Novel variant of the qnrB gene, qnrB12, in Citrobacter werkmanii.

Three epidemiologically and clonally unrelated Citrobacter werkmanii isolates of poultry origin from Germany (Table 1) were investigated for the mechanisms responsible for decreased fluoroquinolone susceptibility. The MICs, determined by broth macrodilution (10), varied between 8 and 256 g/ml for nalidixic acid and between 0.0625 and 1 g/ml for ciprofloxacin (Table 1). CIT2 was additionally resistant to kanamycin. The quinolone resistance-determining regions of gyrA, gyrB, parC, and parE were amplified and sequenced (6). Nalidixic acid-resistant isolate CIT1 showed a Thr83Ile substitution in GyrA, previously reported to be sufficient for decreased fluoroquinolone susceptibility in Citrobacter isolates (11, 12, 16), and a Ser57Thr substitution in ParC, which was also seen in highly ciprofloxacin-susceptible Salmonella isolates (2). Another substitution, Leu88Gln, was detected in ParC of isolate CIT3, for which the MIC of ciprofloxacin was low, 0.0625 g/ml (Table 1). A variable decrease in the (fluoro)quinolone MICs was detected in the presence of the efflux pump inhibitor Phe-Arg-naphthylamide (Table 1), suggesting a variable contribution of an efflux system(s) in determining the level of quinolone susceptibility. PCR-based detection of the genes qnrA, qnrS, and qnrB, as well as of the aminoglycoside/fluoroquinolone-modifying enzyme-encoding gene aac(6 )-Ib-cr (3–8, 14), identified only a qnrB gene in all three C. werkmanii strains. A chromosomal location of this gene appeared most likely since repeated plasmid transformation-conjugation experiments using Escherichia coli CS1562 or E. coli HK225 RifR, respectively, yielded negative results (7), and Southern blot hybridization studies with plasmid profiles and with I-CeuI-digested genomic DNA gave a signal only with the largest I-CeuI fragment (approximately 800 kb) in each strain. Moreover, S1 nuclease digestion, followed by pulsed-field gel electrophoresis (1), did not identify any large qnrB-carrying plasmid which might comigrate with this 800-kb I-CeuI fragment (data not shown). Genomic DNA of isolate CIT1 was digested with restriction enzymes EcoRV, PvuI, SacII, and XhoII. The resulting fragments were religated and subjected to inverse PCR assays (with primers qnrB-inv-fw [5 -CGCACTGTGATTTGACCAATTC-3 ] and qnrB-inv-rv [5 -CGCCATGGAGAGATCACAACT-3 ]). The amplicons were sequenced and assembled. Analysis of a 3,629-bp segment (accession number AM774474) revealed that the qnrB gene was located downstream of four open reading frames, pspF, pspA, pspB, and pspC, encoding phage shock proteins in a genetic background with similarity to Klebsiella pneumoniae plasmid pTN60013 (17). The qnrB gene was identified as a novel variant, designated qnrB12, which codes for a 215-aminoacid protein and exhibits the highest degree of nucleotide (98.9%) and amino acid (99.5%) sequence identity to the qnrB9 sequence from Citrobacter freundii (accession number EF653270). To investigate the genetic background of the qnrB gene in the three Citrobacter isolates, additional PCR assays (with primers qnrB12-complete-fw [5 -TACCGCTGGATCTGCGT GA-3 ] and qnrB12-complete-rv [5 -TGTATGTCACTTTAG CGGCTGAAG-3 ]) were conducted. The resulting 1,060-bp amplicons comprised the entire qnrB gene, as well as 234 bp upstream and 178 bp downstream. Sequencing confirmed the location of the qnrB gene in an identical background. The three qnrB genes exhibited 99% nucleotide sequence identity and 100% amino acid identity. The amplicons were cloned into vector pCR-BluntII-TOPO and transformed into the E. coli TOP10 recipient strain (Invitrogen, Karlsruhe, Germany). In the presence of the cloned qnrB12 gene, a 4-fold increase in the MICs of nalidixic acid and a 32-fold increase in the MICs of ciprofloxacin were detected (Table 1). So far, few reports have suggested a chromosomal location of the commonly plasmid-borne qnr genes (15) and a chromosomally located origin of the genes was only confirmed in Shewanella algae (13). Our hybridization data, in combination with the sequence data of the qnrB12 flanking regions, led to the assumption that a qnrB12-carrying plasmid similar to pTN60013 from K. pneumoniae might have been, at least in