Antimicrobial resistance profiles among Escherichia coli strains isolated from commercial and cooked foods.

A total of 4330 food samples of which microbiological standard for Escherichia coli is negative in Korea were determined for the frequency of E. coli. Ninety six samples (2.2%) were positive for E. coli. Detection rate of E. coli varied significantly by food type and ranged from 0.3% to 10.9%. Seasoned raw meat (yukhoe) and cold bean-soup had the highest prevalence for E. coli (10.9%) followed by gimbap (5.2%), meat broth for cold noodle (2.9%) and sprout (2.1%). E. coli isolates (n=96) were investigated for their phenotypic and genotypic antimicrobial resistance patterns. Seventeen E. coli isolates (17.7%) were resistant to one or more antimicrobial agents tested. High rates of resistance to the following drugs were observed: tetracycline (15.6%), streptomycin (12.5%), ampicillin (10.4%), nalidixic acid (9.4%) and ticarcillin (9.4%). All ampicillin resistant isolates were screened for extended-spectrum β-lactamase (ESBL) production by the combination disk test. None of the E. coli isolates produced ESBLs. Seventeen out of 96 E. coli isolates which were resistant to at least one antibiotic were investigated by PCR for the presence of 3 classes of antimicrobial resistance genes (tetracycline, aminoglycosides and beta-lactams). The tetracycline resistance genes tetA and tetB were found in 7 and 5 isolates, respectively. The aminoglycoside resistance genes, strA/B, aphA1, aadA and aac(3)-IV were found in 9, 5, 2 and 2 isolates, respectively. The beta-lactam resistance gene, bla(TEM) was found in 7 isolates. Results of this study show that 13 E. coli isolates were multidrug resistant (to three or more antibiotics) and 12 isolates carried at least one antimicrobial resistance gene. These isolates can act as the reservoir for antimicrobial resistance genes and facilitate the dissemination of these genes to other pathogenic and commensal bacteria. Adequate intervention to reduce microbial contamination of these foods is strongly recommended.