Quantifying the contribution of four resistance mechanisms to ciprofloxacin MIC in Escherichia coli: a systematic review.

BackgroundThe objective of this study was to examine Streptococcus pneumoniae isolates collected from a longitudinal surveillance program in order to determine their susceptibility to currently used fluoroquinolones and of the frequency and type of mutations in the quinolone-resistant determining regions (QRDRs) of their parC and gyrA genes.MethodsThe Canadian Bacterial Surveillance Network has been collecting clinical isolates of S. pneumoniae from across Canada since 1988. Broth microdilution susceptibility testing was carried out according to the Clinical and Laboratory Standards Institute guidelines. The QRDRs of the parC and gyrA genes were sequenced for all isolates with ciprofloxacin MIC ≥ 4 mg/L, and a large representative sample of isolates (N = 4,243) with MIC ≤ 2 mg/L.ResultsA total of 4,798 out of 30,111 isolates collected from 1988, and 1993 to 2007 were studied. Of those isolates that were successfully sequenced, 184 out of 1,032 with mutations in parC only, 11 out of 30 with mutations in gyrA only, and 292 out of 298 with mutations in parC and gyrA were considered resistant to ciprofloxacin (MIC ≥ 4 mg/L). The most common substitutions in the parC were at positions 137 (n = 722), 79 (n = 209), and 83 (n = 56), of which substitutions at positions 79 and 83 were associated with 4-fold increase in MIC to ciprofloxacin, whereas substitutions at position 137 had minimal effect on the ciprofloxacin MIC. A total of 400 out of 622 isolates with Lys-137 parC mutation belonged to serotypes 1, 12, 31, 7A, 9V, 9N and 9L, whereas only 49 out of 3064 isolates with no mutations belonged to these serotypes. Twenty-one out of 30 isolates with substitutions at position 81 of the gyrA gene had an increased MIC to ciprofloxacin. Finally, we found that isolates with mutations in both parC and gyrA were significantly associated with increased MIC to fluoroquinolones.ConclusionsNot all mutations, most frequently Lys-137, found in the QRDRs of the parC gene of S. pneumoniae is associated with an increased MIC to fluoroquinolones. The high prevalence of Lys-137 appears to be due to its frequent occurrence in common serotypes.

Mechanisms of resistance to ciprofloxacin, ampicillin/sulbactam and imipenem in Acinetobacter baumannii clinical isolates in Taiwan

Sir, Target changes via alterations of DNA gyrase and/or topoisomerase IV, and reduced intracellular drug accumulation, either by reduced drug permeability and/or increased efflux activity, are the two major mechanisms that mediate fluoroquinolone resistance. Resistance to quinolones in Escherichia coli appears to be caused mainly by alterations in the gyrA gene of the DNA gyrase and in the parC gene of the topoisomerase. The mutations are located at the N-terminus of the GyrA protein (residues 67–106 in E. coli) in the quinolone resistance-determining region (QRDR). A similar QRDR has also been reported in parC. We have reported as part of a national surveillance programme in Taiwan 526 E. coli isolates that were classified as fluoroquinolone susceptible according to the NCCLS guidelines. However, the distribution of zone diameters of ciprofloxacin indicated that these isolates consisted of two sub-populations with susceptibility and reduced susceptibility. From five of 44 hospitals surveyed, 193 E. coli isolates were selected for further study. Of these 193 isolates, 20 (10.4%) were with resistance, 44 (22.8%) with reduced susceptibility and 129 (66.8%) with susceptibility with zone diameters of ciprofloxacin 15, 16–29 and 30 mm, respectively, The 20 isolates with resistance, 44 isolates with reduced susceptibility and a random sample consisting of 17 E. coli isolates with susceptibility were assessed for point mutations in the QRDRs of gyrA and parC to determine the genetic bases of fluoroquinolone susceptibility. Based on their mutations in gyrA and parC, the 81 isolates were divided into 13 classes (Table). All resistant isolates in our study had two mutations in gyrA and at least one additional mutation in parC (Table, classes 1–5). A major type of combined mutations, Ser-83 Leu, Asp-87 Asn in gyrA and Ser-80 Ile in parC, was found in 17 resistant isolates (85%). Although single mutations in gyrA have previously been associated with reduced susceptibility to fluoroquinolones in E. coli, the molecular epidemiology of mutations in both gyrA and parC and their associations with reduced susceptibility have not been delineated. All isolates with reduced susceptibility in our study had one mutation in gyrA (Table, classes 6–12). Ser-83 Leu was a major mutation found in the isolates with reduced susceptibility, detected in 39 strains (87%). Isolates with double mutations in gyrA and parC (classes 6–8) had higher MIC90s of fluoroquinolones (0.75 and 2 mg/L of ciprofloxacin and ofloxacin, respectively) than isolates with only one mutation in gyrA (class 9, 0.25 and 1 mg/L of ciprofloxacin and ofloxacin, respectively). These data indicate that additional mutations in parC result in slight increases in MICs, consistent with the results reported for Neisseria gonorrhoeae. No strain possessed a parC mutation without the simultaneous presence of a gyrA mutation. These findings are consistent with fluoroquinolone resistance arising in a multi-step fashion with mutations in gyrA occurring as the first step. Isolates with zone diameters of ciprofloxacin of 15, 16–20 and 21 mm are considered as resistant, intermediate and susceptible, respectively, to ciprofloxacin according to current NCCLS resistance breakpoints. A major concern related to finding a large number of E. coli (22.8%) with reduced susceptibility, especially those with zone diameters of ciprofloxacin between 21 and 29 mm, is the fact that these isolates are unlikely to be detected using current NCCLS guidelines. Increased clinical failures have been reported when human infections caused by Salmonella spp. with reduced susceptibility are treated with fluoroquinolones. Owing to similar concerns over treatment efficacy, the breakpoint for the definition of resistance to fluoroquinolones in N. gonorrhoeae was recently adjusted as follow: isolates with zone diameters of ciprofloxacin 27, 28–40 and 41 mm are interpreted as resistant, intermediate and susceptible, respectively, to ciprofloxacin. Antimicrobial susceptibility test results should not only guide clinicians in determining the most appropriate therapy for their patients, but also assist efforts to monitor and control the spread of resistance. We therefore recommend that current NCCLS breakpoints for fluoroquinolone resistance in Enterobacteriaceae, such as E. coli and Salmonella spp., be modified according to our findings: Correspondence Journal of Antimicrobial Chemotherapy (2001) 48, 931–942

