Abstract:[Background] Escherichia coli is a Gram-negative bacterium widely existing in the gastrointestinal tract and feces of broilers and the soil, water, and air of the farms. [Objective] To investigate the impacts of seasons on the prevalence and bacterial antimicrobial resistance of E. coli. [Methods] A survey was conducted in summer and winter to gather epidemiological data on E. coli in three farms. Questionnaires were used to collect basic information about the farms. E. coli was isolated by agar plates and identified by PCR. The minimum inhibitory concentrations (MICs) of 14 antibiotics on the isolates were determined, and the resistance genes and virulence genes carried by the isolates were identified. Finally, statistical methods were employed to analyze the correlations between seasons and E. coli isolation rate, between seasons and bacterial antimicrobial resistance, as well as between resistance genes and phenotypes. [Results] The total isolation rate of E. coli in the three farms was 94.7% (284/300), and the isolation rate followed a trend of Farm A (100.0%)>Farm C (93.0%)>Farm B (91.0%). The isolation rate of E. coli in each farm showed not significant difference between summer and winter (P>0.05). The isolation rate in Farm A was higher than those in Farm B and Farm C (P<0.05). The isolates of E. coli from the three farms exhibited multi-drug resistance, being highly resistant to ampicillin, cephalothin, kanamycin, tetracycline, ciprofloxacin, sulfamethoxazole, trimethoprim-sulfamethoxazole, and florfenicol. Moreover, the isolates obtained during summer showed a higher rate of resistance than those obtained during winter. A paired t-test was then performed to investigate the relationship between season and bacterial resistance, and only the resistance to polymyxin E showed seasonal differences (P<0.05). The bacterial antimicrobial resistance gene floR had the highest detection rate, followed by qnrS and mcr-1. The detection rates of blaNDM and aac(6')-Ib were low, and cfr was not detected. The correlations between gene detection results and seasons were not entirely consistent among the three farms. A correlation analysis between drug resistance phenotype and genotype revealed positive correlations between floR and resistance to florfenicol, between blaNDM and resistance to ampicillin and amoxicillin/clavulanate, between mcr-1 and resistance to polymyxin E, as well as between qnrS and resistance to ciprofloxacin (all P<0.05). The detection rates of the virulence genes fimC, yijp, matA, and iss were above 70%, and two strains were defined as pathogenic E. coli. [Conclusion] In summary, E. coli colonization in the gastrointestinal tract is not affected by seasons, with multidrug-resistant strains prevalent in poultry farms. The resistance of E. coli isolates in summer was severer than that in winter, with the resistance to polymyxin E linked to seasons. There exist close correlations between bacterial antimicrobial resistance and resistance genes. According to the results, we suggest that monitoring of bacterial resistance should be enhanced during summer. The findings provide guidance for the breeding practices and precise antibiotic usage and a reference for the prevention of E. coli and rational use of antibiotics in poultry.
