Adaptive Mutations and Replacements of Virulence Traits in the Escherichia coli O104:H4 Outbreak Population

Lionel Guy,Lars Engstrand,Ulrika Liljedahl,Jenny Arvén Norling,Ingela Hedenström,Cecilia Jernberg,Siv G E Andersson,Öjar Melefors,Sofie Ivarsson

doi:10.1371/journal.pone.0063027

Lionel Guy, Lars Engstrand + Show 7 more

Open Access

https://doi.org/10.1371/journal.pone.0063027

Copy DOI

Abstract

The sequencing of highly virulent Escherichia coli O104:H4 strains isolated during the outbreak of bloody diarrhea and hemolytic uremic syndrome in Europe in 2011 revealed a genome that contained a Shiga toxin encoding prophage and a plasmid encoding enteroaggregative fimbriae. Here, we present the draft genome sequence of a strain isolated in Sweden from a patient who had travelled to Tunisia in 2010 (E112/10) and was found to differ from the outbreak strains by only 38 SNPs in non-repetitive regions, 16 of which were mapped to the branch to the outbreak strain. We identified putatively adaptive mutations in genes for transporters, outer surface proteins and enzymes involved in the metabolism of carbohydrates. A comparative analysis with other historical strains showed that E112/10 contained Shiga toxin prophage genes of the same genotype as the outbreak strain, while these genes have been replaced by a different genotype in two otherwise very closely related strains isolated in the Republic of Georgia in 2009. We also present the genome sequences of two enteroaggregative E. coli strains affiliated with phylogroup A (C43/90 and C48/93) that contain the agg genes for the AAF/I-type fimbriae characteristic of the outbreak population. Interestingly, C43/90 also contained a tet/mer antibiotic resistance island that was nearly identical in sequence to that of the outbreak strain, while the corresponding island in the Georgian strains was most similar to E. coli strains of other serotypes. We conclude that the pan-genome of the outbreak population is shared with strains of the A phylogroup and that its evolutionary history is littered with gene replacement events, including most recently independent acquisitions of antibiotic resistance genes in the outbreak strains and its nearest neighbors. The results are summarized in a refined evolutionary model for the emergence of the O104:H4 outbreak population.

Highlights

Since the beginning of May 2011 to the beginning of July 2011, we have witnessed the largest outbreak of Shiga toxin-producing Escherichia coli (STEC) reported to date in Europe [1]
Whole-genome SNP Phylogeny We have sequenced the genome of three E. coli strains isolated prior to the outbreak and tested positive in enteroaggregative assays like the outbreak strains
Strain E112/10 was of the same serotype O104:H4 as the outbreak strains, but was isolated a year earlier, in 2010, from a Swede who had been travelling in Tunisia

Summary

Introduction

Since the beginning of May 2011 to the beginning of July 2011, we have witnessed the largest outbreak of Shiga toxin-producing Escherichia coli (STEC) reported to date in Europe [1]. A casecontrol study of 26 infected subjects and 81 control subjects showed that bean sprout consumption explained 100% of the cases [2]. The outbreak resulted in more than 3,000 cases of STEC-diarrhea and more than 800 cases of hemolytic uremic syndrome (HUS), most of which occurred in Germany [1]. Hemolytic uremic syndrome is most often caused by Shiga toxin producing enterohemorragic E. coli (EHEC) strains of serotype O157:H7 that belong to phylogroup E. Early investigations showed that the outbreak strain was of serotype O104:H4 and contained adherence properties similar to enteroaggregative E. coli (EAEC) strains of phylogroup B1 [3,4,5]

Methods

Results

Conclusion