mutations. Mutations in the gyrA gene were present in 49 of 147 strains (33%); double mutations in both gyrA and parC genes were detected in 25 strains (17%). The spectrum of mutations and the corresponding MIC distributions of the various compounds are shown in Table 1. Mutations were significantly more frequent in outbreak-related strains compared with sporadic strains (62% versus 23%; P = 0.005; OR 2.6; 95% CI 1.3–5.5). Double mutations leading to high-level ciprofloxacin resistance (MICs ≥ 32 mg/L) and decreased susceptibility to newer quinolones were found in 26% of outbreak-related strains and 14% of sporadic strains. This is consistent with the suggestion that resistance to fluoroquinolones could be a risk factor for epidemic behaviour in A. baumannii. 1 There were no significant differences in the activity of ciprofloxacin and the various newer quinolones against strains without mutations in gyrA or parC. Against strains with mutations, there was a correlation between the substituting amino acid and the resulting median MIC of the quinolone compounds (Table 1). The newer quinolones were up to three times more active than ciprofloxacin against A. baumannii strains with a single mutation, and up to five times against strains with a double mutation. Vila et al. 2,3 described mutations in gyrA and parC of 15 A. baumannii strains that correlated with resistance to ciprofloxacin. Among 25 Acinetobacter isolates obtained from various locations worldwide, Seward & Towner 7 confirmed mutations in gyrA and parC of 11 ciprofloxacin-resistant isolates and showed that 10 of these isolates were epidemiologically unrelated, as determined by randomly amplified polymorphic DNA analysis. Our findings confirmed that substitution of Ser-83 in GyrA is sufficient to confer ciprofloxacin MICs of ≥4 mg/L. In agreement with other stud

