Thyrotrophin (TSH)-binding proteins in bacteria and their cross-reaction with autoantibodies against the human TSH receptor.

A screen of a range of bacteria normally found in gut flora identified eight with the ability to bind TSH specifically. These included the previously reported Yersinia enterocolitica, Gram-positive, Gram-negative, pathogenic and commensal organisms. Eleven preparations of TSH-receptor autoantibodies strongly able to displace 125I-labelled TSH from the mammalian TSH receptor differed in their ability to displace the tracer from binding to bacterial extracts. None could displace the tracer from E. coli 06-1, four displaced 125I-labelled TSH from E. coli V21/1 and five displaced the tracer from Y. enterocolitica. Of those immunoglobulin preparations which did react with the bacterial protein, their apparent potency compared with that of TSH in displacing tracer from bacterial binders was an order of magnitude greater than with the mammalian receptor. This is consistent with the autoantibodies having a relatively better fit with the bacterial antigen than with the receptor when compared with TSH. The bacterial-binding activity and mammalian receptor-binding activities in each of two samples co-chromatographed on a Remazol yellow GGL-Sepharose affinity column strongly indicated that the same immunoglobulin species reacts with both antigens. These results are consistent with the proposal that a bacterial protein is the primary immunogen for the TSH-receptor antibodies in at least some patients with Graves' disease.

Mucosal association by Clostridium difficile in the hamster gastrointestinal tract.

For many organisms, mucosal association is an important virulence determinant. Although studied in detail for other intestinal pathogens, this aspect of pathogenicity has not been studied for Clostridium difficile. We compared the ability of an avirulent non-toxigenic strain (M-1), a highly virulent toxigenic strain (B-1), and a poorly virulent toxigenic strain (BAT) of C. difficile to adhere to different regions of the gastrointestinal tract of hamsters pre-treated with clindamycin. Strain B-1 associated with the gut mucosa significantly better than strain M-1 (p less than 0.001) for all sites other than the caecum, and achieved significantly higher levels in the caecal contents (p less than 0.001). The same was true when strain B-1 was compared with strain BAT except that there was no significant difference for the large bowel mucosa. To assess the possible role of toxin in promoting mucosal association, e.g., by compromising host defences or exposing masked adherence sites, strain M-1 was given to animals after intra-caecal administration of crude toxin preparations from strain-B1, which were heat-inactivated in control experiments. The addition of this toxin increased significantly the mucosal association of M-1 for the small bowel only, whereas the inactivated toxin had no significant effect. These results imply that there may be intrinsic differences between strains in their ability to colonise and associate with the gut mucosa, which may partly depend on their ability to produce toxin. These differences do not correlate with cell-surface hydrophobicity or the presence of plasmids, flagella or fimbriae.

Clostridium difficile--a spectrum of virulence and analysis of putative virulence determinants in the hamster model of antibiotic-associated colitis.

Each of nine different toxigenic strains of Clostridium difficile was administered orally to groups of hamsters pre-treated with clindamycin and housed individually in sterile isolator boxes. Faecal pellets and caecal contents from well, diarrhoeic, moribund and freshly dead animals were analysed for C. difficile and toxins A (enterotoxin) and B (cytotoxin), and tissue obtained when animals were killed was examined histologically. Not all strains were equally virulent in this model. Four strains of C. difficile killed all animals within 48 h and are designated as highly virulent for hamsters. These strains were clinical isolates from three cases of disease in man and one case in a hamster. Five strains caused death of some animals but only after 5 and upt to 13 days and are designated as less virulent for hamsters. These strains were isolated from asymptomatic infants (2) and household pets (2), and from the environment (1). The surviving test hamsters were killed after 14 days and, in most cases, were colonised by C. difficile, though levels of toxins A and B in caecal contents were low. None of the cultures used for challenge was capsulate or hydrophobic. There was no correlation between virulence and production of toxins A and B in vitro in tryptic-nitrate broth. With two strains examined, there was a correlation between virulence and toxin A (but not toxin B) production in caecal emulsions derived from clindamycin pre-treated hamsters. Caecal contents from the majority of moribund and freshly dead animals had quantities of toxin A sufficient to cause disease or death if given orogastrically. Toxin B was not produced in a fixed ratio with toxin A. The data support the view that high virulence of C. difficile is determined by efficient disease-inducing colonisation of the gut and the ability to generate, rapidly, high levels of toxin A in vivo.

