Within-farm dynamics of ESBL-producing Escherichia coli in dairy cattle: Resistance profiles and molecular characterization by long-read whole-genome sequencing

A longitudinal study was designed in five dairy cattle farms to assess the within-farm dynamics of ESBL-/AmpC-/carbapenemase-producing E. coli and their resistance profiles, along with the genes conferring the resistance phenotypes. Twelve samplings were performed over a period of 16 months, collecting rectal feces from apparently healthy animals in three age groups (calves, heifers, and lactating cows) that were subjected to selective isolation in cefotaxime-containing media. Minimum inhibitory concentrations were determined by broth microdilution for 197 cefotaxime-resistant E. coli (1-3 isolates per age group and sampling date), and 41 of them were selected for long-read whole-genome sequencing. Cefotaxime-resistant E. coli were detected in the five farms, but isolation frequency and resistance profiles varied among farms and age groups. The genetic profiling of a selection of isolates recovered in two of the farms was described in full detail, showing the predominance of a few genomic subtypes of E. coli in one farm (F1) and great variability of strains in another one (F4). Two predominant distinct strains carrying the bla(CTX-M-1) gene in IncX1 plasmids successively spread and persisted in F1 over a prolonged period. In F4, 13 different MLST types carrying a high diversity of ESBL-encoding genes in 6 different plasmid types were observed, probably as the result of multiple source contamination events. In both farms, the presence of certain plasmid types with the same repertoire of ARGs in different E. coli STs strongly suggested the occurrence of horizontal transfer of such plasmids among strains circulating within the farms. Considering the public health importance of ESBL-producing E. coli both as pathogens and as vectors for resistance mechanisms, the presence of beta-lactamase- and other AMR-encoding genes in plasmids that can be readily transferred between bacteria is a concern that highlights the need for One Health surveillance.