An Independent Laboratory Evaluation of the Invisible Sentinel Veriflow E. coli O157:H7 PCR Assay for the Detection of Escherichia coli O157:H7

An Independent Laboratory Evaluation of the Invisible Sentinel Veriflow E. coli O157:H7 PCR Assay for the Detection of Escherichia coli O157:H7 Performance Tested Methods SM Independent Laboratory Validation October 9, 2014 Ben Bastin, Jonathan Flannery, Patrick Bird, Erin Crowley, James Agin, David Goins Q Laboratories, Inc., 1400 Harrison Ave, Cincinnati, OH 45214 An independent validation of the Invisible Sentinel (Philadelphia, Pennsylvania) Veriflow E. coli O157:H7 PCR Assay for the detection of E. coli O157:H7 was conducted at Q Laboratories, Inc. (Cincinnati, Ohio). The study included inclusivity and exclusivity evaluations as well as a matrix study of 3 foods: raw ground beef (80% lean, 325 g), 2% milk (25 g), and leafy spinach (200 g). The inclusivity and exclusivity evaluations consisted of analyzing fifty (50) target isolates of E. coli O157:H7 and thirty-five (35) closely related non-target isolates. For the matrix study, all samples were assayed by the Veriflow E. coli O157:H7 method after 18 hours of incubation at 42.0 C. Twenty (20) replicates of each food matrix were analyzed at a low inoculum level of 0.2-2 CFU/test portion, five (5) replicates were analyzed at a high level of 2-5 CFU/ test portion, and five (5) control replicates were analyzed at 0 CFU/ test portion. The Veriflow E. coli O157:H7 PCR Assay correctly identified all fifty (50) of the E. coli O157:H7 isolates analyzed for inclusivity. For the non-target exclusivity organisms, all thirty-five (35) organisms were correctly excluded. The Probability of Detection (POD) statistical model was utilized to compare the number of presumptive positive results to the number of confirmed positive results. For all 3 matrices, there were no statistically significant differences between the number of presumptive and confirmed positive samples. The results of this evaluation demonstrate the specificity and sensitivity of the Invisible Sentinel Veriflow E. coli O157:H7 PCR assay and its ability to accurately detect Escherichia coli O157:H7 in select foods in less than 24 hours post enrichment for raw ground beef, 2% milk, and leafy spinach. This report presents the analytical results for the comparison of the Invisible Sentinel Veriflow E. coli O157:H7 Assay to the USDA/FSIS-MLG 5.08 [1] and FDA/BAM Chapter 4A [2], as well as inclusivity and exclusivity evaluations of the Invisible Sentinel Veriflow E. coli O157:H7 PCR Assay. All analyses were conducted at Q Laboratories, Inc. (Cincinnati, OH). All test materials, test kits, and propriety media unique to the Invisible Sentinel Veriflow E. coli O157:H7 PCR Assay was provided by Invisible Sentinel (Philadelphia, Pennsylvania). MATERIALS AND METHODS The methodology for this study was followed as outlined in the AOAC Research Institute Performance Tested Methods SM Program protocol: Independent Evaluation of the Invisible Sentinel Veriflow Escherichia coli O157:H7 Method. The study included an inclusivity and exclusivity evaluation, which was conducted in accordance with Appendix J: AOAC INTERNATIONAL Methods Committee Guidelines for Validation of Microbiological Methods

for Food and Environmental Surfaces [3]. In addition to the inclusivity and exclusivity evaluations, a matrix study consisting of three sets of 30 un-paired samples for raw ground beef (80% lean), 2% milk, and leafy spinach. Within each sample set, there were 5 uninoculated replicates (0 CFU/test portion), 20 low level inoculated replicates (0.2-2 CFU/test portion), and 5 high level inoculated replicates (2-5 CFU/test portion). Table A presents the sample preparation guidelines for each matrix. Table A: Matrix Outline Matrix Inoculation Organism Target Inoculation Level Replicates per Method Veriflow E. coli O157:H7 Assay Size Reference Method Reference Method Test Portion Size 0 CFU/ 5 Raw Ground Beef (80% lean) E. coli O157:H7 ATCC 1 43895 0.2-2 CFU/ 2-5 CFU/ 20 5 325 g USDA/FSIS- MLG 5.08 325 g 0 CFU/ 5 2% Milk E. coli O157:H7 ATCC 1 35150 0.2-2 CFU/ 20 25 g FDA/BAM Chapter 4A 200 ml 2-5 CFU/ 5 0 CFU/ 5 Leafy Spinach E. coli O157:H7 ATCC 1 BAA 460 0.2-2 CFU/ 20 200 g FDA/BAM Chapter 4A 200 g 2-5 CFU/ 5 1 ATCC - American Type Culture Collection A background screen of each food matrix was conducted prior to inoculation to determine the possibility of natural contamination of the target organism. For the background screen for raw ground beef, five replicate 325 g test portions were assayed for the natural presence of the target analyte using a USDA/FSIS-MLG 5.08 validated lateral flow device test system for the presumptive detection of E.coli O157:H7. For 2% milk and leafy spinach, the background screen was conducted using five replicate 200 ml or 200 g test portions following the FDA/BAM Chapter 4A. Aerobic Plate Count (APC) analyses were conducted in accordance with the FDA/BAM Chapter 3 reference [4] method to determine the microbial load present in the matrices. Each matrix was inoculated with a different strain of E. coli O157:H7. The inoculum for each matrix was prepared by transferring a pure isolated colony from trypticase soy agar with 5% sheep blood (SBA) into brain heart infusion (BHI) broth and incubating the culture at 35 ± 2 C for 24 ± 2 hours. Following incubation, serial dilutions were performed in BHI to achieve the target inoculum level of 0.2-2 CFU/ test portion for the low level and 2-5 CFU/ test portion for the high level. Twenty (20) sample replicates were evaluated at the low inoculum level and five

