Deducing the Biological Relevance and Transfer of Antimicrobial Resistance Factors in Salmonella enterica serovar Typhimurium.

Abstract

Antimicrobial resistance in Salmonella can be encoded on plasmids or as stable chromosomal elements. To study the nature of resistance transfer by Salmonella enterica serovar Typhimurium strains of animal origin, we examined isolates obtained from pigs in commercial swine operations. We found that one common phage type, DT193, was most often resistant to five antimicrobials (ampicillin, kanamycin, streptomycin, sulfonamides, and tetracycline). Isolates of this phage type typically encoded all of their resistance genes on conjugative plasmids that could be efficiently transferred to E. coli. One of the DT193 isolates (strain UT20), however, carried its resistance factors integrated into its chromosome. Despite the chromosomal location of these resistance factors, they could still be readily transferred by conjugation and resided in transconjugants on an autonomously replicating plasmid. We also found a spontaneous mutant of UT20 that remained resistant, but could not transfer resistance factors by conjugation, and found mutants having different resistance patterns, each having lost regions of the chromosomally encoded resistance element. Furthermore, the method of plasmid transfer was complex: approximately one-half of E. coli transconjugants encoded all five resistances on a ~140 kb plasmid. Approximately half encoded resistance to kanamycin, tetracycline, streptomycin, and sulfonamides on a plasmid of ~126 kb, while less than 2% had a ~109 kb plasmid encoding resistance to ampicillin alone. Restriction mapping of these three plasmids showed them to be closely related, and all encoded their resistances with identical alleles. We propose that recombinational events in the Salmonella host strain produce these three distinct derivative plasmids.

Although the use of antimicrobials in food–producing animals has induced resistance in pathogenic bacteria, it remains unclear whether the cessation of antimicrobial use would eliminate these resistant strains. An important aspect to this question is the growth rate of resistant strains in comparison to similar susceptible strains. To test the fitness for growth of resistant Salmonella, we compared the growth rate in mixed cultures of multi–resistant Salmonella isolates of the DT193 phage type to isogenic fully susceptible derivatives. Competing strains were grown by serial transfer in antimicrobial-free media for 22 days, with each strain in the population quantified every three days. Wild type strains varied greatly in their relative fitness. UT20 displays a great fitness cost when it is in competition against its isogenic, fully susceptible derivative. UT8, whose resistance factors are identical to those of UT20 but are carried on a plasmid, displays an initial fitness cost that is much less severe and appears to be surmountable. This difference in fitness is not due to differences in the genetic backgrounds of the two strains since cured UT20 carrying the plasmid of UT8 also showed only a mild fitness defect. Conjugation within competitions was measured and found not to be an influencing factor on the persistence of resistance factors in a population.

It is our conclusion that the degree to which the fitness burden manifests itself is dependent on the location of the resistant element. Naturally occurring multiple drug resistant (MDR) strains of Salmonella enterica are at a large metabolic disadvantage compared to susceptible isogenic strains if the resistance element is integrated into the chromosome. It appears that if the resistance element is in plasmid form, then the fitness burden is significantly less, as well as being surmountable through means as yet undetermined.