Factors Influencing the Development of Resistance to Fluoroquinolone Antibiotics by Food Borne Bacteria

Progress: Initial work involved collection of libraries of Campylobacter and Salmonella and then analysing these strains for resistance to antibiotics, disinfectants, dyes and organic solvents and for presence of mutations within the QRDR of DNA gyrase and topoisomerase. (See project OD 2004 for the synergy between projects and details on efflux mediated resistance. <P>
The primary mechanism by which Gram negative bacteria become resistant to quinolone antibiotics is mutation(s) in the gyrA gene changing the target protein, DNA gyrase, with high level resistance due to two or more mutations in this gene. Mutations in gyrB and DNA topoisomerase IV genes parC and parE can also be involved in fluoroquinolone resistance. In clinical human and veterinary isolates of Salmonella spp., normally only mutations in gyrA are seen. More recently it was demonstrated that some Salmonella Typhimurium isolates recovered from a patient treated with ciprofloxacin showed resistance to multiple antibiotics including ciprofloxacin associated with over-expression of the AcrAB efflux pump rather than mutations in gyrA. In E. coli the AcrAB efflux pump can be up-regulated by the global regulators such as marRAB and soxR giving rise to low level multiple antibiotic resistance and resistance to organic solvents such as cyclohexane and this is know as the multiple antibiotic resistance or MAR phenotype. Additionally, there is recent evidence that active efflux is the primary mechanism of resistance to ciprofloxacin in Salmonella Typhimurium and in E. coli that DNA gyrase mutations fail to produce clinically significant fluoroquinolone resistance in the absence of the AcrAB efflux pump. <P>
With the above in mind, it was interesting to note that strains showing the higher levels of resistance to ciprofloxacin had both mutations in gyrA and were resistant to cyclohexane. This would suggest the dual involvement of both efflux and mutations in higher-level ciprofloxacin resistance as suggested by previous worker. The involvement of efflux in quinolone and /or cyclohexane resistance of salmonella was determined by examining resistant strains for expression of marA, soxS and acrB genes. The results are being analysed at present at Birmingham University. <P>
In addition to testing for known mutations in gyrA, Wave DHPLC was used to perform detailed analysis of the QRDR of gyrA, gyrB, parC and parE of 210 quinolone resistant strains of Salmonella and has revealed novel mutations. It was also interesting to note that some mutations seemed to be more common in certain serotypes. For instance, 80% of Montevideo serovars contain Asp87-Asn substitutions within GyrA, 65% of Newport serovars contain Ser83-Tyr and 40% of all Typhimurium contain a Ser83-Phe substitution. When the data for all 210 isolates was examined the overall frequency of each substitution was 20%, 14% and 23% respectively. <P>
A chick fluoroquinolone treatment model was established and demonstrated the rapid development of fluoroquinolone resistance in Salmonella Typhimurium mediated by mutations in gyrA for a fully sensitive strain and its’ derivative with a MAR phenotype derived by passage with tetracycline. It was interesting to note a higher level of ciprofloxacin resistance that developed in the derivative with a MAR phenotype compared to the parent strain (MICs of 1 ?g/ml compared to 0.25 ?g/ml for mutation in gyrA only). Mutants arising from the parent strain three weeks after cessation of antibiotic therapy were MAR and had no mutation in gyrA. These were ~ 4 fold less susceptible to ciprofloxacin. In a companion pig study the results were less dramatic, but enrofloxacin treatment was shown to positively select for strains, which already had a mutation in gyrA or were cyclohexane resistant. <P>
For Campylobacter the involvement of mutations in gyrA with quinolone resistance is well established . Analysis of Campylobacter in this project gave evidence of up-regulated efflux, which may be involved in increased resistance to fluoroquinolones. This was investigated. For Campylobacter there was no clear link between resistance to various organic solvents tested and low-level antibiotic resistance. However, detailed analysis of strains showed an association between reduced susceptibility to acridine orange, ethidium bromide and triclosan and reduced susceptibility to some antibiotics including ciprofloxacin, although the differences seen were only ~ 2 to 4 fold (see companion project OD 2004). There was evidence that some (31.3%) of the Campylobacter with the MAR phenotype over-expressed the efflux pump CmeB.
Passage of both Campylobacter and Salmonella in-vitro gave rise to mutants with the MAR-phenotype with increased resistance to unrelated antibiotics including ciprofloxacin. For Salmonella it was interesting to note that MAR strains were more likely to develop gyrA mutations than fully sensitive strains. It was also interesting to note that growth of Salmonella with sub-inhibitory levels of disinfectants increased the subsequent isolation of MAR mutants from agar containing four times the MIC levels of ampicillin, ciprofloxacin or tetracycline for respective strains. For Campylobacter, passage on agar with cefotaxime gave rise to an increase in ciprofloxacin MICs of 0.25 ?g/ml to 8 ?g/ml, a 32-fold increase and this was associated with over-expression of the efflux pump gene cmeB but not cmeD. Passage of Campylobacter with ciprofloxacin gave rise to mutants with increased resistance to ciprofloxacin, which was exclusively associated with mutations in gyrA. Multiple resistant mutants lacked mutations in gyrA and had no changes in expression of efflux pump genes cmeB or cmeF or the major outer membrane porin gene porA.
Data from proteomics studies with Salmonella Typhimurium supported data obtained in the other experiments, and showed that in the presence of ciprofloxacin expression of many proteins was altered. This included TolC, which is the porin component of the acrAB-TolC efflux pump. This may suggest that up-regulation of the AcrAB-TolC efflux pump is in someway mediated by the stress of bacteria growing with ciprofloxacin. Such up-regulation of AcrAB-TolC may give bacteria sufficient resistance to survive what would otherwise be a lethal dose of fluoroquinolone.
In conclusion, these studies have demonstrated the development of fluoroquinolone resistance both in-vitro and in-vivo in both Campylobacter and Salmonella, by mechanisms involving mutations in gyrA and enhanced efflux. The studies demonstrate clearly that fluoroquinolone resistant bacteria can be selected in both the chicken and pig models. Additionally, the studies suggest the double selective pressure of specific antibiotics and specific disinfectants may increase the chance of selecting MAR mutants of Salmonella and these MAR mutants may more readily gain mutations in gyrA than fully sensitive strains when exposed to quinolones. Future work will focus on determining the effect of shorter, but higher doses, of fluoroquinolone treatment in animals or the use of disinfectants that do not select MAR mutants to explore the hypothesis that these will reduce the likelihood of developing resistance.

<P>