- Volume 24, Issue 2, 1987

Introduction of the virulence plasmid, ColV,I-K94, into Escherichia coli strains led to increased sensitivity to erythromycin. This was the result of increased passage of antibiotic into ColV,I-K94+ organisms because the plasmid effect was abolished in bacteria which had been made permeable by chemical treatment. Full sensitivity in ColV+ strains appears to depend on the simultaneous presence of transfer and colicin components. Increased erythromycin sensitivity associated with the plasmid was demonstrated in organisms grown at 37°C; the sensitivity of ColVJ-K94+ organisms grown at 25°C was similar to that of the parent strain. Added Mg++ or Ca++ ions reversed erythromycin inhibition in strains with the basal level of sensitivity (i.e., the Col– parent grown at 25°C or 37°C or the ColV,I-K94+ derivative grown at 25°C) and in those with the plasmid-associated increase in sensitivity. Addition of phosphate or EDTA to broth increased erythromycin sensitivity in Col– and ColV,I-K94+ strains although the latter was affected most. Erythromycin was more inhibitory at pH 8·5 than at pH 7·4. This enhanced activity was more marked against the ColV,I-K94+ strain than against the Col– strain. The effects of growth in phosphate-containing medium and at alkaline pH were partially additive. We suggest that ColV, I-K94+ strains may be sensitive to erythromycin because ColV-specified proteins are extruded by a process of “self-promoted transfer” and that the effects of these proteins on the lipopolysaccharide component of the outer membrane facilitates antibiotic influx.

A method for typing species of Aeromonas by 35S-labelling and sodium-dodecyl sulphate polyacrylamide-gel electrophoresis (SDS-PAGE) of total soluble proteins is described. Radiolabelled-protein profiles from different strains were compared by use of the Dice coefficient. Typability was 100% and discrimination of pairs of differing strains was satisfactory with acceptable reproducibility. The results also yielded information on subgroups of phenotypic species.

Chemotaxis of polymorphonuclear leukocytes (PMNL) in response to cell components of Bacteroides fragilis alone or in combination with Escherichia coli was evaluated. E. coli produced much more powerful chemotactic factors than B. fragilis. The culture filtrate (CF), outer membrane (OM) preparation, and lipopolysaccharide (LPS) of B. fragilis slightly stimulated chemotactic activity of PMNL. The culture filtrate and OM preparation were capable of inhibiting the chemotaxis of PMNL in response to the chemotactic factors of E. coli but LPS of B. fragilis was not able to do so. Reduction by B. fragilis of PMNL chemotaxis in response to E. coli was not specific for B. fragilis but also occurred in the presence of facultative bacteria. In parallel with chemotaxis, lysozyme release, but not β-glucuronidase release, by PMNL was significantly stimulated by E. coli but not by B. fragilis.

Bacteroides antibodies were studied in sera from 74 patients infected with the fragilis group of Bacteroides and 74 healthy control persons, by immunofluorescence of 26 different serotypes of the fragilis group. Antibodies were present at titres of 10–320 in 65 (88%) patients and 50 (68%) controls (p < 0·01). Titres of ⩾80 were demonstrated in sera of 38 (51%) patients and 5 (7%) controls (p < 0·01). Specific IgM antibodies were detected in sera of 42 (57%) patients at a geometric mean titre (GMT) of 30, and 8(11%) controls at a GMT of 11 (p < 0·01). High antibody titres as well as specific IgM were found in 32 (43%) patients, while none of the controls showed such a combination (p < 0·01). The majority of positive patients’ sera (57%) reacted with five or more serotypes, whereas most positive control sera (51%) reacted against only one or two serotypes (p < 0·01). A selected combination of serotypes not reacting with the control sera showed positive reactions with 52 (70%) patients’ sera. These findings may be useful in devising schemes for the serodiagnosis of infection caused by the fragilis group of Bacteroides. However, there are indications of geographic variation in prevalence of serotypes, which may prevent the development of a single universal scheme.

Samples from deep (4–7 mm) periodontal pockets were collected from 17 patients with adult-type periodontal disease and one with the juvenile form of the disease. They were streaked immediately on selective and non-selective media and incubated anaerobically for 96 h. There was a heavy growth of Bacteroides spp. from most samples and 10 representative colonies from each sample were sub-cultured for identification. In a total of 149 isolates from patients with adult-type disease, the commonest species were B. oralis (40), B. asaccharolyticus (35), B. intermedius (31), B. fragilis (12) and B. ureolyticus (10); B. gingivalis was not detected. The distribution of species was not distorted by multiple identical isolates from individual patients. There was a heavy growth of a single species, B. ureolyticus, from the patient with juvenile-type disease.

DNA sequences corresponding to the 4·7-kb gentamicin, tobramycin and kanamycin resistance (Gmr Tmr Kmr) transposon Tn4001 have been detected on a series of nine structurally-related plasmids that mediate this phenotype in Australian isolates of Staphylococcus aureus. Tn4001 sequences have also been demonstrated on the chromosomes of GmrTmr Kmr isolates that do not possess these plasmids, and the exhibited diversity of chromosomal sites occupied by this element implies that Tn4001 has transposed to the chromosome on numerous occasions in vivo. These results suggest that the rapid emergence of nosocomial Gmr Tmr Kmr S. aureus in the early 1980s may have been the result of the transposition of Tn4001 from a chromosomal site to a readily disseminated plasmid.