Escherichia coli is the major causative agent for urinary tract infections (UTIs), and fluoroquinolones (FQ) are commonly used in the treatment of patients with UTIs. The surge in FQ-resistant E. coli is an important public health threat worldwide. We investigated the prevalence and mechanisms of FQ resistance among FQ-resistant E. coli isolated from UTI patients. A total of 131 FQ-resistant E. coli strains were characterized and broth microdilution assay showed that 108 strains (82.4%) were highly resistant to ciprofloxacin (MIC ≥ 32 μg/mL). All strains were analyzed for plasmid-mediated quinolone resistance (PMQR) genes, with 37 (28.2%) testing positive. Among the PMQR genes detected, aac(6')-Ib-cr was the most frequent, found in 30 strains (22.9%), followed by qnrS in 10 strains (7.6%) and qnrB in 1 strain (0.8%). PCR assay showed that all carried acrA, acrB, and tolC genes, but 33 strains (25.2%) revealed at least 4-fold reduction in ciprofloxacin MIC when using PAβN and CCCP as efflux pump inhibitors, indicating the role of the AcrAB efflux pump in ciprofloxacin resistance in these strains. The 19 strains of high-level ciprofloxacin-resistant E. coli were selected to determine the target enzyme alteration by PCR assay and DNA sequencing. Genetic analysis revealed that 16 strains (84.2%) had double mutations in gyrA (S83L and D87 to N or Y) with single mutation in parC (S80I), while 3 strains (15.8%) had double mutations in gyrA (S83L and D87 to N or Y) and parC (S80I and E84 to G or V). The positive efflux activity was linked to increased MIC values of ciprofloxacin (P < 0.001). Overall, the carriage of PMQR genes, efflux activity, and target mutations across E. coli strains contribute to ciprofloxacin resistance, a result that may necessitate a reassessment of the antibiotics in use for empirical UTIs therapy.

To study the temporal evolution of fluoroquinolone non-susceptibility among Streptococcus pyogenes during 2007-10 in Belgium. S. pyogenes (n = 4690) recovered from patients with tonsillopharyngitis or skin, wound or invasive infections were screened for fluoroquinolone non-susceptibility. A selection of fluoroquinolone-non-susceptible strains was investigated for resistance mechanisms: reserpine-sensitive efflux and mutations in topoisomerase genes parC and gyrA. Clonality was determined by emm typing. Fluoroquinolone non-susceptibility (ciprofloxacin MIC ≥2 mg/L) was identified in 535 (11.4%) of 4690 S. pyogenes recovered during 2007-10 in Belgium. The proportion of fluoroquinolone-non-susceptible S. pyogenes increased significantly from 4.3% (2008) to 10.9% (2009) to 21.6% (2010) and coincided with a significant increase in emm6 strains among fluoroquinolone-non-susceptible S. pyogenes. Ciprofloxacin MICs of 2-8 mg/L correlated with first-step ParC substitutions. Two high-level fluoroquinolone-resistant S. pyogenes strains (ciprofloxacin MICs 32 mg/L) showed second-step substitutions in GyrA (Ser-81→Phe or Tyr) in addition to first-step mutations in parC. Reserpine-sensitive efflux was not observed. We report an unprecedented increase in fluoroquinolone-non-susceptible S. pyogenes in Belgium, a country with high quinolone use, as well as emergence of two high-level fluoroquinolone-resistant S. pyogenes strains with second-step mutations in gyrA, warning us of the need for more prudent use of fluoroquinolones and for continued resistance surveillance.

To the Editor: Ciprofloxacin is the treatment of choice of severe nontyphoidal Salmonella infections in adults. Resistance to ciprofloxacin has been found exceptionally in nontyphoidal Salmonella enterica isolates and only in serotypes Typhimurium, Choleraesuis, and Schwarzengrund (1-8). Such isolates have been collected from humans and animals in Europe, Asia, and North America.

