OMAMAN-47-ERPGlutenAssays-MeetingBook

ERP FOR GLUTEN ASSAYS  REVIEW OF 

OMAMAN‐47: QUANTIFICATION OF WHEAT, RYE, AND BARLEY  GLUTEN IN OAT AND OATS PRODUCTS BY ELISA RIDASCREEN® TOTAL  GLUTEN USING   AOAC SMPR 2017.021 FOR QUANTITATION OF WHEAT, RYE, AND  BARLEY GLUTEN IN OATS. 

MEETING ON THURSDAY, DECEMBER 6, 2018

ERP for Gluten Assays

2

Preamble

AOAC Official Method of Analysis℠

RIDASCREEN® Total Gluten

In 2018, R-Biopharm and General Mills initiated a collaborative study of a method for quantitative

measurement of wheat, rye, and barley gluten in oats and oat products following requirements laid

down in AOAC SMPR®2017.021.

Following the AOAC rules for Official Method First Action additional data from an in-house

validation study needs to be submitted.

The following submission contains

RIDASCREEN®Total Gluten OMA Manuscript

Method Safety Checklist

RIDASCREEN® Total Gluten Package Insert (preliminary version for collaborative test)

RIDASCREEN®Total Gluten In-House Validation Report

MSDS for Cocktail (patented) and Ethanol

December 6, 2018

ERP for Gluten Assays

3

Annex B

RIDASCREEN ® Total Gluten

In-house validation report for AOAC OMA First Action

Contents

Abstract

2

1

Scope of Method

1.1

Target Analyte

3

1.2

Matrices

3

1.3

Summary of Validated Performance Claims

3

2

Introduction

2.1

Principle

4

2.2

General Information

4

2.3

Summary of Results

4

3

Materials and Methods

3.1

Test Kit Information

5

3.2

Additional Supplies and Reagents

5

3.3

Apparatus

6

3.4

Standard Reference Material

6

3.5

Standard Solution and Spike Solution

6

3.6

General Preparation

6

3.7

Preparation of Components

7

3.8

Sample Preparation

7

3.9

Analysis

8

3.10 Reading

9

3.11 Calculation

9

3.12 Criteria for Acceptance of Standard Curve

9

4

Summary of Results

4.1

Manufacturer’s In-house Study

10

4.2

Independent Validation studies

32

5

Conclusion

32

6

References

32

7

Acknowledgements

33

Page 1 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

4

Annex B

Abstract

The sandwich ELISA RIDASCREEN ® Total Gluten (R-Biopharm, R7041) employs a combination of four monoclonal antibodies including the R5 to detect intact (non-hydrolyzed) gliadins and related prolamins from rye and barley, High-Molecular Weight (HMW) Glutenin-Subunits (GS) from wheat, HMW-secalins from rye and Low-Molecular-Weight (LMW)–GS from wheat. Samples are extracted by Cocktail solution/80% ethanol and analyzed within 50 minutes. The measurement range is between 5 mg/kg gluten and 80 mg/kg gluten using a calibrator made out of a gluten extract from a mixture of four wheat cultivars. The system showed no cross-reactivity against 83 different food commodities. LoD was found to be between 0.4 mg/kg and 1.9 mg/kg for oat-based food. LoQ was verified at a level of 5 mg/kg. AOAC reference materials mentioned in SMPR ® 2017.021 showed recoveries between 86% and 131% for wheat and barley, while rye showed higher recoveries between 137% and 170%. Recoveries studies with incurred oat cookies and porridge revealed values between 55% and 122%. The precision of the whole procedure is mainly driven by inhomogeneity of samples which was proven by a precision of extraction experiment. A ruggedness study revealed no significant parameters for the extraction or ELISA procedure. The test kit is stable for at least 3 weeks at 37°C. The results of the in-house validation study confirmed that the method is accurate and suitable to measure gluten from all three grain sources with similar recoveries according to SMPR ® 2017.021.

Page 2 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

5

Annex B

1

Scope of Method

1.1

Target Analytes

All gluten fractions from wheat, rye and barley with the exception of D-hordein from barley; In detail these are: intact (non-hydrolyzed) gliadins and related prolamins from rye and barley, High-Molecular Weight (HMW) Glutenin-Subunits (GS) from wheat, HMW-secalins from rye and Low-Molecular-Weight (LMW)–GS from wheat.

1.2 Matrices

In course of this in-house validation, RIDASCREEN ® Total Gluten was tested with the following matrices: oat flour, flaked oats, oat cereals, groats, oat cookies, and oat porridge.

1.3

Summary of Validated Performance Claims

According to the manufacturer, the RIDASCREEN ® Total Gluten detects gluten at 1.9 mg/kg. No cross-reacting substance has been identified by the manufacturer. LoQ was proven to be at 5 mg/kg gluten. AOAC reference materials mentioned in SMPR ® 2017.021 (1) showed recoveries between 86% and 131% for wheat and barley, while rye showed higher recoveries between 137% and 170%. Recovery studies with incurred oat cookies and porridge revealed values between 55% and 122%. The precision of the whole procedure is mainly driven by inhomogeneity of samples. A ruggedness study revealed no significant parameters for the extraction or ELISA procedure. The test kit is stable for at least 3 weeks at 37°C.

