Testing for Non-0157:H7 Shiga Toxin-Producing Escherichia coli (STEC) Serotypes in Ohio
Jodi Taylor
Agriculture Inspection Manager
Ohio Department of Agriculture, Division of Meat Inspection
International Food Protection Training Institute (IFPTI)
2011 Fellow in Applied Science, Law, and Policy: Fellowship in Food Protection
Abstract
The Ohio Department of Agriculture’s Division of Meat Inspection (DMI) carried out preliminary testing of ground beef samples generated at state-inspected, red meat slaughter facilities for the presence of six non-O157:H7 Shiga toxin-producing E. coli (STEC) organisms (known as the “Big 6”). Specifically, these serotypes include O103, O111, O26, O45, O121, and O145. This pilot study used two different screening methods: the BioGX® and the DuPont BAX® assays. The results of this project were inconclusive with regard to quantifying the level of these organisms in state-inspected facilities. However, both the DMI and the Consumer Protection Laboratory (CPL) expressed the opinion that this study suggests that there is much work, including more extensive validation of the aforementioned screening kits and follow-up confirmation tests, which needs to be done prior to the use of these methods for regulatory purposes.
Background
The Ohio Department of Agriculture’s Division of Meat Inspection (DMI) is a state-administered program that is responsible for the regulation and oversight of approximately 280 meat and poultry facilities throughout the state. The DMI’s main objective is to ensure the safety of the meat and poultry products produced by these regulated establishments. The program has been in operation since July 1969, under the authority of the Wholesome Meat Act of 1967. Under this law, state meat-inspection programs must be maintained in a manner that is deemed “at least equal to” the federal program administered by the United States Department of Agriculture’s Food Safety and Inspection Service (FSIS).
Foodborne illness is estimated to cost the U.S. over $152 billion each year (Scharff, 2010) and account for approximately 48 million illnesses, resulting in 128,000 hospitalizations and 3,000 deaths annually (Centers for Disease Control and Prevention, 2012.) One contributing culprit: pathogenic Escherichia coli. Pathogenic E. coli serotypes have been implicated in numerous outbreaks. Symptoms of E. coli infection include stomach cramps, bloody diarrhea, vomiting, and a low-grade fever. In the most severe cases, hemolytic uremic syndrome (HUS) and death can occur.
The DMI began testing for E. coli, specifically O157:H7, in ground beef in October 2002. Beef carcass testing was added in October of 2004, followed by testing of beef manufacturing trimmings and other non-trim, ground beef components in 2007. The DMI subsequently dropped carcass testing in July 2009 due to a lack of scientific data that supported the effectiveness of that program. The funds previously used for carcass testing were reallocated and used to strengthen the ground beef and beef manufacturing trimmings sampling programs. As per the current ground beef program, each inspected facility that is grinding or regrinding beef products is subject to sampling at a minimum of six submissions per year. If a positive finding is reported, that facility is entered into an intensified sampling program consisting of eight follow-up samples taken from consecutive production lots under a “test and hold” mandate. Currently, the DMI has no testing protocol in place for any other pathogenic serotypes of E. coli.
Because there are few commercially available, validated, and reliable analytical methods for the identification of non-O157:H7 STEC organisms, and due to the difficulty of differentiating pathogenic non-O157:H7 strains from non-pathogenic E. coli, mandated testing has been slow to evolve (Eblen, 2007). However, the FSIS has recently proposed legislation mandating that six additional pathogenic strains of E. coli, known as the “Big 6,” be defined as adulterants and thus be included in testing programs for federally regulated establishments. These “Big 6” serotypes of STEC include O103, O111, O26, O45, O121, and O145. In some countries, such as Australia, Argentina, Canada, and the European Union, illnesses associated with these serotypes are at least as prevalent-if not more so-than the better-known O157:H7 STEC (Eblen, 2007). Since 2007, the Centers for Disease Control and Prevention (CDC) has reported only three outbreaks related to non-O157:H7 STEC serotypes, none of which was traced to beef products. In 2010, an O145 outbreak in the U.S. was traced to shredded lettuce (26 confirmed cases, no deaths). In 2011, an outbreak tied to O104 (not one of the “Big 6”) was associated with contaminated sprouts in Germany and France, where 852 total cases resulted in 32 deaths. In 2012, a six-state outbreak of O26 infections in the U.S. was tied to sprouts (14 cases, no deaths) (Centers for Disease Control and Prevention, 2012).
