Prevalence of efflux activity in low-level macrolide-resistant Campylobacter species.

Sir, Campylobacter macrolide resistance can involve an efflux mechanism. This mechanism was evidenced by the restoration of bacterial susceptibility after addition of phenylalanine-arginine b-naphthylamide (PAbN), an efflux pump inhibitor, in the culture medium. In this study we wanted to determine the prevalence of macrolide efflux among low-level resistant strains. From our laboratory collection we identified 30 strains that exhibited a low-level resistant phenotype with an MIC from 4 to 8 mg/L for erythromycin (Table 1). A similar low-level resistant phenotype was observed for these strains with clarithromycin and telithromycin (not shown). These strains were isolated between 1979 and 2005 from man, cattle and poultry and came from different areas in France. Since mutations of the nucleotides 2075 and 2074 on the 23S rRNA genes were shown to account for macrolide resistance in Campylobacter jejuni and Campylobacter coli previously, we examined the nucleotide sequence of these genes in the 30 strains by real-time PCR assay. No mutation was identified in these strains (0/30). To determine the efflux impact in the lowlevel resistant phenotype, we compared antibiotic susceptibility in the presence and in the absence of PAbN, according to the method that we reported previously. All 30 strains showed a significant increase in susceptibility to erythromycin and clarithromycin (Table 1). The same results were obtained with strains 79BM, LV9, 96C208, 98054, 96C151, 47843 and 98178, when the test was performed with susceptibility discs loaded with 5 mg of telithromycin. Conversely, the susceptibility to ciprofloxacin and norfloxacin, two molecules that are also transported by efflux in Campylobacter, was not modified by the presence of PAbN in any of the 30 strains (data not shown). At this point one could consider that the intrinsic antibacterial activity of PAbN can interfere with the macrolide susceptibility restoration. We thus examined the antibacterial activity of PAbN, in agar diffusion assays (Table 1). With 10 mg of PAbN, we observed no noticeable growth inhibition with any of the strains. When the PAbN amount was increased up to 50 mg, the strains exhibited a wide range of susceptibility with inhibition diameters from 6 to 18 mm (Table 1). Antibacterial activity of PAbN could not be correlated with its restoration efficiency as it was assessed by Spearman and Pearson tests (n = 30, P > 0.05). In this study, we showed that the efflux mechanism sensitive to PAbN is the mechanism that was responsible for macrolide resistance in all the 30 low-level resistant Campylobacter tested. This extends previous reports suggesting that low-level resistance could be a general mechanism in C. jejuni and C. coli. No correlation between antibiotic susceptibility and the recently identified CmeF pump was reported; furthermore, we already showed that PAbN was able to decrease the MIC of cmeB mutants. These data suggest that another pump is active in all the low-level resistant Campylobacter. Moreover, this mechanism was found in one strain isolated in 1979 and one in 1985. This might indicate that macrolide efflux could be a conserved mechanism that does not result from the use of new antibiotics in humans or animals. It should now be of major interest to determine if the low-level resistance can allow Campylobacter to acquire a high-level resistance phenotype for macrolides as was recently demonstrated for fluoroquinolone resistance.