The mutS gene product of Escherichia coli and SalmoneUa typhimurium is one of at least four proteins required for methyl-directed mismatch repair in these organisms. A functionaBy similar repair system in Streptococcus pneumoniae requires the hex genes. We have sequenced the S. typhimurium mutS gene, showing that it encodes a 96-kilodalton protein. Amino-terminal amino acid sequencing of purified S. typhimurium MutS protein confirmed the initial portion of the deduced amino acid sequence. The S. typhimurium MutS protein is homologous to the S. pneumoniae HexA protein, suggesting that they arose from a common ancestor before the gram-negative and gram-positive bacteria diverged. Overall, approximately 36% of the amino acids of the two proteins are identical when the sequences are optimally aligned, including regions of stronger homology which are of particular interest. One such region is close to the amino terminus. Another, located closer to the carboxy terminus, includes homology to a consensus sequence thought to be diagnostic of nucleotide-binding sites. A third one, adjacent to the second, is homologous to the consensus sequence for the helix-turn-helix motif found in many DNA-binding proteins. We found that the S. typhimurium MutS protein can substitute for the E. coli MutS protein in vitro as it can in vivo, but we have not yet been able to demonstrate a similar in vitro complementation by the S. pneumoniae HexA protein.

A colorimetric DNA hybridization assay has been developed for the rapid detection of Escherichia coli in foods. The method employs two oligonucleotide probes which are specific for the 16S ribosomal RNA of E. coli. Probes are added to lysates of test cultures and allowed to hybridize to target rRNA if present. The probe-target complex is captured via hybridization to a polystyrene dipstick. The immobilized target is detected using an antibody-horseradish peroxidase conjugate which binds to the immobilized probe-target complex. The probe-target-antibody complex generates a colorimetric signal when exposed to a substrate/chromogen mixture. A total of 233 E. coli isolates representing typical, toxigenic, invasive, hemorrhagic serotype 0157:H7, and other pathogenic strains all resulted in a positive assay signal. Dose-response experiments indicate the sensitivity of the assay is approximately 1 × 106 CFU/ml. Specificity of the assay was determined by testing 207 strains of non-E. coli species at 109 CFU/ml. All of the non-E. coli organisms tested were negative with the exception of Escherichia fergusonii and Shigella species. A total of 345 enriched samples including inoculated, uninoculated, and naturally-contaminated foods was tested for the presence of E. coli by the hybridization assay and a conventional cultural method. The false-negative rate for the hybridization assay was 1.2%. By comparison, the false-negative rate for the culture method in these studies was 23.4%. Based on these data, the DNA hybridization method is significantly more accurate than conventional methods for the detection of E. coli in foods.

No abstract