General Enquiries on the form should be made to:Defra, Strategic Evidence and AnalysisE-mail: StrategicEvidence@defra.gsi.gov.uk

Evidence Project Final Report

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The Evidence Project Final Report is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra websiteAn Evidence Project Final Report must be completed for all projects.

This form is in Word format and the boxes may be expanded, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code FA0146

2. Project titleMethod validation of the real-time PCR approach for the quantitation of horse DNA

3. Contractororganisation(s)

LGC Ltd.,

Malcolm BurnsGavin Nixon

54. Total Defra project costs £ 45,053.00(agreed fixed price)

5. Project: start date 26/02/14

EVID4 Evidence Project Final Report (Rev. 10/14) Page 1 of 18

end date 26/12/14

EVID4 Evidence Project Final Report (Rev. 10/14) Page 2 of 18

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...................................................................................YES NO (a) When preparing Evidence Project Final Reports contractors should bear in mind that Defra intends that

they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the Evidence Project Final Report can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain     

EVID4 Evidence Project Final Report (Rev. 10/14) Page 3 of 18

Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent

non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

Issues surrounding food authenticity and food fraud continue to be of significant interest within the European food industry, partly due to the pressures faced by producers within today’s challenging financial climate and the international nature of modern food production. The recent EU wide issue involving the detection of the undeclared presence of horse-meat in beef products destined for human consumption has emphasised the need for the development of accurate analytical approaches for the quantitative detection of meat species in a sample.

In response to the horse-meat issue, a real time PCR method was developed at LGC for the quantitation of horse DNA (Defra project FA0135). It was demonstrated that that this assay was capable of detecting and accurately determining the amount of horse DNA present in raw beef samples spiked with raw horse meat, and the method was validated through objective assessment of performance characteristics using DNA:DNA ad-mixtures and w/w gravimetric raw horse-meat in raw beef (meat) materials.

In order to enable adoption of the method for use by Public Analysts, further method validation activities were conducted under the remit of the current project (Defra project FA0146), as advised by Defra’s Authenticity Methods Working Group Technical Sub-Group (AMWG-TSG). The work focussed on three key areas: (i) Applicability of the method using different horse samples; (ii) Evaluation of the precision around the 1% threshold level for enforcement activities, and (iii) Assessment of the method’s suitability for use in complex foods.

Horse meat was sourced from two independent horse specimens. Pairwise comparisons of calibrants and test samples derived from the different horse specimens provided evidence that there was no significant difference when using calibration curves based on different horse meat sources.

Based on multiple replicate samples and replicate experiments, the analytical measurement uncertainty at the 1% adulteration level was estimated as 1.58 +/- 0.54% w/w. This compared well to the estimated measurement uncertainty of other real-time PCR approaches at this low level of target analyte.

The potential for significant bias associated with quantitative DNA approaches when applied to differentially and highly degraded DNA (e.g. in processed food products) is a well-known and ubiquitous issue associated with most DNA techniques. In order to give an objective comparison, the performance of the real-time PCR approach for the quantitation of horse DNA developed as part of this project was compared to a proprietary commercially available kit. Both approaches were applied to a range of processed food samples spiked with horse meat. Both approaches gave statistically similar results, demonstrating some small positive bias when applied to beef burgers and cottage pie test samples, through to very large positive bias when tested with corned beef samples, which are known to be highly processed.

Follow-on work is required to help establish inter-laboratory performance characteristics based on a Collaborative Trial of the Standard Operating Procedure of the method developed as part of this project, and to further confirm the suitability of the methodology for quantitation of horse DNA in food samples within the official control and commercial testing networks. A Knowledge Transfer event is required in order to disseminate the results of the real-time PCR approach for quantitation of horse DNA (FA0135) and the subsequent additional method validation aspects (FA0146) to stakeholders representing enforcement and industrial laboratories.

The development and publication of this non-proprietary novel real-time PCR method for the quantitation of horse DNA will inform and strengthen the decision making processes of food companies. Following any successful ring-trial to help further establish inter-laboratory performance characteristics of the method, food companies can apply the methodology to quality management systems within their supply chains and production processes to ensure that potential adulteration is quickly identified and corrective actions are implemented before costly and publically damaging interventions are required such as product recalls. Access to a fully validated, quantitative method for the determination of horse DNA in raw meat samples will enable honest traders to robustly defend their food supply chain.

EVID4 Evidence Project Final Report (Rev. 10/14) Page 4 of 18

GlossaryA260 - Absorbance at 260 nm

A260:230 - Ratio of absorbance at 260 and 230 nm

A260:280 - Ratio of absorbance at 260 and 280 nm

Cq - Cycle of quantification

CV - Coefficient of variation

DEFRA - Department for Environment, Food and Rural Affairs

DNA - Deoxyribonucleic acid

EU - European Union

FSA - Food Standards Agency

gDNA - Genomic DNA

LoD - Limit of Detection

nM – Nano Molar

MB – Mega Base

ml – Millilitre

L - Microlitre

mM - Milli Molar

ng – Nanograms

PA – Public Analyst

PCR - Polymerase Chain Reaction

qPCR - Quantitative real time Polymerase Chain Reaction

R2 - R squared

RQ - Relative Quantification

SDS - Sodium Dodecyl Sulphate

SOP - Standard Operating Procedure

Std. Dev. - Standard deviation

UK - United Kingdom

w/w - Weight for Weight

EVID4 Evidence Project Final Report (Rev. 10/14) Page 5 of 18

Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of

the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Exchange).Scientific objective as set out in the contract

Objective No. Objective Description

Method validation of the real-time PCR approach for the quantitation of horse DNA

00 Defra project inception meetingProject kick-off meeting with Defra and relevant scientific experts to agree on details of the project plan and delivery.

01 Source material and prepare samplesSource appropriate species materials and order in laboratory consumables. Extract DNA and prepare samples and standards accordingly.

02 Method validation using different horse samplesEvaluate performance (trueness and precision) associated with the qPCR assay for quantitation of horse DNA, when calibrants and test samples are derived from different individual horse specimens (i.e. different animals).

