AOAC A1BC-02 Reviewer Forms (May 2, 2023)

AOAC Stakeholder Program on Infant Formula and Adult Nutritionals (SPIFAN)

A1BC-02 REVIEW FORMS

May 2, 2023

AOAC INTERNATIONAL 2275 Research Blvd., Suite 300 Rockville, MD, 20850 USA

dboyd@aoac.org 301.924.7077 x126

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method A1BC-02

Title: Determination of A1 and A2-type Beta-Casein in Infant Formula by Tryptic Digestion and Stable Isotope Dilution LC-MS/MS Author: The A2 Milk Company Reviewer Name: Reviewer 1 Summary of Method: Proteins are dispersed in waster, denatured by heat treatment, and digested by trypsin. Heavy-labelled internal standards are added. Signature peptides are detected by LC-MS/MS and quantified via an external calibration curve. Method Scope/Applicability: Background The Applicability section of the SMPR states that the method should be able to “measure the amount of bovine A1- and A2- type β -casein found in all forms of infant, adult, and/or pediatric formulas (powders, ready-to-feed liquids, and liquid concentrates).” A1- and A2- type β -casein are defined as follows in the SMPR: - A1- type β‐casein. —Full- length variants of β‐casein that contain histidine at amino acid posit ion 67, including, e.g., phosphorylated and glycated proteoforms. - A2- type β‐casein. —Full- length variants of β‐casein that contain proline at amino acid position 67, including, e.g., phosphorylated and glycated proteoforms. The call for methods further specifies: “In collaboration with ISO/TC 34/SC 5 (Milk and milk products) and the International Dairy Federation (IDF), AOAC invites method developers to submit methods for the determination of A1/A2 beta-Casein in both infant formula (according to the SMPR) and in milk and milk products of interest to ISO/TC 34/SC 5 and IDF: Milk powder, Milk protein concentrate, UHT milk (provided samples can be analyzed during an MLT of the method).”

Matrices

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

The title of the method mentions infant formula, but not validation data was provided. It is not known which products were tested and how many SPIFAN categories (infant formula, pediatric formula, toddler formula, adult formula, ready-to-feed liquids and liquid concentrates) they represent. It is not known whether the method was also validated on milk and milk products as specified in the call for methods (this is not strictly required for AOAC SPIFAN).

Analytes

The method uses three main signature peptides to discriminate between β -casein genetic variants:

Peptide A1: capable of detecting A1, B, C, F, and G peptides (all are of the A1-type) Peptide A2: capable of detecting A2, A3, D, and E peptides (all are of the A2-type) Peptide I: capable of detecting peptide I (of the A2-type)

The authors mention that peptides H1 and H2 cannot be detected with this method. However, β -casein H1 should give, upon tryptic digestion, a peptide that is virtually identical to that originating from peptide A2 (one L->I mutation). Also, β -casein H2 should yield a peptide close to peptide I but with an additional Q->E mutation. This peptide would in principle be measured together with peptide I.

S ince β -casein H1, H2, A2, and I are all of the A2 type, this is not an issue, but the text could be modified.

General comments about the method: This method is a straightforward LC-MS/MS method that uses surrogate peptides and heavy labelled internal standards to quantify three genetic variants (or genetic variant pools) of β -casein. However, the method is not specific towards full- length β -casein, as recognized at the beginning of the document (“The method is specific to protein sequences originating from bovine beta- casein, including full-length and part-length beta casein.”). Indeed, N- terminal proteolytic cleavage products of β -casein (proteose peptones PP8_slow and PP5) will be detected by the peptides A1, A2 and I, while C-terminal fragments (γ -caseins) will be detected by the common peptide. This is a major deviation from the SMPR. The method mentions that it is “applicable for determination of 1 – 1,000 mg/100g (reconstituted basis) of all A1 and A2 type beta caseins (except H1 and H2) in infant formulae”. The calibration curve seems to cover a range from ~4 to ~4000 mg/100g (reconstituted basis), which is very close to the analytical range described in the SMPR (3-1000 mg/100g).

No validation data was provided, making it impossible to assess method performance.

