Evaluation of whole-genome sequencing for outbreak detection of Verotoxigenic Escherichia coli O157:H7 from the Canadian perspective

- Evaluation of whole-genome sequencing for outbreak detection of Verotoxigenic.
- Escherichia coli O157:H7 from the Canadian perspective.
- Background: Rapid and accurate identification of Verotoxigenic Escherichia coli (VTEC) O157:H7 is dependent on well-established, standardized and highly discriminatory typing methods.
- Currently, conventional subtyping tests for foodborne bacterial pathogen surveillance are rapidly being replaced with whole-genome sequencing (WGS) in public health laboratories.
- Here we aim to verify the performance of WGS in comparison to pulsed-field gel electrophoresis (PFGE) and multiple-locus variable-number tandem repeat analysis (MLVA) for eight retrospective outbreaks of VTEC O157:H7 from the Canadian perspective.
- To provide a useful and consistent starting point for examining laboratory-based surveillance data for VTEC O157:H7 in Canada, we also aim to describe the number of genetic differences observed among outbreak-associated isolates..
- Conclusions: This study validates the superiority of WGS and indicates the BioNumerics wgMLST schema is suitable for surveillance and cluster detection of VTEC O157:H7.
- Keywords: VTEC, Escherichia coli , O157:H7, Whole genome sequencing, Outbreak, wgMLST, SNVPhyl.
- Full list of author information is available at the end of the article.
- 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0.
- As an inte- gral part of the PulseNet International network, Pulse- Net Canada has made immense strides towards routine implementation of WGS for foodborne pathogen sur- veillance.
- In light of the po- tential severity of these infections, timely response to potential outbreaks of VTEC O157:H7 is critical, making this pathogen a prime candidate for this study..
- Since real-time implementation and routine use of WGS in public health laboratories is highly reliant on standardized data analysis tools [22, 26], we also provide a comparative analysis of two popular methodologies for WGS analyses, including a single nucleotide variant (SNV) based approach (i.e., the in-house developed sin- gle nucleotide variant phylogenomics (SNVPhyl) pipe- line) [34] and a wgMLST approach (i.e., the whole genome multilocus sequence type (wgMLST) schema for Escherichia coli-Shigella (BioNumerics, Applied Maths, Belgium)) [35].
- To the knowledge of the authors, this is the first time a SNV-based method has been directly compared to a wgMLST approach for VTEC O157:H7 outbreak detection..
- To further strengthen the findings of the current study, we also aimed to describe the number of allele or SNV dif- ferences observed among outbreak-related isolates using epidemiological information collected through outbreak investigations and the Public Health Agency of Canada ’ s Outbreak Management Division ’ s WGS validation study..
- outbreak-related.
- For most of the out- breaks investigated, the categorization of isolates by WGS as outbreak-related or not outbreak-related was mostly concordant with the results previously obtained for traditional typing methods PFGE and MLVA (Table 1) except for Outbreaks 2, 3, 4 and 6.
- 23 SNVs/ 23 wgMLST alleles from all other isolates included in the outbreak.
- Interestingly, a previously characterized non-outbreak isolate was observed to group with one of the food isolate outliers identified in Outbreak 6 by few genetic differences (1 SNV/ 3 wgMLST alleles).
- Table 1 Impact of Whole Genome Sequencing on the Categorization of Outbreak-Related Isolates.
- PFGE pulsed-field gel electrophoresis, MLVA multilocus variable-number tandem-repeat analysis, WGS whole genome sequencing.
- The food safety investigation identified several ground beef products produced at the same establishment that were contaminated with VTEC O157:H7.
- Three of the six groupings included one or more clinical isolates (Group 4) as well as some food isolates (Group 3 and Group 6), while the remaining 3 groups only contained food isolates (Group 1, Group 2 and Group 5).
- Table 2 Features of Eight Outbreaks of Verotoxigenic Escherichia coli O157:H7 Characterized by Whole Genome Sequencing in Canada Number of Genetic.