Although testing for Shiga toxin-producing E. coli (STEC) serotypes other than O157:H7 is not currently mandated by the FSIS, the DMI recognizes the potential food safety issues associated with these serotypes. For this reason, the DMI s proposing to perform preliminary testing to serve as a gauge that will aid the agency’s sampling program in the months to come.
Problem Statement
The DMI has a very rigorous and well-executed sampling program for the detection and monitoring of E. coli O157:H7. However, the agency has yet to establish a baseline for the other “Big 6” serotypes. In addition, commercially available test kits are limited, and validation of these kits on field samples has not been well documented. This study will allow the DMI to determine the next steps in addressing this emerging food safety concern, such as the development of an additional testing program or perhaps conducting additional follow-up baseline studies of a broader magnitude.
Research Questions
Have the analytical methods been validated enough to produce accurate results?
Is the DMI able to regulate for these additional E. coli serotypes?
Can these assays be used to determine if non-O157:H7 STEC is present in DMI-inspected beef slaughter facilities?
Methodology
This study focused on the analysis of one hundred 325-gram ground beef samples randomly collected each day of operation, Monday through Thursday, from beef slaughter operations under the jurisdiction of the DMI. All samples were analyzed by the Ohio Department of Agriculture’s Consumer Protection Laboratory (CPL), an ISO 17025 accredited laboratory. Sample collection began on Monday, August 22, 2011, and ended Thursday, September 29, 2011. The management of all participating facilities and the inspectors assigned to those facilities were briefed with the details of the research.
The beef slaughter establishments selected each month for the DMI’s random E. coli O157:H7 ground beef program were further tested for non-O157:H7 STEC. Although samples came to the CPL with the identification of the establishment (as required by the current O157:H7 sampling program), for the non-O157:H7 portion of the testing, the CPL assigned an internal number to the sample which was not associated with the establishment. Therefore, the non-O157H7 STEC portion of the study was conducted blind. The DMI decided that there would be no regulatory ramifications to any participating establishment based on confirmed positive findings for non-O157:H7 STEC during this study.
Each of the 100 ground beef samples were analyzed within 24 hours of collection by the CPL. Any samples in transit for greater than 24 hours were rejected upon receipt by the CPL. Incoming sample conditions that rendered samples unsuitable for testing were noted and added to the “Comments” section of the Laboratory Information Management System (LIMS).
Initially, all samples were prepped by adding 25 grams of the ground beef to 225 mL of mTSB+N (Modified Trypticase Soy broth plus 20 mg/L novobiocin) following the USDA Microbiology Laboratory Guidebook (MLG) chapter 5, original version “Detection and Isolation of Non O157 Shiga Toxin producing E. coli from Meat Products” guidelines. The resulting enrichment was incubated. Following incubation, DNA was extracted from the sample. Any unused enrichment was refrigerated until the analysis was complete.
The initial run used the commercially available STEC screening assay from BioGX®. Each sample of DNA was analyzed on this assay following the parameters detailed in the MICRO-MDPEN-METH-007, Real Time polymerase chain reaction (PCR) detection of Shiga toxin E. coli utilizing the Cepheid Smart Cycler II platform. This screening assay detects the presence of the encoding Shiga toxin 1 and/or Shiga toxin 2gene (stx1 and stx2, respectively) and/or the uidA gene mutation. If negative for all three gene targets, the sample was reported as “None Deteced.” If the sample was positive for any one of the three gene targets, the DNA was further analyzed using the two BioGX® STEC “Big 6” panels. Each panel is set up to screen for three of the “Big 6” serotypes: panel 1 for serotypes O145, O111, and O26 and panel 2 for O45, O121, and O103. If there were any positive results from panel 1 or panel 2, the CPL attempted to culturally confirm the sample.
