Whole-genome sequencing-based antimicrobial resistance and shedding dynamics of Escherichia coli isolated from calves before and after antimicrobial group treatments

Abstract

The fattening of calves is often associated with high antimicrobial use and the selection of antimicrobial resistance (AMR). The objective of this observational longitudinal study was to describe the AMR and strain dynamics, using whole-genome sequencing (WGS), of fecal Escherichia coli in a cohort of 22 calves. All calves received antimicrobial group treatments on Day (D) 1 (oxytetracycline, intramuscularly) and on D4 through D12 (doxycycline, in-feed). Additionally, eight calves received individual parenteral treatments between D7 and D59, including florfenicol, amoxicillin, marbofloxacin, and gamithromycin. Rectal swabs were collected from all calves on D1 (prior to treatment), D2, D9, and D82. The swabs were spread onto Enterobacterales-selective agar, and three E. coli colonies per plate were subjected to WGS. Out of 264 isolates across all calves and sampling times, 80 unique strains were identified, a majority of which harbored genes conferring resistance to tetracyclines, streptomycin, and sulfonamides. The diversity of strains decreased during the in-feed antimicrobial group treatment of the calves. On D82, 90% of isolates were strains that were not isolated at previous sampling times, and the median number per strain of AMR determinants to tetracyclines, florfenicol, β-lactams, quinolones, or macrolides decreased compared to D9. Additionally, clonal dissemination of some strains represented the main transmission route of AMR determinants. In this study, WGS revealed important variations in strain diversity and genotypic AMR of fecal E. coli over time in calves subjected to group antimicrobial treatments.

Importance: The continued emergence and spread of antimicrobial resistance (AMR) determinants are serious global concerns. The dynamics of AMR spread and persistence in bacterial and animal host populations are complex and not solely driven by antimicrobial selection pressure. In calf fattening, both antimicrobial use and carriage prevalence of antimicrobial-resistant bacteria are generally recognized as high. This study provides new insights into the short-term, within-farm dynamics and transmission of AMR determinants in Escherichia coli from the dominant fecal flora of calves subjected to antimicrobial group treatments during the rearing period. The diversity of E. coli strains decreased over time, although, in contrast to previous observations in extended-spectrum β-lactamase-producing Enterobacterales, the predominance of a few clones was not observed. The spread of AMR determinants occurred through the dissemination of clonal strains among calves. The median number per strain of AMR determinants conferring resistance to selected antimicrobials decreased toward the end of the rearing period.

Keywords: Escherichia coli; antimicrobial resistance; cattle; fecal carriage; molecular epidemiology.

Fig 1

Phylogenetic tree representing the genetic relationship between the 155 Escherichia coli isolated from calves, based on core genome MLST complex type (CT; outer circle). Isolate numbers (inner circle) are composed of the calf ID and sampling time, respectively. Each color represents a calf, and the position of each colored band represents the sampling time at which the strain was isolated. Within a CT, when non-identical within-sample isolates (same calf and time) are shown, their difference is indicated as the presence or absence of resistance genes (star), virulence genes (diamond), plasmid incompatibility group (triangle), or O type (circle).

Fig 2

Distribution among calves and dynamics over time of Escherichia coli isolates, with respective core genome MLST CT, sequence type (inner circle color), and determinants of resistance to antimicrobial drugs used during the study (outer circle color). For instance, on D1, calves 606, 602, and 614 carried isolates of different CT belonging to ST58 (same inner circle color) and harbored different resistance determinants’ combinations (different outer circle colors) or lack thereof. In cases of within-sample isolates with identical CT and shown resistance determinants (clonal or not), a single one of the strains is shown for improved clarity. Resistance determinants written in red represent an addition to the temporally previously observed combination of CT and resistance determinants. Connecting lines between isolates at different time points indicate clonal isolates based on CT. Superscripts indicate which calves and at which time interval between samplings received individual antimicrobial treatments: amoxicillin (a), florfenicol (f), gamithromycin (g), and marbofloxacin (m; first and second treatment incidence).

Fig 3

Distribution of the number of E. coli strains according to the number of resistance determinants they harbored against different antimicrobial drugs or classes (tetracyclines, florfenicol, β-lactams, quinolones, and macrolides) at four sampling times. Within a strain, multiple determinants against a single class were counted as one. After the removal of within-sample replicates, the total number of isolates on D1, D2, D9, and D82 was 38, 41, 37, and 39, respectively.

Fig 4

Distribution of the number of calves carrying at least one E. coli harboring selected AMR determinants or combination thereof, at four sampling times. Antibiotic resistance determinants and their function: GyrA_D87N, DNA gyrase subunit A with substitution of aspartic acid (D) to asparagine (N) at codon 87 for quinolone resistance; GyrA_S83L, DNA gyrase subunit A with substitution of serine (S) to leucine (L) at codon 83 for quinolone resistance; ParC_S80I, DNA topoisomerase 4 subunit A with substitution of serine (S) to isoleucine (I) at codon 80 for quinolone resistance; Tet(A), Tet(B), and Tet(C), tetracycline efflux; Tet(M), ribosomal protection for tetracycline resistance; OXA-1, TEM-1, and TEM-35, β-lactamase; FloR, phenicol efflux; and Mph(B), macrolide phosphotransferase.