Discriminatory power of three DNA-based typing techniques for Pseudomonas aeruginosa

We assessed the capacity of three DNA typing techniques to discriminate between 81 geographically, temporally, and epidemiologically unrelated strains of Pseudomonas aeruginosa. The methods, representing powerful tools for hospital molecular epidemiology, included hybridization of restricted chromosomal DNA with toxA and genes coding for rRNA (rDNA) used as probes and macrorestriction analysis of SpeI-digested DNA by pulsed-field gel electrophoresis. The probe typing techniques were able to classify all strains into a limited number of types, and the discriminatory powers were 97.7 and 95.6% for toxA and rDNA typing, respectively. Strains that were indistinguishable on the basis of both toxA and rDNA types defined 12 probe type homology groups. Of these, one contained five strains, three contained three strains each, and eight groups were represented by two strains each. Strains in 10 of the homology groups had the same O serotype. SpeI macrorestriction patterns discriminated between all strains with at least four band differences, which corresponded to a similarity level of 85%. Fifteen pairs of strains were similar at a level of > 75% and differed by only four to seven bands. Of these pairs, 11 belonged to the same probe type homology group, indicating their clonal relatedness. We conclude that macrorestriction analysis of P. aeruginosa with SpeI provides the best means of discrimination between epidemiologically unrelated strains. However, DNA probe typing with either toxA or rDNA reveals information on the strain population structure and evolutionary relationships.