03 Evaluate fitness for purpose for enforcement activities at the 1% thresholdEstimate the measurement uncertainty levels associated with 1% horse in beef ad-mixtures and other samples in view of newly published EU guidance. Provide interim update to Defra.

04 Assess the suitability of the method for use in complex foodsInvestigate the impact of processed food matrices on assay performance by spiking horse meat into complex/processed foods and separated meat content. Estimate performance characteristics of trueness, precision and LOD and compare these to the original raw meat in raw meat results and other horse DNA detection kits.

05 Finalise draft report and SOP and submit to DefraProvision of additional guidance notes (if applicable)Draft report and SOP finalised and submitted to Defra. If any of the above objectives significantly impact upon the published SOP and interpretation of results, the SOP will be amended and additional guidance notes provided as necessary. The report and SOP will also take account of any advice provided by the Authenticity Methods Working Group.

EVID4 Evidence Project Final Report (Rev. 10/14) Page 6 of 18

2.0 The extent to which the Objectives have been metObjective 00 – Defra project inception meetingObjective 00 was successfully completed. A teleconference was held between key members of staff from Defra and LGC to discuss and agree on the sourcing of materials to represent different horse specimens (Tasks 1 and 2); the experimental design to evaluate the uncertainty around the 1% w/w adulteration level (Task 3); and the range and scope of the processed food samples and the levels of spiked in horse meat (Task 4).

Objective 01 – Source material and prepare samplesObjective 01 was successfully completed. Raw muscle tissue (horse and beef meat) was sourced from a known supplier which had been trimmed free of surface inter-muscular fat and connective tissue and authenticated as to species type by qPCR. Mass-based ratio preparations of 100%, 5%, 1%, 0.5% and 0.1% weight for weight (w/w) of raw horse muscle tissue in a raw beef muscle tissue background were produced.

Processed foods, comprising beef lasagne, beef cottage pie, beef burgers and corned beef, were sourced from a reputable supermarket, the meat content separated and used to prepare mass-based ratio ad-mixtures of 5% and 1% w/w of raw/cooked (process state matched) horse muscle tissue in processed meat background.

Objective 02 – Method validation using different horse samplesObjective 02 was successfully completed. Further optimisation was conducted prior to the method validation work to ensure optimal assay performance. The Köppel horse assay probe concentration was increased from 80 nM to 200 nM in order to enhance assay fluorescence signal levels and ensure equivalency between qPCR assays. The adapted Köppel assay was found to perform at least as well as the published assay conditions.

Method validation work investigating the potential impact of different horse specimens upon the quantitative performance of the Köppel/Laube methodology was undertaken. Meat tissue from two individual horse specimens was sourced. Calibration curves were produced based on the two different horse specimens, as were test samples of 1%, 0.5% and 0.1% w/w horse in beef. All pairwise comparisons of calibrant with test sample were conducted. The results indicated that the source of horse calibration material used in the present study did not impact upon the quantitative performance of the Köppel/Laube methodology.

Objective 03 – Evaluate fitness for purpose for enforcement activities at the 1% thresholdObjective 03 was successfully completed. Performance evaluation work utilising a panel of 1 % w/w horse meat in a background of beef meat ad-mixtures generated test data with observed horse adulteration levels below 2% (1.58 % mean value). An estimate of the analytical measurement uncertainty (based on a 95 % confidence interval) was determined for the panel of 1% horse adulteration test samples to be 1.58 +/- 0.54 % w/w raw horse meat in a raw beef background, based on four sample batches evaluated across triplicate experimental plates, with a triplicate PCR replication level within each plate. This indicated that the methodology was fit for purpose at the recommended 1% threshold level used to guide enforcement activities.

Objective 04 – Assess the suitability of the method for use in complex foodsObjective 04 was successfully completed. A study was conducted to assess the applicability of the Köppel/Laube method to testing for horse meat contamination in processed food products, and analytical performance was compared with a commercially available test kit (SureFood® Animal Quant Equus kit). As expected, highly processed food products (e.g. corned beef) were particularly challenging for both methodologies due to low levels of amplifiable template and the presence of potential matrix interferents, but comparable quantitative capabilities across all the spiking levels (1% & 5%) were demonstrated for both approaches. As documented in the published literature, both DNA approaches tended to over-estimate the relative horse meat content in highly processed foods, due to the differentially degraded nature of the DNA. This observation is a common issue with nearly all quantitative DNA approaches when applied to highly processed food samples. Statistical analyses showed that there was no significant difference between quantitative data generated by either methodology. This provided evidence that the quantitative performance of the Köppel/Laube approach was at least as good as a commercially available test kit in terms of fitness for purpose for the evaluation of processed food materials.

Objective 05 – Finalise draft report and SOP and submit to DefraObjective 05 was successfully completed. The Standard Operating Procedure (SOP) and final report were

EVID4 Evidence Project Final Report (Rev. 10/14) Page 7 of 18

submitted to Defra and the AMWG group.

3.0 Materials and MethodsObjective 01 – Source material and prepare samplesSourcing and preparing materialsa) Horse (derived from two separate individuals) and beef muscle tissue samples were sourced from

Kezie Ltd (Duns, UK). The bulk horse materials (Horse A & B) were sourced 6 months apart and labelled by the producer with different ‘use by dates’. The meat samples were surface trimmed and then prepared by removing any separable fat, gristle, etc. retaining the lean meat. The lean meat was cubed, thoroughly homogenised in a clean meat species specific food processor, combined, mixed and the resultant paste stored at -20 oC or -80 oC.

b) Multiple packs of processed food products were sourced from a UK supermarket chain that comprised a selection of the supermarket’s own brand items chosen to represent examples of uncooked, chilled ready meal and canned products:

i. Aberdeen Angus Cottage Pie (~32% beef) – chilled ready meal

ii. Lasagne (~28% beef) – chilled ready meal

iii. Beef Steak Burgers (~94% beef) – uncooked

iv. Corned Beef (~96% beef) - canned

The meat content was separated from non-meat constituents, briefly washed with molecular biology grade water and allowed to drain. The resultant material was thoroughly homogenised in a clean meat species specific food processor, combined, mixed and the resultant paste stored at -20 oC or -80 oC.

