Food Testing Methods
Due to the high infectivity of VTEC, there is no established safe exposure to VTEC; thus, methods for the analysis of foods for the presence of VTEC are typically qualitative and composed of the following stages: Enrichment, Detection, Isolation and Confirmation. Enrichment involves the incubation of the sample in broth media to amplify the target pathogen to levels high enough for reliable Detection and Isolation. This stage is necessary, as, due to the high infectivity of some VTEC strains, a limit of detection approaching 1 cell per analytical unit (10 g to 375 g) must be achievable. Detection is an optional stage in which the presence or absence of biomarkers specific for the pathogen in the enrichment broth is determined. Detection analysis typically involves PCR or serological tests. If pathogen-specific biomarkers are not detected, the sample is considered negative and analysis can be terminated; if pathogen biomarkers are detected, the sample is presumptive positive for the pathogen. Isolation involves the recovery of presumptive pathogen cultures from the enrichment media and purification of that pathogen from other organisms present. This step is typically performed by plating on to agar media and provides pure cultures for Confirmation testing. The presence of the pathogen in the sample is confirmed by determination that one or more isolates recovered from the sample possesses the specific characteristics of the pathogen. For VTEC, this is the isolation of Escherichia coli with the potential to produce verotoxin, which can be determined either by production of the toxin or the presence of its genes. The isolates identified are then available for further characterisation, including subtyping.
Due to the mobility of the phage encoded verotoxin genes, VTEC do not constitute a discrete phylogenetic lineage among E. coli, and there are no phenotypic characteristics unique to VTEC, such as antimicrobial resistance or metabolism, which can form the basis for differential/selective media. However, features specific to certain VTEC subpopulations have been identified and form the basis of analytical methods targeting those subgroups. Methods for VTEC O157:H7/NM can take advantage of relatively high resistance to certain antimicrobials (novobiocin, tellurite, and cefiximine) to increase media selectivity and differentiation based upon the absence of sorbitol fermentation and ß-D-glucuronidase activity and the presence of haemolysis on washed sheep’s blood agar (Beutin et al., 1989; March and Ratnam, 1986; Doyle and Schoeni, 1984). Similarly, the absence of rhamnose fermentation can be used as a differential feature for some O26 strains (Murinda et al., 2004). It should be noted that variability in colony morphology and other phenotypic traits can be observed between strains of the same serogroups, including E. coli O157 (Werber et al., 2011; Gill et al., 2014).
In some jurisdictions, methods of analysis have been adopted which target a subgroup of priority VTEC defined by serogroup and the presence of intimin (eae), for example, the USDA FSIS method MLG 5B (USDA-FSIS, 2018) and the European Union (EU) (ISO, 2012). The serogroups designated a priority in the US are O157, O26, O45, O103, O111, O121, and O145, and the current ISO method targets O26, O103, O111 and O121. MLG 5B utilises immunomagnetic separation (IMS), coupled with acid shock and elevated incubation temperatures to improve selectivity. However, this approach excludes VTEC belonging to other serogroups and LEE negative strains even though such strains have been responsible for foodborne outbreaks and serious illness. Additionally, low recovery rates have been observed with IMS for some targeted VTEC serogroups (Kraft et al., 2017; Hallewell et al., 2017) and the higher enrichment temperatures such as 42°C may not be suitable for all strains, particularly cells that have been physiologically stressed.
Methods for the analysis of foods for VTEC validated for regulatory testing in Canada are published in The Compendium of Methods of Analytical Methods (Health Canada, 2018). These include commercial methods, the standard method of analysis for VTEC O157 (MFHPB-10) and the standard for all VTEC (MFLP-52) (Blais et al., 2014). The development of MFLP-52 was informed by the causal and correlative factors previously established through subject matter expert consensus (VTEC Workshop Report, 2010), and which may need to be updated in light of current public health trends. To ensure the inclusivity of the method the enrichment and isolation conditions represent a compromise between permitting the recovery of diverse VTEC strains, while imposing some selectivity against other microbiota (Gill et al., 2012; Gill et al., 2014). The MFLP-52 method for VTEC begins with broth enrichment, followed by PCR screening for stx1 and stx2, if the enrichment is positive for one or more verotoxin genes the sample is presumptive positive. Presumptive positive enrichments are then plated onto agar media and following incubation up to 60 individual colonies are screened by PCR for stx1 and stx2 to identify VTEC (Blais et al., 2014). Confirmation involves a Cloth-based Hybridization Array System (CHAS) for detecting PCR amplicons for the virulence genes stx1 and stx2, eae, hlyA and serogroup markers for O26, O103, O111, O145, and O157. This approach is intended to be inclusive while supporting rapid identification of markers associated with VTEC causing BD and HUS.