To the Editor: In pneumococci, resistance to fluoroquinolones is associated with chromosomal mutations in the quinolone-resistance–determining regions (QRDR) of type II topoisomerase enzymes, predominantly gyrA and parC. Several mutations have been described in these enzymes, but only a few have been shown by in vitro studies to confer resistance: S81F or Y, C, or I and E85K in gyrA; E474K in gyrB; A63T, S79F or Y or L and D83G or N in parC; and E474K and D435N or H in parE (1–5). Other frequently described mutations are K137N in parC and I460V in parE, which appear to not contribute to fluoroquinolone resistance because they are commonly found in susceptible strains, and no evidence exists for their conferring fluoroquinolone resistance in vitro. We describe here a pneumococcal strain that was isolated from a 66-year-old white man with chronic obstructive pulmonary disease (COPD). The patient was admitted to the hospital with a presumed exacerbation of COPD. He had been discharged from the hospital 2 days earlier, having recovered from a similar manifestation of this disease. His treatment history was 250 mg/day oral levofloxacin for 7 days while in the hospital and levofloxacin for 10 days as an outpatient for a similar lower respiratory tract infection 3 months earlier. On this second admission he was given levofloxacin, 250 mg intravenously, once a day. He was treated with a low dosage because he was in renal failure. The patient continued to worsen and was transferred to the intensive care unit, where ceftriaxone, 1 g intravenously once a day, was given along with levofloxacin. He improved on the combination therapy and was discharged without sequelae. Cultures of the patient's blood and sputum grew Streptococcus pneumoniae. The isolate from blood was resistant to levofloxacin (MIC 8 mg/L) and ciprofloxacin (MIC 8 mg/L), yet susceptible to gatifloxacin (MIC 1 mg/L) and ceftriaxone (MIC 0.38 mg/L), with intermediate resistance to penicillin (MIC 1.5 mg/L). The resistant isolate was of serotype 6A and of multilocus sequence type 376, which is the North Carolina6A-23 clone (http://www.sph.emory.edu/PMEN/index.html). Efflux testing that compared the ciprofloxacin MICs in the presence and absence of reserpine (10mg/L) showed no evidence of an overexpressed efflux pump. We sequenced the QRDRs (gyrA, gyrB, parC, parE) and the entire gyrA and parC genes of the resistant strain isolated from blood by using previously described primers (2). Sequencing showed a S79Y mutation in parC and a Q118K (CAA→AAA) mutation in gyrA. Sequencing of the entire gyrA and parC genes confirmed that no additional amino acid substitutions were outside the QRDRs. The entire gyrA gene PCR product was transformed directly into the susceptible pneumococcal reference strain R6 by a standard transformation protocol (4). Transformants were selected on plates containing increasing concentrations of ciprofloxacin and, in a second step, were transformed with the entire parC gene of the resistant strain. The ciprofloxacin and levofloxacin MICs of R6 transformed with the gyrA gene of the resistant isolate containing the new Q118K mutation were 4 and 2 mg/L, respectively. After additional transformation of these transformants with parC of the resistant isolate containing the S79Y mutation, the selected double transformants exhibited the same MICs as the original clinical isolate (8 mg/L for ciprofloxacin and levofloxacin). The transformation of parC alone conferred an intermediate increase in the MICs (ciprofloxacin 2 mg/L, levofloxacin 4 mg/L). All transformants were confirmed by sequencing. To determine the biologic cost associated with the different resistance mutations in vitro, each fluoroquinolone-resistant mutant was competed against the fluoroquinolone-susceptible parent strain R6 (with an independent streptomycin resistance marker) as described by Johnson et al. (6). The outcome was evaluated as the change in the ratios of the competing strains as a function of the number of generations. Each competition was performed in triplicate by using independent starting cultures of each competing strain. Compared with the wild-type R6 strain, the relative fitness values for the gyrA, parC, and double mutants were 1.06, 1.03, and 0.93, respectively. These data indicate that a single mutation in either parC or gyrA does not impose a substantial fitness burden. In contrast, the double-mutation parC S79Y and gyrA Q118K was associated with a slower growth rate. Similar results of relative fitness for single (parC S79Y and gyrA S81F) and double mutations were observed by Gillespie et al. (7). Development of resistance to fluoroquinolones is a stepwise process, involving spontaneous mutations in the genes encoding the target enzymes DNA gyrase and the topoisomerase IV. Mutants with mutations in 1 of the enzymes are estimated to arise at a frequency of 1 to 10-7 (1). Therefore, fluoroquinolone resistance due to selection of spontaneous mutants during treatment may be related to the number of bacterial cells in the population under selective pressure. Patients with COPD are frequently colonized by high bacterial loads. COPD has been identified in several recent studies as an independent risk factor for fluoroquinolone resistance (8,9). Low doses of fluoroquinolones may also lead to an increased risk for resistance selection (10). Because the Q118K mutation has not been previously described, this new mutation was probably selected by the current or antecedent treatments rather than by an infection with a resistant widely disseminated clone.

Pefloxacin as a Surrogate Marker for Fluoroquinolone Susceptibility for Salmonella typhi: Problems and Prospects.

Evaluation of quinolone resistance–determining region mutations and efflux pump expression in Neisseria meningitidis resistant to fluoroquinolones