Method/Report Authors Markus Lacorn and Thomas Weiss R-Biopharm AG An der neuen Bergstrasse 17 D-64297 Darmstadt, Germany

Submitting Company R-Biopharm AG

An der neuen Bergstraße 17 D-64297 Darmstadt, Germany

Independent Laboratory Not applicable (covered by collaborative study)

Reviewers To be determined

Page 3 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

6

Annex B

2

Introduction

2.1

Principle

The basis of the test is the antigen-antibody reaction. The wells of the microtiter plate are coated with specific monoclonal antibodies against gluten proteins. By adding the standard or sample solution to the wells, present gluten proteins will bind to the specific antibodies. The result is an antibody-antigen complex. In a washing step components not bound are removed. Then antibodies conjugated to peroxidase (enzyme conjugate) are added. This antibody conjugate is bound to the antibody-antigen-complex. An antibody-antigen- antibody-complex (sandwich) is formed. Substrate/chromogen is added after removing of any unbound enzyme conjugate in a washing step. Bound enzyme conjugate converts the chromogen into a blue product. The addition of the stop solution leads to a color change from blue to yellow. The measurement is made photometrically at 450 nm against air. The absorption is proportional to the gluten protein concentration in the sample. The use of wheat flour and gluten in foodstuff is extremely common because of their useful effects on e.g. texture, moisture retention and flavor. Gluten is a mixture of prolamin and glutelin proteins present in wheat, rye and barley. Coeliac disease (CD) is a permanent intolerance to gluten that results in damage to the small intestine and is reversible when gluten is avoided by diet (2). The Codex Alimentarius Commission has stipulated in the „Codex Standard for Foods for Special Dietary Use for Persons Intolerant to Gluten” (3) the limit value for gluten-free food at 20 mg/kg gluten. The official type I method for quantitative gluten determination according to the Codex Alimentarius is an ELISA which uses the R5 antibody (Mendez). This requirement is fulfilled by the sandwich ELISA RIDASCREEN ® Gliadin (Art. Nr. R7001). Since its introduction to the analytical community, the R5 method to quantify gluten led to a strong improvement of the situation for the food industry and CD patients. During the last years some questions arose on the use of the Codex Alimentarius factor of 2 to recalculate from prolamins to gluten, an overestimation of rye and barley, inadequate detection of glutelins, and the inhomogeneous distribution of gluten in oats (4). Especially the inhomogeneity of gluten from wheat, rye or barley in oats led to AOAC Standard Method Performance Requirement (SMPR ® ) 2017.021 which was set up by stakeholders by vote in 2017. RIDASCREEN ® Total Gluten was validated in-house following the AOAC SMPR ® 2017.021 and AOAC Appendix M. Due to the long-lasting experience of the method developer additional experiments will be presented that show a deeper insight into the performance of the method. The system showed no cross-reactivity against 83 different food commodities. LoD was found to be between 0.4 mg/kg and 1.9 mg/kg for oat-based food. LoQ was verified at a level of 5 mg/kg. AOAC reference materials mentioned in SMPR ® 2017.021 showed recoveries between 86% and 131% for wheat and barley, while rye showed higher recoveries between 137% and 170%. Recoveries studies with incurred oat cookies and porridge revealed values between 55% and 122%. The precision of the whole procedure is mainly driven by inhomogeneity of samples which was proven by a precision of extraction experiment. A ruggedness study revealed no significant parameters for the extraction or ELISA procedure. The test kit is stable for at least 3 weeks at 37°C. 2.3 Summary of Results 2.2 General Information

Page 4 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

7

Annex B

3

Materials and Methods

3.1

Test Kit Information

3.1.1 3.1.2

Kit Name

RIDASCREEN ® Total Gluten

Catalogue Number R7041

Ordering Information USA R-Biopharm Inc. 870 Vossbrink Drive

Washington, MO 63090 phone (269)-789-3033 fax (269)-789-3070 e-mail sales@r-biopharm.com Germany R-Biopharm AG An der neuen Bergstrasse 17 D-64297 Darmstadt Germany phone +49-(0)6151-8102-0 fax +49-(0)6151-8102-20 e-mail sales@r-biopharm.de

Further information is available at the web site address: www.r-biopharm.com

3.1.3 Reagents The test kit consists of antibody-coated microwell strips, standards, antibody conjugate, substrate/chromogen, stop solution, buffer, wash buffer (see 3.1.4-3.1.10); All reagents are stable as indicated on the label at 2-8°C (36-46°F). 3.1.4 Antibody-coated microwell strips. 3.1.5 Standards. – Six vials (1.3 mL each, ready to use) gluten proteins in aqueous solution. 3.1.6 Conjugate. – One vial (11 mL, ready to use), peroxidase conjugated antibody solution. 3.1.7 Red Chromogen Pro (Substrate/Chromogen). – One vial (13 mL, stained red). 3.1.8 Stop solution. – One vial (14 mL, contains 1 N sulfuric acid). 3.1.9 Buffer. – One bottle (120 mL, ready to use). 3.1.10 Wash buffer. – One bottle (100 mL, 10fold concentrate).

3.2

Additional Supplies and Reagents

3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 3.2.7

Variable micropipettes (20 – 200 µL and 200 – 1000 µL)

Multistepper pipette and tips for 100 µl

8-channel pipette and tips for 100 µl and 250 µl Graduated pipettes (10 mL and 50 mL) Graduated cylinders (plastic or glass ware)

Distilled water

Ethanol; 99% reagent grade

3.2.8 Cocktail (patented); R7006/R7016 (R-Biopharm AG, Germany); ready to use 3.2.9 Skim milk powder (food quality; gluten-free) 3.2.10 Centrifugal glass vials with a screw top 3.2.11 Fluted paper filter

Page 5 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

8

Annex B

3.3

Apparatus

3.3.1 3.3.2 3.3.3 3.3.4 3.3.5

Microtiter plate spectrophotometer (450 nm)

Scale

Laboratory mincer/grinder, pestle and mortar, or Ultra-Turrax Shaker (e.g. Roto Shaker Genie, Scientific Industries Inc.) Centrifuge (e.g. Minifuge RF, Kendro, Hanau, Germany)

3.3.6 Temperature controlled water bath 50 °C/122°F (e.g. GFL, Burgwedel, Germany)

3.4

(Standard) Reference Material

Oat flour containing wheat or rye or barley gluten at a concentration of 10 mg/kg and 20 mg/kg as mentioned in AOAC SMPR ® 2017.021 (1). Additionally oat blank flour is provided.