sample replicates were evaluated at the high level. Five uninoculated control sample replicates were evaluated at 0 CFU/ test portion. The inoculum was added to a bulk lot of test material, mixed, and held at 2-5 C for 48-72 hours to allow the organism to equilibrate within the matrix. To prepare the 325 g samples of raw ground beef, 25g of the bulk low level inoculum or the high level inoculum was mixed with 300 g of un-inoculated raw ground beef to produce the 325 g test portions. Prior to inoculation of 2% milk, the broth culture inoculum was heat stressed for 10 ± 1 minute at 50 ± 1 C in a water bath. The degree of injury of the culture was estimated by plating an aliquot of diluted culture onto Sorbitol MacConkey Agar with cefixime-tellurite (CT-SMAC) and tryptic soy agar (TSA). The agars were incubated at 35 ± 1 C for 24 ± 2 hours and the colonies were counted. The degree of injury was estimated as: nselect ( 1 ) x100 nnonselect Where n select = number of colonies on selective agar and n nonselect = number of colonies on nonselective agar. To prepare the 200 ml samples of 2% milk for the Veriflow O157:H7 Assay, 25 g samples were directly sub-sampled from the uninoculated, low level inoculated, and high level inoculated bulk samples. To prepare the 200 g samples of leafy spinach for the Veriflow O157:H7 Assay, 200 g samples were directly sub-sampled from the unioculated, low level inoculated, and high level inoculated bulk samples. The level of E. coli O157:H7 in the low level inoculum for all three matrices was determined by Most Probable Number (MPN) on the day of testing. For raw ground beef, the low level MPN was determined by evaluating 5 x 650 g, 20 x 325 g (reference method test portions), and 5 x 130 g inoculated test samples. The high level MPN was determined by analyzing 5 x 325 g (reference method test portions), 5 x 130 g, and 5 x 50 g inoculated test portions. For 2% milk, the low level MPN was determined by evaluating 5 x 400 ml, 20 x 200 ml (reference method test portions), and 5 x 100 ml inoculated test portions. For the high level, 5 x 200 ml (reference method test portions), 5 x 100 ml, and 5 x 50 ml inoculated test portions. For leafy spinach, the low level MPN was determined by evaluating 5 x 400 g, 20 x 200 g (reference method test portions), and 5 x 100 g inoculated test portions. For the high level, 5 x 200 g (reference method test portions), 5 x 100 g, and 5 x 50 g inoculated test portions. The test portion size for the MPN of each matrix is presented below in Table B. Each test portion was enriched with the reference method enrichment broth, at the reference method enrichment dilution scheme, and analyzed by the reference method procedure. The number of positive results from the 3 test levels was used to calculate the MPN using the LCF MPN calculator (version 1.6) provided by AOAC RI. [5] http://www.lcfltd.com/customer/lcfmpncalculator.exe Table B: MPN Sizes Matrix Raw Ground Beef (80% Lean) Reference Method 325 g 2% Milk 200 g Leafy Spinach *Test portions from reference method 200 g Inoculation MPN s Level Level 1 Level 2 Level 3 Low 5 x 650 g 20 x 325 g* 5 x 130 g High 5 x 325 g* 5 x 130 g 5 x 50 g Low 5 x 400 ml 20 x 200 ml* 5 x 100 ml High 5 x 200 ml* 5 x 100 ml 5 x 50 ml Low 5 x 400 g 20 x 200 g* 5 x 100 g High 5 x 200 g* 5 x 100 g 5 x 50 g

Inclusivity and Exclusivity An inclusivity and exclusivity study was conducted in accordance with Appendix J: AOAC INTERNATIONAL Methods Committee Guidelines for Validation of Microbiological Methods for Food and Environmental Surfaces. Fifty (50) target isolates of E. coli O157:H7 were analyzed in the inclusivity evaluation and thirty-five (35) closely related non-target isolates were analyzed in the exclusivity evaluation. For the inclusivity evaluation, all target strains were grown in IS mtsb at 42.0 C for 18-24 hours. Following incubation, serial dilutions of the target organisms were performed in phosphate buffered water (PBW). The cultures were diluted down from a level of 10 to 100x the LOD 50 (LOD 50 = 1 x 10 6 ) and assayed by the Veriflow E. coli O157:H7 method. For the exclusivity evaluation, the non-target strains were grown in BHI at conditions that would produce optimal growth and were analyzed undiluted by the Veriflow E. coli O157:H7 method. The inclusivity and exclusivity strains were blind coded prior to analysis of the Veriflow E. coli O157: H7 cassette. USDA/FSIS-MLG 5.08 Reference Method For the USDA/FSIS-MLG 5.08 method, thirty (30) 325 g test portions were enriched with 975 ml of mtsb (modified tryptic soy broth) that was pre-warmed to 18-35 C, homogenized by stomaching for 2 minutes and incubated at 42 ± 1 o C for 15-24 hours. After incubation, the primary enrichment from each sample was screened using a USDA/FSIS-MLG 5.08 validated lateral flow device test system. Regardless of the screening result, all samples were subjected to isolation by immunomagnetic separation (IMS) by transferring 1.0 ml aliquots of the primary enrichment to a microcentrifuge tube containing a 50 µl suspension of E. coli O157 immunomagnetic separation beads. The solution was placed onto a Labquake agitator and rotated for 10 to 15 minutes at 18-30 C. After rotation, the bead and sample solution was transferred to a MACS large cell separation column and was washed four times with E buffer (pre-warmed to 18-35 C). The final elute was collected with 1 ml of E buffer into a sterile tube. Following the IMS procedure, 0.1 ml aliquot of a 1:10 dilution and a 1:100 dilution of each IMS suspension in E Buffer, were spread plated onto modified Rainbow agar (mrba). A 450 µl aliquot of each remaining IMS suspension was transferred into a microcentrifuge tube and mixed with 25 µl of a 1 N HCl solution. The microcentrifuge tubes were vortexed and placed onto a Labquake agitator and rotated for 1 hour at 18-30 C. After rotating, 475 µl of E buffer was added to each sample tube. The acid treated IMS suspension and a 1:10 dilution of this suspension in E Buffer were plated onto mrba. All mrba plates were incubated for 20-24 hours at 35 ± 2 C. After incubation, plates were observed for typical colonies. The mrba plates containing typical colonies were tested for O157 latex agglutination and up to 5 isolated colonies were streaked to tryptic soy agar with 5% sheep blood (SBA). SBA plates were incubated for 16-24 hours at 35 ± 2 C. After incubation, SBA plates were observed for purity. Isolated colonies from SBA were confirmed positive by conducting a H7 latex agglutination test and a Shiga-toxin analysis using the Premier EHEC ELISA test assay. Final biochemical confirmations were achieved by VITEK GN Biochemical Identification following AOAC OMA 2011.17. [6] FDA/BAM Chapter 4A For 2% milk, 200 ml of sample was aseptically centrifuged at 10,000 ± 100 x g for 10 ± 0.5 minutes. After 10 minutes of centrifugation, the supernatant was decanted and the pellet was resuspended in 225 ml of modified buffered peptone water containing pyruvate (1 x mbpwp). For leafy spinach, test portions consisting of 200 g of product were enriched with 450 ml of

