A severe outbreak of hemorrhagic colitis occurred in London, Ontario, during the month of September 1985. A total of 55 residents and 18 employees of a nursing home developed diarrhea, and 17 residents (age range, 78 to 99 years) died. Specimens from 38 patients, 37 employees and contacts, and 10 autopsies were investigated for all enteric pathogens. Specimens were also planted on MacConkey-sorbitol agar. Fecal extracts were tested on Vero cells for cytotoxin (FVT). Escherichia coli isolates were serotyped and tested for verotoxin and beta-glucuronidase production. Of the 38 symptomatic patients, 26 were positive for FVT, verotoxin-producing E. coli (VTEC), or both. Of the 105 specimens that were examined from these 38 patients, FVT and VTEC were both positive in 30 specimens, FVT only was positive in 13 specimens, and VTEC only was positive in 4 specimens. None of the 27 specimens from 10 autopsies was positive for FVT or VTEC. No other enteric pathogen was found in any of the cases. All asymptomatic individuals were negative for both FVT and VTEC. Of 19 VTEC strains that were isolated, 18 belonged to serotype 0157:H7. These 18 strains and 2 more strains that were obtained from sporadic cases that had occurred within the 2 previous months were found to give similar biochemical reactions in a 36-test identification system. All isolates of serotype 0157:H7 were beta-glucuronidase negative and susceptible to the antimicrobial agents that are used to treat E. coli infections. Testing for FVT and VTEC was found to be the most sensitive and specific technique for the laboratory diagnosis of this disease. Negative sorbitol, positive raffinose, and negative beta-glucuronidase tests appeared to be consistent markers for aiding in the detection of E. coli 0157:H7.

No abstract

Staphylococcus caprae is a coagulase-negative, DNase-positive member of the genus Staphylococcus usually associated with goats, but since 1991 a few laboratories have reported isolating the organism from human clinical specimens. We report on the isolation of 14 strains from human specimens and note that 10 strains were obtained from patients with bone and joint infections. Nine of the 10 infections started with traumatic fractures, and the other was a case of mastoiditis. Seven of these 10 infections were in patients with orthopedic prostheses, which appears to be a risk factor. Three of the 14 strains were from transplant patients. For three of the patients, S. caprae was the only organism isolated. S. caprae may be misidentified because it is not represented in the current MicroScan or Vitek identification systems which are in use in many laboratories, but the organism can be differentiated by a few biochemical tests. S. caprae produces positive results for DNase, pyrrolidonyl aminopeptidase, and acid production from mannitol and maltose; it produces negative results for ornithine decarboxylase and tube coagulase.

The fatty acid compositions of 39 type strains and 529 clinical or reference strains of pathogenic aerobic actinomycetes were analyzed after standardized culture by using the Microbial Identification System (MIS). Library entries for each type strain were created by using the MIS Library Generation Software, and the fatty acid profiles of clinical and reference strains were compared to these library entries. The bacteria separated into two large groups based upon major amounts of branched-chain or of saturated or monounsaturated straight-chain fatty acids. Identification of isolates was possible by using only the type strains for comparison, but fatty acid heterogeneity occurred within most species. The genera Gordona, Mycobacterium, Nocardia, Rhodococcus, Rothia, and Streptomyces, as well as strains of Amycolatopsis orientalis, Dermatophilus congolensis, Nocardiopsis dassonvillei, Pseudonocardia autotrophica, Saccharothrix aerocolonigenes, and Tsukamurella paurometabola, are aerobic actinomycetes that have been implicated in human infections (3, 29, 40). Prior attempts to improve the identification of these aerobic actinomycetes by cellular fatty acid analysis have used various culture media, growth conditions, incubation times, extraction methods, and chromatography techniques, all factors which affect the fatty acids expressed (9, 20, 22, 23, 26). The Microbial Identification System (MIS; Microbial Identification Inc., Newark, Del.) standardizes the chromatography variables associated with quantitative fatty acid analysis and facilitates fatty acid analysis as a means of identifying bacteria (53). We analyzed the type strains of 39 pathogenic aerobic actinomycete species after standardized growth and incubation and created library entries for each species. The fatty acid profiles of 529 clinical and reference strains were then compared to these library entries. Fatty acid analysis is a limited, but practical method for the rapid identification of aerobic actinomycetes, and our results are consistent with some prior claims for the existence of subgroups within defined taxa (8, 33, 39, 42, 48). MATERIALS AND METHODS Bacteria. The sources of the type strains, reference strains, and other strains that were used to establish additional taxa defined by fatty acid profiles are listed in Table 1. Other isolates were obtained from commercial culture collections, the collections of other researchers, and clinical specimens. Actinomadura madurae and Actinomadura pelletieri were excluded from this study due to their fastidious growth requirements and need for prolonged incubation. Culture conditions. All strains, with the exception of Nocardia brevicatena, were grown at 28°C for 96 h in ambient air on four TSBA plates composed of Trypticase soy broth (no. 11768; BBL) and 1.5% agar (no. 11849; BBL). Isolates of N. brevicatena were cultured on plates of TSBA with 1% Tween 80 (T164-500; Fisher Scientific), because we found a slight growth enhancement in the presence of Tween 80. Identification by biochemical tests. Clinical ...