3.5

Standard Solution and Spike Solution

The starting material for spiking consists of a mixture of four commercial wheat flours. It is stored at -20°C. For spiking purposes, the material is extracted with Cocktail (patented) and 80% ethanol. Briefly, 2.5 mL Cocktail (patented) is added to 250 mg of wheat flour mixture and incubated at 50°C for 40 min. Afterwards, 7.5 mL 80% ethanol is added, and the vial is rotated upside down for 60 min at room temperature. The solution is centrifuged at 2500 g for 10 min and filtrated. Based on HPLC analysis (Dr. Katharina Scherf, Leibniz-Institute for Food Systems Biology at the Technical University of Munich, Germany) the prolamin content of the mixture is 51.4 mg/g and the glutenin content is 24.9 mg/g, so in sum 76.3 mg/g total gluten. The total gluten content based on the total protein content multiplied by 80% (see SMPR ® 2017.021) is 75.2 mg/g. This value is used for further calculations, so that the spike solution described above has a concentration of 1.88 mg/mL total gluten. This solution is stable for at least 2 weeks at 20-25°C. The spike solution is diluted appropriately to the desired concentration with RIDASCREEN ® Total Gluten buffer. The starting material for preparation of standards was a mixture of four wheat cultivars (var. Akteur, var. Pamir, var. Julius, var. Tommi ). The materials were also characterized by Katharina Scherf for their gluten contents. 3.6.1 This test should only be carried out by trained laboratory employees. The instructions for use must be strictly followed. No quality guarantee is accepted after expiry of the kit (see expiry label). 3.6.2 Store the kit at 2-8°C (36-46°F). Let all kit components come to room temperature 20- 25°C (68-77°F) before use. Do not freeze any of the kit components. 3.6.3 Return any unused microwells to their original foil bag, reseal them together with the desiccant provided and further store at 2-8°C (36-46°F). The substrate/chromogen is light sensitive, therefore, avoid exposure to direct light. 3.6.4 Carefully dilute the components included in the kit as concentrates; avoid contaminations by airborne grain dust or dirty laboratory equipment. Wear gloves during the preparation and performance of the assay. Clean surfaces, glass vials, mincers and other equipment with 40% ethanol or 2-propanol. Carry out sample preparation in a room isolated from ELISA procedure. Check for gluten protein contamination of reagents and equipment. 3.6.5 Include ready to use standards in duplicate to each run of diluted sample extracts in duplicate. Do not reuse wells of the plate. Use separate pipet tips for each standard 3.6 General Preparation

Page 6 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

9

Annex B

and each sample extract to avoid cross-contamination and pre-flush the tip before pipetting standard or sample extract. Use a multistepper pipet for adding the conjugate, substrate/chromogen and stop solution. Use a single tip for each of these components. 3.6.6 In case that more than three microtiterplate strips (24 wells) are necessary, a second uncoated plate should be used as a pre-plate. Standards and samples are quickly transferred to the coated microtiter plate with an 8-channel pipette. 3.6.7 Components and procedures of the test kit have been standardized for use in this procedure. Do not interchange components between kits of different batches (lot numbers). 3.6.8 Store samples in a cold and dry room protected from light. Ensure that no cross- contamination takes place. 3.6.9 If necessary, check for gliadin contamination of reagents and equipment with the RIDA ® QUICK Gliadin (Art. No. R7003). 3.6.10 Keep in mind that solid samples can be inhomogeneous, therefore ground a representative part of the samples very well and homogenize before weighting. 3.7.3 Buffer -- The buffer (see 3.1.9) is ready to use. Bring the solution to room temperature (20-25 °C; 68-77 °F) before use. Make sure that the buffer is not contaminated with gluten during use. 3.7.4 Sample dilution buffer for further dilution -- Use a solution consisting of 1% Cocktail (patented), 3% of 80% ethanol, and 96% buffer, e.g. 50 µL Cocktail (patented), 150 µL 80% ethanol and 4800 µL buffer. Do not use the diluted samples that were already measured for further dilution since the diluted samples are stable for 30 min only. Re- start the dilution using the extract obtained after filtration. Dilute this extract afresh 1:25 (see 3.8.7) with buffer. For additional dilution, use the sample dilution buffer for further dilution. 3.7.5 Washing buffer -- Provided as a 10-fold concentrate (see 3.1.10). Before use, the buffer has to be diluted 1:10 (1+9) with distilled water (i.e. 100 mL buffer concentrate + 900 mL dist. water). Prior to dilution, dissolve possibly formed crystals by incubating the buffer in a water bath at 37 °C (99 °F). The diluted buffer is stable at 20 - 25 °C (68 - 77 °F) for four weeks. 3.7 Preparation of components 3.7.1 80% aqueous ethanol -- Add 200 mL ethanol to 50 mL distilled water and shake well. 3.7.2 Cocktail (patented) -- The Cocktail is ready to use (see 3.2.8)

3.8

Sample Preparation

3.8.1 Weigh a representative amount (200 g) of oats or oat products and homogenize.

3.8.2 Weigh 1 g ± 0.01 g of homogenized sample to a 50 mL centrifuge tube. Add 10 mL Cocktail (patented), cap the tube vial, mix vigorously, and pay attention to obtain a homogenous suspension. For tannin- and polyphenol-containing matrices (e.g. chocolate, coffee, cocoa, chestnut flour, buckwheat, teff flour, millet or spices) add 1 g gluten-free skim milk powder prior to addition of the 10 mL Cocktail (patented).

3.8.3 Add 30 mL 80 % ethanol, close the tube and mix well. Incubate for 40 min at 50 °C in a water bath.

Page 7 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

10

Annex B

3.8.4 Remove samples from the water bath and shake for 1 h up-side-down or by a rotator at room temperature (20 – 25 °C/ 68 – 77 °F).