Cooked horse meat was prepared for use in cooked food product ad-mixtures. Heat/pressure treatment was chosen to simulate the impact of industrial food processes. Homogenised raw horse muscle tissue was autoclaved at 121 oC for 15 minutes and the cooked material homogenised in a clean meat species specific food processor, combined, mixed and the resultant paste stored at -20 oC or -80 oC.

Raw meat authenticity testingEquus genus and mammalian class identify was authenticated in-house by real-time PCR.

DNA extractionDNA extraction was performed on 1 g tissue samples using the silica-based Kleargene DNA extraction method (LGC Genomics, UK). The extraction process comprised sample homogenisation followed by incubation, lysis of cellular components in an SDS buffer with proteinase K and binding of the isolated DNA to positively charged silica beads. Multiple washing stages were used to clean the DNA which was eluted in 1 ml of elution buffer. DNA yield (A260) and quality characteristics (A260:230 and A260:280) were determined using a Nanodrop ND-1000 Spectrophotometer (Thermo Scientific, UK).

Preparation of unprocessed weight for weight (w/w) ad-mixturesWeight for weight (w/w) ad-mixtures were prepared using homogenised raw horse and beef meat derived from authenticated muscle tissue. The w/w materials were prepared gravimetrically by weighing the required amounts of homogenised horse and beef meats to generate 100%, 30%, 5%, 1% and 0.1% raw horse muscle tissue in a raw beef muscle tissue background. DNA was extracted from these samples using the Kleargene DNA extraction methodology described previously and characterised using a Nanodrop ND-1000 Spectrophotometer (Thermo Scientific, UK).

Preparation of processed weight for weight (w/w) ad-mixturesWeight for weight (w/w) ad-mixtures were prepared using homogenised raw or cooked horse (matched to food process state) and recovered beef meat derived from processed food products. The w/w materials were prepared gravimetrically by weighing the required amounts of homogenised horse and beef meats to generate 5% and 1% raw or cooked horse muscle tissue (matched to cooked state of background material) in a processed beef meat background. DNA was extracted from these samples using the Kleargene DNA extraction methodology described previously and characterised using a using a NanoDrop ND-1000 Spectrophotometer (Thermo Scientific, UK).

Objective 02 – Method validation using different horse samplesBackgroundA relative quantitative-based approach was developed under Defra project# FA01351 to determine horse

EVID4 Evidence Project Final Report (Rev. 10/14) Page 8 of 18

DNA content through the comparative analyses of horse/beef samples using a horse specific qPCR assay (horse DNA only) and universal mammalian assay (total DNA). The approach utilised published hydrolysis probe-based qPCR assays developed by Köppel et al. (2011)2 (horse/donkey/mule growth hormone receptor specific) and Laube et al. (2003)3 (mammalian, poultry and fish myostatin gene).

Following consultation with Defra’s independent Authenticity Methods Working Group – Technical Sub-Group (AWMG-TSG), it was decided to further assess the quantitative capability of the real-time PCR approach by challenging it with samples derived from different horse specimens to demonstrate the scope of the method. Further method validation work was conducted in order to confirm that the source of horse calibrant material did not impact upon assay performance.

Further optimisationAdditional assay optimisation work, that involved increasing the Köppel horse assay probe concentration from 80 nM to 200 nM in order to enhance assay fluorescence signal levels, was undertaken and the optimised assay found to perform at least as well as the published assay conditions, i.e. comparable PCR efficiencies, precision and bias levels. All work conducted within this project was performed under the adapted assay conditions. The SOP was updated accordingly to reflect this.

ConditionsThe Köppel and Laube qPCR assays were performed under standard 2-step real-time PCR cycling conditions (10 min/ 95 oC; 15 s/ 95 oC, 1 min/ 60 oC, 45 cycles). Calibration curves were produced based on serial dilutions of 100% horse DNA (Horse A or B). Triplicate qPCR plates per calibration source (Horse A or B) were undertaken comprising a 20,480 to ~1.25 horse genome equivalents (assuming a haploid genome size of 2474.93 MB) using a seven point standard curve. Test samples consisting of 100%, 1%, 0.5% and 0.1% w/w horse in beef ad-mixtures were generated from Horse A and B as well as appropriate controls per target assay. All samples and controls were represented by a triplicate PCR technical replicate level with the exception of the 100% beef and no template controls at duplicate replication level. Pairwise comparison of horse calibrant and test samples were made in order to evaluate the impact of using calibration curves derived from individual horse specimens.

AnalysesAssay performance-based metrics including PCR efficiency and precision (Table 1) were derived from the dataset and horse gDNA content determined by standard curve analyses which was normalized to total mammalian gDNA content to generate a percentage ratio (% horse adulteration). Estimated % horse adulteration data sets generated using calibrants originating from horse A or B were tested using Analysis of Variance to identify any significant differences between the data sets. Confirmatory data analyses were performed by LGC’s Statistical Team using a log transformed version of the same data set in order to test for the impact of the potential non-normality of the data sets used.

Table 1. Measurements employed to characterise methodology and test materials

Performance Measurements Methodology

PCR Efficiency Assay specific mean PCR efficiency from triplicate plates:% E = 100 x (-1 + 10-(1/calibration curve slope))

Precision Assay specific mean percentage coefficient of variation from triplicate plates:% CV = 100 x (standard deviation/mean)

Bias Assay specific mean bias from triplicate plates:Bias = estimated value/assigned value

Estimated % horse adulteration

Analysis of Variance (ANOVA)

Testing for significant differences between data sets using Excel 2010 (Microsoft UK) statistical functionality

Objective 03 – Evaluate fitness for purpose for enforcement activities at the 1% thresholdBackground

EVID4 Evidence Project Final Report (Rev. 10/14) Page 9 of 18

Standard curve derived horse genome equivalent copy number

Standard curve derived total mammalian genome equivalent copy number

x 100

The EU had recently published guidance 4,5 on how to evaluate test samples for determination of horse-meat content above or below a 1% threshold level used to guide enforcement activities.