Analysis of latex agglutination test for Clostridium difficile toxin A (D-1) and differentiation between C difficile toxins A and B and latex reactive protein.

Virulent toxigenic and avirulent non-toxigenic strains of Clostridium difficile gave a positive result in the latex agglutination test (LAT) for C difficile toxin A (D-1). Similar concentrations of latex agglutinating antigen were produced by these strains in vivo. Positive reactions were also given by C sporogenes, proteolytic C botulinum Types A, B, and A/F, and Bacteroides assaccharolyticus. The latex agglutinating antigen was denatured by boiling for 10 minutes, but not by heating at 56 degrees C for 30 minutes. The reaction was abolished by incubation of test material with crude C difficile antitoxin but not with other clostridial antitoxins or specific antitoxin to C difficile toxin A. The latex agglutinating antigen present in C difficile eluted between 0.39% and 0.47% M sodium chloride, and that produced by the other clostridia, between 0.35% and 0.43% M sodium chloride by fast protein liquid chromatography. The latex agglutinating antigen of C difficile was neither cytotoxic nor mouse lethal and was distinct from toxin A and toxin B. In the analysis of faecal specimens from patients with diarrhoea the latex agglutination test correlated better with the presence of C difficile than with toxin B and detected both toxigenic and non-toxigenic strains. The latex agglutination test should only be used in the laboratory as an alternative to culture for C difficile and not as a method for the detection of C difficile toxins.

Asymptomatic carriage of Clostridium difficile in patients with cystic fibrosis.

Faecal samples from 37 patients with cystic fibrosis and 40 control patients at the Brompton Hospital and the London Chest Hospital were examined for the presence of Clostridium difficile. The organism was isolated from 2 (17%) of control patients who were receiving antibiotics and from one (3.6%) of control patients who had no antimicrobial treatment. Thirty two per cent of the patients with cystic fibrosis excreted C difficile, though none of them had diarrhoea. Two of the three isolates from control patients and nine of the 12 isolates from patients with cystic fibrosis produced toxin B (cytotoxin) in vitro. Toxin B was present in the stools of one of the control patients and three of the patients with cystic fibrosis; toxin A (enterotoxin) was not detected in the faeces of the patients with cystic fibrosis. Two cytotoxigenic strains of C difficile isolated from patients with cystic fibrosis were examined in hamsters; both were virulent, and the animals died.

An in-vitro model of colonisation resistance to Clostridium difficile infection.

To investigate the importance of the normal gut flora in preventing the establishment of Clostridium difficile in vivo we have developed an in-vitro test system based on growth in faecal emulsions. Growth of C. difficile and cytotoxin production are inhibited in faecal emulsions from healthy adults, but not in sterilised emulsions; the importance of viable bacteria in the inhibitory system is evident. Generally, faecal emulsions derived from infants, children and geriatric patients were less inhibitory than those from healthy adults. Those from bottle-fed infants were significantly less inhibitory than those from breast-fed infants. Decreased levels of cytotoxin in the latter group were attributed to the acidic pH of the stools. With the different patient groups studied, faecal samples not inhibitory to C. difficile in vitro were obtained from 21% of patients with antibiotic-associated diarrhoea, 33% of those taking antibiotics but who did not have diarrhoea, 18.7% of those with diarrhoea unassociated with antibiotics, and 79% of those with C. difficile-mediated diarrhoea. In some cases inhibition was due to low faecal pH, as in some infants, and in others to other filterable substances. The degree of inhibition could not be linked to specific volatile fatty acids or enzymes.

Epidemiology of diarrhoea caused by enterotoxigenic Clostridium perfringens.

Enterotoxigenic strains of Clostridium perfringens have recently been implicated in some cases of antibiotic-associated diarrhoea. We present here the results of an epidemiological study of this disease. Five cases of diarrhoea caused by C. perfringens serotype 41 occurred during a 9-week period, and then during a 6-week period there were three cases due to serotype 27 and two due to serotype 24; in all but one case two geriatric wards were involved. In total there were 16 cases in 22 months. All cases were identified by the detection of C. perfringens enterotoxin in the faeces. The mean number of C. perfringens in these cases was 10(8.8) cfu/g of faeces. Of 37 patients who had negative test results for C. perfringens enterotoxin, 18 had positive cultures for C. perfringens, with mean faecal counts of 10(5.3) cfu/g, and nine of these patients had diarrhoea. Thirteen different serotypes were isolated from these 18 patients, including type 41 from seven patients and type 27 from one. Hand carriage of the offending serotype was demonstrated in three of four infected patients, none of four controls and two of 14 ward staff. C. perfringens of serotypes causing disease was isolated from 59% of environmental areas where there was active disease, 27% of areas where there had been disease which had since resolved and 9% of areas where there was no history of disease. The findings imply that cross infection may occur.

