Mechanism Accounting For Fluoroquinolone Multidrug Resistance In Escherichia Coli Isolated From Companion Animals

Pathogenic Escherichia coli (E. coli) are the predominant pathogen isolated inassociation with human and companion animal’s urinary tract infections (UTIs).Fluoroquinolone (FQ) antimicrobial compounds are commonly used to treat UTIs in dogs andcats, with enrofloxacin (ENR), an analogue of ciprofloxacin, being, and the most popular.However, the prevalence of antimicrobial resistance has increased with FQ use. FQ resistancein E. coli occurs in a stepwise fashion and generally is associated with multidrug resistance(MDR). The prevalence of MDR can adversely affect the health of both animal and humans;further, the emergence of resistant isolates in companion animals and the transfer of resistantisolates to humans pose a potential serious risk to public health. The purpose of the presentstudy was to characterize relationships among the different mechanisms in canine and felineuropathogenic E. coli expressing resistance to FQs that ranged from absent to high level, andmultidrug resistant isolates either susceptible or resistant to FQs, and then we hope it canprovide a theoretical basis for the study in E. coli resistance to FQs. All the results are asfollows.1. In vitro antibacterial activity of different quinolones to uropathogenic E. colifrom dogs and catsThe in vitro antibacterial activity studied the in vitro potency and efficacy towarduropathogenic E. coli from dogs and cats of11quinolones grouped according to differentgeneration structure: nalidixic acid (first generation) enrofloxacin, ciprofloxacin, ofloxacin,levofloxacin (second generation), orbifloxacin, marbofloxacin (third generation); gatifloxacin,pradofloxacin and moxifloxacin (fourth generation). In vitro potency measures includedminimum inhibitory concentration (MIC)(MIC50, MIC90); and mutant preventionconcentration (MPC) for FQ susceptible isolates only. Efficacy measures included relativesusceptibility [susceptible MIC breakpoint (MICBP-S)﹕MIC] or resistance [MIC﹕resistantMIC breakpoint (MICBP-R)] and mutant selection window (MSW)(MPC﹕MIC). Usingenrofloxacin as a prototypic FQ. For enrofloxacin susceptible isolates, MIC50and MIC90wasleast for the fourth generation drugs and ciprofloxacin, and greatest for nalidixic acid, enrofloxacin and orbifloxacin (P=0.006). The MPC were least for pradofloxacin (0.29±0.16μg/mL) and greatest for ofloxacin (1.56±0.53μg/mL)(P=0.006). MSW was least forpradofloxacin (54.79±30.45) and greatest for ciprofloxacin (152.29±76.04)(P=0.0024).MICBP-S﹕MIC was greatest for pradofloxacin (190.1±0.61) and least for nalidixic acid(5.33±0.52)(P=0.025). For FQ susceptible isolates, FQs MIC﹕MICBP-Rmay serve as asurrogate for MPC. Furthermore, about98.76%uropathogenic E. coli isolates in this studyshowed cross-resistance among the tested FQs. According to the time-kill studies, the FQstested showed the obvious characteristic of dose/concentraction-dependent.2. The mutations in DNA gyrase and topoisomerase IV, and the prevalence of theplasmid-mediated quinolone resistance (PMQR) genes in uropathogenic E. coli fromdogs and cats.Using enrofloxacin as the prototypic FQ, we found that the number of mutations inchromosomal target mutations correlate well the MICs in a stepwise manner. No gyrBmutation was observed in any of the E. coli isolates. Further, mutations in gyrA, parC andparE genes were not detected in either NDR or SDR phenotype isolates. A single mutation(Ser83Leu) in gyrA increases FQs MIC in susceptible isolates, subsequent mutations DNAgyrase and topoisomerase IV result in resistance that increases from low (enrofloxacin MICs4-16μg/ml) to high level (enrofloxacin MICs≥128μg/ml) with each progressive mutation.These findings demonstrated that MDR-associated FQ resistance in canine and felineuropathogenic E. coli reflects point mutations in DNA gyrase are necessary to achieve aclinical level of FQ resistance. qnrS and aac(6’)-Ib-cr gene were common in PDR isolates (8of13,62%), suggesting they played a role in the formation of MDR and higher-level FQresistant E. coli isolates. Phylogenetic analysis showed that31of52isolates studied belongedto phylogenetic group D,15belonged to group B1,3belonged to group B2, and3belonged