. It was recommended that the real-time PCR method for quantitation of horse DNA be applied in this context to determine the approach’s capability to accurately quantify samples above and below this threshold.

ConditionsThe Köppel and Laube qPCR assays were performed under standard 2-step real-time PCR cycling conditions (10 min/ 95 oC; 15 s/ 95 oC, 1 min/ 60 oC, 45 cycles). Calibration curves were produced based on serial dilutions of 100% horse DNA (single horse specimen). Triplicate qPCR plates were undertaken comprising a 20,480 to ~1.25 horse genome equivalents (assuming a haploid genome size of 2474.93 MB) seven point standard curve. Test samples consisting of 100%, and 1% (A – D) w/w horse in beef ad-mixtures were evaluated as well as appropriate controls per target assay. All samples and controls were represented by a triplicate PCR technical replicate level.

AnalysesHorse gDNA content was determined by standard curve analyses and normalized to total mammalian gDNA content to generate a percentage ratio (% horse adulteration). Estimated % horse adulteration data sets were analysed to determine the analytical measurement uncertainty (based on 95% confidence interval). Confirmatory data analyses were performed by LGC’s Statistical Team using a log transformed version of the same data set in order to test for the impact of the potential non-normality of the data sets used.

Objective 04 – Assess the suitability of the method for use in complex foodsBackgroundPrevious work conducted under Defra project FA0135 had established the basic performance of the methodology under using pure meat ad-mixtures. Real-world application requires that the approach is capable of detecting and ideally quantifying horse DNA within a range of complex and processed food materials. This necessitated an evaluation of complex foods using the test method and a comparison with a commercial test kit in order to validate the results.

ConditionsThe Köppel and Laube qPCR assays were performed under standard 2-step real-time PCR cycling conditions (10 min/ 95 oC; 15 s/ 95 oC, 1 min/ 60 oC, 45 cycles). Calibration curves were produced based on serial dilutions of 100% horse DNA (single horse specimen). Single replicate qPCR plates were performed comprising a 20,480 to ~1.25 horse genome equivalents (assuming a haploid genome size of 2474.93 MB) seven point standard curve. Test samples consisting of 100%, 5% and 1%, w/w raw or cooked horse muscle tissue in a processed beef meat background (cottage pie, lasagne, beef burger and corned beef) were evaluated as well as appropriate controls per target assay. All samples and controls were represented by a triplicate PCR technical replicate level with the exception of the extraction and no template controls at duplicate replication level.

The test panel was co-analysed using the dual-assay SureFood® ANIMAL QUANT Equus, Art. No. S1016 (Congen Biotechnologie GmbH, Germany) as per the manufacturer’s recommended conditions. The test samples were analysed at a triplicate PCR technical replicate level.

AnalysesHorse gDNA content was determined by standard curve analyses and normalized to total mammalian gDNA content to generate a percentage ratio (% horse adulteration). Estimated % horse adulteration data sets were tested using Analysis of Variance (single factor) to identify any significant differences between the Köppel/Laube and SureFood® ANIMAL QUANT Equus approaches. Confirmatory data analyses were performed by LGC’s Statistical Team using a log transformed version of the same data set in order to test for the impact of the potential non-normality of the data sets used.

4.0 ResultsObjective 01 – Source material and prepare samplesMaterial was sourced and prepared as described in the Materials and Methods section.

Objective 02 – Method validation using different horse samplesFurther optimisation was conducted prior to the method validation work to ensure optimal assay performance. The Köppel horse assay probe concentration was increased from 80 nM to 200 nM in order to enhance assay fluorescence signal levels and ensure equivalency between qPCR assays. The adapted Köppel assay was found to perform at least as well as the published assay conditions as demonstrated by

EVID4 Evidence Project Final Report (Rev. 10/14) Page 10 of 18

comparable PCR efficiencies, precision and bias levels. All work conducted within this project was performed under the adapted assay conditions, and the SOP updated accordingly.

Additional validation work was undertaken to investigate whether the source of horse calibration material impacts upon the quantitative performance of the Köppel/Laube methodology through the action of sequence variation within the assay target regions. The genetic regions targeted by the two component assays are known to be highly conserved and expected to contain few polymorphic sites. Figure 1 and Figure 2 illustrate the low levels of predicted sequence variability associated with the Laube and Köppel qPCR assays when targeting Equus callabus. The reverse primer utilised within the Laube assay shows a single base mismatch with all the horse myostatin gene sequences identified. However, this mismatch appears to be conserved across the species and centrally located within the primer binding sequence which should minimise any quantitative impact whilst ensuring broad meat species applicability.

10 20 30 40 50 60 70 80 90. . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . .

AY840554 (2342 to 2438) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T T T T T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241334 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241335 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G A A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241336 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241337 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G A A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241338 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241339 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241340 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241341 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241342 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TEU241343 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C C C T G G T A TEU241344 (134 to 230) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TGQ183900 (3009 to 3105) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TKC708231 (3009 to 3105) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TKC708232 (3009 to 3105) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TKC708233 (3009 to 3105) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TNM_001081817 (509 to 605) T T G T G C A A A T C C T G A G A C T C A T C A A A C C C A T G A A A G A C G G T A C A A G G T A T A C T G G A A T C C G A T C T C T G A A A C T T G A C A T G A A C C C A G G C G C T G G T A TPrimers & Probe T T G T G C A A A T C C T G A G A C T C A T . . . . C C C A T G A A A G A C G G T A C A A G G T A T A C T G . . . . . . . . . . . . . . . . . . . . . . . A T G A A C C C A G G C A C T G G T A T

Figure 1. ClustalW alignment of horse myostatin sequences (each row represents a unique accession number) covering the Laube PCR amplicon region using BioEdit Sequence Alignment Editor v7.1.11 (Ibis Biosciences, Carlbad, USA). Database search limited to Equus callabus species within the NCBI ‘nr Nucleotide Sequence Database’ (National Center for Biotechnology Information, Bethesda, USA)

10 20 30 40 50 60 70 80 90 100. . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . . . . | . .

XM_001498656 (1687 to 179 C C A A C T T C A T C A T G G A C A A C G C C T A C T T C T G T G A G G C A G A T G C C A A A A A G T G C A T C C C C G T G G C C C C T C A C G T C G A G G T T G A A T C A C G T G T C G A G C C A A G C T T T A A CXM_005604292 (1609 to 171 C C A A C T T C A T C A T G G A C A A C G C C T A C T T C T G T G A G G C A G A T G C C A A A A A G T G C A T C C C C G T G G C C C C T C A C G T C G A G G T T G A A T C A C G T G T C G A G C C A A G C T T T A A CAF392878 (600 to 706) C C A A C T T C A T C A T G G A C A A C G C C T A C T T C T G T G A G G C A G A T G C C A A A A A G T G C A T C C C C G T G G C C C C T C A C G T C G A G G T T G A A T C A C G T G T C G A G C C A A G C T T T A A CPrimers & Probe C C A A C T T C A T C A T G G A C A A C G C . . . . . . . . . . . . . . . . . . . . . . . . . A A G T G C A T C C C C G T G G C C C C T C A . . . . . . . . . . . . . . . . C G T G T C G A G C C A A G C T T T A A C

Figure 2. ClustalW alignment of horse growth hormone receptor sequences (each row represents a unique accession number) covering the Köppel horse PCR amplicon region using BioEdit Sequence Alignment Editor v7.1.11 (Ibis Biosciences, Carlbad, USA). Database search limited to Equus callabus species within the NCBI nr Nucleotide Sequence Database (National Center for Biotechnology Information, Bethesda, USA)

The study evaluated the impact of horse specimen on the quantitative performance of the methodology through the use of calibrants and test samples generated from two different horses (A & B). Table 2 demonstrates the similar values of the performance characteristics associated with calibration curves derived from horse specimen A or B, supporting the assumption that different tissue sources will not impact upon assay performance

Table 2. Assay performance metrics comparing calibration data generated using horse specimen A and B. Performance metrics generated based on data from triplicate reactions within three plates (n=3) comprising a 7 point serial dilution series of 100% horse DNA (20,480 to ~5 horse genome equivalents)

qPCR Assay Horse Mean R2 Mean PCR Efficiency

Köppel HorseA 0.998 95.150B 0.996 96.182

Laube Mammalian

A 0.997 100.209B 0.997 96.790

The two sets of calibrants were successfully used to quantify the % horse adulteration levels in a range of ad-mixtures prepared around the 1% threshold level. Figure 3 demonstrates that test samples analysed

EVID4 Evidence Project Final Report (Rev. 10/14) Page 11 of 18

using a specific calibration set do not exhibit any obvious tissue source related differences in the experimentally determined adulteration level estimates (similar results and overlapping confidence intervals).

Tissue Comparison Study

Assigned % Horse Adulteration

Obs

erve

d M

ean

% H

orse

Adu

ltera

tion

0.1 0.5 10.0

0.5

1.0

1.5

2.0

Horse A Calibrants /Horse A Ad-mixturesHorse B Calibrants /Horse A Ad-mixturesHorse A Calibrants /Horse B Ad-mixturesHorse B Calibrants /Horse B Ad-mixtures

Figure 3. Comparison of % horse adulteration data sets generated using horse A and B specimens. Data from three replicate plates (n=3) based on a seven point calibration curve of 100% horse DNA (20,480 to ~5 horse genome equivalents) with w/w horse in beef ad-mixture test samples (1%, 0.5%, 0.1%) per target assay at a triplicate technical replicate level. Error bars represent +/- 2 SD.

Statistical analysis (Analysis of Variance) was conducted to identify whether the data sets generated using calibrants derived from horse specimen A or B were statistically different. Table 3 shows that significant differences existed between test samples (p-value <0.001) as expected, there was no significant difference between the results based on the calibration sets (p-value 0.97), and the test samples were performing consistently between the two calibration sets (p-value 0.99). These results indicated that the source of horse calibration material did not impact upon the quantitative performance of the Köppel/Laube methodology. The statistical results were confirmed by independent analyses performed by the LGC Statistical Team based on log transformed data, with no appreciable differences in results.

Figure 3 shows large differences between the three test samples (0.1, 0.5 and 1% w/w samples) as expected, but no differences between the use of Calibrant A and Calibrant B for a particular ad-mixture (A or B) within each of the test samples. There appears to be a difference between ad-mixtures within a test sample, which is most likely to reflect the physical inaccuracies of repeatably preparing the gravimetric standards at such low level concentrations. Note that this difference is consistent between ad-mixtures within each of the test samples, but this does not impact upon the estimation of the horse content of the samples using Calibrants A or B.

Table 3. Analysis of Variance applied to pooled data set by combining % adulteration data derived using calibration set A or B.

Source of Variation SS df MS F P-value F critTest Sample 89646.36 7 12806.62 313.84 <0.001 2.31Calibration set 0.08 1 0.08 0.00 0.97 4.15Interaction 44.73 7 6.39 0.16 0.99 2.31Within 1305.80 32 40.81

EVID4 Evidence Project Final Report (Rev. 10/14) Page 12 of 18

Total 90996.97 47

Objective 03 – Evaluate fitness for purpose for enforcement activities at the 1% thresholdFitness for purpose at the recommended 1% threshold level used to guide enforcement is critical to successfully rolling out the Köppel/Laube approach to the PA community. Work was conducted to evaluate the performance of the methodology at the 1% threshold level by testing a small panel of 1% horse muscle tissue in a background of beef muscle tissue ad-mixtures. Four independent 1% w/w samples were evaluated across triplicate experimental plates, with a triplicate PCR replication level within each plate (Figure 4).