1 x mbpwp. All test portions were incubated statically for 5 hours at 37 ± 1 o C followed by the addition of Acriflavin, Cefsulodin, and Vancomycin supplements. Samples were re-incubated for an additional 18-24 hours at 42 ± 1 o C. Following the primary enrichment of leafy spinach, an optional immunomagnetic separation (IMS) procedure was conducted due to the high background flora. A 1 ml aliquot was transferred into a 1.5 ml microcentrifuge tube containing 20 µl of immunomagnetic separation beads. The microcentrifuge tubes were rotated on the Labquake rotator for 10 minutes, removed and placed onto a magnetic stand. The tubes were inverted three times and allowed to sit on the magnetic stand for three minutes. The contents of the microcentrifuge tube were decanted making sure that the pellet of beads was not disturbed. The microcentrifuge tubes were removed from the magnetic stand and1 ml of ISO-IMS wash buffer was added to each tube ensuring the bead pellet was fully reconstituted. The tubes were placed back onto the magnetic stand and inverted three times and allowed to sit for three minutes. This wash process was repeated three more times for a total of four washes. Following the fourth wash, the microcentrifuge tubes were removed from the magnetic stand and the bead pellet was reconstituted with 1.5 ml of ISO-IMS wash buffer. Serial dilutions (1:10) of the 2% milk enrichments or the bead solution from the leafy spinach IMS procedure were performed in phosphate buffered water (PBW). A 50 µl aliquot of the serial dilutions were plated in duplicate onto CT-SMAC and Rainbow Agar (RBA) and incubated for 18-24 hours at 37 ± 1 o C. After incubation, plates were observed for typical colonies. Plates containing typical colonies were screened for the O157 antigen by latex agglutination. Up to 10 isolated colonies that screened positive were streaked to SBA and tryptic soy agar with yeast (TSA/YE) to verify purity and incubated for 18-24 hours at 37 ± 1 o C. Following incubation, a ColiComplete (CC) disc was placed into the heaviest area of growth on the TSA/YE plates and incubated for an additional 18-24 hours at 37 ± 1 o C. After incubation, SBA plates were observed for purity. Isolated colonies from SBA were confirmed positive for the presence of H7 by latex agglutination. The CC discs on the TSA/YE plates were observed for typical reactions (blue color change with no fluorescence under long wave UV) and a spot indole test was conducted. Final biochemical confirmations were obtained by VITEK GN Biochemical Identification following AOAC OMA 2011.17. Invisible Sentinel Veriflow E. coli O157:H7 PCR Method For raw ground beef, 1625 ml of pre warmed (42 ± 2 o C) IS mtsb was added to each 325 g test portion, homogenized by stomaching for 2 minutes, and incubated at 42 ± 2 o C for 18 hours. After 18 hours of incubation, a 500 µl aliquot was transferred to individual 1.5 ml sampling tube for analysis by the Veriflow E. coli O157:H7 PCR assay. For 2% milk, 225 ml of pre warmed (42 ± 2 o C) IS 1 x mbpwp was added to each 25 g test portion, homogenized by stomaching for 2 minutes, and incubated at 42 ± 2 o C for 18 hours. After 18 hours of incubation, a 500 µl aliquot was transferred to individual 1.5 ml sampling tube for analysis by the Veriflow E. coli O157:H7 PCR assay. For leafy spinach, 200 g test portions were enriched with 450 ml of pre-warmed (42 ± 2 o C) IS 1 x mbpwp and incubated at 42 ± 2 o C for 18 hours. After 18 hours of incubation, a 500 µl aliquot was transferred to individual 1.5 ml sampling tube for analysis by the Veriflow E. coli O157:H7 PCR assay. After a 500 µl aliquot of the enrichment was transferred to a 1.5 ml sampling tube containing Buffer A, the tubes were inverted to mix samples. Sample tubes were boiled in a dry bath incubator for 10 ± 0.5 minutes at 100 C ± 1 C and then allowed to cool to room temperature (20-25 o C) for 10 ± 0.5 minutes. From the cooled sample, a 5 µl aliquot was transferred to a thawed PCR reagent tube containing 45 µl of Master Mix. All sample PCR

