|dc.description.abstract||Antimicrobial resistance is an emerging aspect of antimicrobial therapy. However, few studies have documented this relationship in vivo. The purpose of this study was to describe the relationship between antimicrobial therapy and emerging antimicrobial resistance within the normal flora. The dog was chosen as the model because it is the target species to which therapy is directed and it has a close relationship to humans. Fecal E. coli were chosen as the sentinel organism because of ease of access, the fast reproductive time, high rate of mutability and it is the common cause of disease, particularly urogenital, in dogs and humans. Amoxicillin and enrofloxacin were chosen
as the target drugs because they are common choices for treatment of E. coli associated disease and different resistance mechanisms were anticipated.
We hypothesized that amoxicillin resistance would be short-lived, oriented toward beta-lactams and plasmid mediated, whereas enrofloxacin resistance would persist, but be limited to enrofloxacin. Amoxicillin or enrofloxacin was administered to 8 healthy, antimicrobial-free purpose-bred dogs; no drug was administered to 8 control dogs until resistance was expressed in their fecal E. coli. The drug was then discontinued and monitoring continued until resistance was either resolved or 28 days had passed. Representative bacterial isolates expressing resistance were collected from each dog per group per time point. Each isolate was serotyped and tested for virulence. Each isolate was characterized for the degree of resistance and whether the resistance was to multiple drugs. In addition, representative isolates from each group were serotyped.
Close to 100% of fecal E. coli isolates rapidly became resistant to both drugs. Amoxicillin resistance resolved in 7-9 days. In contrast, all E. coli were eradicated in 4 dogs receiving enrofloxacin. In the remaining dogs, resistance resolved in 11-21 days but did not resolve in one dog by the end of the study. Resistance to both drugs was at least 16 fold higher than the breakpoint (high level). Resistance associated with enrofloxacin but generally not amoxicillin was multidrug resistance (MDR). Amoxicillin therapy induced resistance to penicillins, selected cephalosporins but not carbapenems. Beta-lactamase resistance was due to TEM β-lactamase (TEM). It was horizontally transferred, with the exception of extended-spectrum β-lactamase enzymes (ESBLs), which were detected only in selected isolates. Amoxicillin therapy induced variable phenotypes and genotypes. One dog receiving amoxicillin developed MDR (including enrofloxacin) resistance. For enrofloxacin treated dogs, genotypes within phenotypes were less variable. All isolates expressed TEM β-lactamase mediated by a non-transferrable mechanism. Resistance to fluoroquinolone (FQ) was mediated by double mutations in both gyrA and parC. Mutation in a global regulator SoxS was found and our data indicate that this oxidative stress response regulator may impact MDR mediated by enrofloxacin in fecal E. coli. Among serotypes, more variety occurred in non-MDR than in MDR or in control dogs, with clonality associated with enrofloxacin but not amoxicillin resistance. Virulence factors were limited to cnf-1 and cnf-2 which were detected in only 5 isolates.
In conclusion, these studies indicated that a high level of antimicrobial resistance rapidly evolved in E. coli from dogs treated with either amoxicillin or enrofloxacin. However, resistance resolved more quickly with amoxicillin, and was associated with multiple drugs for enrofloxacin. Non-ESBL beta-lactamase resistance occurred only with amoxicillin and was transmissable. We report for the first time (1) a substitution of alanine for glutamate at codon 84 of parC gene is associated with enrofloxacin in E. coli and (2) a point mutation leading to a substitution of serine for alanine at codon 12 of a general regulator soxS in MDR is associated with enrofloxacin therapy.||en