1% Threshold Study

1% Horse in Beef Ad-mixtures

Obs

erve

d M

ean

% H

orse

Adu

ltera

tion

Ad-mixt

ure A

Ad-mixt

ure B

Ad-mixt

ure C

Ad-mixt

ure D

0.0

0.5

1.0

1.5

2.0

2.5

% HorseAdulteration

1% Threshold

Figure 4. An evaluation of 1% horse adulteration test samples. Mean % adulteration values were calculated from triplicate data sets. Error bars represent +/- 2 SD. Labels A – D represent individual 1% w/w gravimetric preparations that were made separately.

An estimate of the analytical measurement uncertainty (based on a 95 % confidence interval) was determined for the panel of 1% horse adulteration test samples to be 1.58 +/- 0.54 % w/w raw horse meat in a raw beef background (calculated from the mean of samples A – D). This estimate was based on four independent 1% w/w gravimetric samples across three replicate plates, with three PCR replicates within a plate. Note that the observed percentage horse adulteration values ranged from 1.32 % and 1.93 %, a positive bias which may represent an artefact of the sample preparative approach employed. The log transformed data gave a measurement uncertainty estimate that was not appreciably different from this value. This would indicate that the methodology is fit for purpose at the established 1% threshold guidance level for enforcement activities.

Please note that this measurement uncertainty estimate is based on specific experimental conditions and only represents the analytical measurement uncertainty involved in the laboratory analysis, including replicate PCRs, replicate plates, and replicate independent samples. This estimate of the analytical measurement uncertainty does not take into account measurement uncertainty from upstream sources such as sampling, which are outside the remit of the current project. The estimate of the measurement uncertainty of 1.58 +/- 0.54% equates to a coefficient of variation (CV) of approximately 17%. This compares well with estimates quoted in the published literature when using real-time PCR for trace detection. For example, in relation to using real-time PCR for estimation of durum wheat adulteration around the 3% legislative limit, published literature has cited values of 3.0 +/- 1.6% 6. This equates to a CV of around 33%. Published literature for quantitation of meat species have also stated that it was sometimes difficult to distinguish between 50% and 90% adulterated samples, and use of some methods previously

EVID4 Evidence Project Final Report (Rev. 10/14) Page 13 of 18

Comparing the Quantitative Performance of Alternative Methodologies

Test Sample

Obs

erve

d %

Hor

se A

dulte

ratio

n

1% Horse

in B

eef B

urger

1% Horse

in C

orned Bee

f

1% Horse

in C

ottage P

ie

1% Horse

in Las

agne

5% Horse

in B

eef B

urger

5% Horse

in C

orned Bee

f

5% Horse

in C

ottage P

ie

5% Horse

in Las

agne

0

5

10

15

20

40

50

60

Köppel/LaubeSureFood®ANIMAL QUANTEquus Kit

investigated in the published literature did not prove robust enough for reliable quantitation 7.

Objective 04 – Assess the suitability of the method for use in complex foodsThe successful transfer of the methodology to real-world applications necessitates that the approach is capable of detecting and ideally quantifying horse DNA within a range of complex and processed food materials. As it is well known that meat species quantification in processed food materials is particularly challenging, it was decided to compare the quantitative capability of the Köppel/Laube approach to a leading commercially available test kit brand (SureFood® Animal Quant Equus, Congen Biotechnologie GmbH, Germany) to generate meaningful results. A challenging test panel comprising raw/cooked horse meat in a background of raw/processed beef meat was evaluated by both approaches and the quantitative data (generated from single replicate experiments) presented in Figure 5 and statistically analysed using an Analysis of Variance to determine whether the two test approaches performed comparably (i.e. no significant difference between data sets, Table 4)

Figure 5. A comparison of quantitative test results derived from the Köppel/Laube and SureFood® Animal Quant Equus methodologies. Mean data values from single replicate plates (n=1) based on triplicate PCR’s within a plate are shown. Both methodologies used a standard curve-based approach in combination with mammalian and horse specific assays to determine horse meat content relative to total meat content.

Figure 5 displays the comparable quantitative performance associated with both methodologies (based on the mean values derived from a triplicate level of PCR replication within a plate). This figure highlights the challenge of preparing representative highly processed food ad-mixtures typified by food products such as corned beef which shows an over estimation of horse meat content in the corned beef ad-mixtures, i.e. 60% and 49% adulteration observed with 5% w/w horse in corned beef ad-mixtures. Corned beef is a highly processed food subjected to high pressure/temperature treatment that degrades gDNA and impacts on the levels of amplifiable DNA. The loss of amplifiable DNA due to food processing can be observed in Figure 6 that shows the estimated mean reference gene copy numbers associated with the test samples and illustrates the higher amplifiable DNA content found in raw meat products. For example, the 1% horse in beef burger contains more than 2,500 copies of detectable template per 50 ng, whereas highly processed found products such as corned beef was found to contain less than 50 copies of detectable template. Previous work by researchers such as Hird et al 8 supports these observations and highlights the impact of food processing on the likelihood of detecting trace contamination within the modern food chain.

EVID4 Evidence Project Final Report (Rev. 10/14) Page 14 of 18

Comparison of Total Amplifiable DNA Content by Alternative qPCR Assays

Test Sample

Estim

ated

Mea

n R

efer

ence

Gen

e C

opy

Num

ber (

n=3)

1% Horse

in B

eef B

urger

1% Horse

in C

orned Bee

f

1% Horse

in C

ottage P

ie

1% Horse

in Las

agne

5% Horse

in B

eef B

urger

5% Horse

in C

orned Bee

f

5% Horse

in C

ottage P

ie

5% Horse

in Las

agne

1

10

100

1000

10000

Köppel/LaubeSureFood®ANIMAL QUANTEquus Kit

Figure 6. A comparison of total amplifiable DNA content by alternative reference qPCR assays (both targeting mammalian genes) from the Köppel/Laube and SureFood® Animal Quant Equus methodologies (single replicate plates, triplicate measurements). Standard curve-based approach used to determine total amplifiable meat content. Log10 scale used to aid visual interpretation of the data. Error bars represent +/- 2 SD.