Sir, Streptococcus pneumoniae is the most common bacterial pathogen of the respiratory tract, primarily causing pneumonia, meningitis and acute otitis media. Infections caused by this pathogen are being further aggravated by the spread of resistance to several classes of antimicrobial agents including b-lactams, macrolides and tetracycline. Recently, respiratory fluoroquinolones, including levofloxacin, gemifloxacin and moxifloxacin, have been used for the treatment of community-acquired pneumonia. Respiratory fluoroquinolones, especially moxifloxacin, are still highly effective against S. pneumonia. However, with increasing use of respiratory fluoroquinolones worldwide, the number of clinical isolates of S. pneumoniae with fluoroquinolone resistance continues to increase. Fluoroquinolones inhibit DNA gyrase and topoisomerase IV, which are involved in bacterial DNA replication. Both enzymes are composed of two subunits: GyrA and GyrB in DNA gyrase, and ParC and ParE in topoisomerase IV. The most extensively described mechanism of resistance to fluoroquinolones is the development of point mutations in a particular region of each enzyme subunit, known as the quinolone resistance-determining region (QRDR). In clinical isolates of S. pneumoniae, resistance mutations in gyrA and parC are more common than in gyrB and parE. The most prevalent resistance mutations are S81F/Y or E85K in gyrA and S79Y/F in parC, all of which confer high-level resistance to ciprofloxacin and levofloxacin (MIC ≥16 mg/L) and low-level resistance to moxifloxacin (MIC 4–8 mg/L). There are high-level moxifloxacin-resistant clinical strains, but their resistance mechanisms are unknown. To study the progression of moxifloxacin resistance in S. pneumoniae, mutants of S. pneumoniae R6 were generated in a stepwise fashion by selecting colonies that grew at various antibiotic concentrations on gradient plates at each stage; the highest concentrations for each stage were 0.25, 8 and 16 mg/L. Three selection cycles were required to obtain the high-level moxifloxacin-resistant mutants, designated R6M-1, -2 and -3 respectively (Table 1). These moxifloxacin-resistant mutants also showed resistance to ciprofloxacin and levofloxacin. Interestingly, the increases in resistance were not equal for each antibiotic. The levels of resistance increased with the acquisition of additional mutations in gyrA, parC and parE (Table 1). The highly resistant R6M-3 strain contained common mutations in gyrA and parC, but it also had a mutation in parE (P454S in ParE). The mutation in ParE (P454S) was previously described in a clinical isolate in combination with mutations in ParC (S79Y) and GyrA (S81F); however, there was no evidence that this mutation was involved in fluoroquinolone resistance. PCR fragment transformation was used to test whether the ParE (P454S) mutation was involved in moxifloxacin resistance. The primers parE398 (5′-AAGGCGCGTGATGAGAGC-3′) and parE483 (5′-TCTGCTCCAACACCCGCA-3′) were used to amplify the QRDRs of parE from the highly moxifloxacin-resistant strain R6M-3 and from R6WT (where WT stands for wild-type). The resulting fragments (290 bp) were transformed into the strain R6M-2 by following the procedure described by Joloba et al. Transformants were selected in the presence of 16 mg/L moxifloxacin, and transformation was confirmed by PCR and sequence analysis. Thirty colonies were obtained following transformation with the PCR fragment carrying the ParE P454S mutation, and no colony was obtained following transformation with the PCR fragment without mutation. Sequencing analysis showed that all of the transformants carried a gene encoding the P454S mutation in ParE. Transformation efficiency was calculated to be 3.0×10 according to a previously described

Background and AimRecently, the extensive use of quinolones led to increased resistance to these antimicrobial agents, with different rates according to the organism and the geographical region. The aim of this study was to detect the resistance rate of Klebsiella pneumoniae Iraqi isolates toward quinolone antimicrobial agents, to determine genetic mutations in gyrA and parC, to screen for efflux-pump activity, and to screen the presence of plasmid-mediated quinolone resistance (PMQR) genes.MethodsForty-three K. pneumoniae isolates were confirmed phenotypically and genotypically by Vitek 2 system and species specific primers by PCR using the targeting rpo gene followed by sequencing. Antibiotic susceptibility test was carried out using disc diffusion method. Quinolone resistant isolates were subjected to ciprofloxacin MIC testing, and cartwheel method to screen for efflux pump activity. The presence of the plasmid mediated quinolone resistance genes qepA, qnrB, qnrS, and aac(6)Ib was tested by PCR. Sequencing of gyrA and parC was performed.ResultsWe observed a high rate of resistance to ceftriaxone, gentamicin ciprofloxacin, and levofloxacin. Low rate of resistance was detected against amikacin and azithromycin. Ciprofloxacin MIC results revealed that 96.1% of the isolates had MICs >256 µg/mL, 83.4% had MICs >512 µg/mL while 34.6% had MIC >1024 µg/mL. Testing of isolates against ciprofloxacin mixed with EtBr at various concentrations resulted in decreased resistant. Sequencing results showed that Ser83Leu was the most common mutation in gyrA that was observed in all quinolone resistant isolates, followed by Asp87Asn. Ser80Ile mutation in parC was observed in 77.7% of the tested isolates. The prevalence of PMQR genes was 92.5% aac (6)-Ib, 51.8% qnrB, 40.7% qepA, and 37% qnrS.ConclusionQuinolone resistance is common in K. pneumoniae isolates in Baghdad. The frequent mutation in gyrA and parC, and the presence of PMQR genes is alarming.