The method should first be modified so that it only measures full- length β -casein genetic variants, as described in the Applicability statement of the SMPR. Comprehensive validation data should be provided before method performance could be properly evaluated.

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Detailed comments about the method:

Apparatus and reagents

e.g. should be used instead of i.e.

Reagent preparation

(a) replace concentration with concentrated

(e) trypsin is usually reconstituted in an acidic environment to avoid autolysis, but not here. Was this tested? Why did the authors use this reconstitution reagent?

Standard preparation

(c) peptide intermediate solutions the 10 mg/L (1 st row) corresponds in fact to a 100 mg/L solution. Please correct this line and other lines if necessary Clarify that 4 x 100 μL means 100 μL of each of the four surrogate peptides The 0.01 mg/L solution is not used later in the document

SIL table: what does “SIL0” mean? It might be a typo. Please correct.

Beta casein stock solution. The procedure asks 1 mg of beta- casein to be dissolved in approx. 100 μL to reach 10,000 mg/L. However, the last sentence says that it can be scaled down to 1mL. Please correct. Beta casein spiking intermediates. The 1 mg/L intermediate (third row, column “source”) is not described in the document.

Sample preparation

The peptide low, medium and high spikes are inverted.

The method does not require sample reduction (nor alkylation). While there are no disulfide bridges in beta casein, overall sample reduction might increase protease accessibility. Was this tested? If yes, were the results comparable to the presented procedure?

Determination

There is no formic acid in mobile phase B. Is this correct? Classically, FA is added to both mobile phases to keep the environment acidic. Can the authors explain why it was not added here? The compound-dependent MS parameter table could benefit from some explanations (what is “.1” and “.2”?, what does “-H” correspond to?) The last sentence of this section describes a calibration curve with area ratios (of peptide vs. internal standard) plotted against concentration. The concentration of the internal standard is not taken into account,

3

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

assuming it is the same throughout, but spikes with protein or peptides (see section “sample preparation”, point (f)) leads to a decrease in the internal standard concentration. This should be taken into account.

Appendix B

Glycosylation (a biologically relevant process) should not be confused with glycation (a process-induced modification) Pros/Strengths: • Method can distinguish surrogate peptides originating from the different genetic variants • Simple sample preparation Cons/Weaknesses • Method is not able to distinguish between full- length and partially hydrolyzed β -casein • In absence of validation data, the sensitivity of the method and its capability to effectively detect the different genetic variants at low concentrations and in a complex matrix can be questioned • The sum of A1, A2 and I peptides is probably not equal to the common peptide (see point 1 above), how are those different values used? • The specificity of each peptide peaks might be strongly influenced in processed raw materials and finished products due to protein modification (protein glycation). How does this relate into numbers. Supporting Data • General Comment: No method performances were shown; a detailed validation report is needed to properly assess this method.

4

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Performance Characteristics:

Analytical Range SMPR:

β -casein, each, expressed per 100 g of reconstituted product: 3–1000 mg/100 g

Report:

close to range – but no validation data are shown

LOQ SMPR:

β -casein, each, expressed per 100 g of reconstituted product: 3 mg/100 g

Report:

~4 mg/100 g for each β -casein, based on the lowest point of the calibration curve.

Data are not shown. This should be evaluated in matrices: the LOQ of each genetic variant should be assessed in presence of the other variants , in particular, the LOQ of β -casein A1 should be assessed in A2-type matrices.

Accuracy/Recovery SMPR: 80-110%

Report:

70-120% for protein spike recoveries

Data are not shown. How was this evaluated? In what matrix/matrices? At which levels? Is the recovery comparable across the analytical range?

Precision (RSD r ) SMPR:

15% for concentrations between 3 and 50 mg/100g 10% for concentrations above 50 mg/100g

Report:

Data are not shown. What is the RSDr in infant formulas and adult/paediatric nutritionals?

Reproducibility (RSD R ) SMPR:

20% for concentrations between 3 and 50 mg/100g 15% for concentrations above 50 mg/100g

Report:

Not applicable here (Single Laboratory Validation data)

Specificity SMPR:

Not mentioned

Report:

No discussion and data are not shown. Representative chromatograms of the processed matrix ingredients could be added. Specificy could be evaluated whenever possible (for internal standards, and for A1 peptide in A2-type matrices at least).

