Surveillance of Enterococcus spp. reveals distinct species and antimicrobial resistance diversity across a One-Health continuum

Abstract

For a One-Health investigation of antimicrobial resistance (AMR) in Enterococcus spp., isolates from humans and beef cattle along with abattoirs, manured fields, natural streams, and wastewater from both urban and cattle feedlot sources were collected over two years. Species identification of Enterococcus revealed distinct associations across the continuum. Of the 8430 isolates collected, Enterococcus faecium and Enterococcus faecalis were the main species in urban wastewater (90%) and clinical human isolates (99%); Enterococcus hirae predominated in cattle (92%) and feedlot catch-basins (60%), whereas natural streams harbored environmental Enterococcus spp. Whole-genome sequencing of E. faecalis (n = 366 isolates) and E. faecium (n = 342 isolates), revealed source clustering of isolates, indicative of distinct adaptation to their respective environments. Phenotypic resistance to tetracyclines and macrolides encoded by tet(M) and erm(B) respectively, was prevalent among Enterococcus spp. regardless of source. For E. faecium from cattle, resistance to β-lactams and quinolones was observed among 3% and 8% of isolates respectively, compared to 76% and 70% of human clinical isolates. Clinical vancomycin-resistant E. faecium exhibited high rates of multi-drug resistance, with resistance to all β-lactam, macrolides, and quinolones tested. Differences in the AMR profiles among isolates reflected antimicrobial use practices in each sector of the One-Health continuum.

Figure 1

(A) Sampling locations in the province of Alberta, Canada, for the isolation of Enterococcus spp. (B) Proportion of Enterococcus species isolated from beef processing (abattoir and retail meat), bovine feces, feedlot catch basin, surface water (natural water resources), urban wastewater and human clinical isolates (Calgary Laboratory Services, Calgary, AB), including both selective (erythromycin-resistance) and non-selective isolates, as determined by sequencing of the groES – groEL intergenic region and comparison against the Enterococcus spp. groES-EL intergenic region custom database.

Figure 2

(A) Overall frequency of AMR phenotypes in enterococci species isolates from across the one-health continuum. Upper panels for E. hirae, E. faecalis and E. faecium indicate prevalence of resistance to various numbers of drugs, while lower panels indicate drug resistance class profiles; (B) Correlation between isolation sources, Enterococcus species and antimicrobial resistance phenotypes (a measure of association was determined through Pearson’s correlation analysis); (C) Comparison of proportion of phenotypic AMR between enterococci isolated on antibiotic-free and macrolide (erythromycin) selective media. BP, beef processing (abattoir & retail meat); BF, bovine feces; CB, catch basin; NWS, natural water source; WW, waste water; clinical NS, clinical non-sterile; clinical S, clinical sterile; clinical VRE, clinical vancomycin resistant. ND (not detected/not determined/no data) indicates lack of data-points due to the absence or low detection of a particular Enterococcus species in a sector of continuum.

Figure 3

Phylogenomic relatedness tree constructed based on analysis of single-nucleotide polymorphisms (SNPs) of the core genes of E. faecalis genomes (n = 366) isolated from various environments and samples related to beef production and processing systems, environment, and human-related sources. The genomes were compared using E. faecalis OG1RF genome (GenBank accession # NC_004668/CP002621.1) as a reference. Phylogenomic analysis of the WGS data was conducted using SNVPhyl and the resulting newick tree file was visualized in FigTree v1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/) to obtain phylogenomic dendrogram.

Figure 4

Phylogenomic relatedness tree constructed based on analysis of single-nucleotide polymorphisms (SNPs) of the core genes of E. faecium genomes (n = 342) isolated from various environments and samples related to beef production and processing systems, environment, and human-related sources. The genomes were compared using E. faecium DO genome (GenBank accession# CP003583.1) as a reference. Phylogenomic analysis of the WGS data was conducted using SNVPhyl and the resulting newick tree file was visualized in FigTree v1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/) to obtain phylogenomic dendrogram.

Figure 5

Detection of antimicrobial resistance genes (ARGs) in whole genome sequenced E. faecalis (A) and E. faecium (C) and their correlation plots respectively (B,D) to demonstrate association between the isolate sources and ARGs. A measure of association was determined through Pearson’s correlation analysis.