Co-existence of plasmid-mediated quinolone resistance determinants and mutations in gyrA and parC among fluoroquinolone-resistant clinical Enterobacteriaceae isolated in a tertiary hospital in Warsaw, Poland

Ninety-three Malaysian extended-spectrum β-lactamase (ESBL)-producing Klebsiella pneumoniae isolates were investigated for ciprofloxacin resistance. Two mismatch amplification mutation (MAMA) assays were developed and used to facilitate rapid detection of gyrA and parC mutations. The isolates were also screened for plasmid-mediated quinolone resistance (PMQR) genes including aac(6′)-Ib-cr, qepA, and qnr. Ciprofloxacin resistance (MICs 4– ≥ 32 μg/mL) was noted in 34 (37%) isolates, of which 33 isolates had multiple mutations either in gyrA alone (n = 1) or in both gyrA and parC regions (n = 32). aac(6′)-Ib-cr was the most common PMQR gene detected in this study (n = 61), followed by qnrB and qnrS (n = 55 and 1, resp.). Low-level ciprofloxacin resistance (MICs 1-2 μg/mL) was noted in 40 (43%) isolates carrying qnrB accompanied by either aac(6′)-Ib-cr (n = 34) or a single gyrA 83 mutation (n = 6). Ciprofloxacin resistance was significantly associated with the presence of multiple mutations in gyrA and parC regions. While the isolates harbouring gyrA and/or parC alteration were distributed into 11 PFGE clusters, no specific clusters were associated with isolates carrying PMQR genes. The high prevalence of ciprofloxacin resistance amongst the Malaysian ESBL-producing K. pneumoniae isolates suggests the need for more effective infection control measures to limit the spread of these resistant organisms in the hospital.

To investigate the prevalence of fluoroquinolone resistance in Streptococcus pyogenes and its in vitro selection by ciprofloxacin and the respiratory fluoroquinolones, levofloxacin and moxifloxacin. S. pyogenes (n = 5851) recovered from pharyngitis and invasive infections during 2003-06 in Belgium were screened for fluoroquinolone non-susceptibility (ciprofloxacin MIC > or =2 mg/L) and further studied for mutations in the topoisomerase genes, reserpine-sensitive efflux, clonality by PFGE and emm typing. Fourteen well-characterized fluoroquinolone-non-susceptible or -susceptible isolates were exposed stepwise to increasing levels of ciprofloxacin, levofloxacin and moxifloxacin. Selected mutants with increased MICs were analysed for resistance mechanisms. Mutation frequencies at 2x and 4x MIC of moxifloxacin and levofloxacin were estimated for a clinical emm6 parent strain carrying mutations in both parC and gyrA. Prevalence of fluoroquinolone-non-susceptible S. pyogenes (n = 437; 7.47%) increased significantly from 2.08% and 5.08% to 13.11% during 2003-05 and decreased to 8.93% in 2006 (chi(2) test; P < or = 0.001). emm6 constituted 80.09% of the total fluoroquinolone-non-susceptible isolates. Of the 71 S. pyogenes sequenced, 70 harboured first-step parC or gyrA mutations correlating with ciprofloxacin MICs 2-8 mg/L. Reserpine-sensitive efflux was not observed. One emm6parC mutant (Ser79Ala) also showed a second-step mutation in gyrA (Ser81Tyr), with MICs of ciprofloxacin, levofloxacin and moxifloxacin of 32, 8 and 1 mg/L, respectively. Mean mutation frequencies under moxifloxacin selection were 500- to 30 000-fold higher for this strain than those for an emm6 control strain. Selection of the emm6 double mutant with moxifloxacin generated a mutant with a moxifloxacin MIC of 64 mg/L and a levofloxacin MIC of 128 mg/L, and an additional Asp83Tyr substitution in ParC. We report an emergence of levofloxacin and high-level ciprofloxacin resistance associated with a second-step gyrA mutation in a clinical emm6 S. pyogenes. The observed high mutation frequency and in vitro selection of high-level resistance to the respiratory fluoroquinolones in the emm6 double mutant is of concern.