Genomic Surveillance of Enterococcus faecium Reveals Limited Sharing of Strains and Resistance Genes between Livestock and Humans in the United Kingdom

Abstract

Vancomycin-resistant Enterococcus faecium (VREfm) is a major cause of nosocomial infection and is categorized as high priority by the World Health Organization global priority list of antibiotic-resistant bacteria. In the past, livestock have been proposed as a putative reservoir for drug-resistant E. faecium strains that infect humans, and isolates of the same lineage have been found in both reservoirs. We undertook cross-sectional surveys to isolate E. faecium (including VREfm) from livestock farms, retail meat, and wastewater treatment plants in the United Kingdom. More than 600 isolates from these sources were sequenced, and their relatedness and antibiotic resistance genes were compared with genomes of almost 800 E. faecium isolates from patients with bloodstream infection in the United Kingdom and Ireland. E. faecium was isolated from 28/29 farms; none of these isolates were VREfm, suggesting a decrease in VREfm prevalence since the last UK livestock survey in 2003. However, VREfm was isolated from 1% to 2% of retail meat products and was ubiquitous in wastewater treatment plants. Phylogenetic comparison demonstrated that the majority of human and livestock-related isolates were genetically distinct, although pig isolates from three farms were more genetically related to human isolates from 2001 to 2004 (minimum of 50 single-nucleotide polymorphisms [SNPs]). Analysis of accessory (variable) genes added further evidence for distinct niche adaptation. An analysis of acquired antibiotic resistance genes and their variants revealed limited sharing between humans and livestock. Our findings indicate that the majority of E. faecium strains infecting patients are largely distinct from those from livestock in this setting, with limited sharing of strains and resistance genes.IMPORTANCE The rise in rates of human infection caused by vancomycin-resistant Enterococcus faecium (VREfm) strains between 1988 to the 2000s in Europe was suggested to be associated with acquisition from livestock. As a result, the European Union banned the use of the glycopeptide drug avoparcin as a growth promoter in livestock feed. While some studies reported a decrease in VREfm in livestock, others reported no reduction. Here, we report the first livestock VREfm prevalence survey in the UK since 2003 and the first large-scale study using whole-genome sequencing to investigate the relationship between E. faecium strains in livestock and humans. We found a low prevalence of VREfm in retail meat and limited evidence for recent sharing of strains between livestock and humans with bloodstream infection. There was evidence for limited sharing of genes encoding antibiotic resistance between these reservoirs, a finding which requires further research.

Keywords: Enterococcus faecium; One Health; genome sequencing; livestock; vancomycin resistant; vancomycin-resistant.

FIG 1

Maximum likelihood tree based on SNPs in the genes core to the 1,442 E. faecium isolates in the collection. Colored rings from inner to outer indicate isolate source, BAPS group, phenotypic resistance to ampicillin (black, resistant; white, susceptible/intermediate), and phenotypic resistance to vancomycin (black, resistant; white, susceptible). Outermost ring indicates the clades, with the gaps in Clade A1 corresponding to BAPS5.

FIG 2

Network analysis of E. faecium isolates from farms and wastewater treatment plants. Lines are drawn between farms/plants sharing isolates less than seven SNPs (A) or 35 SNPs (B) apart, indicating approximately 1 and 5 years of evolution, respectively. Lines only show binary links between farms, with arbitrary line lengths. Blue square, wastewater treatment plant; pink circle, pig farm; light blue circle, cattle farm; black circle, turkey farm; yellow circle, chicken farm.

FIG 3

Principal component analysis based on the presence/absence of E. faecium accessory genes across the study collection. Panels A and B show principal component (PC) 1 (x axis) against PC2 (y axis) labeled by BAPS group (A) or isolate source (B). Panels C and D show PC3 (x axis) against PC4 (y axis) labeled by BAPS group (C) or isolate source (D). Principal components 1, 2, 3, and 4 explain 46%, 5%, 3%, and 2% of the variation within the accessory genome, respectively.

FIG 4

Genome-wide association study to detect genes that were overrepresented in E. faecium isolates from livestock and humans. (Left) maximum likelihood tree based on SNPs in the core genes of farm animal and human invasive isolates (n = 849). First two columns show the source species (red, human; pink, pig; light blue, cattle; black, turkey; yellow, chicken) and BAPS group (dark blue, BAPS1; gray, BAPS2; red, BAPS3; light green, BAPS4; pink, BAPS5; orange, BAPS6; yellow, BAPS7; dark green, BAPS8; light blue, BAPS9). Further columns are based on the accessory genome and show the presence (red) or absence (white) of genes strongly associated with humans (A) or animals (B). Columns are ordered based on the pattern of presence and absence. Scale bars indicate ∼30,000 SNPs.

FIG 5

Genetic analysis of antibiotic resistance genes in E. faecium. (A) Venn diagram showing antibiotic resistance genes identified in each reservoir. (B) Bubble graph showing the prevalence of genes in isolates from humans versus animals. Lower graph shows an expanded view of low-prevalence genes from the upper graph. Size of the bubble represents the number of isolates that the gene was identified in. Bubbles are colored by antibiotic class.

FIG 6

Comparison of antibiotic resistance gene variants in E. faecium isolates from humans, livestock and wastewater. Each figure section shows a phylogenetic tree of an antibiotic resistance gene based on SNPs. Pie charts show the source of the isolates in which each gene variant was found (see key for colors), and the size of the pie chart indicates the number of isolates that variant was identified in (see key).