New phenotypic typing scheme for group B streptococci.

AIMS: To develop a new typing system for group B streptococci based on 35S-methionine-labelled protein profiles of bacterial proteins. METHODS: 377 clinical isolates of group B streptococci were examined by incorporation of 35S-methionine into bacterial proteins under strict anaerobic conditions. After sodium dodecylsulphate-polyacrylamide gel electrophoresis, autoradiography was performed. The patterns produced were visually analysed and categorised into clusters of organisms based on the pattern of band production between 32-46 kilodaltons. RESULTS: 294 of the typed strains classified into seven different groups designated a-g. 32 strains failed to incorporate 35S-methionine sufficiently to be grouped and 11 strains did not fall into one of the seven identified groups. Typability, reproducibility, and discrimination of the system was evident. CONCLUSIONS: This typing system may help to distinguish between colonising and invasive strains of the organism.

Detection of the C protein gene among group B streptococci using PCR.

AIM: To develop a polymerase chain reaction (PCR) for the specific detection of the C protein gene in strains of group B Streptococcus. METHODS: A single primer pair derived from the nucleotide sequence of the IgA binding beta antigen of the C protein complex permitted the specific amplification of a 592 base pair DNA fragment from the C protein gene. After 35 cycles of amplification this product could be detected by agarose gel electrophoresis. Southern blot hybridisation confirmed that this product was the C protein gene. RESULTS: PCR detected the C protein gene in 75 (63%) of 119 strains of group B streptococci analysed. The product was not detected in other Gram positive organisms, showing that this PCR assay was highly specific. The sensitivity of the assay was satisfactory to a dilution of 1 in 10,000 of extracted DNA. CONCLUSIONS: The C protein of group B streptococci is associated with neonatal sepsis. The specific detection of the C protein gene by PCR may help identify which strains are likely to be associated with infection by the organism.

Assessment of two methods for rapid intrapartum detection of vaginal group B streptococcal colonisation.

AIMS: To compare two methods for the rapid detection of intrapartum vaginal carriage of group B streptococci (Streptococcus agalactiae) with standard culture techniques and to establish their suitability for routine use. METHODS: Vaginal swabs from 266 patients in labour were incubated in glucose broth in an anaerobic atmosphere for four to six hours. The Wellcogen Strep B latex particle agglutination test kit was subsequently used for antigen detection. In the second part of the study swabs from 117 women were assessed for the presence of group B streptococci using the ICON STREP B immuno-concentration assay (Hybritech). Both methods were compared with standard semiquantitative culture on Columbia horse blood agar and Islam's medium. RESULTS: In the first study vaginal carriage of group B streptococci was shown in 38 of 266 (14.3%) patients by culture. Latex particle agglutination with the Wellcogen kit detected 30 of these positive results (sensitivity 78.9%, specificity 100%). In those patients with moderate to heavy colonisation (> 10(4) colony forming units per millilitre) antigen was detected in all (26/26) culture positive patients (sensitivity 100%, specificity 100%). In the second study 16 (13.7%) patients were culture positive. The ICON test detected 11 positive results (sensitivity 68.8%, specificity 100%) and for heavy colonisation (10(5) cfu/ml) detected nine of nine cases (sensitivity 100%, specificity 100%). The ICON test took 10 to 15 minutes to perform. CONCLUSION: These tests are potentially useful for the rapid detection of group B streptococci vaginal colonisation in labour, particularly heavy colonisation. Both tests are insufficiently sensitive to replace standard culture methods.

Differentiation of vaccine and wild mumps viruses using the polymerase chain reaction and dideoxynucleotide sequencing.

Parts of the F gene from 16 mumps viruses derived from vaccines and clinical isolates were amplified using the polymerase chain reaction and their nucleotide sequences were determined. Over a region of 111 nucleotides, eight regions of variability were detected with a maximum of six (5.4%) changes occurring between any two virus strains. The Jeryl Lynn and Urabe vaccine strains were clearly different from each other and from wild virus isolated from cases of non-vaccine-associated mumps. In contrast, viruses isolated from the cerebrospinal fluid and throat in cases of meningitis and parotitis following vaccination with the Urabe strain were identical to this strain. We conclude that the vaccine was the source of these infections.