Lineage-specific variation in frequency and hotspots of recombination in invasive Escherichia coli

Abstract

Background

The opportunistic bacterium Escherichia coli can invade normally sterile sites in the human body, potentially leading to life-threatening organ dysfunction and even death. However, our understanding of the evolutionary processes that shape its genetic diversity in this sterile environment remains limited. Here, we aim to quantify the frequency and characteristics of homologous recombination in E. coli from bloodstream infections.

Results

Analysis of 557 short-read genome sequences revealed that the propensity to exchange DNA by homologous recombination varies within a distinct population (bloodstream) at narrow geographic (Dartmouth Hitchcock Medical Center, New Hampshire, USA) and temporal (years 2016 – 2022) scope. We identified the four largest monophyletic sequence clusters in the core genome phylogeny that are represented by prominent sequence types (ST): BAPS1 (mainly ST95), BAPS4 (mainly ST73), BAPS10 (mainly ST131), BAPS14 (mainly ST58). We show that the four dominant clusters vary in different characteristics of recombination: number of single nucleotide polymorphisms due to recombination, number of recombination blocks, cumulative bases in recombination blocks, ratio of probabilities that a given site was altered through recombination and mutation (r/m), and ratio of rates at which recombination and mutation occurred (ρ/θ). Each sequence cluster contains a unique set of antimicrobial resistance (AMR) and virulence genes that have experienced recombination. Common among the four sequence clusters were the recombined virulence genes with functions associated with the Curli secretion channel (csgG) and ferric enterobactin transport (entEF, fepEG). We did not identify any one recombined AMR gene that was present in all four sequence clusters. However, AMR genes mdtABC, baeSR, emrKY and tolC had experienced recombination in sequence clusters BAPS4, BAPS10, and BAPS14. These differences lie in part on the contributions of vertically inherited ancestral recombination and contemporary branch-specific recombination, with some genomes having relatively higher proportions of recombined DNA.

Conclusions

Our results highlight the variation in the propensity to exchange DNA via homologous recombination within a distinct population at narrow geographic and temporal ranges. Understanding the sources of the genetic variation in invasive E. coli will help inform the implementation of effective strategies to reduce the burden of disease and AMR.