Whole genome comparative analysis shows limited Escherichia coli co-occurrence within co-habiting humans and dogs.
11 affiliations
- 1Quadram Institute Bioscience, Norwich Research Park, Norwich, NR4 7UQ, UK.
- 2Centre for Microbial Interactions, Norwich Research Park, Norwich, NR4 7UG, UK.
- 3Department of Electrical Engineering and Electronics, University of Liverpool, Liverpool, L69 7ZF, UK.
- 4Department of Clinical Science and Services, Royal Veterinary College, Hertfordshire, Hatfield, AL9 7TA, UK.
- 5Centre for Interdisciplinary Research in Animal Health (CIISA), Faculty of Veterinary Medicine, University of Lisbon, Lisbon, Portugal.
- 6Associate Laboratory for Animal and Veterinary Sciences (AL4AnimalS), Lisbon, Portugal.
- 7I-MVET, Research in Veterinary Medicine, Faculty of Veterinary Medicine, Lusófona University - Lisbon University Centre, Lisbon, Portugal.
- 8Department of Pathobiology and Population Sciences, Royal Veterinary College, Hertfordshire, Hatfield, AL9 7TA, UK. [email protected].
- 9Quadram Institute Bioscience, Norwich Research Park, Norwich, NR4 7UQ, UK. [email protected].
- 10Centre for Microbial Interactions, Norwich Research Park, Norwich, NR4 7UG, UK. [email protected].
- 11University of East Anglia, Norwich Research Park, Norwich, NR4 7TJ, UK. [email protected].
Abstract
Background
Dogs are one of the most popular companion animals in the UK, with an estimated 13.5 million within UK households. Within these households, human-animal physical interactions may lead to inter-species bacterial sharing, including Escherichia coli. E. coli is an opportunistic pathogen with a broad host range and can contribute to the spread of antimicrobial resistance (AMR). This work aimed to understand the extent of E. coli co-occurrence between hosts within the same household, and between dogs and humans.Results
E. coli was non-selectively cultured from faecal samples obtained from 24 households with at least one human (n = 31) and one dog (n = 27) sample available at the same timepoint. Up to eight presumptive E. coli colonies were selected per sample and underwent short read whole genome sequencing. In total, 217 E. coli genomes were classified into 53 known sequence types (STs) and one unknown ST. Within samples, up to four different STs were present within each human and up to three different STs within a dog. Using a five single nucleotide polymorphism (SNP) threshold, there were 20 E. coli co-occurrences between hosts, including the same and different host species. Four instances involved potential cross-species co-occurrence within the same household. A genome-wide association study using a combined unitig and gene presence and absence approach identified one gene significantly more common in human derived samples; in silico characterisation revealed this gene is likely to have a function related to Type-F conjugative transfer system pilin assembly. Overall, 48.8% of the genomes had at least one genetic determinant associated with AMR, and 17.5% were classified as multidrug resistant. E. coli from humans carried more AMR determinants than E. coli from dogs.Conclusion
There was limited evidence of E. coli co-occurrence between different host species within households, and only one host-associated genetic feature. Humans harboured E. coli associated with more AMR determinants than those from dogs, suggesting that E. coli from humans presented a larger AMR source and AMR risk to other hosts in this dataset. Our findings can help inform risk analysis of E. coli sharing and transmission between humans and their companion dogs.
Related
- Whole genome sequencing and tuberculosis surveillance and control: a critical view from resources limited settings
- Whole-genome sequencing reveals inter-household networks of gut-colonising ESBL-producing Escherichia coli in two rural Malawian districts
- Community and clinical resistomes and mobilomes: a comparative genomic snapshot of Escherichia coli and Klebsiella pneumoniae
- First report of co-occurrence of Anopheles gambiae and An. arabiensis (Diptera: Culicidae): comparative blood meal sources and infection rates in Ethiopia.
- Patient and alignment factors associated with the occurrence and subtype differentiation of degenerative medial meniscus tears: a weighted comparative analysis.
BibTeX
@article{astorga2026whole,
title = {{Whole genome comparative analysis shows limited Escherichia coli co-occurrence within co-habiting humans and dogs.}},
author = {Gabriel Astorga and Samuel J Bloomfield and John Wain and Leonardo de Oliveira Martins and Dipali Singh and Sumeet K Tiwari and George M Savva and Anette Loeffler and Juliana Menezes and Constança Pomba and Siân Marie Frosini and Alison E Mather},
journal = {BMC genomics},
year = {2026},
month = aug,
doi = {10.1186/s12864-026-13236-2},
url = {https://europepmc.org/article/MED/42958017},
note = {CC BY},
}
RIS
TY - JOUR
TI - Whole genome comparative analysis shows limited Escherichia coli co-occurrence within co-habiting humans and dogs.
AU - Gabriel Astorga
AU - Samuel J Bloomfield
AU - John Wain
AU - Leonardo de Oliveira Martins
AU - Dipali Singh
AU - Sumeet K Tiwari
AU - George M Savva
AU - Anette Loeffler
AU - Juliana Menezes
AU - Constança Pomba
AU - Siân Marie Frosini
AU - Alison E Mather
PY - 2026
DA - 2026/08/04
JO - BMC genomics
DO - 10.1186/s12864-026-13236-2
UR - https://europepmc.org/article/MED/42958017
AB - Background Dogs are one of the most popular companion animals in the UK, with an estimated 13.5 million within UK households. Within these households, human-animal physical interactions may lead to inter-species bacterial sharing, including Escherichia coli. E. coli is an opportunistic pathogen with a broad host range and can contribute to the spread of antimicrobial resistance (AMR). This work aimed to understand the extent of E. coli co-occurrence between hosts within the same household, and between dogs and humans.Results E. coli was non-selectively cultured from faecal samples obtained from 24 households with at least one human (n = 31) and one dog (n = 27) sample available at the same timepoint. Up to eight presumptive E. coli colonies were selected per sample and underwent short read whole genome sequencing. In total, 217 E. coli genomes were classified into 53 known sequence types (STs) and one unknown ST. Within samples, up to four different STs were present within each human and up to three different STs within a dog. Using a five single nucleotide polymorphism (SNP) threshold, there were 20 E. coli co-occurrences between hosts, including the same and different host species. Four instances involved potential cross-species co-occurrence within the same household. A genome-wide association study using a combined unitig and gene presence and absence approach identified one gene significantly more common in human derived samples; in silico characterisation revealed this gene is likely to have a function related to Type-F conjugative transfer system pilin assembly. Overall, 48.8% of the genomes had at least one genetic determinant associated with AMR, and 17.5% were classified as multidrug resistant. E. coli from humans carried more AMR determinants than E. coli from dogs.Conclusion There was limited evidence of E. coli co-occurrence between different host species within households, and only one host-associated genetic feature. Humans harboured E. coli associated with more AMR determinants than those from dogs, suggesting that E. coli from humans presented a larger AMR source and AMR risk to other hosts in this dataset. Our findings can help inform risk analysis of E. coli sharing and transmission between humans and their companion dogs.
SN - 1471-2164
ER -