reagent tubes were cycled according to the parameters specified in the package insert. The cycling parameters for the amplification procedure are presented in Table C. Following the amplification procedure, 4 drops of Buffer MP were added to each sample PCR tube. The entire contents of the PCR tube (200 µl) were transferred to the sample window of a Veriflow E. coli O157:H7 assay cassette. The test was allowed to develop for 2 minutes ± 15 seconds, and then 4 additional drops of Buffer MP were added to the sample window of the Veriflow assay cassette and allowed to develop for 1 minute ± 15 seconds. The cassette lever was retracted and the test result was visually interpreted. Regardless of the presumptive results, each sample was confirmed by the appropriate reference method. Final biochemical identification using the VITEK 2 GN Biochemical Identification card, following AOAC OMA 2011.17 Table C: PCR Amplification Parameters for the Veriflow E. coli O157:H7 Method Stage 1 Description Temperature Time Number of Cycles Incubation 37 C 10 minutes 1 Stage 2 Description Temperature Time Number of Cycles Denaturing 95 C 10 minutes 1 Stage 3 Description Temperature Time Number of Cycles Denaturing 94 C 2 minutes 1 Stage 4 Description Temperature Time Number of Cycles Denaturing 94 C 15 seconds Annealing 55 C 15 seconds 40 Extension 70 C 30 seconds Stage 5 Description Temperature Time Number of Cycles Cooling 4 C 1 RESULTS Inclusivity and Exclusivity For the inclusivity evaluation, 50 of 50 isolates were correctly identified as E. coli O157:H7 by the Veriflow E. coli O157:H7 assay. For exclusivity evaluation, all 35 isolates were correctly excluded by the assay. Detailed results about the inclusivity and exclusivity testing are presented in Table 1 and 2 of the Appendix. Matrix Study For each matrix, the probability of detection (POD) was calculated as the number of positive outcomes divided by the total number of trials [7]. The POD was calculated for the candidate presumptive results, POD CP, the candidate confirmatory results, POD CC, the difference in the candidate presumptive and confirmatory results, dpod CP, presumptive candidate results that confirmed positive, POD C, the reference method, POD R, and the difference in the confirmed candidate and reference methods, dpod C. POD analyses were conducted for the Veriflow E. coli 8

O157:H7 samples and compared to either the USDA/FSIS MLG 5.08 or the FDA/BAM Chapter 4A depending on the matrix. A summary of the presumptive and confirmed results for the matrices are presented in Table D below. A detailed summary of the inclusivity and exclusivity results are presented in Tables 1 and 2 of the Appendix. Detailed results of the matrix background screens and APC results are presented in Table 3 of the Appendix. The heat stress data is presented in Table 4 of the Appendix. MPN results for each matrix are presented in Tables 5-7 of the Appendix. A detailed summary of results for each matrix is presented in Tables 8-10 and a summary of POD analyses [8] are presented in Tables 11 and 12 of the Appendix. As per criteria outlined in Appendix J of the Official Methods of Analysis Manual, fractional positive results were obtained. Table D: Summary of Results Inoculation Level Ground Beef (80% Lean) Veriflow E. coli O157:H7 PCR Assay Presumptive Confirmed USDA/FSIS-MLG 5.08 Uninoculated 0/5 0/5 0/5 Low 9/20 9/20 7/20 High 5/5 5/5 5/5 Inoculation Level 2% Milk Veriflow E. coli O157:H7 PCR Assay Presumptive Confirmed FDA/BAM Chapter 4A Uninoculated 0/5 0/5 0/5 Low 11/20 11/20 8/20 High 5/5 5/5 5/5 Inoculation Level Leafy Spinach Veriflow E. coli O157:H7 PCR Assay Presumptive Confirmed FDA/BAM Chapter 4A Uninoculated 0/5 0/5 0/5 Low 8/20 8/20 7/20 High 5/5 5/5 5/5 Raw Ground Beef For the low inoculation level of the Veriflow E. coli O157:H7 Assay, there were 9 presumptive positives and 9 confirmed positives following the USDA/FSIS-MLG 5.08 reference method confirmation procedure. There were 7 observed positives for the reference method. For the high inoculation level of the Veriflow E. coli O157:H7 Assay, there were 5 presumptive positives and 5 confirmed positives following the USDA/FSIS-MLG 5.08 reference confirmation procedure. There were 5 confirmed positives following the reference method. For the low inoculation level, a dpod C value of 0.10 was obtained with a 95% confidence interval of (-0.19, 0.37), indicating no significant difference between the candidate and reference method. A dpod CP value of 0.00 was obtained with a 95% confidence interval of (-0.28, 0.28), indicating no significant difference between the candidate presumptive and confirmed results. For the high inoculation level, a dpod C value of 0.00 was obtained with a 95% confidence interval of (-0.43, 0.43), indicating no significant difference between the candidate and reference