The Analysis of Variance test (Table 4) shows that there was no significant difference between quantitative data generated by either methodology. This reinforced that the quantitative performance characteristics of Köppel/Laube approach were comparable to a proprietary leading commercial SureFood® Animal Quant Equus kit, and therefore deemed as suitable and exhibiting similar performance characteristics as this commercial kit. The statistical results were confirmed by independent analyses performed by the LGC Statistical Team based on log transformed data, which also showed no appreciable differences between the two methods.

Table 4. Analysis of Variance (single factor, 95 % confidence interval set) statistical test function within Microsoft Excel 2010 (Microsoft UK, Reading, UK) applied to combined Köppel/Laube and SureFood® Animal Quant Equus data sets (observed % horse adulteration, n=1)

Source of Variation SS df MS F P-value F critBetween Groups 32.1243 1 32.1243 0.1184 0.7359 4.6001Within Groups 3799.8696 14 271.4193

Total 3831.9939 15

5.0 DiscussionActivity surrounding the topics of food authenticity and food fraud continues to grow within the European food industry partly due to the pressures faced by producers within today’s challenging financial climate and the international nature of modern food production. The recent EU wide issue involving the detection of the undeclared presence of horse-meat in beef products destined for human consumption has emphasised the need for the development of accurate analytical approaches for the quantitative detection of meat species in a sample.

In response to the EU horse-meat issue, a real time PCR method was developed at LGC for the quantitation of horse DNA (Defra project FA0135). It was demonstrated that that this assay was capable of detecting and accurately determining the amount of horse DNA present in beef sample mixtures, and the

EVID4 Evidence Project Final Report (Rev. 10/14) Page 15 of 18

method was validated through an objective assessment of performance characteristics using DNA:DNA ad-mixtures and w/w raw horse-meat in raw beef (meat) materials.

In consultation with Defra’s Authenticity Methods Working Group Technical Sub-Group (AMWG-TSG), it was decided to validate the method further to help facilitate adoption of the method by Public Analysts and industrial stakeholders. The current project (Defra project FA0146) focussed on the three key areas of: (i) method validation using different horse samples; (ii) evaluating fitness for purpose for enforcement activities at the 1% threshold and (iii) assessing the suitability of the method for use in complex foods.

The broad applicability of the method to horse meat quantitation is dependent upon the universality of assay response to horse samples derived from multiple sources. Additional method validation activities focussed on establishing the potential impact of different horse specimens upon the quantitative performance of the Köppel/Laube methodology. In-silico analyses indicated that the genetic markers employed within the two component assays are highly conserved within the Equus callabus species and unlikely to contain polymorphic sites that could impact upon PCR performance. Experimental work comprising calibrant and test materials prepared from multiple horse specimens supported this assumption and demonstrated that the source of horse calibration material does not impact upon the quantitative performance of the Köppel/Laube methodology. Assay performance metrics such as PCR efficiency and analytical results were found to be comparable between experiments utilising different sources of calibration materials. Statistical analyses found the quantitative data sets generated using either calibration materials to not be significantly different (0.97 p-value) which supports the universality of the approach. Additional non-biological factors are likely to influence quantitative performance of real-time PCR assays, inclusive of such factors as the DNA extraction method used to generate test and calibrant materials, which should be adequately considered by the analyst.

Following recent EU-RL guidance concerning the analysis of test samples for horse-meat content above or below a de-minimus threshold of 1%, a panel of 1% w/w ad-mixtures was evaluated to ensure fitness for purpose at the guidance threshold level for enforcement activities. The observed % horse adulteration values ranged from 1.32 % to 1.93 %, a positive bias which may reflect the difficulties in repeatably preparing gravimetric samples for use in trace level detection. An estimate of the mean and associated analytical measurement uncertainty of the result (based on four 1% w/w samples, assessed across three replicate plates with three PCR replicates per plate) was determined for the panel of 1% horse adulteration test samples to be 1.58 +/- 0.54 % w/w raw horse meat in a raw beef background. Based on related published literature in the field concerning the use of real-time PCR for quantitation of trace levels of ingredients in foods, the measurement uncertainty estimate appeared reasonable. This would indicate that the methodology is fit for purpose at the recommended 1% threshold level used to guide enforcement activities.

All DNA approaches for the identification and quantitation of meat species exhibit the performance characteristic of increased measurement uncertainty when applied to processed food samples. This is due to a combination of factors including the DNA target chosen, the integrity and quality of the DNA in the processed food material, levels of DNA inhibitors, and a general lack of guidance and agreement on how meat speciation results should be expressed. Food processing can impact on the performance of PCR assays (through the presence of PCR inhibitors and DNA degradation) and hence the quantitative capability of a real-time PCR based test. Commercially sourced samples, representative of those used in the first UK-wide survey of food authenticity in processed meat products, comprising ready meals, raw and canned products, were used to help characterise and model the effects of food matrices on the horse meat quantitation assay. The study assessed the applicability of the Köppel/Laube method to testing for horse meat contamination in processed food products and compared analytical performance with a commercially available proprietary test kit (SureFood® Animal Quant Equus kit). As expected, highly processed food products (e.g. corned beef) were particularly challenging for both methodologies due to low levels of amplifiable template and potential matrix interferents, but demonstrated comparable quantitative capabilities across all the spiking levels (1 % & 5 %).

Both approaches demonstrated up to two fold positive bias in the estimation of horse meat content when applied to beef burgers and cottage pie test samples, up to four fold positive bias when applied to lasagne test samples, and up to 20 fold positive bias when tested with corned beef samples. The real-time PCR method measures the amplifiable horse DNA relative to the total amplifiable mammalian DNA in a sample, and when the DNA template has been differentially degraded (as with highly processed food samples) any quantitative DNA approach can demonstrate significant bias when used to estimate the meat content of such a sample. As expected, highly processed food products (e.g. corned beef) were particularly challenging for both methodologies due to low levels of amplifiable template and the presence of potential matrix interferents, but comparable quantitative capabilities across all the spiking levels were demonstrated for both approaches. Statistical analyses showed that there was no significant difference between quantitative data generated by either methodology.

The additional method validation studies conducted under this project have demonstrated that under the validation parameters investigated, the methodology is capable of reliably detecting and/or quantifying

EVID4 Evidence Project Final Report (Rev. 10/14) Page 16 of 18

horse DNA content across a range of adulteration levels within raw and certain processed food product types.

