Emergence of the plasmid-mediated quinolone resistance gene qnrS1 in Escherichia coli isolates in Greece.

Three major groups of the plasmid-mediated quinolone resistance (Qnr) determinants have been identified so far in Enterobacteriaceae: the QnrA group, which includes 6 variants, the QnrB group, which includes 19 variants, and the QnrS group, which includes 3 variants (5). Although Qnr proteins produce only low-level resistance, they provide a favorable background for higher resistance to occur at quinolone concentrations that would be lethal in their absence, through secondary changes in DNA gyrase and topoisomerase IV, porin, or efflux systems (6). The purpose of our study was to investigate the presence and dissemination of the qnr genes among ciprofloxacin-resistant Escherichia coli isolates from different hospitals in Greece, a region with a relatively high frequency of quinolone resistance (1), where qnr genes had not been reported previously. A total of 113 nonrepetitive ciprofloxacin-resistant E. coli clinical isolates were taken at random from the laboratory collections of four unrelated hospitals in northern and central Greece between 2006 and 2007 and analyzed for qnr genes. Ciprofloxacin MICs were estimated by the Etest (AB Biodisk, Solna, Sweden) and the agar dilution method according to Clinical and Laboratory Standards Institute (CLSI) guidelines (3) by using the breakpoints 1 and 4 g/ml for susceptibility and resistance, respectively. E. coli ATCC 25922 was used as a control in all susceptibility assays; positive controls for the genes qnrA, qnrB, and qnrS were kindly provided by J. SanchezCespedes. Susceptibilities to all antimicrobials tested were defined according to the CLSI interpretative criteria (3). PCR was performed with primers amplifying all known qnr gene variants. The primer pair for the gene qnrA was 5 -AGA GGATTTCTCACGCCAGG-3 and 5 -CCAGGCACAGATC TTGAC-3 (yielding a 580-bp product), that for qnrB was 5 GGGTATGGATATTATTGATAAAG-3 and 5 -CTAATCC GGCAGCACTATTA-3 (yielding a 264-bp product), and the primer pair for qnrS was 5 -GCAAGTTCATTGAACAGGG T-3 and 5 -TCTAAACCGTCGAGTTCGGC-3 (yielding a 428-bp product). The gene gyrA was amplified with primers 5 -TTAATGATTGCCGCCGTCGG-3 and 5 -TACACCGG TCAACATTGAGG-3 (yielding a 648-bp product) and parC was amplified with primers 5 -AAACCTGTTCAGCGCCGC ATT-3 and 5 -GTGGTGCCGTTAAGCAAA-3 (yielding a 395-bp product) to evaluate possible coexisting chromosomal mutations. The corresponding specific PCR products were sequenced by LARK Technologies, Essex, United Kingdom. For the qnr-positive isolates, synergy experiments were also performed using ciprofloxacin and the efflux pump inhibitor CCCP (carbonyl cyanide m-chlorophenylhydrazone) (8) to check the contribution of efflux pump overexpression to ciprofloxacin resistance. Pulsed-field gel electrophoresis (PFGE) analysis of XbaI-digested genomic DNA was performed, and the banding patterns of the strains were compared visually according to the criteria proposed by Tenover et al. (9). Filter mating experiments were performed with qnr-positive isolates by using E. coli 26R793 (lac negative and rifampin resistant) as the recipient. Transconjugants were selected on MacConkey agar plates containing 100 mg of rifampin/liter and 6 mg of nalidixic acid/liter, tested for qnr genes by PCR, and analyzed for plasmids by alkaline lysis. Eleven of the 113 E. coli isolates (10%) derived from three independent hospitals in Thessaly (Larissa, central Greece) and Macedonia (Thessaloniki, northern Greece) and exhibiting nine unrelated PFGE strain patterns were qnr positive. The ciprofloxacin MICs for these isolates were 16 to 128 g/ml; the characteristics of the isolates are presented in Table 1. One Qnr-positive isolate (isolate 3) was an extended-spectrum -lactamase producer carrying the gene blaCTX-M-15. No synergy between CCCP and ciprofloxacin in any isolate was observed. Sequencing of the PCR products showed that all 11