICA

Different methods and tests have been used to evaluate the pathogenic potential of distinct Y. enterocolitica serotypes and biotypes. We tested a total of 60 Y. enterocolitica strains, being 25 of human origin (serotype O3 biotype 4 and serotype O5 biotype 1); 6 of animal origin (serotype O3 biotype 4); 19 isolated from the environment (serotype O5.27 biotypes 1 and 2); and 8 isolated from food (serotype O5 biotype 1 and serotype 05.27 biotype 1). The methods used were based on plasmid gene expression (autoagglutination, calcium-dependence at 37 degrees C and Congo Red absorption tests), chromosomal gene expression (assays for pyrazinamidase activity, salicin fermentation and esculin hydrolysis), and invasion of HEp-2 cells. All but one of the Y. enterocolitica O3 strains, were found to be potentially pathogenic when submitted to the pyrazinamidase-salicin-esculin tests. In contrast, the results obtained with the assays related to plasmidial gene expression were not so uniform, probably due to plasmid loss. The least homogeneous results were obtained with the HEp-2 cell invasion test. Y. enterocolitica O5 behaved in a uniform manner when tested with the first two groups of tests (based on chromosomal and plasmidial gene expression), but not when tested with the HEp-2 invasion assay. The strains of serotype O5.27 biotype 1 presented a uniform behavior hen submitted to the chromosomic-related tests, showing no pathogenicity. However, they did not provide conclusive results with the tests related to plasmidial gene expression or HEp-2 cell invasion. We conclude that the tests related to chromosomal gene expression (esculin-salicin-pyrazinamidase) are simple and highly effective for the detection of potentially pathogenic Y. enterocolitica isolated from clinical cases.