togroup A. No consistent pattern emerged between phylotype and mechanisms of resistance.We did not find the ST131isolate due to the limited sample size.3. The expression of efflux pumps gene acrB and related regulatory genesThe leading multiple-drug acrB gene was expressed in all isolates tested, with expressionbeing greater in ENRR-MDR isolates; expression then decreased in the following order:ENRS-MDR, SDR, and NDR isolates. Overexpression of the acrB gene was not evident inany NDR and SDR isolates, while overexpression was present in20of27ENRR-MDRisolates and6of11ENRS-MDR isolates. A significant difference could not be detected in theexpression level of acrB between NDR or SDR isolates but tended to increase forENRS-MDR isolates (P=0.073), whereas differences were significant between all ENRSandENRR-MDR isolates (P<0.001). In addition, a significant difference was found between high-, moderate-, and low-level resistance, and no FQ resistant isolates (P=0.014). Theexpression of acrB increased with the severity of the resistance phenotype, in that the moresevere the MDR phenotype, the higher the probability of the acrB expression for the majorityMDR isolates. In other word, as MICs increase, acrB activity and the number of drug classescontributing to the MDR phenotype increasesFor marA gene, up-regulation was found in6ENRR-MDR isolates and3ENRS-MDRisolates, including two acrB and one SoxS up-regulated ENRS-MDR isolates. Additionally,soxS was overexpressed in3ENRR-MDR isolates, which was related to the overexpressedacrB. ompF gene was up-regulated in8isolates, while it was down-regulated in44isolates(24ENRR-MDR isolates versus20ENRS-MDR isolates).4. The expression of multidrug efflux pumps in multidrug resistant E. coli underthe stress of different FQsqRT-PCR indicates that four efflux pumps tested are constitutively expressed in allENRR-MDR and ENRS-MDR isolates, especially for emrE and acrB. Expression was greaterin ENRR-MDR followed by ENRS-MDR isolates. However, when stressed by the addition ofFQs all genes were induced to increase expression. Numerically the expression of emrE, acrB,macB and cmr increased1.12-5.4folds,1.23-3.60folds,1.40-6.58folds and1.52-5.91foldsin ENRS-MDR isolates, respectively. The corresponding magnitude of increase of fourpumps in ENRR-MDR were1.22-5.87folds,1.04-2.94folds,1.34-6.29folds and1.47-5.77folds, respectively. Different FQs have different stress for the expression of pumps;enrofloxacin has the greatest effect, followed by ciprofloxacin, orbifloxacin, marbofloxacin,moxifloxacin, gatifloxacin and pradofloxacin. It is consistent with the activities of differentFQs. In addition, the expression are different for four pumps, expression was greatest foremrE, followed by acrB, macB and cmr. Our study is the first to demonstrate the markedincrease in expression of emrE compared to other efflux pumps. Exposure to a FQ caused arapid and marked increase in the expression of all four pumps studied. This suggests thatcontinued use of FQ as routine therapy for UTI in dogs or cats should be discouraged.5. In vitro selection of pradofloxacin resistance in E. coli isolated from companionanimalsResistant uropathogenic E. coli strains can be selected by pradofloxacin, ciprofloxacin,and marbofloxacin in vitro after one or two steps. It is more difficult to select resistantmutants by pradofloxacin than by ciprofloxacin and marbofloxcain. Pradofloxacin target thegyrA subunit of DNA gyrase first as similar with ciprofloxacin and marbofloxacin, and thenthe number of mutations in target genes correlates well with the MICs and the level ofstepwise resistance. In addition,4resistant mutants selected by ciprofloxacin and mabofloxacin demonstrated a mutation at position12in global regulator gene SoxS,respectively, whereas it is not detected in the mutants selected by pradofloxacin.Overexpression of acrB gene is found in all the resistant mutants compared with parentisolates, and acrB activity increases as MIC increase, the more severe the MDR phenotype,the higher probability that acrB expression increased. These findings demonstrated thatMDR-associated pradofloxacin resistance in canine and feline uropathogenic E. coli reflects acombination of point mutations and enhanced efflux pump activity.