Strains of species in the Bacillus cereus group are potentially enterotoxic. Thus, the detection of all B. cereus group strains is important. As 16S ribosomal DNA sequence analysis cannot adequately differentiate species of the B. cereus group, we explored the potential of the groEL gene as a phylogenetic marker. A phylogenetic analysis of the groEL sequences of 78 B. cereus group strains revealed that the B. cereus group strains were split into two major clusters, one including six B. mycoides and one B. pseudomycoides (cluster II) and the other including two B. mycoides and the rest of the B. cereus group strains (cluster I). Cluster I was further differentiated into two subclusters, Ia and Ib. The sodA gene sequences of representative strains from different clusters were also compared. The phylogenetic tree constructed from the sodA sequences showed substantial similarity to the tree constructed from the groEL sequences. Based on the groEL sequences, a PCR assay for detection and identification of B. cereus group strains was developed. Subsequent restriction fragment length polymorphism (RFLP) analysis verified the PCR amplicons and the differentiation of the B. cereus group strains. RFLP with MboI was identical for all the B. cereus group strains analyzed, while RFLP with MfeI or PstI classified all B. cereus and B. thuringiensis strains into two groups. All cluster II B. mycoides and B. pseudomycoides strains could be discriminated from other B. cereus group bacteria by restriction analysis with TspRI.

A I O . 2000. PCR±RFLP analysis of the vrrA gene and cerAB gene was used to investigate the genomic diversity in 21 strains of Bacillus anthracis and 28 strains of Bacillus cereus, and was compared with results obtained by ribotyping and enterobacterial repetitive intergenic consensus±PCR (ERIC±PCR) analysis. VrrA-typing divided the B. anthracis into four groups. Except for one Pasteur vaccine strain, the vrrA PCR±RFLP pro®les of the B. anthracis were separated into three groups, which were different from those of the B. cereus strains. Ribotyping separated the B. anthracis isolates into seven ribotypes, and a common fragment of an approximately 850 bp band from the ERIC±PCR ®ngerprints separated most B. anthracis strains into two groups. VrrA/cerAB PCR±RFLP, ribotyping and ERIC±PCR generated 18, 22 and 23 types, respectively, from B. cereus strains. The results suggest that a combination of all three methods provides a high resolution typing method for B. anthracis and B. cereus. Compared with ribotyping and ERIC±PCR, PCR±RFLP is simple to perform and has potential as a rapid method for typing and discriminating B. anthracis strains from other B. cereus group bacteria.

A Random Amplified Polymorphic DNA (RAPD) fingerprinting method was developed to differentiate isolates of Salmonella serotype typhi (S. typhi) and other Salmonella isolates. A panel of five primers was used to examine 63 isolates of Salm. typhi, including 56 strains isolated in Taiwan and seven strains obtained abroad. Twenty‐one RAPD types were revealed using the RAPD fingerprinting method. An RAPD with primer 6032 yielded a polymorphism in a 350 bp fragment that differentiated the attenuated vaccine strain Salm. typhi Ty21a from the rest of the Salm. typhi strains. Strains of Salm. typhi were divided into five types with primer D14307. Primer D14307 also proved capable of discrimination among 65 other Salmonella isolates representing 42 different serotypes. The bacterial DNA used in this RAPD protocol was obtained using a commercially available DNA extraction kit (GeneReleaser). The DNA of various strains of Salmonella from this simple extraction procedure could be discriminated within a few hours using the RAPD technique.

No abstract

In January and February of 1981, 85 cases of enteritis caused by Salmonella muenchen were reported from Ohio, Michigan, Georgia, and Alabama. Initial investigation failed to implicate a food source as a common vehicle, but in Michigan 76 per cent of the patients, in contrast to 21 per cent of the control subjects, admitted personal or household exposure to marijuana (P less than 0.001, relative risk = 20). Marijuana samples obtained from patients' households contained as many as 10(7) S. muenchen per gram. The outbreak-related isolates of S. muenchen were sensitive to all antibiotics and were phenotypically indistinguishable from other S. muenchen. Plasmid fingerprinting, however, revealed that all isolates related to marijuana exposure contained two low-molecular-weight plasmids (3.1 and 7.4 megadaltons), which were absent in control strains. Plasmid analysis of the isolates showed that the outbreaks in Ohio, Michigan, Georgia, and Alabama were related, and analysis of isolates submitted from various other states demonstrated that cases associated with marijuana may have been dispersed as far as California and Massachusetts.