5

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• System suitability:

The method could be expanded to describe system suitability tests (acceptance criteria section) more specifically

1. Is the Validation Study Report in a format acceptable to AOAC? No . The validation report was not transmitted. 2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? Almost . See points above. The calculations need to be clarified to explain how to calculate the relative amounts of A1, A2, and I, how to calculate A1-type and A2-type, and how to use the value of the common peptide. 3. Are the figures and tables sufficiently explanatory without the need to refer to the text? No . Validation data and chromatograms are missing. 4. Are all the figures and tables pertinent? No . Figures (electropherograms) and tables (detailed results) are missing. 5. Could some be omitted and covered by a simple statement? No . The ones that are currently in the document should remain, but extra figures and tables should be added. 6. Are the references complete and correctly annotated? No . The references are only details. All details are missing from the text. 7. Does the method contain adequate safety precaution reference and/or statements? Yes . A generic sentence is added, which is sufficient for this type of procedure as there is not extraordinary danger associated with the method.

Recommendation: The authors should modify their protocol to only measure full- length β -casein and provide:

- A validation report - Full validation data - Representative chromatograms - Clarification on the points listed above before this method can be moved forward to first action status.

6

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method A1BC-02

Title: Determination of A1 and A2-type Beta-Casein in Infant Formula by Tryptic Digestion and Stable Isotope Dilution LC-MS/MS Author: A2 Milk Company Reviewer Name: Reviewer 2 Summary of Method: Powder samples are reconstituted, denatured at elevated temperature, and SIL IS’s for each specific peptide are added. The resulting sample undergoes tryptic digestion, which yields specific peptides for A1-type , A2-type, I-type (an A2-type) beta casein. A peptide attributable to all beta-casein variants is also liberated from the enzymatic digestion. An acidic solution in mixed aqueous/organic solvent is used to stop the digestion. The solution is diluted and analyzed via LC-MS/MS using specific quantifier and qualifier transitions for the native and SIL peptides. Quantitation is accomplished using internal standard calibration with conversion to each protein variant using the molar masses of the peptide and their corresponding protein variant. Method Scope/Applicability: Method scope states the assay is applicable to infant nutritional products and is selective to bovine beta-casein, to the exclusion of sheep and goat sourced milk. General comments about the method: The method uses commonly employed digestion techniques to liberate unique peptides for the variants of interest. Method Clarity: The method steps are generally clearly understandable. The apparatus, reagent, and determination sections provide adequate detail. • Reagent Preparation (c) Digestion Buffer. This buffer is described as 100 mM ammonium bicarbonate. The instructions state to add 7.91 g ammonium bicarbonate to a total of 100 mL, which would create a 1M ammonium bicarbonate solution. Clarification of this step is recommended. • Standard Preparation (c) Peptide intermediate solutions. Each row in the Name column should be 10X higher, based on the stock solution concentration (1000 mg/L) and dilution (100 µL each solution to a total 1000 µL volume). Clarification of this table is recommended. • The table for beta-casein spiking intermediates appears to be missing the description for creating the 1 mg/L intermediate. Clarification of this table is recommended. Pros/Strengths: • The method uses selective peptides to measure the specific beta-casein variants. • Use of SIL internal standards are beneficial. There were several typographical errors noted in the method:

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Cons/Weaknesses • Is an 11-point calibration curve required? • The activity of the trypsin is not specified.

• The method states to utilize the purity when calculating the beta-casein spiking stock solution concentration. The Sigma sourced reference material uses gel electrophoresis for purity measurement on the certificate of analysis (COA). Is COA purity adequate or is there additional characterization required? • Protein and peptide matrix spikes appear to have a higher total volume relative to un-spiked samples. To make the calculation more consistent it is recommended that all samples (spiked and un-spiked) have the same total volume. • The presented equation in the Calculations section appears incomplete.

