Further Characterization of Three Yersinia enterocolitica Strains with a Nalidixic Acid–Resistant Phenotype Isolated from Humans with Diarrhea

Introduction

Y ersiniosis, a foodborne disease, is characterized by symptoms including diarrhea, fever, abdominal pain, and vomiting (Drummond et al., 2012). Fluoroquinolones (FQ) are used for the treatment of this infection, particularly in immunocompromised individuals (Capilla et al., 2004). In Gram-negative bacteria, three mechanisms of FQ resistance have been described including target gene mutation(s); reduction in drug accumulation; and plasmid-mediated quinolone resistance (PMQR). DNA gyrase and DNA topoisomerase IV are tetrameric enzymes and are the targets for quinolone drugs. Resistance has been attributed to chromosomal mutations in the corresponding subunit-encoding genes (gyrA, gyrB for DNA gyrase and/or parC and parE for topoisomerase IV) (Ruiz, 2003). Within these loci, the quinolone-resistance-determining region (QRDR) is a mutational hot spot (Fàbrega et al., 2010). Several mutations have already been identified in gyrA in Yersinia enterocolitica, the most common of which lead to amino acid substitutions at Ser-83 and Asp-87, conferring resistance to nalidixic acid (Sánchez-Céspedes et al., 2003; Capilla et al, 2004; Sihvonen et al., 2011).

Compared to other members of the Enterobacteriaceae, quinolone resistance in Y. enterocolitica is not commonly encountered. In this study, chromosomal- and plasmid-mediated resistance mechanisms were investigated in three nalidixic acid–resistant Y. enterocolitica recovered from humans in Switzerland. The data presented extend our understanding of this resistance type in these foodborne pathogens.

Materials and Methods

Results and Discussion

The three Y. enterocolitica strains were isolated from humans between 2006 and 2010, in Switzerland. Two of the isolates were bio/serotypes 4:O3 (506-06 and 10-2307) and are classified as pathogenic. The remaining isolate was of biotype 1A (871-07). Isolates 506-06 and 10-2307 harbored the large pVYe plasmid of 67-kb (data not shown) and were positive for the ail and pYad markers, features associated with pathogenic Y. enterocolitica bio/serotypes (Lantz et al., 1998; Falcão et al., 2004; Thisted Lambertz et al., 2006). Isolate 871-07 did not contain this plasmid, consistent with the lack of ail and pYad markers, and the 1A biotype.

All three clinical isolates were resistant to nalidixic acid by disk diffusion with resistance profiles including 506-06: AmpKfNaSSxt; 871-07: AmcAmpKfNa; 10-2307: AmpKfNa. Two of the three clinical isolates (506-06 and 871-7, Table 1) were resistant to nalidixic acid with MIC>256 mg/L. All three clinical isolates were susceptible to ciprofloxacin and none contained PMQR markers (data not shown).

The amino acid substitutions associated with nalidixic acid resistance are summarized in Table 2. A mutation in gyrA was identified in isolate 871-07 consistent with the substitution of Ser-83-Ile, and an Asp-87-Tyr substitution was identified in isolate 506-06, similar to the change in Y. enterocolitica 8081. Mutations in these loci contributed to a nalidixic acid–resistant phenotype, as determined by E-test (Table 1). Furthermore, no mutations in gyrB, parC, or parE were detected.

Measurement of MICs to nalidixic acid and ciprofloxacin in the presence of phenylalanine-arginine-β-naphthylamide contributed to an increase in susceptibility with the exception of the reference strain, in respect of ciprofloxacin (Table 1). These findings suggest that efflux activity in this bacterium contributes to quinolone/FQ resistance, a feature that confirms earlier observations (Capilla et al., 2004).

Compared to other Enterobacteriaceae, small numbers of Y. enterocolitica resistant to nalidixic acid have been reported to date (Sánchez-Céspedes et al., 2003; Capilla et al., 2004; Fàbrega et al., 2010; Sihvonen et al., 2011; Fredriksson-Ahomaa et al., 2012). Although target gene mutations in gyrA are the primary mechanism of resistance to nalidixic acid, efflux pump activity, also contributes to support this phenotype. This finding supports data reported earlier (Capilla et al., 2004). Importantly, all three clinical isolates remain susceptible to ciprofloxacin.