The greatest potential for the application of next generation sequencing (NGS) to the analysis of food for VTEC is the genomic characterisation of isolates. As the cost and speed of NGS has fallen the technology has become economically competitive with other methods of isolate characterisation particularly as once a genome has been sequenced the same data can be interrogated for the presence of multiple elements or subjected to multiple forms of subtyping analysis. An additional advantage is a greater flexibility in the analysis of isolates: rather than relying solely on rigidly validated "wet lab" methods, NGS analysis can be adapted ad hoc for the determination of gene markers that may be relevant to a particular food safety event, or in response to criteria indicated by changing public health trends. Regulatory agencies in the international community have been actively developing strategies for the implementation of WGS technology in support of regulatory food inspection objectives through the detection, identification and characterization of priority bacterial pathogens such as VTEC (Tong et al., 2015; Lambert et al., 2015; Lambert et al., 2017, Carrillo et al., 2019).
CFIA laboratories have developed a practical process in which genomic DNA isolated from single colonies is sequenced using the Illumina MiSeq platform, followed by analysis of the sequence data in a fully integrated process for the determination of key genomic markers (Lambert et al., 2015; Blais, 2017). A bioinformatics pipeline named GeneSeekR has been designed to determine salient features such as identity (e.g., species, serotype), risk characterization attributes (e.g., virulence, verotoxin sub-types), molecular type (single nucleotide polymorphism and multi-locus sequence typing analyses) and "value-added" markers (e.g., antibiotic resistance profile for surveillance purposes). Quality assurance metrics are included to indicate the reliability of each analysis, including indicators for sequence data quality and bioinformatics performance. The recently developed ConFinder tool (Low et al., 2019) is designed to identify the presence of contaminating DNA from multiple strains of the same bacterial species. A key output is the Record of Genomic Analysis (ROGA) featuring a standardized reporting format intended to meet the needs of the end-user community (i.e., Food Safety Science Directorate, risk assessors and recall specialists). This technology is currently being used by CFIA to generate WGS data for food isolates in real time. The implementation of this new program provides an unprecedented degree of resolution in the analysis of foodborne bacterial isolates, enabling their timely identification and risk profiling at a cost similar to traditional methods, and has the potential to replace lengthy biochemical characterization and typing procedures used in contemporary food-testing laboratories.
A key capacity afforded by the WGS approach is the ability to rapidly identify important attributes associated with risk, such as the serotype and verotoxin subtypes of VTEC isolates, without the need to forward isolates to specialized reference centres for lengthy analytical procedures. Serotype can be readily determined using the SeroTypeFinder tool (Joensen et al., 2015) hosted by the Centre for Genomic Epidemiology (www.genomicepidemiology.org). Currently, the most widely used standard method for verotoxin subtyping is a PCR technique developed by the Statens Serum Institute (SSI), which differentiates subtypes using subtype-specific primer pairs (Scheutz et al., 2012). More recently, a validated in silico technique based on WGS analysis using an algorithm simulating the SSI PCR process, termed V-Typer, has been incorporated in a bioinformatics pipeline routinely used at CFIA for the characterization of VTEC isolates (Carrillo et al., 2016).
An emerging application of WGS technology is the use of genomic information to inform tailoring of methods of analysis to improve the probability of recovery of specific pathogen strains. For example, it has been demonstrated that investigation of foodborne VTEC outbreaks can be supported by the prediction of antimicrobial resistance of the pathogen strain from genomic data to indicate antimicrobial agents which can be used to customise the selectivity of culture media (Knowles et al., 2016; Blais et al., 2019). AMR analyses can be conducted with the use of publicly available tools such as ResFinder (https://cge.cbs.dtu.dk/services/ResFinder/). Such an approach has been demonstrated to be effective in the recovery of different VTEC strains from ground beef samples containing high levels of background bacteria (Blais et al., 2019). When applied in outbreak investigations for the detection and isolation of specific VTEC strains, this approach can potentially overcome the limited selectivity of the current enrichment conditions.