• Details about how the results are calculated in the SMPR units (mg/100g reconstituted product) is not described. • A description of how the common beta-casein peptide is used in the analysis is not provided. • Protein spike recovery acceptance criteria are 70% - 120%, which is a larger range than SMPR 2022.005 (80% - 110%). • Glycosylation of measured peptides is possible, likely on the serine and threonine residues. How does the method address this bias risk? Supporting Data • General Comment: The supplied documentation included only the method. No SLV was found. The reviewer asked AOAC staff if an SLV report for this submission was available, with the reply that only the method was provided. In this instance no assessment of the method performance can be made.

Method Optimization:

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Performance Characteristics:

Analytical Range:

LOQ:

Accuracy/Recovery:

Precision (RSD r ):

Reproducibility (RSD R ):

• System suitability:

3

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

1. Is the Validation Study Report in a format acceptable to AOAC?

2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)?

3. Are the figures and tables sufficiently explanatory without the need to refer to the text?

4. Are all the figures and tables pertinent?

5. Could some be omitted and covered by a simple statement?

6. Are the references complete and correctly annotated?

7. Does the method contain adequate safety precaution reference and/or statements?

Recommendation: The submitted method shows potential to meet the SMPR. The reviewer looks forward to evaluating data from the validation report, when it is available.

4

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method_ Amino 05

Title: LC- MS/MS method for Residue Testing for Variants of Bovine β -Casein Protein

Author: A2 Milk Company

Reviewer Name: Reviewer 3

Summary of Method: Samples are dissolved in water, thermally denatured, and spiked with stable isotope labelled internal standard peptides. The mixture is digested using trypsin which liberates four measured peptides – one for A1-type β -caseins, a second for A2-type β -caseins (excluding I), a third for I β -casein, and a fourth as a generic ‘total β -caseins’ peptide used for quality control. The concentrations of the peptides are measured by stable isotopic dilution LC-MS/MS from which the equivalent amount of full-length β -casein protein can be calculated. Method Scope/Applicability: This method is applicable to the determination of A1 and A2-type bovine β -caseins in infant nutritional products. All isoforms/variants are detectable excluding H1 and H2. All phosphoforms are detectable, and all glycoforms are detectable so long as glycosylation does not occur within the region of the quantified peptide. The method is specific to protein sequences originating from bovine beta casein, including full-length and part-length beta casein. General comments about the method: Each is measured using unique peptide generated via trypsin and converted back to intact full length β -casein protein. The method can overestimate the amount of intact β -casein (A1, A2, I β -casein and total β -casein) due to contributions from proteose peptones (e.g., PP5) and γ -casein etc., as they also contain the same signature peptides. These proteose peptones and γ -casein are generally found in high levels in whey. Can be an issue for example, if A2 infant formula products uses mixed/pooled whey source, then proteose peptones and γ -casein can potentially be quantified as intact A1 β -casein.

Method Clarity: The method is clearly written and straightforward and easy to follow.

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Pros/Strengths: • The method measures A1, A2, as well as i- β -casein variant and total β -casein in infant formula • Uses commonly available instrumentation to most food testing labs. Cons/Weaknesses • Signature peptides are “custom synthesized”.

Supporting Data • General Comment: Largely absent

Method Optimization:

• Performance Characteristics: Expected performance characteristics provided as part of system suitability, however full characterization of validation parameters as part of method development not given.

Analytical Range:

LOQ:

Accuracy/Recovery:

Precision (RSD r ):

Reproducibility (RSD R ):

• System suitability: Run acceptance criteria given for calibration curve and quantifier and qualifier concentrations, protein and peptide spike recoveries, and inhouse-QC sample

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

1. Is the Validation Study Report in a format acceptable to AOAC? No.

2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? Yes.

3. Are the figures and tables sufficiently explanatory without the need to refer to the text? Yes.

4. Are all the figures and tables pertinent? Appendix B, C, and D are valuable background information however not suitable for inclusion in method document but would be useful discussion in journal publication of method.

5. Could some be omitted and covered by a simple statement? Yes.

6. Are the references complete and correctly annotated? References not formatted appropriately.

7. Does the method contain adequate safety precaution reference and/or statements? General safety precaution given. Not detail on specific hazards, however.

Recommendation:

3