3.8.5 Centrifuge for 10 min at least 2500 g (alternatively: 2 mL of the extract can be centrifuged with high speed for 10 min in reaction caps by using a micro centrifuge). Afterwards filter the supernatant (with fluted paper filter).

3.8.6

This extract can be stored at room temperature for at least seven days.

3.8.7 Dilute the sample 1:25 with buffer (see 3.7.3), e.g. 40 μL extract + 960 µL buffer (1:1000 final sample dilution). Use diluted supernatant immediately in the assay within 30 min (use 100 μL per well in the assay). 3.8.8 If further dilution is required a solution consisting of 1% Cocktail (patented), 3% of 80% ethanol, and 96% buffer, e.g. 50 µL Cocktail (patented), 150 µL 80% ethanol and 4800 µL buffer. Do not use the diluted samples that were already measured for further dilution since the diluted samples are stable for 30 min only. Re-start the dilution using the extract obtained after filtration. Dilute this extract afresh 1:25 (see 3.8.7) with buffer. For additional dilution, use the sample dilution buffer for further dilution. 3.9.1 Bring all reagents to room temperature (20 - 25 °C / 68 - 77 °F) before use. Carefully follow the recommended washing procedure. Do not allow microwells to dry between working steps. 3.9.2 Do not use more than three strips (24 wells) at a time. In the case of more than three strips, a second uncoated plate (e.g. low binding from Greiner bio-one Cat.-No. 655101) should be used as a pre-plate to avoid a time shift over the microtiter plate. All standards and samples are pipetted into the uncoated plate (at least 150 µL) and then quickly transferred to the coated microtiter plate with an 8-channel pipette. 3.9 Analysis

3.9.3 It is recommended to pipette the conjugate, the substrate/chromogen and the stop solution with a multi-channel or stepper pipette to avoid a time shift over the plate.

3.9.4 Insert a sufficient number of wells into the microwell holder for all standards and samples to be run in duplicate. Record standard and sample positions.

3.9.5 Add 100 μL of each standard solution (see 3.1.5) or prepared sample to separate duplicate wells and incubate for 20 min at room temperature (20 - 25 °C / 68 - 77 °F).

3.9.6 Pour the liquid out of the wells and tap the microwell holder upside down vigorously (three times in a row) against absorbent paper to ensure complete removal of liquid from the wells. Fill all the wells with 250 μL washing buffer (see 3.7.5) and pour out the liquid again. Repeat two additional times.

3.9.7 Add 100 μL of the ready-to-use enzyme conjugate (see 3.1.6) to each well and incubate for 20 min at room temperature (20 -25 °C / 68 -77 °F).

3.9.8 Pour the liquid out of the wells and tap the microwell holder upside down vigorously (three times in a row) against absorbent paper to ensure complete removal of liquid from the wells. Fill all the wells with 250 μL washing buffer (see 3.7.5) and pour out the liquid again. Repeat two additionally times.

3.9.9 Add 100 μL of the reddish substrate/chromogen solution (see 3.1.7) to each well and incubate for 10 min at room temperature (20 - 25 °C / 68 - 77 °F) in the dark.

Page 8 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

11

Annex B

3.9.10 Add 100 μL of the stop solution (see 3.1.8) to each well. Mix gently by shaking the plate manually.

3.10

Reading

3.10.1 Read the results with a microtiter plate reader. Measure the absorbance at 450 nm. Read within 10 minutes after addition of stop solution. 3.10.2 The dilution factor 1000 which results after sample preparation has already been considered for the standard concentrations. The concentration of the sample can directly be read from the standard curve. 3.10.3 A further dilution and new detection of samples is necessary for absorbance readings (A450 nm) > standard 6. Please follow instructions given in 3.8.8. Do not use diluted samples that were already measured for further dilution.

3.11 Calculations

3.11.1 Determine the gluten content of each duplicate sample wells by reference to a calibration curve measured by the actual test run utilizing special computer software; plot absorbance of standards vs gluten content of standards. It is recommended to using the RIDA ® SOFT Win (R-Biopharm AG, Z9999) with 4-parameter logistic regression analysis (5).

3.12 Criteria for Acceptance of Standard Curve

3.12.1 The course of the standard curve is shown in the Quality Assurance Certificate enclosed in the test kit. Absolute absorbances may vary between different runs (e.g. due to different temperatures or analysts). However, the shape of the standard curve should be similar to the one given in the Quality Assurance Certificate. Minimum requirements are: 3.12.2 OD at 450 nm for standard 6 higher than 1.2 3.12.3 OD values for standards should continuously increase with higher concentrations, especially when comparing standard 1 (0 mg/kg) and standard 2 (5 mg/kg) 3.12.4 An OD value for standard 1 which is much higher than the OD value stated in the certificate, could be an indication for errors during pipetting or incubation or contamination. 3.12.5 For a trained technician, coefficients of variations (CVs) for duplicates of standards and samples on the plate are usually less than 10% except for standard 1 (0 mg/kg) where CVs up to 30% can frequently occur. 3.12.6 All requirements should not be judged isolated but as a whole; if one criterion does not fit to the above-mentioned criteria, it is recommended not to repeat the analytical run directly but check for pipetting errors and results from control samples first.

Page 9 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

12

Annex B

4 Summary of Results and Discussion

4.1

Manufacturer’s In-house Study

4.1.1 General remarks The manufacturer’s in-house validation scheme followed the AOAC SMPR ® 2017.021, AOAC Appendix M (6), CLSI guideline EP25-A for stability testing of in-vitro diagnostics (7) and long- lasting practical experiences of the method developers from other AOAC approvals e.g. AOAC Performance Tested Method ℠ no. 101501 for RIDASCREEN ® FAST Milk (8). Selectivity Study: Antibodies and Analytical Targets The sandwich ELISA RIDASCREEN ® Total Gluten contains four different monoclonal antibodies. One of them is the R5 monoclonal which was raised against secalins. The specificity of this antibody is described elsewhere (9). The second antibody was raised against a HMW-GS peptide which contains the toxic sequence GYYPTS (10). The third and fourth antibodies were raised against an LMW-GS extract that was obtained from Katharina Scherf (Leibniz-Institute for Food Systems Biology at the Technical University of Munich, Germany). While the R5 and the HMW-GS antibody are coated on the microtiterplate and are in the conjugate (homologous system), two LMW-GS antibodies were used, because it was not possible to set up a homologous system. One of them is coated on the plate while the other one is in the conjugate. 4.1.2

Figure 1. Reactivity of the combination of all four antibodies against different fractions from wheat or rye or barley gluten; all fractions were tested at concentrations between 5 ng/mL and 80 ng/mL.

