Global Phylogeny and F Virulence Plasmid Carriage in Pandemic Escherichia coli ST1193

Abstract

Lower urinary tract, renal, and bloodstream infections caused by phylogroup B2 extraintestinal pathogenic Escherichia coli (ExPEC) are a leading cause of morbidity and mortality. ST1193 is a phylogroup B2, multidrug-resistant sequence type that has risen to prominence globally, but a comprehensive analysis of the F virulence plasmids it carries is lacking. We performed a phylogenomic analysis of ST1193 (n = 707) whole-genome sequences from EnteroBase using entries with comprehensive isolation metadata. The data set comprised isolates from humans (n = 634 [90%]), including 339 (48%) from extraintestinal infection sites, and isolates from companion animals, wastewater, and wildlife. Phylogenetic analyses combined with gene detection and genotyping resolved an ST1193 clade structure segregated by serotype and F plasmid carriage. Most F plasmids fell into one of three related plasmid subtypes: F^-:A1:B10 (n = 444 [65.97%]), F^-:A1:B1 (n = 84 [12.48%]), and F^-:A1:B20 (n = 80 [11.89%]), all of which carry the virulence genes cjrABC colocalized with senB (cjrABC-senB), a trademark signature of F29:A^-:B10 subtype plasmids (pUTI89). To examine the phylogenetic relationship of these plasmids with pUTI89, complete sequences of F^-:A1:B1 and F^-:1:B20 plasmids were resolved. Unlike pUTI89, the most dominant and widely disseminated F plasmid that carries cjrABC-senB, F plasmids in ST1193 often carry a complex resistance region with an integron truncation (intI1_Δ745) signature embedded within a structure assembled by IS26. Plasmid analysis shows that ST1193 has F plasmids that carry cjrABC-senB and ARG-encoding genes but lack tra regions and are likely derivatives of pUTI89. Further epidemiological investigation of ST1193 should seek to confirm its presence in human-associated environments and identify any potential agricultural links, which are currently lacking. IMPORTANCE We have generated an updated ST1193 phylogeny using publicly available sequences, reinforcing previous assertions that Escherichia coli ST1193 is a human-associated lineage, with many examples sourced from human extraintestinal infections. ST1193 from urban-adapted birds, wastewater, and companion animals are frequent, but isolates from animal agriculture are notably absent. Phylogenomic analysis identified several clades segregated by serogroup, all noted to carry highly similar F plasmids and antimicrobial resistance (AMR) signatures. Investigation of these plasmids revealed virulence regions with similarity to pUTI89, a key F virulence plasmid among dominant pandemic extraintestinal pathogenic E. coli lineages, and encoding a complex antibiotic resistance structure mobilized by IS26. This work has uncovered a series of F virulence plasmids in ST1193 and shows that the lineage mimics the host range and virulence attributes of other E. coli strains that carry pUTI89. These observations have significant ramifications for epidemiological source tracking of emerging and established pandemic ExPEC lineages.

Keywords: Escherichia coli; antimicrobial resistance; integron; pandemic; plasmids; virulence.

FIG 1

Phylogenetic core SNV maximum-likelihood tree of E. coli ST1193 genome sequences. The tree is rooted on a representative of ST14, E. coli assembly ESC_NB5808AA_AS. Data placed at isolate tips include (in order) isolate metadata (country, source, and year) and genotypic information (serogroup [O:H], F plasmid subtype, intI1 BLASTn hit size, and presence of senB, fyuA, and irp2), HierCC data at H0 (related isolates at H0 are highlighted), and ARG presence. Highlighted data included the most common country source (Australia), sources other than human, the most common intI1 BLASTn hit, and F plasmid subtypes in their respective rings. The tree scale is in substitutions per site. A group of O75:H5 isolates that were notably distant from the rest of the serogroup is highlighted in pink.

FIG 2

Comparative linear alignments of plasmids from ST1193 with pUTI89. BLASTn identity is split between the two top and bottom sets of alignments, with the range indicated to the right of each pair. Note that some ORF annotations are hidden due to overlap, particularly for IS elements. Hits were restricted to 1,000 bp or more.

FIG 3

Circular whole-genome sequence alignments of isolates predicted to host F⁻:A1:B1, F⁻:A1:B10, and F⁻:A1:B20 plasmids against (A) pUTI89, (B) pAVS0096-a, (C) pMVC530_A, and (D) pMVC465_A. Ordering of isolate whole-genome sequences is consistent between the four panels. BLASTn hit range for the alignment rings was set to between 70% and 100% identity. Gene size for intI1 is shown as the BLASTn hit.

FIG 4

Phylogenetic core SNV maximum-likelihood tree of the Australian E. coli ST1193 O75:H5 clade. The tree is midpoint rooted. The tree scale is in substitutions per site. Sequence ESC_FB3738AA_AS was used as the reference. Metadata and genotypes are presented at the tree tips, with genotypes shown in darker (positive) and lighter (negative) colors on the external rings. Notable clades are also highlighted different colors, with corresponding clade (C) numbers 1 to 10.

FIG 5

Phylogenetic core SNV maximum-likelihood tree of E. coli ST1193 H10 cluster 11740. The tree is midpoint rooted. The tree scale is in substitutions per site. ESC_BA5200AA_AS was used as the reference. Metadata and genotypes are shown at the tree tips, with genotypes shown in darker (positive) and lighter (negative) colors on the external rings. The intI1 data are presented as BLASTn hit size.