Re-analysis of an outbreak of Shiga toxin-producing Escherichia coli O157:H7 associated with raw drinking milk using Nanopore sequencing

Abstract

The aim of this study was to compare Illumina and Oxford Nanopore Technology (ONT) sequencing data to quantify genetic variation to assess within-outbreak strain relatedness and characterise microevolutionary events in the accessory genomes of a cluster of 23 genetically and epidemiologically linked isolates related to an outbreak of Shiga toxin-producing Escherichia coli O157:H7 caused by the consumption of raw drinking milk. There were seven discrepant variants called between the two technologies, five were false-negative or false-positive variants in the Illumina data and two were false-negative calls in ONT data. After masking horizontally acquired sequences such as prophages, analysis of both short and long-read sequences revealed the 20 isolates linked to the outbreak in 2017 had a maximum SNP distance of one SNP between each other, and a maximum of five SNPs when including three additional strains identified in 2019. Analysis of the ONT data revealed a 47 kbp deletion event in a terminal compound prophage within one sample relative to the remaining samples, and a 0.65 Mbp large chromosomal rearrangement (inversion), within one sample relative to the remaining samples. Furthermore, we detected two bacteriophages encoding the highly pathogenic Shiga toxin (Stx) subtype, Stx2a. One was typical of Stx2a-phage in this sub-lineage (Ic), the other was atypical and inserted into a site usually occupied by Stx2c-encoding phage. Finally, we observed an increase in the size of the pO157 IncFIB plasmid (1.6 kbp) in isolates from 2019 compared to those from 2017, due to the duplication of insertion elements within the plasmids from the more recently isolated strains. The ability to characterize the accessory genome in this way is the first step to understanding the significance of these microevolutionary events and their impact on the genome plasticity and virulence between strains of this zoonotic, foodborne pathogen.

Keywords: Foodborne outbreak; Genomic epidemiology; Illumina; Nanopore; Prophage comparison; STEC O157:H7; WGS.

Figure 1

A maximum-likelihood phylogeny showing both Illumina derived and nanopore derived SNP-typing results for samples sequenced in this study.

Figure 2

Easyfig alignment showing the chromosome and loci of prophages in all samples sequenced in this study. Stx-encoding prophage, Red; Prophage-like region, Blue; Locus of Enterocyte Effacement (LEE), Green and other non-stx-encoding prophages, Black.

Figure 3

Neighbour joining tree based on Jaccard distances of stx-encoding prophages of publicly available samples and the outbreak samples sequenced in this study. Prophages are coloured by sub-lineage of STEC O157:H7. Sub-lineage Ia, Green; Ib, Yellow; Ic, Red; I/IIa, Blue; I/IIb, Grey; IIa, Orange; IIb, Black and IIc, Purple.

Figure 4

Mid-rooted neighbour-joining trees of Jaccard distances showing prophages from samples sequenced in this study with prophages from BA000007 (Sakai) (A), STEC O157:H7 strain 9000 (B) and STEC O157:H7 strain 397,404 (C). In each diagram prophages grouped by green are prophages shared in samples and reference genome; red are reference genome only and yellow are prophages unique samples sequenced in this study.

Figure 5

Easyfig alignment of compound prophage 5 with sample 432,300 (top) acting as a reference. Sample 432,299 (bottom) detailing a 47.3kbp deletion.

Figure 6

Easyfig alignment showing exemplar IncFIB plasmids from samples 432,301 (top) and 811,035 (bottom).