OD 450 nm

3.0

2.5

Wheat prolamins

Wheat LMW-GS

2.0

Wheat HMW-GS

Rye prolamins

1.5

Rye glutelins

Barley prolamins

1.0

Barley glutelins

0.5

0.0

0

20

40

60

80

100

Concentration (ng/ml)

Page 10 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

13

Annex B

In the following, these antibodies are characterized for their binding activities against all fractions from gluten. These fractions were also obtained from Katharina Scherf (Leibniz- Institute for Food Systems Biology at the Technical University of Munich, Germany). The preparation is described elsewhere (11). Figure 1 shows the reactivity of the combined system, where three monoclonal antibodies (R5, HMW-GS antibody and LMW-GS antibody 1) are bound to the microtiter plate and three antibodies (R5, HMW-GS antibody and LMW-GS antibody 2) are in the conjugate. With the exception of barley glutelins (mainly consisting of D-hordeins) all prolamins from wheat, rye, and barley as well as all important glutelins are detected at a comparable level. In detail, prolamins from wheat, rye, and barley as well as wheat HMW-GS, wheat LMW-GS, and rye glutelins are measured at a comparable OD value. To investigate the reactivity of each single antibody in the system towards the different gluten fractions, homologous systems were produced for this experiment. As can be seen in figure 2, the R5 monoclonal antibody reacted as expected from the literature with barley prolamins, rye prolamins, and rye glutelins at a higher dose-response compared to wheat prolamins (gliadins). HMW-GS from wheat were also detected at a low dose-response. Glutelins from barley and wheat LMW-GS were not detected (not shown).

Figure 2. Reactivity of the R5 antibody system against different fractions from wheat or rye or barley gluten; all fractions were tested at concentrations between 5 ng/mL and 80 ng/mL; the R5 was coated to microtiter plates and used in the conjugate.

OD 450 nm

4.0

3.5

Wheat prolamins

3.0

Wheat HMW-GS

2.5

Rye prolamins

2.0

Rye glutelins

Barley prolamins

1.5

1.0

0.5

0.0

0

20

40

60

80

100

Concentration (ng/ml)

The main epitope of the R5 antibody is the pentapeptide QQPFP (12), which is highly repetitive on prolamins from wheat, rye and barley (Data base: UniProt). The γ-40k-secalins and γ-75k-secalins which are usually attributed to rye prolamins are also present in the rye glutelins fraction in high abundance (personal communication Dr. Katharina Scherf, Leibniz- Institute for Food Systems Biology at the Technical University of Munich, Germany). These

Page 11 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

14

Annex B

proteins contain the QQPFP epitope and are thus also detected by the R5 antibody. Barley glutelins, LMW-GS, HMW-GS from wheat and HMW-secalins from rye (attributed to the rye glutelins fraction) do not contain the QQPFP epitope. The low response to HMW-GS from wheat might be due to some impurity of the fraction or to a minor cross-reactivity. The monoclonal antibodies that were raised against wheat LMW-GS, show a high reactivity towards LMW-GS but also reactivity towards wheat prolamin (fig. 3) in the heterologous system. All other fractions were not detected (not shown). The exact epitopes of the LMW-GS antibodies are unknown. The protein sequences from LMW-GS are quite different to all other gluten proteins (Data base: UniProt). The low response to prolamins from wheat might be due to some impurity of the fraction or to a minor cross- reactivity. Figure 3. Reactivity of the LMW-GS antibody system against different fractions from wheat; all fractions were tested at concentrations between 5 ng/mL and 80 ng/mL; the LMW-GS antibody 1 was coated to microtiter plates and LMW-GS antibody 2 was used in the conjugate.

OD 450 nm

4.0

3.5

3.0

Wheat prolamins Wheat LMW-GS

2.5

2.0

1.5

1.0

0.5

0.0

0

20

40

60

80

100

Concentration (ng/ml)

The homologous system based on the HMW-GS monoclonal antibody shows the highest reactivity towards the wheat HMW-GS fraction, followed by rye glutelins and prolamins from wheat and rye (fig. 4). All other fractions were not detected (not shown). The HMW-GS antibody was raised against a peptide containing the GYYPTS sequence. It is assumed that this is also the main epitope of this antibody. This sequence is repetitively present on HMW-GS from wheat and HMW-secalins from rye (Data base: UniProt), which are attributed to the rye glutelins fraction. The low response to prolamins from wheat and rye might be due to some impurity of the fractions or to a minor cross-reactivity.

Page 12 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

15

Annex B

Figure 4. Reactivity of the HMW-GS antibody system against different fractions from wheat or rye; all fractions were tested at concentrations between 5 ng/mL and 80 ng/mL; the HMW-GS antibody was coated to microtiter plates and used in the conjugate.

OD 450 nm

3.5

3.0

Wheat prolamins

2.5

Wheat HMW-GS

Rye prolamins

2.0

Rye glutelins

1.5

1.0

0.5

0.0

0

20

40

60

80

100

Concentration (ng/ml)

The barley glutelins (D-hordeins) are quite different to all other gluten proteins (Data base: UniProt). It is therefore not surprising that none of the antibodies raised against rye secalins (R5), LMW-GS or a HMW-GS peptide reacts noticeably with this fraction. However, D-hordeins represent only about 5% of gluten proteins in barley (13).