method. A dpod CP value of 0.00 was obtained with a 95% confidence interval of (-0.43, 0.43), indicating no significant difference between the candidate presumptive and confirmed results. Detailed results of the POD analyses are presented in Tables 11 and 12 of the Appendix. 2% Milk For the low inoculation level of the Veriflow E. coli O157:H7 Assay, there were 11 presumptive positives and 11 confirmed positives following the FDA/BAM Chapter 4A reference confirmation procedure. There were 8 observed positives for the reference method. For the high inoculation level of the Veriflow E. coli O157:H7 Assay, there were 5 presumptive positives and 5 confirmed positives following the FDA/BAM Chapter 4A reference confirmation method. There were 5 confirmed positives following the reference method. For the low inoculation level, a dpod C value of 0.15 was obtained with a 95% confidence interval of (-0.15, 0.41), indicating no significant difference between the candidate and reference method. A dpod CP value of 0.00 was obtained with a 95% confidence interval of (-0.28, 0.28), indicating no significant difference between the candidate presumptive and confirmed results. For the high inoculation level, a dpod C value of 0.00 was obtained with a 95% confidence interval of (-0.43, 0.43), indicating no significant difference between the candidate and reference method. A dpod CP value of 0.00 was obtained with a 95% confidence interval of (-0.43, 0.43), indicating no significant difference between the candidate presumptive and confirmed results. Detailed results of the POD analyses are presented in Tables 11 and 12 of the Appendix. Leafy Spinach For the low inoculation level of the Veriflow E. coli O157:H7 Assay, there were 8 presumptive positives and 8 confirmed positives following the FDA/BAM Chapter 4A reference confirmation procedure. There were 7 observed positives for the reference method. For the high inoculation level of the Veriflow E. coli O157:H7 Assay, there were 5 presumptive positives and 5 confirmed positives following the FDA/BAM Chapter 4A reference confirmation method. There were 5 confirmed positives following the reference method. For the low inoculation level, a dpod C value of 0.05 was obtained with a 95% confidence interval of (-0.23, 0.32), indicating no significant difference between the candidate and reference method. A dpod CP value of 0.00 was obtained with a 95% confidence interval of (-0.28, 0.28), indicating no significant difference between the candidate presumptive and confirmed results. For the high inoculation level, a dpod C value of 0.00 was obtained with a 95% confidence interval of (-0.43, 0.43), indicating no significant difference between the candidate and reference method. A dpod CP value of 0.00 was obtained with a 95% confidence interval of (-0.43, 0.43), indicating no significant difference between the candidate presumptive and confirmed results. Detailed results of the POD analyses are presented in Tables 11 and 12 of the Appendix. OBSERVATIONS The results of this study demonstrated the reliability of the Veriflow E. coli O157:H7 method when compared to the traditional reference methods for the detection of E. coli O157:H7 in select food matrices. The Veriflow E. coli O157:H7 method offers a significant savings in time compared to the USDA/FSIS-MLG 5.08 and the FDA/BAM Chapter 4A reference methods by producing accurate presumptive results less than 24 hours compared to 96 hours for the reference methods. The Veriflow E. coli O157:H7 cassette produced distinct positive lines allowing fast, accurate, and easy interpretation of the cassette. For 2% milk samples, a significant reduction in time to results was observed with the Veriflow E. coli O157:H7 method by eliminating the

cumbersome centrifugation process required by the FDA/BAM Chapter 4A reference method. Approximately 3 hours were saved by using the candidate method enrichment process versus the FDA/BAM Chapter 4A reference method. The method protocol is easy to follow and simplifies the process for the rapid detection of E. coli O157:H7 in food matrices. REFERENCES [1] U.S. Department of Agriculture Food Safety and Inspection Service Microbiology Laboratory Guidebook, Revision 5.08 Isolation and Identification of Escherichia coli O157:H7 from Meat Products and Carcass and Environmental Sponges (2014) http://www.fsis.usda.gov/wps/wcm/connect/51507fdb-dded-47f7-862dad80c3ee1738/mlg-5.pdf?mod=ajperes (Accessed September 2014) [2] U. S. Food and Drug Administration Bacteriological Analytical Manual, Chapter 4A Diarrheagenic Escherichia coli, (July 2014) (Accessed September 2014) http://www.fda.gov/food/foodscienceresearch/laboratorymethods/ucm070149.htm [3] Official Methods of Analysis of AOAC INTERNATIONAL (2012) 19th Ed., Appendix J: AOAC INTERNATIONAL Methods Committee Guidelines for Validation of Microbiological Methods for Food and Environmental Surfaces, AOAC INTERNATIONAL, Gaithersburg, MD, http://www.eoma.aoac.org/app_j.pdf (Accessed September 2014) [4] U. S. Food and Drug Administration Bacteriological Analytical Manual, Chapter 3 Aerobic Plate Count (January 2001) (Accessed September 2014) http://www.fda.gov/food/foodscienceresearch/laboratorymethods/ucm063346.m [5] Least Cost Formulations, Ltd., MPN Calculator Version 1.6 www.lcfltd.com/customer/lcfmpncalculator.exe (Accessed September 2014) [6] AOAC International, Evaluation of the VITEK 2 Gram-Negative(GN)Microbial Identification Test Card: Collaborative Study, Journal of AOAC International, 95, 3, (2012). [7] Wehling, P., LaBudde, R., Brunelle, S., Nelson, M. Probability of Detection (POD) as a Statistical Model for the Validation of Qualitative Methods. Journal of AOAC International, Vol. 94, No. 1, 2011. [8] Least Cost Formulations, Ltd., AOAC Binary Data Interlaboratory Study Workbook Version 2.3 (2014) -http://lcfltd.com/aoac/aoac-binary-v2-2.xls (Accessed September 2014)

