Genomic epidemiology of enteropathogenic Escherichia coli in southwestern Nigeria

Abstract

Background: Enteropathogenic Escherichia coli (EPEC) are etiological agents of diarrhea. We studied the genetic diversity and virulence factors of EPEC in southwestern Nigeria, where this pathotype is rarely characterized.

Methodology/principal findings: EPEC isolates (n = 96) recovered from recent southwestern Nigeria diarrhea case-control studies were whole genome-sequenced using Illumina technology. Genomes were assembled using SPAdes and quality was evaluated using QUAST. Virulencefinder, Ectyper, and ResFinder were used to identify virulence genes, serotypes, and resistance genes. Multilocus sequence typing was done by STtyping. Single nucleotide polymorphisms (SNPs) were called out of whole genome alignment using SNP-sites and a phylogenetic tree was constructed using IQtree. Thirty-nine of the 96(40.6%) EPEC isolates were from diarrhea cases diarrhea. Nine isolates from diarrhea patients and four from healthy controls were typical EPEC, harboring bundle-forming pilus (bfp) genes whilst the rest were atypical EPEC. There were 15 EPEC-EAEC hybrids. Atypical serotypes O71:H19 (16, 16.6%), O108:H21 (6, 6.3%), O157:H39 (5, 5.2%), and O165:H9 (4, 4.2%) were the most prevalent; only 8 (8.3%) isolates belonged to classical EPEC serovars. The largest, ST517 clade harbored multiple siderophore and serine protease autotransporter genes and included an O71:H19 subclade <10 SNPs apart, representing a likely outbreak involving 15 children, four with diarrhea. Likely outbreaks, of typical O119:H6(ST28) and atypical O127:H29(ST7798) were additionally identified.

Conclusion/significance: EPEC circulating in southwestern Nigeria are diverse and differ substantially from well-characterized lineages seen previously elsewhere. EPEC carriage and outbreaks could be commonplace but are largely undetected, hence, unreported, and require genomic surveillance for identification.

Fig 1. Maximum-likelihood whole-genome SNP-based tree of EPEC isolates from this study juxtaposed against source (location, age and health status of patient) and classification (typical/ atypical, classical or non-classical serovar, intimin type, phylogroup and ST) criteria.

Data are available in interactable format at https://microreact.org/project/hnPuLKV1d7pnskcG146MJj-epec-isolates-and-their-locations-2025.

Fig 2. Maximum-likelihood whole-genome SNP-based tree and results of phylogenomic analysis of 61 EPEC isolates from Nigeria in this study and 67 international strains.

The inner ring represents the isolation sources and location, the middle ring represents the strain types, with the locally isolated strains in yellow, outer ring represent the phylogenomic lineages, ND; not determined Local isolates, NC; not classified international strains. Isolates from the Gambia that are closely related to isolates from primary health care centres in this study are marked with red crosses. Data are available in interactive format at: https://microreact.org/project/usSgwCwVrm1AZ76kHqfk1d-epec-local-and-international-strains.

Fig 3. EPEC hybrids carrying pAA genes that define the EAEC pathotype (air, eilA, aggR, aap, aggA, agg4A, aar and capU) and/or ETEC (ltcA)-defining virulence genes.

Fig 4. Characterization of a cluster of ST517 EPEC recovered in Ile-Ife. (a). Epicurve showing the dates of enrollment of children with diarrhea (top) and those from whom ST517 isolates were recovered (bottom). (b). Virulence and antimicrobial resistance genes in ST517 isolate from Ife that, based on sharing <10 SNPs, are believed to constitute an outbreak. (c). Age distribution of individuals from whom EPEC included in the tree in (b) were recovered with the red line representing the median age for each category. (d). The nutritional status of children from whom isolates in (b) above were collected.