Candida albicans cells were treated with alkali and acid to extract preferentially the cell wall α-mannan. Cells were recovered at three stages, as extraction proceeded from mild to more extensive: Alk-1, Alk-2 and Alk + Acid. Yeast adhesion to human epithelial cells was then examined with an in-vitro adherence assay. Yeasts from all three stages of extraction adhered in significantly lower numbers to buccal mucosal cells than did unextracted yeasts. Adhesion was as low for Alk-1 cells as for those submitted to more complete mannan extraction. When yeast cells from all three stages were treated with Concanavalin A (Con A), a lectin probe with strong affinity for yeast α-mannans, and then subjected to the adherence assay, there was no significant change in adhesion. When yeast agglutinability by Con A was examined in tests with treated and untreated yeast cells, abundant agglutination occurred only with the untreated cells. However, Alk-1 cells, though lacking in adhesive capacity towards mucosal cells, showed significant agglutination. The results suggest that candidal adhesion is mediated by an alkali-soluble, mannan-containing moiety(ies) which appears to be lost early in the extraction process. Blockage of this moiety by Con A inhibits the adhesion of unextracted cells. Extracted cells lack this moiety but still possess enough structural mannan for Con A recognition and agglutination.

Using an in-vitro adherence assay it was observed that the number of Candida albicans cells that attached to individual buccal mucosal cells varied greatly. Three mucosal-cell characteristics—state of aggregation, size and viability—that might influence yeast adhesion in vitro were studied. The number of attached yeast cells per mucosal cell varied from 0 to 32. The majority of buccal cells (88%) had none or very few yeasts attached, whereas a minority of cells (12%) bound more than one half of all the attached yeasts. In donors whose buccal cells had large numbers of attached yeasts this percentage increased and the number of cells with no attached yeast cells fell. Cells of an intermediate size (36–70μm) had a greater affinity for yeasts than did cells of other sizes. Buccal cell viability appeared not to be necessary for adhesion of yeasts. No significant differences were observed in the number of yeast cells attached to single buccal cells compared with attachment to buccal cells within sheets. It would appear, therefore, that there are distinct subpopulations of epithelial cells with high and low affinity for attachment by C. albicans in vitro. Mucosal cell size or viability might influence this affinity.

Human IgM, IgG and IgA responses after infection with Yersinia enterocolitica serovar O3 were studied by immunoblotting sera against whole-cell homogenates of a plasmid-containing strain of Y. enterocolitica O3 and a plasmid-free strain derived from it; each strain was grown in conditions expressive for the plasmid. The antibodies observed were directed against several plasmid-encoded polypeptides. The response against different bacterial components decreased uniformly with time and the persisting antibody production was directed against several epitopes. Strong reactions to the prominent plasmid-specified antigens of mol. wts (103) 26, 34, 45 and 52·5 were found more often with IgG-class antibodies than with IgM or IgA; the latter immunoglobulins recognised, respectively, antigens of mol. wt (103) 26 and 45 (IgM) and 26, 34 and 52·5 (IgA). Immunoblotting of sera from patients with yersinia-triggered reactive arthritis did not reveal any antigens that were involved additionally or specifically. However, IgA-mediated recognition of certain antigens of mol. wts (103) 26, 34 and 52·5 tended to persist longer in the arthritic patients.

Three strains of gram-negative bacteria—one each of Escherichia coli, Klebsiella pneumoniae and Enterobacter sp.—were treated with anti-lipopolysaccharide hyperimmune equine plasma (anti-LPS) or nun-immune control plasma and examined by scanning electronmicroscopy. Within a few minutes of treatment with anti-LPS, bacteria were agglutinated. Evidence of cell membrane destruction was observed shortly thereafter and total cell disintegration and disruption occurred within 1–2 h. In contrast, non-immune plasma had no effect on cell morphology. This confirms the findings in previous microbiological studies that specific antibodies in anti-LPS bind to lipopolysaccharide (LPS endotoxin), and thereby initiate the destruction of gram-negative bacteria.

Two chlamydial enzyme immunoassays, Chlamydiazyme® and IDEIA,® were evaluated in a mouse model of chlamydial genital-tract infection. The Chlamydiazyme assay and the IDEIA were assessed on specimens from 10 and 11 mice, respectively. The animals were infected with Chlamydia trachomatis, strain SA2 f, and the results obtained by these methods on vaginal specimens taken on 4 or 5 occasions during 41–42 days were compared with those obtained in cell culture and to a less extent by the MicroTrak® direct immunofluorescence test. In comparison with culture, the Chlamydiazyme assay had a sensitivity of 62% and a specificity of 92%; IDEIA had a sensitivity of 76% and a specificity of 94%. These assays sometimes did not detect chlamydiae in specimens taken immediately before specimens which proved positive by culture and the immunoassays were less sensitive if swabs were taken after those for culture. The IDEIA also failed to detect chlamydiae in the late stage of the murine infection when chlamydial elementary bodies were seen by immunofluorescence. The implications of the observations for investigations in the human field as well as for further studies in the mouse are discussed.

A strain of Proteus mirabilis associated with chronic urinary tract infection was found to produce an EDTA-sensitive IgA protease that cleaved the IgA heavy chain into two fragments at sites different from those attacked by other microbial IgA1 proteases. Another 14 P. mirabilis strains of diverse type and from various clinical conditions also produced a similar IgA protease. This enzyme may be a virulence determinant of P. mirabilis.