Appendix

Table 1: Inclusivity Strain Results Organism Source Source # Origin Result Organism Source Source # Origin Result Escherichia coli O157:H7 MSU 1 TW00116 Human + Escherichia coli O157:H7 Q Labs 2 QL2-705 Beef Trim + Escherichia coli O157:H7 MSU 1 TW00975 Human + Escherichia coli O157:H7 MSU 1 DEC3B Human + Escherichia coli O157:H7 MSU 1 TW02302 Hamburger + Escherichia coli O157:H7 MSU 1 DEC3C Human + Escherichia coli O157:H7 MSU 1 DEC4C Buffalo + Escherichia coli O157:H7 MSU 1 DEC3D Human + Escherichia coli O157:H7 MSU 1 TW05356 Human + Escherichia coli O157:H7 Q Labs 2 QL2-370 Beef Trim + Escherichia coli O157:H7 MSU 1 TW07587 Human + Escherichia coli O157:H7 MSU 1 DEC4A Cow + Escherichia coli O157:H7 Q Labs 2 QL2-710 Beef Trim + Escherichia coli O157:H7 MSU 1 DEC4B Human + Escherichia coli O157:H7 Q Labs 2 QL2-207 Ground Beef + Escherichia coli O157:H7 ATCC 4 BAA 460 Human Feces + Escherichia coli O157:H7 NCTC 3 13125 Not Available + Escherichia coli O157:H7 MSU 1 DEC4D Cow + Escherichia coli O157:H7 NCTC 3 13126 Not Available + Escherichia coli O157:H7 MSU 1 DEC4E Human + Escherichia coli O157:H7 NCTC 3 13127 Not Available + Escherichia coli O157:H7 ATCC 35150 Human Feces + Escherichia coli O157:H7 NCTC 3 13128 Not Available + Escherichia coli O157:H7 Q Labs 2 QL2-202 Ground Beef + Escherichia coli O157:H7 MSU 1 TW04863 Human + Escherichia coli O157:H7 Q Labs 2 QL2-203 Ground Beef + Escherichia coli O157:H7 ATCC 4 43888 Human Feces + Escherichia coli O157:H7 ATCC 4 51659 Clinical Isolate + Escherichia coli O157:H7 ATCC 4 43889 Human Feces + Escherichia coli O157:H7 Q Labs 2 QL2-205 Ground Beef + Escherichia coli O157:H7 ATCC 4 43890 Human Feces + Escherichia coli O157:H7 Q Labs 2 QL2-206 Ground Beef + Escherichia coli O157:H7 ATCC 4 43894 Human Feces + Escherichia coli O157:H7 NCTC 3 12900 Not Available + Escherichia coli O157:H7 ATCC 4 43895 Raw Hamburger + Escherichia coli O157:H7 Q Labs 2 QL2-214 Beef Trim + Escherichia coli O157:H7 ATCC 4 51657 Clinical Isolate + Escherichia coli O157:H7 MSU 1 DEC3E Human + Escherichia coli O157:H7 ATCC 4 51658 Clinical Isolate + Escherichia coli O157:H7 Q Labs 2 QL2-701 Beef Trim + Escherichia coli O157:H7 Q Labs 2 QL2-204 Ground Beef + Escherichia coli O157:H7 Q Labs 2 QL2-704 Beef Trim + Escherichia coli O157:H7 ATCC 700531 Clinical Isolate + Escherichia coli O157:H7 MSU 1 DEC3A Human + Escherichia coli O157:H7 ATCC 700599 Salami + Escherichia coli O157:H7 Q Labs 2 QL2-706 Beef Trim + Escherichia coli O157:H7 ATCC 700728 Not Available + Escherichia coli O157:H7 Q Labs 2 QL2-707 Beef Trim + Escherichia coli O157:H7 ATCC 700927 Not Available + Escherichia coli O157:H7 Q Labs 2 QL2-708 Beef Trim + Strains obtained from: 1 Michigan State University Culture Collection; 2 Q Laboratories, Inc. Culture Collection; 3 National Culture Type Collection; 4 American Type Culture Collection

Table 2: Exclusivity Strain Results Organism Source Source # Origin Result Organism Source Source # Origin Result Escherichia coli O26:H11 MSU 1 DEC10E Cow - Escherichia coli O121 MSU 1 TW07614 Human - Escherichia coli O26 MSU 1 DEC9F Human - Escherichia coli O121 MSU 1 TW08023 Human - Escherichia coli O26 MSU 1 TW04284 Human, Child - Escherichia coli O121 NCTC 2 9121 Not Available - Escherichia coli O45 MSU 1 TW07947 Human - Escherichia coli O142 NCTC 2 10089 Not Available - Escherichia coli O45 MSU 1 TW14003 Human - Escherichia coli O145 NCTC 2 10279 Not Available - Escherichia coli O45 MSU 1 TW10121 Human - Escherichia coli O145 MSU 1 TW07596 Human - Escherichia coli O55:H6 MSU 1 DEC1A Human - Escherichia coli O145 MSU 1 TW01664 Human - Escherichia coli O91 NCTC 2 9091 Not Available - Escherichia coli O146 NCTC 2 10677 Not Available - Escherichia coli O103 MSU 1 TW07971 Human - Escherichia coli 0163 NCTC 2 11021 Feces - Escherichia coli O103 MSU 1 TW07697 Human - Alcaligenes faecalis ATCC 3 8750 Not Available - Escherichia coli O103 MSU 1 TW11239 Human, Child - Citrobacter freundii ATCC 3 8090 Not Available - Escherichia coli O111 MSU 1 DEC8D Human, Infant - Enterobacter aerogenes ATCC 3 13048 Sputum - Escherichia coli O111:H12 MSU 1 DEC6A Human, Infant - Escherichia blattae ATCC 3 29907 Cockroach - Escherichia coli O111:H8 MSU 1 DEC6C Human - Escherichia fergusonii ATCC 3 35469 Human Feces - Escherichia coli O113 NCTC 2 9113 Not Available - Escherichia hermannii ATCC 3 33650 Human Toe - Escherichia coli O115 NCTC 2 10444 Calf - Escherichia vulneris ATCC 3 29943 Human Wound - Escherichia coli O117 NCTC 2 9117 Not Available - Salmonella enterica subsp.enterica serovar ATCC 3 10708 Not Available - Choleraesuis Escherichia coli O118 NCTC 2 9118 Not Available - Strains obtained from: 1 Michigan State University Culture Collection; 2 National Culture Type Collection; 3 American Type Culture Collection

