Summary. Most toxigenic strains of Clostridium di#iciie produce two toxins : an enterotoxin (toxin A) and a cytotoxin (toxin B). Only one strain (strain 8864) has been reported to produce toxin B but no toxin A. Serogroup F strains (44) of C. dzficile, often isolated from asymptomatic infants, have been examined for toxin production. These strains, which were from distinct geographical and clinical sources, did not produce any detectable toxin A in citro when examined in three distinct immunoassays. Nevertheless, all the strains produced a cytotoxin. Immunological differences between the cytotoxin of the serogroup F strains and that produced by C. dz3ciie strain VPI 10463 (serogroup G) were demonstrated with monoclonal antibodies specific for either the toxin B produced by C . dificiie strain VPI 10463 or C. sordellii lethal toxin (LT). Polymerase chain reaction amplification with primers derived from C. dtficile strain VPI 10463 toxin A and B genes showed that serogroup F strains seem to possess a toxin B gene homologous with that of strain VPI 10463 and at least fragments of the toxin A gene. When axenic mice were inoculated with serogroup F strains, the animals survived; they did not develop diarrhoea and no toxin A could be detected in their faeces. However, cytotoxin was detected. Furthermore, these mice were protected against subsequent challenge with the otherwise lethally toxigenic C . dzflcile strain VPI 10463. The serogroup F strains appeared to be homogeneous and distinct from other C. diflcile strains with regard to toxin production.

This study was undertaken to examine toxin production by Cltidium diWjcile 8864, a naturally occurring isolate that has been reported to produce toxin B in the absence of toxin A. To date, this is the only strain of C. difficile reported to produce only one of the toxins. The results of our initial studies with antibodies against toxins A and B confirmed these observations. Toxin B from strain 8864 and from VPI strain 10463, a strain that produces high levels of both toxin A and toxin B, was purified to homogeneity by sequential anion-exchange chromatography on DEAE-Sepharose CL-6B, gel filtration on Ultrogel AcA22, and immunoadsorption chromatography, and their toxic activities were compared. Our results showed that toxin B from strain 8864 and toxin B from C. dificile VPI strain 10463 were comparable in their cytotoxic activities and that the 8864 toxin B was more lethal. In addition, we observed that toxin B from strain 8864 was weakly enterotoxic, which may explain the ability of this strain to cause intestinal disease in hamsters treated with antibiotics. Analysis with specific antibodies showed that the toxin B molecules from these strains were highly related but contained distinct epitopes. The results of hybridization studies with probes specific for the toxin B gene of VPI strain 10463 demonstrated differences between the toxin B genes of the two strains. In addition, probes specific for the toxin A gene of VPI strain 10463 showed that strain 8864 contains a region which shows identity with the S' end of the toxin A gene but not the region of the gene which encodes a hydrophobic region and the repeating units.

In order to improve our understanding of the role of Clostridium difficile in infants we characterised the strains isolated from this population. The production of toxin A and toxin B was studied. The toxin A, playing a major role in the disease, was searched for in faecal samples. The serogroup of the isolates was determined because some serogroups have been shown to be more pathogenic than others. Over a 9-month period, 102 faecal samples from 102 hospitalised infants (0-12 months) were analysed and 26% of the children were colonised with C. difficile. Fifteen isolates secreted neither toxin A nor B (62.5%). Nine isolates were toxigenic and secreted both toxins (37.5%). Of the eight toxigenic strains tested, six were from serogroup H and two serogroup K. Of the 13 nontoxigenic strains tested, 8 belonged to serogroup D, 2 to serogroup X, and 1 each to serogroup A, serogroup B and serogroup C. Three infants out of 102 studied had toxin A in their faeces. In summary, the infants can be colonised by (1) nontoxigenic strains, most of them from nonpathogenic serogroup D, without toxin A in the faeces; (2) toxigenic strains of virulent serogroups H and K, with or without toxin A in the faeces. Although some infants had diarrhoea, none needed a specific treatment for C. difficile. No specific C. difficile pathology could be retained and different mechanisms are advanced to explain this absence of pathogenicity.

Three groups of six healthy adult volunteers were randomly assigned to a treatment with 400 mg of oral cefpodoxime proxetil, oral cefixime, or placebo per day for 10 days. Informed consent was obtained from all volunteers. Clostridium difficile was not detected in the feces of any subject before treatment or at any time in the subjects in the placebo group. C. difficile was, however, detected in all subjects treated with cefpodoxime proxetil and in five of six treated with cefixime. Genomic DNA restriction patterns showed that the strains of C. difficile differed from one volunteer to another. Two subjects both shed different strains at different times during the 25-day surveillance period. All isolates were resistant to cefixime and cefpodoxime (MIC for 90% of strains, 256 and 512 mg/liter, respectively). Antibiotic activity was found in the feces of one volunteer treated with cefpodoxime proxetil and of four volunteers treated with cefixime. It was inversely correlated with the presence of fecal beta-lactamase activity. Intestinal side effects were limited to modifications of stool consistency, which occurred in only 3 of the 12 treated volunteers and did not lead to cessation of treatment. These modifications were significantly associated with the presence of fecal antibiotic activity (P less than 0.05) but not with the shedding of toxigenic or nontoxigenic strains of C. difficile or with the presence of toxin A in feces, which was detected only in one perfectly healthy treated volunteer.

In an attempt to understand more completely why patients treated with phenothiazines (chlorpromazine and cyamemazine), methotrexate, and certain antibiotics such as clindamycin have an increased risk of developing pseudomembranous colitis, the production of toxins A and B by Clostridium difficile in the presence of these drugs was measured in vitro as well as in vivo by using axenic mice. None of the drugs tested increased the production of toxins either in vitro or in vivo.