- Number of Isolates Sequenced in the Study.
- The clinical iso- late with no exposure information available was ob- served to group with two of the Brand B food isolates in Group 3 by very few genetic differences (1 SNVs/ 0 wgMLST alleles) suggesting this clinical case was likely related to the outbreak.
- This observation was not unexpected as these food isolates were obtained from frozen beef products with different production dates than those represented in the rest of the cluster and had PFGE pattern combinations and MLVA profiles that were distinct from the other outbreak isolates..
- Since SNVPhyl and wgMLST measure isolate relatedness based on different aspects of the genome (i.e., SNVs vs..
- WGS has the capacity to identify isolates demonstrating variant PFGE patterns/MLVA profiles that would otherwise be overlooked with traditional subtyping methods for cluster detection, and as a result, impacted the categorization of cases for 50% of the out- breaks investigated in this study.
- Similarly, WGS proved useful in identifying a non-outbreak isolate that grouped tightly with Outbreak 4 (0–1 SNVs/ 0–8 wgMLST alleles), however, this isolate was observed to have a related MLVA profile (based on PulseNet Canada standardized interpretation guidelines) to that of the other isolates included in this outbreak..
- Had MLVA been routinely used to characterize all VTEC O157:H7 in Canada, this isolate would have been detected prior to the WGS era.
- At the time of the inves- tigation, MLVA was primarily used to provide enhanced resolution to selected outbreaks but was not officially implemented for routine characterization of VTEC O157:H7 in Canada until 2013.
- a previously charac- terized non-outbreak clinical isolate with a variant PFGE pattern and unrelated MLVA profile was observed to group with one of the food isolate outliers from Out- break 6 by few genetic differences (1 SNV/ 3 wgMLST alleles).
- Interestingly, both isolates were collected within the same time frame and in the same province.
- More recently, WGS elucidated a possible link- age between a single case of Listeriosis in the United States and a cross-border recalled lettuce product, which was also identified as a novel vehicle for Listeria at the time of the investigation [36].
- Two outbreaks examined in the study, Outbreak 3 and Outbreak 6, consisted of both clinical and food iso- lates possessing multiple PFGE patterns and MLVA pro- files.
- As well, multiple isolates were identified for both Outbreak 3 (n = 2 clinical isolates) and Outbreak 6 (n = 4 food isolates) that did not group with any isolates in the respective outbreak based on WGS, which could suggest these isolates were unre- lated.
- the case(s) bore a PFGE pattern different from that of the food iso- late, and therefore, was excluded.
- or that the product sam- pling and testing that was completed as part of the food safety investigation went beyond the scope of the impli- cated food product.
- The final scenario is not uncommon during food safety investigations as a variety of inputs and lot codes are often tested to determine the scope of the contamination.
- Due to lack of resources, not all sporadic iso- lates collected during the time frame of the outbreaks examined could be sequenced and was another limitation.
- of the study.
- The detection of potentially related clusters of isolates from multiple years by WGS raises the question of whether narrow and predefined search windows (i.e., illness onset observed within the same 60-day period of time) will still be appropriate for WGS-based analyses, and forces consideration of the broader impacts on la- boratory-based surveillance with respect to how clusters are identified.
- In the present study, this was illustrated by the closer dis- tances seen between Outbreaks 3 and 6 (linked to ground beef ) and Outbreaks 4 and 7 (linked to lettuce)..
- Since WGS has taken center stage, there has been no shortage in the development of open-source bioinformatics tools, SNV-pipelines or commercially available software for ana- lyzing WGS data in the context of foodborne pathogen surveillance and cluster detection .
- In this study, two popular analysis methods were assessed, in- cluding a SNV-based approach (i.e., SNVPhyl) and wgMLST (the analysis method of choice for PulseNet International), to assess their utility and comparability for outbreak detection using eight well-characterized retro- spective outbreaks of VTEC O157:H7 in Canada.