While testing of VTEC O157 is a routine practice for local, private and hospital laboratories, there are no standardized guidelines for when to test for other VTEC serotypes. In late 2016 and early 2017, Canada experienced an outbreak of VTEC O121 due to contaminated flour (Morton et al., 2017). Due to the lack of standardized guidance on when to test for non-O157 VTEC, concerns regarding missing cases that were part of this investigation were raised by provincial public health laboratories and epidemiologists. As such, interim guidelines were provided to public health laboratories on when to test for non-O157 VTEC to ensure that cases associated with this outbreak were identified. National guidelines to address this lack of standardization have been published by the Canadian Public Health Laboratory Network (Chui et al., 2018).
A challenge to monitoring VTEC in Canadians is the introduction of the use of culture-independent diagnostic tests (CIDT) to diagnose foodborne infections. While traditional methods require the isolation of a foodborne organism from a specimen sample submitted by a patient, these rapid tests identify the causative organism but do not result in an isolate for further subtyping (e.g., identification of serotype) or testing (e.g., antimicrobial susceptibility). As the uptake of CIDT increases, surveillance systems in Canada and internationally will experience erosion of the capacity to conduct surveillance of foodborne pathogens. Particularly vulnerable is the granularity required to distinguish between sporadic and outbreak associated cases. A proposed solution to the increased use of CIDT is conducting reflexive culture – using a CIDT for a panel of enteric pathogens and then culturing positive samples, in order to obtain isolates. These isolates would then be available for further testing and to inform surveillance systems. However, in countries where the use of CIDT has become prevalent and reflex culture has been introduced, several roadblocks have been identified in making this approach a success. The primary issue identified is the cost of the additional laboratory work required for culture and the lack of clarity on who is responsible for covering this cost. If physicians are not requesting cultures to be performed, the costs associated with adding this test are not reimbursed to the local laboratories. If local laboratories do not culture the samples, then these will be required to be sent to public health laboratories, adding a level of complexity due to shipping timelines and costs.
- Methods of analysis for VTEC in food typically have four parts: Enrichment in broth media; Detection of biomarkers for VTEC; Isolation of VTEC from the enrichment; and Characterisation of the VTEC isolates.
- Due to the high infectivity of VTEC and the potentially low levels in contaminated foods, methods of analysis for VTEC in foods must have a limit of detection approaching 1 cell per analytical unit, which may be as large as 375 g. To achieve this sensitivity amplification of VTEC by enrichment is a necessity.
- Isolation of VTEC is necessary to confirm that the virulence markers detected in enrichment represent a viable cell and provide isolates for characterisation.
- Enrichment and Isolation remain challenging as there is currently no selective or differential media available which is specific for VTEC. VTEC are not members of a discrete phylogenetic lineage and so there are no phenotypic characteristics unique to VTEC which can form the basis for differential/selective media.
- There are phylogenetic groups within the VTEC which can be selected/differentiated on the basis of antimicrobial resistance or substrate utilisation.
- Methods for the analysis of foods for VTEC validated for regulatory testing in Canada are published in The Compendium of Methods of Analytical Methods. The standard method of analysis for VTEC O157 is MFHPB-10 and the standard for all VTEC is MFLP-52.
- Next generation sequencing offers numerous advantages and cost savings in the characterisation and subtyping of VTEC isolates.
- In Canada, clinical testing for VTEC O157 is a routine practice. Clinical testing for non-O157 VTEC is more challenging and there has been a lack of established standards. Recently national guidelines to address this lack of standardization have been published by the Canadian Public Health Laboratory Network.
- The increasing use of culture-independent diagnostic tests (CIDT) to diagnose clinical cases of infection by VTEC and other foodborne pathogens may erode public health surveillance. As CIDT does not provide either isolates or data from which subtyping can be conducted to identify clusters or conduct source tracking.