Table 3: Matrix Background Results APC 1 E. coli O157:H7 Pathogen Screen 2 Matrix (CFU/g) Replicate A Replicate B Replicate C Replicate D Replicate E Raw Ground 3.6 x 10 4 None Detected/ None Detected/ None Detected/ None Detected/ None Detected/ Beef 325 g 325 g 325 g 325 g 325 g 2% Milk 3.0 x 10 1 None Detected/ None Detected/ None Detected/ None Detected/ None Detected/ 200 ml 200 ml 200 ml 200 ml 200 ml Leafy Spinach 1.7 x 10 5 None Detected/ None Detected/ None Detected/ None Detected/ None Detected/ 200 g 200 g 200 g 200 g 200 g 1 APC conducted in accordance with FDA BAM Chapter 3 2 E. coli O157:H7 screen conducted in accordance with an AOAC RI approved LFD (MLG 5.08) or via the FDA/BAM Chapter 4A. Table 4: Inoculum Heat Stress Results for 2% Milk Inoculating Organism Agar CFU Percent Injury E. coli O157:H7 ATCC 35150 TSA: trypticase soy agar CT-SMAC: Sorbitol MacConkey Agar with Cefixime-Tellurite TSA 3.0 x 10 9 74.0 % CT-SMAC 7.8 x 10 8

Table 5: MPN * Summary Table Inoculum Level Raw Ground Beef (80% Lean) E. coli O157:H7 ATCC 43895 Low (0.2-2 MPN/) A B C D E 650 g + - + + - 5 x 325 g (Reference Samples) 7/20 130 g + - - + - MPN/Test portion 0.49 Low Conf. Limit MPN/ 0.25 High Conf. Limit MPN/ 0.85 High (2-5 MPN/) A B C D E 5 x 325 g (Reference Samples) 5/5 130 g + + + + + 50 g - - + + + MPN/Test portion 3.05 Low Conf. Limit/ 1.30 High Conf. Limit/ 7.19 * MPN was calculated for the low inoculum level by analyzing five 650 g, five 130 g, and the 20 low level reference method samples (325 g) using the LCF MPN Calculator version 1.6 http://www.lcfltd.com/customer/lcfmpncalculator.exe. For the high inoculum level, the MPN was calculated by analyzing five 130 g, five 50 g, and the 5 high level reference method samples (325 g) using the LCF MPN Calculator version 1.6 http://www.lcfltd.com/customer/lcfmpncalculator.exe

Table 6: MPN * Summary Table Inoculum Level 2% Milk E. coli O157:H7 ATCC 35150 Low (0.2-2 MPN/) A B C D E 400 g + - + + - 5 x 200 g (Reference Samples) 8/20 100 g - - + - + MPN/Test portion 0.54 Low Conf. Limit MPN/ 0.29 High Conf. Limit MPN/ 0.92 High (2-5 MPN/) A B C D E 5 x 200 g (Reference Samples) 5/5 100 g + + + + + 50 g + + + - + MPN/Test portion 3.72 Low Conf. Limit/ 1.54 High Conf. Limit/ 8.97 * MPN was calculated for the low inoculum level by analyzing five 400 g, five 100 g, and the 20 low level reference method samples (200 g) using the LCF MPN Calculator version 1.6 http://www.lcfltd.com/customer/lcfmpncalculator.exe. For the high inoculum level, the MPN was calculated by analyzing five 100 g, five 50 g, and the 5 high level reference method samples (200 g) using the LCF MPN Calculator version 1.6 http://www.lcfltd.com/customer/lcfmpncalculator.exe

Table 7: MPN * Summary Table Inoculum Level Leafy Spinach E. coli O157:H7 ATCC BAA 460 Low (0.2-2 MPN/) A B C D E 400 g + - - + + 5 x 200 g (Reference Samples) 7/20 100 g - - - - + MPN/Test portion 0.44 Low Conf. Limit MPN/ 0.21 High Conf. Limit MPN/ 0.76 High (2-5 MPN/) A B C D E 5 x 200 g (Reference Samples) 5/5 100 g - + + + + 50 g + + - + - MPN/Test portion 1.84 Low Conf. Limit/ 0.91 High Conf. Limit/ 3.75 * MPN was calculated for the low inoculum level by analyzing five 400 g, five 100 g, and the 20 low level reference method samples (200 g) using the LCF MPN Calculator version 1.6 http://www.lcfltd.com/customer/lcfmpncalculator.exe. For the high inoculum level, the MPN was calculated by analyzing five 100 g, five 50 g, and the 5 high level reference method samples (200 g) using the LCF MPN Calculator version 1.6 http://www.lcfltd.com/customer/lcfmpncalculator.exe

Table 8: Results of the Veriflow E. coli O157:H7 Assay for Raw Ground Beef (80% Lean, 325 g) Raw Ground Beef (80% Lean) E. coli O157:H7 ATCC 43895 Low Level 0.49 (0.25, 0.85) MPN/ Sample # Veriflow E.coli O157:H7 Assay Presumptive Confirmed USDA/FSIS-MLG 5.08 1 - - - 2 + + - 3 - - + 4 - - + 5 + + - 6 - - - 7 - - + 8 + + - 9 + + - 10 - - - 11 + + + 12 - - - 13 + + - 14 + + - 15 - - - 16 - - - 17 + + + 18 + + - 19 - - + 20 - - + Total 9/20 9/20 7/20 High Level 3.05 (1.30, 7.19) MPN/ 1 + + + 2 + + + 3 + + + 4 + + + 5 + + + Total 5/5 5/5 5/5 Uninoculated 1 - - - 2 - - - 3 - - - 4 - - - 5 - - - Total 0/5 0/5 0/5

Table 9: Results of the Veriflow E. coli O157:H7 Assay for 2% Milk (25 g) 2% Milk E. coli O157:H7 ATCC 35150 Low Level 0.54 (0.29, 0.92) MPN/ Sample # Veriflow E.coli O157:H7 Assay Presumptive Confirmed FDA/BAM Chapter 4A 1 - - - 2 + + + 3 - - - 4 + + + 5 - - - 6 + + + 7 + + + 8 - - - 9 + + - 10 - - - 11 - - + 12 + + - 13 + + - 14 - - - 15 + + + 16 - - + 17 + + - 18 + + - 19 + + + 20 - - - Total 11/20 11/20 8/20 High Level 3.72 (1.54, 8.97) MPN/ 1 + + + 2 + + + 3 + + + 4 + + + 5 + + + Total 5/5 5/5 5/5 Uninoculated 1 - - - 2 - - - 3 - - - 4 - - - 5 - - - Total 0/5 0/5 0/5