- Again, this was not un- anticipated as the methods evaluate different aspects of the genome.
- In contrast, SNVPhyl examines single nucleotide changes associated with the shared genome (i.e., SNVs that are common to both the reference and the collection of mi- crobial genomes under analysis), and therefore, the num- ber of genetic sites available for comparison is highly dependent on the reference selected and the overall gen- etic diversity of the dataset.
- To the knowledge of the authors, the use of wgMLST for surveillance and cluster detection has not yet been documented for VTEC O157:H7 in the scientific lit- erature, however, the technique has been found useful for other foodborne pathogens including L.
- In a recent publication by Holmes et al., it was demonstrated that the core genome multilocus sequen- cing typing approach (cgMLST) in BioNumerics v7.6 generated comparable resolution to that of the Public Health England single nucleotide polymorphism (SNP) semi-automated bioinformatics pipeline.
- Although a slightly lower level of resolution is not unex- pected as a smaller proportion of the genome is examined for cgMLST opposed to SNV-based approaches, the po- tential for isolates to appear more related than they actu- ally are would suggest cgMLST is not optimal for continuous monitoring activities such as surveillance and/.
- Based on the findings of the present study and ex- perience investigating a wider variety of clusters and outbreaks in Canada, VTEC O157:H7 isolates associ- ated with the same outbreak had 10 or fewer wgMLST allelic differences (<.
- This is reminiscent to the interpretation of PFGE results to define “matches”, which evolved from the use of Tenover ’ s criteria (i.e., number of bands different) to organism-specific guidelines highly informed by the context of the isolates in question and historical baselines [50 – 52].
- Ultimately, a standardized strain nomenclature could be used to automate cluster detection, more specifically, as the ini- tial step in the cluster detection process to identify closely-related isolates.
- This retrospective study serves as proof of principle and clearly demonstrates increased genetic resolution of WGS for cluster detection and outbreak support of VTEC O157:H7 in Canada over traditional subtyping methods (i.e., PFGE and MLVA).
- However, until a global allele database has been established, this study validates the current BioNumerics v7.6 wgMLST schema based on 17,380 loci as a suitable version for surveillance and cluster detection of VTEC O157:H7 in the interim, and has been implemented as the primary tool for surveillance and cluster detection of VTEC O157:H7 within the PulseNet Canada network.
- As with the implementation of any new method, cost considerations must be taken into account and will be highly dependent on the resources and capabilities of the laboratory itself.
- Forty-one non-outbreak related isolates occurring within 60 days of the selected outbreaks were included in the study based on molecular subtype (PFGE and/or MLVA).
- coli O157:H7 infection during routine public health follow up.
- All data was ana- lyzed centrally by the Public Health Agency of Canada’s Outbreak Management Division..
- Centralized analysis of PFGE results was per- formed following published criteria [52] and relatedness of MLVA profiles was measured using PulseNet Canada standardized interpretation guidelines: isolates differing by no more than one repeat at ≤3 loci or up to 3 repeats at a single locus and no variation at VNTR 34 (i.e., the second integer in the MLVA profile), are considered related [5]..
- Whole-genome sequencing was primarily performed at the National Microbiology Laboratory in Winnipeg;.
- The quality of the raw sequence reads was evaluated using FastQC, Kraken, and BioNumerics (BN) v7.6.
- Greater than or equal to 45% of the reads must have mapped to Escherichia coli for the isolate to be sub- mitted for the final stage in the quality assessment.
- (i.e., at least 90% of the 2, 513 loci contained in the BioNu- merics v7.6 core loci schema, which is synchronized with the Enterobase schema [48], must have been identified) and a genome size between 5.0 and 5.6 M base (Mb) pairs were required.
- “good quality”, the corresponding locus or gene must have been identified by both allele-calling algorithms.
