Genotypic analysis of Shiga toxin-producing Escherichia coli clonal complex 17 in England and Wales, 2014-2022

Abstract

Introduction. Shiga toxin-producing Escherichia coli (STEC) are zoonotic, gastrointestinal pathogens characterized by the presence of the Shiga toxin (stx) gene. Historically, STEC O157:H7 clonal complex (CC) 11 has been the most clinically significant serotype; however, recently there has been an increase in non-O157 STEC serotypes, including STEC O103:H2 belonging to CC17.Gap statement. STEC O103:H2 is an STEC serotype frequently isolated in England, although little is known about the epidemiology, clinical significance, associated public health burden or evolutionary context of this strain.Aim. Surveillance data and whole-genome sequencing data were analysed to determine the microbiological characteristics and public health burden of CC17, including the clinically significant serotype O103:H2, in England and Wales.Methodology. Isolates of E. coli belonging to CC17 (n=425) submitted to the Gastrointestinal Bacteria Reference Unit from 2014 to 2022 were whole genome sequenced, integrated with enhanced surveillance questionnaire data and analysed retrospectively.Results. Overall, diagnoses of CC17 infection increased every year since 2014. Most cases were female (58.5%), with the highest proportion of cases belonging to the 0-4 age group (n=83/424, 19.6%). Clinical presentation data identified diarrhoea (92.1%), abdominal pain (72.4%) and blood in stool (55.3%) as the most frequent symptoms, while 20.4% cases were admitted to hospital and 1.3% developed haemolytic uraemic syndrome. The five most common established serotypes were O103:H2 (64.5%), O123:H2 (11.1%), O151:H2 (6.6%), O71:H2 (3.3%) and O4:H2 (2.6%). The majority of CC17 isolates (78.6%) had the stx1a/eae virulence gene combination. Nine outbreak clusters of STEC infections that were mainly geographically dispersed and temporally related were identified and associated with foodborne transmission.Conclusions. Nationwide implementation of PCR to detect non-O157 STEC and improvements to algorithms for the follow-up of PCR-positive faecal specimens is recommended. Enhanced surveillance is necessary to assess the incidence of CC17 infection and overall burden of this CC within the UK population.

Keywords: Shiga toxin-producing Escherichia coli O103:H2; epidemiology; genome sequence; molecular microbiology; surveillance.

Fig. 1.. (a) Annual number of cases belonging to CC17 reported to UKHSA from 2014 to 2022, originating from England and Wales (n=425). (b) Cases per year of E. coli belonging to CC17 (n=425) with stx subtype breakdown are also represented as the different stacks for each bar. EPEC, enteropathogenic E. coli

Fig. 2.. The number of STEC O103:H2 cases (n=257 cases) depicted by solid black line, and total CC17 cases (n=425) depicted by dashed line, by months in England and Wales.

Fig. 3.. (a) Age–sex distribution of total CC17 cases reported to UKHSA [isolates originating from England and Wales (n=424)], where date of birth, sample receipt date and sex were available. (b) Age–sex distribution of STEC O103:H2 cases in England and Wales, where date of birth, sample receipt date and sex were available (n=256). Red indicates female data, and green indicates male data.

Fig. 4.. Population structure of CC17 in England and Wales (n=420) (rooted at midpoint) curated in IQTree2 through an alignment of variant positions against the reference genome AP010958.1 Escherichia coli O103:H2 str. 12009 and visualized in ITOL v6. The ST of isolates is indicated based on the coloured ranges highlighting the isolate labels and clade. In order from inwards to outwards, the year isolates were received is indicated by the first inner colour strip (blue) after the ST labels, while coloured strips designating the stx subtype and continent travelled follow the year. The outer coloured strip ring denotes the nine different SNP clusters identified and grouped according to SNP address (within five SNPs) and SNP distance analysis.

Fig. 5.. AMR profiles of CC17 isolates (n=425). In silico detection of AMR genes was performed using GeneFinder and the UKHSA AMR gene database. AMR profiles are displayed using Upset in R.