Microbial and metabolic profile of achlorhydric stomach: comparison of pernicious anaemia and hypogammaglobulinaemia.

The microbial flora and some of its metabolites and enzymes in the stomach were compared in patients with achlorhydria, pernicious anaemia, and primary hypogammaglobulinaemia and in patients with dyspepsia with normal gastric acidity. Detailed analysis of the flora of the gastric juice and of the mucosa from the antrum, body, and fundus in six patients with hypogammaglobulinaemia (mean pH 8.2), seven patients with pernicious anaemia (mean pH 7.3), and five patients with dyspepsia (mean pH 1.9) yielded 22 different genera of bacteria, mainly from the patients with achlorhydria, the most common being streptococci, micrococci, staphylococci, veillonella, and lactobacilli. A similar flora was found associated with the mucosa at all three sites. Various metabolites were also looked for. beta Glucoronidase and C14 lipase were found in patients with hypogammaglobulinaemia but not in those with pernicious anaemia or dyspepsia. Volatile fatty acids were not found. Relatively high concentrations of ethanol were found in the patients with hypogammaglobulinaemia compared with those with pernicious anaemia (p = 0.02). Similar concentrations of dimethylamine were found in all three groups, but the concentrations of trimethylamine were much higher in patients with pernicious anaemia and hypogammaglobulinaemia. The high concentrations of some microbial enzymes and ethanol differentiated the group with hypogammaglobulinaemia from the rest, and these may bear some relation to the high incidence of gastric cancer in patients with hypogammaglobulinaemia.

Protection of hamsters against Clostridium difficile ileocaecitis by prior colonisation with non-pathogenic strains.

Prior colonisation of clindamycin-treated hamsters with non-toxigenic strains of C. difficile protected them from subsequent colonisation with a toxigenic pathogenic strain. In total, 13 of 18 'protected' hamsters survived for up to 27 days whereas all 27 animals challenged with the toxigenic strain alone died within 48 h. Protection was not evident if a heat-killed suspension was used or if the colonising non-toxigenic strain was first removed with vancomycin. No antitoxic activity could be detected in the faeces of animals colonised with the non-toxigenic strains. Other species of clostridia did not protect against the lethal effects of subsequent exposure to the toxigenic strain. Conversely, non-toxigenic strains would not protect the animals from the lethal effects of a different clostridial pathogen, C. spiroforme. In most cases, even in the protected animals, the toxigenic strain eventually became dominant and caused disease, with translocation across the gut wall occurring early in the disease process. It was also shown that a non-toxigenic strain of C. difficile can adhere to gut mucosa. It is proposed that the protection afforded by the non-toxigenic strains may be due to competition for ecological niches.

Bacterial colonization and nitrite concentration in the achlorhydric stomachs of patients with primary hypogammaglobulinaemia or classical pernicious anaemia.

The degree and type of bacterial colonization was similar in achlorhydric patients with either severe primary hypogammaglobulinaemia or classical pernicious anaemia. This suggests that antibodies do not play a significant part in controlling the gastric flora in achlorhydric patients. The nitrite concentration in gastric juice was higher in the hypogammaglobulinaemia patients, raising the possibility that these patients may have very high levels of nitrite-derived mutagenic compounds in their gastric juice. This may account for the high incidence of gastric cancer in these patients.

Epidemiology of Clostridium difficile in infants.

The epidemiology of Clostridium difficile was studied prospectively in 451 newborn infants by daily screening of fecal samples. Colonization rates in three postnatal wards ranged from 2% to 52%. Many colonizations were sporadic, but on two wards there was evidence of clustering. On one of these occasions prospective environmental sampling yielded C. difficile organisms from a potential common source. Mothers were shown not to be the sources of their infants' organisms. Both toxin-producing and non-toxigenic strains were common; differentiation according to toxin type was epidemiologically useful. Cross contamination is the most likely explanation of the spread of C. difficile among hospitalized infants; the organism could spread among adults who are at risk of developing antibiotic-associated colitis in a similar manner.