Whole Genome Sequencing of Avian Pathogenic Escherichia coli Causing Bacterial Chondronecrosis and Osteomyelitis in Australian Poultry

Abstract

Bacterial chondronecrosis with osteomyelitis (BCO) impacts animal welfare and productivity in the poultry industry worldwide, yet it has an understudied pathogenesis. While Avian Pathogenic Escherichia coli (APEC) are known to be one of the main causes, there is a lack of whole genome sequence data, with only a few BCO-associated APEC (APEC_BCO) genomes available in public databases. In this study, we conducted an analysis of 205 APEC_BCO genome sequences to generate new baseline phylogenomic knowledge regarding the diversity of E. coli sequence types and the presence of virulence associated genes (VAGs). Our findings revealed the following: (i) APEC_BCO are phylogenetically and genotypically similar to APEC that cause colibacillosis (APEC_colibac), with globally disseminated APEC sequence types ST117, ST57, ST69, and ST95 being predominate; (ii) APEC_BCO are frequent carriers of ColV-like plasmids that carry a similar set of VAGs as those found in APEC_colibac. Additionally, we performed genomic comparisons, including a genome-wide association study, with a complementary collection of geotemporally-matched genomes of APEC from multiple cases of colibacillosis (APEC_colibac). Our genome-wide association study found no evidence of novel virulence loci unique to APEC_BCO. Overall, our data indicate that APEC_BCO and APEC_colibac are not distinct subpopulations of APEC. Our publication of these genomes substantially increases the available collection of APEC_BCO genomes and provides insights for the management and treatment strategies of lameness in poultry.

Keywords: ST117; ST57; ST69; ST95; avian pathogenic E. coli (APEC); bacterial chondronecrosis; bacterial osteomyelitis; lameness; microbial genomics; poultry production.

Figure 1

Pangenome-derived core-genome phylogeny visualizing relatedness of 205 APEC_BCO strains. Tip points are colored based on the site of isolation for a given strain, while primary tip labels adjoined with dotted lines indicate strain names and are colored based on sequence type. The next outermost band details the ST and e-serotype combinations. Phylogroups for individual strains are denoted on the outermost lettered band adjacent to broad, colored bars highlighting the general correspondence of clades and phylogroups. The tree is midpoint rooted.

Figure 2

Carriage of IncF RSTs by MLST. Cell values represent the frequency (as a count) with which an IncF RST was identified within a given ST. Cells are colored based on their column-wise representation within a given ST (as a percentage), where the shades of red and their corresponding percentages shown in the key (top right).

Figure 3

Virulence traits identified in APEC_BCO genomes. A pangenome-derived core-genome phylogeny is presented atop a table visualizing virulence genotypes. Tip points are colored based on the site of isolation for a given strain, while primary tip labels adjoined with dotted lines indicate strain names and are colored based on sequence type to allow viewers to follow the uppermost band denoting primary STs within the collection. The next outermost band details the ST and e-serotype combinations. Phylogroups for individual strains are denoted on the outermost lettered band, which details the best-viewed digital format. Below, virulence gene names are shown to the right of a given row alongside the frequency of a given gene (as both a count and percentage) across the APEC genomes. Virulence gene (row) order is clustered using strain-wise Euclidean distances of virulence profiles, such that genes which tend to co-occur are grouped together. The clusters were then highlighted and numbered sequentially (left).

Figure 3

Virulence traits identified in APEC_BCO genomes. A pangenome-derived core-genome phylogeny is presented atop a table visualizing virulence genotypes. Tip points are colored based on the site of isolation for a given strain, while primary tip labels adjoined with dotted lines indicate strain names and are colored based on sequence type to allow viewers to follow the uppermost band denoting primary STs within the collection. The next outermost band details the ST and e-serotype combinations. Phylogroups for individual strains are denoted on the outermost lettered band, which details the best-viewed digital format. Below, virulence gene names are shown to the right of a given row alongside the frequency of a given gene (as both a count and percentage) across the APEC genomes. Virulence gene (row) order is clustered using strain-wise Euclidean distances of virulence profiles, such that genes which tend to co-occur are grouped together. The clusters were then highlighted and numbered sequentially (left).

Figure 4

Box and jitter plot visualizing pairwise SNP distances (y-axis) between strains sharing a sequence type (x-axis). (A) Comparison between APEC_BCO and APEC_colibac collections, (B) within APEC_colibac collection, and (C) within APEC_BCO strains. Strain pairs of only three predominant STs are shown and colored by ST; ST117 (red), ST95 (yellow), and ST57 (green). Note that only strain pairs differing by 100 or fewer SNPs and those of ST117, ST57, and ST95 are included in this diagram—a similar figure containing all strain pairs within these three STs are shown in Figure S2.