Table 10: Results of the Veriflow E. coli O157:H7 Assay for Leafy Spinach (200 g) Leafy Spinach E. coli O157:H7 ATCC BAA 460 Low Level 0.44 (0.21, 0.76) MPN/ Sample # Veriflow E.coli O157:H7 Assay Presumptive Confirmed FDA/BAM Chapter 4A 1 - - + 2 + + - 3 + + - 4 - - + 5 - - - 6 + + - 7 - - - 8 - - - 9 + + + 10 - - - 11 + + - 12 - - + 13 - - + 14 + + - 15 - - + 16 - - - 17 - - - 18 + + - 19 + + + 20 - - - Total 8/20 8/20 7/20 High Level 1.84 (0.91, 3.75) MPN/ 1 + + + 2 + + + 3 + + + 4 + + + 5 + + + Total 5/5 5/5 5/5 Uninoculated 1 - - - 2 - - - 3 - - - 4 - - - 5 - - - Total 0/5 0/5 0/5

Table 11: Veriflow E. coli O157:H7 Assay, Candidate vs. Reference POD Results Matrix Raw Ground Beef (80% Lean) Analysis Time Point 18 Hours Strain E. coli O157:H7 ATCC 43895 MPN a / Candidate Reference N b x c d e POD C 95% CI X POD R 95% CI dpod C f 95% CI g - 5 0 0.00 0.00, 0.43 0 0.00 0.00, 0.43 0.00-0.43, 0.43 0.49 (0.25, 0.85) 20 9 0.45 0.26, 0.66 7 0.35 0.18, 0.57 0.10-0.19, 0.37 3.05 (1.30, 7.19) 5 5 1.00 0.57, 1.00 5 1.00 0.57, 1.00 0.00-0.43, 0.43 2% Milk 18 Hours E. coli O157:H7 ATCC 35150-5 0 0.00 0.00, 0.43 0 0.00 0.00, 0.43 0.00-0.43, 0.43 0.54 (0.29, 0.92) 20 11 0.55 0.34, 0.74 8 0.40 0.22, 0.61 0.15-0.15, 0.41 3.72 (1.54, 8.97) 5 5 1.00 0.57, 1.00 5 1.00 0.57, 1.00 0.00-0.43, 0.43-5 0 0.00 0.00, 0.43 0 0.00 0.00, 0.43 0.00-0.43, 0.43 E. coli O157:H7 Leafy Spinach 18 Hours ATCC BAA 460 0.44 (0.21, 0.76) 20 8 0.40 0.22, 0.61 7 0.35 0.18, 0.57 0.05-0.23, 0.32 1.84 (0.91, 3.75) 5 5 1.00 0.57, 1.00 5 1.00 0.57, 1.00 0.00-0.43, 0.43 a MPN = Most Probable Number is based on the POD of reference method test portions using the LCF MPN calculator version 1.6, with 95% confidence interval b N = Number of test potions c x = Number of positive test portions d POD C = Candidate method confirmed positive outcomes divided by the total number of trials e POD R = Reference method confirmed positive outcomes divided by the total number of trials f dpod C = Difference between the confirmed candidate method result and reference method confirmed result POD values g 95% CI = If the confidence interval of a dpod does not contain zero, then the difference is statistically significant at the 5% level

Table 12: Veriflow E. coli O157:H7 Assay, Presumptive vs. Confirmed POD Results Matrix Raw Ground Beef (80% Lean) Analysis Time Point 18 Hours 2% Milk 18 Hours Strain E. coli O157:H7 ATCC 43895 E. coli O157:H7 ATCC 35150 MPN a / Presumptive Confirmed N b x c d e POD CP 95% CI X POD CC 95% CI dpod CP f 95% CI g - 5 0 0.00 0.00, 0.43 0 0.00 0.00, 0.43 0.00-0.43, 0.43 0.49 (0.25, 0.85) 20 9 0.45 0.26, 0.66 9 0.45 0.26, 0.66 0.00-0.28, 0.28 3.05 (1.30, 7.19) 5 5 1.00 0.57, 1.00 5 1.00 0.57, 1.00 0.00-0.43, 0.43-5 0 0.00 0.00, 0.43 0 0.00 0.00, 0.43 0.00-0.43, 0.43 0.54 (0.29, 0.92) 20 11 0.55 0.34, 0.74 11 0.55 0.34, 0.74 0.00-0.28, 0.28 3.72 (1.54, 8.97) 5 5 1.00 0.57, 1.00 5 1.00 0.57, 1.00 0.00-0.43, 0.43-5 0 0.00 0.00, 0.43 0 0.00 0.00, 0.43 0.00-0.43, 0.43 E. coli O157:H7 Leafy Spinach 18 Hours 0.44 (0.21, 0.76) 20 8 0.40 0.22, 0.61 8 0.40 0.22, 0.61 0.00-0.28, 0.28 ATCC BAA 460 1.84 (0.91, 3.75) 5 5 1.00 0.57, 1.00 5 1.00 0.57, 1.00 0.00-0.43, 0.43 a MPN = Most Probable Number is based on the POD of reference method test portions using the LCF MPN calculator version 1.6, with 95% confidence interval b N = Number of test potions c x = Number of positive test portions d POD CP = Candidate method presumptive positive outcomes divided by the total number of trials e POD CC = Candidate method confirmed positive outcomes divided by the total number of trials f dpod CP = Difference between the candidate method presumptive result and candidate method confirmed result POD values g 95% CI = If the confidence interval of a dpod does not contain zero, then the difference is statistically significant at the 5% level