4.1.3 Selectivity Study: Cross reactivities The cross reactivity panel (table 1 and 2) is slightly modified but still according to Koerner et al. (14) and represents the opinion of the AOAC Allergen community. Some commodities were added to the list due to long lasting experiences of the manufacturer. In total, 83 commodities were tested. The commodities were extracted once like a sample with Cocktail/80% ethanol (3.8) and were tested in the RIDASCREEN ® Total Gluten. Some commodities were tested by adding skim milk powder to mask the interference by polyphenols (3.8). Tables 1 and 2 clearly show that no cross-reactivity (OD of sample > OD of standard 2) exists against the 83 tested commodities. A lot of these commodities are used to compose alternative food for CD patients.

Page 13 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

16

Annex B

Table 1. Compounds tested for cross-reactivity and results of OD readings of these extracts compared to standard 1 (0 mg/kg gluten) and standard 2 (5 mg/kg gluten); compounds marked with an asterisk were extracted with addition of 1 g of gluten-free skim milk powder to mask polyphenols and other interfering substances.

OD Standard 1

OD Standard 2

OD Sample

Cereals Amaranth Arrowroot

0.143 0.187 0.143 0.187 0.106 0.161 0.161 0.187 0.187 0.228 0.160 0.143 0.106 0.143 0.187 0.161 0.161 0.161 0.161 0.161 0.161 0.161 0.161 0.161 0.161 0.111 0.161 0.228 0.228 0.111 0.143 0.143 0.161 0.161 0.161 0.228

0.414 0.444 0.414 0.444 0.429 0.426 0.426 0.444 0.444 0.445 0.443 0.414 0.429 0.414 0.444 0.426 0.426 0.426 0.426 0.426 0.426 0.426 0.426 0.426 0.426 0.390 0.426 0.445 0.445 0.390 0.414 0.414 0.426 0.426 0.426 0.445

0.112 0.116 0.225 0.114 0.255 0.159 0.177 0.141 0.243 0.265 0.238 0.130 0.214 0.234 0.242 0.186 0.152 0.151 0.153 0.151 0.178 0.173 0.168 0.173 0.192 0.113 0.201 0.238 0.219 0.212 0.200 0.269 0.164 0.176 0.218 0.255

Buckwheat, flour*

Chestnut, flour*

Corn starch

Millet, flour* Quinoa, flour

Rice flour, sweet Rice flour, white

Tapioca, flour

Teff, flour*

Beans/ Lentils/ Peas Black bean, flour

Chick peas Fava beans

Garbanzo beans

Guar gum

Green beans

Green pea, flour

Lentil, flour

Lima bean, flour

Lupine, flour

Pea flour, yellow

Peanut, raw

Peanut, roasted

Romano bean, flour

Sorghum, flour

Soya, flour

Soya milk

Soya protein

White bean, flour Seeds Carob, seedlings

Flax seed Pistachio

Poppy seed

Sesame, flour

Sunflower kernel

Page 14 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

17

Annex B

Table 2. Compounds tested for cross-reactivity and results of OD readings of these extracts compared to standard 1 (0 mg/kg gluten) and standard 2 (5 mg/kg gluten); compounds marked with an asterisk were extracted with addition of 1 g of gluten-free skim milk powder to mask polyphenols and other interfering substances.

OD Standard 1

OD Standard 2

OD Sample

Nuts Almond, raw

0.143 0.143 0.143 0.161 0.161 0.161 0.228 0.161 0.143 0.143 0.161 0.228 0.187 0.106 0.187 0.143 0.143 0.187 0.143 0.106 0.106 0.143 0.187 0.106 0.161 0.161 0.161 0.161 0.161 0.161 0.143 0.143 0.143 0.143 0.143 0.143 0.143

0.414 0.414 0.414 0.426 0.426 0.426 0.445 0.426 0.414 0.414 0.426 0.445 0.444 0.429 0.444 0.414 0.414 0.444 0.414 0.429 0.429 0.414 0.444 0.429 0.426 0.426 0.426 0.426 0.426 0.426 0.414 0.414 0.414 0.414 0.414 0.414 0.414

0.124 0.132 0.176 0.130 0.149 0.147 0.219 0.144 0.139 0.176 0.215 0.211 0.240 0.325 0.208 0.202 0.246 0.147 0.315 0.149 0.154 0.287 0.147 0.357 0.165 0.150 0.145 0.184 0.168 0.228 0.141 0.143 0.176 0.177 0.144 0.225 0.235