Sixty of the original 100 DNA samples were also screened utilizing the DuPont BAX® STEC screening assay (40 samples were inadvertently discarded). The BAX® STEC screening assay screens the samples for the stx gene as well as the Intiman Gene (eae). The BAX® STEC screen assay regards a sample positive only if both the stx and eae genes are positive. If only one of the targets (stx or eae) is present, the sample is considered negative. Samples that tested positive on the BAX® STEC screening assay were then tested on the BAX® Panel 1 (E. coli O26, O111, and O121) and Panel 2 (E. coli O45, O103, and O145). If the sample is positive for the stx, eae, and wzx gene, CPL attempted to culturally confirm the sample (FSIS, 2011).
All potential positive samples for serotypes O145, O111, O26, or O103 were culturally confirmed using immunomagnetic separation (IMS) beads. Currently, the magnetic beads for serotypes O45 and O121 are not commercially available. If any samples were positive for any of the aforementioned four serotypes, the IMS was carried out on the automated Dynal Bead Retriever. Following the IMS, the samples were plated on each of the following: one Rainbow agar plate, one CHROM agar E. coli (CHROME), one Eosin Methylene Blue agar plate (EMB), and one Horse blood agar plate. The plates were read following the incubation period, and a total of 10 to 20 typical colonies (for E. coli) were selected for further analysis. The DNA was extracted from each colony and again analyzed using the PCR panels (from either the BioGX® or the BAX® kit). If both panels were negative, the analysis was stopped, and the samples were reported as “None Detected.” If any panel was positive, the DNA was tested against the serotype-specific antisera. If found to be positive, the sample was analyzed using the VITEK® 2 a microbial identification system. If identified by the VITEK® 2 as an E. coli, the sample was reported as positive for non-O157:H7 E. coli.
Results
This section will summarize the results for both the BioGX® and BAX® methods as well as the limitations encountered with these testing methods. Additionally, due to the labor-intensive measures required to culturally confirm a screen positive, coupled with the high probability of not being able to locate the presumptive colony, the samples identified as potential positives via both testing methods were not culturally confirmed by CPL.
In both testing methods, the sample was enriched by adding 25 g of ground beef to 225 ml of mTSB containing 20 mg/L of novobiocin, which is an antibiotic that is used in the screening process for E. coli O157:H7, as it has been shown to have no effect on the growth E. coli O157:H7. This particular enrichment was used for enriching samples for non-O157 STEC, also following the USDA-MLG chapter 5, original version “Detection and Isolation of Non O157 Shiga Toxin producing Escherichia coli from Meat Products.” As a result of this study, the CPL hypothesized that the concentration of novobiocin used was too high and could have potentially inhibited the growth of any low-level population of non-O157H7 E. coli. The FSIS has since decreased the amount of novobiocin used in the enrichment to 8 mg/L with reference to the current USDA-MLG version (Chapter 5B.01).
The initial BioGX® analysis revealed four potential positive samples for one of the “Big 6” serotypes: two for O26, one for O45, and one for O121. The confirmation steps outlined in the “Methodology” section were carried out, and ultimately eliminated two of the original four potential positive samples (both O26 samples were confirmed to be “None Detected”). The remaining two potential positive samples were unconfirmed for their respective serotypes (O45 and O121), since, again, the magnetic beads for serotypes O45 and O121 were not commercially available. Based on these findings, the testing process resulted in a 2% positive rate.
The main limitation of this initial BioGX® testing method is that the kit does not test for the presence of the Intiman Gene (eae), a protein essential for the intimate attachment and the formation of attaching and effacing lesions on gastrointestinal epithelial cells (Centers for Disease Control and Prevention, 2011). Because the “Big 6” serotypes must be either stx1 or stx2 positive and eae-positive, this initial testing method fell short of being able to definitively screen samples as potentials for one of the “Big 6.”
In addition, the CPL experienced problems with the software used to interpret the results. Users encountered problems with the filter that accounts for other organic material present in the sample, which made distinction of a potential positive difficult.
During this research project, the DMI learned that the FSIS would not be using the BioGX®, but, rather, a customized method from DuPont. The CPL obtained several BAX® test kits from DuPont and analyzed the samples that were retained after the BioGX® testing was complete (60 samples total).
