Phylogenetic intermixing reveals stable fly-mediated circulation of mastitis-associated bacteria in dairy settings

Abstract

Stomoxys flies are common blood-feeding pests on dairy farms and are suspected carriers of pathogenic bacteria due to their close association with manure and cattle hosts. While prior studies have used amplicon sequencing and culture-dependent methodologies to characterize the composition of the Stomoxys microbiota, little is known about strain-level acquisition of mastitis-causing bacteria from manure by Stomoxys or the functional diversity of Stomoxys-associated taxa. In this study, we address these key knowledge gaps by using whole genome sequencing to provide the first comparative genomic analysis of Stomoxys-derived Escherichia coli, Klebsiella pneumoniae, and Staphylococcaceae isolates. Our results show that fly and manure isolates collected from the same farm system are phylogenetically interspersed, with subsequent pairwise genome alignments revealing near-identical strains and plasmids shared between the two sources. We further identify a phylogenetic clade of Mammaliicoccus sciuri containing known mastitis agents associated with both flies and manure. Functional analysis reveals that this clade is highly enriched in xylose metabolism genes that are rare across other M. sciuri lineages, suggesting potential niche differentiation within the genus. Collectively, our results provide strong evidence for the acquisition of fecal-associated bacteria by adult Stomoxys flies, confirming the link between biting muscid flies and manure habitats. The intermixing of fly and manure isolates in clinically relevant taxonomic groups strongly suggests that flies serve as carriers of opportunistic mastitis-causing or other fecal-borne pathogens and may serve as important vehicles of pathogen dissemination across the dairy farm environment.IMPORTANCEBovine mastitis causes up to $32 billion dollars in losses annually in the global dairy industry. Opportunistic intramammary pathogens can be transmitted through incidental contact with bacteria in environmental reservoirs like manure. However, factors affecting the abundance, persistence, and spread of these bacteria are not well understood. Our research shows that mastitis pathogens are present in the guts of blood-feeding Stomoxys (stable) flies, which develop in cow feces and bite cows. Genomic analysis of isolates from flies, manure, and mastitis cases reveals that strains and antimicrobial resistance genes are shared between these sources. Further analysis of fly gut isolates shows virulence factors and possible niche specialization, identifying fly-associated clades with known mastitis agents from mastitic cows. This strongly suggests that Stomoxys flies play a role in the carriage and circulation of bovine mastitis pathogens from manure in dairy settings.

Keywords: antimicrobial resistance; cattle; comparative genomics; host-microbe interactions; insect pests; microbiome; pathogen ecology; virulence.

Fig 1

Genome assembly statistics. (A) Histogram depicting the distribution of N50 contig lengths across sequenced genomes. (B) Histogram depicting the number of BUSCO (Benchmarking Universal Single-Copy Orthologs) detected across sequenced genomes. (C and D) Histograms depicting the calculated contamination (C) and completeness (D) levels across sequenced genomes. (E) Violin plots showing the distribution of CDS (coding sequences) across sequenced taxonomic groups. N represents the number of genomes within each taxonomic group.

Fig 2

Phylogenetic intermixing of sequenced E. coli strains. (A) Midpoint rooted maximum-likelihood phylogenetic tree based on a core gene alignment of sequenced E. coli isolates. The rectangular node to the left of the strain name indicates the origin of each isolate as follows: flies (red), manure (blue), and mastitis (yellow). The colored bar to the right of the strain name indicates the MASH phylogroup of each isolate: B1 (green, top), A (light red), E (orange), D (dark blue), and B2 (cyan, bottom). (B) Heatmap depicting the presence-absence matrix (blue indicates presence) of select ARGs (Resfinder) and VFs (VirulenceFinder). *esp refers to strains positive for either espA, espB, or espJ. (C) Bar graph showing the total number of VFs detected in each strain. Stacked colors represent counts of different classes of VFs as follows from light to dark: adhesion, effector delivery system (EFD), invasion, iron uptake, toxin, and other VFs.

Fig 3

E. coli population structure and phylogenetic placement of sequenced E. coli strains. A midpoint-rooted maximum-likelihood phylogenetic tree was constructed based on a core gene alignment of 410 E. coli genomes. Isolates sequenced in this study are marked with a black circle. The node color indicates the origin of each isolate as follows: flies (red), manure (blue), mastitis (yellow), other host (nonclinical) (gray), and other host (clinical) (orange). The phylogroup A E. coli clade is highlighted in red; the phylogroup B1 E. coli clade is highlighted in green.

Fig 4

Phylogenetic intermixing of sequenced K. pneumoniae strains. (A) Midpoint-rooted maximum-likelihood phylogenetic tree based on a core gene alignment of sequenced K. pneumoniae isolates. The rectangular node to the left of the strain name indicates the origin of each isolate as follows: flies (red), manure (blue), and mastitis (yellow). (B and C) Heatmap depicting the presence-absence matrix (blue indicates presence) of select ARGs (Resfinder) (B) and VFs (VFDB) (C). The predicted capsular polysaccharide type is provided to the right of the heatmaps.

Fig 5

K. pneumoniae population structure and phylogenetic placement of sequenced K. pneumoniae strains. A midpoint-rooted maximum-likelihood phylogenetic tree was constructed based on a core gene alignment of 281 K. pneumoniae genomes. Isolates sequenced in this study are marked with a black circle. The node color indicates the origin of each isolate as follows: flies (red), manure (blue), mastitis (yellow), and other (gray).

Fig 6

M. sciuri genomic analysis. (A) Midpoint-rooted maximum-likelihood phylogenetic tree based on a core gene alignment of 281 M. sciuri genomes, including the 35 fly- and 11 manure-derived strains sequenced in this study. Sample origins are colored as follows: flies (red), manure or cattle gastrointestinal tract (blue), bovine mastitis (yellow), cattle skin or nasal cavity (purple), and other (gray). The two fly-associated clusters of interest are highlighted respectively on the tree. (B) PCA clustering of genomes based on gene content as determined by Panaroo. Ellipses were calculated with a 95% confidence interval for the five clusters identified through k-means analysis. (C) Heatmap of Chi-squared standardized residuals (10,000 Monte Carlo simulations) showing the association of isolates from different sources with each k-means cluster. The sample origins were further hierarchically clustered based on Euclidean distances, as shown by the dendrogram. (D) Dot plot of Chi-squared standardized residuals (10,000 Monte Carlo simulations) showing the association of select ECs with M. sciuri k-means clusters. The dot plot shows KEGG ECs for clusters 1 and 4 with a calculated standardized residual value greater than 3.5 and a prevalence greater than 80%. The full list of residual values for all ECs is available in Table S6.

Fig 7

Staphylococcaceae ARG analysis. (A) Bar plot showing the counts of different ARGs found either on the same contig as a plasmid rep gene (light green) or on a contig without a rep gene (dark green). (B) Bar plot showing the counts of ARGs found on the same contig as different plasmid replicon types, as identified by PlasmidFinder. (C) Heatmap depicting the global percent nucleotide similarity of plasmid-associated contigs encoding the lnuA gene. Red indicates a greater percent nucleotide similarity, while blue indicates a lower percent similarity. The associated strain name is labeled on both the rows and columns, with AM399082.1 and AM399081.1 included as reference plasmids. Both rows and columns were hierarchically clustered based on Euclidean distance, with only the column dendrogram shown. The origin and taxonomic identification of the strains are marked below the dendrogram. Four strains (AS0311, AS0554, AS0569, and AS1261) were isolated from samples collected from the Dairy Cattle Center (DCC), a smaller satellite farm of the main sample collection site.