Antibiotic resistance among Escherichia coli and Salmonella isolated from dairy cattle feces in Texas.

In several developing countries, studies on antimicrobial resistance among bacteria from food animals are rare mostly because of under-resourced laboratories. The objective of this study was to develop and field-test a low cost protocol to estimate the isolate- and sample-level prevalence of resistance to critically important antibiotics among Escherichia coli and Salmonella isolated from dairy cattle feces. Using a predesigned protocol, fecal samples were collected to isolate non-type-specific E. coli and Salmonella using selective media without antibiotic supplements. Besides, samples were screened for E. coli and Salmonella isolates not susceptible to third-generation cephalosporins and quinolones using selective media supplemented with cefotaxime (1.0 μg/mL) and ciprofloxacine (0.5 μg/mL), respectively. All bacterial isolates were further tested for antibiotic susceptibility using disk diffusion. Bacterial isolates not susceptible to third-generation cephalosporins were tested for extended spectrum beta-lactamase (ESBL) phenotype using the combination disk test. Molecular methods were performed on selected bacterial isolates to identify and distinguish genetic determinants associated with the observed phenotypes. Among 85 non-type-specific E. coli isolated from MacConkey agar without antibiotics, the isolate-level prevalence of resistance to tetracycline was the highest (8.2%). Among 37 E. coli recovered from MacConkey agar with cefotaxime, 56.8% were resistant ceftriaxone. Among 22 E. coli isolates recovered from MacConkey agar with ciprofloxacin, 77.3% and 54.5% were resistant to nalidixic acid and ciprofloxacin, respectively. Sixteen Salmonella were isolated and only one demonstrated any resistance (i.e., single resistance to streptomycin). Among E. coli isolates not susceptible to ceftriaxone, an AmpC phenotype was more common than an ESBL phenotype (29 versus 10 isolates, respectively). Whole genome sequencing showed that phenotypic profiles of antibiotic resistance detected were generally substantiated by genotypic profiles. The tested protocol is suited to detecting and estimating prevalence of antimicrobial resistance in bacteria isolated from food animal feces in resource-limited laboratories in the developing world.