Evaluation of the taxonomic utility of six-enzyme pulsed-field gel electrophoresis in reconstructing Salmonella subspecies phylogeny.

Pulsed-field gel electrophoresis (PFGE) remains an important tool in the molecular epidemiological evaluation of strains emerging from disease outbreak clusters. Recent studies of Escherichia coli O157:H7 and Salmonella Enteritidis have noted marked improvements in the discriminatory power of PFGE when combining band profiles from up to six restriction enzyme datasets into a single concatenated analysis. This approach has provided more accurate assignments of genetic relationships among closely related strains and allowed effective phylogenetic inference of host and geographical reservoirs. Although this approach enhances epidemiological congruence among closely related strains, it remains unclear to what extent six-enzyme PFGE pattern similarity reiterates evolutionary relatedness among more distantly related Salmonella strains (i.e., serovar or subspecies levels). Here, taxonomic accuracy of six-enzyme PFGE is tested phylogenetically across two distinct Salmonella enterica populations-Salmonella reference collection B (SARB), representing the breadth of taxonomic diversity of S. enterica subspecies I only, and Salmonella reference collection C (SARC), comprising the seven disparate subspecies of S. enterica plus S. bongori. Cladistic analysis of SAR strains revealed substantial polyphyly between the two strain collections such that numerous SARB strains clustered more closely with diverged SARC subspecies rather than with other members of subspecies I. Also, in several cases, SARC sibling strains from the same subspecies were evolutionary obscured-broken into distant locales on the most parsimonious six-enzyme trees. Genetic diversity among SARB and SARC strains was comparable at 45% and 47%, respectively, while congruence testing revealed discordance among individual enzyme datasets. While six-enzyme PFGE is effective in ascertaining accurate genetic relationships for more closely related strains (e.g., strains within the same serovar), reconstitution of evolutionarily meaningful strain groupings may be elusive for Salmonella at the serovar level and above. Thus, caution is warranted when applying PFGE with a limited number of enzymes as the primary phylogenetic marker in these instances. Published by Elsevier B.V.