Almond, roasted

Cashew, raw

Hazelnut, flour Hazelnut, raw Hazelnut, roasted

Walnut, raw

Macadamia, raw

Meat Beef and pork hash

Chicken Sausage

Turkey han

Spices Anise*

Basil*

Caraway*

Cinnamon*

Cloves*

Coriander *

Curcuma*

Curry*

Fennel*

Garlic*

Ginger*

Majoram*

Mustard Powder*

Mustard * Nutmeg* Paprika* Pepper*

Salt*

Other Apple fibre

Apricot, dried fruit

Carrageen

Casein Cacao* Coconut Coffee*

Page 15 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

18

Annex B

Egg powder

0.143 0.106 0.161 0.161 0.161 0.161 0.106 0.228 0.228 0.143

0.414 0.429 0.426 0.426 0.426 0.426 0.429 0.445 0.445 0.414

0.208 0.133 0.142 0.181 0.171 0.175 0.172 0.208 0.166 0.191

Fig, dried fruit Orange juice

Pineapple, dried fruit

Papaya, dried fruit

Potato, flour

Skim milk powder Stabilizer xanthan Sugar beet sirup

Tea, black*

4.1.4 Ruggedness Study These experiments were undertaken to show the influence of parameters on test kit results. These parameters are known to be subject of variation during use of the test kit. The parameters tested for their ruggedness were, [ 1] extraction time at 50°C (first step in water bath), [2] water bath temperature for first incubation step, [3] incubation time for extraction for the second incubation step at room temperature, [4] pipetted volumes for standards and samples, [5] incubation time for all ELISA steps, [6] incubation temperature of the ELISA procedure, and [7] plate shift due to pipetting velocity. Table 3. Variation of water bath temperature from 40°C up to 60°C for extraction at 40 min extraction time (first extraction step); variation of incubation time from 30 min up to 50 min at 50°C extraction temperature (first extraction step); variation of incubation time from 50 min up to 70 min at room temperature (second extraction step); 3 different incurred and processed samples were tested; samples were incurred with wheat, or rye, or barley; results are given in mg/kg gluten. Temperature water bath 1 st Incubation time 2 nd Incubation time Control 40°C 60°C 30 min 50 min 50 min 70 min Sample A (wheat) 33.9 31.2 33.3 29.2 28.7 28.2 30.1 Sample B (rye) 39.0 34.4 36.6 35.4 36.7 38.7 38.6 Sample C (barley) 34.3 31.8 33.8 36.4 41.4 33.3 37.7 Table 4. Variation of pipetted volumes for all pipetting steps from 90 µl to 110 µL; variation of incubation time from 18 min to 22 min for the first and second ELISA incubation step combined with variation of incubation time from 9 min to 11 min for the third ELISA incubation step; variation of incubation temperature from 20°C to 30°C for the whole ELISA procedure; 5 different samples were tested; samples contained wheat, or rye, or barley; results are given in mg/kg gluten. Control 90 µl 110 µl 18 min 22 min 20 °C 30°C Sample A (blank) 2.24 3.29 2.9 2.42 2.9 2.68 1.09 Sample B (wheat) 23.6 28.4 24.5 22.0 24.5 23.3 24.0 Sample C (barley) 31.2 27.7 27.9 27.6 27.9 24.8 42.2 Sample D (wheat) 38.1 39.1 38.1 33.0 38.1 39.6 40.1 Sample E (rye) 33.0 31.9 32.7 29.0 32.7 32.2 39.7

Page 16 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

19

Annex B

Table 3 shows that there is no consistent tendency or trend in the results when extraction parameters were varied within an expectable range, regardless of the contaminating grain Table 4 shows that there is no consistent tendency or trend in the results when ELISA parameters were varied within an expectable range. The reaction of every sandwich ELISA starts with the addition of the first standard / sample. In case that many samples are pipetted onto the same plate, a considerable time shift (different effective incubation times on the plate) can occur between the first well and the last well. For the plate shift experiment, the same sample extract was pipetted on a whole plate, including a standard curve. Standards and the sample extract were pipetted using a normal pipet. The tip for the pipet was exchanged and flushed after every two wells, simulating the pipetting of different sample extracts and standards in duplicates. Each single well with sample extract was evaluated separately (see table 5). Table 5. Pipetting scheme (vertical) for characterization of a shift of results due to pipetting velocity. 1 2 3 4 …. 12 A STD 1 STD 1 sample 5 sample 6 ….. sample 70 B STD 2 STD 2 sample 7 sample 8 …. sample 72 C STD 3 STD 3 sample 9 sample 10 …. sample 74 D STD 4 STD 4 sample 11 sample 12 …. sample 76 E STD 5 STD 5 sample 13 sample 14 …. sample 78 F STD 6 STD 6 sample 15 sample 16 …. sample 80 G sample 1 sample 2 sample 17 sample 18 …. sample 82 H sample 3 sample 4 sample 19 sample 20 …. sample 84 Figure 5 shows the graphical analysis of the plate shift experiment. Overall the plate shift due to the pipetting velocity is quite low. From the formula within figure 5, it can be calculated that 17 mg/kg gluten is achieved after about 42 replicates when the true concentration of the sample was 20. Due to the fact that the method developers advise to use a pre-plate in case of more than 3 stripes, the plate shift effect is negligible. The same is true for an experiment with a horizontal pipetting direction (results not shown).

mg/kg gluten

30

25

y = -0.0727x + 20.045

20

15

10

5

0

0

12

24

36

48

60

72

84

replicate (no.)

Figure 5. Results (mg/kg) from vertical pipetting scheme for one sample extract in 84 single replicates using a micropipette with tip exchange and tip flushing before each duplicate.

Page 17 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

20

Annex B

4.1.5

Accelerated Stability Study

Three independent lots of RIDASCREEN ® Total Gluten were tested to ensure consistent manufacturing between lots and to estimate the possible shelf life of the test kit by performing an accelerated stability study over 3 weeks at 37°C storage temperature (7). For the pilot scale lots the maximum shelf life is normally set to 9 months when stored at 2 – 8°C. Table 6 shows the comparison of three different lots stored at 37°C and tested over a period of 3 weeks. For each lot and time point five different samples were tested. As can be seen in table 6 there is no indication of a significant and consistent trend of results.

Table 6. Accelerated stability study at 37°C over a period of 3 weeks using 3 different pilot scale lots; five different samples were tested; results given as mg/kg gluten.

Week at 37°C

Lot 1

0

1

2

3

sample A sample B sample C sample D sample E Lot 2 sample A sample B sample C sample D sample E Lot 3 sample A sample B sample C sample D sample E

3.04 17.0 16.1 25.2 21.7 3.06 14.9 23.5 28.7 24.5 2.38 17.9 22.6 33.1 24.6

3.54 15.3 17.6 27.5 22.1 3.99 21.4 18.0 28.3 23.4 4.20 28.4 15.8 27.3 19.9

4.28 22.2 15.2 28.6 29.1 4.68 20.6 15.5 25.9 19.0 4.81 23.1 16.4 43.8 20.8

5.93 21.1 15.1 31.6 18.1 7.34 23.9 15.9 33.1 20.9 5.86 23.2 14.5 35.9 18.0

The quality assurance at R-Biopharm AG consists of a final in-process testing in the production department with all final components followed by the official product batch release testing by the quality assurance department and additional real time stability testing after regular storage intervals (usually approx. 3 to 4 months). To show that the proposed shelf life was correct, the testings are normally prolonged for an additional 6 months for each lot. Real-time stability tests for RIDASCREEN ® Total Gluten are still running until 15.08.2020 when a stability of 24 months at 2 – 8°C will be tested for all three lots.

