Background: Resistant Escherichia coli from livestock may be found in food. Retail chicken meat contaminated with ESBL-producing bacteria has been documented in several countries. In addition, ciprofloxacin resistance is found to be high in Belgian livestock. Importantly, the mechanisms of fluoroquinolone resistance in veterinary and human clinical isolates are the same. Whole genome sequencing of resistant bacteria can identify the burden of antibiotic resistance and determine relatedness of resistant bacteria in farms.
Materials/methods: In the i-4-1-Health project, presence of ESBL-producing (ESBL-E. coli) and ciprofloxacin-resistant (CiproR-E. coli) E. coli in fecal samples (FecalSwab, Copan Italy) from Belgian broiler (n=15) and pig (n=15) farms was investigated longitudinally. Per farm, three measurements with a six-month interval were performed between September 2017 and April 2019. When present, three ESBL-E. coli and three CiproR-E. coli from each farm and each measurement were sequenced (Illumina MiSeq). Bioinformatic analysis was performed with BioNumerics (Applied Maths). Core-genome multilocus sequence typing (cgMLST) was used to determine clonal relatedness. A threshold of 20 allelic differences for genetic distances was used to distinguish epidemiologically related/unrelated isolates.
Results: The most common ESBL-gene in ESBL-E. coli from broilers (37%) and pigs (38%) was blaCTX-M-1. Of all CiproR-E. coli, 14% showed plasmid-mediated quinolone resistance (oqxAB, qnrS1, qnrB19), 97% showed mutations at hotspots S83L, D87N in GyrA and/or S80I in ParC and 8% also carried an ESBL-gene. Plasmid-mediated colistin resistance was detected among ESBL-E. coli and CiproR-E. coli in 4 pig farms (mcr-2.1 n=2, mcr-1.1 n=1, mcr-5.1 n=2, mcr-9.1 n=5) and 1 broiler farm (mcr-9.1 n=1). The predominant sequence type was ST1158 (7%) in broilers and ST744 (9%) in pigs. Reoccurrence of related ESBL-E. coli and/or CiproR-E. coli within a farm over different measurements was observed in 22/30 farms, indicating recirculation of isolates within farms (Figure-1). Clonally related E. coli were detected among different broiler farms, but not among pig farms.
Conclusions: Recirculation of clonally related E. coli suggests that resistant strains reside in the farm environment. The occurrence of related strains among different broiler farms suggests a common reservoir (e.g. at top of the pyramid in the broiler production system) or transmission of resistant bacteria.