The results obtained from the BAX® method revealed a total of six potential positives. Of those, five could be classified as “true” “Big 6” potential positives. These results represent an 8.6% positive rate (based on the smaller sample size of 58). The sixth sample, although positive for stx1 or 2, was eae-negative, meaning that although the sample may be a pathogenic non-O157:H7 STEC, it isn’t one of the “Big 6.”
While performing the testing, the CPL experienced interference with other E. coli organisms, making it difficult to single out the “Big 6” organisms. The CPL also experienced difficulty with the software needed to execute the analyses, and technical support from DuPont was not readily available. Currently, the software is being modified to improve both the sensitivity and specificity.
Of the four potential positives identified with the BioGX® method, three (O121 and both O26 samples) were also identified with the BAX® method. The O45 serotype was eliminated using the BAX® method, as the serotype tested negative for the eae gene. Three additional samples were identified as potentially positive using the BAX® method (i.e., both stx1 and/or stx2 and eae-positive) that were not identified as such using the BioGX® method (in fact, all three samples tested negative for stx1 and stx2). At this point, it is not known why those four samples were missed by one method but identified by another. Both assay methods require more extensive validation in the laboratory.
Below is a table summarizing all potential positive findings by method and specific serotype:
Conclusions
Based on this evaluation of the two commercially available screening methods, along with the challenge of isolating potentially positive colonies needed to culturally confirm a non-O157:H7 STEC, the DMI concludes that more work needs to be done before incorporating these assays into a regulatory sampling program to monitor non-O157:H7 STEC organisms in the meat industry. Some critical changes have been made to the official USDA-MLG method since this study was conducted (e.g., the decrease in the amount of novobiocin used in the extraction step). More validation studies need to be carried out in the laboratory to evaluate the commercial assays that are currently available in the market before the DMI can officially start testing meat samples for non-O157 STEC. The CPL is also in the process of evaluating the current USDA-MLG method (Chapter 5B.01) that uses custom-made PCR reagents (rather than a commercial kit) for detection of STEC as well as the use of “modified” cultural methods to isolate the STEC. All steps after that selection are futile if the original pathogen was missed during this selection process.
Recommendations
The DMI recommends conducting additional research and validation before these assays are used in mandatory testing programs. If the BioGX® and BAX® kits are validated either by the CPL or another official certifying, body such as AAOC International, the DMI will consider additional testing to again try to ascertain if these non-O157:H7 organisms are present in red meat slaughter facilities under DMI jurisdiction. As mentioned earlier, the CPL is also considering verifying the USDA-MLG method for use in the laboratory.
The results of this study suggest that the FSIS should consider working in conjunction with state programs to conduct validation studies in state-accredited (ISO 17025) labs to attempt to identify and address weaknesses in both the screening and confirmation steps.
Acknowledgments
This project would not have been possible without the dedication and support of Dr. Joanne Brown, my mentor, and the entire staff of the International Food Protection Training Institute (IFPTI), to which I am very grateful for this opportunity. I would also like to express thanks to Maya Achen, Christina Reed, and Renee Schmauch with the CPL, whose time and efforts were invaluable in completing the testing phase of this research. Finally, I would like to thank Dr. Mike Hockman and my fellow coworkers at the DMI for allowing me the time away to complete this Fellowship.
Corresponding Author
Jodi Taylor, Ohio Department of Agriculture, Division of Meat Inspection
Email: Taylor@agri.ohio.gov
References
Centers for Disease Control and Prevention (2012). Outbreaks by pathogen: Escherichia coli (E. coli). Retrieved from http://www.cdc.gov/outbreaknet/outbreaks.html#pathogen
Eblen, Denise (2007). USDA, FSIS, OPHS. Public health importance of non-O157 Shiga toxin-producing Escherichia coli (non-O157 STEC) in the US food supply. Retrieved from
http://www.fsis.usda.gov/PDF/STEC_101207.pdf
Scharff, Robert (2010). Produce Safety Project. Health-related costs from foodborne illness in the United States. Retrieved from http://www.producesafetyproject.org/admin/assets/files/Health-Related-Foodborne-Illness-Costs-Report.pdf-1.pdf
USDA FSIS Laboratory Guidebook (2011). MLG 5B.01 Detection and isolation of non-O157 Shiga-toxin producing Escherichia coli (STEC) from meat products.