ABSTRACT
The aim of this study was to identify the strengths and weaknesses of five DNA fingerprint methods for epidemiological typing of Streptococcus pneumoniae. We investigated the usefulness of (i) ribotyping, (ii) BOX fingerprinting with the BOX repetitive sequence of S. pneumoniae as a DNA probe, (iii) PCR fingerprinting with a primer homologous to the enterobacterial repetitive intergenic consensus sequence, (iv) pulsed-field gel electrophoresis of large DNA fragments, and (v) restriction fragment end labeling to detect restriction fragment length polymorphism of small DNA fragments. Twenty-eight S. pneumoniae strains isolated from the blood and/or cerebrospinal fluid of 21 patients were analyzed. Genetic clustering among the 28 strains was independent of the DNA fingerprint technique used. However, the discriminatory power and the similarity values differed significantly among the individual techniques. BOX fingerprinting, pulsed-field gel electrophoresis, and restriction fragment end labeling provided the highest degree of discriminatory power. Furthermore, the ease with which computerized fingerprint analysis could be conducted also varied significantly among the techniques. Ribotyping, BOX fingerprinting, and restriction fragment end labeling were very suitable techniques for accurate computerized data analysis. Because of their high discriminatory potential and ease of accurate analysis, we conclude that BOX fingerprinting and restriction fragment end labeling are the most suitable techniques to type pneumococcal strains.
