Combination of pulsed-field gel electrophoresis (PFGE) and single-enzyme amplified fragment length polymorphism (SAFLP) for differentiation of multiresistant Salmonella enterica serotype typhimurium

Objective   To assess the combined application of plasmid profile typing, pulsed-field gel electrophoresis (PFGE) and PCR-based single-enzyme amplified fragment length polymorphism ( S AFLP) for the differentiation of 18 multiresistant (MR) and one drug-sensitive strain of Salmonella enterica serotype typhimurium from humans and food animals.
Methods   Strains were phage typed and tested for resistance to a panel of antimicrobial agents. Strains were also tested for the ability to transfer resistance either directly or by mobilization to standard strains of Escherichia coli K12. Plasmid DNA was extracted from both drug-resistant donor strains and from drug-resistant exconjugants. Total genomic DNA was characterized by PFGE following digestion with the restriction endonuclease Xba I. The resultant patterns were categorized and analyzed by dendrogram analysis using the Dice coefficient and by data clustering using unweighted pair-group arithmetic averaging (UPGMA). Isolates were also characterized and categorized by SAFLP. The levels of discrimination achieved by each method were assessed individually and in combination.
Results   Plasmid DNA was detected in all of the 18 MR isolates but, not in the drug-sensitive isolate. Using PFGE, 19 different profiles were identified, falling into eight major categories. However, by SAFLP, only eight profiles were observed. Subsequent investigations have demonstrated epidemiologic relationships within at least one of these SAFLP profile groupings.
Conclusions   These studies have demonstrated that PFGE and SAFLP can be used independently for the differentiation of MR S. Typhimurium from humans and food animals. However, when used in combination, SAFLP can provide a format for broad epidemiologic groupings. These groupings can be further subdivided by PFGE to provide detailed information on putative strain relationships at the genotypic level.