DNA approaches tend to exhibit significant bias when applied to highly processed food materials due to the differentially degraded DNA. In this report, the real-time PCR assay demonstrated some over-estimation of the horse meat content of very highly processed food samples, to the same level as a leading commercially available test kit.

Follow-on work is required to help establish inter-laboratory performance characteristics based on a Collaborative Trial of the Standard Operating Procedure and to further confirm the suitability of the methodology for horse meat quantitation within the PA and commercial testing networks. A Knowledge Transfer event is required in order to disseminate the results of the real-time PCR approach for quantitation of horse DNA (FA0135) and the subsequent additional method validation aspects (FA0146) to stakeholders representing enforcement and industrial laboratories.

6.0 Main implicationsThe adulteration of meat products, whether deliberate or adventitious, remains an important food labelling issue with serious food authenticity and possible religious implications. The additional method validation work conducted within this project has further established the methodology as an alternative non-proprietary approach that is well suited to the detection and quantitation of horse adulteration in a sample. The Laube/Köppel methodology is fit for purpose and well positioned for subsequent inter-laboratory development and evaluation work within the combined framework of Defra/FSA, the UK PA network and UK food industry (e.g. through Knowledge Transfer events and a Collaborative Trial).

Following generation of inter-laboratory performance characteristics as part of a Collaborative Trial of the method, additional objective data may be generated to further support the fitness for purpose of the method. The addition of an appropriate method to the testing toolkit available to PAs and food testing laboratories will inform and strengthen the decision making processes of food companies and regulators. Food companies could apply the methodology to quality management systems within their supply chains and production processes to ensure that potential adulteration (deliberate or adventitious) is quickly identified and corrective actions (supply or production-based) are implemented before costly and publically damaging interventions are required such as product recalls. Regulators will also be able to confidently enforce labelling in cases where this method identifies a non-compliant result. Access to a fully validated, quantitative method for the determination of horse DNA in raw meat samples will help enable honest traders to robustly defend their food supply chain.

7.0 Possible future workOptions for consideration include:

1. Delivering a Knowledge Transfer Event/Workshop to disseminate and further develop expertise within the Public Analyst network and food industry. Any dissemination events should be made available to PA and industrial representatives in order broaden access to Defra funded outputs. Defra’s AMWG have approved the motion to take these project results forward as a Knowledge Transfer exercise. Following consultation with stakeholders, LGC is well positioned to develop a range of w/w meat adulteration reference materials to improve standardisation and comparability within the area.

2. Organise and participate in an international Collaborative Trial of the method in order to assess the precision (between laboratory reproducibility) of the method, and involve a number of UK official control laboratories and official control laboratories from other EU member states.

3. Expand the scope of the Köppel/Laube approach to include additional commercially/societally important test species such as pork and lamb using currently available published qPCR assays. The development and validation process will build upon the current horse meat quantitation test and expand the range of species that can be reliably quantified (all genomic-based targets).

4. Transferring a panel of simple to use relative quantitative assays (such as those validated under option 3) into a validated Pre-Spotted Plate format (similar to that employed by the EU-RL GMFF) in consultation with EU-RL and associated regulatory/industry bodies. The availability of meat speciation and quantitation test plates would reduce associated costs and improve access to qPCR-based diagnostic testing.

EVID4 Evidence Project Final Report (Rev. 10/14) Page 17 of 18

References to published material9. This section should be used to record links (hypertext links where possible) or references to other published

material generated by, or relating to this project.

References1 LGC. Method development for the quantitation of equine DNA and feasibility of establishing

objective comparisons between measurement expression units (DNA/DNA compared to w/w tissue) - FA0135. FA0135, (2013), http://randd.defra.gov.uk/Document.aspx?Document=11331_Final_report_FA0135.pdf.

2 Köppel, R., Ruf, J. & Rentsch, J. Multiplex real-time PCR for the detection and quantification of DNA from beef, pork, horse and sheep. Eur Food Res Technol 232, 151-155, doi:10.1007/s00217-010-1371-y (2011), http://dx.doi.org/10.1007/s00217-010-1371-y.

3 Laube, I. et al. Methods for the detection of beef and pork in foods using real-time polymerase chain reaction. International Journal of Food Science & Technology 38, 111-118, doi:10.1046/j.1365-2621.2003.00651.x (2003), http://dx.doi.org/10.1046/j.1365-2621.2003.00651.x.

4 European Union Reference Laboratory for Animal Proteins in feedingstuffs. Addendum to the EURL-AP protocol : Cut-Use of the EURL-AP reference material to check the 1% level threshold of horse meat in beef. (2013), http://eurl.craw.eu/img/page/sops/Protocol-reference%20material-final.pdf.

5 European Union Reference Laboratory for Animal Proteins in feedingstuffs. Detection of horse DNA using real-time PCR. EURL-AP recommended protocol. (2013), http://eurl.craw.eu/img/page/sops/Protocol%20for%20detection%20of%20horse%20DNA%20using%20real-time%20PCR.pdf.

6 Wiseman, G., Bowler, P. & Ohara, K. Q01085: The quantitative determination of common wheat in durum pasta using real-time PCR method-validation. Q01085, (Food Standards Agency, 2004).

7 Hird, H., Saunders, G., Popping, B., Garrett, S. & Wiseman, G. Q01084/87/88/89/90: Final optimisation and evaluation of DNA based methods for the authentication and quantification of meat species. Q01084/87/88/89/90, (Food Standards Agency, 2005), http://www.foodbase.org.uk//admintools/reportdocuments/267-1-488_QO1084_species_id_final_report.pdf.

8 Hird, H. et al. Effect of heat and pressure processing on DNA fragmentation and implications for the detection of meat using a real-time polymerase chain reaction. Food additives and contaminants 23, 645-650, doi:10.1080/02652030600603041 (2006), http://dx.doi.org/10.1080/02652030600603041.

EVID4 Evidence Project Final Report (Rev. 10/14) Page 18 of 18