4.1.6 Lot-to-lot comparison Data showing lot-to-lot comparability are available in this report: (1) chapter 4.1.5 - Accelerated Stability Study (2) chapter 4.1.7.7 - Precision of laboratory-internal reproducibility (3) chapter 4.1.7.3 - Trueness

Page 18 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

21

Annex B

4.1.7

Matrix Study

4.1.7.1

Estimation of Limit of Detection (LoD)

To estimate the matrix-specific LoD, blank samples for oat cereals, flaked oats, and oat flour were extracted 10 times and analyzed using 4-parameter logistic regression. Table 7 clearly shows that it was not possible to obtain true blank samples, as is often the case for oat-based samples. Mean concentrations varied between 1.9 mg/kg gluten and 4.1 mg/kg gluten. Therefore it was not reasonable to use the widespread formula LoD = mean concentration of blank samples + 3-fold standard deviation of these samples. For a true blank sample, the mean concentration is zero and the LoD is the threefold standard deviation.

Table 7. Estimation of LoD in cereals, flour, and flakes by extracting 10 replicates (arranged in two columns) and estimating concentrations by using a 4-parameter logistic regression curve fitting; for each matrix and test kit lot the mean concentration, standard deviation, mean + 3- fold SD, and 3-fold SD are given in mg/kg gluten.

Lot 1

Lot 2

mg/kg

Flour

mg/kg

mg/kg

mg/kg

1 2 3 4 5

3.21 3.12 2.39 2.49 2.94

3.16 3.09 3.18 4.76 3.03

2.38 2.42 2.00 1.91 2.24

2.53 2.20 2.06 4.09 2.26

mean

3.14 0.64 5.05 1.92

2.41 0.62 4.27 1.86

SD

mean + 3 x SD

3 x SD

Flakes 1

2.58 2.73 2.82 2.73 2.51

3.01 2.92 3.16 2.76 2.83

1.76 1.82 1.84 2.02 1.91

2.13 2.02 2.00 1.80 1.89

2 3 4 5

mean

2.67 0.13 3.05 0.38

1.87 0.10 2.17 0.30

SD

mean + 3 x SD

3 x SD

Cereals 1

3.67 4.70 3.69 4.38 4.02

4.08 4.40 4.34 3.87 3.52

3.07 4.46 4.40 4.48 3.64

3.77 4.07 4.36 4.60 3.07

2 3 4 5

mean

4.07 0.38 5.22 1.15

3.99 0.58 5.73 1.74

SD

mean + 3 x SD

3 x SD

Page 19 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018

ERP for Gluten Assays

22

Annex B

For this approach one needs to assume that the standard deviation of blank samples and samples with a low but measurable contamination is comparable. Using this approach the estimated LoDs for cereals, flour, and flakes were 1.7 mg/kg, 1.9 mg/kg, and 0.4 mg/kg, respectively (table 7). For confirmation, rice-based gluten-free foods were analyzed (table 8), since it is much easier to obtain rice-based foods completely free of gluten. Indeed, the rice-based foods showed gluten concentrations near to zero. Using the established LoD formula (mean + 3 x SD), an overall LoD of 2.3 mg/kg was estimated, matching the estimates for the oat-based foods.

Table 8. Estimation of LoD in rice-based gluten-free food matrices by extracting 6 replicates (arranged in two columns) and estimating the concentrations by extrapolation; for each matrix the mean concentration, standard deviation, and LoD = mean + 3-fold SD are given in mg/kg gluten.

Gluten-free matrix Food Cake Mix Food Cake Mix Food Cake Mix Pizza Crust Mix Pizza Crust Mix Pizza Crust Mix Rice Flour Blend Rice Flour Blend Rice Flour Blend

mg/kg

mg/kg

mean

SD

LoD

0.06 2.19 0.45 0.62 0.20 0.14 -0.25 0.08 0.11 0.20 1.51 0.53 0.45 0.22 -0.11

0.68 0.73 0.85 1.69 0.25 0.22 0.62 0.31 0.22 1.57 0.73 0.79 0.20 0.11 0.25

0.82

0.72

3.0

0.52

0.60

2.3

0.18

0.29

1.0

Bisquit Mix Bisquit Mix Bisquit Mix Brownie Mix Brownie Mix Brownie Mix

0.89

0.55

2.5

0.19

0.18

0.7

0.43

0.64

2.3

All results

4.1.7.2 Verification of Limit of Quantification (LoQ) As described in chapter 4.1.6.1, it was not possible to obtain true blank samples for flours, cereals and flakes. It was therefore decided to verify the LoQ by a spiking experiment at low level of 2 mg/kg gluten and/or 4 mg/kg gluten. The experiment was repeated in a second lot. As shown in table 9, mean values for blank samples were as expected from the LoD experiment. When spiking 2 mg/kg gluten to cereals, the mean recoveries (corrected for blank samples) were in the acceptable range between 67% and 83%. For the 4 mg/kg spike level, the results were more consistent at 84% which was also mirrored by CVs at or below 15%. Due to the lower gluten values for blank flours and flakes, only 4 mg/kg were spiked. The mean recoveries ranged from 64% up to 104%. CVs were also at or below 15%. This shows that quantification at the 5 mg/kg gluten level is possible with acceptable precision.

Page 20 of 33

RIDASCREEN ® Total Gluten In-house validation report 2018-10-21

December 6, 2018