- The E_FM and V_FM in- dexes correspond to the expectancy and variance values under the null hypothesis (H 0 ) at α = 0.05, which implies there is no correlation between the topologies of the SNVPhyl tree and wgMLST tree as one is simply a ran- dom shuffle of the other.
- Additional file 1: BioProject and BioSample information for the 250 isolates used in the study.
- PHAC: Public Health Agency of Canada.
- wgMLST: Whole genome multilocus sequence typing.
- WGS: Whole genome sequencing.
- We thank the National Microbiology Laboratory Enteric Diseases, and Science Technology and Core Services personnel for their technical assistance, with a special thank you to Adrian Zetner in the Bioinformatics Section for assisting with the generation of the figure used in this manuscript.
- The datasets generated and/or analyzed during the current study are available in the National Center for Biotechnology Information (NCBI) repository under BioProject PRJNA481261 and BioProject PRJNA454819 (CFIA isolates only)..
- Drafting of the manuscript: JR, RK, RM, and CN.
- The impact of multilocus variable-number tandem-repeat analysis on PulseNet Canada Escherichia coli O157: H7 laboratory surveillance and outbreak support .
- Pulsed field gel electrophoresis: a review of application and interpretation in the molecular epidemiology of infectious disease.
- Public Health Genom.
- Whole-genome sequencing for national surveillance of Shiga toxin – producing Escherichia coli O157.
- Utility of whole-genome sequencing of Escherichia coli O157 for outbreak detection and epidemiological surveillance.
- Multilocus variable-number tandem repeat analysis distinguishes outbreak and sporadic Escherichia coli O157:H7 isolates.
- Highly diverse variable number tandem repeat loci in the E.
- coli O157: H7 and O55: H7 genomes for high-resolution molecular typing.
- enterica Serovar Dublin: development of a method suitable for inter- laboratory surveillance and application in the context of a raw Milk cheese outbreak in France in 2012.
- Real-time whole-genome sequencing for routine typing, surveillance, and outbreak detection of verotoxigenic Escherichia coli.
- Characterization of foodborne outbreaks of Salmonella enterica Serovar Enteritidis with whole-genome sequencing single nucleotide polymorphism-based analysis for surveillance and outbreak detection.
- dissection of an Escherichia coli O157:H7 outbreak by single nucleotide polymorphism analysis.
- Implementation of whole genome sequencing (WGS) for identification and characterization of Shiga toxin-producing Escherichia coli (STEC) in the United States.
- Genomic epidemiology: whole- genome-sequencing – powered surveillance and outbreak investigation of foodborne bacterial pathogens.
- Salmonella whole genome sequencing implementation group.
- Identification of Salmonella for public health surveillance using whole genome sequencing.
- opportunities of whole genome sequencing (WGS) for foodborne pathogen surveillance, source tracking and risk assessment.
- Vision for the implementation of whole genome sequencing (WGS) for global food-borne disease surveillance.
- The utility of multiple molecular methods including whole genome sequencing as tools to differentiate Escherichia coli O157: H7 outbreaks.
- Usefulness of high-quality core genome singlenucleotide variant analysis for subtyping the highly clonal and the most prevalent Salmonella enterica serovar Heidelberg clone in the context of outbreak investigations.
- Salmonella Thompson outbreaks associated with the consumption of chicken shawarma and the usefulness of genome sequencing in the investigation.
- Escherichia coli – Shigella Schema for whole genome sequence typing Release Note.
- Whole genome sequencing for genomics-guided investigations of Escherichia coli O157: H7 outbreaks.
- Implementation of nationwide real-time whole-genome sequencing to enhance listeriosis outbreak detection and investigation.
- A genomic overview of the population structure of Salmonella.
- Expert opinion on the introduction of next-generation typing methods for food- and waterborne diseases in the EU and EEA

Xem thử không khả dụng, vui lòng xem tại trang nguồn
hoặc xem Tóm tắt